I think that Carnie makes a very poor case for the innateness being too specific to Language. I take some issue with the name Universal Grammar because I think it ends up painting a strange and inadequate picture of what goes on in our heads.
Consider the theory of "parameters" -- that we have four innately available options for word order in sentences, and that we select one of them when we learn a language. It very strongly seems to me that this is not exactly how we go about actually learning. I think we are able to rationally identify patterns and produce sentences according to them, but it is reason that is innate to us, not the patterns.
Carnie postulates the "logical problem of language acquisition", which is that we are able to learn languages even though we never have enough input to be sure that we have all the general facts. If there are infinite possibilities, it would seem that every decision would be paralyzing -- we need restrict the possible associations and make language learnable.
Here, he proposes Universal Grammar as a solution. I would say yes if that means a rational capacity to restrict possible associations, but no if that means that we have patterns in our minds such as subject-object-verb order. Making generalizations and abstractions with varying degrees of certainty is a hallmark of what we do as humans, and not only in the realm of language. We never have enough input in science to be completely certain, either, yet we are able to advance theories and build on existing ones satisfactorily.
With this in mind, what did you think of the hypothesis that parts of speech are acquired through experience, and understood through different distributions? I found it problematic that Carnie proposed that parts of speech are understood through different heuristic models (and probabilistic distributions), but that language itself was heavily influenced by something innate (and that Carnie did not robustly disprove the statistical probability model that opposed the UG one).
ReplyDeleteHey Antonio! Loving the title of this post, haha :D When reading Carnie, I also was uncomfortable with the quick 'argument' he presented in favor of Universal grammar in the first chapter. One thing to think about in favor of this idea of universal grammar, however, is how we divide up the world conceptually. The Structuralists such as Sassure argue that there is no inherent categorization present in the world. That is to say, this platonic idea of a 'tree' (for instance) as a discrete form is a conceptual distinction that is mirrored, and is perpetrated by language. We could theoretically divide up our world in any number of ways. The tree bleeds into the world, so to speak. To bring this back to UG, perhaps S-O-V ordering isn't innate to human brains. But why is the *idea* of a 'subject,' a 'verb,' and an 'object' present across all languages, regardless of whether they're related or not? Isn't there something to be said for the fact that we all categorize and divide up the world in similar ways? Our ability to communicate, understand, and interact with each other hinges upon our ability to conceptualize of the world in a similar way. We might not have the same words to describe a tree, or even the same way of placing a 'tree' in a sentence, but we are have a conceptual idea of a tree as a discrete entity and it's this basic paradigm that I think constitutes some form of universal grammar.
ReplyDelete