To Understand It On Its Own Terms

Philosophy and Phenomenological Research 35 (1974): 246-56

Denis Dutton



We commonly hear it said that a work of art must be understood “on its own terms,” and that phrase is used in other contexts as well; people, especially people very different from ourselves, are said to have to be understood on their own terms. But what is the meaning of the expression “on its/their own terms?” Note that we do not say of every possible object of understanding that it must be understood on its own terms. The statement, “Chemistry must be understood on its own terms,” sounds odd. The “on its own terms” appears to lack point. It invites the retort, “Good grief, on what other terms had you expected to understand it?!” We seem to be brought to this: where the exhortation “understand it on its own terms” is to be used in reference to some object, there must be some reasonable way of understanding that object alternative to understanding it on its own terms. We may ask the child its own childish reasons why it keeps clobbering little sister (“She took my dolly!”), or we may account for its actions in terms the child does not comprehend (“sibling rivalry”). We may on occasion even do the same sort of thing with adults, perhaps giving a reason out of Freud for something done by a person who never heard of Freud or psychoanalysis. But naturally, distinctions between “their terms” and “our terms” become of extreme importance when we attempt to understand the peoples of other cultures. In such cases there literally are different systems of terms, that is, different languages, within which we must make sense of activities.

It is often the mark of an object to which “understand it on its own terms” might apply that it is in one way or another an expressive object, hence the familiar use of the phrase in connection with works of art. Now it would be tempting to argue that since whatever is said is said in a language, the import of “understand it on its own terms” is then that something expressed by, or in, an object is not only understood, but is understood in the language in which it is expressed. But there are problems with talk about language in this context. It seems clear that understanding a small child on its own terms means understanding what its acts and words mean in its own childish language. (Incidentally, it is worth noting here that questions of our understanding a newborn infant on its own terms do not really arise: there are no terms, that is to say, there is no language, within which the baby makes sense of itself or its world.) Similarly, an anthropologist who wishes to understand a primitive tribe on its own terms would presumably be faced with first learning the terms, i.e., the language, of that tribe. And though much the same sorts of considerations can arise in aesthetic matters, there are pitfalls where one tries to apply talk about language to art. One often hears it loosely said that Palestrina or Titian speak particular artistic languages different, it is supposed, from the languages of Schumann or Picasso. What is generally meant by this is that artists (usually) work within the limits inherent in the prevailing styles of their epochs. This is of course true: what it makes sense for Palestrina to do in a composition is not what it makes sense for Schumann to do; Schumann’s audience found meaning in some music which Palestrina’s audience would have found nonsensical. Thus when we say, for example, that Rachmaninoff ignored musical developments of the twentieth century and chose to write in the “language” of Tschaikovsky and Borodin, we mean that his compositional style was that of his artistic forebears.

However, I would prefer to avoid such a loose use of “language.” Talk of language more properly implies not only a community of language users, but also a system of public, regular meanings. For instance, late medieval art must be understood in terms of its own highly formal system of symbolism. Here we are really justified in speaking of a “language of art” because there are defined rules for the use of images in painting: the diamond represents the Virgin; the red rose, the blood of the martyrs; the black dog, envy, and so forth. Wagner’s use of leitmotivs would be another example, where the appearance of a particular motive in the score is intended to represent musically a character or idea in the drama. Such clear cases of the employment of an aesthetic language are, nevertheless, not common in the history of art. More usually, talk about language in art is talk about style in art. Where we are advised, on the other hand, to understand a work of art on its own terms, what is most frequently meant is not that we learn some artistic language. Ordinarily, “understand it on its own terms” has to do with grasping the point of an art object or art form.


