TOWARD A THEORY OF MOTIVE ACQUISITION1

DAVID c. MCCLELLAND
Harvard University

T OO little is known about the processes of

personality change at relatively complex
levels. The empirical study of the problem
has been hampered by both practical and theoreti-
Neither theory would provide much support for the
notion that motives could be developed in adult-
hood without somehow recreating the childhood
conditions under which they were originally formed.
cal difficulties. On the practical side it is very Furthermore, psychologists have been hard put to
expensive both in time and effort to set up system- it to find objective evidence that even prolonged,
atically controlled educational programs designed serious, and expensive attempts to introduce per-
to develop some complex personality characteristic sonality change through psychotherapy have really
like a motive, and to follow the effects of the edu- proven successful (Eysenck, 1952). What hope is
cation over a number of years. It also presents there that a program to introduce personality
ethical problems since it is not always clear that change would end up producing a big enough effect
it is as proper to teach a person a new motive as to study?
it is a new skill like learning to play the piano. Despite these difficulties a program of research
For both reasons, most of what we know about has been under way for some time which is at-
personality change has come from studying psycho- tempting to develop the achievement motive in
therapy where both ethical and practical difficul- adults. It was undertaken in an attempt to fill
ties are overcome by the pressing need to help some of the gaps in our knowledge about personal-
someone in real trouble. Yet, this source of in- ity change or the acquisition of complex human
formation leaves much to be desired: It has so characteristics. Working with n Achievement has
far proven difficult to identify and systematically proved to have some important advantages for this
vary the "inputs" in psychotherapy and to measure type of research: The practical and ethical prob-
their specific effects on subsequent behavior, except lems do not loom especially large because previous
in very general ways (cf. Rogers & Dymond, research (McClelland, 1961) has demonstrated the
1954). importance of high n Achievement for entrepre-
On the theoretical side, the dominant views of neurial behavior and it is easy to find businessmen,
personality formation suggest anyway that acquisi- particularly in underdeveloped countries, who are
tion or change of any complex characteristic like interested in trying any means of improving their
a motive in adulthood would be extremely difficult. entrepreneurial performance. Furthermore, a great
Both behavior theory and psychoanalysis agree that deal is known about the origins of n Achievement
stable personality characteristics like motives are in childhood and its specific effects on behavior
laid down in childhood. Behavior theory arrives so that educational programs can be systemat-
at this conclusion by arguing that social motives ically planned and their effects evaluated in terms
are learned by close association with reduction in of this knowledge. Pilot attempts to develop
certain basic biological drives like hunger, thirst, n Achievement have gradually led to the formula-
and physical discomfort which loom much larger tion of some theoretical notions of what motive
in childhood than adulthood. Psychoanalysis, for acquisition involves and how it can be effectively
its part, pictures adult motives as stable resolu- promoted in adults. These notions have been sum-
tions of basic conflicts occurring in early childhood. marized in the form of 12 propositions which it
1
1 am greatly indebted to the Carnegie Corporation of is the ultimate purpose of the research program to
New York for its financial support of the research on which test. The propositions are anchored so far as
this paper is based, and to my collaborators who have possible in experiences with pilot courses, in sup-
helped plan and run the courses designed to develop the porting research findings from other studies, and
achievement motive—chiefly George Litwin, Elliott Danzig, in theory.
David Kolb, Winthrop Adkins, David Winter, and John
Andrews. The statements made and views expressed are Before the propositions are presented, it is neces-
solely the responsibility of the author. sary to explain more of the theoretical and practical
321
322 AMERICAN PSYCHOLOGIST

background on which they are based. To begin needs to be briefly summarized here to provide a
with, some basis for believing that motives could theoretical underpinning for the attempts at motive
be acquired in adulthood had to be found in view change to be described. It starts with the proposi-
of the widespread pessimism on the subject among tion that all motives are learned, that not even
theoretically oriented psychologists. Oddly enough biological discomforts (as from hunger) or pleas-
we were encouraged by the successful efforts of ures (as from sexual stimulation) are "urges" or
two quite different groups of "change agents"— "drives" until they are linked to cues that can
operant conditioners and missionaries. Both signify their presence or absence. In time clusters
groups have been "naive" in the sense of being of expectancies or associations grow up around af-
unimpressed by or ignorant of the state of psycho- fective experiences, not all of which are connected
logical knowledge in the field. The operant condi- by any means with biological needs (McClelland
tioners have not been encumbered by any elaborate et al., 1953, Ch. 2), which we label motives. More
theoretical apparatus; they do not believe motives formally, motives are "affectively toned associative
exist anyway, and continue demonstrating vigor- networks" arranged in a hierarchy of strength or
ously that if you want a person to make a re- importance within a given individual. Obviously,
sponse, all you have to do is elicit it and reward the definition fits closely the operations used to
it (cf. Bandura & Walters, 1963, pp. 238 ff.). measure a motive: "an affectively toned associative
They retain a simple faith in the infinite plasticity cluster" is exactly what is coded in a subject's
of human behavior in which one response is just fantasies to obtain an n Achievement score. The
like any other and any one can be "shaped up" strength of the motive (its position in the indi-
(strengthened by reward)—presumably even an vidual's hierarchy of motives) is measured essen-
"achievement" response as produced by a subject tially by counting the number of associations
in a fantasy test. In fact, it was the naive belonging to this cluster as compared to others
optimism of one such researcher (Burris, 1958) that an individual produces in a given number
that had a lot to do with getting the present of opportunities. If one thinks of a motive as an
research under way. He undertook a counseling associative network, it is easier to imagine how
program in which an attempt to elicit and re- one might go about changing it: The problem
inforce achievement-related fantasies proved to be becomes one of moving its position up on the hier-
successful in motivating college students to get archy by increasing its salience compared to other
better grades. Like operant conditioners, the mis- clusters. It should be possible to accomplish this
sionaries have gone ahead changing people because end by such tactics as: (a) setting up the net-
they have believed it possible. While the evidence work—discovering what associations, for example,
is not scientifically impeccable, common-sense ob- exist in the achievement area and then extending,
servation yields dozens of cases of adults whose strengthening, or otherwise "improving" the net-
motivational structure has seemed to be quite radi- work they form; (b) conceptualizing the network
cally and permanently altered by the educational —forming a clear and conscious construct that
efforts of Communist Party, Mormon, or other labels the network; (c) tying the network to as
devout missionaries. many cues as possible in everyday life, especially
A man from Mars might be led to observe that those preceding and following action, to insure that
personality change appears to be very difficult for the network will be regularly rearoused once
those who think it is very difficult, if not impos- formed; and (d) working out the relation of the
sible, and much easier for those who think it can network to superordinate associative clusters, like
be done. He would certainly be oversimplifying the self-concept, so that these dominant schemata
the picture, but at the very least his observation do not block the train of achievement thoughts—
suggests that some theoretical revision is desirable for example, through a chain of interfering associa-
in the prevailing views of social motives which link tions (e.g., "I am not really the achieving type").
them so decisively to early childhood. Such a This very brief summary is not intended as a
revision has been attempted in connection with the full exposition of the theoretical viewpoint under-
research on n Achievement (McClelland, Atkinson, lying the research, but it should suffice to give a
Clark, & Lowell, 19S3) and while it has not been rough idea <?f how the motive was conceived that
widely accepted (cf. Berelson & Steiner, 1964), it we set out to change. This concept helped define
 MOTIVE ACQUISITION 323

