In lieu of an abstract, here is a brief excerpt of the content:

Crispin Sartwell SUBSTANCE AND SIGNIFICANCE: A THEORY OF POETRY Jean-Paul Sartre once said that what distinguishes the writer of poetry from the writer of prose is that the poet "considers words as things and not as signs."1 I think that this claim embodies a deep insight into the nature of poetry, and I want to develop it into a reasonably precise account of what poetry is. The immediate problem that arises from Sartre's view is that it seems mysterious how anyone could regard words as objects without being fundamentally confused. For it is clear that Sartre thinks that the poet regards words as material objects. But a word, or a sequence of words such as a poem, is not a material object. A word or a poem can appear in many places at the same time; it can be inscribed, read, recited, or mentally traversed. The poem "itself" is not identical with any of these particular manifestations. I can burn my copy of Shakespeare's sonnets, but I cannot burn Shakespeare's sonnets; a poem is, in the phrase of Arthur Danto, "logically incombustible."2 Shakespeare's manuscript is lost; the sonnets live on. My recitation of a poem ends, say, at three o'clock, but the poem itself does not cease to exist at three o'clock. If identical objects cannot have divergent properties, then the poem is not identical with any of its particular manifestations. Nor can we identify a poem as the scattered, diffuse object consisting of all its inscriptions or recitations; to do so would be to identify the poem with a (perhaps massive) repetition of itself. A sonnet has fourteen lines, but the set of all the inscriptions ofShakespeare's "Shall I compare thee to a summer's Philosophy and Literature, © 1991, 15: 246-259 Crispin Sartwell247 day" is an item with an extraordinarily large number oflines. The poem itself does not grow as the set of its inscriptions grows. The poem, to repeat, is not a material object; rather, it is that which all the manifestations of it have in common; it is the structure or sequence which all the inscriptions and recitations share in virtue of which they are inscriptions and recitations of the poem. Now I do not intend to tackle the ontological problems surrounding poetry in any full-fledged way; to do so would strand us in extremely rugged metaphysical terrain. The ontological status of a poem might be in some ways analogous to the ontological status of a number, a property, and other purportedly "abstract" objects. And one cannot distinguish works ofpoetry from works ofprose (and, again, the project here is to make such a distinction) ontologically. Moby Dick is logically incombustible, as are, discouragingly enough, government reports. I propose, then, to adopt a familiar way of talking about the poem itself in distinction from its inscriptions, recitations, and mental traversais; I shall term a poem a type of which there are tokens. The tokens are the inscriptions, recitations, and so forth; the type is what these inscriptions and recitations have in common. The type/token terminology was first adopted by Peirce, and has been taken up by Wollheim, among many others.3 In itself, such terminology does little to disperse various ontological problems, but I am going to proceed in the touching faith that such puzzles can eventually be solved, because the theory of poetry I will develop depends on some such distinction being made out. And though the theory could perhaps be reconstructed even on a fairly thoroughgoing nominalism, I am going to treat poems as abstract entities , because as far as I can see, nominalism of the type endorsed by Nelson Goodman, for example, introduces more difficulties than it solves. Goodman identifies inscriptions ofa poem (for example) as items which display sameness ofspelling, which, relative to a system of notation, correspond exactly with regard to the sequence of letters, spaces, and punctuation marks.4 Among other things, this makes poetry (or any linguistic item) impossible to translate into different languages; it even makes it impossible to update, say, The Canterbury Tales into contemporary spelling. Now in fact there are severe and special translation problems...


Additional Information

Print ISSN
pp. 246-259
Launched on MUSE
Open Access
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.