First Published in Journal of Aesthetic Education. Vol. 4, No. 2, Special Issue: Curriculum and Aesthetic Education (Apr. 1970) pp. 21-36.

The thesis of this article can be simply stated: If we are to build pro ­ grams in aesthetic education that are both scholarly and democratic, we'll have to learn how to think in other categories - that is, we must learn how to work at least part of the time in modes of thought other than our own. By "scholarly," in this context, I mean that whatever is taught should be true to its subject matter; and by "democratic" I mean that students and teachers in a general educational program should be free to confront the major alternative conceptions of art, of artistic learn ­ ing, and the nature of the human - indeed, of aesthetic education itself.

Certainly a major approach to curriculum-building and instruction is what has come to be called the behavioral conception of education. As a matter of fact, a 1969 Pre-Conference Institute of the National Art Education Association provided a week of advanced training in the be ­ havioral approach to art education. (Initial training was offered in four Regional Institutes in 1968.) The behavioral approach has its focus on the operational definition of educational objectives; that is, the desired outcomes of instruction are specified in the form of descriptions of ob ­ servable "terminal behaviors" of students. Interest in this approach is apparently widespread in the schools among professionals in all subject ­- matter areas.

Over the last several years I have worked with art teachers, curriculum specialists, scholars, and professionals in the arts attempting to develop curricula which relate the arts. My impression is that the arts and hu ­ manities may have special problems with the behavioral approach. Now I do believe that educators can profit from the study and application of this approach. But I am also concerned with some of its limitations. More generally, I am motivated to discover why the literature of aes ­ thetic education has thus far failed to present various concepts of educa ­ tion in the arts and humanities in a way that relates these concepts to the problems of educational change in the schools.

One requirement for educational change is agreement among curriculum innovators on what shall be taught. Now there is no dearth of powerful ideas on what constitutes an "aesthetic education," "human ­ istic learning," the "good life," and so on. But fundamental disagree ­ ment on the nature of the subject appears to be a characteristic trait of these ideas. Answers to the question "What is art?" are problematical in a way that answers to the question "What is science?" rarely are. (Evidently, scholars had little difficulty in agreeing on what the "new math" is, and it was a relatively short step to introducing the notion of sets, group theory, and even calculus into the public schools.) To say that the arts deal with values and the sciences deal with facts is a gross oversimplification. But the sciences, conceived as structures of knowl ­ edge or as ways of achieving knowledge, do lend themselves to the kind of objective treatment by curriculum innovators that cannot be applied to the arts and humanities. Objectivity, here, requires that the very multiplicity of competing ideas, values, attitudes, beliefs, styles, qualities, etc., be recognized as an essential "content" in deciding what shall be taught.

Other practical questions, whose answers provide the necessary condi ­ tions for educational change, are surely the following: To whom shall the content be taught? By whom? Inwhat order? By what methods? Under what conditions? Upon what criteria? With what objectives? And once again, in contrast to the sciences, viable yet alternative answers to these questions are abundantly available and curriculum innovation must proceed in light of these essentially competing positions.

I have said "must proceed." Actually the literature of the field gives evidence that we are not proceeding in this manner. I believe there are at least two explanations for this state of affairs. First, the scholar, re ­ searcher, or philosopher, in attempting to work out a coherent theory of, say, aesthetic education, is (quite rightly) pursuing his own conceptual interests. In so doing, the criterion of coherence forces him either to assimilate seemingly incompatible concepts and evidence into his theory or else reject them on logical or empirical grounds. The resulting array of disparate theories from different writers confronts the would-be in ­ novator with the apparently impossible job of justifying on theoretical grounds his allegiance to one theory rather than another.

A second explanation is that political expediency typically forces school personnel to set aside - or ignore - theoretical and scholarly con ­ siderations, especially in those school systems where innovation is given high priority. The new curriculum can then only reflect the collective ideas, talents, and interests of the local teachers. The resulting eclecti ­ cism is justified on "practical" grounds.

In the first instance, we have valuable concepts unrealized in practice; in the second instance, practices uninformed by concepts.

