For as long as I have been teaching poetry—some 13 years—I have faced a quandary: what is it that enables me to find endless pleasure, fascination, and delight in reading and rereading a good poem, and plenty to say about it—and how can I get this ability across to my students? I have come to believe, not that my students know nothing about poetry, but that they have been taught the wrong thing, a mistake that will be the subject of my paper: the failure to distinguish between literature and communication, between art and message. If the poem is a message, its meaning is rarely clear, and, if it seems clear, its importance is a puzzle. "What is the poet trying to say in this poem?" the hardworking teacher asks a silent room, as if the poet had a verbal impediment inhibiting clarity of self-expression. Alternatively, the poem’s obscurity as message is attributed to the threat of censorship or the mystical complexity of the matter. Yet whatever the reason might be for the supposed message’s obscurity, that same reason is instantly defeated if and when we do discover the message. Alternatively, we may opt for relativism: the poem means whatever you think it means. But in that case, why bother teaching or studying at all? Your own personal beliefs are all you know on earth and all you need to know.
A poem is not a message. Rather, a poem is an aesthetic arrangement of language to program a coherent pattern of surprises. The language of a poem is not communicative discourse, but, as Aristotle might say, an imitation of communicative discourse—so that appropriate reading requires that we simultaneously recognize what the discourse would purport to say were it truly communicative, and remember that it is not communicative, so that our attention is then shifted from the end to the means. The subtle distinction between communication and the semblance of communication defines and constitutes the literary text and therefore must lie at the heart of any truly appropriate and effective approach to poetry. Without it, we are doomed to reinforce literary incompetence rather than to foster literary competence. The rest of this paper will consider the implications of this distinction with regard to poetry, starting with a respectful critique of the influential essay to which our panel’s title alludes, "Closing Statement: Linguistics and Poetics," by the forebear upon whose giant shoulders we stand today, the linguist Roman Jakobson.
Jakobson’s opening thesis is that poetry is indeed an appropriate matter of study for linguists. This is possible because the poetic function is shared between poetry and all other uses of language as one of the six functions of verbal communication. Poetry, then, merely foregrounds this poetic function relative to the other functions of language. Each function is associated with an element of any act of verbal communication—the addresser, the addressee, the code (or language), contact between addresser and addressee using the same code, and a context to which the message refers. The poetic function is the "focus on the message for its own sake." Every communicative act has a poetic function—there is always some attention to the qualities of the message itself—but, in poetry, more attention is paid to the message’s qualities than to any other element. Jakobson then goes on to explain how poetry gives predominance to the poetic function, by means of parallelisms and oppositions—kind of equivalence—at every level of language, from sound to sense to syntax to rhetorical tropes (metaphor, simile, antithesis) to image and ideation. Now, the principle of equivalence normally governs the moment-to-moment selection of the appropriate element in producing an utterance—the right sound, word, phrase, etc.—from among all comparable possibilities. This comparing, contrasting, and selecting is what Saussure calls the "associative function" of language. The elements are then strung together one after the next to form the utterance. In poetry, according to Jakobson, the realm of the principle of equivalence is expanded so as also to govern this combining of elements into a sequence—giving the temporal dimension of poetic language, as a result, its characteristic repetition and parallelism. Jakobson does not explain why this projection of the principle of equivalence from selection to combination should be the way that poetry gives predominance to the poetic function and thus shifts the focus of attention to the message "for its own sake." Presumably, this expansion of the realm of equivalence or parallelism causes the utterance to have qualities of uniformity that make it palpable or opaque rather than merely communicative and transparent. Meaning is always achieved through difference. If the play of differences in an utterance is undermined by the prevalence of equivalence, then attention is drawn away from the object of reference, or meaning, and toward the means of reference, the physical material of language itself.
