[The following is a significantly revised, expanded version of a comment posted at Kasey Mohammad's Lime Tree blog on 6/22/07, part of a discussion pertaining to issues of "literary value" and "competence." Thanks to Ammiel Alcalay, Jasper Bernes, Joshua Clover, Simon DeDeo, Norbert Francis, Asher Ghaffar, David Lau, Aryanil Mukherjee, Dale Smith, and Barrett Watten for public and back-channel responses that prompted me to elaborate on certain points. KJ]
I've noticed for some time, and commented, on a few occasions, that there seems to be a lack of knowledge of (or interest in) traditional prosody amongst a good segment of the "post-avant" community. The same goes for grammar and linguistics. Truth is, the rare times I've seen post-avant critics make use of grammatical analysis, I've noticed that the result is usually quite pedestrian and gratuitous (prepositions! direct objects! noun clauses!) Often, the result is preposterously wrong.
Thus, here I offer something that may have some productive relation to issues of poetic competence. Let's see...
I recently read the very interesting interviews conducted by Aryanil Mukherjee with Charles Bernstein and Mark Wallace, at the superb Bengali on-line magazine Kaurab.
In their interviews, Bernstein and Wallace are asked about Language and "post-language" poetry in relation to "grammar," and each poet responds, in rather unambiguous register, to the effect that power relations of the dominant social order are reflected in grammar's conventional, basic structures. Bernstein, for example [more on his specific list of categories below], says the following:
AM: You and other Language Poets have professed that grammar structures tend to support the power structures of Western societies. Could you explain that with an example?
CB: Grammar, vocabulary, diction, form, and style reflect the power relations in a society. You can't change the society by changing your grammar[,] but any radical social, economic, or cultural change must necessarily come to terms with its rhetorics and its metaphors. [nota bene: No example, as requested, is forthcoming from Bernstein. KJ]
And Wallace, within a long answer on Language poetry and its relationship to younger post-language poets, makes the following echoing (and in context, approving) remark:
"In particular, many language poets have noted the way [sic] in which grammar structures tend to support the power structures of western societies."
I was surprised to read these comments. Such claims of insidious correspondence between normative grammatical relations and false ideological effects were, of course, a staple of early Language poetry politics. But I hadn't seen the "grammar principle" repeated anywhere for a long time, and I assumed it had more or less been abandoned as a naive, quasi-Skinnerian or Tel Quelian holdover from the late-60s, when Langpo theory was taking its baby steps. After all, other hyperbolic notions from Langpo's utopian phase (ending ca. first Gulf War) that made claims for linguistic structures reflecting and reproducing "capitalist" categories (e.g., the equivalence between "reference" and "commodity fetishism") have pretty much been left behind in a quiet that would make a Quietude poet blush. Even as the poetry, in its generic contours, has remained little changed, and even as the scorn for poems written in "the language of social and linguistic norms" has gone on... 
In any case (and not as any kind of linguist), I'd assumed some kind of unstated, humble assent had developed amongst post-avant poets: an assent around the notion that fundamental linguistic structures -- following long-accepted research in the field -- are *not* "socially" generated at their core, but are, rather, a deep function of mind, part of the genetic endowment of the species -- or at very least an expression of subconscious cognitive operations that by and large subtend social/political domains. I'd assumed, in other words, there had been an at least grudging recognition that Chomsky's dominant theory of Universal Grammar -- or else more recent research from Cognitive Linguistics on governing semantic frames that are largely shaped in childhood -- made any "Marxist" proposals about *grammar and syntax proper* as some kind of ideologic superstructural effect a problematic wager, to say the least.
But since I seem to have been mistaken, I want to ask, with Chomskian theory primarily in mind, first: In what sense, exactly, as Bernstein and Wallace have it, would underlying structures of grammar mirror and reinforce existing social orders? Does grammar do this at phonological, morphological, and syntactical levels? If so, how is linguistic phenomena in these areas -- the most traditional areas of grammatical study -- shaped by ideological forces? What, precisely, is the ideological source and register of such shaping? What would be a standard, reified syntactic structure, for example (to re-enter Aryanil Mukherjee's query), that could be seen as instantiating some collective expression of "false consciousness"? These would be questions that are begged by Bernstein and Wallace's apparent assumptions.
