Some Notes on Languages of Art

Citation preview

Some Notes on Languages of Art Nelson Goodman The Journal of Philosophy, Vol. 67, No. 16. (Aug. 20, 1970), pp. 563-573. Stable URL: http://links.jstor.org/sici?sici=0022-362X%2819700820%2967%3A16%3C563%3ASNOLOA%3E2.0.CO%3B2-X The Journal of Philosophy is currently published by Journal of Philosophy, Inc..

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/jphil.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org Fri May 18 08:42:23 2007

SOME NOTES ON LANGUAGES O F ART

563

usually saved from such peril by the conventions (such as stages, frames, pedestals, and versification) that operate to inform us that we are confronted by art, not nature, so that we are never really deluded by the fictive construction and thus receive no false knowledge from it. T h e connection between these last remarks and the earlier parts of my paper are, I trust, clear: the distinction that Goodman blurs in identifying literary art with written inscriptions is, in effect, the distinction between fiction and history, and between art and nature. T h e ordinary verbal actions-speech or writing--of other men are a part of nature, as much a part of the natural environment as the behavior of sun, stars, rocks, and trees; and those verbal actions are as much a part of history as all the other actions of our fellow creatures. Furthermore, our lives do depend upon our searching, testing, making delicate discriminations, and discerning subtle relationships in the utterances of these fellow creatures. T h e cognitive explorations thus directed, however, are not always rewarded with discoveries and revelations; for these utterances are often obscure and fragmentary, or predictable and monotonous. If we wish to assure ourselves of a more satisfying cognitive experience with language, we will turn to poems and novels-knowing, of course, that the speech of men in nature and history is distinct from the language of art. BARBARA HERRNSTEIN SMITH

Bennington College SOME NOTES ON LANGUAGES O F ART

I

AM grateful to the three major participants in this symposium * for their time and attention, their compliments, and their perceptive criticism. On the whole, there is more agreement than disagreement among us; and perhaps the following remarks will help to resolve some points of difference. There is an obvious sense of nature which any aspiring monist, myself included, would have to grant embraced cultural artifacts, including artworks themselves, along with all other material objects and physical events. My intention was merely to indicate that I am not using nature in that sense here, but in the other familiar sense that contrasts nature with artifacts, the natural object or event with the artificial one that may be taken to represent it. T h a t contrast, it seems to me, or the relation between its two elements, is something that in other terms occupies Goodman throughout Languages of Art. *Symposium on Languages of Art held at the Sixty-sixth Meeting of the American Philosophical Association, Eastern Division, on December 29, 1969, in New York City. See this JOURNAL, this issue, pp. 531-563.

564

THE JOURNAL O F PHILOSOPHY

I. REPLY TO RICHARD WOLLHEIM Mr. Wollheim quite correctly suggests that by describing a predicate like "is a unicorn-picture'' as an unbreakable one-place predicate, I meant to preclude treating it as a two-place predicate and thus to block the fallacious inference from " p is a unicorn-picture" to "there is something that p is a picture of." I did not mean to restrict our freedom to treat such predicates as conjunctive; we may, for example, construe " p is a unicorn-picture" as elliptical for " p is a picture, and p is of-a-unicorn." Under that construction, we may logically infer that there is something that is a picture, and also that there is something that is of-a-unicorn, but not that there is anything other than the p we started with. We may likewise, as Wollheim has pointed out to me, infer from " p is a Pickwick-asclown picture" that p is a clown-picture. Since such predicate-splitting is not prohibited, the epithet "unbreakablev-intended merely to reinforce "one-placev-is misleading and should be omitted. On the other hand, I do not see the reason for Wollheim's statement: "In order to classify chairs as Louis XIV chairs or Louis XV chairs, we do not have to recognize them as chairs. But in order to classify Louis XIV pictures or Louis XV pictures, we must first recognize them as pictures." It seems to me that a person might easily learn to classify objects in a room as Louis XIV pictures, Louis XV pictures, chairs, and tables without having learned to classify things as pictures or as articles of furniture. It also seems to me that a person might easily learn to classify things as pictures or as articles of furniture without learning how to sort these classes into any subclasses. Sometimes we learn first how to apply general labels and only later how to apply more specific ones; sometimes we learn how to apply specific labels and only later how to apply more general ones; and this I think is as true for pictures as for furniture or anything else. I agree with Wollheim that much of what is said in the book about representation-as and fictive representation can be applied to perception. For example, the analysis of representation-as may well help in clarifying the notion of seeing-as; and rather than say that someone has been seeing ghosts, we might rather say that he has been ghost-seeing. On the other hand, I cannot follow Wollheim's discussion of a picture of a sheaf-of-corn that represents Christ. He says that since we cannot see Christ in the picture, the picture does not represent Christ as a sheaf of corn. While what we 'can see in' a picture seems to me far from clear, I should suppose that if we cannot see Christ in a picture of a sheaf of corn, neither can we see

