Art, Behavior, and the Anthropologists

Current Anthropology 18 (1977): 387-407.

Denis Dutton

with the sciences or with the humanities? Most attempts to settle this question involve comparing these disciplines with the natural sciences on the one hand and with history on the other. If we take history as paradigmatic of the various forms of humanistic inquiry, we will certainly find many illuminating comparisons to be drawn between it and the social sciences, but history is not the only humanistic inquiry. In fact, there exists another whole realm of the humanities that has been almost universally neglected as an area which might provide revealing comparisons with sociology and anthropology: that is aesthetic criticism. In the discussion which follows, I want to begin to remedy some of that neglect.

The reluctance to use aesthetic criticism as a point of departure for discussing logical features of the social sciences stems largely from the ignorant belief (shared by more than a few social scientists) that criticism is a purely evaluative form of inquiry and therefore not comparable with a systematic study of human social activity. This, of course, is far from the case: evaluation is only a part (and often a minor part) of the critic’s primary business, which is to understand human artistic achievement through the careful description, analysis, and interpretation of aesthetic objects and activities. It is these latter aspects of critical inquiry which are so highly relevant to the ways in which anthropologists attempt to understand human social life. The present discussion cannot, however, treat all of the many aspects of criticism and the social sciences. Instead, attention will be concentrated on one of the most universally accepted notions in the practice and methodology of social science: the idea that social activities can be understood to exhibit underlying social functions which can be recognized and comprehended by a scientifically trained observer, but which are neither intended by those being observed nor recognized by them. I begin with one of the clearest and most concise statements of this position in the literature of social science methodology.

In his widely studied and influential Social Theory and Social Structure, Merton (1968: 105) distinguishes what he calls the “manifest” from the “latent” functions of a social activity:

Manifest functions are those objective consequences contributing to the adjustment or adaptation of the [social] system which are intended and recognized by participants in the system; Latent functions, correlatively, being those which are neither intended nor recognized.

This distinction enjoys very broad acceptance among social scientists, many of whom have suggested, moreover, that concern with latent functions is what distinguishes the work of modern social scientists from that of traditional historians. Whereas, it is argued, historians have supposedly been satisfied to study social activity in terms of the actors’ avowed purposes and recognized intentions for performing the activity, Merton claims that “the distinctive intellectual contributions of the sociologist are found primarily in the study of unintended consequences (among which are latent functions) of social practices....”

One of the examples Merton provides to illustrate the distinction between manifest and latent functions is the case of the Hopi rain dances. He tells us (pp. 118-19) that no one who wishes to understand the Hopi ceremonials will get very far if he limits himself to their expressed purposes, their manifest functions; he will be forced to write them off as mere superstition:

But with the concept of latent function, we continue our inquiry, examining the consequences of the ceremony not for the rain gods or for meteorological phenomena, but for the groups which conduct the ceremony. And here it may be found, as many observers indicate, that the ceremonial does indeed have functions-but functions which are non-purposed or latent.

Ceremonials may fulfill the latent function of reinforcing the group identity by providing a periodic occasion on which the scattered members of a group assemble to engage in a common activity. As Durkheim among others long since indicated, such ceremonials are a means by which collective expression is afforded the basic sentiments which, in a further analysis, are found to be a basic source of group unity.

It is this notion that the point of a tribal activity — by whomever’s description — might be something unintentional, unrecognized, or “non-purposed” which interests me. That people’s activities can result in unintended consequences goes without saying. It is also clear that certain aspects of intentional activity itself can be described as “unintended” or “non-purposed.” But Merton’s distinction and the sort of sociological analysis it purports to engender go beyond merely acknowledging these elementary facts. By applying Merton’s differentiation between manifest and latent functions to various examples from anthropology and from art criticism, we shall be able to see that there are some crucially important conceptual distinctions which his analysis overlooks.

Suppose we have a situation where Smith wishes to bring about condition B by performing activity A. Suppose further that in doing A Smith also brings about condition C. Does it follow that Smith intended to bring about C? The answer to that question, of course, depends on the relation C bears to A and B. For instance, Smith, who has overslept, wants to get to work on time, and in order to do so he drives to work at 80 miles per hour. His high speed is consciously intended — “intended and recognized,” in Merton’s phrase — and we may imagine its having certain results which Smith does not anticipate, results “unintended and unrecognized.” Smith’s car may throw a piston rod, or he may be given a ticket for driving recklessly. Smith does not want or intend to throw a rod or get a ticket. There are, however, other possibilities here. For example, there may be aspects of the situation which Smith recognizes but does not intend. He does not intend to wear down the tread of his tires, but he recognizes that this will happen on his way to work. Conversely, there may be aspects of situation which he does not recognize, in the he does not consciously contemplate them, but which are nevertheless intentional on his part. It may never occur to him that he stays on the right side of the road on his way to work; he does so, and quite intentionally, without ever considering it. A further complication arises in connection with the fact that different descriptions may be applied to the same piece of behavior. Smith may describe what he is doing as “trying to get to work on time” or “just driving a bit over the speed limit,” while the policeman who tickets him may describe it as “driving recklessly.” From the fact that Smith intended to drive at 80 miles per hour and the fact that driving at such a rate of speed may be “reckless driving,” it does not follow that he intended to drive “recklessly” (though in the present example this may assume an exceptional amount of ignorance on his part). Still, he cannot be said to have done something intentional which quite unintentionally resulted in his driving recklessly, in the way that it quite unintentionally resulted in his throwing a rod; to the contrary, he intentionally did something which, as it happens, constitutes reckless driving in the eyes of another.