Or, second, assuming that many post-avant poets may now be basing their case on Cognitive Linguistic ideas, the main offshoot and challenge to Universal Grammar since the 1980s, I would ask: In what sense (following CL) would grammar, as the subconscious mapping of temporal and spatial frames that govern speaking and understanding, represent ideational refractions of this or that set of power relations? Is it that these mental processes are primarily enacted, at depth, by sociological stuff? How so, if so? These, also, would be questions that are begged.
Social relations and their effects on discourse and communication are, of course, very real and of obvious import to poets. But that these are set within and deeply graphed across foundational psychological matrices (if linguistic science is to be believed) is also very real and of obvious import to poets, too -- particularly, perhaps, to those still tendering simplistic and supercilious polemics contra "the language of...linguistic norms" (whatever, exactly, that might mean).
The study of language is a pretty precise and rigorous science, so facile pronouncements about its operations made by poets, especially when these get proffered as key elements of a poetic politics, should be explained more specifically and at some length. (It's possible that Bernstein and Wallace are using the term "grammar" somewhat loosely and confusedly, and that they mean something like discourse analysis, or something in relation to the sub-area of pragmatics. Or perhaps, since the term appears in passing, they are talking about "metaphors," in the sense of George Lakoff's recent popular writings -- though one wonders [a central bemusement] how disjunctive techniques in service of textual opacity are at all required to expose the often-harmful uses of metaphors... Whatever the case might be, a bit more rigor in explanation would be helpful.)
In the meantime, social-constructivist critic/poets of the by-now-official "avant garde" might do well to ask themselves the following, ever more enmeshed as their practices are with the prescriptive rules of academic careerism and the standard syntaxes of authorial position-taking: To what extent -- to reorient Bernstein's remark -- might "grammar, vocabulary, diction, form, and style" within the post-avant subculture more directly and relevantly reflect a range of power relations at work in the literary field itself? And what complicity might "avant" poets share in maintaining and extending such relations? These matters, I'd argue, are much more immediately critical to politically minded poets than any purported homologies between grammatical conventions and cultural hegemonies.
Getting back, though, to my above-mentioned surprise at all this, perhaps I shouldn't be so surprised at all. Because if one asks what there is in Langpo and its "post" literature that takes stock of key findings in linguistics over the past five decades and really attempts to deal with these in relation to the reflectionist schemas of the group's heroic stage (a stage perhaps still more extant than realized!), well, what one realizes is that there is little to nothing available at all.
And one must ask how it is that the epistemic revolution initiated by Chomsky in language science seems to be, even now, of decidedly minor importance to the Language group and its progeny. 
Could it be there's been a tacit awareness that tenets of that revolution are at potential variance with certain post-avant dogmas regarding the political implications of "radical" formalist poetry? Or might it more simply and banally be that there generally exists -- to use Kasey Mohammad's phrasing from his original post -- little else in avant poetry than a "bare sufficiency of competence" when it comes to linguistics? Which may be, possibly (I suggest this sincerely), a good enough reason to drop the term "Language" until the situation changes?
I'd propose that some of the elementary issues touched on here have imperative weight for poets of the innovative sphere; indeed, the problems that are posed centrally pertain to the very ontological roots and formal practices of the milieu. I'd suggest, in fact (knowing I'm hardly alone in feeling so), that should these problems not be soon engaged with a strong measure of candor and care, a sagging theoretical habitus is in stress of coming down, and not too distantly, in a heap of semi-theological straw.
 Bernstein's phrase from his 2003 parsonical speech "Enough," a barely veiled attack on the Poets Against the War project. My response to his comments can be found here.
 In fact, in a recent e-mail to me, Ron Silliman states categorically that Chomsky's work holds very little of interest to poets. Silliman, one of those post-avant writers who have offered decidedly banal and sometimes inept sentence-parsing reads, would argue that Saussure, Jakobson, or the later Wittgenstein, say, are of greater worth. Which is fine, in a quaintly atmospheric sort of way, but it leaves a yawning area of major questions awkwardly waiting...