Because of my recent experiences with the problems of curriculum development in aesthetic education, I do not now believe that this dilemma can be resolved within the disciplines of aesthetics, philosophy, or learning theory. Nor do I believe that professionals in the creative and performing arts or the humanities are likely to resolve this dilemma. Rather, I believe that new strategies need to be worked out in the area of curriculum theory and development; strategies capable of relating the work of scholars and teachers, artists and students, education spe ­ cialists and administrators. (1) In short, these strategies need to have intellectual and creative as well as practical dimensions. Such cur ­ riculum strategies are difficult to develop because of the many concepts they inevitably involve. So I would like to begin by identifying what I take to be their minimal logical requirements.

A useful way to arrive at an understanding of the logical features of good theorizing is to analyze examples of what everyone will agree is patently bad theorizing. Specifically, what are some of the obviously incorrect ways in which words or concepts may be joined or related? An example of a logically unsound utterance in ordinary language is:

"Mangoes and fruit will be served for dessert."

Only a speaker un ­ familiar with English or with tropical fruit would make such a state ­ ment. Likewise, he would avoid such conjunctions as the following: poi and food, man and hombre, Guernica and painting. Taking each conjunction in turn we have examples of what might respectively be called a confusion in

(1) level of abstraction (one concept includes the other),

(2) co-equivalence or overlapping meaning (the words refer to the same entity), and

(3) specificity (a confusion in relating concrete and abstract nouns).

Native speakers of English are not expected to make such mistakes in ordinary conversation.

In talking about curriculum matters, however, these three "errors of conjunction" are easily and perhaps frequently made. Current state ­ ments of objectives for aesthetic education may be used to illustrate the need for careful conceptual analysis in curriculum-building. Consider this statement:

The curriculum for aesthetic education may be defined as all planned efforts to develop in students

(1) an understanding of the arts,

(2) the skills of artistic production and performance, and

(3) skills in the analysis and evaluation of art and aesthetic experience.

Now that we are armed with some notion of the kinds of errors that are possible when one conjoins curriculum goals, as in this example, we may ask whether this tripartite definition relates its components in a conceptually adequate manner. We may ask, first, how understandings are related to skills, or (to put it in different words) how knowing about art is related to knowing how to produce art. Does one concept logically entail the other in its meaning? Is one ability a necessary condition for the achieve ­ ment of the other ability? And what behaviors may be said to exemplify each of these objectives? It becomes clear that satisfactory answers to these questions are required before a logically adequate curriculum strategy or theory can be constructed. Itshould also be clear that more than definitions are needed to answer these questions; what is also needed is an elucidation of the central concepts and an analysis of the behavior or experience to which these concepts refer.

When curriculum specialists turn to learning theory, epistemology, and aesthetics for assistance, they risk committing other logical errors. This is because the potential for ambiguity multiplies when concepts from more than one discipline are related. For example, one may properly speak of the atomic number of oxygen but not of the atomic number of truth or of painting. This error could be called the "error of confound ­ ing universes of discourse." It is perpetrated when concepts are removed from their original theoretical contexts and made to appear in a "super theory" without having established their new logical and empirical credentials. These concepts-become-words retain a semblance of method ­ ological precision without its substance.

But what about the words "production" and "performance" and the words "analysis" and "evaluation" in our definition of aesthetic educa ­ tion? Can they be taken as labels for concepts that can be related and yet avoid the mixing of universes of discourse? To state this differently, can we so qualify the ordinary meanings of these words so as to develop a consistent and meaningful theory? And, at a more abstract level, is it possible to relate theoretically the concepts of "experience" and "be ­ havior"? The most powerful theories we have in education are be ­ havioristically based; yet people in the arts would argue that aesthetic experience is the heart of the arts - not mere behavior. However, be ­ havior does seem to be all we can observe and measure objectively; with ­ out measurement or observation of some kind, it would be difficult to estimate the success (or lack of it) of any program in aesthetic education one might devise.

Keeping in mind these logical strictures upon curriculum strategy or theory, and holding in abeyance the issue of "experience" vs. "behavior" in the arts, I would now like to suggest some possible sources of be ­ havioral models for building objectives and methods of study in aes ­ thetic education.