It is remarkable the extent to which such basic distinctions as these are obscured by the form of sociological analysis advocated by Merton and accepted by a large number of social scientists, including a great many anthropologists. And yet it is distinctions such as these which I would think essential to an understanding of individual and social action and activity. Merton presents us with two alternatives in analyzing the functions or meaning of social conduct: the functions of social conduct are either covered by the expressed intent of the actors or are unintentional and probably recognized only by an outsider trained to perceive and identify them, the professional sociologist or anthropologist. Merton tells us that this dichotomy forms the foundation for much of the work of social science, and in fact he goes on at great length to explain how social scientists going back even before Veblen have employed a distinction between manifest and latent function, even though they have not used that specific terminology. Here I do believe Merton is right: it is a pattern among social scientists to try to dig deeper than the expressed intentions of the actors in a social situation and to regard whatever important meanings they discover as unintentional where those meanings are not covered by the explicit intentions of the actors.

Take, for example, Bunzel’s classic monograph, The Pueblo Potter (1929). She did her investigation of Pueblo pottery before Merton’s theoretical writing appeared, but despite the high quality of much of the research, she often displays attitudes toward the practices of primitive artisans that adumbrate the analytic distinctions advocated by Merton. Many of Bunzel’s descriptions and explanations reveal an implicit limitation on the space she is willing to allow for the concept of intention in Pueblo artistic practice. Here are some excerpts from the conclusion of her book 1929: 87, italics added) :

Of the principles of design which produce these characteristic forms the makers are, with but few exceptions, entirely unconscious. Their unconscious sensitivity to form is especially marked in the modeling of vessels. With no more definite guide than their perception of form, they reproduce accurately and without hesitation the characteristic vessel of their group. . . .

Except at Zuni, where designs are consciously built up of recognized elements, artists are equally unconscious of the principles governing the structure of their ornamental designs. . . . Everything, including the terminology of design, leads to the conclusion that decorative style is the product of unconscious and nonrational mental processes. Frequently the acceptance of the prevailing style is in distinct contradiction to the artist’s conscious and expressed intent.

Since the artist is largely unaware of the processes determining the character of her designs, her conscious preoccupation is chiefly with matters of technical structural, nondecorative) perfection. Theoretically there are no limitations upon the whim of the artist. We may conclude that the uniformity of style in any group is not due . . . to conscious desire on the part of the artists. . . .

Bunzel finds it surprising that the Pueblo artists can maintain such “uniformity of style,” inasmuch as they seem to be “entirely unconscious” of the “principles of design” used to produce pots of the particular style. What she means by this, however, is that she asked for an explanation of the stylistic principles employed by the various groups of artists (each Pueblo tribe having its own individual style) and except at Zuni none was forthcoming. Moreover, the Indians were completely unconcerned with style; many insisted that they were not trying to adhere to any style, but merely wanted to express themselves in their own ways. In Bunzel’s analysis it follows from this that adherence to style is unintentional (see also Dutton 1974).

But in what possible sense could such adherence to style be unintentional? Let us distinguish “explicit” from “implicit” intentions. Explicit intentions are those recognized by an actor in connection with his activity. Thus it is an explicit intention of Smith to drive at high speed in order to arrive at work on time. We note with no surprise, however, that in driving to his job Smith stays on the right side of the road, watches the center line, manages to keep right when cornering, and so on. These actions are not inadvertent by-products of his intentional activity. They are rather a part of his whole activity. Though they are not consciously contemplated, they are not unintentional; they are in fact implicitly intentional.

There are three important senses in which activities may be characterized by implicit intentions. The first of these is understood by noting that there is much in what people do — especially in activities involving the attainment of a specified end within a conventional or habitual frame of performance — which is implicitly intentional by virtue of its being taken for granted in performance, though it is not covered by an expression of the immediate purposes of the individual actor in what he is doing. Bunzel sees a “conflict” between each potter’s “conscious and expressed intent” to create pots in her own individual way and the fact that the pots appear so much alike, that they are all made in accordance with the prevailing style of the particular tribe. There need be no conflict here, any more than there need be a conflict between, say, a poet’s expressed intent “to communicate my thoughts and emotions in my own individual way” and the fact that his poetry observes the grammatical conventions of his native language. When an artist goes about his aesthetic activity and in so doing observes certain artistic conventions — adhering to a particular grammar and syntax, staying within the limitations of a prevailing style, and so forth — we cannot call such behavior unintentional. Again, Bunzel makes much of the Indians’ inability to provide anything resembling a rigorous account of the principles of their pottery style. It does not follow from this, however, that they are simply “unaware” of those principles or that they follow them unintentionally. Similarly, the poet may have never learned the rules of grammar in such a way as to be able to explain them in detail, yet he does understand them in the sense that he can write grammatically correct sentences, spot mistakes, and sort out good from bad sentences. Such ability is not, to use phrase, an “unconscious sensitivity” either to the forms of language in the case of the poet or to the forms of pots in the case of the Indians: it is highly conscious, though it may involve little or no ability to formulate or articulate grammatical or stylistic rules per se.

In some instances, the case of the man who drives to the right on the road for example, we expect that the actor can give a rational account of those elements in what he is doing which do not form part of his explicit intention. In others, such as providing a formulation of grammatical rules, a clear account may not be forthcoming. In either case, however, adherence to a pattern of activity is implicitly intentional in this first sense, that is to say, it is understood in action though not immediately contemplated, not “before the mind.” (For another discussion which seriously suffers from a failure to maintain this distinction, see [1973]. In Bateson’s treatment of primitive art, elements I have placed under this sense of “implicit intention” are unhelpfully annexed-along with perceptual mechanisms, dreams, and repressed to the “unconscious.")

