Philosophy and Literature 27.1 (2003) 188-195
[Access article in PDF]
Wittgenstein and Poetic Language
B. R. Tilghman
ACCORDING TO RICHARD KUHNS there are certain important parallels between symbolist poetic theory, especially that of Paul Valéry, and Wittgenstein's picture of language developed in the Tractatus Logico-Philosophicus. 1 It strikes me that these parallels are strained and certainly unhelpful in illuminating either what Valéry says about poetry or his poetic practice. In fact, I find certain things in Valéry that are closer in spirit to Wittgenstein's view of language in the Philosophical Investigations. 2 My concern here is not to criticize Kuhns's thesis nor to provide a systematic exposition of Valéry's thoughts about poetry so much as it is to use some of what they say as stepping stones to show the relevance of Wittgenstein's work for understanding at least some of what goes on in poetry.
Kuhn says that "Linguistic forms fascinated both Valéry and Wittgenstein, and the possibility that language might achieve necessity in that it produced sentences held together by bonds of absolute security and irrevocability operates in the logical forms of the Tractatus as it does in Valéry's art. A poem, Valéry believed, must be wrought with such necessity of part to part, such fastness of interior design, that the reader be compelled by his reading to recognize all the implications of the sentences." 3 I set aside trying to make sense of the idea of language achieving necessity and assume that the analogue of the "bonds of absolute security and irrevocability" in the Tractatus is that a perspicuous [End Page 188] symbolism should immediately make clear all the logical implications of any proposition. The analogy between whatever it is that counts as poetic implications and logical implication, however, seems impossibly lame. The word "implication" applied to the sentences of poetry is not a technical term as is the word when used in logic and its meaning may not be altogether clear. Nevertheless, some of what it may mean can be brought out, I believe, in terms of examples. Consider T. S. Eliot's line that "April is the cruelest month, breeding lilacs out of the dead land . . . ," the full sense—implications—of the words are not realized until they are seen in the context of the rest of the poem where we can understand them as making a point about the constrained and empty lives of the dwellers in the waste land. The description of the spring as cruel will help to serve as the poem develops to reveal to us the resistance of people to open themselves to human emotion and reaction. There are too many differences between this kind of thing and logical implication to make the analogy helpful.
At one point Valéry says that "A poem is really a kind of machine for producing the poetic state of mind by means of words." 4 To that statement Kuhns comments that "Valéry's idea that a poem is a machine for producing poetic states is analogous to Wittgenstein's idea that logic is a machine for making sense, that there exists a structure which is the structure of possible meanings." 5 Valéry apparently wanted a poetry in which there is something akin to a necessary connection between the images and tropes within a poem so that there is only one way to read it. Although it is by no means clear what success in such an enterprise would be, some sense can be made of the idea of a poem as a structure of meanings held together by a kind of necessity. I shall try to make this clear. A poem invites us to move in various directions as the force of some figure of speech comes home to us or various associations dawn on us. In music a certain chord progression seems to demand a resolution into the tonic as its completion. This could be described as a kind of necessity, but, of course, not a logical one. Given our...