Yesterday afternoon, I wrote a brief message linking to a video of a lengthy lecture given by Noam Chomsky in Paris four years ago. In this extended post, I am going to lay out what I understood to be the core of Noam’s remarks. Why on earth am I doing this? The answer is this: What I care most about in the world is deep, clear ideas, elegantly expounded so that one can see and appreciate their power, simplicity, and beauty. Chomsky’s life work in the field of Linguistics has all of these qualities. As I listened to him, I could see, through his words, the power, the elegant simplicity, of his theories, and so I want to try to share with you what I heard.
Now, let us be clear. I know next to nothing about Linguistics. I am a complete novice on the subject. I may very well get something wrong, and I will surely fail to capture the complexity of Noam’s thought. But it is beautiful, in much the same way that the central argument of the Critique of Pure Reason is beautiful, in the way that great mathematics is beautiful.
Chris writes, “In 2003 Charlie Rose asked Chomsky "if this was your last day on earth, would you like what is mentioned about you to focus on your political works or linguistic contributions". Chomsky, shrugging and laughing said "to tell you the truth I honestly don't care". I can easily believe that Noam said that, and indeed meant it. This post is in no way an attempt to prioritize his Linguistic theories over his political commentary. It is simply an appreciation, an homage, to the beauty of his ideas.
Chomsky’s theories overturned several widely held assumptions about the origins of language, about language acquisition, and about the purpose of language. As I see it, his analyses rest on three very simple but powerful observations. First, grammatical sentences are potentially infinite, or at least unbounded, in length. There is no limit to how long a sentence can be. [At a minimum, one can, by simple concatenation, extend a sentence indefinitely by adding “and” followed by another phrase.] It follows from this that the ability to form infinitely long grammatical sentences could not have developed little by little, through extension of already existing well-formed sentences. Thus, it could not be that first human beings acquired the ability to form simple sentences. Then, through experience or genetic mutation, they acquired the ability to form somewhat longer sentences. And so forth until finally they had mastered the ability to form sentences of unlimited length. It follows from this, in contradiction to very widely held views, that human language ability is not a progressive enlargement of animal communication capacities.
Think about that for a moment. We are all familiar with the ability of animals to communicate: bees doing their tail-wagging “dance” to report on the location of pollen, whales and elephants “talking” over long distances at low frequencies, hunting packs of dogs or prides of lions communicating as they quite intelligently pursue prey [as I have seen them do on safari]. The paleontological record shows that early hominids had the ability to make simple stone tools as long as a million years ago, an ability obviously transmitted from generation to generation by some sort of communication. What more natural than to see human language as a slow evolution from these behavioral skills? But Chomsky’s point, made several times in the lecture, is that this must be wrong, because no extension or evolution of these animal behaviors could lead to the capacity for infinitely long strings of symbols. All the available archaeological evidence suggests that language was actually created, invented, or developed no more than 100,000 years ago, probably in pretty much its present form.
Chomsky’s second observation, famously deployed in his classic critique of the behavioral theory of language acquisition advanced by B. F. Skinner, is that it is impossible to explain a child’s language acquisition purely as a response to external stimuli. This is true for three reasons. First, if we take seriously Skinner’s notion of stimulus and response, there are simply not enough stimuli [in the form of uttered speech] in the life experience of a little baby to account for the acquisition of language at the age of one or two. Second, no amount of stimulus in the form of speech in the vicinity of the baby can explain the child’s eventual development of the ability to form new sentences that he or she has not heard before, and perhaps could not have heard before. Third, the actual sensory environment of the child is what Chomsky, quoting William James, calls in the lecture a “buzzing, blooming confusion,” and it is simply impossible on Skinnerian terms to explain how babies unerringly pick out of that auditory chaos the instances of language whose presence Skinner supposed serve as the stimuli in a stimulus/response behavioral event.
Chomsky’s third observation – with which I was not already familiar and which struck me as extraordinarily powerful – is that if we take from the theory of evolution the basic insight that the human capacity for language must be grounded in some genetic mutation, then it is obvious that this mutation occurred in the genome of a single individual, who was thereby equipped with the capacity for language acquisition and use. But in order for this mutation to survive, it must have conferred some competitive advantage to the individual. And since he or she would be the only human being in the world with the capacity for language, the fundamental adaptive advantage of the mutation must have derived from the new ability to think, NOT from an improved ability to communicate!! BECAUSE WITH WHOM WOULD HE OR SHE COMMUNICATE?
Chomsky now assumes [taking his guidance from Galileo, rather elegantly] that the mutation giving rise to the capacity for language must have been very simple. He suggests in the lecture that the most elementary, primitive innovation conferred by the mutation was the ability to take two elements and form from them a simple unordered set. He names this operation “merge,” and with considerable formal flair, he proceeds to show how the operation of merge can, recursively, give rise to sentences of any desired length and syntactic complexity.
Well, there is vastly more in the lecture, which in turn was only a cursory overview of a lifework. But perhaps this is enough to indicate something of its elegance and beauty.
Noam is a classy dude.