Chomskyans like Steven Pinker make two claims that at first blush seem to make strange bedfellows. On the one hand, they argue that the differences between the natural languages humans use to communicate hide a common universal structure of thought or universal grammar, such that, in Pinker's words, learning to speak a natural language is a question of translating universal "mentalese" into whatever particular language one is brought up speaking.  On the other hand, Pinker thinks grammar conservatism, the urge to differentiate between correct and incorrect usage of a language, is mere snobbery.

On further examination, of course, the positions are not at all contradictory.  Obviously if the languages we learn to speak are all outward manifestations of a universal mentalese, then it hardly makes a difference whether we say "aren't "or "ain't" or use object pronouns in subject clauses, since all these expressions are equally ephemeral expressions of a common linguistic bedrock. I wonder, then, if the complementary pairing is equally true. If one remains unconvinced by the arguments for the generative grammar, will one also be less tolerant of deviations from standard usage, more inclined to grammatical snobbery?

Test case of one: I remain deeply skeptical of the notion that using a natural language means translating from a universal mentalese, and am much more inclined to accept at least some degree of linguistic relativism: the idea that the languages we speak have the capacity to color how we perceive the world.  And sure enough, I am an unrelenting grammar snob.  Many have been the occasions when I have stopped myself in mid-correction of my son (or worse, one of his friends), as he sows his sentences with superfluous "likes" or starts a phrase at the end with a "me and him," and wondered self-critically if my urge to wrench his speech into proper usage is nothing other than a symptom of the onset of curmudgeonly age, ill-befitting a more tolerant political disposition.

But perhaps there is something to this parallel after all.  If Pinker feels free to embrace all usage as equal, it is precisely because, for a universal grammarian, all usage is equal—equally unimportant, that is. But if one remains unanchored by the comforting certainty that humans share a common linguistic bond, does not the urge to cling to local rules and structures begin to make more sense? And this would be the case even if one accepted, as I do, that in the long run one's own language is doomed to the same inexorable fate that all languages succumb to: evolution, transformation, or even eventual extinction. In fact, it is precisely because I know this that I cringe all the more when I hear Lady Gaga sing "you and me could write a bad romance." (With grammar like that, what else could it be?)

So, in a nutshell: yes, different linguistic communities have different standards, and there is no neutral arbiter to distinguish one as superior to another.  Yet the same linguistic communities are often part of larger communities with institutions dedicated to the impossible task of freezing language long enough to establish standards and defend them, futilely, against their inevitable demise.  While by definition conservative and certainly more readily available to children of privilege, I do not believe these standards automatically translate into a political conservatism, just as being grammatically-challenged is no guarantee of a progressive political character.  Rather, I believe that some torchbearers in any linguistic community, the educators being the most likely candidates, will inevitably cling to these standards and endeavor to preserve them.  Just as inevitably, usage will adjust, and the standards will eventually follow suit.  

As with ethical arguments, then, the assumption that the nonexistence of objective, independent absolutes leads to moral anarchy is false; people are and will remain committed to ethical norms as determined largely by their communities, and these norms can and do change, but only with difficulty. The fact that these norms are not the expression of an underlying universal law does not make them any weaker, though. Like grammar conservatism in the face of linguistic relativism, they are probably all the tenacious because they are on their own.