Music in Theory and Practice:

A Behavioral Viewt

by

JayRahn

If one regards a theory of music merely as a framework for

studying music, then one might acknowledge that several such theories

have already been advanced. The attempts of Guido Adler, Hugo

Riemann, Frank Harrison, Claude Palisca, Mantle Hood, Alan Merriam

and Charles Seeger are among the better known recent efforts to

construct a unified framework for the study of music. l However, such

attempts have resulted in views that are somewhat eclectic and

disjointed: eclectic, in that there appears to be no particular unity

among the various components of the individual frameworks, and

disjointed, insofar as relations among the various components of the

• I wish to thank J. D. Keehn of the Psychology Department, Atkinson

College, York University for his many detailed suggestions on the first
draft of this study. Any flaws that have survived his close scrutiny are
my own.

* Guido Adler, "Umfang, Methode und Ziel der Musikwissenschaft,"

Vierteljahrschrift fur Musikwissenschaft, vol. 1, no. 1 (1885), pp. 5-20;
Hugo Riemann, Grundriss der Musikwissenschaft (Leipzig: Verlag Quelle
und Meyer, 1914); Frank LI. Harrison, Mantle Hood and Claude V.
Palisca, Musicology (Englewood Cliffs, N. J.: Prentice-Hall, 1963);
Alan P. Merriam, The Anthropology of Music (Evans ton: Northwestern
University Press, 1964), pp. 17-35; Charles Seeger, "Toward a Unified
Field Theory for Musicology," Selected Reports, vol. 1, no. 3 (1970),
pp. 171-210. The latter lists several other efforts in the same vein.

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 106

individual frameworks are far from clear. Such endeavours also do not

meet the needs that I would envision for a theory of music in that they

largely ignore the connection between theory and practice. By contrast,

in the present study, I advance a theoretical outlook that, I hope, is

more thoroughly unified and integrated and that bears directly on the

practice of music.

Behaviorism.

An important feature of the outlook which I develop here is

that it adopts a stance which seems to have been, both intellectually and

emotionally, quite alien to many, if not most, students of music,

namely, behaviorism. By and large, the various orientations that have

been favored by musical scholars of late seem to be clearly opposed to

the various positions adopted by behaviorists. I am referring here to

views of music that are derived, directly or indirectly, from

introspectionism, phenomenology, Gestalt psychology, structuralism,

and so forth.

One can search far and wide in the fields of musicology,

ethnomusicology, musical semiotics, musical information theory,

musical aesthetics and music theory without encountering a formulation

that even dimly resembles a behavioral outlook. Indeed, apart from

certain specialized topics in music psychology, music education and

music therapy, behavioral approaches have been largely overlooked in

serious studies of music. I feel that this neglect of behavioral

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 107

approaches can be attributed to, among other factors, the diversity that

one encounters among behavioral formulations.

As G. E. Zuriff has shown in his recent comprehensive survey

of behaviorism, behaviorists have not always agreed with each other. 3

Whereas there are certain issues about which behaviorists are virtually

unanimous, indeed adamant, there are others that are a source of division

within the behaviorist camp. Accordingly, rather than providing a

detailed summary of the various behaviorist positions on issues relevant

to music, I attempt in what follows to show what sort of sense a

behaviorist such as I might make of musical phenomena. In so doing,

I give an account of music that accords with my own current position,

which I would identify as behaviorist, and I try, as well, to point out

certain areas where there would be major divergences among

behaviorists. Moreover, instead of presenting an account of

behaviorism followed by an application of the outlook to musical

situations, I attempt to show behaviorism "in action" in various

musical contexts from the outset of my discussion.

A Behavioral View of Various Musical Activities.

From a behavioral perspective, playing or singing a given tone

is a physical act, as is putting a note onto paper, iron filings onto

2 G. E. Zuriff, Behaviorism: A Conceptual Reconstruction (New York:

Columbia University Press, 1985).

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 108

tape, or a series of 0's and l's onto a computer disc. Getting a student

to perform, write, record or programme a tone in a given way is

achieved by arranging physical aspects of the student's situation in such

a manner that the task is accomplished (e.g., by means of praise,

censure, explanation or demonstration). In this way, a behavioral

orientation opens up the domain of musical phenomena about which

one might theorize. In a behavioral theory of music, one need not feel

that one is trapped inside musical tones; instead, one can deal seriously

with overt acts that are involved in performing, creating and teaching

music.

