Psychology A Science in Conflict

Howard H.
Kendler
SOCIAL SCIENCES AND HISTORY division REF
The Chicago Public Library
Received-
BF Kendlert Howard
38 Harvardf 1919—
• K43
Psychology
JHE CHICAGO PUBLIC LIBRARY

EDUCATION & PHILOSOPHY
© THE BAKER ft TAYLOR CO.

Digitized by the Internet Archive
in 2018 with funding from
Kahle/Austin Foundation
https://archive.0rg/detaiis/psychoiogysciencOOOOkend
Psychology: A Science in Conflict
Psychology:
A Science in Conflict
HOWARD H. KENDLER
University of California, Santa Barbara
New York Oxford

OXFORD UNIVERSITY PRESS
1981
Copyright © 1981 by Oxford University Press, Inc.
Library of Congress Cataloging in Publication Data

Kendler, Howard Harvard, 1919—
Psychology; A Science in Conflict
Bibliography: p. Includes index.
1. Psychology—Philosophy. I. Title.
[DNLM: 1. Psychology—Methods. BF38.5 K33p]
BF38.K43 150M 80-22613
ISBN 0-19-502900-3 ISBN 0-19-502901-1 (pbk.)
Printing (last digit ):987654321
Printed in the United States of America

Preface
This book is a direct outgrowth of my career in psychology. I be¬

came committed to psychology after experiencing the excitement
of doing original research under the tutelage of Solomon E. Asch.
I chose to go to the University of Iowa for graduate training be¬
cause of my interest in Gestalt psychology. There, more by acci¬
dent than by design, I took an advanced seminar with Kenneth W.
Spence that inspired me to do research with him. It gradually be¬
came apparent that the differences between Gestalt theory and
the neobehaviorism of the Hull-Spence-Miller paradigm were not
simply a consequence of competing theoretical assumptions. To
fully understand the differences, one had to comprehend varying
conceptions of science, strategies of research and theorizing, and
even views about the role of science in society.
My methodological interests expanded during World War II
when I served as an army clinical psychologist. Being a psycho¬
therapist, interpreting a variety of psychological tests, participat¬
ing in psychiatric staff meetings, and supervising a ward of pa¬
tients forced me to confront new methodological issues that had
not been met in the laboratory.
My interest in methodological problems in psychology soon
developed into a professional responsibility. In order to help unify
a graduate program at New York University, I was asked to pre¬
pare a course (Psychology 201: Gonceptual Problems in Psychol¬
ogy ) to be required of all incoming students that would acquaint
them with the methodological structure of psychology. My plan
was to reveal the alternative methodological decisions that have
been and can be made in psychology and to analyze the implica¬
tions of these decisions on research, theory, and social applica-
vi Preface
tion. I aspired to do this in a detached manner in order to illumi¬

nate the various alternatives without prescribing any particular
approach. Admittedly this neutral stance was difficult to main¬
tain because of my commitment to the particular set of methodo¬
logical assumptions and strategies that governed my own research
program. But realizing that there was no one valid approach to
psychology I tried to inhibit any missionary zeal to persuade
others to adopt my orientation. And I might add that this zeal
subsided as a result of repeatedly teaching Psychology 201.
My position on many of the issues analyzed in this book have
changed from the time the course was initially taught at New
York University. That was inevitable. The vigorous and exciting
discussions and debates with the talented N.Y.U. students forced
me to rethink my positions and modify some of my views. I also
profited from teaching the course at the University of California,
Berkeley; The Hebrew University of Jerusalem; and the Univer¬
sity of California, Santa Barbara. One methodological assumption,
however, has remained firm. The final support for any theoretical
or interpretive psychological statement must be linked to empiri¬
cal evidence. This does not imply that all questions, especially
those directed at many significant problems of human existence
and social organization, can be answered by empirical means
alone. Empiricism has definite limits. But those limits cannot be
overcome by a priori philosophical reasoning.
Numerous debts have to be acknowledged. Being a Fellow
at the Center for Advanced Studies in the Behavioral Sciences
helped me plan this book. Tracy S. Kendler offered many cogent
criticisms and suggestions at various stages of the project. Gerald
Zuriff submitted a penetrating critique of the manuscript that
proved to be most valuable in the preparation of the final version.
Discussions with Boy Lachman, usually dominated by disagree¬
ments, proved enlightening. Many useful suggestions were of¬
fered by David Messick, John Foley, and T. S. Krawiec. Naturally,
I am obligated to many scholars whom I cite. In some cases, I de¬
cided against citing a specific reference because of the context in
which a particular idea was embedded. Isolating the notion that
interested me might be seen as a distortion or an oversimplifica-
Preface vtt
tion. Rather than risk a debate about whether I was misstating a

partieular position, I thought it best to deal with the idea alone.
Finally, to proteet the innocent, I must acknowledge full responsi¬
bility for the views expressed in this book.
Many different secretaries worked on this project. The major
contributions were made by Christine Whitehead, JeNeal Brad¬
ford, Carolyn Scherr, and Chris Clark. I appreciate the care they
devoted to their task. Finally, I would like to express my indebt¬
edness to Melanie Miller at Oxford University Press.
Santa Barbara Howard H. Kendler

January 1981
N - *•.
NMt****
"■' ■' ft^ ** * •
'V >,'«• • < K ■'; %, .» ,*yi

1
r «i l.’Oi. ’»T • .V uJtnii ff ■‘’ .’ ', 'h-
i' ■ 4 ii4^ Vt< ..

» - ■ t' ^ ^ yt •
.n V , <- ^ \ 1 ■^ * st» I Cl f'
'St!..’ 1 !i-
. , -i 1
• f . ,’ •» t, .
■<i * 1' 4 ;■! V ; . .'• -. l , *1 .. ,.

,• •■'.*•
• .4 « >
^
it i- ‘ . >♦■-
H,;, '/ ’hr
%
.1 .1. ^ . 1 •• ■ 1^'
< I j f’
'■ <^’4 i. i
i
'I. , •■ ■ . f ; , ‘p-VCt,
^.■' '4 ■ ” ■■ » '• -■■ ♦' -»■■'
. -■» ■»=■■»
n ■i‘fi
*,
;■ ' -■, . 1 ■
V"
. • f A
!»: -4 fit&ititvvi u<
»
.; ^ * V’ *f4
• \ tfiiii'itfiiîi
't Jf ' m
- -». •'
Contents
1. Plans and Purposes, 3
2. The Data Base and Subjeet Matter of Psyehology:

Part One, 13
3. The Data Base and Subject Matter of Psychology:

Part Two, 66
4. On Understanding: Part One, 100
5. On Understanding: Part Two, 152
6. Psychology and Values: Epistemological Considerations, 217
7. Psychology and Values: Psychological Considerations, 269
8. Future Trends in Psychology, 299
9. Postscript, 371
References, 372
Index, 385
To those echoes of the past
that still ring clearly.
Dedicated to the memory of my brother
Joel J. Kendler (1918-1935)
Plans and Purposes 1
1. Several interpretations of the nature of psychology are possible. An
understanding of contemporary psychology demands an appreciation
of both the different methodological assumptions that underlie various
interpretations and the distinctive kinds of knowledge they produce.
Such clarification can best be accomplished by analyzing psychology
as it is rather than as it should be. In this chapter the orienting atti¬
tudes of the proposed epistemological analysis are described.
PROBLEMS
Psychology is an ambiguous science. Arguments rage about the

nature of the discipline: what it is and what it should be. This
book aims to elarify some of the basic issues of this eontinuing
debate by analyzing three interrelated problems: the nature of
psyehology’s empirieal methods, the criteria employed to arrive at
psyehologieal truths, and the prineiples that govern the appliea-
tion of psyehologieal knowledge.
To reveal the structure of psyehology requires a systematic
plan, and yet no predetermined set of logical or empirical opera¬
tions are available for sueh a task. Therefore, one is forced to
make personal deeisions to guide a rational analysis while recog¬
nizing that these decisions cannot command acceptance in the
same sense that a logical proof or an experimental faet can.
To compensate for this inevitable arbitrariness of any meth-
odologieal plan, it is necessary to make elear its goals and orient-
3
4 Psychology: A Science in Conflict
ing attitudes. My intention is to oflFer a detaehed analysis of the

strueture of eontemporary psychology that will identify both
common themes and discordant notes. This analysis will be guided
by five predilections, which will presently be elaborated. First,
psychology will be described as it is and not in any idealized ver¬
sion. Second, the goal will be illumination and not prescription.
Third, the methodological analysis will be that of a psychologist
rather than of a philosopher of science. Fourth, since knowledge
is an end-product of human behavior, the psychology of the sci¬
entist and the sociology of the community of psychologists will at
times be considered when analyzing the nature of psychology.
Fifth and last, my aim is not to achieve 'unanimous agreement
about the epistemological decisions that are appropriate for psy¬
chology, an unrealistic and perhaps undesirable goal. Instead, my
desire is to encourage collective understanding—the comprehen¬
sion by all members of the community of psychologists—of the
competing and contrasting methodological decisions that different
groups of psychologists have made and can make in the future.
These five predilections will now be clarified and justified:
1. Contrasting views of psychology. The history of psychol¬
ogy testifies clearly to the fact that different conceptions of psy¬
chology are possible. More often than not, these contrasting posi¬
tions have been justified on the basis of representing the real es¬
sence of psychology. Probably the most controversial question,
and the one with the widest-ranging impact, has been whether
psychology is the science of the “mind” or “behavior.” Is there a
true answer to such a question? Does some Platonic notion of psy¬
chology exist that has to be discovered? I think not. The nature
of psychology can best be clarified by being both democratic and
operational. Psychology is what psychologists do. If psychologists
do basically different things, then there are basically different
kinds of psychology.
2. Methodological decisions: illumination versus prescription.
Methodological analyses of psychology in the past have fre¬
quently resulted in prescriptions that unhesitatingly offer eternal
truths as to how research is to be done, what dependent variables
to measure, what problems to investigate, what psychological
processes are fundamental, what questions are relevant, what

answers are trivial, and what constitutes true explanation. In eval¬
uating these prescriptions, some modestly proposed, others messi-
anically, it must be noted that no universally accepted guidelines
presently exist that specify the proper procedures a scientist
should follow in the pursuit of ‘‘truth.” From data collecting to
theoretical interpretations, the scientist is constantly required to
make many decisions that cannot be defended by appeals to truth
or logic. To understand what these decisions are, what considera¬
tions encouraged their adoption, and what their consequences
are, can be helpful in identifying and clarifying significant meth¬
odological issues in contemporary psychology. The aim of such an
analysis will be illumination, not persuasion.
3. Different perspectives in analyzing the methodology of
psychology. Though I am not a philosopher of science, this book
might suggest that I am functioning as one. Such a conclusion
would be based upon the assumption that an analysis of the struc¬
ture of psychology is intrinsically a philosophical task regardless
of who performs it. Such an assumption can be questioned. Al¬
though the philosophical and psychological orientations overlap
in both sphere of interest and method of analysis, they are never¬
theless far from equivalent in technique or purpose. Without
denying Feigl’s point that “philosophical analysis [is] continuous
with scientific research” (1967, p. 160) one can nevertheless rec¬
ognize that a philosophical perspective will yield a different view
of psychology than will a research orientation. Perceiving psycho¬
logical issues primarily within an empirical context, and for the
moment the term empirical will be used in the broad sense of re¬
ferring to any form of observation, encourages the isolation and
identification of problems capable of empirical resolution. This
does not mean that all methodological issues, even those that
demand only linguistic clarification, will be fitted into an empiri¬
cal mold, however inappropriate. Instead, many issues that will
engage our attention will be perceived more in terms of their em¬
pirical substance than in terms of their philosophical content or
background. For example, the mind-body problem, which has
been pursued interminably in the history of philosophy, has led
to numerous competing resolutions such as interactionism, paral¬

lelism, epiphenomenalism, monism, and the identity hypothesis.
From the viewpoint of many contemporary psychologists, these
philosophical positions are irrelevant to the science of psychology,
while to others they represent unnecessary prejudgments about
empirical issues. Although such views may be considered by some
philosophers to be excessively parochial, they do suggest that a
methodologically oriented analysis of the nature of psychology by
a psychologist will differ from that conducted by a philosopher.
This is not surprising considering that the methods, purposes,
and goals of the two disciplines differ.
The thrust of these comments should be obvious. This book is
written from the perspective of a psychologist whose purpose is to
reveal the structure of psychology. For the purpose of this book,
being a a professional psychologist and at best an amateur philos¬
opher has two advantages. Viewing psychology from within al¬
lows one to detect problems and issues not apparent from with¬
out. Ignoring subtle philosophical issues prevents one from be¬
coming entangled in those that have no bearing on one s action as
a psychologist. In sum, my analysis may not be sufficient for the
philosopher, hut it may be necessary for the psychologist.
4. Psychological insights into methodological issues. The rela¬
tionship between a methodological analysis of science and the
psychological analysis of a scientist is a source of much confusion.
Two contradictory, oversimplified views illustrate how science is
perceived when objective methodological criteria or subjective
psychological factors are emphasized to the exclusion of the
other. At one extreme is the completely rational view of science
that states that from theoretical formulation to empirical investi¬
gation (or vice versa) every decision made by a scientist can be
judged by objective ‘logical” considerations. The entire scientific
enterprise is governed by completely rational rules that, if under¬
stood and implemented, will inevitably lead to “truth.” Therefore
to understand psychology one needs only to identify the rational
rules that govern the scientific enterprise and determine whether
they are being obeyed by the psychologist. The contrasting view
is that the practice of science is governed by the personal tastes of
the individual scientist. The selection of one theory over another

is not determined by rules of evidence but instead by personal
preference. Understanding science depends not upon a logical
analysis of its methods but instead upon a psychological compre¬
hension of its practitioners.
Both views, objective and subjective, are oversimplified and
obviously wrong, although they have been attributed to sophisti¬
cated and well-known interpretations of the philosophy of sci¬
ence, e.g., logical positivism is perceived as having tried to un¬
cover an infallible recipe for conducting science in such a way
that progress is guaranteed; Thomas Kuhn, in his widely read
The Structure of Scientific Revolutions (1962) is interpreted as
espousing the view that science is without rules and that only
personal tastes dictate the acceptance and rejection of theories.
In addition to being distorted by oversimplification, these ob¬
jective and subjective characterizations of science suffer from two
fundamental misconceptions. First, they fail to recognize that the
scientific method consists of a variety of activities not necessarily
governed by the same standard of conduct. Although rules of
logic govern one activity such as deducing the empirical conse¬
quences of theoretical assumptions, personal taste will determine
the choice of a particular research problem (e.g., mother-child
relationships or hypothalamic functioning) to investigate. Second,
some scientific activities involve both rational choices and per¬
sonal preferences. For example, in selecting a criterion for expla¬
nation, one may adopt a deductive framework with its implicit
logical demands while simultaneously rejecting the deductive
rigors of mathematics for a looser informal logic in order to ob¬
tain a wider explanatory scope for one’s tentative theoretical
hunches.
In discussing the relationship between a methodological anal¬
ysis of science and the psychological analysis of the scientist, it
becomes essential to clarify the meanings of subjective and ob¬
jective scientific standards. On the most fundamental level, sci¬
ence is a sociopsychological enterprise, which simply means that
all its activities are expressions of principles that govern human
behavior. But this does not mean therefore that all decisions are
subjective in the sense of reflecting personal taste. Consider a

game of chess. A person decides to play chess in preference to
checkers. Without going too deeply into the psychological reasons
underlying his choice, we can nevertheless attribute it to a per¬
sonal preference. Attributing the choice to personal preference
does not mean that it is arational or irrational in the sense that
the choice is without reason or unreasonable. The person finds
chess more interesting and challenging, he enjoys reading chess
magazines, he is an outstanding player, his father encouraged
him to play chess, etc. All these reasons can be summarized in
the simple statement that he prefers chess to checkers and there¬
fore when given a choice he opts for the former. Although his
choice is reasonable, it is not demanded by some explicit rule. An
example of rule-governed behavior would be the restrictions
placed upon the chess player when he moves the various pieces,
e.g., bishop can be moved only along a diagonal while rooks are
limited to moves along ranks and files. Whereas the choosing to
play chess instead of checkers was an example of behavior con¬
trolled by personal preference, the decision to move a bishop
along a diagonal is dictated by a socially adopted rule. Society
demands that a bishop be moved along a diagonal and if one
insists on moving it along a file, he will be prevented from playing
the game that is conventionally called chess. Socially adopted
rules clearly separate chess from non-chess, chess from checkers,
and checkers from non-checkers.
The rules governing the movement of chess pieces do not
control what pieces will be moved, only that they be shifted
from one square to another in a specified way. The actual move¬
ment of pieces is not simply an expression of preference because
some moves (e.g., opening the game by moving the King’s Rook
pawn) are counterproductive if one aims to win a chess game.
The actual moves of a reasonable sophisticated player are con¬
trolled by principles of strategy. Although these principles are
not as automatic and restrictive as the rules that govern the
movement of pieces, they nevertheless place restraints upon be¬
havior.
My intention is not to suggest that the scientific method is
completely analogous to chess but instead to illustrate different

kinds of behaviors that operate in science as well as in chess. Be¬
cause science is completely and exclusively a human enterprise,
the actions of scientists can be viewed solely within a sociopsy-
chological framework. This does not mean, however, that all sci¬
entific behaviors are similar. As a first approximation I have sug¬
gested that a distinction can be made among three kinds of be¬
havior that lie along a dimension of increasingly restrictive rules
of conduct. At one extreme is behavior that is regulated by no
rule but is simply an expression of preference. At the other ex¬
treme are rigid rules that must be strictly obeyed, while in be¬
tween are behaviors governed by principles of strategy that allow
some options.
The acknowledgment of science as a sociopsychological en¬
terprise and the distinction among different kinds of preference-
governed and rule-governed behaviors provide a frame of ref¬
erence to discuss the role of the psychology of the scientist in
analyzing the science of psychology. Although, in principle,
deeper understanding of any human activity can be achieved by
comprehending its psychological origins, this generalization has
certain limits. Popper (1949), for example, thought it necessary
to “distinguish sharply between the process of conceiving a new
idea, and the methods and results of examining it logically.” In
the same vein, Reichenbach (1938) used the terms conteoct of
discovery and context of justification to distinguish between the
form in which knowledge is subjectively created and the form in
whieh the knowledge is communicated to others. The purpose of
this distinction is to separate from empirical propositions, ex¬
pressed either as statements of fact or as theoretical principles,
those events that are irrelevant to the task of judging the proposi¬
tion’s meaning and validity. For example, when evaluating the
electromagnetic theory of light, one does not have to understand
all those psychological characteristics of Maxwell, such as his per¬
sonality and thinking style, that were responsible for his famous
formulation. To understand and judge the electromagnetic theory
of light requires only knowledge of the theory itself. Consider an
extreme example. A psychologist, under the influence of alcohol.
formulates a new theory of reinforeement. Let us assume that the

efiFeets of aleohol played a major role in his “ereative” aet in the
sense that the novel notion would not have oceurred if he were
sober. The theory eannot properly be rejeeted beeause the the¬
orist was drunk. Sueh information is irrelevant, even when we
know that alcoholie eonsumption ean interfere with problem
solving. The theory must be judged within the eontext of justifiea-
tion in its ability to integrate the faets of reinforeement.
Now that some limit has been set upon the psyehological
analysis of the seientifie enterprise, attention will be direeted to
two ways in whieh the strueture of psyehology ean be elarified by
eonsidering the psyehology of the seientist. The first coneern is
with the motives and aspirations of the seientist that influenee his
adoption of speeifie methodologieal assumptions about the nature
of psyehology: its subjeet matter, mode of explanation, soeial
role, ete. Psyehology has beeome what influential psychologists
have thought it should be. If the notion is rejected that various in¬
terpretations of the nature of psychology can be evaluated by
how close they come to some single Platonic ideal, which must
and can be revealed, then the competing conceptions can be
better understood when the personal psychological reasons for
their adoption become known. An attempt to understand the psy¬
chology of the psychologist need not be as difficult or complex as
it would appear initially. Psychologists have usually been unin¬
hibited about expressing the philosophical commitments under¬
lying their views of science and society, and these commitments
in turn have typically shaped their conception of psychology.
Thus in trying to understand the structure of psychology, it not
only becomes desirable but also necessary to comprehend the mo¬
tives and aspirations of psychologists.
Nobody should be surprised to discover that conceptions of
psychology have not always been articulated with such clarity
that no doubt remains about their meaning. To overcome this
ambiguity historians and methodologists frequently project their
own biases into an original formulation, thus achieving explicit¬
ness at the expense of accuracy. An alternative method of clarify¬
ing a vague position is to interpret it consistently with the au-
thors scientific style. An example of such an effort is Ghiselin’s

(1969) attempt to illuminate Darwin’s evolutionary theory by re¬
constructing Darwin’s methodological commitments to the hypo-
thetico-deductive approach. Such efforts are psychological in na¬
ture because a person’s methodological preferences are revealed
by the analysis of the total range of his scientific behavior.
The second manner in which psychological analysis can be
brought to bear upon methodological clarification is to consider
behavioral evidence that is relevant to the scientist’s data-collect-
ing activities. For example, the question has been raised as to
whether observations are biased in the direction of what the sci¬
entist expects or hopes to observe. In support of certain method¬
ological positions (Hanson, 1958; Kuhn, 1962) the results of psy¬
chological experiments investigating perceptual processes have
been cited. If evidence about the perceptual behavior is available,
it would seem only proper that such information be used to
clarify methodological issues. The question is not whether it is
proper to use psychological evidence in analyzing methodological
issues but instead whether the information is being used properly.
This brief discussion relating psychological and methodologi¬
cal analyses does not exhaust all possible sources of interaction.
The point here is that in some, but not all, cases methodological
discussions can profit by considering the psychology of the
scientist.
5. Collective understanding hy psychologists of different
methodological positions. What is badly needed in contemporary
psychology is a collective understanding about the present struc¬
ture of psychology. Some perhaps do not share this view or value
its goal. But it should be emphasized that collective understand¬
ing does not require unanimous agreement about what method-
ology is appropriate for psychology. Understanding, for example,
the reasons for adopting conscious experience as the dependent
variable in psychology does not require accepting such a decision.
But intelligently rejecting this choice does demand an apprecia¬
tion of the reasons for its acceptance. Collective understanding of
the different methodological decisions not only should expose psy¬
chologists to alernatives that have been ignored or misunderstood
but also holds the promise of elevating the quality of method¬

ological controversies.
PROLOGUE
The plan of this book is to analyze successively three basic and

interrelated questions that face all psychologists: What is the sub¬
ject matter of psychology? What are the criteria for understand¬
ing psychological events? What ethical principles underlie the
application of psychological knowledge?
There is no one valid answer to these questions. Instead each
question poses a “choice point” that permits alternative responses.
These responses in turn raise further options. In essence, psy¬
chologists are confronted with a methodological maze that ulti¬
mately can lead to incompatible conceptions of psychology. The
purpose of this book is to explore the various paths through this
maze and to disentangle the numerous methodological and em¬
pirical issues associated with each path. This examination of the
structure of psychology, which reveals a diversity of methodologi¬
cal assumptions and ultimate goals, inevitably raises questions
about the future of psychology. This topic will receive attention
in the last chapter.
Now that the orienting attitudes and the goals that will
guide this effort to reveal the structure of contemporary psychol¬
ogy have been presented the analysis can begin.
The Data Base and
Subject Matter of Psychology:
Part One
1. Psychology can be viewed as the science of the mind or as the sei-

ence of behavior or as some combination of both. Whatever concept
is chosen the data base is the observation of the scientist. The observa¬
tional base of studying conscious experience or behavior is intrinsically
subjective. Although equally subjective, the study of behavior is qual¬
itatively different from the study of the mind. Behavior can be ob¬
served by many individuals thus permitting intersubjective agreement
about behavioral events. In contrast, intersubjective agreement about
conscious experience is doubtful because the event is available only
to the scrutiny of the experiencing individual.
2. Behavior, an observable event independent of theoretical precon¬

ceptions, can be defined directly as an activity of an effector or indi¬
rectly as an event that modifies the environment. This dual-classifica¬
tion system is capable of incorporating certain forms of “meaningful”
behavior.
3. The psychologist who wishes to systematically study consciousness

has several options ranging from using trained introspectors to naive
observers and from analyzing the mind into basic elements or pro¬
cesses to considering it a holistic entity. The history of psychology
suggests that phenomenological research is incapable of achieving
consensual agreement about the nature of mental events. Methodologi¬
cal preconceptions about the manner in which the mind is to be in¬
vestigated determine what is observed.
18
4. The question is raised as to whether the unresolved phenomeno¬

logical controversies in experiential psychology are methodologically
similar to unresolved theoretical disputes in behavioral psychology. This
similarity is denied because the roots of the experiential disagreements
result from observational differences while those of behavioral psy¬
chology are a consequence of theoretical limitations. In other words,
the observer and what is observed are separable in behavioral psy¬
chology but not in experiential psychology.
PSYCHOLOCrS SUBJECT MATTER .
Although academic psychology has passed its hundredth anniver¬

sary, agreement about its proper subjeet matter has yet to be
reaehed. Is this state of affairs due to the absence of a sufficiently
penetrating and persuasive analysis that someday will be forth-
eoming? Or does the failure result from the problem itself: the
subject matter of psyehology eannot be determined by any logieal
or faetual means that in the foreseeable future will lead to any
eonsensus among the members of the eommunity who eall them¬
selves psyehologists. By identifying and analyzing the deeisions
that have led to various eoneeptions of psyehology, this ehapter
will offer arguments in favor of the seeond alternative.
The possibility will also be raised that a universal eoncur-
renee about the subjeet matter of psyehology may be unachiev¬
able in principle beeause different subjeet matters are possible.
Although sueh a conelusion might sadden those who desire an
integrated diseipline, a frank reeognition of its intrinsie disunity
would do much to minimize those unnecessary and destructive
debates that have plagued psyehology’s past.
CONSCIOUS EXPERIENCE AND BEHAVIOR
Two extreme views have prevailed about the subjeet matter of

psyehology: it should be solely eonseious experienee or solely be-
The Data Base and Subject Matter of Psychology: 1 15
havior. Numerous compromises have been offered that, in one

form or another, express the notion that psychology should study
both. The roots of this argument are deeply embedded in the his¬
tory of philosophy, an examination of which would no doubt help
clarify the issues but would not lead to any meaningful resolution.
For that, psychology must examine itself and discover the alterna¬
tive solutions within its own historical and methodological frame¬
works.
Descartes (1596-1650) most clearly represents the dualistic
heritage that influenced the development of the science of psy¬
chology. For Descartes the human being is a union of a psycho¬
logical mind with a mechanical body. The mind is a thinking sub¬
stance, conscious of itself and capable of being studied by self¬
observation. The body requires for its study the methods of natu¬
ral science. Whereas the mind is free, the body is constrained by
principles of physics. To overcome the separation between the
mind and the body, Descartes postulated a mechanism for their
interaction: the pineal body, which lies in the brain, is controlled
by the mind, and in turn controls the body.
Descartes’ conceptions touch upon fundamental epistemo¬
logical and methodological problems that have continuously re¬
mained at center stage in the history of psychology. First, by
postulating two worlds, one of the mind, the other of the matter,
Descartes raised the question whether knowledge about each de¬
rives from different sources. Second, the Cartesian system antici¬
pated the distinction between conscious experience and publicly
observed behavior as the subject matter for the science of psy¬
chology. Third, the issue of appropriate methods for investigating
mind and body was raised by Descartes with the suggestion that
different procedures are required: introspection, observation from
the inside, is needed to examine the mind while observation from
the outside, the procedure of natural science, is required for the
study of the body. Fourth, Descartes recognized the significance
of the problem of the relationship between mind (conscious ex¬
perience) and body (behavior). Although his proposed solution
based upon the activities of the pineal gland possesses only his¬
torical interest, it must be recognized that the problem itself is
still important and unresolved. Finally, Descartes suggested, con¬

sistent with Platonism, that the mind contained innate ideas such
as notions of time and space as well as the existence of God.
Whether ideas were innate or acquired became a source of debate
among philosophers, and this argument was transformed for psy¬
chologists into a debate concerning the relative contributions,
within their discipline, of heredity and environment.
The history of philosophy relevant to psychology since the
time of Descartes has been concerned for the most part with
variations of his ideas. Mv attention now will be focused on the
source of empirical knowledge of mind (conscious experience)
and body (behavior), and the subject matter of psychology—prob¬
lems that are distinct but unfortunately have become muddled.
Data Base Versus Subject Matter
Many philosophers who rejected Cartesian dualism, particularly

some of the British empiricists, stressed the priority of phenome¬
nal experience—raw observations, sense data—as the foundation of
all knowledge. Berkeley (1710), for example, concluded that
events, objects, and sensations do not have ‘'any existence out of
the minds or thinking things which perceive them.” This kind of
philosophical speculation led to the conundrum familiar to most
college freshmen: “Is there a sound in an uninhabited forest when
a tree falls to the ground?” To anticipate future discussion the
question becomes clarified when the distinction is made between
physical and psychological measurement, between sound waves
and sensations of sound. Within this context the appropriate an¬
swer becomes “Yes and No!”
From Berkeley’s skepticism not too much intellectual distance
had to be traversed to arrive at Mach’s conclusion that all sci¬
ences, including physics and psychology, share a common data
base—the sensory experience of the scientist. Mach’s judgment
anticipated positions adopted by apparently irreconcilable orien¬
tations: structuralism and behaviorism. Titchener (1910) states:
“It is plain that all sciences have the same sort of subject matter;
they all deal with some aspect of the world of human experience.”
Pratt, expressing a position consistent with methodological be¬
haviorism (Bergmann, 1956), a position that asserts that behavior
can be investigated and explained without direct examination of
mental states, writes: “No science is capable of definition in terms
of subject matter, for the subject matter of all sciences is the
same. Some person, some scientist, makes a report on something
which he has observed . . . The physical sciences, the biological
sciences, and the mental sciences all make use of the same ob¬
servational data. Psychology is a division of scientific labor, not a
subject matter” (1939, pp. 22-32).
How can such agreement be achieved by methodological
positions—structuralism and behaviorism—that expound such con¬
trasting views of the subject matter of psychology? This agree¬
ment is partly illusory because it stems from the failure to distin¬
guish between two possible meanings of the term subject matter.
Both Titchener and Pratt use subject matter to refer to the data
base of science, and both reach the same conclusion that the rock-
bottom foundation of all sciences is the phenomenal experience
of the observing scientist. But subject matter can be interpreted
within another context, that of characterizing the phenomena that
identify a discipline. Just as a course in French has a different
subject matter than a course in German, so does physics possess
a different subject matter than psychology. Although students in
introductory French and German courses are both learning,
among other things, to associate foreign words with their English
equivalents, the types of words involved are different. Similarly,
investigating the speed with which balls roll down an inclined
plane or the trajectory of a spacecraft involves a different set of
events than the study of a dog salivating to a tone or an adult
human solving a mathematical problem. The former deals with
the movements of inanimate objects while the latter deals with
the behavior of organisms, a discrimination that can be made in
a highly reliable fashion.
It could be argued, although not very persuasively, that the
behavior of organisms cannot be distinguished fundamentally
from physical principles that govern objects in motion because
ultimately behavioral prineiples ean be redueed to laws of phys-

ies. Although a diseussion of reduetionism will be postponed until
later (Chapter 4), it should be sufEeient for the time being to
state that even if the optimistie elaims of reduetionism were to be
aeeepted, the diserimination between physical and behavioral
phenomena—their observational characteristics—is still highly re¬
liable. The concern with differences in subject matter, from the
viewpoint of characterizing a set of phenomena, is not one of
theoretical reduetionism but instead one of being able to dis¬
criminate reliably between distinctive features of different sets of
phenomena. (Editors of physical and psychological journals
rarely if ever get confused about which 'discipline a manuscript
properly belongs to.)
Subject matter, defined as the distinctive characteristics of a
set of phenomena, is being emphasized because it identifies a cen¬
tral problem in the history of psychology. Whether psychology
should have one or two subject matters—private conscious experi¬
ence and/or objective behavior—has been a source of controversy
for the major part of its history. Although both the structuralist
and behaviorist schools agree that scientific knowledge is based
upon a foundation of phenomenal experience, a discrimination
was made between two kinds of such experience: the observations
of one’s private experience, such as feeling happy, and the obser¬
vation of the public behavior of another organism, the recall of a
list of words.
In preparing to deal with the practical problem of selecting
an appropriate subject matter for psychology, it might be useful
to shift from a historical-philosophical framework in judging the
mind-body problem to the perspective of an empirically oriented
psychologist who, while freeing himself of methodological pre¬
conceptions, seeks to assess the research potential of his discipline
as the science of the mind or of behavior. The examination from
this vantage point is not designed to resolve perennial disagree¬
ments concerning the mind-body problem either in psychology or
philosophy, but instead to identify the options that are available
to a psychological investigator who aspires to draw conclusions
acceptable to his peers.
As noted, scientific choices are frequently an expression of
‘Volitional decisions.” These decisions can exert profound eflFects

upon the structure of a science when they lead to a “volitional
bifurcation” (Reichenbach, 1938), a choice between irreconcil¬
able alternatives as was presumably the case when Wundt opted
for psychology to be a science of the mind while Watson chose
behavior to be the dependent variable in psychology. The term
“volitional bifurcation” is particularly appropriate to characterize
such decisions because it emphasizes the divergent directions that
such choices force a science to take.
The Distinction Between Private Experience

and Public Behavior
Is the distinction between mind, for the moment equated

with phenomenal experience, and behavior, defined as publicly
observed actions, valid or does it represent a conceptual confu¬
sion? One way of dealing with such a question is to transform it
into a query about the independence of two kinds of observa¬
tions: mental events and publicly observed behavior. This ques¬
tion has two components: (1) Can conscious experience be dis¬
tinguished from public behavior? (2) If so, does it differ in its
epistemological characteristics? Although these two issues cannot
be completely isolated they will be analyzed separately.
The Observational Base of Experience and Behavior
The distinction between conscious experience and public be¬

havior can be viewed from two observational perspectives: the
person who is examining himself and the person who is examining
another.
I can reliably distinguish between my observations of my
conscious experience and my observations of my public behavior
(sometimes only with the assistance of recording instruments) —
my feelings of affection and my affectionate behavior, my intense
anger and my angry outbursts. In other words, I am aware of the
difference between my inner experience directly accessible only
to myself and the behavioral acts that I and others can observe.
Unlike the case of a person distinguishing between his own
conscious experience and public behavior, the individual who is
examining another person can justify the distinction under con¬

sideration only by inference. I can directly observe another per¬
son’s public behavior but can only infer his inner experience from
introspective reports or some other form of public behavior.
One can easily be persuaded that the distinction is valid be¬
cause one can observe it directly, plus the fact that others report
a similar discrimination. On this basis it can be argued that it
would be wise for psychologists to accept the validity of the dis¬
tinction between private experience and public behavior without
qualifications and leave any subtle epistemological issues to
philosophers.
Accepting such a pragmatic conclusibn appears both reason¬
able and desirable as long as one does not overextend its implica¬
tions. First and foremost one must recognize that the ready ac¬
ceptance of the consensual agreement about the distinction be¬
tween one’s own private experience and public behavior in no
way elevates the observational base of private experience into the
public domain. No matter how many individuals report that they
have private experience, the fact remains that any particular pri¬
vate experience is accessible to only one person, the experiencing
individual. To highlight this point consider the hypothetical exam¬
ple of a person who denies experiencing private feelings or
thoughts. Although one would be encouraged to reject such a
statement as mistaken or pathological one must recognize that
such a rebuttal is not based upon direct contradictory observa¬
tions. In other words, I cannot look into his mind to obtain con¬
trary evidence. I cannot therefore conclude that his mindlessness
is factually incorrect in the same sense that I can reject statements
about his public behavior. If this mindless chap for example
maintained that his reaction time fails to be increased by alco¬
holic consumption, I can prove him wrong (or perhaps correct)
by a simple laboratory demonstration. We could both observe his
public behavior but only he could observe his own private ex¬
perience. Thus the assumption that all individuals have a mental
life becomes less compelling when we seriously entertain the pos¬
sibility of mindless persons. In the final analysis, the only mind
whose existence I can be sure of is my own.
Within this context acceptance of the distinction between

human minds and public behavior becomes a matter of choice
since it is not demanded by logic, and its factual basis could con¬
ceivably be open to an unresolvable dispute. I may be overcau¬
tious in not rejecting the possibility of mindless humans and
overconcerned with the problem of “other minds”; yet a greater
danger than excessive prudence is that of disregarding special
methodological problems entailed in researching other minds.
In any case, the psychologist, unlike the philosopher, must go
beyond his analysis of the mind-body problem; he must cope with
it both empirically and theoretically. In regard to the private ex¬
perience/public behavior distinction, he can deny, ignore, or deal
with it.
I choose the last alternative not because it is demanded by
the dictates of science but because it simply seems to be a rea¬
sonable decision. I do not wish to be pushed from the reasonable
concession that other minds cannot be directly observed to what
I consider the unreasonable doubt that other minds exist. In short,
I accept the existence of my own mind and intuitively trust the
verbal reports of the existence of other minds. But I cannot ignore
the decisionlike quality of my choice, a recognition that makes me
particularly aware of problems associated with ‘'other minds” and
of methodological decisions that differ from my own.
The Epistemological Status of Experience and Behavior
Accepting the distinction between “subjective” experience

and “objective” behavior raises the question as to whether these
two kinds of events are epistemologically different. To clarify the
issue consider a pedestrian example: I generate an image of a rat
responding at a choice-point of a T-maze. Compare such an
image with that of my observing a “real” rat in the “same” kind
of situation. In one fundamental sense the “observations” of the
imaged rat and the real rat are the same. Both observations are
strictly subjective in that they are both experientially private,
directly accessible to no one else but myself.
Although fundamentally similar in terms of their intrinsic
subjectivity, the two kinds of observations differ in an important

pragmatic way. The observation of the image of the rat is strictly
intrasubjective in the sense that it is encased in my conscious
experience, incapable of being perceived by anyone else. I am the
only source of knowledge about his choice behavior, whether he
turns left or right. I can transmit to somebody else information
about the choice, but then the information becomes second-hand,
no longer based on direct observation.
Observing a real rat at a choice-point allows for reaching
inêrsubjective agreement. A group of observers with adequate
vision, a knowledge of what is left and right, and a reasonable
level of sobriety, should have no difficulty in reaching a unani¬
mous agreement about the rat’s choice. Of course, factors could
operate to prevent achieving unanimous agreement: the criterion
of a right or left response could be poorly defined thus producing
disagreement when the rat darts into the right alley and then re¬
traces and enters the left alley. Or false evidence could be re¬
ported: an observer could be a compulsive liar or a theoretically
biased observer might “unconsciously” record a response to the
left when in fact the rat went right.
These possible sources of disagreement or distortion of inter-
subjective evidence have important philosophical implications
and raise practical problems for the experimenter. But these prob¬
lems are of limited significance to the experimenter because they
are so easily handled. Any competent and honest investigator will
have no difficulty in arranging an experimental situation in which
observers, regardless of theoretical persuasion, can achieve com¬
plete agreement about the choices of rats in a T-maze. And if in¬
competence or dishonesty were operating, subsequent replications
would overcome the mischief created by the initial misleading
results.
This analysis suggests that observations of “inner” and “outer”
events are equally subjective but qualitatively different. Because
of the nature of the world and the structure of our bodies, agree¬
ment about the occurrence of outer events can be achieved more
easily than those of inner events. It is not the experience itself
that distinguishes private from public observations but instead
the availability of the observed event to the personal scrutiny of

other observers. When an event can, in principle, be observed by
more than one person, socially agreed upon criteria can be
adopted to encourage observational agreement.
Admittedly, my treatment of the pragmatic advantages of
observing ‘'outer” as compared to “inner” events suggests the ex¬
istence of a real world—an assumption that most scientists make
implicitly or explicitly—but this need not lead to any conceptual
confusions in regard to research or theorizing.^
Implications of the Distinction Between Experience and Be¬

havior. A misconception common among many psychologists who
accept the distinction between phenomenal experience and public
behavior, and most do, is that the science of psychology is re¬
quired to investigate both. Accepting the distinction implies noth¬
ing more than the distinction itself, i.e., private experience and
public behavior are distinguishable, and each or both are possible
subject matters for psychology. It does not follow that both are
required to be investigated. A number of arguments could be
oflFered to encourage the investigation of one subject matter to the
exclusion or subordination of the other. One could argue: that
public behavior can be profitably analyzed by the “methods” of
natural science, or that human consciousness distinguishes man
from animals and therefore should be the fundamental subject
matter of psychology, or that both subject matters need be studied
but with the realization that one is basic to the other in the sense
that knowledge of it can lead to accurate predictions of the other,
or that one should be studied because strategic considerations
suggest that the understanding of one subject matter precedes the
comprehension of the other, and so on.
All of these arguments can be defended rationally and con¬
vincingly. One cannot, however, demand acceptance of one and
the rejection of the others on the basis of logic although at some
future date, with the benefit of hindsight, the relative fruitfulness
of the different orientations might become apparent.
1. The thorniest methodological problems created by the assumption of

an external reality are involved with the criteria for explanation (see p. 117).
Once again this analysis identifies a significant volitional de¬

cision that psychologists have to make at the very beginning of
their eflForts. This decision typically is made with little or no
thought, being more a result of educational training than personal
deliberation. Regardless of the basis of its adoption, the decision
exerts a profound effect on the character of a psychologist’s
psychology.
POSSIBLE SUBJECT MATTERS IN PSYCHOLOGY
Up to now my treatment of two possible Subject matters for psy¬

chology^ has been very general. No attempt was made to describe
in any detail these subject matters—private experience and public
behavior—beyond delineation for the purpose of mere identifica¬
tion. Private experience refers to the raw phenomenal feelings^
thoughts, and images that are directly observable only to the ex¬
periencing individual Public behavior refers to the actions of an
individual organism that can be observed by others.
To understand these concepts and speculate about their pos¬
sible interrelationships requires a more analytic treatment of pri¬
vate experience and public behavior. In the history of psychology
2. A modern cognitive psychologist might dispute the notion that psy¬

chologists are limited to a choice between two subject matters: conscious
experience and behavior. He could argue that the true subject matter of
psychology is “the functional organization of the mind, which strictly
speaking cannot be reduced to either conscious experience or behavior. The
mind’s organization, according to this argument, is not totally revealed in
conscious experience and so it could not be reflected in introspection. In a
similar vein, even though behavior can serve as an indicator of the mind s
organization, the two are not identical. The fallacy in this argument stems
from the conflation of a theoretical statement with an observational one.
“The functional organization of the mind” is not simply an observational
statement but instead is a theoretical inference in the same sense as is the
following statement: “The subject matter of psychology is the formation and
strengthening of stimulus-response associations.” “Behavior and conscious
experience ” refer directly to certain lands of observations; the functional
organization of the mind” and “stimulus-response associations” refer to the¬
oretical inferences about observational events.
the initial concern was with the mind followed by a shift of in¬
terest to behavior. This analysis begins in the opposite order first
because the problems of analyzing public behavior are simpler
than treating private experience and second because some special
problems of analyzing private experience can be illuminated by
knowledge of the methods used to describe behavior.
The Nature of Behavior
A one-sided debate has prevailed in psychology about the mean¬

ing of behavior. Critics of behaviorism insist that behavior is an
ambiguous concept and therefore cannot serve as a satisfactory
dependent variable for psychology. Behaviorists, for the most
part, ignore these criticisms and go about their task of systemati¬
cally studying behavior. If precise empirical laws involving be¬
havior as a dependent variable can be discovered, then the con¬
cept of behavior is sufficiently clear. No additional rebuttal
appears necessary to the argument that behavior cannot serve as
a dependent variable for psychology. In reaction to such a de¬
fense critics of behaviorism suggest that perceiving behavior only
in the context of specific response measures, such as the amount
of salivation or frequency of bar pressing, is at the root of the
behaviorist’s failure to appreciate the ambiguity within a general
concept of behavior that presumably applies to all forms of be¬
havior. Thus although specific illustrations of behavior can be
oflFered, no systematic, consistent definition can be formulated.
There are many problems in defining behavior, more than the
typical behaviorist realizes, but the significance and extent of
these difficulties depend on whether the concept of behavior is
treated solely as a dependent variable completely isolated from
eflForts to explain it or as something embedded in a theoretical
network.
In order to analyze the meaning of behavior it first becomes
necessary to eliminate the confusion that springs from linking be¬
havior in some fashion to conscious experience. A decision to in¬
vestigate behavior does not imply that mental events are unreal.
Nor does it suggest that behavior can be employed as a mirror to

reflect the inner workings of the mind. The adoption of behavior
as the dependent variable has absolutely no implications for the
role of conscious experience in psychology or for the relationship
between experience and behavior.^
Early behaviorists, such as Watson, equated behavior with
bodily movement. The notion of bodily action as the criterion of
behavior immediately raises the issue of intent. Is a person be¬
having when his body is moved forward by some external force
such as a fierce wind? Or is there any difference between the be¬
havior of two persons who descend a flight of stairs, one by step¬
ping down on the successive steps, the other by tripping at the
head of the stairs and falling to the bottom? If these responses are
treated solely as dependent variables, then the questions posed are
not concerned with defining behavior but instead with explaining
it. If one were to research the behavior of these two persons, defin¬
ing behavior only in terms of an end product of going from loca¬
tion A to B, one would discover that the same change in position
could be governed by different principles. Getting from the top
of a staircase to the bottom need not result from a common cause,
and no one who measures a particular form of behavior is auto¬
matically committed to the assumption that the same processes
operate whenever a common response occurs. The questions con¬
cerning intent appear trivial to the empirically oriented behavior-
ist, who would immediately recognize that the pattern of behavior
of being pushed by the wind is different from intentionally mov¬
ing forward, or that falling downstairs involves different responses
3. It is also important to recognize that the choice of behavior as the

subject matter of psychology has no implication for other methodological
decisions such as the meaning of concepts, the nature of explanation, or the
role of science in society. Historically, the selection of behavior as the de¬
pendent variable in psychology has been associated with the behaviorist
movement, which seeks to make psychology a natural science in the same
sense that biology, chemistry, and physics are. In the ensuing discussion of
the nature of behavior, the relationship between behavioral phenomena and
natural science principles will sometimes be mentioned; however, this does
not imply that the choice of behavior as the subject matter of psychology
demands a natural-science orientation.
from walking downstairs. Although these questions may be of in¬

terest to philosophers, they are of no immediate significance to
the psychologist who studies behavior with a natural-science
orientation.
A much more relevant concern with employing behavior as
a dependent variable is the ease with which behavioral descrip¬
tions can be contaminated by phenomenological interpretations.
Take, for example, the following behaviors: a boy throws a snow¬
ball at his teacher, a child writhes in pain in a dental chair. The
behavior that is being observed in the first example is the throw¬
ing of a snowball that covers a trajectory from the boy’s hand to
the teacher’s head; no intent is perceived. Similarly, in the second
example the child is squirming, but no pain is observed. The ob¬
server reads into these situations an explanation of the behavior
in line with his own past experience and phenomenology. If a boy
hits somebody with his snowball he probably intended to, and if
a child is writhing in the dental chair he is no doubt experiencing
pain. Of course, the boy could have had defective aim and hit the
wrong target, or perhaps he threw the snowball without any in¬
tent to hit anybody. The child may be writhing not from pain but
perhaps from restlessness. The initial explanations oflFered are rea¬
sonable in the sense that they are probably true. Human beings
learn to interpret behavior on the basis of their own phenomeno¬
logical experiences, which they attribute to the person whose be¬
havior they seek to explain. These experiences aid, at times, in
interpreting the world, but such attributions are neither essen¬
tially true nor necessarily equivalent to a natural-science theory.
The problem raised by contaminating a description of be¬
havior with phenomenological states is not a result of some in¬
trinsic defect with behaviorism as a methodology but rather a
consequence of the psychological difficulty of limiting descriptions
of behavior to directly observed characteristics of responses. The
difficulty is not insurmountable, but it can generate confusion and
self-deception, especially when the mentalistic descriptions of be¬
havior are uncritically accepted as explanations.
A final criticism of the decision to employ behavior as the
sole dependent variable in psychology is that it eliminates from
study such significant human events as eontemplation, fantasy,

and dreaming. What would life be without an inner life? This
question raises profound issues that are better postponed until the
eoneepts behavior, understanding, and values are elarified. First
I will turn to the analysis of the meaning of behavior. It should
be recognized that the agreement among behaviorists to study
behavior does mean that they also share other methodological
assumptions and strategic decisions as to how behavior should be
investigated and understood. My sole concern will be with be¬
havior itself as a dependent variable for the seience of psychology.
Defining Behavior
Behavior, as defined by some dietionaries and the early

behaviorists, refers to any aetivity of the organism. A moment’s
refleetion reveals the unmanageable breadth of this definition.
Numerous activities oceur simultaneously ranging from move¬
ments of neural impulses along sensory nerves to gross muscular
movements; from biochemieal ehanges in the retina to hormonal
secretions in the pituitary; from eoping with the physieal environ¬
ment to interaeting with the soeial environment. To sear eh for
laws or principles that would apply to sueh a variety of phenom¬
ena is an invitation to disaster. A more reasonable approaeh is to
seleet a cireumseribed definition of behavior that would have a
fair chance of serving as a dependent variable for a large set of
fruitful empirical relationships. Such a definition emerges not so
much from rational prejudgments as from laboratory procedures.
Conditioning, a popular methodology for most early behaviorists,
was partieularly influential in explieating the meaning of behavior.
There are basieally two kinds of behavioral measures in eon-
ditioning—direet and indirect. Classical conditioning illustrates
the former; ehanges in the aetivity of some effeetor organ, gland,
or musele, are direetly measured. Pavlov demonstrated that a
hungry dog exposed to repeated pairings of a tone with food
would finally salivate at the sound of the tone alone. This salivary
action, the conditioned response, is measured in terms of the
amount of glandular seeretion. Another example that uses a direet
measure is the classical conditioning paradigm involving the pair¬

ing of a light (conditioned stimulus) with a faradic shock (un¬
conditioned stimulus) applied through an electrode attached to
a leg of a dog. After repeated pairings the light alone elicits mus¬
cular contractions in the leg. In both cases of direct measurement
of salivation and muscular contraction, the performance of a spe¬
cific organ is observed.
Indirect measures of behavior are illustrated in the instru¬
mental conditioning procedures made famous by B. F. Skinner
with the use of the so-called Skinner Box. An animahs response
is instrumental in producing a reward, e.g., a pellet of food is
delivered after a rat presses a lever or a pigeon pecks a disk. Al¬
though muscular changes are inevitably involved in such behav¬
ior, they are nevertheless ignored in the psychologist’s observa¬
tions. Electronic gadgetry records the movement of the lever or
the disk, and such changes define the response. In other words,
the behavior of the organism is measured indirectly in terms of its
effect upon the environment.
Thus, conditioning provides two clear kinds of behavioral
measures that lend themselves to high intersubjective agreement:
a direct measure that reports effector activity and an indirect
measure that describes what a response accomplishes.^
Evaluation of the Direct and Indirect Classification Schema
Are these measures of behavior satisfactory? Satisfactory for

what? They are certainly satisfactory for providing reliable mea¬
sures of conditioning, and satisfactory also for the accumulation
of a large amount of information about conditioning. Heart rate,
pupillary dilation, and other measures of effector functioning are
4. Some behaviorists would question the validity of the distinction be¬

tween direct and indirect measures because both ultimately depend upon
their environmental effects. A muscular movement, or the amount of saliva¬
tion, is determined by its effect on some recording instrument. This reserva¬
tion ignores the point that direct measures reflect specific effector activity
whereas indirect measures do not. This distinction has both empirical and
theoretical significance.
so Psychology: A Science in Conflict
commonly used behavioral indices of attention, emotion, social

processes, etc. Indirect measures, in turn, blanket the entire spec¬
trum of experimental psyehology, e.g., key-pressing responses in
a concept identification problem, picking up and holding a snake
in desensitization therapy.
Is satisfactoriness, in the sense of being intersubjeetively re¬
liable, the only standard by which direct and indirect measures
of behavior can be judged? Two other criteria—inclusiveness and
fruitfulness—might be considered in evaluating the adequacy of
behavior measures. As used here, an inclusive classification
schema would eneompass the entire range of behavioral events;
no form of behavior would be eliminated from consideration. It
might be argued that meaningful behavior is excluded from the
direct-indirect behavior classification schema.
Inclusiveness of the Dual-Classification Schema:

Meaningful Behavior
What does “meaningful behavior” mean? Something is mean¬
ingful if it makes sense, if it can be translated into some language
that is comprehensible. Behavior can be interpreted as meaning¬
ful in several different ways.
Operational Meaning Versus Theoretical Meaning. Perhaps

the most important distinction in analyzing the meaning of mean¬
ingful behavior is that between operational and theoretical mean¬
ing. This distinction can be clarified by reviewing the criticism
leveled against associationistic theories of learning by Gestalt
psychologists who argued that the theories were based upon arti¬
ficial tasks that lacked meaning to the experimental subjects.
Thorndike's research on animal intelligence (1898) is one ex¬
ample. To study the course of learning of eats, dogs, and chick¬
ens, Thorndike designed a puzzle box that enclosed an animal
and gave it full view of food located on the outside. The animal
could escape and obtain the food by making an instrumental
response such as pulling a loop, pressing a lever, or stepping on a
platform. Aeeording to Thorndike, the subject learned to make
The Data Base and Subject Matter of Psychology: 1 SI
the correct response “by trial and error with accidental success”
(1898, p. 105). A cat, for example, tried a variety of responses
such as pressing against the door, sticking its paw through the
slats, and other “erroneous” acts that failed to open the door.
Then by “accident” it would emit the appropriate instrumental
response, escape, and consume the food.
Gestalt psychologists argued that there was no meaningful
relationship between the act of pressing the platform and escap¬
ing from the box. There was no more reason to press on the
platform than to insert the paw through the slats or lie down.
Instead, meaningful behavior depends upon being able to per¬
ceive the rational relationship between the structure of the prob¬
lem and the behavior required to solve it. Problem solving de¬
pends upon understanding, not “accidental success.” An example
of a meaningful problem is the task employed by Kohler (1925)
in studying insight. In one of his experiments a banana was sus¬
pended from the top of a cage out of a chimpanzee’s reach. Sev¬
eral boxes were scattered around the floor. The animal repeatedly
jumped for the fruit without success. Eventually these futile leaps
ceased, giving way to restless pacing. Then according to the re¬
port, the pacing stopped, and after some apparent contemplation
the animal pushed the boxes below the banana and stacked them,
enabling him to climb to the top and grab the fruit.
Kohler used the term insight to describe the problem-solving
behavior of his chimpanzees. He assumed that insight was a con¬
sequence of a sudden change in the chimpanzee’s perception of
the problem. Instead of seeing the boxes and banana as isolated
objects, the chimpanzee suddenly perceived them as related parts
of the problem. As a result of insight (perceptual reorganization)
problem solution occurred.
Gestalt psychologists essentially argued that a qualitative dif¬
ference existed between learning based upon an arbitrary asso¬
ciation and learning resulting from understanding rational rela¬
tionships. An organism from his knowledge of the world had no
reason to expect that the pressing of the platform would lead to
escape, but he could anticipate that stacking the boxes would
enable him to grab the bananas. The implication of this position
is that a theory of learning based upon the formation of arbitrary

associations that occur in puzzle boxes, and conditioning as well,
would be inadequate to the task of explaining learning by under¬
standing.
Although the Gestalt position has theoretical merit, it does
not follow that methods of measuring behavior based upon
“meaningless” behavior cannot be applied to the study of mean¬
ingful” behavior. Such a conclusion results from conflating the
issue of the validity of the theoretical distinction between two
kinds of learning with the methodological issue of the inclusive¬
ness of the direct-indirect classification of behavior. The dual¬
classification schema refers to observable'^ characteristics of be¬
havior (operational), not to causal factors (theoretical). There is
no epistemological difference between observing a cat pressing a
platform or a chimpanzee stacking a set of boxes even though the
principles governing each form of behavior may differ. Meaning¬
fulness, in the context discussed, is not an observable property
of behavior but instead is an interpretation of the behavior that is
observed. In other words, the insight of Kohler’s apes was not in
the responses of stacking the boxes and grabbing the banana but
in the theory proposed to explain that behavior.
If the theoretical distinction between meaningful and mean¬
ingless processes in learning is made, then methods of measuring
meaningfulness can be developed. It is important to recognize
that the operational measures of meaningfulness are not equiva¬
lent to the theoretical processes that are assumed to influence
them. Consider the case of clustering in free recall (Bousfield,
1953). Clustering can be demonstrated by presenting in random
order sixty words, each word from one of four different concep¬
tual categories, (e.g., animals, vegetables, men’s names, and pro¬
fessions), to subjects. If immediately thereafter the subjects
would list as many of the sixty words as they could recall, their
lists would fall into clusters of concept-related words (e.g., lion,
tiger, giraffe, pig; corn, pea, carrot; Dick, Jack, Paul; etc.). As¬
suming that the input order of the words from the different con¬
ceptual categories is random, the degree of clustering is deter¬
mined by how closely the order of the subject’s recall coincides
with the distinct conceptual categories. The measure of clustering
The Data Base and Subject Matter of Psychology: 1 SS
is a dependent variable that presumably reflects a rational prin¬

ciple that operates to determine memory. In the case of cluster¬
ing the rational principle is the linguistic organization of words
into conceptual categories. A subject, however, could behave
rationally in a manner that would reduce the amount of cluster¬
ing. For example, he might recall the sequence ‘"plumber, Adam,
garlic” because the name of his plumber is Adam whose breath
frequently smells of garlic. For this subject the idiosyncratic ra¬
tional principle underlying the sequence of his recall of these
three words exerts more of an effect in determining this sequence
of behavior than does the rational principle of clustering. This
does not imply that clustering is a defective measure. It only
emphasizes the distinction between clustering as an operationally
defined attribute of behavior and as one of many theoretical pro¬
cesses that organize information into meaningful patterns.
Does our analysis of meaningful behavior extend beyond the
laboratory to behavior in everyday life? Can the direct-indirect
classification be applied, for example, to the behavior of a para¬
noid? His meaningful, but unjustified, paranoiac reactions can be
measured indirectly by the discrepancy between his persecutory
responses (e.g., hiding because “X is trying to kill me”) and the
objective social situation (e.g., X is his therapist who is at¬
tempting to help him). Direct measures may be possible in the
future if, and when, neurophysiological correlates of paranoia are
discovered.
In sum, the analysis offered distinguishes between opera¬
tional and theoretical definitions of ""meaningful behavior.” This
advances the argument that the direct-indirect behavioral classifi¬
cation is sufficiently broad to include meaningful behavior of the
sort that is an expression of rational processes. However, it is im¬
portant to recognize that no guarantee can be offered that mea¬
suring techniques designed to reflect rational processes will yield
fruitful empirical relationships and lead to satisfactory theories.
Scaling techniques to measure behavior are always experimental
and tentative in nature, and their usefulness or ""validity” cannot
be prejudged. The value of such methods can only be discerned
in relation to their empirical and theoretical yield.
The claim that the dual classification system is adequate for
meaningful behavior does not imply the absence of practical and

technical difficulties. Consider the question of when does a child
emit his first word. A word in this context is not merely a sound;
it possesses referential meaning. When does the sound mama,
which is a common babble, first acquire meaning? The absence
of criteria that would discriminate between the babble mama and
the first occurrence of the word mama does not, however, reflect
any intrinsic limitations of indirect measures of behavior; instead,
it illustrates one of the many complexities of language develop-
ent. In principle, there is no reason to believe that the discrimina¬
tion between babbling and word formation can never be made,
although more complicated measures of behavior (perhaps com¬
binations and sequences of direct and indirect measures) may
have to be developed. In addition, it may be discovered that the
transition between the babble mama and the word mama is
gradual, not discontinuous.
Phenomenological (Subjective) Meaning. The meaning of

meaningful behavior that has just been described refers to a
theoretical position that postulates the operation of certain ra¬
tional processes that govern behavior. These rational processes, if
they are to meet the demands of the dual classification system,
must yield predictions or estimates of behavior that are intersub-
jectively defined. Another possible interpretation of meaningful
behavior stems not from the perspective of the observing scientist
but instead from that of the organism who is behaving. This kind
of meaning emphasizes the meaning a situation has for the organ¬
ism. It can be labeled a phenomenological or subjective meaning,
and like theoretical meaning, can be explicated by reference to
the Gestalt interpretation of insight. Recall that Gestalt psycholo¬
gists suggested that a perceptual reorganization is required to
achieve insight; the essential elements of the problem have to be
perceived as related, not isolated, events. One possible interpre¬
tation of this assumption is that only by understanding the in¬
ternal viewpoint of an organism is it ever possible to understand
its actions. This position insists that the conscious experience of
the organism cannot be bypassed if behavior is to be fully com¬
prehended.
Although at first glance phenomenological (subjective)

meaning appears to be excluded from behavioristic psychology, a
moment’s reflection would suggest otherwise. The central issue is
the manner in which the phenomenological meaning is ascer¬
tained. Behavioristie psychology accepts the distinction between
intrasubjective and intersubjective evidence and insists that psy-
ehologieal data meets the requirements of the latter. If phenome¬
nological meaning is defined intrasubjectively, unavailable to the
inspeetion of others, then it fails to meet the demands of the
direct-indireet classification schema and therefore is excluded
from behavioristic psychology. As previously noted, behavioristic
psychology need not deny the existence of conscious experience,
but it does recognize methodologieal limitations associated with
direct investigations of the mind. If phenomenological meaning is
viewed from a different methodological perspective, not from
direct raw experience but instead as an inferred ''state of mind,”
then its admissibility into a behavioral psychology based upon
pubhc observation becomes possible and, as some would argue,
desirable. Verbal reports of inner states are the most likely candi¬
dates, but certainly not the only, upon which to infer a model of
the mind. Such reports of a person’s inner state are epistemologi¬
cally no different from any other kind of publicly observable be¬
havior, but it must be recognized that the verbal report, not the
conscious experience, is the basic sense observation.
The position of raw phenomenal experienee in a behavioral
psyehology that demands publicly observable events as the foun¬
dation of its knowledge is a souree of much controversy as well as
confusion in the history of psychology. Too often behaviorism is
eonceived of as a monolithic orientation instead of as a group of
methodological positions that share common principles. The fun¬
damental assumption is that the subject matter of psychology is
publicly observed behavior. Adopting that position leaves a wide
variety of options in regard to the treatment of phenomenal ex¬
perience. At one extreme would be an epiphenomenal behavior¬
ism that rejects as unnecessary all inferences from mental pro-
eesses; at the other extreme would be a subjective behaviorism
that strives to explain behavior in an objeetive fashion on the
basis of a model of the mind. In between would be a variety of
forms of pragmatic behaviorism that would infer phenomenal

processes only when the demands for explaining a particular kind
of behavior requires it.
Transcendental Meaning. One kind of ‘meaningful” behavior

remains to be discussed, transcendental meaning. This concept re¬
fers to the general notion that life has a meaning: human beings
individually and collectively strive toward certain goals and seek
appropriate standards of conduct. Such ambiguous and popular
phrases as “life’s meaning,” “self-fulfillment,” “meaningful rela¬
tionship,” and others are used to express the idea that life has a
purpose. Transcendental meaning, in the above sense, expresses a
common theme in most, if not all, religious doctrines. Within this
context, the statements implied by transcendental meaning are
not falsifiable in that empirical evidence can, in principle, be ob¬
tained to deny their validity.
Transcendental meaning is obviously antithetical to a behav-
iorist psychology, which is designed to employ natural-science
methods in order to arrive at psychological truths. However, like
phenomenological meaning, it is important to recognize that some
feature of transcendental meaning might be able to be incorpo¬
rated within an empirically oriented behaviorism. To do this, one
must understand that a behaviorist methodology is not wedded to
a strong environmental or hereditary position; although leading
behaviorists, for social and personal reasons, have tended to align
themselves with the former. When the problem of transcendental
meaning is removed from its ethical and spiritual context, it can
be interpreted in terms of genetic predispositions in regard to
human motivation and social behavior. If the idea is accepted
that human behavior is not completely malleable, then one can
suggest that certain classes of behavior are more likely to occur
and be maintained than others. Thus the empirical question can
be asked as to whether certain ethical positions and forms of
social organization are more in tune with the genetic preprogram¬
ming of human behavior than are others. Presumably, if such
questions are stripped of ethical and cultural biases—and one can¬
not overestimate the difficulty of accomplishing this—then em-
The Data Base and Subject Matter of Psychology: 1 S7
pirical issues will remain, the implications of which can be

checked by investigations employing direct and indirect be¬
havioral measures.
A Closing Comment on Inclusiveness. In concluding this dis¬

cussion of the inclusiveness of the dual-classification schema, it
should be noted that although direct and indirect measures of be¬
havior emerged from the early behaviorists’ devotion to condi¬
tioning, there are no a priori reasons to believe that such kinds of
measures are inappropriate for investigating any other form of
behavior. One could suggest that measures of behavior that are
used to study ‘‘meaningful behavior” should emanate from theo¬
retical considerations instead of being borrowed from a research
problem that was initially conceived as being devoid of cognitive
involvement. Such a criticism is based upon the faulty notion
that the dual-classification system represents an a priori prescrip¬
tion as to how the initial work in conditioning could, and should,
be conducted. In actual fact, the direct and indirect conceptions
of behavior did not emerge from rational considerations but in¬
stead from an ad hoc methodological analysis of how behavior
was defined in conditioning. The fact that the dual-classification
schema is an effective representation of behavior in conditioning
in no way limits its applicability to other forms of behavior.
Fruitfulness of the Dual-Classification System
The answer to the question about whether direct and in¬

direct measures of behavior are fruitful depends upon a criterion
of fruitfulness. Various orientations in contemporary psychology
do not share a common goal for psychological inquiry. Conse¬
quently, before the query about fruitfulness can be answered, an
appropriate yardstick must be adopted. Such yardsticks will be
the topic of discussion in Chapters 4 and 5. For the time being,
however, the conclusion can be drawn that direct and indirect
measures have served the aims of two different streams of be¬
havioristic psychology, the one concerned with formulating ab¬
stract theories of behavior, the other with describing behavior
with an eye to control it. This, of course, does not rule out the pos¬
sibility that a more suitable and fruitful behavioral classification
than the direct-indirect system can be developed in the future.
Some Final Comments about Behavioral Measures. Two addi¬

tional points must be made. The first concerns the relationship
between direct and indirect measures of behavior, the second
concerns a possible misinterpretation of the sequence of topics of
this methodological analysis.
Distinguishing between two forms of measuring behavior im¬
mediately raises the question of the relationship between the
two and the possibility that one is more basic than the other.
These are not methodological issues but empirical and theoretical
ones. The empirical relationship between direct and indirect mea¬
sures can only be revealed by empirical investigations that moni¬
tor changes in both types of responses. For example, in instru¬
mental conditioning it would be possible to discover the relation¬
ship, as conditioning progresses, between muscular movements
and a lever-pressing response. Which measure is more basic is
purely a theoretical issue assuming that one accepts as the goal
of scientific inquiry the formulation of a satisfactory theory. The
more basic measure, for any empirical phenomenon, is the one
that is employed in a theory that can explain the empirical laws
involving the other measure. For the present, however, the ques¬
tion of which is more basic is not an immediate issue and would
be best sidetracked until it becomes one. It would seem that cer¬
tain empirical problems are more easily investigated with one
than with the other measure. Pragmatic considerations often
serve as the best guide for methodological decisions.
The reader may get the impression that all psychologists are
confronted with an orderly sequence of methodological issues
beginning with a choice of subject matters, proceeding to a selec¬
tion of a suitable dependent variable, and followed by the adop¬
tion of a criterion of understanding, and so on. The intent of my
analysis is not to construct a methodological maze with a prede¬
termined succession of choice points that must be traversed by
all psychologists. That would be misleading. Methodological
questions are highly interrelated, and the order of making deei-

sions ean vary from psyehologist to psyehologist. In some cases,
all the significant decisions are made simultaneously, while in
other cases some decisions are isolated from others. My aim is not
to prescribe the order in which the significant methodological de¬
cisions should be made. Instead, my concern is to rationally re¬
construct psychology in a systematic fashion so that the funda¬
mental decisions are laid bare. It is conceivable that a similar
analytic eflFort could be made that would proceed with an en¬
tirely different order of issues. Starting from the data base of
psychology seems to be an appropriate beginning.
The Nature of Conseiousness
The behaviorist revolution in America was successful in directing

psychologists’ attention from the study of the mind to behavior.
Although the study of consciousness (sometimes referred to as
phenomenology or experientialism) was discouraged, it was never
abandoned. A topic that permeates human existence and is of
central concern to both the worlds of art and religion inevitably
had to attract the interest of some psychologists.
Interest in the systematic study of consciousness has waxed
and waned over the years. It was given a boost more than a
decade ago when an internal immigration to the examination of
the mind took place in an attempt to escape from the apparently,
or truly, insoluble problems of the ‘"real” world. Fascination with
“mind-expanding” drugs and a reawakened interest in religion
were expressions of this renewed interest in consciousness. At the
same time, modern behaviorists, freed of the need to polemicize
the conflict between subjective experience and public behavior,
began to view the problem of the subject matter of psychology
more tolerantly. The issue shifted from whether conscious ex¬
perience should be investigated to whether it could be investi¬
gated without sacrificing a natural-science approach. This is the
issue that will now be analyzed, not to legislate the fundamental
ingredients of science in order to excommunicate “unscientific”
approaches, but rather to identify conflicting methodological posi¬

tions that establish different standards for evaluating knowledge.
As was the ease for behavior, consciousness will be treated
as an entity in itself. My first eoncern will be with pure phenome¬
nology. Then the possible relationships between conscious experi¬
ence and behavior or neurophysiology will be explored.
Two related questions come to the fore when eontemplating
the study of mind. Can conscious experience, whieh up to this
point has been dealt with in a global manner, be analyzed into
some fundamental seheme analogous in inclusiveness to direct
and indirect measures of behavior? How can the mind be sys¬
tematically studied if a given mental event is accessible to only
one observer?
Options in Experiential Research
My qualifications to cope with the problems of describing

conscious experience may be immediately challenged on the basis
that for sueh a task only an experieneed phenomenologist is com¬
petent. However, my aim is not to propose a phenomenological
description that will lay bare the essential eharacteristics of the
mind. That is an empirieal problem to be solved by phenomeno-
logieal research. Instead, my purpose is to examine the problem
of deseribing eonseious experience in some systematic fashion. If
the issues are not strictly parochial then presumably they ean be
analyzed by a general psychologist.
Numerous options confront the investigator who initiates ex¬
periential researeh. Not only can he look at the mind in different
ways, but he also ean sense different mental events. The phenom¬
enologist can insist that the perception of mental events demands
a trained observer who ean impartially sense the evidenee of im¬
mediate experienee. Or he can adopt an opposing view by “as¬
suming” that eonseiousness ean only be aceurately observed in a
naive and relaxed fashion in the absence of any presuppositions.
He ean divoree his self-observation completely from physical
stimulation or systematically attempt to relate the two. The
mind ean be conceived as eonsisting of basie elements of experi-
ences, such as sensations, images, and feelings; or as an activity in

which all psychic events eontain an aet and an objeet; or as an
undifferentiated, inehoate mass of experiences that can be de-
seribed and interpreted both in an analytic and holistic fashion.
This brief enumeration of possible methods and conceivable
mental events demonstrates, as any history of psyehology (e.g..
Boring, 1950) will show, that different frames of reference, both
in regard to the method of observation and the kinds of mental
events observed, are possible.
What are the implications of these differences? Will these
differences inevitably produee conflieting pictures of the mind or
merely varying views that can potentially be eombined into a
single integrated portrait? The answer to these questions will de¬
pend upon the criteria adopted to evaluate the validity of various
conceptions of the mind.
Phenomenology or Phenomenologies?
In judging the adequacy of different conceptions of eon-

sciousness one can pursue a dialectieal approaeh and conelude
that a universal agreement about the nature of the mind is in
principle beyond reach. Three completely different reasons for
this conelusion are possible. First, everybody is endowed with a
unique mind, and therefore, the formulation of general principles
that apply to all minds is impossible. Second, phenomenal experi-
enee is not independent of the method used for observation; and
consequently, different methods of phenomenologieal research
must yield different deseriptions. Third, the language used to
deseribe mental experience will influence its eontent, and even if
a common method is utilized, descriptions employing different
languages will yield fundamentally different pietures of the mind.
Arguments that invoke the notion that some goal is in prin¬
ciple unaehievable usually leaves this writer cold because of their
strong commitments to untested prejudgments and to the limits
they place upon human ingenuity. Perhaps such restrictive posi¬
tions should be reserved for obvious logical impossibilities sueh
as the eonstruetion of a triangular rectangle. A more reasonable
42 Psychology: A Science in Conjlict
way of judging the possibility of formulating a single valid eon-

eeption of mental events would be to base one’s inference on his¬
torical evidence. When this is done the conclusion drawn is that a
universally agreed-upon representation of the mind, although not
impossible, is highly improbable.
Such a position would accept the notion that general mental
principles may operate but recognize the overwhelming obstacles
that are in the path of achieving consensual agreement about
their identification. The conceptual framework one adopts to ex¬
amine the mind influences the verbal description of consciousness,
and consequently divergent views will rpult as long as different
presuppositions are adopted. In opposition to such a position one
can argue that observing the mind, or any other phenomenon, de¬
mands some prejudgments in order to make empirical observa¬
tions. But when such prejudgments are considered as heuristic
rather than final positions, they can be modified, and even aban¬
doned, in the face of evidence. If observations of mental events are
pursued with diligence, one would hope that the empirical results
will ultimately reshape the initial prejudgements in the direction
of developing methods of observations that will be capable of
achieving consensual agreement.
Does history offer any optimism for the claim that an open-
minded, diligent pursuit of the problems of the mind will yield a
picture of the mind capable of achieving consensual agreement?
Although past failures cannot rule out future successes, prior con¬
troversies about the nature of the mind do not leave much room
for optimism that the study of the mind as an entity in itself will
ultimately yield any generally agreed-upon interpretation.
Titchener proclaimed that his sensationalistic orientation to
the study of the mind was ''simply an heuristic principle, ac¬
cepted and applied for what it is worth in the search for the
mental elements” (1909, p. 34). Although suggesting an open-
minded view, the important restriction of the methodological
"search for the mental elements” should not be overlooked. In the
controversies that structuralists in general and Titchener in par¬
ticular engaged in, disagreements with their own positions were
attributed to faulty methods of self-observation by their critics.
This does not mean that the structuralists were incapable of
changing their descriptions in light of empirical evidence. Modi¬

fications of descriptions of the mind occurred frequently, and
disagreements among structuralists were common—e.g., Nafe
(1927) rejected Titchener s distinction between feelings and sen-
sations, concluding that the former mental event could be re¬
duced to the latter. But such modifications and disagreements did
not extend beyond the methodological commitment to analyze
consciousness into basic elements or processes.
It will be informative to briefly review some controversies
about the nature of the mind in order to achieve an appropriate
perspective for judging the source of such disputes. Brentano
conceptualized consciousness as consisting of psychical acts in¬
stead of the elementary sensations of the structuralists.^ An im¬
plication of Brentano’s view is that psychical events can be ana¬
lyzed into act and content, a view that did not jibe with the
introspective reports of others. Whereas Brentano maintained all
psychical acts had to have a content, others maintained that af¬
fective acts could be experienced independent of any content. To
be specific, Brentano felt you had to be happy about something;
others (e.g., Kiilpe, 1909) concluded you can just experience
happiness. Titchener suggested that Brentano’s conception of the
mind was generated by an inappropriate methodology: “The act-
and-content psychology . . . [is] a psychology not of observation
but of reflection” (Titchener, 1909, p. 53). That is, Brentano was
not accurately observing conscious experience; his reports were
contaminated by an inappropriate perspective.
Another disagreement exhibited a similar pattern; the partici-
5. AlthjDugh the distinction between conceptualizing consciousness as

sensations or as acts is embedded in subtle and complex philosophical and
phenomenological issues (Boring, 1950), I feel that this distinction repre¬
sents a difference similar to that between stimulus and response; sensations
are psychical stimuli that represent the mental environment, while acts are
psychical responses that refer to the activity of the mind. For a sophisticated
behaviorist turned naive phenomenologist, it would appear that a complete
understanding of consciousness requires not only an adequate description of
the content of mental experience but also a comprehension of the principles
governing the production of these sensations (Kendler, 1970; Skinner, 1969).
In other words, there are two aspects of consciousness: its content and its
activity.
pants explained away conflicting phenomenological descriptions

by attributing them to defective methods of self-observation.
Boring recounts the vitriolic controversy between Wundt and
Stumpf on the topic of tonal distance.
The clash seemed to have arisen because Stumpf leaned heavily
upon his own musical sophistication, while Wundt relied on the
laboratory results with apparatus and the psychophysical methods.
Whatever is obtained under unprejudiced, carefully controlled
experimental conditions must be right, Wundt virtually said. If
the laboratory yields results that are obviously contrary to expert
musical experience, they must be wrong, was Stumpf’s rejoinder
(Boring, 1950, p. 365).
William James, who some consider the greatest introspector
of all times,® noted difficulties in observing the transition between
successive experiences in the stream of thought. “When the rate
is slow, we are aware of the object of our thought in a compara¬
tively restful and stable way. When rapid, we are aware of a
passage, a relation, a transition from it, or between it and some¬
thing else.” Consciousness, “like a bird’s life, seems to be made
up of an alternation of flights and perchings.” James thus distin¬
guishes between the substantive and transitive parts of the stream
of thought.
Now it is very difficult, introspectively, to see the transitive parts
for what they really are . . . The rush of the thought is so head¬
long that it almost brings us up at the conclusion before we can
arrest it . . . The attempt at introspective analysis in these cases
is in fact like seizing a spinning top to catch its motion, or trying
to turn up the gas quickly enough to see how the darkness looks
(1890, pp. 243-44).
The fleeting quality of the transition that James describes
does not reflect what truly is in consciousness, Titchener argued,
but instead defective observation. “If James had looked away
6. This claim, in itself, highlights fundamental difficulties in interpreting

phenomenological research. What is James’ presumed superiority a result of?
An abnormal sensitivity to conscious events, an ability to have more subtle
experiences than others, or an exceptional writing style that encourages an
empathic response in the reader? Phenomenological research would have
difficulty in disentangling these various possibilities.
from ‘awareness of object’ and ‘awareness of relation’ and had

looked toward the actual content of consciousness, we should not
have heard of the top and the gas jet” (1909, p. 29).
The discussion about the nature of the mind will be concluded
by referring to the debate about “imageless thought” (Woodworth,
1938). Several experimenters independently reported that some
subjects, when responding to simple word-association tasks or
when interpreting Nietzsche’s aphorisms, failed to observe any
images in their consciousness. Wundt considered these results
to be invalid; images do occur while thinking. Imageless thought,
for Wundt, was an outcome of a defective methodology; one can¬
not observe thought while thinking. Nevertheless, thought ex¬
periments were conducted in Titchener’s laboratory. The results
indicated the presence of images in consciousness. Titchener con¬
cluded that imageless thought was a consequence of defective in¬
trospection. Woodworth, in reviewing the controversy, suggested
that “the whole question may well be shelved as permanently de¬
batable and insoluble” (1938, p. 788). The controversy about
imageless thought led to a disenchantment with the introspective
method and encouraged the emergence of behaviorism.
The Source of Unresolved Phenomenological Controversies.

