Goodman's Account of Representation N. G. E. Harris The Journal of Aesthetics and Art Criticism, Vol. 31, No. 3. (Spring, 1973), pp. 323-327. Stable URL: http://links.jstor.org/sici?sici=0021-8529%28197321%2931%3A3%3C323%3AGAOR%3E2.0.CO%3B2-E 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 23:43:28 2007
N. G . E . H A R R I S
Goodman's Account of Representation
Goodman's account of the concept of "representation," which plays such an important part in his book Languages of Art,l is defective and I shall show how. Let us start by asking, "With what sort of representation is Goodman concerned?" Is he concerned with whatever might be referred to as representation, or is he concerned only with some limited form of representation? T h e answer is that in the parts of the book I shall be considering, whenever Goodman talks of "representation," he is talking of "pictorial representation," or the analogue of this in arts other than painting. Or at least that is what he always takes himself to be talking of. In a footnote on page 4 he says: "What I am considering here is pictorial representation, or depiction, and the comparable representation that mav occur in other arts. . . . Some writers use 'representation' as the general term for all varieties of what I call symbolization or reference, and use 'symbolic' for the verbal and other nonpictbrial signs I call nonrepresentational." So although Goodman would say that the word-token "Socrates" written on the blackboard refers to Socrates, he would not want to say that the word-token represents Socrates. not her example of his usage of "represents" contrasts more clearly with ordinary usage: H e would not want to say that if I were to use pieces of chalk to stand for battalions in a tactical I M A I N T A I N THAT
N. G. E . HARRISis lecturer i n philosophy i n the University of Dundee and was an assistant editor of The Philosophical Quarterly (19'70-2).
briefing, I am using the pieces of chalk to represent battalions. For although in the ordinary language usage of "represents" the pieces of chalk represent battalions, they do not pictorially represent them. Since Goodman intends to use the term "representation" consistently to refer to pictorial representation, he has no need to use the word "pictorial" in front of the word "representation," and in general does not do so. But the danger of doing this is that one may occasionally slip back into the everyday usage of "representation" and make a claim which appears on the face of it highly plausible if not platitudinous, until one puts the word "pictorial" back in. I shall suggest that on at least one occasion Goodman falls into this trap. I n chapter 1, in addition to talking about representation, Goodman also talks about realism. H e switches from one to the other a number of times and the result is a confusing picture. T h e two concepts are linked in that only if a picture represents something can it be realistic. But the opposite implication does not hold. A picture can clearly be representational without being realistic. I n other words the fact that a picture is being representational is a necessary but not a sufficient condition of its being realistic. Since the two concepts are linked in this way, they may be used to throw light on one another, in that if something is a necessary condition for a picture to be representational, then it will be a necessary condition for it to be realistic, and if something is a sufficient condition for a picture to be realistic, it will be a sufficient condi-
N.
tion for it to be representational. Goodman's discussion of realism in chapter 1 is largely concerned with establishing that imitation is not a necessary condition of realism and, if he is right, then it will follow that imitation is not a necessary condition of representation. But this is hardly a very exciting conclusion. What it fails to show is that resemblance is not a necessary condition of representation. For resemblance is a much wider notion than imitation. A may still resemble B, although A does not imitate B. For instance, there are many ways in which, say, Constable's landscapes resemble one another, but none need be a copy or imitation of any of the others. Imitation is but one amongst many sorts of resemblance. Furthermore, those who want to claim that resemblance in some sense is a necessary condition of pictorial representation do not want to restrict resemblance to imitative resemblance. Yet Goodman on occasion appears to do just that. After arguing that imitation is not a necessary condition for realism, he claims on page 34: "This leaves unanswered the minor question what constitutes realism of representation. Surely not, in view of the foregoing, any sort of resemblance to reality" (my italics). Now if Goodman consistently chose to use the term "resemblance" to refer only to imitative resemblance, then we might allow that he had proved his point even though that point is much weaker than the one he might prima facie be taken as making. But Goodman is not consistent. In the first section of chapter 1 he quite clearly uses resemblance in a way in which imitation is not involved, for amongst other things he says, (i) resemblance is reflexive; (ii) a picture of Marlborough Castle resembles other pictures more than Marlborough Castle. So despite Goodman's claims to the contrary, I think that whatever the merits of his discussion of realism as a discussion of realism, it throws no light at all on the nature of representation. And since I am primarily interested in representation, I shall not consider his discussion of realism further, except for one incidental matter. Goodman seems to allow for the possibil-
G.