the goals of the techniques of change, such as tractive" rather than "additive." After we have
reducing the effects of associative interference from demonstrated a substantial effect with some 10-12
superordinate associate clusters. But what about factors working to produce it, our plan is to sub-
the techniques themselves? What could we do tract that part of the program that deals with each
that would produce effective learning of this sort? of the factors to discover if there is a significant
Broadly speaking, there are four types of empirical decline in the effect. It should also be possible
information to draw on. From the animal learning to omit several factors in various combinations to
experiments, we know that such factors as repeti- get at interactional effects. This will obviously
tion, optimal time intervals between stimulus, require giving a fairly large number of courses
response, and reward, and the schedule of rewards in a standard institutional setting for the same
are very important for effective learning. From kinds of businessmen with follow-up evaluation of
human learning experiments, we know that such their performance extending over a number of
factors as distribution of practice, repetitions, years. So obviously it will be some time before
meaningfulness, and recitation are important. each of the factors incorporated into the proposi-
From experiences with psychotherapy (cf. Rogers, tions which follow can be properly evaluated so
1961), we learn that warmth, honesty, non- far as its effect on producing motive change is
directiveness, and the ability to recode associations concerned.
in line with psychoanalytic or other personality The overall research strategy also determined the
theories are important. And, from the attitude- way the attempts to develop the achievement
change research literature, we learn that such vari- motive have been organized. That is to say, in
ables as presenting one side or two, using reason order to process enough subjects to permit testing
or prestige to support an argument, or affiliating the effectiveness of various "inputs" in a reasonable
with a new reference group are crucial for develop- number of years, the training had to be both of
ing new attitudes (cf. Hovland, Janis, & Kelley, short duration (lasting 1-3 weeks) and designed
19S3). Despite the fact that many of these vari- for groups rather than for individuals as in person-
ables seem limited in application to the learning to-person counseling. Fortunately these require-
situation in which they were studied, we have tried ments coincide with normal practice in providing
to make use of information from all these sources short courses for business executives. To conform
in designing our "motive acquisition" program and further with that practice, the training has usually
in finding support for the general propositions that also been residential and voluntary. The design
have emerged from our study so far. For our problems introduced by the last characteristic we
purpose has been above all to produce an effect have tried to handle in the usual ways by putting
large enough to be measured. Thus we have tried half the volunteers on a waiting list or giving them
to profit by all that is known about how to facilitate a different, technique-oriented course, etc. So far
learning or produce personality or attitude change. we have given the course to develop n Achievement
For, if we could not obtain a substantial effect in some form or another some eight times to over
with all factors working to produce it, there would 140 managers or teachers of management in groups
be no point to studying the effects of each factor of 9-25 in the United States, Mexico, and India.
taken one at a time. Such a strategy also has For the most part the course has been offered by
the practical advantage that we are in the position a group of 2-4 consultant psychologists either to
of doing our best to "deliver the goods" to our executives in a single company as a company train-
course participants since they were giving us their ing program, or to executives from several different
time and attention to take part in a largely untried companies as a self-improvement program, or as
educational experience.2 part of the program of an institute or school de-
Our overall research strategy, therefore, is "sub- voted to training managers. The theoretical prop-
2 ositions which follow have evolved gradually from
Parenthetically, we have found several times that our
stated desire to evaluate the effectiveness of our course these pilot attempts to be effective in developing
created doubts in the minds of our sponsors that they did n Achievement among businessmen of various
not feel about many popular courses for managers that
no one has ever evaluated or plans to evaluate. An atti- cultural backgrounds.
tude of inquiry is not always an asset in education. It The first step in a motive development program
suggests one is not sure of his ground. is to create confidence that it will work. Our
324 AMERICAN PSYCHOLOGIST