One could add countless other examples of unresolved contro¬
versies about reports of self-observations. History demonstrates
that no general phenomenological method was ever adopted that
could resolve these controversies for the entire psychological
community. In light of this historical evidence one can conclude
that pure phenomenology, unlike natural science, can never yield
universally valid conclusions. That is, any proposed picture of the
mind will never be accepted as “universally” valid in the same
sense that evolutionary theory or the theory of relativity is. The
justification for this conclusion could be either a radical subjec¬
tivist assumption that every phenomenological description is valid
regardless of its incompatibility with others or a methodological
position that decrees that every method employed to observe con¬
sciousness is based on some presuppositions and these presup¬
positions will inevitably influence what will be observed.
In rebuttal, the argument can be advanced that this pessi-
mism about phenomenology is based upon a fundamental mis¬

understanding. A universally valid description of human con¬
sciousness is a false goal because human minds do differ in their
sensitivities, content, feelings, etc., and therefore differences in
phenomenological descriptions are bound to occur and persist
just as do individual differences in behavior.
In support of this line of argument one can offer the results
of psychophysical investigations that meet the methodological
standards of a behavioristic psychology. They reveal that indi¬
viduals do differ in their sensory capacities. Those with acute
color vision can form a language community (Koch, 1964) and
discuss among themselves esoteric aspect's of color sensations and
delicate differences in hue that would possess no meaning for
others. Phenomenology within this context, it can be argued, is
no different from behavioristic psychology, which obviously rec¬
ognizes individual differences.
Although this argument, at first glance, appears reasonable,
it overlooks an important difference between pure phenomenol¬
ogy and behavioristic psychology. Whereas the abnormal color
sensitivities can be publicly displayed by psychophysical meth¬
ods, experiential sensitivities (e.g., beauty) that have not been
correlated to publicly observed stimulus events cannot. This does
not automatically mean that such sensitivities do not operate, but
one cannot be sure that they do. Is the “hypersensitive” art critic
at a gallery exhibition who is reporting rhapsodic experiences in
fact having them or indulging in social gamesmanship encour¬
aged by prior social reinforcements? In sum, unresolvable con¬
troversies in experiential psychology do not possess the same
epistemological status as individual differences do in behavioral
psychology; they fail to meet the standards of intersubjective
agreement.
Another possible interpretation of the phenomenological dis¬
putes is that they represent unresolved theoretical differences,
which are certainly not peculiar to phenomenology since they
occur with great frequency in behavioral psychology. One can
point to the controversies of latent learning, perceptual defense,
one-trial learning, continuity versus noncontinuity conceptions of
discrimination learning, racial diflFerences in intelligence, operant

conditioning of verbal responses, the organization of attitudes,
and many others, all of which have been contested within be¬
havioral psychology and all of which have failed to yield an un¬
qualified conclusion supported by consensual agreement of the
scientific community. In rebuttal, a behavioral psychologist would
argue that theoretical differences are not methodologically equiv¬
alent to disputes about basic observational events. The latent-
learning controversy, in which I was involved during the early
period of my research career, entailed a host of theoretical issues
about which the competing theories (of animal learning), the
Hull-Spence S-R model and the Tolman cognitive map formula¬
tion, were not articulated with sufiBcient clarity to yield unam¬
biguous implications for all of the empirical problems studied.
Some of the concepts that proved to be unclear were zero rein¬
forcement, drive satiation, perceptual awareness, and fractional
anticipatory goal responses. The point at issue is whether the
theoretical disputes in behavioral psychology spring from dis¬
agreements about basic empirical observations, e.g., whether a
rat turned left or right in a T-maze or entered or did not enter a
blind alley in a multi-unit maze. If these observations are the
source of the disagreements, as they are in phenomenological
psychology (e.g., does one experience images when responding
to the word chair or grass in a simple word-association task?),
then the conclusion can be drawn that the unresolvable disputes
in phenomenological psychology do not set off this discipline from
behavioral psychology.
In the larger context of the philosophy of science, we have
become enmeshed in two controversial issues: whether observa¬
tional terms can be distinguished from theoretical terms and
whether we can ever know if one theory is true and another is
false. These are exceedingly controversial and complex issues in
the philosophy of science. Consistent with the plan of this book
I will avoid the philosophical problems that are not immediately
relevant to the problems under scrutiny. This does not mean that
psychologists do not have much to learn from various concep¬
tions of the philosophy of science, past and present, but instead
it means that for the active researcher and theorist, viewing the
methodological issues within a restricted psychological frame¬
work may have special advantages/
Observational Purity
My present concern will be with the problem of whether a

scientist’s perceptions are immaculate, whether his observations
are free of influences from his experiences and preconceptions
(the issue of the ‘'truth” of theories will be treated in Chap¬
ters 4 and 5). To cope with this general problem effectively it
must be analyzed into separate issues.‘This can be done by ex¬
amining the statement that denies that perceptions can be im¬
maculate: “Observer and observed are not separable” (Kessel,
1969, p. 1002).
There is no question that an observer’s experience and pre¬
conceptions will determine what is observed. This can be accom¬
plished in three major ways: (1) observers can observe different
situations; (2) observers can attend to different events in the
same situation; and (3) their preconceptions can determine what
they observe when attending to the same event. Let me now
analyze each of these alternatives and then examine their impli¬
cations for the problem of achieving consensual agreement about
basic observational evidence.
It may appear trivial to note that a scientist’s interest will
determine what he will observe. But it does represent an exam-
7. A common ploy in some current methodological discussions (e.g.,

Koch, 1964) is to disparage a particular psychological orientation (e.g., be¬
haviorism) by claiming that it is based upon an inappropriate philosophical
position (e.g., logical positivism) that has long been rejected by the philo¬
sophical community. The implications that flow from such a criticism are:
(1) if a particular conception of science is not completely right, it must be
completely wrong and hence possesses no value for psychology; (2) a
unanimity of opinion about specific philosophical positions prevails among
philosophers of science; (3) different philosophies of science and method¬
ological orientations in psychology represent fixed positions resistant to all
change; (4) some philosophical paradigm is appropriate to all sciences at
all stages of development; and (5) psychologists would be well-advised to
seek Good Housekeeping-type Seals of Approval for their methodological
positions from philosophers of science who are currently in vogue.
pie of how observations are influenced by the observer. The ob¬

server s interest patterns are basic to the division of psychology
into different research areas: learning, cognition, personality, so¬
cial behavior, and so on. They are responsible for what events a
psychologist observes: pigeons pecking disks, children choosing
cards in a visual-learning discrimination task, patients ventilat¬
ing their emotional hang-ups in psychotherapeutic sessions, and
social groups behaving in natural settings.
Observers in the same situation can also observe different
events. This point has been demonstrated repeatedly in discrimi¬
nation-learning problems in which organisms are required to
choose one of two visual displays. Such learning can be acceler¬
ated or retarded depending upon whether or not the subject is
observing the features of the stimulus displays that are correlated
with reinforcement (e.g., Ehrenfreund, 1948). In a similar vein it
has been demonstrated that older children perform in a superior
manner to younger children in a perceptual-learning task because
they scan the visual displays more systematically and hence are
better able to detect the important similarities and differences
(Vurpillot, 1968).
Now if we shift our perspective from that of an experimental
subject observing stimulus displays to that of an experimenter
observing behavior, we find similar processes at work. In the
same experimental situation different experimenters can be ob¬
serving different events. A noted example in the history of learn¬
ing theory is the different observations made in instrumental bar¬
pressing conditioning situations by a '‘Skinnerian” and a "Guth-
rien”—the former observing whether the bar is depressed and the
latter, the subject’s pattern of movement when depressing it. This
difference in what is observed is an expression of the theoretical
preconceptions of the observer as to what are the significant data.
I learned early in my research career that what the experi¬
menter observes, or fails to observe, can be at the heart of a
theoretical dispute. One aspect of the latent learning controversy,
which raged during the 1940s and early 50s, illustrates this point.
The empirical question at issue was whether rats who traversed
a multiunit maze but who were not fed when they reached the
goal box learned as much about the maze pattern as those who
were fed. The initial California latent learning studies (Blodgett,

1929; Tolman & Honzik, 1931) oflFered an affirmative response;
when food was introdueed into the goal box for the previously
nonfed subjeets, their maze performanee (blind alley entranees)
on the next trial matehed those who had eonsistently reeeived
food. Sueh evidence was interpreted to be consistent with Tol-
man’s cognitive theory of animal learning, which postulated that
reward was not necessary for learning. The subjects not fed in
the goal box learned as much about the structure of the maze as
did the food-rewarded group; the inferior performance of the
former during the time they were not fed was due to the absence
of an incentive to utilize their knowledge. '
An underlying assumption of this analysis was that the equal
performance of blind-alley entrances exhibited by the two groups
following the introduction of food for the previously nonfed ani¬
mals reflected equal amounts of learning (Kendler, 1952). 1 en¬
couraged a doctoral student, Joseph Kanner, to test this assump¬
tion. His first chore was to replicate the California latent learning
phenomenon, no easy task considering that other attempts had
failed (MacCorquodale & Meehl, 1948; Reynolds, 1950). After
several failures Kanner (1954) nevertheless was able to discover
some of the fundamental variables in the California latent learn¬
ing phenomenon and was then able to replicate the original find¬
ing. Of great interest were the striking differences observed, aside
from blind-alley entrances, in the maze behavior of the food and
nonfood groups, the latter exhibiting marked signs of ‘Tearful¬
ness” suggesting that the motivational incentive conditions of the
two groups were strikingly different. Additional research (Kan¬
ner, 1958) indicated that contrary to Tolman's original interpre¬
tation the two groups had not acquired “equal knowledge” about
the structure of the maze.^
8. Kanner’s study (1958) had little impact upon the latent learning
controversy because at the time of its publication, interest in the problem
had petered out as a consequence of the exhaustion of the combatants. As a
result, a study for which I had predicted instant notoriety largely went un¬
noticed. Kanner still has 273 reprints of the 300 I encouraged him to
purchase.
The point of the above is not to revive interest in the latent

learning controversy (although it is still relevant to some funda¬
mental theoretical issues) nor to argue in favor of one position
over the other. All the issues in the controversy were not resolved
because the competing theoretical formulations were much vaguer
than initially thought and some of the participants failed to dis¬
tinguish between separate theoretical problems. The latent learn¬
ing data and the attempts to interpret them revealed the stark
inadequacies of the competing theoretical structures. The psy¬
chological community, rightly or wrongly, finally decided that
possible payoffs in new data and theoretical clarification did not
justify continued efforts. In short, the controversy died at its own
hands.
A close examination of this historically interesting dispute
reveals that it did not concern disagreements about the same ob¬
servational events. True, some phenomena were difficult to repli¬
cate, not because of any observational confusion as to what be¬
havioral event occurred, but because of the combined effects of
ignorance about the influence of some experimental operations
and sloppy methods of replication. And the compulsion of many
experimenters to prove competing theories wrong rather than
clarify theoretical issues added to the general confusion. But fun¬
damentally, the core of the difficulty lay with the imprecision of
the competing formulations and the primitive state of knowledge
about the effects of a large number of experimental operations,
not with the theoretical pollution of observational statements.
Two ways in which a scientist’s experiences and preconcep¬
tions can influence his observations have been noted. He can
select different phenomena to observe, or he can limit his atten¬
tion to certain features of an experimental situation while ignor¬
ing others, perhaps of greater potential importance. The first does
not pose any special obstacle to the scholar who seeks to identify
the different observations to which the different conceptions of
behavior are linked. Distinguishing an operant-conditioning con¬
ception of behavior from one based upon a Piagetian model re¬
quires reconstruction of the observations from entirely different
situations. Doing this may cause difficulties because of inade-
quate reporting of what is actually observed or contamination of

observations with theoretical preconceptions. But with diligence
and persistence, and the cooperation of the scientists who work
within these conceptual frameworks, it should be possible for a
scholar to identify the different observations made that are basic
to these theoretical differences and to reveal the nature of the
links (e.g., clear, vague, or nonexistent) between observational
and theoretical statements.
Similarly, the observational support of conceptions that use
the same experimental paradigm but that are interested in differ¬
ent events should provide no insurmountable obstacle for the
scholar who desires to trace these conceptions back to observa¬
tional events. Again it need be noted that the point of this analy¬
sis is not to argue that such observational reconstructions are easy
or always possible. The argument advanced is that, in principle,
in either of the two cases described—events in different empirical
situations or different events in the same empirical situation—
there should be no confusion between the two sets of observa¬
tions underlying the different theoretical conceptions.
The influence of different backgrounds of scientists upon
their observations of the same event is at the heart of the prob¬
lem of distinguishing between observational and theoretical state¬
ments, i.e., deciding whether observational statements are un¬
avoidably polluted by theoretical preconceptions. If they are then
judging the relative merits of competing formulations that seek
to interpret the “same’’ observations could be a hopeless task be¬
cause the observers would be unable to agree about the basic
evidence for which they are offering competing formulations.
Behavioral psychology would be in almost the same predicament
as experiential psychology; consensual agreement about basic
facts would seem to be an unobtainable goal.
The study of perception provides information that is relevant
to the methodological issue of whether observational statements
must inevitably be contaminated or whether consensual agree¬
ment about observational events is a realizable goal for behav¬
ioral psychology. The problem in deciding between these two al¬
ternatives is determining what perceptual evidence is relevant
and what interpretation is appropriate.
The Data Base and Subject Matter of Psychology: 1 5S
Figure 2.1
Numerous perceptual phenomena demonstrate that different

observers looking at the same stimulus display will perceive dif¬
ferent events. A reversible figure-ground design is one example
of a visual pattern that can generate different perceptions. Figure
2.1 illustrates a stimulus display that can either be perceived as a
vase or two faces in profile. What is it really? More about that,
later.
Kuhn (1962), in his influential book The Structure of Scien¬
tific Revolutions, cites an experiment of Bruner and Postman
(1949), to demonstrate the influence of a person’s preconceptions
on his observations. College students were briefly shown indi¬
vidual playing cards with instructions to identify them. Mixed in
with the regular cards were incongruous ones such as a black

four of hearts. Most students identified such a card as a red four
of hearts or a black four of spades. That is, their perceptions were
not determined exclusively by the physical characteristics of the
cards but also by the conceptual categories that they had previ¬
ously learned, black spades and red hearts.
A similar finding was obtained in an experimental demonstra¬
tion of the influence that perception exerts on memory (Car¬
michael, Hogan, & Walter, 1932). Different subjects were shown
the same outline figure (e.g., two circles connected by a straight
line) each coupled with a different label (e.g., eyeglasses or
dumbbells). Later the subjects were asked to draw from memory
the figures they had been shown. Their drawings tended to be
distorted in the direction of the label that was assigned to the
drawing, i.e., the drawings were not remembered as they really
were but rather were distorted in the directions as to how they
were conceptualized.
Many other perceptual phenomena can be reported that are
consistent with the principle that preconceptions can influence
observations. We must conclude that the principle that an ob¬
server’s perceptions can be influenced by his past training and
present set is beyond debate. Quite obviously the facts of psy¬
chology are consistent with the notion that a scientist’s observa¬
tions can be contaminated by his preconception.
The significant question that should be raised about these
and related studies demonstrating that a person’s preconceptions
can influence his observations is whether they are representative
of the observations of a scientist. Is there empirical data from
behavioral research similar to the reversible figure-ground rela¬
tionship in which one observer can perceive one event, a vase,
while another an entirely different one, two profiles? It would be
mistaken to argue that the reversible figure-ground problem is
such an example. One must distinguish in behavioral research be¬
tween the experimenter and the subject. Although subjects, for
brief periods of time, might observe different events (a vase or
two profiles), two experimenters would have no difficulty in
agreeing that the stimulus display was a pattern with borders
that serve as outlines of either a vase or two profiles. This stimu¬

lus display, as described by the observations of the experimenter,
will evoke a perception of either a vase or two profiles when ob¬
served by a subject for a brief time period. And when subjects
continue to fixate upon the reversible figure-ground display, they
observe a sequence of perceptions of a vase and two profiles in
which the two percepts alternate. But from the viewpoint of the
experimenter no confusion has ever occurred as to whether the
stimulus display was a vase or two profiles. It was neither. It is
only what appears in Figure 2.1, which can be easily described and
reproduced without difficulty. Thus the analysis of the reversible
figure-ground phenomenon itself provides absolutely no difficulty
for research psychologists who aspire to achieve consensual (in¬
tersub jective) agreement as to the nature of an environmental
situation and the resultant behavior.
An analysis of the Bruner-Postman study (1949), which
demonstrates that an observer’s preconceptions can distort his
perceptions, also raises the question as to whether such an experi¬
ment is an appropriate model of scientific observations. The dra¬
matic effects obtained occurred when the anomalous playing card
was brieffy exposed. Longer exposure times enabled most subjects
to correctly identify the anomalous card. One can speculate as to
what would have been the results if the experiment had been de¬
signed to reflect actual research practices. Presumably, efforts
would have been made to observe the significant event, the red
four of spades, under optimal conditions that would have en¬
couraged accurate identification. And if an error were made, and
no doubt observational errors are more likely to occur when
events are at odds with prevailing expectations, then the self-
correcting procedures of good research practice such as the use
of several observers and/or the replication of the study would un¬
cover the observational error. In sum, although preconceptions
may influence empirical observations they do not necessarily
have to, especially when precautions are taken.
Although careful research practices are important in reduc¬
ing the possible contamination of the experimenter’s observations
by his preconceptions, it must be noted again that the proper
perspective with which to interpret the relevance of the Bruner-

Postman study to the problem of immaculate perceptions is not
from the position of the subject but rather from that of the experi¬
menter. From this perspective it should be noted that Bruner
and Postman did not encounter any difficulty in specifying clearly
to the psychological community the actual characteristics of the
anomalous display that was employed in their study. Any experi¬
menter would be capable of replicating their anomalous playing
card.
An obvious implication of this line of analysis is that a clear
demarcation between observational statements and theoretical
statements is always possible. Yet I am rdluctant to accept such a
strong conclusion. I am not qualified to judge the appropriateness
of such a distinction for physics, the science that usually serves
as the model for philosophical analyses. Although logical posi¬
tivism, the dominant school of the philosophy of science in the
second quarter of the twentieth century, proposed an absolute
distinction between theory and observation, its position has since
been frequently and vigorously attacked. Many distinguished
philosophers of science (e.g.. Popper and Lakatos) deny that
such a distinction is possible.
In spite of this I can offer two reasons to encourage psychol¬
ogists to try to maintain the distinction between observational
and theoretical events. The first has to do with the nature of the
distinction from the perspectives of the philosopher and the psy¬
chologist, and the second with practical considerations.
Linguistic Representation of Scientific Observation
The observational-theoretical distinction for the philosopher

is embedded in a complicated linguistic problem that may be
best ignored, at least for the present, by the psychologist. The
problem is essentially one of whether a language can be formu¬
lated that accurately mirrors the observations of the scientist, free
of any bias that theoretical preconceptions might encourage. Log¬
ical positivists thought that such an ideal language was possible.
They assumed that fundamental to all observations were a finite
set of elementary sensations linked either to psychological or

physical measurement. Complex observations (e.g., rate of re¬
sponding, aggression), according to this conception, were essen¬
tially combinations of these elementary observations. A language
consisting of terms coordinated to these elementary sensations
could therefore, in principle, describe accurately any set of scien¬
tific observations.
Although efforts to construct such a pure observational lan¬
guage were attempted, none succeeded. The reasons for these
failures are open to debate. Some argue that the psychological
assumption that complex ‘'scientific” perceptions are simply com¬
binations of elementary sensations is invalid; holistic perceptions
are different than the sum of their parts. Others question whether
all scientific observations can be reduced to a finite set of ele¬
mentary sensations. Is there not an infinite number of possible
elementary sensations when observing an empirical situation?
Consider what would be involved in a complete description of an
operant-conditioning situation or an experiment designed to in¬
vestigate obedience (e.g., Milgram, 1963). If one wanted to de¬
scribe every detail of the entire situation one would be involved
in a never-ending task. Still others would argue that the so-called
elementary sensations are not the basic elements of observation;
instead, they are the products of perceptual analysis. Our obser¬
vational knowledge emerges from perceptions of integrated enti¬
ties (e.g., levers, individuals) that can be analyzed into smaller
components.
It should be apparent that the linguistic orientation pursued
by many philosophers in their attempt to analyze the validity of
the distinction between observational and theoretical events is
most complex, containing many important empirical questions
that presumably (or perhaps just hopefully) will be illuminated
by future research in both perception and psycholinguistics.
However, the important issue of an observational-theoretical dis¬
tinction need not be put aside until a deeper understanding of
the psychological problem is achieved. The needs of psycholo¬
gists can be met by viewing the problem of the observational-
theoretical distinction from a purely pragmatic viewpoint. What
procedures can be followed that would yield the highest degree

of observational (consensual) agreement among researchers?
What methods can be pursued that will permit one psychologist
to reconstruct the observations that are basic to another’s con¬
clusions?
Possible Volitional Decisions in Dealing with Problems

of Observational Purity
If one takes the position that an observer and what he ob¬

serves are inseparable, then one must be pessimistic about
achieving any consensual agreement aniong psychologists about
empirical evidence except in those cases when the two observers
share common theoretical commitments. This extreme position is
obviously wrong, as any historical review of research areas such
as memory, sensation, ethology, and others would show. True,
there are theoretical arguments about what experimental design
is most appropriate to test a particular hypothesis. And as already
mentioned, disputes exist as to what results a particular experi¬
mental design produces. But one must also consider the numerous
instances where scientists with different theoretical commitments
are able to communicate in a completely satisfactory manner the
observations of experimental design, procedure, and results with¬
out ambiguity; when replications are desired they can be exe¬
cuted without confusion. The debate about the validity of the
distinction between theoretical and observational terms has been
one-sided, overconcerned with the difficulties encountered in
maintaining the distinction at the expense of ignoring the more
frequent occasions when such difficulties are not met.
In my own experiences as a research psychologist, I have not
encountered insuperable difficulties in distinguishing between my
own observational and theoretical statements. I have also been
able to make a similar distinction with regard to efforts of a large
number of other researchers. This is not to say that in my case,
or in theirs, the theoretical concepts are always unambiguous and
clearly coordinated to experimental operations. Nor is it to deny
that some psychologists hopelessly conflate theoretical statements
and empirical evidence either because they ignore or reject the

distinction. The argument that is being advanced is that consen¬
sual agreement, in principle, can be attained about empirical ob¬
servations free of theoretical contamination, assuming that in¬
vestigators are interested in achieving such a goal. To be specific,
no fundamental difficulty will be encountered when one reports
the rate of responding under different schedules of reinforcement,
the stimulus display confronting preschoolers in their discrimina¬
tion-learning task and the mean number of trials they require to
execute a reversal shift, the description of a set of line drawings
and the percentage of subjects of different age groups that make
categorical, overgeneralized, and overdiscriminated classifica¬
tions. Some of these observational terms (e.g., overgeneralized,
overdiscriminated) have their origins in theoretical concepts, but
their operational meaning can be traced back to their observa¬
tional foundation (Kendler, 1980). No confusion need occur
about these observational events among observers of different
theoretical persuasions. And even when theoretical implications
get embedded in empirical descriptions (e.g., the Ptolemaic-
inspired statement that ‘"the sun rises in the east”) one can fre¬
quently link the statement to an observational base.
It may be that the kinds of research problems that have been
described are not representative of all areas of psychology. Per¬
haps reading EEG records is contaminated by theoretical precon¬
ceptions. Discriminating between grammatical and ungrammatical
sentences may be another illustration of theoretical pollution of
observational statements. Theoretical commitments may taint ob¬
servations in Rorschach testing, and probably did in my efforts as
a clinical psychologist. I believe, however, that improved record¬
ing procedures and increased concern with problems of reliability
among observers would reduce, if not eliminate, theoretical pollu¬
tion of observational statements.
In spite of my belief that careful empirical practices can pro¬
duce immaculate perceptions,^ I cannot offer any infallible meth-
9. There is another sense in whieh theoretical preconceptions can be¬

come imbedded in observational statements, but this issue is unimportant.
If an observer reports that a subject had a reaction time of a certain mag-
odological rule that will guarantee, either for the empirical inves¬
tigator or the scholar who is attempting to reconstruct knowl¬
edge, the complete separation of observational events from theo¬
retical conceptions. Therefore, I am forced to entertain the possi¬
bility that perhaps in some empirical areas that the observational-
theoretical distinction may be difficult, or even impossible, to
maintain.
Accepting this possibility raises the question as to whether
the distinction loses all relevance to scientific practice. If some
notion is not completely right must it be completely wrong? Con¬
sider this problem within the practical framework of building a
boat with primitive equipment that is irtadequate to the task of
preventing all leaks; one would nevertheless make every eflFort to
minimize potential leaks. Is not the scientist confronted with a
similar problem? His tools of epistemological analysis are inade¬
quate but he nevertheless strives, as best he can, to keep his ob¬
servational statements uncontaminated.
Nagel expresses a similar point of view:
It would be idle to pretend, however, that there are no difficulties

in drawing a distinction between observational and theoretical
statements; and I certainly do not know how to make such a dis¬
tinction precise. Nevertheless, I do not consider that this distinc¬
tion is therefore otiose any more than I believe that the fact that
no sharp line can be drawn to mark off day from night or living
organisms from inanimate systems makes these distinctions empty
and useless (1971, p. 19).
The argument that a sharp distinction between theoretical

and observational statements should serve as an ideal of scientific
practice, even if unrealizable, would seem to be at odds with my
disclaimer about oflFering methodological prescriptions for all to
nitude, the accuracy of that observation is based upon the validity of the
theoretical principles of physics, which are assumed to account for the func¬
tioning of the equipment. In this context every observational statement is
influenced by theoretical preconceptions. But these theoretical preconceptions
are not those under investigation and hence are not relevant to the significant
issue of whether observational statements are contaminated by those theo¬
retical principles that are being investigated.
follow. Such a recommendation, strietly speaking, is not being

made. A researeher must make some deeisions to govern his sei-
entifie activities. Considering the general philosophieal and meth-
odologieal problems that have been diseussed in regard to the
distinction between theoretieal and observational statements, I
eonelude that a produetive strategy to pursue is to strive to main¬
tain the distinetion as best one ean, an eflFort that I have not
found diffieult.
Other volitional deeisions are possible. One possibility would
be that sinee no rule ean be stated that elearly separates obser¬
vational statements from theoretieal preconeeptions, attempts at
maintaining sueh a diserimination will inevitably prove mislead¬
ing and confusing; science would be unavoidably distorted by
attempts to maintain a fietitious distinction. Adopting this line of
reasoning suggests that one should go about one’s scientifie busi¬
ness without any self-conseious eflForts to distinguish between
observational and theoretieal statements. Such a methodological
decision cannot be faulted if one denies the reality or the useful¬
ness of the observational-theoretieal distinetion. But one ean raise
the question whether sueh a deeision encourages an exeessive
amount of subjectivity in science and ereates unneeessary ob¬
stacles for scientists who are trying to discover the relative merits
of competing theories.
This is an appropriate point in my analysis of the observa¬
tional purity of behavioral psyehology to return to the question
that initiated such a discussion: Was the inability to aehieve con¬
sensual agreement about mental events by phenomenologieal
(experiential) psyehologists eomparable to the unresolved theo¬
retieal differences in behavioral psyehology? If one accepts the
assumption that the observational-theoretieal distinction in be¬
havioral psyehology is relevant and useful, even if not always
sharp and elear, then one must answer the question in the nega¬
tive. Phenomenological psychology is apparently eonfronted, at
least for the present, with the insolvable task of isolating its ob¬
servational events from the presuppositions of the methods used
to observe them. One ean speeulate that different methods of self¬
observations do not represent different perspeetives for observing
the same event but instead they bring into existence diflFerent
phenomenal events. If true, then we really do not have diflFerent
phenomenological methods but instead diflFerent phenomenolo-
gies in the sense that diflFerent experiential worlds are created by
diflFerent methods of phenomenological exploration.^®
In rebuttal, it might be argued that the same situation pre¬
vails in behavioral psychology; theoretical preconceptions de¬
termine what is observed. For example, if one assumes that move¬
ments are learned in instrumental conditioning, then one observes
the manner in which an animal presses a bar; but if one assumes
that acts are acquired, then one notes only whether the bar is de¬
pressed. Such an argument overlooks' fundamental diflFerences
between behavioral and experiential observations. In the example
cited it would be possible to observe both events (movements
and acts) simultaneously, achieve inter subjective agreement
about the observational events, and ultimately determine the rel¬
ative merits of the competing conceptions of instrumental con¬
ditioning. None of these would be possible in pure phenomenol-
ogy; the intrasubjectivity of the experiential event prohibits re¬
cordings of the event, intersubjective agreement about its occur¬
rence, and resolution of theoretical diflFerences.
If one denies the pertinence of the observational-theoretical
distinction to behavioral psychology, one need not be compelled
to accept a methodological equivalence between pure phenome-
nology and behavioral psychology. One can acknowledge that the
theoretical disputes in experiential psychology bear a resem¬
blance to those in behavioral psychology in that they emerge
from contamination of observational data by theoretical precon¬
ceptions. At the same time one can recognize that the contamina¬
tion can be qualitatively diflFerent or significantly greater in one
case than in the other. If this is true, then one would expect that
history would demonstrate that agreement about observational
events would be more diflBcult to achieve in experiential psychol¬
ogy than in behavioral psychology. I would suggest that histori¬
cal evidence supports this expectation.
10. This speculation must remain a conjecture because no method of

phenomenological analysis is available to evaluate its validity.
In shifting the frame of referenee from methodologieal anal¬

ysis to historieal examination, one must entertain the possibility
that improved methods of phenomenological research are pos¬
sible. One need not accept the convenient assumption that experi¬
ential analyses are infallible and incorrigible. One can suppose
the opposite; a person's self-observations can be mistaken and are
correctable. The concept of stimulus error identified a possible
source of error in observing mental events—confusing physical
characteristics with psychological experience. One can also argue
that just as illusions occur in the perception of the outer world
so are they possible when observing the inner world. Conceiv¬
ably, a tense person can “mistakenly” believe he is experiencing
inner serenity when in fact he is feeling a sense of relaxation fol¬
lowing physical exertion.
Many, if not all, phenomenological methods operate on the
assumption that self-observations can be fallible and corrigible.
In essence, graduate training in departments of psychology with
a structuralist orientation, as was the case at Cornell when Titch-
ener was chairman, was designed in part to teach future psychol¬
ogists methods of self-observation that would be free of errors. It
is interesting to note that a similar problem is posed for the pa¬
tient in psychoanalysis who must learn to examine and report his
phenomenal experience in an “accurate” and forthright way.
Even if one adopts the position that self-observation should be
executed in a naive fashion, in the absence of preconceptions (as¬
suming that is not a presupposition), one will nevertheless learn,
as is suggested by the author’s experience (self-observation can
be fun!), that increased sensitivity to the events of phenomenal
experience seems to occur with increased practice.
In essence, the suggestion that the “accuracy” of self-obser¬
vations can be improved with training supports the notion that
introspection is fallible and corrigible. The question is immedi¬
ately raised as to whether persistent efforts to improve observa¬
tions of mental events would not yield a level of consensual
agreement about the nature of the mind that heretofore has not
been achieved.
The discussion up to this point does not offer any optimism
about such a possibility. The reason seems obvious. The assump-
tion of fallibility and corrigibility applies only to experiential data

but not to the methods employed to obtain them. Although it
might be possible to resolve diflFerenees about experiential data
when introspeetors are in agreement about the appropriate
method of self-observation, no standards are presently available
to resolve diflFerenees about whieh introspeetive method is the
valid one. If we aeeept the historieal conclusion that diflFerent
methods of phenomenological research provide diflFerent pictures
of the mind, we are once again led to the conclusion that we do
not really have diflFerent methods of phenomenology, but diflFerent
phenomenologies. Within this context, the conclusion seems in¬
evitable that consensual agreement abouf the nature of conscious
experience by methods of pure phenomenology is an unobtain¬
able goal.
Shared Experience
Before leaving the analysis of the mind as an independent en¬

terprise, reference to the important concept of shared experience
needs to be made. Interestingly, this concept, which is of great
importance for humanistic psychologies, can be elucidated by ref¬
erence to some comments of Titchener, the American leader of
structuralism who said:
I turn now to the topic of visual imagery, which is always at my

disposal and which I can mould and direct at will. I rely, in my
thinking, upon visual imagery in the sense that I like to get a
problem into some sort of visual schema, from which I can think
my way out to which I can return (1909, p. 10).
He goes on to discuss the experiential meaning of “visual

schema”:
The term visual schema” is, of course, itself equivocal. Those of

you whose minds are built on the same general plan as my own
will know well enough what it means. But I must warn the others,
to whom this sort of imagery is unknown, not to think of a geo¬
metrical figure printed black on white, or of anything a hundredth
part as definite (ibid., p. 12).
Later on, after offering other examples of his visual imagery,

Titchener concludes: ‘‘All this description must be either self-
evident or as unreal as a fairy-tale” (ibid., p. 14). In other words,
what he is proclaiming is that if the reader has had phenomenal
experiences similar to Titchener's, the reader would “understand”
Titchener s mind, at least that portion involved with visual im¬
agery. Without such experience, Titchener s mind would remain
beyond others’ personal apprehension.
Shared experience then becomes a vehicle for “understand¬
ing” another person’s private experience. The concept of shared
experience, however, does not overcome any of the methodologi¬
cal diflBculties previously discussed in regard to attaining consen¬
sual agreement about the basic data of private experience. An
obvious question is how does one know that the same kind of ex¬
perience is being shared. One evidently does not look into an¬
other person’s mind and observe that his own experience is being
shared by another. The psychological basis of the shared experi¬
ence is that a verbal description of one’s own experience (or that
of a character in a novel) can arouse in another an empathic
response, which signifies that he has had a similar experience.
The sharing itself cannot be proved to have occurred by methods
of pure phenomenology unless the weak methodological require¬
ment of merely the testimony itself is accepted as sufficient.
The concept of “shared” experience (in quotes to emphasize
the speculative nature of the sharing) has significance outside of
the range of phenomenology. The concept represents a commonly
used index of understanding the behavior of others, e.g., if a
clinician can empathize with his patient he can understand the
patient’s actions. Also, “shared” experience can presumably func¬
tion as a causal link in a behavioral chain, e.g., if a person can
empathize with someone he will be altruistic toward him (Krebs,
1975). In sum, “shared” experience as a phenomenological event
has implications for the understanding and control of behavior,
which will be returned to later (page 94).
The Data Base and
Subject Matter of Psychology:
Part Two
1. Behavior and conscious experience can each be investigated inde¬

pendently of each other or in combination. The resulting information
from any one of these three approaches (behavior, experience, behav¬
ior and experience) can serve as the subject matter of psychology.
2. Phenomenologists, in their effort to reveal the nature of the mind,

can rely exclusively on verbal descriptions, or they can attempt to
coordinate these descriptions to public events. When coordination is
successful “quasi-objective” knowledge results. Such knowledge does
not guarantee a veridical account of consciousness, but it does con¬
tribute a degree of objectivity not possessed by the phenomenological
report itself.
3. Behavioral psychology can deal with phenomenal experience using

a variety of strategies ranging from choosing to ignore it to construct¬
ing a theoretical model of the mind that meets the demands of meth¬
odological behaviorism.
4. An operational approach in behavioral psychology is desirable if

it is understood that the entire meaning of a concept is not limited to
its operational meaning. An operational definition links a concept to
its observational base thus facilitating communication about empirical
knowledge. The empirical meaning of a concept is represented by the
total number of empirical relationships within which the operationally
66
defined concept is involved. The intuitive meaning of a concept refers

to the hunches and ideas a scientist has when attempting to think
creatively about a concept while its theoretical meaning is contained
in the totality of theoretical statements that are proposed to account
for the empirical relations involving the concept.
5. The kinds of knowledge yielded by studying public behavior and

private experience are fundamentally different although future tech¬
nical developments may make it possible to investigate conscious ex¬
perience in an objective fashion. At present, psychologists are con¬
fronted with convergent and divergent paths in investigating conscious
experience and behavior. What path one selects will depend on both
the goal one sets for psychology and the criterion one accepts for
understanding.
RELATIONSHIPS BETWEEN BEHAVIORAL

AND EXPERIENTIAL PSYCHOLOGY
My analysis of the subject matters of psychology, behavior and

private experience, has suggested that the study of the former
can stand alone as an independent natural-science discipline.
This means that it would be possible to interpret behavior, even
human behavior, on the basis of empirical relationships and theo¬
retical construction without any reference to direct observations
of mental events. A broad consensual agreement among a com¬
munity of behavioral psychologists could be achieved about the
meanings of the underlying observational and theoretical state¬
ments. Again, this does not mean that such statements would be
completely free of ambiguity but rather that with scholarly dedi¬
cation possible sources of ambiguity or confusion would ulti¬
mately be revealed. Thus it should be possible for one behavioral
psychologist to understand another in spite of theoretical differ¬
ences. And when clarification is not easily forthcoming, historical
pressures inherent in the scientific enterprise (e.g., experimental
replications, epistemological clarification) will bring to the sur-
face the facts and confusions inherent in unresolved theoretical

disputes. Behavioral psychology has the capacity to be “objective”
in the sense of dealing with events, observational and theoretical,
that are available to public scrutiny.
The idea that it is possible to adopt objective behavior as a
dependent variable without any reference to mental events does
not imply that such a decision is strategically desirable, only that
it is methodologically feasible. Introspective reports need not
threaten behavioristic purity if the reports themselves are con¬
sidered basic observational data and not direct reflections of
conscious experience. The point at issue for the behaviorist is to
discover whether such reports possess heuristic value as sources
of significant data or fruitful hypotheses.
Experiential psychology, with conscious experience as its
subject matter, can also stand alone as an independent discipline.
However, its foundations appear too insubstantial to support a
discipline capable of achieving broad consensual agreement
about the facts of conscious experience and the means by which
theoretical differences to account for them can be resolved. By
broad consensual agreement is simply meant that the accord ex¬
tends across boundaries of the variety of methodological orienta¬
tions that have and can be employed in pure phenomenological
research. Both historical evidence and epistemological analysis
support the opinion that consensual agreement is an unobtainable
goal for pure phenomenology. As a consequence, pure phenome-
nology is incapable of employing a natural-science approach.
Is such a verdict final? Are mental events forever beyond the
pale of objective (intersubjective) knowledge? I think not. But
only when one accepts the intrinsic methodological limitation of
raw phenomenal experience can one creatively cope with it. Ac¬
knowledging the intrasubjective isolation of experiential data en¬
courages one to develop means to compensate for its privacy.
What is obviously needed is an epistemological prosthetic that will
provide public support for the tenuous methodological underpin¬
nings of private experience. This can be accomplished by relating
verbal descriptions of private experience to other publicly ob¬
servable events. By doing so, confidence can be increased that the
verbal reports that presumably reflect private experiences do so

with some degree of accuracy.
Phenomenal Experience and Public Events
A simple example of the advantages that accrue from relating

experiential reports to public events can be demonstrated by ref¬
erence to the popular pastime of wine tasting. Suppose some self-
assured connoisseur, who happens to be a dean, in his infinite wis¬
dom proclaims a particular wine to be “amusing but honest.
Viewed within the framework of pure phenomenology we can
either accept the description as possessing some degree of valid¬
ity in that he is having a distinctive experience, or we can enter¬
tain the suspicion that he is showing off. The significant point is
that if we do not proceed beyond the evidence of his introspec¬
tive report, we will forever remain in the dark about which of
these two alternatives is correct.
One way of resolving the issue is to challenge him to a blind¬
fold test (assuming he does not rule on salary raises) in which he
is instructed to identify from a variety of wines those that are
amusing but honest. You would be led to the conclusion that the
dean is a phony about wine tasting if his judgments are incon¬
sistent: several different wines are amusing but honest on one
occasion but not on another. Suppose, however, he is consistent:
of all the wines he swishes in his mouth, he identifies only two, a
Portuguese rose and a California rose, as amusing but honest. If
the blindfold test is executed effectively so that the only cue to
his response is the taste of the wine, one is encouraged to con¬
clude that these wines produced a distinctive experience for him.
No doubt, such an example will be immediately labeled, by
some, as trivial—unrelated to significant human experiences such
as the feeling of self-worth, human dignity, and other experiential
concepts that are thought to be at the core of human existence.
This kind of reaction obstructs methodological clarification and
frees those who deal with life’s vital issues from meeting demand¬
ing standards of evidence when drawing conclusions.
The point of my wine-tasting example was not to suggest

that all problems of phenomenologieal researeh would eollapse
to a psyehophysical approach in which verbal reports of con¬
scious experience are related to publicly observable characteris¬
tics of the environment. Obviously that is not the case for several
reasons that will subsequently be identified. For the moment, the
important consideration is that the problems of experiential re¬
search take on an entirely new cast when considered within the
context of relating phenomenal experience to publicly observable
events (e.g., different kinds of wine).
When the dean is successful in consistently selecting the
same wines as amusing and honest, our'confidence increases that
his introspective reports reflect phenomenal events. The con¬
sistent relationship between the reports and publicly observable
events indicates that the former are not solely linguistic expres¬
sions in that they are unrelated to conscious experience and are
simply an expression of social gamesmanship. Of course, the char¬
acterization amusing hut honest itself is not demanded by the
underlying experience. It could be that another individual, a new
assistant professor, whose discriminations were as consistent as
those of the dean, would describe the two roses as light, mod¬
erately full-bodied, and relatively dry. The difference between the
two characterizations could possibly be linguistic in that different
words are used for the same or similar experience. Such a possi¬
bility could be explored by subjecting the two to additional wine¬
tasting tasks to determine the similarities in their discriminative
abilities along with attempts to discover whether the linguistic
dimensions used to represent the dean's experience (e.g., amus¬
ing, honest) are translatable into the terms used by the assistant
professor (e.g., lightness, body, dryness).^
1. Cain (1979) has reported that odor identification is vastly improved

when the observer has labels to identify his olfactory experiences. This
raises the interesting phenomenological question as to whether the labels
influence the experience directly by maldng the olfactory sensations more
distinctive or serve only as convenient terms with which to associate un¬
changing experiences. Because the experiences cannot be publicly observed,
the question does not lend itself to a universally acceptable answer by
methods of pure phenomenology.
I am not trying to suggest the rough outlines of a proposed

research project in experimental phenomenology nor to imply
that at the end of such a project the strong conclusion could be
drawn that two individuals are sharing a common phenomenal
experience. I am only suggesting that a quantum leap in reliable
knowledge is achieved when phenomenological reports are in
agreement with each other and correlate with some publicly ob¬
servable external referent. To return to an underlying theme,
such a conclusion is based upon the assumption that introspective
reports go beyond phenomenal experience itself. The raw ex¬
perience is transformed into an introspective report, and a funda¬
mental task for experiential psychology is to tease out from the
introspective report the purely linguistic components {amusing
but honest versus light, moderately full-bodied, and relatively
dry) that are unrelated to the phenomenal experience itself. By
discovering that a relationship prevails between the report and a
pubhcly observable event, one surmises that the verbal reports
are not only linguistic; they do, in fact, reflect some characteris¬
tics of phenomenal experience. And perhaps the most compelling
argument in favor of such an interpretation is that one could pos¬
sibly predict the kinds of verbal descriptions that would be forth¬
coming when novel forms of stimulation (e.g., new wines) are
presented. If the self-observers who had previously agreed now
also report similar introspections, then confidence increases in the
similarity of their experiences.
One must be exceptionally guarded, as I have tried to be, in
drawing any conclusion about the similarity of the phenomenal
experiences of two different individuals. There are two main rea¬
sons for this. First, and foremost, the phenomenal experiences of
the two individuals cannot be directly compared, and thus at
best, the presumed similarity must remain a tenuous inference.
Second, the implication of the absence of a psychophysical rela¬
tionship is ambiguous in regard to drawing inferences about phe¬
nomenal experience. Consider once again my hypothetical dean.
Suppose his introspective report, amusing but honest, was not
correlated with any particular wines. Could one therefore deny
that he had a distinctive phenomenal experience when he uttered
such statements? One could not! Perhaps his verbal reports re-
72 Psychology: A Science in Conftici
fleeted a distinctive experiental state, but such states are a func¬

tion of a constellation of variables and therefore are not simply
tied to a taste of a particular wine. Another possibility is that the
dean is indulging in deception; he is merely using words for their
social effects, and they have nothing to do with his inner experi¬
ence. I cannot deny the former interpretation although my intui¬
tion favors the latter.
In essence, if a psychologist opts to become a phenomenolo-
gist in order to reveal the nature of the mind, he has a choice be¬
tween being a pure phenomenologist who depends exclusively on
verbal communication to reflect conscious experience or being an
experiential psychologist who seeks to overcome the intrinsic
limits of the privacy of the mind by coordinating introspective
reports to public events. The knowledge obtained from these two
different positions is qualitatively different; a shift from the first
orientation to the second results in a shift from purely subjective
knowledge to quasi-objective information.
The conclusion that the relationship between introspective
reports of conscious experience and publicly observable en¬
vironmental events represents only quasi-objective knowledge ap¬
plies only to the efforts of experiential psychology. The qualifier
quasi is added to emphasize the point that such knowledge, as al¬
ready argued, has limitations for drawing definitive conclusions
about the mind. The same kind of information—introspective re¬
ports related to environmental events—is objective (intersubjec-
tive) for the behavioral psychologist because he is not burdened
with the task of revealing the nature of the mind. His task is
restricted to behavioral analysis, and introspective reports can be
considered simply another form of public behavior, free of any
inferential content for the nature of the mind. The same knowl-
edge a psychophysical relationship—has different meanings de¬
pending upon whether the investigator is studying the mind or
behavior.
Some phenomenological experiences need not or cannot be
coordinated solely to environmental events. Physiological mea¬
sures can also provide the coordination required to qualify “con¬
scious experience as quasi-objective knowledge. This point is
illustrated by the phenomenon of meditation.
In meditation one restriets one’s pereeptual experienee and

eoneentrates fully upon an apparently insignifieant event, sueh as
eounting breaths; or pondering intuitively rather than logieally a
riddle, known as a koan (e.g., “How ean I attain enlightenment by
driving on the Los Angeles freeway?”); or repeating eertain sig¬
nificant words over and over again; or contemplating an object
such as a vase for long periods of time; or repeating monotonous
chants. Confirmed meditators describe their phenomenal experi¬
ence as pleasurable detachment, no mind, one-pointedness, inner
serenity, and other characterizations that suggest a lack of respon¬
siveness to the outside world and an expanded consciousness of
the self.
Obviously the question arises, particularly among the tough-
minded variety of behaviorists, whether such phenomenally de¬
scriptive terms mean anything? Do meditators merely learn new
verbal responses or do they actually acquire new experiences? To
answer this question a number of investigations have sought to
determine whether meditation has an eflFect on physiological pro¬
cesses. In one well-known study (Wallace & Benson, 1972), rep¬
resentative of others, an affirmative answer was found. Electrical
skin resistance, a commonly used measure of “emotionality” (the
higher the skin resistance, the greater the “relaxation”; the lower
the measure, the greater the “tension” and “anxiety”) changes
during periods of meditation. For a representative experienced
meditator, skin resistance rises rapidly when meditation begins
and drops when it terminates. Other findings are that meditation,
compared to periods of nonmeditation, decreases oxygen con¬
sumption, carbon monoxide elimination, the rate and volume of
respiration, heart rate, and blood lactate level. The intensity of
slow alpha waves also increases during meditation. The entire
pattern of physiological changes resulting from meditation indi¬
cates a wakeful, highly relaxed, hypometabolic state.
The epistemological status of such results are no different
from those of the psychophysical relationship between introspec¬
tive reports and wine tasting. Assuming that phenomenal experi¬
ence is related to physiological processes, one is encouraged to
use such evidence to demonstrate that meditation in fact pro¬
duces a distinctive kind of mental event. If this physiological
pattern is distinctively diflFerent from that exhibited during sleep

or hypnosis, and some investigators deny this, one can also con¬
clude that the meditative experience has unique phenomenal
properties that distinguish it from other relaxed states. However,
consistent with my previous analysis, relating physiological events
to phenomenal reports does not guarantee that persons who are
exhibiting a common physiological pattern are in fact having the
same mental experience; nor can it be demonstrated that those
who are not exhibiting a common physiological pattern are also
not experiencing similar mental events. Thus the conclusion must
be drawn, as in the case for environmental-mental relationships,
that physiological events coordinated with phenomenal experi¬
ence produce only quasi-objective information.
What is obviously true is that the physiological pattern is not
the conscious experience. If, for example, a group of meditators
described their experience as inner serenity and such a report was
perfectly correlated with a distinctive pattern of physiological
and environmental measures, nonmeditators would know nothing
about how the experience of inner serenity feels. The subjective
experience of inner serenity cannot be confused with public
events to which reports of the experience are coordinated. This
distinction has been touched upon by many philosophers. Schlick
made a distinction between to have an experience (e.g., inner
serenity) and to have knowledge (e.g., the introspective report
of inner serenity is related to a distinctive pattern of physiological
and environmental events). Dewey similarly distinguished be¬
tween having and knowing. One can have a toothache and one
can know that toothaches are “caused’’ by a pattern of neuro¬
physiological events.
Behavioral psychology is concerned with knowing about be¬
havior while experiential psychology is involved with understand¬
ing having. Understanding, in this context, can be achieved in
two possible ways. One can understand a phenomenal experience
(e.g., toothache, inner serenity) by having it, or one can under¬
stand a phenomenal experience by knowing the publicly observ¬
able conditions that produce it. I do not understand labor pains
in the former sense but do in the latter sense.
A pure phenomenologist could turn his back on possible rela¬

tionships between phenomenal experience and environmental or
physiological events because such information, he could argue, is
irrelevant to the task of discovering a veridical account of human
experience. Therefore, adopting the epistemological standards of
the natural sciences is self-defeating because the richness of
everyday experience cannot be revealed by the scientific method,
by knowing.
If one acknowledges that a veridical account of human ex¬
perience is possible, but that it cannot be evaluated by standards
of empirical verification (or falsifiability), then one must propose
some method capable of capturing the quality of phenomenal
experience with great fidelity. But what method? This question
comes back to the problem of establishing rules by which one
can judge the relative merits of different methods of observing
phenomenal experience. Numerous methods have been employed,
ranging from highly structured analytic procedures to those that
encourage a naive, natural, ‘‘presuppositionless” attitude.
In the absence of any objective criteria, the pure phenome¬
nologist must resort to subjective standards. Since the knowledge
forthcoming from the enterprise of pure phenomenology is com¬
pletely insulated from any refutation, one can conclude that it is
unscientific, a judgment that some phenomenologists would
proudly accept. For our purposes not much is gained from the
attribution “unscientific” because of the lack of agreement that
prevails about demarcation lines between science and nonscience.
Of greater relevance to psychology than explicating the boundary
between science and nonscience is examining the possible rela¬
tionship between purely subjective and objective knowledge.
Consider the hypothetical case of describing conscious experi¬
ence by a method of pure phenomenology. Suppose the effort
produces an ideally descriptive language that coordinates distinc¬
tive phenomenal events (e.g., ecstasy, happiness) to distinctive
public events (e.g., patterns of physiological events). If each
term of the phenomenal vocabulary that was generated by pure
experiential research was coordinated to a particular pattern of
public events, one would be in a position to formulate a theory
in which phenomenological constructs, defined in terms of coor¬

dinated public events, could serve both as ‘causal” agents for
subsequent behavior (e.g., empathy toward members of minority
groups will reduce prejudice) and as a dependent variable (e.g.,
empathic feelings toward others can be acquired by special train¬
ing procedures). In addition, the phenomenological language,
with its coordinated public measures, could serve as a bridge in
the attempt to relate, both in terms of similarity and differences,
one phenomenal experience with another. Even experiences that
could not be shared (e.g., maternal and paternal love, male and
female orgasms) could be analyzed in terms of their phenomenal
similarity.
At this point warnings and qualifications are demanded to
protect my methodological flanks. I have not argued that the re¬
sults of pure phenomenological research, even of the highest
quality, can be incorporated into an objective psychological the¬
ory involving environmental, behavioral, and physiological mea¬
sures. I have only suggested that they might be. In other words,
no insurmountable obstacle precludes the possibility that ac¬
counts of human experience from pure phenomenological research
could be integrated with an objective psychology. I mention this
possibility, although I am not optimistic about its occurrence in
the foreseeable future, in order to identify potential points of
contact between the apparently incompatible methodological
orientations of pure phenomenology and behavioral psychology.
The most likely possibility is that these distinctly different kinds
of psychological orientations will remain disparate, thus con¬
tributing to a fundamental disunity within psychology.
If extremist methodological positions are rejected by expe¬
riential and behavioral psychology, then a number of different
lines of reconciliation become possible. Experiential psychology,
without abandoning its primary concern with mental events, can
nevertheless acknowledge the possible relevance of environmen¬
tal, behavioral, and physiological events. If this is done then some
methodological decision must be made concerning the relative
value of subjective verbal reports in comparison to objective in¬
dices of environmental, behavioral, and physiological events for
understanding the mind. By considering the objective indices, the
experientialist acknowledges the insufficiency of verbal reports

alone. An obvious strategy for an experientialist to pursue is the
one already described—a quasi-objective approach in which the
accuracy of phenomenal reports are judged within the context of
their coordination to publicly observed events. Such events pro¬
vide a framework for developing a mutually agreed-upon vocab¬
ulary to describe phenomenal experience. The more objective
measures one can relate to private experience, the greater is the
likelihood of achieving consensual agreement about the nature
of the mind. At present, this opinion represents more an article
of methodological faith than a demonstrated fact. The current
interest in the relationship between physiological states and men¬
tal events should demonstrate soon whether this expectation is
justified.
Phenomenal Experienee and Behavioral Psyehology
Behavioral psychology can deal with phenomenal experience in a

number of diflFerent ways that will now be described and evalu¬
ated. The major decision is either ignore it or deal with it. Jus¬
tification of the first position can take one of three forms.
Metaphysical Behaviorism
Metaphysical behaviorism (Bergmann, 1956), denies the ex¬

istence of mental states. Watson’s aggressive polemics against
structuralism finally encouraged him to favor this extreme posi¬
tion. It is one thing to point out the methodological difficulties
inherent in analyzing subjective experience; it is quite another
matter to deny its existence. This distinction seemed to have been
lost in Watson’s energetic attacks on the conception of psychol¬
ogy as the science of the mind.
Epiphenomenalism
By adopting the philosophical position of epiphenomenalism

some psychologists feel justified in ignoring conscious experience.
A major assumption of this position is “that experience is simply

a nonfunctional derivative of certain physiological processes tak¬
ing place in the nervous system” (Wagoner & Goodson, 1976).
According to this argument then, conscious experience could not
be responsible for changes in behavior any more than the move¬
ment of a person’s shadow is responsible for his walking. If true,
why bother with conscious experience when analyzing behavior?
The trouble with this gambit is that it oversimplifies the
problem of causation, prejudges empirical issues, and ignores
practical considerations. The assumption that conscious experi¬
ence does not play a functional role in determining behavior is
hopelessly enmeshed in the problem of causation, a topic that
has failed to yield any universally accepted solution since the
days of the ancient Greek philosophers. It is important to recog¬
nize that accepting the notion that conscious experience is an ex¬
pression of neurophysiological events and has no existence apart
from them does not rule out the possibility that knowledge about
experience, regardless of its inferential status, can account for
behavior or even neurophysiological events. The macro genetic
concept of gene had no existence apart from the microgenetic
sequence of rungs on a DNA ladder; yet we know that the for¬
mer concept of gene effectively helped predict a variety of he¬
reditary phenomena.
Another tack that can be taken in questioning the nonfunc¬
tional status of conscious experience that is consistent with epi-
phenomenalistic doctrine is to note that assuming experience to
be a “derivative” of neurophysiological processes does not neces¬
sarily imply that the two are equivalent. Being embarrassed is not
identical to the neurophysiological events to which the embar¬
rassment is coordinated. This does not mean that embarrassment
exists apart from the neurophysiological changes but only that
the two are operationally distinct because they are based upon a
different set of observations. From this point of view, one can
argue that conscious experience under certain circumstances
(e.g., limited knowledge of neurophysiological events, the nature
of the behavior that is to be predicted) will be more closely cor¬
related with certain forms of behavior than available neuro¬
physiological information.
The relationship between conscious experience and behavior

can be perceived by the psychologist as a purely empirical prob¬
lem. Rather than adopt a traditional philosophical resolution of
the mind-body problem such as epiphenomenalism, parallelism,
or identity, or some other position, a more productive strategy
would be to transform specific experiential-behavioral problems
into theoretical questions with determinate empirical conse¬
quences. Accepting this approach would discourage embarking
on a search for a single general answer to experiential-behavioral
relationships that may be unobtainable simply because the nature
of the relationship may vary for diflFerent kinds of experiences and
physiological events. The empirical approach, at least for the
time being, would focus on specific experiential-behavioral rela¬
tionships with the realization both that the conclusion drawn
about one need not apply to others and that any specific conclu¬
sion is susceptible to modifications with additional information.
The simplest, and perhaps the most compelling, argument
against ignoring introspective reports is that significant data can
be excluded. Even a shadow, a noncausal agent of behavior,
can provide useful information. A detective who wants to watch
the movements of a person under surveillance, without being
seen himself, can focus on his target’s shadow and thus be able
to predict his movement.
The Incompatibility Between Phenomenology

and Behaviorism
Instead of rejecting conscious experience because of its un¬

real or nonfunctional status, one can choose to ignore it because
of the conviction that a behavioral analysis will inevitably be¬
come corrupted, contaminated, and confused when consideration
is given to mental processes. Even in the absence of interest in
phenomenology many behaviorists seem unable to resist em¬
ploying phenomenological interpretations of behavioral events.
Tolman (1932), a sophisticated behaviorist, created theoretical
ambiguities by using phenomenological terms to express his be¬
havioral hypotheses. Even Skinner (1969), a hard-line behavior¬
ist, succumbed to phenomenological descriptions when he stated
that when contingencies of reinforcement are arranged in a hos¬

pital for psychotics the “patients make fewer demands on the
staflF and yet display as much dignity and happiness as their
pathology permits.” The solution to the threat of phenomenology
would be to adhere to a strict behavioral analysis similar to that
required for the treatment of the psychology of infrahuman or¬
ganisms. Only a rigorous mentally free strategy in psychology, the
argument goes, will be capable of meeting the standards of a
natural-science methodology.
Arguments Against a “Mentally Free”

Strategy in Behaviorism
A “mentally free” strategy is unnecessary and self-defeating.

The objectivity of a behavioral psychology can be preserved by
treating introspective reports as directly observable events. Such
reports can and have had heuristic value for developing a behav¬
ioral psychology.
Introspective reports can be time-savers. Psychophysical
methods and operant conditioning methods involving simple yes
or no responses or key-pressing reactions can reveal that a human
responds to the two lines of the Miiller-Lyer illusion as unequal
in length. But is it necessary to go through so much trouble to
discover what a simple introspective report would reveal in a
second? This does not mean that psychophysical and operant
conditioning methods might not be useful later on when precise
psychophysical laws are desired, but rather, that introspective
reports can provide useful information in identifying interesting
problems, especially in the field of perception.
Introspective reports can be a source of useful hypotheses.
Learning a list of nonsense syllables (e.g., BEW, ZUR, VOD)
was once thought to be acquired by rote memory. Each syllable
was presumably memorized as a meaningless sequence of letters
until the subject mastered each item. Introspective reports of
many subjects would not reflect such processing. Instead, one
strived to make sense out of nonsense; nonsense syllables were
encoded into words and meaningful relations were established
among them. For example, the above string of three nonsense

syllables could be encoded as part of a sentence: BEWare of
ZURich and VODka. If introspective reports are systematically
obtained from adult human subjects, the results of most, if not
all, would be at odds with a simple rote-learning model.
Phenomenological Interpretations of Behavioral Theories
Behavioral theories are not required to be consistent with

phenomenal experience. They have to be compatible with publicly
observable evidence. The reason for rejecting a simple rote¬
learning model is not that it fails to jibe with phenomenal evi¬
dence but instead that it is inconsistent with available data (e.g..
Postman, 1971; Tulving & Osier, 1967). Recent theoretical devel¬
opments have suggested that an adequate theory of serial learn¬
ing must include mechanisms of encoding and organization, pro¬
cesses that are suggested by introspective reports.
But why should not the demand be made that behavioral
theories reflect phenomenal experience? The reason is obvious.
Processes that determine behavior are not always represented in
phenomenal experience.
A familiarity with the history of psychology of thinking will
immediately dispel the illusion that introspective reports, which
presumably reflect consciousness, can provide a complete ac¬
count of human thinking. One of the most significant problems
in developmental psychology, the relationship between language
and thought, provides ample evidence that introspective reports
are unable to reflect the cognitive processes that are operating
(e.g., Anglin, 1977). This conclusion is true for adult behavior as
well. For example, Binet (Reeves, 1966) concluded from the
analysis of introspective reports of two expert arithmetical cal¬
culators and several creative writers that unconscious processes
were operating in thought. The same conclusion is supported by
the introspective analysis of thought by psychologists of the
Wurzburg school. Little conscious content in thought was noted.
In certain tasks, such as associating a superordinate word with a
subordinate one (e.g., chair-furniture), thought occurred with-
out any conscious deliberation. “One does one’s thinking before

one knows what he is to think about” (Boring, 1950, p. 404). In
the more demanding task of concept identification, subjects ap¬
pear unaware of their cognitive processes. Hull (1920) found
that some of his subjects could classify linear patterns (Chinese
ideograms) into correct conceptual categories without being con¬
sciously aware of the eritical cue. Even creative efforts at the
highest intellectual level can suddenly occur without any pre¬
meditation. Henri Poincare, the famous French mathematician,
described an episode in which he decided to go on a vacation
after working unsuccessfully for some time on a mathematical
problem. Travelling from his home td another town, he com¬
pletely forgot about his mathematieal problem until he arrived
at his destination and boarded his bus. At the moment he put his
foot on the bus, the solution to the problem that had plagued
him for weeks suddenly became clear.
The above evidence is a compelling argument against basing
theoretical processes of thinking exclusively on reports of con¬
scious experience. Something more is needed.
Another limitation of introspective reports is that they can
be misleading. This is a cardinal assumption of many, if not most,
personality theorists and clinical practitioners. A person can con¬
sciously believe that his behavior is determined by one factor
when in reality another factor is responsible. Posthypnotic sug¬
gestions illustrate one such example: a person falsely explains a
partieular act because he is unaware of the command given to
him while under hypnosis. Numerous kinds of phenomena, which
have been described by concepts such as repression, rationaliza¬
tion, ambivalence, amnesia, projection, motivated forgetting, re¬
action formation, and overreaction, all suggest that conscious ex¬
perience can be a misleading source of information for under¬
standing behavior. Although psychologists disagree as to the
appropriate theoretical conceptualization of such phenomena, the
evidence, some experimental (e.g., Latane & Darley, 1968; Latane
& Rodin, 1969) but mostly clinical, denies the assumption that
introspective reports will accurately reflect behavioral causation.
Do I contradict myself? I first suggest that introspective re-
The Data Base and Subject Matter of Psychology: 2 8S
ports can have heuristic value and then argue that important
behavioral processes can be absent from such reports and that
they can actually be misleading. Although these two conclusions
are in opposition they are not contradictory for the simple reason
that no single relationship operates for all cases in which com¬
parisons are made between introspective reports and behavior.
The two in some cases can dovetail, in others be related, and still
in others be antagonistic. A uniform strategy for dealing with
introspective reports in a behavioral analysis would therefore be
unwise.
Behavioral psychologists have a variety of strategies to
choose from in dealing with introspective evidence. They range
from one extreme of ignoring it to the other of using it for con¬
structing a theoretical model of conscious experience. In evaluat¬
ing such strategies three important principles must be remem¬
bered for dealing with introspective evidence within a behavioral
framework: (1) the ultimate value of the strategy must be dem¬
onstrated rather than prejudged; (2) phenomenal experiences
must be considered an inferred construct and not basic datum;
and (3) introspective evidence need not reflect fundamental be¬
havioral processes (and vice versa). In regard to the final point,
Tolman, one of the leading theorists in the early history of cogni¬
tive psychology, oflFered the following warning about the value of
introspection for formulating theoretical constructs, which he re¬
ferred to as intervening variables:
Is not introspection after all, at least in the case of men, a signifi¬
cant method by which one can get at and define these intervening
variables in a direct and really reliable fashion? I doubt it. I be¬
lieve that introspection is a form of social response—a type of
final behavior . . . one which has very complicated conditions.
. . . The very essence of introspection lies in the fact that it is a
response to audiences—external and internal. And such being the
case, it seems less likely to mirror most types of intervening vari¬
able so directly and correctly as do more gross nonsocial forms
of behavior (1936, p. lOI).
Although Tolman can be accused of overstating his ease by

suggesting a universal prineiple that belittles the heuristie value
of introspective reports, his comments nevertheless serve as a use¬

ful warning to those who believe that introspective reports can
serve as the royal road to the understanding of human behavior.
One last point. This discussion has revolved around the rela¬
tionship between a subject’s introspective report of his phe¬
nomenal experience and his behavior. Another mind-behavior
relationship prevails in behavioral psychology; the phenomenal
experience of the psychologist serves as the source of theoretical
ideas. Tolman, who expressed disdain for the theoretical value of
introspective reports of subjects, admitted to exploiting his own
phenomenal experience in order to develop theoretical constructs
of the learning behavior of rats. Unless a human’s phenomenal
experience has some special advantage for interpreting rat be¬
havior, Tolman could be accused of being inconsistent. If intro¬
spective reports, which he argued are responses to an internal as
well as an external audience, are inferior to gross behavior for
suggesting theoretical notions, why should his own experience
possess a special value for the behavior of rats?
Tolman’s apparent inconsistency is a side-issue; the signifi¬
cant problem concerns the merits of a theorist’s phenomenal ex¬
perience as a source of theoretical notions. This problem is
equivalent to that of evaluating the theoretical potentialities of
introspective reports of subjects; pragmatic considerations are the
only reasonable guidelines. It would be difficult to justify any a
priori argument that mining ones own mind possesses some in¬
trinsic validity for formulating behavioral hypotheses.
Mining one’s own mind might be helpful in generating theo¬
retical insights. The effectiveness of the strategy could probably
be increased by appreciating its limitations. An obvious limitation
for constructing theories of infrahuman behavior (as well as that
of young children) is that of anthropomorphism—attributing
mechanisms of behavior reffected in the theorist’s consciousness
that are not available to the subject. Darwin, who attributed hu¬
manlike feelings and highly intelligent behavior to dogs, illus¬
trates one such example. My own opinion is that Tolman (1936)
and Krechevsky (1932) made similar errors on a much smaller
scale. Although they both tried to be operational in defining their
mentalistic-sounding concepts of cognitive maps (Tolman, 1932)
and hypotheses (Krechevsky, 1932), their theoretical terms rap¬

idly acquired surplus meaning that exceeded by far the learning
capacities of their rodent-subjects.
Two other risks in extrapolating theoretical processes from
one s own inner experience should be mentioned. These also can
be discussed within the context of Tolman’s theory of animal
learning, but unlike anthropomorphism, these interfere with in¬
terpreting human behavior. First, phenomenally descriptive con¬
cepts (e.g., cognitive maps) are intuitively appealing, providing
a sense of understanding that is not matched by their explanatory
power; in Dewey’s terminology (page 74) a sense of having is
mistakenly interpreted as knowing. Second, is the tendency for
theories involving terms that reflect mentalistic processes to ig¬
nore principles of performance. This was pointedly expressed by
Guthrie, who accused Tolman’s cognitive theory of leaving the
rat "‘buried in thought” (Guthrie, 1952, p. 143). Tolman’s rats,
like Hamlet, had the capacity to think but not to act.
A similar problem prevails in psychotherapy. Psychothera¬
pists frequently assume that if a patient becomes consciously
aware of his psychological problems, his pathological behavior
will disappear. Unfortunately, insight into one’s own difflculties
often seems to have no effect upon behavior. Therapists who pos¬
tulate that behavioral improvement follows the attainment of
insight can, and do, defend their position by arguing that if be¬
havior fails to improve following insight, “true” insight has not
been achieved. (Theoretical defense mechanisms will be dis¬
cussed in the next chapter.)
Once again, let me clarify my basic point. Formulations that
seek to interpret behavior with theoretical concepts that presum¬
ably reflect phenomenological processes are not intrinsically de¬
fective. Rather, they involve potential sources of difficulties that
can be best avoided by becoming aware of them.
In concluding my discussion of the role of phenomenal ex¬
perience for behavior theory, whether based on the conscious
experience of subjects or the theorist, one particular problem
must not be overlooked. A frequently repeated truism for many
psychologists is that the key to understanding a person’s behavior
is to know how he perceives the world. The whole problem of
psychological explanation would be solved, the truism goes, if

only we had a key to unlock the secrets of the mind. There is no
question that if we had direct access to a person’s conscious ex¬
perience our ability to predict his behavior would improve.
Nevertheless, such information would be far from perfect for the
reason already mentioned; there is not necessarily a one-to-one
correspondence between inner experience and outer behavior.
But even of greater significance is the problem of understanding
why a person’s conscious experience is the way it is. Surely we
ean better understand a person’s bizarre behavior when we ap¬
preciate that he is a victim of paranoid thoughts? But that for
such a person a vast explanatory problem still remains is sum¬
marized by the simply-stated question, “Why does he have para¬
noid thoughts?” This question, if one desires to understand the
behavior of paranoids, with the aim of reducing the incidence of
paranoia, is of much greater importance than discovering and
describing how a paranoid feels. Bergmann expressed a similar
view when discussing the concept of psychological environment,
which represents a subject’s personal perception of the world:
But even so, what is the predictive value of the suggestive meta¬
phor “psychological environment”? Is it not the business of science
to ascertain which objective factors in the past and present states
of the organism and its environment account for the difference in
response, so that we can actually predict it instead of attributing
it, merely descriptively and after it has happened, to a difference
in the psychological environment? (1943, p. 133).
In sum, one can strategically opt for interpreting behavior

within a phenomenological framework. To achieve success, more
is required than formulating theoretical concepts that reflect
inner experience.
Consensual Agreement about Basie Observations

in Behavioral and in Experiential Psyehology
The conclusion thus far is that two distinct and legitimate subject
matters are available for study, conscious experience and objec-
tive behavior. The methodological principles governing behav¬

ioral and experiential research diflFer in regard to the capacity of
each discipline to approach consensual agreement about basic
observations; because the basic observations of behavioral psy¬
chology are open to public (intersubjective) scrutiny, consensual
agreement is more easily achieved than is the case for experiential
psychology.
In my attempt to structure the fundamental methodological
issues I have ignored two problems related to consensual agree¬
ment of basic observations: operationism in behavioral psychol-
ogy and the direct comparison of the conscious experience of two
persons in experiential psychology. The foundation has been laid
for the profitable discussion of these two problems.
OPERATIONISM
My analysis of the distinction between observational and theo¬

retical statements has been approached pragmatically within the
operating procedures of behavioral psychologists who are inter¬
ested in communicating experimental procedures and results, and
theoretical notions. Although recognizing that terms that describe
the scientist’s observation can be contaminated by theoretical
preconceptions, in the absence of any precise linguistic rules to
distinguish between theoretical and observational statements, a
psychologist’s efforts should be governed by the ideal of main¬
taining the distinction. In actual practice, this means the psy¬
chologist is confronted with the problem of conveying sufficient
information about his research so that others will know exactly
what he did and, if desired, replicate his efforts. Although mis¬
understandings can and do occur, such failures must be judged
against the frequency of successful communications.
Operational Meaning
In essence, what is being argued is that the operational approach

is not only desirable but also feasible. Such a statement demands
analysis, both because of the variety of meanings that have been

assigned to operationism and the failure of many operationalists
to appreciate that operational meaning does not encompass all of
the meaning contained in a scientific concept.
In its most limited sense operationism represents a simple
recipe to further meaningful discourse about empirical results.
Misunderstanding could be reduced and, it is hoped, eliminated
by tracing the meaning of a term used to report experimental
evidence back to its observational base via the operations of the
scientist. For example, operationally the statement that “The two
groups of children were equated for intelligence with the Pea¬
body Picture Vocabulary Test” means that a particular test of
intelligence was administered properly and the statistical evalua¬
tion revealed that the obtained diferences between the IQ scores
of both groups could reasonably be attributed to chance.
Although the statement that “The two groups of children
were equated for intelligence with the Peabody Picture Vocabu¬
lary Test” is simple and would constitute a minute portion of an
entire experimental report, it is nevertheless an exceedingly com¬
plex proposition in that it summarizes a large number of experi¬
mental operations and implicit assumptions. The operations range
from the numerous procedures required to administer the intelli¬
gence test to the appropriate statistical evaluation to appraise the
difference in IQ scores. Can such a brief statement accurately
summarize all these observations? Yes. The communication is
taking place between psychologists who are presumably suitably
trained to understand the observational references of such a sim¬
ple statement. If questions or confusions arise (e.g.. What were
the variances of the IQ scores for each group? Did some of the
subjects suffer from a language handicap that would make the
intelligence test inappropriate?) direct communication between
the author and the reader may be required for a “complete” un¬
derstanding of the observational base of the experimental report.
Conceptualizing operationism in this limited sense (e.g., the
operational definition of intelligence as given in the description
of the technique of intelligence testing) seems to reduce it to the
trivial. Quite obviously, prior to the advent of operationism scien-
tists were being operational. Scientific communication demands

it. If this is all that operationism means, why was a fuss created
when operationism appeared on the psychological scene in the
1930s? If we remember that structuralism and other methodologi¬
cal orientations employing mentalistic concepts were plagued
with imprecision and unresolved theoretical disputes, then we
can understand the enthusiastic reception operationism initially
received. A faith was generated, particularly among psychologists
whQ aspired to a natural-science methodology, that conceptual
ambiguities would dissipate, theoretical differences would dis¬
appear, and progress would be ensured. This hope was based
upon a misconception of operationism. As Bergmann aptly noted,
‘'The root of the trouble was that some psychologists in their en¬
thusiasm mistook the operationist footnote for the whole of the
philosophy of science” (1954, p. 210).
Critics of operationism also exaggerated its role; shortcom¬
ings were attributed to operationism for its inability to solve
problems for which it was not designed to cope. In tlie field of
intelligence, which suffers from more conceptual confusions than
it needs or deserves, the role of operationism is frequently mis¬
understood. One example is the criticism that operationally de¬
fined intelligence—the intelligence test measures—both ren¬
ders the formulation impregnable to attack and contributes
nothing toward the validation of the intelligence scale (Penning¬
ton & Finan, 1940).
Empirical Meaning
There are several kinds of meanings of intelligence of which op¬

erational meaning is only one. The validity of an intelligence test
refers to empirical meaning, how test scores are related to other
operationally defined concepts such as academic success, achieve¬
ment tests, and job performance. A clearly operationally defined
concept need not have empirical meaning. That is, operationism
does not guarantee the fruitfulness of a concept. For example,
cephalic index, the ratio multiplied by 100 of the maximum
breadth of the head to its maximum length, is operationally de¬

fined when the teehniques of measuring maximum breadth and
length are elearly stated. Nevertheless, its elegant elarity has
failed to give psyehology a eoneept of any importanee for the
simple reason that no one has been able to involve the eoneept of
eephalie index in any significant psychological law.
The fact that cephalic index is operationally defined but em¬
pirically empty in no way detracts from the communication value
of its operational meaning. Knowing that an operationally defined
concept is empirically empty provides significant information. It
indicates at that time that the concept is not fruitful, and the
investigator has the option of dropping'the concept, persisting in
his attempts to discover some empirical meaning or modify its
operational meaning.
The possibility of altering a concept's operational meaning
has been overlooked by some critics of operationism. A common
complaint against operationism is that it tends to freeze the
meaning of a concept before it could be developed into a fruitful
construct. Such a criticism is inappropriate if the proper scope of
operationism is realized. An operational definition is not perma¬
nent. Its meaning can evolve by the addition of new or the elimi¬
nation or modification of old operations. Nevertheless, at any
given time its operational meaning can be discerned:
A statement such as this, . . . Operationism requires that all the

conditions be taken into account, can be quite misleading. For a
construct which unwittingly leaves out a relevant factor (or de¬
termining condition) just leads to a different formulation of the
empirical laws. For example, by telling us what manipulations he
performs, what pointers he reads, . . . and what computations
he carries out with the numbers thus obtained . . . , a primitive
physicist would give us a methodologically correct definition of
his empirical construct “density of a liquid.” And this in spite of
the fact that he might not have given any attention to the tem¬
perature in his laboratory, one of the conditions upon which, and
as we know, and as he might not know at that state of his investi¬
gation, the results of his manipulations and computations depends.
The point is that we are able to trace back the terms of his Ian-
guage to the immediately observable. He has laid down all the

eonditions under which he is going to say: “This liquid has the
density 1.3.” Therefore, we know what he means, and that is aU
general methodology can insist upon at this level of the so-called
operational definition of empirical constructs (Bergmann &
Spence, 1941, p. 3).
Intuitive Meaning
Another criticism frequently heard is that operationism places an

unnecessary burden on the creative scientist. Demanding a set of
operations when one is forming vague ideas of a concept may be
too much of a burden for the creative process. Again we find a
criticism that is not aimed at the proper target. Operationism
does not offer a prescription for productive thinking. Operational
definitions are the results of creative thought, not necessarily the
means by which it is accomplished. A psychologist may have only
a vague, imprecise notion of a concept (e.g., intelligence, motiva¬
tion ) when first thinking about it. When he is “thinking through”
the concept, indulging in infrasubjective communication, the re¬
quirements of operational definitions need not be considered, and
certainly not be met. In actual fact, it is not inconsistent for a
researcher who is investigating intelligence to use an IQ test as
his operational definition of intelligence while simultaneously
entertaining an intuitive conception of intelligence that is much
broader and more complex, involving such notions as symbolic
representations, cognitive structures, intellectual operations, and
other concepts that one might consider to be fundamental in in¬
tellectual functioning. To underline this distinction intuitive
meaning must be contrasted with operational meaning.
Critics of operationism argue that the richness of the intui¬
tive meaning of a concept is destroyed by the narrowness of its
operational meaning. Such a conclusion emerges from the mis¬
conception that the two kinds of meaning need to be at odds
when in fact they can be complementary. Intuitive notions can
be enriched by surveying relevant empirical evidence, which of
necessity contains the ingredients of operational definitions. Al¬

though an investigator may not self-consciously formulate an
operational definition when conducting research, such a definition
can be extracted from an examination of his procedures.
I find it difficult to imagine that a psychologist interested in
interpreting intelligence within a broad theoretical framework
would fail to be stimulated by the wealth of data that has been
collected from intelligence testing designed to predict academic
and job success. The evidence obtained as a result of wide-scale
intelligence testing reveals important developmental changes that
must be accounted for by any theory of intelligence.
Theoretical Meaning
At this point a fourth meaning of a concept, theoretical meaning,

must be added to conclude this discussion of operational defini¬
tions. Again using intelligence as an example it can be viewed
neither as: (1) an operational definition referring to the proce¬
dures of a particular instrument of measurement; nor as (2) an
empirical law relating operationally defined concepts (e.g., in¬
telligence test scores with academic success); nor as (3) a set of
intuitive refiections and ideas entertained by a scientist when
thinking creatively (hopefully) about the psychological nature
of intelligence; but instead as (4) an explicit theory of intellectual
functioning. In essence, the theoretical meaning of intelligence
is contained in the totality of theoretical statements that are pro¬
posed to account for the empirical relations involving intellectual
functioning. Such a theory, if it aspires to any breadth, would
seek to go beyond the facts of intelligence testing. The intelli¬
gence test movement, it must not be forgotten, was not conceived
for the purpose of collecting data about fundamental processes
of intellectual functioning or for formulating a general theory of
intelligence. A broad theory of intelligence would have to cope
with many processes such as symbolic representation, conceptual
development, memory, intellectual operations, and inference. The
theory would have to account for the facts of learning sets, hy-
The Data Base and Subjeet Matter of Psychology: 2 93
pothesis testing, conservation, discrimination-shift behavior,

stages of information processing, and so on. Within such a broad
theoretical and empirical context, intelligence means much more
than what the intelligence test measures. Whereas operational
meaning is designed solely for purposes of communication, theo¬
retical meaning serves the need for understanding.
After distinguishing between operational, empirical, intui¬
tive, and theoretical meanings, it should be apparent why Berg-
mann assigned operational meaning the status of a footnote in
the philosophy of science. A primary aim of science is to provide
understanding, and the really difficult tasks in achieving this goal
are the formulation of some guiding notions (intuitive meaning),
collection of significant facts (empirical meaning), and develop¬
ment of fruitful hypotheses (theoretical meaning). In contrast,
operational meaning assumes minor importance. But it should be
understood that the search for these four kinds of meaning are
not independent efforts, insulated from each other. Rather they
represent different components of a single venture, and even
though operational meaning may be of secondary importance
when compared with the other forms of meanings, it neverthe¬
less serves an essential function in the entire enterprise of be¬
havioral psychology.
There is no doubt that much criticism and resistance to op¬
erational definitions was encouraged by the exaggerated claims
and oversimplifications advanced by its proponents. Even P. W.
Bridgman, the physicist-philosopher who fathered operationism
(1927), disassociated himself from the inflated claims of naive
operationists by admitting to feelings of having sired a “Franken¬
stein.”
I abhor the word operationalism or operationism, which seems to

imply a dogma, or at least a thesis of some kind. The thing I en¬
visaged is too simple to be dignified by so pretentious a name;
rather, it is an attitude or point of view generated by continued
practice of operational analysis (1954, p. 224).
At the same time he recognized a fundamental opposition to

even his modest conception of operationism:
Any person can make an operational analysis, whether or not he

accepts what he supposes to be the thesis of “operationalism,” and
whether or not he thinks he is wasting his time in so doing. So far
as the operationalist” is to be distinguished from the “nonopera-
tionalist,” it is in the conviction of the former that it is often
profitable and clarifying to make an operational analysis, and also,
I suspect, in his private feeling that often the ' nonoperationahst”
does not want to make an operational analysis through fear that
it might result in a change in his attitude (1954, pp. 224-26).
Bridgman’s clinical judgment could be extended a bit further.