Art and music instruction in American public schools has been carried forth largely with the images of the creative studio artist and the per ­ forming musician acting as models of behavior for both student and teacher. Educators' images of the student as artist and the student as musician have not always reflected the complex and diverse creative processes of contemporary artists and musicians, and perhaps there are valid educational reasons for keeping them separate to some degree. The ideal of the artist-teacher, too, has its strengths and limitations. But whatever the historical reasons for such a development, the task of building a curriculum which has pretensions beyond the training of skills in artistic production and performance (to include, for example, teach ­ ing for understanding and appreciation of the arts) , must necessarily in ­ clude a consideration of many more possible kinds of models. Of course it does not follow that a curriculum of aesthetic education ought to con ­ tain everything it is possible to include in it. Yet the virtue of consider ­ ing all possibilities is that additional or more appropriate models might be found to approximate more closely the kinds of behaviors we expect to be the outcome of aesthetic education.

For example, if the ability to analyze and evaluate art is projected as a desirable skill to be learned, then obviously the art critic could be taken as a model for this behavior. But consider for a moment some other relevant models. The curator, the anthropologist, the aesthetician are each involved in the analysis and evaluation of the plastic arts. Yet each of these specialists exhibits his own distinctive procedures, techniques, and skills. Careful attention to the range of behaviors it is possible to include under the rubric "analysis and evaluation" might well protect us from oversimplifying our new models. We have, I believe, tended to oversimplify the model we presently employ - the artist.

It is also true, however, that the many kinds of professionals in the arts exhibit similar as well as different kinds of behavior, if we examine the range of behaviors within each professional category. To what extent one can identify "basic" or underlying linguistic and nonlinguistic be ­ haviors that these professionals share in common depends, in part, on the level of abstraction one employs in describing behavior in the arts. For example, the art historian, psychologist, and sociologist of art share identical activities of defining, hypothesizing, arguing, and reasoning at one level of description. Yet they are readily distinguishable at other descriptive levels; e.g., the kinds of artistic evidence they seek and the sorts of conclusions their inquiries yield.

In the initial stages of the behavioral models approach to curriculum building it is useful to identify each professional by name, e.g., dancer, composer, poet. Adequate representation of characteristic activities in the arts and humanities is thus assured. Once a pool of descriptions of activities is generated, however, more precision will be gained if these specific descriptions are clustered under more abstract descriptions with no reference to professional names. (2) Such items as "The student is creating a dance," ". . . composing a song," " . . . writing a poem," would therefore be grouped under the description "The student is pro ­ ducing an artwork." (Of course, descriptions of the activities involved in creating a dance could also be generated and would facilitate plan ­ ning at the instructional level.) Thus while the range of distinctive behaviors of professionals are kept in view, the general descriptions of "basic behavior" in the arts and humanities may then be considered as possible objectives for aesthetic education.

Perhaps the chief virtue of this "models" approach is that it allows for a clear distinction between what the case in arts education is today and what could be the case in the future, so that curricular decisions as to what should be the case are recognized for what they are - not historical or sociological facts, nor theoretical speculation, but value judgments which require justification. For example, the tendency in visual art education in recent times has been to assume that "knowing how" to paint is more valuable because somehow more fundamental than "knowing that" x, y, and z are true about painting and its history. In behavioral terms this is to assume that producing works of art is educationally more valuable than explaining works of art. We need make no such assumption. By setting aside at this stage the question of what behaviors ought to be taught and learned, we can more readily appreciate the fact that production and explanation are equally "basic"; i.e., their descriptions are at the same level of abstraction.

Ambiguity in curriculum-planning is further reduced when the behavioral redefinition of traditional goals is attempted. For example, art educators typically seek to develop a student's "creative ability." But what sorts of behavior are desired? Just what skills does the art teacher wish to impart? One wonders how much longer such techniques as easel painting will be considered important when the multimedia skills of film-making and editing, light projection, script-writing, acting, and dancing - not to mention the more orthodox techniques of combining materials - are now required to produce current avant-garde art objects and events. And what kinds of skills will be appropriate for the art of the future? Perhaps skills involving electronic circuitry, subliminal perception, laser beams, and computer programming. The task of identifying possible behavior objectives clearly calls for predictions of what might happen in the arts of the future.

