Linguistics and Poetics Revisited: Response

Derek Attridge
Rutgers University
Email: Derek Attridge

Three papers, each of which raises far more questions than I could address in ten minutes, and which taken as a group raise even more. I shall be very selective in this response.

All three papers are concerned with the development of an adequate theoretical account of poetic form as part of the total working of the poem; all find traditional accounts wanting. Amittai Aviram finds fault with the notion of "‘imitative form’ -- the idea that sound effects are there to express content," which he sees as the "guiding norm in the field of stylistics." Richard Cureton calls for a theory of "both literary form and literary content, not just a theory of one or the other, or some selected aspect of both." Steven Willett notes that "we lack a robust heuristics to show how meter means." I share this dissatisfaction, and I am pleased that such intellectual energy and ingenuity is being devoted to finding a way forward. Of course, in such short presentations, we could hear only a brief sampling of the kind of work that is being carried out, and many assertions remain unproven, many concepts unexplained. But rather than focus on these areas of inevitable inexplicitness, I would like to raise a larger issue that binds all three talks, and that I believe has to be thought through rather carefully before we can make any progress toward solving the problems that have been identified.

A theoretical account is an attempt to achieve generality of a kind that will provide explanations of specific instances. A theoretical account that is worth achieving is one whose explanations will be useful. This means that it is important to establish an appropriate level of generality, the one that will make possible the most useful explanations. Moreover, the adjective useful is meaningless unless we specify usefulness to whom for what. Let us assume that those of us engaged in the development of poetic theory are aiming primarily to be useful to today’s readers, writers, performers, and teachers of poetry by enhancing and enriching what they do. (I realise this is only an assumption; a poetic theorist may in fact wish to be useful to linguists, to philosophers, to cultural critics, or to other groups; and different aims may result in different levels of generality being appropriate.)

There are a number of distinct modes of generality, of course. For instance, there is cultural generality. Should a theory of poetic form aim at validity across all cultures? Steven Willett addresses his critique of musico-temporal theories of poetic rhythm to those who attempt such transcultural generalizations. He adduces Ancient Greek, Japanese, and Chinese examples to show the inadequacy of a general theory of meter that identifies it with regular beating. What such examples do not prove, of course, is that a theory aiming at a narrower degree of cultural generality (a theory of modern European meter, or of English meter, for instance) should avoid the identification of meter with regular beating. Because we have words that we apply to the productions of different cultures -- words like "poetry," "meter," even "art" -- it does not follow that they name the "same" entity or practice. And even if there is a traceable historical connection between the cultures, no transferability can be assumed. One of the longest legacies of misunderstanding in the field of English prosody has been the assumption that terms which were found to be appropriate for classical Greek meter would necessarily be appropriate for later western meters.

The other two papers are less clear in their assumptions about cultural generality: Amittai Aviram makes no explicit reference to the cultural breadth of his claims, while Richard Cureton operates both at the most grandly general level of "the human mind" or "human cognition" and its "major products" and more narrowly at the level of the western tradition, as when he cites Shakespeare’s Sonnet 18 as one of "our canonical lyric poems." The more all-encompassing his claims, it seems to me, the less useful his theory is to the reader of poetry in a particular tradition. It would help us understand both these arguments, I think, if both were more explicit about this question -- and we would know whether Professor Willett’s critique of some of the claims of universal generalization could be construed as relevant to them.

Another type of generality is historical generality. Should a theory of poetic form, even if limited to a single culture or a relatively homogeneous group of cultures, aim at validity across the entire historical span for which evidence is available? Once again, the answer will depend on the usefulness of the explanations produced by the different degrees of generality (and this in turn will depend on the actual historical changes which have occurred). It is not theoretically impossible that Old English and twentieth-century verse have so little in common that their shared properties reveal very little about the way they each work. Professor Willett, I would argue, reveals the dangers of historical generalization when he states that the history of the emergence of common measure in English proves that its two shorter lines do not end with unrealized beats. (Since I was responsible for that rather horrible term, which I no longer use, I feel obliged to say something on this point!) As Saussure so clearly pointed out, diachronic facts do not determine synchronic facts. Linear prosodic history, which traces the succession of forms, often fails to ask why a certain form, whatever the reasons for its emergence, attains cultural centrality, when a myriad others don’t. The four-three-four-three stanza in English, as thousands of musical settings show, provides a satisfying variation on the four-four-four-four stanza, because the final beats of the second and fourth lines, by virtue of being mental and somatic rather than acoustic, contribute to the experience of closure.

