« The Relevance of Aesthetic Experience to the Examination of Phenomenal Consciousness | Main | »
Wednesday, September 01, 2004
What Ever Happened to Exemplification?
Posted by Brian Soucek on September 1, 2004 at 02:06 AM in music, representation | Permalink
In Languages of Art, and more concisely in his chapter from Reconceptions, "How Buildings Mean," Nelson Goodman proposed three ways that works of art might have meaning: namely, through denotation, exemplification, and expression. Within the arts, denotation and expression have both been exhaustingly (if inconclusively) discussed. But of the three, exemplification was the form of reference that Goodman himself identified as one of the "symptoms of the aesthetic". Yet as far as I can see, the concept figures little in recent discussions. So what has happened to the notion of exemplification?
As Goodman defines it, exemplification is reference plus possession. In his words, "Such reference runs, not as denotation does, from symbol to what it applies to as label, but in the opposite direction, from symbol to certain labels that apply to it or to properties possessed by it." If a fabric swatch exemplifies a certain color--yellow, say--then it both refers to the property yellow and itself possesses that property. Not all of the swatch's properties are exemplified by it, however. As Goodman once wrote, "The lady who ordered dress material ‘exactly like the sample’ did not want it in two-inch square pieces with zigzag edges.”
So why is this concept so important for aesthetics? For one thing, because it allows works to mean without requiring that they refer to anything outside of themselves. It thus avoids a standard formalist prohibition on reference "outside the frame". And yet it does so without sacrificing the notion that works of art might mean something.
I'd like to suggest that this is particularly true in music. Peter Kivy has offered forceful arguments for his belief that "music alone" has no semantic content. But his arguments have all been directed at opponents (Richard Kuhns, Arthur Danto, Jerrold Levinson, Schopenhauer) who treat music as representing something or other. Levinson, for example, has suggested that violently passionate music might characterize passion as violent. Kivy's response is that some music certainly contains violence or passion. But as he says in Philosophies of Arts, "Surely it makes no sense to claim that that, ipso facto, commits one to saying that the music represents those things, or is about them, or that those things are its 'content' in the semantic sense of the word. They are merely its 'content' in that they are 'in' the music, as the colors are 'in' the painting: they are part of its 'musical content.'"
I wonder if exemplification might answer Kivy's concerns. After all, Kivy is right to say that violence--or passion, or joy--are 'in' the music. They are properties possessed by it. And that is one of the requirements of exemplification. The other, though, is reference, and this to me seems to be the element Kivy's account is missing. We would somehow misunderstand a violently passionate work if we missed the fact that it had that property. The work refers to its passion; it shows it forth (like the swatch exhibits its yellowness). It doesn't likewise refer to all of its properties--say, the fact that it is made up of 1,245 notes. Of course we could imagine another work that DID refer to the number of notes it is made up of, thereby exemplifying that number. In fact, I think that musical analysis does exactly this: it tries to determine which of the work's properties are exemplified. That's how we might answer the question parallel to Goodman's: How does music mean?
These are just sketchy ideas, but perhaps they are enough to make us wonder why Goodman's concept of examplification hasn't been further pursued.
Comments
Posted by: Michal Gal | Sep 4, 2004 7:12:24 PM :
Brian,
I completely agree: exemplification is central and useful both in general account on art and in specific accounts on artworks. But I think it is useful not only for naturally abstract art, like music, but also, and more interestingly so, for discussing a thematic art.
You say that exemplification "allows works to mean without requiring that they refer to anything outside of themselves". And this explains how abstract art, like music, can mean. First, an artwork does refer to an external thing by exemplifying-—it refers to the label/property that applies to it. While the artwork is denoted by the label it exemplifies, it refers to the group of objects to which the artwork belongs due to being applied to by the label.
One of the main aims for which Goodman uses exemplification is to dissolve the distinction between internal and external properties of artworks, and to prove all of them external. "If an external property is one that relates the picture or object to something else, then colors and shapes obviously must be counted as external," he says in "Ways of Worldmaking" (p.62), "for the color or shape of an object not only may be shared by other objects but also relates the object to others having the same or different color or shape." Exemplified properties are the semiotic counterpart of what Formalists deemed internal or pure properties. Through exemplification Goodman shows that even "pure" works symbolize. The exemplified properties can be also expressive, like "violently passionate"—in that case they are exemplified metaphorically.