Art educators also speak about developing a student's "understanding" of modem art. Does the student understand Expressionist art when he begins to paint that way or when he can analyze verbally what he has done in his paintings? When he can visually identify slides of great paintings in the Expressionist tradition or when he can write an essay on the history of this art style? The point of the behavioral approach is not that we must legislate the meaning of the words "create" and "understand"; rather it is that we should specify and differentiate the kinds of results we wish to achieve in particular courses of instruction in aesthetic education. For if the objectives of instruction are left vague or ambiguous, then almost any kind of activity may pass for teaching or learning.

Ideally, the models strategy would produce a taxonomy of activities that indicates the number of choices we have at any particular level of abstraction among the many possible behavioral objectives described. We can thereby avoid the errors of conjunction: i.e., selecting and conjoining non-behavioral statements that contain words with overlapping or synonymous meanings, or that mix abstract and concrete nouns or conceptions drawn from several disciplines.

These logical niceties notwithstanding, a ready rejoinder is due for our consideration. As the curriculum-builder selects objectives he is, of course, rejecting other objectives. And if he is to proceed in a rational manner, he must deal with a variety of problems involving the justification of his judgments - the relating of facts to theory, of decisions to value premises, of concepts to "leading ideas" or organizing principles. What becomes apparent in surveying the literature is that there are also alternative organizing principles or bases for building curricula. While "behavior" provided the basis for the models strategy outlined in the preceding discussion, other candidates are "knowledge" and "experience," concepts with a much longer history in curriculum development. On what grounds are we to select and relate or reject these concepts as possible bases for developing a coherent strategy?

In response to this question we might begin by examining the premises of behaviorism and noting its limitations. First, it must be acknowledged that behaviorism, the most powerful orientation in contemporary American psychology, strongly influences theory and practice in American education. This is no doubt due to the fact that the nature of human learning is a central concern of both psychology and education. Behavioristic conceptions of the human and learning dominate both fields to such an extent that disciples find it difficult to entertain seriously the alternative conceptions available to them. And, apart from those rare occasions when non-behavioristically oriented theorists debate doctrinal differences with behaviorists, (3) the ideational premises of workers in both fields go relatively unexamined, whether in research, teaching, or other professional activities.

The classical behaviorism of Watson (4) was built on the criterion of objectivity: only data gathered by independent observers of the same event or object would be considered as scientific. In striving to emulate the physical sciences, Watson argued that psychological concepts and laws should be expressed in terms of physical stimuli and the resulting muscular reactions and environmental effects. Thus, a truly scientific psychology was to have no room for subjective subject matter, and methods of self-observation or introspection were to be rigidly excluded. Apart from this methodological behaviorism, which conceived of the human as a series of stimulus-response sequences, a metaphysical behaviorism can be found. Watson's materialism is revealed in his belief that thinking was really covert speech - "talking to oneself' - and therefore a form of behavior; tongue and throat movements would be detected if only sensitive enough instruments could be devised. That is to say, what is real is physical.

The work of Watson, and later Guthrie, (5) drew strong support and inspiration from Pavlov's physiological studies of the conditioned reflex in animals. An "unconditioned stimulus" (e.g., food offered a hungry dog) causes an "unconditioned response" (e.g., salivation). Now, if a second stimulus (e.g., a ringing bell) is presented repeatedly just prior to the unconditioned stimulus, soon the subject will exhibit a conditioned reflex (e.g., salivation) even when the original stimulus is no longer offered. In other words, a response can be learned. The conditioned reflex principle became the behavioristic model for learning, human and otherwise. (Since controlled conditions are necessary for these experiments, quite naturally the bulk of behavioral research consists of experiments in animal learning.) Thus, psychology was reduced to a study of behavior which was further reduced to a study of learning.