The nonoperationalist may resist an operational analysis not for
fear of a ehange of attitude but rather Because of an unshakable
faith in the validity of his intuitive coneeptions. He is convinced
that all operational definitions will fall short of the true meaning
of his intuitive notions and, if employed, might lead to a prema¬
ture ‘disproof” of an idea that is intrinsically sound. The non¬
operationalist is correct in believing that operational definitions
cannot possibly reflect the full range of his intuitions for the sim¬
ple reason that the latter are mueh broader and more intrieate
than the former. But what he fails to realize, or refuses to ac¬
knowledge, is that intuitive notions can only evolve into viable
theoretical conceptions via the employment of some operational
definitions. Theoretical development eannot occur in the absence
of empirical evidence, which of necessity, implicitly or explicitly,
contains operational definitions. In essence, the nonoperationalist
employs a proteetive shield to defend his intuitive ideas from
possible refutation and by so doing stunts their scientific growth.
DIRECT COMPARISON OF THE CONSCIOUS

EXPERIENCES OF DIFFERENT INDIVIDUALS
The major obstacle to achieving eonsensual agreement in expe¬

riential psychology is that conscious experience is private, acces¬
sible to only the experiencing individual. As a result, even when
the subjective reports of two individuals are in essential agree¬
ment and are buttressed by a similar pattern of common physio-
logical measures and/or environmental events, a nagging doubt

prevails about whether the two are really sharing a common ex¬
perience. There appears to be no way of getting around the pos-
sibihty that common verbal descriptions coupled with common
objective indices need not reflect common experiences (or even
the obverse that diflFerent introspections combined with different
objective measures may occur in combination with a common
experience). It is because of this possibility that phenomenologi¬
cal investigations seeking to correlate introspective evidence with
publicly observable behavior and/or physiological events are
characterized as quasi-objective. The inability to directly examine
the conscious experience of different individuals necessitated the
adjective quasi.
The question that I would like to raise now is whether the
highest level that can possibly be achieved by a natural-science
approach to experiential psychology is quasi-objective. To an¬
swer this question one must face up to the problem of whether it
can be demonstrated directly that two individuals are sharing a
common experience. If so, experiential psychology could achieve
a level of objectivity that is presently unobtainable.
One answer to the question concerning the feasibility of di¬
rectly comparing the inner experiences of two individuals is that
it cannot be done; the privacy of mind cannot be penetrated by
any physical means. An alternative response is that although a
comparison is presently impossible, technical developments could
make it feasible.
Over the years I have presented this issue to graduate classes
concerned with methodological issues in psychology. The ques¬
tion is posed in the context of a “game” I played with my younger
son, Kenneth, who at the age of seven responded to questions
about visual images by admitting that he was capable of seeing
“clear pictures” in his head. On one bitterly cold winter day I
asked him to close his eyes and imagine High Lake, an idyllic
spot in Vermont where we spent our summers. I then closed my
eyes and said, “Tm ‘seeing’ High Lake also. Are our ‘pictures’ of
High Lake the same?” (I forget his answer although I sometimes
wonder whether questions of this sort encouraged him to become
a psychiatrist.) After describing this anecdote I challenged my

class to design an experiment, unimpeded by any technical limita¬
tions, capable of answering the question that was posed to my
son.
My own position has been that the direct comparison of inner
states is neither logically nor physically impossible but rather, at
present, technically impossible. Consistent with this position, the
following research design was proposed replete with science-
fiction apparatus.
The first piece of equipment that is needed is an apparatus
that publicly and accurately displays a person’s visual image. To
demonstrate its accuracy, the subject, by appropriate means,
would be placed in a state in which he would be unaware of
whether he is observing his own image or observing his image on
the display. By appropriate psychophysical techniques it could be
demonstrated that the subject cannot discriminate between the
two.
Once the adequacy of such equipment is demonstrated then
the significant comparison can be made. Ken and I would simul¬
taneously image High Lake from a common vantage point. Our
brains would be connected to the visual display that could accu¬
rately reproduce either image. We then, individually, would be
subjects of a psychophysical experiment to determine whether we
can discriminate our own image from that of the other. By sys¬
tematic research of this sort it should be possible to determine
whether common visual images do occur and with what fre¬
quency.
Most students accept the appropriateness of this design al¬
though a majority argue, as is always the case in designing ex¬
periments, that some variation in procedure would be preferable.
These reservations, however, are not critical because they do not
question the fundamental assumption that such research is feasi¬
ble given that certain technical breakthroughs were achieved.
Some argue, for reasons that are unclear, that the proposed kind
of research is physically or logically impossible. Perhaps their
disagreement stems from a confusion of peripheral issues with the
fundamental problem of the direct comparison of conscious ex-
periences of diflFerent individuals. In order to minimize that con¬

fusion the three following clarifications are offered: (1) The
proposed experimental design is not based on the assumption that
visual images are “floating around the head” and are perceived
by a “mind’s eye.” Although visual images can be observed phe¬
nomenologically their occurrence depends upon neurophysiologi¬
cal events. (2) The proposed design does not imply that all prob-
problems of comparing phenomenological experiences such as
feelings, will be as “simple” as investigating shared imagery. But
the basic argument is that if private visual images can be made
public, appropriate procedures could be developed for other
forms of private experience. (3) The proposed design in no way
implies that the linguistic problem of describing inner experience
will automatically be solved by demonstrating that similar images
or feelings can be shared. The problem of coordinating a linguis¬
tic system to represent the basic observations of what were pre¬
viously private events is no different from making observational
statements about public events. In essence, what would be ac¬
complished if the proposed science fiction project were successful
would be the eradication of the epistemological boundaries be¬
tween what is currently described as private experience and
public behavior. The problems of developing an appropriate
language system for publicly observed “private” events and de¬
termining whether such a system would be insulated from theo¬
retical preconceptions bring us full circle to the problems of
discussing observational statements about public behavior. The
issues are identical and, therefore, the answers should be the
same.
SUMMARY
Has anything been gained by this excursion into science-fantasy-

land? Not if the entire issue of the relationship between behav¬
ioral and experiential psychology is simply put aside until the
miracle of achieving direct access into the minds of others is
realized, assuming it ever will be. The point of my analysis is to
lay out the possible methodological choices confronting the psy¬

chologist when deciding upon which subject matter(s) to inves¬
tigate. The conclusion reached is that the independent entities of
public behavior and conscious experience yield two fundamen¬
tally different kinds of knowledge: the criterion of consensual
agreement for observational statements is attainable in behavioral
but not necessarily in experiential psychology. Such a conclusion
in no way justifies restricting the subject matter of psychology to
public behavior. It does suggest, however, that consensual agree¬
ment among phenomenologists about the facts of conscious expe¬
rience will at best be limited to communities of psychologists that
share a common parlance and investigatory procedures.
Psychologists are not confronted with an ‘‘either-or ’ choice
when choosing between public behavior and private experience
as subject matters. One can seek to investigate both, but if one
strives for methodological consistency one must adopt a single
epistemological framework to judge the validity of all kinds of
observational statements. An obvious and common choice has
been to select the public framework of behavioral psychology
while viewing private experience as an inferred state from di¬
rectly observed events (introspective reports, responses to per¬
sonality inventories, environmental stimulation, etc.). Viewing
conscious experience as a theoretical construct, in preference to
ignoring it or conceiving it as a directly observable event, must
be considered a strategic decision. Its value must be demon¬
strated rather than taken for granted.
Within the epistemological framework of a psychology that
adopts public behavior as its subject matter, the phenomenal
experience of the psychologist possesses no intrinsic validity for
any formal interpretation of behavior. That is, the validity of a
theoretical construct cannot be justified on the basis of any ap¬
parent correspondence with the phenomenal experience of the
theorist. It has for behavioral psychology no factual content al¬
though it certainly can serve as a source, and perhaps a fruitful
one, of potential hypotheses. This is cogently represented by the
title of Max Meyers book (1921), The Psychology of the Other
One, that illustrates the methodological isolation of the behav-
ioral psychologist from the subject matter he investigates and

interprets.
Up to this point, conscious experience, as reported by sub¬
jects or directly observed by psychologists, has been perceived as
ancillary to the major task of understanding publicly observed be¬
havior. Without changing the underlying epistemological frame¬
work of behavioral psychology, it becomes possible to expand the
discipline to include the analysis of experiential processes, to in¬
vestigate “subjective behavior.” This can be accomplished by
“objectifying” subjective experience—by relating experiential re¬
ports to publicly observable events, particularly physiological
processes, because of the presumed intimate relationship between
the two. The underlying assumption of such an eflFort is that the
accuracy with which introspective reports reflect conscious ex¬
perience is in some way proportional to the number and variety
of independent publicly observed events to which they are coor¬
dinated. These correlated public events, then, serve as anchor
points to coordinate a phenomenally descriptive language with
conscious experience.
In sum, this analysis oflFers both divergent and convergent
paths for psychology, with its dual subject matter, to pursue. The
particular choice selected will be determined not only by the
goals psychologists set for their efforts but also by the conceptions
ihey adopt for understanding psychological events, a topic that
is examined in the next two chapters.
On Understanding:
Part One
4
1. Understanding is a psychological concept th^t refers to the manner
in which empirical phenomena are comprehended. Psychologists have
employed different criteria to achieve understanding: deductive ex¬
planation, interpretive consistency, comprehension by behavioral con¬
trol, and intuitive understanding.
2. Deductive explanation is achieved when an event is logically de¬

duced from a set of assumptions that have empirical implications.
Deductivism as a form of understanding must be distinguished from
deductivism as a research strategy that encourages the formulation of
formal deductive models at all stages of theoretical development.
There is no methodological recipe that indicates at what stage of theo¬
retical development a formal deductive model should be introduced.
3. Successful deductive theories cannot be automatically created by

an “inductive logic.” Although logic can get one from theoretical as¬
sumptions to empirical statements, it cannot provide passage from
empirical statements to theoretical assumptions.
4. The natural-science approach to psychology has usually been asso¬

ciated with the use of deductive explanation to achieve understanding.
Some psychologists have argued that the scientific method must be
modified to meet the needs of the social, behavioral, and humanistic
sciences. If the scientific method is considered as a systematic proce¬
dure used to arrive at warranted theoretical statements regardless of
content, then the method transcends the borders of various scientific
disciplines.
100
On Understanding: 1 101
5. Deductive theories can be conceptualized as representing reality or

as convenient fictions that enable discrete events to be interpreted in
a coherent manner. In both cases, the structure of the deductive
theory is essentially the same. Nevertheless, the strategic consequences
of neurophysiological theories, as contrasted to “black-box” theories,
may be quite different, particularly in regard to the resolution of theo¬
retical disputes.
6. Reductionism has several meanings in regard to the reduction of

behavioral to neurophysiological events: relating behavior to neuro¬
physiological variables, adopting a model of behavior based on physio¬
logical evidence, and coordinating a behavior theory to a physiological
theory. The theoretical reduction of psychology to neurophysiology is
not necessarily inevitable.
7. The “truth value” of deductive theories can be evaluated within

three different frameworks. The traditional view assumes that scien¬
tific knowledge is accumulated in a progressive, linear fashion and
that theoretical disputes are resolved by critical experiments that vali¬
date one theory and disprove its competitor. The subjective view, an
extreme version of Kuhn’s analysis of the history of science, assumes
that there are repetitive patterns of normal science and revolution.
During normal science, a global methodological-theoretical orientation
—known as a paradigm—guides empirical and theoretical development.
A revolution occurs when a prevailing paradigm is supplanted by a
new one. This revolution does not occur suddenly and decisively but
instead takes many years to accomplish because the competing para¬
digms are incommensurable, and the choice of one paradigm in pref¬
erence to another is subjectively determined. The historical view as¬
sumes that competing research programs are commensurable, and the
adoption of one program in preference to another can be defended on
the basis of explicit standards such as one program is progressive
(e.g., predicting novel events) while another is degenerating (e.g.,
adding ad hoc assumptions). The historical view is the most appropri¬
ate framework for evaluating deductive theories in psychology.
THE SEARCH FOR MEANING
Aristotle, in the opening sentence of his Metaphysics, postulates

that “all men by nature desire to know.” Today, many psycho¬
logical theories (e.g., Maddi, 1970), make a similar assumption,
viz., humans possess a motive for meaning—the need to under¬
stand and interpret their world in a consistent manner. Clinical
evidence suggests to these theorists that failure to achieve a
meaningful existence is a prime cause of psychological distur¬
bances. The laboratory also provides evidence for a need to know.
Experiments that investigate cognitive processes find that hu¬
mans persist in their attempts to extract regularities from se¬
quences of environmental events. One example is the behavior of
human subjects who are confronted with the task of predicting
which of two alternative events will occur. When the events are
randomized the subject, unaware of the chance order, persists to
an extraordinary degree in producing hypotheses to account for
the sequence of occurrences (Feldman, 1963). When the binary
events are generated by some rule that is not excessively com¬
plex, the principle that underlies the regularities is usually dis¬
covered with little difficulty.
My intent is not to argue for any universal human motive to
seek meaning. I personally find it difficult to deal with a hypothe¬
sis that ignores environmental, or hereditary, components in be¬
havior. At the same time, I am willing to admit (at least I like
to believe) that psychologists as a group are motivated by a need
to understand the problems with which they deal.
Many scientists accept the notion that a valid interpretation
of the world can only be achieved by the scientific method. Im¬
plicit in this position is the denial that a “true” understanding can
be reached in nonscientific disciplines—art, the humanities, the-
ology, philosophy—and that only one form of understanding pre¬
vails in science. My position will be more subjective. A sense of
understanding can be achieved in a variety of ways. That under¬
standing can only be achieved by empirical methods is essentially
denied by those, including myself, who seek to understand sci¬

ence by epistemological analyses.
To understand understanding, within contemporary psychol¬
ogy it becomes necessary to identify and analyze different modes
of “explanation.” If we accept the notion that understanding can
be achieved by different kinds of knowing, we can more easily
understand the apparent diversity of “explanations” that are
offered by different psychologists. For example, childhood autism
has been interpreted by some as a frustration response to a mech¬
anistic environment (Murray, 1974) and by others as resulting
from a particular schedule of reinforcement (Ferster, 1961). In
evaluating such contrasting views one could rationally argue for,
or arbitrarily select, a model of scientific explanation that would
serve as a standard to judge interpretative efforts. Although such
standards are helpful for one’s own work, they can nevertheless
render incomprehensible the efforts of others who fail to adopt a
similar interpretive framework. Since our aim is illumination, not
persuasion, four different kinds of understanding will be identi¬
fied: deductive explanation, interpretive consistency, comprehen¬
sion by behavioral control, and intuitive understanding. Though
not necessarily mutually exclusive, they are each characterized
by a dominant feature that makes them distinctively different.
DEDUCTIVE EXPLANATION
The form of understanding usually associated with natural-science

methodology is deductive, which assumes that natural events can
be interpreted in the form of logical order. Understanding is
achieved when an event is deduced from one or more general
propositions. In essence, the deductive process is analogous to
mathematical proof with the precision varying from mathematical
verification to the rigorous use of ordinary language.
The medium for deductive explanations is a “deductive the¬
ory”—a system of propositions that are logically organized and
are coordinated to empirical events in such a manner that legiti¬
mate deductions (predictions or postdictions) about the phe-
nomena can be made. The essential eharacteristics of a deductive

theory possess an empirical content and are capable of generating
deductions that could, in principle, be at odds with the evidence.^
In line with my refusal to don the robes of the philosopher of
scienee, I will neither suggest that all deduetive theories share a
fundamental conceptual strueture nor adopt what appears to be
a more reasonable position that they ean assume a variety of
theoretical forms (Suppe, 1974).
Deductivism as a Form of Understanding and

as a Theoretical Strategy
To clarify the discussion of deductive explanation a distinetion

must be made between two different features of deductivism.
Deductivism can be considered as a methodological decision to
adopt a deductive model of understanding. It can also be con¬
sidered as a strategy to formulate highly formal deductive models
at aU stages of theoretieal development. Deductivism as a model
of understanding does not neeessarily imply deductivism as a
methodological strategy. The conflation of the two is illustrated
in the efforts of Clark L. Hull, who passionately argued for the
neeessity of hypothetic-deductive theories of behavior. Of interest
is the following quotation from his autobiography:
The study of geometry proved to be the most important event of

my intelleetual life; it opened to me an entirely new world—the
fact that thought itself could generate and really prove new rela¬
tionships from previously possessed elements. Later in the writing
of a prep school paper in English composition, I tried to use the
geometrical method to deduce some negative propositions regard¬
ing theology (Hull, 1952, p. 144).
1. The discussion of this crucial point must be postponed until later

(pages 135-151). For the time being a strongly worded interpretation of
this essential characteristic is that the deductive theory could be falsified;
weakly expressed it means that the deductive theory could be embarrassed
by certain kinds of empirical evidence.
Although Hull obviously understood the distinetion between

logieal proof and empirieal predietion, his failure to mention this
differenee refleets the high premium he always plaeed in his
theoretieal efforts on a formal deductive structure, often to the
detriment, some believe, of empirical content. That is, sometimes
he was more concerned with formalism than empiricism.
An overall evaluation of Hull’s efforts, I believe, would indi¬
cate that he exerted salutary effects upon theoretical practices in
psychology, particularly in regard to the explication of the sup¬
porting logic underlying theoretical predictions. I am reminded
of a comment made to me by M. E. Bitterman, a frequent and
severe critic of Hull’s work. He took upon himself, during one
sabbatical, the task of reading successive volumes of the Psycho¬
logical Review in order to obtain deeper insights into the history
of psychology. He remarked that Hull’s articles in the 1930s “were
like a breath of fresh air” when contrasted with the muddy con¬
ceptualizations of that and previous eras. In his early articles
Hull formulated theoretical models of “complex” forms of animal
learning based upon principles of conditioning. The models were
cleverly conceived, possessing clear deductive implications that
served to guide future research. The whole effort was modest in
scope and lucid in content. Later on his theoretical formalism
seemed to get out of hand. Elegance of formal style began to
operate as an end in itself; the empirical evidence required to
support the formal superstructure tended to be ignored. The
prime example of this excessive formalism was the Mathematico-
Deductive Theory of Rote Learning: A Study in Scientific Meth¬
odology (Hull, Hovland, Ross, Hall, Perkins & Fitch, 1940), which
was expressed in symbolic logic, mathematics, and stilted English
prose. Hull was so impressed with its formal structure that he
was led to make the following “modest” evaluation:
At the very least [it] is believed to represent in a elear manner

the form which the more scientific work on behavior of the future
should take (1952, p. 158).
History offered another evaluation. The Mathematico-Deductive

Theory of Rote Learning was largely ignored and quickly for-
gotten. It generated little researeh and failed to eneourage theo-

retieal attempts with the same formal eharacteristics.
Why the abysmal failure? First and foremost, the Mathe-
matico-Deductive Theory of Rote Learning was based upon an
inadequate assumption, viz., prineiples of eonditioning eould
alone handle the faets of verbal learning. The question raised is
whether this oversimplified and distorted view was encouraged
by the ceremonial formalism that dominated Hull’s thinking. My
historical interpretation suggests an affirmative answer; concern
with theoretical style was at the expense of empirical content.
This, however, need not be the inevitable consequence of the
search for deductive explanation. Deductive explanations can be
mediated by modest theoretical proposals as Hull and many of
his co-workers demonstrated (e.g., Spence, 1936; Miller, 1944).
Spence, I believe, was much more sensitive to the fine nuances of
experimental data than was Hull and more aware of the provi¬
sional nature of psychological theorizing. Spence felt no compul¬
sion to offer anything resembling a final theoretical solution.
Imprecision or incompleteness in model building did not bother
him as long as his conceptions could generate empirical predic¬
tions, the defining characteristic of a deductive model.
The criticism of excessive formalism could be directed, but
with less force, against Hull’s most significant effort. Principles of
Behavior (1943). The title is misleading in that the book deals
primarily with conditioning phenomena. (Hull thought they
would provide sufficient evidence for principles that would apply
to all forms of behavior, including social.) Principles of Behavior
proved to be successful both in identifying significant variables
and proposing fruitful hypotheses about learning. The book gen¬
erated much experimental research, probably more than any other
single publication in the history of psychology. In spite of its
heuristic value the argument could be advanced that Hull’s theory
was degraded by his overconcern with ceremonial formalisms.
Formal considerations outweighed empirical ones in the selection
of primary assumptions. Take, for example, his treatment of the
delay of reinforcement gradient mainly as a function of passage
of time. No consideration was given to the animal’s reactions
during the delay period, a factor that was later demonstrated to

be significant (Spence, 1956). One possible reason for HulFs
oversight is that it enabled him to express the delay of reinforce¬
ment principle in the elegant mathematical style that he desired.
One could generally agree with the criticism of ""ceremonial
formalism” that has been directed at Hull but still side with his
strategy. Psychology, at that time, had a pressing need for theo¬
retical hypotheses that could in principle be falsified. Although
Hull’s formalism may have been excessive for his own theoretical
task, it was pitched at the right level to attract the attention of
the psychological community who had to be ""shown” the virtues
of deductive explanation. (A disconcerting aspect of the history
of psychology is that all its ""great men” overstated their cases—as
did many who were not so great!) Hull’s efforts, in fact, served
as an impetus to the development of mathematical models of
learning. Although the post-Hullian mathematical models of
learning were much more sophisticated, particularly in regard to
the clarity of their empirical implications, they also can be faulted
for sacrificing empirical content for formal elegance. At the time
of their formulation many mathematical interpretations of prob¬
ability learning, concept identification, prisoner-of-war game, and
selective attention, seemed to be primarily formal exercises that
distorted by oversimplification the behavioral events they sought
to explain.
As an aside it should be noted that diametrically opposed
evaluations of a common methodological effort are possible with¬
out necessarily being contradictory. Particular mathematical mod¬
els of learning can be faulted as theoretically naive and of limited
empirical relevance and at the same time be praised for encour¬
aging the raising of explanatory standards in psychology.
The point of my historical excursion has not been to evaluate
the relative benefits of deductive precision or explanatory breadth
but instead to underline the difference between the formal as¬
pects of deductive explanation and the strategic methods used to
achieve that goal. Hull, in essence, argued for the strategy of
rigorousness at all stages of theoretical development. In contrast,
I have suggested (Kendler & Kendler, 1975) that deductive am-
biguity in the early stages of theoretical development, when the

empirical evidence is sparse and many theoretical alternatives are
possible, may be a more productive strategy than premature pre¬
cision that is overcommitted to one of several reasonable theo¬
retical options. Feigl (1970), a distinguished philosopher of sci¬
ence, concludes: *lt is a matter of controversy just how fruitful
or helpful strict axiomatizations are for the ongoing creative work
of the theoretical scientists.”
Other strategies are also possible. One of the great surprises
of my professional life occurred in 1970 during a conversation
about the unity of psychology with Abraham H. Maslow, who
was one of the major promoters of humanistic psychology and for
whom I served as an undergraduate assistant at Brooklyn College
in the late 1930s. I was at that time preparing an essay, along
with Janet T. Spence, entitled Tenets of Neobehaviorism, for a
memorial volume dedicated to Kenneth W. Spence (Kendler &
J. T. Spence, 1971). In analyzing the epistemological assump¬
tions of neobehaviorism, we identified the deductive model of
explanation as a primary one. I raised the question to Maslow
whether humanistic psychology aspired to this kind of explana¬
tion and was flabbergasted to hear an affirmative answer. He sug¬
gested that it would be achieved when the physiological aspects
of consciousness were more fully understood and thought the re¬
search of the kind in the book Tart edited. Altered States of Con¬
sciousness (1969), would lead ultimately to deductive theories in
humanistic psychology. As far as I know, Maslow never put such
notions in print, and there is much published in humanistic lit¬
erature that is at odds with it. Whether his comments in our con¬
versation were an expression of a temporarily expansive and opti¬
mistic feeling or the beginning of a new direction in his thinking is
something that will never be known. But his position at that time
does illustrate the possibility of adopting a goal of deductive ex¬
planation while simultaneously being concerned with issues (e.g.,
the nature of human experience, self-actualization) that appear
to be far removed from deductively structured formulations. Al¬
though Maslow could have been accused of harboring inconsis¬
tent concerns (deductive explanations versus a humanistic inter-
pretation) the fact of the matter is that deductivism, as a method

to achieve understanding, does not specify at what stage of a re¬
search program the demands of deductive explanation need be
considered and at what stage they should be met. The sole re¬
quirement seems to be that the goal is acknowledged, which im¬
plies an intention to achieve it. Conceivably more than a lifetime
of work might be required to obtain and transform information
about conscious experience and physiological processes into a
deductive theoretical system relevant to humanistic psychology.
Although I find it difficult to imagine how one could adopt
the goal of deductive explanation while ignoring it completely in
day-to-day practice, I cannot formulate any rule that would spec¬
ify exactly when one should become concerned about the deduc¬
tive capacities of one’s theoretical conceptions. (It should be clear
by now that the standard I am using for a deductive principle is
minimal: some logically determined prediction, expressed in ordi¬
nary language that could be embarrassed by inconsistent findings.)
The difficulty in specifying at what stage of a research pro¬
gram deductive practices should begin, and the level of rigor
they should possess, should not cause any surprise. Constructing
a successful deductive theory is purely an exercise in creative
thought. At the present we are unable to extract principles from
our knowledge of creative thinking that would suggest a step-by-
step plan for constructing a successful deductive theory. It is ques¬
tionable whether psychological principles of such power could
ever be formulated because the growth of knowledge, including
the formulation of novel theories, may be beyond prediction
(Popper, 1957). In any case, at least for the present, if not for¬
ever, we cannot evaluate the strategy employed to construct a
deductive theory. We are limited to judging the explanatory abil¬
ity of the theory itself.
Social criteria can be applied with reliability to assess the
logical correctness of a particular prediction. Although mathe¬
maticians may entertain some disagreement about the logic of a
specific deduction, for the most part, consensual agreement about
its correctness can be achieved. The further one gets from forms
of mathematical representation, or the closer one gets to ordinary
language as the medium for deductive explanation, the greater

will be the disagreement about the logical legitimacy of a particu¬
lar explanatory deduction. But these disagreements are usually
resolvable when the issues are clarified by discussion. And even
when a deadlock is reached historical pressures will begin to op¬
erate to demonstrate either that the deduction is true or false or
so ambiguous as not to deserve further consideration.
Inductive Processes in Deductive Theorizing
Is there an inductive logic that can guide the efforts of a theorist

in his attempt to formulate valid assumptions, that is, postulates
that will generate predictions in line with empirical evidence?
The answer depends upon what is meant by the term logic. If the
term is used in the conventional sense, reasoning according to a
set of explicit rules, then the notion of an inductive logic that
inevitably leads from empirical evidence to valid theoretical as¬
sumptions must be rejected. Although one can logically get from
theoretical assumptions to empirical statements, one cannot logi¬
cally get from empirical statements to theoretical assumptions.
The main reason is that theories say more than the data they
purport to explain. They not only summarize available evidence,
but they also possess implications for data not yet obtained. A
theory is more than the evidence it is designed to explain; the
data are less than the theory that interprets them. In essence, a
theory has something extra and that extra emerges from the crea¬
tive efforts of the theorist.
It is a mistaken notion to liken the problem of theory con¬
struction to an inductive logical problem that can be solved by
deductive means. Identify a preselected number from one to eight
after asking only three questions that can be answered by “Yes”
or “No.” By halving remaining alternatives (Five or above? Seven
or above? Eight?) the correct answer must be forthcoming after
the third query. Although the solution appears to be achieved by
inductive means, in reality it represents the deductive conse¬
quences of a simple probability model.
Empirical theories do not deal only with specific events or a
On Understanding: 1 Ill
finite set of events. A theory limited to the interpretation of the

influence of practice on the acquisition of classical conditioned
responses deals with an infinite number of empirical events. One
can never exhaust all possible experimental designs involving all
combinations of conditioned stimuli, unconditioned stimuli, and
response measures. Nor can one exhaust the infinite number of
conditioning trials. Thus no matter how well a given theory can
account for available evidence, no guarantee is possible that it
will be successful in interpreting new experimental data.
Hull (1943), influenced by Tolman’s tactics of theorizing
(1936), operated on the assumption that an inductive logic could
be used to determine the mathematical properties of theoretical
concepts. By investigating the influence of practice (number of
reinforced trials) in some ideal experimental situations in which
the influences of other significant variables were presumably con¬
trolled, it would be possible to conclude that the observed be¬
havior reflected without distortion the properties of the theoreti¬
cal process of habit formation. Hull, in fact, formulated his law
of habit growth from curve-fitting procedures applied to results
of parametric studies that measured behavioral changes (e.g.,
amplitude of the conditioned response) as a function of the
amount of training.
In principle there is nothing wrong with the procedure of
using experimental results as direct sources of theoretical notions.
But there is a world of diflFerence between considering the results
as logical inductions and as mere suggestions about the properties
of the theoretical process. In the former case one concludes a fait
accompli: the theoretical principle is valid and nothing remains
but to spin out its implications. In contrast, considering the evi¬
dence only as suggestive and tentative encourages one to criti¬
cally examine the adequacy of the theoretical assumption and
prevents one from becoming overcommitted to it.
Explanation in Physics and Psychology
Experimental psychologists have frequently been accused of suc¬

cumbing to scientism, the adoration of the methods of the natural
sciences, a false god for psychology. The argument offered is that

the methods of the natural sciences may not be appropriate for
those of the social, behavioral, or humanistic sciences; the “scien¬
tific method” must be adapted to the special needs of each dis¬
cipline.
The essence of such an argument revolves about the mean¬
ing of method. If one understands method to represent specific
kinds of research tools and techniques, then quite obviously each
scientific discipline must develop its own particular investigatory
procedures. Psychology needs standardized tests and operant con¬
ditioning apparatus; physics requires cloud chambers and low
temperature laboratories. However, if 'scientific method is con¬
sidered a systematic mode for arriving at warranted empirical
conclusions, then the method transcends the borders of various
scientific disciplines. Regardless of the particular procedures of
investigation used, the criteria of empirical truth reuiains the
same.
Psychologists who employ the deductive model of under¬
standing, for the most part, freely admit to the adoption of what
they consider to be the natural-science approach to psychology.
The accusation that their behavior as psychologists is being moti¬
vated by their desire to emulate physics appears to them to be
irrelevant to the problem of arriving at warranted conclusions. If
one maintains, they would argue, that psychology requires a dif¬
ferent conception of understanding than does physics, one is obli¬
gated to articulate the difference between “natural science” un¬
derstanding and “psychological” understanding. Although it is
difficult, if not impossible, to draw a precise demarcation line be¬
tween science and nonscience, this does not suggest that the divi¬
sion between the two should be blurred beyond any distinction.
Productive methodological discussions within psychology will
become impossible if the term science is robbed of all its signifi¬
cant meanings. An analogous, though extreme, example occurs in
a discussion about art in Tom Stoppard’s play Travesties.
Tzara: Doing the things by which is meant Art is no longer con¬

sidered the proper concern of the artist. In fact it is frowned
On Understanding: 1 ns
upon. Nowadays, an artist is someone who makes art mean
the things he does. A man may be an artist by exhibiting his
hindquarters. He may be a poet by drawing words out of a
hat. In fact some of my best poems have been drawn out of
my hat which I afterwards exhibited to general acclaim at
the Dada Gallery in Bahnhofstrasse.
Carr: But that is simply to change the meaning of the word Art.
Tzara: I see I have made myself clear.
Carr: Then you are not actually an artist at all?
Tzara: On the contrary. I have just told you I am.
Carr: But that does not make you an artist. An artist is someone
who is gifted in some way that enables him to do something
more or less well which can only be done badly or not at all
by someone who is not thus gifted. If there is any point in
using language at all it is that a word is taken to stand for a
particular fact or idea and not for other facts or ideas. I
might claim to be able to fly . . . Lo, I say, I am flying.
But you are not propelling yourself about while suspended in
the air, someone may point out. Ah no, I reply, that is no
longer considered the proper concern of people who can fly.
In fact, it is frowned upon. Nowadays, a flyer never leaves
the ground and wouldn’t know how. I see, says my somewhat
baffled interlocutor, so when you say you can fly you are
using the word in a purely private sense. I see I have made
myself clear, I say. Then, says this chap in some relief, you
cannot actually fly after all? On the contrary, I say, I have
just told you I can. Don’t you see my dear Tristan you are
simply asking me to accept that the word Art means what¬
ever you wish it to mean; but I do not accept it (Stoppard,
1975, pp. 38-39).
The natural-science oriented psychologist favors the deduc¬

tive model of understanding because of its value in achieving
publicly warranted empirical conclusions; the explanatory state¬
ments contained in a deductive explanation are available to all
rather than being merely in the minds of the person asserting
them.
Even if one were interested in uncovering the motive for
selecting a deductive model of understanding, other hypotheses,
besides the frank desire to mimic physics, are possible. For ex-
ample, Brodbeck (1962) argues that aeeepting the deductive

model does not mean that physics is being imitated but instead
that physics is being admired. The most compelling point about
the deductive model of explanation, she argues, is in its conclu¬
siveness. The persuasiveness of deductive explanation became ap¬
parent to me when a schoolboy. Gravitation finally became under¬
standable when I was exposed to the story of the discovery of a
planet in 1843. It had been noted that the planet Uranus followed
a peculiar and variable path around the sun. Astronomers de¬
duced from the theory of gravitation that some unknown body
must be exerting a gravitational pull on Uranus. As a result, the
location of a heretofore unknown planetary body, later to be
named Neptune, was predicted and discovered.
Particularly important for later discussion is the distinction
between two different aspects of the “conclusiveness” or “persua¬
siveness” of a deductive explanation: the public logico-empirical
component and the private intuitive sense of understanding. It is
difficult for me to recall exactly what weight each component had
in my achieving a sense of understanding of gravitation except to
note that the latter component was operative. That is, the deduc¬
tive explanation was phenomenologically fulfilling; it generated a
compelling sense of understanding. I am not suggesting that
every confirmed prediction gives rise to as dramatic a case of
intuitive understanding as did the discovery of Neptune for me.
Nor am I implying that for all individuals deductive explanations
are intuitively persuasive. My proposal is that deductivism, as a
public method of arriving at warranted conclusions about empiri¬
cal events, can provide a phenomenologically satisfying sense of
understanding.
The Role of Laboratory Experiments

in Deductive Theorizing
The example of the discovery of Neptune should put to rest any

argument that deductive explanations are possible only for an ex¬
perimental science. Darwin's theory of evolution is another case.
According to Ghiselin (1969), Darwin, in his naturalistic observa¬

tions of animal and plant life as well as coral reefs, was guided by
the goal of deductive explanation. Although the logical form of
his theories can be criticized, they did basically represent a de¬
ductive system that was testable by empirical evidence. Darwin,
himself, argued in favor of the deductive model of explanation
by emphasizing the importance of hypotheses in the collection of
empirical evidence. In one of his letters he wrote:
About thirty years ago there was much talk that geologists ought
only to observe and not theorize, and I well remember someone
saying that at this rate a man might as well go into a gravelpit
and count the pebbles and describe the colours. How odd it is
that anyone should not see all observations must be for or against
some view—if it is to be of any service (Darwin, 1903, p. 195).
Perhaps Darwin’s most acute statement concerning the rela¬

tionship between theory and observation was, “I have an old be¬
lief that a good observer really means a good theorist” (1903,
p. 176).2
Analysis of clinical data can also be organized into a deduc¬
tive explanatory system; an example is the attempt to explain the
changes over the years that have occurred in the incidence of
peptic ulcers among men and women. Studies conducted before
1900 indicated that peptic ulcers occurred more frequently among
women. About 1910 the incidence for both sexes became equal;
since 1910 there has been a progressive rise in the occurrence of
peptic ulcers among men, with a decrease in the incidence among
women. In the 1960s four times as many men as compared to
women had peptic ulcers. It has been suggested (Rosenbaum,
1967) that these changes can be accounted for by changes in so¬
cial attitudes about dependency and aggression for men and
2. Darwin was at times inconsistent in describing his scientific strategy.

In an autobiographical sketch, he confessed to working “on true Baconian
principles, and without any theory collected facts on a wholesale scale”
(Darwin, 1887). Nevertheless, in numerous other comments, including the
above two, and more importantly in his actions, Darwin (Ghisehn, 1969)
demonstrated the significant role that hypotheses with deductive conse¬
quences played in his scientific efforts.
116 Psychology: A Science in Conjiict
women; the inhibition of the needs to express dependency and

aggression encourages peptic ulcers.
In the freer and more independent socioeconomic society of
the nineteenth century, men were not inhibited about expressing
dependency and aggression. It was socially acceptable for them
to be active and aggressive in work and to be mothered by their
wives at home. In contrast, women could satisfy their dependency
needs at home, but being aggressive was considered unfeminine
and therefore unacceptable. After the turn of the century, more
men found jobs in large organizations with many superiors, and
consequently they were forced to inhibit aggression. It also be¬
came less fashionable to be dependent, ’both in the eyes of other
men as well as women. On the other hand, it became more so¬
cially acceptable for women to express aggression along with
dependency. It follows from such an analysis that if the wom¬
en’s liberation movement is successful in freeing women from
their conventional social roles, then the incidence of peptic ulcers
among them will increase. Discouraging dependency while not
modifying dependency needs is one important factor. In addition,
opening jobs to women in large organizations that inhibit aggres¬
sive behavior is another variable that should, if Rosenbaum’s the¬
ory (1967) is valid, increase the incidence of peptic ulcers among
women.
This is not to say that the logical rigor of the dependency-
aggression formulation as it applies to the women’s liberation
movement is free of all ambiguity nor that clearcut measures of
needs to be dependent and aggressive are readily available. Nor
does it imply that if operational definitions were available to test
the theory and that if the predicted increase in the incidence of
peptic ulcers among women occurred, then the theory would be
confirmed, and if not, falsified (pages 142-151). Nor is it being
stated that deductive theories of natural-occurring events in the
behavioral sciences can achieve the same predictive capacity as
theories of experimental phenomena. The only point being made
here is that a deductive theory that applies to clinical events in a
historical perspective can be formulated and tested.
The major thrust of my analysis is that the deductive model
On Understanding: 1
of understanding can be adopted for behavioral events whether

m the form of experimental, clinical, or naturalistic evidence. If
one accepts the notion that understanding is achieved by deduc¬
tive means, then a way can be found to formulate psychological
hypotheses that can be checked against observational events.
The Reality of Deduetive Theories
Should deductive theories be conceptualized as representing real¬

ity or treated as man-made fictions? Arguing for the former posi¬
tion, William Whewell (1794-1866), the English philosopher ex¬
pressed a realist position:
The cultivation of ideas is to be conducted as having for its object

the connexion of facts; never to be pursued as a mere exercise of
the subtilty of the mind, striving to build up a world of its own,
and neglecting that which exists about us. For although man may
in this way please himself, and admire the creations of his own
brain, he can never, by this course, hit upon the real scheme of
nature. With his ideas unfolded by education, sharpened by con¬
troversy, rectified by metaphysics, he may understand the natural
world, but he cannot invent it. At every step, he must try the
values of the advances he has made in thought, by applyine his
thoughts to things (1858, pp. 184-85).
Louis L. Thurstone (1887-1955), the famous mathematical

psychologist, takes an opposite view:
The constructs in terms of which natural phenomena are com¬

prehended are man-made inventions. To discover a scientific law
is merely to discover that a man-made scheme serves to unify,
and thereby to simplify, comprehension of a certain class of natu¬
ral phenomena. A scientific law is not to be thought of as having
an independent existence which some scientist is fortunate enough
to stumble on. A scientific law is not a part of nature. It is only
a way of comprehending nature (1935, p. 44).
This argument has been with us for centuries. It is still being

pursued, perhaps because it can be so much fun, especially over
cocktails. The fundamental issue within the pragmatie framework

of this book is whether the choice to view a deductive theory as
a reflection of reality or as pure fiction has any effect upon one’s
activities as a scientist. Let us examine some of the implications
of adopting either of these apparently competing views.
Adopting the realist position requires the assumption of a
reality existing beyond one’s sensory observations. This conjec¬
ture ean be aceepted without neeessarily becoming entrapped in
metaphysical quicksand. One can, in a sophisticated manner, ac¬
cept a naive realist view that a reality exists beyond sensory ob¬
servation but admit to the impossibility of demonstrating it. If
only this is done then the practical effect of the difference be¬
tween the two views, uncovering truth or superimposing order,
becomes negligible. By examining most scientists’ empirical and
theoretical practices, one would find it difficult to distinguish be¬
tween those who intuitively hold an “uncovering” versus “super¬
imposing” orientation toward theory construction.
If it has no effect why do many scientists bother with this
issue? Courant (Courant & Robbins, 1941), a famous applied
mathematieian, offers a clinical judgment: realists eling to their
metaphysical assumption because it is a “psychological hardship”
for them to aecept theoretical constructs as fictions; they require
a crutch of reality to satisfy their need to think in concrete terms.
Courant’s conjecture can be considered relevant to the rela¬
tive worth of psychological interpretations based upon “real”
neurophysiological events (e.g., hypothalamic functioning) or
“fictitious” processes (e.g., cognitions, habits) related to environ¬
mental faetors, both past and present (e.g., amount of training).
Do neurophysiological explanations of psychological events (e.g.,
auditory attention as a function of neurological events in the
cochlear nerve and the auditory cortex) provide a depth of under¬
standing that is unmatched by purely environmental-behavioral
conceptions (e.g., auditory attention as a function of the pattern
of stimulation) ? Or, on the contrary, are these two kinds of knowl¬
edge fundamentally equivalent? To answer these questions one
must analyze: (1) the epistemological status of psychophysiologi-
cal and environmental-behavioral concepts and (2) the meaning
of reductionism.
Physiological Versus Black-Box Theories
Many psychologists concerned with neurophysiological bases

of behavior feel their task is difFerent from the black-box theorist
who seeks to formulate deductive theories that employ abstract
theoretical constructs to bridge the gap between environmental
manipulations and the behavior of organisms. Believing that they
are dealing with the ‘Teal causes’’ of behavior, they feel that iden¬
tifying the permanent physiological structures involved in behav¬
ior, and ascertaining their mode of functioning, is sufficient. De-
ductivism, as a means of explanation, seems to play at best an
insignificant role in the theoretical construction. They see as being
different the efforts of the behavior theorist, who formulates ab¬
stract theoretical constructs defined in terms of environmental
manipulations (e.g., drives, organization, mental structures, asso¬
ciation) that account for the behavioral laws observed.
The difference between the two approaches is usually high¬
lighted at colloquia. The biopsychologists begin with a minimal
introduction that identifies the behavior (e.g., the maintenance of
bodily temperature) and the physiological structures involved,
and with an enthusiastic description of the experimental equip¬
ment. Then a series of slides, always at least one too many, are
flashed on the screen that illustrate the empirical relationships
between the behavior and the functioning of the physiological
structures. Since the facts speak for themselves, conclusions are
kept to a minimum. Facts, facts, and more facts are the order of
the hour.
In contrast, the colloquia of the behavioral psychologists usu¬
ally include a lengthy introduction containing a modest statement
of the profound significance of the research topic, an extolment of
the theory that guided the research, and a disparagement of “su¬
perficial” competing formulations. Next comes a “justifiably” ex¬
tended explication of a new experimental design that veritably
controls all extraneous variables, followed by a few slides report¬
ing data. A final conclusion involves a quaint combination of of¬
fensive and defensive remarks about the theoretical and empirical
aspects of the talk, a bit of philosophy of science, and optimism
about future research and theoretical developments. Although

facts are reported the speculation-fact ratio is often embarrass¬
ingly high.
Perhaps the caricatures are excessive, but they drive home
the point that the eflForts of biopsychologists are usually much less
abstract than those of the behavioral theorist. This diflFerence may
result from a selection process that attracts abstract thinkers to
behavioral psychology, or from the demands for different think-
ing styles in each field, or, most likely, from a combination of both
factors. The significant issue is whether each field obtains distinc¬
tive kinds of knowledge or is just that each encourages different
research strategies and theoretical practices.
Hypothetical Constructs Versus Intervening Variables
MacCorquodale and Meehl (1948) sought to distingish be¬

tween physiological and purely behavioral theories by contrasting
two types of theoretical constructs, hypothetical constructs and
intervening variables. The former “involve the hypothesization of
an entity, process, or event which is not itself observed . . .
[while the latter] do not involve such hypothesization.” These
writers demand “of a theory . . . that those elements which are
hypothetical . . . have some probability of being in correspon¬
dence with the actual events underlying the behavior phenom¬
enon . .
To appreciate the implications of their suggestion one must
decide whether the proposed hypothesization of an entity, process,
or event is intrinsic to the deductive capacity of the formulation
or merely an unessential, speculative by-product. To answer this
question consider the example MacCorquodale and Meehl chose
to buttress their distinction; they refer to some of Clark Hull’s
theoretical articles and state:
We suspect that Professor Hull himself was motivated to write

these articles because he considered that the hypothetical events
represented in his diagrams may have actually occurred and that
the occurrence of these events represents the underlying truth
about the learning phenomena he dealt with (1948, p. 104-5).
Although it may be clinically justified to assume that Hull

had conceptions of physiological mechanisms coordinated to some
of his theoretical constructs, the meaning of these constructs—
operational, empirical, and theoretical (pages 87-94)—are not
tied to the personal cogitation of the theorist. The deductive ca¬
pacity of these theoretical constructs when embedded in their
theoretical structures is not formally changed one whit by the
theorists’ “physiologizing.” Consider the confusion that would re¬
sult if it were. At the time Hull was speculating about physiologi¬
cal processes coordinated to his theoretical constructs, Spence,
who played the role of a junior partner in the Hull-Spence theo¬
retical enterprise, was denying that these constructs had any
physiological implications. Should the concept of habit and other
concepts that Hull thought might be related to physiological
events have two or possibly more interpretations, not depending
on their stated meanings but instead on the intuitive physiologi¬
cal conceptions of particular individuals? Or consider the other
side of the coin; an intervening variable, a theoretical construct
designed to summarize environmental-behavioral relationships, is
found to be coordinated to some specific neurological process.
Should that be a source of embarrassment because of the pre¬
sumed nonphysiological character of the intervening variable?
Of course not! It would be a fortunate case of serendipity that
would enlarge the empirical scope of the theory involving that
intervening variable. In sum, unnecessary confusion can result by
impregnating the theoretical constructs with the physiological in¬
tuitions a theorist may be entertaining.
The distinction between hypothetical constructs and inter¬
vening variables achieves its greatest importance when referring to
theoretical formulations that are, or are not, clearly embedded in
physiological processes. Examples are the many etiological theo¬
ries of schizophrenia that seek to explain genetic involvement
(Heston, 1970). One such formulation (Stein & Wise, 1971) hy¬
pothesizes that a genetically determined metabolic disorder pro¬
duces a chemical compound (6-hydroxydopamine or a closely re¬
lated substance) that damages the reinforcement system, which
in turn produces two behavioral consequences: (1) a disruption
of the continuity of goal-directed behavior and (2) a reduced ca¬

pacity to experience pleasure. These two behavioral consequences,
it is suggested, resemble two major symptoms of schizophrenia:
(1) diminished capacity to maintain goal-directed behavior, espe¬
cially in regard to continuity of thought processes, and (2)
anhedonia, a deficiency in experiencing pleasure. Such a theory
with its primary assumption of a malfunctioning of the reinforce¬
ment process has definite psychophysiological implications for the
eiBFect of certain biochemical interventions upon the functioning
of the reward centers in the hypothalamus. Repeated injections of
6-hydroxydopamine in rats reduced by 60 percent the rate of bar
pressing reinforced by electrical stimufation to the hypothalamic
rewards centers. The negative effect exerted by 6-hydroxydopa¬
mine on the reinforcement mechanism can be counteracted by
the tranquilizing drug chlorpromazine, which is effective in elimi¬
nating some schizophrenic symptoms. The rate of bar pressing is
restored to normal levels when chlorpromazine is injected into
rats previously treated with 6-hydroxydopamine.
The intent of this discussion is not to justify the use of rats
for research on schizophrenia nor to examine the validity of the
Stein-Wise model (which has only been briefly sketched) but in¬
stead to differentiate between those physiological conjectm:es that
function as integral components of a theory and those that do not.
A purely behavioral theory of schizophrenia could be formulated
that assumes a dysfunctioning of the reinforcement process. The
Stein-Wise model does something more. It not only postulates that
schizophrenia results from a genetically determined dysfunction¬
ing of the reinforcement process, but it also specifies a physiologi¬
cal locus (hypothalamus) and a biochemical basis (6-hydroxy¬
dopamine or a closely related substance). If these assumptions
were proven false, the guts of the theory (how much more physio¬
logical can one get!) would be destroyed. What benefits accrue
from specific physiological postulates if they increase the proba¬
bilities of being wrong? A theory of schizophrenia that restricted
itself solely to the behavioral aspects of reinforcement would not
be embarrassed to find that genetic involvement is not mediated
by a metabolic disturbance, that the reinforcement process is not
localized in the hypothalamus, and that 6-hydroxy dopamine, or a

related substance, is not involved.
Presumed Advantages of Physiological Hypotheses
Three possible reasons could be given for proposing physio¬

logical hypotheses of behavior as integral assumptions within a
theory. One, consistent with the commonly held belief of many
biopsyehologists, is that the real causes of schizophrenia are
physiological in nature, and eonsequently, a theory that truly
explains schizophrenia must be expressed in terms of physiologi-
eal processes. The trouble with such an assumption is that it
opens up a homefs nest of metaphysical problems associated
with the concept of real causality. If some can argue that the real
cause of behavior occurs at the physiological level, others, with
equal justification, can claim that ‘‘true explanations” reside in
physiochemieal or even atomic and subatomic interpretations. In
addition, the concept of a true explanation might also exert a
suffocating effect on research and theorizing by inhibiting fur¬
ther efforts toward a deeper understanding, a point made by
Orville Wright in 1903: “But if we all worked on the assumption
that what is accepted as true is really true, there would be little
hope of advanee” (Wright, 1953, p. 314).
The second justification for proposing physiological hypothe¬
ses as an integral part of a behavioral theory is that it increases
the empirical content of the theory. Risking disproof by framing
one’s psychological theory in physiologieal terms enables one to
reap a greater theoretical bounty. In addition to deducing conse¬
quences of environmental manipulations, psychophysiological re¬
lationships can also be explained. And by so doing, the theory
holds greater promise for precise theoretieal predictions as well
as behavioral control. If schizophrenia were caused by a dys-
funetion in the reinforcement meehanism, knowing its physiologi¬
cal locus and biochemical features would improve the chances of
modifying the disordered behavior.
Although a greater empirical content can result from physio¬
logical hypotheses, such a consequence need not follow. The
crucial factor is the empirical realm to whieh a purely behavioral

theory and a biopsychological conception is relevant. If a bio-
psychologieal theory interprets the known environmental-behav¬
ioral relationships as well as additional physiologieal events, then
the biopsychological formulation obviously has the advantage of
greater empirical breadth. However, a purely behavioral theory
designed to explain environmental influences might be capable of
interpreting facts outside the realm of a biopsychological theory.
Suppose a physiologieal model of schizophrenia could explain all
schizophrenic behavior (e.g., anhedonia, odd verbal and eogni-
tive behavior, hallueinations) in terms of underlying neurological
and biochemical events. In addition, it would suggest effective
treatments that would eliminate schizophrenie behavior. Such a
formulation, however, would not identify the environmental fac¬
tors (e.g., low socioeconomie conditions, unfavorable interper¬
sonal relationships) that eneourage sehizophrenia in those who
are genetically vulnerable (Meehl, 1962). Although identifying
the environmental inputs that influence the onset of schizophrenia
would be greatly facilitated by knowledge of underlying physio¬
logieal proeesses, such knowledge is not automatieally included
in biopsychological formulations. Consequently, the claim that
physiological theories always possess a superior explanatory ea-
pacity as compared to purely behavioral formulations must be
rejected.
The third argument to favor a physiological over a black-box
conception is based upon strategic considerations. Looking under¬
neath the skin for causal factors has a greater likelihood of achiev¬
ing suecess. One “knows,” unless one is a mystic, that behavior is
a consequence of processes and events within the organism; some¬
thing there must be related to the psyehological phenomena that
one seeks to understand. The events outside the skin are so vast,
and so difficult to control and to synthesize into some coherent
explanatory system, that the chances of successful theorizing are
unpromising.
Black-box theories have been dominant in most areas of psy¬
chology: social, learning, memory, to name a few. All of these
areas have been plagued by persistent and unresolvable theoreti-
cal disputes (e.g., attitudinal consistency, incremental versus one

trial learning, the number of memory stores). One of the main
reasons that these major theoretical disputes go unresolved^ is
that black-box theories do not provide sufficient constraints to al¬
low for resolutions of theoretical disputes. The environmental
operations are too far removed from the underlying biological
mechanisms that are responsible for behavior. As a consequence
numerous theoretical options are available to handle embarrass¬
ing data. The result is that the core assumptions of black-box
theories are able to survive regardless of what results are ob¬
tained, thus effectively preventing one theory from achieving
dominance over its competitors. In contrast, physiological theo¬
ries that identify the operation of permanent structures or bio¬
chemical reactions as being responsible for given forms of behav¬
ior have relatively fewer options in defending their truth value
when their predictions fail. One can always express reservations
about the experimental procedures employed, but when repeated
attempts fail to provide any evidence in favor of the hypothesized
physiological involvement, one is forced to abandon one’s conjec¬
tures. When, however, positive evidence is obtained, one becomes
confident that the theoretical assumptions are generally correct,
even though many unanswered questions remain.
The main support for the methodological conclusion that
theoretical controversies are more easily resolved with physio¬
logical as compared to black-box theories is not based upon logi¬
cal considerations but instead upon historical evidence. Black-box
theories, in the past, have been notoriously unsuccessful in re¬
solving their theoretical disputes. But reservations about the im¬
plications of this historical evidence can nevertheless be expressed.
Past failures need not prevent future successes. Black-box theories
could become more sophisticated and precise thus enabling theo¬
retical resolutions to become possible. In addition, it should be
recognized that even if the methodological advantages of physio¬
logical theories are acknowledged, it does not follow that they
3. Several important methodological issues, which will be analyzed later
(pages 142-151), are involved in evaluating the relative merits o£ compet¬
ing theories and the survival of one over others.
need to be formulated at the beginning of all research programs.

It may be strategically desirable to gather systematic environ¬
mental-behavioral relationships to guide one s future physiologi¬
cal hypotheses. That is, ignoring physiological processes while
environmental-behavioral relationships are being gathered may
prove to be the most direct approach to a physiological under¬
standing of behavior. One could point to the history of sensory
psychology where psychophysical relationships proved useful in
developing psychophysiological theories.
Realistic Versus Black-Box Theories
Some would disagree with the above analysis by arguing

that physiological theories oflFer a qualitatively diflFerent kind of
knowledge than that provided by black-box conceptions. This lat¬
ter position (e.g., Harre, 1970) would maintain that a nonrealistic
theory organized in a hypothetic-deductive fashion is, episte¬
mologically speaking, not equivalent to a “realistic” theory that
attributes phenomena to the operation of permanent structures
(e.g., hypothalamus, auditory cortex, blood chemistry). Instead
of assuming, as purely abstract deductive theories do, that logical
order reflects natural order, the realist position maintains that
empirical events are truly explained only when the operation of
the permanent structures responsible for the phenomena are de¬
scribed. This position suffers from an ambiguity in the meaning
of permanent structure and from inconsistencies stemming from
the presumed qualitative difference between structural and de¬
ductive explanations.
The realist position, in its strong form, suggests that a true
explanation depends on knowing how a “real” structure operates
to produce a given phenomenon. If one considers, for example,
the theoretical concept of gene it becomes difficult to know ex¬
actly when the requirement of a permanent structure is met:
some bodily entity, a mechanism within the reproductive system,
a germ cell, the chromosome, the DNA molecule? Before the
DNA molecule was decided upon as the unit of heredity, accurate
deductive predictions were made about a variety of genetic phe-
nomena (e.g., human eye eolor). The fundamental change that

did occur as the gene acquired a more permanent form was an
enlargement in the empirical realm of genetic theory as well as a
potential increase in its ability to control genetic phenomena. But
these changes are quantitative, not qualitative. A common mis¬
conception is that knowledge of the operation of a permanent
structure, like the DNA molecule, removes the need for deductive
logic in explaining events. Since the structure causes an event no
other explanation is required. However, the discovery of the DNA
molecule raises a host of questions concerning the internal func¬
tioning of the molecule as well as its interaction with other bio¬
logical structures and processes. In answering these questions
theoretical speculations are required and their truth value is de¬
termined by how well their deductions coincide with empirical
findings.
To bring this discussion closer to psychology, consider the
consequences of the technical breakthrough that enabled the elec¬
trical activity of a single neuron to be recorded. Although this
development permitted one to observe the operation of a funda¬
mental permanent structure, the need for abstract theorizing ac¬
tually increased in order to explain how behavior was influenced
by the interaction among independent neurons. Regardless of how
deep one penetrates into the physiological substrata of behavior,
abstract theoretical problems will always remain. Contemporary
physics, with its persistent search for subatomic particles that can
never be directly observed, supports this expectation.
The lack of a functional diflFerence between realistic and
black-box theories is highlighted when total ignorance reigns
about the identity or functioning of a structure responsible for a
set of phenomena. In such a case, according to the realist position
an iconic model should be invented that speculates about the
structure responsible for a given event. Is anything gained by this
methodological prescription except the suggestion that ultimately
it would be wise to look for the psysiological correlates of behav¬
ior? The prescription can backfire when, in the absence of any
firm notions of the relevant structure, a rash of unfounded specu¬
lations are encouraged that may carry an aura of respectability
and a source of persuasion, but nothing else. These kinds of con¬

jectures have been derisively treated as attempts to change the
meaning of the C.N.S. (central nervous system) to the Conceptual
Nervous System. Iconic models, however, can serve some benefi¬
cial functions. They can facilitate theoretical efforts by concretiz¬
ing abstract relationships. They can guide future research when
technical developments enable the direct investigation of the op¬
eration of the hypothesized structure. They can also encourage
the psychological theorist to employ available physiological evi¬
dence to guide and restrict his speculations. But these functions
do not create any qualitative difference between an iconic model
and a pure black-box conception. There is always the possibility
that a theoretical construct within a black-box conception can be
coordinated to a physiologieal structure even though the theorist
had no such intention.
The key question in this debate is what determines the truth
value of iconic models prior to their success, or failure, to be
coordinated to a permanent structure. The answer would seem to
be their deductive capacity. If they can interpret empirical events,
their truth value is enhanced. If they are unsuccessful, modifica¬
tion or rejection of the model is in order. If later events prove
that the structural speculations are justified, then the iconic model
would have a greater explanatory capacity than an equivalent
black-box theory. If, however, the structural speculations fail to
gain any empirical support, but the explanatory capacity of the
iconic model in regard to environmental-behavior relationships is
supported, then the basic model would be retained while the
structural speculations would be discarded. Thus, the conclusion
is that physiological theories, either with clearly measurable
physiological implications or purely speculative ones, are episte¬
mologically similar to black-box conceptions because both em¬
ploy the same litmus test of truth value, viz., their capacity to
generate empirically supported deductions.
The conelusion that physiological and black-box theories are
epistemologically similar does not mean that they do not differ
strategically. Historical evidence, as noted, suggests that when
physiological operations are involved theoretical controversies are
more easily resolved than when the eompeting eoneeptions eon-

tain only abstraet eonstruets. The strategic advantage of physio¬
logical theories that are based upon physiological operations do
not necessarily, if at all, extend to physiological speculations. To
merely assume that some phenomena result from structures within
the body does not automatically remedy the excessive lack of con¬
straints that characterize black-box conceptions. Some precision
is required to benefit from physiological assumptions. Although
successful strategies cannot be spelled out either in advance or in
detail, one useful rule of thumb is that if physiological hypotheses
are to pay off by increasing the precision of psychological theoriz-
ing, then they must be coordinated to experimental operations
that either intervene in bodily function or measure the ongoing
activity of a given structure. The technological revolution in bio¬
psychology has drastically diminished the justification for specu¬
lating in general about physiological functioning in the absence
of any experimental intervention.
Empirical and Theoretical Reductionism
Now that epistemological and strategic considerations have

been considered in regard to the comparison between physio¬
logical and black-box theories, the concept of reductionism can
be analyzed. Like all important methodological concepts, reduc¬
tionism suffers from a multiplicity of meanings. As a first approxi¬
mation let me describe reductionism as used in psychology as the
attempt to interpret psychological phenomena in terms of under-
lying physiological processes. This possible kind of interpretation
can be accomplished in several different ways: (1) relating be¬
havior directly to physiological variables, (2) adopting a model
of behavior based on physiological evidence, or (3) coordinating
a I. ehavior theory to a physiological theory.
The most common type of psychophysiological interpretation
occurs when behavior is related to some physiological character¬
istic of an organism. Some examples are: intellectual retardation
as a function of phenylketonuria (a disorder that is transmitted as
a simple recessive genetic characteristic involving a disturbance
ISO Psychology: A Science in Conflict
in protein metabolism and that can be alleviated by appropriate

diet), rate of bar pressing as a function of reinforcement site in
the hypothalamus, language behavior as a function of the size and
site of cortical damage, and behavior of the heroin addict as a
function of methadone treatment. Nonetheless, these simple psy-
chophysiological laws can be considered as empirical relation¬
ships in which the dependent variable, behavior, is related to an
independent variable representing some aspect of physiological
functioning. As already argued, the epistemological characteristics
of such laws are no different from those in which behavior is
found to be a function of environmental stimulation (e.g., visual
acuity as a function of level of illumination) or of past behavior
(e.g., criminal behavior as a function of criminal record). They
differ only in the characteristics of the independent variable
(physiological, environmental, or behavioral).
Another way of interpreting behavior within a physiological
framework is to use physiological knowledge and speculations as
ideas for formulating behavioral assumptions: the characteristics
of the theoretical assumptions are suggested by physiological pro¬
cesses. An example is the attempt to formulate hypotheses about
cognitive development with the assistance of facts and conjectures
about neurological development (H. H. Kendler & Kendler, 1975).
The ‘ physiologizing,” strictly speaking, is not part of the theory of
cognitive development; it represents an independent model that
is not directly involved in the explanatory capacity of the theory
(Lachman, 1960). The question that has been raised about such
formulations is whether the physiological speculations are useful
in the formulation of fruitful hypotheses and in making the theo¬
ries more precise. The tentative answer offered is that their use¬
fulness is proportional to their ability to initiate physiological in¬
vestigations.
The final manner in which psychology is interpreted in terms
of physiology is that of theoretical reductionisms: a theory of be¬
havior is subsumed under a theory of physiology, and as a result,
the physiological theory becomes capable of explaining psycho¬
logical phenomena. Although the fundamental notion of theoreti¬
cal reductionism from behavior to physiology appears simple, a
On Understanding: 1 ISl
careful scrutiny of the problem suggests otherwise. The first point,

previously made, is that complete predictability does not neces¬
sarily occur when successful theoretical reductionism takes place.
A physiological theory of schizophrenia, which would designate
all the physiological processes involved in the disorder, would
not necessarily identify all the environmental determiners even
though such a task would be made much simpler. A similar situa¬
tion would prevail for a theory of color vision that would give a
complete physiological interpretation of color vision without ac¬
counting for the totality of empirical relationships between light
waves and color sensations.
Subsumed can have another meaning, a much more demand¬
ing one. A theory of behavior is subsumed under a theory of
physiology when each term of the psychological theory is coordi¬
nated with one of the physiological theory. If this is accom¬
plished, as in the previously given examples of schizophrenia,
then the physiological theory can completely predict the psycho¬
logical phenomena, e.g., the relationship between interpersonal
stress and schizophrenia. By coordinating the physiological con¬
structs with the psychological constructs one bridges the gap be¬
tween physiological and psychological knowledge.
Theoretical reductionism, in which the terms of a physiologi¬
cal and a psychological theory are coordinated, is presently more
in the realm of science fiction than that of reality. To accomplish
reductionism of this sort requires highly formalized theories of
both physiology and behavior. Perhaps the closest psychophysi¬
ologists have come to this standard is in the area of sensory
physiology where psychological phenomena (e.g., Mach Bands)
have been coordinated to neurological processes (e.g., lateral in¬
hibition ). But even such successes fall short of the formal elegance
demanded by theoretical reductionism although these efforts nev¬
ertheless do encourage the belief that the goal of theoretical re¬
ductionism between psychology and physiology is within striking
distance for some sensory phenomena.
Although one can specify the criteria of theoretical reduc¬
tionism, one cannot guarantee that reducing psychology to physi¬
ology can be accomplished. Kaplan vigorously expresses this view-
1S2 Psychology: A Science in Conflict
point when answering the question of whether psychology is

reducible to physiology (or chemistry or physics):
An affirmative answer ... as a matter of doetrine, seems to be

. . . both unwarranted and scientifically useless. . . . As a meth¬
odological presupposition ... it provides a valuable perspective
in which the behavioral scientist can see continuing possibilities
for turning to his own use the findings of other sciences. But the
significance of such possibilities appears only as they are actual¬
ized; otherwise their affirmation expresses only a hope. ... It
may well be that the psychologist can derive his whole discipline
from neurology, biochemistry, and the rest; it is not destructive
skepticism but productive pragmatism to say, ‘I’d like to see him
do it!” (1964, pp. 124-25).
Such a position might be considered inconsistent with the

doctrine of a unity of sciences in which it is assumed that the
various disciplines can be arranged hierarchically (e.g., physics,
chemistry, biology, psychology) to reflect the potential theoreti¬
cal reduction of any discipline to a more basic one. This assump¬
tion implies a determinism across disciplines^ by accepting the
notion that a less basic discipline (e.g., psychology) can in prin¬
ciple be explained by a more basic discipline (e.g., biology). By
refusing to acknowledge the inevitability of theoretical reduction-
ism from psychology to physiology, does not one concede the
possible operation of some emergent process that precludes theo¬
retical reductionism? Not necessarily. Unsuccessful attempts at
theoretical reduction to physiology need not result only from un¬
predictable emergent qualities of psychological events. It could
result from imperfect knowledge of psychology or physiology or
from inadequate mathematical tools with which to perform the
theoretical reduction. In the absence of a successful reduction it
is difficult to conceive how the question of reductionism from psy¬
chology to physiology can be resolved conclusively; we cannot in
principle deny the possibility of reductionism, but neither can we
4. One can reject this assumption without abandoning determinism.

One can adopt an antireductionist position that assumes each discipline pos¬
sesses emergent properties that cannot be predicted from a “lower” disci¬
pline while still accepting strict determinism for events within a discipline.
On Understanding: 1 13S
be certain that it is possible. Theoretical reductionism appears

to be an admirable operating assumption but not necessarily a
valid one.
Arguments have been advanced that insist that emergent
psychological properties preclude the theoretical reduction of
psychology to physiology. One such argument emerged from
Brentano’s phenomenological distinction between intentional and
nonintentional acts. Turner summarizes the problem this way:
[An intentional act] is a mental activity culminating in a deci¬

sion which at the behavioral level is the occasion for some
behavior selected from a set of behavioral options. In the lan¬
guage of intention, then, this activity underlies the motive for the
particular decision. But the actual behavior reflecting the particu¬
lar decision requires a different type of description, and as such,
is reducible to physicalistic terms. On the other hand, the phrase
“wanting to go to the movies” designates an intentional act in
which the wanting is about going to the movies but which itself
is not an objectively describable act. The two languages are quite
different. One is mentalistic and intentional, the other is physical¬
istic and nonintentional, with neither a logical nor empirical
bridge between the two. If one now wishes to maintain that psy¬
chology should concern itself with motivational explanation in the
intentional language, and this would distinguish it logically from
the physical sciences, then, to be sure, the idea of reduction
would be incompatible with our understanding of psychology and
physiology (1965, p. 347). .
Turner later argues, and I agree, that one is not required to

accept the insulation of intentions as mental events from publicly
observable physical events, behavioral or physiological. Intentions
can be treated as theoretical constructs inferred from environ¬
mental-behavioral relationships, e.g., individual A develops an
intention to attend the movies every Saturday night. Or intention
can be treated as a dispositional concept inferred from behavior,
e.g., individual B has an intention to go to the movies because he
asks the question, “Are you interested in going to the movies to¬
night?” They can also be treated as physiological processes if it
were possible to coordinate specific, for example, cortical events
with introspective reports. This line of analysis suggests that the

so-called mental intention can possibly be treated either as a mo¬
tivational construct or a physiological construct. Within this con¬
text the reduction of the concept of intention, as an inferred
phenomenological event, becomes a reasonable possibility. The
previous point that psychology as a behavioral science can be re¬
duced to physiology can be extended to the relationship between
phenomenology on the one hand and behavioral psychology
and/or physiology on the other. Such reductions are within the
realm of the possible, although one cannot guarantee that they
will necessarily occur. And it is important to recognize that the
kind of reduction of intention that is being discussed is concerned
with explaining phenomena in a deductive sense, not in describ¬
ing its experiential characteristics. This point returns us to the
previous analysis of knowing and having, the former being con¬
cerned with causes, the latter with experience.
A pragmatic and fruitful way of viewing the problem of
emergent properties of psychological phenomena is to recognize
that it is fundamentally a theoretical issue rather than a purely
empirical one (Nagel, 1961). Assume that a particular psycho¬
logical theory postulates emergent perceptual or cognitive prop¬
erties that cannot be derived from physiological theory. One
should not fault the theory because of its reductionistic “failure.”
An evaluation should be based upon its explanatory capacity. But
no matter how successful the theory may be, one cannot conclude
that its emergent behavioral assumption, which is not deducible
from physiological knowledge, is necessarily “true.” The possi¬
bility always remains that a subsequent formulation might achieve
a reductionistic breakthrough by successfully deducing the to¬
tality of psychological laws from physiological assumptions. Al¬
though at present behavioral psychology need not fear being en¬
tirely consumed by physiological psychology, one must recognize
that in principle this is always a possibility. Consequently, a wise
methodological strategy to pursue is to ignore a priori claims that
psychology is an emergent discipline or that it will ultimately be
subsumed under physiology. This does not imply that either al¬
ternative cannot be adopted as a reasonable strategy as long as
On Understanding: 1 1S5
the question of the ultimate success of physiological reduction is

left open.^
The Evaluation of Deductive Theories
Of all problems associated with deductive explanations none is

more complex than that of judging the "truth value” of a theory.
Philosophers and scientists have continually debated this issue
and at the present moment there is no reason to believe they will
ever achieve a consensus. Psychologists, in my estimation, have
an easier task than philosophers in dealing with this problem be¬
cause their needs are less demanding. Instead of having to cope
with the subtle details of formulating specific epistemological
principles for judging theories, the needs of the psychologist can
be met by establishing general guidelines that will assist in the
evaluation.
To begin our discussion of assessing theories that seek de¬
ductive validation, it will be best to describe in bold strokes three
different conceptions of developmental changes in science: tradi¬
tional, subjective, and historical. Each conception implies a differ¬
ent perspective for evaluating deductive theories.
The Traditional View
The traditional interpretation emerges from the commonsense

conception of science that the accumulation of scientific knowl¬
edge proceeds in a progressive, linear fashion. All knowledge is a
direct outgrowth of previous knowledge. When theoretical dis¬
agreements arise a crucial experiment is designed that is acknowl¬
edged by all to represent a fair test of the competing formula¬
tions. The crucial experiment yields unambiguous results as to
which theory is right and which is wrong. In the finest tradition
of Hollywood movies the two rival theorists accept the empirical
verdict and compliment each other for their unswerving devotion
5. Problems of reducing psychology to physiology and emergentism are
returned to in the final chapter, which discusses future trends in psychology.
to the search for truth. This scenario is repeated whenever theo¬

retical disagreements occur, and as a result, science inevitably
moves forward with a continual increase in empirical knowledge
and theoretical understanding.
The main difficulty with the traditional conception of the
growth of knowledge is that it fails to fit the facts of history.
Sometimes competing theorists cannot agree upon an appropriate
crucial experiment. Other times crucial experiments turn out not
to be crucial because ad hoc interpretations allow a theory to in¬
corporate unanticipated results. In addition, the history of science
provides numerous incidents where significant evidence, reliably
demonstrated, was ignored for long periods of time simply be¬
cause it did not fit into prevailing theoretical conceptions.
The Subjective View

The most compelling argument against the traditional con¬
ception of scientific theory is contained in Thomas Kuhn's The
Structure of Scientific Revolutions (1962). The preparation of this
book was encouraged by James Conant when he was president of
Harvard University. He believed that we needed “a widespread
understanding of science in this country, for only thus can science
be assimilated into our secular pattern” (Conant, 1947). Conse¬
quently, undergraduate education should involve some training
in science, even for those who were not science majors. Conant
persuaded Thomas Kuhn, then a doctoral student in the field of
theoretical physics, to accept the responsibility for teaching a his¬
tory of science course to the nonscientist. When preparing the
course Kuhn was surprised to discover that scientific develop¬
ment and practices were radically diflFerent from what he was led
to expect from his long-standing avocational interest in the phi¬
losophy of science. Instead of finding that scientific progress was
based on the accumulation of individual discoveries and theoreti¬
cal refinements, Kuhn perceived a repetitive pattern in the history
of science consisting of two markedly different enterprises: nor¬
mal science and revolution.
Normal science refers to the accumulation of knowledge
On Understanding: 1 1S7
within a widely adopted global orientation, labeled by Kuhn as a

paradigm. Indicating that the paradigm is incapable of any com¬
plete formulation, Kuhn suggested that it consists of a “strong
network of commitments—conceptual, theoretical, instrumental,
and methodological and quasi-metaphysical.” Paradigms serve as
the source of the methods, problem-field, and standard of solu¬
tion accepted by any mature scientific community at any given
time. Normal science, according to Kuhn, consists of empirical
and theoretical eflForts within the paradigm that gather additional
relevant information and improve the match between theoretical
prediction and data.
A paradigm possesses two related social functions. First, it at¬
tracts to its fold a group of adherents; second, it provides them
with a number of unanswered questions that keeps them busy ex¬
perimenting and theorizing (and, of course, publishing). The
paradigm provides an intellectual structure that integrates im¬
portant facts but at the same time generates numerous research
opportunities. Therefore Kuhn describes normal science as a
“mopping-up” activity.
In the com-se of this “mopping-up” activity anomalous re¬
sults occur that cannot simply be incorporated into the prevailing
paradigm. Thus the stage begins to be set for the second stage of
historical development—the reuoZwtion-during which time a pre¬
vailing paradigm is overthrown by a new one. Revolutions are not
achieved by a simple and rapid shift in consensual agreement
about which theory is valid. That is, a prevailing paradigm does
not collapse in the face of embarrassing data thereby paving the
way for the adoption of a new, more adequate paradigm. What
actually happens in a revolution is that the facts embarrassing to
an existing paradigm are discovered. The empirical area of the
anomaly is then extensively investigated, and modifications in the
paradigm are suggested in the hope of preserving its validity.
Typically these modifications are not completely successful, but
nevertheless the existing, though inadequate, paradigm survives.
The adherents of the existing paradigm learn to live with or, per¬
haps more properly speaking, to ignore the inconsistent results.
However, the anomalous results do inspire some young investi-
gators, whose emotional attaehments to the prevailing paradigm

are neither deep nor enduring, to formulate a eompeting para¬
digm that ean absorb the anomalous data. The struggle between
the old and new paradigms requires many years to resolve, and
its final resolution is more a result of the mortality of man than of
any set of data or logieal analyses. The older paradigm typieally
is not relinquished by those who were trained and who worked
within it. They refuse to understand the new paradigm and
argue against its validity. The younger people reeognize the ‘‘su¬
periority” of the new paradigm, adopt its standards and eoncep-
tions, and investigate the unanswered questions it generates. With
time the eontrol of the science passes from the hands of the older
investigators with their “outmoded” paradigm to the younger
scientists with their “superior” one. A stage of normal science sets
in and prevails until a new set of anomalous data are discovered.
Then the era of a revolution begins again.
According to the Kuhnian analysis science is not a cumulative
discipline as is suggested by the traditional conception; scientific
revolutions are “non-cumulative developmental episodes in which
an older paradigm is replaced in whole or in part by an incom¬
patible new one” (Kuhn, 1962, p. 91). In other words the new
paradigm does not emerge from the old.
This historical analysis, at first glance, seems to characterize
much of the history of psychology. One can cite the paradigmatic
shifts from structuralism to behaviorism, from stimulus-response
to cognitive psychology, from cognitive dissonance theory to at¬
tribution theory, and so on. In each case a shift takes place in a
methodological orientation that involves changes in the investiga¬
tory procedures, problem areas, and even in the criteria of proof.
In spite of its apparent relevance to psychology Kuhn’s hy¬
pothesis about the alternating historical sequence of normal sci¬
ence and revolution may not be applicable. Kuhn’s historical
model emerged primarily from his analysis of the development of
physics from the time of Copernicus. One can argue that psy¬
chology has yet to reach a comparable stage of evolution. Prop¬
erly speaking, psychology is in a preparadigmatic stage because
collective agreement has yet to be achieved about the nature of
its proper methods and its significant problems. Physics, at one

time, was at a preparadigmatic stage in which different concep¬
tions of the fundamental methods of investigation competed for
collective acceptance. The natural-science method finally won
out, and today one can refer to a single discipline of physics in
which essential agreement reigns about significant problems,
methods of investigation, and concept of truth. This does not
mean that those physicists who are interested in the philosophical
underpinnings of their discipline are in total concurrence. What
it does mean is that such philosophical disagreements that do
exist are not so profound as to affect the day-to-day activities of
physicists. One could not infer the philosophical commitments a
physicist held about his discipline, assuming he held any, from
the examination of his professional activities. In contrast, the pro¬
fessional behavior of a psychologist is highly related to his meth¬
odological commitments. By examining the professional behavior
of the structuralist, the behaviorist, the psychoanalyst, and the
humanistic psychologist, one could infer their epistemological
conception of psychology.
It is quite possible, as has been suggested in discussing the
subject matter and data base of psychology (Chapter 2), that a
paradigmatic stage of development is beyond the reach of psy¬
chology because of its concern with intrinsically different kinds
of phenomena. Although it is reasonable to hope that the facts of
behavior, phenomenal experience, and physiological events will
someday be deductively integrated, no a priori reason exists to
insure success. Another obstacle to achieving the paradigmatic
stage of historical development is the one presently being dis¬
cussed—the adoption of a common criterion of truth. Again there
is no reason to believe that psychologists will ever achieve con¬
sensual agreement.
In essence what is being suggested is that the historical
forces within psychology differ radically from those that were
operative in physics, and therefore the transition from a prepara¬
digmatic to a paradigmatic stage of development will never be
made. One obvious difference is that the high level of predict¬
ability achieved by early theories of mechanics and astronomy.
which psychological theories have been unable to mateh, encour¬

aged the domination of natural-science methodology by crushing
methodological orientations that eschewed predictability as a sig¬
nificant component of the criterion of truth. A look at contempo¬
rary psyehology would suggest that even if psychology had
achieved the predictive successes of early physics, pressures
would still be operating against a natural-scienee approach that
relies on understanding of the sort that deductive explanation
achieves. Psyehology plays many roles in modern society and
meets a variety of individual needs: a behavioral science that
strives to operate within a natural-science framework, a source of
moral preeepts, a guide-to-living, a key to understanding human
experience, a source of entertainment, and a sophisticated form
of commonsense. Even if natural-science psychology were more
successful, the desires of many psychologists and many segments
of society who seek answers to “psychologicaF questions that a
natural-science approach ignores or dismisses would remain un¬
satisfied. Thus psychology is predestined, within a Kuhnian frame
of reference, to remain in a preparadigmatic stage because so¬
ciety demands more of psychology than a natural-science ap¬
proach can deliver.
If the prepadadigmatic-paradigmatic historical sequence is
to be applied to psychology it must be limited to that segment
that seeks to operate according to natural-science principles. If
this is done it still remains unclear, partly due to the ambiguities
inherent in the concept of paradigm® whether psychology is at
the preparadigmatic stage or whether some harder areas (e.g.,
sensory physiology or theories of pereeption, learning, and mem¬
ory) have progressed to the paradigmatic stage. We need not be
too concerned with whether certain fields of natural-science psy¬
chology fulfill the requirements of a paradigmatic shift as con¬
ceptualized by Kuhn. Some similarity seems obvious. The goal of
this present analysis is not to pin down the exact meaning of
Kuhn’s historical analysis but instead to clarify the process by
6. Kuhn (1974) confesses that the popularity of the paradigm con¬

cept results in part from its vagueness, which allows readers to interpret its
meaning according to their own biases.
On Understanding: 1
141
which deductive theories are evaluated. And it is within this con¬

text that Kuhn, along with other philosophers of science (Hanson,
1958; Polyani, 1958) who have emphasized subjective factors in
science, has oflFered an important and influential suggestion.
Kuhn s views, which have ehanged over the years as a result
of intense critieism (e.g., Lakatos & Musgrave, 1970) and his own
efforts toward precision (Kuhn, 1970), can be expressed either in
a strong or weak fashion. The strong position, whieh emphasizes
the view that all facts are theory laden, is expressed in Kuhn’s
(1962) own words: “The competition between paradigms is not
the sort of battle that can be resolved by proofs” (p. 147) but is
more like a eonversion experienee” (p. 150). In essence, this sub¬
jective view of paradigmatic choice^ argues a relativist position
that the superiority of one paradigm over a competing one is fun¬
damentally a matter of personal taste. The weak form of Kuhn’s
position is that personal taste enters into a eontroversy between
paradigms but does not exclusively determine its resolution. Some
rational, objective evaluative procedures are invoked when choos¬
ing between competing paradigms.
The subjective view is presently identified with the strong
interpretation of Kuhn s position. In essence, this position assumes
that competing paradigms are ineommensurable in that one can¬
not translate the meaningful content of one paradigm into the
language of the other. As a result, the relative superiority of one
paradigm over the other cannot be ascertained except by a purely
subjective judgment.
If the eoncept of a paradigm is used broadly to cover general
methodological orientations such as structuralism and behavior¬
ism as well as specific theoretical conceptions such as the inter-
7. It must be emphasized that I am not equating Kuhn’s position with

a subjeetive method of evaluating deductive explanations although Kuhn’s
first edition of The Structuve of Scientific Revolutions (1962) did strongly
imply this attitude and was most instrumental in encouraging such a view
among many psychologists. For the record it should be noted that Kuhn
(1970) has retreated from such a relativist position by acknowledging
paradigmatic-independent facts and the validity of the notion of scientific
progress.
ference and duplex (short- and long-term storage systems) theo¬

ries of memory, then the kinds of paradigms being compared
must be considered in evaluating the subjective view. Although
the subjective view may be relevant for comparing the general
methodological orientations of structuralism and behaviorism, it
may be inappropriate for judging the merits of interference and
duplex theories of memory. As the last chapter suggested, no
compelling logic or evidence can be offered to support the con¬
tention that psychology must be the science of the mind (struc¬
turalism) or of behavior (behaviorism). In the final analysis it
becomes a matter of personal choice as to which volitional deci¬
sion will be adopted. This does not deny the possibility of recon¬
ciling a mind-oriented with a behavior-oriented psychology; it
merely asserts that a decision between the two orientations may
have to be made before the potentials of such a possible resolu¬
tion can be judged.
The appropriateness of the subjective view of theoretical
differences, such as those that distinguish between the interfer¬
ence and duplex theories of memory, appears less compelling than
it does for the structuralism versus behaviorism issue. Although
interference and duplex theories of forgetting have tended to
focus on different empirical problems (e.g., retroactive inhibition
studies versus free recall) and even to harbor somewhat different
conceptions of proof (pp. 358-370), a fundamental overlap of
empirical interest nevertheless exists. This common empirical in¬
terest can form the basis for judging the relative merits of the
worth and validity of the competing conceptions. This conclusion
can best be explicated by considering the historical view of judg¬
ing deductively formulated theories.
The Historical View
The traditional and subjective views of evaluating competing

theoretical conceptions employ different procedures for reaching
a verdict. The former offers a simple decision rule that can be
applied uniformly by all scientists. Within a historical framework
a particular theory is dissected and an unequivocal decision is
On Understanding: 1
reached as to whether it is congruent with the available data. If

it is, the theory is verified; if not, it is falsified. The essential in¬
gredients of the traditional view are deliberation followed by a
clear verdict that compels consensual agreement among the com¬
munity of scientists.
The subjective view is completely at odds with the tradi¬
tional conception. A paradigm consisting of both methodological
commitments and a set of theoretical assumptions is viewed within
a historical context. The decision to accept or reject is not con¬
strained by any publicly adopted rules of evaluation; the choice
of one paradigm over a rival does not occur ^ by deliberation and
interpretation" (Kuhn, 1962, p. 121). All facts are paradigm-
dependent and therefore cannot function as critical information
for evaluating competing conceptions.
The historical view, now under scrutiny, differs from both
the traditional and subjective conceptions but contains important
ingredients of each. In line with the subjective view, it denies that
at any stage of development a specific theory or general paradigm
can be analyzed and evaluated according to a set of uniform rules
that will yield a verdict acceptable to all scientists. Consistent
with the traditional view, it accepts the notion of paradigm-
neutral data that allow for deliberation and argumentation about
the relative merits of competing conceptions. Competing para¬
digms are commensurable, according to the historical view, and
the acceptance of one paradigm can be defended on the basis of
explicit standards. Yet, in contrast to the traditional conception,
the historical view does not suggest that only one set of explicit
standards are available that all members of the scientific com¬
munity will or must adopt.
Emphasizing the historical dimension for evaluating a para¬
digm is certainly not a novel notion. Many philosophers and psy¬
chologists have insisted that only this kind of analysis can yield
a deep understanding of a paradigm. But acknowledging the
significance of viewing a paradigm within a historical context will
function as an empty slogan unless criteria are suggested to judge
paradigmatic achievements.
Lakatos (1970a, 1970b) has suggested a frame of reference
with which to judge competing paradigms. He takes his cue from

a paraphrase of Kant’s famous dictum, “Philosophy of science
without history of science is empty; history of science without
philosophy of science is blind.”
According to Lakatos the proper unit for historical analysis
of scientific development is a research program, not a theory. A
research program contains an infinite number of specific theories
with a common cluster of core assumptions and a heuristic for
selecting research problems. For example, the neobehavioristic
(Hull-Spence) research program contained core assumptions such
as the fundamental unit of behavior (stimulus-response associa¬
tion), two basic theoretical processes (habit and drive), drive
energizes habit structures, each drive has its own distinctive cue
(drive stimulus), total drive is a function of both physiological
needs and incentive conditions, behavior is guided by anticipa¬
tory goal responses (rg’s), and the facts of conditioning are a
fruitful source for formulating general principles of behavior.^
The core assumptions of cognitive psychology are more difficult
to identify because of the influences from diverse historical tradi¬
tions—mentalism, perception, linguistics, computer science, and
Tolman’s (1936) theory of animal learning. From my vantage
point, cognitive psychology possesses core assumptions such as
organisms acquire internal representations (information) of the
physical world that can be transformed; transformations occur in
stages (levels); problem solving involves different plans (strate¬
gies ) for employing available information; performance is a joint
function of principles of cognition and action; and phenomenol-
ogy, perception, and memory are fruitful sources of general prin¬
ciples of behavior. And finally an important heuristic in achieving
an understanding of a psychological event is to represent it in a
flow diagram.
In a similar vein, core assumptions can be identified for
8. This list of core assumptions is neither meant to be exhaustive nor

definitive. There is no simple rule to identify eore assumptions or to distin¬
guish them from peripheral postulates. A historical analysis will presumably
identify a set of notions that are adopted by a variety of specific theories
that spring from a eommon researeh program.
On Understanding: 1
145
operant-conditioning approach, attribution theory, social-learning
theory, and other current research programs. The important con¬
sideration at present is not the explication of the core assumptions
of various research programs but the justification of considering
the research program as the basic unit of historical development.
Why should not specific theories such as Hull’s (1943) Principles
of Behavior or Tolman s (1936) Purposive Behavior in Animals
and Men or information-processing models of memory be consid¬
ered the proper subject of methodological analysis? The reason
is that such formulations possess a unique explanatory capacity of
their own, but at the same time they are historically related to
subsequent theories. The reason for this is that a theory can never
be falsified by any experimental evidence. As already noted, tests
of theories always involve ancillary assumptions about the appro¬
priateness of the empirical procedures used. Results contrary to
theoretical expectations can be attributed to the inappropriate¬
ness of the ancillary assumption rather than to any fundamental
inadequacy of the theoretical assumption. An example of this is
the controversy surrounding the interpretation of an experiment
(Campione, 1970) designed to test the dual-stage theory of devel¬
opmental changes in discrimination-shift behavior (H. H. Kendler,
& Kendler, 1972). The results obtained failed to reveal any devel¬
opmental trends and thus were interpreted to be at odds with the
dual-stage theory that predicted” ontogenetic changes. Propo¬
nents of the theory (H. H. Kendler, Kendler, & Ward, 1972)
argued, however, that improper experimental procedures were
used because special training techniques were instituted for slow
learners, most of whom were from the youngest age group. They
maintained that the dual-stage theory demanded that the subjects
of different age groups receive essentially similar training, and
thus the special procedures employed violated an assumption of
the theory the experiment sought to test. Of course this dispute
can be, and was, put to further empirical test (H.H. Kendler,
Kendler, & Ward, 1972), but the important point for our discus¬
sion is that the theorist always has the option in the face of ‘em¬
barrassing data to reject the ancillary assumptions of the experi¬
ment.
The theorist has another defensive maneuver in the face of

embarrassing evidence. He can make ad hoc modifications in
his theory to accommodate the embarrassing data. An unlim¬
ited number of codifications become possible because a theory
consisting of a set of logically independent postulates, can only
be tested globally. The theorist, therefore, when attempting to
accommodate inconsistent data, has numerous options ranging
from changing any one of several assumptions to modifying the
postulated relationships among them. Strictly speaking, one might
argue that these theoretical changes do not modify a theory but
in reality abandon it. Theories that generate different deductions
are different theories. Although different from the viewpoint of
their explanatory capacity, the two theories are obviously related
historically, the newer one emerging from the older one. In the
historical sense the two successive theories are components of
what Lakatos considers a research program—a set of core assump¬
tions that yield a variety of related theories generated in an at¬
tempt to interpret an expanding set of data.
Partly as the result of the traditional conception of historical
development in science, a negative attitude in some quarters pre¬
vails toward defense tactics of preserving a research program by
either rejecting ancillary assumptions or instituting ad hoc theo¬
retical modifications. Rather than admit that his theory has been
disproved by negative evidence, the theorist who rejects ancillary
assumptions or reformulates his conception is considered to be
indulging in some sort of skullduggery. He maintains his "‘theory
by invoking escape clauses. Since this can go on indefinitely he
really is not abiding by the rules of the game of science. Bluntly,
he is cheating. Another view is possible; the tenacious theorist
who refuses to admit defeat is serving an important function in
science. If a research program were abandoned at its first expo¬
sure to embarrassing data, we would never know if it could have
been salvaged by appropriate revisions. Few, if any, scientists
would disagree with the notion that some tinkering with a theory
is appropriate when initially confronted with anomalous evidence.
But how much tinkering is permissible? What decision rule can
be formulated to indicate that further revisions will be a waste of
On Understanding: 1
147
time and effort? When can the theorist and the entire scientific
community decide that a research program should be abandoned?
Implicit in these questions are two related issues: (1) the
evaluation of a research program and (2) the decision to aban¬
don it. Lakatos (1970a, 1970b) has suggested that a research pro¬
gram can be judged in terms of its historical development, par¬
ticularly in terms of whether it is progressive or degenerating.
Consider the specific case of an anomalous result that cannot be
integrated within a particular theory of a given research pro¬
gram. A progressive change would occur when an ad hoc modifi¬
cation of the theory not only copes with the anomalous result but
also predicts the occurrence of a novel event. In contrast, a re¬
search program is degenerating if an ad hoc modification does
nothing more than accommodate the anomaly. Thus a progressive
program is one in which modifications in the theoretical structure
are accompanied by an expansion in its explanatory capacity
while a degenerating program is one that repeatedly adds ad hoc
theoretical assumptions in order to cope with embarrassing evi¬
dence.
However, the evaluation of the historical development is
more complex than has just been suggested. Progressive changes
can occur when dealing with some anomalous results but not
with others. As a result, a simple verdict of "progressive” or "de¬
generating” may be hard to arrive at when evaluating a particu¬
lar research program at a given time in its history. In addition, a
research program is not evaluated in isolation. It is evaluated in
comparison to competing programs. When compared to degen¬
erating programs a particular research program that achieves
modest progressive successes may be strongly supported but may
be largely ignored when pitted against a competitor that is
achieving dramatic progress. Finally, it must be emphasized that
the characteristic that Lakatos selects to define progressive—the
prediction of novel events—need not be accepted as an ultimate
criterion. One might argue that theoretical precision is more im¬
portant than empirical generality. There is nothing intrinsic in
the scientific enterprise to suggest that one is more important
than the other, and consequently, one cannot demand that pro-
gressive and degenerating be defined either in terms of predicting

novel events or theoretical precision, or some optimal combina¬
tion of both. I, for one, in vie'wing the history of psychology and
noting the narrowness of many rigorously stated theories (e.g.,
mathematical models of learning), believe that it may be more
strategic to value empirical generality over rigorous mathematical
precision. But the opposite view is certainly reasonable. Nothing
can match the confidence in a theory when its successful predic¬
tions are rigorously stated. It will not help to debate this issue be¬
cause both theoretical precision and empirical generality are
goals of deductive theorizing. The major strategic issue is whether
one of the goals should be valued more at different stages of
theoretical development.
Now that a frame of reference for evaluating competing re¬
search programs has been sketched out, we can return to the
question of when a research program should be abandoned be¬
cause it is hopelessly degenerated. The key to answering such a
question is contained in the realization that an infinite number of
ad hoc modifications can be made in a degenerating research pro¬
gram. One's estimate that additional ad hoc modifications will not
salvage a degenerating research program must fall short of cer¬
tainty. This does not mean that a theorist should blindly persist
in his attempts to transform a degenerating formulation into a
progressive one. He can reasonably reach a point when he con¬
cludes: "'Enough is enough. There is no sense in continuing to
patch up my formulation. It is likely that it is fundamentally
wrong."
The decision that a research program should be abandoned
will be reached at different times by different researchers. No
psychologist should be surprised to discover that adherents, as
compared to critics, of a particular research program would be
more reluctant to conclude that it should be abandoned. Nor
should psychologists experience difficulty in understanding that
the tendency to persist in trying to salvage a research program
will vary among its adherents because such a decision results
more from psychological than logical factors. If we accept the no¬
tion that a scientist’s behavior will be guided by his personality
On Understanding: 1
characteristics, then we can expect the stubborn scientist, assum¬

ing he is intellectually resourceful, to persevere more than the
pliant one.
The fact that agreement will not prevail concerning the de¬
cision to abandon a research program does not place any real ob¬
stacle in the path of scientific progress. Each choice, persist or
abandon, serves a useful function. The first will provide informa¬
tion about the fruitfulness of a previously untried revision. As
previously indicated, a possibility always remains that a research
program can be turned around from a degenerating to a progres¬
sive one. The choice to abandon indicates that a former adherent
has reached the considered judgment that his research program is
hopelessly degenerated, and therefore the scientifie community is
advised to expend no further effort to salvage it.
The preceding analysis sounds strikingly subjective, no dif¬
ferent in quality than one that belittles the role of objective evi¬
dence in judging the relative merits of competing theories. Such
a view misses the fundamental point of my analysis. One must
distinguish between two positions: (1) all data are theory laden,
and hence competing research programs are incommensurable
and (2) data need not be contaminated by theoretical precon¬
ceptions and therefore research programs can be compared al¬
though difference in opinion can prevail about the appropriate
criteria to be employed. In the first case objective evaluations are
impossible while in the second they are within the realm of the
possible.
The employment of the historical framework does not guar¬
antee that those who are acquainted with research programs cur¬
rently attempting to interpret the same set of phenomena will ar¬
rive at equivalent evaluations. Because subjective factors are in¬
volved, both in the selection of the criterion on which judgments
will be based as well as in the decision about the future potential
of the programs, differences of opinion will likely occur. But these
differences in opinion can, by sound and diligent scholarship, be
traced back to the making of subjective decisions and the ra¬
tionale for their adoption.
This point becomes clearer when the historical frame of ref-
erence is contrasted with the traditional and subjective concep¬

tions. The traditional framework suggests that a clearcut evalua¬
tion of competing theories is always possible and thus encourages
unrealistic expectations. When the results of an experiment that
is considered critical fail to produce consensus about its theoreti¬
cal implications, the ^^traditional scientist is confused by, and in¬
capable of eoping constructively with, the unresolved dispute. He
persists in perceiving the theoretical controversy in the false light
of an unrealistic epistemological framework. As a consequence,
endless, and in most ways fruitless, debates are generated about
the specifications of a truly critical experiment, the one valid form
of a given theory, and the unswerving demands of science. Con¬
fusion, not clarification, is the final outcome of a traditional epis¬
temological analysis.
A subjective analysis of a theoretical dispute also will not
achieve any fundamental clarification although it may not be as
hampered by the pursuit of false issues as is a traditional analysis.
But it will overlook issues that could clarify the dispute instead
of treating the controversy simply as an expression of personal
taste. The important point is that if one equates scientific under¬
standing with deductive explanation, then there is a basis for ex¬
pecting that an epistemological analysis will clarify the issues at
stake among different research programs instead of denying such
a possibility. In essence, the subjective view denies the possibility
of a rational understanding because it is based upon a nihilistic
epistemology.
I do not wish to belittle the difficulties inherent in an histori¬
cal analysis. The scholarship it demands is of a much higher de¬
gree than that required by the traditional and subjective analysis.
This level of scholarship, however, can be approached if one does
not fall victim to the expectations of agreement demanded by the
traditional view and the anticipation of disagreement encouraged
by the subjective view.
No simple recipe can be offered that will ensure a construc¬
tive and reasonable historical analysis. What is required, aside
from profundity and respect for research programs other than
one’s own (qualities that may be in short supply) is an accurate
scoreboard of what the data are as well as knowledge of the logi¬
cal consequences of the competing theoretical positions. This
does not mean that if either is ambiguous a constructive analysis
is precluded. It is not ambiguity that interferes with a historical
analysis but rather the inability or the refusal to recognize it.
Once it is identified its implications for the future of the research
program can be analyzed and evaluated. By understanding the
present one is in a better position to plan for the future.
On Understanding:
Part Two
1. Interpretive consistency is a form of understanding that offers a

coherent view of a set of empirical events. It differs from deductive
explanation in that the precision of its logical organization falls short
of yielding empirical predictions capable of being falsified. Subjective
judgments play a significant role in determining whether given inter¬
pretations have met the criterion of consistency. Interpretive con¬
sistency meets the needs of human science but fails to meet the
standards of natural science.
2. Managing the occurrence of a given phenomenon produces a form

of understanding known as behavioral control. Operant methodology,
which has been highly successful in controlling certain forms of be¬
havior, operates on a probabilistic, not strictly deterministic, basis.
The search for techniques of behavioral control generates a research
strategy that, unlike the strategy employed by deductive theorists,
focuses on the influence of powerful variables and on the develop¬
ment of rigorous experimental procedures that eliminate the effects of
extraneous variables. Such a strategy may be particularly effective at
this stage in the history of psychology.
3. Intuitive knowing is a form of understanding resulting from a sub¬

jective feeling of comprehension. The psychological factors producing
intuitive knowing must be distinguished from its epistemological
status as an independent form of scientific understanding. Intuitive
knowing frequently anticipates deductive explanation but by itself
fails to meet the standards of understanding demanded in the natural
152
sciences. Intuitive knowing is employed as a criterion of understand-
ing more frequently than psychologists like to admit, especially when
dealing with complex topics that involve both psychological and phil¬
osophical components (e.g., “the image of man’).
4. The inability of psychologists to agree upon a basic mode of under¬

standing has encouraged the suggestion that juridical methods be ap¬
plied to the resolution of scientific controversies. The idea that scien¬
tific truth can be adjudicated is rejected although judicial procedure
may have relevance to the application of scientific knowledge.
DEDUCTIVE EXPLANATION AND