initial efforts in this area were dictated by the it can change. In common-sense terms, belief in
simple practical consideration that we had to "sell" the possibility and desirability of change are
our course or nobody would take it. We were tremendously influential in changing a person.
not in the position of an animal psychologist who So we have used a variety of means to create
can order a dozen rats, or an academic psychologist this belief: the authority of research findings on
who has captive subjects in his classes, or even a the relationship of n Achievement to entrepre-
psychotherapist who has sick people knocking at neurial success, the suggestive power of member-
his door every day. So we explained to all who ship in an experimental group designed to show
would listen that we had every reason to believe an effect, the prestige of a great university, our own
from previous research that high n Achievement genuine enthusiasm for the course and our convic-
is related to effective entrepreneurship and that tion that it would work, as expressed privately
therefore business executives could expect to profit and in public speeches. In short, we were trying
from taking a course designed to understand and to make every use possible of what is sometimes
develop this important human characteristic. regarded as an "error" in such research—namely,
What started as a necessity led to the first the Hawthorne effect, experimenter bias, etc., be-
proposition dealing with how to bring about motive cause we believe it to be one of the most powerful
change. sources of change.
Proposition 1. The more reasons an individual Why? What is the effect on the person, theo-
has in advance to believe that he can, will, or retically speaking, of all this goal setting for him?
should develop a motive, the more educational at- Its primary function is probably to arouse what
tempts designed to develop that motive are likely exists of an associative network in the achievement
to succeed. The empirical support for this proposi- area for each person affected. That is, many
tion from other studies is quite impressive. It con- studies have shown that talk of achievement or
sists of (a) the prestige-suggestion studies show- affiliation or power tends to increase the frequency
ing that people will believe or do what prestigeful with which individuals think about achievement or
sources suggest (cf. Hovland et al., 1953); (b) the affiliation or power (cf. Atkinson, 1958). And the
so-called "Hawthorne effect" showing that people stronger the talk, the more the relevant associative
who feel they are especially selected to show an networks are aroused (McClelland et al., 1953).
effect will tend to show it (Roethlisberger & Such an arousal has several possible effects which
Dickson, 1947); (c) the "Hello-Goodbye" effect would facilitate learning: (a) It elicits what exists
in psychotherapy showing that patients who merely in the person of a "response" thus making it easier
have contact with a prestigeful medical authority to strengthen that response in subsequent learning.
improve significantly over waiting list controls and (b) It creates a discrepancy between a goal (a
almost as much as those who get prolonged therapy "Soll-lage" in Heckhausen's—1963—theory of mo-
(Frank, 1961); (d) the "experimenter bias" tivation) and a present state ("Ist-lage") which
studies which show that subjects will often do represents a cognitive dissonance the person tries
what an experimenter wants them to do, even to reduce (cf. Festinger, 1957); in common-sense
though neither he nor they know he is trying to terms he has an image clearly presented to him
influence them (Rosenthal, 1963); (e) the goal- of something he is not but should be. (c) It tends
setting studies which show that setting goals for to block out by simple interference other associa-
a person particularly in the name of prestigeful tions which would inhibit change—such as, "I'm
authorities like "science" or "research" improves too old to learn," "I never learned much from
performance (Kausler, 1959; Mierke, 1955); (/) going to school anyway," "What do these academics
the parent-child interaction studies which show that know about everyday life?" or "I hope they don't
parents who set higher standards of excellence for get personal about all this."
their sons are more likely to have sons with high After the course has been "sold" sufficiently to
n Achievement (Rosen & D'Andrade, 1959). The get a group together for training, the first step in
common factor in all these studies seems to be that the course itself is to present the research findings
goals are being set for the individual by sources in some detail on exactly how n Achievement is
he respects—goals which imply that his behavior related to certain types of successful entrepreneurial
should change for a variety of reasons and that performance. That is, the argument of The
 MOTIVE ACQUISITION 325

Achieving Society (McClelland, 1961) is presented this time, they are a little confused anyway as to
carefully with tables, charts, and diagrams, usu- exactly what we mean by the term. So we tell
ally in lecture form at the outset and with the help them they can find out for themselves by learning
of an educational TV film entitled the Need to to code stories written by others or by themselves.
Achieve. This is followed by discussion to clear They take the test for n Achievement before this
up any ambiguities that remain in their minds as session and then find out what their own score
far as the central argument is concerned. It is is by scoring this record. However, we point out
especially necessary to stress that not all high that if they think their score is too low, that
achievement is caused by high n Achievement— can be easily remedied, since we teach them how
that we have no evidence that high n Achievement to code and how to write stories saturated with
is an essential ingredient in success as a research n Achievement; in fact, that is one of the basic
scientist, professional, accountant, office or person- purposes of the course: to teach them to think
nel manager, etc.; that, on the contrary, it seems constantly in n Achievement terms. Another
rather narrowly related to entrepreneurial, sales, or aspect of the learning is discriminating achieve-
promotional success, and therefore should be of ment thinking from thinking in terms of power
particular interest to them because they hold jobs or affiliation. So usually the elements of these
which either have or could have an entrepreneurial other two coding schemes are also taught.
component. We rationalize this activity in terms Proposition 3. The more thoroughly an indi-
of the following proposition. vidual develops and clearly conceptualizes the
Proposition 2. The more an individual perceives associative network defining the motive, the more
that developing a motive is consistent with the likely he is to develop the motive. The original
demands of reality (and reason), the more educa- empirical support for this proposition came from
tional attempts designed to develop that motive are the radical behaviorist Skinnerian viewpoint: If the
likely to succeed. In a century in which psycholo- associative responses are the motive (by defini-
gists and social theorists have been impressed by tion), to strengthen them one should elicit them
the power of unreason, it is well to remember that and reinforce them, as one would shape up any
research has shown that rational arguments do response by reinforcement (cf. Skinner, 1953).
sway opinions, particularly among the doubtful or But, support for this proposition also derives from
the uncommitted (cf. Hovland et al., 19S3). Real- other sources, particularly the "set" experiments.
ity in the form of legal, military, or housing rules For decades laboratory psychologists have known
does modify white prejudice against Negroes (cf. that one of the easiest and most effective ways
Berelson & Steiner, 1964, p. 512). In being sur- to change behavior is to change the subject's set.
prised at Asch's discovery that many people will If he is responding to stimulus words with the
go along with a group in calling a shorter line names of animals, tell him to respond with the
longer than it is, we sometimes forget that under names of vegetables, or with words meaning the
most conditions their judgments conform with opposite, and he changes his behavior immediately
reality. The associative network which organizes and efficiently without a mistake. At a more
"reality"—which places the person correctly in complex level Orne (1962) had pointed out how
time, place, space, family, job, etc.—is one of the powerful a set like "This is an experiment" can be.
most dominant in the personality. It is the last He points out that if you were to go up to a
to go in psychosis. It should be of great assistance stranger and say something like "Lie down!" he
to tie any proposed change in an associative net- would in all probability either laugh or escape as
work in with this dominant schema in such a way soon as possible. But, if you say "This is an
as to make the change consistent with reality experiment! Lie down!" more often than not, if
demands or "reasonable" extensions of them. The there are other supporting cues, the person will
word "reasonable" here simply means extensions do so. Orne has demonstrated how subjects will
arrived at by the thought processes of proof, logic, perform nonsensical and fatiguing tasks for very
etc., which in adults have achieved a certain long periods of time under the set that "This is an
dominance of their own. experiment." At an even more complex level,
The next step in the course is to teach the sociolologists have demonstrated often how quickly
participants the n Achievement coding system. By a person will change his behavior as he adopts a
326 AMERICAN PSYCHOLOGIST