For an illustration of what this last remark means let us look at the changing critical attitude toward piano transcriptions. In the nineteenth century it was common practice for virtuoso pianists such as Liszt to present on recital programs transcriptions of works originally composed for other media: Beethoven symphonies, Schubert songs, fantasies on themes from operas and so on. Until recently many modern critics have written off such transcriptions as pale reflections of the originals at best, tawdry junk at worst. Indeed, one will likely be forced to such a conclusion so long as one insists on seeing such works as keyboard imitations of their prototypes. As more and more musicians are beginning to discover, however, transcriptions are not always merely imitations of the original compositions, but are — especially in the case of Liszt — often themselves imaginative works of the highest order. Of course, if one misconstrues Liszt’s intentions in his Don Juan Fantasy or his pianistic renderings of Schubert Lieder as attempts to imitate Don Giovanni or Die Forelle, then one will probably conclude that these works are nothing but kitsch. But the point of transcription is not imitation: it is recreation of the prototype within the possibilities and limitations afforded by the piano. Thus it is not surprising that Liszt’s transcriptions occasionally violate the letter of the prototype in the effort to express its spirit. Understanding Lisztian piano transcription on its own terms then means recognizing what the point of the practice is and judging any example accordingly. Not that one need accept even the point of the practice as valid or worthwhile: the general practice may be as open to criticism as any instance of it. What must be resisted is criticism which ignores or mistakes the point of an artistic practice.

But, as in the case of piano transcription, what the point of a practice, activity, custom, or institution may be is sometimes not obvious. An interesting example is provided by looking at the work of an anthropologist, Robert Redfield’s The Primitive World and Its Transformations. In a discussion of what he describes as the “traditional and uncritical” character of primitive thought, Redfield says that tribal people often manage to see meaning in even the most rote activity. As evidence for this claim he cites the remarks of another anthropologist who has authored a work on Pueblo pottery art.

Ruth Bunzel, studying Pueblo potters, found that the Indian woman who was in fact copying the designs of other potters with only the smallest variation was unaware that she copied, condemned copying as wrong, and had a strong conviction that she was in fact inventive and creative.1

Redfield has the gist of what Bunzel reports, though he is mistaken if he means to imply that it was just one woman who apparently suffered from such delusions. In point of fact, all the Pueblo woman potters told Bunzel copying was wrong and most were supposedly laboring under the mistaken “conviction” that they were “inventive and creative.” The question I want now to ask is this: what warrants the anthropologist in insisting that those Pueblo designs are copies of one another? It may well be that the seemingly insignificant “small variations” Bunzel noticed are anything but insignificant to the potters themselves. Bunzel describes the disclaimers of the Indians as a “very simple and rather amusing” failure to align ideals of creativity with practices of plagiarism and she blames the situation in part on “sterility of imagination.”2

But consider the following: imagine taking a Hopi Indian to Carnegie Hall on successive nights to hear Vladimir Ashkenazy and then Artur Rubinstein perform the same Schubert sonata. If he claimed that Rubinstein was suffering from “sterility of imagination” because of the fact that he had merely “copied” Ashkenazy, we would naturally attribute this to the Hopi’s ignorance of European music: he mistakes the pianist’s purpose, which is not to create a new piece of music but to interpret an old one. One shows that one misses the point if one tries to criticize piano performance according to criteria applicable to composition. Now as it happens, a similar interpretation is compatible with the data which Bunzel reports in her monograph on Hopi pottery art. It may well be that painting a pot is for a Hopi potter more like the performance of a score — the execution of an established pattern or motif — than it is like the invention of a new composition. Such an interpretation offers an important advantage over the account provided by Bunzel and Redfield: it obviates the need either to characterize the Indians as being unimaginative plagiarists or to suppose they are afflicted by some sort of mass delusion.

There might be some who would allow that I am right in my analysis of the preceding example but who would yet insist that such misinterpretation as that of the anthropologists could be recognized and eventually rectified by simply amassing more data. To this I would reply that quantity of data is not always the issue. If one thinks that the point of decorating a pot must be to invent a new design, attention to details may just serve to confirm what plagiarists the Hopi potters “really” are. Similarly, the history of aesthetic criticism is full of stories such as that generations of sensitive and knowledgeable men wrote off the late quartets of Beethoven as the ravings of a deaf and demented man just as the art of piano transcription has been dismissed in our own time. One can know all the notes (or all the pot designs) and go on missing the point indefinitely. What then might have saved Bunzel and Redfield from having made this mistake? A central issue raised by the example involves Redfield’s having called the Pueblo potters “uncritical” in their attitude toward their work. In fact, he uses the example precisely to illustrate the generally “uncritical” nature of primitive thought and it is not altogether surprising that he finds Bunzel’s monograph ideal to support his claim. This in turn takes us to an issue which is part of what lies at the bottom of the problems with Bunzel’s work: the question of criticism.

