From Languages of Art to Art in Mind

Citation preview

From Languages of Art to Art in Mind Dominic M. McIver Lopes The Journal of Aesthetics and Art Criticism, Vol. 58, No. 3. (Summer, 2000), pp. 227-231. Stable URL: http://links.jstor.org/sici?sici=0021-8529%28200022%2958%3A3%3C227%3AFLOATA%3E2.0.CO%3B2-Y The Journal of Aesthetics and Art Criticism is currently published by The American Society for Aesthetics.

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/tasfa.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.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org Tue Jul 31 23:42:57 2007

Symposium: The Legacy of Nelson Goodman

Lopes, From Languages of Art to Art in Mind 227

Dominic M. McIver Lopes From Languages of Art to Art in Mind

straight if we call answers to question (5) accounts of depiction. Depiction is a species of a generic relation of representation, borne by things as diverse as lyrics, shopping lists, and mental states. Answers to any of the five questions fall under the more general heading of accounts of pictures." The questions are independent of each other in the sense that an answer to one may or may not entail an answer to any other. For example, the classic resemblance theory of depiction is in the first instance an answer to question (5). A picture depicts such-and-such because its viewers recognize its resemblance to such-and-such. But, if correct, this claim also distinguishes pictures from descriptions, which we do not grasp by recognizing resemblances. When our answer to one question answers some of the others, it is tempting to run some of the five questions together. Goodman's opponents often yield to the temptation and fail to appreciate that Goodman is concerned to answer only questions (3) and (4). Pictures differ from descriptions because they belong to analog (or syntactically and semanti" cally dense) schemes, and they differ from phenomena such as diagrams and maps because they belong to relatively replete schemes. Briefly, a scheme is analog if any difference makes a difference. Scheme A is replete relative to scheme B if more character-constitutive features are representationally relevant in A than in B . T h i s does not constitute a definition of pictures, nor does it distinguish them from all other symbols, such as sculptures or movies. Moreover, analogicity and relative repleteness are purely formal properties and appear to have no implications for accounts of depiction. Goodman's account of depiction may be described as minimalist at best. Pictures denote or refer. Among his undisputed contributions is his discussion of what he calls the routes of reference-denotation, predication, exemplification, and expression. Pictures represent in all these ways. But Goodman is not interested in the roots of reference-how referential relationships are established.5 His remarks about what determines what a picture represents are entirely negative: resemblance is neither necessary nor sufficient for depiction.6 Beyond rejecting the resemblance theory, Goodman refuses to give "general instructions for determining what a work describes or represents."' Every system of pictures includes a plan of correlation mapping

Nelson Goodman's legacy to aesthetics is not what one might think. Languages of Art, written in the 1960s, opens with the following statement of purpose: "Investigations in structural linguistics in recent years need to be supplemented with an intensive examination of nonverbal symbol systems."l The book goes on to classify and differentiate symbol systems, giving detailed attention to pictures and musical scores, contrasting them with language. However, it is now clear that the concentration on language characteristic of philosophy in the 1960s was preliminary to a broader study of mind and cognition. Goodman's statement of purpose needs updating: Investigations in mind and cognition in recent years need to be supplemented with an intensive investigation of the arts as media for cognition. Goodman's legacy in part is preparing the ground for such an investigation. As evidence for this, I propose to take a new look at Goodman's writings and at recent cognitive science, focusing on his account of pictures.

Most commentaries on Languages of Art have been directed at Goodman's discussion of pictures. Almost all essay a refutation of it. But despite this widespread denunciation, I believe its pivotal elements have slowly gained acceptance. One might ask how this can be so. One possibility is that the view denounced is not, or not entirely, the view proposed. Goodman's is not a convention theory of how pictures represent. It accommodates and even invites a naturalized account of depiction, which draws upon the psychology of perception. There are several questions an examination of pictures may seek to answer. Goodman lists three: ( 1 ) how to define pictures, (2) how to distinguish pictures from all other types of symbols, and (3) how to distinguish the pictorial from the descriptive.* Another question clearly addressed in Goodman's work is (4) how to distinguish pictures from related symbols, specifically maps and diagrams. A final question concerns (5) what determines what scenes and objects each pictorial symbol represents. It will help to keep matters