The second sense of implicit intention involves the diverse ways in which the same intentional activity can be described. Smith describes his behavior as to get to work on time” or “just driving a bit over the speed limit.” It does not follow from this that since driving at 80 miles per hour is reckless driving, Smith therefore explicitly intended to drive recklessly. He may be surprised at the charge and may strenuously deny that his behavior amounts to reckless driving. Yet even here we could not say that his behavior simply results in reckless driving in the same way that it results in a thrown piston rod. This is because his behavior, seen from a perspective different from his own, reckless driving. In terms of the descriptions involved, he did not explicitly intend to “drive recklessly,” but he did intend to do certain things which a police officer deems to fall under that description. We thus cannot say that his reckless driving is an unintended by-product of certain intentional activities; quite to the contrary, those activities are precisely what the police mean by “reckless driving.”

Does this mean that any intentional behavior must be understood as unequivocally intentional under any possible description beyond the actor’s own? Of course not; but, on the other hand, there are anthropologists who, going to the other extreme, write as though behavior can only be understood as intentional under the actor’s own description and that behavior under any other description ought to be regarded as strictly unintended. Consider for a moment the possible answers one might give to the question, Do primitive peoples create works of art? Most primitive languages have no word which can be considered as a proper equivalent to our word “art.” In noting this we are reminded that art as an autonomous category of human activity is a mainly, though not an exclusively, European notion. Thus much of what we would call the “art” of primitive peoples comes to us tied with play, as a part of religion, in the production of cooking utensils or hunting gear, and so forth. That the primitive has no concept for art does not mean, however, that he cannot intentionally create it. He may describe what he is doing as “making a pot” or “carving a handle” and may even deny that his product possesses any special merit, but this does not detract from the possibility of great aesthetic achievement in his activity. Far from being an inadvertent byproduct of other purposes and activities, his intention to create something which we may correctly describe as a work of art is implicit in the care, sensitivity, and intelligence which he brings to his task. These elements of skill and care are precisely what give us reason to speak of his activity as “artistic,” and the way they reveal themselves in the finished product justifies us in calling it a work of art. This, after all, is what constitutes our notion of art, and it makes it possible for us correctly to apply to a pot or knife handle a description which the primitive does not have available in his own vocabulary.

Merton, as noted, treats the expressed intentions of individuals performing an activity (the manifest functions) as exhaustive of the meaning of that activity for its performers and regards other important meanings which it may have in their lives or for their society as unintended (latent functions). But it is not enough to take an aspect of what people do, redescribe it in a language they do not know or use, and call that re-described aspect of their activity “unintentional.” For example, Merton says that a latent function of the Hopi rain dances may be that they provide “a periodic occasion on which the scattered members of the group assemble to engage in a common activity” and that they thus may be “a basic source of group unity.” Never does Merton, nor do any of the other methodolo-gists who have made use of this “classic” case, provide a shred of evidence that the Hopi themselves are unaware that the dances not only purportedly bring rain but are important for these other reasons too. It is simply assumed that they are unaware of these things and do not intend them. This cannot be assumed, but must be shown; and in fact it never is.

In terms of thediscussion so far presented, we can see two possibilities which Merton’s analysis ignores. In the first place, such functions as those called “latent” by Merton may be taken for granted by the Indians and seldom, if ever, considered by them. In this way the functions may be implicitly intentional in the first sense outlined above. Second, if the Indians ever did think or talk of these matters, we could well expect that they would not do so in the language of the sociologist. No Hopi would speak of the dances as important because they “reinforce group identity by providing a periodic occasion on which the scattered members of the group assemble to engage in a common activity” (Merton’s description), but Hopi might well think to themselves or say to each other how important it is that the Hopi people come together or how nice it will be to see friends or relatives at a forthcoming dance. It is not at all difficult to imagine their being aware that the dances are important to keep the Hopi nation together as well as to make rain. Though the terms “group unity” or “means by which collective expression is afforded . . . basic sentiments” may be foreign to the Indians, it does not follow that these things, when accomplished by the rituals, are accomplished unintentionally; it does not follow simply from the fact that the Hopi describe and explain them somewhat differently than the anthropologists that they are unaware of the importance of these factors in their social life.

This is not a trivial point: much of what passes as the analysis of latent functions in sociology and anthropology is simply redescription in abstruse jargon of aspects of life which ordinary people are often quite aware of and may take for granted. A good example of just this sort of thing is provided by who in his discussion of the function distinction follows Merton in claiming that the sociologist makes his “discoveries” by “tracing” the “unintended consequences of our actions.” Here is a brief example he provides: . . is doubtful if the Bible Belt farmers realize that their religious rituals promote social cohesion. . . . The Kansas farmers would probably be baffled and astonished at the news that they are ritually reaffirming the ties of the group” (p. 220). Of course, unless they have sociology or anthropology degrees from their Bible Belt colleges, they may be baffled by the phrases “ritual reaffirmation” or “social cohesion,” but it is probable that few of them would deny that, beyond the obvious religious significance, revival meetings are important “just so folks can get together.” They might further be willing to acknowledge the importance of the meetings in developing and maintaining the strength of their church and their community. But what does the tedious jargon of the social scientist mean, if not just that? Again, it must be stressed that I am not claiming that the Bible Belt farmers have contemplated every aspect of their revival meetings which the considers in his analysis. They may have, though in very different terms, and there is no ground to suppose that they do not “realize” those aspects of their activities which the sociologist takes to be latent functions or that they would be “baffled"or “astonished” by the sociologist’s analysis once they had grasped his jargon. (This issue is insightfully dealt with by

It is interesting to note here that speaks of “tracing” the unintended consequences of action. This would seem to call to mind the work of the natural scientist tracing the connections between events in a causal chain, but examples involve no tracing in this sense at all; they are in fact merely matters of taking an intentional activity which has been redescribed in abstruse language and supposing it to be unintentional under the new description. My point is that the activity may well be implicitly intentional even under the new description in the sense that the actor’s intentional behavior may constitute something covered by the sociologist’s or anthropologist’s redescription. This does not in itself necessarily involve the discovery of matters of fact, but may be little more than the invention of words. Though neither the Indians nor the farmers may know much sociology, we are given no evidence that they know as little of their own purposes and motivations as Merton and would encourage us to think.