UNDERSTANDING
According to many tough-minded scientists, or at least those who

believe they are, all that needs to be said about understanding
has already been said. Scientifie understanding, as represented by
deduetive explanation, is the only valid kind of understanding.
All other forms generate deceptions.
The trouble with this conelusion is that it dismisses signifi-
eant attempts at eomprehension both within and outside of sci-
enee. The human mind is capable of achieving understanding
through other means than deduetive explanation. In an effort to
eomprehend the current scene in psychology, attention must now
be direeted toward those other forms of comprehension that psy¬
chologists employ to arrive at their conception of truth.
INTERPRETIVE CONSISTENCY
A set of propositions is logically consistent when no contradiction

can be derived from it. In this sense a deductive model of expla¬
nation demands eonsisteney. It must provide a coherent view of
a set of events; it eannot allow for ineonsistent implications as
would be the ease for a theory of aggression that asserted that
man is the only member of the animal kingdom who murders for
the sake of killing while simultaneously predieting that members
of other speeies also behave similarly.
Consisteney, in the logieal sense, must be distinguished from
truth, in the empirical sense. A mathematical system, for exam¬
ple, could be logically consistent but have no truth value because
it is not coordinated to events in the empirical world. But more
significant for our present concern is the case of formulations that
have apparent empirical relevance but whose empirical implica¬
tions cannot be rigorously tested. That is, some psychological in¬
terpretations are formulated in such a manner, not necessarily de¬
liberately, as to protect them from possible falsification. In spite
of this, the formulation provides an apparently consistent inter¬
pretation of a particular empirical realm. This is not to say that
the conception can be expressed in a strict logical form. Instead,
it is limited to informal conceptions expressed in common lan¬
guage that give the impression of hanging together. The most
notable examples of interpretive consistency in the absence of de¬
monstrable empirical validity are various psychoanalytic interpre¬
tations of personality development.
If Freudian theory, for example, were judged by the de¬
mands of a deductive model of explanation, it would be found
wanting. Nagel summarizes his methodological analysis of Freud¬
ian theory by concluding that “as a body of doctrine for which
factual validity can be reasonably claimed, I can only echo the
Scottish verdict: Not proven” (1959, p. 55). It would be difficult
to disagree with such a conclusion if one insists that factual valid¬
ity can only be achieved by a formulation that meets the de¬
mands of a deductive model of explanation. Freudian theory falls
short of this requirement because it lacks any precision in its de¬
ductive capacity; any and all kinds of behavior appear consistent
with it. One reason for this is that Freudian theory is filled with
processes and mechanisms (e.g., id versus superego) that operate
antagonistically. Because their mode of functioning is not pre¬
cisely stated, a Freudian theorist, with a modicum of ingenuity,
has license to offer a consistent interpretation of any form of be¬
havior occurring under any set of conditions. If the pleasure prin-
ciple proves inadequate to explain a given form of behavior, the
death wish can be invoked. In addition, the same antecedent con¬
dition can predict all possible outcomes. A man’s unresolved Oe¬
dipus complex can explain such widely diverse behaviors as
homosexuality, marrying a woman who resembles his mother,
marrying a woman who is totally unlike his mother, not marry¬
ing, or avoiding all sexual contacts.
Postulating opposed processes by itself does not necessarily
produce laxness in predictability. Antagonistic processes operate,
for example, in motivational theories of hunger, which postulate
excitatory and inhibitory centers in the hypothalamus (Stellar,
1954; Teitelbaum, 1966) that, when operative, have the eflFect of
encouraging or discouraging eating. Because these centers are as¬
sumed to have diflEerent modes of control, one could manipulate
antecedent conditions in order to predict consequent behavior.
Freudian theory is incapable of doing this. No objective criteria
are available to determine the relative strengths of the antago¬
nistic processes and their mode of interaction. As a result rigorous
predictions about future behavior cannot be made, but consistent
interpretations of past behavior can be proposed. Rabbi David
Small, a fictional detective, expresses the same idea when ques¬
tioning the explanatory ability of psychologists :
In my experience . . . [being a psychologist] does not necessarily

confer expertise in understanding the motives of men, only some
skill in designing explanations of their behavior, which may or
may not be true and which can’t be proved one way or another
(Kemelman, 1976, p. 64).
Popper’s emphasis on the principle of falsifiability in order to

demarcate empirical science from other kinds of systematic asser¬
tions (e.g., metaphysics, pseudosciences) appears particularly rel¬
evant to an evaluation of Freudian theory. As already noted, sin¬
gle bits of evidence cannot falsify a theory. All tests of theories
involve ancillary assumptions that presumably specify those in¬
vestigatory procedures that are appropriate for testing the theory.
If the results obtained are inconsistent with the theory, the fault
may lie with the ancillary assumptions, not the theoretical princi-
pies. Or the theoretical principles may be considered defective

while the ancillary assumptions are accepted as valid. In either
case, the presumably embarrassing evidence encourages modifi¬
cations in the research program such as in a clearer specification
of the ancillary assumptions or in the revision of the theoretical
postulates, both of which lead to further empirical testing. Thus
even though the theory itself may not be falsified in the strict
sense of the word, alteration in the research program is demanded
in light of evidence. And, as a consequence, a continuing dia¬
logue is encouraged between the research program and empirical
evidence.
Freudian theory fails to meet this relaxed standard of falsifi-
ability, which requires a continuing interaction between theory
and data. The theory is so loosely organized and ambiguously
stated that it is incapable of generating unequivocal empirical
predictions either of experimental results or clinical evidence.
Consequently, in the hands of a Freudian the theory is easy to
‘‘confirm” but impossible to refute. Because of this deficiency
Popper draws a shocking conclusion: psychoanalytic theory does
not differ from the pseudosciences of astrology and phrenology.
Such a conclusion would no doubt offend Freudians, and perhaps
even some astrologists and phrenologists. What must be under¬
stood however is that such a conclusion does not equate the qual¬
ity of the content of Freudian theory with that of astrology and
phrenology. The only implication of Popper’s analysis is that the
epistemological forms of these various conceptions—psychoana¬
lytic theory, astrology, phrenology—are equivalent in the sense
that they are all incapable of being falsified. One could accept
such a conclusion while simultaneously maintaining that Freud’s
contribution to the science that searches for deductive explana¬
tions of psychology was profound. In spite of the inadequacies of
his theoretical formulations, Freud was instrumental in contrib¬
uting important orienting attitudes and keen empirical insights
that exerted salutary effects on the development of psychology as
an empirical science, e.g., a deterministic orientation for all forms
of behavior, recognition of unconscious processes (motives of
which the behaving individual is unaware), prevalence of con-
flict behavior, and a variety of defense mechanisms to reduce or

avoid aversive feelings and events (Kendler, 1974). The implica¬
tion of this analysis is that although a particular formulation is on
the wrong side of the tracks in regard to the criterion of falsifi-
ability, it does not follow that it is completely irrelevant and use¬
less to an empirical science that seeks deductive explanation. A
formulation that cannot be falsified can develop into a falsifiable
formulation or inspire the creation of one.
Since there appeared to be a natural affinity between Freud¬
ian theory of personality and the experimental psychology of
learning because each emphasizes the role of past experience in
their interpretation of behavior, the hope was harbored that psy¬
choanalytic concepts could be resystematized into a coherent or¬
der that would meet the requirements of deductive explanation.
Although research was done to illustrate experimental analogues
of Freudian mechanisms of behavior (e.g., fixation, regression),
and surveys were conducted to exemplify developmental aspects
of psychoanalytic theory (e.g., infantile sexuality, dreams as wish-
fulfillments), no real progress was made in improving the de¬
ductive capacity of the formulation. In one survey of objective
studies of psychoanalytic concepts, a pessimistic conclusion was
expressed about the chances of Freudian theory being remolded
into a deductive form:
Other social and psychological sciences must gain as many hy¬

potheses and intuitions as possible from psychoanalysis but . . .
further analysis of psychoanalytic concepts by nonpsycho analytic
techniques may be relatively fruitless so long as those concepts
rest in the theoretical framework of psychoanalysis (Sears, 1943,
p. 143).
This conclusion has stood the test of time. Little has been ac¬
complished in increasing the falsifiable potential of Freudian
theory.^ Instead the course of historical events has underlined the
1. This conclusion has been challenged. Silverman (1976) insists that

his research could have yielded results inconsistent with psychoanalytic
theory. Meehl (1970) has suggested some Freudian hypotheses that are
falsifiable. Both positions are certainly defensible because, as noted pre-
hopelessness of such endeavors. Freudian theory has inspired a

variety of competing theories, similar in epistemological form but
drastically different in content. Both Adler and Jung rebelled
against Freud’s emphasis on sex. Adler assumed that the funda¬
mental driving force in human behavior was a self-assertive drive,
not the sex-oriented libido of Freud. Jung adhered to the concept
of the libido but did not restrict it to sexual pleasure. The libido
underwent changes with age: in infancy it is directed toward
nourishment; in childhood toward play; after puberty toward sex¬
uality; and later in life toward spiritual values. Although all these
formulations can provide an apparently reasonable and consistent
interpretation of human behavior, their relative merits cannot be
judged in any objective fashion by empirical means.
Natural Science Versus Human Science
Why has this kind of theorizing arisen in psychology and what

can be done about it? In answer to the first question the distinc¬
tion between two fundamentally different conceptions of science
should be noted (Berlyne, 1975). In one conception, prevalent in
America and common in hard science and experimental psychol¬
ogy courses, is the conventional view that science is a method of
controlled empirical inquiry designed to interpret and predict
events. To draw warranted conclusions from such inquiries one
must meet the demands of deductive explanation. Another con¬
ception of science, more common in Europe, is much broader and
looser in that it refers to systematic scholarship that offers co¬
herent interpretations of phenomena.
The looser sense of science can take many different forms
viously, a formulation that fails to meet the standards of falsifiability can

be modified to one that can. In estimating this possibility one must not only
consider the adequacy of Silverman’s (1976) research program and the
potential of Meehl’s (1970) suggestions but also the willingness and capacity
of some segment of the psychoanalytic community to do at least some of the
research and theorizing needed to make the psychoanalytic research pro¬
gram meet the standards of falsifiabifity.
On Understanding: 2
159
and is responsible for the application of the term science to a va¬
riety of disciphnes including history, law, literary criticism, and
even religion (e.g., Patai, 1971). One reasonable interpretation of
this looser sense of science is that it refers to a qualitatively dif-
erent discipline than the hard sciences. An important exponent
of such a view was Giambattista Vico (1668-1744), a Neapolitan
philosopher of history, social theorist, and jurist. Rather than den¬
igrate the level of understanding that history offers in comparison
to physics, Vico (Gardiner, 1967) took the position that an his¬
torian can achieve a more profound and intimate understanding
o IS subject matter than can a physical scientist. Vico denied
that the scientific method used to study the physical sciences was
the only valid method of scientific inquiry and insisted that there
are different kinds of sciences, i.e., history requires a method that
IS distinct from physics. One important feature of this difference
IS the kind of understanding that can be achieved by these two
isciphnes. Because the physicist is external to the inanimate sub¬
ject matter he studies, he cannot achieve the personal under¬
standing that IS ultimately available to the historian. The histo¬
rian, a human studying human events, is capable of empathizing
with those who made history and therefore can achieve an inti-
rnate acquaintance with his subject matter that is denied the
physical scientist.
If this is true then one would like to know the exact details
of the methods used to achieve understanding of the ‘‘human”
sciences such as history and the “litmus test” by which one dis¬
tinguishes between true and false interpretations. No critical test
akin to deductive “validation” through empirical tests is offered.
Only important interpretative safeguards are suggested. The his¬
torian must assume a proper perspective when viewing past
events. He cannot project his own consciousness or the standards
o conduct, values, and motivations of his society onto historical
personages. In some fashion he must reconstruct the psychology
and sociology of past eras and interpret historical events within
that “valid” framework.
Vico also viewed history in a highly integrative fashion, in
terms of both ahistorical and historical features. A society cannot
be properly understood if it is treated as a set of isolated systems;

legal, ethical, artistic, political, and so on. The entire society
forms a coherent pattern in which all component systems exert
a reciprocal influence. In a similar vein Vico argued that societal
evolution develops in a strictly deterministic fashion. He pro¬
posed an interesting cyclical hypothesis to account for the histori¬
cal evolution of societies. In brief he postulated that societies de¬
velop through several stages, from loosely federated groups of
patriarchal families to larger integrated oligarchies. Human rea¬
son ultimately dominates, and democratic ideals and values are
expressed. Enlightened social, legal, and political institutions are
created, but they contain the seeds of their own destruction. The
questioning attitudes that democracy encourages become focused
on the values of democracy itself. As a result the social institu¬
tions are corrupted and become weaker, and finally the society
undergoes decay as a function of internal dissolution and/or ex¬
ternal conquest. Primitive conditions return and the historical cy¬
cle begins anew.
Vico’s ideas were more or less ignored until the nineteenth
century and not fully appreciated until recent times. His notions
anticipated many important social philosophers and theorists, in¬
cluding Karl Marx. My concern is more with their epistemo¬
logical structure than their theoretical content, and from this
perspective, the central issue revolves about the method of as¬
certaining the validity of conceptions such as Vico’s. No doubt
the cyclical conception of societal evolution can be faulted be¬
cause of its failure to jibe with certain historical facts, perhaps
those associated with the development of India and China. I can¬
not help but feel that in the hands of an ingenious theorizer the
argument could be offered that apparent inconsistencies revealed
upon close examination really do not invalidate Vicos formula¬
tion because they either result from improper interpretations or
ignore potential theoretical modifications.
Are not such “theoretical defense mechanisms” equivalent to
those used for formulations that explicitly adopt the deductive
model of explanation? Does not the historical framework sug¬
gested by Lakatos (1970a) in his proposed analysis of progressive
On Understanding: 2 2^^
and degenerating aspects of theories essentially endorse those de¬

fense tactics designed to maintain the validity and integrity of a
theory? I think not. The crucial difference is that Lakatos's frame
o reference demands that specific implications of a formulation
e falsifiable, a demand not met by conceptions such as Vico’s. A
case in point is the Marxist economic interpretation of history.
Popper described Marxism (and also Freudian theory) as a
myth because it was detached from the empirical world in that
its implications were beyond falsification. If one examines the his¬
tory of Marxist principles, with the endless exegetical discussions
and disputes concerning their meaning and validity, and the per¬
fect survival rate of the conception in the hands of its adherents,
one is likely to conclude that Marxism as a theory of history and
economics is fundamentally different from the theories in physics,
like those of Newton and Einstein. The difference, it must be un¬
derlined, IS not in terms of axiomatic elegance but in deductive
consequences.
One possible reason for the difference between the predictive
capacities of physical theories and historical conceptions of social
change is that the latter are not always governed by lawlike reg¬
ularities. Physical systems, especially when placed under the scru¬
tiny of experimental analysis, are insulated from unpredictable
outside influences. This is not true for the historical development
of a society. Unpredictable events can shape the future course of
a society. For example, scientific breakthroughs, which triggered
the industrial revolution and the atomic age, are themselves es¬
sentially unpredictable. How can social change be predicted if a
major cause of it cannot?
Another factor that operates against predictability of social
change is the accidental interaction among independent systems.
Take the case of the person who runs out of his house to prevent
a stray dog from digging up his lawn and breaks his leg when he
slips on a banana peel dropped out of a garbage pail by an inex¬
perienced replacement of the regular garbageman who was called
to jury duty. In order to predict the accident one would have
had to anticipate, among other events, the coincidence of the ac¬
tivities of the stray dog, the jury duty of the regular garbageman.
2g2 Psychology: A Science in Conflict
and the sloppiness of his replacement. Do not important histori¬

cal events have the same unpredictability? Was the disastrous in¬
volvement of the United States in Vietnam equally unpredictable?
Was it not the result of the coincidence of numerous umelated
events such as the assassination of President Kennedy, the macho
personality of President Johnson, and the naivete of Secretary
McNamara s faith that systems analysis and computer technology
could predict the behavior of the Vietnamese?
The questions posed identify a clash between two competing
'In principle’' arguments. One suggests that the power of the sci¬
entific method can be applied successfully to any natural occur¬
rence while the other maintains that limits to predictability oper¬
ate. Consistent with previously expressed reservations of m
principle” arguments, sides in this controversy will not be takem
This does not mean that the problem should be ignored. It should
be realized that successful predictions of events become more
difficult as the number of operating variables increase and as the
interactions among them become more complex. Predicting the
time it will take a leaf to fall when in a windstorm is a much
more difficult task than when it is in a vacuum. Regardless of ‘ m
principle” arguments it would be reasonable to expect that social
theories have a greater difficulty of meeting demanding standards
of deductive explanation than do physical theories.
Whatever explanatory limits may be placed upon theories
that seek to account for societal evolution, it must be understood
that they do not necessarily apply to all historical changes. In my
previous discussion of the change in the incidence of peptic ul¬
cers in men and women during the last 100 years (page 115), it
was noted that deductions from an informally stated theory met
the standards of falsifiability. A much more interesting and sig¬
nificant example of a historical theory, and one that has special
relevance to psychology both in form and content, is Darwins
theory of natural selection. Perhaps the future will prove, if it has
not already been demonstrated, that theoretical psychologists,
such as Hull (1943) and Lewin (1935), made a strategic error
when they sought conceptual guidance from the physical theories
of Newton and Galileo rather than from Darwin.
On Understanding: 2 16S
Darwinian Theory: A Historical Theory

with Deductive Capacity
Darwin s theory of evolution has been criticized as a gigantic

tautology (the fittest survive because they are the fittest) and
faulted because it can explain but not predict. Both views are in¬
correct because the formulation does have the capacity to predict
future events. This point was overlooked, even by Darwin him¬
self, probably because the retrospective aspect of the theory had
such profound implications that its predictive capacity was ig¬
nored. For the prevailing view that the variety of animal and
plant species represents the unalterable work of God, Darwin
substituted the conception that species evolve over time by virtue
of environmental pressures that favor the survival of one life
form over another. Another probable reason for disregarding the
predictive capacity of natural selection theory is that the biologi¬
cal mechanism responsible for variations within a species was un¬
known. With the advent of genetic theory this gap was filled,
and consequently, the deductive capacity of natural selection
theory increased enormously. But the significant point is that
Darwinian theory in its original form provides the basis for pre¬
diction of future events. One can cite experimental evidence that
is consistent with the predictive capacity of the original formu¬
lation.
One such example is the well-known case of two varieties of
the peppered moth in England that appeared to have different
survival rates in different environments. One form of the pep¬
pered moth is light colored with dark spots scattered irregularly
over the body and wings. The other form, due to the presence of
the melanin pigment, is much darker. On trees covered with li¬
chens the light varieties of the peppered moth are effectively
camouflaged, but the darker forms stand out and consequently
are more subject to predation from birds. The opposite relation-
ship prevails in industrial areas where tree trunks are darkened
from soot; the light-colored forms are more visible. In short, the
lighter form has greater selective advantage in nonindustrial areas
while the darker forms have the advantage in the industrial areas.
One would therefore prediet that due to the proteetive coloration

the survival rate of light and dark peppered moths would vary in
industrial and nonindustrial regions. This hypothesis was tested
by releasing an equal number of light and dark moths in each en¬
vironment and after a given amount of time recapturing as many
as possible. In line with the expectations a greater percentage of
the dark forms were recaptured in industrial areas while the op¬
posite occurred in the nonindustrial areas (Baker & Allen, 1971).
One can also cite evidence that extends beyond a given gen¬
eration to support natural selection theory. A mixture of grass
and clover was planted in a field, which was later divided into
halves by erecting a fence. For a three-year period one side was
cut a few times during the summer for hay while the other side
was heavily grazed by cattle. Plants from each side of the fence
were then dug up and transplanted in an experimental garden.
After a period of growth under uniform conditions it was found
that the grasses and clover from the un grazed half produced
plants that for the most part were erect and vigorous while the
high proportion of plants from the grazed side were short and
rambling. From natural selection theory one would have pre¬
dicted this outcome: in the ungrazed field natural selection fa¬
vored the tall plants, which because of their height would com¬
pete successfully for sunlight with the short, rambling forms, but
in the grazed half the opposite relationship would prevail because
low and small plants had a greater chance of not being consumed
by the cattle (Baker & Allen, 1971).
Neo-Darwinism, the theory of natural selection supplemented
by modern genetic formulations, has a greater explanatory capac¬
ity than is suggested by the kind of experimental evidence just
cited. For example, it predicts experimental data that express the
operation of sexual selection, a concept that was original with
Darwin but was unrelated to any specific mechanism of heredi¬
tary transmission. If sexual mating is encouraged by some physi¬
cal characteristic then the genetic basis of the characteristic will
increase over successive generations. Female fruit flies, for exam¬
ple, mate more readily with red-eyed males than with white-eyed
ones. Consequently, in a controlled laboratory population in
On Understanding: 2
165
which the percentage of red- and white-eye genes are known
one would predict that the proportion of red-eye genes will in-
crease m successive generations, a prediction that has been con¬
firmed (Baker & Allen, 1971).
Hn. S”® guidance provided by natural selec-

of thp unraveling the mystery surrounding the distribution
allv^L for sickle-cell anemia, a usu-
disorder m which the red blood cells become sickle
ped when deprived of oxygen. It was discovered that both
parents, usually black, of children suffering from sickle-cell ane¬
mia possessed the “sickle-cell trait”; only some of their red blood
s ecame sickle shaped when deprived of oxygen. Children
bom to such parents exhibit the basic Mendelian proportions of
p percent normals (nonsicklers), 50 percent with the sickle-cell
trait, and ^ percent with sickle-cell anemia, who usually die in
early life. One would expect, according to principles of natural
selection, that the sickle-cell gene would decrease in successive
generations because most of the offspring who inherit this gene
roin each parent would fail to reproduce. The sickle-cell gene
has decreased among American blacks with its rate of occurrence
as expected, becoming less for successive generations (with more
effective genetic counseling the decreasing rate could become
arger). Environmental conditions, however, can influence gene
requencies as demonstrated by the distribution of the sickle-cell
pne m certain parts of Africa where a high incidence of malaria
IS prevalent. The heterozygous person with a sickle-cell gene
from one parent and a normal gene from the other possesses
peater resistance to malaria than the homozygous individual who
inherits a nonsickling gene from each parent. Whereas the fre¬
quency of the sickle-cell gene is approximately 5 percent among
Amencan blacks, it has been found to be as high as 45 percent in
certain lowland sections of Africa along rivers that are infested
with the malaria parasite (Dunn, 1985). Because the heterozy¬
gous individual with the sickle-cell trait is more resistant to mL
laria than is the homozygous nonsickling person, it can be pre¬
dicted Aat the sickle-cell gene will be preserved at a higher
proportion in malaria-infested regions than in other areas.
The final example of the predietive capacity of natural selec¬

tion theory will involve science fiction. One of the most compelling
arguments in favor of natural selection is its retrospective expla¬
nation of the absence of saltatory differences among species—
'nature” has no gaps. Since the diversity of species is accounted
for by random variation and natural selection, the difference
between species, according to the theory, should not be extreme
but rather gradual. The ‘"gap” between large classes of organisms,
such as reptiles and mammals, is occupied by transitional forms
with characteristics of both, e.g., the duckbill platypus lays eggs
like a reptile yet nurses its young like a mammal. Its body tem¬
perature also varies, not as much as that of a reptile but consid¬
erably more than that of a typical mammal like a dog.
The slight variations that natural selection predicts among
neighboring species is dramatically illustrated by the differences
among the Galapagos finches, who presumably descended from a
single species that emigrated to the Galapagos from the mainland
of South America. The ancestor species was probably a seed¬
eating ground finch, but as a result of individual variation and
natural selection fourteen different kinds of finches developed
(Lack, 1947)—some ground dwellers, others tree dwellers—each
with a somewhat different diet thus reducing the competition for
survival. The significant point is that the unique combination of
"chance” individual variations and environmental diversity rep¬
resented by a variety of ecological niches favored the develop¬
ment of novel forms of finches. Although natural selection theory
could not predict the exact forms the new species would assume,
the expectation was that their divergence would be gradual.
From this sort of analysis one would make the prediction
that if another planet with life was discovered that operated on
the same biological principles as does this planet, the speciation
of that planet would differ markedly from ours. The random vari¬
ation of offsprings combined with environmental pressures of
natural selection on that planet would produce a unique constel¬
lation of variables, and as a consequence its life forms would di¬
verge significantly from ours.
This extended discussion of natural selection theory was de-
On Understanding: 2
167
signed to demonstrate that Darwins formulation achieved the

standards of deductive explanation. One might disagree with
such a claim by noting that: (1) natural selection theory was,
and still is, incapable of predicting future evolutionary changes
in species or the creation of new species, and (2) natural selec¬
tion theory, as originally formulated by Darwin, achieved the
status of a deductive theory only with the development of genetic
theory.
^ Both of these criticisms are oflF the mark. Admittedly, Dar¬
win’s original formulation fell short of deductive elegance, but at
all stages of its development it could be put to the empirical test,
which in principle could yield embarrassing evidence. As already
noted, the original formulation was testable in regard to the as¬
sumption concerning differential survival as a function of environ¬
mental pressures. For a theory to be empirically meaningful does
not require that all its assumptions be directly testable. A theory
is a global conception when it comes to prediction. If a formula¬
tion can yield some deductive consequences that are empirically
testable, then it has explanatory capabilities.
Natural selection theory appears to be a classic example of a
theory that has developed progressively. Genetic theory did not
save the formulation but rather represents a consistent extension
of it. In other words, the deductive possibilities of natural selec¬
tion theory were not created by genetic theory but instead en¬
larged by it. Differential survival as a function of environmental
adaptation (a key assumption in Darwin’s theory) became sub¬
ordinate, with the advent of genetic knowledge, to the differential
reproduction of genetic material. Differential survival is mainly
responsible for differential reproduction of genetic material, and
thus selection becomes the mechanism by which systematic, in¬
heritable changes occur during successive generations. And finally
we can look to the future with some optimism (and trepidation)
that increases in the knowledge of the chemistry and physics of
genetic material will allow us to investigate directly the problems
of evolutionary changes in species and even the creation of new
species.
My excursion into natural selection theory was to demon-
strate that theories designed to aceount for historical phenomena

are capable of being formulated in a manner that yields deduc¬
tive consequences. After contrasting Darwinism with Freudianism
and Marxism I concluded that the empirical implications of the
first met the standards of falsifiability while the latter two did
not. Why was Darwinism successful in meeting such standards
and achieving general acceptance in the scientific community
while Freudianism and Marxism failed in comparison? An obvi¬
ous answer is that Darwinism is fundamentally valid, in the de¬
ductive explanatory sense, while Freudianism and Marxism are
not. Perhaps such an evaluation is the most appropriate response,
but nevertheless, it may be incomplete.
Deductive Explanation and Interpretive

Consistency as Forms of Understanding
Without implying that Freudianism and Marxism contain the

kernel truths that Darwinism possesses, the suggestion can be
made that different orienting attitudes toward the nature of un¬
derstanding influenced the development of these various concep¬
tions. These differences could be a consequence of many factors
including general methodological stances that prevailed in some
segments of the behavioral and social sciences at the time as well
as the intrinsic difficulties involved in isolating and controlling
the relevant variables in these fields. In addition, a combined
epistemological-psychological factor can be mentioned. This fac¬
tor rests upon a division between two kinds of knowledge, one of
which can be described as a conception while the other, the me¬
dium for deductive explanation, is designated as a theory. The
psychological underpinnings of this distinction are expressed in
the contrast between the pursuit of meaning and the search for
causes.
A conception, in this methodological analysis, is an awareness
of a general schema that governs the functioning of a given sys¬
tem. One example of a conception is Vico’s analysis of the con¬
tinuity in the rhythm of history as expressed in the rise and de-
On Understanding: 2
169
dine of human societies. Similarly, Oswald Spengler (1950) per¬

ceived the history of cultures as an expression of a cohesive force
that creates a configuration in which all of its components-reli-
gion, art, economics, politics—reflect a specific quality of its peo-
pla Spengler assumed that cultures are self-contained systems
and, as such, obey certain laws of historical development and de-
c ine. Arnold Toynbee (1946-57), in turn, examined twenty-one
civilizations and adduced a particular pattern of development and
decline that presumably characterized all societies. To this list
can be added Marx’s historical materialism, which attributed to
economic processes the key role in historical development; and
Freud’s psychoanalytic theory, which postulated that psychosex-
ual development underlies personality formation.
I maintain that all of these conceptions illuminate, in a pro¬
found sense, the phenomena they sought to clarify. This illumina¬
tion of historical events, and I think it is appropriate to consider
personality development within that context, was achieved by
providing a cohesive formulation of events that had been for the
most part ignored. In each case, from Vico to Freud, an explana¬
tory vacuum was replaced by an imaginative interpretation that
had the appearance of both consistency and validity. Although
none of these interpretations met the standards required of a de¬
ductive explanation, they nevertheless provided some sense of
understanding by identifying a problem considered by some to
be pointless and by others to be inexplicable. Thus the rise and
decline of societies, the economic force in social organization,
and the psychological significance of dreams—all hidden by ne¬
glect or oversight—were brought forward and recognized as sig¬
nificant issues. In addition, possible interpretations were offered.
Darwin’s theory of natural selection could be considered a
conception in this context. He perceived the problem of variation
of species in relation to environmental diversity and formulated a
possible interpretation of the phenomenon. But his conception
was transformed into a theory of natural selection that, as has
been argued, meets the standards of a deductive explanatory sys¬
tem. Thus the fundamental difference between conception and
theory is that the latter possesses a predictive component. The
absence of this component does not justify dismissal of such con¬

ceptions as myths if we mean by such a designation purely a
product of one’s fancy. Conceptions further understanding both
by directing attention to a significant phenomenon and by sug¬
gesting a possible interpretation. In this sense the term concep¬
tion could not properly be applied to astrology or phrenology be¬
cause neither deals with a significant empirical relationship.
There is no evidence to justify the belief that the position of
the stars or the shape of a person’s skull is related to his behavior.
Misconception would be a more appropriate characterization.
In spite of identifying significant empirical relationships,
Freudian theory could not generate unambiguous predictions
about specific empirical events. It did suggest in a general way
that children engage in erotic behavior, an observation that has
been supported by empirical evidence (e.g.. Sears, 1943). But the
psychosexual theory of personality development implied more to
infantile sexuality than simple self-initiated genital activity. The
fundamental concern of the theory revolved about concepts such
as stages (oral, anal, and genital) of psychosexual development,
Oedipus, Electra, and castration complexes, and penis envy. The
empirical consequences flowing from the theoretical relationships
among these concepts appear to be beyond falsification. Failure,
for example, to obtain evidence of penis envy may be due to the
operation of some unconscious process that suppresses any ob¬
servable consequence of the mechanism. Regardless of the evi¬
dence it is always possible to argue, as devout Freudians do, that
penis envy does in fact operate.
The significant question is why so many formulations in the
so-called “human sciences” (social and behavioral) assume the
empirically unassailable position that is characteristic of psycho¬
analytic theories. The answer would seem to lie in the function
such formulations serve in satisfying the need for understanding.
Some psychologists pursue meaning while others search for causes.
Meaning in this sense is expressed by a consistent interpretation
(i.e., a conception) that provides a compelling sense of under¬
standing whereas the identification of cause is achieved by a suc¬
cessful deductive theory. The two are not mutually exclusive
On Understanding: 2
when an “objective” deductive theory offers a “subjective” com-

pe ing sense of understanding. However, in the absence of a suc¬
cessful deductive theory a consistent interpretation can provide
a satisfying sense of understanding. And what is more important
even when a deductive theory is available some psychologists will
nd the meaning offered by a consistent interpretation more ap-
pealing than a causal interpretation. ^
Vico was explicit in affirming that the criterion of under¬
standing human events was different from those employed to in¬
terpret physical phenomena. He suggested that the difference re¬
volved about the position of the interpreter in relation to the
events he seeks to understand; the physical scientist is detached
whde the scientific investigator of human affairs, such as the his-
tonan, is involved. The kind of understanding is different be¬
cause t e ehavioral scientist can draw upon his own personal ex¬
perience in comprehending human events. As noted, Vico argued
that the intimate acquaintance a historian can have with his sub¬
ject matter enables him to achieve a deeper understanding than
what a physical scientist can possibly achieve.
Numerous social and behavioral scientists have argued along
the lines of Vico in maintaining that understanding human events
involves a different process than comprehending natural phe¬
nomena. More often than not this claim for a distinctive method
m the human sciences is made in the absence of any description
of what that special method is.
Criteria of Interpretive Consistency
There are obviously numerous forms of understanding hu¬

man events aside from deductive explanations. The line of argu¬
ment being pursued here is that one general “method” of compre¬
hension, known as interpretive consistency, is different from
eductive explanation but is nevertheless capable of providing
a compelling sense of understanding. It may very well be that the
rules underlying this kind of understanding cannot be clearly
explicated because, unlike the method of deductive explanation,
interpretive consistency is not evaluated by a set of explicit judg-
ments that yields a consensual agreement about its truth value.

In the absence of any precise set of regulations to judge inter¬
pretive consistency the most one can do in describing this kind of
understanding is to characterize its procedures. One possible
starting point is Vico’s suggestion that historical events become
comprehensible when placed in the context of the experiencing
individuals. This orientation is related to the notion of verstehen,
a concept that has been suggested by numerous social theorists
as representing the method of achieving understanding of human
events (Abel, 1948). In essence, verstehen is a cognitive operation
that the human scientist performs in order to understand human
behavior. He accomplishes this by attempting to share another
person’s state of mind—a task, the success of which is difficult,
and sometimes impossible, to judge as my discussion of “shared
experience” suggests (page 64).
Consider this simple case of achieving verstehen (Abel,
1948). During a freezing spell in early spring a man observes his
neighbor rise from his desk by a window and walk to the wood¬
shed where he starts chopping wood. After completing the task
the neighbor carries the wood into his house and places it in the
fireplace where he lights it. He then returns to his desk and re¬
sumes his daily writing chores.
The explanation of the neighbor’s behavior appears obvious;
he felt chilly and lit a fire to warm up. This interpretation is ar¬
rived at by filling in a psychological gap between two events that
the observer thinks are important, the dropping temperature and
the heat produced by fire. The psychological gap is filled by the
observer postulating that if he had been in his neighbor’s shoes
he would have started a fire. According to this analysis, verstehen
is achieved when we perceive a person’s action as consistent with
what we would have done under similar circumstances.
Although plausible, the interpretation of the neighbor’s be¬
havior may not be correct. Perhaps he likes the cold temperature
in his study but lights the fire to show off his fireplace to an antici¬
pated visitor, or perhaps he is suffering financial difficulties and
unconsciously hopes that the fire will “accidentally” produce a con¬
flagration that will enable him to collect some insurance money.
On Understanding: 2 2^^
Conceivably these alternative hypotheses could be tested in a de¬

ductive fashion by gathering additional information-e.g., the
neighbor s own explanation for lighting the fire, the arrival of any
visitors, the neighbor s financial condition, his behavior in similar
situations m the future. The point however is that the initial ex¬
planation is so compelling that other interpretations and further
evidence are not sought. In other words, verstehen provides a
phenomenologically satisfying interpretation that rings "true” and
resists rejection.
It is necessary to distinguish between the cognitive operation
of verstehen and the content and quality of the interpretation that
produces verstehen. Certainly the interpretation of the neighbor s
behavior in no way approaches the complexity and profundity of
the explanations offered by Vico, Spengler, Toynbee, and Freud.
The point here however is that a similarity prevails in regard to
how the understanding is achieved; an event is understood by fit¬
ting it into the observer s explanatory framework. This framework
can vary from a naive, but consistent, intuitive conception of the
sort that generated the explanation of the neighbor’s behavior to
a sophisticated theoretical Freudian paradigm. Between these ex¬
tremes is Vico’s historical position, which postulates that the ac¬
tions of historical figures are understood by reconstructing their
consciousness.
The use of an explanatory framework to interpret phenomena
obviously does not distinguish understanding achieved by inter¬
pretive consistency from deductive explanations. Theories, re¬
gardless of what criteria are employed to evaluate them, provide
conceptual frameworks that influence the interpretation of events.
The fundamental difference between understanding achieved by
interpretive consistency and deductive explanation is the criteria
adopted for evaluating proposed interpretations. In the case of
the anecdote of ^die neighbor s behavior and Vico’s conception of
the human sciences, the criteria for understanding are clearly dif¬
ferent from those used for deductive explanation. Vico explicitly
makes this point, and consequently if one acknowledges his dis¬
tinction, no confusion need arise as to the general procedure that
is to be employed to evaluate whether a given theoretical pro-
posal is seeking deduetive explanation or interpretive consistency.

This is not to say that a simple “litmus test” can be applied to an
interpretation to determine whether it meets either standard.
Some ambiguity exists, but the important factor is that the two
kinds of interpretations are sufficiently different so that no confu¬
sion between the two need occur if the theorist explicitly states
what his explanatory goals are.
I am reminded of a conversation with a distinguished histo¬
rian who in response to my query about how he judged historical
explanations responded simply “Aesthetics!” He elaborated by in¬
dicating that one interpretation was favored over another if it
created a sense of understanding that was deeper, fuller, and
more coherent than the other. This sense of understanding was
not equivalent to a deductive explanation because he thought
that a search for such understanding in history was a highly ro¬
mantic and unrealistic enterprise for a variety of reasons includ¬
ing the incompleteness of historical data, the fallibility of most
historical sources, and the inability to devise a satisfactory deduc¬
tive test of a theory. In addition, he was willing to argue that a
purely deductive explanation, even if it were possible, could fall
short of a satisfactory level of understanding by failing to offer a
“meaningful” interpretation.
The meaning of “meaningful” in this sense may be clarified
by shifting our attention to early developments in mathematical¬
learning theory. Two similar mathematical representations of sim¬
ple learning processes were offered by Bush and Mosteller (1951)
and Estes (1950), each possessing essentially the same deductive
consequences. They differed however in that Estes s theory was
supplemented by a “meaningful” interpretation of why the mathe¬
matical processes operated as they did. He suggested that learn¬
ing occurred in a manner consistent with Guthrie s brand of
stimulus-response associationism (Hilgard & Bower, 1975). Ac¬
cording to Estes’s formulation the experimental situation con¬
sisted of a population of stimulus elements, which were sampled
on successive trials. The sampled elements were attached to the
correct response. Learning was complete when all the stimulus
elements were associated with the correct response.
If the predictive consequences of the two mathematical mod-
els are the same, then one could argue that the supplementary in¬
terpretation provided by the stimulus-sampling notion was super¬
fluous. But viewing theories within a historical perspective would
that factors other than their deductive capacities can play
a role in their development. The stimulus-sampling interpretation
of learning has instrumental value; it provides the theorist with
an analogy for interpreting a variety of phenomena within a com¬
mon theoretical framework. Thus the supplementary interpreta¬
tion aids the theorist in his attempt to expand the empirical do¬
main of his formulation. In this function the stimulus-sampling
notion operates as a tool of thought for seeing similarities not
revealed by the mathematical representation.
At this point it will be advantageous to back off from specific
points and regain the appropriate perspective for viewing the
fundamental methodological issue. The line of analysis being pur¬
sued is that a difference exists between interpretive consistency
and deductive explanation. This difference is fairly obvious when
the criteria of the two are clearly distinguished as is the case for
Vico, who postulates different forms of understanding (verstehen
versus deductive explanation) for the human and physical sci¬
ences. The distinction becomes much hazier when an explicit dif¬
ference is denied as is the case when a theorist claims that the
standards of deductive explanation are met but the implications
of his conception, upon examination, prove to be immune to falsi¬
fication. It should be clear from previous discussions that the ar¬
gument is not that deductive explanations cannot be achieved in
the human and biological sciences but instead that other kinds of
understandings are possible. One major task in comprehending
psychological knowledge is to identify the criteria of understand¬
ing that is sought. Although many psychoanalytic theorists would
argue that their conceptions are of the natural-science variety, a
historical analysis of the claim suggests otherwise. This historical
failure does not deny that psychoanalytic explanations can be in¬
tellectually convincing.
Over and above all of the other virtues of (Freud’s) theory stands
this one—it tries to envisage a full-bodied individual living partly
in the world of reality and partly in the world of make-believe,
beset by eonflicts and inner eontradietions, yet eapable of rational

thought and aetion, moved by forces of which he has little knowl¬
edge and by aspirations which are beyond his reach, by turn con¬
fused and clearheaded, frustrated and satisfied, hopeful and de¬
spairing, selfish and altruistic; in short, a complex human being.
For many people, this picture of man has an essential validity
(Hall & Lindzey, 1957, p. 72).
And once this ‘"picture” is judged valid it becomes immune to

falsification.^
The concept of interpretive consistency, if it is to be fully
understood, must await more complete psychological analysis. It
represents a cognitive knowing that is achieved by a coherent in¬
terpretation. Quite obviously individual diflFerences play an im¬
portant role; what is “aesthetic,” or encourages “verstehen,"" or
produces a consistent interpretation for one person will not neces¬
sarily be equally eflFective for another. For this reason the level of
consensual agreement that is attainable in deductive explanations
will not be approached by understanding achieved through in¬
terpretive consistency. The fact, however, that a particular con¬
sistent interpretation is not widely accepted in no way denies the
reality of its psychological validity for those who adopt it.
One reason for adopting interpretive consistency as a crite¬
rion for understanding in preference to deductive explanation is
that the demands of the human sciences are diflFerent from those
of the natural sciences. These demands can be met by suggesting,
as Vico did, a conception of understanding that differs from that
of the natural sciences. It would be possible to argue that although
deductive explanation is the deepest kind of understanding that
can be achieved, meeting such standards is an impossibility for
many phenomena in the behavioral sciences. To be specific, no
true deductive explanation can ever be confirmed, or even ap¬
proached, for certain kinds of phenomena such as Richard Nixon's
2. A reaction to Freudian theory that one hears at cocktail parties, is

that Freud did not create a new science of man but instead a new picture
of man. This statement, translated into the present nomenclature, would
read: “Freud did not create a new theory of man but instead a new con¬
ception of man.”
On Understanding: 2 2^^-
behavior in the Watergate AflFair or the true eauses of the Viet¬

nam War. The only realistic and rational approach to such phe¬
nomena, one can argue, is to realize the impossibility of achieving
deductive explanation but at the same time to be guided by the
ideals of such interpretations. That is, one can formulate a con¬
sistent interpretation based upon psychological and/or historical
theories that have yielded predictions about other phenomena
susceptible to deductive confirmation.
So we see that understanding via interpretive consistency can
be arrived at by opposite routes; from the conviction that diflFer-
ent rules for understanding must be invoked in the human sci¬
ences to the belief that deductive explanations are incapable of
being achieved but possible to approach.
One other psychological factor should be mentioned for the
ready acceptance of interpretive consistency as an index of under¬
standing. Attempts at understanding phenomena frequently begin
by identifying a variable or process that is assumed to be the
causal agent. This tendency is exemplified by commonsense in¬
terpretations of the sort that are expressed in oversimplified max¬
ims such as Practice makes perfect and *'Spare the rod and spoil
the child.” Practice and punishment are potent psychological
variables but neither can bear the entire explanatory burden for
even the simplest type of behavior. Whether the practice is being
reinforced and what the organism’s reaction to noxious stimula¬
tion is must also be considered. Although the elaborate and pro¬
foundly influential formulations of Marx and Freud are not com¬
parable to these maxims in form or content, they nevertheless
importance of a single factor (e.g., economics,
sex). This overemphasis can be considered a natural by-prOduct
of human thought. When analyzing multivariable phenomena
people tend to overestimate the influence of certain variables to
the exclusion of others. This tendency toward oversimplification
encourages the adoption of interpretive consistency as a mode of
understanding. By ignoring or minimizing the influence of factors
other than the one originally focused on, the theorist is forced to
spin out an interpretive web that maintains the priority of the
favored variable while simultaneously protecting the entire for¬
mulation from deductive embarrassment.
The Relative Value of Interpretive Consisteney

and Deduetive Explanation
The reader could easily get the impression that an invidious
comparison has been made between the understanding that is
achieved by deductive explanation and that which results from
interpretive consistency. The major thrust of this analysis has
been to identify two qualitatively different kinds of understand¬
ing. This is not to say that a particular formulation cannot shift
over time from one kind to the other, from deductive explanation
to interpretive consistency or vice versa. Nor does it imply that it
will always be easy to establish a clear dfemarcation line between
these two kinds of understanding. My suggestion is that the pos¬
sibility of formulating a falsifiable implication distinguishes de¬
ductive explanation from interpretive consistency. No doubt some
differences of opinion can occur in deciding on which side of the
demarcation line a particular formulation falls. It seems that epis¬
temological analysis is inevitably burdened by fuzzy boundary
lines between significant distinctions. Nevertheless, the distinc¬
tions between the two forms of understanding can still serve a
useful purpose in the epistemological clarification of different
theoretical formulations.
My major defense against the accusation that an invidious
comparison has been made is that the distinction need not be con¬
sidered within a judgmental perspective. Those who seek pre¬
dictability will choose to operate within a deductive explanatory
framework. But it must be recognized that the ultimate achieve¬
ment of satisfactory deductive explanations cannot be guaranteed
for all phenomena. Failures could result either from the limita¬
tions of the human intellect or from the intrinsic indeterminism
of given phenomena.^ In either case the most that can be hoped
3. There is no way of proving indeterminism. The ability to prediet

phenomena demonstrates a deterministic system. The absence of predict¬
ability however does not imply the absence of determinism. It can simply
reflect inadequate knowledge or a defective theory. The deterministic prin¬
ciple can be supported by successful theories, but it cannot be falsified by
unsuccessful formulations. In essence, determinism is not a thesis about the
nature of the world but instead a guiding research strategy.
On Understanding: 2 I’/g
for in understanding certain phenomena, ranging from political

revolutions to the creative eflForts of geniuses, is interpretive con¬
sistency. Consequently, interpretive consistency is, for some, a
more valued form of understanding than is deductive explanation
simply because it represents the form of comprehending ‘'truly
significant human events.’’
BEHAVIORAL CONTROL
One approach to understanding an event is to discover the fac¬

tors that control its occurrence, i.e., behavior is understood when
it is controlled. Some of those who achieve a sense of understand¬
ing by meeting the demands of interpretive consistency or deduc¬
tive explanation cannot comprehend how anyone could presume
to understand a phenomenon by controlling its occurrence. How
can a trivial demonstration of control provide the same deep un¬
derstanding conveyed by a conceptual superstructure that inter¬
prets a given fact as an expression of abstract principles? To be
specific, how could the control of a phobic act, such as fear of
snakes, be equated with the rich interpretation offered by Freud¬
ian theory? And finally, can anyone insist that we do not under¬
stand planetary motion because we cannot control it?
Two separate issues are embedded in this set of questions,
one of conscious experience associated with a sense of under¬
standing, the other of methodological preference for certain cri¬
teria of understanding. If a “sense of understanding” is viewed as
a subjective state, then one cannot refute a psychologist’s insis¬
tence that he understands behavior when it can be controlled in
the absence of any theoretical interpretation. We cannot deny
his subjective reality. If we shift gears and view the problem from
a methodological perspective, then the selection of behavioral con¬
trol as a criterion of understanding should become comprehensi¬
ble. After considering the vast effort expended in psychological
theorizing and the meager payoff in rigorous and successful con¬
ceptions, it becomes easy to comprehend how some psychologists
would opt for control as a sign of understanding. The choice of
behavior control becomes even clearer when it occurs in the

American cultural setting that values a pragmatic view of science.
Behavioral control is commonly associated with eflFective
training methods. The essence of this mode is expressed in John
Watson’s famous boast:
Give me a dozen healthy infants, well formed, and my own spec¬

ified world to bring them up in, and I’ll guarantee to take any
one at random and train him to become any type of specialist I
might select—a doctor, lawyer, artist, merchant-chief, and yes,
even beggar-man and thief, regardless of his talents, penchants,
tendencies, abilities, vocations and race of his ancestors (Watson,
1926, p. 10).4
In reacting to this and other statements of Watson, Skinner

(1959) noted that ‘'polemics led [Watson] into extreme positions
from which he never escaped,” a tendency that Skinner, him¬
self, has exhibited. Skinner and some of his colleagues have, at
times, encouraged the belief that their behavioral management
techniques are capable of strict deterministic control of behavior
comparable in power to Watson’s claim. Their confidence and opti¬
mism were encouraged by their successes in “shaping” the behav-
4. The purpose of citing this quotation is to illustrate, clearly and dra¬

matically, an example of behavioral control. It must be mentioned, how¬
ever, that this quote is taken out of context, its usual fate. The quotation is
frequently employed to discredit Watson in particular and behaviorism in
general. The sentence that follows, which is usually ignored, states, “I am
going beyond my facts and I admit it, but so have the advocates of the
contrary and they have been doing it for many thousands of years.” A
tolerant interpretation of Watson’s remarks is that he was attempting to
counteract the extreme position that genetic factors solely determined a per¬
son’s behavior by an extreme environmentalism that was as equally justified
(or unjustified).
Environmentalism is not a logical by-product of methodological behav¬
iorism. Methodological behaviorism espouses a public, objective approach in
psychology, making no substantive assumptions about the manner in which
genetic and environmental factors influence behavior. The interest of early
behaviorists in environmental control of behavior was not an expression of a
methodological commitment, but instead a reflection of influential philo¬
sophical and sociological forces in American psychology (Kendler, 1979).
lor of rats and pigeons during operant eonditioning and learning

of eomplex trieks (e.g., pigeons playing Ping-Pong). It beeame a
matter of faith among some operant eonditioners that shaping, if
properly and persistently applied, eould fashion behavior in any
desired form. Those who were wedded to this notion did not fully
appreeiate that the sueeesses aehieved were to some extent a fune-
tion of the tasks seleeted and the narrow range of organisms used.
A eeiling on the effeetiveness of shaping teehniques beeame
apparent when the Brelands (Breland & Breland, 1961) attempted
to shape raeeoons to drop coins in a piggy bank for a behavioral
display in a municipal zoo. The subjects tended to rub the coins
together and dip them in the slot in a manner that raccoons use
to wash food. The instinctive pattern of behavior prevented the
shaping procedure from achieving the desired goal.
Many other instances can be cited to demonstrate that oper¬
ant methods of controlling behavior developed in the laboratory
have definite limits. At one time hopes were harbored that myna
birds could be taught to communicate in English, but such naive
aspirations were abandoned when it was realized that emitting
appropriate sounds did not qualify as learning a language. Shap-
ing techniques were also thought to hold the key for curing schizo¬
phrenics and autistic children. Such high hopes were dashed when
it became apparent that operant methodology was not as effective
in changing psychotic to normal behavior as it was for training
pigeons to play Ping-Pong. (Pigeon Ping-Pong, in truth, was a
pale imitation of human Ping-Pong. The impact of the demon¬
stration rested not on how well pigeons played the game but only
that they could play it at all.) Operant conditioners have backed
off from their early hopes of curing psychotic behavior, justifying
their efforts by the claim that psychotic patients in behavior man¬
agement programs ‘make fewer demands on the staff and yet
display as much dignity and happiness as their pathology per¬
mits.” (Skinner, 1969).
The failures of operant methodology to control certain forms
of behavior do not invalidate the principle that regulating re¬
sponse-reinforcement contingencies is an effective method of
managing behavior. They merely show that such a method of
controlling behavior is not sufficient by itself. Other faetors must

be eonsidered, partieularly the genetic predispositions and physio-
logieal eapaeities required for eertain forms of behavior.
If the eontrol of behavior by operant methods is viewed
within the eontext of the entire range of phenomena to which it
has been applied, then the conclusion must be drawn that the
shaping technique is not a strictly deterministic mode of eon-
trolling behavior. Failures ean and do oceur. The ehances of suc-
eess vary with the organism and the task. Information about the
sueeess rate of behavior modifieation teehniques suggests that op¬
erant eonditioning methodology represents a probabilistic mode
of controlling behavior. Complete success in eliminating unde¬
sirable forms of behavior, ranging from a phobia to compulsive
smoking, is never achieved. The most that ean be said for these
operant methods is that they have a certain probability of success.
Theoretical and Practical Orientations

in Controlling Behavior
An interesting methodologieal eonfliet, which has important im¬

plications for the control of behavior, has oceurred in the psychol¬
ogy of learning between traditional theorists who employ abstract
concepts (Hull, 1943; Spence, 1956; Tolman, 1932) and those
who aecept control of behavior as their primary goal. Proponents
of traditional theories of learning tested their deductive powers
on the performanee of groups. Their strategy was to formulate
basic principles that applied to all organisms and leave to later
the diseovery of idiosyneratic factors that influence individual
performance. Sueh general formulations could therefore only offer
probabilistic information for those who sought to employ the
theory for the purpose of eontrolling individual behavior. This,
as noted, does not distinguish the knowledge underlying dedue-
tive theories from the information used by operant conditioners.
Yet, prineiples of operant conditioning proved far more effective
for controlling behavior than did deductive theories. Why? The
answer seems to lie in the diflFerences in strategy employed
by those who were formulating deductive theories (e.g., Hull)
and those who were concerned with controlling behavior (e.g.,
Skinner).
Traditional deductive theorists and control-oriented operant
conditioners perceived the relationship between experimental
operations and behavior differently. Both the theorist and the
operant conditioner manipulated experimental conditions, but the
theorist’s actions were governed by the nature of his theoretical
constructs while those of the operant conditioner were determined
by the behavior of his subject. The theorist tended to accept the
control exerted by his experimental manipulations especially when
deductions coincided with expectations. In contrast, the behavior-
controllers persistently attempted to improve upon the control
exerted by environmental manipulations. This encouraged the op¬
erant conditioner to focus on individual behavior while the theo¬
rist was concerned only with group averages.
This primary concern with individual behavior underlies the
operant conditioner s rejection of traditional statistical evaluations.
If one aspires to understand behavior, one must deal, the argu¬
ment goes, with the form in which it occurs (individual behav¬
ior) and not as some abstraction (a group average). Resorting to
statistical evaluation is an admission of ignorance about factors
that control behavior and an expression of despair that effective
management can never be achieved. Rather than succumb to this
pessimism one can actively seek to identify the conditions that
control the behavior of individual organisms. The search for un¬
derstanding then becomes the search for control.
Although the argument in favor of investigating individual
behavior appears reasonable, even though one may not fully agree
with it, the interesting question remains as to the relationship that
prevails between individual and group data. Sidman offers a
clear-cut, but puzzling answer: “The two types of data represent,
in a real sense, different subject matters” (1960, p. 53). Others
have worked primarily with individual data without adopting this
extreme view. Ebbinghaus, for example, studied only his own be¬
havior with the purpose of uncovering general principles of hu-
man memory. His goal was no diflFerent from modern memory

theorists who utilize between-group experimental designs. One
ean also refer to the eflForts of psychophysicists who test individ¬
ual subjects repeatedly. Similar designs are also being used in
information-processing studies (e.g., Sternberg, 1969) and ex¬
periments on the effects of meditation (e.g., Wallace & Benson,
1972), and again one finds no implication that such data need be
set apart from group averages. One can also note that the be-
tween-subjects design is not completely foreign to the house
organ of operant conditioners, the Journal of the Experimental
Analysis of Behavior, even though it is relatively rare, much be¬
low the percentage that appears in traditional APA and Psycho-
nomic Society’s journals (Hilgard & Bower, 1975). Finally it
should be mentioned that some operant conditioners have be¬
come enticed by theoretical issues (Honig, 1962; Terrace, 1968),
particularly those surrounding discrimination learning. Although
they employ operant conditioning methods and their basic data
consists of cumulative records, they are nevertheless able to make
contact with theories based upon evidence from between-group
studies involving discrete trials (Spence, 1936, 1937). Obviously
in this context one cannot conclude that two kinds of data, group
and individual, are fundamentally different.
Nor can one maintain the sharp distinction between the two
when viewed from the observational level. Average results are
simply a mathematical combination of individual events. The two
are related in the sense that from knowledge of statistical aver¬
ages combined with measures of variability it becomes possible to
infer the observational basis of individual behavior. Thus from
both theoretical and observational perspectives a strict segrega¬
tion between group and individual data, as suggested by Sidman,
appears unwarranted.
There is another way of looking at the difference between
group and individual data that is in terms of strategy. A detached
view of psychology very often leads to the conclusion that differ¬
ent experimental practices (e.g., free responding versus separate
trials in conditioning) are equally justified in the sense that none
is logically demanded. Although true, this conclusion does not
mean that each alternative will be equally fruitful. Research prac¬

tices, themselves, are experimental in nature because the kind
and quality of knowledge they will yield cannot be completely
anticipated. One can, nevertheless, argue that some practices will
be more fruitful than others in generating important information.
It is within this context that Sidman’s distinction gains its best
defense.
The history of behavioral psychology suggests that its phe¬
nomena are forbiddingly complex. One could argue that the in¬
ability of the discipline to achieve a widely accepted theoretical
interpretation of any broad area of behavior from learning to
perception or from personality to social behavior results from the
indefinite number of causal agents that operate in the simplest
experimental task. The diflBculty of achieving interpretations that
cross situational boundaries without losing explanatory precision
simply results from the tremendous changes in causal patterns
from one situation to the next.
The fundamental tactic in arriving at a broad interpretation
of behavior may be in coping effectively with the complexity of
the pattern of causal agents that operate in various experimental
designs. The strength of the operant approach may be in its ca¬
pacity to simplify the causal patterns of behavior more effectively
than other orientations. This, it can be argued, is accomplished in
two ways: (1) by exerting more effective control than others over
experimental procedures thus reducing the effects of extraneous
variables and (2) by identifying crucial processes in behavior
that are relevant to a wide range of situations.
In regard to the first, a comparison between the experimental
practices of operant conditioners and traditional learning theorists
tends to support the view that the former group exerted much
tighter control (D’Amato, 1970). Being primarily concerned with
arranging satisfactory tests of their model, traditional learning
theorists, such as Hull, Spence, and Tolman, frequently ignored
ways of improving the precision with which stimulus patterns
were presented and responses measured. They employed much
cruder equipment—puzzle boxes, multiunit mazes, T-mazes, dis¬
crimination-learning apparatus than did operant conditioners.
Of course, this diflFerence could be considered the consequence of

Skinners early fascination with gadgetry (Skinner, 1976), which
was expressed in its final form in the sophisticated electronic tech¬
nology of the operant-conditioning laboratory. But, it should be
remembered, skilled technicians were available also to learning
theorists. It is mainly that the problems operant conditioners
thought essential in their pursuit of behavioral control encour¬
aged the use of highly sophisticated equipment. Another aspect
of the looser control exerted by theorists is the relatively brief
time their subjects were confronted with an experimental task,
usually no more than a few trials a day extended over a period
of a week or so. These brief encounters were in marked contrast
to the intensive experimental experience of operant subjects,
lengthy daily experimental sessions spanning weeks and even
months.
Admittedly, it is difficult to draw a firm historical conclusion
about the exact cause of the difference in experimental practices
between learning theorists and operant conditioners. It could be
that the aim of exerting tight control over behavior infiuenced the
empirical methodology of operant conditioning in the direction of
its goal. It could be argued that the tight experimental control
emanated primarily from the identification of the potent variable
of response-reinforcement contingencies that made attempts at
behavior management effective. Whatever the reasons, the im¬
portant point is that operant conditioning outstripped theoretical
orientations in its ability to control behavior, even though it has
fallen short of the strict deterministic management to which it
aspired.
The conclusion of this analysis is that individual behavior
and group behavior, expressed in cumulative records and mean
performance, do not represent qualitatively distinct sets of events
but instead consequences of different orienting attitudes. These
consequences can be compared and the conclusion can be drawn
that one sort of data (e.g., individual cumulative records) is more
significant than others and that the strategy of behavioral control
is superior to that of formulating deductive theories. Such an
argument returns us to the problem of defining significance, a
task that demands specifying the aims of a research program.
This circularity, however, can be broken by considering two ques¬
tions : (1) what is the relative success of diflFerent response mea¬
sures (e.g., cumulative records, average group performance) in
achieving the goals for which they were selected (e.g., behavioral
control, theoretical interpretation) ? and (2) how useful is a par¬
ticular response measure for the goals of other research programs?
My impression of the evidence is that cumulative records, on both
counts, would receive a higher score than average group perfor¬
mance. This does not mean that a science of behavior employing
only cumulative response curves would be complete. Far from it.
Cumulative records certainly do not represent all of behavior.
Operant conditioners initially ignored acquisition processes
in conditioning. By using every trick of shaping, operant condi¬
tioners got their subjects to behave as rapidly as possible at as¬
ymptotic (“steady state”) performance levels. Acquisition, there¬
fore was perceived as an uninteresting problem, representing an
obstacle to be overcome in order to get at the critical problem of
achieving steady-state performance. In addition cumulative rec¬
ords have not proved to be particularly effective in unravelling
the mysteries of many research areas (e.g., cognitive processes,
social behavior). But these shortcomings must be judged in terms
of the present state of knowledge in psychology and not some
future ideal. In this context the significance of the operant re¬
search strategy with its emphasis on the behavior of individuals
in the form of cumulative response records cannot be overesti¬
mated. From animal training to child rearing to educational tech¬
nology to behavior modification to psychopharmacology, operant
methods have contributed important procedures for effective be¬
havior management. And the fundamental reason for its success,
in the last analysis, may be that the search for effective methods
of behavioral control is a particularly effective strategy in behav¬
ioral psychology. The overwhelming number of causal agents that
operate even in the simplest situation makes the problem of the¬
ory construction grotesquely difficult. When concerned with con¬
trolling behavior one learns very quickly whether one is on the
right track. The feedback is rapid, especially when compared with
the complicated validation procedures associated with theorizing.

One can argue, at least at this stage of its history, that striving for
behavioral control may be the optimal strategy for the compli¬
cated science of psychology.
INTUITIVE KNOWING
Comprehension by intuitive knowing is strikingly different from

the understanding achieved by deductive explanation, interpre¬
tive consistency, and behavioral controlj^ Whereas the claims of
understanding produced by deductive explanation, interpretive
consistency, and behavioral control can be examined by others,
intuitive knowing is purely a personal experience. It represents
‘'pure understanding” in the absence of any knowledge.
In essence, intuitive knowing is a subjective psychological
phenomenon in which an individual experiences insightful appre¬
hension or comprehension of an event. There is no reason to be¬
lieve either that different individuals achieve such insightful ex¬
periences in the same way or that their intuitive revelations have
common phenomenal characteristics. In fact, the subsequent dis¬
cussion will suggest the opposite.
Some psychologists, although admitting that the subjective
phenomenon of intuitive knowing occurs, would nevertheless con¬
sider it irrelevant to an analysis of science. Science, for them, is
essentially a discipline that was created in order to avoid drawing
conclusions based on subjective judgments. Warranted conclu¬
sions are justified only when the evidence and logic supporting
them are open to public purview. Intuitive knowing is locked in
an individual’s mind and therefore does not qualify as a scientific
conclusion. This concept should be discarded.
Such a view is unrealistic, especially when applied to psy¬
chology. It ignores an apparently significant cognitive phenome¬
non that may be central to methodological issues and contro¬
versies. More importantly, the blunt rejection of intuitive knowing
as a form of understanding is based on the naive belief that sci¬
ence always operates with a simple set of objective standards.
Even if intuitive knowing were rejected as an appropriate scien¬

tific mode of comprehension it would still be of interest for the
role it presumably plays in achieving more acceptable forms of
understanding. One of the many examples that can be cited to
illustrate the role of intuitive knowing in theorizing is Einstein’s
description of his own efforts that led to the theory of relativity:
During all those years there was a feeling of direction, of going

straight toward something concrete. It is, of course, very hard to
express that feeling in words; but it was decidedly the case, and
clearly to be distinguished from later considerations about the
rational form of the solution (Wertheimer, 1945, p. 184).
In a general sense Einstein acknowledged an intuitive know¬

ing of a theoretical solution prior to his being able to formulate a
rational explanation. This understanding encouraged him to pur¬
sue the problem to its successful solution. Numerous other exam¬
ples of intuitive knowing in the absence of an explicit rational
proof can be mentioned. Already reported was Poincare’s (page
82) conviction that his sudden solution to a logical problem was
correct even though he had no time to check the proof. On a
more personal note, examples of intuitive knowing from my own
experience can be described. After struggling unsuccessfully for
two days in an attempt to design a “critical” test of two compet¬
ing interpretations of a latent-learning study (Spence & Lippit,
1946), I had a dream from which I awoke with the realization
that I had the key to the design of the experiment even though
the details had yet to be worked out. The next morning the ex¬
periment was planned and later executed (Kendler & Mencher,
1948). On another occasion, I recall treating a young soldier of
twenty who was depressed and remorseful. He had just suffered
a death in the family similar to one I had experienced as an
adolescent. His reactions resembled mine. I intuitively “knew”
how he felt and my empathic reaction seemed to facilitate the
treatment. He began to realize that the unfortunate event itself
was the cause of his difficulty and was not a result of any per¬
sonal inadequacy.
Psychological Versus Epistemological

Analysis of Intuitive Knowing
Four examples of intuitive knowing have been offered. The first

three represent a phenomenological experience that preceded
problem solving: the formulation of a theory, a logical proof, and
an experimental design. The intuitive knowing in all of these ex¬
amples is not to be confused with the successful problem solving
itself. In all three cases the intuitive knowing fell short of prob¬
lem solving in the objective sense of providing a rational solution.
But in each case the intuitive knowing represents a subjective
preawareness of an objective solution. In more analytical terms,
within a phenomenological frame of reference the preawareness
seemed to possess a cognitive and motivational component, with
the former anticipating the solution and the latter encouraging its
achievement.
My interest in the phenomenological aspects of intuitive
knowing is tangential to the central issue of its status as an inde¬
pendent form of scientific comprehension. I am not really inter¬
ested in the argument that intuitive knowing is a necessary link
in arriving at scientific comprehension. That is an empirical prob¬
lem that properly belongs to the psychology of creative thinking.
Intuitive knowing as a means to an end is a problem for psychol¬
ogy; intuitive knowing as an end in itself is a problem for episte¬
mology.
The fourth example of intuitive knowing occurred as an em-
pathic reaction to a patient’s problem. Empathy, the “projection”
of one’s experience into another’s consciousness, illustrates some
of the epistemological problems associated with intuitive know¬
ing as a form of understanding. Some psychotherapists seek to
empathize with patients because of the conviction that such
identification is either essential to, or helpful for, effective treat¬
ment. The belief also prevails that the ability to empathize with
others underlies moral development; the golden rule, “Do unto
others as you would have them do unto you,” cannot be obeyed
On Understanding: 2 j^qj^
unless one has the eapacity to empathize with other human be¬
ings. Great literature, some would suggest, is ereated when the
author enables one to empathize with his eharaeters and thereby
share their experienees while simultaneously learning about one¬
self. Dostoievski and Bellow, for me, are partieularly eflFeetive in
this respeet.
My impression is that the intuitive knowing experienced when
reading literature is comparable to the experience of a clinician
when he empathizes with a patient. The relationship between the
two kinds of experience posed an interesting puzzle for me when
I was trying to decide whether to take an undergraduate major in
psychology. I was at that time enrolled in an introductory psy¬
chology course that had as its text John F. DashielFs Fundamen¬
tals of Objective Psychology. I was surprised and pleased with the
book because the intimate relationship between psychological
principles and experimental evidence was a new and intriguing
concept to me. All in all, psychology in 1938 appeared less vague
and more objective than other so-called social sciences, and from
my viewpoint, more appealing. This impression, however, was
badly shaken when I attended a symposium in which most of the
psychology department participated. The topic was “Human Na¬
ture, and without exception, each psychologist, to clarify his or her
position, quoted a passage from some great novel or poem. This
was confusing because I expected that their arguments would be
buttressed by experimental evidence. If literature was to be em¬
ployed to interpret psychological principles perhaps my interest
in psychology would be better served by majoring in English?
But that alternative had an apparent drawback because at that
time the leading lights of the English department were offering
psychoanalytic interpretations of literature!
In restrospect, that symposium was not as confusing as it ini¬
tially appeared. If one accepts the position that different modes
of understanding are possible, it should not be too surprising to
discover that more than one kind can be employed especially for
different occasions. Although one may be committed to explain
the maze behavior of rats by deductive explanation, one could
easily shift to the mode of intuitive knowing when confronted
with the imposing task of interpreting human nature. An interest¬

ing sidelight to this issue is contained in Skinner s autobiography
when he discusses his own search for understanding:
I had apparently failed as a writer but was it not possible that

literature had failed me as a method? One might enjoy Proust’s
reminiscences and share the emotional torment of Dostoevski’s
characters but did Proust or Dostoevski really understand? (1976,
p. 291).
The toughest-minded psychologist of all times, at least in the

opinion of some, once entertained the ^possibility that literature
provided the key for understanding human behavior!
Criterion of Intuitive Knowing
Perhaps the most significant question concerning the concept of

intuitive knowing is the criterion to be employed for determining
whether understanding has been achieved. The fundamental ques¬
tion in this regard is whether intersubjective agreement should be
sought or whether a radical subjectivism is sufficient. One exam¬
ple of an extreme subjective position is characterized by the self-
acknowledged sensitive clinician who has complete confidence in
his ability to intuitively understand his patients. Because his ex¬
periences are so compelling, he sees no need to support them by
additional evidence. If the intuitions of others are at variance
with his own, then they must be defective. Although such arro¬
gance may be difficult to stomach, we should not lose sight of the
fact that in other fields, such as aesthetics, it is generally acknowl¬
edged that some individuals possess exceptional aesthetic sensi¬
tivities, superior to most others. Why cannot the same relation¬
ship prevail in intuitively knowing another person, in sensing his
inner world of experience? The reasons for the expert’s greater
sensitivity may be difficult to identify, but its existence, for many,
is nevertheless obvious.
The above argument has two components that require sepa¬
rate treatment. One eoncerns the possibility that individuals vary
On Understanding: 2
in their capacity for intuitive understanding, which in a clinical

situation may take the form of successfully empathizing with the
patient. The second involves the evidence that is demanded to
demonstrate that such understanding has been achieved. The
answer to the first question should not generate any disagree¬
ment because it would seem obvious that individuals do differ in
their capacity to achieve insight into a variety of intellectual and
psychological problems. The second question gets to the key
epistemological issue about which there is much fundamental dis¬
agreement in contemporary psychology—the criterion of intuitive
knowing.
Intuitive Knowing in Art and Science
Although the similarity in the inner experiences underlying

aesthetic sensitivity in literature and empathic identification in
clinical psychology has been noted, this similarity does not de¬
mand, or even suggest, that the two experiences be evaluated in
the same way. Aesthetic sensitivities can be questioned and re¬
jected, but they cannot be denied. For example, the judgment
that a particular work of pop art is beautiful can be considered to
border on the absurd, but one cannot deny that an observer is
perceiving beauty when he insists he is. “Beauty is in the eye of
the beholder.”
Even when the evaluation of beauty is limited to those who
are acknowledged to be experts, either as critics or as artists, com¬
plete agreement about what works of art possess aesthetic value
will not be forthcoming. And when comparisons are made in the
aesthetic judgments of “experts” from diferent eras, the disagree¬
ments become even more marked; works of art that are consid¬
ered to reflect the essence of beauty in one era are repulsive or
deathly dull to another. Many of El Greco’s contemporaries who
were considered to be his equal or superior are presently ne¬
glected. Such historical differences cannot be attributed to prog¬
ress either in the form of the increased creativity of artists or in
the sensitivity of critics. The notion that modern music is more
beautiful than classical or baroque or that artistic criticism is con-
tinuously improving would be vigorously challenged, A more rea¬

sonable conclusion is that the concept of artistic progress, in a
sense comparable to scientific progress, is inappropriate. Art does
not progress but rather it is expressed in constantly changing
forms. This does not mean that technical procedures (e.g., quality
of clay or paints) cannot improve (or deteriorate) or that within
a given time span (e.g., the Renaissance or Elizabethan) an out¬
burst of artistic creativity cannot take place. Ultimately, what it
does mean is that beauty is intuited, not measured by rational
procedures.
This conclusion is easily misinterpreted. It does not deny
the possibility that aesthetics can be investigated in an objective
fashion. One could strive to discover characteristics of art objects
and observers that yield aesthetic reactions. If such an effort pro¬
duced an adequate theory, that theory would explain aesthetic
reactions but would not measure beauty in the same way that a
scale measures weight. The reason is that the weight of an object
is independent of the observer—its beauty is not.
The changing tastes in art do not necessarily support a com¬
pletely relativistic view of aesthetic value and the implication
that by appropriate training a person could be made to perceive
beauty in any object. One possibility is that the innate structure
of our receptors and nervous system places some limit on the
range of objects to which we can have aesthetic reactions. Leon¬
ard Bernstein expresses, somewhat dramatically, the same notion
about the tonal foundations of beautiful music:
My words are poor, my diagrams even poorer, but this is one

thing I intuitively know to be true, and I will put my hand in the
fire for it, that whatever that creative mystery is ... it cannot
exist or come to be unless it is inextricably rooted in the rich earth
of our innate response in the deep unconscious regions where the
universals of tonality . . . reside (1976, p. 417).
Like Chomsky, on whom his arguments are based, Bernstein

tends to ignore environmental influences. Although innate factors
influence aesthetic reactions, they do not dictate completely the
nature of beauty. Within the range of potentially beautiful ob-
jects great individual differences will occur making it necessary to

consider the properties of both the object and observer when for¬
mulating a theory of aesthetics. The significant point is that a
valid psychological theory of aesthetics is possible, but such a
theory should not be confused with an independent and absolute
scale of beauty.
The thesis that the aesthetic value of an object can be appre¬
hended but not verified does not imply that aesthetic judgments
are necessarily impulsive. Intuitive reactions need not occur in
a flash; they can occur gradually, the understanding becoming
deeper and enlarged. They need not be mystical and inexpressi¬
ble; they can be a product of lengthy contemplation. The diffi¬
culty in characterizing intuition is due to the problems inherent
in phenomenological descriptions (pages 39-64). These diffi¬
culties become disturbing only when viewed within a scientific
framework that seeks precision, reliability, and validity.
Philosophers have dealt with the problems of aesthetics since
the time of Plato, and even though they have sought an answer
to the question, “What is beauty?’’ they do not feel that they have
failed by not achieving a universally acceptable answer. Their
discipline is fundamentally undisciplined in the sense of demand-
ing precise criteria to judge conflicting answers to a common
question. But this need not be perceived as a limitation. Although
the question of “What is beauty” has not been answered in a
manner acceptable to all, the philosophical considerations have
provided an interesting history of ideas and generated a variety
of questions (e.g., the relationship between aesthetics and moral¬
ity) that would not at first glance appear to be related to the
problem of aesthetics. And if the past indicates the future we can
anticipate that the present unresolved controversies in the philos¬
ophy of aesthetics will continue to be debated, new questions will
will be raised, and further disagreements will be generated.
The person who seeks a “valid” interpretation of beauty
within the complete flow of world history is confronted with ap¬
parent chaos. This condition need not be considered to be de¬
fective or unfortunate. Regardless of whether beauty is under¬
stood in any formal objective sense, one should recognize that the
time, effort, and money that people expend in their pursuit of

artistic enjoyment is highly worthwhile; it is one of the most grat¬
ifying aspects of life. One could also speculate that if other areas
of life were as successful in meeting the needs of people as is art,
our present societies would be more attractive. And finally, one
could raise the issue of whether artistic enjoyment might not be
diminished if we had a better understanding of the psychology of
aesthetics; perhaps destroying the mystery in the apprehension
of beauty would lessen the enjoyment.
The analysis of aesthetic sensitivity as a form of intuitive
knowing highlights problems with this form of understanding in
psychology. Aesthetic sensitivity possesses an intrinsic subjectivity
that cannot be overcome. Although one can offer a rational de¬
fense of one’s aesthetic taste, or even an empirical theory to ex¬
plain it, one cannot assert its truth. However, in psychology intui¬
tive knowing can acquire truth value by becoming a component
in an empirical relationship. For example, the ability to empa¬
thize, which is presently being considered as a form of intuitive
knowing, can be investigated to determine whether it influences
the outcome of psychotherapy. Because empathy itself represents
a presumed accurate projection of one’s own consciousness (e.g.,
that of the therapist) onto another (e.g., the patient)—an event
locked into the consciousness of two individuals—it becomes nec¬
essary to relate these intrasubjective states to publicly observed
events. A number of studies have attempted to use a variety of
objective measures to infer empathy (e.g., rating transcripts of
psychotherapeutic sessions, psychological tests, introspective re¬
ports ) and evaluate psychotherapy. In general, the results suggest
that "‘the therapist's empathy . . . facilitates the patient's gain
from psychotherapy" (Luborsky, Chandler, Auerbach, Cohen, &
Baehraeh, 1971, p. 153).^
Intuitive knowing can also be incorporated into a natural-
5. It is not the intent of this analysis to get embroiled in the empirical

problem of the effectiveness of psychotherapy and the role that empathy
plays. My concern is with the concept of intuitive knowing as a source of
understanding and as an empirical variable in psychological investigations.
The analysis of the concept of empathy serves that purpose.
science orientation that seeks deductive explanation. The history
of the Gestalt principle of pragnanz, or good form, illustrates this
development. The perceptual principle of pragnanz states that
stimulus elements that compose a good form tend to be perceived
together. For example Figure 5.1 A is perceived as a pattern con¬
sisting of a half circle intersected by a straight line, as in B, and
not as one containing the components of C.
But what is a good form? Although the definition of good
form is admittedly vague” (KoflFka, 1935), good form itself was
considered easy to intuit. Gestalt psychologists were willing to
accept this kind of understanding because of the confidence they
placed in the capacity of phenomenological analysis to reveal
theoretical principles of perception. Identifying good and poor
examples of good form demonstrated to them that the concept
was valid. But this kind of understanding did not satisfy some
American psychologists who were sympathetic to Gestalt concep¬
tions but not to the subjective quality of their theoretical prin¬
ciples. They reformulated the law of good form into an objective
statement: patterns with good form are redundant, i.e., they con¬
tain surplus information.
Figure 5.2 illustrates the relationship between redundancy
and form quality. Both patterns have been drawn on graph paper
within an area of 100 small squares, 10 rows by 10 columns. Pre¬
dicting the shape of the form by knowledge of its parts is an
easier task for pattern A than for pattern B. This can be demon¬
strated by instructing subjects, prior to their perceiving the en¬
tire figure, to “move,” beginning with the upper left-hand corner,
square by square across the page guessing before the next square
is uncovered whether it is black or white. Subjects will have a
higher percentage of correct hits with pattern A than with pat¬
tern B. Guessing would be required at the beginning for both
patterns, but one would very rapidly learn to guess correctly for
the upper three rows of A because they are all white. The upper-
left black square of the rectangle would be guessed incorrectly,
but soon the errors would disappear as the subject perceived the
entire pattern. Pattern B would generate significantly more errors
because of its irregular contours and unsymmetrical shape. Fig-
Figure 5.1 A stimulus pattern (A) illustrating the principles of good form.
You tend to perceive this pattern as made up of the components in B, not
those in C.
A B
Figure 5.2 Which form has a higher amount of '‘goodness? As the text
explains, form A is more redundant and hence is a “better^' figure.
ure A is redundant, e.g., if you know the upper half you can pre¬
dict the lower half. In essence, good forms are redundant because
the whole can be predicted from a few parts while poor forms
are unpredictable. This principle can be represented mathemati-
cally (Attneave, 1954; Garner, 1962). My concern, however, is
with the development of the meaning of the concept of good
form (pragnanz). Initially it was based upon intuitive knowing,
but later it could be described by objectively defined constructs.
This represents another example of intuitive knowing anticipating
future developments in a natural-science oriented psychology.
Intuitive Knowing and Understanding
Intuitive knowing has been viewed both as an interesting psycho¬

logical phenomenon and as a form of understanding. In the for¬
mer case it has been noted that the phenomenon occurs both in
art and science and is capable of being empirically analyzed. In
the latter case it has been suggested that intuitive knowing fre¬
quently anticipates deductive explanation but by itself fails to
meet the standards of understanding demanded in the natural

seiences.
There is more reliance on intuitive knowing as an indepen¬
dent form of understanding in psychology than is usually ac¬
knowledged. Intuitive knowing is often employed, consciously or
unconsciously, as an ancillary, alternative, or fundamental kind
of understanding. As an ancillary form of understanding intuitive
knowing can be regarded as an essential component within a mul¬
tidimensional interpretive framework. One example of this is the
answers given to a question I once asked a group of clinical psy¬
chologists: “Would you consider a theory of personality to pro¬
vide a satisfactory sense of understanding if it accurately pre¬
dicted your patient’s behavior but failed to provide any insight
into his inner life?” To a woman they all answered, “No!” Some
believed that a deductive theory was incomplete if the proposed
theoretical processes failed to characterize the inner life of pa¬
tients. Others thought that a deductive theory and a phenomeno¬
logical theory could be independent, each serving a separate
need.
Intuitive Knowing as a Fundamental Mode

of Understanding
Intuitive knowing as a fundamental mode of understanding

is rarely discussed among psychologists in an open and frank
manner. It plays a much more important role than most psycholo¬
gists are willing to admit. The position is implicit in the oft-
repeated phrase, “Psychology is an art.” One possible interpreta¬
tion of this statement is that the effective psychologist possesses a
special knack of knowing others, just as the esteemed art or
literary critic with his highly developed aesthetic sensibilities
knows what is good art. Each deals with profoundly human ex¬
periences that, from the viewpoint of the experiencing psycholo¬
gist or critic, are basically ineffable. Each must communicate
these experiences as best as possible. This can be achieved by
adopting communicative criteria that are reasonable rather than
invent ones that are unattainable. Thus we must identify those
individuals with exceptional capacities for intuitive knowing and

trust their judgments, even though in some cases we can enter¬
tain reservations. And as already noted, within communities of
like-minded critics and psychologists, certain individuals attain a
position of authority so that their judgments serve as the criterion
for that group. In essence, their judgments are considered valid
although the means by which the conclusions are reached remain
unclear.
The proponents of intuitive knowing as the sole criterion of
understanding can oflFer two lines of argument to defend their
position. One is that intuitive knowing is employed in science
and logic as a criterion of understanding. For example, it has re¬
cently been suggested (Kolata, 1976) that contemporary mathe¬
matics is confronted with a crisis because extremely lengthy and
complex proofs, which are becoming more common, are too in¬
volved to be properly understood. Chemoff (1976), a mathemati¬
cian, denies that this is the case. He admits that computational
errors and logical blunders can be made in “monster proofs,” but
for the most part they can be assimilated by mathematicians with¬
out confusion and diflBculty. His position is as follows:
The point is simply this: a human mathematician does not attain

an understanding of a proof merely by checking that all the indi¬
vidual steps have been strung together according to the rules. On
the contrary, such detailed mechanical plodding is neither neces¬
sary nor sufficient. What is crucial is to see through the tech¬
nicalities to grasp the underlying ideas and intuitions, which often
can be expressed concisely and even pictorially. Once the gestalt
is perceived, the competent technician can fill in as much formal
detail as needed (1976, p. 276).
The basic point expressed here is that intuitive knowing pos¬

sesses a psychological primacy in achieving understanding. If it
can play this role in mathematics why cannot it play the same
role in psychology?
The second line of argument to support the use of intuitive
knowing as the criterion of understanding is analagous to the one
offered in support of interpretive consistency. There are some sig-
nificant phenomena that cannot be interpreted by natural-science

methodology. Why not employ intuitive knowing in attempting
to understand them rather than accept the notion that they are
incomprehensible? Intuitive knowing is effective in dealing with
problems of everyday life. Why not exploit it in formulating com¬
prehensive psychological theories, especially those concerned with
understanding the human condition? Psychologists, if they really
are serious about interpreting the full range of human experience
and behavior, must resort to methods of understanding appropri¬
ate to their task.
Intuitive Knowing and the Image of Man
Intuitive knowing appears to be at the roots of the claim that

