Being numerous: The twentieth-century poetic imagination of the ground of social life

This dissertation addresses the current impasse in the historiography of twentieth-century Anglo-American poetry. Critics of the period have identified two traditions: a "right" (descended from Yeats, Eliot or Stevens) that defends ontotheological accounts of value, withdrawing into atavistic sciences and archaic institutions; and a "left" of empiricist and experimental poets (descended from Williams, Zukofsky or Stein) that adopts ironic and secular accounts of agency and identity. There is frequent debate about which poet belongs to what tradition and about which tradition is to be preferred, but there is little disagreement about the division itself.;I argue that this opposition is false from the outset, and that we can unify our poetic history by revising our object of analysis. Where critics have been concerned to give better and more precise descriptions of the poem as an object, I identify a central strain of radical and rationalist poetics for which the poem is more or less irrelevant to poetic intentions. The poets I consider here (the Language poets, George Oppen, Frank O'Hara and William Butler Yeats) are typical rather than marginal in that they understand "poetry" as a kind of faculty or capacity, rather than an object or performance. For them, "poetry" identifies a form of creative potentiality that minimally distinguishes what a person is.;The historical study of poetics as what I will call a discourse of potential, rather than a tradition of actual texts, complicates the work of critics who have sought to understand the politics of poetry by examining books as actual or symbolic commodities or discerning the varieties of poetic "difficulty" as strategies for securing cultural prestige. Where these accounts attempt to understand the poet's work primarily in terms of the poet's material and political interests, I show that poets profess a prior and importantly immaterial interest in the pre-political problem of acknowledgment: not the question of how the privileges of identity are to be justly distributed, but of how to secure epistemologically the ground of identity; not just of how to do things with persons, but how to know that a person is there at all.