With the two kinds of implicit intention I have so far discussed, the situation has formulated such that either we have an aspect of an activity which is not covered by the actor’s explicit intentions inasmuch as it is taken for granted by the actor, or we have a situation where an intentional activity is merely redescribed and called unintentional under the new description. In both of these situations nothing I have suggested would indicate that the actor himself would be unwilling to acknowledge that the implicit intention in question does indeed make up part of his whole intention in doing what he does. The Pueblo potter might well be willing to agree, after the point has been carefully explained to her, that it is part of her intention to follow the prevailing style of her tribe in making her pottery (and Bunzel never indicates that she made any particular attempt to explain this to the Indians). The same may go for the-second of implicit intention: there is no reason to assume, as Merton appears to and as does quite expressly, that people who engage in a social activity will deny the social scientist’s analysis of their activity once they have understood his jargon. Now I wish to push this point even farther and in so doing introduce a third and final sense of implicit intention which is not adequately recognized by Merton’s function distinction.

Suppose we find in a group activity such as the Hopi rain dance a socially important aspect of the activity which persists, is seen time and again in the performance of the activity. Suppose further that we confront the Indians themselves with our interpretation of what they are doing: we explain that besides the supposed rainmaking power of the dances we have also observed that they are a basic source of group unity, that they serve to allow for the organized release of aggressions, or whatever. And suppose the Indians understand full well what we are saying but vehemently deny that the rituals are important in this respect-claiming, perhaps, that the Hopi nation is united by blood ties, that they do not have any aggressions to release, and that the dances are only important to the society in that they effect the rains on which the welfare of the tribe depends. If we are certain enough of our interpretation, are we justified in saying now that the function of the dances in promoting group unity is latent in Merton’s sense? In such a case would we not be driven to the conclusion that these important and persistent nonmeteorological functions of the dances are “neither intended nor recognized"? Here again I shall argue that the answer is no.

It is a general presupposition of critical practice that when one notices a particularly significant aspect of a work of art one always treats it as intentional unless one has good reason not to. For example, there are many significant relationships between the many leitmotivs of Wagner’s Ring cycle. Suppose Wagner himself denied that there were meaningful relations between these melodic figures and chord progressions and claimed that any we thought we discerned were mere accident, that our critical speculations were on a level with observing the shapes of animals accidentally produced by cloud formations. We might accept Wagner’s denial and proceed on the hypothesis that the connections we find so telling in relation to the whole of the Ring are mere accidents. But such a hypothesis soon begins to look rather implausible: such “accidents” happen with all too much regularity in Wagner’s music-dramas. In point of critical fact, such important melodic and harmonic relationships as are found in the Ring cannot with any plausibility at all be treated as inadvertent, but must be understood as part of the achievement of Wagner’s musical art. To take another example, Arvin in his introduction to Moby Dick, quotes a letter in which Melville claims that in writing the novel he was but “vaguely"conscious that it would be possible to give the book “an allegorical construction.” Arvin insists that such a remark must have been made tongue-in-cheek, and it may well have been, but I do not think that we need be driven to that conclusion. Even if Melville were telling the truth, we would still want to say that the possibilities for allegorical readings of Moby Dick are part of what makes the novel great: and it is not great by accident-those possibilities are part of Melville’s literary achievement.

Of course, we cannot rule out the possibility of happy accidents in art. We must not forget, however, that in the practice of aesthetic criticism one never attributes some excellent quality of an art work to chance unless there is strong evidence for doing so, and even the disavowals of artists are often not strong enough evidence. We can imagine stories: a drop of ink spatters on the score just before it is sent to the printer, and centuries later musicologists write learned papers on how Scarlatti prefigures Schoenberg in one of his sonatas. Such mistakes (which doubtless do exist) must, however, come to us as isolated instances. If Scarlatti sonatas are persistently seen to contain Schoenbergisms in characteristic passages, then stories of periodic accidents with the inkwell will leave us unconvinced.

I have stressed the word “achievement.” Works of art are among the greatest of human achievements, and treating them and the elements which make them up as intentional is part of what is meant by respecting them. Such respect may hold regardless of facts of the artist’s biography, regardless of his apparent ignorance of the possibilities for interpretation implicit in his work, regardless even of what he may say in repudiation of his work or its interpreters. Anything truly accidental produced by the artist or the work would not be part of the achievement of the work, but would properly be said to be an extraneous result, a byproduct, of the work. Achievement is always relative to the possibility of failure in art; achievement and failure are concepts internal to the whole enterprise of art. To say, moreover, that some aspect or feature of an art object or body of art is particularly valuable or significant (except for cases which can be confirmed as the result of pure chance) is almost inevitably to respect that aspect or feature as intentional with the object or body of art.

Let us approach the issue in more general terms. In criticism one commonly distinguishes the purposes of the artist from the purposiveness of the art work. Earning a living may have been purpose in composing a piece, and Wagner may have had no other purpose in mind for his leitmotivs than to announce the entry of characters and ideas in the drama. Melville may have had no other purpose in writing Moby Dick than to tell a good whaling tale. But the larger significance we find in these works or their elements, the purposes and meanings they exhibit that go quite beyond the explicit intentions of the artist, these we do not regard as unintentional or inadvertent. To the contrary, the purposiveness which we find in great works of art is in criticism treated as implicitly intentional activity on the part of the artist. This is the third sense of implicit intention I have in mind: the purposiveness which human activity, aesthetic or otherwise, reveals. Such purposiveness, though it may not be covered by the explicit intentions of the actor, is understood as intentional by virtue of its persistent intelligence and directedness. The obvious purposiveness which human activity can reveal often speaks louder, as it were, than the avowed intentions or purposes of the actors. In criticism when we talk about the meaning of the work beyond what the artist says he intends, we do not do so with the idea that such meaning is unintended. The purposiveness of the work reveals clear marks of intention and may thus be regarded as part of the achievement of the artist, even despite his disavowals.