Other Musical Activities.

In a behavioral outlook, hearing a tone can be regarded as a

physical process that begins at the point where given sound wav

excite regions of the basilar membrane and provoke one's neurons to

fire at certain rates and in certain patterns. Imagining a sound i

considered to be a physical process akin to perceiving a sound, with the

important difference that there is no sound wave that corresponds to th

imagined sound in the immediate temporal and spatial vicinity of the

imaginative act. Remembering a tone is viewed as being much lik

perceiving or imagining a tone, with the difference that the remembere

sound is not perceived at the time of the remembering, but wa

perceived or imagined sometime in the past. Finally, one can note tha

"pre-composing," analyzing, and even theorizing about music might b

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 109

regarded as physical acts that involve the manipulation, and even the

creation, of symbols, which may or may not correspond to tones that

have been performed, composed, improvised, taught, perceived,

imagined or remembered by oneself or others. In this way, then, certain

private musical activities, which are often considered to be "occult"

(i.e., non-physical), can be included within a behavioral account.

Private Musical Activities.

In the case of perceiving, imagining, remembering, pre-

composing, analyzing or theorizing about music, a large part of one'

activity is private; that is, a great amount of one's activity is n

observed directly by others. However, with regard to perception, fo

example, certain parts of such processes (e.g., various excitations of th

basilar membrane) are physical rather than occult and have bee

observed by others. Furthermore, there is no prima facie reason

believe that other components of perception or other private proces

(e.g., imagining and remembering) are not physical. Nevertheles

internal, private processes are regarded in different ways by differen

sorts of behaviorists. This point can be illustrated by the various wa

in which behaviorists of differing stripes might account for an episo

where a person matches a given tone with his or her voice.

Suppose that a physical tone is produced and a person in the

immediate vicinity of the physical tone produces, by singing, an echo

"response" that matches the frequency of the original tone (or

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 110

"stimulus"). In a common-sense account, one might say that the

person "heard" the tone and adjusted the muscles of his or her vocal tract

in order to produce the matching tone. However, a so-called "radical

behaviorist" of the Watsonian variety might leave the "hearing" of the

stimulus-tone out of account and regard the incident as one in which the

stimulus-tone was a cause of the response-tone. Disregarding, for the

time being, other causes which such a radical behaviorist might cite,

one can note that an account of this sort would have the merit of being

quite parsimonious and that every aspect of the account would

correspond to an observable thing (or event). Both the stimulus-tone

and the response-tone (and, in principle, the relevant muscular

movements of the singer) would be directly observable by the

investigator, whereas any private "hearing" of the tone would be left out

of the account. Indeed, some radical behaviorists would deny that there

was any sort of internal event that corresponded to the common-sense

notion of what is entailed in "hearing a tone." However, here one

should distinguish carefully between the radical behaviorism of Watson,

the founder of the school, and that of Skinner, the foremost theorist of

the school.

The Watsonian version of radical behaviorism, now much

discredited for its denial of consciousness, has just been described,

whereas Skinner's radical behaviorism, which admits consciousness (and

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 Ill

consequently, "hearing") is implicit in much of what fol

this regard, one should observe that whereas a follower o

myself, would admit conscious acts, the preference wou

them as effects rather than as causes.

A so-called "methodological behaviorist" might ackn

that something more than the necessary adjustments of t

happened "inside" the singer between the time of the stim

the time of the response-tone, but would disregard this "s

because it was not observed. Some methodological behavi

say that the "something else" that happened was a ph

whereas other methodological behaviorists would sa

happened was non-physical and yet other methodological

would remain non-committal on the issue of physicality

all methodological behaviorists would be united in d

whatever might have happened between the time of the

and the time of the response-tone for the methodologic

what happened was not observed. Indeed, certain m

behaviorists would insist that what happened was

unobservable.

Other behaviorists would specify not only that something

more than muscular movements in the vocal tract happened between the

^ See J. D. Keehn, "Consciousness and Behaviourism," British Journal

of Psychology, vol. 55, no. 1 (1964), pp. 89-91.