new role set (as a parent, a teacher, a public of- acts have limited influence because they often de-
ficial, etc.). In all these cases an associative net- pend on reality supports (as in typewriting), but
work exists usually with a label conveniently at- learned thoughts (symbolic acts) can occur any
tached which we call set and which, when it is time, any place, in any connection, and be applied
aroused or becomes salient, proceeds to control to whatever the person is doing. They are more
behavior very effectively. The purpose of this part generalizable. Acts can also be inhibited more
of our course is to give the subjects a set or a easily than thoughts. Isak Dinesen tells the story
carefully worked out associative network with ap- of the oracle who told the king he would get his
propriate words or labels to describe all its vari- wish so long as he never thought of the left eye of a
ous aspects (the coding labels for parts of the camel. Needless to say, the king did not get his
n Achievement scoring system like Ga1", P, etc.; cf. wish, but he could easily have obeyed her prohibi-
Atkinson, 1958). The power of words on con- tion if it had been to avoid looking at the left eye
trolling behavior has also been well documented of a camel. Thoughts once acquired gain more con-
(cf. Brown, 1958). trol over thoughts and actions than acquired acts
It is important to stress that it is not just the do because they are harder to inhibit. But why do
label (n Achievement) which is taught. The per- they gain control over actions? Are not thoughts
son must be able to produce easily and often the substitutes for actions? Cannot a man learn to
new associative network itself. It is here that our think achievement thoughts and still not act like
research comes closest to traditional therapy which an achiever in any way? The question is taken
could be understood as the prolonged and laborious up again under the next proposition, but it is well
formation of new associative networks to replace to remember here that thoughts are symbolic acts
anxiety-laden ones. That is, the person over time and that practice of symbolic acts facilitates
comes to form a new associative network covering performing the real acts (cf. Hovland, 1951, p.
his relations, for example, to his father and mother, 644).
which still later he may label an "unresolved The next step in the course is to tie thought to
Oedipus complex." When cues arise that formerly action. Research has shown that individuals high
would have produced anxiety-laden associations, in n Achievement tend to act in certain ways. For
they now evoke this new complex instead, blocking example, they prefer work situations where there
out the "bad" associations by associative interfer- is a challenge (moderate risk), concrete feedback
ence. But all therapists, whether Freudian or on how well they are doing, and opportunity to
Rogerian, insist that the person must learn to take personal responsibility for achieving the work
produce these associations in their new form, that goals. The participants in the course are there-
teaching the label is not enough. In fact, this is fore introduced to a "work" situation in the form
probably why so-called directive therapy is in- of a business game in which they will have an
effective: It tries to substitute new constructs opportunity to show these characteristics in action
("You should become an achiever") for old or more specifically to develop them through prac-
neurotic or ineffective ones ("rather than being tice and through observing others play it. The
such a slob") without changing the associative game is designed to mimic real life: They must
networks which underlie these surface labels. A order parts to make certain objects (e.g., a
change in set such as "Respond with names of Tinker Toy model bridge) after having estimated
vegetables" will not work unless the person has a how many they think they can construct in the
whole associative network which defines the mean- time allotted. They have a real chance to take
ing of the set. The relation of this argument is over, plan the whole game, learn from how well
obvious both to Kelly's (1955) insistence on the they are doing (use of feedback), and show a paper
importance of personal constructs and to the profit or loss at the end. While they are surprised
general semanticists' complaints about the neu- often that they should have to display their real
rotic effects of mislabeling or overabstraction action characteristics in this way in public, they
(Korzybsky, 1941). usually get emotionally involved in observing how
But, theoretically speaking, why should a change they behave under pressure of a more or less "real"
in set as an associative network be so influential work situation.
in controlling thought and action? The explana- Proposition 4. The more an individual can link
tion lies in part in its symbolic character. Learned the newly developed network to related actions, the
 MOTIVE ACQUISITION 327

more the change in both thought and action is likely tion that underlies the practice of most therapy
to occur and endure. The evidence for the impor- when it involves working through or clarifying,
tance of action for producing change consists of usually in terms of a new, partially formed con-
such diverse findings as (a) the importance of struct system, old memories, events from the last
recitation for human learning, (b) the repeated 24 hours, dreams, and hopes of the future. Again,
finding that overt commitment and participation theoretically, this should serve to enlarge and
in action changes attitudes effectively (cf. Berelson clarify the associative network and increase the
& Steiner, 1964, p. 576), and (c) early studies by number of cues in everyday life which will rearouse
Carr (cf. McGeoch & Irion, 1952) showing that it. The principle of symbolic practice can also be
simply to expose an organism to what is to be invoked to support its effectiveness in promoting
learned (e.g., trundling a rat through a maze) is transfer outside the learning experience.
nowhere near as effective as letting him explore For some time most course participants have
it for himself in action. been wondering what all this has to do with them
Theoretically, the action is represented in the personally. That is to say, the material is intro-
associative network by what associations precede, duced originally 6n a "take it or leave it" objective
accompany, and follow it. So including the acts basis as something that ought to be of interest to
in what is learned enlarges the associative network them. But, sooner or later, they must confront the
or the achievement construct to include action. issue as to what meaning n Achievement has in
Thus, the number of cues likely to trip off the their own personal lives. We do not force this
n Achievement network is increased. In common- choice on them nor do we think we are brain-
sense terms, whenever he works he now evaluates washing them to believe in n Achievement. We
what he is doing in achievement terms, and when- believe and we tell them we believe in the "obsti-
ever he thinks about achievement he tends to think nate audience" (cf. Bauer, 1964), in the ultimate
of its action consequences. capacity of people to resist persuasion or to do
So far the course instruction has remained fairly in the end what they really want to do. In fact,
abstract and removed from the everyday experi- we had one case in an early session of a man who
ences of businessmen. So, the next step is to apply at this point decided he was not an achievement-
what has been learned to everyday business activi- minded person and did not want to become one.
ties through the medium of the well-known case- He subsequently retired and became a chicken
study method popularized by the Harvard Business farmer to the relief of the business in which he
School, Actual examples of the development of had been an ineffective manager. We respected
the careers or firms of business leaders or entrepre- that decision and mention it in the course as a
neurs are written up in disguised form and assigned good example of honest self-evaluation. Never-
for discussion to the participants. Ordinarily, the theless, we do provide them with all kinds of
instructor is not interested in illustrating "good" or information as to their own achievement-related
"bad" managerial behavior—that is left to partici- behavior in the fantasy tests, in the business game,
pants to discuss—but in our use of the material, in occasional group dynamics session—and ample
we do try to label the various types of behavior opportunity and encouragement to think through
as illustrating either n Achievement and various what this information implies so far as their self-
aspects of the achievement sequence (instrumental concept is concerned and their responsibilities to
activity, blocks, etc.), or n Power, n Affiliation, etc. their jobs. Various devices such as the "Who am
The participants are also encouraged to bring in I?" test, silent group meditation, or individual
examples of managerial behavior from their own counseling have been introduced to facilitate this
experience to evaluate in motivational terms. self-confrontation.
Proposition 5. The more an individual can link Proposition 6. The more an individual can per-
the newly conceptualized association-action complex ceive and experience the newly conceptualized
(or motive) to events in his everyday life, the more motive as an improvement in the self-image, the
likely the motive complex is to influence his more the motive is likely to influence his juture
thoughts and actions in situations outside the train- thoughts and actions. Evidence on the importance
ing experience. The transfer-of-training research of the ego or the self-image on controlling behavior
literature is not very explicit on this point, though has been summarized by Allport (1943). In recent
it seems self-evident. Certainly, this is the proposi- years, Rogers and his group (Rogers, 1961; Rogers
328 AMERICAN PSYCHOLOGIST