The "conditioning" experiments of classical behaviorism produced much objective data but little in the way of a coherent theory with predictive principles. To remedy this situation, neo-behaviorists, such as Hull, (6) turned in the thirties to the "hypothetico-deductive method" of theory-building in physics as a model for theory-building in psychology. More accurately, neo-behaviorists turned to philosophers of science - especially to logical positivists, operationalists, and pragmatists for their model. While the S-R framework was retained, strict translatability of lawlike theoretical statements into observation statements was now demanded: dependent variables were operationally defined in terms of a measurable index of response, while independent variables were defined in terms of stimulus. (Some theorists introduced intervening variables, conceived as representing "internal" behavior, expressed as a mathematical function of other variables.) Several forms of behaviorism developed, e.g., the molar and molecular theories, but basic doctrine centered on the verification principle: the meaning of a theoretical statement is its method of verification. Or, in the words of Carnap, a logical positivist interested in the philosophy of psychology, "Every sentence of psychology may be formulated in physical language." (7)

Among leading contemporary psychologists, B. F. Skinner is the be ­ haviorist par excellence, in that he has not compromised (as, in his view, many have) the scientific criterion of objectivity by allowing "theoretical constructs," "intervening variables," or other non-empirical entities to enter into his work. In his words, "I am a radical behaviorist in the sense that I find no place in the formulation for anything which is mental."(8) For Skinner "psychology is a science of the behavior of organisms ... it is a part of biology, a natural science for which tested and highly successful methods are available."(9)

Unlike other behavioristic learning theories, the power of Skinner's theory has been demonstrated with students in school settings in his pioneering work with teaching machines. Since the principle of pro ­ grammed instruction is, I believe, the logical method upon which much of the work of behavioristically oriented educators is implicitly based, it is worth drawing out the educational implications of a standard defini ­ tion of teaching machines. Here is such a definition: "devices which

(1) present a unit of verbal or symbolic information visually, usually in question form;

(2) provide the student with some means of respond ­ ing to each unit; and

(3) inform the student as to the correctness of each response."(l0)

Ideally, then, content to be taught should be broken down into be ­ havioral units that students can readily master, from simple to more difficult materials. Each correct response to a question or task is "rein ­ forced" by informing the student that his response is correct. Learning is therefore equated with a specified performance of some kind and educational objectives are translated into desirable terminal behaviors.

It should be noted, here, that one of the striking weaknesses of our art and music programs is a general failure to articulate the cumulative learning that we wish students to achieve at each level from kindergarten through twelfth grade. For example, one is just as likely to see a third ­ grader engaged in junk printing as a sixth- or eighth-grader. More generally, if an independent observer were to view separately each of the great variety of art activities to be found in an elementary school, it is unlikely that he could reconstruct the curriculum plan from these observations alone. It is unlikely due to the fact that seldom is an art curriculum built on a sequence of learnings such that the learning of one skill or concept is identified as a necessary condition for the learning of a second skill or concept, and so on. Such a curriculum in the arts would require that the curriculum-builder describe the terminal be ­ haviors that he wished the learner to display at the end of the instruc ­ tional unit that he could not have displayed at the beginning of the unit, and that the sequence of units be built on the same principle. Perhaps the primary virtue of the behavioral approach to education is that objectives must be related to instruction and evaluation.

What, then, are the disadvantages of such an approach? It depends upon the issues with which one is concerned. Negative evaluations (not to say, attacks) may be gleaned from various sources, but our brief review will touch on some of the philosophical and psychological prob ­ lems raised.

One line of reasoning running through contemporary defenses of behaviorism begins with the rhetorical question: What else can one observe but behavior? It is rhetorical because any answer one offers must be framed in terms of observations. This restriction on what data count in the biological and social sciences and in education has been challenged by theorists as divergent in other respects as Carl Hempel, the philosopher of science, and Carl Rogers, the psychologist. One conclusion Hempel draws from his analysis (and rejection) of the verifi ­ cation principle of meaning is that "the content of a statement with empirical import cannot, in general, be exhaustively expressed by means of any class of observation sentences."(11) In Hempel's view, "the state ­ ments of empirical science have a surplus meaning over and above what can be expressed in terms of relevant observation sentences,"(12) that surplus residing in the totality of their logical relationships to all other statements in an empiricist language. Rogers appeals to quite different data. In his view, phenomenological inquiry "can investigate all the issues which are meaningless for the behaviorist - purposes, goals, values, choice, perceptions of self, perceptions of others, the per ­ sonal constructs with which we build our world, the responsibilities we accept or reject, the whole phenomenological world of the individual with its connective tissue of meaning. Not one aspect of this world is open to the strict behaviorist. Yet that these elements have significance for man's behavior seems certainly true."(13)