Having been reminded of the prevailing attitude (in the main European tradition) of art critics toward art, let us turn to some remarks made by anthropologists and others in the course of a symposium on tribal art held in London in 1957. First a statement by Murray (1961: 97-98) on African carving:

The obvious competence of a “good” Yoruba carver and his justified confidence in his mastery suggest that his point of view is one which Yoruba society finds desirable; that the finished, precise style of Yoruba carving is what Yoruba find aesthetically satisfying. The esteem in which such works are usually held and the fact that they are more likely to be found in the houses of important than poor men is confirmation of this. Modern European artists, however, find such carvings too slick and prefer what must be, to Yoruba, inferior works. In the tentativeness, amateurishness, and what are probably blunders of such works the European finds sensitiveness and feeling.

Notice that Murray is not placing a particular Yoruba carving before us to tell us that some feature we may find attractive in it is accidental or a happy mistake on the part of the carver, one the carver himself would have wished to avoid. No, he is claiming that in a whole school of art what the artists consider valuable is something critics of another culture care little or nothing about, and that conversely what those critics value so highly as “sensitiveness and feeling” is really inadvertent blundering on the part of the artists. If anything, however, the evidence Murray supplies for his views seems only to support the belief that rich have no better taste in aesthetic matters than the rich do elsewhere. That the native rich of Nigeria prefer shiny carvings does not prove that shiny carvings are aesthetically excellent to Nigerians of taste any more than the preference of the American rich for Cadillacs proves those cars-to be aesthetically excellent to Americans of it may only show that shiny carvings and shiny Cadillacs are ways for the rich to display their wealth. Murray’s argument depends on a comparison of the preferences of rich Nigerians with the preferences of European aesthetes; it ought to be based on a comparison between the aesthetes of both these cultures (this is a mistake scrupulously avoided by Child and Siroto That fallacy interests me less, however, than Murray’s contention that Europeans could consistently find the mistakes of Yoruba carvings so full of “sensitiveness and feeling.” Presumably, no Yoruba artist ever tries on purpose to carve amateurishly, and intention is incompatible with the concept of blundering, so what Europeans find admirable and valuable in Yoruba is, on this account, quite unintentional on the part of the carvers. Yet the argument implies that there is something here which is done consistently, that there is a persistent quality in some Yoruba carvings which Europeans respond to, and insofar as Europeans can lay claim to finding value in some consistently quality in Yoruba art, then we are justified in doubting Murray’s assertion that such qualities are not actually intended.

Again, it is entirely possible that any particular work of art could contain blunders which we might misconstrue as showing “sensitiveness and feeling.” But to find beauty of accidental origin in a work of art can be no more likely than to find it in a stone picked up on the beach. In fact, we might expect it to be less likely, inasmuch as the artist accepts or rejects accidents as they occur in the process of creation. In this sense, he takes responsibility for whole finished project, whether he goes about his work with the meticulous care of a Raphael or the cunning recklessness of a Jackson Pollock. In any event, the beauty we find in a Renaissance painting and the beauty we find in a stone washed up on the sand are of very different kinds. The former is an achieved beauty, the latter is not. We cannot praise the ocean for having shown “sensitiveness and feeling” in producing a stone we think beautiful: the ocean is not a conscious agent and could never be said, in any ordinary rational sense, to achieve anything in the way in which an artist does. Yet there are anthropologists who talk of the ways in which primitive peoples produce works of art as though it were best to regard them in the manner in which one would regard the ocean’s depositing debris on the shore: blindly, mindlessly, blunderingly, without artistic purpose or intention. In this regard, it should not surprise us in the least that, in his valuable analysis of the 1967 Wenner-Gren conference on “Primitive Art and Society,” Jones mentions that for some of the participants the tribal artist’s “intention was not merely an irrelevant category; it was almost a red herring.”

An example of this attitude occurs in the remarks made by Leach in the course of the previously mentioned 1957 symposium. Leach tells us that another of the symposium’s participants, Sir Herbert Read,

. . . feels that we can perceive in primitive art objects something which, as it were, is inherently there. Personally, I do not hold that view. I consider all these ethnographic objects, if they do not come from European culture, as far as we are concerned, “found objects.” I recollect an exhibition of Graham Sutherland’s work some years ago in which he showed among his pictures bits of driftwood because they had stimulated him in painting some of his pictures. In the same way these objects of what we call primitive art are exciting to us. They stimulate us as “found” objects, not because we perceive in them aesthetic qualities already there.

Now, I do not deny that it is possible to try to regard objects through the treating its contents as But who on earth would want to? And what purpose might it serve anyway, save eliminating from the study of primitive art such troublesome notions as purpose, intention, and achievement? However, the systematic elimination of these very concepts marks a tendency more or less present in a great many works of anthropology and social science in general (and, indeed, it seems to be behind the whole attempt to distinguish latent from manifest functions).

In a more recent paper Leach continues in a similar vein by trying to demonstrate the appalling and incredible “hypothesis,” as he calls it, “that when we westerners affirm that some completely alien cultural product is a work of art this is a response to the fact that elements in the iconography have touched off animal feelings which have been deeply repressed by taboo” (p. 232). The only examples he provides deal with sexual taboo. Once again, we are to be persuaded that “we westerners” are incapable of a direct response to the aesthetic properties intrinsic to objects of tribal art. But if what excites us in primitive art are merely those things which “touch our repressed animal instincts, why bother at all with Acoma pottery, Tahitian or Benin bronzes, when the local newsstand offers so much more? Leach’s idea has a quaintly Victorian ring: that what really appeals to us most about savage things is their elemental sexiness.