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 112

appearances of the two tones, but would go so far as to assert without

apology that what happened was physical, and would take pains to

discern the specific physical process that was involved. Many

psychologists who study the physiology of perception could be

numbered among such behaviorists, though other sorts of behaviorists

would denigrate their efforts as unnecessary or, in view of the extremely

small amount that is known about the physiology of hearing,

excessively optimistic.

Finally, another sort of behaviorist might assert that

something more than adjustments of muscles in the vocal tract

happened between the occurrence of the two tones, and like the

methodological behaviorists, might regard the "something" that

happened as either physical or non-physical. However, rather than

leaving the account at the point where a methodological behaviorist

would leave it, such behaviorists would go on to specify that whatever

happened could be regarded as an "intervening variable," and such a

behaviorist might attempt to formulate a transformation rule on the

basis of which one could describe relations between the stimulus-tone

and the response-tone. A large number of so-called "cognitive

psychologists," who might themselves resist, and even resent, being

described as behaviorists, could be included in this category.

From the preceding account, it should be clear that

behaviorists of all sorts are methodologically rather fussy. Although

they might diverge from one another with regard to the sorts of

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 113

interpretations they would place on specific evidence and the sorts of

evidence they might consider relevant to understanding a given event,

their attitudes towards evidence as such are quite uniform. The

emphasis in a behavioral account is on direct evidence, that is, on

things and events that are directly observed, or, in principle, observable,

by the investigator. Such indirect evidence as the introspective reports

that are favored by many non-behaviorists is avoided by behaviorists

because of its putatively unreliable or even fictive character. If an

introspective report of a putatively private process is accepted by a

behaviorist, it is treated at face value, that is, as a response (e.g., a

verbal response) in its own right, rather than as indirect evidence for

some "deeper," "more important," or "more basic" process. And even if

a behaviorist of the cognitive variety posits an intervening variable,

such an entity is generally regarded as a hypothetical construct, and,

beyond this hypothetical status, there is generally no particular

commitment to its reality as an entity.

It should be clear that a behavioral approach might be

disturbing to those who hold a certain common-sense view of music.

What seems to be one of the most treasured notions in the workaday

discourse of musicians, namely, the notion of the "musical ear,"

appears in a behavioral account to refer to a) nothing at all, b) a merely

hypothetical construct, c) a process, whether physical or non-physical,

that is admitted to be known in a highly incomplete manner, or d) an

effect rather than a cause. In this way, the musical ear seems to

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 114

disappear if one adopts a behavioral orientation just as "earthiness" and

"fieriness" disappeared in the transition from Aristotelian science to

modern science. However, in the case of behaviorism, as has been the

case in modern science, an apparent loss is accompanied by an

explanatory gain, and this gain is explanatory in the deepest, causal

sense of the word.

Causal Explanation.

If behaviorism merely represented a methodological stance

towards private activity, it might be an interesting philosophical

curiosity. However, from the vantage-point of the behaviorist position

on certain methodological issues, several far-reaching findings have

resulted. Among the most important of these is that one can effectively

account for why a living being (either a human or one of the lower

animals) acts in the way that it does entirely in terms of specific aspects

of its inherited, genetic make-up, and more importantly for a

behaviorist, in terms of the history of its activity as an individual. In

particular, animate creatures tend to act in certain ways because of the

immediate consequences of their having acted in those ways during the

past.

In a behavioral analysis, certain activities are recognized as

being inherited and other activities are acknowledged to be spontaneous.

In line with the view of post-Darwinian biology, inherited activities are

regarded as having consequences for the survival of the species and are

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 115

accordingly selected by the environment. In addition, many activities,

both inherited and spontaneous, can be viewed as shaped or modified by

the environment during a being's lifetime. Depending on the

consequences of such activities, their frequency, strength and other

characteristics might change.

From a behavioral point of view, if one wants to know why,

for example, Debussy wrote a given piece as he did, one would inquire

not only into his genetic make-up (e.g., his musical talent) but also

into the consequences of his previous activities of a closely related sort

(e.g., what a non-behaviorist might describe as Debussy's previous

musical "experiences"). Similarly, if one would like to know why the

Lapps sing joiks in the way that they do, one would inquire into their

biological make-up, which would be expected to exhibit much the same

range of variation as that which is found in other human groups, and

into their individual "histories" of activity.

Problems in Applying a Behavioral Framework.