Surely the definitive phenomenological refutation of behavioral learn ­ ing theory (or at least those whose genesis is to be found in the condi ­ tioned reflex theory) is offered by the French philosopher, Maurice Merleau-Ponty, in his book The Structure of Behavior. (14) In rejecting the pholosophically naive realism of both common sense and much scientific thought, Merleau-Ponty argues the impossibility of reducing the whole of behavior to a sum of real parts - to physiology. While contiguity of stimulus and response can account for trial and error be ­ havior, it does not account for learning.

The decisive factor [for learning] lies in the manner in which fortuitous contiguities are utilized by the organism, in the elaboration which the organism makes them undergo [p. 100].

Even in these cases least favorable to our interpretation, either the "trials" involve no internal law, but then they never result in learning; or there is learning, in which case the organism must be capable, on the one hand, of creating relations among the different possible "solutions" and, on the other hand, among all of them and the "problem," by which their value is mea ­ sured. Even when the relation of signal to goal is a relation of pure succes ­ sion - we can presume that learning does not consist in recording contiguities. It is necessary that the succession become a "succession for the organism" [p. 101].

Scientific objectivity, for Merleau-Ponty, does not consist in restrict ­ ing one's attention to physical stimuli and behavioral responses, but rather requires attending to all the properties of learning phenomena. To talk of "color," "pressure," or of the "structure of a situation and its meaning," is for the Frenchman on an equal footing. The only appropriate scientific question is to ask "whether they are truly consti ­ tutive of the objects intended (vises) in an inter-subjective experience and necessary for their definition." [p. 102]

Merleau-Ponty's phenomenological analysis of human learning does not explain the higher in terms of the lower, nor the lower in terms of the higher processes:

Behavior, inasmuch as it has a structure, is not situated in either of these two orders. It does not unfold in objective time and space like a series of physical events; each moment does not occupy one and only one point in time; rather, at the decisive moment of learning, a "now" stands out from a series of the groupings which have preceded it as it engages and anticipates the future of the behavior; this "now" transforms the singular situation of the experience into a typical situation and the effective reaction into an aptitude . . . [p. 125].

Merleau-Ponty might well have been talking about learning as the residual consequence of resolving "problematic situations," of relating "means" to "ends" in human conduct. Indeed, it is instructive to compare the above passages with remarkably similar ones to be found in Dewey's writings.

On Dewey's experimentalist analysis of the pattern of inquiry, "Or ­ ganic interaction becomes inquiry when existential consequences are anticipated; when environing conditions are examined with reference to their potentialities; and when responsive activities are selected and ordered with reference to actualization of some of the potentialities, rather than others, in a final existential situation . . . "( 15) This passage appears in Logic: The Theory of Inquiry, a work that culminates his career-long study of "how we think."(16) An elaboration of Dewey's basic categories of existence is to be found in Experience and Nature, (17) where the contrast between the tenets of behaviorism in psychology and his position is striking. While behaviorism is "mind" -less and without "con ­ sciousness" - these words do not appear in the lexicon of S-R learning theory - Dewey characterizes the referents for these terms in their organic setting in experience:

While on the psycho-physical level, consciousness denotes the totality of actualized immediate qualitative differences, it denotes, upon the plane of mind, actualized apprehensions of meanings, that is, ideas. There is thus an obvious difference between mind and consciousness; meaning and an idea. Mind denotes the whole system of meanings as they are embodied in the workings of organic life; consciousness in a being with language denotes awareness or perception of meanings. It is the perception of actual events, whether past, contemporary or future, in their meanings, the having of actual ideas. The greater part of mind is only implicit in any conscious act or state; the field of mind - of operative meanings - is enormously wider than that of consciousness . . . Mind is, so to speak, structural, substantial; a constant background and foreground; perspective consciousness, a series of heres and nows . .. [p. 303].