Now, if even exemplified ("internal") properties are symbolic, refer to group of things, there is no philosophical difference between thematic and abstract art. And we can see what is the importance of the notion of exemplification, which, you rightly claim, is unjustly neglected in the aesthetics discourse. Exemplification is the aesthetic symbolic function of the work that shows the properties of the work that belong to it as an artwork, the properties that are candidates for appreciation. These are the properties, as you emphasize, that the work shows forth, exhibits, as if saying “look at this”. I think that this is mostly helpful dealing with non-abstract artworks, which Danto and Goodman show as sometimes hard to identify. Thus exemplification touches one of the main issues in aesthetics, specifically in the 20th-21st centuries: not only what we should look at when engaging with artworks, but also how artworks engage with us.
Posted by: Brandon Cooke | Sep 19, 2004 7:04:25 PM :
Brian,
I too have wondered why Goodman's thoughts on exemplification have not been explored more by aestheticians. I actually read a paper on metaphorical exemplification at the 2000 BSA meeting, and so I agree that there seems to be some promise in the notion for dissolving some problems in aesthetics.
I'm worried about a few of the claims you make, though. Some of the worries are exegetical, and others philosophical. In Languages of Art, Goodman argues that expression should be understood as metaphorical exemplification, not literal exemplification (p. 85) There seem to be some good philosophical reasons for this that get to the core of the problem of expression: on the one hand, we experience artworks as manifesting emotional qualities, yet on the other, they don't seem to be the sort of thing (i.e. things with minds) that can bear emotional qualities. Goodman's solution: artworks possess expressive properties metaphorically.
I think it's important to qualify the mode of exemplification in this way. You write, in sympathy with Kivy, that the passion is 'in' the music. Certainly that is the phenomenologically correct description. But without the qualification you present yourself with a dilemma: attribute mental properties to a thing without a mind, or give emotion predicates special aesthetic senses (Zangwill proposes such a thing in his paper on Realism and Metaphor). This second option strikes me as a bad one, too, for the simple reason that emotion predicates have the same structural connections to other concepts in our experience of art as they do in their 'standard' applications to persons.
Goodman does indulge in talk about properties throughout LoA, but argues (p. 54) that exemplification of predicates and other labels is more basic than property exemplification. I'm still not sure whether employing the notion of exemplification commits one to this sort of nominalism.
It seems to me that the biggest challenge for the metaphorical exemplification story is to give an account of what makes certain metaphorical predications appropriate. This is a challenge for most theories of metaphor generally, and the difficulty of giving a convincing answer has led some (wrongly, I think) to conclude that apparently metaphorical predications are not after all metaphorical. Goodman gives some thought to this problem but I don't find has answer satifying. He rejects the possibility that they apply in virtue of some critical pronouncement. He writes that "a symbol must have every property [literally or metaphorically] it expresses; what counts is not whether anyone calls the picture sad but whether the picture is sad, whether the label 'sad' does in fact apply. 'Sad' may apply to a picture even though no one ever happens to use that term in describing the picture; and calling a picture sad by no means makes it so." (p. 88) This passage, incidentally, strikes me as being in tension with things Goodman says in Ways of Worldmaking, but nevermind. The more important worry is that Goodman does not offer any positive account of the aptness of certain metaphorical labels.
Posted by: Theodore Gracyk | Oct 2, 2004 4:22:13 PM :
Brian Soucek wonders whether Peter Kivy might endorse exemplification as a way of giving semantic content to "music alone." Short of calling him on the phone and asking him, I will venture that Kivy would and would not agree. That is, I think Kivy does admit to exemplification of some properties (as opposed to their instantiation!), but Kivy wouldn't count this as a case of semantic content without the addition of verbal or written text. But then again, shouldn't we say the same of Goodman's fabric swatch?
Here's my thinking: Kivy would agree that a piano performance might instantiate a specific musical work, simply by having a performance that complies with the score. But this falls short of expemplification, as Goodman presents it. (E.g., Glenn Gould's performance of the Goldberg Variations during a concert performance has certain properties possessed by that work, but the performance isn't a SAMPLE of the Goldberg Variations.)
Yet Kivy does recognize that some musical works involve structural representation (e.g., when Bach's melody rises on the word "heaven" when Bach provides music for a religious text). This is not to be confused with “Hearing in” the music with the aid of a title or another text (e.g., hearing the train engine in Honegger's "Pacific 231).
But if a rising melody is going to be a structural representation, then the melody must exemplify rising. And the rise in the melody must be their apart from the text. The text differentiates a mere rise in melody from one that exemplifies, but this is not an objection to exemplification. If I'm at the fabric store, certain social conventions determine that some swatches of fabric exemplify, while other swatches of identical fabric are merely scraps for the garbage.
Posted by: Derek Allan | Feb 27, 2005 7:19:08 AM :
Over at the AESTHETICS-L list we are currently having a discussion about Goodman's 'Languages of Art' in case anyone would like to join in.