Jakobson avoids the question of whether poetry is in fact a kind of communication or a noncommunicative use of language. He does say, at the outset, that "The truth values ... as far as they are—to say with the logicians—’extralinguistic entities,’ obviously exceed the bounds of poetics and of linguistics in general." Linguists almost universally assume that language can only be used for communication—that communication is implied in the very term language—and this assumption is shared by the teachers who ask what the poet is trying to say. Jakobson’s formulation of a "message ... for its own sake" confuses the matter further: if the utterance is "for its own sake," that is, not for the sake of conveying meaning, then it is not really a message, though it may take a form that imitates a message, a point Jakobson does not note.
What we have before us, in fact, is a paradox. A message can only be a message if it has a referential context. But if the poetic function predominates and our attention is supposed to be drawn to the message "for its own sake," then the referential context is nullified, unreal, and thus the message is virtual, not actual. Literature works because we recognize what the message would be if it were a real message, but, at the same time, are also aware that it is not a message but the imitation of a message, revealing to us the qualities inherent in messages themselves and thus simultaneously barring any possibility of the message’s actual validity. If I turned to you and said, "Let’s go then," there must be a place to which to go, and you can either assent and come with me or decline and stay here. But if the speaker of a poem says, "Let us go, then, you and I," we do not decide to go or not go, we wait to find out what the imaginary situation is and who the imaginary addresser and addressee might be. Yet this imaginary situation would be impossible did we not also recognize the likeness to a real invitation to accompany someone to a real place.
Communication, the passage of actually valid messages, requires (a) a referential context, a real world to which words can refer; and (b) a communicative context of exchange between real people that can at least hypothetically go on indefinitely—that is, for which there is no necessary and definitive closure, only the happenstance choices to stop exchanging messages for the moment or the unpredictable accidents of interruption or death. A statement is either true or false relative to the real world. A question is meant to be answered (unless it is a rhetorical question, which is a statement, command, or exclamation in disguise). A command elicits either compliance or refusal. An exclamation elicits appropriate emotions and sometimes actions.
The literary text by definition does not refer to a real referential context but alludes to a referential context to which it would refer if it were communication. There are no real Three Bears, but one should know what a girl is like and what bears are like, in general, in order to appreciate the story of Goldilocks. A literary text is frozen in its form and closed off from the ongoing exchange of communication. It is the mere fact that it can be repeated verbatim without becoming less appropriate to its moment that distinguishes a literary text as such. The appropriate response to a literary text is not to take it as a message—not to look for the Three Bears—but to grasp the pattern of the whole and to enjoy its program of surprises and, often, its poignancy—in other words, to engage in aesthetic pleasure. Thus, whereas a statement in communicative discourse can be judged as true or false, a statement in a literary text is neither true nor false, it is either appropriate or inappropriate, artistically convincing or unconvincing, when it occurs in its fixed place in the text.
Now, because the literary text alludes to normal reference rather than engaging in it, the function of the literary utterance is then to illustrate the workings of the language of which it is made rather than to communicate anything. Just as a scientific model of the weather imitates the weather in order for us to study its causes and mechanisms—but it does not actually chill us with snow—so literature is an imitation of real language and discourse that exposes how real language and discourse work. The literary text invites reading and rereading because its referential effects cannot be resolved into truth or falsehood by means of ongoing communicative exchanges. Thus, wereas the answer to the question, "Why does it say this?" when applied to a statement in a communicative message is either "Because it’s true" or "Because the speaker is lying," the answer to the same question when applied to a statement in a literary text is an interpretation, and requires a study and recognition of the pattern of the whole to understand the function of each part in contributing to that pattern.
The paradox of literature—that it must imitate a message, but, in so doing, cannot actually be a message, and therefore undermines the perfection of its own imitative powers—this paradox is a manifestation of the paradoxes that govern life itself. Knowledge, truth, and being are always paradoxical, inhabiting the contradictions between principle and empirical truth, theory and practice, general and specific, mind and brain, inside and outside, etc. So representations of knowledge, truth, and being are paradoxical as well. In normal communication, the paradoxes of being are effaced, put off, in the interests of practical business and effective communication. Since communication continues indefinitely, the resolution of paradoxes is put off into the indefinite future. But in literature, the paradoxes are exposed because the text is bounded. The paradoxes cannot be deferred through the process of an ongoing communicative exchange, but instead become apparent to the reader, merely because the literary text is destined to indefinitely extended contemplation rather than the passing business of communication. The paradoxes of life are painful and frustrating, but literature puts the paradoxes into play, as abused children may make the very themes of abuse into objects of play. Such play neither expresses nor resolves the problems of living a paradoxical existence, but play maintains our sanity and our ability to go forward.