SOME NOTES ON LAXGUAGES OF ART

565

him in a picture of a lamb; yet surely we can say that Christ is often represented as a lamb. I n his discussion of the relationship between representing a soandso and being a soandso-picture, Wollheim correctly reports me as sajing, in the early pages (pp. 28, 29) of my book, that if p is a soandso picture and p denotes a suchandsuch, then p represents a suchandsuch as a soandso. At that stage, I was tentatively taking representation to be denotation by a picture. Later (especially pp. 41-43 and V,l), I point out that denotation by a picture is not always representation; and this has the retroactive effect of modifying the formula in question to read: if p is a soandso-picture and p rep~ e s e n t sa suchandsuch, then p represents a suchandsuch as a soandso. Similar modifications are implied for related passages. None of them seems to affect Wollheim's discussion appreciably; but the point needs to be understood for the investigation of further questions he raises. T o one of these questions-whether a soandso-picture that represents a suchandsuch is therefore a suchandsuch-as-a-soandso picture-I think the answer is negative. There is a difference between being a lion-picture representing Churchill and being a Churchillas-a-lion-picture, whether or not the latter represents Churchill. Whether a picture represents a suchandsuch may under some circumstances have some bearing upon, but does not determine, whether it is a suchandsuch-picture. Wollheim fears that such an answer will leave denotation as the core of only one sense of representation and would endanger "the elegant explanation of the difference between representation and expression, in terms of the 'difference in direction', i.e., denotation us. converse denotation." In saying that denotation is the core of representation, I meant only that denoting is the core of representing. Converse denoting is the core of being a suchandsuchpicture in that the picture is denoted by "suchandsuch-picture." Denoting is thus involved in both cases, but runs from the picture to the object in the first case and from the predicate to the picture in the second. I n much the same way, converse denotation is the core of expression. Being black or small, furthermore, does not-any more than does representing a black or small thing-guarantee that a picture is a black-thing-picture or a small-thing-picture. Representation of Churchill by a black ink drawing, or by a small painting, does not constitute representing Churchill as black, or as small. Rather, Churchill is represented as black by a black-man-picture (which may

566

THE JOURNAL OF PHILOSOPHY

itself be ochre), or as small by a small-thing-picture (which may itself be large, such as a large baby-picture). Apparently not observing this distinction carefully enough, Wollheim obscures the crucial difference between representing a black horse by a white speck and representing the horse as a white speck. T h e white speck that represents a black horse in a photograph is not necessarily or even normally a white-speck-picture or a black-horse-as-white-speck-picture, and does not necessarily or even normally represent the black horse as a white speck or, for that matter, as a black horse. T h e main thrust of Wollheim's paper, however, seems to be in his question, "How, or by what (kind of) criteria do we apply predicates in classifying pictures into kinds, or (more succinctly) how do we decide what, in one sense of the term, they represent?" His answer is that we do this according to what we can see in the pictures. As I have already indicated, the notion of what we 'can see in' a picture seems to me too unclear to provide any illumination. Mr. Wollheim says further that "what we can see in a picture and [the artist's] intention are linked in that, by and large, an intention to represent x must express itself through the making of a picture in which !we can see x." But of course a picture may be a soandso-picture or represent a suchandsuch contrary to the artist's intentions or, as with pictures produced by an automatic camera or by nature, in the absence of any relevant intention. In the setting up of symbol systems, as in the building of bridges, intentions are indeed usually involved; but in both cases, we can study the results independently of the thoughts of the makers. 11. REPLY TO BEN JAMIN