E.
HARRIS
ity of representation being a matter of degree in the way that realism would seem to be a matter of degree. I am not going to concern myself with the question whether realism is a matter of degree, beyond saying that it seems to me at least a plausible thesis. But the question whether representation is a matter of degree is one to which I should give a negative answer. T o me a picture either represents something or it represents nothing, and a picture that represents a certain thing represents that thing no more and no less than does any other picture that represents that thing. A picture that represents the Taj Mahal may represent the Taj Mahal very badly, very unrealistically, but it will be no less a representation of it than any other painting that represents the Tai Mahal. I have said %at Goodman seems to allow for the possibility of representation being a matter of degree. Where is my evidence for saddling him with this dubious doctrine? On page 6 he considers the question, "Is it perhaps the case that if A denotes B, then A represents B just to the extent that A resembles B?" That is, is it the case that the more A resembles B, the more A represents B? Admittedly, he then goes on to say: "I think even this watered-down and innocuous-looking version of our initial formula betrays a grave misconception of the nature of representation." But the suggested misconception he goes on to spell out is not that there are no degrees of representation, but that resemblance is not a relevant consideration. And it takes him fourteen pages to do this. Yet had he thought, as I do, that there is no question of there being degrees of representation, the "watered-down" thesis could have been demolished in as many lines. A second and more important passage occurs on page 230 where Goodman says of symbol schemes, "the difference between the representational and the diagrammatic is a matter of degree." And since he goes on to talk of some symbol schemes being more diagrammatic than others, he seems to allow that some symbol schemes are more representational than others. And firom this it would seem to follow that two pictures of
Goodman's Account of Representation the same subject constructed in different symbol schemes may represent that subject to different extents. But it should be remembered that what Goodman means is that the two pictures may pictorially represent the subject to different extents. Now I am willing to allow that two pictures may differ in degrees of pictorial representation, but if they do they will differ only in the extent to which their representation is pictorial. This is perhaps what Goodman himself means, but he would express it by saying that whilst the two pictures denote the same subject, they represent it to different extents. Goodman's terminology is clearly very different from that normally used, but his use of other terminology might be justified if ordinary usage was either confusing or otherwise defective, or if Goodman's own terminology was particularly illuminating. But to my mind neither is the case. Indeed, I want to claim that his usage of the word "represents" is misleading. For the use of a single word, and the importance he attaches to obtaining an analysis of the concept, give an appearance that the concept is a relatively primitive one. But, as I think is clear from his own analysis, this is not so. Goodman's concept of "representation" is the conjunction, or perhaps what one might better call the intersection, of two concepts, that of "picturing" and that of "denoting," i.e., "representing" in the ordinary usage. I shall show later that unfortunately Goodman fails even to give a satisfactory characterization of the composite concept of "pictorial representation." First, I want to say something more about his attempted refutation of the thesis that a condition of A pictorially representing B is that A resembles B. He quickly disposes of the possibility of the condition being a sufficient one, for there are many cases of resemblance which certainly do not involve pictorial representation. Thereafter he discusses whether resemblance can be a necessary condition of representation. What is striking about his discussion as a whole is his unquestioning use of the concept of resemblance. I have already pointed out how he uses "resemblance" on some occasions as
synonymous with "imitation," and on other occasions uses it more widely. The trouble with the word is that we often use it in contexts where much can be tacitly assumed. We may say that a child resembles his father, "Isn't he like his dad?" But what we may mean is something like "He resembles his father in facial characteristics," or we may mean something quite different, such as "He resembles his father in being an enthusiastic football club supporter." Which is meant will normally be clear from the context. Although we may often use "resembles" as though it were a two-place predicate, really it is a three-place one. The places are to be filled by (1) a thing of some kind, (2) the thing which is said to be like it, (3) the feature or features with respect to which (2) is like (1). How the last of the three places is to be filled is often left unstated because it is obvious. But Goodman consistently treats "resembles" as though it were a two-place predicate and this leads him into error. He attacks the thesis that " 'A represents B' implies 'A resembles B' " by finding cases where we should want to say that A represents B but where A does not resemble B with respect to certain features, such as being a geometrical projection of B. But whilst this will be a refutation of the thesis that " 'A represents B' implies 'A resembles B with respect to being a geometrical projection of B,' " it does not refute all theses of the form " 'A represents B' implies 'A resembles B with respect to features C.' " Now in Western art it appears to be conventional to take the artist's word for what it is he is representing. And this might appear to give Goodman a chance to show that in representation no resemblance in any significant features need be involved. Consider the following example. Suppose a painter of religious scenes always paints a new-born lamb in his pictures and that we know from his written or spoken remarks that he takes this lamb to represent the devil. And suppose that it is not the case that we can identify the lamb with the devil by some other means, such as that the lamb is performing an action known to have been performed by the devil. Consider
N . G .