Any activity, if it is to be recognizable, isolable, or definable, is something which is carried out according to rules, norms, or standards, carried out within certain limits. A performance of an activity is liable to be judged as having been done rightly or wrongly, well or poorly. Values are important here, and opinions come into play in the criticism of performances. But it is not that in the criticism of performances of activities we only express our opinions about the qualities of performance. Frequently we are interested in whether a given performance is to count as a performance of a particular activity at all. For instance, in the present case we are in part concerned with whether the Pueblo potters can in general be described as copying one another. In this case merely observing some movements of an Indian with a paint brush may not be enough: to understand what is being done we are going to have to know more. What is significant here, I believe, is this: any person who undertakes to perform an activity in so doing lays himself open to the possibility of criticism of a certain sort. The nature of the activity — that is, what the activity is — will determine what sort of criticism one is liable for in the performance of it. And from the other direction, the kind of criticism relevant to an activity can be used to tell us what that activity is.

Here is an illustration of what I mean. We might imagine an anthropologist — or a philosopher — who, when confronted with the phenomenon of two pianists playing the same piece of music from memory in a recital, would claim that these two performances must count as instances of the same activity however differently we might evaluate each. “So long as the two pianists are playing the same notes from memory,” we might hear him say, “then we are faced with what is essentially the same activity, regardless of how we may judge the value of the different performances.” I wish to deny this in the following way. Suppose we go in the afternoon to a recital given in the parlor of Miss Snodgrass’s home by her pupils. During the recital we hear a ten-year old girl perform an easy Scarlatti sonata. Upon leaving, you remark on how rhythmically distorted the little girl’s performance was, to which I reply, “Well, I admit it wasn’t very good, but at least give her credit for having played all the notes.” That evening we go to a concert hall where we each pay scalper’s prices to hear Vladimir Horowitz play a program which happens to include the same sonata. After the concert you complain that Horowitz’s performance of the Scarlatti was altogether too percussive, to which I respond, “Well, I admit it wasn’t very good, but at least give him credit for having played all the notes.” This contrast is devised to bring out the issue of what is to count as relevant criticism. Applied to the performance of a professional virtuoso the remark is a joke; and it is a joke exactly because it does not apply. In the case of Horowitz it goes without saying that he plays all the notes. And that it “goes without saying” is the whole issue, for not everything that can be truly said of a performance will be relevant as criticism of that performance. If we pay premium prices for seats we expect to hear something more than all the notes, and it is that something more which is really at stake. On the other hand, having played all the notes may be reason for the most glowing praise of a child; in any event, it is highly relevant in the consideration of any performance of a Scarlatti sonata by a ten-year-old. What I wish to indicate is that in this example we are not confronted with different instances of the same kind of activity (the moving of hands and fingers, the making of certain sounds represented by the same score, etc.). Rather, we see from the fact that since in each case what is relevant to say at all in criticism differs, we must therefore be faced with two different kinds of activity. It is not so much what in fact happens in either specific instance which is important (we can imagine Horowitz missing notes and the little girl playing them all), but what ordinarily counts as important in what happens.

Let us now return to the Pueblo potter. Bunzel observes two Hopi women painting pots. She notices the designs being painted are “the same with only the smallest variation.” But already the interpretive project is going awry. To understand how, you must imagine the anthropologist saying in a tone which suggests a significant discovery, “Why, those designs are almost the same!” To this, the correct response is, “Of course, they’re the same, but that’s not the point,” just as your response to me outside the concert hall should have been, “Of course, he played all the notes, but that’s not the point.” Again, it is not false to say that the designs are almost identical; it is pluperfectly clear that they are. The problem lies in imagining that it is important to say so. To go back to an earlier example, it is as though that Hopi Indian left Carnegie Hall exclaiming in amazement, “Why, he just played the same notes as the pianist we saw last night!” Such a remark, though true, hardly serves to advance an understanding of the essential nature of the activity in question. Moreover, if we think such a fact to be significant we may be woefully misled. Analogously, if we think it relevant or important in the present context to remark that the Indian women are painting the same design and leave it at that, it naturally follows that some or all of them are copying and furthermore, given their loud and insistent claims to originality, it follows that they are, as Bunzel supposes, self-deluded. But if my analysis is right, Bunzel’s assertion that the two designs are the same, while true, is literally beside the point.