Professor Aviram’s approach illuminates the issue of historical generalization in a different way. He begins with his own experience of reading a good poem. As theorists of poetry we all no doubt began there, since it was this experience -- its pleasure, its mystery -- that called out for explanation. The danger of such a starting point, of course, is that it’s very difficult not to generalize freely from this experience, to take it as paradigmatic rather than as the product of a very specific cultural, and indeed personal, history. Professor Aviram’s thesis that "poems do not communicate" should perhaps be rewritten as "the poems that I like, when I read them in the way I prefer to read poetry, do not communicate -- and I’m prepared to argue that this is a good way to think about poetry in the late twentieth-century western world". It is certainly a refrain we’ve heard a great deal in the twentieth century, from Richards’s "pseudo-statements" and McLeish’s "a poem should not mean but be" to Derrida’s notion of "suspended reference," and it is presumably no accident that it comes to be prominent in the wake of the poetry of Mallarmé, Rilke, and Eliot. But is it appropriate for Piers Plowman or Paradise Lost or The Prelude, whose authors may well have felt they had something important to communicate, and many of whose readers may have agreed? (It may be the case, of course, that all we contemporary westerners can aim at is a theory of the contemporary, western reading of poetry; or to put it another way, after Mallarmé and Eliot we can no longer read Milton and Wordsworth in the manner their contemporaries read them; but this needs to be spelled out.)

Richard Cureton is also unclear as to whether his is a theory of the contemporary experience of poetry, or about an experience that transcends historical change, that would in principle be true of Langland’s (and his early readers’) engagements with verse as much as Ashbery’s (and his readers’). In his "Poetic Paradigm" he differentiates the twentieth century sharply from earlier centuries, which makes it seem that our understanding of poetry must be different from that of our predecessors, yet he writes as if he were capable of escaping the march of the centuries to achieve some kind of ahistorical insight. (I used to think of him as primarily a Jakobsonian, but he seems to me now more of a Hegelian.)

Among the other kinds of generality we might have considered if there had been time is that of the generic field. Once we have specified cultural and historical range, we need to ask what is the usefulness in trying to develop a single theory of literature, or of poetry, or of any particular subdivision of poetry? Richard Cureton seems to aim at the totality of literature, or perhaps language, while Amittai Aviram moves between literature (or perhaps published texts, since he talks of texts which are "closed off from the ongoing exchange of communication") and poetry. (Roman Jakobson notoriously privileged lyric poetry in his poetics.)

Surveying the wide terrain over which all three of these papers range -- and I should add that in this they are certainly following in Jakobson’s footsteps -- my response is to call for more modest aims, and more piecemeal methods, in the hope of achieving more concrete results. What if we did not assume that there was one way in which form and content relate, in a single entity called "poetry"? What if we assumed, in fact, that the single word poetry obscured a variety of cultural practices, and in any reading of a single text, a variety of responses and experiences? It might be more productive to examine one aspect of the way modern European poetry functions in late twentieth-century culture (taking into account such issues as the difference between reading and listening that Professor Willett so usefully reminds us of), and perhaps to try to relate this to earlier practices, or practices in other cultures -- but without assuming beforehand that we will find anything comparable in them. Of course, as Professor Cureton reminds us, we cannot isolate one aspect entirely from all the other phenomena with which it is caught up, but I believe we have a better chance of producing something useful if we proceed cautiously and pragmatically, expanding our horizons gradually instead of beginning at the outer limit and trying to make everything cohere at once.