The behaviorist's insistence on the relevance of particular

details of past experience to current and future activity poses a number

of potential problems. First, since the entire history of an individual's

previous activity in all its detail is regarded as being, in principle,

potentially relevant to an account of that individual's later activity, it

might be very difficult to explain any particular act completely because

of gaps in one's knowledge of previous acts and their consequences.

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 116

The behaviorist's response to this situation is to insist on the greater

importance of more proximate activity and stimulation, that is, activity

and stimulation that are most similar to the activity and stimulation

being considered.

Second, determining which aspects of previous acts are

relevant to an account of a particular act might be extremely

problematic in view of the potential relevance of all aspects of the

individual's entire history of activity. For this reason, the behaviorist

undertakes controlled experimentation, changing one proximate variable

at a time, in order to study relations between specific causes and effects.

Third, predicting that a particular act will be the result of a

complex history of activity is clearly impossible in view of the absence

of complete knowledge of the causes that might determine that act.

Accordingly, behaviorists frame their causal formulations in terms of

probabilities, insisting that what is being caused is not, properly

speaking, an individual act, but rather stability or change in the

disposition to act in a certain way.

Fourth, in view of the suspect status of indirect evidence in a

behaviorist outlook, it might seem that historical research, where

indirect evidence looms very large, would be impossible within a

behavioral framework. However, a behavioral historian need not rely

on indirect evidence to any greater extent than those who hold differing

views. Further, the central preoccupation of behavioral methodology

need not be construed as an insistence on a restriction to what is directly

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 117

observed; instead, behavioral methodology can be understood as

insisting, as in A. J. Ayer's version of logical positivism, on what is

in principle observable. And in this regard, behaviorists can be

considered to make far fewer assumptions than their non-behaviorist

counterparts.4

Finally, a behavioral formulation would seem to be unable to

account for large-scale social or cultural activity. However,

behaviorists acknowledge that certain activities can be shared and

different members of a community can be regarded, in the absence of

contradictory evidence, as sharing similar histories of activity. By way

of summary, then, one can note that in a behavioral account of the

musical activity of Debussy or of Lapps, factors of talent and particular

aspects of proximate past activity and stimulation, whether individual

or shared, and whether observed or observable, would be considered to

alter the probability that a given act might take place.

Advantages of a Behavioral Approach.

Beyond its methodological scrupulousness and the growing

body of causal findings to which its methodological approach has given

rise, there are several advantages to adopting a behavioral approach. In

the first place, behaviorism yields causal accounts that demonstrably

"work" when applied in the "real world." Most especially, the

4 A. J. Ayer, The Central Questions of Philosophy (Harmonds worth:

Penguin, 1976), pp. 22-26.

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 118

behaviorist program has given rise directly to certain techniques of

teaching and therapy that have been spectacularly successful. The

relevance of these techniques to music education and music therapy is

immediately obvious and their application in these fields, not

surprisingly, has become commonplace. In addition, one can note that

the specifically behavioral notion of "discrimination" is widely applied,

whether deliberately or inadvertently, in the ordering of materials from

simple to complex or from easy to difficult in the most successful

curricula for performance, ear training (or dictation) and sight singing.

And in this regard, one can observe that a corresponding ordering of

topics from elementary to advanced has been a more or less constant

feature of music theory texts since Antiquity.

Behavioral accounts of activity in general go far to show that

behavioral principles operate whether one applies them deliberately or

not. In other words, whether or not one is a behaviorist, one's activity

can be effectively accounted for in terms of behavioral theory. Indeed, I

would maintain that the most successful performers, improvisors and

composers have been the best "musical engineers" in the sense that they

have controlled the responses of their listeners (including themselves) to

the greatest extent, whether or not any of the parties were, to use a non-

behaviorist term, "aware" of the control being exerted.

Another advantage to adopting a behavioral approach is that

one need not thereby deny many of the most engaging findings of other

areas of inquiry. As was indicated above, behaviorism is compatible

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 119

with post-Darwinian biology, and hence, by extension, the natural

sciences in general. In addition, the bulk of economics is consistent

with a behavioral outlook as are large tracts of political science.^ In

this way, the theory and practice of arts administration and the music

industry need not be cut off from other musical endeavors. Moreover,

some of the more promising approaches to sociology, anthropology,

semiotics, linguistics and philosophy are not only compatible with a

behavioral approach but also directly inspired by it. 6 Accordingly, the

ethnomusicologist, musical semiologist, and aesthetician of music need

not, in principle, be alienated from his or her field by virtue of adopting

a behavioral approach. Additionally, in its insistence on a thoroughly

probabilistic outlook, behaviorism is highly compatible with musical

information theory. Nevertheless, behaviorism seems to have been

remarkably unsuccessful in attracting many adherents in the fine arts

and humanities, and I feel that an important factor in this regard has

been the great emphasis accorded to causality in behavioral accounts.