Poetry plays by turning our attention specifically to paradoxes involving the physical material of language, its sound—sound effects, meter. Prose fiction turns its attention to paradoxes in other linguistic processes—syntax/sequence and means of representation—plot, point of view, character, etc.
Poetry best shifts our focus toward the physical material of language by making audible rhythm. The result is that the mind must be split between two incommensurable aspects of language—sound and meaning. This incommensurability is a paradoxical fact about language in general, but poetry draws attention to that fact rather than obscuring it. This is an instance of how literature exposes paradoxes otherwise hidden. A crucial paradox about language is that it must seem natural in order to maintain transparency but is actually arbitrary and conventional. Poets and readers will exploit the built-in attention to sound in poetry to play with the paradox of language. One way is the pretense that language is not arbitrary but natural. Rhymes may pretend a certain Cratylism (the belief that words sound like what they mean), as if the association of rhyming sounds arose necessarily from a connection in meaning. That very pretense is also exposed as an illusion in the very act of being put forward.
For whatever reason, the idea that poetry exposes the illusion of naturalness in language—that is, deconstruction—has been considered a sort of deflation or devaluation of poetry. Nothing could be further from the truth. To recognize that a magic trick is an illusion does not deprive us from wonder at the trick—on the contrary, it makes that wonder possible. Just as when we see close up that a Van Gogh painting of irises is mere blobs of paint, and then are wonderstruck as we step back and these blobs become a convincing representation of the flowers, so in poetry, the passage between the two sides of the poetic experience, real sound and imaginary reference, is endlessly fascinating. This dualism is a manifestation of the wonder and poignancy of the paradox of being itself—like the passage from brain to mind, the passage from empirical knowledge to theoretical principles, or the passage from moral ideals to actions and judgments. Accordingly, poets and readers often exploit these facts to make poetry not only an imitation of discourse but an imitation of an imitation in an endless mise-en-abime, a dazzling hall of mirrors of self-reflexivity.
To attempt to close the gap between these two sides of poetry, or of being, far from being a revelation of poetry, actually destroys its wonder. Such is almost the case in Jakobson’s actual comments on Poe in the same essay, which fall under the heading of "imitative form"—the idea that sound effects are there to express content. This assumption, along with the confusion of poetry with message, is the guiding norm in the field of stylistics, resulting, typically, in a superficial reading of the content, a neglect of its intricate complexities. Indeed, while the details Jakobson notices are dazzling, he does not assimilate them into a reading that attends to the narrative and thematic elements of the poem, and implicitly trivializes these as the purport of a simple message for which the intricate sound effects are the means. But Poe’s "the pallid bust of Pallas just above my chamber door" is delightful precisely because of the implied assertion of the impossible—that pallor and Pallas have anything really to do with each other—and, meanwhile, the poem’s narrative content foregrounds the theme of illusion, as a bird that can only mechanically repeat the same word is made to seem as if it were responding prophetically to the speaker’s questions—even as the speaker’s own narrative allows us to see through this very illusion.
The dualism of sound and meaning is crucial to the life and pleasures of poetry because otherwise, poetry devolves into a communicative message—whether personal, moral, political, or metaphysical—as if the content were either patent and trivial or true and authoritative—as if the poet were either a self-involved cryptographer, a morally superior being, a propagandist, or a mere dupe to ideology. More recent efforts to cross the gap, inspired by contemporary brain science, strike me as posing the same problem to poetry’s fundamental reason for being. The dualism underlying poetry is not a problem to be solved, but the paradoxical essence of its literariness and the cause of the very pleasures to which we must open the doors for ourselves and our students.