The Journal of Aesthetics and Art Criticism pictures onto what they represent, but Goodman is silent about what determines or constrains these plans of correlation. Others have been tempted to fill the lacuna, drawing on resources internal or external to Goodman's account of picture^.^ In particular, it is widely held that Goodman favors a convention theory of depiction. Some go so far as to ascribe to him the view that pictures are verbal symbols!g But Goodman denies that he holds either view, insisting that "there is no vocabulary of picturing as there is of saying," and depiction does not depend on "rule-f~llowing."~~ Indeed, the convention view is incompatible with the claim that pictures belong to analog schemes, as conventions are rules operating upon disjoint and differentiated characters.! 1 That Goodman refuses to give an account of depiction does not mean that he thinks none can be given, though, to be sure, certain of his more notorious statements support such a conclusion. He writes, for example, that "almost any picture may represent almost anything," and "the choice among systems is free."12 I urge care in the interpretation of these statements, however. If there are constraints on what can depict what, they are at bottom psychological ones. Our visual systems are constructed so that we can see only certain objects or events in certain pictures: there are limits to the plasticity of vision.1" propose that Goodman's statements do not rule out there being such constraints. In an interesting discussion of the philosophy of psychology Goodman warns us about "universalism"-making necessary conditions of psychological generalizations.14Silence on the psychological limits on human picture-interpretation may be read as a product of anti-universalism. On one hand, the claim that anything may depict anything is true but uninteresting if read as a metaphysical or epistemic possibility-there are no a priori constraints on picture-object mappings. There could be creatures with perceptual systems so malleable that they could see a pink elephant in Wivenhoe Park. On the other hand, if read as a plain empirical generalization about the human capacity to interpret pictures, the claim that anything can depict anything is at best tendentious. Although the latter, tendentious, reading is standard among Goodman's commentators, Goodman gives no argument committing us to it. As Catherine Elgin states in Reconceptions, "to deny

that resemblance is the basis for pictorial representation is not to say that anything can be a picture of anything else. There may well be limits on the structure and complexity of systems we can master."]5 Nothing in Goodman's view rules out a naturalized approach to depiction, which takes what pictures represent to be determined crucially by the operation of perceptual mechanisms with some plasticity. On the contrary, having established that depiction is a species of the broader phenomenon of representation or intentionality, the obvious next step is to explore the connections between depiction and other kinds of representation, particularly mental representation.16 (After all, a naturalized approach is now widely accepted even in the philosophy of language. For example, the reference of certain kinds of expressions is grounded in the contents of perceptual states. To understand a demonstrative such as "that book" one must be in perceptual contact with it.) This is not Goodman's next step, but it is one for which his minimalist account of depiction has prepared us and makes room.17 11. OBJECT-PRESENTING EXPERIENCE

The complaint that Goodman's view of pictures implies that pictures are conventional, or even verbal, is unfounded. But a more subtle objection may still be raised. The refusal to answer question (5) prematurely rules out plausible answers to the other questions. Goodman distinguishes pictures from other kinds of symbols by their formal properties of analogicity and relative repleteness alone.l8 But there are possible representations that belong to analog and relatively replete schemes that we would not deem to be pictures because they lack an essential feature of pictures. When we look at pictures and understand them correctly we characteristically have "object-presenting experiencesn-we see in pictures the scenes they represent.lqhis phenomenon distinguishes pictures both from verbal descriptions and from quasi-pictorial symbols such as maps and diagrams. Goodman suggests in passing that objectpresenting experience is a consequence of familiarity: pictures in familiar systems trigger object-presenting experience^.^^ This will not do. Object-presenting experiences are also triggered by pictures in unfamiliar systems. Picasso's por-