an understanding of human psychology must be based upon an
adequate ‘"image of man” (Chein, 1972; Rychlak, 1968). Without
a proper conception of humans, it is claimed, research findings
will unavoidably be incomplete and misleading. Equating hu¬
mans with machines will inevitably distort human psychology.
Pavlov committed this fatal error when he stated:
Man is, of course, a system—roughly speaking, a maehine—like

every other system in nature subject to the uneseapable and uni¬
form laws of all nature; but the human system, in the horizon of
our contemporary scientific view, is unique in being most highly
self-regulatory. Among the products of man^s hands, we are al¬
ready familiar with machines which regulate themselves in various
ways. From this standpoint, the method of investigating the sys¬
tem of man is preeisely the same as that of any other system:
decomposition into parts, study of the signifieance of each part,
study of the connexions of the parts, study of the relations with
the environment, and finally the interpretation on this basis of its
general workings and administration, if this be within the capaeity
of man (1941, p. 144).
The conception of man as a machine appears in a variety of

forms all of which are equally objectionable to those who insist
that such interpretations strip humans of their intrinsic humanity
and thus distort the image of man. The opposition to machinelike
conceptions of human behavior should not be confused with dis-
agreements eoneerning the relative merits of analytie or holistie

approaches. One might fault Pavlov for his emphasis on analysis
but nevertheless agree with the essence of his remarks that hu¬
mans can be best conceptualized as some kind of mechanism.
Such disagreements really revolve about the kind of machine that
best represents human psychology. The machine models can vary
from slot machines to electronic computers. The psychological
models that these machine models suggest cover the range from
single-unit stimulus-response theories, which assume environ¬
mental inputs determine behavioral outputs, to the information¬
processing conceptions, which postulate numerous mediating
mechanisms that transform incoming information into an orga¬
nized body of knowledge that guides subsequent behavior. All
these formulations, favored by a majority of experimental psychol¬
ogists, it is argued, ignore the “identity” of human subjects, that
unique quality of human self-awareness that cannot reside in any
machine or subhuman animal. In attempting to explicate what
“identity” means, Rychlak asserts that,
. . . subject behavior has to be conceptualized identically with

experimenter behavior (as they are both human beings, etc.).
Hence if the experimenter is capable of formulating hypotheses,
designing experimental tests of these suggestions, and then eval¬
uating the outcome of such efforts according to an arbitrary level
of statistical significance, so too can we think of the subject as
formulating hypotheses, designing tests of what direction he
wants the course of his life to take, and then judging the outcome
of his self-induced efforts to make things end up where he would
like them to be, based on what he favors as a worthwhile even¬
tuality (1976, p. 221).
One interpretation of the above is that the experimenter,

aware of his own identity, or what some might prefer to label
self or ego, fails to attribute the same experience to his subject.
This inner experience of a self-identity possesses a reality for the
experimenter, and to ignore it for the subject creates a false “im¬
age of man.”
A person s identity has several meanings. One extreme would
be to equate it with an attribute of a person s inner experienee,

sueh as his self-image, with the additional assumption that it pos¬
sesses an existenee independent of outer behavior. Interpreting
identity within this context, at least for me, achieves a reasonable
level of intersubjective agreement required to pursue the impli¬
cations of the maxim that an appropriate “image of man” is a pre¬
requisite for understanding human psychology. Thus if the con¬
cept of identity is essential in the psychological analysis of
humans, this means that an adequate theory of behavior will
never be developed as long as conscious experience is ignored as
a set of core observations. According to this argument, methodo¬
logical behaviorism, which asserts that behavior can be investi¬
gated and explained without direct examination of mental states,
is doomed to failure. This predicted failure would apply to the
eflForts of the methodological behaviorist who acknowledges the
significance of inner experience and attempts to create a theo¬
retical model of the mind to represent its influence on behavior
(page 35). It would also apply to the methodological behavior¬
ist who decides that the methods and subject matter of phenome¬
nology and behavioral science are separate and distinct, neither
being superior to the other, just diflFerent (page 67). In sum,
one must incorporate the direct observations of inner experiences
into any satisfactory account of human psychology. In addition,
these inner observations must include that component of inner
experience that represents the identity of an individual. Both of
these conclusions, one methodological, the other substantive, have
emerged from an intuitive conviction that they are correct. It
would be impossible to prove by explicit logical means at this
stage in psychology’s history, that methodological behaviorism is
doomed to total failure or that a combined phenomenological-be¬
havioral approach that does not involve a concept of identity will
prove inadequate. It must be recognized that the claim made
here is not the modest one that it may be more strategic to in¬
corporate inner experience in one’s approach to psychology or
that the concept of identity will prove to be fruitful. Instead what
is being stated is that one intuitively knows that both of these al¬
ternatives are valid.
Self-fulfillment. The concept of identity or ego or self can be

interpreted in a much more idealistic or mystical fashion. It can
refer to the intrinsic potential of a person, the kind of individual
he really is and should be. This kind of conception, which is com¬
mon within certain religious traditions, is captured in the follow¬
ing tale, from Hasidic Jewish lore:
Before his death. Rabbi Zusya said, “In the coming world, they
will not ask me: ‘Why were you not Moses?’ They will ask me:
‘Why were you not Zusya?’ ” (Howe, 1976, p. 642).
Religions usually specify what it means to be fulfilled. If one fully

accepts religious authority then one cannot question its capacity
to describe the meaning of self-fulfillment and to determine when
it has been achieved. Is it possible to ascertain self-fulfillment in
the absence of such authority? Maslow thought so. His concept
of self-fulfillment was based on the hypothesis that humans pos¬
sess a hierarchy of inborn needs, the highest of which is self-
actualization: “The desire to become more and more what one is,
to become everything that one is capable of becoming” (1954,
p. 92).
Maslow offered empirical evidence to buttress his conception
of self-actualization. He provided examples (e.g., Beethoven, Lin¬
coln, Einstein, Eleanor Roosevelt) of historical as well as con¬
temporary personalities who had achieved creative fulfillment of
their potentialities. In addition he listed experiential characteris¬
tics of self-actualizers. One is a peak experience, during which
time a person is completely integrated and fuses into one with
the world; is at the height of his or her powers; lives only in the
present; and becomes free of inhibitions, sincere, emancipated
from the thoughts of the past and the future. Peak experiences
are not situationally defined; they can occur during moments of
artistic creativity, simple reflection, or when giving birth.
The key question is whether the concept of self-actualization
was demanded by empirical considerations or instead has infil¬
trated humanistic psychological theories because of ethical con¬
siderations. In answering this question it becomes important to
recognize that the concept of self-fulfillment antedates psychol-
ogy. It emerges from the powerful philosophieal tradition of hu¬

manism that is not committed to a faith in an Almighty but
nevertheless insists that humans are spiritual beings whose behav¬
ior cannot be reduced to the demands of the flesh.
This general humanistic conception, as well as the specific
one proposed by Maslow, contains two basic assumptions, each
with a different epistemological characteristic. The first is that
human motivation cannot be reduced to animal motivation. The
second is that humans seek to realize their own intrinsic potential
that represents their true self.
The first position is in line with much evidence of species-
specific behavior. Since the behavior of all species differs in some
way from each other it should come as no surprise to discover
that some of the motives of homo sapiens are distinguishable
from those of other animal species. These motivational differences
presumably could be expressed in theoretical formulations with
determinate empirical consequences. Thus it should be possible
to establish points of contact between humanistic motivational
theories and more traditional ones. Theoretical differences among
such conceptions could be evaluated by the deductive mode of
achieving understanding.
When it comes to the assumption that humans possess a mo¬
tive for seeking self-fulfillment, the chances of a theoiêtical reso¬
lution occurring between humanistic and natural-science formu¬
lations appear dim. The apparent incompatibility does not stem
from the basic notion that individual humans have different po¬
tentialities. This widely accepted idea is a key assumption in
the model of aptitude testing. Individuals have many potentialities
(aptitudes) that are, or fail to be, realized depending upon envi¬
ronmental pressures and opportunities. Self-fulfillment is based
upon essentially the same notion although the nature of the po¬
tentialities are neither as precisely defined nor as easily measured
as conventional aptitudes. Self-fulfillment represents a broader,
more complex concept referring not only to specific talents but
also to personal and social values. Presumably, if deemed desir¬
able, humanistic psychologists could develop quantitative mea¬
sures of these potentialities.
The critical epistemological issue associated with the concept

of self-fulfillment concerns the criteria employed to identify those
potentialities that would be self-actualizing. Conventional apti¬
tude testing does not have such a problem. A person has many
aptitudes, but none is more truly representative of the individ¬
ual than is another. They vary only in their probability of being
realized. A vocational guidance counselor may encourage his cli¬
ent to reach his potential in regard to one aptitude in preference
to others because of practical, economic, or social reasons but not
because that particular potentiality reflects ‘what one is.”
To make the issue of the criteria of self-fulfillment more con¬
crete, let us consider the case of Eleanor Roosevelt who, accord¬
ing to Maslow, achieved self-actualization. Would she have been
considered self-actualized if she had divorced Franklin Roosevelt
because of his marital infidelities, become a Republican, and es¬
poused laissez-faire capitalism as the only way to help the poor
and disadvantaged? Or would have Presidents Kennedy and
Johnson been self-actualized if they had avoided getting their
country involved in the Vietnam War? And finally the interesting
question can be posed as to whether Woody Allen, a self-admitted
anhedonic and widely acclaimed humorist and movie director, is
self-actualized.
The thrust of these questions is obvious. Is self-fulfillment a
phenomenon that is basically empirical in nature or does it de¬
pend on value judgments? How would it be possible to resolve
a dispute between psychologists who diflFered as to whether a par¬
ticular person (e.g., Jimmy Carter, Henry Kissinger) was self-
actualized? In answering these questions it is useful to look at the
concept of self-fulfillment as employed by political theorists.
According to the Marxist view, self-fulfillment can only be
achieved in a socialistic society. The institution of private prop¬
erty with its encouragement of competition and acquisition of
material goods prevents members of a capitalistic society from
reaching their human potential. Hitler argued that Aryans could
achieve self-realization only in a Nazi society. With the same line
of argument, but with obviously different ethical commitments
and political outcomes, others insist that only in a political de-
mocracy can humans realize their potentialities. Within the con¬

text of political theory it would appear that the interpretation of
self-fulfillment depends on preferences for difiFerent forms of po¬
litical and economic organizations. In psychology, the concept of
self-fulfillment has also been impregnated with value judgments.
Although apparently concerned with empirical issues in his analy¬
sis of human motivation, Maslow was in fact making moral judg¬
ments when formulating his concept of self-actualization.
The issues in this kind of debate about the behavioral and
experiential characteristics that define self-fulfillment are com¬
plex. They will be treated in Chapter 7. For the time being I have
adopted the position that one cannot by^ empirical means alone
demonstrate that certain kinds of behavior and experiences con¬
stitute the true or real potential of humans. Our empirical knowl¬
edge suggests that humans have a wide (not limitless) variety of
potentialities and that environmental circumstances will deter¬
mine which ones will be achieved. The choice of any criterion of
self-fulfillment always involves a value judgment.®
How can one justify the value judgments that lie at the core
of the concept of self-fulfillment? Maslow suggests that empirical
evidence dictates his view; all individuals have a motive to ac¬
tualize themselves and will succeed if their more basic drives are
satisfied. Our analysis leads to a different conclusion. Moral val¬
ues for Maslow appear to be self-evident truths arrived at by in¬
tuitive knowing. The empirical evidence (examples of self-
actualized people and experiential qualities of a peak experience)
Maslow offers is selected to rationalize his moral preferences.
The conclusion that the ‘Image of man” as conceived by hu¬
manistic psychology is based upon intuitive knowing and is not
solely an empirically based concept does not in any way attenuate
its social significance. An “image of man” in the sense of charac¬
terizing the potential of human beings, plays a central role in the
organization and functioning of any society because it identifies
“good” and “bad” behavior. All societies require some conception
of the “image of man” to support their legal and ethical systems.
6. This does not mean that all eonceptions of self-fulfillment are equally
aehievable or ethieally equivalent.
The significant question is whether it is the responsibility or ob¬

ligation of psychologists to offer such conceptions when they are
not demanded by empirical evidence.
The above analysis of the image of man” is not offered as a
criticism of intuitive knowing as a form of understanding. Intui¬
tive knowing is beyond criticism if considered as the ultimate
type of understanding. The implication of the analysis is that
those who employ intuitive knowing to arrive at conclusions
should not confuse it with other forms of understanding. In this
regard, one must recognize that when intuitive knowing is de¬
fended as a legitimate mode of understanding it must stand
alone. To defend the choice of intuitive knowing by insisting that
it has some built-in validity because it frequently anticipates de¬
ductive explanations would be self-defeating. The difficulty with
such a line of argument is that at times intuitive knowing fails to
coincide with the requirements of deductive explanation although
these failures are usually not publicized by those who expound
the virtues of intuition. Most persons, at some time in their lives,
intuitively knew that the earth was ffat or that heavier objects
fall in space more rapidly than lighter ones. And some theorists
will confess that, at one time or another, they have formulated
models they “knew” to be intuitively true only to discover later
that their faith in intuition was misplaced.
One, of course, can take the position that intuitive knowing
is the mode of understanding for problems that cannot be re¬
solved by deductive explanation, such as the nature of the “image
of man.” This position creates an inconsistency because if one ac¬
cepted the goal of understanding via deductive explanation, then
one would argue that the “image of man” should be contained in
a psychological theory that can explain psychological events. In
other words, it is the purpose of psychological theory to discover
the “image of man” rather than simply proclaim it.
At this point we return once again to the evaluation of hu¬
manistic theories of personality, such as Maslow’s. Do they de¬
ductively integrate some or all of the facts of personality, or do
they provide an intuitively appealing picture coupled with moral
guidance?
Intuitive Knowing Versus Interpretive Consistency
Before elosing this section on intuitive knowing further clarifica¬

tion is required of the distinction between it and interpretive con¬
sistency. The distinguishing characteristic of the latter is that the
interpretation is complex, consisting of a pattern of assumptions
that are rationally interlinked, as is the case for Freudian and
Marxist theories. Intuitive knowing, in contrast, is less systematic,
relying more on understanding through apprehension than ra¬
tionalistic means. Another way of expressing this is to state that
the understanding achieved by interpretive consistency depends
on elaborate rational justification while understanding acquired
by intuitive knowing is essentially impressionistic in nature. This
does not mean, of course, that they do not share any common
properties or that they are always mutually exclusive. Both are
fundamentally subjective in the sense that they rely more on per¬
sonal conviction than public acceptance. In a similar vein, both
are more psychological in nature than is the understanding
achieved through deductive explanation and behavioral control.
Interpretive consistency and intuitive knowing can be con¬
sidered as separate stages in arriving at a broad interpretation
that is not designed to meet the demands of deductive explana¬
tion. Interpretive consistency may represent a rationalization of
intuitive knowing. Apprehending the significance of sex in per¬
sonality development or of economic forces in history could be
the first step in the formulation of Freudian and Marxist inter¬
pretations.
CONSENSUAL AGREEMENT AND

MODES OF UNDERSTANDING
My major thesis is that psychologists employ different standards

of understanding and that the frank admission of the criteria uti¬
lized would do much to clarify methodological and theoretical
disputes that plague psychology. Rather than insist that one form
of understanding is superior to others, it would be more useful to
recognize that different forms of understanding can be employed
and each can be defended for certain reasons and rejected for
The four forms of understanding that I have analyzed can be

diflFerentiated in terms of the public nature of criteria they em¬
ploy. Deductive explanation and behavioral control emphasize
public criteria that yield a higher level of consensual agreement
than do interpretive consistency and intuitive knowing. This does
not mean that consensual agreement is always forthcoming if de¬
ductive explanation or behavioral control is adopted as a criterion
of understanding. There are many reasons that these forms of
understanding cannot guarantee total agreement. First of all, the
meaning of the term consensual agreement is not as clear as some
would like to believe. Consensual agreement among whom?
Everybody? Like it or not, scientists in general and psycholo¬
gists in particular belong to an elitist community that demands of
its members a degree of training and aptitude not shared by
others. Even within the community of psychologists some are
more qualified to judge the validity of a particular statement than
others. The deductive explanation expressed in mathematical
logic can only be evaluated by competent mathematicians. State¬
ments about the control of brain waves can only be assessed by
those who can read EEC records. Finally, there are observational
and theoretical disagreements that require time, sometimes years
or even generations, to be resolved or clarified. But in spite of
these problems one should recognize that consensual agreement
about the validity of statements is often within reach when de¬
ductive explanation or behavioral control is used as a criterion of
understanding even though such statements may have to be modi¬
fied in the future in the face of new evidence. For example, the
statements that the perception of Mach Bands is due to retinal
inhibition is a valid statement, as is also the proposition that the
resistance to experimental extinction in pigeons can be controlled
by manipulating the schedule of reinforcements. Even a psychol¬
ogist who thought such statements were trivial, or at best periph-
eral to the major interests of human psychology, could, I believe,

support the validity of such propositions if he or she were willing
for a moment to adopt the criterion of deductive explanation or
behavioral control. The same would not be true for statements
such as male homosexuality results from an unresolved Oedipus
complex or the “image of man' is intrinsically good. The emphasis
on public criteria that characterizes deductive explanation and
behavioral control is absent from interpretive consistency and in¬
tuitive knowing. As a result, consensual agreement seems impos¬
sible to achieve for those who employ interpretive consistency or
intuitive knowing as a criterion of understanding. Consensual
agreement is possible for a segment of those communities that
share a common conceptual frame of reference. Psychologists
who accept interpretive consistency as an appropriate form of
understanding but do not subscribe to orthodox psychoanalytical
theory would reject the validity of the hypothesis that male homo¬
sexuality is due to an unresolved Oedipus complex. Similarly,
psychologists who adopt an intuitive-knowing frame of reference
would not necessarily agree that the “image of man" is intrinsi¬
cally good. It should be recognized that this failure to achieve
consensual agreement as already noted, does not invalidate the
criterion of understanding employed. It can be argued that con¬
sensual agreement is not the ultimate criterion of truth. In the
human sciences the sensitivity of a single individual, like Sigmund
Freud or Jean-Paul Sartre, may be more significant than the con¬
sensual agreement of thousands of experimental psychologists.
The point of the present discussion is not to resolve these is¬
sues simply because they cannot be resolved, only clarified. But
the discussion does lead to an entirely different method of achiev¬
ing consensual agreement about scientific conclusions. The basic
notion is to apply the methods of judicial processes—the adver¬
sary system that operates in courts of law—to the resolution of
scientific controversies. The exact procedures suggested differ
over a wide range, from simple jury trials to cases before a Su¬
preme Court. The issues considered appropriate for judicial reso¬
lution vary from the determination of valid conclusions to the
proper application of research findings. The former topic, the de-
On Understanding: 2
21S
termination of valid conclusions, is relevant to the issue of under¬
standing because judicial verdicts could provide the basis for
achieving stable beliefs concerning empirical and theoretical
knowledge. For many, especially those who received a tradi-
honal education about the structure of science, the notion that
judicial processes can reveal scientific truth represents not only
an absurdity but a threat to the very foundation of science. If
nothing else Galileo’s experiences with a judicial system illus¬
trate the incompatibility of legal and scientific procedures and
The trial of Galileo, which led to the verdict that the sun re¬
volves around the earth, need not be considered as representative
of all judicial-scientific courts. The verdict in Galileo’s trial would
have probably been dififerent if the judges were qualified scien¬
tists instead of churchmen who believed that knowledge comes
from God or at least from religious authority. It can also be ar¬
gued that the judicial system of the sort that common and Roman
law encourages bears an intrinsic similarity to the scientific
method. I recall an attempt to disabuse introductory psychol¬
ogy students of common misconceptions about the nature of psy¬
chology. A warning was given that the course would be limited
to conclusions warranted by evidence and logic.” A perceptive
student noted that such a criterion had as much relevance to law
as to science.
The most revolutionary justifieation for arriving at warranted
scientific conclusions by judicial means is that the behavioral and
soeial seiences demand it. This thesis is advaneed by Levine
(1974), who questions the adequacy of experimental methods
and statistieal inference to deal with signifieant soeial phenomena
as they oecur in a natural setting. A research project, for exam¬
ple, cannot employ random assignment when comparing the be¬
haviors of ‘ sehizophrenics and the brain damaged, ... the alien¬
ated with the unalienated, . . . aehievement test scores of blaek
and white children.” With such interests a researcher is eon-
fronted with a ehoice between employing inadequate natural-
scienee methods or developing “additional methods of approach
that are in their own right rigorous but appropriate to the sub-
stantive problems with which they deal.” Levine denies “that the
model of the laboratory experiment is the best if not the only way
of becoming certain of truth” and suggests that a “method based
on legal proceedings . . . may be suitable for many of psychoh
ogy's problems, in particular for field, clinical, and evaluative
studies.”
Levine should be given credit for offering an alternative to
more traditional scientific methods of arriving at warranted con¬
clusions. It is easy to criticize behavioristic methodologies for
their failure to provide unequivocal information that can be used
to solve significant social problems. It is also easy to state that
other methods are required. But what nqethods? At least he has
the courage to try to answer that question.
Levine’s position, however, is based on two questionable
premises. First, he equates natural-science methodology with
laboratory experimentation. Second, he denies that laboratory
phenomena in the behavioral sciences can reflect principles that
govern naturally occurring events. The history of physics fails to
jibe with these assumptions. Many warranted conclusions, emerg¬
ing within a deductive explanatory framework, have been arrived
at by the observations of natural events such as the motion of as¬
tronomical bodies. In addition, a productive interaction has fre¬
quently occurred between observations of natural events and the
results of laboratory studies. The history of the biological sci¬
ences fails to justify an antithesis between laboratory and natu¬
rally occurring events. Darwinian theory of evolution and its
subsequent impact on the experimental analysis of genetic de¬
termination underline the reciprocal relationship that can occur
between laboratory phenomena and natural events. I would like
to suggest that laboratory studies of individual differences and
the related development of psychological tests have led to a fun¬
damental understanding of individual differences in behavior in
social situations such as in schools and industry. There seems to
be no justification to limit the natural-science method to the
laboratory.
The failure to generalize from laboratory studies to complex
social situations may not represent any intrinsic limitation of the
On Understanding: 2
215
former but may instead be a consequence of any number of fac¬
tors such as lack of ingenuity in designing laboratory studies, in¬
adequate theorizing, unrealistic expectations, and a desire to solve
social issues that do not lend themselves purely to empirical reso¬
lution. We must also realize that some real-life events are too
complex for simple interpretations. Physicists cannot predict the
exact course that a leaf will take when descending in a wind-
stonn. Psychologists should not try to outperform physicists in
their theoretical predictions. And like physicists as well as weather¬
men, physicians, and surgeons, psychologists must frequently
couch their predictions in probability statements. Finally, there
may be problems, and social psychologists have exhibited a pen¬
chant for getting involved in them, that cannot be resolved by
empirical means. Solutions to many fundamental social issues that
surround problems such as poverty, race relations, minorities, and
others, will not emerge from empirical efforts alone.
I am not arguing that if we exclude unrealistic expectations
and avoid empirically unresolvable problems that theories based
on laboratory experiments will be able to make valid leaps to
real-life situations. I doubt if that will happen. My intuition is
that a closer reciprocal interaction between the laboratory and
natural events, of the sort that has taken place between ethology
and experimental comparative psychology, will be more produc¬
tive. Social psychologists too rapidly climbed on the method¬
ological bandwagon of small-group research believing that it
would automatically yield the basic principles of social behavior.
For many the behavior of the small group became an end in it¬
self, an encapsulated world isolated from the real one. Small-
group research generated many studies and much amusement
but little concern with whether it was dealing with social pro¬
cesses that operated in the natural environment. Yet there is no
reason to deny the possibility that laboratory studies can contrib¬
ute some understanding of individual and social events occurring
in the natural environment.
The core issue is whether truth can be adjudicated. In one
sense it can, if we accept the notion that the decision of a science
court will be legally binding on all. This would be one method of
arriving at stable beliefs. Once and for all, it might be argued,

such nagging problems as the nature of intelligence and language
acquisition and the validity of cognitive dissonance theory could
be put to rest. Are such hopes reasonable or realistic? The Su¬
preme Court dictates the meaning of the United States Constitu¬
tion but that does not truly guarantee agreement because dissi¬
dent views are not squelched, and the possibility is always open
to modify or change the decision by new arguments or by chang¬
ing the membership of the court. Imagine what a Chomsky or a
Skinner would do if the scientific court ruled against them. It is
difficult to believe that they would modify their scientific judg¬
ment one whit. Judicial processes would.be unsuccessful in coun¬
tering the principles inherent in Galileo's statement: "Tn questions
of science the authority of a thousand is not worth the humble
reasoning of a single individual."
The distinction between employing a judicial-type system to
arrive at scientific truth and adopting a particular social policy
must be kept in mind. Levine fails to make this distinction per¬
haps because he does not believe that separation of facts from
values is possible. However, my point is that the adversary sys¬
tem is not only counterscientific but also incapable of resolving
theoretical controversy and establishing stable beliefs about psy¬
chological knowledge. In contrast, the adversary system may
prove to be an effective procedure for applying scientific knowl¬
edge in a political democracy. Before this topic can be discussed
it will be necessary to analyze the role of value in psychology,
the subject matter of the next two chapters.
Psychology and
Values: Epistemological
Considerations
1. A methodological analysis of modern psychology demands a clari¬

fication of the relationships between empirical evidence and moral
judgments.
2. Any single overall statement about the relationship between sci¬

ence and values (e.g., “science is value free,” “science and values are
inextricably intertwined”) must be rejected because the relationship
between the two is not the same for different scientific activities (e.g.,
data collecting, applying scientific knowledge).
3. The methodological position that moral principles are logically dis¬

sociated from empirical evidence has been challenged by the view
that the social and behavioral sciences operate within the value orien¬
tation of their host society thus preventing the isolation of facts from
the ethical preconceptions that determine their ultimate use. This posi¬
tion is analyzed in relation to the controversies surrounding sociobiol¬
ogy and intelligence testing. The conclusion is drawn that facts and
values are logically dissociated but empirical evidence is relevant to
choices among competing social policies that share common ethical
goals.
217
EPISTEMOLOGY AND ETHICS
Epistemology is concerned with knowledge while ethics deals

with morality. A popular philosophical opinion is that a sharp di¬
vision exists between engendering knowledge and evaluating its
moral implication. Consequently, the inclusion of the topic psy¬
chology and values in a book dedicated to the methodological
analysis of psychology may appear inappropriate. Without neces¬
sarily denying the lack of a direct relationship betwêen episte¬
mology and ethics, the argument can be advanced that a method¬
ological analysis of psychology would, of necessity, be incomplete
if the topic of ethics were ignored. Problems of morality are
frequently involved in the discovery and application of psycho¬
logical knowledge. Many psychologists have chosen their profes¬
sion because of their deep conviction that psychology can con¬
tribute to the building of a better society. For them, psychological
knowledge is intimately related to ethical issues. The fundamen¬
tal problem, which they cannot ignore, is the appropriate role
that ethical decisions should play in the profession of psychology.
Finally, ethics can properly be considered a part of psychology
simply because moral behavior, like any other form of behavior,
is subject to a psychological analysis. An understanding of mod¬
ern psychology demands the clarification of the relationship be¬
tween psychology and values.
THE ROLE OF VALUES IN SCIENCE
Two statements that have generated much confusion and mischief

are “Science is value-free” and “Science and values cannot be dis¬
sociated.” Each statement, in its own peculiar way, has distorted
the role of psychology in society and thus has hampered psychol¬
ogy’s contributions.
The global beliefs that science is value-free and science and
values are inextricably intertwined tend to overlook the fact that
scientific effort consists of a variety of activities, a few of which
Psychology and Values: 1 219
are selecting research problems, forming hypotheses, collecting

data, determining the nature of an empirical law, interpreting
the theoretical significance of data, applying knowledge to solve
a practical problem. For example, the attempt to reduce ethnic
and racial antagonisms is usually based upon the value judg¬
ments that such social conflicts should be eliminated and that
members of minorities should be treated fairly. The involvement
of value judgments is appreciably less when dealing solely with
a factual problem, e.g., investigating ontogenetic changes in clus¬
tering. However, it must be recognized that even when such ap¬
parently ‘neutral” problems are investigated, values are involved
because science itself possesses a value system of its own. One
obvious value is that of honesty. Science is a social effort that de¬
mands of its practitioners accurate reporting of data. In science it
is not a matter of honesty being the best policy, it is the only pol¬
icy. Science would inevitably collapse if dishonesty were ram¬
pant. This demand for veracity does not imply, as noted in the
previous discussion of understandings that biases in favor of one
theory over another should not occur. In fact, one can suggest
that persistence in exploring the full potentialities of theoretical
positions is a value implicit in the scientific method in its search
for truth. I am, however, not interested in a complete listing of
all the values that are inherent in the scientific method, an effort
that would inevitably generate marked disagreements.^ The sig¬
nificant question is whether any value is intrinsic. If so, the no¬
tion that science is value-free must be rejected. To justify this
conclusion it is sufficient to identify the essential value of honesty
in science. I do not know of any disagreement with the principle
that it is morally wrong for a scientist to fake or distort data, or
even to treat evidence in a cavalier fashion. This does not mean
that all scientists abide by this mandate of honesty. Nor does it
mean that dishonesty will always be punished or even detected,
or that when it occurs it will lead to a permanent distortion of
scientific knowledge. In the long run, dishonest reports are cor-
1. For example, no disagreements would prevail among scientists about

their rights to question reigning theoretical assumptions, but at the same
time many would dispute the rights of all scientists to have free access to
scientific information when defense secrets are involved.
rected in the light of further research. What the value of honesty

does mean is that for an overwhelming majority of scientists hon¬
esty is a moral imperative that effectively controls their behavior
when reporting results.
In essence, in science honesty is good and dishonesty is bad.
But what is good and what is bad? This question lends itself to
two possible interpretations. What are the criteria that we employ
when we characterize some behavior as good or bad? What are
the psychological reasons that some behaviors are ‘"good” while
others are "Tad”? The first question merely asks for an operational
definition while the second seeks a theoretical explanation. For
the present my concern is with the first query because I am try¬
ing to understand what is meant by a value so that I can com¬
prehend its relations to fact. The second question will be analyzed
later when attention shifts to the analysis of the relationships that
might prevail between values and facts.
Some would argue that ethical terms, such as good, valuable,
or desirable, are indefinable. I interpret such a position to mean
that goodness, for example, is a rock-bottom experience that can¬
not be analyzed into more basic experiences. However, for my
purposes, a behavioral analysis of ethical terms in preference to a
phenomenological one has advantages in identifying what is con¬
sidered to be good and bad, or better and worse. From this point
of reference goodness represents a choice and judgment. Honesty,
for example, is chosen in preference to dishonesty because it is
judged to be good while dishonesty is judged to be bad. When
discussing ethical values within this framework, the operational
meaning is expressed by choice behavior and its linguistic justifi¬
cation. Admittedly, this simple definition of a value judgment is
not very satisfying because it neither illuminates nor is sufficiently
restrictive. But the purpose of an operational definition is not to
clarify but rather to identify (page 88). The fact that my defini¬
tion of a value applies equally well to the trivial preference for
chocolate to vanilla ice cream as it does to the profound choice
between good and evil should not be disconcerting because a
sharp dividing line may not exist between a value and a prefer¬
ence; they merely vary in importance.
Now that the nature of value judgment has been identified

or, perhaps more properly speaking, approximated, we can pro¬
ceed to the crucial issue of the relationships between facts and
values. In the analysis of this problem it is important to distin¬
guish between the logical and psychological aspects of the rela¬
tionship. In the case of the former the issue is whether facts logi¬
cally demand the adoption of certain values. In other words, are
facts simply descriptive or prescriptive? The second relationship,
the psychological, treats ethical commitments as a form of behav¬
ior that can be empirically investigated to determine why they
were adopted and what their consequences are for the individual
and society. Although the two issues are distinct we will discover
that their subtle interactions often make it difficult to maintain
their independence.
THE LOGICAL DISSOCIATION BETWEEN

VALUES AND FACTS
Passmore (1953), in attempting to answer the general question,

“Can the social sciences be value-free?” suggests that facts and
values can and should be dissociated. In explicating his position
he considers a problem of social policy, “Ought we to abolish
class distinction?” Let us assume that in attempting to answer
this question data are obtained that indicate that the abolition of
class distinctions decreases servility and diminishes cultural vari¬
ety. Although agreeing about the validity of the empirical evi¬
dence, two individuals could reasonably disagree about the social
policy that should be adopted, i.e., whether to retain or eliminate
class distinctions. “Our policy is determined in the light of facts,
but is not deduced from them” (Passmore, 1953, p. 675). In
short, a particular social policy does not follow logically from the
data. The only moral justification for adopting a policy of pre¬
serving or abolishing class distinctions, after considering the evi¬
dence, is the acceptance of an ethical principle that values cul¬
tural diversity more than social equality or vice versa. Within
such a moral framework one could argue logically for the preser-
vation or for the abolition of class distinctions. But the core of

that logical argument resides in the moral premise rather than in
the empirical evidence. By being morally committed to a princi¬
ple that values cultmal diversity or abhors servility, one could
support a social policy concerning the abolition of class distinc¬
tions. Without the moral premise the empirical evidence is purely
descriptive; it does not prescribe social policy concerning class
distinctions.
The argument that moral principles are logically dissociated
from facts, as expressed by Passmore, is vulnerable to several ob¬
jections. First, the empirical problem raised—pitting cultural di¬
versity against social equality—is a trivial one that does not rep¬
resent crucial human events involving life and death, freedom
and slavery, happiness and despair that demand consideration of
the factual basis of values. Second, the detached view that is en¬
couraged in deciding between cultural diversity and equality is
illusory. All science, but especially social science, proceeds within
a social context in which pure objectivity cannot operate. The at¬
tempt to invoke a detached, impartial view is essentially a politi¬
cal act designed to achieve certain social goals while simultane¬
ously thwarting others. Third, the empirical study of human
behavior will reveal a source of moral values. Knowledge of
human behavior essentially endorses a moral viewpoint. Fourth,
regardless of attempts to analyze the nature of moral discourse,
the central issue for mankind is to learn to act rightly. No mat¬
ter how persuasive an analysis of metaethics may be, if it does
not provide a practical ethics to govern behavior it will have
failed. Although these four arguments against the position that
values cannot be deduced from facts are epistemologically inter¬
related, they nevertheless represent distinct positions that can
best be analyzed separately.
Moral Values and Significant Human Events
There is no question that the values assigned to cultural diversity

and servility appear insignificant when compared to more impor¬
tant values associated with human existence, e.g., life, liberty.
and the pursuit of happiness. One can argue that a social policy
that endorses death, imprisonment, and denies the right to achieve
happiness is morally repugnant. The social consequences of such
policies, as in Nazi Germany, are so abhorrent that one need not
condemn them by reference to any higher moral authority or
principle. However, a more careful analysis suggests otherwise.
No matter how repugnant Nazi atrocities were, their moral re¬
jection is not fundamentally factual. Taking a life need not be
considered morally wrong. Many individuals who found Nazism
morally abhorrent would nevertheless condone a person’s right to
commit suicide, especially when succumbing to a painful terminal
illness. Some would also condone killing in a war that they
thought justified, as was the case for most Americans in World
Wâr II. Imprisonment would also be morally acceptable if it were
employed against a person who if free would probably commit a
felony. Nor would it be morally objectionable to deprive a sadist
of any opportunity to seek “happiness” by satisfying his motive.
The point is that facts of murder, imprisonment, and denial of
the pursuit of happiness do not logically lead to their moral re¬
jection. One must recognize that the abhorrence of Nazi atrocities
is not an expression of a true moral judgment but, instead, what
is considered to be a sound moral judgment. One could describe
the ethical principles that one has adopted and then recognize
that they are incompatible with Nazism. This leads to the sound
moral judgment that Nazism should be opposed because it is
morally repugnant. Just as was the case for deciding either to
abolish or to retain class distinctions, the facts themselves do not
lead logically to a moral imperative. The moral position one
adopts, whether one deals with “trivial” or “significant” facts, has
its origins in the ethical framework within which the facts are
judged.
Values and Political Preconceptions
It can be argued that the example oflFered by Passmore (1953),

possibility of a detached choice between social
equality and cultural diversity, essentially distorts the nature of
the options. It suggests a rational choice between two simple al¬

ternatives where in fact the choice is really complex because two
diflFerent political and economic systems are at stake. Although
many members of the middle class perceive the option of cultural
diversity as particularly attractive due to their fondness for a va¬
riety of ethnic and artistic traditions, they nevertheless would be
reluctant to admit that they harbor undemocratic attitudes that
condone servility. The manner in which Passmore presents the
choice eflFectively hides the fundamental conflict that is at stake—
a choice between an economically democratic society and an ex¬
ploitative one in which some members are forced to behave like
servants, if not slaves. In fact, the suggestion that cultural diver¬
sity has to be purchased at the price of individual servility is a
factual distortion. It may be true for a capitalist economy but cer¬
tainly not for a socialist society in which ethnic cultural integrity
and social democracy go hand in hand. Although the facts do not
demand the adoption of the goal of ethnic diversity or social
equality, facts associated with the social, economic, and political
organization of a country will determine whether such a choice
needs to be made. Thus this argument would suggest that Pass¬
more’s general conclusion that facts and values can be completely
dissociated is misleading because it ignores the nature of the so¬
ciety in which ethical decisions are made.
Although Passmore’s hypothetical example of the choice be¬
tween ethnic diversity and social equality was useful in introduc¬
ing the epistemological problem of the relationship between facts
and values it now becomes necessary when analyzing political in¬
fluences to deal with real problems that permit the discussion of
specific factual and ethical issues. Two such problems, both of
which have generated intense emotional debates, are sociobiology
and intelligence testing.
Ethical Issues in Sociobiology
The appearance of Sociobiology: The New Synthesis by the

zoologist Edward O. Wilson (1975) touched off a storm of de¬
bate by suggesting that many significant forms of human social
behavior have genetie roots rather than being shaped totally by
environmental factors. Wilson supported his speculations about
the genetic determination of human social behavior with evi¬
dence obtained from animal and insect societies. For example, al¬
truistic behavior is exhibited by ‘‘^soldier” termites who, in re¬
sponse to ants that attack their colony, will explode themselves
and spray poison over the invaders. From such evidence it is sug¬
gested that genetic factors, admittedly to a lesser degree, also op¬
erate in human social behavior. For example, the self-sacrificing
behavior of mothers toward their children may not be simply an
expression of effective moral training but, instead, of genetic in¬
fluences that have helped our species to survive.
One might think that the suggestion that human social be¬
havior is to some extent genetically determined would surprise
and offend no one. However nature-nurture debates have a his¬
tory of arousing strong emotional reactions due in part to past
participants offering an oversimplified view of genetic and envi¬
ronmental contributions to complex forms of behavior. Gaylord
G. Simpson, the distinguished evolutionist, offers an appropriate
warning:
Much of the critical discussion of sociobiology . . . has been an¬

other form of the nature-nurture debate, a discussion that has
proved futile and indeed meaningless because that is not a legiti¬
mate either-or question. Man is not born a tabula rasa, nor is born
a programmed automaton. When the argument approaches that
extreme polarization, it is sensible to say, ‘‘A plague o’both your
houses” (1977, p. 774).
It is one thing to argue that human social behavior is rigidly

and inevitably controlled by genetic factors but quite another
thing to state that genetic factors predispose humans to acquire
certain social behaviors in preference to others. In the former
case the argument is based upon the assumption that humans are
programmed automatons” while in the latter case the suggestion
is that genetic and environmental factors interact. It is not my
purpose to debate the issue as to which of these two positions
Wilson adopts although some would insist that it is the latter
(e.g., Wade, 1976). Neither will it be my eoncern to judge whether

Wilson and other soeiobiologists have misused animal analogies
in interpreting human social behavior although my impression is
that they have. My sole concern is with the ethical implications
of empirical laws that demonstrate that genetic factors influence
behavior.
If one accepts the position that facts and values are disso¬
ciated, then one can conclude that the implication of a genetic
predisposition does not logically lead to the adoption of a particu¬
lar value. In primitive societies men tended to engage in strenuous
activities that involved lengthy periods of travel (e.g., hunting)
while women usually indulged in less physically taxing endeavors
such as preparing food and caring for the young (Murdock, 1937).
This division of labor would have survival value for the group
because important jobs (e.g., fighting, hunting, nursing) are as¬
signed to those possessing aptitude for them. It does not, how¬
ever, follow logically that such a division is morally justified un¬
less one accepts the survival of the group as the ultimate value.
And even if one did accept such a moral position it could be
argued that a division of labor based strictly on gender would be
counterproductive because some women, especially if they re¬
ceived appropriate training, would be more effective at some
strenuous activities than some men. Today in a modern industrial
society the justification for assigning jobs on the basis of sex, even
when one adopts survival as a moral imperative, becomes less
warranted. Most professional, business, and political positions do
not require physical strength; society’s future is served better by
considering women along with men for its most important tasks.
And even when specifying the requirements for a job that re¬
quires strength, such as a ditchdigger or stevedore, sex need not
be considered even though the probability would be greater that
a larger proportion of men than women would meet the desired
standards.
The preceding analysis, it should be recognized, does not so
much encourage the elimination or preservation of sex discrimina¬
tion as argue that the biological facts associated with being male
and female do not logically lead to any moral position regarding
the social treatment of the sexes. And this conclusion applies with
equal force to situations in which survival of the individual or the
society would be enhanced if sexual discrimination was practiced.
History is filled with incidents in which individuals (e.g., early
Christian martyrs) or groups (e.g., the Jews at Masada) decided
that some moral values were more important than survival itself.
Political Involvement in Science
The argument against the position that facts and values are
logically independent is that it fails to represent the realities of
science. According to the Sociobiology Study Group, a group of
scientists who vehemently criticized Wilson’s Sociobiology, scien¬
tific knowledge and political preconceptions are intertwined:
Our central point is that sociobiology—like all science—proceeds

in a social context: “pure objectivity” is as much a myth for socio¬
biologists as for science reporters. All attitudes toward sociobiol-
ogy—ours as much as any—reflect certain political preconceptions
which need to be made explicit. The weaker the restraint of fact,
and the closer the subject to immediate human concern, the
greater the influence of these preconceptions (Alper et al., 1976,
pp. 424-27).
Although strongly and sharply stated the position neverthe¬

less contains many ambiguities, not the least of which is the
implication that political preconceptions are inevitably involved
in scientific eflForts. The argument is reminiscent of the discus¬
sion surrounding the issue of “immaculate perceptions” in scien¬
tific observations: whether pure “objectivity” can be achieved or
whether theoretical preconceptions determine what is observed
(pages 48-64).
There is no doubt that political preconceptions can be in¬
volved in the interpretation of factual data. One case in point is
that of Trofim Lysenko, the Russian agronomist, who claimed that
by changing the environment of the seeds of spring wheat it was
possible to impart to them the genetic characteristics of winter
wheat. He interpreted his findings as supporting the doctrine of
inheritance of acquired characteristics, which had been discred¬

ited by modern geneticists. Lysenko argued that his theoretical
interpretation was consistent with Marxism, and in this view he
received support from Stalin and the Soviet Central Committee.
Only after the death of Stalin were Soviet scientists able to pub¬
licly reject his genetic theory and its presumed tie to Marxian
orthodoxy.
American geneticists have also been influenced by political
preconceptions. Charles B. Davenport in 1917 drew erroneous
conclusions about human race crossings from data obtained with
hens (Provine, 1973). It had been discovered that crossing Leg¬
horns, who had been bred to lay eggs but not to brood, with
Brahmas, who had been bred to lay eggs and brood and hatch
them before laying more, resulted in offspring that lacked the ad¬
vantages of each strain. The hybrid were neither good layers nor
brooders, thus failing to serve the needs of chicken farmers with
or without artificial incubators.
Davenport assumed that natural selection encouraged the
development of genetic traits that were harmoniously adjusted
both with each other and to the environment. Crossbreeding, rea¬
soned Davenport from the evidence obtained with Leghorns and
Brahmas, created the danger of disharmony among genetic char¬
acteristics such as might occur when a member of a human race
with large stature had a child with one from a race with small
stature. A disharmonious relationship might occur between the
inherited stature and size of the circulatory system of the off¬
spring; the circulatory system might be too large to fit into a small
stature. Davenport suggested that similar unfortunate results
might occur from crossbreeding between blacks and whites. One
often sees in mulattos an ambition and push combined with in¬
tellectual inadequacy which makes the unhappy hybrid dissatis¬
fied with his lot and a nuisance to others.” He concluded: “Mis¬
cegenation commonly spells disharmony of physical, mental and
temperamental qualities and this means also disharmony with en¬
vironment. A hybridized people are a badly put together people
and a restless, ineffective people” (quoted in Provine, 1973,
p. 791). Although Davenport was careful to avoid labeling any
race as inferior to another, other genticists of that era supported

legislation against interbreeding between white and blacks to pre¬
vent the degradation of the population.
In viewing these two examples of the political influence on the
science of genetics, it would be wise to avoid, or at least de-empha-
size, the murky problem of explaining the scientist’s behavior.
Did Lysenko and/or Davenport and their co-workers deliberately
try to deceive their audience, or were they merely ""unconscious”
victims of a dominant political atmosphere that induced them to
perceive and interpret evidence in a biased fashion? I know of no
simple litmus test that could discriminate between these two pos¬
sibilities. For the present discussion it is not necessary to attempt
such a dilBcult clinical judgment. The point at issue is whether
political preconceptions can determine scientific conclusions, and
the evidence suggests an affirmative answer. In fact. Provine’s
analysis, which traces the history of geneticists’ attitudes toward
the biological consequences of racial interbreeding, reveals an in¬
teresting sequence of events.
Geneticists in England and the United States clearly reversed

their published remarks on the effects of race crossing between
1930 and 1950. The reversal occurred in two steps. First came
the change in the 1930’s from a condemnation of wide race
crosses to an agnostic view. The second change, from the agnostic
view to the belief that wide race crossings were at worst biologi¬
cally harmless, took place during and shortly after World War II.
(1973, p. 796).
These ""reversals,” however, ""occurred in the light of a little

new compelling data from studies of actual human crossings.”
According to Provine, the major reason that caused geneticists
to change their minds,
was the revulsion of educated people in the United States and

England to Nazi race doctrines and their use in justifying exter¬
mination of Jews. Few geneticists wanted to argue, as had the
Nazis, that biology showed race crossings was harmful. Instead,
having witnessed the horrible toll, geneticists naturally wanted to
argue that biology showed race crossings was at worst harmless.
No racist nation could misuse that conclusion. And geneticists did

revise their biology to fit their feelings of revulsion (ibid., p. 796).
From his historical analysis, Provine concludes:
It is necessary and natural that changing social attitudes will in¬

fluence areas of biology where little is known and the eonelusions
are possibly socially explosive. The real danger is not that biology
ehanges with society, but that the public expects biology to pro¬
vide the objective truth apart from social influences. Genetieists
and the publie should realize that the science of genetics is often
closely intertwined with social attitudes and political considera¬
tions (ibid., p. 796).
Provine’s conclusions are consistent With those of the Socio¬

biology Study Group, but the reasons for the agreement are not
spelled out. Is it because political biases cannot be unravelled
from empirical knowledge? The position that political beliefs can¬
not be separated from empirical evidence is diflficult to defend
against the fact that some scientists refused to succumb to politi¬
cal and other pressures. Several leading Russian geneticists stuck
to the traditional view that acquired characteristics cannot be in¬
herited. Similarly, in spite of dominant biases in the United States
against black and white interbreeding, several American geneti¬
cists questioned the data upon which this social prejudice was
based (Provine, 1973). Were these protests of the Russian and
American geneticists merely an expression of competing political
philosophies, or were they determined by the demands of scien¬
tific traditions in evaluating evidence? The latter interpretation
appears appropriate. Lysenko’s opponents were not necessarily
anticommunists; they disagreed, apparently justifiably, with his
interpretation of his data while failing to perceive any incom¬
patibility between Marxism and traditional genetic theory. In a
similar vein, several American geneticists took reasonable excep¬
tion to the kind of evidence that was offered about the ill effects
of race crossings. Some even recognized the difficulty of drawing
conclusions from data or race crossings for social policy; if racial
crossings had some possible negative biological or psychological
effect would that justify legal sanctions against racial interbreed-
Psychology and Values: 1
2S1
ing? Adherence to democratic ideals, in which individuals are free

to choose their own mate, can be considered to be more valuable
than minimizing possible negative consequences of racial inter¬
breeding.
The line of argument that scientists are incapable of resisting
accepted social dogma collapses when attention is focused on the
great leaps forward in the history of science. Although the Roman
Catholic Church in 1616 denounced the Copernican system as
dangerous to their faith, Galileo published a book for the layman
that supported its validity while simultaneously questioning the
Ptolemaic system, which the church embraced. Charles Darwin
a owed the weight of evidence to change his preconceptions
about the origin of species. Darwin had studied for the ministry,
and while on the Beagle expedition he wrote notes in his diary
that were m line with a creationist interpretation (Ghiselin, 1969).
But as we know from history the data he collected forced him to
adopt an evolutionary theory. Although all scientists may not be
able to resist political or religious preconceptions, some obviously
can. If this is so, one must reject the view that ‘pure objectivity”
is a myth. Certainly scientists possess preconceptions—theoretical
political, religious-but these “subjective” factors do not always
operate in an all-powerful fashion in the final determination of
empirical evidence and its theoretical interpretation.
The evidence offered to support the position that preconcep¬
tions in the form of political and religious influences can be re¬
sisted also suggests that political biases can be unravelled from
empirical knowledge. Again it is important to note that the argu¬
ment is not that these influences do not have an effect upon the
course of science but instead that their influences can be mini¬
mized and overcome by the force of data and reason. If that be
the case then the position that powerful political or religious pre¬
conceptions prevent the dissociation of facts from values must
also be rejected.
The preceding analysis, which maintains that facts and values
are not logically related, fails to answer the question as to why
some scientists insist that they are. Is this insistence simply a con¬
fusion about the intrinsic properties of the scientific endeavor, or
does it represent a philosophical position that is extrinsic to the

scientific eflFort itself? Although the answer to this question de¬
pends on clinical judgment, and therefore one must be very cau¬
tious in one's interpretation, it does appear that the latter factor
plays the dominant role for many scientists in current contro¬
versies surrounding the implications of genetic influences on hu¬
man behavior. One can distinguish between those researchers
who frankly wish to politicize the issue from those who attempt to
separate the empirical evidence from social policy. For example,
Richard Lewontin, one of the leaders of the Sociobiology Study
Group, states: “Nothing we can know about the genetics of hu¬
man behavior can have any implications for human society'
(quoted in Wade, 1976). This statement is consistent with the
epistemological position supported by the analysis that facts and
values can be dissociated. Nevertheless Lewontin's position di¬
verges from that neutral stance by adding an empirical generali¬
zation to the epistemological assumption. Inevitably human ge¬
netic research will be used for some political goal or as Lewontin
states: “The process (of doing research in human genetics) has
social impact because the announcement that research is being
done is a political act” (ibid.). In essence, he argues that research
in human genetics “is bound to produce a pseudo science that
will inevitably be misused” (ibid.). Such research, as in Wilsons
Sociobiology, “represents an effort to cloak in modern terminol¬
ogy the age-old political doctrine that the main features of human
social existence are biologically determined” (Alper et al., 1976).
Several assumptions underlie this argument: (1) research on
human genetics will be misinterpreted and/or distorted; (2) the
misinterpretation and/or distortion will produce undesirable ef¬
fects; and (3) the evidence, whether misinterpreted or distorted,
plays an influential role in political decision-making.
In evaluating this analysis consider the effect of different so¬
cial philosophies “based on” evolutionary theory, which assumes
a higher reproduction rate of fit as compared to unfit organisms.
Most people are familiar with the employment of evolutionary
theory to justify social practices (social Darwinism) that would
be morally repugnant to most present-day Americans. In the nine¬
teenth century men of wealth and political power argued against
233
child labor laws, public education, compulsory safety regulations

in factories and mines, and other social welfare legislation be¬
cause such laws went ‘against nature” as implied by evolutionary
theory. Society, like the jungle, was perceived as an arena in
which individuals competed against each other to survive and re¬
produce. Only if complete freedom was permitted could natural
laws operate. Welfare legislation encouraged the survival of the
less fit.
Evolutionism, however, was employed not only to justify the
status quo but also social reform. Karl Marx wanted to dedicate
the first volume of Das Kapital to Darwin. He and other social
reformers interpreted Darwin’s theory as indicating that social
organizations were constantly undergoing change and that there¬
fore political governments and economic processes were alterable.
The forces of history would lead to progressive changes in social
organizations that would serve the needs of the people to survive
and reproduce.
Let us assume that a particular scientist is a firm believer in
either laissez faire capitalism or Marxism. If he accepts the notion
that facts and values can be dissociated, he could not conclude,
as a scientist, that evolutionary theory supports his political posi¬
tion or that of his opponent. However, extrascientific considera¬
tions could encourage him to assume a stance at odds with his
belief as a scientist that evolutionary theory has no logical im¬
plication for social policy. He could conclude that his metaethical
analysis that facts and ethics are logically independent is too sub¬
tle to be appreciated by laypersons and even some scientists. His
political opponents in the scientific community would surely mis¬
use evolutionary theory to support their political position to the
detriment of his own. His only defense is to retaliate by demon¬
strating that evolutionary theory can be interpreted to support
his own political doctrine. To himself he might admit that his
roles as a scientist and a politically active citizen are in conflict;
but he might also believe that for the good of society, as well as
for the welfare of science, it becomes necessary for scientists to
accommodate to the demands of political action by discarding
their metaethical conclusion.
The difficulty with this rationalization is that scientists with
opposed political convictions can argue in the same manner to

support their thesis. As a result the scientific evidence becomes
unnecessary baggage in the political dispute. The scientific evi¬
dence, as shown with evolutionary theory, can be used to support
any political doctrine and social policy. The likely result of this
kind of debate in which scientific evidence is slanted to support
competing political philosophies is that the public will learn to
look with suspicion on the social-policy recommendations of scien¬
tists. Similar consequences will result when politically active scien¬
tists assume the methodological stance that scientific knowledge
is inextricably intertwined with ethical judgments. Thus when
scientists assign political implications to^ scientific evidence, re¬
gardless of their methodological positions concerning the logical
relationship between faets and ethics, the evidence of necessity
will be unconvincing because contradictory interpretations will
be offered simultaneously.
My analysis supports the position that facts and values can
be decoupled while simultaneously noting that scientists differ in
their desire to politicize the relationship between empirical evi¬
dence and social policy. Those who are interested in the depoliti-
eization process are committed to the idea that there are objective
laws of nature independent of political beliefs. Politically oriented
scientists need not necessarily deny this assumption, but they
could insist that scientific evidence must be viewed within a
political framework to guarantee the attainment of proper soeial
goals.
A Methodological Analysis of Problems

Associated with the Investigation of
Racial Differences in Intelligence
This analysis of some of the reactions to sociobiology high¬

lights the major issues associated with the relationships among
facts, values, and politics. But many subtle points remain. To fur¬
ther clarify the general problem and deal with remaining issues,
it will be useful to analyze a perennial and painful problem in
psychology—racial differences in intelligence. In treating this topic
it is necessary to discuss: (1) the meaning of intelligence, (2) the

meaning of race, (3) the ethical justification for investigating
racial diflFerences in behavior, (4) the ethical implications of ob¬
tained racial differences in behavior, and (5) an appropriate
framework for judging research on racial differences.
The Meaning of Intelligence
Two views have dominated in the attempt to specify the

meaning of intelligence. The one that both has historical prece¬
dent and approximates commonsense interpretations is that iri-
telligence, which might be tentatively conceptualized as the power
to understand and reason, is some innate physiological entity that
presides over intellectual functioning. This view dates back to
Aristotle, who assumed that humans had many souls, each with a
different function. The lowest soul, shared by all living things,
supervised vegetative functions. The highest soul, intellect, pos¬
sessed only by humans, controlled reason.
In contrast to the position that intelligence is some innate
entity hidden in a deep recess of the body is the interpretation
that intelligence is an attribute of behavior. According to this
view, the first step in understanding intelligence, its causes as well
as its behavioral consequences, would be to develop some mea¬
sure of it. Once such a measure is fashioned, it can be further re¬
fined and improved.
The second approach has proved to be more practical. Al¬
though it appears reasonable to assume that some neurological
entity, obviously the brain, controls intellectual functioning, it is
not completely clear what particular patterns of neural anatomy
and biochemical functions are involved. In addition, it must be
recognized that whatever neurophysiological unit serves as the
seat of intelligence, it does not operate in a static manner. The
brain itself can be modified by a variety of external inputs rang-
ing from the quality of nourishment to different forms of environ¬
mental stimulation. One must also realize that any search for the
neurophysiological substrata of intelligence would have to be
guided by some behavioral evidence. Any anatomical site or
neural process that did not influence commonly acknowledged

forms of intellectual behavior would automatically be suspect as
the “seat” of intelligence. In sum, the notion that there are “real
neurophysiological substrates of intelligence is reasonable, but the
direct search for such processes when psychologists first became
interested in intelligence testing in all likelihood would have been
unproductive.
Measures of Intelligence. An early attempt to construct a

“mental test” was made by James McKeen Cattell in 1890. He
gave a series of ten tests to undergraduates at Columbia Univer¬
sity. The tests were designed mainly to^ test their sensory and
motor skills: reaction time for sound, dynamometric pressure,
estimation of 10-second interval, and others (Tuddenham, 1962).
The selection of such tests could be defended by assuming that
the more sensitive and responsive a person is the greater would
be his mental ability. Is this assumption justified? The criterion
employed to answer this question holds the key to understanding
the meaning of intelligence in contemporary psychology.
CattelFs test encouraged others to develop similar measures.
Many psychologists of that era were convinced that as a conse¬
quence of Cattell’s effort a significant breakthrough was taking
place in the measurement of mental abilities. However, their opti¬
mism was short-lived. The tests were found wanting when evalu¬
ated by the newly developed Pearson correlation method. These
tests were found not to be correlated with behavior that could be
called “intelligent.” For example, Cattell’s test was uncorrelated
with college grades although the grades in different courses were
correlated among themselves. How could a valid test of mental
ability be unrelated to behavior that is presumed to be an expres¬
sion of intellectual abilities? Failure to offer a satisfactory answer
to this and related questions resulted in a disenchantment with
these tests. By 1905 when Alfred Binet and his collaborator, Theo¬
dore Simon, produced their test of intelligence, interest in the
Cattell-type test was practically dead.
It is important to recognize a fundamental difference be¬
tween the efforts of Cattell and those of Binet and Simon. The
assumption that mental ability was a function of sensitivity and

responsiveness served as the sole guideline for Cattell in construct¬
ing his test. No apparent consideration was given to the problem
of whether scores on such a test would be correlated with any in¬
tellectual form of behavior. Binet and Simon, in contrast, set out
to construct a test that would be correlated with academic suc¬
cess. They were given the specific assignment by the French Min¬
ister of Public Instruction to identify those children who could
not profit from regular academic training. With this clear goal in
mind and the ingenuity to formulate a dimension of intelligence
in the form of a mental age scale, Binet and Simon were able to
identify students whose academic potential was limited. It was
not a perfect or foolproof test in that errors could be made either
in the assessment of an individuahs academic potential or in the
administration of the test and interpretation of the results. Never¬
theless, it was far superior to any other technique for evaluating
children’s ability to be educated. In addition, later versions and
modifications of the Binet test proved to be extremely useful in
selecting individuals for a variety of industrial, commercial, and
military positions.
It must be recognized that however eflFective one might con¬
sider intelligence tests to be in measuring academic ability, it
does not follow that one can demand such tests be adopted as the
best or true measure of intelligence. As already noted, intelli¬
gence tests measure certain attributes of behavior, and one can
always argue that those attributes selected may be inappropriate
or incomplete. The problem is not unlike that of selecting indices
for defining the economic concept of gross national product, which
is not a problem of discovering the ‘real” index but instead one
of inventing a convenient and useful measure for representing a
nation s productivity. The ultimate justification for a particular in¬
dex is the contribution it makes to the gathering of a significant
body of knowledge. But no matter how systematic and extensive
these empirical laws may be one can always insist that some sig¬
nificant aspect of “productivity” is ignored. For example, one
could insist that a crucial element in the gross national product
is the variety of social services offered, and any index that ignores
that feature will provide a distorted view of the economy. Simi¬

larly, one can argue that because schools tend to encourage con¬
formity, linking intelligence to academic performance will inevi¬
tably generate measures of intelligence that ignore intellectual
creativity.
These objections can be countered by the argument that the
initial operational definition of a concept, such as gross national
product or intelligence, need not be considered so frozen as to
preclude further modification. The concepts can be expanded to
include an additional attribute (e.g., creativity). Or one could in¬
vestigate how that particular attribute (e.g., creativity) is related
to the original concept (e.g., IQ). The history of intelligence test¬
ing illustrates a continued concern with understanding the nature
of intelligence test scores as well as repeated attempts to develop
a “better’" measure of intelligence. Factor analysis, the statistical
technique that breaks down complex behavioral patterns into
basic components, demonstrates that a common factor is involved
in responding to a widely diverse set of intelligence test items
thus supporting the notion that there is such a behavioral char¬
acteristic as general intelligence. More recently, a greater interest
has been expressed in the relationship between traditional tests
of intelligence and theoretical interpretations of intellectual de¬
velopment. But, in the final analysis, many would argue that the
greatest justification for the traditional intelligence test to be
viewed as a “real,” or more properly a reasonable, test of intelli¬
gence is the substantial correlations intelligence test scores have
with criteria of educational and occupational achievement.
As already suggested, one can reject, for a variety of reasons,
the notion that the concept of intelligence can be equated with
scores obtained on traditional intelligence tests. One of many
arguments is that of Medawar, an English Nobel Laureate in the
field of medicine, who insists that it is impossible to attach a
single-number valuation to intelligence because of its intrinsic
complexity. Among the elements of intelligence are,
speed and span of grasp (of understanding), the ability to see

implications and conversely to discern non sequiturs and other
fallacies, the ability to discern analogies and formal parallels be¬

tween outwardly dissimilar phenomena or thought struetures, and
much else besides. One number will not do for all of these . . .
(Medawar, 1977, p. 13).
To support his argument Medawar notes the confusions gen¬

erated by numerous attempts to attach single-number valuations
to complex concepts in other sciences. One such example involved
the quality of soil in terms of its agricultural worth. A single¬
number valuation could not incorporate the variety of attributes
that played a significant role: porosity, particle size and shape,
water content and hygroscopy, hydrogen-iron concentration, and
material flora.
Medawar s argument may appear telling but it is really not
very compelling. Psychologists are not as simple-minded as Meda¬
war, and many other biological and physical scientists, would like
to believe they are. Most psychologists, since Binet, share the
view that a single number fails to convey a complete description
of a person’s intelligence. Factor analysis has revealed specific
factors (e.g., verbal, spatial, numerical, memory) in addition to a
general factor. The Wechsler Adult Intelligence Scale (Matarazzo,
1972) distinguished between verbal and performance intelligence,
as well as components within each. Although a multinumber valu¬
ation of intelligence possesses certain advantages over single¬
number valuation, it does not follow that the latter is useless. As
already noted, single-number valuations of general intelligence
are useful predictors of academic and vocational achievement.
And it should be recognized that a more sophisticated multinum¬
ber valuation has emerged from the original pragmatic decision
to characterize intelligence with a single number.
Any decision to employ the traditional IQ test as a research
instrument to investigate racial differences in intelligence is not
binding for psychologists in particular and society in general. It is
reasonable, although not necessarily compelling, to argue that the
traditional intelligence test possesses too many limitations (e.g.,
excessive linkage to academic success, lack of theoretical ratio¬
nale) to be used as a measuring instrument in investigating the
socially explosive issue of possible racial diflFerence in intelligence.

The interesting question is whether such a position implies any
entailed decision. In the refusal to accept the traditional intelli¬
gence test as a valid measure and in the absence of offering an
alternative measuring instrument that meets standards of reli¬
ability and validity, is it not incumbent upon such a person to re¬
main mute in regard to the nature of the possible relationship
between race and intelligence? It would seem odd, at least to me,
for any individual who essentially argues that the empirical ques¬
tion cannot be properly invesigated to insist that there are, or are
not, real differences in intelligence among races.
The Meaning of Race
Just as it is possible to reject IQ tests as measures of intelli¬

gence one can also refuse to employ racial distinctions in research
by arguing that the concept of race is too vague to be used in
scientific research. Races are not clearly definable biological en¬
tities, and hence an inevitable distortion will be created by treat¬
ing them as such.
The charge that the concept of race is too ill-defined to serve
any useful research function can be countered. Such a position
perhaps is an expression of a desire for race to mean what com¬
mon sense would like it to mean rather than what the facts
demand it to mean. No satisfactory definition of race can be forth¬
coming if one makes an a priori demand that the division be¬
tween races should be absolutely sharp, i.e., races should be
“pure.” Humans from all parts of the world are interfertile, and
throughout the course of history much interbreeding has occurred
among different subgroups that previously had been reproduc-
tively isolated. Consequently, the concept of race cannot be de¬
fined in terms of characteristics of individuals but must instead
be defined by genetically determined attributes (e.g., morpho¬
logical, serological) attributes of subpopulations. By accepting
such an operational definition we will then be able to assimilate
the fact that members of the same race are not genetically iden¬
tical but can vary markedly among themselves. In addition, and
241
most important, with such a definition the similarities among

members of the same race can be ascertained and the difiFerence
between races can be measured.
The Ethical Justification for Investigating Racial

Differences in Behavior
Is it proper for scientists to investigate possible racial diflFer-

ences in behavior, especially those involving highly valued assets
such as intelligence, when the results could be socially disruptive
and oflFensive to individual groups? Do not scientists have an ethi¬
cal obligation to abstain from such investigations? There is no
answer, certainly not a simple one, to these questions. The ques¬
tions, however, can be clarified by discussing certain issues.
Although many scientists would like to believe that a ‘"true”
democracy would aflFord them complete freedom to investigate
any problem, some restrictions are always imposed. Democratic
societies, as well as professional organizations, have a right, as
well as a duty, to impose limitations on research that threatens
the well-being of its members. It would be difficult to persuade a
majority of the scientific community that their rights are being
threatened if an individual researcher is denied the privilege of
inflicting pain and suflFering on human subjects in order to study
sadomasochistic behavior. Obviously no sharp boundary exists,
socially and professionally, between acceptable research and un¬
acceptable research. This point is testified to by the controversy
that surrounded the ethical justification of the Milgram study
(Milgram, 1963) in which subjects were instructed to administer
shocks to other humans. Although no shocks were actually ad¬
ministered, the subjects thought they were, and many, during
the course of the experiment suflFered intense emotional reac¬
tions. Was such research ethically justified? Some psvcholosfists
thought not. ^
Investigating racial dijBFerences in intelligence between races
or among distinctive social groups can be considered socially de¬
structive. W^hen I was in Israel for an academic year during the
mid-1970s I heard numerous rumors that the government was
discouraging research that was focussed on the differences in in-

telligenee test scores and academic performance of Ashkenazi
Jews on the one hand, and Sephardic and Oriental Jews on the
other. Ashkenazi Jews are European Jews, or their descendents,
who emigrated to Israel mainly from Poland, Russia, and Ger¬
many. Sephardie Jews are descendents of those Jews who were
expelled from Spain in 1492 and dispersed themselves mainly
throughout the Mediterranean basin. Oriental Jews, in turn, are
presumably deseendents of those Jews who stayed in the general
vicinity of the Holy Land after the destruetion of the second
Temple in 70 a.d. Although the dividing line is not always sharp,
espeeially between the Sephardim and Orientals, several different
criteria including surnames, language spoken, and country of ori¬
gin can be employed for classifying Israeli Jews into these cate¬
gories.
Ashkenazis are disproportionately represented in the profes¬
sions, the upper soeioeconomic class, and in higher education.
Could this be a consequence of differences in IQ? When I in¬
formally posed this question to a number of Israeli psychologists
the response was affirmative, but the explanation of the cause of
the difference varied. Some attributed it to environmental fac¬
tors; the eultural heritage of European Jews with their greater
emphasis and achievements in edueation gave them an advantage
over Sephardic and Oriental Jews. Others suggested that it was
genetic; although they were all Jews the migration of the Jews
throughout history had produced distinetive breeding populations.
It is interesting to note that the eontroversy about differenees in
intelligence between different groups of Jews was investigated in
Brooklyn, New York (Gross, 1967) where a community of Ortho¬
dox Ashkenazis and Sephardim lived. A 17-point difference in the
IQ’s of the two groups was found, and a controversy was created
as to whether the differenee in favor of the Ashkenazis was due to
possible environmental or genetic differences (Jensen, 1973).
It seems eminently reasonable for a garrison state like Israel
to avoid supporting research that might prove soeially divisive
and thus endanger its chances of survival. In order to resist the
threatened destruetion by its enemies, it is necessary for Israel to
Psychology and Values: 1 24S
maintain a harmonious relationship among its di£Ferent social

groups, at least to the extent that its members cooperate among
themselves when serving in the armed forces. To obtain evidence
that would encourage feelings of inferiority among the Sephardic
and Oriental Jews might make them less eflFective citizens and
soldiers.
The significant point here is that it does not follow that sci¬
ence, with its need for freedom of inquiry, is necessarily threat¬
ened by the failure of a government to support certain kinds of
research. Governments, as well as the scientific community itself,
can impose restrictions based upon countervailing values such as
human welfare and society’s survival. Admittedly, the scientific
efiPort can be threatened if such terms as human welfare and so¬
ciety's survival are used loosely by a government that seeks to
prevent the gathering and dissemination of information that is
embarrassing to its reputation. It is presumed that in a political
democracy with a free press such political shenanigans will ulti¬
mately be revealed to the government’s disadvantage at the next
election.
More important than the question of government support is
the issue of whether a government has a right to ban certain
kinds of research. Again the existence of countervailing values of
the freedom of scientific inquiry must be recognized. Research
that is potentially dangerous to the health of a community or
damaging to the environment would appear to be legitimate tar¬
gets for governmental restrictions. But in such cases it is not the
question (e.g., the nature of certain chemical reactions) being in¬
vestigated that is rejected but instead the hazardous method by
which it is being studied. If less dangerous procedures were in¬
vented the question could then be pursued.
Cannot it be argued that some problems should be prohibited
from being studied? If a government decides not to support re¬
search on racial differences in intelligence because of potential
social disruptions does not the government have the obligation to
ban such research? If the research is going to be socially disrup¬
tive, does it matter who supports it? Although some might find
this argument attractive the distinction between the government’s
policies of refusing to finanee a research project and legally ban¬

ning it should be reeognized.
A government has an obvious right to order its priorities in
the kinds of researeh it desires to support. But that right, in a
politieal democraey, does not provide them with any consequent
privilege to impose restrietions on the freedom of seientists to pur¬
sue problems that they believe are important. Although many psy-
ehologists in Israel and the United States entertain moral qualms
about investigating raeial differences in intelligenee and/or be¬
lieve it to be a theoretieally insignificant issue, they nevertheless
would objeet strenuously to any governmental restrictions. Such
a governmental decree would be pereeived as a direet threat to
freedom of inquiry.
Many arguments can be offered in opposition to any ban of
research that might demonstrate that raees do indeed differ in
eertain charaeteristies. In general, the position ean be taken that
whatever benefits might acerue in reducing racial tension would
be offset by the damage done to a free soeiety in general and the
scientifie enterprise in particular. On a more analytieal level, the
justification of the government for banning the investigation of
certain problems because of their social implications would be
denied by the premise that faets and values can be dissociated
thus making it impossible for any factual evidence to invalidate
a moral prineiple. No matter what findings were obtained about
raeial differences in intelligenee they eould not invalidate the
moral principle of equal treatment for all before the law. In addi¬
tion, those psychologists who might have moral qualms about en¬
gaging in sueh research could nevertheless be suffieiently tolerant
to entertain the possibility that some who do can be above moral
reproaeh. While black-white differenees in intelligence were some¬
times investigated to justify racial prejudice in the form of un¬
equal soeial treatment (Kamin, 1974), sueh motivation need not
be involved.
It is of interest to note that Arthur Jensen’s coneern with
black-white differences in intelligence emerged as a result of an
initial eonvietion “that the mueh higher ineidence of [intellectual]
retardation among ehildren of low SES [socioeconomie status].
and particularly among minority children, was the fault of IQ
tests and also, possibly, of the schools” (1974, p. 224). The results
he obtained in pursuing this hypothesis encouraged him to reject
it. But even though he ultimately adopted the hypothesis that ge¬
netic factors were responsible for some of the diflFerence between
the IQ scores of blacks and whites, he did not employ such data,
as did other psychologists and politicians, to justify school segre¬
gation or other forms of diflFerential treatment:
It is unjust to allow the mere fact of an individual’s racial or social

background to affect the treatment accorded him. All persons
rightfully must be regarded on the basis of their individual qual¬
ities and merits, and all social, educational, and economic institu¬
tions must have built into them the mechanisms for insuring and
maximizing the treatment of persons according to their individual
behavior (Jensen, 1969, p. 78).
The implications of this discussion of the ethical issues sur¬

rounding the researeh topic of racial differences in intelligenee is
that moral questions eannot be avoided. At the same time a single
moral position cannot be reeommended or expeeted. And it should
be realized that moral questions are involved in all choices be¬
tween different scientifie activities though we are not always con¬
sciously aware of them. An experimental psychologist who chooses
to investigate the different kinds of memory stores rather than
employ his talents to improve educational practices is making a
moral judgment that the former aetivity is more important than is
the seeond. The clinieal psychologist also makes a moral decision
when he employs unproven therapeutic techniques in preference
to diseovering whether they are truly effective.
However, when the psychologist, or other scientist, gets em¬
broiled in a highly controversial issue that many believe possesses
numerous social implications (e.g., racial differences in intelli¬
gence, the employment of a neutron bomb) he rapidly becomes
aware that he is confronted with a moral dilemma. Whatever ac¬
tions he takes he is almost eertain to be accused of moral delin-
queney or moral cowardiee.
What guidelines ean be offered to the scientist who contem-
plates becoming involved in socially controversial questions? An

assiduous commitment to the ideal of scientific integrity, critical
thinking, and a great deal of courage should stand him in good
stead. But it must be recognized that specific actions do not flow
automatically from the adherence to the ideals of scientific integ¬
rity and the ability to think critically. Two scientists, of equal in¬
tegrity and ability might choose entirely different moral, empiri¬
cal, and theoretical positions in regard to a socially controversial
issue. Except for the moral issue, the situation is not unlike that
of the two scientists who disagree about the interpretation of a
common phenomenon. Their theoretical differences encourage
them to investigate different facets of the same general phenome¬
non. The resolution of the controversy, if deductive explanation
(pages 142-151) is sought, and assuming that the competing in¬
terpretations are empirically meaningful, will be achieved only in
a historical context. The moral position adds an additional in¬
gredient, which can and does cloud the empirical and theoretical
issues for some investigators but need not for all.
The Ethical Implications of Racial

Differences in Behavior
In the case of the debate surrounding racial differences in

intelligence there appear to be two main sources of confusion
generated by moral and related political considerations. One has
already been alluded to previously—the conflation of factual state¬
ments with moral judgments. A common misconception is that if
racial differences were found they would indicate the existence
of racial inferiority and ipso facto be incompatible with demo¬
cratic principles. As a result many people, especially psychologists
involved with social issues, harbor the hope that all races are
equal in regard to psychological abilities. Their hope, consciously
or unconsciously, becomes a hypothesis that they defend vigor¬
ously. There is nothing wrong in defending such a hypothesis as
long as it is recognized that it need not be true and that the func¬
tioning of a democratic society does not depend upon it. Admit¬
tedly, if the hypothesis were found to be true, fewer social prob¬
lems would be created than if it were proved to be false.
If one fully appreciated the influences of genetic variations, it
would come as no surprise that races could differ in psychological
abilities. Natural selection, in addition to the effects of mutation
and selective migration, can exert powerful effects on the biologi¬
cal structure of members of different breeding populations that
were equivalent in intelligence at one time; but because of natural
selection, which might encourage a higher reproductive rate of
brighter members of one race, genetic differences could ulti¬
mately develop. One could argue that the wide variations in the
environments of the breeding populations of the world combined
with the markedly different cultural traditions associated with
mate selection and differential reproduction rates of different seg¬
ments of society would make racial differences more likely than
not. But in the final analysis, this is an empirical problem that re¬
quires evidence, not prejudgments.
The possible occurrence of racial differences raises the threat
that the concept of racial superiority will be invoked thus encour-
support for racial segregation, persecution, and
even genocide. These worries cannot be ignored but they are not
necessarily warranted. Racial differences and racial superiority
are not equivalent terms, and the former neither implies the latter
nor does it imply any form of differential social treatment. Con¬
sider cases in which races differ in regard to certain biological
attributes. Oriental infants, for example, are more sensitive to
alcoholic ingestion than are Caucasians (Wolff, 1972). Can it be
argued, therefore, assuming that alcoholism is considered to be a
liability, that Orientals, with a lower rate of alcoholism, are supe¬
rior to Caucasians because their chances of being victimized by
alcoholism are less? The notion of Oriental racial superiority would
be immediately rejected because a reduced susceptibility to alco¬
holism would not be accepted as an index of racial superiority. If
one desired to use the term “superior” the only conclusion one
could draw was that Orientals are superior to Caucasians in re¬
sisting alcoholism, and the evidence cited indicates that this su¬
periority is to some extent an expression of genetic factors.
Consider, for the moment, the assumption that genetic factors
are related to performance in athletic events. A person who is tall,
can leap high, and has good vision has a much better chance of
becoming a successful basketball player than one who is short, a

poor jumper, and near-sighted. Various racial groups differ in
their body size and proportions, bone density, and visual acuity.
Is it reasonable to assume that members of different racial groups
have different probabilities of excelling in certain sports because
of genetic factors? For example, is it beyond belief that blacks
have a greater chance of becoming outstanding basketball players
than Japanese? Blacks tend to be taller and have better vision. In
contrast, the more compact build of Japanese may be responsible
for their superior performance in gymnastics. Simple principles of
physics dictate that a person who has long legs and a short torso
will have more difficulty in developing putstanding skills in gym¬
nastics than one with a more compact build resulting in a lower
center of gravity.
Because racial differences is such a touchy topic, I am forced
to qualify my comments in order to guard against the polite accu¬
sation of being an extreme biological determinist and the insult¬
ing charge of being a racist.^ Social motives and environmental
2. Racist is a pejorative term the meaning of which has expanded in

recent years. Initially it referred to a person who harbored prejudice against
members of certain races (or breeding populations such as Jews) and en¬
couraged the social policy of differential treatment. Two examples of racism
practiced against blacks were school segregation and refusal to permit them
to be employed in major league baseball. Today it is not an uncommon
practice to be labeled a racist if one either investigates racial differences in
behavior or suggests that such differences might exist even while explicitly
affirming that it is nevertheless morally repugnant to practice social dis¬
crimination on the basis of race. Andrew Young, the former United States
Ambassador to the United Nations, added a new dimension of racism. A
racist is a person who cannot empathize with members of another race. Such
a subjective definition obviously makes it difficult to identify a racist. I
would suspect that everybody to some degree, including Andrew Young, is
a racist according to this definition. It is difficult, if not impossible, for many
to understand how it feels to be a member of an ethnic group, hke the
Armenians and Jews, who were victims of genocide. How does it feel to
know that at some other time in another place one would have been slaugh¬
tered simply because of his ethnic origins? I would also suspect that many
ethnic minorities are insensitive to the phenomenological experiences asso¬
ciated with being a WASP (white Anglo-Saxon Protestant). A WASP friend
of mine, raised in an ethnic neighborhood in New York City, bitterly com-
variables are not being ruled out as significant factors influencing

athletic achievements. One can argue that basketball is a recrea¬
tion that is particularly suited to the concrete playgrounds of the
black ghettos. And the large number of black superstars in basket¬
ball has encouraged black youths to excel in this sport. But to
acknowledge the influence of environmental factors in no way
denies the important role genetic factors can play. The “either-or”
orientation inevitably produces distortion and confusion; a com¬
plete account of any behavior must include both genetic and en¬
vironmental variables.
The main point is that there may be racial differences in
behavior and such differences may have a significant genetic
component. Therefore for a given form of behavior it can be
stated that one race outperforms or is superior to another. But
there is no evidence to support the contention that one race is
superior in all forms of behavior. Consequently it is factually
wrong, and if one accepts a democratic ethic, morally repugnant,
to employ the concept of racial superiority. And finally, the ob¬
vious point bears repeating that any known obtained racial dif¬
ference in behavior represents a divergence between the two
distributions. Thus for any given attribute of behavior, racial
membership could not be used by itself as a selective index for
predicting a specific level of performance.
An Appropriate Framework for Judging