Leach is certainly correct in his more modest claim that we frequently do not understand the full, including social and religious, meaning of many primitive works of art. But simply because obscure social or religious dimensions of a primitive work of art may be foremost in the mind of its creator, it does not follow that the purely formal beauty of the object is inessential to him or is an unintentional by-product of his effort to make an object with social or religious importance, or that it cannot be immediately recognized by us. There is much that we may, through our ignorance, miss in primitive art; surely this does not give us reason to suppose that we are condemned ever to regard primitive art objects merely as symbols for sex organs or pieces of pretty driftwood. There are consistently perceived qualities of aesthetic excellence in schools or genres of primitive art, and such excellences simply cannot be understood as unintentional. Leach wishes to deny Read’s contention that we can perceive in primitive art aesthetic qualities which are “inherently there.” But it is unthinkable that the aesthetic qualities which we recognize in primitive works of art are not inherent in those works in the sense that they are, at the very least, implicitly intentional.

One large issue at stake here is, of course, the question of whether or not primitive art can be said to be wholly different in kind from Western art. Perhaps it is a matter of wanting to maintain a monopoly on an academic territory, but some is a different kind of activity, employing greatly different aesthetic criteria, from art in the European tradition. Thus when, to cite a typical example, Haselberger innocently remarked in the pages of this journal that “a Melanesian carves a paddle and strives to give it beautiful form,” Inverarity (19613369) responded by asking rhetorically, “Beautiful by whose standards-the Melanesian’s, the author’s, those of the Western world, or whose?”, as though to suggest these are obviously widely divergent in the matter. In point of fact, though it is impossible to begin to do justice to the topic in this article, there is much evidence to indicate that cross-cultural differences in aesthetic preference and criteria for excellence have been vastly overrated, both in and out of anthropology.

Again, it is in this context worth noting a recent anthology of articles on anthropology and primitive art (Otten 1971). In her introduction, the editor dismisses those “well-meaning teachers and art critics” who “equate the arts of non-literate peoples with those of late civilization in which writing has permeated the culture as the ordinary medium of communication and information storage.” She goes on to explain why primitive art is so different: “In pre-literate or protoliterate culture, the art symbol becomes fact; that is, it simultaneously represents, defines and manifests its referent. In such cultures art objects and events serve as media for information storage, rather than books” (p. xiv). These remarks do less to establish the distinction between European and tribal art than to show the author’s ignorance of the Western tradition in art. Consider the case of European painting: it is really only in the last hundred-odd years, that is, since the advent of photography, that it has not served as an “ordinary medium of communication and information storage.” As for symbols’ becoming facts, as for their defining and manifesting what they represent, we need look no further than the whole history of Western art from Giotto’s madonnas to soup cans. The symbolic and iconographic aspects Otten deems fundamental only to primitive art are essential to the entire Western tradition of aesthetic creativity, secular no less than religious. In light of her stress on the importance of the symbolic functions of tribal art and its to “reinforce cultural values,” one might have thought she would have included for comparison’s sake an article on these aspects of European art, perhaps by an art or cultural historian such as or Huizinga. Instead she begins the collection with a paper analyzing the graphic behavior of a chimpanzee. I find something repellent in the idea of a book purporting to discuss the aesthetic life of tribal peoples which uses as its point of departure an analysis of the scribblings of an ape.

Consider, by way of contrast, the following comments penned by Durer on the occasion of a visit to Brussels in 1520 during which he viewed some of the booty brought back by the conquerors of Mexico:

I also saw the things brought to the King from the new golden land [Mexico]: a sun all of gold, a whole fathom wide; and a moon all of silver of the same size, and two rooms full of all kinds of weapons and arms, arrows, of real wonder; and strange dress, bed-linen and all manner of other domestic things, that one marvelled more to behold than any ingenious invention. . . . I have never seen in all my life anything that has moved my heart so much . . . and I have wondered at the of men of foreign lands. I cannot express the feeling I had.

Durer’s response consists in an immediate and profound sense of wonder based on a recognition of the marvelous achievements of the New World artists. Why do anthropologists so often appear to resist a principle taken for granted by and indeed accepted by European critics since antiquity? Why do they so often refuse either to see or to talk about the objects of their study — primitive art works, rituals, institutions, and so forth — as having implicit in them the possibility of achievement? Previous examples I have provided have tended to suggest this view. The Yoruba carver. the confused and uncritical Pueblo potter whose creations arise from unconscious and nonrational mental processes, or the nameless primitive whose perceptions of his work are so remote from ours that whatever artistic merit we feel we see is no more “inherently there” in the object than it is in pretty driftwood, not to rain dancer — all of these characterizations contribute to an overall picture of the savage as dull, plodding, or semiconscious.