^ See, for example, B. F. Skinner, Science and Human Behavior (New

York: MacMillan, 1953).
" George Caspar Homans, Social Behavior: Its Elementary Forms, 2nd
ed. (New York: Harcourt, Brace, Jovanovich, 1974); Marvin Harris,
Cultural Materialism: The Struggle for a Science of Culture (New York:
Vintage, 1980); Charles Morris, Foundations of the Theory of Signs
(Chicago: University of Chicago Press, 1938); B. F. Skinner, Verbal
Behavior (Englewood Cliffs, N.J.: Prentice-Hall, 1957); Willard Van
Oman Quine, Word and Object (Cambridge, Mass: M.I.T. Press, 1960).

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 120

Causal and Descriptive Accounts.

Perhaps because of their striking success in framing causal

accounts for a wide range of activity, behaviorists seem to have

denigrated the central role that is played by description in any serious

empirical inquiry. For example, the writings of B. F. Skinner, one of

the foremost behavioral theoreticians, contain many derogatory

references to description (or "topography") as opposed to causal

explanation.7 However, I would submit that a clear preference for a

causal theory over a descriptive theory is somewhat difficult to

maintain, for both sorts of theory are properly to be evaluated in terms

of their consistency and parsimony, the clarity (e.g., operationality) of

their terms, their verifiability and falsifiability, their scope, and so

forth. Indeed, it is not entirely clear that any distinction whatever need

be made between causal and descriptive theories if one has already

distinguished between relatively good and bad theories.

If one strips away the often embarrassing determinisms (e.g.,

"the chord of Nature") from many music theories, one finds that one is

frequently left with highly effective frameworks for describing pieces.

Indeed, I feel that much of what passes as music theory can be regarded

most charitably and fruitfully as a valuable framework for the

description of relations among tones in individual works.

7 B. F. Skinner, About Behaviorism (New York: Vintage, 1974), pp.

29, 71-75.

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 121

In the case of music, and in the case of much of the subject

matter of the fine arts and humanities, one is confronted with enormous

problems of description by virtue of the high degree of idiosyncracy and

complexity in the entities with which one is dealing. Those aspects of

a late Beethoven quartet that serve to distinguish it as a particular work

seem much more difficult to describe than the relevant aspects of an

experimental set-up in a study of operant conditioning. And indeed,

arriving at a convincing interpretation of an individual work or group of

works can represent an outstanding achievement apart from any

relevance that the description might have for a causal account of the

activity that surrounds it

I do not mean to deny that causality might enter into even the

most hermetic analysis of a work of art. Rather, I would insist that the

sorts of causal connections which might be invoked or implicit in such

an account could be quite low in level or even trivial (e.g., on the order

of noting that a given pair of tones produces a given interval as opposed

to asserting that the two tones in question were caused by a certain

independent variable or were the proximate cause of a dependent variable

more surprising than their highly predictable acoustical effect). In

addition, if one wishes to connect one's description of a piece with a

causal account that is more than trivial, one's description might have to

be extremely detailed, for otherwise a Beethoven quartet might be treated

erroneously as the equivalent of a far different work. In this regard, it

seems to me that one of the primary functions of music theory is

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 122

descriptive in this sense, for in order to determine why Beethoven might

have written a piece in a certain way or why listeners to the work might

have responded to it in a certain way or why it might be advantageous

for a composer to take the work as a point of departure for a new

composition, one must first determine just what is involved in the

work. And a consideration of this point leads to a hypothesis that I feel

is both truly musical and truly behavioral.

A Musico-Behavioral Hypothesis.