How could an anthropologist avoid such difficulties? The answer is by carefully attending to the forms of criticism employed in relation to the activity in question which are used by the people who perform it. As was noted before, Redfield uses Bunzel’s report as a basis for his claim that primitive peoples are uncritical. But it is unimaginable that a school of art which has produced such astonishingly fine pieces as the Pueblo potters does not have within it, internal to it, rigorous standards of criticism. The potters themselves may not be able to explain these standards well, but that does not entail that the standards do not exist. Something must count as a good design, something as a poor one, to the artisans, and it is crucial that an investigator know what it is. Missing the point comes about because of a failure here to attend to that which would have indicated what the activities in question really are, i.e., the notions which count as relevant in the criticism of performances of these activities. Until this has been accomplished, the anthropologist does not yet know what it is he is investigating.


When we speak of “the point of an activity” we mean that which is isolated as the most important, the most valuable, the most significant aspect of the activity; we mean that which constitutes the main or primary meaning of the activity; we mean that which indicates the essential nature of the activity. Now quite obviously the point of an activity is something which may prove hard to come by; there are all sorts of ambiguities which arise when we attempt to determine what is important in what we or others do. Here is another example, taken again from the work of Ruth Bunzel, this time discussing the Hopis’ close Pueblo neighbors, the Zuni. She tells us that for all their emphasis on the carrying out of religious rituals these Indians lack an intense religious feeling. Zuni prayer, according to Bunzel, “is not a spontaneous outpouring of the heart. It is rather the repetition of a fixed formula.” Her remarks apparently presuppose an opposition between something’s being an outpouring of the heart and its being the repetition of a fixed formula.3 Interestingly, she is disputed on this very question by Li An-Che, a Chinese anthropologist who lived with and studied the Zuni in 1935. “Why,” asks Li, “should a ‘spontaneous outpouring of the heart’ be antithetical to ‘the repetition of a fixed formula’?”4 The argument, as things develop, is over what we are to take as the essential nature of the activity in question. The argument is over the point of Zuni prayer. On one side, Bunzel claims that Zuni prayer amounts in essence to the accurate recitation of formulae. On the other side, Li Finds not only that Zuni worship is characterized by deep religious intensity, but also maintains that the evocation of such passion is the central aim of Zuni prayer.

Now of course it is true that we live in a culture which embraces much rote, meaningless activity, from the church pew to the assembly line. In our culture the idea of repeating a fixed formula seems naturally to suggest an activity devoid of intense involvement or expressiveness. However, anyone familiar with the spirit of Confucian thought will not be surprised that a Chinese anthropologist might want to challenge the assumption that frequent repetition and passionate intensity are necessarily incompatible in the exercise of religious ritual. Li claims to have found that the Indians have what is to us an almost uncanny ability to breathe life and meaning into the “same” prayer each time they “recite” it. I use quotation marks here because if Li is right it follows that in an important sense Zuni religious practices are not a matter of “reciting the same prayer.” Another comparison with music will be helpful in understanding this. A musical score can be thought of as something like a fixed formula and a musician’s job as like the repetition of a fixed formula. Yet in music this repetition is never supposed to rule out a performance’s being a “spontaneous outpouring of the heart.” On the contrary, such a quality is precisely what is sought after in performance. If a pianist is having an off night, is not in the mood, or whatever, we say his rendition degenerates into mechanical repetition of the notes. But indeed, a truly great artist, a Schnabel or a Rachmaninoff, is capable of infusing each performance of a piece with such freshness and spontaneity that it is even odd to speak of him as “repeating a formula.” In such cases the performance is not properly understood as the mere recitation of notes. Hence the phrase “piano recital” while it may well describe an afternoon of piano playing by Miss Snodgrass’ pupils, has always struck me as strangely inappropriate when applied to a performance by, say, Glenn Gould. An artist does not recite, he interprets; some would say he recreates. It is, moreover, the mark of a discriminating listener that he is able to hear and understand the difference between the recitation of a hack and the interpretation of an artist. This in turn connects with Li’s ultimate comment on Bunzel: he politely suggests that she does not know enough of Zuni life to be able to discern the intensity with which these Indians carry out their sacred rites. Were she studying a Christian sect, he says, she would be “more apt to make a refined judgement.”5