& Dymond, 1954) have measured improvement in do not encourage n Achievement in their top execu-
psychotherapy largely in terms of improvement tives (Andrews, 1965), there is obviously a problem
of the self-concept in relation to the ideal self. here to be worked through if n Achievement is to
Indirect evidence of the importance of the self- survive among thoughts centered on dominance.
schema comes from the discussion over whether a The problem is not only rationally discussed. It
person can be made to do things under hypnosis is acted out in role-playing sessions where Mexicans
that are inconsistent with his self-concept or values. try, and often to their own surprise fail, to act like
All investigators agree that the hypnotist can be the democratic father with high standards in the
most successful in getting the subject to do what classic Rosen and D'Andrade (19S9) study on
might normally be a disapproved action if he makes parental behavior which develops high n Achieve-
the subject perceive the action as consistent with ment. Any technique is used which will serve
his self-image or values (cf. Berelson & Steiner, to draw attention to possible conflicts between
1963, p. 124). n Achievement and popular or traditional cultural
The same logic supports this proposition. It values. In the end it may come to discussing
seems unlikely that a newly formed associative parts of the Bhagavad Gita in India, or the Koran
network like n Achievement could persist and in- in Arab countries, that seem to oppose achievement
fluence behavior much unless it had somehow "come striving or entrepreneurial behavior.
to terms" with the pervasive superordinate network Proposition 7. The more an individual can per-
of associations denning the self. The logic is the ceive and experience the newly conceptualized
same as for Proposition 2 dealing with the reality motive as an improvement on prevailing cultural
construct system. The n Achievement associations values, the more the motive is likely to influence
must come to be experienced as related to or con- his future thoughts and actions. The cultural
sistent with the ideal self-image; otherwise associa- anthropologists for years have argued how impor-
tions from the self-system will constantly block tant it is to understand one's own cultural values
thoughts of achievement. The person might be to overcome prejudices, adopt more flexible atti-
thinking, for example: "I am not that kind of tudes, etc., but there is little hard evidence that
person; achievement means judging people in terms doing so changes a person's behavior. What exists
of how well they perform and I don't like to hurt comes indirectly from studies that show prejudice
people's feelings." can be decreased a little by information about eth-
Closely allied to the self-system is a whole nic groups (Berelson & Steiner, 1963, p. 517), or
series of networks only half conscious (i.e., cor- that repeatedly show an unconscious link between
rectly labeled) summarizing the values by which attitudes and the reference group (or subculture to
the person lives which derive from his culture and which one belongs—a link which presumably can
social milieu. These values can also interfere if be broken more easily by full information about it,
they are inconsistent with n Achievement as a especially when coupled with role-playing new atti-
newly acquired way of thinking. Therefore, it has tudes (cf. Berelson & Steiner, 1963, pp. 566 ff.).
been customary at this point in the course to intro- The theoretical explanation of this presumed ef-
duce a value analysis of the particiants' culture fect is the same as for Propositions 2 and 6. The
based on an analysis of children's stories, myths, newly learned associative complex to influence
popular religion, comparative attitude surveys, thought and action effectively must somehow be
customs, etc., more or less in line with traditional, adjusted to three superordinate networks that may
cultural anthropological practice (cf. Benedict, set off regularly interfering associations—namely,
1946; McClelland, 1964). For example, in Amer- the networks associated with reality, the self, and
ica we have to work through the problem of how the social reference group or subculture.
being achievement oriented seems to interfere with The course normally ends with each participant
being popular or liked by others which is highly preparing a written document outlining his goals
valued by Americans. In Mexico a central issue and life plans for the next 2 years. These plans
is the highly valued "male dominance" pattern may or may not include references to the achieve-
reflected in the patriarchal family and in the ment motive; they can be very tentative, but they
macho complex (being extremely masculine). are supposed to be quite specific and realistic; that
Since data show that dominant fathers have sons is to say, they should represent moderate levels of
with low n Achievement and authoritarian bosses aspiration following the practice established in
 MOTIVE ACQUISITION 329

