From Chris Andrews' review of Donald L. Shaw's Spanish American Poetry after 1950: Beyond the vanguard:
[Shaw] points out, for example, how [Olga] Orozco's Esbozos frente a un modelo (Sketches in Front of a Model), where the idea that writing poetry is like trying to "translate a text written in a constantly changing code", echoes Borges's famous statement in La Muralla y los libros (The Wall and the Books): "this imminence of a revelation which does not occur is, perhaps, the aesthetic phenomenon".*The young ultraist Borges believed that metaphors are the primordial element of poetry, and that they should be both novel and "effective" (eficaz), but by the 1950s he had decided that the real primordial element is rhythm and that all good metaphors are variations on familiar ones. Shaw suggests that Borges had lost faith in metaphor's capacity to "open up new dimensions of reality". But perhaps he decided that novelty and effectiveness are inversely proportional: the familiar metaphors have become familiar because they carry more cognitive freight more reliably. According to George Lakoff and Mark Turner, there are nine correspondences underlying the "life is a journey" metaphor (people leading lives are travellers, their purposes are destinations, and so on), all of them fairly obvious. When Andre Breton writes of a tongue that is a "stabbed wafer" (in "Union libre"), the correspondences that readers come up with are likely to be idiosyncratic.
I am also more and more certain that rhythm is the primal element of poetry. You lose rhythm, you lost everything. In the womb, we sense rhythm before we see anything. Andrews' idea about familiar metaphors chimes with Louise Gluck's decision, as described in Proofs and Theories, to use common diction. The familiar and the common are not only more reliable vehicles of meaning, they are also more universal. They resist mere novelty. They challenge the poet to deepen his game.