a particular picture by this artist, say one of Christ feeding the five thousand. Now this would presumably be described bv Goodman as a case in which the devil is represented as a lamb. T h e part of the picture that consists of the lamb-picture would be described by him as a lamb-picture that denotes the devil and, taking his sense of "denotes," I should agree with this. Now it might seem to follow that here we have a case of pictorial representation where the picture resembles in no significant respect what it denotes. But I think the answer to this is that whilst the lamb-picture represents (in the ordinary sense of this word) the devil, it does not pictorially represent the devil, and it is only pictorial representation that I would want to hold involves resemblance in significant features. So I want to say that if X is represented as Y by Z, in general X will not be pictorially represented by Z. Now this is consistent with most of what Goodman says in his section on "representation-as" (chapter 1, p. 6). But prima facie it appears to conflict with what he says on page 30. For he says there: "To represent the first Duke of Wellington is to represent Arthur Wellesley and also to represent a soldier, but not necessarily to represent him as a soldier; for some pictures of him are civilian-pictures." Now suppose we had a civilian-picture of the Duke of Wellington. This will not be a ~ i c t u r eof the Duke as a soldier, so it is not a picture that denotes him and is a soldier-picture. But it does denote him, so it is not a soldier-picture. But Goodman wants to say that thepicture represents a soldier. If he is being consistent with his use of the term "represents," he must mean that it "pictorially represents" a soldier, and this will conflict with what I want to say, for I want to say that it is in this sort of case that pictorial representation is not involved. But Goodman cannot be consistent here in his use of "represents." For he cannot mean that the civilian-picture pictorially represents a soldier for he would then be committed to saying that it is a soldier-picture, and hence that it is both a soldier-picture and not a soldier-picture. In order to avoid this contradiction,
E.
HARRIS
Goodman must mean by saying that the civilian-picture represents a soldier, merely that it denotes a soldier, not that it pictorially represents one. But this will mean, as I have suggested, that he is inconsistent in his use of the term "represents." This is an illustration of the risk i n taking "represents" as a technical term for "pictorially represents." Finally, I shall show that Goodman's definition of "represents" in chapter 6 is defective in that it rules out cases which we should definitely want to describe as cases of pictorial representation, and not as cases of description, which it would seem is what they should be labeled if we use his definitions. On page 226 Goodman claims that for a symbol to be representational it must either belong to a symbol scheme that is syntactically dense or to a syntactically dense part of a scheme that is partially dense. And he defines what it is to be syntactically dense as follows: "A scheme is syntactically dense if it provides for infinitely many characters so ordered that between each two there is a third" (p. 136). So it follows that any symbol from a symbol scheme which is not syntactically dense will not pictorially represent anything. I shall give two examples of symbol schemes which are not syntactically dense at all, but in which one might very well construct a symbol, i.e., a picture, which I think we should all want to say was a pictorial representation. (1) Suppose that an artist forms mosaics entirely from one centimeter squares of plastic which are each uniformly colored in one of ten colors. And suppose that working from these materials he produces on a square floor with fifteen meter sides a picture which could be very easily recognized as being of, say, the Statue of Liberty. Would we really want to follow Goodman and deny that such a picture pictorially represents the Statue of Liberty? (2) People have produced, and even exhibited, pictures which have been
Goodman's Account of Representation
327
produced on a typewriter. If in prosuch pictures from being classed ducing such a picture one rigidly amongst pictures that depict? maintains standard spacing between lines, then the symbol scheme will be s ~ ~ ~ a cdense. ~ G aBut, ~ ~ ~INeIson Goodman, languages of Art (Indimnapoagain, would we want to rule out all lis, 1968); see esp. chaps. I and 6.