learning about n Achievement of choosing the it is more likely to influence thought and action
moderately risky or challenging alternative. The than if it is not aroused.
purpose of this document is in part to formulate for As described so far the course appears to be
oneself the practical implications of the course devoted almost wholly to cognitive learning. Yet
before leaving it, but even more to provide a basis this is only part of the story. The "teachers" are
for the evaluation of their progress in the months all clinically oriented psychologists who also try
after the course. For it is explained to the partici- to practice whatever has been learned about the
pants that they are to regard themselves as "in type of human relationship that most facilitates
training" for the next 2 years, that 10-14 days is emotional learning. Both for practical and theo-
obviously too short a time to do more than con- retical reasons this relationship is structured as
ceive a new way of life: It represents the residential warm, honest, and nonevaluative, somewhat in the
portion of the training only. Our role over the manner described by Rogers (1961) and recom-
next 2 years will be to remind them every 6 months mended by distinguished therapists from St.
of the tasks they have set themselves by sending Ignatius 8 to Freud. That is to say, we insist that
them a questionnaire to fill out which will serve the only kind of change that can last or mean
to rearouse many of the issues discussed in the anything is what the person decides on and works
course and to give them information on how far out by himself, that we are there not to criticize
they have progressed toward achieving their goals. his past behavior or direct his future choices, but
Proposition 8. The more an individual commits to provide him with all sorts of information and
himself to achieving concrete goals in life related emotional support that will help him in his self-
to the newly formed motive, the more the motive confrontation. Since we recognize that self-study
is likely to influence his future thoughts and may be quite difficult and unsettling, we try to
actions. create an optimistic relaxed atmosphere in which
Proposition 9. The more an individual keeps a the person is warmly encouraged in his efforts and
record of his progress toward achieving goals to given the opportunity for personal counseling if
which he is committed, the more the newly formed he asks for it.
motive is likely to influence his future thoughts and Proposition 10. Changes in motives are more
actions. These propositions are both related to likely to occur in an interpersonal atmosphere in
what was called "pacing" in early studies of the which the individual feels warmly but honestly
psychology of work. That is, committing oneself supported and respected by others as a person
to a specific goal and then comparing one's per- capable of guiding and directing his own future
formance to that goal has been found to facilitate behavior. Despite the widespread belief in this-
learning (cf. Kausler, 1959), though most studies proposition among therapists (except for operant
of levels of aspiration have dealt with goal setting conditioners), one of the few studies that directly
as a result rather than as a "cause" of performance. supports it has been conducted by Ends and Page
At any rate, the beneficial effect of concrete feed- (1957) who found that an objective learning-theory
back on learning has been amply demonstrated by approach was less successful in treating chronic
psychologists from Thorndike to Skinner. Among alcoholics than a person-oriented, client-centered
humans the feedback on performance is especially approach. Rogers (1961) also summarizes other
effective if they have high n Achievement evidence that therapists who are warmer, more em-
(French, 1958), a fact which makes the relevance pathic, and genuine are more successful in their
of our request for feedback obvious to the course work. Hovland et al. (1953) report that the less
participants. 8
In his famous spiritual exercises which have played a
The theoretical justification for these propositions key role in producing and sustaining personality change in
is that in this way we are managing to keep the the Jesuit Order, St. Ignatius states: "The director of the
newly acquired associative network salient over the Exercizes ought not to urge the exercitant more to poverty
next 2 years. We are providing cues that will or any promise than to the contrary, nor to one state of
regularly rearouse it since he knows he is still part life or way of living more than another . . . [while it \&
proper to urge people outside the Exercizes] the director
of an experimental training group which is sup- of the Exercizes . . . without leaning to one side or the
posed to show a certain type of behavior (Proposi- other, should permit the Creator to deal directly with the-
tion 1 again). If the complex is rearoused suf- creature, and the creature directly with his Creator and
ficiently often back in the real world, we believe Lord."
330 AMERICAN PSYCHOLOGIST

manipulative the intent of a communicator, the routine of everyday life. So far as we know there
greater the tendency to accept his conclusions. is no scientific evidence to support this proposition,
There is also the direct evidence that parents of though again if one regards Jesuits as successful
boys with high n Achievement are warmer, more examples of personality change, the Order has fre-
encouraging and less directive (fathers only) quently followed the advice of St. Ignatius to the
than parents of boys with low n Achievement effect that "the progress made in the Exercizes
(Rosen & D'Andrade, 1959). We tried to model will be greater, the more the exercitant withdraws
ourselves after those parents on the theory that from all friends and acquaintances, and from all
what is associated with high n Achievement in worldly cares." Theory supports the proposition
children might be most likely to encourage its in two respects: Removing the person from every-
development in adulthood. This does not mean day routine (a) should decrease interfering associa-
permissiveness or promiscuous reinforcement of all tions (to say nothing of interfering appointments
kinds of behavior; it also means setting high and social obligations), and (b) should heighten
standards as the parents of the boys with high the salience of the experience by contrast with
n Achievement did but having the relaxed faith everyday life and make it harder to handle with
that the participants can achieve them. the usual defenses ("just one more course," etc.).
The theoretical justification for this proposition That is to say, the network of achievement-related
can take two lines: Either one argues that this associations can be more strongly and distinctly
degree of challenge to the self-schema produces aroused in contrast to everyday life, making cogni-
anxiety which needs to be reduced by warm sup- tive dissonance greater and therefore more in need
port of the person for effective learning to take of reduction by new learning. By the same token
place, or one interprets the warmth as a form of we have found that the dramatic quality of the
direct reinforcement for change following the experience cannot be sustained very long in a
operant-conditioning model. Perhaps both factors 12-18-hour-a-day schedule without a new routine
are operating. Certainly there is ample evidence attitude developing. Thus, we have found that a
to support the view that anxiety interferes with period somewhere between 6 to 14 days is optimal
learning (cf. Sarason, 1960) and that reward shapes for this kind of "spiritual retreat." St. Ignatius
behavior (cf. Bandura & Walters, 1963, pp. 283 ff.). sets an outside limit of 30 days, but this is when
One other characteristic of the course leads to the schedule is less intensive (as ours has some-
two further propositions. Efforts are made so far times been), consisting of only a few hours a day
as possible to define it as an "experience apart," over a longer period.
"an opportunity for self-study," or even a "spiritual Proposition 12. Changes in motives are more
retreat" (though that term can be used more ac- likely to occur and persist if the new motive is a
ceptably in India than in the United States). So far sign of membership in a new reference group. No
as possible it is held in an isolated resort hotel principle of change has stronger empirical or his-
or a hostel where there will be few distractions torical support than this one. Endless studies
from the outside world and few other guests. This have shown that people's opinions, attitudes, and
permits an atmosphere of total concentration on beliefs are a function of their reference group and
the objectives of the course including much informal that different attitudes are likely to arise and be
talk outside the sessions about Ga% Ga~, P, and sustained primarily when the person moves into
other categories in the coding definition. It still or affiliates with a new reference group (cf.
comes as a surprise to us to hear these terms sud- Berelson & Steiner, 1963, pp. S80 ff.). Many
denly in an informal group of participants talking theorists argue that the success of groups like
iaway in Spanish or Telugu. The effect of this Alcoholics Anonymous depends on the effectiveness
retreat from everyday life into a special and with which the group is organized so that each
specially labeled experience appears to be twofold: person demonstrates his membership in it by
It dramatizes or increases the salience of the new "saving" another alcoholic. Political experience
associative network and it tends to create a new has demonstrated that membership in small groups
reference group. like Communist or Nazi Party cells is one of the
Proposition 11. Changes in motives are more most effective ways to sustain changed attitudes and
likely to occur the more the setting dramatizes behavior.
the importance of self-study and lifts it out of the Our course attempts to achieve this result (a)
 MOTIVE ACQUISITION 331