Symposium: The Legacy of Nelson Goodman trait of Gertrude Stein was taken to look like her, albeit not in the manner expected. (The problem with the resemblance theory is not that pictures do not look like their subjects but that they do so in indefinitely many ways.) Moreover, descriptions in one's native language belong to a familiar system but do not generate object-presenting experiences. What Goodman must do is connect the formal traits of pictures to their capacity to produce object-presenting experience. The trouble is that analogicity and relative repleteness appear to be neither necessary nor sufficient for objectpresenting experience. The best explanation of object-presenting experience is surely that our grasp of what pictures represent depends in a special way on the operation of perception. I believe that Goodman's view contains the resources needed for a partial response to this objection. The response is partial because concessions must be made. There is no denying that the capacity to induce object-presenting experience is an essential feature of representational pictures, even if its character is not well understood. This means that analogicity and relative repleteness are not sufficient to distinguish the pictorial. Some symbols belonging to analog and relatively replete systems are not pictures because they do not trigger the required experience. Finally, part of an explanation of the phenomenology of pictures must lie in a naturalistic account of depiction. It is partly because pictorial interpretation is the exercise of perceptual skills that pictures induce object-presenting experiences. (As we have seen, this concession does no damage to Goodman's view, at least on one reading of it.) All this granted, a case can be made that the formal properties of pictures do play a role in an adequate account of object-presenting experience. In making this case I shall bring out another way in which Goodman's account of pictures leads us toward issues in the philosophy of mind. 111. EXPERIENCE: ANALOG AND NONCONCEPTUAL CONTENT

Goodman devised the notion of analog representation in order to distinguish pictures from descriptions and also to explain why pictures are not allographic and hence can be faked. Yet the notion has found its home not in aesthetics but

Lopes, From Languages of Art to Art in Mind 229 in cognitive science. Paired with the notion of digital representation, it provides for one standard typology of computers.21 Like computers, minds are representation machines. Thus, if there is a fundamental difference between analog and digital representations, then it is a matter of fundamental importance whether minds operate upon one, the other, or both.22 In particular, the notion of analog representation is frequently used to distinguish perceptual experiences from other mental states. Mental states come in varieties and one of the tasks of cognitive science is to explain this. The challenge of explaining how some come to be conscious is at present a particularly pressing one. But conscious mental states fall into at least two rough groups, or perhaps clusters, with perceptual experiences the paradigm of one and beliefs the paradigm of the other. The distinction is partly phenomenological: what it is like to see that the cheese is moldy is not what it is like to believe that the cheese is moldy. A less contentious way to draw the distinction is by showing that experiences and beliefs differ in their representational properties. Several writers have proposed that the content of experiences is nonc o n c e p t ~ a lSince . ~ ~ the content of beliefs is conceptual, experience is not a kind of belief.24 Representational states are conceptual if being in them requires possession of concepts of the properties they represent. If believing that the cheese is moldy requires that the believer possess concepts of moldiness and of cheese, then the state's content is conceptual. If being in a representational state does not require the possession of concepts of what is represented, then the content of the state is nonconceptual. To possess a concept one must meet at least First, to possess a two minimal condition~.~5 concept of a property or an individual one must be able to identify instances of it with some reliability. Concept possession involves a recognition ability. Second, to possess a concept of a phenomenal property one must experience a similarity between instances of the property when presented with them. One does not possess a concept of red if one does not experience red things as similar in respect of color. Additional strictures are required to individuate concepts in a more fine-grained manner, but what these strictures are is a matter of some controversy. It is sufficient for present purposes to say that pos-

The Journal of Aesthetics and Art Criticism session of a concept of a property entails ( I ) an ability to recognize instances of it and (2) experiencing instances of it as phenomenally similar. It is not generally acknowledged that there are two types of arguments showing that the contents of perceptual experiences are nonconceptual. A good example of the first, discussed by Mark DeBellis, concerns the cognition of musical sounds.26Many people do not have well-trained musical ears. They may be able to recognize and sing a melody while neither recognizing that two of its notes are identical nor experiencing them as similar. Such listeners do not possess a concept of the note, yet they hear the note-it is presented to them in experience-and they can discriminate it from other notes in the melody. The content of such a listener's experience is nonconceptual. I would like to make two observations about this argument. First, the argument does not claim that no hearers possess a concept of the note. The content of a representation is nonconceptual if possessing a concept of the property represented is not required. Naive listeners hear notes but do not possess concepts of them. Musically trained listeners do possess note concepts. We may say that it is strongly contingent whether a hearer does or does not possess such a concept. By this I mean that we could possess the concepts (through ear training), and this would not entail our having quite different cognitive capacities. Second, this is a case of nonconceptual representation of digitally represented properties. Pitches are analog, but notes are digital. As we shall see, these points are not unrelated. The second argument is that the content of perceptual experience is nonconceptual because, for creatures like us, it is analog. As Christopher Peacocke puts it, there are many dimensions-hue, shape, size, direction-such that any value on that dimension may enter the fine-grained content o f an experience. In particular, an experience is not restricted in its range o f possible contents to those points or ranges picked out by concepts-red, square, straight ahead-possessed by the perceiver.27