BORETZ

Mr. Boretz argues for a phenomenalistic rather than a physicalistic conception of the work of art. In general I regard the choice between a phenomenalistic and a physicalistic system as quite free, but I choose to treat works of art, or their instances, as physical rather than phenomenal in order to discourage what seems to me the mistaken and prevalent identification of the aesthetic with the immediate and uninterpreted. Boretz is not guilty of that error. Rather he envisages, apparently, treating works of art and physical objects as both, though different, constructs from phenomena. I doubt if the net results of this direct route of construction would be very different from those of proceeding from phenomena to physical objects and then to works of art. "Sounds," Boretz further contends, "are not parts of music. . . And neither are paint, pigment, or canvas parts of paintings, nor

.

SOME NOTES ON LANGUAGES OF ART

567

masses of bronze parts of sculptures, nor pages and letters parts of poems." I see nothing wrong in saying that a given bronze object is a sculpture. This is entirely compatible with the object's being something else as well, and by no means implies that to be a bronze object is to be a sculpture. A bronze object is a sculpture insofar as it serves as an aesthetic symbol; it may also serve as a nonaesthetic symbol in other contexts, or function in entirely nonsymbolic ways. It is a hammer by virtue of some of its properties, a sculpture by virtue of others. It is a hammer when pounding a nail, a sculpture when serving as a symbol of a certain kind. This will not suit Mr. Boretz. What counts, he insists, is not what a work refers to but what it is in itself. I am afraid he overlooks the fact that, according to my treatment, the properties that anything exemplifies or expresses, far from lying outside it, are properties it possesses. Talk about those properties is talk about what the work is. Why not, then, speak simply of properties possessed rather than of properties exemplified and expressed? Because not all the properties the object possesses, but only those it exemplifies or expresses when functioning as a symbol of a certain kind, are relevant to it as a work of art. T h e capacity for hammering is extraneous to the bronze as a sculpture. Now we may, if we like, take the work of art not to be the concrete object that as an aesthetic symbol exemplifies or expresses certain properties, but to be the abstract complex consisting of the properties it thus symbolizes. T h e sculpture will then not also be a hammer; rather, the sculpture and the property of being a hammer will be universals that happen to be manifested in one piece of bronze. This version seems to me to have some inconveniences and no compelling advantages, but is in principle interchangeable rather than incompatible with taking the work to be the concrete object-or, as in the case of music or poetry, a class of events or objects. An ardent formalist, Boretz objects not so much to my theory of expression as to my discussing expression at all. For him, the actual structure of the work is all that matters; the work is fully determined by specifying that structure in detail, and calling the work sad is romantic poppycock. Feelings a work may express, like things or events it may describe, are beside the point. My purpose, of course, was not to advocate or decry formalism or expressionism in music, or realism or abstractionism in painting, or any other artistic doctrine, but merely to explicate the nature of representation, exemplification, expression, and other modes of symbolization. Nevertheless, I think expression has an importance, even for a

