often needs qualification, as when we find ourselves needing to describe someone as a “stepmother, surrogate mother, adoptive mother, foster mother, biological mother,” etc., which happens when the various models don’t converge. Lakoff’s point is that “the concept mother is not clearly defined, once and for all, in terms of common necessary and sufficient conditions. There need be no necessary and sufficient conditions for motherhood shared by… biological mothers, donor mothers…, surrogate mothers…, adoptive mothers, unwed mothers who give their children up for adoption, and stepmothers.” I propose that “poetry,” like “mother,” is a “cluster model,” and that the availability of widely varied privileged models for poetry, combined with the impossibility of giving necessary and sufficient conditions that cover all cases of poetry, makes a practice of metacraft incumbent on all of us who write poetry.
Recognizing poetry as a cluster model, and consequently recognizing the variety of privileged models available, helps explain the Creeley anecdote: Creeley and Nims, one sees, privileged one model, and the questioner privileged another. Recognizing poetry as a cluster model also means that no model of poetry is validated by correspondence with some real and eternal Platonic ideal: to reiterate Lakoff’s words, “the concept [poem] is not clearly defined, once and for all, in terms of common necessary and sufficient conditions.” No one’s model is right unconditionally or universally. Not Creeley’s, not Helen Vendler’s or Paul Muldoon’s, not yours, not mine. Creeley and Nims can laugh together at the questioner because they privilege the same model, but not because their model is the “right” or “true” model. Recognizing poetry as a “cluster model” reveals that what is at stake in my writing poetry is not only how robustly I realize my privileged model of “a real poem” (i.e. how I write), but also which model I privilege (i.e. what I write).
No cognitive model of poetry is more widely accepted than that based on a contrast between prose, presented “continuously” on the page, and poetry, broken into lines. This “lineation model” shows up in ways as varied as the familiar joke about converting prose into poetry by expanding the margins, and J. V. Cunningham’s assertion that “as prose is written in sentences, without significant lineation, so poetry is written in sentences and lines.” However common this model may be, though, it is not comprehensive. To note one obvious exception, there is by now a long tradition of the “prose poem,” whose very name indicates both its claim to be poetry and its refusal to privilege the cognitive model that would make lineation definitive of poetry. History, too, says that the lineation model can’t be comprehensive, and was not always privileged. Recall that the Iliad and the Odyssey, those most canonical of canonical poems, were composed orally, by illiterate singers. Our sense of line is orthographic: a line for us is a typographical convention, something that, even if it represents something metrical, is realized as something fundamentally visual, something that occurs on the page, in writing. If our sense of lineation were not orthographic, our jokes about composing poetry by expanding the margins would make no sense. But the Homeric singers could not have been thinking in such terms. The “line” for them was aural, not visual, and oral, not written; it was metrical, with no orthographic aspect at all. Homeric singers didn’t make a contrast between poetry and prose, so such a contrast couldn’t have been the model for Homeric poetry.
That has everything to do with the Claudia Rankine book. Don’t Let Me Be Lonely is printed as prose, but the book is tagged for marketing purposes ambiguously as “lyric essay / poetry.” It couldn’t have been written if Rankine had accepted as her model for poetry the contrast between poetry and prose. The page referred to above, the list of pharmaceutical companies, defies the lineation model. Yet lists have value for us, including potential emotional value: the most obvious example of a list laden with emotional value is the Vietnam Veterans Memorial in Washington, D.C., which simply lists names. Rankine’s list turns out to be a list of the “thirty-nine drug companies [that] filed suit in order to prevent South Africa’s manufacture of generic AIDS drugs,” a suit that attempts to enforce the companies’ claim to own, as “intellectual property,” antiretrovirals, thus protecting their own profits, though doing so would entail the deaths of millions of people, the great majority of “the five million South Africans infected by the HIV virus.” Rankine laments in the poem that “it is not possible to communicate how useless, how much like a skin-sack of uselessness I felt.” The list of pharmaceutical companies comes close to communicating that, though. When I face the Vietnam Veterans Memorial, I understand the irrecoverable loss of thousands of lives of individual human persons in a different way than before, with an emotional immediacy that my general awareness of the fact of those deaths does not possess; similarly, when I see Rankine’s list of pharmaceutical companies, I understand her uselessness because I recognize it as my own uselessness in the face of, and my complicity in the fact of, colonialist plunder of material wealth and scorn for human life. I understand my uselessness and complicity in a different way than before, with far greater immediacy. If we do not wish for that possibility to be excluded from poetry, then we cannot accept as given or fixed the lineation model as our way of thinking of poetry.
The lineation model is inadequate, but what about the model Aristotle proposes in the Poetics? He specifically states that the lineation model won’t do: “the distinction between historian and poet is not in the one writing prose and the other verse.” Even in verse, Herodotus still would be history, and the difference between history and poetry “consists really in this, that the one describes the thing that has been, and the other a kind of thing that might be,” which makes poetry “more philosophic and of graver import than history, since its statements are of the nature of universals, whereas those of history are singulars.” Call this the “algebraic model.” If the lineation model establishes parameters for poetry by differentiating it from prose, the algebraic model constructs poetry by differentiating it from history. History documents the facts, accounting for what actually occurred. Poetry portrays the necessities and principles that underlie the facts, accounting for what did not in fact occur but might have, and may yet. Poetry’s contrast with history resembles algebra’s contrast with arithmetic. The arithmetic equation 2 + 2 = 4 tells me that the two bananas I had today for breakfast and the two I had yesterday total four bananas. The algebraic equation x + .02x = y tells me how much any salary would be after a two percent raise. It is hypothetical, in the logical form of material implication: if a particular thing happens in particular conditions, then the result will be such-and-such. It tells me that if I made $100,000 last year and got a two-percent raise, I’d make $102,000 this year; and it tells me that, even though I didn’t make $100,000 last year, and I didn’t get any raise.
This “algebraic model” gives a way of finding King Lear, a play about events that never happened, and persons who never existed, more edifying than The Decline and Fall of the Roman Empire, a history of events that did happen and were performed by persons who did exist. Or again, of finding The Waste Land, though Eliot’s characters were invented, more edifying than Democracy in America, though Tocqueville’s characters were “real people.” Yet, for all its virtues, the algebraic model cannot be all things to all poems. Where, to name one instance, would the Divine Comedy, with its many “real people,” even people of Dante’s personal acquaintance, fit in this scheme?
Jena Osman could hardly have included in The Network the etymology charts of which the page referred to above is an instance, if she accepted Aristotle’s model as her own. Osman wants to include in her book, and to emphasize, the factuality of language. As she herself formulates things on the first page of her book, “Rather than invent a world, I want a different means to understand this one. I follow Cecilia Vicuña’s instruction to use an etymological dictionary: ‘To enter words in order to see.’” Osman’s book enacts a premise formulated in this way by Jan Zwicky: “Few words are capsized on the surface of language, subject to every redefining breeze. Most, though they have drifted, are nonetheless anchored, their meanings holding out for centuries.” Words, though they change, do not change randomly, so any word contains in itself a form of history, is itself a kind