The richness and fine grain of experience depends on the acuity of one's organs of perception, not the limits of one's store of concepts. Vision rep-

resents countless shapes and shades of color, and our experiencing them does not depend on our being able to recognize them reliably or to experience them as similar. Again, the analog content of vision is nonconceptual even though a perceiver may in fact possess some or all of the concepts in question. However, it is weakly contingent whether a perceiver does or does not possess concepts of all properties along an analog continuum. To do so, a perceiver would have to come equipped with quite different cognitive abilities from ours. A creature whose color experiences were conceptual would have to possess a repertoire of color concepts large enough to track every discriminable difference in color.** The second argument captures the distinction between experience and belief at a more fundamental level than the first argument. The analog character of experience far outstrips our conceptual abilities. We would have to have very different cognitive abilities were this not so. Perceptual experience differs from belief not just (if at all) in its sensational, qualitative character but in its representational content. The reason is that perceptual experience is analog. Thus, given that pictures trigger experiences as of what they represent, it is no accident that they comprise an analog system of representation. Their analogicity makes them appropriate vehicles for object-presenting experiences. The notion of the analog, first devised as a strictly formal property of symbol systems, also plays a part in explaining why some symbols provide for objectpresenting experiences. Goodman's Languages of Art gave us a clear grasp on how systems of representation differ from each other but said little about why we should have several systems in our repertoire. To appreciate this, we must examine the different contributions they make to cognition. What is remarkable about pictures and secures them a place in our symbolic and cognitive repertoire is, first, that we can interpret them principally through the exercise of perceptual mechanisms and, second, that they have a kind of content that can enter directly into experience. As we continue to investigate pictures and their place in our cognitive lives, we do well to keep the seminal contributions of Languages of Art in mind.

Symposium: T h e Legacy of Nelson Goodman D O M I N I C M. M C I V E R L O P E S

Department of Philosophy Indiana University-Kokomo Kokomo, Indiana 46904-9003 internet: [email protected] 1.Nelson Goodman, Languages ofArt, 2nd ed. (Indianapolis: Hackett, 1976). p. xi. 2. Nelson Goodman and Catherine Elgin, Reconceptions in Philosophy and Other Arts and Sciences (London: Routledge, 1988), p. 129; see also Oliver R. Scholz, "When Is a Picture?" Synthese 95 (1993): 95-106. 3. For a helpful discussion of the relationship between accounts of depiction and accounts of pictures, see Richard Wollheim, "Representation: The Philosophical Contribution to Psychology," in The Mind and Its Depths (Harvard University Press, 1993), pp. 159-170. 4. For a useful extension of this notion, see Neil McDonell, "Are Pictures Unavoidably Specific?" Synthese 57 (1983): 83-98. 5. Nelson Goodman, Of Mind and Other Matters (Harvard University Press, 1984), p. 55. 6. Goodman, Languages ofArt, p. 5. For a contrary interpretation, see Craig Files, "Goodman's Rejection of Resemblance," The British Journal ofAesthetics 36 (1996): 398412. 7. Goodman, Of Mind and Other Matters, p. 84. 8. Douglas Arrell, "What Goodman Should Have Said About Representation," The Journal of Aesthetics and Art Criticism 46 (1987): 41-49, draws on resources internal to Goodman's account. Views that draw on external resources include Richard Wollheim, "Nelson Goodman's Languages of Art," in On Art and the Mind (Harvard University Press, 1974), pp. 290-314; Jenefer Robinson, "Two Theories of Representation," Erkenntnis 12 (1978): 37-53; and Dominic Lopes, Understanding Pictures (Oxford: Oxford University Press, 1996). 9. For example, Catherine Lord and JosC A. Bernardete, "Baxandall and Goodman," in The Language ofArt History, ed. Salim Kemal and Ivan Gaskell (Cambridge: Cambridge University Press, 1991), p. 76. 10. Goodman, OfMind and Other Matters, pp. 10 and 12. Goodman does hold that realistic pictures are those belonging to familiar systems. But if this suggests anything, it suggests only a convention theory of realism, not a convention theory of depiction. Unrealistic pictures and pictures in unfamiliar systems depict. See Dominic Lopes, "Pictorial Realism," The Journal ofAesthetics andArt Criticism 53 (1995): 277-285. 11. Goodman and Elgin, Reconceptions in Philosophy, p. 110. 12. Goodman, Languages ofArt, pp. 38.40, see also p. 231. 13. The claim that depiction is constrained by the limits of the visual system does not entail that what we see in a picture is not partly a matter of learning and culture. Vision is plastic.