TABLE 1
VARIABLES CONCEIVED AS ENTERING INTO THE MOTIVE CHANGE PROCESS

B
Input or independent variables Intervening variables Output or dependent variables

1. Goal setting for the person (PI, Pll) Arousal of associative network Duration and/or extensiveness of
(salience) changes in:
2. Acquisition of n Achievement asso- Experiencing and labeling the associa- 1. n Achievement associative net-
ciative network (P2, P3, P4, PS) tive network work
3. Relating new network to superordi- Variety of cues to which network is 2. Related actions: use of feedback,
nate networks linked moderate risk taking, etc.
reality (P2) Interfering associations assimilated or 3. Innovations (job improvements)
the self (P6) bypassed by reproductive interfer- 4. Use of time and money
cultural values (P7) ence 5. Entrepreneurial success as denned
4. Personal goal setting (P8) by nature of job held and its
5. Knowledge of progress (P3, P4, P9) rewards
6. Personal warmth and support (P10) Positive affect associated with net-
7. Support of reference group (Pll, work
P12)

Note.—Pl f Pll, etc., refer to the numbered propositions in the text.

by the group experience in isolation—creating the lack of space, partly because we are in an even
feeling of alumni who all went through it together; earlier stage of examining and classifying the effects
(b) by certain signs of identification with the of our training 1 and 2 years later preparatory to
group, particularly the language of the coding sys- conceptualizing more clearly what happens. It will
tem, but also including a certificate of membership; have to suffice to point out that we plan extensive
and (c) by arranging where possible to have par- comparisons over a 2-year period of the behaviors
ticipants come from the same community so that of our trained subjects compared with matched
they can form a "cell" when they return that will controls along the lines suggested in Column C.
serve as an immediate reference group to prevent What the table does is to give a brief overall
gradual undermining of the new network by other view of how we conceptualize the educational or
pressures. treatment process. What is particularly important
In theoretical terms a reference group should is that the propositions refer to operationally
be effective because its members constantly provide defined and separable treatment variables. Thus,
cues to each other to rearouse the associative after having demonstrated hopefully a large effect
network, because they will also reward each other of the total program, we can subtract a variable
for achievement-related thoughts and acts, and and see how much that decreases the impact of
because this constant mutual stimulation, and re- the course. That is to say, the course is designed
inforcement, plus the labeling of the group, will so that it could go ahead perfectly reasonably with
prevent assimilation of the network to bigger, very little advanced goal setting (PI), with an
older, and stronger networks (such as those objective rather than a warm personal atmosphere
associated with traditional cultural values). (Pll), without the business game tying thought
In summary, we have described an influence to action (P9), without learning to code n Achieve-
process which may be conceived in terms of ment and write achievement-related stories (P3),
"input," "intervening," and "output" variables as without cultural value analysis (P7), or an iso-
in Table 1. The propositions relate variables in lated residential setting (PI, Pll, P12). The
Column A via their effect on the intervening vari- study units are designed in a way that they can
ables in Column B to as yet loosely specified be- be omitted without destroying the viability of the
havior in Column C, which may be taken as evi- treatment which has never been true of other
dence that "development" of n Achievement has studies of the psychotherapeutic process (cf. Rogers
"really" taken place. The problems involved in & Dymond, 19S4).
evaluation of effects are as great and as compli- But is there any basis for thinking the program
cated as those involved in designing the treatment, works in practice? As yet, not enough time has
but they cannot be spelled out here, partly for elapsed to enable us to collect much data on long-
332 AMERICAN PSYCHOLOGIST

terra changes in personality and business activity. matically decided to quit his excellent job and go
However, we do know that businessmen can learn into the construction business on his own in a big
to write stories scoring high in n Achievement, way. A man with some means of his own, he had
that they retain this skill over 1 year or 2, and that had a very successful career as employee-relations
they like the course—but the same kinds of things manager for a large oil firm. His job involved
can be said about many unevaluated management adjusting management-employee difficulties, nego-
training courses. In two instances we have more tiating union contracts, etc. He was well-to-do,
objective data. Three courses were given to some well thought of in his company, and admired in
34 men from the Bombay area in early 1963. It the community, but he was restless because he
proved possible to develop a crude but objective found his job increasingly boring. At the time
and reliable coding system to record whether of the course his original n Achievement score was
each one had shown unusual entrepreneurial activ- not very high and he was thinking of retiring and
ity in the 2 years prior to the course or in the living in England where his son was studying. In
2 years after course. "Unusual" here means essen- an interview, 8 months later, he said the course
tially an unusual promotion or salary raise or had served not so much to "motivate" him but
starting a new business venture of some kind. Of to "crystallize" a lot of ideas he had vaguely or
the 30 on whom information was available in 1965, half consciously picked up about work and achieve-
27% had been unusually active before the course, ment all through his life. It provided him with
61% after the course (x2 = 11.2, p < .01). In a a new language (he still talked in terms of stand-
control group chosen at random from those who ards of excellence, blocks, moderate risk, goal
applied for the course in 1963, out of 11 on whom anticipation, etc.), a new construct which served
information has so far been obtained, 18% were to organize those ideas and explain to him why
active before 1963, 27% since 1963. he was bored with his job, despite his obvious
In a second case, four courses were given success. He decided he wanted to be an n-Achieve-
throughout 1964 to a total of 52 small businessmen ment-oriented person, that he would be unhappy
from the small city of Kakinada in Andhra in retirement, and that he should take a risk, quit
Pradesh, India. Of these men, 25% had been his job, and start in business on his own. He acted
unusually active in the 2-year period before the on his decision and in 6 months had drawn plans
course, and 65% were unusually active immedi- and raised over $1,000,000 to build the tallest
ately afterwards (x2 = 17.1, p < .01). More con- building in his large city to be called the "Everest
trol data and more refined measures are needed, Apartments." He is extremely happy in his new
but it looks very much as if, in India at least, activity because it means selling, promoting, trying
we will be dealing with a spontaneous "activation" to wangle scarce materials, etc. His first building
rate of only 25%-35% among entrepreneurs. is partway up and he is planning two more.
Thus we have a distinct advantage over psycho- Even a case as dramatic as this one does not
therapists who are trying to demonstrate an im- prove that the course produced the effect, despite
provement over a two-thirds spontaneous recovery his repeated use of the constructs he had learned,
rate. Our own data suggest that we will be un- but what is especially interesting about it is that
likely to get an improvement or "activation" rate he described what had happened to him in exactly
much above the two-thirds level commonly reported the terms the theory requires. He spoke not about
in therapy studies. That is, about one-third of a new motive force but about how existing ideas
the people in our courses have remained relatively had been crystallized into a new associative net-
unaffected. Nevertheless the two-thirds activated work, and it is this new network which is the new
after the course represent a doubling of the normal "motivating" force according to the theory.
rate of unusual entrepreneurial activity—no mean How generalizable are the propositions? They
achievement in the light of the current pessimism have purposely been stated generally so that some
among psychologists as to their ability to induce term like "attitude" or "personality characteristic"
lasting personality change among adults. could be substituted for the term "motive" through-
One case will illustrate how the course seems out, because we believe the propositions will hold
to affect people in practice. A short time after for other personality variables. In fact, most of
participating in one of our courses in India, a the supporting experimental evidence cited comes
47-year-old businessman rather suddenly and dra- from attempts to change other characteristics.
 MOTIVE ACQUISITION 333

