The argument here is in effect that no vernacular variant of any language could possibly embody a mistake.
This is the very epitome of attacking a strawman. The article didn't come close to arguing that "no vernacular variant of any language could possibly embody a mistake."
As a matter of fact, the author specifically and painstakingly drew a distinction between what he considers simply mistake-based variants of languages (which do exist) and new language dialects with different, but well-specified grammars (of which he considers AAVE to be one).
The author drew a distinction between making a grammatical mistakes, and a vernacular dialect. He didn't claim that there were some vernacular dialects that embodied mistakes and others that didn't. I can't imagine any linguist who would.
What I'm saying is that all vernaculars have consistently observed (surely you don't literally mean well-specified) grammatical rules. Their speakers couldn't understand one another otherwise. So if your definition of mistake = inconsistency, then (by definition) you'll never find a vernacular that embodies mistakes.
A linguist would argue that it's a mistake only if an adult speaker learned the language after the critical period and didn't manage to grasp all of the rules.
If a child (or someone who grew up in the language environment) made the "mistake" then it's an example of language change in action... unless the child is so young that it's simply a language learning artifact such as "I goed to the store today, daddy" (which, incidentally, is logically consistent with other past-tense forms yet ungrammatical b/c of social norms)...
The author drew a distinction between making a grammatical mistakes, and a vernacular dialect. He didn't claim that there were some vernacular dialects that embodied mistakes and others that didn't. I can't imagine any linguist who would.
Fair enough. When you said "vernacular variants" (as opposed to dialects) it sounded as if you were including all vernacular variants including simple grammatical mistakes, sporadic use of slang, etc.
What I'm saying is that all vernaculars have consistently observed (surely you don't literally mean well-specified) grammatical rules.
I agree that all vernacular dialects have consistent rules. And you're right, I did mean consistently observed, not well-specified.
This is the very epitome of attacking a strawman. The article didn't come close to arguing that "no vernacular variant of any language could possibly embody a mistake."
As a matter of fact, the author specifically and painstakingly drew a distinction between what he considers simply mistake-based variants of languages (which do exist) and new language dialects with different, but well-specified grammars (of which he considers AAVE to be one).