In general it may be said that part of what criticism involves is grasping the possibility of achievement in an activity and recognizing the extent to which that possibility has been realized. Not everything produced in an activity is an achievement internal to its nature; marble dust on the studio floor and broken piano hammers are, for instance, the external by-products of artistic endeavor. An acquaintance with facts about the quantity and composition of dust produced or number of hammers broken will never in itself bring one to an understanding of the art of Donatello or understanding their art means understanding the profound aesthetic achievement of their statues or sonatas. This again is not simply a matter in criticism of determining what the explicit intentions or purposes of the artist are and assessing his creation accordingly. The critic does not ask the artist “What do you intend to achieve in your work?” and treat the answer as setting outer limits for what can be deemed as having been intended, and thus achieved, in the creation of the work of art. The point is that works of art can reveal to us their purposiveness: can show subtleties of nuance and relationship which cannot but be understood as resulting from the working of human intelligence. Such purposiveness, moreover, can be recognized without resort to (or even in spite of) the artist’s expressed purposes. Durer could see and understand Mexican art as revealing the “subtle ingenuity of men in foreign lands” despite his ignorance of their religion or society and despite the fact that he had no knowledge of the artists’ explicit present discussion, the creation of works of art-primitive or European-is not different in principle from understanding other human activities. Human activity must be specified here, because it is human activity-as opposed, for example, to the activity of ocean waves-which involves the concepts of purpose, purposiveness, intention, and achievement. Yet here is what is so curious: it is not as though the anthropologists do not enjoy the art objects — the pots and carvings — in question. They do; but they seem reluctant to see and speak they admire in them as having been achieved. To recognize an achievement is necessarily to that extent to respect the human intelligence which has achieved: this is a far cry from the usual attitude toward a piece of driftwood.

A way of dealing with Merton’s application of his notion of latent function to the Hopi dances, then, would be to ask, is it plausible to speak of the rain dances as being means by which the Hopi people achieve social unity? Can we speak of the dances in this respect as showing a sense of purpose or purposiveness which can ultimately be credited to the intelligence of the Hopi people? Or is the social unity produced by the dances merely a byproduct of a passionate desire on the part of the Hopi to effect rainfall? Here I believe the evidence speaks quite clearly against Merton. As I have noted, analysis in terms of latent function would proceed from the fact that the Indians themselves may deny that the dances have anything beyond a meteorological use to the claim that the promotion of social cohesion is “neither intended nor recognized” by them. Of course, the denial-if it be honest-shows that, in some sense at least, the social unity brought about must be unrecognized, but what of the claim that it is unintended? A perusal of the literature on the Hopi reveals an intense concern on their part with form and style of performance of the ceremonials. When the dancers have finished their performance, their fellows do not wait for it to rain before praising or blaming them for having brought off the ceremony or not: criticism of the most discriminating sort is already going on during the performance. Moreover, it has been repeatedly noticed that one of the chief marks of a “good” citizen among the Hopi is participation in tribal affairs. Individuals will be praised by other members of the tribe for “having a good heart” and “always taking part in the dances.”

It can be further noted — and here is a fact never once alluded to by the many methodologists who have made use of the Hopi rain dance as an example of a latent function — that the Hopi do have purely social dances on their calendar. Now surely, the social dances promote social unity as well as the rain dances. So indeed does the common language of the Hopi, and so does their shared religion. The rain dances can be regarded as one of the many traditions and institutions which have helped to keep the tribe together in the face of white influence. Again, this tendency to promote the living of forms of social life which keep the group together is clearly a persistent strategy which pervades the whole of Hopi culture; it is not a byproduct of one of their institutions. Seen in the context of the whole of Hopi culture, the strategy of always doing that which tends to keep the tribe strong and united is implicitly intentional; it clearly reveals purposiveness.

If Merton’s suggestion were correct, we would have to suppose that in years when rain was sufficient, people would not be subject to blame for not engaging in the dances, but this is hardly the case. Moreover, tribe members are criticized for dancing carelessly or inexpertly, and if it later rains in spite of this, that is no excuse. Granted that the explicit intention of the Hopi in the rain dance is to produce rain, the fact that the rain dances, in addition to many other forms of Hopi life, tend toward the promotion of strength and unity among the Hopi peoples cannot be plausibly regarded as simply unintentional. It is in fact part of the great achievement of the Hopi that they have withstood better than any other North American tribe the onslaught of European civilization. That achievement could not have come about except through the workings of human intelligence, and the purposiveness directed toward the end of keeping the Hopi nation a united whole reveals that intelligence.

Merton’s distinction between manifest and latent functions deals, of course, with more than the element of intention. Merton is also interested in indicating the extent to which there may be aspects of what we do which go unrecognized by us. Here there is surely a use for something like his distinction, though it can hardly be claimed that the sort of inquiry it tends to engender is from this standpoint a novel or unique intellectual contribution of the social sciences. Far from it: the notion that there may be more significance in what he does than an ordinary individual or an artist recognizes for himself is at least as old as Plato. In general I think it can be said that frequently there is more purposiveness in our activities than we fully recognize, just as there is often more in a work of art than the artist himself completely sees. But Merton’s function distinction goes beyond the limited and useful aim of articulating the larger or more subtle patterns of meaning of what we do which go past us in our unreflective ordinary lives. Merton conflates the unrecognized purposes in what we do with the merely inadvertent byproducts of what we do, inasmuch as it is all to be understood as simply “unintentional.” Yet every example he gives is open to the sort of criticism I have applied to his analysis of the Hopi dances.

Ultimately we are confronted here with two very different ways of talking about the same phenomena. The object of Merton’s way of talking would seem to go beyond even the conflation of important social meanings of social activities with their inadvertent byproducts His version of functionalism would appear to obviate the very need to incorporate the categories “inadvertent” and “intentional” into the vocabulary of social science at all. Thus the social scientist is understood to be centrally concerned to discover the functions of social activities and only incidentally interested to know whether the functions are manifest or latent. Whereas I have argued in favor of speaking about the meanings of the Hopi dances as matters of intelligent human activity, the form of discourse implied by Merton’s methodology would seem to encourage the investigator to regard these meanings as involving some sort of mindless physical process. This, however, is not precisely the case: what is really wanted is to discover the operation of some sort of mindless social process. And this finally takes us to the root of the problem, the source of so much of the confusion, nonsense, and vacuity of contemporary social science: the desire to regard the social activities and achievements of human beings in the way that a chemist or biologist regards a physical or organic process.