A comparison of these passages reveals the great similarity of their respective analyses of "mind," "consciousness," "organism," "behavior," and "situation"; and it would be easy to find passages in Merleau- Ponty's Phenomenology of Perception (18) which correspond to these by Dewey:

. . . To perceive is to acknowledge unattained possibilities; it is to refer the present to consequences, apparition to issue, and thereby to behave in defer ­ ence to the connections of events. As an attitude, perception or awareness is predictive expectancy, wariness. Since potential consequences also mark the thing itself, and form its nature, the event thus marked becomes an object of contemplation; as meaning, future consequences already belong to the thing. The act of striving to bring them existentially into the world may be com ­ muted into esthetically enjoyed possession of form [Experience and Nature, p. 182]. Primarily meaning is intent and intent is not personal in a private and exclusive sense [p. 180].  

In staking out their respective "intentional" and "transactional" posi ­ tions, the philosophic motive of both Merleau-Ponty and Dewey was to avoid the traditional pitfalls of realism, idealism, and dualism. How ­ ever, the educational motive here is to indicate that their shared "experi ­ ential" conception of the human and learning (19) offers the curriculum ­ planner a major alternative to the reductive method of behavioristic psychology and education. This alternative could (not unfairly) be called experientialism.

Because behaviorism and experientialism hold radically opposed root conceptions of the human, it is to be expected that their application as alternative bases for curriculum development in aesthetic education will lead to differing notions of the job to be done. For the behaviorist, it is clear that the scholar, teacher, or curriculum specialist must pre ­ scribe a curriculum consisting of a series of behavioral objectives for the student. For the experientialist, on the other hand, it is the student's perceptions, anticipations, purposes, and problems that decide the mean ­ ing and value of elements in his learning situation. Inthe formal setting of the school, others may hypothesize what particular sequence of ex ­ periences would be appropriate for him, but he remains the arbiter of the success of the educational process. Thus, the two bases differ even on questions of what constitutes "the curriculum" and who qualifies as a "curriculum planner." The full and impartial recognition of the con ­ sequences of building curricula on one or another base would therefore require that curriculum innovators think critically in categories other than their own. An overall strategy would not only involve participants in the modes of thought of the theorist, scholar, and artist but also of those who work in the setting of the school, the student, teacher, and administrator.

To summarize: I have

(1) outlined the logical requirements for any consistent strategy for curriculum innovation,

(2) proposed a be ­ havioral models strategy for aesthetic education,

(3) examined the premises of behaviorism from the point of view of what I have called the "experiential" position.

I will conclude by indicating what I believe to be some of the problems inherent in attempting innovation by means of deducing curricular implications from a single theory.

I see the chief problem, here, as a dilemma: the dilemma of neu ­ trality vs. commitment in curriculum innovation. A curriculum theory, whatever else it might include, may be said to constitute a system of statements which addresses itself to these standard questions: What shall be taught? To whom? By whom? In what order? By what methods? Under what physical, social, intellectual conditions? To what degree of mastery, as measured by what criteria or standards? And so on. Since obviously more than one answer can be given to each of these questions, a logically consistent system of answers or statements would seem to entail that

(1) elaborated answers cannot be built upon ad hoc notions introduced piecemeal with disparate justifications drawn from a variety of sources, and

(2) the system must be organized or grounded in a value premise, doctrine, philosophy, or world view where all elements of the theory are deduced from this all-embracing position.

This conception of curriculum theorizing as the task of moving from a leading idea to the logical consequences of that idea (in terms of instruction and school organization) has perhaps been the strongest force in curriculum innovation in the history of American public schools. From Kilpatrick's idea that "we learn what we live" to Bruner's idea that "any subject can be taught effectively in some intellectually honest form to any child at any stage of development," curriculum theories tend to be viewed as essentially competing systems to be advanced, re ­ placed, or ignored. This monolithic deductive model of curriculum theory requires a leading idea that is so powerful and imaginative and persuasive that it is capable of providing an answer to any curriculum question put to it, since contradictory answers from competing theories or other sources must of logical necessity be discounted. To carry out such a large-scale innovation in the schools requires, also, a commit ­ ment on the part of large numbers of educators at every level. And this is at once the strength and weakness of the monolithic model. On the one hand, it would seem that wholehearted commitment to one idea only - not to say, indoctrination - is essential to widespread innova ­ tion in the schools. (It is difficult if not impossible to imagine know ­ ingly committing oneself to contradictory ideas.) On the other hand, a fatal weakness is the historical inflexibility of "true believers" when faced with new scientific evidence, philosophic and political analyses, and changing social and ethical values that come after the adoption of a monolithic theory and commitment to its logical consequences. Unfortunately there will always be at least some important new data, ideas, and values that cannot be absorbed into the established theory without distorting the meaning of either the theory or the new data, ideas, and values. Such discontinuity invariably leads to a breakdown of both theory and practice, and the resulting educational revolution brings with it a new monolithic theory.