Research on Racial Differences
One way of viewing the controversy about racial differences

in intelligence is to consider it as a dispute between two compet¬
ing research programs (page 144)—one hypothesizing a genetic
involvement, the other suggesting a purely environmental inter-
plained that as a child his schoolmates did not understand him. Unfortu¬
nately, the term racist today is widely employed as a weapon to intimidate
political opponents. Because the term possesses so many different meanings,
it is not clear what moral crime, if any, the victim of the accusation
committed.
pretation. This approach would encourage the detachment of

ethical issues from empirical ones. But there is no way to force a
scientist to decouple political considerations from factual ones.
As a result, within the scientific community the continuing debate
about political implications of scientific evidence remains irre¬
solvable simply because difiFerent rules for accepting evidence
and drawing conclusions are employed. But the political impli¬
cations of scientific evidence extend beyond the scientific com¬
munity. In a political democracy the problems will be disposed
of by political institutions and processes, not by the actions of
scientists. The only reasonable expectation a scientist can hope
for is to receive a fair hearing about hi§ scientific interpretation
and political position. From my vantage point I would argue that
the social and moral issues are more effectively handled within a
democratic system when they are distinguished from the scientific
ones. In an attempt to demonstrate this view, consider the prob¬
lem of racial differences in intelligence within the framework of
competing research programs.
As previously noted, to actively deal with this issue one must
accept the notion that the concepts of intelligence and race can
be both meaningful and reliably measured. One can decide to re¬
ject either or both assumptions and thus reject the results of em¬
pirical studies that employ these concepts as presently measured.
It is assumed that this scientifically conservative stance requires
that one remain mute about the true relationship between race
and intelligence. If, however, one accepts these measures, not
necessarily as final ones but instead as reasonable indices of in¬
telligence and race that will likely be improved upon in the fu¬
ture, then certain methodological precautions must be appreciated.
First, a major methodological booby trap must be recog¬
nized. One cannot naively assume that a critical experiment can
be designed, the results of which will lead to the unequivocal con¬
clusion that there is, or there is not, some genetic involvement in
the obtained difference of approximately 15 points in the mean
IQ of whites and blacks (Shuey, 1966). It has been previously
argued (page 145) that even under ideal conditions, such as ma¬
ture theories and precise experimental methods, crucial experi-
ments are not possible. Data from experiments that are presumed
to be critical of a theory can be either incorporated into the the¬
ory by ad hoc assumptions or rejected by ad hoc arguments that
insist the study failed to meet the empirical specifications of the
theory.^
In the case of the competing research programs postulating
a genetic or purely environmental interpretation of black-white
differences in intelligence, the ideal of a critical-type experiment
is far removed from any reasonable approximation because of our
ignorance about several crucial factors: a full understanding of
possible environmental influences in IQ, a measure of an environ¬
ment’s level of intellectual stimulation, a satisfactory theory of
intelligence, knowing precisely the genetic determinants of in¬
telligence, and so on. As a result of a lack of sophistication the in¬
terpretations of any research finding are indecisive in regard to
the fundamental issue that is being debated. If this be true, two
questions immediately come to mind: “How can one make judg¬
ments about the validity of the competing theories?” and “What
is the sense of doing such research if individual studies cannot
yield unambiguous results?”
In regard to the first question, it does seem reasonable to as¬
sume that either the genetic-involvement or environmental-disad¬
vantage theory is valid or, as some might prefer to state, has a
greater verisimilitude. In order to make a decision in regard to
the competing theories, different participants in the controversy
have employed different criteria to support their judgments.
3. Jensen (1975) in his efforts to argue that genetie factors should be

considered as a possible cause of some of the difference in IQ scores be¬
tween whites and blacks points out the difficulty of proving a genetic hnk.
He notes that there is no definitive proof that the mean difference between
the heights of Pygmies and Watusis is genetic in origin. These two groups
five in markedly different environments, consume different diets, and prac¬
tice different social customs. The presumption is that genetic factors are
involved because of the magnitude of the difference in the average height
relative to the standard deviation within each group plus the fact that height
is a physical trait with high heritability. But this presumption is not equiva¬
lent to “proof” because it would be possible to maintain that some unex¬
plored environmental factor is responsible for the obtained difference.
The Box-Score Method. One procedure employed to evaluate

competing formulations is the “box-score” method, named after
the system used in sporting events, like baseball, to determine
which team outperforms the other (e.g., scores more runs) and
therefore is victorious. The box-score method applied to the con¬
troversy surrounding racial difference in IQ would involve listing
all relevant studies and determining which position, the environ¬
mental disadvantage or genetic hypothesis, was favored by more
studies. Even if it were possible to determine reliably whether a
given study supported one of the two competing theories, the box-
score method would have the grave defect of treating all studies
as equally well designed and equally informative. Some studies
throw more light upon a given question than do others and this
difference cannot, and should not, be ignored. Some findings have
been replicated while others have not. A comparison of positive
results of competing research problems may reflect no more than
the relative number of investigators committed to each position or
to the intensity of their motivation. It would be a pathetic com¬
mentary on the science of psychology if truth were determined by
the box-score method.^
The Fatal-Flaw Method. Another procedure, the opposite of

the box-score method, can be described as the “fatal-flaw” orien¬
tation. Instead of considering all studies of equal weight, it rejects
all investigations that fail to control all possible variables. That is,
it demands that only those studies that are truly crucial be con¬
sidered in judging competing formulations. An example of the
fatal-flaw orientation is Leon Kamin’s controversial The Science
and Politics of IQ, which draws the conclusion that “there exists
4. It is a frankly embarrassing commentary on psychology to discuss the
box-score method as a possible procedure for evaluating theoretical disputes.
Fundamentally it appears so antithetical to the rational spirit of science,
which demands an integrative analysis of all data. But the box-score method
requires mention because psychologists do employ it. At papers and lectures
I have heard psychologists support one theoretical position over another be¬
cause the results of a majority of studies favor it. Textbooks frequently report
box scores in tabular form, listing all relevant studies and the theoretical
position each favors. The theory that has the most supportive data wins!
no data which would lead a prudent man to accept the hypothesis

that IQ scores are in any degree heritable” (1974, p. 1). One re¬
viewer argues that Kamin’s conclusion that IQ dijBFerences have no
genetic basis rests on a questionable strategy: “He evidently as¬
sembled a checklist of ‘fatal flaws’ and applied it to each study as
an isolated event” (Scarr-Salapatek, 1976, p. 99). Because no ex¬
perimental design is capable of controlling all possible variables
that might be considered relevant by each of the competing theo¬
ries, then it becomes possible by the fatal-flaw method to reject
any empirical result by arguing it is a consequence of some un¬
controlled variable. If the fatal-flaw method is applied to each
competing research program, the conclusion will be encouraged
that the theoretical issue at stake (Do human breeding popula¬
tions differ in terms of their mean intelligence?) is not presently,
or in the future, resolvable. However, if the fatal-flaw method is
applied asymmetrically it becomes a convenient weapon to dis¬
miss an opponent’s research program.
Kamin argues that IQ tests have served as instruments to op¬
press the poor by attributing their economic deprivation to “fixed
biological causes.” Kamin’s aim in writing his book (1974) was
“not only to contribute to the scientific knowledge, but also to in¬
fluence policy makers, and, perhaps, some scientists who do not
recognize that their science and their policies are not clearly sepa¬
rable.” This latter statement is ambiguous because it fails to dis¬
tinguish between the belief that in regard to the IQ controversy
politics and science have not been kept separate or that they can¬
not be kept separate. In any case, assuming a political stance en¬
courages the use of the fatal-flaw method because it facilitates
rejection of evidence that is perceived as contrary to one’s politi¬
cal commitments.
The Historical Method. Although the box-score and fatal-flaw

methods offer a definitive verdict, the verdict itself is suspect be¬
cause of the limitations of the rational foundations of the meth¬
ods. In contrast, the historical method of evaluating competing
research programs does not offer a definitive rule by which one
research program can be considered superior to another, but it
does oflFer a rational procedure for reaching a judgment about the

relative merits of competing formulations. The historical method
(Lakatos, 1970a, 1970b), as applied to the specific issue of racial
diflFerences in IQ and the larger problem of the genetic and en¬
vironmental components in intelligence test scores, would sug¬
gest that the competing formulations be analyzed with two trends
in mind: progressive and degenerating (page 147). Progressive
implies advancement in knowledge either in the form of predict¬
ing new discoveries, particularly of a novel nature, or in improv¬
ing the deductive precision of the theoretical model. A degenerat¬
ing course, in contrast, is characterized by a failure to yield new
facts coupled with a proliferation of ad hoc assumptions for in¬
terpreting recalcitrant findings.^
The historical method itself is not necessarily committed to a
specific or to a single criterion. What the method does imply is
that whatever criterion is employed should be made explicit.
It should be noted that even though the goals of theoretical pre¬
cision and empirical breadth might point a research program in
different directions the two objectives are not necessarily incom¬
patible. Actually, theoretical precision can encourage the discov¬
ery of novel facts.
It is particularly important when employing the historical
method as the basis for comparing competing research programs
to critically assess proposed explanations of anomalous results.
Although research programs have the potential to have their as¬
sumptions modified, or even to have new ones added, the theorist
who is trying to incorporate recalcitrant data has an obligation to
practice some restraints. Consider the previously mentioned study
(Gross, 1967) that found a 17-point difference between the IQs
of Brooklyn-born boys who were descendents of Ashkenazi and
Sephardic Jews. The family environment of the two groups were
carefully examined to discover the source of this discrepancy.
5. P. Urbach has employed Lakatos’s frame of reference in evaluating

eompeting genetic and environmental interpretations of intelligence. His
artiele published in the British Journal of the Philosophy of Science, 1974,
25, 99-135, 235-59 is entitled “Progress and Degeneration in the IQ
Debate.”
Only one significant factor could be found—parental attitudes to¬

ward making money. Three times as many Sephardic mothers
stated that they wanted their sons to be “wealthy” while twice as
many Ashkenazi mothers said that earnings were “unimportant.”
Havighurst (1970) who accepts the very reasonable notion that
the education of socially disadvantaged children would profit
from the knowledge and the utilization of kinds of rewards and
punishments that operate in minority subcultures, interprets the
diflFerence in the intelligence of the two groups of Jews as the con¬
sequence of di£Fering parental attitudes toward making money.
Although this hypothesis presently cannot be rejected, the ques¬
tion should be entertained as to whether similar environmental
influences have been or can be demonstrated to produce such
large IQ diflFerences. It is absurdly simple to invent an environ¬
mental hypothesis to account for any difference in IQs between
distinct social groups because they inevitably have different ex¬
periences. It should, however, be incumbent upon those who ad¬
vance ad hoc conjectures to ask themselves: what evidence is
available and what study can be done to support my speculation?
Social Policy and Empirical Finding
The epistemological principle under examination here is that

facts and values are logically dissociated. Consequently if genetic
differences were found between different breeding populations or
ethnic groups the findings would have no direct implications for
social policy.
This principle was ignored during the days of enforced seg¬
regation of black and white students in the public schools of many
southern states. White supremacists, including some psychologists,
argued that the obtained difference in mean IQ of whites and
blacks was genetic in origin and therefore justified school segre¬
gation. Those who rebutted their position by arguing that the
difference was environmental in origin fell into a dialectical trap.
By implication their argument placed the issue of school segrega¬
tion on the basis of race into the arena of fact when it properly
belongs to the field of morality. Accepting a democratic ethic
makes it morally repugnant to treat individuals diflFerently solely

on the basis of their raeial or ethnie membership. All individuals
in a demoeratie soeiety, regardless of any physieal differenees, are
entitled to be treated with an equality of eoneern. Those who in
past deeades argued that raeial segregation was unjustified be-
eause the raeial differenees in IQ were due to environmental in-
fluenees would be placed in an embarrassing position of endors¬
ing segregation if it were later discovered that the difference had
a genetic basis.
One could also argue against the views of the racial segre¬
gationist by pointing out a logical flaw in the argument. If a dif¬
ference in IQ justifies segregation, then^racial segregation is un¬
justified because the difference in IQ between the means of the
races is less than the differences within races. Consequently, a
strict adherence to IQ standards for admission into different
schools would automatically destroy racial segregation. This is,
however, a logical, not a moral, refutation of the segregationist’s
position.
Segregating students into special classes or schools need not
be considered morally repugnant. It is one thing to assign a stu¬
dent to a remedial class because he does poorly on an academic
aptitude test; it is quite another matter to force him to go to a
special school because of his racial or ethnic heritage. The former
in no way conflicts with the ethics of a social democracy because
the individual is being treated the same as everybody else with
low academic aptitudes. In addition, the segregation on the basis
of aptitude is implemented for the welfare of the individual con¬
cerned although an unfortunate social stigma may be attached to
his limitations. The important point is that such a child is not be¬
ing forced to bear that stigma unfairly because of ethnic or racial
considerations.®
6. There appears to be a profound psychological difference at least for

the social observer, if not for the victim, between a person being maltreated
because of his racial or religious background as compared to being abused
because of some political transgression or quirk of fate. It seemed to have
been morally more obnoxious for a black to be victimized politically in
South Africa than in Idi Amin’s Uganda. In the case of South Africa the
Psychological Aspects. If any possible difference in IQs be¬

tween races has no logical implication for social policy why does
the issue have such an emotional impact? There are many rea¬
sons, too many to identify and discuss for our purposes. One that
has already been mentioned is that many people, scientists among
them, frequently from both extremes of the political spectrum, do
not recognize or cannot accept the logical separation of facts from
social policy. For them any difference in intelligence among races
or ethnic groups would encourage or justify differential treatment
of members of each group. Thus the question of racial differences
in intelligence is not a factual problem but a political issue and, as
such, generates much emotion.
If we shift from a logical to a psychological framework we
have to recognize that the behavioral attribute of intelligence can¬
not be viewed by many in the relatively detached manner that
physical and or other psychological characteristics are considered,
e.g., eye color, height, musical and clerical aptitudes, and athletic
abilities. It is psychologically easier to have a handicap of 15 in
golf than to have a handicap of 15 IQ points (although one golf
buff I know would gladly exchange IQ points for a reduction in
his golf handicap). For the individual who is a member of a racial
or ethnic group that scores lower on intelligence tests, the emo¬
tional reactions can be intense. The lower mean score carries a
stigma of inferiority, and when this is combined with inferior liv¬
ing standards including such vital factors as housing, medical
care, and educational facilities, an angry resentment, understand¬
ably, becomes a common reaction. As long as one is operating
within a psychological framework, it should be realized that these
reactions of inferiority and animus are socially determined, not a
necessary by-product of racially linked genetic factors. Whether
suffering of the blaek results from something over whieh he has no eontrol
while in the case of Uganda, assuming the victim was not a member of a
persecuted tribe, the punishment, even though unjustified, could in principle
have been avoided if the victim had behaved differently. Perhaps that is
why the genocide of the Armenians and European Jews appears to be more
of a heinous crime than government-inspired murders of political prisoners
even though the consequence for the victim is the same.
there is or is not a racially linked genetic factor in hereditary dif¬

ferences among humans, people with lower intelligence would be
disgruntled if their living conditions were shockingly different
from those with greater aptitude. There is no doubt that social
problems are exacerbated when obvious cues associated with race
and ethnic groups are involved, but the basic point is that social
difficulties produced by racial differences in IQ scores are to some
extent socially determined. In other words, the hierarchy of social
prestige and rewards associated with IQ scores is not necessarily
unmodifiable; it is amenable to change so that the diverse talents
of different groups are recognized, esteemed, and rewarded. For
example, if the discrepancy between the employment rates of
black and white youth were minimized or eliminated, presumed
differences in intelligence would probably not be as explosive an
issue as it is today.
Because of individual differences we are all inferior, but we
need not get upset by it. Einstein was an inferior violinist. Wil¬
liam James wanted to be a painter but was convinced he lacked
talent. Many distinguished professors are professors because they
could not make it as concert pianists or professional athletes or
standup comedians. Consider the popular sport of basketball.
Many successful college players are unsuccessful in pursuing a
professional career. Strictly speaking, most of them are probably
inferior basketball players in comparison to those who succeed,
and no doubt some of those who fail suffer from a sense of in¬
feriority. And it may very well be that these feelings of inade¬
quacy are more prevalent among white than black players. This
speculation is suggested from the fact that 70 percent of basket¬
ball players in the elitist National Basketball Association are
blacks, a proportion that far exceeds the 11 percent that blacks
represent in the general population. Although there is no defini¬
tive proof that the greater probability of blacks to excel at basket¬
ball is genetically determined, the hypothesis nevertheless re¬
mains plausible. Such a hypothesis does not generate as much
emotion as one that assumes a genetic base for racial differences
in intelligence. But it should be recognized that if the observed
racial difference in the proportion of professional-grade basketball
players among blacks and whites is genetically determined, then a

higher proportion of whites may suflFer a sense of inferiority in
regard to their basketball ability.
In sum, feelings of inferiority are an expression of the human
condition. They can in principle be generated by genetic factors
that are involved with behavioral differences among races and
ethnic groups. Their destructive effects upon individuals can be
minimized or eliminated by social engineering that values diver¬
sity and minimizes suffering. The major empirical point here is
that in light of available evidence the formulation of genetic^ hy¬
potheses to account for some behavioral differences among races
or ethnic groups is reasonable.
Some psychologists, and many laypersons, insist that geneti¬
cally rooted differences in behavior are impossible. By assuming
this stance they are accepting a factual statement as valid, e.g.,
all racial and ethnic groups have equal native intelligence. From
this it follows that all racial and ethnic groups (e.g., blacks. Chi¬
canes) should be proportionally represented in the professions
and other positions demanding high intelligence. The absence of
this proportional representation implies that some form of dis¬
crimination is operating. Since this discrimination is justified nei¬
ther by the distribution of talent nor by democratic principles of
fair treatment, some rectification of the unbalanced representation
of minority groups^ in prestigious positions should and must take
7. Interactionist is a more accurate term because genetic hypotheses

that attempt to account for racial differences in behavior typically assume
that environmental factors also can account for some of the difference (Jen¬
sen, 1977).
8. Minority status is not only dependent on the proportion a particular
racial, ethnic, or sexual group represents in the total population, but also
upon whether the group is underrepresented in the professions and political
positions and appointments. Consequently, blacks, Chicanos, and American
Indians are considered minorities in regard to admission policies for college
and professional schools, whereas Jews are not. Orientals who tend to be
overrepresented in medical schools and somewhat underrepresented in law
schools are sometimes considered minorities and at other times not. Females,
who strictly speaking, represent a majority of the population, are treated in
affirmative action programs as members of a “minority.”
place. In addition to this plea for social fairness, one can also
argue that a balanced representation of racial and ethnic groups
“is a basic requirement for a democratie republic. Security of
person and property requires that each community have its share
of lawyers, teaehers, doetors, and professional experts” (Garcia,
1978).
One method of achieving balaneed representation of “each
community” is to employ an admission program into professional
sehools that will produce racial, ethnic, and sexual representation
in different professions approximately proportional to their num¬
bers in the population. Those in favor of sueh a plan frequently
employ such terms as goals and timetables to indicate that some
method that favorably weights minority membership should be
employed to increase the representation of members of that group.
The exaet proportion, as well as the target date to achieve it, is
somewhat elastie, but the purpose is clear that systems that do
not encourage the admission of minority groups are to be changed
to those that do. Some eritics, as well as some defenders, of the
proposal to have balanced representation have employed the
blunter, more preeise term quota to designate the goal of the pro¬
posed admission poliey—an exact proportion of a certain minority
must be admitted.
One procedure used to achieve balanced representation en¬
titles members of the minority group to receive special considera¬
tion for admission by being permitted to meet standards lower
than those for “majority” students. When this is done the problem
of separating questions of facts from those of values becomes very
subtle. For the most part values and facts have been conflated by
both proponents and erities of affirmative action programs. Al¬
though policy questions associated with these programs will ulti¬
mately be answered by judicial, legislative, and administrative
actions, psychologists can play a helpful role in making these po¬
litical decisions. A useful starting point for describing this role
will be a review and analysis of the well-publicized Bakke case.
The Bakke Case. Allen Bakke, a white applicant to the medi-

eal school of the University of California, Davis, was rejected
both in 1973 and 1974 even though he would have been admitted
on the strength of his qualifications if he were black. (Bakke
scored in the 90th percentile in the medical college admission
test while the mean of the admitted minority students was below
the 50th percentile.) Sixteen of 100 places in the first-year class
had been set aside for disadvantaged students. The term disad¬
vantaged was not defined by economic criteria, it was based on
race and ethnic background.
The California Supreme Court ruled that this special admis¬
sion plan was unlawful since Bakke had been discriminated
against because of his race, in violation of the constitutional
guarantee of the Fourteenth Amendment that no state should
‘‘deny to any person within its jurisdiction the equal protection
of the laws.” It should be noted that this ruling was not against
the special admission plan for minorities itself but was instead
against the manner in which it was formulated. If minority status
did not have a racial criterion the special admission policy would
have been legal. If, for example, minority status was defined in
terms of economic level, then the program would have been per¬
ceived by the court as constitutional. However, the university, in
appealing the court’s decision, noted that the entire purpose of
their affirmative action program would be defeated. If an eco¬
nomic criterion was employed less affluent whites would be sub¬
stituted for more affluent whites (thus creating a somewhat dif¬
ferent moral and legal issue). As a result of defining disadvan¬
taged status in economic terms, the University of California ar¬
gued that significant minority representation of blacks in medical
school education would not be achieved.
The United States Supreme Court, during 1978, in a set of
split (5 to 4) and hair-splitting decisions ruled that Bakke was re¬
jected from the medical school at the University of California,
Davis, on unconstitutional grounds and insisted he be admitted
into the next class. The court decided that setting aside a precise
number of places for racial minorities is unacceptable. The Su¬
preme Court of the United States, in these two decisions, agreed
with the ruling of the California Supreme Court. A sharp area of
disagreement occurred in the United States Supreme Court ruling
that race could be considered as one of several factors in the

university’s admission policy, whereas the California Supreme
Court had ruled that race could not be a selective factor. The five
United States Supreme Court Justices did not agree as to how
race should be used as a selective admission factor. Three differ¬
ent goals were mentioned in the various decisions: to redress past
discriminations, to increase the number of minority physicians,
and to attain a diverse student body.
My intent is not to get involved in the legal and political de¬
bate surrounding the rulings in the Bakke case, rather it is to dis¬
entangle ethical and empirical components from arguments about
aflBrmative action programs. One of the central assumptions of
many who defend affirmative action programs is that in order to
redress past discriminations it becomes necessary to do away with
“undemocratic” selection procedures that operate against mem¬
bers of minorities. The argument advanced is that some minority
students, such as blacks, are disadvantaged in that they have not
shared in the past benefits received by majority students. Conse¬
quently, the special admission policy, which allows race to be a
selective factor, compensates for past inequalities and thereby es¬
tablishes a truly fair competition among applicants.
An analogy of a foot race has been used to highlight the is¬
sues in special admission policies. Proponents of affirmative ac¬
tion argue that the use of aptitude tests operates as a handicap
against blacks because the tests do not take into consideration
the socially imposed disadvantages of blacks. Without a special
admission policy the blacks would be starting in the competition
to get into medical and other schools behind majority applicants.
Because aptitude tests do not consider blacks’ disadvantaged sta¬
tus they are biased and essentially undemocratic.
Three Interpretations of a Democratic Ethic as Applied to

Test Bias. In an attempt to analyze the meaning of test bias.
Hunter and Schmidt (1976) concluded that the meaning de¬
pends upon the ethical position one associates with democracy. A
psychological test could be unbiased according to one ethical po-
sition but biased in reference to another. To demonstrate this,

Hunt and Schmidt distinguish among three distinct ethical posi¬
tions that would be supported by some faction as being consistent
with democratic ethics. These three ipositions—unqualified indi¬
vidualism, quotas, qualified individualism—sere useful for our
analysis of the relationship between facts and values as raised
in the Bakke case.
The ethical position, characterized as unqualified individual¬
ism, is the one that many associate with the traditional demo¬
cratic value of judging individuals as individuals and not as mem¬
bers of any groups: racial, ethnic, religious, or sexual. Thus,
admission into professional school or advancement to a better
job should be determined by information that would be corre¬
lated with potential success. The persons selected would always
be those with the highest predicted performance. According to
this ethical position there are two ways in which an institution
like a medical school or business organization could behave im¬
morally. First, it may knowingly use an invalid predictor in pref¬
erence to a valid one—appearance may be employed in prefer¬
ence to a score obtained on a valid aptitude test. Second, it may
knowingly fail to employ all relevant knowledge for predicting
success. For example, it may not use race, sex, or ethnic group
membership when these distinctions have been demonstrated to
be valid predictors of future performance in addition to other
useful measures. Although many tend to resist the notion, in cer¬
tain situations it may be possible to make a better prediction of
future success knowing whether an applicant is black or white,
or male or female. This possibility would have to be determined
by empirical evidence.
The ethical position, described as the quota system, postu¬
lates that society should be divided up into certain well-defined
groups, such as black, white, Chicano, etc., and that each group
has a right to receive its fair share of society’s rewards. There are
many different ways that this system can be implemented both
in terms of the manner in which society is to be sliced into dif¬
ferent segments (race, sex, ethnic background, etc.) as well as
what rewards are to be divided (e.g., governmental positions, ad-
missions into colleges and professional schools, etc.). In terms of

the former problem certain “minorities,” such as blacks, women,
and Chicanos, have been most vocal in demanding their rights,
but it should be recognized that other minorities can and will ar¬
gue that they have an equal claim to the rewards of society. Re¬
ligious, sex, and age groups are but a few of the infinite number
of segments into which society could be divided. These groups
most probably will not be mutually exclusive, which creates prob¬
lems as to how the overlapping memberships (e.g., black and fe¬
male) will be dealt with when the system is implemented. In
terms of the second problem, the division of social rewards among
different groups, numerous alternatives are possible. Should the
quota system be limited to governmental jobs? Elected positions?
Corporations and private universities? Small businesses? Any kind
of quota system would be confronted with innumerable prob¬
lems, not least of which would be the opposition of groups who
favor a quota system but believe that they are not properly rep¬
resented. These problems, which represent serious and difficult
issues, do not in any a priori manner invalidate a quota system as
an ethical guide to social organization. One should recall that
great difficulty was experienced, and much time was needed, in
designing the variety of political procedures used in present-day
democratic societies. If a quota system is deemed appropriate
some acceptable arrangement can probably be developed.
Qualified individualism, in contrast to unqualified individual¬
ism, highlights an important ethical consideration. Should a per¬
son’s race, religion, ethnic background, or sex enter into con¬
sideration for selection for a position even if it improves the
prediction of success? Qualified individualism says “No”; unquali¬
fied individualism says “Yes.” The justification for qualified
individualism is that it is ethically abhorrent to deny a person
his complete individuality even when knowledge of his group af¬
filiations will increase the power to predict his or her behavior.
Let us consider two cases to highlight the distinction be¬
tween unqualified and qualified individualism. Several decades
ago it was common practice for departments of psychology to
discriminate against women applicants to graduate school. In
support of this policy the argument was advanced that women

were poorer risks for obtaining a doetorate and pursuing a pro¬
fessional eareer with vigor beeause it was assumed that they
would abandon their eareers at the first ehance of marriage and
raising a family. If, in faet, women were poorer risks, it still
would be immoral for a supporter of qualified individualism to
seleet a male applieant with slightly less promising aptitude
seores • than a female applieant even though the former would,
eonsidering all faetors, possess a greater probability of eom-
pleting his graduate education and vigorously pursuing a pro¬
fessional eareer. A proponent of unqualified individualism, while
selecting the male applicant, could acknowledge that soeial forces,
ineluding prejudiees, were responsible for making women poorer
risks. He could nevertheless justify his selection by stating his
only interest is in pieking the best prospeet.
Another example that is useful to eonsider is the high sehool
traek eoaeh who must ehoose between a white or blaek sprinter
who are running equally well. The eoaeh must eonsider not only
their present performance but also their future potential because
they have yet to aehieve their best times. Beeause blaeks have a
greater probability of excelling in the sprints, it is eommon prae-
tice among high school coaches to seleet blaeks for sprinters and
encourage white runners to speeialize in longer distanees. This
example again illustrates the basie ethical conflict between advo-
eates of unqualified and qualified individualism, with the former
opting for the blaek and the latter being unable to make any de¬
cision in the absenee of further information.
The three positions deseribed do not exhaust all possible eth-
ieal stands eoncerning the appropriate moral treatment of mem¬
bers of diflFerent soeial groups. However they are suflBeient to
illustrate some of the methodologieal issues assoeiated with ethi¬
cal decisions. First of all, it must be reeognized that these three
ethieal positions ean be justified in their own right as moral
imperatives, or they can be defended in terms of some higher
moral commitment. In the ease of the former eaeh of the three
positions can be adopted because of its intrinsie ethical righ¬
teousness that provides a moral authority of its own. Or each can
be defended as a moral derivative of a specific abstract ethical

principle. For example, a person who believes that the guiding
moral principle in social organization should be the achievement
of the greatest good for the greatest number of citizens could
reach the logical conclusion that selecting individuals in the most
effective manner possible would be consistent with his fundamen¬
tal conviction. A person who adopted the quota system might ar¬
rive at his position by a commitment to the ethical concept of
social fairness: members of a socially identifiable group should
not perceive themselves as deprived or inferior. The position of
qualified individualism, it could be argued, is derived logically
from a more abstract democratic commitment to social equality
independent of race, religion, ethnic group, or sex.
If any one of these three positions is accepted as an ethical
imperative, it becomes impervious to any rational analysis that
could yield a verdict against it. Hunter and Schmidt (1976), who
bring to bear statistical analyses in their methodological dissec¬
tion of ethical stands associated with test bias, conclude that the
disputes among the different positions cannot be “objectively re¬
solved.” To support their conclusion they admit that they were
unable to reach agreement about what ethical position should be
adopted: “Each person must choose as he sees fit.”
If, however, the three orientations are perceived neither as
ethical imperatives nor as logical implications from competing
moral principles, but instead as positions derivable from an
agreed-upon higher value (such as an acceptance of a democratic
ethic), then objective procedures might be useful in deciding
which position should be adopted. The first question that should
be posed is whether only one of the three positions (unqualified
individualism, quotas, or qualified individualism) is logically de¬
rivable from a commonly accepted higher moral principle. For
example, according to the principle that fairness or justice is the
ultimate moral value, and assuming these concepts are clear, it
might be demonstrated logically that one of the three ethical po¬
sitions discussed is fairer or more just than the others. Therefore,
a controversy among three individuals, each favoring a different
ethical stand, could be resolved by their adopting a higher moral
value that is demonstrated to be logically consistent with one of

the three competing ethical stands.
Another possibility for rationally resolving an ethical conflict
would be to answer an empirical question that is embedded in
the controversy. For example, two individuals might both favor
the general ethic that society should be organized around the
utilitarian principle of the greatest good for the greatest number
of people, but nevertheless disagree as to whether unqualified in¬
dividualism or the quota system best meets these specifications.
For them there is no logical resolution of their disagreement—
only an empirical one. The proponent of unqualified individual¬
ism would argue that his position will produce the most talented
person for each position, and therefore society will be most effec¬
tively served. In the case of medical service the patient will re¬
ceive the best possible care because the physicians will be the
most competent that society can provide. The advocate of quotas
might question these assumptions by denying that the medical
competence of the practicing physician will be different if quotas
are employed. He might even advance the argument that medical
service would actually be improved by instituting a quota system.
Psychotherapy for some minority members may be more effective
when carried out by therapists of the same ethnic group. Conse¬
quently, if members of minority groups are underrepresented in
psychiatry and clinical psychology, then the effectiveness of such
treatment for minority members would be less than if quotas
were employed.
If evidence were obtained on the quality of medical practi¬
tioners admitted on the basis of a quota system for minorities,
and the effectiveness of psychotherapy when the therapist shares
minority status with his patient was demonstrated, then it could
be decided which principle, unqualified individualism or quotas,
yields the greatest utility. By agreeing upon the basic moral com¬
mitment, empirical evidence can demonstrate which subsidiary
moral beliefs produce consequences that are in line with the val¬
ues of the ethical imperative. In the present example, which only
touches upon a few of many complex empirical issues, the results
need not favor only one of the two competing ethical positions.
For example, if it were discovered that students who are admitted

on the basis of minority status with lower aptitude scores become
inferior surgeons but better psychotherapists, then the advocates
of unqualified individualism and quotas, who share the common
moral imperative of utilitarianism, would be forced to adjust their
position in light of the evidence. Both would favor a quota system
for potential psychiatrists but not for potential surgeons.
Psychology and
Values: Psychological
Considerations
1. The claim that a moral basis for behavior can be derived from em¬
pirical evidence is examined. The argument is made that when psy¬
chologists attempt to establish an ethical imperative on the basis of
psychological data they are essentially offering themselves, or the dis¬
cipline of psychology, as a moral authority. The conclusion is drawn
that psychology is unable to offer an unqualified authority upon which
to base an ethical system to govern human behavior.
2. Practical considerations demand that society develop an effective

moral code. The major point here is that ethical beliefs can be
founded on, but not determined by, psychology. Psychology can pro¬
vide information about the empirical consequences of a social policy
but cannot serve as an authority for its adoption or rejection. For that,
political processes are needed.
3. Psychology can play a useful social role in evaluating the empirical

consequences of different social policies thereby helping political insti¬
tutions to make appropriate decisions. To do this effectively psycholo¬
gists must function as detached scientists and not involved advocates.
269
HUMAN BEHAVIOR AS A SOURCE

OF MORAL ABSOLUTES
By accepting the ethical principle that the status of truth is

denied to moral absolutes, one can easily fall victim to the be¬
lief that all moral choices are ultimately arbitrary, that is, capri-
eious or without reason. Such a conclusion is at odds with our
contention that factual evidence is relevant to choices among
eompeting social programs (e.g., negative income tax versus di¬
rect welfare payments) that share common ethical goals (e.g.,
the elimination of suffering and the encouragement of economic
independence). Admittedly, the kind of research that would be
required to evaluate the empirical consequences of a negative in¬
come tax versus direet welfare payments would be exceedingly
eomplex, and no doubt debate would rage about the correct in¬
terpretation of any results. In addition, opinions would differ as
to whether each program was fairly represented by the experi¬
mental operations employed as well as to the appropriateness of
the criteria used to measure "‘suffering” and “economic indepen¬
dence.” However difficult such research may be, and however
many empirieal issues may remain unresolved, the important
point is that research can illuminate some of the empirieal con-
sequenees of eaeh program and thereby provide information con¬
cerning which program comes eloser to desired ethical outcomes.
In essenee then, relevant facts can determine moral choices when
subsidiary moral principles are involved (e.g., “poor people should
be given minimal financial assistance while being motivated to
improve their plight” versus “the basic economic needs of all per¬
sons should be satisfied”) and when agreement prevails about
moral absolutes (e.g., the elimination of suffering and the en¬
couragement of economic independence).
The conclusion that factual evidence is relevant to certain
moral choices might eneourage one to reconsider the premise that
moral absolutes cannot be deduced from or confirmed by empiri¬
cal evidence. Is it not possible that some set of facts could justify
271
a standard of morality by which ethical imperatives could be
JudgedP The argument can be advanced that an empirically
based universal standard of morality has not been found because
we have been misguided in our search. We have sought a univer¬
sal standard of morality in some divine authority or governmen-
tal decree whereas it actually resides within human behavior it¬
self. By examining human behavior it should be possible to
extract an empirically based standard of morality. Consider this
possibility in relation to the problem of moral development.
Without becoming committed to any particular interpreta¬
tion of moral development (e.g., Kohlberg, 1973) or to any as¬
sumption concerning the relationship between moral thought and
moral action, it appears reasonable, in light of available evidence,
to conclude that developmental differences can be used to evalu¬
ate moral beliefs and behavior. It should be possible to construct
a test of morality similar in design to that of the intelligence test.
An individual’s moral quotient could be determined in a fashion
analagous to his intelligence quotient. In addition to serving as
a measure of a person’s intelligence, an intelligence test can also
serve as a measure of a problem’s difficulty. A higher mental age
IS required to solve the problem “In what way are a tree and a fly
alike?’’ than the problem “In what way are a pear and a peach
alike? Therefore the first problem is more difficult than the sec¬
ond. In the same manner ethical principles can be evaluated by
the level of development that is required for their adoption. In
essence, the suggestion is being offered that the age standard
method that Binet applied to the measurement of intelligence can
also be employed to measure morality. This suggestion raises the
question of whether the moral age required to comprehend an
ethical imperative cannot provide the empirical basis for favoring
one ethical imperative over another. For example, the moral im¬
perative of “Do unto others as you would have them do unto
you demands a higher level of moral development than does the
moral commitment of “Obey the law.’’ Does it not follow that the
former principle is superior to the latter? If it is agreed that the
first is better than the second, we would have an empirical base
to justify an ethical imperative.
In order to pinpoint the crucial epistemological issue of

whether a factual basis can be discovered for ethical imperatives,
I will assume, for the moment, that it would be possible to de¬
velop a scale of moral development that would transcend cul¬
tural differences. It would therefore be possible to evaluate dif¬
ferent societies, as well as religions, in terms of the level of moral
development they have achieved and to which they aspire. Ac¬
cording to one line of reasoning then I have in my possession
psychological facts that demonstrate that some moral imperatives
are superior to others.
Although it is true that a scale of moral development could
be employed to evaluate ethical imperatives in terms of their
‘maturity,” it does not necessarily follow that the more mature
principle is better than the less mature one. For example, in eval¬
uating intellectual functioning it is one thing to conclude that for¬
mulating a theory of relativity requires more intelligence than
tuning a piano or repairing a defective plumbing system, but is
quite another thing to say that the former is better for society
than are the latter. Many people believe that highly developed
intellectual functioning has been a curse on society, producing
more harm than good. Humankind would be happier, they feel,
if high level brainpower had less influence in our society, e.g.,
the ordeal of Vietnam may have been avoided if academics
had not been so influential in government. In sum, more intellect
does not mean better intellect.
A similar argument can be advanced against the position that
the level of moral development is a valid indicator of moral su¬
premacy. No matter how mature one must be to behave accord¬
ing to the golden rule, “Do unto others as you would have them
do unto you” it does not follow that such an ethical principle is
intrinsically better than the less mature moral exhortation, “An
eye for an eye, and a tooth for a tooth” Good and bad, or better
and worse, are rock-bottom decisions; no amount of facts can dic¬
tate that the former moral principle is intrinsically better than the
latter.
The rejection of moral development as the ultimate litmus
test for moral imperatives does not mean that a standard of moral
27S
development cannot be used in analyzing ethical controversies.

e conflict between the Western democracies and Nazi Ger-
many involved fundamental moral differences that made the con¬
flagration of World War II unavoidable. One could argue, prop¬
erly I believe, that the collision was not simply an expression of
a cultural relativism that produced different moral codes, which
m the lina analysis were simply an expression of arbitrary moral
choices. The Nazi’s morality can be described in psychological
terms as pathological and immature, a description that can iustifv
Its rejection. But such a rejection is not based upon a compelling
logic that demands its rejection in the same sense that one rejects
the conclusion that two plus two are five.
Employing moral development as a frame of reference for
evaluating ethical imperatives is similar to Maslow’s efforts to dis-
cover ‘moral principles common to the entire human species,
which can be scientifically confirmed” (Goble, 1971, p. 91). in¬
stead of using moral development as a criterion for judging ethi¬
cal imperatives, Maslow resorts to his own theory of motivation
which postulates a hierarchy of inborn needs the highest of which
is the need for self-actualizationd
Maslow suggests that by examining self-actualizing people
it becomes possible to discover a scientifically valid ethical sys¬
tem. His rationale for this argument is that self-actualizing indi¬
viduals are the healthiest, having achieved the highest level of
human potential;
You can find the values by whieh mankind must live, and for
whieh man has always sought, by digging into the best people in
^ believe . . . that I ean find ultimate values which are
right for mankind by observing the best of mankind. ... If
under the best eonditions and in the best speeimens I simply
stand aside and describe in a seientifie way what these human
vaês are, I find values that are the old values of truth, goodness,
and beauty and some additional ones as well-for instance, eaietv
justiee, and joy (1961, pp. 5-6). >
205-209 for a discussion of self-actualization in the con¬

text of intuitive knowing as a form of understanding.
The link between the empirieal evidenee and the ‘ right set
of values is not as direet as some would like to believe. In faet,
it is suggested that Maslow’s argument represents a simple tau¬
tology. Maslow seleets self-aetualizers who share his own per¬
sonal value system. He labels them as the “best,” and therefore
their ethieal commitments—his own, basically—become the ulti¬
mate values which are right for mankind.”
As noted previously (page 207), the concept of self-actual¬
ization is impregnated with value judgments. According to Plato,
people would realize their potentialities by conforming to the
ideal model of the human soul. Thomas^ Aquinas suggested that
human fulfillment could be secured by the practice of virtue and
by allegiance to the church and its sacraments. Johann Fichte
(1762-1814), a German idealist, provides an interesting and in¬
structive example of the potentials of a concept such as self-
actualization. While retaining the concept of self-realization
Fichte radically changed its meaning. Initially self-realization was
to be achieved by the acceptance of an austere moral code, then
by accepting the will of God, and finally by identifying with the
will of German nationalism, the latter formulation anticipating
the extreme national chauvinism of Nazism. And today, the Red
Brigade justifies its ruthless terrorism (e.g., “Kill a cop and go
home for dinner”) by appealing to the need for self-fulfillment.
Reforming present states, they argue, either in the direction of
liberal capitalism or democratic socialism, will not be sufficient
for developing a state dedicated to personal self-realization. Pres¬
ent states must be abolished, destroyed before that goal can be
achieved (Sheehan, 1979).
This analysis suggests that Maslow deluded himself into be¬
lieving that his proposed ethical system was demanded by psy¬
chological facts in contrast to other systems that depend on some
outside authority such as the church or the state. In truth, he was
only substituting the authority of the psychologist—in this case,
himself—for that of God or government. The difficulty with Mas-
low’s position, it must be underlined, is not with the values Mas¬
low chose to adopt, because psychologists, like other people, are
entitled to choose values they consider best. The fault is with the
manner in which Maslow chose to justify his value system. By in¬

sisting that his value system is demanded by psychological facts
and is a consequence of a purely scientific analysis, Maslow mis¬
led himself as well as his audience.
What should be realized is that when psychologists attempt
to support a moral position by empirical evidence, they really are
offering themselves, or the discipline of psychology, as a moral
authority. If one is inclined to believe in God, it becomes easy to
accept God as a source of moral authority. There is no confusion
as to whether God has the ‘right” to speak with authority. Such
is not the case for the psychologist, no matter how eminently
clever he may be in persuading the public that his proposed set
of ethical imperatives emerge from psychological evidence, the
truth of the matter is that his ethics express his own personal
qualified he may feel to play the same role. Regardless of how
preference. Psychologists and other behavior scientists will never
agree as to what constitutes a valid set of ethical imperatives
simply because of the logical impossibility of crossing the gap be¬
tween facts and values.
If this analysis is correct, it becomes incumbent upon behav¬
ior scientists to speak to the public about moral issues with great
care and a strong sense of responsibility. Many years ago Kurt
Lewin advised Jewish parents to provide their children with cul¬
tural and religious training so that they could effectively adjust
to their Jewishness and cope with any expression of anti-Semitism
that they might encounter. For many, this seemed to be a reason¬
able recommendation especially when empirical evidence was of¬
fered to support it. But a closer analysis suggests that Lewin
should have been more precise in his recommendation, particu¬
larly because he was expressing his own personal values. Even
accepting the evidence, which some would not, that religiously
trained Jewish children are better adjusted than those who are
not, one should make a distinction between the psychological
recommendation that “Jewish children ought to receive religious
training” and “Jewish children should receive religious training
if personal adjustment is considered a more important value than
one’s moral view of organized religion.” The significant point is
that psychologists must be careful not to use the reputation of

their profession for the purpose of advocating their own personal
values. It would also be helpful if the public could feel confident
that when psychologists discuss ethical problems they would
abide by rigorous standards of accuracy in reporting relevant em¬
pirical findings. The public has, with reason, become skeptical
about claims of psychologists who argue for the adoption of par-
tieular value systems. And the more psychologists persist in these
attempts the more cynical society will become about the entire
enterprise of psychology.
PRACTICAL AND CRITICAL ETHICS
The preceding discussion of psychology and values can be faulted

for not coming to grips with the most important problem of eth¬
ics: providing guidance to help individuals act properly. Without
a moral code to steer people in the right direction, society will in¬
evitably collapse under the pressures of the conflicting needs, de¬
sires, and passions of its members. One can also suggest that the
absence of an ethical commitment is destructive to the individual
himself, regardless of the demands of society. Some psychologists
(e.g., Frank, 1972; Maslow, 1984) have argued that valueless¬
ness—the lack of a sense of what is right and wrong, what is de¬
sirable and undesirable-represents the major psyehological dis¬
turbance of our time. At one extreme, it generates a social malaise
that alienates individuals from their society while at the other ex¬
treme it creates an irrational violence that seeks to destroy the
entire society. Without purpose individuals become demoralized
and society drifts aimlessly, incapable of coping effectively with
its current problems while being unable to plan for its future.
My concern here is not with the empirical justification of
the above generalizations but rather to clarify the distinction be¬
tween practical and critical ethics. Great moral philosophers from
Aristotle on considered the problems of ethics to be of a practical
concern because they were concerned with distinguishing right
from wrong. Contemporary moral philosophers, particularly Brit-
ish and North American, have assumed a detached critical stance

of concern not with identifying right from wrong but instead with
analyzing the nature of moral arguments. This critical approach
has led to the conclusion that science in general, and psychology
in particular, can offer no unqualified authority upon which to
base an ethical system. If this critical conclusion is combined
with the reasonable acknowledgment that ethical commitments
profoundly infiuence individual behavior and the functioning of
an entire society, then psychologists are caught on the horns of a
disturbing dilemma; ethical commitments are psychologically im¬
portant, but moral truth is alien to the science of psychology.
In the face of such a dilemma what can the psychologist do?
One possibility is that he can encourage individuals to adopt
ethical imperatives on the basis of faith in sources outside of psy¬
chology. Such an action, it must be recognized, is consistent with
the critical conclusion but is certainly not demanded by it. Be¬
cause psychology cannot offer valid ethical imperatives and yet
moral guidance is needed, it does not logically follow that psy¬
chologists should encourage individuals to seek values through
religious faith or political fervor. Another possibility, on prag¬
matic grounds, is to renounce the critical conclusion that psychol¬
ogy cannot recommend the ethical imperatives that humanity
needs; any critical conclusion, no matter how impeccable the
logic may be, that counters the fundamental needs of humans
and their society must be rejected. An additional alternative is to
finesse the disturbing dilemma by failing to recognize it. By un¬
critically insisting that psychology offers special insights into
moral truth, one can serve the public interest while simultane¬
ously enhancing one’s own sense of social responsibility and self-
righteousness.
Those who demand intellectual rigorousness need not aban¬
don the field of ethics to those who do not. But they do have to
contemplate the disturbing thought that their methodological
stance creates a moral vacuum that could be employed to jus¬
tify even nihilism. Since no moral imperative is valid all moral
positions can be denied. Once this conclusion is accepted an
opening wedge is created, which might lead to the stark conclu-
sion that prevailing political institutions are unjustified and there¬

fore should be destroyed. Unfortunately, one need only point to
the spread of violence and destructiveness of the past decades,
and the nihilistic rationalizations of it, to justify one’s worst fears
about possible side-effects that are created when science fails to
offer any ethical code to govern human behavior.
How can this dreary consequence of the denial of the validity
of moral principles be coped with? As noted, one could agree
with the epistemological conclusions but for pragmatic reasons
decide not to abide by it. The world does not offer any ultimate
morality but it needs one, and therefore it must be created by
some act of faith in religion, political philosophy, or psychology.
A more realistic choice, and intellectually more justified, is to
recognize that moral imperatives are not to be found in nature;
our demands for the ultimate values of life must go unheeded.
For some, the ethical plight of humanity demonstrates that life is
absurd; for others, it highlights the animalistic origins of homo
sapiens. In either case, there is no escape from this moral pre¬
dicament; we can only hope to adjust to it. If one takes seriously
the consequences of this critical analysis, then self-deception is
not a viable alternative; a rational solution remains the only
possibility.
The Role of Empirical Evidence in Moral Decisions
If there is no authority for moral values, and the moral values in¬
fluence the behavior of individuals and the functioning of society,
then humans must create their own values if they wish to control
their own destiny. It is not that the facts of psychology detncind
that individuals adopt a set of values and certainly not that they
adopt a specific ethical system. But it should be recognized that
individual and social behavior are different when moral codes are
operating than when they are not. For example, psychopathic
personalities without a concept of justice, cheat, deceive, and in
general exploit their fellow humans. In sharp contrast are individ¬
uals whose behavior is guided by moral principles. Similarly, the
279
social eflFects of economic poverty depend upon the ethical values

of its victims. The destructive consequences are less when the
commitments to moral standards are strong (Lewis, 1966).
The empirical consequences of moral beliefs do not provide
any simple objective index of their acceptability. But such infor¬
mation can be important to individuals for approving or rejeeting
them. In some cases, the empirieal consequenees of certain moral
principles eneourages their universal adoption. Toilet training,
although not generally recognized, poses a moral issue. Ought
parents toilet train their ehildren? Should not the parent have the
freedom to decide whether a child should control his biological
functioning? Should not the child have the final say? These issues
are not debated simply because rejecting toilet training as a moral
responsibility produces consequenees that are aversive to society,
parents, and children themselves. As a consequence, toilet train¬
ing is acceptable not because it is morally right or valid but in¬
stead because it produces a state of affairs, both individually and
soeially, that is more desirable than the condition that would re¬
sult from failing to employ it.
In contrast, capital punishment, even assuming the available
evidence about empirical consequences were elear, inevitably
generates mueh debate and disagreement about its moral justifi-
eation. There is little likelihood that uncontroversial evidence
portraying its empirical consequences would generate universal
agreement about its desirability as a social policy. If it were dem¬
onstrated that capital punishment deterred crimes such as mur-
der (e.g., among ‘lifers” in prison or burglars), some would
nevertheless oppose it because they argue that it erodes human
compassion. Essentially, the argument would be that a lack of
compassion toward others exerts a more deleterious effect on so¬
ciety than the total number of murders committed by “lifers” and
burglars. Evidenee, of course, could be obtained on these pre¬
sumed deleterious side-effeets, and sueh data may prove effective
in changing attitudes toward capital punishment. But because
there is no way of achieving a universally agreed upon trade-off
function between the number of murders prevented and the dele¬
terious side-effeets of capital punishment, it would be naive to ex-
280 Psychology: A Science in Conftict
pect anything approaching a unanimous decision as to whether

eapital punishment should be adopted as a soeial poliey. And it
should also be recognized that new eonditions eould be ereated
that would encourage many to ehange their positions. For exam¬
ple, many who had eonsidered eapital punishment to be morally
repugnant ehanged their minds as a result of the rash of hijaek-
ings designed to trade hostages for imprisoned hijaekers. Their
reasoning is that capital punishment for terrorists would be effee-
tive in redueing hijaeking and the concomitant woundings and
killings of innocent hostages beeause hijaeking for the purpose of
freeing imprisoned terrorists would be unwarranted.
The major point here is that ethieaf beliefs ean be founded
on but not determined by psyehology. Psyehologists ean provide
information about the empirieal eonsequenees of a given soeial
policy (e.g., capital punishment) but eannot serve as an author¬
ity for its adoption or rejeetion. In other words, the information
psyehology provides ean help a person deeide what poliey to
support but eannot demand whieh poliey to support. The same
empirieal outeome eould eneourage one person to favor eapital
punishment and another person to oppose it. This analysis has
two important implieations.
First, in the absence of a supreme moral authority to justify
ethieal imperatives, difference of opinion will surely reign about
what is right and wrong, and/or good and bad. If freedom of dis-
eussion prevails many general ethieal systems and speeifie moral
positions will be advanced. The former will include philosophical
positions enunciating different ethieal ideals (e.g., goodness, hap¬
piness, pleasure, utility, justiee, individual self-realization, fair¬
ness, elimination of individual suffering) while the latter will
involve speeifie moral stands about soeial polieies (e.g., gov¬
ernmental support of abortion, affirmative aetion, building of a
neutron bomb). This diversity of moral eommitments appears in¬
evitable as long as individuals are free to ehoose, and probably is
inevitable even if they are not. In addition, it should be noted
again that an individual’s or an entire society’s eommitment to
one system or position should not be viewed as permanent be¬
eause new evidence may encourage a ehange of attitude.
Second, a rational resolution of basic difiFerences in ethical

principles and moral positions becomes possible. To achieve this
goal one has to abandon the ideal of discovering an unambiguous
answer to a moral conflict that will be accepted by all. If one
limits the meaning of rational to the development of reasonable
procedures that can be employed to make choices between com¬
peting social policies that express fundamental moral principles,
then rational resolutions of moral conflicts are possible.
The entire American political system, including the execu¬
tive, legislative and judicial branches, provides rational proce¬
dures to resolve ethical disputes. Sometimes arguments arise as
to whether the rational procedures are being followed as is the
case when one branch of government is accused of assuming the
powers of another. But for the most part, the entire system has
usually been able to solve its internal procedural problems and
has provided a means of resolving ethical disputes such as that of
segregation. The one notable exception is the moral issue of slav¬
ery, which required the Civil War for its resolution.
Some would argue (Roche, 1968) that the issue of slavery
had to be resolved by war because it was solely a moral issue and
therefore not strictly a juridical problem. Such a position is diffi¬
cult to understand because a juridical issue was involved in the
defining of the black slave as a ‘"person” or “property.” According
to the Fifth Amendment no person “shall be deprived of life, lib¬
erty, or property without due process of law.” If the black slave
was property then slave-owners could not be deprived of their
wealth. If the black slave was a person then he could not be de¬
prived of his “life” and “liberty” and therefore could not remain
as a slave.
In principle, the question of slavery could have been solved
by political processes. This was prevented because commitments
to moral positions on the slavery question were stronger than
those related to maintaining the integrity of the federal govern¬
ment. In contrast, segregation, a significant moral issue, was re¬
solved without war by the greater willingness of the population
to abide by governmental decrees in 1954 than in 1861.
Political processes then can be considered as mechanisms for
resolving moral controversies although it must be recognized that

the moral position adopted by the society is not considered a
moral imperative either in the sense that it is valid or that indi¬
viduals must believe in it. The only demand is that individuals
abide by it. At the same time one has the right to employ appro¬
priate political procedures to overturn the moral decision that
the political processes have generated. No moral decision is pro¬
tected from change or ultimate rejection. For example, rulings of
the Supreme Court have been rejected by subsequent rulings,
congressional laws, and amendments to the Constitution.
What can the psychologist contribute to the resolution of
moral conflicts in the political arena? My belief is that the con¬
tributions can be both substantial and helpful as long as psy¬
chologists avoid conflict of interests between their ideological
commitments and scientific responsibilities. Obviously, psycholo¬
gists, like everyone else, have political beliefs and ethical commit¬
ments. It would be naive to believe that they do not, and unfair
to insist that they should not. The significant problem is whether
their ideological positions can and should be kept separate from
their scientific activities.
It seems quite clear that in some cases psychological evi¬
dence can be insulated from the influence of ideological commit¬
ments. One example of this is the actions of a psychologist (T. S.
Kendler) who became involved in the legal issue resulting from
the passage of a statute in Hawaii in 1943 that prohibited the
study of foreign languages (other than English) before a child
reached the age of ten or completed the fourth grade. The law
also prohibited the teaching of a foreign language to a child un¬
der fifteen years of age who received a below-average grade in
English composition. This act was based on the legislative finding
that “the study and persistent use of foreign languages by chil¬
dren of average intelligence in their . . . early years definitely
detract from their ability to understand and assimilate their nor¬
mal studies in the English language . . . may and do, in many
cases, cause serious emotional disturbances, conflicts and malad¬
justments” (T. S. Kendler, 1950, p. 505). The legal justification of
the statute was that it sought to improve the education of the
children and protect their welfare.
This law was considered by some to be directed against the

eflForts of ethnic groups, such as the Japanese, to maintain their
cultural heritage by teaching their children the language and tra¬
ditions of their ancestors. The law also threatened some forms of
religious training that depended on knowledge of another lan¬
guage such as Hebrew or Latin. The Commission on Law and
Socal Action of the American Jewish Congress decided to file a
brief in the United States Supreme Court questioning the consti¬
tutionality of the Hawaiian statute and requested assistance from
Tracy Kendler, then in their employ, to obtain information rele¬
vant to the presumed deleterious effects of learning and using a
second language.
A review of the literature revealed several studies that were
inconsistent with the legislative findings of the deleterious effects
of second-language learning. Schiller (1934) found that Jewish
children from homes in which Yiddish was the primary language
performed on English verbal tests as well as did Jewish children
from homes in which English was predominant. Symonds (1924)
found the Chinese students were not harmed by attendance at
foreign language schools. It was also pointed out that a study
(Spoerl, 1943) that reported a superior adjustment of monoglot
college students as compared to bilingual students need not be
interpreted as demonstrating the detrimental effects of bilingual¬
ism; a more reasonable interpretation is that bilingualism and the
greater degree of maladjustment are both consequences of envi¬
ronmental factors associated with second-generation American
status. Thus the conclusion was drawn that the legislative findings
that bilingualism produced maladjustment and educational re¬
tardation were empirically unfounded.
A cynic might very well question whether the above exam¬
ple demonstrates the possibility of separating ideological commit¬
ments from scientific efforts. After all, the psychologist in this
example was merely doing the bidding of the American Jewish
Congress who, because of their own interests, was concerned only
with data that were opposed to the findings of the Hawaiian leg¬
islature. But this example was deliberately selected to make clear
the scientific responsibilities of a psychologist who provides evi¬
dence relevant to social policy. Certainly it is legitimate for an or-
ganization such as an American Jewish Congress to defend its

own social and political interests. In doing so, it possesses a
clear right to hire scientists to provide information that can be
used in its legal, political, and social activities. It is also legitimate
for a scientist to accept a position in an organization that advo¬
cates certain social policies because no necessary conflict of inter¬
est need arise. If the organization insists that the scientist provide
evidence that is not scientifically justified, then an obvious con¬
flict would be created that would make it impossible for a scien¬
tist to maintain her integrity. But in this particular case the
scientist could make a judgment, and hpr opinion could be ex¬
posed to the scrutiny of other scientists, as to whether the
presumed deleterious eflFects of bilingualism were factually jus¬
tified. Admittedly, the issue could be clarified by additional
research that would isolate the important variables and yield
more precise conclusions. But it is not incumbent on an organiza¬
tion, such as the American Jewish Congress, that espouses partic¬
ular social policies to solve complex psychological problems. Nor
is a psychologist working in their employ required to achieve a
complete understanding of the empirical issues before drawing
any scientific conclusion. The fundamental demand is that in
evaluating the empirical issue a psychologist should not com¬
promise her scientific standards. When a psychologist reports
data to a court or a legislative hearing, she has the obligation to
society, as well as a responsibility to her profession, not to offer
testimony of a dubious scientific value.
In contrast one can consider the testimony of social scientists
in the famous Brown v. Board of Education case in which the
Supreme Court ruled that segregating students on the basis of
race was unconstitutional. One bit of evidence offered was a
survey of opinions of anthropologists, sociologists, and psycholo¬
gists (Deutscher & Chein, 1948) as to the psychological conse¬
quences of enforced segregation. A large majority (90 percent)
of those who responded to the poll believed that enforced segre¬
gation had a detrimental effect on black children even in cases
where the educational facilities provided were equal to those pro¬
vided white children. A slightly smaller percentage (83 percent)
believed that school segregation had detrimental effects on white
children.
Several problems are raised by such testimony. One is whether
any evidence, empirical findings or expert testimony, is required.
Some persons would consider that the problem of segregation is
primarily an ethical issue, and therefore psychological data are
not centrally relevant. This position argues that the basic assump¬
tion of a democratic society is that all members of society should
be treated equally. Segregation obviously implies unequal treat¬
ment, which is immoral regardless of any empirical evidence.
Although this position is attractive in its simplicity and ethi¬
cal sensitivity, it is unrealistic. First, segregation, in spite of the
fact that the United States is a democratic ^ society, was condoned
by the Supreme Court with its "separate but equaU doctrine. Sec¬
ond, since the time of Louis Brandeis, who included social and
economic data in his brief to defend the Oregon ten-hour per day
work law for women, social science evidence has influenced judi¬
cial decision (T. S. Kendler, 1950). If the "separate but equal”
doctrine were to be challenged it would appear strategic to offer
data that would contradict its major assumption.
Once the argument shifts from the level of ethical impera¬
tives to that of the empirical consequences of particular laws, one
is forced to accept the verdict of the empirical evidence. If later
evidence contradicts the initial data, it becomes incumbent upon
those who justified the original policy to reverse their policy rec¬
ommendations. To be specific, suppose that at some future time
American social scientists change their political attitudes from
their present "liberal” stance to a more "conservative” or "reac¬
tionary” position, or that the refractoriness of the problems of in¬
terracial tensions encourages them to respond to a new survey in
support of the position that school segregation need not have det¬
rimental psychological or social consequences. Such data would
certainly be rejected as biased by those who initially offered sur¬
vey data to demonstrate the harmful effects of school segregation.
But can it not be argued that the original survey was biased be¬
cause the results could have been influenced by the political and
ethical commitments of social scientists? Could one have any
confidence in the eonelusion that their responses to the questions

of the survey were independent of their politieal desires? Sueh an
aceusation could be rebutted by arguing that it would have been
easy, as I suspeet it would, to obtain empirieal evidence that
school segregation exerted harmful effects, both to the education
of the children and to their personal and social adjustment. But
sueh a rebuttal essentially rebukes those who offer the results of
the survey. Why offer dubious evidence when rigorous scientific
standards could have been met. Is it not the responsibility of the
seientist to offer unbiased evidence? If he cannot he should re¬
frain from testifying. If psychology is taplay a significant role in
evaluating the empirical consequences of different soeial policies,
and thereby assist political institutions in executing their respon¬
sibilities, then its evidenee must be trusted. Unless that confidence
is earned it will not be fortheoming, and the potential value of
psychology to society will have been lost.
Riehard C. Atkinson, a psyehologist who was recently Direc¬
tor of the National Science Foundation, strongly believes that
confidence is laeking in psyehology’s ability to offer unbiased sci¬
entific evidence. In an address to the American Psychological As¬
sociation, he notes that:
there are many reasons behind this country’s current negative atti¬
tude toward the social sciences. But I want to emphasize that we,
as social scientists, have contributed to this state of affairs. We
have done so by not being careful enough in drawing a sharp dis¬
tinction between our role of scientists versus another role—that of
political advocate and policy-maker. The psychologist’s job as a
scientist is to search for data, principles, and laws that enlarge
our understanding of psychological phenomena. But too often, in
reporting research findings, we become advocates for a particular
public policy. There is no reason why psychologists should not
advocate political viewpoints, but they should advocate them only
as individual citizens. The psychologist’s role as a scientist is to
set forth the facts, and to set forth those facts in as value free a
fashion as possible. It is the job of the citizens of this country and
their elected representatives to use those facts in making policy
decisions. Too often I have witnessed psychologists, speaking on
education, child rearing, social institutions, and mental health,
using what they claim is research evidence as a disguise for ad-
vocating a particular public policy. Psychologists and social scien¬
tists, more so than other scientists, need to carefully distinguish
between providing scientific data and making policy. If a psy¬
chologist is fascinated by political power and the ability to shape
public policy, he or she should run for elective office and not dis¬
guise political efforts by cloaking them in the framework of psy¬
chological research. I recognize that it is difficult, if not impos¬
sible, to present scientific findings in a value-free fashion. But
every effort must be made to do so. Otherwise psychology will
come to be regarded as a social force rather than a scientific dis¬
cipline. If that should occur, psychology’s potential for helping to
solve society’s problems will be lost.
Let me add a proviso to these remarks. Obviously psycholo¬
gists must be prepared to offer their best judgment on policy
issues, even when that judgment is based on data and theory of
questionable validity. But in giving such advice, care must be
taken to emphasize the limitations of the scientific evidence and
to explain that there may be other tenable interpretations. Most
disturbing to me is the psychologist who, on the basis of the flim¬
siest data, makes pronouncements as though they were backed by
the full weight of science (1977, pp. 207-8).
Similar concern has been expressed about psyehology’s role

in legal proeeedings: ‘'At present, it is still possible for the soeial
psychologist to ‘hoodwink a judge who is not overwise’ without
intending to do so; but suceesses of this kind are too eostly for
seienee to desire them” (Cahn, 1955, p. 166). Coneern is also ex¬
pressed for the law: “Reeognizing as we do how sagaeious Mr.
Justiee Holmes was to insist that the constitution be not tied to
the wheels of any economic system whatsoever, we ought to keep
it similarly uncommitted in relation to the other soeial scienees”
(ibid., p. 167).
The Conflation of Ethical Decisions

and Empirical Evidence
Two major factors encourage seientists to conflate ethical deci¬

sions and empirical data. One results from either failing to grasp
the distinetion or from refusing to aeknowledge it, the other from
rejecting it. Consider action research as an example that appears

to represent the first kind of confusion.
Action research (Chein, Cook, & Harding, 1948) was en¬
couraged by Kurt Lewin who attempted to combine the goal of
understanding social behavior with the solution of practical prob¬
lems. Examples of action research were attempts, begun several
decades ago, to reduce interracial hostilities in public schools,
housing developments, and recreational centers where blacks and
whites intermingled. Such a goal obviously need not generate any
conflict between scientific responsibilities and social and political
involvements. Reducing interracial friction is both an interesting
theoretical problem and a widely accepted moral goal in a demo¬
cratic society.
An epistemological analysis of doing action research on re¬
ducing interracial hostilities indicates that two separate decisions
are involved: (1) an ethical decision that reducing interracial
hostilities is virtuous and (2) a scientific decision that empirical
methods can be developed that will enable one to understand and
control the phenomenon. In justifying this kind of research Chein,
Cook, and Harding (1948) tend to fuse these two separate issues
so that the reader is encouraged to believe that by doing so a sci¬
entific ideal is achieved that is denied to those who try to keep is¬
sues of ethics separate from those of facts.
Such an approach [action research] to scientific endeavor, one

which is aimed at the discovery of the determining conditions of
events, is obviously ideal for the scientist whose life as a scientist
is integrated with his life as a citizen, who wishes to pursue a
scientific way of life and at the same time to devote his energies
toward civic betterment.
It is with these considerations in mind that one can, perhaps,
best understand the field of action research. It is a field which
developed to satisfy the needs of the socio-political individual who
recognizes that, in science, he can find the most rehable guide to
effective action, and the needs of the scientist who wants his
labors to be of maximal social utility [italics mine] as well as theo¬
retical significance (Chein, Cook, & Harding, 1948, pp. 43—44).
Although these authors aeknowledge the influenee subjeetive

faetors ean exert on the colleetion and interpretation of data from
action research, they do not appear to recognize that the meaning
of maximal social utility can be influenced, even completely de¬
termined, by subjective factors. Perhaps this oversight is due, in
part, to their belief that they comprehend the motivation of the
scientist:
The scientific way of life is governed by three broad classes of

interacting motives: curiosity, the desire to know what is going
on when one s back is turned, where one’s vision cannot easily
reach, or where a situation is too complex for clear viewing;
practicality, the desire that the results of one’s labor, search, and
enquiry should be useful and significant, that they should “^make
a difference”; and intrinsic orderliness, the desire that the masses
of accumulated data be reduced to a comprehensible order and
that the complexities which have been unraveled in the satisfac¬
tion of one s curiosity be not again obscured by the imposition
upon the data of an arbitrary order (ibid., 1948, p. 43).
Unless one insists in an a priori manner that this motivational

pattern defines a “true” seientist it is most doubtful that it repre¬
sents an accurate pieture of all scientists simply because the mo¬
tivations of scientists vary. Many scientists with whom I am
acquainted appear to be motivated solely by sheer curiosity;
practical applications do not even enter their minds. Others seem
more eoncerned with collecting facts, the more the better.
Debating the issue of the motivational patterns of scientists
is not my aim beyond questioning the validity of the above quo¬
tation. The significant point is that by postulating the “true” mo¬
tivational pattern of scientists, action researchers are encouraged
to believe that their practical concerns are purely an expression
of scientific motives and by so doing they gloss over complex
ethical issues. Are they truly functioning as scientists when they
select the criterion of social utility or are they behaving as politi¬
cal advocates? If action research simply represents an attempt to
employ scientific method to discover the most effective programs
to implement a given social policy, then the problems of facts
and ethical judgments are not conflated. If, however, action re¬
search implies a scientific mechanism by which valid ethical
judgments can be identified, then the two distinct issues of facts
and ethical decision have been confused.
There is obviously no way to achieve total agreement in the

scientific community as to the proper method to treat questions
of facts and ethics. There are those who will remain convinced
that psychologists have the right, as well as the duty, to advocate
social policies and ethical imperatives in the name of the science
of psychology. There are others who will insist that a tough-
minded epistemological analysis will fail to yield any scientific
validation procedure that is able to identify ethical imperatives,
and therefore when psychologists do make such claims they are
misleading their audience. And there will be those who will
knowingly or unknowingly ignore the problem. Although one can
not demand that a given viewpoint about the relationship be¬
tween values and facts be accepted by all psychologists, if for no
other reason than the command will go unheeded, one can never¬
theless encourage psychologists, when commenting on public pol¬
icy, to make their own viewpoint very clear. By so doing they
will be fulfilling their responsibilities to psychology and to so¬
ciety. It might be argued that the lay public will not be able to
comprehend such a subtle philosophic issue, but if one accepts
Atkinson's observation that some believe that psychology is being
used falsely as an advocate for public policy, then it must be con¬
cluded that at least some segments of society believe that they do
understand the problem. If one has faith both in democracy and
the power of education, public discussion of the problem of the
relationship between facts and values may produce a level of un¬
derstanding beyond what many would expect. I believe that so¬
ciety will resolve the issues raised in this chapter and hope that
rational procedures will be employed.
PSYCHOLOGY AND SOCIAL PROBLEMS
The contributions of psychology to the solution of social problems

have been commented upon but not treated in any detail. Two
major kinds of contributions are possible: revealing the conse¬
quences of a given social policy and developing effective methods
of moral training.
291
Assessing Social Policy
Assessing the results of a soeial poliey ean be aeeomplished either

by evaluating the results of a given poliey that is presently in
foree (e.g., busing) or by exeeuting a pilot study to estimate the
eflFeets of a program that is being eonsidered for adoption (e.g.,
negative-income tax). In either case, it must be remembered that
the findings of such studies would be irrelevant to those who are
committed to an unqualified and unchangeable ethical view
about the policy under consideration (e.g., achieving school inte¬
gration is an ethical imperative, all forms of welfare are bad).
For those who maintain a pragmatic view, probably a large ma¬
jority for most controversial social programs, the outcome of such
research will influence their support for the social policy that is
being evaluated. To meet the needs of society the final outcomes
of such evaluative studies should probably include a host of be¬
havioral measures reflecting those effects that some believe to be
of prime importance. For example, in assessing the results of
busing information should be obtained about the reduction of in¬
terracial strife, the effectiveness of education for black and white
students, the financial costs, and so forth. It also should be noted
that evaluative research could provide information (e.g., indenti-
fication of personalities and intellectual characteristics of children
that are associated with positive outcomes of busing) that might
suggest modifications of the busing program that would encour¬
age greater public support.
Another example of evaluation research contributing to the
possible amelioration of a social problem is the determination of
the contributions of TV news coverage of violent crimes to sub¬
sequent crimes of a similar sort. The evidence is that contagious
violence does occur (Berkowitz, 1970), but much more needs to
be learned about its underlying mechanisms. As usual, there will
be those who will insist that, whatever its consequences, no at¬
tempt should be made to censor TV news for the simple reason
that freedom of news broadcasting is a more important value than
reducing violent crime. But it is possible that even these reserva¬

tions might dissipate in light of clarifying evidence. If we were
able to fully grasp the true nature of the facts about contagious
violence, TV networks might be willing to police themselves and
refuse to show news episodes that might trigger violent behavior.
The idea that evaluation research could prove useful in solv¬
ing problems of busing and contagious violence would be con¬
sidered by some as the height of optimism, if not outright folly.
Much research has been done on busing and contagious violence,
and we are no nearer a satisfactory resolution of these issues
than we were prior to the research. Such jesearch, the argument
goes, inevitably will produce ambiguous and conflicting results
that will be incomprehensible to those who sincerely seek infor¬
mation relevant to forming an opinion about the issue. Although
there is a factual basis to these accusations, they must be under¬
stood in the context in which evaluative research is conceived and
conducted and the results disseminated. For the most part indi¬
vidual researchers have decided to investigate the problem within
the limits of available resources. The communication of the re¬
sults by the news media to society has frequently been mislead¬
ing; accuracy has been sacrificed for dramatic effects. No serious
effort has been made to educate the populace, as well as mem¬
bers of the government, as to the appropriate methods for inter¬
preting evaluation research. What is obviously needed is a more
systematic method for planning, executing, and disseminating the
results of socially significant evaluation research. There are many
institutional and organizational forms that such a systematic
method could take, but it would be premature to offer any exact
recommendation. Some warnings, however, are in order. If one
accepts the notion that the evaluative research should approach
the ideals of objectivity, then it is recommended that those who
conduct the research not be wedded or opposed to the policy un¬
der investigation. This concern, for example, would rule out the
governmental agency that is responsible for implementing a social
policy from evaluating it. Just as the ombudsman is independent
of the governmental apparatus that he is called upon to investi¬
gate so should researchers investigating the social policies be free
Psychology and Values: 2 29S
of the influence of those who have a stake in the success or fail¬
ure of that policy. Whether this requires that all evaluation re¬
search be done by nongovernmental organizations or whether it
demands that several independent groups should research the
same policy in order to guard against possible sources of bias is
a problem that can be illuminated later.
The potential value of doing pilot research on social pro¬
grams being considered for adoption would probably exceed by
far the worth of evaluations of programs in operation. One of the
inescapable consequences of technological advances and social
changes in this complex modern society are unanticipated side ef¬
fects, usually negative in nature. The automobile, for example,
has led to air pollution, traffic fatalities, highway-scarred country¬
side, traffic jams, and an energy crisis. Would the automobile
have been aborted if its negative side effects had been known at
its conception? Probably not, but its development might have
been planned more wisely. Social planning, based on data from
pilot studies, could have minimized the terrible price society is
currently paying for the combustion-engine automobile.
Social programs designed to eliminate society’s ills have also
produced negative side effects. Welfare programs designed to as¬
sist persons have at times encouraged excessive dependence and
demoralization. Public housing programs have destroyed neigh¬
borhoods they were designed to help. Medicaid programs have
produced scandalous overcharging, unanticipated taxing of physi¬
cians’ services and hospital facilities, and skyrocketing costs.
Must we always remain in the dark about consequences of
an innovative invention or social program? Need unwanted side
effects be discovered only after they occur? Since science is a
method for foretelling the future, can it not be employed to judge
the effectiveness of programs that have yet to be tried? This no¬
tion is presently gaining favor and many psychologists and other
behavioral scientists are actively engaged in evaluating social
programs that are being considered for adoption. One example
has been the attempt to predict the consequences of a negative-
income tax. The basic notion of such a program is that the income
tax system is extended downwards so that the families falling be-
low a poverty level receive cash (negative taxes) to compensate

for their lack of earnings. As the family income rises above the
poverty level, as a result of a member getting a job, the tax pay¬
ments do not necessarily stop. Instead they are reduced so that
the total income will always increase when the family's earned
income is augmented. Many programs at present automatically
remove a family from welfare rolls if a member accepts a posi¬
tion even though the wage received is less than the total welfare
payment. Under such conditions a person on welfare is reinforced
for refusing a position. A negative-income-tax program, in con¬
trast, encourages job acceptance becausê it leads to increased
income.
But is it not possible that the income guaranteed by a nega¬
tive-income-tax program reduces the motivation to find a job? Al¬
though a job will necessarily increase a family s total income per¬
haps the added money will not be sufficient to encourage a
person to seek a job when he has the option of not working? One
attempt to estimate whether a negative-income-tax plan would
destroy the incentive to work was conducted in a number of cities
in the New Jersey-Pennsylvania area (Elesh, Ladinsky, Lefcowitz,
& Spilerman, 1971). Families on a negative-income-tax program
were compared with an equivalent group with no supplementary
income. The families on the program were free to do whatever
they wanted with their payments. The major finding was that the
earned income of the experimental and control groups was the
same thus indicating that the negative-income-tax plan did not
reduce the incentive to work. The only effect noticed was that
the total hours worked by families of the experimental group
were less than those of the control group, apparently because
people in the negative-income-tax plan took a longer time to find
a job to their liking.
An experiment like this does not answer all questions about
the negative-income-tax program. In fact it must be recognized
that one study, or even many, will not be able to answer all ques¬
tions. One must be realistic about the contributions of applied
social experimentation of the sort that has just been described.
Inevitably, the amount of information will be limited. The under¬
lying assumption of applied social experimentation is that a wiser
decision can be made about social programs in the light of evi¬

dence, unavoidably incomplete, than in the absence of any data.
It may appear that a sharp break in the discussion has oc¬
curred—an involvement with social policy has suddenly been sub¬
stituted for a concern with ethical issues. In reality, a logical
transition has occurred between the discussion of practical ethics
and an examination of applied social experimentation with its
goal of evaluating present and future social programs. Essen¬
tially, social policies are an expression of practical ethics. The
adoption of new social programs is supported and encouraged
because their goals represent desired values, i.e., social conse¬
quences that are good or at least better than those that result
from currently operating policies. The role that psychology can
play in practical ethics is to reveal, in an impartial manner, the
consequences of different social programs that are designed to
achieve certain moral ends. However, to repeat an important
point, the results of such research will only be considered rele¬
vant to the social issue if the efforts of the psychologists and
other behavioral scientists have the trust and respect of the so¬
ciety. If their efforts are perceived as a form of advocacy of their
own political and social values, then their findings will be, and
should be, ignored. In the long run a backlash could occur that
would confirm Atkinson’s prediction: “To permit psychology to
be misused as an advocate for public policy will lead inevitably
to the demise of the field” (1977, p. 210).
For a variety of reasons it would be naive to consider ap¬
plied social experimentation as a panacea for all of the social
ills. Although many of the major problems confronting this so¬
ciety (e.g., poverty, deteriorating education, pollution, interracial
strife, population control, drug abuse, job monotony, aging, crimi¬
nality ) could be better understood and maybe solved, or at least
meliorated, by research, one must entertain the possibility that
some problems may prove to be intractable to any kind of solu¬
tion in the foreseeable future. But that in itself would be a posi¬
tive contribution in that it would encourage a more realistic
perception of the problem and perhaps even stimulate new ap¬
proaches in adjusting to it.
It must also be recognized that a “multiethical” society such
as the United States will always have fundamental moral con-

fliets (e.g., abortion) among difFerent groups; inevitably some
segments of soeiety of the population will become offended by
and resent the solution that is adopted. However, even in this re¬
spect, research might suggest techniques that will be helpful in
reducing intergroup friction.
Some might fear that applied social experimentation will
threaten basic liberties. For example, research might indicate that
the most effective method of coping with terrorism would be to
sacrifice the liberty that guarantees protection ‘'against unreason¬
able searches and seizures.” However, to perceive the applied so¬
cial experimentation as the source of the threat would be mis¬
taken. Applied social experimentation provides knowledge; society
determines how it is to be used.
Moral Training
The final point in my discussion of the potential contributions of

psychology to the solution of ethical problems is a simple one, al¬
though of great theoretical and empirical complexity. If one ac¬
cepts both the notion that a society can adopt ethical modes of
behavior and that individuals can be trained to behave in a man¬
ner consistent with those modes, then psychologists should be
able to develop effective methods of moral training. What form
these training techniques will assume must be left open-ended
particularly in regard to such fundamental issues as to whether
training in a school setting can be substituted for family training.
But it is certainly within the realm of the possible that educa¬
tional procedures can be designed that will decrease the incidence
of moral delinquency expressed both in criminal and ruthless be¬
havior. Again, the basic assumption here is that if psychologists
are provided with a clear conception of some final behavioral per¬
formance that society deems desirable, then techniques can be
developed to increase the probability of occurrence of the de¬
sired behavior. The only obstacle to such an optimistic expecta¬
tion would be unmodifiable forms of unethical behavior that are
genetically determined. Although future research might demon¬
strate that such a pessimistic view is appropriate for some limited
forms of criminal behavior, it seems inconceivable that it would
apply to the entire spectrum of a socially defined range of im¬
morality.
One point frequently overlooked in moral training is that
certain goals may be more easily attained than others. One such
example, which stems from a dispute in the history of ethics, is
whether happiness constitutes a more appropriate ethical im¬
perative than ''the absence of suflFering.” Should a society be "en¬
gineered so that its members will be happy or would it be a
more reasonable goal to attempt to eliminate "suflFering.” Some
psychologists would immediately dismiss such a question as
meaningless because of the inability to define precisely such
phenomenological states such as "happiness” and "suffering.”
This reaction could be considered excessively harsh and rigid be¬
cause it should be possible to use such phenomenological de¬
scriptions when developing intersubjectively reliable behavioral
measures of "happiness” and "suffering.”
This problem is similar to the dispute that revolves about the
concept of mental health.” Should it be defined as a list of posi¬
tive behavioral characteristics, or is it more reasonable to define
mental health as the absence of pathological reactions such as
hallucinations, incapacitating anxiety, and persistent and inap¬
propriate depression? Although a positive conception of mental
health appears at first glance to represent a more desirable goal
than the mere absence of pathological reactions, it soon becomes
apparent that the positive conception of mental health is bur¬
dened with complex ethical issues associated with "goodness”
and "desirability” plus an intrinsic inelasticity to cope with the
profound behavioral variability among humans. Thus any positive
conception of mental health might be impractical and perhaps
even threatening to a conception of individual freedom that rec¬
ognizes the immense diversity of human behavior. In the same
vein, the human condition may make the romantic notion of hap¬
piness for all a naive hope rather than a realistic goal. Perhaps
the "best” society a democracy can engineer (and it is more than
likely that other forms of government would not do as well)

would be one based upon the ethieal goal of eliminating “suf-
r • »
rering.
As a psychologist, I do not wish to be placed in the position
of supporting or favoring a pessimistic view of the human poten¬
tial. The only thrust of my argument is that psychology, when
considering ethical behavior, cannot limit its sights exclusively to
problems of moral training. The obvious point that all moral
goals are not equally achievable must be recognized and its im¬
plications fully appreciated. By recognizing this point and its in¬
trinsic relationship to the problem of ethical training, psychology
will be in the best position to oflFer positive contributions to the
solution of some of the most pressing ethical issues of our time.
The effectiveness of psychologists’ contributions will depend
upon their ability to disabuse themselves of the notion that they
can dictate what is good and bad. Philip Handler (1980), when
president of the National Academy of Sciences, warned that,
“Scientists best serve public policy by living within the ethics of
science, not those of politics.” Psychologists must always be
aware of the threat of a tyrannical morality to a democracy by
those who are convinced that they alone know what is right—
whether they be religious zealots, political fanatics, or theoretical
psychologists.
Future Trends
in Psychology
1. A fundamental disagreement prevails among psyehologists about

the nature of their diseipline and the role it should play in soeiety.
These differences represent irreconcilable views that prevent psychol¬
ogy from operating as a unified science or profession.
2. In the coming years the following trends are expected to develop

or continue: (a) Neurophysiological interpretations of behavior will
increase in popularity because of recent biological and technological
developments and also because of the increased disenchantment with
“black-box” behavioral theories, which due to insufficient constraints
do not provide for clear-cut resolutions of theoretical differences. This
trend toward neurophysiological interpretations does not spell the de¬
mise of “black-box” formulations because they can provide a functional
analysis of behavior that can guide neurophysiological research and
theorizing, (b) In order to avoid doing research that yields trivial re¬
sults psychologists will try to develop strategies that increase the prob¬
abilities of fruitful investigations (e.g., conduct research within a de¬
velopmental or comparative framework), (c) The failure of experi¬
mental social psychology to yield major theoretical breakthrough has
encouraged social research in natural settings. Although such research
may not be theoretically more significant, it promises to have greater
social utility. Of particular importance is evaluation research, which
has the potential for providing appropriate information for the selec¬
tion of effective social policies and programs, (d) Two controversies,
holism versus atomism and rationalism versus empiricism, that have
299
persisted throughout the history of psychology will continue to occupy

the center stage of psychological theorizing. Neither controversy lends
itself to any simple resolution because each position in both conflicts
represents an important research and theoretical strategy that has
some merit, (e) Although the effectiveness of psychotherapy, “talk
therapy,” has frequently been questioned, its popularity will continue
because it serves many personal needs ranging from offering the pa¬
tients “psychological health” or “spiritual guidance” to providing them
with an intrinsically interesting or entertaining experience. In the con¬
text of a national health program, psychotherapy has the greatest
chance of receiving government support when it demonstrates its cost
effectiveness in a medical setting, (f) TJie information-processing
paradigm, although experiencing an unrivalled success in capturing
the allegiance of a large proportion of experimental psychologists in¬
terested in learning, memory, perception, and thinking, is presently
confronted with experimental and theoretical problems that do not
lend themselves to any simple solutions. The absence of a cohesive
theory has resulted in a proliferation of models that share, at best, a
common language but no core assumptions. The explanatory capacity
of information-processing models have frequently been overstated
when theoretical conjectures are offered in the guise of substantive
formulations. The use of the computer analogy to explain human be¬
havior must continue to be considered a strategy the value of which
has yet to be convincingly demonstrated. In essence, the information¬
processing paradigm is an experiment in theorizing, the results of
which are not yet in.
THE PRESENT STATE OF PSYCHOLOGY
This examination of the epistemological structure of psychology

has revealed a diversity of basic assumptions and ultimate goals.
What does this methodological state portend for the future of
psychology?
Although the question seemed eminently appropriate for the
concluding chapter of a book that seeks to reveal the nature of
Future Trends in Psychology SOI
psychology, an answer is not easily forthcoming. Pessimistic eval¬

uations of its future by some discerning critics contrasted with
enthusiastic promises of human salvation by some of its practi¬
tioners, plus claims of psychology’s social value along with fears
of its social consequences, all suggest that prediction of psychol¬
ogy’s future is a complex and risky task. The task can be made
easier by lowering ones sights from that of an aspiring oracle
who strives to predict the exact shape that psychology will as¬
sume in years to come to that of a modest prognosticator who at¬
tempts to identify some of the directions that psychology will
take. By adopting the more modest option, the task of previewing
the future becomes manageable.
THE STATUS OF CONTEMPORARY PSYCHOLOGY
Most psychologists feel both secure and confident about the fu¬
ture of their discipline. Its growth has been phenomenal during
the past decades, illustrated by the increase of the membership of
the American Psychological Association (APA) from less than
1,000 in 1925 to well over 40,000 today. Concomitant growths
have also occurred in the number of publications and journals, as
well as in departments and schools that are offering doctorates in
psychology. But hidden behind these figures are sources of dis¬
sension that seek to split psychology, at least into two, if not
more, separate and independent disciplines. Many natural-science
oriented psychologists became disenchanted with the APA follow¬
ing World War II because of what they perceived as an over¬
concern with professional issues at the expense of scientific af¬
fairs and standards of scholarship. As a result a new organization
was formed in 1959, the Psychonomic Society, designed not to
compete with the APA for its membership but rather to sever ties
with most of the psychologists who belonged to APA. The aim of
the Psychonomic Society was to limit its membership to those
psychologists who demonstrated proven research skills of the sort
that was consistent with a natural-science approach. The Psycho¬
nomic Society deliberately avoided matters of professional inter-
ests, limiting its concern mainly to the holding of annual meet¬

ings to encourage communication about empirical and theoretical
matters in the behavioral sciences. Although initially most mem¬
bers of the Psychonomic Society held joint membership in the
APA, the continued professionalization of that organization en¬
couraged some older psychologists to terminate their APA mem¬
bership and some younger ones to refuse to join.
The founders of the Psychonomic Society had no desire to
abandon the name of psychology or its traditions. The most con¬
troversial step in founding the organization was the selection of
the name psychonomic in preference to a title that would include
the term psychology. They desired a naipe that would set them
apart from the APA but at the same time would not deny their
psychological heritage. If the name Society of Experimental Psy¬
chology, or some variation of it, were available, it would have
been adopted. But that option was precluded by the existence of
an honorary organization, the Society of Experimental Psycholo¬
gists, organized by Edward B. Tichener in 1927.
Although the Psychonomic Society was perceived by many
outsiders as an effort of ivory-tower psychologists to turn their
backs on problems of society, the charge was unfounded. Experi¬
mental psychology from its beginning in the United States had
an interest in applied fields such as psychological testing, educa¬
tion, advertising, industry, and medicine (McKinney, 1976).
Many of the founders of the Psychonomic Society had been
active in applied programs during World War II, particularly
those associated with the selection and training of Air Force per¬
sonnel. Although members of the Psychonomic Society might dif¬
fer as to when it was strategic to appl)^ methods and knowledge
of psychology to the solution of practical problems, the traditions
of natural science denied any unbridgeable chasm between pure
and applied research. In fact, many “psychonomes would argue
that the distinction between pure and applied research is tenuous
at best: Can the development of the transitor or a psychological
test be simply classified as either an expression of pure or applied
research?
One can point to several other conflicts that signified a fun-
Future Trends in Psychology SOS
damental divisiveness in contemporary psychology that were not,

as some claimed, simply a contest for professional power or a dis¬
agreement about applied psychology. Departments of psychol¬
ogy, after struggling to achieve their academic independence,
found themselves struggling to retain their integrity. Some no¬
table failures occurred. At one distinguished university two de¬
partments had to be formed: a department of social relations,
which attracted psychologists with a social-science bent, and a
department of psychology, which primarily contained tradition-
oriented experimental psychologists in the fields of sensation, per¬
ception, and learning. The department of social relations, con¬
sisting mainly of personality and social psychologists, as well as
anthropologists and sociologists, felt it necessary to provide train¬
ing in traditional experimental psychology but difficulties arose
when some of the experimental psychologists and their students
found it intolerable to be isolated from the methodological and
historical moorings they coveted.
Throughout the country departments of psychology were at¬
tempting all sorts of maneuvers to reconcile the differing concep¬
tions of graduate education appropriate for experimental, clinical,
and social psychologists. In some cases the clinical programs
were dropped or transferred to the schools of education. Many
decided to avoid the conflicts by not having a clinical training
program. One alternative offered was to give a doctorate in psy¬
chology (D.Psy.) that did not require a research thesis. All
these academic machinations, expressions of a fundamental dis¬
cord between clinical and experimental psychology, were not
passively accepted by clinicians. Those who felt that their future
was being controlled by psychologists who, at best, did not un¬
derstand the needs of clinical psychology and, at worst, did not
understand psychology established independent schools of psy¬
chology that were not affiliated with universities but nevertheless
granted doctorate degrees.
More recently, physiological psychologists, many of whom
prefer the designation of biopsychologists or neuroscientists, have
revealed a new source of friction. They find that their ties to biol¬
ogy in particular, and to neurosciences in general, are much
greater than to their nonphysiologically oriented colleagues. In a

few cases independent departments of biopsychology have been
established while in other cases special training programs in bio¬
psychology have been created that have encouraged the isolation
of biopsychologists from other psychologists.
This ongoing tendency toward fission in psychology can be
simply viewed as a consequence of the age of specialization that
characterizes all sciences. From physics to biology there are fields
of specialization that appear to have nothing in common with
other areas. Although specialization (perhaps overspecialization)
is a source of divisiveness in contemporary psychology it is not
the main source of turmoil. The roots are deeper within the fun¬
damental nature of psychology.
About a decade ago I delivered a talk (Kendler, 1970) en¬
titled “The Unity of Psychology” to the Division of General Psy¬
chology of APA, whose members consider themselves to be
involved with problems and issues that transcend the boundaries
of the numerous fields of psychology. In this talk the diversity of
volitional decisions that psychologists have made in choosing
their subject matter and the mode of understanding was rec¬
ognized but hope was expressed that the integrity of the field, in
spite of the fundamental differences in orientation, could be
achieved:
The unity of psychology is not beyond the realm of possibility

and consequently it cannot be considered to be completely unreal.
Whether a unity is achieved or not will be dependent on the
psychological community accepting a criterion of explanation that
wiU permit a framework by which the relative merits of com¬
peting interpretations can be judged, not necessarily to provide a
single unequivocal overall evaluation but instead to identify the
assets and liabilities of each. It may seem naive to think that a
common explanatory frame of reference can be agreed upon by
such an already fragmented community but a look at the annals
of more mature sciences testifies to the ability of historical pro¬
cesses to settle what initially appeared to be unresolvable con¬
troversies (ibid., pp. 45—46).
My methodological preferences then were exhibited by suggest¬

ing that, “the most likely candidate for a mutually acceptable ex-
Future Trends in Psychology 805
planatory frame of reference is one that requires a deductive

component”
In retrospect I must confess to being naive and unrealistic. I
harbored the optimistic belief that the natural-science method in
psychology would ultimately achieve a level of success that would
encourage the abandonment of competing methodological ap¬
proaches. Methodological controversies had occurred during the
early stages of physics, chemistry, and biology, but the natu¬
ral-science methodology prevailed because of the successes it
generated.
My optimism was unjustified for two reasons. Natural-science
methodology was not as successful in discovering and integrating
the facts of psychology as I had anticipated. Even if it were more
successful the need for other methodological approaches (e.g.,
humanistic, literary) would remain. This point struck home re¬
cently when, in search of a paperback, I wandered over to a large
drugstore in an airline terminal. The display was divided into
seven major sections, one of which was psychology. I failed to
recognize most of the books except those in the general areas of
psychoanalysis and humanistic psychology. The titles suggested
that these books possessed a variety of appeals: recipes for ad¬
justment, acquaintance with other minds, knowledge of psycho¬
analysis and humanistic psychology, disclosures about abnormal
behavior, descriptions of encounter groups, insights into moral
beliefs, and fun and games. If natural-science psychology, with
its demanding requirements for warranted conclusions, were
more successful, some of these books would lose their appeal.
Psychological games and simple recipes for living, regardless of
their justification, are attractive to many people as is demon¬
strated by the persistent interest in astrology. These needs will be
satisfied by authors who, because of conviction or profit, will con¬
tinue to produce their wares in the absence of any critical natu¬
ral-science standards to evaluate their efforts.
But returning to the more basic question, why has the prog¬
ress of natural-science psychology been disappointing? Back in
1929, Boring, in concluding his famous book A History of Experi¬
mental Psychology "'confesses to a certain disappointment . . .
that experimental psychology had not accomplished more than it
has in seventy years of life” (p. 659).^ Fifty years later many dis¬
tinguished experimentalists would still bemoan the laek of im¬
pressive progress and perhaps even aflBrm William James’s ehar-
aeterization of psyehology as a ‘"nasty little subjeet.”^ Boring
offered two explanations for the lack of progress in psychology.
One was that psychology had no great men like Darwin, the
other that psychology was too involved in philosophy: “Psychol¬
ogy ought to fare better when it completely surrenders its philo¬
sophical heritage, in fact as well as in voiced principle, and pro¬
ceed, unimpeded by a divided soul, about its business” (1929,
p. 661).
Neither reason is compelling. The absence of a genius may
reflect the refractoriness of psychological problems to any broad-
range theoretical solutions. In other words, psychology does not
provide opportunities for gifted individuals to achieve genius sta¬
tus.^ Perhaps the overconcern with methodological issues ex¬
presses the unique problems of psychology that are not shared by
other disciplines. In addition, it can be argued that the method¬
ological discussions have not been too excessive but instead too
superflcial.
In contrast to Boring’s disappointments, Sigmund Koch’s
(1974) view of psychology is exhilaratingly depressed. He notes
that John Stuart Mill in 1843 suggested that the “backward state”
of the psychological sciences, which Mill labeled the “moral sci¬
ences,” could “only be remedied by applying to them the methods
of the physical sciences.” Koch believes Mill’s recommendations
have been implemented over the past 100 years and “the Millian
hypothesis has been fulsomely disconfirmed.”
Koch backs up his conclusions with a variety of arguments.
1. Boring (1929) dates the beginning of experimental psychology to

