this post is one of hopefully many that will track my progress in writing my Senior Thesis for my degree in Philosophy at Vassar College.
Steven Pinker is the author of the bestselling survey-of-the-field The Language Instinct. Like other “popular science” books, Pinker’s is a book that is light on raw data but attempts to compensate for this in easily-digested arguments and succinct evaluations of the various different theories and experiments that amount to the ‘state of the science’.
On pages 56-59, Pinker attacks probably the most strident philosophical objection to a cognitive-science perspective on language, namely that language determines thought, or language determinism. As Pinker would have it, language determinism (as typified by the Sapir-Whorf Hypothesis) is the belief that thought is actually shaped and molded by the language of the thinker. Pinker argues that this implies that thought is dependent on language, rather than vice versa. He proceeds from this point onwards via an argument against absurdity: how can this possibly be true if 1) we invent new words for new thoughts, and 2) we oftentimes say what we do not seem to mean?
The first conclusion is a little more difficult than the second, so I’ll address it later. The idea that we might sometimes say what we do not mean is a classical case of what Pinker himself calls a “conventional absurdity.” By latching onto a commonly-held emotion or memory, Pinker considers a problem solved. The implications borne out by his conclusion, however, are ignored. If there are thoughts that cannot be articulated into words, what are they? Well, you can’t say, because they cannot be articulated. So then why does Pinker talk about the thoughts as if they are meaningful? Here’s the crucial point: meaningful thought must be able to be articulated into words. This does not make thought dependent upon language; it makes the meaning of thought contingent upon language being possible.
Pinker, a scholar no doubt well-trained in the field of science, is probably just unused to picking out the “berries” as he strolls through the woods, to borrow one of Wittgenstein’s metaphors. There is a lot more that can be said about this first objection, but we’ll move on to the second.
Pinker seems to raise a valuable point: how can thought be dependent or determined by language if new words are invented? This is to say: if we were struck at one point with the inability of our current language to capture the notions of a thought we had, a mind determined by language would be unable to improvise. But here is where Pinker makes his biggest mistake. First, he argues that a language determinist sees language as the same as thought (=identity), and then argues that thought is dependent upon thought. However, two objects cannot have the relation of identity and also of dependence. They can, however, have the relation of co-dependency.
Pinker’s argument fails because it puts forward a philosophical argument (language=thought) and then disproves an entirely different one, namely language>thought. It is ridiculous to presume that language as an entity possesses some sort of superior relationship to thought, as another metaphysical entity. However it is equally if not more ridiculous to see thought as wholly independent of (and therefore having no relation to) language.