s~~

THE JOURNAL OF PHILOSOPHY

formalist, that Boretz overlooks. Admittedly, saying that a musical work is sad-like saying that it is long or rather loud-is pretty vacuous; but this is because sadness, like length and loudness, is a commonplace, obvious, and general property, so that its ascription to a work provides little interesting information. And while sadness-like length and loudness-is indeed a consequence of the detailed make-up of the work, the converse does not hold. Works differing widely in detail may all have the same property of sadness; and the common structural property, the literal correlate of sadness, is not easily specified. T o describe a work or passage as muscular, electric, spatial, curvilinear, brittle, or floating may be to describe metaphorically some recondite and highly important structural features. T h e notion of the structure of a work is as specious as the notion of the structure of the world. A work, like the world, has as many different structures as there are ways of organizing it, of subsuming it under categorial schemata dependent upon some or other structural affinities with and differences from other works. What seems a definitive structural description may ignore some of the most significant patterns. We may overlook trees for woods, or woods for trees, or groves for either; we may overlook themes for notes; and we may examine every part of a picture without grasping subtle, or even gross, features of composition or design. Understanding a work involves the discovery, the recognition, of unobvious patterns. T h e requisite breaching of barriers established by habit and literal language often occurs through the importation of schemata from a foreign realm. New likenesses and differences, new relationships and patterns, are thus revealed, and are described by the metaphorical application of these alien terms. Theoretically, these metaphors can be supplanted by complex literal descriptions; but even as metaphors, they are in effect descriptions of structural features. Briefly, the feelings a work expresses are properties it has, not because the work literally has feelings, but because the feelingterms applied are metaphorical descriptions of structural (or other) properties the work has and exemplifies. Only at the risk of overlooking important structural features of a work can a formalist ignore what the work expresses. In discussing misical notation, Boretz urges that what is specified in standard scores are not specific pitches but rather certain bands of pitches and that, moreover, the score is intended and used as a design, determining relationships rather than specific pitches or even ranges. Nothing I have said is at all incompatible with taking note marks

SOME NOTES ON LANGUAGES OF ART

569

as prescribing pitch (and duration) within a range rather than uniquely. Precision or uniqueness is not a requirement for notationality. Indeed, I have pointed out that characters in a notation may have very broad compliance-classes, so long as these classes are differentiated and do not intersect. Now suppose, going even further, we take the score as prescribing no pitches whatsoever, even within ranges, but only relationships among pitches; that is, as requiring, for example, that the second note should be one range higher than the first, the third two ranges lower than the second, and so on. A score might then be transcribed into numerals as follows: .,I, -2,3,4, -1 .... The performer can now begin with any note he likes, play any note from the next higher range, and so on. Yet we still have a system meeting all the requirements for a notation. Work-preservation in any chain of correct steps between score and performance is assured. And incidentally, my notorious remark that a single deviation from the score disqualifies a performance as a genuine instance of the work will still hold good, though a single error will now consist not in playing a note different from the one specified, but in playing a note not standing in the specified relationship to what precedes it. Suppose, though, a score is taken rather as a suggestive pattern, prescribing no explicit kind or degree of conformity in a performance. Then, clearly, notationality is lost. Nothing in my theory precludes such an interpretation. I have not been concerned with deciding on the proper interpretation of any vocabulary of symbols but rather with setting forth the conditions under which interpreted systems are notational. Under some interpretations, the standard language of musical scores is notational, and music is allographic; under other interpretations, the language is nonnotational, and music is not allographic. Which interpretation is best is a question I am neither equipped nor required to answer. Nevertheless, since that question is intimately related to the question of the identity of a musical work, and since my saying that the identity of a musical work consists in exact compliance with the score has aroused the ire of many musicians, I should like to point out (in answer rather to them than to Boretz) that the characteristics that make a performance of a work are not to be confused with the characteristics that make a good performance unless we are ready to accept all performances of any work as good performances of it.