Gustav Fechner’s publication of Elemente der Psychophysik in 1860.
2. Many psychologists mistakenly remember James’s disparaging com¬
ment as that “nasty little science.” When he completed The Principles of
Psychology he informed his publisher that his book proved that there was no
science of psychology (Watson, 1978).
3. Boring (1950) in his second edition of A History of Experimental
Psychology abandoned his great-man theory of scientific progress; great men
were not “causes of progress” but only “agents.”
Future Trends in Psychology S07
Psychology cannot be considered a coherent field, either theo¬

retically or substantively—a view that is, in some sense, in agree¬
ment with my own. Failure to achieve theoretical coherence is
not particularly damning because, as Koch recognizes, contempo¬
rary physics also does not qualify as a coherent discipline.
Can psychology be considered a science? Koch, admitting
that the term science has not yet been, and probably never will
be, precisely explicated, nevertheless decides that physics obvi¬
ously qualifies for that designation but psychology does not. He
retreats somewhat from this latter conclusion by admitting that
biopsychology approximates a science but then suggests that it
properly belongs to biology and would benefit from such an af¬
filiation. He ignores the contributions of psychology to the suc¬
cesses of biopsychology-the empirical laws that have provided
direction to psychophysiological research and a sophisticated
methodology, particularly in regard to measuring behavior. Only
when nonpsychologists attempt to measure behavior do the
achievements of psychologists become obvious. Consider the con¬
tributions of psychological methods to pharmacology in evaluat¬
ing the influence of drugs on behavior and to ethology in clarify¬
ing such concepts as sensitive period and imprinting. Regardless
of its future departmental affiliations, biopsychology will lose
much if it ignores psychological measures of behavior as well as
the laws that relate behavior to environmental events and past
experience.
Koch’s thesis is that one cannot create a field of science by
edict as behaviorists have attempted. He accuses them of being
more concerned with their commitment to science than to their
subject matter. According to Koch this criticism is no less appro¬
priate for information-processing models than it is for stimulus-
response conceptions;
The strategy now becomes that of turning to the revered natural

sciences, especially the engineering disciplines, not merely for pre¬
fabricated methods but for prefabricated answers as well. We
thus get the models based on computer simulation (in some of
which the computer is the psychological subject) or on trans¬
positions to the events generated by actual human subjects of a
variety of developments in applied mathematics, ranging from in¬

formation theory to the theory of games. In this way does the
science of psychology maintain consistency with its history; by
headlong retreat from the psychological subject immediately upon
the long delayed moment of reconfrontation (1974, p. 18).
Because the knowledge emanating from what I call here

natural-science psychology is fundamentally spurious, according
to Koch, it generates a degrading image of man. Those who ex¬
pect that Koch’s concern with the concept of the image of man,
as discussed earlier (page 202), would lead him to be more
sympathetic to humanistic psychology will be sadly disappointed.
Koch’s initial contact with humaniîc psychologists led him
to characterize them as “a motley group with heterogeneous in¬
terests” that was “not a Torce’ at all but, rather a large number of
individuals who would have . . . difficulty communicating with
each other and who stand for nothing focal, other than a feeling
of disaffection from the emphasis of recent American psychol¬
ogy” (ibid., p. 34). His final verdict for the human potential
movement, which, properly speaking, can only be considered as
a segment of humanistic psychology, is that “it challenges any
conception of the person that would make life worth living, in a
degree far in excess of behaviorism” (ibid., p. 37). Depth psy¬
chology fares no better; it is emphatically dismissed because of
its obvious scientific inadequacies:
The widely shared idea that psychoanalysis is a “science” which

can cumulate and progress by a logic of verification similar to that
of physics or even biology is absurd. As a result of all these gen¬
erous efforts, a jungle of phony discourse has grown up in this
area which would be unique in its capacity to obfuscate the en¬
thusiasts who assent to it, were it not for the superior virtuosity
of Marshal McLuhan, or perhaps certain Existential philosophers
(ibid., p. 31).
Having shown to his own satisfaction the dismal condition of

contemporary psychology, Koch raises the question of what
should be done to improve the state of psychology. Suggesting
remedies proves to be more difficult than offering criticisms.
Koch admits to being unable to “oflFer a eonstruetive and merry

eoda. He thinks that, We are ... at a grave impasse in the
history of seholarship—indeed, the history of intelligenee. I am
sanguine enough to believe it a temporary one. But I have no
recipe for its removal (ibid., p. 38). He does make one strong
reeommendation that the term psychology should be replaeed
with the phrase the psychological studies. This substitution would
underline the faet that psyehology is not a eoherent diseipline
with one methodologieal approaeh; “The psyehologieal studies,
if they are really to address the historieally constituted objeetives
of psyehologieal thought, must range over an immense and dis¬
orderly speetrum of human aetivity and experienee” (ibid., p. 26).
To prepare for this examination Koeh eneourages a ehange in
attitudes:
We who are psychologists or humanists must become for a while

not psychologists or humanists, but men. Let the teaching of the
psychological studies and the humanities be a matter of men ex¬
ploring the meanings of human experience, actions, and artifacts
at their most value-charged reaches, among men. Let the teacher
be wiser, more able than the student to discriminate finely and
value precisely within important segments of human reality (ibid.,
p. 39).
It would be unfortunate, and unwise, to dismiss Koch’s criti¬

cisms if for no other reason than that they articulate a disenchant¬
ment with psychology felt by many people, including psycholo¬
gists and their students. Are such feelings justified? A simple “Yes”
or ‘No response is inappropriate because independent issues are
involved, each demanding a qualified answer.
If a simple litmus test is not available to identify a true sci¬
ence it would appear both foolish and unproductive to pursue the
question of whether Koch’s negative verdict about psychology is
correct. If one accepts the notion that the assignment of the
designation science to a discipline is appropriate when progres¬
sive increase in a body of factual knowledge takes pi ace with a
concomitant integration of those facts, then the conclusion can be
drawn that psychology would qualify as a science. My support of
SIO Psychology: A Science in Conflict
this thesis is no doubt influenced by my decision to write an in¬

troductory psychology text (Kendler, 1963, 1968, 1974) that was
based upon the assumption that a scientific text was possible.
Executing the task confirmed the initial judgment. The following
research areas, certainly not all that could be mentioned, contain
a body of facts that are integrated either in the form of deductive
models of explanation or by systematic methods to control the
phenomena’s occurrence: visual and auditory sensitivity, color vi¬
sion, visual space perception, size constancy, acquisition and ex¬
tinction of classical condition responses, schedules of reinforce¬
ment in operant conditioning, reinforcement by brain stimulation,
effects of punishment, ‘‘instinctive” behavior, organization and
memory, developmental changes in human learning, language
acquisition, functional fixedness, human aggression, individual dif¬
ferences and psychological testing, human sexuality, child rearing,
behavior modification, genetic basis of behavior disorders, psy¬
chopharmacology, and attitude change.
If one examines these topics within a historical perspective,
an appropriate conclusion would be that knowledge of the em¬
pirical relationships has increased and understanding has im¬
proved. In short, scientific progress has occurred. A person siding
with Koch’s position could refute this claim of scientific progress
by arguing that criteria employed of scientific knowledge is too
lenient, that the evidence cited emphasizes both the lack of co¬
herence of psychology and the failure of psychology to deal with
significant aspects of being human such as personal experience
and artistic achievements.
According to Koch the analytic pattern that characterizes
science requires:
(a) the disembedding from a domain of phenomena of a small

family of “variables” which demarcate important aspects of the
domain’s structure, when that domain is considered as an ideal¬
ized, momentary static system; and (b) that this family of vari¬
ables be such, by virtue of appropriate internal relations, that it
can be ordered to a mathematical or formal system capable of
correctly describing changes in selected aspects of the state of the
system as a function of time and/or system changes describable
as alterations of the “values” of specified variables (ibid., p. 23).
Koch notes that this description characterizes such ‘ simple”
physical systems as the laws of the pendulum and the motions of
falling bodies, achievements that required “a prolonged develop¬
ment of ancillary knowledge, culminating in an act of genius . .
My reading of Koch’s two requirements of scientific status would
exclude Darwin s formulation of his evolutionary theory. They
would also exclude other parts of biology, particularly those that
are in their early stages of development (e.g., brain chemistry).
Koch’s requirements appear too stringent; they emphasize char¬
acteristics of final goals of scientific knowledge while ignoring in¬
termediate points.
Perhaps the key criticism directed at contemporary psychol-
Koch is its failure to deal with subject matters that he
considers to be related to the most distinctive and most highly
valued aspects of human activity, i.e., conscious experience, par¬
ticularly that associated with aesthetic creativity and apprecia¬
tion. A natural science of phenomenology, as has been suggested
(pages 94-99), may be unachievable simply because the ob¬
server and what he observes cannot be separated. If aesthetic
creativity and appreciation must be understood in terms of hu¬
man experience, and not in terms of a detached symbolic system
that organizes the factual domain of aesthetic behavior, then per¬
haps Koch is asking too much of natural-science psyehology. Saul
Bellow, the distinguished American novelist and Nobel Laureate,
offers an interesting commentary upon this issue. In his novel
Humboldfs Gift he describes the desire of his protagonist, Charles
Citrine, a writer, to understand the desires of others.
To do this one had to remove all personal opinions, all interfering

judgments; one should be neither for nor against this desire. In
this way one might come gradually to feel what another soul was
feeling. I had made this experiment with my own child Mary.
For her last birthday she desired a bicycle, the ten-speed type. I
wasn’t convinced that she was old enough to have one. When we
went to the shop it was by no means certain that I would buy it.
Now what was her desire, and what did she experience? I wanted
to know this, and tried to desire in the way that she desired. This
was my kid, whom I loved, and it should have been elementary
to find out what a soul in its fresh state craved with such inten-
sity. But I couldn’t do this. I tried until I broke into a sweat,

humiliated, disgraced by my failure. If I couldn’t know this kids
desire could I know any human being? I tried it on a large num¬
ber of people. And then, defeated, I asked where was I anyway?
And what did I really know of anyone? The only desires I knew
were my own and those of nonexistent people like Macbeth or
Prospero. These I knew because the insight and language of
genius made them clear (1975, p. 416).
The suggestion is clear. Maybe the humanities, literature in

particular, can only reveal what Koch longs to discover, the sig¬
nificant experiences of being human. Although the potential suc¬
cess of a natural-science approach to phenomenological experi¬
ence, not merely the causes and consequences of it, but rather a
valid description of experience itself, cannot be ruled out, but
neither can its success be guaranteed (pages 94-97). Bellow
may be correct in concluding that only the language of the artis¬
tic genius can convey to a receptive audience an intuitively valid
description of the experience of others.
Whatever the future may hold for experiential psychology, it
would appear that the suggestion of substituting a department of
psychological studies for the traditional department of psychology
as it is now constituted, represents more of a hollow hope than a
productive alternative. Even if one assumed that the intellectual
interaction between humanists and psychologists would be benefi¬
cial to the latter, and perhaps even to the former, there is no
guarantee that such benefits would result from institutionalizing a
department of psychological studies. History has demonstrated
that the intellectual stress and strain among psychologists, result¬
ing from their diflFerent conceptions of “science” has often been
sufficiently great to divide them into separate departments or even
institutions. Forcing individuals from widely divergent disciplines
to share a common department is no guarantee that presumed
similarities will appear more apparent, attractive, and significant
than the obvious differences. If there is any justification for a de¬
partment of psychological studies, then it should be possible to
devise a set of core courses that should be required for all its stu¬
dents, from those who are fascinated by Anthony Trollope’s char-
Future Trends in Psychology 3IS
acterizations to those who seek to understand the interrelation¬

ships between linguistie abilities and problem solving during early
ehildhood.
The obvious reservations expressed about the potential con¬
tributions of a department of psychological studies stems from the
conclusion of my analysis that psychology is a multidisciplinary
study with diflFerent subject matters and a variety of epistemologi¬
cal assumptions about what constitutes ‘"truth/’ To attempt a rec¬
onciliation among these diverse orientations would be tantamount
to conducting a game between a chess and a checker player with¬
out changing the rules of either game. It cannot be done. Recog¬
nizing and accepting the fundamental differences in contempo¬
rary psychology may be a wiser alternative than attempting to
combine artistic and scientific traditions and “methods” that at
best could yield a homogenized product without the positive at¬
tributes of either.
FUTURE TRENDS OF PSYCHOLOGY
An understanding of the past provides a glimpse into the future.

Ry identifying significant historical trends and charting their
courses, one can predict the directions that psychology will take
in decades to come.
Neurophysiological Interpretations of Behavior
Methodological arguments have previously been advanced (pages

119-135) questioning the inevitability, and even the desirability,
of neurophysiological interpretations of behavior and experience.
These arguments have been challenged by two historical forces—
the obvious dramatic improvements in research technology (e.g.,
monitoring the activities of individual neurons, techniques for
measuring minute biochemical changes) and increases in biologi¬
cal knowledge both of which have stimulated biopsychological re¬
search and theorizing. Another reason, less obvious, for the in-
creased popularity of the neurophysiological approach is the fail¬

ure of “blaek-box” behavior theories—formulations that employ
abstraet intervening variables to bridge the gap between environ¬
mental manipulations and the behavior of the organism—to oflFer
satisfaetory aceounts of the phenomena they sought to explain.
The attempt to explain the psyehology of learning by these black¬
box theories offers an instruetive case history that in many ways
parallels attempts in other areas, sueh as cognition, perception,
and motivation, to formulate intervening-variable type coneep-
tions. These formulations of the learning proeess led to many
heated eontroversies sueh as latent learning, incremental versus
one-trial learning, eontiguity versus reinforeement, and seleetive
attention. These disputes proved unresolvable in the sense that
the evidenee did not elearly favor one formulation over the eom-
peting one as was the ease for the Coperniean and Ptolmaie
eoneeptions, and the evolutionary and ereationist eonceptions. A
eommon opinion, which I shared, about these unresolvable eon¬
troversies in the psyehology of learning was that the diffieulties
emanated from the ambiguity of the competing theories:
It appeared to many, a decade ago, that certain theories notably

those of Hull and Tolman, were engaged in mortal combat in the
arena of hard data. Now that the dust has settled, it seems that
the combatants were more often shadow boxing. Damage to the
theoretical positions did occur hut in many instances the wounds
were self-inflicted. However, these theoretical disputes neverthe¬
less did serve a purpose. They provided much interesting data;
but more important, they revealed the stark inadequacies and
limitations of existing learning theories (Kendler, 1959, p. 43).
The solution to this theoretical impasse seemed obvious.

Learning theories had to be made more precise. This need for
preeision was expressed in mathematical models of the learning
proeess employing stoehastie processes (e.g., Estes, 1950). But it
beeame apparent over the years that these mathematical repre¬
sentations, although they served as an effective antidote to the
ambiguities indigenous to general learning theories (Hull, 1943;
Tolman, 1932), were forced to saerifice the generality of the for-
mulations they were designed to replace. No mathematical model

was proposed that effectively accounted for the diversity of con¬
ditioning phenomena, much less the data from more complex
learning tasks. Specific models for specific phenomena became the
order of the day, and it became abundantly clear that if black¬
box theorists desired precision the purchase price was generality.
For most theorists the price was too high as evidenced by the grad¬
ual decrease of interest in these highly specialized models. At
best, these models offered limited answers to narrow questions.'^
In order to explain the failure of general learning theories to
be specific and specific mathematical models to be general the
structural characteristics of black-box theories must be exam¬
ined. They contain three major components: a set of independent
variables, a related network of theoretical constructs, and rules
for theoretical derivations of the to-be-expected behavior. The
actual control exerted on the subject’s behavior resides primarily
in the regulation of the independent variables such as the nature
of the discriminanda, their temporal relationships, the nature and
size of the incentive, the management of the subject’s background
and home environment, and so on. It is generally acknowledged
that there are many fluctuating and steady conditions within both
the environment and the subject that influence the behavior and
over which the experimenter has little or no control or knowledge.
Such a theoretical system can be aptly described as being em¬
pirically open in the sense that variables not accounted for by the
theory exert important influences on behavior, i.e., a lot of “noise”
affects the results. In addition, and perhaps of greater impor¬
tance, is the capacity of black-box theories to accommodate em¬
barrassing data by ad hoc modifications. There appear to be no
end to the escape clauses that can be invoked by black-box theo¬
ries to handle any phenomenon. This empirical openness and lack
4. Although failing to achieve a broad conception of the learning pro¬

cess, these mathematical models did exert a salutary effect on theoretical
practices. In addition to making precise predictions, some encouraged an
interest in trial-to-trial changes in the learning process that was largely
ignored by general theories because of their primary concern with group
differences.
of theoretical constraints place a definite limit on the explanatory

capacity of general black-box theories.
To overcome the deficiencies of black-box formulations the
influences of extraneous variables must be reduced and the theo¬
retical constraints must be increased. Perhaps this can only occur
by changing their very character—a shift of interest from environ¬
mental-behavioral relationships to neurophysiological-behavioral
dependencies. A sensible assumption, if behavior is not assumed
to be an emergent product of a vitalist force, is that the causal
basis of behavior resides within the physiology of an organism.
The excessive noise that is characteristic of black-box research
can be reduced by manipulating neurophysiological variables, the
effects of which are more easily assessed than are the influences
of environmental variables. This shift of focus can also reduce
the theoretical flexibility of such theories. It is much easier to ob¬
tain negative evidence against a hypothesis concerning the opera¬
tion of some physiological process than it is to discover that a set
of abstract theoretical processes are incapable of explaining a set
of behavioral phenomena. One example is the current controversy
as to whether the contingency or contiguity between the condi¬
tioned and unconditioned stimuli is responsible for the formation
of classical conditioned responses. Perhaps only neurophysiologi¬
cal evidence can provide an answer (Mpitsos, Collins, & Mc¬
Clellan, 1978).
In sum, the thrust of this analysis is that black-box theories
have a limited capacity to resolve the theoretical issues that they
have generated. To enlarge this capacity requires that the theo¬
retical issue be translated into a neurophysiological question.
In support of a neurophysiological strategy one can contrast
the history of theories involving environmental-behavior depen¬
dencies with those that have been concerned with neurophysio-
logical-behavior relationships. If one considers a progressive in¬
crease in knowledge as the hallmark of a scientific advance, then
one would have to conclude that biopsychology has progressed
more than has black-box psychology. An inordinate amount of
“progress” occurring in black-box psychology results from the re¬
jection of a research program from one empirical domain (e.g.,
conditioning) in favor of a research program from a new empiri¬

cal domain (e.g., memory). The shift of interest is justified by the
argument that the new empirical realm is more significant than
the old one.^
In contrast, biopsychology does not push aside its empirical
legacy. “Real” structures (e.g., cones, brain) cannot be as easily
dispensed with as can the “fictitious” concepts (e.g., field, associa¬
tion) of environmental-behavioral psychology. Color-vision the¬
ory, when approached in a neurophysiological manner, cannot
avoid coping with the activities of the cones in the retina, bipolar
cells, cells in the lateral geniculate body, the visual pathways,
and of course, the visual centers of the brain. They are here to
stay, and their functioning must be included in any theoretical
account of color vision.
Although it is easy to identify the deficiencies of black-box
theories and offer a biopsychological orientation as a remedy, it
is difficult, if not impossible, to suggest the manner in which the
transition between the two approaches is to be implemented. As
a general strategy one can recommend that neurophysiological
hypotheses should be encouraged when practicable, with the full
realization that the qualifying phrase, when practicable, is highly
ambiguous. By examining some alternative tactics the meaning of
this phrase can be clarified.
Reservations have already been expressed about broadly
speculative conceptions that have been more involved with the
conceptual nervous system than the central nervous system. A dis¬
tinction should be made between those conceptions that encour¬
age a false sense of understanding while exerting no impact on
research and those that suggest fruitful hypotheses and research
programs. Hebb (1949) represents a superior effort of this latter
type. He formulated a broad-based conception that combined
knowledge of both physiology and behavior. His ideas exerted
both a direct impact on research (e.g., the effect of early experi-
5. A commentary on this kind of “progress” is expressed in the ob¬

servation that one generation of natural scientists stands on the shoulders of
the previous generation while behavioral scientists step on the backs of their
predecessors.
ence on later learning, perceptual development) and theory (e.g.,

central processes and behavior). Although Hebb’s formulation
with its emphasis on cell assemblies influenced future concep¬
tions, its impact was greater in terms of the questions raised than
the answers offered. In contrast are the efforts of Ratliff (1965)
who dealt with a much narrower range of phenomena (Mach
bands) but offered precise neurological answers to specific per¬
ceptual questions that in some form will probably stand the test
of time. No invidious comparison is implied because both efforts
are important but for different reasons. What is being suggested
is that in the present era some strategies possess greater potential
than do others. The impressive recent developments in neurology,
biochemistry, genetics, pharmacology, endocrinology, and related
technology, encourage an intimate relationship between biology
and psychology in both empirical research and theoretical con¬
struction. The knowledge of psychology and biology has reached
a level that justifies conceptualizing problems, when possible,
within a biopsychological perspective capable of immediately
yielding researchable questions. This view does not eschew broad
integrative biopsychological theories but instead argues that the
best strategy to achieve such goals is from the ‘Tottom up” in¬
stead of from the “top down.” Starting from the top has tended to
generate broad speculations that are incapable of deductively in¬
tegrating available evidence from physiology and psychology. The
tactic that appears to possess more promise is to proceed upward
from low-level theories—initially formulate, as did Ratliff, biopsy¬
chological models for fairly circumscribed empirical realms that
include both psychological and physiological events. The expecta¬
tion and hope is that independent low-level models would merge
into broader conceptions by creative theorizing.
Two important qualifications are now in order. First, it is not
being suggested that every empirical issue in psychology should
be approached within a biopsychological framework. What is be¬
ing argued is that over the years a biological framework has be¬
come increasingly relevant to all of psychology—from sensory
psychology at one extreme to social behavior at the other—and
this trend will continue in the future for natural-science psychol-
ogy. Today one cannot understand sensory processes without con¬

sidering biological mechanisms, and the evidence suggests that
the same condition will prevail, in the very near future, for other
fields as well (e.g., perception, learning, behavior pathology). Al¬
though this trend does not demand that theoretical notions always
be expressed within a biopsychological framework, it does imply
that graduate psychology students receive fundamental training
in neurophysiology, not necessarily to be prepared to do biopsy¬
chological research but instead to understand future develop¬
ments in their own specialized fields.
Second, the expressed reservation about black-box theories
should not be interpreted as an epistemological prohibition against
such theorizing nor does it necessarily deprecate environmental-
behavioral research or, for that matter, phenomenological anal¬
yses. Consider the history of theories of color vision. The theo¬
retical problems of this research area were posed by a careful
phenomenological analysis of human color experiences. Under¬
standing of color vision was further enhanced by environmental-
behavioral research in which psychophysical methods revealed
numerous relationships between physical characteristics of light
waves and attributes of color experience. In addition, environ¬
mental-behavioral research also discovered significant phenomena
relevant to color vision theories, e.g., color mixing, negative after¬
images, simultaneous contrast. The entire set of environmental-
behavioral data, however, failed to provide any clear-cut verdict
in favor of one of the two dominant theoretical orientations: the
Young-Helmholz model or the Hering formulation. With the ad¬
vent of single-neuron recordings and other technical advances the
task of evaluating the competing research programs assumed a
clearer perspective because the theoretical issues could be inter¬
preted in terms of neuronal functioning at different levels of the
visual pathways. The Young-Helmholtz theory appears relevant
to receptor functioning (MacNichol, 1964) while the Hering for¬
mulation seems applicable to the functioning of bipolar cells and
cells of the lateral geniculate body (De Valois, Abramov, & Ja¬
cobs, 1966; De Valois & Jacobs, 1968). Although the original theo¬
ries, which guided much productive environmental-behavioral re-
search, had to be reformulated, they nevertheless served as useful

guideposts for further theoretical development. Phenomenological
analyses and black-box theories, without question, serve impor¬
tant functions in the quest for a natural-science understanding of
psychological events, but their value will be best appreciated if
they are perceived not as final destinations but as significant land¬
marks along the way.
Two major arguments can be directed at the preceding anal¬
ysis, which encourages a transition from environmental-behavioral
to neurophysiological-behavioral formulations. The first is that
this strategic recommendation expresses an unreasonable im¬
patience. Use of the historical method to evaluate competing re¬
search programs demands that sufficient time be available for
controversies to be solved. The decades that have passed since
numerous black-box research programs have been locked in dis¬
pute have not been sufficient for a satisfactory resolution to take
place. Perhaps their resolution will be forthcoming with the ad¬
vent of newer mathematical systems that are developed for psy¬
chological data and of better experimental methodologies that
will exert more powerful controls. In essence, this argument
denies a limit to the explanatory powers of environmental-behav¬
ioral theories and encourages a persistence that for the most part
has been absent in psychology with its penchant for novelty.
The second argument against the strategy of encouraging
biopsychological formulations is that it promises more than it can
deliver. The notion that all of the causal agents for behavior are
within the skin does not reduce much of the noise that permeates
black-box research. The human brain consists of approximately
10 billion neurons woven into an incredibly concentrated and in¬
tricate pattern. Six hundred million synapses exist within one
cubic inch of brain tissue. Each cortical neuron, depending on its
location, has from 6,000 to 60,000 synapses. The potential num¬
ber of different combinations of individual brain cells involved in
various activity patterns defies imagination. The number of differ¬
ent combinations possible among only five cells would far exceed
a trillion. Is it not the height of optimism to believe that the mys¬
teries of such a system can be unravelled? How can the human
brain understand itself?
Future Trends in Psychology ^21
Although these rebuttals do not dissuade me from eneour-

aging a neurophysiologieal strategy to aehieve the goals of a
natural-seienee psyehology, they eannot be rejeeted and should
not be ignored. The major reason for oflFering this methodologieal
analysis is not to persuade psyehologists, ineluding students, about
what to do but instead to make them aware of the alternatives
before eommitting themselves to a partieular strategie deeision.
Although there will be those who will opt for a blaek-box strat-
Ggy, the historieal trend toward an expansion of the empirieal
realm of biopsyehology appears inevitable. The question is how
rapidly will this trend aeeelerate and how sueeessful will its ef¬
forts be?
The Fruitfulness of Developmental

and Comparative Researeh
Nobody, in my experienee, has ever denied that mueh of psyeho-

logieal researeh is trivial. That is to say nothing would have been
lost if the researeh had not been done and that having been done
It exerts no influenee on psyehology, now or in the future. With
the explosion in the number of new psyehologieal journals, the
eomplaints inerease about the amount of unneeessary researeh.
Although general agreement prevails about the high inei-
denee of trivial studies, there is little eonsensus about the defining
eharaeteristics of sueh researeh. One ean easily get the impression
that the bottom-line meaning of triviality revolves about the para-
digmatie eommitments of the aceuser. Researeh that one does, or
that is eondueted within the general orientation that one has
adopted, is automatieally signifieant. Everything else is trivial. A
different attitude, whieh some would eonsider more enlightened
and others would eonsider more biased, is that certain research
areas yield significant evidence while others inevitably produce
trivial results. At one extreme is the position that only socially
significant research dealing with real problems of life is impor¬
tant and therefore most laboratory studies are trivial; their only
justification is to meet the academic demands of publish-or-
perish.” At the other extreme would be the view that socially sig-
nificant problems are too eomplex and too value-ridden to be

profitably investigated; tough-minded and tightly eontrolled in¬
vestigations are needed to provide a solid empirieal base on
whieh a seientifie psyehology ean be erected.
There are numerous other frames of references that are em¬
ployed to distinguish trivial from significant research but none
have gained universal or even majority acceptance. Yet the excess
of trivial research appears real to most psychologists. Only a small
proportion of papers are cited with any frequency, and an even
smaller percent are attended to after the passage of a decade or
more.
Is there any strategy that can be pursued that would reduce
the incidence of trivial research? In answering such a question, it
must first be realized that clear-cut criteria that identify signifi¬
cant investigations are impossible. The significance of any empiri¬
cal study will depend upon a variety of influences that cannot be
adequately assessed at the time the investigation is being planned.
The results that will be obtained, the potential of the research
program in which the study is embedded, the impact of subse¬
quent studies to determine whether the original investigation was
“properly” designed, the future importance of the general re¬
search area, and other factors will all have an influence on the
potential significance of any set of findings. In addition, it must
be recognized that some trivial research is unavoidable because
some conjectures, reasonable at the time they are formulated, will
not lead to any payoff. Trivial research is also encouraged by in¬
stitutional pressure on individuals to do ‘ scholarly work. And
finally it must be acknowledged that psychology provides an
abundance of research problems. New research paradigms, which
tend to be generated at a rapid rate, each contain a host of prob¬
lems that simply involve systematic variations of the major vari¬
ables. Anybody who desires to do research, for any reason, will
experience no difficulty in selecting a problem to investigate.
In spite of the intrinsic vagueness of the term trivial, and the
operation of several forces that encourage insignificant research,
an argument can be advanced that strategic guidelines are avail¬
able to increase the probability of significant research outcomes.
Nature can provide guidance, admittedly imperfect, to encourage
important investigations; any phenomenon that exhibits develop¬
mental or comparative significance holds promise of being impor¬
tant. This argument in favor of a developmental and/or compara¬
tive research strategy rests upon a simple assumption that the
structure and functioning of organisms undergo important changes
during their ontogeny and phytogeny. When behavior is related
to these changes they tend to possess a built-in significance. The
question Why directed at the obtained relationship will prob¬
ably focus on a fruitful question.
Two examples in support of the significance of comparative
and developmental research are Beach’s (1947) research on the
evolutionary changes in mammalian sexual behavior as a func¬
tion of hormonal secretions and Money and Ehrhardt’s (1972)
developmental analysis of the sexual behavior of human her¬
maphrodites with the same genetic sex but diflFerent sex typings.
Both studies provide empirical evidence that are a consequence
of potent variables and as a result possess an importance that is
not fully dependent on the interpretive framework in which they
are embedded. If the hypotheses that Beach, and Money and
Ehrhardt offered as interpretations of their findings were ulti¬
mately rejected, the findings, assuming their reliability, would
still be significant. In contrast are the numerous findings that are
wedded to a specific empirical procedure for their theoretical jus¬
tification. When the realization occurs that the research effort is
failing to live up to its expectations, the empirical findings are
discarded to be remembered only as a historical anomaly. One
such example, which began at the beginning of the twentieth cen¬
tury, was the use of multiunit maze learning of animals as a
source of theoretical principles of learning. When maze learning
was investigated in great detail, it was discovered that a large
number of specific factors influenced the results (Woodworth,
1938). Some of the factors were: (1) anticipatory errors (an er¬
ror, for example, a right turn, that anticipates the final correct re¬
sponse that leads to the goal); (2) position habits (some rats
exhibit a preference for right or left turns presumably because of
their anatomical structures); (3) centrifugal swing (when swing-
ing around a corner a rapidly running rat tends to hug the wall
to which he is brought close encouraging him to enter the alley,
correct or not, that the wall leads to). The eonsequence of all the
research on multiunit maze learning was that the large number of
specific factors obscured or distorted whatever basic principles
were operating. Multiunit learning was abandoned as a research
area for the psychology of learning.
The argument that a natural significance attaches to develop¬
mental and comparative research might be considered to be an¬
other form of the previous position in favor of a biopsychological
orientation. Biopsychology investigates the relationship between
behavior and neurophysiological events' and in essence the same
relationship is involved in the developmental and comparative
approaches. Although this may be potentially true it does not
necessarily follow that problems in developmental and compara¬
tive psychology must always be conceptualized within a bio¬
psychological framework. The Piagetian research program is a
striking example of a developmental approach that avoids getting
involved in specifying, or even speculating about, the neuro¬
physiological substrate of behavior. Only an acknowledgement is
made to the involvement of maturational factors in ontogenetic
changes in cognitive development. One can express reservations
about the Piagetian research program; yet one would be hard put
to deny the significance of both the obtained ontogenetic changes
in cognitive functioning and the general theoretical proposal that
the sequence of cognitive stages is invariant.
Similarly one can point to phylogenetic differences in learn¬
ing set data. Harlow (1949) extended traditional discrimination¬
learning procedures to a series of successive problems each with a
different pair of discriminanda. With monkeys as subjects he
noted an astonishing change in learning proficiency from slow,
gradual learning during the initial problems to the rapid, practi¬
cally one-trial learning after three hundred problems. The ability
to learn-how-to-learn (i.e., to establish learning sets) is influenced
by phylogenetic differences; primates are markedly superior to
carnivores who, in turn, exceed by far the performance of rodents
(Warren, 1965). Although, at present, no black-box theory of
Future Trends in Psychology ^25
learning sets has gained wide acceptance, the data themselves

will, in all likelihood, retain their significance because of their
natural importance. And it is likely that their significance will be
enhanced as biopsychological procedures improve. The learning-
set paradigm represents a sensitive tool to analyze the relation¬
ships between neurophysiological structures and problem-solving
behavior. ^
Developmental and eomparative research designs are obvi¬
ously no guarantee that a particular investigation will be well
executed or conceptually significant. The history of developmen¬
tal psychology is filled with uninspired investigations in which
convenient behavioral measures are related to chronological age.
In absence of any theoretical context or compelling justification
for employing the response measure, the resulting evidence usu¬
ally proves insignificant. But at the same time it should be noted
that even in the absence of any grand theoretical design, empiri¬
cal relationships between age and behavior can be important.
The eflForts of Gesell and his co-workers (1940, 1946) have pro¬
vided interesting information about ontogenetic changes in the
human child that has proved useful both to parents who desire to
know whether their child is progressing normally and to behavior
scientists who seek to understand developmental processes.
The developmental method of analyzing behavior also pro¬
vides a powerful tool for analyzing the complex nature-nurture
interactions. Consider the problem of trying to tease out environ¬
mental and hereditary influences on gender-related cognitive func¬
tioning. Presumably females, on the average, exceed males in
several language skills while males, in contrast, surpass females in
a variety of visual-spatial tasks (Wittig & Petersen, 1979). Is this
a result of genetic influences or sex typing? Such a question would
be impossible to answer by comparing groups of adult men and
women. But if the problem is approached within a developmental
framework, it becomes possible to identify prenatal and postnatal
genetic and environmental influences, and their interactions, that
lead to the observed adult diflFerence. And when the developmen¬
tal analysis of one species is compared with other species for be¬
havior patterns that are shared (e.g., gender-related sex behav-
ior), the obtained inforaiation beeomes even more enlightening

in unraveling hereditary and environmental influenees.
If the suggestion is eorreet that developmental and eompara-
tive research possesses a built-in significance, except when con¬
ducted in a mindless fashion, then one can expect an increase in
popularity of these fields in the future. Of course it should be rec¬
ognized that the notion that developmental and comparative re¬
search is specially important is not a novel idea. Fundamentally
these research areas represent an offshoot of a Darwinian orienta¬
tion because ontogenetic and phylogenetic differences in behavior
typically possess functional significance. This is why both kinds
of research were encouraged initially by the functionalists and
later by the behaviorists. However, this analysis goes one step fur¬
ther than simply admitting the importance of developmental and
comparative research. My point is that these areas possess a spe¬
cial theoretical significance in psychology. Neisser (1974) ex¬
pressed this point when he bemoaned the fragmented state of the
information-processing research program and suggested that a
developmental orientation might help to integrate it.
Developmental psychology has been a rapidly expanding
field in recent decades, due in part to the infiuence of Piaget but
also to the demands of society in coping with problems intrinsi¬
cally developmental in nature such as the education of the young
and the care of the aged. These problems cannot be approached
without an appreciation of the developmental processes that are
involved.
Although comparative psychology has not expanded as rap¬
idly as has developmental psychology, partly because of its inti¬
mate relationship to biopsychology with its great demands for
technical skills, it nevertheless has achieved a more central impor¬
tance in theoretical conceptions including that of human behavior.
We are now more aware of the importance of species-specific be¬
havior than we were decades ago and are more sophisticated
about the relationship between human and infrahuman behavior.
If we take a hint from biological research and the Darwinian
framework, we can avoid either extreme position that insists there
is no difference or no similarity between human and infrahuman
behavior. Continuities and diseontinuities will be found, and their

extent will be determined by the speeifie kind of behavior that is
being analyzed.
In eonelusion, forees within psyehology demand that special
attention be paid to developmental and comparative psychology.
These areas contain empirical relationships that express the op¬
eration of basic processes that will have to be incorporated into
any natural-science theory aspiring to some degree of generality.
Evaluation Research
Psychological methodology has outstripped psychological knowl¬

edge. The arsenal of mathematical weapons and modes of mea¬
suring behavior appear more sophisticated than the information
that these techniques have produced. The area of psychology that
is perhaps most aptly described by this commentary is social psy¬
chology, the field frequently viewed as being in a ‘crisis state.”
Many social psychologists admit to being disenchanted because
the hopes their discipline initially engendered have not even be¬
gun to be realized. Not only were social psychologists going to
participate in the exciting intellectual task of discovering the re¬
ciprocal influences that take place between individuals and so¬
ciety, but they also would contribute knowledge that would help
implement needed social change.
Several reasons could be (and some have been) cited for the
failures of social psychology—the limited explanatory capacity of
black-box theories, the conflation of facts and values, the intracta¬
bility of social problems, the resistance to social change, and
others. For the moment attention will be shifted to a strategic
decision made by many social psychologists that may have con¬
tributed to their failure.
Psychology as a discipline, from its very inception, attached
great importance to the experimental method as the appropriate,
or at least the most fruitful, procedure to achieve understanding.
One tends to forget that although Wilhelm Wundt (1832-1920),
the father of the independent discipline of psychology, extolled
the virtues of the experimental method to study conscious experi¬

ence, he nevertheless thought it inappropriate for analyzing social
phenomena such as religion and customs that were treated in his
book Volkerpsychologie [Ethnic psychology].
American social psychologists, however, were unwilling to
accept the verdict that laboratory research was incompatible with
their subject matter. Kurt Lewin and his associates rejected this
assumption by observing social interactions under controlled ex¬
perimental conditions. One of the pioneering investigations was a
laboratory study designed to investigate the influence of “demo¬
cratic” and “authoritarian” leadership on group behavior. Groups
of 10-year-old boys were encouraged \o form clubs ostensibly
for the purpose of making theatrical masks. In some groups the
leader was coached to behave “democratically,” to be friendly,
and to encourage group discussion and decisions while in others
the leader was to be “authoritarian” and to deliberately dictate the
group’s policies and actions. Although the productivity of the
“democratic” groups was not consistently superior to that of
the “authoritarian” groups, they were friendlier, more coopera¬
tive, and less hostile (Lewin, Lippitt, & White, 1939). In referring
to this and similar laboratory experiments in social psychology, I
concluded:
Their major significance . . . was a ground-breaking demonstra¬

tion of the feasibility of adjusting the “social environment” experi¬
mentally and measuring group behavior objectively. This was an
important contribution; the advancement of any science has been
highly correlated with the ability to investigate phenomena in a
controlled experimental situation (Kendler, 1963, p. 55).
This conventional conclusion appeared beyond debate when

written. But today the value of small-group experimental research
is being seriously questioned. The editors, associate editors, and
consulting editors of three major social psychological journals
were requested “to name up to five empirieal studies which had
made a significant contribution to the field of soeial psychology.
Significant was purposely left undefined to allow respondents to
use their own definitions” (Diamond & Morton, 1978, p. 217).
Although the respondents did not exhibit a high degree of
agreement about which studies were ‘‘significant,” they did agree

about why “significant” studies were important; they opened up
new research areas and encouraged further explorations. Asch's
(1956) experimental analysis of conformity received the largest
number of citations; it stimulated many studies designed to in¬
vestigate the influence of specific variables—situational, person¬
ality, group, cultural—on the tendency of individuals to conform
to group pressures.
The second most popular study was the Festinger and Carl-
smith’s (1959) forced-compliance experiment. According to the
cognitive-dissonance model forcing a communicator to deliver a
deceitful message creates a dissonance between his actions and
his attitudes. The dissonance is assumed to be greater when the
reward for compliance is small (one dollar) as compared to large
(twenty dollars). In the latter case the communicator can ratio¬
nalize his deception by the sizable reward gained, but he cannot
admit to lying for a paltry sum of one dollar. Consequently, the
prediction, which was confirmed, was that the subjects who com¬
ply for one dollar would reduce the dissonance by changing their
attitude toward the “deceptive” message, convincing themselves
that it possesses some truth.
The cognitive-dissonance research program dominated ex¬
perimental social psychology for many years, stimulating hun¬
dreds of experiments. Many perceived the cognitive-dissonance
conception as a powerful theory that proved the scientific matu¬
rity of social psychology. Gradually, however, intrinsic ambigui¬
ties within the theory became obvious, and debates began to rage
about the appropriate interpretations of experimental results.
When interpreting the results of a given experiment, marked dis¬
agreements occurred as to whether dissonance was truly created
and if so, whether it was reduced. It appeared to some (Berko-
witz, 1969) that the theory had developed into a gigantic tautol-
ogy; a given behavior is a consequence of dissonance because
dissonance causes the behavior. What were obviously needed
were independent measures of dissonance arousal and dissonance
reduction.
The basic ambiguity inherent in the concepts of dissonance
arousal and dissonance reduction led to the degeneration of the
S30 Psychology: A Science in Conflict
cognitive-dissonance research program. Ad hoc interpretations of

apparently inconsistent results were easy to concoct while increas¬
ing precision of the theoretical concepts failed to materialize.
Competing formulations could do no better in accounting for the
wealth of data generated by the cognitive-dissonance model. The
general pattern of historical development of the cognitive-disso¬
nance research program resembled that of black-box learning con¬
ceptions; exciting new theories stimulated much research, the
results of which ultimately demonstrated the inadequacies of the
formulation.
The important methodological conclusion is that the ''signifi¬
cant” experimental social psychology studies (Diamond & Morton,
1978) achieved prominence not because they were part of a re¬
search program that represented a theoretical breakthrough but
rather beeause they stimulated a great deal of research. But as
already noted, the ability to generate additional research may not
be suflBeient justification for a theoretical effort. Should it not be
demanded that some of these efforts pay off in explaining a set
of experimental phenomena as well as in interpreting similar be¬
havior in its natural setting?
The optimism that accompanied the entrance of social psy¬
chology into the laboratory, in restrospect, seems unjustified. The
most that could be said for the significant social psychology ex¬
periments is that they called attention to, and provided tentative
insights into, apparently important processes of social behavior,
which is no mean achievement, but still far removed from the
original goals. But in some cases, experimental social psychology,
failed to do even this. One notable example is the research on an
experimental task, known as the prisoner’s dilemma, in which two
participants have an opportunity to profit from cooperation or
suffer losses from competition: "What was originally a provoca¬
tive stimulus to looking at conflict and bargaining in real social
life in a new way became a highly technical specialty in its own
right, a 'scientific’ game adrift from social reality” (Smith, 1976,
p. 440).
The failure of experimental social psychology to achieve its
anticipated goals has encouraged a reexamination of the empiri¬
cal techniques of social psychology as well as the means by which
the discipline can best serve the needs of society. In regard to the
first issue the thesis has been advanced that social psychology is
primarily a historical inquiry because basic social forces cannot
be replicated in the laboratory. As noted previously, a tendency
has prevailed in certain segments of psychology to treat experi¬
mental and empirical as synonymous. For a discipline to be scien¬
tific, the experimental method, according to some, must be the
dominant mode of inquiry. This position flies in the face of the
history of science; Darwin demonstrated the power of naturalistic
observations when combined with creative thought. One could
also cite the achievements of paleontology and geology, both of
which greatly depend on historical analysis.
A methodological prescription is not being offered to the ef¬
fect that the historical method is either the appropriate or the
best method for social psychology. Evolutionary theory, as well
as paleontology and historical geology, has been supplemented in
significant ways by experimental efforts. It would be foolish to
suggest that at all times one specific empirical method should be
the only or the most fruitful method for social psychology. But it
would not be foolish to consider the relative accomplishments of
historical and experimental social psychology while attempting to
estimate their future value, especially for the near term.
To be specific, consider for a moment, the potential contribu¬
tions of the analysis of civil strife by the political scientist, Gurr
(1970), with the efforts of laboratory research to verify cognitive
dissonance theory. By collecting and analyzing data from a large
number of sources of civil strife in 114 countries and formulating
summary measures of the strife (e.g., pervasiveness, intensity,
duration) as well as of the characteristics of the societies (e.g.,
economic development, type of political system, geocultural re¬
gion), Gurr proposes a theory of civil strife that emphasizes the
inffuence of relative deprivation:
People become most intensely discontented when they cannot get

what they think they deserve, not just what they want in the ideal
sense; and when they feel that they are making inadequate pro¬
gress toward their goal, not whether they have actually attained
them or not. . . . Underlying the relative deprivation approach
to civil strife is the frustration-aggression mechanism, apparently
a fundamental part of our psychological makeup. When we feel

thwarted in an attempt to get something we want, we are likely
to become angry, and when we become angry the most satisfying
inherent response is to strike out at the source of frustration
(1970, p. 596).
The model identifies the general patterns of social conditions

that cause civil strife and hypothesizes about the possible influ¬
ences of minority—group status, political objectives, amount of
education, and other variables. If Gurr’s model were subjected to
the same empirical barrage as was the cognitive-dissonance for¬
mulation, ambiguities would inevitably appear. The history of the
frustration-aggression hypothesis has repeatedly demonstrated the
need for additional clarification. Today many would argue that a
social-learning theory of aggression, which does not postulate
frustration to be a causal factor, handles available evidence more
eflFectively (e.g., Christy, Gelfand, & Hartmann, 1971). The con¬
cept of relative deprivation, like that of cognitive dissonance, is
also burdened with difficult measurement problems.
One can suggest, however, that Gurr’s model—emerging di¬
rectly from data from a natural setting—in contrast to the cogni¬
tive-dissonance model, can be applied more easily and eflFectively
to the solution of a pressing social problem. Although the prob¬
lem of civil strife may not be any more fundamental to a general
theory of social psychology than is attitudinal consistency, it no
doubt represents a much more critical problem in contemporary
societies. If the explanatory limits that characterize black-box
learning theories apply with equal, or greater, force to theoretical
models in social psychology, perhaps the eflForts of social psy¬
chologists should be tilted more in the direction of contributing
to the solution of social problems. At this stage in its history social
psychology may be more eflFective in evaluating social programs
than in formulating abstract theories of social behavior. And for
its own good, as well as that of society, social psychology would
be better oflF doing what it can do best. The era of unqualified
support of research programs generated solely by a scientist’s cu¬
riosity is over. Governmental funding appears now to be guided
by the pragmatic value implicit in William James’s blunt, but elo-
quent, dictum: ‘'Truth is the cash value of an idea.” Within this

perspective, evaluation of social programs would represent a
much higher priority than would formulations of abstract theories
of social psychology that have extremely limited explanatory ca¬
pacities and range of application.
The recommendations that emerge from this analysis are
simple in their implications but complex in their implementa¬
tion. Society has many problems that demand solutions. Social
psychology possesses the technology to assist in this effort by vir¬
tue of its capacity to evaluate objectively the consequences of dif¬
ferent social treatments. In order to achieve this potential it be¬
comes necessary to reorient interest away from purely laboratory
phenomena to natural-occurring events. This does not necessarily
imply abandonment of the experimental for the historical method.
The two methods can be combined as was done in the study that
evaluated the negative income tax (page 294). The methods can
also be combined by relating studies in the natural setting to
those in the laboratory. However, one cannot assume that labora¬
tory studies automatically reflect processes operating in real-life
situations. Instead, such relationships must be demonstrated.
A naively romantic attitude that assumes research alone will
provide the necessary information for society to choose among
several competing social programs is doomed to disappointment.
The previous analysis of ethical commitments in interpreting the
values of a social program argues against such a simplistic notion.
But the conclusion was drawn that evaluation research could
make a major social contribution if value judgments were isolated
from facts and research findings were communicated in an un¬
biased fashion.
Two Unending Controversies:

Holism Versus Atomism
and Rationalism Versus Empirieism
Up to now, in an attempt to identify fundamental historical trends

that will shape the future of psychology, attention has been fo-
cused on substantive topics: biopsychology, developmental and

comparative psychology, and evaluation research. This analysis
will now be concerned with two methodological controversies
that have their roots in the history of philosophy instead of in re¬
search methods in psychology. These two disputes have had per¬
vasive influences in the history of psychology and have occupied,
in one form or another, the center stage of important theoretical
controversies. Undoubtedly they will persist because each posi¬
tion represents an important research and theoretical strategy. In
addition, the disputes are frequently involved with metaphysical
and ethical considerations, and as a result neither controversy
lends itself to any simple resolution. To understand their future
impact upon psychology requires clarification of the methodologi¬
cal issues involved in each debate.
Atomism Versus Holism
Without suggesting that the concepts of atomism and holism

are each limited to a single meaning, one can say that the task of
science includes both an atomistic and holistic component—to
analyze a phenomenon, or set of phenomena, into a basic set of
variables or elements (atomistic) and to synthesize such elements
into a valid interpretation (holism). From this descriptive per¬
spective these two components, atomistic and holistic, are neither
incompatible nor mutually exclusive. Instead they are comple¬
mentary because each serves an essential need of science.
Fundamental errors can be made with both analysis and
synthesis. A phenomenon can be incorrectly analyzed into errone¬
ous elements. William James accused the structuralists of com¬
mitting such an error, the psychologist’s fallacy,” when they re¬
ported discrete elements in conscious experience. These elements
(or processes), he argued, were not originally in conscious ex¬
perience but rather were put there by the structuralists’s precon¬
ceptions about the nature of consciousness. Similarly, early be-
haviorists accepted the atomistic assumption that the conditioned
response was the unit of complex behavior. Pavlov stated, and
Watson accepted, the position ‘‘that the different kinds of habits
based on training, education and discipline of any sort are noth¬

ing but a long chain of conditioned reflexes” (Pavlov, 1927, p. 395).
This view was adopted more as a matter of wishful thinking than
as a consequence of empirical evidence. It was based upon an
oversimplified conception of conditioning and a lack of knowl¬
edge about complex behavior. Of course, analytic attempts need
not be failures. The interpretation of Mach bands (Ratliff, 1965)
in terms of retinal inhibition is one of the notable examples of
success in psychology.
Two points should be stressed in evaluating the potential of
an atomistic approach that seeks to analyze a given phenomenon
into basic elements. First, success is not guaranteed. Analyzing
conscious experience into basic processes proved to be a failure
as a result both of apparently contradictory evidence and of meth¬
odological limitations of introspection. The theoretical analysis of
complex behavior into individual conditioned responses failed be¬
cause of its inability to explain complex forms of behavior such as
memory for meaningful material that depended upon hierarchical
forms of representation (e.g., Bower, 1970).
Although success is not guaranteed with an atomistic ap¬
proach, neither is failure preordained. One can not conclu¬
sively rule out, even though it appears highly improbable, that
some elementary model of phenomenology or conditioning could
achieve success where its predecessors failed. In other words,
specific analytical attempts fail but not the analytical method.
Consequently, a wholesale and a priori rejection of an atomistic
approach is unjustified. Holistic conceptions denying the rele¬
vance of an atomistic approach have surfaced in many forms in
the history of biology and psychology, often justified by the meth¬
odological thesis that the analytical dissecting methods of physics
and chemistry are inappropriate for the life sciences (e.g., Gold¬
stein, 1939). The accomplishments of Crick and Watson, who
analyzed genetic transmission into a code consisting of four basic
chemicals, should give pause to anyone considering the adoption
of a general antiatomistic approach in the life sciences.
Two general strategies can operate when attempting to syn¬
thesize the results of analysis. One is that the basic elements will
SS6 Psychology: A Science in Conflict
be sufficient to explain the operation of the entire system, the

other is that they will not. An example of the former, in which the
analysis alone provides the grist for the synthesis, is Boyle's law:
At a constant temperature the volume of a confined gas decreases
in inverse proportion to the pressure exerted on it. In this law the
component variables (temperature, volume, pressure) are suffi¬
cient to describe the entire system without reference to any holistic
principle. The dark adaptation curve of the human eye illustrates
another case in which the entire system functions in a manner
that can be described as ‘‘the whole is the sum of its parts.” The
curve of dark adaptation, which represents the weakest light a
subject can see after he has been in the ârk for various periods
of time, has two sections. In the first section, it drops steeply and
levels oflF after five minutes. The second section drops almost as
sharply at the beginning, then levels ojBF and drops very slowly
for more than thirty minutes. The two sections result from the
dffierent operating characteristics of rods and cones with the first
section representing cone adaptation and the second, rod adap¬
tation.
Sometimes the entire system does not appear to operate in
terms of only its component parts, i.e., the operative character¬
istics of the component parts do not appear sufficient to explain
the operation of the entire system. Perceptual phenomena are
often cited as illustrating the inadequacy of an atomistic ap¬
proach; what one perceives is not simply an aggregate of the
component elements but instead the entire configuration deter¬
mines how the parts will appear. Figure 8.1 is a pattern of four
dots that is perceived not only as four dots but also as a square.
Gestalt psychologists suggest that the perception of the square
illustrates the maxim: “The whole is more than the sum of its
parts”; the perception of the dots is altered as a function of the
configuration into which they enter.
Analytical terminology can also be employed to describe the
perceptual phenomena generated by Figure 8.1. The perception
of Figure 8.1 is a function of the four dots and the relationships
among them, i.e., the pattern of stimulation is analyzed into two
components, the individual dots and their positions in relation to
each other.
Future Trends in Psychology 3S7
Figure 8.1
Is one description—the holistic or atomistic—preferable to the

other? On a very fundamental level they are equivalent in that
they both identify the same set of variables, the four dots and the
interrelationships among them. However equivalent they may be,
the two interpretations can have markedly diflFerent influences on
methodological and theoretical decisions, even when the impor¬
tance of analysis is recognized in a holistic approach or the im¬
portance of synthesis in an atomistic orientation. For example.
Gestalt psychologists, who emphasized the significance of con¬
figurations in psychological events, made it clear that analysis
should be encouraged, as long as the intrinsic principles of the
whole were not ignored. Similarly structuralists and stimulus-
response psychologists, who expounded the virtues of analysis,
nevertheless stressed the need for synthesis and even employed
holistic language that suggested that parts were altered when em¬
bedded in a configuration. Titchener, when considering the prob-
SS8 Psychology: A Science in Conflict
lems of how the basic processes of consciousness combine to form

a unitary experience, concluded that the effects of interaction
were significant; the elementary processes '‘flow together, mix to¬
gether, overlapping, reinforcing, modifying or arresting one an¬
other, in obedience to certain psychological laws” (1899, p. 17).
Similarly, Hull postulated that individual stimuli when presented
together "interact ... in such a way that each receptor dis¬
charge changes all the others to a greater or lesser extent . . .
This type of action [afferent interaction] is particularly important
because the mediation of the responses of the organisms to dis¬
tinctive combinations or patterns of stimuli, rather than to the
components of the patterns, is presumably dependent upon it”
(1943, p. 385).
Regardless of how much emphasis each placed upon a holis¬
tic or atomistic approach. Gestalt psychologists, structuralists, and
stimulus-response psychologists acknowledged that the two orien¬
tations are complementary. However, the question can be raised
as to whether the sentiments expressed paid only lip service to a
scientific ideal rather than made a commitment to cope with the
intricacies of the atomistic or holistic task. Overemphasizing
either an atomistic or holistic approach tends to discourage a
theorist from dealing constructively with the problems of the
other orientation (analysis or synthesis).
Gestalt psychologists seemed for the most part satisfied to
demonstrate, as is the case for the perception of Figure 8.1, that
the whole is something other than just the sum of its parts. The
demonstration provided many with an intuitive understanding
(page 188) that the holistic interpretation was sufficient; further
analysis and research were not needed.
The eagerness to accept a holistic interpretation without ex¬
pending any effort to analyze the problem is nicely illustrated in
KohleFs (1938) research on transposition, the phenomenon in
which an organism transfers a relational solution from one dis¬
crimination problem to a similar one. Ghickens were trained to
peck for food at the lighter of two gray cards. After the subjects
had learned to consistently select the lighter gray they were
tested with a new pair containing the previously correct gray and
Future Trends in Psychology SS9
a still lighter one. If they had learned a speeifie habit to eaeh of

the gray eards during the initial diserimination, Kohler argued
that “it would be quite ineomprehensible” for the animals to
choose the lightest gray because they had been previously re¬
warded for selecting the gray that was now the darker one. The
results showed that a majority of choice (68 percent) was to
the lightest gray thus justifying a holistic interpretation that the
chickens had learned, not two separate habits, but instead a rela¬
tionship (e.g., select the lighter one) between the two stimuli.
A stronger commitment to analysis than Kohler had would
have encouraged certain questions. Why were 32 percent of the
responses to the previously rewarded stimulus, the darker one?
What would have occurred if the difference between the pair of
gray cards was greater or lesser than that employed by Kohler?
Is it truly impossible that a theory based upon the learning of in¬
dividual habits could explain Kohler’s findings? Spence (1937),
an S-R psychologist and therefore committed to an analytical ap¬
proach, entertained such questions and formulated an ingenious
conditioning model to explain transposition. His model was based
on the assumption that in Kohler’s experiment two specific habits
had been acquired; thus, his effort, if nothing else, contradicted
Kohler’s conclusion that transposition could not be interpreted in
terms of a simple habit model. Spence’s model did more in that
it explained the influence of the difference between training and
test stimuli. However Spence’s conception proved inadequate to
the task of interpreting all transposition data mainly because the
phenomenon proved to be much more complex than the initial
evidence suggested. The major point here is that the study of
transposition, as a case history in science, demonstrates how the
ready acceptance of a holistic position can discourage further
analytic attempts.
One can also point out that an extreme atomistic approach
can discourage holistic (relational) considerations. Even though
Titchener acknowledged that basic sensory experiences could
modify each other, he never attempted to work out the principles
of synthesis. His obsession with analysis prevented him from cop¬
ing with the possible interrelationships among basic sensory ex-
periences. Similarly, Hull’s use of afiFerent interaction to deal with

perceptual interaction eflFects, of the sort that interested Gestalt
psychologists, was at best a recognition of the problem, but it
accomplished nothing in contributing to its solution. Like Titche-
ner, Hull’s strong atomistic orientation hampered him in coping
with possible interaction effects.
Emergentism
A discussion of holism versus atomism cannot be com¬

pleted without some reference to emergentism—the assumption
that properties of wholes cannot be understood from properties
of their component parts. Gestalt psychologists essentially adopted
this position in their treatment of the phi-phenomenon. In
Wertheimer’s demonstration of apparent movement two lights
were successively projected through two slits, one vertical and
the other about 25 degrees from the vertical. When the time in¬
terval between the two lights was approximately 60 milliseconds,
the subject perceived the light move from one position to the
other. Because there was no physical movement it was concluded
that the perceived movement emerged from the perceptual con¬
figuration. The phi-phenomenon was an illustration of the funda¬
mental “formula” of Gestalt theory:
There are wholes, the behavior of which is not determined by

that of their individual elements but where the part-processes are
themselves determined by the intrinsic nature of the whole
(Wertheimer, 1938, p. 2).
This holistic interpretation cannot simply be justified by a

discrepancy between psychological and physical measurement,
i.e., apparent movement occurs in the absence of physical move¬
ment. Psychological and physical measurement need not be per¬
fectly matched because they reflect the operation of two different
systems.
Understanding the significance of the phi-phenomenon re¬
quires distinguishing between its negative and positive theoreti¬
cal implications. Too often this is not done, and as a consequence,
evidence that is embarassing to one theoretical position is auto¬

matically assumed to be in support of another.
The data of the phi-phenomenon are inconsistent with the
structuralist hypothesis that perceptions are combinations of ele¬
mentary sensations. If this were true how could movement be per¬
ceived in the phi-phenomenon when only stationarity is sensed?
Structuralism could not offer a satisfactory answer to this question
but that does not preclude a satisfactory analytical interpretation.
It should be remembered that the research strategy of structural¬
ism was first to discover the basic elements (or processes) of con¬
sciousness, then formulate principles that govern the synthesis of
basic elements, and finally to relate the analysis of experience to
physiological events. This research program foundered on the first
task because of the inability to formulate phenomenological cri¬
teria for designating an element of experience to be basic. The
failure of structuralism to provide a satisfactory account of the
phi-phenomenon did not stem directly from a failure of synthesis
but instead from insurmountable methodological problems of a
descriptive phenomenology (page 45). Assuming that a satis¬
factory method of treating phenomenological data could be devel¬
oped, the likelihood of a satisfactory atomistic account of the phi-
phenomenon cannot be ruled out. Certainly an analytical neuro¬
physiological theory would be within the realm of the possible.
The holistic interpretation of the phi-phenomenon based upon
“the intrinsic nature of the whole” suffers from both incomplete¬
ness and vagueness. It essentially states that the phi-phenomenon
does not require any explanation because it is irreducible to any
analytical interpretation. Should not an explanation, especially if
a deductive criterion is adopted, say something more? The no¬
tion that the whole determines the functioning of its parts is at
best obscure and at worst mystical. What is meant by the state¬
ment that the whole determines the functioning of its parts? Are
not wholes always parts of larger wholes? Can you explain the
behavior of a husband without considering the activities of the
wife? And should not the behavior of the children be considered?
His own parents? The extended family? Society as a whole?
Neighboring societies? The international situation? This line of
reasoning leads to the eonelusion that the world, or perhaps the

entire universe, determines the behavior of a husband. Although
some would consider this to be an unreasonable example to illus¬
trate the limitations of the holistic approach, it should be noted
that holists in the history of biology constantly expanded the size
of the system they thought necessary to consider. Initially it was
argued that the fundamental biological unit was not the cell but
the organism as a whole. Then it was argued that the organism
in relation to its surrounding environment was unanalyzable
(Phillips, 1976). This inherent tendency in holism to enlarge the
size of the unanalyzable system is at the basis of William James’s
conclusion that holism is an impossible orientation because, “ev¬
erything would have to be known before anything is known!”
Bertrand Russell takes a similar dim view: “If all knowledge was
knowledge of the universe as a whole, there would be no knowl¬
edge.”®
Holism, when it takes on an exaggerated opposition to atom¬
ism, can assume a mystical posture. This was done by Hans
Driesch (1908) who, in order to account for the whole determin¬
ing the functioning of its entire self, postulated the existence of a
vital spirit that guided the development of embryos. An extreme
holism that takes the position that a whole is irreducible backs
itself into a corner where it is confronted with a choice between
mysticism or self-denial. Either a mystical force like an entelechy
must be postulated to control the activities of the entire configura¬
tion, or the whole must be analyzed into specific entities and in¬
teractions that determine its operation. When the holistic commit¬
ment is excessive the tendency is to adopt a mystical force that
binds the whole into an unanalyzable configuration. Phillips
(1976) suggests that this tendency has operated in some concep¬
tions of humanistic psychology. Accepting Kurt Goldstein’s The
Organism (1939) as the precursor of humanistic psychology,
Phillips believes that the conception of a human being as a “func¬
tional whole” is a reasonable orientation when interpreted as em¬
phasizing interrelationships among various components of the
6. Both quotations are cited in Phillips (1976).

Future Trends in Psychology 84S
human organism. But, as exemplified in the following quotation,

when such a view becomes exaggerated enough to deny the rele¬
vance of the atomistic approach of the natural sciences, the holis¬
tic position crosses the boundary of science, however vague it
may be, and enters into the realm of transcendentalism:
Our procedure is rooted in a more profound eonvietion: this is a

conviction that a state of greater perfection can never he under¬
stood from that of less perfection, and that only the converse is
possible. It is very feasible to isolate parts from the whole, but a
perfeet whole can never be eomposed by synthesizing it from less
perfect parts (Goldstein, 1939, p. 515).
A Logical Consideration
Debates about holism and atomism tend to ignore an impor¬

tant logical problem. Can consciousness be reduced to a set of
neural events? The answer is “No” and “Yes” depending upon
what is meant by “reduced.” One cannot explain the experience
of happiness by deducing it directly from neurophysiological hy¬
potheses. The conclusion of a valid deduction cannot contain an
expression that does not appear in the premises (e.g.. All animals
are mortal, all humans are animals, therefore all humans are mor¬
tal ). However if a theory of consciousness were coordinated to a
theory of neurophysiological processes, then it would be possible
to explain happiness as a consequence of neurophysiological
events (e.g.. Happiness is an experience that results from x neuro¬
physiological condition, x neurophysiological conditions are pres¬
ent in individual A, therefore individual A is happy). Within this
context radical atomists delude themselves by believing that con¬
sciousness can simply be reduced to a set of neural events while
extreme holists deceive themselves by denying the possibility that
conscious experience can be analyzed in terms of neural events.
Summary
My methodological critique of holism may appear to be more

severe than my critique of atomism. However true that may be, it
would be a mistake to conclude that I believe holism represents

a greater danger than atomism or that in the history of psychol¬
ogy holism has done more damage. The major thesis advanced is
that analysis and synthesis are complementary activities in sci¬
ence, and the emphasis on one to the exclusion of the other will
misdirect theoretical efforts. A radical atomism can be blind to
significant interaction effects while an unbridled holism can be
convinced that analysis is an inappropriate procedure in the life
sciences.
It is tempting after dealing with the complexities of the rela¬
tionship between holism and atomism to conclude, as did Gestalt
psychologists, that analysis can be fruitful as long as the organi¬
zational principles of the whole are not ignored. Such advice says
much less than it appears to. At best, it encourages one to be
brilliantly prescient, a recommendation which one cannot dis¬
agree with but might experience difficulty achieving. At worst, it
deludes one into believing that there is a simple orderly proce¬
dure to cope with the complexities of the analytic-synthetic prob¬
lems of science. There is not.
Rationalism Versus Empiricism
Just as it is futile to argue about whether holism or atomism

is the appropriate orientation in psychology, so is it fruitless to
debate whether rationalism or empiricism is the proper frame¬
work for all of psychology or for any specific research problem.
Rationalism and empiricism will be discussed from both epis¬
temological and psychological perspectives. This does not imply
that epistemology is independent of psychology because episte¬
mology is a human enterprise subject to psychological analysis
(Piaget, 1978). But these two domains can be treated separately
if one makes the subtle distinction between the origin and ac¬
quisition of knowledge.
The epistemological viewpoint of empiricism assumes that
knowledge is derived from observation while rationalism sup¬
poses reason alone can yield truths about the world. If one ac¬
cepts the notion that scientific knowledge is based on observa-
tions, then empiricism is an essential component of the scientific

enterprise. But is it the only one? Not if the observations of in¬
dividual facts and empirical laws are viewed as isolated events
unrelated to each other. These events by themselves do not pro¬
vide a coherent picture of the world; that can only be achieved
by rational means. In other words, rational theory is required to
integrate and interpret empirical observations. Rationalism and
empiricism therefore supplement each other. This view is neatly
expressed by Giorgio de Santillana: “The true scientist has an em¬
piricist conscience and a rationalist imagination” (cited in Wil¬
liams, 1967).
The psychological meaning of rationalism and empiricism
has to do with an organism’s acquisition of knowledge. Is all of
our knowledge a result of our experiences (learning), or are we
born with innate ideas such as God exists or with a blueprint for
a universal grammar? The concept of innate ideas can be inter¬
preted in one of two ways. One, impossible to accept, is that a
neonate possesses a full-blow notion of an Almighty or a detailed
knowledge of a universal grammar appropriate for all languages.
Its linguistic and cognitive limitations prevent us from seriously en¬
tertaining such a hypothesis. The other possible meaning, difficult
to reject, is that the human organism is genetically prepro¬
grammed to develop neurophysiological structures that are essen¬
tial for the acquisition of certain knowledge and forms of behav¬
ior. Our possession of two eyes assists us in developing a veridical
concept of three-dimensional space; the structure of our genitalia
predisposes humans to certain forms of erotic behavior and cer¬
tain ideas about sex; the structure and functioning of our brain
enables us to form a concept of the number system and to respond
to certain sequences of sounds as music; and so on. There is
nothing that we do or think that in some way is not an expression
of our neurophysiological heritage. The theoretical issue is not
whether we have “innate ideas” in the sense of genetic prepro¬
gramming, but instead the nature of the influence of genetic pre¬
programming on a given form of behavior. There is no question
that in the history of American psychology a greater emphasis
has been placed on learning as contrasted with innate factors in
explaining most forms of behavior. The most extreme anti-innate

position was expressed by Watson in his boast that a child could
be selected at random and trained to be any kind of a specialist
(page 180). To a significant extent this overemphasis of the
power of learning was an expression of the political commitment
that genetic limitations need not obstruct the realization of equal-
itarian ideals. Watson ultimately rejected his extreme environ¬
mentalism, and later learning theorists acknowledged the signifi¬
cance of maturational limitations on the effectiveness of training.
For example, the concept of reading readiness implies that regard¬
less of how potent training procedures are, they are ineffective
for the child who has not yet reached an âppropriate maturational
level. Although not usually recognized as such the concept ex¬
presses a rationalist position in the rationalist versus empiricist
debate. Essentially it states that we are born with an innate no¬
tion about reading that begins to operate when a certain matura¬
tional level is reached.
Up to this point, my analysis of the rationalism versus em¬
piricism conflict raises no significant epistemological or empirical
issue. The scientific effort contains both an empirical and rational
component. From the psychological point of view the controversy
at best represents the degree of emphasis one should place on ge¬
netic and environmental influence. Differences of opinion con¬
cerning the relative importance of these factors or the interaction
between them can ultimately be investigated empirically.
Thus it should be possible to limit the differences between
rationalistic and empirically oriented conceptions to those that
could be empirically resolved. One notable exception would oc¬
cur if a rationalist strictly adhered to the epistemological tradi¬
tion, insisting that a rational interpretation of a phenomenon, in¬
dependent of empirical evidence, was sufficient. For the most
part psychologists have explicitly rejected this position even
though their efforts, at times, appear to be consistent with it.
Chomsky (1975), perhaps the most distinguished spokesman for
a rationalist position, insists that his own interpretation of lan¬
guage acquisition is as much subject to empirical test as is any
empiricist (learning) conception. In principle, such a claim can-
not be denied. However, in practice, an exaggerated commitment

to one side of the rationalism-empiricism controversy, as is the
case in the holism-atomism conflict, can create a climate of con¬
fusion that impedes resolutions of empirical questions.
Schlesinger (in press) argues that Chomsky unfairly demands
more empirical evidence to support the empiricist position than
he requires for his own rationalist assumptions. It is not the ra¬
tionalist assumption that innate factors are involved in language
acquisition that is the cause of the asymmetry, but instead Chom¬
sky’s implementation of this view that encourages his own posi¬
tion to be impregnable to empirical attack while simultaneously
creating a “straw-man” interpretation of the empiricist (learning
theory) approach. This is accomplished by insisting that in order
to explain human language acquisition it becomes necessary to
postulate innate “fixed and highly restricted schemata which
come into operation under limited conditions of exposure to
data” (Chomsky, 1975, p. 154). As a result, a system of gram¬
matical rules that represents knowledge of language develops
from these postulated “schemata.” This claim is not expressed
within any specific theory; it is still only “a program of research”
(Chomsky, 1968, p. 70). The trouble with it, Schlesinger argues,
is that it fails to yield any empirically refutable hypotheses.
Chomsky argues in favor of a universal grammar citing as an ex¬
ample the fact that English-speaking children unerringly make
use of structure-dependent grammatical rules that they have had
no opportunity to learn. Because it is inconceivable to think that
children are preprogramed to learn only English, these grammati¬
cal rules must be universal. But if a universal grammar is not
known, it becomes impossible to know whether any given lan¬
guage departs from its principles. And even if a set of languages
revealed a grammar common to all, would the discovery of an¬
other language that obeyed a discrepant principle deny the as¬
sumption of innate schemata? No, suggests Schlesinger. The
meaning of the innate schemata would most likely be expanded
so that the universal grammar would have an option to contain
principle A or its opposite.
Chomsky’s argument in favor of his position over the learning-
theory approach can be characterized as “negative theorizing.”