Nevertheless, the propositions should hold best FESTINGER, L. A theory of cognitive dissonance. New
more narrowly for motives and especially the York: Harper & Row, 1957.
FRANK, J. Persuasion and healing. Baltimore: Johns
achievement motive. One of the biggest difficulties
Hopkins Press, 1961.
in the way of testing them more generally is that FRENCH, E. G. Effects of the interaction of motivation
not nearly as much is known about other human and feedback on task performance. In J. W. Atkinson
characteristics or their specific relevance for success (Ed.), Motives in fantasy, action and society. Princeton,
in a certain type of work. For example, next to N. J.: Van Nostrand, 1958. Pp. 400-408.
nothing is known about the need for power, its HECKHAUSEN, H. Eine Rahmentheorie der Motivation in
relation to success, let us say, in politics or bargain- zehn Thesen. Zeitschrift fur experimentelle und ange-
wandte Psychologic, 1963, X/4, 604-626.
ing situations, and its origins and course of develop-
HOVLAND, C. I. Human learning and retention. In S. S.
ment in the life history of individuals. It is pre- Stevens (Ed.), Handbook of experimental psychology.
cisely the knowledge we have about such matters New York: Wiley, 1951.
for the achievement motive that puts us in a po- HOVLAND, C. I., JANIS, I. L., & KELLEY, H. H. Communi-
sition to shape it for limited, socially and individu- cation and persuasion: Psychological studies of opinion
ally desirable ends. In the future, it seems to us, change. New Haven: Yale Univer. Press, 1953.
research in psychotherapy ought to follow a similar KAUSLER, D. H. Aspiration level as a determinant of per-
course. That is to say, rather than developing "all formance. Journal of Personality, 1959, 27, 346-351.
KELLEY, G. A. The psychology of personal constructs.
purpose" treatments, good for any person and any
New York: Norton, 1955.
purpose, it should aim to develop specific treat- KORZYBSKI, A. Science and sanity. Lancaster, Pa.: Science
ments or educational programs built on laboriously Press, 1941.
accumulated detailed knowledge of the character- MCCLELLAND, D. C. The achieving society. Princeton,
istic to be changed. It is in this spirit that the N. J.: Van Nostrand, 1961.
present research program in motive acquisition has MCCLELLAND, D. C. The roots of consciousness. Princeton,
been designed and is being tested out. N. J.: Van Nostrand, 1964.
MCCLELLAND, D. C., ATKINSON, J. W., CLARK, R. A., &
LOWELL, E. L. The achievement motive. New York:
REFERENCES
Appleton-Century, 1953.
AIXPORT, G. W. The ego in contemporary psychology. McGEOCH, J. A., & IRION, A. L. The psychology of human
Psychological Review, 1943, 50, 451-478. learning. (2nd ed.) New York: Longmans, Green,
ANDREWS, J. D. W. The achievement motive in two types 1952.
of organizations. Journal of Personality and Social Psy- MIERKE, K. Wille und Leistung. Gottingan: Verlag fiir
chology, 1965, in press. Psychologic, 1955.
ATKINSON, J. W. (Ed.) Motives in fantasy, action and ORNE, M. On the social psychology of the psychological
society. Princeton, N. J.: Van Nostrand, 19S8. experiment: With particular reference to demand charac-
BANDURA, A., & WAITERS, R. H. Social learning and per- teristics and their implications. American Psychologist,
sonality development. New York: Holt, Rinehart & 1962, 17, 776-783.
Winston, 1963. ROETHLISBERGER, F. J., & DicKSON, W. J. Management
BAUER, R. A. The obstinate audience: The influence and the worker. Cambridge: Harvard Univer. Press,
process from the point of view of social communication. 1947.
American Psychologist, 1964, 19, 319-329. ROGERS, C. R. On becoming a person. Boston: Houghton
BENEDICT, RUTH. The chrysanthemum and the sword. Mifflin, 1961.
Boston: Houghton Mifflin, 1946. ROGERS, C. R., & DYMOND, R. F. (Eds.) Psychotherapy
BERELSON, B., & STEINER, G. A. Human behavior: An and personality change. Chicago: Univer. Chicago Press,
inventory of scientific findings. New York: Harcourt, 1954.
Brace, 1964. ROSEN, B. C., & D'ANDRADE, R. G. The psychosocial
BROWN, R. W. Words and things. Glencoe, 111.: Free origins of achievement motivation. Sociometry, 1959,
Press, 1958. 22, 185-218.
BURRIS, R. W. The effect of counseling on achievement ROSENTHAL, R. On the social psychology of the psycho-
motivation. Unpublished doctoral dissertation, Univer- logical experiment: The experimenter's hypothesis as un-
sity of Indiana, 1958. intended determinant of experimental results. American
ENDS, E. J., & PAGE, C. W. A study of three types of Scientist, 1963, 51, 268-283.
group psychotherapy with hospitalized male inebriates. SARASON, I. Empirical findings and theoretical problems
Quarterly Journal on Alcohol, 1957, 18, 263-277. in the use of anxiety scales. Psychological Bulletin,
EYSENCK, H. J. The effects of psychotherapy: An evalua- 1960, 57, 403-415.
tion. Journal of Consulting Psychology, 1952, 16, SKINNER, B. F. Science and human behavior. New York:
324. Macmillan, 1953.