570

THE JOURNAL OF PHILOSOPHY 111. REPLY TO BARBARA HERNSTEIN SMITH

~Mrs.Smith rises as the champion of unwritten literary works. She complains that according to my view a work never written down, no matter how often recited, would never exist at all. She feels I have treated oral utterances, along with silent readings, quite shabbily. I plead not guilty. For what seem to me quite sound reasons, I have indeed treated oral utterances of literary works quite differently from, but no less respectfully than, performances of musical works. Let me explain. I have suggested that we have a choice of three ways of construing the relationship between the written and oral versions of a literary work: (a) to take the oral utterances as compliants of-i.e., as denoted by-the written inscriptions; (b) to take the written inscriptions as compliants of the oral utterances; (c) to take utterances and inscriptions both as equally instances of the text-in a broadened sense of "text."

Alternative a is analogous to my treatment of music, where the performances comply with-are denoted by-the score. Why not choose this course, then? For two reasons: first, that since verbal as contrasted with musical inscriptions have other compliants such as objects and events, this would confuse matters by giving verbal inscriptions two different sets of compliants; second, that since inscriptions and utterances perform the same functions of telling stories, describing scenes, etc., treating the relation between them as asymmetrical seems arbitrary and deforming. Both reasons weigh also against alternative b and in favor of c-that is, of taking inscriptions and utterances of words as instances of the same symbols, with the difference between the written and the oral counting for no more than differences among fonts of type or among voices of speakers. Along with this first decision goes another. Obviously, and Mrs. Smith agrees here, a literary work cannot be identified with the objects and events that are denoted by the words. Rather, the work must be identified with the words themselves, written or oral. Thus utterances along with inscriptions are instances of the work. My treatment of literature therefore differs from my treatment of music not in downgrading the verbal utterances but in upgrading the

SOME NOTES ON LANGUAGES OF ART

57I

verbal inscriptions. In music, only performances, not inscriptions of the score, count as instances of the work. I n literature, on the other hand, the inscriptions and utterances are peers as instances of the work. An unspoken work exists in its inscriptions alone, an unwritten work in its utterances alone; and Mrs. Smith's defense of the unwritten work is unnecessary. Treating poetry and music in the same way, as Mrs. Smith advocates, would not at all affect the status I assign to recitations and performances as instances of a work. Rather, either (1) to the horror of musicians, copies of the score will have to be considered as no less instances of a symphony than are performances, or (2) printed or written copies of a poem ~villhave to be considered not as instances of the poem but as merely prescribing or defining what auditory events are the only instances of the work. Both courses disregard an important difference: that whereas visual characteristics of the score that do not affect performance are irrelevant to a musical work, certain visual characteristics of the text (from simple lineation to the complex layout of some poems by e. e. cummings) are, often, quite apart from any effect upon recitation, integral to a poem. Silent recitations of a poem or silent performances of a score by a musician might be admitted along with spoken recitations and played performances as instances of works of poetry and music. Still, if a copy of the text exists, a poem that is never recited either silently or aloud has an actual instance, whereas even if a copy of the score exists, a song that is never sung silently or aloud has none. Mrs. Smith's second complaint is that I count all texts, including grocery slips, as literary works. I do not. T o say that all literary works are texts (written or spoken) is not to say that all texts are literary works. I did not undertake to define the special kinds of texts that are poems, novels, etc.; and the nearest I came to distinguishing literary works in general from other texts is in suggesting certain 'symptoms of the aesthetic'. These symptoms provide a guide rather than a definition. Mrs. Smith, on the contrary, seems to propose defining literary works as fictive texts. But obviously falsity is neither enough nor required to make a text a literary work. The definition goes wrong both in including all lies as literary works and in excluding all histories and biographies. And Franklin's Autobiography, while not a literary work if true, would become one if falsified by inserting "not"s or changing names and dates. All that Mrs. Smith really wants to urge, I think, is that what matters in the case of literary works, fiction or nonfiction, is not

THE JOURNAL OF PHILOSOPHY

572

what if anything they denote but what their structure is and what they evoke or suggest. Early in my book (p. 26), I made much the same point concerning the pictorial arts:

. . . the

world of pictures teems with anonymous fictional persons, places, and things. The man in Rembrandt's Landscape with a H u n t s man is presumably no actual person; he is just the man in Rembrandt's etching. In other words, the etching represents no man but is simply a man-picture, and more particularly a the-man-in-Rembrandt's-Landscape-with-a-Huntsman picture. And even if an actual man be depicted here, his identity matters as little as the artist's blood-type. Furthermore, the information needed to determine what if anything is denoted by a picture is not always accessible. . . . But not only where the denotation is null or indeterminate does the classification of a picture need to be considered. For the denotation of a picture no more determines its kind than the kind of picture determines the denotation. Not every man-picture represents a man, and conversely not every picture that represents a man is a man-picture. And in the difference between being and not being a man-picture lies the difference, among pictures that denote a man, between those that do and those that do not represent him as a man.

Later (253-254), prominence of what is exemplified or expressed finds full recognition as a symptom of the aesthetic in general. But this one feature will not serve by itself to define literary or other works of art. As she wants to identify art with fiction, Mrs. Smith wants to identify science with fact. Though granting many of the oftenoverlooked affinities between art and science discussed in my book, she writes: "what we acquire knowledge of in a work of art is primarily the work itself" and later "the objects of our aesthetic experiences are artificial worlds, fictive natures, and . . . the consequences of knowing them are confused at one's peril with the consequences of knowing nature proper." But if a work of art is artificial, so is a scientific system. What each reveals is a world of its own. The idea that science simply describes nature is no more tenable than the idea that pictures simply mirror it. As I have argued elsewhere,l there is no such thing as 'nature proper', no one way the world is, nothing already formulated or framed and waiting to be transcribed. All we ever get is one among the many ways the world is. Furthermore, while I quite agree that knowing a work of art or a scientific system is not for the sake of anything else, still 1 "The

48-56.

Way the World Is," Review of Metaphysics, XIV,1 (September 1960):

NOTES AND NEWS

573

nothing can be known in isolation; for knowing-by sense, emotion, or intellect-involves discriminating, comparing, contrasting, and so relating what is experienced to what lies beyond it. In science, unlike art, we are indeed usually less concerned with what is exemplified or expressed than with what is asserted. But to say that truth is asked of science but not of art is far too simpleminded a way of drawing the distinction between them. Much else besides-and sometimes even rather than-truth is asked of science. Moreover, when we examine our tests for truth in science we find them far from alien to tests for quality in art. Mrs. Smith's remark that I almost snub the artist is nicely offset by Mr. Wollheim's remark that I give the artist something like his due. NELSON GOODMAN

Harvard University NOTES AND NEWS The Department of Philosophy at the University of Colorado announces the appointment of Leonard G. Boonin, former University of Missouri faculty member, as associate professor of philosophy; Professor Boonin is a specialist in the philosophy of law. The Department of Philosophy at Emory University is pleased to announce the following appointments, effective September I, 1970: Dr. William F. Edwards, Professor of Philosophy; Dr. Nicholas G. Fotion, Associate Professor; Drs. William Kerr and Dieter Turck, Visiting Assistant Professors. On November 19 and 20, 1970,The University of Illinois at Chicago Circle is sponsoring a conference on "Philosophy and the Black Liberation Struggle." The conference will be opened by an address by a member of the Chicago Black Panthers. The major speakers will be: Professor Albert Mosley, Federal City College, Washington, D.C., who will speak on "Social and Political Concerns in Black Philosophy" Professor Bernard Boxill, the University of California, at Santa Barbara, who will speak on "Reparations" Professor Kenneth Mills, Yale University, who will speak on "Dialectics of Black Liberation" Mr. William M. C. Okadigbo, Catholic University of America, who will speak on "Towards a Universal Ideology of Black Liberation" The conference is open to the public and is made possible through a grant from the Carnegie Corporation. Further information may be secured by writing the Philosophy Department, University of Illinois at Chicago Circle, Box 4348, Chicago, Illinois 60680.