What can the curriculum-builder learn from the historical record? First, that educators committed to a monolithic theory find it difficult to understand, test, or relate old or new findings from diverse disciplines, since their criterion for what is educationally useful is derived from their theory; second (and paradoxically), that commitment to a lead ­ ing idea seems required for large-scale educational innovation. He might conclude, then, that instead of building yet another such theory, he should seek a strategy or approach capable of producing curricula which functionally relate competing value premises or leading ideas and findings from various disciplines. This would allow individuals at the teaching, supervisory, and administrative levels with diverse skills and orientations in the arts to fully commit themselves to at least some aspects of the program. And to insure the vitality of such curricula, the strategy would also provide continuous evaluation at all levels for periodic modification of the aesthetic education program.

___________________________________________________________________________

Notes:

(1) Such a strategy, consisting of playing a "curriculum development game" with alternative behavioral, experiential, cognitive, and linguistic bases for building an aesthetic education program, was presented as a working paper at the CEMREL conference of August 1968 at Aspen, Colorado, and more formally as a Phi Delta Kappa Lecture at the Tenth Annual Research Symposium, Salt Lake City, April 12, 1969. The present article draws upon a portion of this material.

(2) Another advantage in dropping references to professionals at this stage is that the distinction between pre-professional training in the several arts in elec ­ tive courses and instruction for aesthetic education in the general curriculum is easier to maintain.

(3) See T. W. Wann (ed.), Behaviorism and Phenomenology: Contrasting Bases for Modern Psychology (Chicago: The University of Chicago Press, 1964).

(4) J. B. Watson, Psychology from the StandPoint of a Behaviorist (Philadel¬phia: Lippincott, 1919).

(5) S. Smith and E. R. Guthrie, General Psychology in Terms of Behavior (New York: Appleton, 1921).

(6) C. L. Hull, Principles of Behavior (New York: Appleton-Century, 1943).

(7) R. Carnap, "Psychology in Physical Language," Erkenntnis, III(1932-33).

Reprinted in Logical Positivism, A. J. Ayer (ed.), (Glencoe, Illinois: Free Press, 1959), p. 165.

(8) Comment of Skinner in a symposium whose proceedings are published in Behaviorism and Phenomenology, p. 106.

(9) Ibid., p. 79.

(10) Edward B. Fry, Glen B. Bryan, Joseph W. Rigney, Teaching Machines: An Annotated Bibliography, Suppl. No.1, Audio-Visual Communication Review, Vol. 8, 1960.

(11) Carl G. Hempel, "The Empiricist Criterion of Meaning," in Logical Positivism, p. 122.

(12) Ibid., p. 123.

(13) Behaviorism and Phenomenology, p. 119.

(14) Maurice Merleau-Ponty, The Structure of Behavior, trans. Alden Fisher (Boston: Beacon, 1963).

(15) John Dewey, Logic: The Theory of lnquiry (New York: Henry Holt, 1938), p. 107.

(16) John Dewey, How We Think (Boston: D. C. Heath, 1910).

(17) John Dewey, Experience and Nature (New York: Dover, 1958). First printed in 1929.

(18) Maurice Merleau-Ponty, Phenomenology of Perception (London: Rout ­ ledge and Kegan Paul, 1962).

(19) There are, to be sure, striking differences between Dewey's general orientation and other aspects of the phenomenological movement, especially as represented by Sartre, Heidegger, and their intellectual progenitor Husserl.