Although this discussion has been limited to one influential concept in the methodology of social science, many other comparisons of the same sort could be carried out between the kinds of description and analysis utilized in the social sciences and those used in aesthetic criticism. There is no reason to believe that it is possible to come to a better understanding of human social life by borrowing forms of discourse and attitudes from the natural sciences than by borrowing them from art or literary criticism. And though, as so many critics of the func-tionalist school in anthropology and sociology have pointed out, the concern over function rather than intention may not have any undisputed counterpart in the natural sciences (though it is taken from biology), it is certain that the original motive in discounting people’s intentions in what they do and in stressing the “functions” of social activities was to make the social studies more scientific. The reason is very simple: no area of investigation has ever developed into a science until it was able to isolate a limited number of theoretical concepts and apply them to its subject matter in order to make sense of it in a systematic way. Chemistry has its concepts of atomic structure and periodicity, biology its concepts of phylogeny, mutation, selection, and so forth. Until these disciplines developed these theoretical notions they could not make systematic sense of their subject matters in the way that we demand of a true science. 4

Thus if we take human social life as the field of inquiry for the social studies, we immediately see that there will be serious obstacles facing the attempt to isolate a few key concepts which will give us the ability to make sense of social life in any systematic way. The difficulty is that people have many ways of doing what they do, and literally countless reasons for acting as they do. Obviously, if the social studies are ever to develop into sciences, we shall have to disregard people’s expressed intentions in their activities and come to a few universal causes (or perhaps a single cause) underlying the great diversity of overt intention. And so it has been in the short and confused

history of the social sciences: one social “scientist” tells us that all human activity is reducible to sexual impulse, another to the quest for power, another to the desire for self-preservation, or increased pleasure, and so forth. Whenever we encounter methodologists who wish to transform the social studies into sciences, we find discomfort with-even refusal to the fact of infinite diversity of human motivation and purpose. Rather than being seen as something to be explored and enjoyed, this diversity is seen as a threat to systemization, as something which must be cut through or pushed aside in favor of a comprehensive “anthropological theory.” But this theory will of necessity be expressed in terms of underlying latent functions of social activities. Here lies the fundamental appeal of the notion of a latent function: that it gives the investigator a warrant to disregard the great diversity of human purpose in order to build a scientific theory. Hence, though it does not in itself stand as a general sociological or anthropological theory of action, acceptance of the function distinction is a necessary condition for the formulation of such a theory.

In contrast to a widespread attitude in the social sciences, criticism delights in the limitless variety of human action; it even demands it. The activities of artists, though they happily include some of the most sublime human achievements, are not logically different from ordinary human activity. In his effort to understand the nature of aesthetic activities and artifacts, the critic has available to him established vocabularies and forms of discourse. The foregoing discussion has examined one aspect of these forms of discourse in relation to an issue in social science methodology, but there are many other areas where this approach would bear fruit. If the methodologists of anthropology were thus to pay as careful attention to the language and structure of criticism as they do to the logical model offered by natural science, the result would be a far better understanding of their field of inquiry.5


1. In an extremely sensitive, informed, and elegantly written article on meaning in African art, Fagg has made what I take to be essentially the same point: “Philistines (of whom there are a number in anthropology and the other social sciences) often naively argue, bowing to the prohibition on value judgments, that it is useless to try to apply artistic criteria to the arts of the tribal peoples because these peoples simply do not have the concept of art. In fact, of course, what they do not have is the literary concept of art which has dominated western art in the last two centuries and is associated with the rise of aesthetics (which etymologically ought to be about feeling) — a concept which is at war with the very nature of visual art. The philistine case is, paradoxically, destroyed precisely by the fact that the tribal peoples express their concept of art not in language but in art itself (a super-language). Even the Greeks, for all their preoccupation with verbal analysis, seem to have apprehended this, having no separate or art in our sense but making do with the word or skill of all kinds (techne)."

2. If there has been so little empirical research into this whole question, it may be because those investigators most qualified to settle it are among those least likely to bother asking it in the first place. Child and Siroto have produced a very useful little study (1965); see also Vandenhoute (1948) which is glossed by Gerbrands (1971). Attention should be paid to Firth’s excellent essay (1963, chap. 5) and Fagg's two superlative addresses. There is much of relevance to be found in Carpenter (1969) and Redfiled (1971). Despite my criticisms voiced here and elsewhere, Bunzel's monograph on Pueblo pottery art is still one of the best ethnographies on tribal art in existence and stands as a sort of field confirmation of theses promulgated years ago by such formalist aestheticians as Bell (1914).

3. A remark made almost 30 years ago by Collier is apposite: “It a somewhat incredible fashion among anthropologists to remark that the whole of this unified multitude of Pueblo sacred drama is nothing more than an operation to make the corn grow, or to bring an emotion of security to the afraid and insecure. Account for Chartres Cathedral thus, for the Bhagavad-Gita, for Michelangelo, Plato, Aeschylus, for the Christ, and for music . . . , and let one say that all is Freudian projection and mere imitative magic, if he thinks his hypothesis requires this bankruptcy of perception of him.” Alas, unlike zoot suits and bobby socks, this is a fashion still with us.

4. Jarvie (1973: 12) tells us, for example, that Radcliffe-Brown thought of his brand of functionalism as “a scientific approach to the study of society; therefore he rejected as unenlightening all historical, psychological, environmental, evolutionary, or biological explanations. He pressed this point further. What such a scientific approach disclosed was that there were underlying patterns or principles in each social organization. Such unchanging patterns of relationships he came to call the social structure. It was his hypothesis that if the social structures of societies were compared it would be possible to classify them into a small number of groups and discover the general principles governing their operation. These would be the laws of a scientific anthropology.”

5. I would like to acknowledge the helpful comments of Daniel Moerman, Martha Morris, Ronald Stockton, and Frank Wayman on earlier versions of this article.