The assumption that language is innate, according to Chomsky,
can only be faulted by a satisfactory learning theory. In essence,
Chomsky’s position is that rationalism is validated by the failures
of empiricism. Whatever the inadequacies of learning theories,
and however unsuccessful they may or will be, their failures do
not confirm a rationalist position. Only a rationalist theory can
accomplish that. Schlesinger (in press) stresses this point in his
conclusion that, “Rationalism . . . becomes the last refuge for
those too tired to search any further.”
The methodological asymmetry between the demands of a
rationalist and empiricist approach is greatest in relation to the
demand for supporting evidence. A rationalist psychological the¬
ory tends to fall back on a rationalistic epistemological position
in order to support its position. In contrast, the empiricist ap¬
proach by its very nature encourages the search for supporting
empirical evidence although, admittedly, decades ago learning
theorists seemed to interpret language acquisition more in terms
of theoretical analogies than independent empirical evidence. But
now, primarily as a function of Chomsky’s challenges, more fruit¬
ful empirical methodologies have been developed that have forced
psycholinguists, even those trained in a rationalist tradition, to
buttress their theoretical notions with relevant evidence. And
now that the empirical realm of language acquisition is expand¬
ing (Schlesinger, in press), Chomsky’s extreme rationalist position
is withering in the face of evidence that suggests the importance
of experience in language acquisition.
It is instructive to note that Schlesinger does not adopt the
extreme learning-theory approach that encouraged Chomsky’s
original rationalistic position. Instead, he assumes an interaction-
ist position that encourages a program that would include three
areas of research:
A. Linguistic research with the aim of discovering generalizations
about the principles underlying human language (putative
linguistic universals).
B. Psychological research aiming at the discovery of other cogni¬
tive domains in which principles revealed in A operate.
C. Psychological research in which explanations are constructed

and tested of how the prineiples revealed in A develop
through interaction with the environment (ibid.).
The purpose of this discussion is neither to become involved

in the empirical issues of the rationalist-empiricist debate con¬
cerning language acquisition nor to defend Schlesinger’s concep¬
tion of a fruitful research program. The major point is that all
psychological activities are to some extent a function of complex
interactions between genetic preprogramming and environmental
influences. Because no simple or single answer is possible to this
general problem, the rationalist-empiricist controversy will inevi¬
tably persist. In order to prevent the debate from engulfing us in
confusion it becomes imperative to recognize the distinction be¬
tween its epistemological and empirical meanings and to become
aware of the possible methodological asymmetries that certain po¬
sitions, either rationalist or empiricist, can promote. Although the
present discussion of psycholinguistics noted the asymmetries that
a rationalist position can encourage, the reverse can operate
within an extreme environmentalism (page 252). In regard to
epistemology the greatest confusion will be created by those who
in self-delusion believe that they are operating as empiricists but
in reality are functioning as full-blown rationalists. In regard to
empiricism clarity will be enhanced by those who espouse the
significance of innate or environmental processes without deny¬
ing, or overlooking, the eflFect of complementary genetic or envi¬
ronmental factors.
Psychotherapy
One of the great mysteries in psychology is the popularity of psy¬

chotherapy, “talk therapy,” which aims to modify a person’s
behavior and experience by communication techniques. For the
most part, available evidence fails to demonstrate convincingly
the effectiveness of various forms of psychotherapy. Yet the sup¬
ply of patients, and even therapists, appears inexhaustible. In a
recent advanced undergraduate class half of my students were
involved in some form of psychotherapy, either as a patient or in

a co-counselor relationship in which a pair of students exchange
roles as counselor and client.
Why is psychotherapy so popular? If a certain form of be¬
havior is widely practiced some form of reinforcement is appar¬
ently operating. The absence of overwhelming evidence for psy¬
chotherapy’s effectiveness does not deny the possibility that the
patient is being reinforced in some manner for his participation.
Over the years I have asked both patients and therapists
about their evaluations of psychotherapy as a result of their per¬
sonal experiences. The following anecdotes highlight both the dif¬
ficulties of judging the effectiveness of psychotherapy and the va¬
riety of incentives that might be operating.
Many years ago at a cocktail party I met a man in his
mid-forties who was married to an heiress and therefore able
to pursue his personal interest in psychoanalysis. In our conversa¬
tion he mentioned that he had been in analysis for the past twelve
years. Surprised, and without any forethought, I exclaimed, “Any
improvement?” He responded matter-of-factly, “Oh, no, but I
would hate to think what I would be like if not for my analysis!”
Recently, in response to my query as to how a psychothera¬
pist knew whether her counseling was effective for her patients,
mostly male university students, the answer given was that prior
to the counselling most of the students were virgins, afterwards
none were. (The striking nature of the data left me so speechless
that I forgot to ask what specific therapeutic procedure was
employed.)
A successful and widely respected artist, on a well-known
talk show, readily admitted that his analysis had not modified his
neurotic behavior, but he was nevertheless convinced that it un¬
leashed his creative potential.
A distinguished psychologist concluded that his own psycho¬
analysis did not do any more for him than a good friend could
have done. His analyst retorted that it was interesting to note
that his famous patient did not have a good friend.
Finally, Woody Allen, in one of his early recordings of a
night club performance, questioned whether the fees to his psy-
Futuve Ttends in Psychology
351
choanalyst should be listed on his income tax as a medical ex¬

pense or a religious contribution. It can also be asked whether,
for some patients, psychotherapy should be considered a recrea-
tional expense?
The issue raised by these anecdotes is whether psychother¬
apy, in many cases, is truly a means to the end of achieving ‘psy¬
chological health” or is an end in itself. To be specific, one does
not take an antibiotic, have a hernia repaired, or receive treat¬
ment to recover from alcoholism for the same reason that one
goes to church, reads a book, or forms a friendship. The treat¬
ment itself IS not an integral part of living but instead is a means
y which one can live one s life in a desired fashion.
Back (1972) suggests that experiential groups (personal
growth groups, T-groups, sensitivity training groups, etc.) fre¬
quently serve as ends, not just means. He speculates that their
popularity is due to certain social forces unique to the United
States; Europeans have not been attracted to experiential groups
in any degree approaching that of the Americans. Back believes
experiential groups are popular in the United States because of a
combination of the mobility, affluence, and secularization of the
society. The mobility of American society frustrates many per¬
sons need for close personal relationships. Affluence provides
money and time to seek satisfactions that money by itself fails to
bring. The secularization of American society fails to provide the
opportunities for strong emotional and mystical experiences that
normally occur in a ritualistic setting. Furthermore, experiential
groups can be fun and can lead to fun relationships. And if the
person who wants to participate cannot accept a hedonistic inter¬
pretation of experiential groups, he can always justify them in
terms of the need for personal growth.
The same characterization can be applied to many cases of
one-on-one psychotherapy even to the extent of facilitating the
formation of close personal relationships with the therapist and a
community of others who undergo psychotherapy. Another fac¬
tor, which may be the most important, is that psychotherapy
clarifies the human condition. For many people Freud offered a
picture with an essential validity (p^^gc 175). However persua-
sive Freud’s conception is one should realize some people will be

more drawn to other conceptions of the human condition such as
those offered by Jung, Erikson, Maslow, Peris, or Werner Erhard.
Therapists have offered a variety of pictures of the human condi¬
tion varying in complexity, depth, and ethical commitments, and
they all have found some audience that consider their interpreta¬
tion to possess an ‘"essential validity.” The wide popularity of
psychotherapy may be due in part to the search of many for a
compelling picture of the human condition that they have been
unsuccessful in finding outside of the therapeutic situation.
Albee (1977) suggests that historical forces have elevated
the position of the psychotherapist to one of major social impor¬
tance in American society. The early growth of capitalism de¬
manded changes in human motivation involving both a repression
and control of pleasure seeking, and a positive evaluation of hard
work, thrift, long-range goals, plus a tough-minded rationalism.
These changes are still operative in upwardly mobile groups in
the United States and in less well-developed societies that are
making the transition from an agrarian-village to an industrial-
urban society.
This motivational pattern, similar to the Protestant ethic of
Max Weber (1904-5, 1958), has been attenuated in large sec¬
tions of the American middle-class. Albee suggests that this is
due, in part, to the economic needs of the capitalistic system,
which requires impulse buying and self-gratification to consume
the excessive amount of manufactured goods that are not really
needed. Several generations of people have already been raised
in this consumer-oriented society who believe that it is all right
to yield to impulse, to buy without guilt, and to consume with¬
out shame” (Albee, 1977).
With the deterioration of the Protestant ethic has also come a

deterioration in rationalism and science as well as faith in tradi¬
tional religious beliefs. Fifty years ago most people put their faith
either in religion or in science as the best source of explaining
away their existential anxieties and uncertainties. With both of
these systems now in jeopardy, people are desperate for new an¬
swers to the eternal existential questions. As a result, we are
Future Trends in Psychology S5S
witnessing the rise of the psychotherapist as the new shaman,

explanation giver, and guru. Psychology flourishes as the new
myth system (ibid., p. 151).
An exponent of psychotherapy could readily acknowledge

that Albee is a keen observer of the social scene and accept the
validity of the distinction between psychotherapy as a method of
healing and as an oracle for answers to eternal existential ques¬
tions and still conclude, ""So what?’" If psychotherapy serves two
important functions, so much the better! The popularity of psy¬
chotherapy demonstrates its capacity to satisfy these two needs
(some would argue they are really one). But even those who ac¬
cept the distinction and question the healing ejBFectiveness of psy¬
chotherapy as well as the propriety of providing spiritual guidance
would find it diflScult to restrict the practice of psychotherapy. To
deny a person’s freedom to receive psychotherapy as a form of
spiritual guidance might prove to be more difficult than to re¬
strict his opportunity to receive it as a form of treatment. In the
former case the right to a religious freedom is fundamental in
American society, and from the political perspective the right to
seek spiritual guidance from a psychotherapist is as justified as
receiving it from a priest, minister, or rabbi. In the case of psy¬
chotherapy as a means of healing one can suggest that psycho¬
therapists should be able to provide evidence to their patients
that their techniques can be effective. But even this demand
could be considered excessive because of a variety of factors not
the least of which is the absence of any clear-cut, universally ac¬
cepted criterion of mental health.
Psychotherapy can be perceived in another context, that of a
social need rather than that of an individual choice. Sooner or
later, the United States, like other Western democracies, will in¬
stitute some form of a national health program. Many profes¬
sional psychologists will argue that a national health program
should include psychotherapeutic assistance to those who need it.
But since the cost of this program is to be shared by the entire
population, questions will be raised about the effectiveness of
psychotherapy as practiced by psychiatrists, psychologists, and
other professionals, as well as by untrained amateurs such as
those who do co-counselling. Questions will also be raised about

whether society should foot the bill for a “demoralized individ-
uabs search for “spiritual guidance.” Should a government get di¬
rectly involved in a person s search for spiritual values with the
threat it poses to the separation of religion and government? Or
is “spiritual guidance” a “right” of an individual as is physical
health and therefore should be included in any national health
program? Even if the response to the second question is affirma¬
tive, the question will be raised as to what kind of guidance
should be offered: the “Protestant ethic,” human growth poten¬
tial, a simple hedonism, or some other form?
The questions posed have obvious' ethical components, but
they also involve empirical issues that demand clarification. The
efficacy of psychotherapy has been subjected to empirical evalua¬
tion for decades. The results have not been convincing, one way
or the other. Some would even maintain that the problem is too
complex for available techniques, an admission that would be
self-defeating if one favored the inclusion of psychotherapy in a
national health program.
The failure of past research to offer a definitive evaluation of
psychotherapy may reflect the effects of an inappropriate research
strategy. Initially the global questions of “Is psychotherapy effec¬
tive?” or “Which kind of psychotherapy is best? were asked.
With time, more specific issues have been raised as to the effects
of different variables (e.g., empathic ability of therapist, simi¬
larity between therapist and patient) and outcome measures
(e.g., phenomenological reports, job performance) without re¬
gard to the particular form of psychotherapy that is being prac¬
ticed. The increasing sophistication of research techniques plus
a stronger demand from a tax-conscious society for a cost-
effectiveness evaluation of governmental programs brings the
expectation that future research on psychotherapy will be more
informative than it has been in the past.
Two fundamental research topics, outcome measures and
cost-effectiveness, promise to attract the attention of investiga¬
tors in the future. These issues aie highlighted by the results of
a recent survey that sought to evaluate the outcome of psycho-
Future Trends in Psychology
therapy and counseling (Smith & Glass, 1977). The conclusion

drawn was that, ‘‘On the average the typical therapy client is bet¬
ter off than 75 percent of untreated individuals.” Different types
of therapy did not have much of an effect. Four major outcome
measures were employed: fear-anxiety reduction, self-esteem, ad¬
justment, and school/work achievement. Although these mea¬
sures were not strictly comparable, the findings gave “a credible
impression that fear and self-esteem are more susceptible to
change in therapy than are the relatively more serious behaviors
grouped under the categories adjustment’ and achievement.’ ”
Some psychologists would disagree as to what constitutes the
more serious” outcome measures, but the important point is that
the results suggest the interesting notion that psychotherapy is
more effective in changing the mental state of a patient, as mea¬
sured by personality inventories (e.g., sobriety, disruptive behav¬
ior, job performance), than in changing overt behavior. This
finding raises many interesting questions surrounding the rela¬
tionship between “subjective” and “objective” outcomes. Is it pos¬
sible that only the verbal descriptions of one’s inner experience are
being modified in psychotherapy, not the inner experience itself?
Would not neurophysiological measures be helpful in trying to
answer such a question? Investigating the relationships among
the various outcome measures is an essential step in the evalua¬
tion of psychotherapy.
Even if the general implications of Smith and Glass’s conclu¬
sions are accepted, the demonstration of psychotherapy’s effec¬
tiveness is not sufficient evidence for it to be included in a
national health program. Is it cost-effective? Gallo (1978) raises
an interesting point when considering this question. Estimating
that the average patient in the Smith and Glass study spent $510
for psychotherapy, a figure that would be considered a discount
in New York and Los Angeles areas, Gallo remarks:
Unfortunately, all we know in this case is the cost. Wê certainly

do not know how the patients’ adjustment would have been af¬
fected had they chosen to spend their $510 on a trip to Hawaii,
dancing lessons, a new wardrobe, plastic surgery, or a sail boat.
There are no data on alternative courses of action (1978, p. 516).
Gallo’s position may appear extreme because it suggests that re¬

search to discover the cost-effectiveness of psychotherapy could
be never-ending. If a trip to Hawaii is not as effective as psycho¬
therapy, how about Bora-Bora? However demanding Gallo may
appear the significant point is that when evaluating any thera¬
peutic techniques the consequences of other alternatives must be
considered. And the less effective a treatment is, the more likely
it is that other alternatives will be examined.
If psychotherapy proves more effective than competing pro¬
cedures, then the question will be raised as to how it can be most
economically employed. What level of training, if any, is re¬
quired? Are some individuals because of their personality capable
of being good therapists with a minimal training? Gannot much
of the need for psychotherapy be satisfied by teachers, ministers,
nurses, friends, and self-help programs? These questions may ap¬
pear unfair, imposing excessive demands upon psychology that
exceed those directed at other professions, like medicine. Al¬
though this objection may possess some validity it should be rec¬
ognized that the effectiveness of psychotherapy does not ap¬
proach that of many medical procedures. However unfair one
may feel such demands are, they nevertheless will be made, and
psychologists will be well advised to be prepared for them.
Some problems of measuring the cost effectiveness of psy¬
chotherapy in a national health program may be solved by
considering psychotherapy within the context of physical, not
psychological health (Olbrisch, 1977). A number of studies have
demonstrated the effectiveness of psychotherapy in hospital set¬
tings and the economic savings resulting from it. Patients under¬
going heart surgery appear to be greatly benefited from relaxa¬
tion therapy. A variety of surgical patients ranging from children
undergoing tonsillectomies to women anticipating gynecological
operations have been helped by psychotherapy. Some objective
outcome measures that have been used in judging the effective¬
ness of psychotherapy are the number of days required for post¬
operative care before discharge, the amount of pain medication
required, the incidence of postoperative psychosis, and the amount
of vomiting in the recovery room. In some of these studies, it
Future Trends in Psychology
must be noted, successful psychotherapy was carried out by

nurses and anesthesiologists.
The evidence (Olbrisch, 1977) that psychotherapy can meet
the criterion of cost eflFectiveness in a hospital setting suggests an
expansion of psychological services in the health sciences. Psy¬
chological, or what some would prefer to call behavioral, medi¬
cine could be widely practiced, depending upon the development
of new procedures and their empirical validation.
The question has been raised whether a cost-conscious so¬
ciety will be willing to pay for psychological services that clearly
do not pass some stringent criterion of cost eflFectiveness. This
query was directed at psychotherapy that was mainly concerned
with spiritual guidance and achieving “inner serenity” in the ab¬
sence of significant behavior change. However, at the same time,
the suggested eflFectiveness of psychotherapeutic procedures in
surgical cases raises the question not of whether national health
service can aflFord a psychological program but whether it can af¬
ford not to have one.
A major problem in a national health program would be cre¬
ated by individuals who overutilize services and facilities with¬
out any apparent medical disability. This would add to excessive
costs while simultaneously diminishing the eflFectiveness of the en¬
tire program (Cummings, 1977). The early identification of these
overutilizers and their eflFective treatment would be an enormous
contribution to a national health program. Similarly, solving prob¬
lems created by such prevalent disorders as alcoholism and hy¬
pertension would also reduce the drain on limited resources and
thereby help elevate the level of service. All of these problems,
and others of equal importance, have not escaped the attention
of psychologists. Challenges imposed by a national health pro¬
gram may initiate research programs that will yield more eflFec-
tive techniques of coping with these problems than have occurred
in the past. A national health program will confront psychologists
with an opportunity to convince society, and perhaps even them¬
selves, that they have an important role to play in the health
sciences.
The Information-Processing Paradigm
If we use Kuhn’s eoneept of paradigm loosely, and some (Mas-

terman, 1970) would insist that it has never been used in any
other way, we ean discuss a recent movement in psychology as a
single force although in reality it possesses much less unity than
other popular paradigms of the past, e.g., Freudian psychoanalysis,
Hull-Spence neobehaviorism, or cognitive dissonance theory. Since
the historical process that produced the information-processing
paradigm has yet to be completed it ^.becomes difficult, if not
impossible, to project its force and direction into psychology s
future. Historical analysis should not begin until history is com¬
pleted. At the same time any discussion of future trends in psy¬
chology would be incomplete if reference to the information¬
processing paradigm were not made. Having observed the rise
and fall of several paradigms in psychology from 1938 on, I
am impressed with the unrivalled success of the information¬
processing orientation in capturing the minds and hearts of such
a large percentage of investigators in the general research area
that used to be referred to as learning, memory, and problem
solving.
The information-processing paradigm emerged from a con¬
cern with cognitive psychology, the domain of which is primarily
man’s higher mental processes (Lachman, Lachman, & Butter¬
field, 1978). Many historical forces operated to initiate and sus¬
tain the revolution in which the information-processing paradigm
overthrew, in the sense of capturing a greater following, two
dominant behavioristic orientations: operant behaviorism, which
was initiated by Skinner, and neobehaviorism, which owes its
methodological and theoretical stance mainly to Hull and Spence.
Neobehaviorism had reached its apogee of popularity during
the 1950s and soon thereafter it became apparent to many, in¬
cluding its proponents, that Hull’s optimistic notion of achieving
a general theory based upon principles of conditioning, would
not be realized. Spence, in particular, acknowledged that theories
of animal learning would not suffice to explain human behavior,

repeatedly emphasizing that the boundary conditions of his dis¬
crimination learning theory (1936) excluded articulate human or¬
ganisms. In addition, when suggesting how a neobehavioristic ori¬
entation could handle human perception and cognition, Spence
(1950) explicitly acknowledged the necessity for postulating pro¬
cesses not required for the interpretation of infrahuman behavior.
This acknowledgement, in no way, undermined the strategy of
formulating a theory of animal behavior based upon principles
of conditioning. The absence of inconsistency was based upon
both strategic and theoretical considerations. If one cannot formu¬
late a satisfactory theory of the behavior of a rat how could one
possibly hope to explain human behavior? Because of evolution¬
ary considerations would not a theory of human behavior be as¬
sisted by knowledge of animal behavior? Should not the formula¬
tion of a theory of rat behavior provide methodological insights
into the construction of a conception of a human behavior?
Although one need not consider that affirmative answers are de¬
manded, it is difficult to conclude that such answers are un¬
reasonable.
Within the neobehaviorist’s camp attempts were made to
combine assumptions from conditioning theory with new prin¬
ciples and extend them to cognitive behavior (e.g., Osgood, 1957;
H. H. Kendler & Kendler, 1962). In certain ways these attempts,
which were characterized as mediational stimulus-response mod¬
els, represented a transition from association theory to informa¬
tion-processing formulations since they postulated theoretical pro¬
cesses (e.g., the transformation of incoming stimuli into symbolic
representations) that intervened between stimulus (input) and
response (output) and served as mechanisms for rational, intel¬
ligent behavior. But for the most part this evolutionary change in
neobehaviorism to account for human cognitive behavior was
ignored, discounted, or flatly rejected by those who adopted
the information-processing paradigm. Members of revolutionary
movements in psychology tend to be sustained more by what
they believe is wrong in a competing formulation rather than
what they have proved to be right in their own formulation. This
need for a revolutionary identity clearly distinct from all other

movements led many of those committed to an information¬
processing approach to ignore diflferences among behaviorisms
and to treat the entire behaviorist orientation as being equivalent
to operant behaviorism, the form most opposed to their own
views.
Of the various efForts to identify the historical ancestry of the
information-processing paradigm, the eflForts of Lachman, Lach-
man, and Butterfield (1978) appear the most successful although
they too exhibit the tendency to conflate operant behaviorism
with neobehaviorism. They cite six major influences on the devel¬
opment of the information-processing paradigm: neobehaviorism,
verbal learning, human engineering, communication engineering,
computer science, and linguistics. In each case the information¬
processing paradigm adopted some of the positions of its intellec¬
tual ancestors while rejecting others. For example, in discussing
the influence of neobehaviorism, Lachman, Lachman, and Butter¬
field conclude that the information-processing paradigm adopted
their positive attitudes toward nomothetic explanation, empiri¬
cism, laboratory experimentation, and operationism to describe
empirical investigations, but rejected neobehaviorism’s commit¬
ment to an all-encompassing theory, explanation by extrapola¬
tion, animal experimentation, a primary concern with learning,
the concept of conditioned associations, an adherence to logical
positivism, and an antimentalism. This simple listing highlights
the complexity of the historical background of the information¬
processing viewpoint by emphasizing the diversity of its origins
as well as the intricacies of the influences that flow from each in¬
tellectual ancestor. Because of this historical complexity and be¬
cause of the rejection of the ideal of an overall global theory, the
information-processing paradigm was not as integrated as its pre¬
decessors, thus allowing for modifications of strategic decisions
and theoretical assumptions in light of their effects. In contrast,
operant behaviorism and neobehaviorism were locked into posi¬
tions that were diflScult to abandon regardless of the consequences.
A price was paid for the great flexibility of the information¬
processing paradigm. The absence of a clear-cut theoretical goal
to encompass a circumscribed empirical realm plus the ease with

which the information-processing conception could be applied to
practically all forms of behavior led to a gigantic mushrooming
of research findings that lacked any general coherence aside from
the terminology that was employed to report and interpret them.
To this diversity of content has been added a range of explana¬
tory standards that varies from rigorous mathematical deduction
to intuitive interpretations represented solely by a flow chart.
The core concepts of the information-processing model have
also tended to become reified. When these concepts are used to
describe a phenomenon a belief is encouraged that an explana¬
tion has been offered where in fact only a pretheoretical model
has been suggested. These models are theoretical trial balloons
that tentatively interpret available evidence and guide future re¬
search. Pretheoretical models are a common gambit in psycho¬
logical theorizing and serve an essential purpose. However, there
are dangers, as the history of psychology testifies, when they are
judged to be more than a “pilot study” in theorizing. Hull (1935),
for example, offered a pretheoretical model that illustrated how
novel combinations of stimulus-response associations could ex¬
plain reasoning in animals. Could is not, however, synonymous
with would. Hull perceived Maier’s finding (1929) that rats
could reason as a challenge to his stimulus-response association-
ism. He met this challenge by offering a persuasive associationistic
interpretation. His overconcern with theory at the expense of data
backfired. The phenomenon of reasoning was not as simple as
Maier originally reported or as Hull thought (T. S. Kendler, &
Kendler, 1967). What appeared to Hull, and his many adherents,
as an incorporation of the facts of reasoning into a stimulus-
response theory proved to be an ingenious pretheoretical model
but nothing more.
The general strategy of forming pretheoretical models to su¬
perimpose on a relatively unexplored empirical area cannot prop¬
erly be criticized. Theoretical conjectures are the life blood of a
science. But a problem can be created by such a strategy when
the model is perceived more as a valid interpretation than as an
interesting speculation. Because the information-processing para-
digm has such theoretical flexibility and empirical breadth, as

well as a rich and persuasive terminology, the border between
intuitive understanding and deductive explanation can, and has
(Kendler, 1978), become badly blurred. This blurring has been
encouraged by the greater devotion of some adherents of the
information-processing paradigm to model building than to col¬
lecting relevant facts.
The explanatory capacities of information-processing concep¬
tions based upon the operation of computers have been a special
target for criticism. These models essentially liken human cogni¬
tive operations to functions in computer programs. One of the
most vociferous critics of computer simulation of human cogni¬
tion is Joseph Weizenbaum (1976), a distinguished computer
scientist. Two of his major criticisms are (1) the operation of a
computer is a poor base from which to infer the intelleetual func¬
tioning of humans, and (2) the claims of computer models to ex¬
plain human behavior are inflated.
The first criticism essentially states that humans and com¬
puters should not be viewed as members of the same genus
since different principles govern their functioning. This kind of
criticism is equivalent to the reservations expressed about inter¬
preting the operation of the central nervous system as a telephone
switchboard. Misconceptions, the argument goes, will inevitably
spring from an inappropriate analogy. This criticism in no way
questions the general value of computers as research tools and
the specific contribution that can be made by translating a com¬
plex theory model into a computer program in order to discover
the theoretical implications of the model. The thrust of the criti¬
cism is that humans and computers are constructed differently
and therefore operate on different principles.
This criticism is rejected by computer enthusiasts who argue
that the analogy is justified, at least tentatively, because both the
computer program and the human thought process essentially are
expressions of a set of symbol manipulating rules. If we discover
that the output of a computer program matches that of a human
in a problem-solving task can we not conclude that both are em¬
ploying the same set of rules? The answer is obviously “No,” but
at the same time the possibility that it could be true in a given

case cannot be rejected. The problem is not unlike that of discov¬
ering whether two individuals arriving in New York from Boston
reached their destination in the same way. Both could have come
by car but they need not have.
The question of whether the same set of rules underlie the
similarity between the output of a computer program and human
performance in a cognitive task is obviously an empirical and
theoretical problem that cannot be solved by a priori arguments,
no matter how persuasive they may appear. In considering this
task certain problems need be recognized.
First, the concept of matching outputs is not as clear as it ini¬
tially appears. One of the early simulations of human cognition
attempted to model human proofs of theorems in symbolic logic
(Newell & Simon, 1961). Although some success was achieved in
matching the introspective reports of subjects with that of the
computer output, obvious discrepancies occurred between the
language of the subject, which was idiomatic English (e.g., “Tm
looking at the idea of reversing these two things now”) and the
telegraphic, logical communication of the computer (e.g.. Goal
6: Apply R3 to LI). In addition, the computer sometimes pro¬
vided more information than the human subject and at other
times less. If matching computer outputs to human performance
serves as the litmus test to evaluate the computer simulation how
are these discrepancies to be evaluated? What is the dividing line
between a successful and unsuccessful match? This question can
be bypassed by those who seek to understand human cognition
by computer simulation because their only concern is to identify
the best simulation of a given cognitive task and to improve upon
it in the future. But if one adopts a questioning attitude about the
general approach of computer simulation of human cognition
then the issue of defining an acceptable match between computer
and human outputs cannot be ignored.
The problem of specifying the criteria of a reasonable match
between humans and computers becomes excessively complex
when ordinary language serves as the output. Ideas can be ex¬
pressed in an infinite variety of forms in common language. Ma-
chine translation of one language into another has not been as

effective as initially antieipated because of the diffieulties eomput-
ers have with dealing with idiomatie expressions. An amusing
example that highlights this problem comes from a program de¬
signed to translate English into Russian (Laehman, Lachman,
& Butterfield, 1978). The English sentenee, ‘'The spirit is willing
but the flesh is weak” was translated into Russian essentially as,
“The vodka is fine, but the meat is tasteless.” Perhaps it would be
strategic for computer simulation of intelleetual tasks to restriet
their outputs to highly struetured behavioral measures such as
choice responses, reaetion times, and so on.
The seeond problem involved in the effort to simulate eogni-
tive behavior by eomputer programs is the extreme speeifieity of
the interpretations. The program applies to a single empirieal re¬
sult, or at best, a limited set of related findings. This speeifieity
eneourages the conelusion, no matter how suceessful the simula¬
tion is, that the eomputer operations do not neeessarily refleet un¬
derlying psyehologieal proeesses. Sueh a conelusion can be best
refuted by demonstrating a general explanatory eapacity for a
given eomputer model. If a program that simulates human per-
formanee in problem A ean be suceessfully applied to problem B,
then the eonfidenee in the validity of the computer interpretation
of problem A, as well as for B inereases.^ The more phenomena
the general program ean explain the more eompelling will the po¬
sition be that the similarity obtained between the outcomes of the
eomputer program and human performanee is an expression of
eommon underlying proeesses. In other words, eomputer simula¬
tion of human eognition must be treated as a theoretical endeavor
and judged in the same way as any other theoretieal effort.
7. The similarity of an epistemological criterion of a general theoretical

explanation and a psychological criterion of understanding should be noted.
Wertheimer (1945) insisted that a child’s ability to discover the area of a
particular rectangle did not necessarily imply that he understood the prob¬
lem. However, if the child was able to indicate how the area of a parallelo¬
gram was to be computed, then one could have confidence that he really
understood the area of a rectangle. In sum, a transfer criterion is appropriate
for both theoretical and psychological conceptions of understanding.
The extreme speeifieity of eomputer simulation programs en¬

couraged Simon and Newell (1964, 1971) to formulate a program
known as the General Problem Solver to cope with three major
cognitive tasks: chess playing, discovering proofs in logic, and
solving cryptarithmetic puzzles (e.g., discover numerical equiva¬
lent of letters in words when DONALD + GERALD = ROBERT
where D = 5). The General Problem Solver contained pro¬
grammed instructions that are assumed to be equivalent to com¬
mon problem-solving processes that guide human thought. The
General Problem Solver operates serially, coping with one major
process at a time. For example, one process is to organize the
problem-solving task into a set of subgoals that ultimately will
lead to problem solution. Once a particular subgoal is selected
then a means-end analysis begins to operate that specifies the dif¬
ference between the existing situation (e.g., the position in a
chess game at the end of the sixth move) and the desired situa¬
tion (e.g., the control of the four center squares) so that the rele¬
vant heuristic procedure can be employed to reach the subgoal.
The solution is not guaranteed; it may prove impossible to gain
control of the center of the board from a particular position. If
that be the case, a new subgoal is selected (e.g., secure the safety
of one’s king) and a means-end analysis of that subgoal is begun.
Proponents of the strategy of simulating human thought with
general conceptions, such as the General Problem Solver, that
contain no reference to the content of specific tasks viewed their
efforts as a major breakthrough in cognitive psychology. Twenty
years ago these efforts were considered to “have already produced
a rigorous, detailed explanation of a significant area of human
symbolic behavior” (Newell & Simon, 1961). Although much un¬
finished business still had to be completed, an optimism reigned
that the goal of a general theory of human thought and problem
solving would inevitably be achieved in the near future. This op¬
timism proved to be unjustified. Nevertheless, the computer simu¬
lation research program, in spite of its failure to achieve its stated
goal, is nevertheless defended as being more successful than com¬
peting research programs (e.g., phenomenological and behavior¬
istic interpretations). However, even this defense can be called
into question by raising the issue of whether the General Problem

Solver is really a theory.
Weizenbaum (1976) argues that the General Problem Solver
is not really a general theory but instead ‘‘is nothing more than a
programming language in which it is possible to write programs
for certain highly specialized tasks.’' That is, the GPS does not
have any independent status that enables it, in some logical fash¬
ion, to generate specific programs for the different cognitive tasks.
Gannot, however, the general theory be abstracted from the pro¬
grams that are employed for the various cognitive tasks? Smith is
skeptical:
%
Anyone who has tried to read a theory embedded in a program

knows how difficult it is to arrive at the general principles of the
theory. More likely than not, whatever general principles there
are, are buried in the mass of details needed to make the theory
sufficient, i.e., to make the program run. More generally, as the¬
ories become increasingly sufficient they must by nature contain
more details, and consequently become less transparent (1978,
p. 49).
A theory expressed in the form of a computer program.

Smith notes, may be so inaccessible that it becomes impossible to
know what constitutes inconsistent evidence. If this be the case
then some method must be developed to extract the major theo¬
retical assumptions of a computer theory from the coordinated
procedures that are needed to make the program sufficient. Per¬
haps this will prove difficult or impossible, or new epistemologi¬
cal procedures will be developed to evaluate general theories em¬
bedded in computer programs. For the moment, at least, the
capacity of computer theories to interpret a comprehensive range
of human cognition has yet to be demonstrated.
The problem of theoretical specificity versus generality is not
limited to the segment of the information-processing approach
that employs simulation techniques. It appears to be a core prob¬
lem within the entire information-processing approach. The goal
of an all-encompassing global theory was initially rejected by the
information-processing paradigm for apparently good reasons
(Lachman, Lachman, & Butterfield, 1978). Psychologists were

notoriously unsuccessful in formulating comprehensive theories
of behavior, and it appeared sensible to limit theorizing to cir¬
cumscribed empirical areas until a level of success was achieved
to justify a broader conception. Although the impression prevails
that the information-processing approach has generated a com¬
prehensive interpretation of cognitive behavior, the actual theo¬
retical achievements deny it. A common paradigmatic language
is not equivalent to a core set of assumptions with a wide range
of implications. In essence, the information-processing concep¬
tions are not strictly theories in the sense that they coherently
organize a factual domain but instead are ‘‘experiments in con¬
ceptualization. Their authors are stretching the limits of the in-
formation-processing paradigm to see if it provides the concep¬
tual tools with which to attack the timeless questions humankind
has raised about its own mental functioning” (Lachman, Lach¬
man, & Butterfield, 1978).
The work of Kintsch and Van Dijk (1978), modestly entitled
“Toward a model of text comprehension and production,” is an
example of a conceptual experiment, or what has been described
as a pretheory that aims to interpret text comprehension and pro¬
duction. In concluding their article they acknowledge serious lim¬
itations and omissions: “It deals only with semantic representa¬
tion not with the text itself,” “The model stops short of full
comprehension,” “The component that interprets clusters of prop¬
ositions as facts . . . is . . . missing,” “General world knowledge
organized in terms of frames ... is another missing link.”
Acknowledging, “these serious limitations” Kintsch and Van
Dijk justify their efforts by the conviction that as long as “com¬
prehension” is considered “as one undifferentiated process . . .
it is simply impossible to formulate precise, researchable prob¬
lems.” In sum, specific hypotheses that would generate research
are needed.
One can fault neither the efforts of Kintsch and Van Dijk
(1978) nor the general strategy that they have adopted. But their
“conceptual experiment” does highlight the problem of theoretical
breadth currently confronting the information-processing para-
digm. In light of the obvious failures of global theories in the past,

it appeared strategic to eschew general theories of information
processing. As a consequence of this strategy a plethora of specific
models were created that threw little light upon cognitive func¬
tioning in any complex real-life task because of the absence of
any general conception as to how the specific processes and struc¬
tures were integrated. When the limitations of these specific mod¬
els became apparent, the conceptual tools within the information¬
processing paradigm, mainly a large set of assumed processes
and mental structures that appeared to be intuitively valid, al¬
lowed for the ready formulation of general theories that opened
up a ‘wealth of interesting significant research problems” and
made up “in promise for what it [lacked] ... in completeness”
(Kintsch & Van Dijk, 1978).
The key problem is whether the promises will be realized. Is
a general information-processing theory possible, or will the cog¬
nitive approach suffer the fate of predecessors that attempted to
formulate a natural-science type theory for a broad range of be¬
havior? One factor that dampens the optimism is the inability of
research within the information-processing paradigm to resolve
fundamental issues such as whether information is processed in a
serial or parallel fashion. This difficulty might appear insignificant
because a broad theory of cognition could be formulated that per¬
mits some assumptions to remain indeterminate. If physicists
need not resolve the issue of whether light should be represented
as waves or corpuscles, why must psychologists be forced to
choose between serial and parallel processing? Although such a
defense may provide some comfort, the difference between the
explanatory capacities of the physical theory of light and of gen¬
eral information-processing models destroys the appropriateness
of the analogy. The fundamental issue is whether there are ceil¬
ings to the explanatory powers of black-box theories, regardless
of the research program in which they are embedded. Although
information-processing conceptions have provided broader in¬
sights into human memory than previous associationistic theories,
their potential, like all environmental-behavioral conceptions, are
severely limited by the lack of direct control over the events and
processes that are basic to the phenomena that are to be ex¬

plained. The ingenuity of human reasoning to spin out a variety
of pretheoretical models combined with this lack of control effec¬
tively prevents a choice between competing assumptions. Conse¬
quently fundamental theoretical problems remain unresolved and
broad theories unattainable.
If the possibility of an explanatory limitation is acknowl¬
edged, then alternatives to broad theories of cognition based on
environmental-behavioral laws should be considered. One, al¬
ready suggested (page 316), is that the unresolved theoretical
disputes be transformed into neurophysiological problems. If this
can be accomplished then the more direct control that is exerted
over the phenomenon increases the likelihood of resolving the
theoretical issue. It does not follow that all theoretical disputes
can be recast into neurophysiological questions. Before this can
be accomplished additional knowledge about neurophysiology, or
even environmental-behavioral events, will be required. In some
cases the transformation may never be accomplished. The impor¬
tant methodological point is that black-box theorists need to re¬
consider the optimistic notion that their methods will lead ulti¬
mately to broad-range theories of the sort that have been achieved
in physics and biology. By acknowledging the possibility of in¬
trinsic limitations they can cope more productively with the type
of theoretical debates that in the past have generated much re¬
search but not theoretical determinate outcomes. By recognizing
theoretical limits one may be in a better position to identify when
a particular controversy has reached a point where additional re¬
search will prove fruitless in resolving the theoretical issues.
Rather than embark on a wild-goose chase that ultimately leads
to discarding theoretical issues on the scrap-heap of history, it
would be wise to put aside the problem until further knowledge
and advanced technology provide the means by which the ques¬
tions can be answered. Perhaps, if this strategy were followed sig¬
nificant theoretical issues and related research would not be so
rapidly forgotten from one generation to the next.
What does the future hold for the information-processing
paradigm? Will it be more successful than its predecessors in the
arena of natural-science behavioral theories? Yes, if success is

judged in terms of longevity. The life span of the information¬
processing conception will most likely exceed those of other para¬
digms because of its greater theoretical flexibility, its capacity to
generate a wide range of questions, and the availability of an ap¬
parently infinite number of experimental tasks to answer them.
Longevity, however, is not the only criterion that can be applied
to behavioral theories, and certainly not the most important. Can
the information-processing paradigm produce a coherent theory
of cognition or even of the basic subareas of cognition such as
memory or language? The history of black-box theories encour¬
ages a pessimistic view that broad-gauged theories of informa¬
tion processing, especially those that do justice to the capacities
of human intelligence in real life, will achieve any compelling
success. And by implication, the hope of extending the informa¬
tion-processing conception to other fields of psychology (e.g., per¬
sonality, psychopathology, motivation) appears at best remote.
Some may convince themselves that information processing is at
the core of all human activities; it is questionable that such an as¬
sumption, even in its broadest sense, is sufiicient to account for
affective and motivational events, as well as personality develop¬
ment.
Anticipating possible difficulties that the information-process¬
ing conception might encounter can serve constructive purposes.
In this attempt to read the future, however, one should not forget
that information processing, in the final analysis, “is a paradigm,
an approach to constructing theories, a style of theorizing. It can¬
not be correct or incorrect, only more or less productive” (Hayes,
1978). A final evaluation will depend both on the productivity of
the information-processing paradigm and competing paradigms.
Postscript
The unity of psychology has all but collapsed. Psychology is a

multidisciplinary field with different segments employing irrecon¬
cilable orientations. As a result bitter disputes have occurred con¬
cerning the proper methodological position that psychology should
adopt. Inevitably, these disputes have spilled over into controver¬
sies about the appropriate professional roles of psychologists.
These differences are unavoidable considering the fundamental
nature of psychology. A choice of competing methodological al¬
ternatives cannot be made by purely rational means although so¬
ciety may encourage one form of psychology at the expense of
others because of the manner in which society interprets its social
responsibilities. The best that can be hoped for within psychology
is a mutual understanding of the competing methodological posi¬
tions and an appreciation of the decisions that led to their adop¬
tion. The purpose of this inquiry has been to achieve that goal.
S71
References
Abel, T. The operation called verstehen. American Journal of Sociologu,

1948, 54, 211-218.
Albee, G, The protestant ethic, sex, and psychotherapy. American Psycholo¬
gist, 1977, 32, 150-161.
Alper, S., Beckwith, J., Chorover, S. L., Hunt, J., Inouye, H., Judd, T.,
Lange, R. V., & Sternberg, P. The implications of sociobiology.
Science, 1976, 192, 424-427.
Anglin, J. M. Word, object, and conceptual development. New York: Nor¬
ton, 1977.
Asch, S. E. Studies of independence and conformity: I. A minority of one
against a unanimous majority. Psychological Monographs, 1956, 70,
No. 9 (Whole No. 416).
Attneave, F. Some informational aspects of visual perception. Psychological
Review, 1954, 61, 183-193.
Atkinson, R. C. Reflections on psychology’s past and concerns about its
future. American Psychologist, 1977, 32, 205-210.
Back, K. W. Beyond words. The story of sensitivity training and the en¬
counter movement. New York: Basic Books, 1972.
Baker, J. J. W., & Allen, G. E. The study of biology (2nd ed.). Reading,
Mass.: Addison-Wesley, 1971.
Beach, F. A. Evolutionary changes in the physiological control of mating
behavior in mammals. Psychological Review, 1947, 54, 279-315.
Bellow, S. Humboldt’s gift. New York: Viking, 1975.
Bergmann, G. Psychoanalysis and experimental psychology: A review from
the standpoint of scientific empiricism. Mind, 1943, 52, 122-140.
Bergmann, G. Sense and nonsense in operationism. Scientific Monthly, 1954,
79, 210-214.
Bergmann, G. The contributions of John B. Watson. Psychological Review,
1956, 63, 265-276.
Bergmann, G., & Spence K. W. Operationism and theory construction.
Psychological Review, 1941, 48, 1-14.
Berkeley, G. The principles of human knowledge. Dublin: Jeremy Pepyat,
1710.
372
References 37S
Berkowitz, L, Social Motivation. In G, Lindzey & E. Aronson (Eds.), The

Handbook of Social Psychology (Vol. 3, 2nd ed.). Reading, Mass.:
Addison-Wesley, 1969.
Berkowitz, L. The contagion of violence: An S-R mediational analysis of
some effects of observed aggression. In W. J. Arnold & M. M. Page
(Eds.), Nebraska Symposium on Motivation. Lincoln: University of
Nebraska Press, 1970.
Berlyne, D. E. Humanistic psychology as a protest movement. Paper pre¬
pared for the 4th Banff Conference on Theoretical Psychology. Con¬
ceptual Issues in Humanistic Psychology. Centre for Advanced Study
in Theoretical Psychology, Banff, Alberta, October 12-16, 1975.
Bernstein, L. The unanswered question: Six talks at Harvard. Cambridge:
Harvard University Press, 1976.
Blodgett, H. C. The effect of the introduction of reward upon the maze per¬
formance of rats. University of California Publications in Psychology,
1929, 4, 113-134.
Boring, E. G. A history of experimental psychology. New York: Appleton-
Century, 1929.
Boring, E. G. A history of experimental psychology (2nd ed.). Englewood
Cliffs, N.J.: Prentice-Hall, 1950.
Bousfield, W. A. The occurrence of clustering in the recall of randomly
arranged associates. Journal of General Psychology, 1953, 49, 229-
240.
Bower, G. H. Organizational factors in memory. Cognitive Psychology, 1970,
1, 18-46.
Breland, K., & Breland, M. The misbehavior of organisms. American Psy¬
chologist, 1961, 16, 681-684.
Bridgman, P. W. The logic of modern physics. New York: Macmillan, 1927.
Bridgman, P. W. Remarks on the present state of operationalism. Scientific
Monthly, 1954, 79, 224-226.
Brodbeck, M. Explanation, prediction, and “imperfect” knowledge. In
H. Eeigl & G. Maxwell (Eds.), Minnesota Studies in the Philosophy
of Scienee (Vol. 3). Minneapolis: University of Minnesota Press,
1962.
Bruner, J. S., & Postman, L. On the perception of incongruity: A paradigm.
Journal of Personality, 1949, 18, 206-223.
Bush, R. R., & Mosteller, F. A. A mathematical model of simple learning.
Cahn, E. Jurisprudence. New York University Law Review, 1955, 30, 150-
169.
Cain, W. S. To know with the nose: Keys to odor identification. Science,
1979, 203, 467-470.
Campione, J. C. Optional intradimensional and extradimensional shifts in
children as a function of age. Journal of Experimental Psychology,
1970, 84, 296-300.
S74 References
Carmichael, L., Hogan, H. P., & Walter, A. A. An experimental study of the

effect of language on the reproduction of visually perceived form.
Journal of Experimental Psychology, 1932, 15, 73-86.
Chein, I. The science of behavior and the image of man. New York: Basic
Books, 1972.
Chein, I., Cook, S. W., & Harding, J. The field of aetion research. American
Psychologist, 1948, 3, 43-50.
Chernoff, P. R. Understanding mathematical proofs: conceptual barriers.
Science, 1976, 193, 276.
Chomsky, N. Review of Skinner’s Verbal behavior. Language, 1959, 35,
26-58.
Chomsky, N. Language and mind. New York: Harcourt, Brace and World,
1968.
Chomsky, N. Reflections on language. New York: Pantheon, 1975.
Christy, P., Gelfand, D. M., & Hartmann, D. P. EflFects of competition-
induced frustration on two classes of modeled behavior. Develop¬
mental Psychology, 1971, 5, 104-111.
Conant, J. On understanding science. New Haven: Yale University Press,
1947.
Courant, R., & Robbins, H. What is mathematics? (4th ed.). New York:
Oxford University Press, 1941.
Cummings, N. A. The anatomy of psychotherapy under national health in¬
surance. American Psychologist, 1977, 32, 711-718.
D’Amato, M. Experimental psychology. New York: McGraw-Hill, 1970.
Darwin, F., & Seward, A. C. (Eds.). More letters of Charles Darwin (Vol.
1). London: John Murray, 1903.
Darwin, F. The life and letters of Charles Darwin, including an autobio¬
graphical chapter (3 vols.). London: John Murray, 1887.
Deutscher, M., & Chein, I. The psyehological effects of enforced segrega¬
tion: A survey of social science opinion. Journal of Psychology, 1948,
26, 259-287.
De Valois, R. L., Abramov, L, & Jacobs, G. H. Analysis of response patterns
of LGN cells. Journal of the Optical Society of America, 1966, 7,
966-977.
De Valois, R. L., & Jacobs, G. H. Primate color vision. Science, 1968, 162,
533-540.
Diamond, S. S., & Morton, D. R. Empirical landmarks in social psychology.
Personality and Social Psychology Bulletin, 1978, 4, 217-221.
Driesch, H. The Science and Philosophy of the Organism. London: A. and
C. Black, 1908.
Dunn, L. C. Heredity and evolution in human populations (Rev. ed.).
Cambridge: Harvard University Press, 1965.
Ehrenfreund, D. An experimental test of the continuity theory of diserimina-
tion with pattern vision. Journal of Comparative Psychology, 1948,
41, 408-422.
References 875
Elesh, D., Ladinsky, D., Lefcowitz, M. J., & Spillerman, S. The New Jersey-
Pennsylvania experiment: A field study in negative taxation. In
L. L. Orr, R. E. Hollister, & M, J. Lefcowitz (Eds.), Income main¬
tenance: Interdisciplinary approaches to research. Chicago: Mark¬
ham, 1971.
Estes, W. K. Toward a statistical theory of learning. Psychological Review,
1950, 57, 94-107.
Feigl, H. The mental and the ‘"physical.” Minneapolis: University of Min¬
nesota Press, 1967.
Feigl, H. The orthodox ’ view of theories: Remarks in defense as well as
critique. In M. Radner & S. Winokur (Eds.), Minnesota Studies in
the Philosophy of Science (Vol. 4). Minneapolis: University of
Minnesota Press, 1970.
Feldman, S. E. Probabilistic hierarchies to ambiguous concept classes, four-
nal of Experimental Psychology, 1963, 65, 240-247.
Ferster, C. B. Positive reinforcement and behavioral deficits of autistic chil¬
dren. Child Development, 1961, 32, 437-456.
Festinger, L., & Carlsmith, J. M. Cognitive consequences of forced com¬
pliance. lournal of Abnormal and Social Psychology, 1959, 58, 203-
210.
Frank, J. D. The bewildering world of psychotherapy. lournal of Social
Issues, 1972, 22, 27-43.
Gallo, P. S., Jr. Meta-analysis—A mixed meta-phor? American Psychologist,
1978, 33, 515-516.
Garcia, J. What lies ahead after Bakke ruling? Los Angeles Times, Tanuarv
7, 1978, Part II, p. 4.
Gardiner, P. Vico, Giambattista, S. V. In P. Edwards (Ed.), The Ency¬
clopedia of Philosophy (Vol. 8), New York: Macmillan & The Free
Press, 1967.
Garner, W. R. Uncertainty and structure as psychological concepts. New
York: Wiley, 1962.
Gesell, A., Halverson, H. M., Thompson, H., Ilg, F. L., Costner, B. M.,
Ames, L. B., & Amatruda, C. S. The first five years of life: a guide
to the study of the preschool child. New York: Harper & Row, 1940.
Gesell, A., Ilg, F. L., Ames, L. B., & Bullis, G. E. The child from five to ten.
New York: Harper, 1946.
Ghiselin, M. T. The triumph of the Darwinian method. Berkeley: Univer¬
sity of California Press, 1969.
Goble, F. The third force. New York: Pocket Books, 1971.
Goldstein, K. The organism. New York: American Book, 1939.
Gross, J. Learning readiness in two lewish groups. New York: Center for
Urban Education, 1967.
Gurr, T. R. A comparative study of civil strife. In H. D. Graham & T. R.
Gurr (Eds.), The history of violence in America (Rev. Bantam ed.).
New York: Bantam, 1970.
376 References
Guthrie, E. R. The psychology of learning (Rev. ed.). New York: Harper,

1952.
Hall, C. S., & Lindzey, G. Theories of personality. New York: Wiley, 1957.
Hanson, N. R. Patterns of discovery. Gambridge: Gambridge University
Press, 1958.
Harlow, H. F. The formation of learning sets. Psychological Review, 1949,
56, 51-65.
Harlow, H. F. The nature of love. American Psychologist, 1958, 13, 673-
685.
Harre, R. The principles of scientific thinking. London: Macmillan, 1970.
Hayes, P. J. Gognitivism as a paradigm. The Behavioral and Brain Sciences,
1978, 1, 238-239.
Hebb, D. O. The organization of behavior. New York: Wiley, 1949.
Heston, L. L. The genetics of schizophrenic and schizoid disease. Science,
1970,167, 249-255.
Hilgard, E. R., & Bower, G. H. Theories of Learning (4th ed.). Englewood
Gliffs, N.J.: Prentice-Hall, 1975.
Honig, W. K. Prediction of preference, transposition, and transposition-
reversal from the generalization gradient. Journal of Experimental
Psychology, 1962, 64, 239-248.
Howe, I. World of our fathers. New York: Harcourt Brace Jovanovich, 1976.
Hull, G. L. Quantitative aspects of the evolution of concepts. Psychological
Monographs, 1920, 28, 123.
Hull, G. L. The mechanism of the assembly of behavior segments in novel
combinations suitable for problem solution. Psychological Review,
1935, 42, 219-245.
Hull, G. L. Principles of behavior. New York: Appleton-Gentury, 1943.
Hull, G. L. Autobiography. In G. A. Murchison (Ed.), A history of psy¬
chology in autobiography (Vol. 4). New York: Russell & Russell
Press, 1952.
Hull, G. L., Hovland, G. I., Ross, R. T., Hall, M., Perkins, D. T., & Fitch,
F. B. Mathematico-deductive theory of rote learning. New Haven:
Yale University Press, 1940.
Hunter, J. E., & Schmidt, F. L. Gritical analysis of the statistical and ethical
implications of various definitions of test bias. Psychological Bulletin,
1976, 83, 1053-1071.
James, W. A. The principles of psychology (Vol. 1). New York: Holt, 1890.
Jensen, A. R. How much can we boost IQ and scholastic achievement?
Harvard Educational Review, 1969, 39, 1-123.
Jensen, A. R. Educability and group differences. New York: Harper & Row,
1973.
Jensen, A. R. What is the question? What is the evidence? In T. S. Krawiec
(Ed.), The psychologists (Vol. 2). New York: Oxford University
Press, 1974.
References 377
Jensen, A. R. Race and mental ability. In F. J, Ebling (Ed.), Racial •varia¬

tion in man. New York: Halsted Press, 1975.
Jensen, A. R. Cumulative deficit in IQ of blacks in the rural south. Develop¬
mental Psychology, 1977, 3, 184-192.
Kamin, L. J. The science and politics of IQ. Potomac, Md.: Erlbaum, 1974.
Kanner, J. H. A test of whether the “nonrewarded” animals learned as much
as the rewarded animals in the California latent learning study.
Kaplan, A. The conduct of inquiry. San Francisco: Chandler, 1964.
Kemelman, H. Wednesday the Rabbi got wet. New York: Morrow, 1976.
Kendler, H. H. Some comments on Thistlewaite’s perception of latent learn¬
ing. Psychological Rulletin, 1952, 49, 47-51.
Kendler, H. H. Learning. In P. R. Farnsworth & Q. McNemar (Eds.),
Annual review of psychology. Stanford: Annual Reviews, 1959.
Kendler, H. H. Rosie psychology. New York: Appleton-Century-Crofts, 1963.
Kendler, H. H. Rosie psychology (2nd ed.). New York: Appleton-Century-
Crofts, 1968.
Kendler, H. H. The unity of psychology. The Canadian Psychologist, 1970,
11, 30-47.
Kendler, H. H. Rasic psychology (3rd ed.). Menlo Park, Cal.: Beniamin,
1974.
Kendler, H. H. A seductive paradigm. In J. Cotton & R. L. Klatzky (Eds.),
Semantic Factors in Cognition. Hillsdale, N.J.: Lawrence Erlbaum,
1978.
Kendler, H. H. Behaviorism and psychology: An uneasy alliance. Paper
delivered at the 1979 meeting of The American Psychological Asso¬
ciation, September, 1979.
Kendler, H. H. Developmental changes in classificatory behavior. Child De¬
velopment, 1980, 51, 339-348.
Kendler, H. H., & Kendler, T. S. Vertical and horizontal processes in prob¬
lem solving. Psychological Review, 1962, 69, 1-16.
Kendler, H. H., & Kendler, T. S. From discrimination learning to cognitive
development: A neobehaviorist odyssey. In W. K. Estes (Ed.),
Handbook of learning and cognitive processes (Vol. 1). Hillsdale,
N.J.: Lawrence Erlbaum, 1975.
Kendler, H. H., Kendler, T. S., & Ward, J. W. An ontogenetic analysis of
optional intradimensional and extradimensional shifts. Journal of Ex¬
perimental Psychology, 1972, 95, 102-109.
Kendler, H. H., & Mencher, H. C. The ability of rats to learn the location
of food when motivated by thirst—An experimental reply to Leeper.
Kendler, H. H., & Spence, J. T. (Eds.). Essays in neobehaviorism. New
York: Appleton-Century-Crofts, 1971.
Kendler, T. S. Contributions of the psychologist to constitutional law. The
American Psychologist, 1950, 5, 505-510.
S78 References
Kendler, T. S., & Kendler, H. H. Experimental analysis of inferential be¬

havior in children. In L. P. Lipsitt & C. C. Spiker (Eds.), Advances
in child development and behavior (Vol. 3). New York: Academic
Press, 1967.
Kessel, F. S. The philosophy of science as proclaimed and science as prac¬
ticed: “Identity” or “dualism”? American Psychologist, 1969, 24,
999-1005.
Kintsch, W., & Van Dijk, T. A. Toward a model of text comprehension and
production. Psychological Review, 1978, 85, 363-394.
Koch, S. Psychology and emerging conceptions of knowledge as unitary. In
T. W. Wann (Ed.), Behaviorism and phenomenology. Chicago:
University of Chicago Press, 1964.
Koch, S. Psychology as science. In S. C. Brown (Ed.), Philosophy of psy¬
chology. New York: Barnes & Noble, 1974.
Koffka, K. Principles of Gestalt psychology. New York: Harcourt, Brace,
1935.
Kohlberg, L. Implications of developmental psychology for education-
examples of moral development. Educational Psychologist, 1973, 10,
2-14.
Kohler, W. The mentality of apes. New York: Harcourt, Brace, 1925.
Kohler, W. Simple structural functions in the chimpanzee and in the chicken.
In W. D. Ellis (Ed.), A source book of Gestalt psychology. London:
Routledge & K. Paul, 1938.
Kolata, G. B. Mathematical proofs: the genesis of reasonable doubt. Science,
1976,192, 989.
Krebs, D. Empathy and altruism. Journal of Personality and Social Psychol¬
ogy, 1975, 32, 1134-1146.
Krechevsky, I. “Hypotheses” versus “chance” in the pre-solution period in
sensory discrimination-learning. University of California Publications
in Psychology, 1932, 6, 27-44.
Kuhn, T. S. The structure of scientific revolutions. Chicago: University of
Chicago Press, 1962.
Kuhn, T. S. The structure of scientific revolutions (2nd ed.). Chicago:
University of Chicago Press, 1970.
Kuhn, T. S. Second thoughts on paradigms. In F. Suppe (Ed.), The struc¬
ture of scientific theories. Urbana: University of Illinois Press, 1974.
Kiilpe, O. Outlines of psychology (E. B. Titchener, Trans.). New York:
Macmillan, 1909.
Lachman, R. The model in theory construction. Psychological Review, 1960,
67, 113-129.
Lachman, R., Lachman, J. L., & Butterfield, E. C. Cognitive psychology
and information processing: An introduction. Hillsdale, N.J.: Law¬
rence Erlbaum, 1978.
Lack, D. Darwins finches. Cambridge: Cambridge University Press, 1947.
Lakatos, I. History of science and its rational reconstruction. In R. C. Buck
References S79
& R. S. Cohen (Eds.), Boston Studies in the Philosophy of Science

(Vol. 8). Dordrecht, Netherlands; Reidel, 1970a.
Lakatos, I. Falsification and the methodology of scientific research pro¬
grammes. In I. Lakatos & A. Musgrave (Eds.), Criticism and the
growth of knowledge. Cambridge: Cambridge University Press,
1970b.
Lakatos, I., & Musgrave, A. Criticism and the growth of knowledge. Cam¬
bridge: Cambridge University Press, 1970.
Latane, B., & Darley, J. M. Group inhibition of bystander intervention in
emergencies. Journal of Personality and Social Psychology, 1968, 10,
215-221.
Latane, B., & Rodin, J. A lady in distress: Inhibiting effects of friends and
strangers on bystander intervention. Journal of Experimental Social
Psychology, 1969, 5, 189-202.
Levine, M. Scientific method and the adversary model: Some preliminary
thoughts. American Psychologist, 1974, 29, 661-676.
Lewin, K. A dynamic theory of personality. New York: McGraw-Hill, 1935.
Lewin, K., Lippitt, R., & White, R. K. Patterns of aggressive behavior in
experimentally created “social climates.” Journal of Social Psychology,
1939,10, 271-299.
Lewis, O. The culture of poverty. Scientific American, 1966, 2J5(4), 19-25.
Luborsky, L., Chandler, M. Averbach, A. H., Cohen, J., & Bachrach, H. M.
Factors influencing the outcome of psychotherapy: A review of quan¬
titative research. Psychological Bulletin, 1971, 75, 145-185.
MacCorquodale, K., & Meehl, P. E. On a distinction between hypothetical
constructs and intervening variables. Psychological Review, 1948,
55, 95-107.
MacCorquodale, K., & Meehl, P. E. On the elimination of cul entries with¬
out obvious reinforcement. Journal of Comparative and Physiological
Psychology, 1951, 44, 178-183.
McFarland, M. W. (Ed.). The papers of Wilbur and Orville Wright
(Vol. 1). New York: McGraw-Hill, 1953.
McKinney, F. Fifty years of psychology. American Psychologist, 1976, 31,
834-857.
MacNichol, E. F., Jr. Three pigment color vision. Scientific American, 1964,
21 J(b), 48-56.
Maddi, S. The search for meaning. In M. Page (Ed.), Nebraska Symposium
on Motivation. Lincoln: University of Nebraska Press, 1970.
Maier, N. R. F. Reasoning in white rats. Comparative Psychology Mono¬
graphs, 1929, No. 29.
Maslow, A. H. Motivation and personality. New York: Harper & Row, 1954.
Maslow, A. H. “Eupsychia—the good society.” Journal of Humanistic Psy¬
chology, 1961, 1(2), 1-11.
Maslow, A. H. Religions, values and peak experiences. Columbus: Ohio State
University Press, 1964.
880 References
Masterman, M. The nature of a paradigm. In I. Lakatos & A. Musgrave

(Eds.), Criticism and the growth of knowledge. Cambridge: Cam¬
bridge University Press, 1970.
Matarazzo, J. D, Wechslers measurement and appraisal of adult intelligence.
Baltimore: Williams & Wilkins, 1972.
Medawar, P. B, Unnatural science. The New York Review of Books, 1977,
24(1), 13-18.
Meehl, P. E. Some methodological reflections on the difiiculties of psycho¬
analytic research. In M. Radner & S. Winokur (Eds.), Minnesota
Studies in the Philosophy of Science (Vol. 4). Minneapolis: Univer¬
sity of Minnesota Press, 1970.
Meehl, P. E. Schizotaxia, schizotypy, schizophrenia. American Psychologist,
1962,17, 827-838.
Meyer, M. Psychology of the other-one. Columbia, Missouri: Missouri Book
Co., 1921.
Milgram, S. Behavioral study of obedience. Journal of Abnormal and Social
Psychology, 1963, 67, 371-378.
Miller, N. E. Experimental studies in conflict. In J. M. Hunt (Ed.), Per¬
sonality and the Behavior Disorders. New York: Ronald Press, 1944.
Money, J., & Ehrhardt, A. A. Man and woman, boy and girl. Baltimore:
Johns Hopkins University Press, 1972.
Mpitsos, G. J., Collins, S. D., & McClellan, A. D. Learning: A model sys¬
tem for physiological studies. Science, 1978, 199, 497-506.
Murdock, G. P. Comparative data on the division of labor by sex. Social
Forces, 1937,15, 551-553.
Murray, M. The treatment of autism: A human protest. Journal of Human¬
istic Psychology, 1974,14, 57-59.
Nafe, J, P. The psychology of felt experience. American Journal of Psychol¬
ogy, 1927, 39, 367-389.
Nagel, E. Methodological issues in psychoanalytic theory. In S. Hook (Ed.),
Psychoanalysis, scientific method, and philosophy. New York: New
York University Press, 1959.
Nagel, E. Theory and observation. In E. Nagel, S. Bromburger, & A. Griin-
baum (Eds,), Observation and theory in science. Baltimore: Johns
Hopkins Press, 1971.
Neisser, U, Review of W, G. Chase (Ed.) Visual information processing.
Proceedings of a symposium. Science, 1974, 183, 402-403.
Newell, A., & Simon, H. The simulation of human thought. In Current
trends in psychological theory. Pittsburgh: The University of Pitts¬
burgh Press, 1961.
Olbrisch, M. E. Psychotherapeutic interventions in physical health. American
Psychologist, 1977, 32, 761-777.
Osgood, C. E. A behavioristic analysis of perception and language as cogni¬
tive phenomena. In J. S. Bruner et al. (Contributors), Contemporary
References 381
approaches to cognition: A symposium held at the University of

Colorado. Cambridge: Harvard University Press, 1957.
Passmore, J. A. Can the social sciences be value-free? In H. Feigl &
M. Brodbeck (Eds.), Readings in the Philosophy of Science. New
Patai, R. Tents of Jacob: The diaspora—yesterday and today. Englewood
Cliffs, N.J.: Prentice-Hall, 1971.
Pavlov, I. P, Conditioned reflexes (G. V. Anrep, Trans.). London: Oxford
Pavlov, I. P. Conditioned reflexes and psychiatry. New York: International
Publishers, 1941.
Pennington, L. A., & Einan, J. L. Operational usage in psychology. Psy¬
chological Review, 1940, 47, 254-266.
Phillips, D. C. Holistic thought in social science. Stanford: Stanford Uni¬
versity Press, 1976.
Piaget, J. What is psychology? American Psychologist, 1978, 33, 648-652.
Polyani, M. Personal knowledge. Chicago: University of Chicago Press,
1958.
Popper, K. R. The poverty of historicism. London: Routledge & K. Paul,
1957.
Popper, K. R. The logic of scientific discovery. New York: Basic Books,
1949.
Postman, L. Organization and interference. Psychological Review, 1971, 78,
290-302.
Pratt, C. C. The logic of modern psychology. New York: Macmillan, 1939.
Provine, W. B. Geneticists and the biology of race crossing. Science, 1973,
182, 790-796.
Ratliff, F. Mach bands. San Francisco: Holden-Day, 1965.
Reeves, J. W. Thinking about thinking. New York: George Braziller,
1966.
Reichbenbach, H. Experience and prediction. Chicago: University of Chi¬
cago Press, 1938.
Reynolds, B. A repetition of the Blodgett experiment on “latent learning.”
Roche, J. P. Constitutional law: Distribution of powers, s. v. In D. Sills
(Ed.), International Encyclopedia of the Social Sciences (Vol. 3),
pp. 300-307. New York: Macmillan & The Free Press, 1968.
Rosenbaum, M. Psychosomatic medicine. 11. Psychophysiological disorders.
In A. M. Freedman, H. 1. Kaplan, & B. J. Saduck (Eds.), Compre¬
hensive textbook of psychiatry. Baltimore: Williams & Wilkins, 1967.
Rychlak, J. E. A philosophy of science for personality theory. New York:
Houghton Mifflin, 1968.
Rychlak, J, E. Personality theory: Its nature, past, present and future?
Personality ir Social Psychology Bulletin, 1976, 2, 209-224.
3S2 References
Scarr-Salapatek, S. Review of L. J. Kamin’s The science and politics of IQ.

Contemporary Psychology, 1976, 21, 98-99.
Schiller, B. Verbal, numerical and spatial abilities of young children.
Archives of Psychology, 1934, No. 161, 69.
Schlesinger, I. M. Steps to language. Hillsdale, N.J.: Lawrence Erlbaum,
in press.
Sears, R. R. Survey of objective studies in psychoanalytic concepts. New
York: Social Science Research Council Bulletin (No. 51), 1943.
Sheehan, T. Italy; Behind the ski mask. The New York Review of Books,
1979, 26(13), 20-26.
Shuey, A. M. The testing of Negro intelligence (2nd Ed.). New York:
Social Science Press, 1966.
Sidman, M. Tactics of scientific research. New York: Basic Books, 1960.
Silverman, L. H. Psychoanalytic theory: “The reports of my death are
greatly exaggerated.” American Psychologist, 1976, 31, 621—637.
Simon, H. A,, & Newell, A. Information processing in computer and man.
American Scientist, 1964, 52, 145-159.
Simon, H. A,, & Newell, A. Human problem solving: The state of the theory
in 1970. American Psychologist, 1971, 26, 145-159.
Simpson, G. G. An adversary view of sociobiology (Review of M. S. Ahlin s
The use and abuse of biology). Science, 1977, 195, 773-774.
Skinner, B. F. Are theories of learning necessary? Psychological Review,
1950, 57, 193-216.
Skinner, B. F. John Broadus Watson, behaviorist. Science, 1959, 129, 197-
198.
Skinner, B. F. Contingencies of reinforcement: A theoretical analysis. New
Skinner, B. F. Particulars of my life. New York: Knopf, 1976.
Smith, E. E. Theories of semantic memory. In W. K. Estes (Ed.), Hand¬
book of learning and cognitive processes (Vol. 6). Hillsdale, N.J.:
Lawrence Erlbaum, 1978, pp. 1-58.
Smith, M. B. Social psychology, science, and history, so what? Personality
and Social Psychology Bulletin, 1976, 4, 438-444.
Smith, M. L., & Glass, G. V. Meta-analysis of psychotherapy outcome
studies. American Psychologist, 1977, 32, 752-760.
Spence, K. W. The nature of discrimination learning in animals. Psycho¬
logical Review, 1936, 43, 427-449.
Spence, K. W. The differential response in animals to stimuh varying within
a single dimension. Psychological Review, 1937, 44, 430—444.
Spence, K. W. Cognitive versus stimulus-response theories of learning.
Spence, K. W. Behavior theory and conditioning. New Haven: Yale Uni¬
versity Press, 1956.
Spence, K. W., & Lippitt, R. An experimental test of the sign-gestalt theory
References S83
of trial and error learning. Journal of Experimental Psychology, 1946,

36, 491-502.
Spengler, O. The decline of the West. New York: Knopf, 1950.
Spoerl, D. T. BilinguaHty and emotional adjustment. Journal of Abnormal
and Social Psychology, 1943, 38, 35-57.
Stein, L., & Wise, C. D. Possible etiology of sehizophrenia: Progressive dam¬
age to the noradrenergic reward system by 6-hydroxy dopamine.
Science, 1971,171, 1032-1036.
Stellar, E. The physiology of motivation. Psychological Review, 1954, 61,
5-22.
Sternberg, S. Memory seanning: Mental proeesses revealed by reaetion-
time experiments. American Scientist, 1969, 57, 421-457.
Stoppard, T. Travesties. New York: Grove Press, 1975.
Suppe, F. The structure of scientific theories. Urbana: University of Illinois
Press, 1974.
Symonds, P. M. The effects of attendance at Chinese language schools on
ability with English language. Journal of Applied Psychology, 1924,
8, 411-424.
Tart, C. T. (Ed.). Altered states of consciousness. New York: Wiley, 1969.
Teitelbaum, P. The issue of operant methods in the assessment and control
of motivational states. In W. K. Honig (Ed.), Operant behavior:
Areas of research and application. New York: Appleton-Century-
Crofts, 1966.
Terrace, H. S. Discrimination learning, the peak shift, and behavioral con¬
trast. Journal of Experimental Analysis of Behavior, 1968, 11, 727—
741.
Thorndike, E. L. Animal intelligence: An experimental study of the asso¬
ciative processes in animals. Psychological Monographs, 1898, 2(8).
Thurstone, L. L. Vectors of the mind. Chicago: University of Chicago Press,
1935.
Titchener, E. B. An outline of psychology. New York: Macmillan, 1899.
Titchener, E. B. Lectures on the experimental psychology of the thought
processes. New York: Macmillan, 1909.
Titchener, E. B. Text-book of psychology. New York: Macmillan, 1910.
Tolman, E. C. Purposive behavior in animals and men. New York: Century,
1932.
Tolman, E. C. Operational behaviorism and current trends in psychology.
Proceedings of the 25th Annual Celebration of the Inauguration of
Graduate Studies, The University of Southern California. Los An¬
geles: The University of Southern California Press, 1936, pp. 89-103.
Tolman, E. C., & Honzik, C. H. Introduction and removal of reward and
maze performance of rats. University of California Publications in
Psychology, 1930, 4, 257-275.
Toynbee, A. J. A study of history (Abridgement in two volumes by D. C.
Somervell). London: Oxford University Press, 1946-57.
384 References
Tuddenham, R. D. The nature and measurenient of intelligence. In L. Post¬

man (Ed.), Psychology in the making. New York: Knopf, 1962.
Tulving, E., & Osier, S. Transfer effects in whole/part free-recall learning.
Canadian Journal of Psychology, 1967, 21, 250-262.
Turner, M. B. Philosophy and the science of behavior. New York: Appleton-
Century-Crofts, 1965.
Vurpillot, E. The development of scanning strategies and their relation to
visual differentiation. Journal of Experimental Child Psychology,
1968, 6, 632-650.
Wade, N. Sociobiology: Troubled birth for a new discipline. Science, 1976,
’ 191, 1151-1155.
Wagoner, K. S., & Goodson, F. E. Does the mind matter? In M. H. Marx
& F. E. Goodson (Eds.), Theories in comtemporary psychology
(2nd ed.). New York: Macmillan, 1976. ‘
Wallace, R. K., & Benson, H. The physiology of meditation. Scientific
American, 1972, 226(2), 85-90.
Warren, J. M. Primate learning in comparative perspective. In A. M. Schrier,
H. F. Harlow, & F. Stollnitz (Eds.), Behavior of nonhuman primates
(Vol. 1). New York: Academic Press, 1965.
Watson, J. B. Experimental studies on the growth of the emotions. In
G. Murchison (Ed.), Psychologies of 1925. Worcester, Mass.: Glark
Watson, R. I. The great psychologists (4th ed.). Philadelphia: Lippincott,
1978.
Weber, M. The protestant ethic and the spirit of capitalism (T. Parsons,
Trans.). New York: Scribner’s, 1958 (Originally published 1904-05).
Wêizenbaum, J. Computer power and human reason. San Francisco: W^. H.
Freeman, 1976.
Wertheimer, M. Gestalt theory. In W. D. Ellis (Ed.), A source book of
Gestalt psychology. New York: Harcourt, Brace & World, 1938.
Wertheimer, M. Productive thinking. New York: Harper, 1945.
Whewell, W. Novum organon renovatum, being the second part of the
philosophy of the inductive sciences (3rd ed.). London: Parker &
Son, 1858.
Williams, B. Rationalism. In P. Edwards (Ed.), The encyclopedia of phi¬
losophy (Vol. 7). New York: Macmillan & The Free Press, 1967.
Wilson, E. O. Sociobiology: The new synthesis. Gambridge: Belknap Press
of Harvard University Press, 1975.
Wittig, M. A., & Petersen, A. G. (Eds.). Sex-related differences in cognitive
functioning. Developmental differences. New York: Academic Press,
1979.
Wolff, P. H. Ethnic differences in alcohol sensitivity. Science, 1972, 175,
449-450.
Woodworth, R. S. Experimental psychology. New York: Holt, 1938.
Index
action research, 288-89 source of morality, 270-76; study

Adler, Alfred, 158 of, as discipline, 67—68; as subject-
aesthetics, 194-96 matter of psychology, 13-39
Albee, G., 352-53 behavioral control, 170-88
Allen, Woody, 207, 350-51 behaviorism, 16-18, 25-28; be¬
Alper, S., 227, 232 havioral control and, 180n.; bio-
American Psychological Association psychological theory compared
(APA), 301-2 with, 119—20, 122—24; conscious¬
animal behavior: latent learning in, ness in, 39; experiential psychol¬
49-51; meaningfulness in, 30-32; ogy and, 67-87; intuitive knowing
neobehaviorist view of, 359 and, 204; phenomenological mean¬
anthropomorphism, 84 ing in, 35-36; theory and obser¬
Aquinas, Thomas, 274 vation in, 62
Aristotle, 102, 235 Bellow, Saul, 191, 311-12
Asch, Solomon E., 329 Benson, H., 73
Atkinson, Richard C., 286-87, 290 Bergmann, G., 86, 89-91, 93
295 Berkeley, George, 16
atomism, 334-44 Bernstein, Leonard, 194
Auerbach, A. H., 196 bilingualism, 282-84
Binet, Alfred, 81, 236-37
Bachrach, H. M., 196 biopsychology, 119-33, 303-4, 307;
Back, K. W., 351 developmental and comparative
Bakke, Allen, 260-62 research and, 324—26; neurophysi¬
Beach, F. A., 323 ology in, 313-21
behavior: deductive models of, Bitterman, M. E., 105
105—6; developmental research on, black-box theories, 119-29, 314-17,
325—26; epiphenomenalist view of, 319, 320
78; genetic differences in, 259; blacks, 245, 249; in athletics, 258-
intelligence as, 235—37; moral 59, 265; Bakke case and, 261-62;
training for, 296-97; neobehavior¬ effects of segregation upon, 284;
ist view of, 359; neurophysiological as slaves, 281. See also race
interpretations of, 313-21; physi- Boring, E. G., 44, 82, 305-6
ology of, 119, 121-26; racial dif¬ box-score method, 252
ferences in, 241-49; reductionism Brandeis, Louis, 285
in explanations of, 131-35; rules Breland, K., 181
governing, 8-9; sociobiological Breland, M., 181
explanations of, 226-27, 233; as Brentano, 43, 133
S85
386 Index
Bridgman, P. W., 93-94 Davenport, Charles B,, 228-29

Brodbeck, M,, 114 deductivism, 103-10, 153; evaluation
Brown v. Board of Education (U.S., of theories of, 135-51; as form of
1954), 284 understanding versus research
Bruner, J. S., 53—56 strategy, 104-10; inductive
Bush, R. R., 174 processes in, 110-11; interpretive
Butterfield, E, C., 360, 367 consistency and, 168-79; in
natural science model, 111-14;
Cahn, E., 287 role of experimentation in, 114-
Cain, W. S., 70n. 17; as theory, 117-35
Carlsmith, J. M., 329 Descartes, Rene, 15-16
Cattell, James McKeen, 236-37 developmental research, 323-27
cephalic index, 89-90 Dewey, John, 74, 85
Chandler, M., 196 Diamond, S. S., 328-30
Chein, I., 288-89 Driesch, Hans, 342
Chernoff, P. R., 201
Chomsky, Noam, 194, 346-48 Ebbinghaus, Hermann, 183-84
clustering, 32-33 Ehrhardt, A. A., 323
cognitive-dissonance theory, 329-32 Einstein, Albert, 189, 258
cognitive maps, 84-85 emergentism, 340-43
cognitive psychology, 24n., 144 empirical meaning, 89-91
Cohen, J., 196 empirical reductionism, 129-35
comparative research, 323-27 empiricism, 5, 344-49
computers, 362-66 epiphenomenal behaviorism, 35
Conant, James, 136 epiphenomenalism, 77-79
conceptions, 168—70 epistemology: of empiricism and
conditioning, 28-29 rationalism, 344-49; ethics and,
conscious experience: behaviorist 218; of experience and behavior,
views of, 79-80; direct compari¬ 21-24; of intuitive knowing, 190-
sons of, 94-97; epiphenomenalist 92
view of, 78; experimental research error, in self-observation, 63
on, 40-41; having and knowing Estes, W. K., 174
in, 74—75; as quasi-objective ethics, 218; of investigations of
knowledge, 72; relationship be¬ racial differences in behavior,
tween behavior and, 25-26; as 241-52; psychology as source of,
subject-matter of psychology, 13- 276-77; racial distinctions and,
24 262-67. See also morality
consciousness, 39-65 evaluation research, 291-93, 327-33
consensual agreement, 210-16 evidence, 146; moral choice and,
consistency, 153-54 270-74, 278-90
Cook, S. W., 288-89 evolution, 163-68, 214, 232-34, 311
Courant, R., 118 experience: shared, 64-65. See also
conscious experience; phenomenal
Darwin, Charles, 162, 163, 231, 311; experience
anthropomorphism by, 84; de¬ experiential groups, 351-52
duction used by, 114—15, 167, experiential psychology, 13-14, 39;
169; Marx and, 233; observation behavioral psychology and, 67-
by, 331 87
Darwinism, 163-68, 214, 326 experimental psychology: on con¬
Dashiell, John F., 191 sciousness, 40-41; consensual
Index
387
agreement in, 213—15; observa¬ historical method, 253-55
tional bias in, 48-56; present state history: as science, 159, 161, 171
of, 302-3, 305-6; scientism in, holism, 334-44
111-12 Holmes, Oliver Wendell, 287
experimental social psychologv, 328- Howe, Irving, 205
31, 333
Hull, Clark L., 82, 120-21, 145, 162,
explanation; biopsychological, 123- 185, 361; on afferent interaction,
24; deduction in, 103-10, 153; 338, 340; on deductive theorizing,
induction in, 110-11; natural sci¬ 104-7; on induction. 111; neo¬
ence model of, 111-14 behaviorism and, 358, 359
humanistic psychology, 108, 307,
fatal-flaw method, 252-53 342
feelings, 43 Hunter, J. E., 262-63, 266
Feigl, H., 5, 108 hypotheses, 115, 121-26
Fichte, Johann, 274 hypothetical constructs, 120
Festinger, Leon, 329
formalism: deductivism in, 105-7
iconic models, 128
Freud, Sigmund, 177, 351-52
imageless thought, 45
Freudianism, 154-58, 161, 168-70
175-76 induction, 110-11
information-processing paradigm.
Gahleo, 213, 216, 231 358-70
Gallo, P. S., 355-56 insight, 31, 85
genetics, 126-27; Darwinism and, intelligence; intuitive meanings of,
91-92; operational meaning of,
163-65, 167; developmental re¬
search and, 325—26; in explana¬ 89; race and, 234-68; theoretical
meaning of, 92-93
tion of schizophrenia, 121-22; of
intentions, 133-34
learning abilities, 345-49; Lysenko
interpretive consistency, 153-79;
dispute in, 227-28; race and, 228-
intuitive knowing and, 210
30, 240, 247—59; in sociobiology,
225-27 intervening variables, 83, 121
Gesell, A., 325 introspective reports, 80-84
intuitive knowing, 188-210
Gestalt psychology, 31-32, 34, 336-
38, 340 intuitive meaning, 91-92
Ghiselin, M. T., 11, 115
Glass, G. V., 355 James, William, 44-45, 258, 306,
Globe, F., 273 332-34, 342
Goldstein, Kurt, 342-43 Jensen, Arthur, 244-45, 251n.
Goodson, F. E., 78 Jews, 242-43, 254-55, 275, 283
Gurr, T. R., 331-32 Jung, Carl G., 158
Guthrie, E. R., 85
Kamin, Leon, 252-53
Hall, G. S., 175-76 Kanner, Joseph, 50
Handler, Philip, 298 Kant, Immanuel, 144
Harding, J., 288-89 Kaplan, A., 131-32
Harlow, H. F., 324-25 Kemelman, H., 155
Havighurst, 255 Kendler, Howard H,, 304-5, 314,
Hayes, P. J., 370 328
Hebb, D. O., 317-18 Kendler, Tracy S., 282, 283
Hering, 319 Kessel, F. S., 48
388 Index
Kintsch, W., 367, 368 measurement: of behavior, 38-39; of

knowing, intuitive, 188-210 intelligence, 236-40; of morality,
Koch, Sigmund, 306-12 271-72
Kohler, W., 31, 32, 338-39 Medawar, P. B., 238-39
Krechevsky, L, 84-85 meditation, 72—74
Kuhn, Thomas, 53, 358; subjectivism Meehl, P. E., 120, 157n., 158n.
of, 7, 101, 136-38, 140-41, 143 mental health, 297
metaphysical behaviorism, 77
methodology, 4-12; of behavioral
Lachman, J. L., 360, 367 measures, 38—39; of black-box
Lachman, R., 360, 367 versus physiological hypotheses,
Lakatos, I., 143-44, 146, 147, 160- 125—26; deduction in, 103-4;. of
61 direct comparison of conscious ex¬
language: computers and, 363-64; perience, 95—97; of evaluation of
genetic programming for, 345, competing theories, 252-55; of
347; in representation of observa¬ observation, theory in, 61-62; ob¬
tion, 56-58; social policy decisions servational bias and, 48-56; of
on, 282-84 phenomenological research, 68—
latent learning controversy, 49-51 69, 76-77; in physics and psy¬
learning: empiricist and rationalist chology, 111-12; unity of psychol¬
approaches to, 345-49; mathe¬ ogy and, 305; used in operant
matical models of, 107; maze re¬ conditioning, 183-86
search on, 323-24; meaningful be¬ Meyer, Max, 98-99
havior in, 32; operant conditioning Milgram, Stanley, 241
versus, 185-86; rote model of, Mill, John Stuart, 306
80-81; unresolvable controversies mind: in cognitive psychology, 24n.;
in, 314-15 Descartes on, 15—16; inability to
Levine, M., 213-14, 216 directly observe, 20-21; phe¬
Lewin, Kurt, 162, 275, 288, 328 nomenological observation of, 42-
Lewontin, Richard, 232 45; as subject-matter of psychol¬
Lindzey, G., 175-76 ogy, 13-14
logic, 110 Money, J., 323
logical positivism, 56-57 morality: behavior as source of,
Luborsky, L., 196 270-76; empirical evidence in
Lysenko, Trofim, 227-30 decisions of, 278-90; psychology
as source of, 276—77. See also
MacCorquodale, K., 120 ethics
Mach, Ernst, 16 moral training, 296-98
Maier, N. R. F., 361 Morton, D. R., 328-30
Marx, Karl, 160, 169, 177, 233 Mosteller, F. A., 174
Marxism, 161, 168, 207, 228, 230
Maslow, Abraham H., 108, 205-9, Nafe, J. P., 43
273-75 Nagel, E., 60, 154
mathematical models, 107, 314—15 negative-income-tax programs, 293-
mathematics, 201 94
meaning: empirical, 89—91; intuitive, Neisser, U., 326
91-92; operational, 87—89; phe¬ neobehaviorism, 358-60
nomenological, 34—36; search for, neo-Darwinism, 164-65
102-3; theoretical, 92-93 neurophysiology of behavior, 313-21
meaningful behavior, 30-37 Newell, A., 365
Index
S89
objectivity, 227, 231
ductionism in, 129-35; social
observation: of behavior, 27; bias in, problems and, 290-98; theories in,
48-56; consensual agreement on, 119—29. See also individual sub¬
86—87; in distinction between pri¬ disciplines and theoretical orienta¬
vate experiences and public be¬ tions
havior, 19-21; importance of Psychonomic Society, 301-2
theory in, 115; linguistic repre¬ psychotherapy, 85, 349-57
sentation of, 56—58; of mental public events, 69-77
events, 42; theoretical biases in,
58-64
operant behaviorism, 358, 360 quasi-objective knowledge, 72, 95
quota systems, 263-64, 266-68
operant conditioning, 181-87
operational meaning, 87-90, 93, 220
operationism, 87-94 race: genetic theories of, 228-31;
intelligence differences by, 234—
paradigms, 137-44, 358 68. See also blacks
Passmore, J. A., 221-24 rationalism, 344-49
Pavlov, Ivan P., 28, 202-3, 334-35 Ratliff, F., 318
perception: observational bias and, reductionism, 18, 129-35
52-56 Reichenbach, H., 9
phenomenal experience, 69-86 reinforcement, 122
phenomenological (subjective) research programs, 144-49
meaning, 34-36 Roosevelt, Eleanor, 207
phenomenology, 39-48; causes of Rosenbaum, M., 116
disputes within, 68; observations rote learning, 80-81
and methods in, 61-64; as science, Russell, Bertrand, 342
311-12 Rychlak, J. F., 202
Phillips, D. C., 342
physics, 111-14, 138-39 Santillana, Giorgio de, 345
physiology, 121-35 Schiller, B., 283
Piaget, Jean, 324, 326
schizophrenia, 121-24
Poincare, Henri, 82, 189 Schlesinger, I. M., 347-49
Popper, Karl R., 9, 155, 156, 161 Schlick, 74
Postman, L., 53-56 Schmidt, F. L., 262-63, 266
pragmatic behaviorism, 36 science: action research and, 288-
priignanz (good form), 197-99 89; consensual agreement in,
Pratt, C. C., 17
210-16; deductivism in, 111-14;
preconceptions, 53-56 differing conceptions of, 158-62;
Provine, W. B., 229-30
psychology, 307-12; sensory ex¬
psychoanalytic theory, 156-58, 175- periences in, 16-17; as sociopsy-
76, 308
chological enterprise, 6-11; values
psychology: consensual agreement in, 218-68; views of change in,
in, 211—16; in debate on social 135-51
policy, 280, 282—90; discrimina¬ scientific method, 102, 112, 159
tion within, 264—65; ethics in, Sears, R. R., 157
218; explanation in, 111-14; fu¬ segregation, 255-56, 284-85
ture trends in, 313-70; of intuitive self-actualization, 205-8, 273-74
knowing, 190—92; as moral author¬ self-observation, 44, 61-64
ity, 275-77; paradigms in, 138- sensations, 43, 57
40; present state of, 300-313; re- shared experience, 64-65
390 Index
Sidman, M,, 183-85 of, 369; predictability of, 161-68;

Silverman, L, H,, 157n., 158n, role of experimentation in, 114-17
Simon, H. A., 365 Thorndike, E. L., 30-31
Simon, Theodore, 236-37 Thurstone, Louis L., 117
Simpson, Gaylord G., 225 Titchener, Edward Bradford, 16-17,
Skinner, B. F., 29, 79-80, 180-81, 42-45, 64-65, 302, 337-40
186, 192, 358 Tolman, E. G., 79, 111, 144, 145,
slavery, 281 185; animal learning theory of, 50;
Smith, E. E., 366 on introspection, 83-85
Smith, M. B., 330 Toynbee, Arnold, 169
Smith, M. L., 355 transcendental meaning, 36-37
social policy, 255-68, 279-96 Trollope, Anthony, 312-13
social problems, 290-98 Turner, M. B., 133
social psychology, 327-33
sociobiology, 224-27, 231, 234 understanding: consensual agree¬
Sociobiology Study Group, 227, 230, ment in, 210-16; control as, 179-
231 80; deduction in, 103-51; inter¬
Spence, Janet T., 108 pretive consistency in, 79, 153-
Spence, Kenneth W., 108, 121, 185; 79; intuitive knowing and, 188,
deductive models used by, 106; 199-200
neobehaviorism and, 358-59; on Urbach, P., 254n.
operationism, 90-91; on transposi¬
tion, 339 values: of psychologists, 276; in sci¬
Spengler, Oswald, 169 ence, 218-68. See also ethics;
Stein, L., 121-22 morality
stimulus error, 63 Van Dijk, T. A., 367, 368
Stoppard, Tom, 112-13 verstehen, 172-73, 175, 176
structuralism, 16-18, 340-41 Vico, Giambattista, 159-61, 168-
Stumpf, Garl, 43 69, 171-73, 175, 176
subjective behaviorism, 35
subjective (phenomenological) Wagoner, K. S., 78
meaning, 34-36 Wallace, R. K., 73
Symonds, P. M., 283 Watson, John B., 19, 26, 77, 180,
334, 346
Tart, G. T., 108 Weber, Max, 352
test bias, 262-66 Weizenbaum, Joseph, 362, 366
theoretical meaning, 92-93 Wertheimer, M., 189, 340, 364n.
theoretical reductionism, 129-35 Whewell, William, 117
theory: in behavioral control, 182- Wilsom, Edward O., 224-27, 231
88; as bias in observation, 58-62; Wise, G. D., 121-22
black-box models in, 314—17, 319, Woodworth, R. S., 45
320; conceptions versus, 168-70; Wright, Orville, 123
deduction in, 103-9, 117-51; Wundt, Wilhelm, 19, 44, 327-28
falsification testing of, 155-58;
induction in, 110-11; interpretive Young, Andrew, 248n.
consistency within, 154-55; limits Young-Helmholtz theory, 319
f
‘<k:
.: ••
s^T.
I
i
Unique in analyzing the nature of psychology, this book does not
prescribe a “correct interpretation” but rather illuminates the dif¬
ferent kinds of psychology that currently exist. It offers an under¬
standing of contemporary psychology by revealing the assumptions
of different interpretations and the distinctive kinds of knowledge
they produce. Three basic, interrelated topics are analyzed succes¬
sively: the subject matter of psychology; the criteria for understand¬
ing psychological events; and the ethical principles underlying the
application of psychological knowledge. These topics confront psy¬
chologists with a methodological maze which can ultimately lead to
incompatible conceptions of psychology. The author explores the
various routes through this maze, and disentangles the numerous
methodological and empirical issues associated with choices. In
examining the structure of psychology—which in turn reveals the
diversity of methodological assumptions and ultimate goals—the
book inevitably raises questions concerning the future of psychology,
a topic which is discussed in the final chapter.
“This is a fine work. It is a book for all serious students of psy¬

chology, particularly for those of us who should look from time to
time beyond our immediate, narrow concerns .... it is timely and
provocative. It provides a telling statement, well worth reading and
debating.” Robert E. Silverman, York University
“Will be of value to courses on philosophy of science in psychology,

or to any course that attempts to deal with psychology’s issues. I
certainly intend to use it myself, because while no student should
be expected to agree in all respects with Kendler’s positions on
issues, every student should be expected to come to terms with these
issues.” Wendell R. Garner, Yale University
“One of the most important psychology books of the last 25 years.”

William A. Hunt, Loyola University of Chicago
The Author
Howard H. Kendler is Professor of Psychology at the University of
California, Santa Barbara. He is the author of Basic Psychology,
Third Edition, and co-editor of Essays in Neobehaviorism (with ]. T.
Spence).
Oxford Universit
0195029003
Jacket design by Joy Taylor 04/20/201 / 2:0/-3

Psychology A Science in Conflict

Uploaded by

Copyright:

Available Formats

Psychology A Science in Conflict

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Psychology A Science in Conflict

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What are some of the topics analyzed in the book?

What are some of the topics analyzed in the book?

Howard H.

JHE CHICAGO PUBLIC LIBRARY

© THE BAKER ft TAYLOR CO.

New York Oxford

Library of Congress Cataloging in Publication Data

Printing (last digit ):987654321

Printed in the United States of America

This book is a direct outgrowth of my career in psychology. I be¬

tion. I aspired to do this in a detached manner in order to illumi¬

tion. Rather than risk a debate about whether I was misstating a

Santa Barbara Howard H. Kendler

"■' ■' ft^ ** * •

'V >,'«• • < K ■'; %, .» ,*yi

i' ■ 4 ii4^ Vt< ..

.n V , <- ^ \ 1 ■^ * st» I Cl f'

■<i * 1' 4 ;■! V ; . .'• -. l , *1 .. ,.

^.■' '4 ■ ” ■■ » '• -■■ ♦' -»■■'

1. Plans and Purposes, 3

2. The Data Base and Subjeet Matter of Psyehology:

3. The Data Base and Subject Matter of Psychology:

4. On Understanding: Part One, 100

5. On Understanding: Part Two, 152

6. Psychology and Values: Epistemological Considerations, 217

7. Psychology and Values: Psychological Considerations, 269

8. Future Trends in Psychology, 299

Psychology is an ambiguous science. Arguments rage about the

ing attitudes. My intention is to oflFer a detaehed analysis of the

processes are fundamental, what questions are relevant, what

to numerous competing resolutions such as interactionism, paral¬

the individual scientist. The selection of one theory over another

subjective in the sense of reflecting personal taste. Consider a

completely analogous to chess but instead to illustrate different

formulates a new theory of reinforeement. Let us assume that the

thors scientific style. An example of such an effort is Ghiselin’s

but also holds the promise of elevating the quality of method¬

The plan of this book is to analyze successively three basic and

1. Psychology can be viewed as the science of the mind or as the sei-

2. Behavior, an observable event independent of theoretical precon¬

3. The psychologist who wishes to systematically study consciousness

4. The question is raised as to whether the unresolved phenomeno¬

PSYCHOLOCrS SUBJECT MATTER .

Although academic psychology has passed its hundredth anniver¬

CONSCIOUS EXPERIENCE AND BEHAVIOR

Two extreme views have prevailed about the subjeet matter of

havior. Numerous compromises have been offered that, in one

still important and unresolved. Finally, Descartes suggested, con¬

Data Base Versus Subject Matter

Many philosophers who rejected Cartesian dualism, particularly

ultimately behavioral prineiples ean be redueed to laws of phys-

‘Volitional decisions.” These decisions can exert profound eflFects

The Distinction Between Private Experience

Is the distinction between mind, for the moment equated

The Observational Base of Experience and Behavior

The distinction between conscious experience and public be¬

examining another person can justify the distinction under con¬