Central to a behavioral formulation is the notion that the

consequences of past activities can increase or decrease the probability

of such activities recurring in much the same form in the future. In

behavioral terms, a consequence that increases the probability of an

activity is a "reinforcer." Certain reinforcers (e.g., nourishment) are tied

to the survival of the species and can be regarded, accordingly, as

inherited. Others (e.g., currency in a monetary economy) are acquired

during one's lifetime, often in connection with inherited reinforcers. It

is fairly clear that certain consequences extrinsic to musical activity

(e.g., food and money) substantially reinforce (or in non-behaviorist

terms, "motivate") musical activity. However, I do not know of any

detailed formulation according to which consequences that are intrinsic

to musical activity have been advanced as reinforcers of musical

activity.

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 123

It seems plausible that the sheer exercise of fine motor activity

might reinforce the activities that precede it Indeed, such an hypothesis

is consistent with the observation that performers and composers often

have acquired habitual activities that seem to serve as a preparation to

playing, singing and setting notes onto paper. It would also seem that

the very fact that sounds are immediately produced as a consequence of

fine motor activity (as in the case of most performance) serves as a

reinforcer of that activity- witness the effect of the reverse, situation

where, for example, a piano key does not produce a sound when

depressed, and, as predicted by behavioral theory, frustration and anger

result. In addition, playing the correct notes would seem to be highly

reinforcing, as in the case where one's seeing of notes in a score

corresponds with both one's touch on an instrument and one's hearing

of the resulting tones (or even with one's imagining of such tones as in

the case where pianists practice on a silent keyboard). However, I do

not feel that such intrinsic reinforcers can fully account for humanity's

seemingly universal and frequent pursuit of musical activity, especially

in the instance of sheer hearing. Accordingly, I would include among

the most important reinforcers of musical activity (including the

maintenance of auditory attention) properties of the sounds themselves.

Whereas one might cite as reinforcing aspects of musical

sounds certain putative "emotions," "associations," "images" and so

forth that music has been considered to arouse, it is difficult, if not in

principle impossible, to study such things, nor is it entirely clear that

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 124

they are universal features of music. However, sheer structure would

appear to be an invariant property of musical sounds, and, indeed, of

things in general. No matter how modest a thing might be, it has

structure, that is certain amounts and kinds of complexity. A thing's

complexity might vary in amount and kind, but it is necessarily present

in the thing. Whether a particular structure is reinforcing might be a

matter of inheritance or lived "experience." However, even if it is a

matter of inheritance its manifestation might be shaped by experience.

Surely, one might be hungry without having an appetite for reptile

meat-not right now, thank you very much. And in the same way, one

might "crave" musical structure but not hunger after a particular kind or

amount of complexity. Moreover, just as one might be satiated by

food, and hence seek out a moderate amount of nourishment on a given

occasion or cease eating when one is full, so too might one tend to seek

out certain amounts of musical complexity at any given time or stop

listening altogether. Indeed, studies of musical style would suggest that

the kinds and amounts of complexity that have reinforced people have

tended to be normative in a given situation, and that these amounts and

kinds of complexity have varied greatly and coherently from situation to

situation. Further, it would seem that, in behaviorist fashion, one

must look to the environments in which music has been produced,

which include the sounds that have been "in the air" at any given time,

in order to determine the factors that have given rise to particular

amounts and kinds of musical complexity on any given occasion. And

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms
 125

according to the present formulation, such a determination would be of

central importance not only to the scholar but also to the practitioner of

music insofar as, for example, composers and performers can be

considered to shape musical structure and the responses of their

listeners.

Conclusion.

By way of conclusion, one can note that the austere and

exacting methodological outlook of behaviorism yields a view of th

world that is highly rich and variegated.^ Such an outlook discourages

one from positing, a priori, potentially stultifying dualisms, e.g.

between humanity and the rest of Nature. If one adopts a behavioral

framework, one can travel rather freely among the major curren

disciplines of scholarship. If one introduces sounds and structure into a

behavioral formulation, one can undertake the traditional enterprises of

musical scholarship within a behavioral context. And if one does so

one might begin to entertain the prospect of a theory of music that not

only accounts for, but predicts and controls musical practice.

8 See, for example, Gilbert Ryle, On Thinking (Oxford: Basi

Blackwell, 1979), and B. F. Skinner, Notebooks (Englewood Cliffs, N.J.
Prentice-Hall, 1980).

This content downloaded from 128.151.124.135 on Sat, 16 Mar 2019 00:26:17 UTC
All use subject to https://about.jstor.org/terms