Lopes, From Languages of Art to Art in Mind

231

14. Goodman, Of Mind and Other Matters, p. 16. 15. Goodman and Elgin, Reconceptions in Philosophy, p. 115. 16.For example, Christopher Peacocke, "Depiction," Philosophical Review 96 (1987): 383-410; Daniel Gilman, "Pictures in Cognition," Erkenntnis 41 (1994): 87-102; Lopes, UnderstandingPictures; Robert Hopkins, Picture, Image, and Experience (Cambridge: Cambridge University Press, 1998). 17. Such an enterprise need not lead us back to the resemblance theory. Indeed, reasons to doubt a resemblance view of depiction parallel reasons to doubt a resemblance view of the intentionality of visual experienceas well. See Tyler Burge, "Cartesian Error and Perception," in Subject, Thought, and Context, ed. Philip Pettit and John McDowell (Oxford: Oxford University Press, 1986). pp. 128-129; and J. B. Maund, "Representation, Pictures, and Resemblance," in New Representationalism~,ed. Edmond Wright (Aldershot: Avebury, 1993). pp. 49-52. 18. "Representation depends upon certain syntactic and semantic relationships among symbols rather than upon a relationship between symbol and denotatum." Goodman, Languages of Art, p. 227. 19. Of course, such an experience need not and usually will not dispose one to believe one is seeing the scene represented. 20. Goodman, Languages ofArt, p. 39. 21. David Lewis, "Analog and Digital," Nolis 5 (1971): 321-327; and John Haugeland, "Analog and Analog," in Mind, Brain, and Function, ed. J. I. Biro and R. W. Shahan (Brighton: Harvester, 1982), pp. 213-225. 22. See Elliot Sober, "Mental Representations," Synthese 33 (1976): 101-148; and Goodman, Of Mind and Other Matters, p. 16. 23. For example, Christopher Peacocke, "Perceptual Content," in Themes from Kaplan, ed. J. Almog, J. Perry, and H. Wettstein (New York: Oxford University Press, 1989). pp. 298-317; Adrian Cussins, "The Connectionist Construction of Concepts," in The Philosophy of Artificii Intelligence, ed. Margaret Boden (Oxford: Oxford University Press, 1990). pp. 368440; JosC Luis Bermudez, "Nonconceptual Content: From Perceptual Experience to Subpersonal Computational States," Mind and Language 10 (1995): 333-352; and William Alston, "Perception and Conception," in Pragmatism, Reasons, and Norms, ed. Kenneth R. Westphal (Fordham University Press, 1998). pp. 59-87. 24. This is not to deny that belief may influence the character of sensory experience. Perception may be both plastic and nonconceptual. 25. Mark DeBellis, Music and Conceptualization (Cambridge: Cambridge University Press, 1995). 26. Ibid., chap. 3; see also Alston, "Perception and Conception," pp. 77-80. 27. Christopher Peacocke, A Study of Concepts (MIT Press, 1992). p. 68; see also Peacocke, "Perceptual Content," pp. 298-317; Alston, "Perception and Conception," pp. 80-81. 28. Of course, color experience might fail to be analog, if we had very poor hue color acuity.