Anyone will naturally show more discrimination when judging practices which are a familiar part of his own culture. But I would also want to resist the temptation to suppose here that this whole issue revolves around a particular anthropologist’s ability, or lack of it, to catch subtle nuances in the performance of certain activities being observed. Bunzel and Li, after all, are not disagreeing about the intensity of a single instance of Zuni prayer but about the point of the whole enterprise, and I would want to stress that this question is certainly not necessarily a matter of mere opinion. There is a way — in principle at least — of settling the issue and that involves invoking the procedures discussed above in connection with the Pueblo potters. If we want to decide whether the point of Zuni prayer is generally to execute a formula or to express one’s deepest religious feeling, we ought to ask a Zuni what constitutes a good prayer. Perhaps he will not cooperate with an answer, in which case we will be forced to infer from other facts what for the Zuni is a good performance of a prayer, what a poor one. At any rate, until we understand the kinds of criticism which the Zuni apply to the exercise of religious rituals, we do not yet know what those activities really are. For instance, if all our inquiries and surmises lead us to the conclusion that a well performed prayer is one in which all the correct words are spoken in the right order, that is, if the Zuni ordinarily praise and blame individuals for uttering all the sacred words in the right order, and for nothing else, then we would have strong evidence for believing Bunzel to be on target in her description of the activity.

If, however, the Zuni never mention saying all the words in the proper order as grounds for praising an individual, but rather say things like, “Palowahtiwa goes through the prayers very well. He concentrates on the words and never lets bad thoughts enter his heart,” then we would have grounds for supposing that Li’s analysis is the true one. (My own knowledge of the literature on the Pueblo Indians leads me to suspect strongly that the Zuni would indeed no more praise an adult worshipper for merely saying all the sacred words in the correct order than we would praise a professional piano virtuoso for playing all the notes of a Chopin mazurka in the right order. A similar parallel could, I believe, be drawn between novice Zuni worshippers and novice pianists, where such questions as mere accuracy might be more relevant.)6 We can, of course, easily invent further possible complications which would make it difficult to decide the issue, but I never wished to claim that anthropology must be a simple matter. I only wished to show what direction such inquiry ought to take; that it should try to determine in the first instance what the essential point of each of the activities it observes is and that this can be accomplished by attending carefully to the standards of criticism commonly applied to performances of the activity by members of the tribe studied.

To conclude: I began by trying to give some indication of what is meant by the plea to understand an activity on its own terms. This, I argued, usually means that we must understand the activity according to its point. Furthermore, the point, the essential meaning, of an activity is determined by reference to the kinds of criticism ordinarily applied to the activity. This last contention reflects then upon the first. It follows that to understand an activity on its own terms is to understand it according to those terms of criticism which are its own. Such understanding involves recognizing what is relevant or appropriate to speak of in praising or blaming a performer of the activity. The relevant sorts of criticism tell us what the activity is, and understanding it on its own terms thus means treating it as what it is.


1. Robert Redfield, The Primitive World and Its Transformations (Ithaca: Cornell University Press, 1953). p. 15. The Hopi and Zuni are the two major tribes of Pueblo Indians of the Southwest.

2. Ruth L. Bunzel, The Pueblo Potter (New York: AMS Press, 1969), pp. 51-54. This is a reprint of Bunzel’s doctoral dissertation, Columbia University, 1929. Despite serious conceptual confusion and inconsistency, not only with regard to matters discussed here but also in connection with the relation between freedom of artistic expression and limitations of prevailing style, the monograph remains the best available work on Pueblo pottery art.

3. Ruth L. Bunzel, introduction to Zuni Ceremonialsm, Forty-seventh Annual Report, Bureau of American Ethnology, (Washington: Smithsonian Institution, 1932), p. 493. This monograph was used by Ruth Benedict as the basis for her famous description of the Zuni in Patterns of Culture (Boston: Houghton Mifflin, 1934).

4. Li An-Che, “Zuni: Some Observations and Queries,” American Anthropologist 39 (1937), p. 64.

5. Ibid.

6. See also John W. Bennett, “The Interpretation of Pueblo Culture: a Question of Values,” Southwestern Journal of Anthropology 2 (1946), pp. 361-74.