Louis Pasteur (1822-1895)
Let’s begin with a ridiculous question: there are many possible grammars, most of which are too complicated for humans to speak, and yet all the thousands of grammars used by humans just happen to be grammars that we can learn to speak, and learn quite readily. How did that improbable truth come about?
Presumably everybody can see immediately that only learnable grammars will be used and passed down through the generations, so of course we all use learnable grammars. There is, however, a hidden assumption in this explanation. Grammar itself must have been selected to fit the human cognitive environment. It could not be an innate. If it were somehow imposed on humans from the outside, the odds that the grammar built into our brain would be usable would indeed be so small as to be miraculous.
I have begun this report so strangely, in response to a paper that has been posted on the internet by Hubert Haider and titled An anthropic principle in lieu of a “Universal Grammar”. The ‘anthropic principle’ of the title is a proposed answer to the question that there are many conceivable universes where the arbitrary values of fundamental constants in physics are different, and in most of these universes people are impossible. How did the improbable come about so that, of all the possible universes, we got one in which human life is also possible?
At this point it is tempting to soar off on a tangent about the anthropic principle and the nature of the universe, but frankly I consider Dr. Haider’s introduction of the anthropic principle a red herring. In his notes, Haider says he had worked out his evolutionary argument and then a friend suggested it related to some outré ideas among physicists. I think on close reading, the ideas are fundamentally different and so I’m skipping the physics tangent.
Haider’s basic idea builds on Michael Arbib’s notion of the language-ready brain and Terrence Deacon’s notion of the co-evolution of language and brain, only Haider comes at them from a different angle. Arbib’s language-ready brain is ready to generate speech according to rules while Haider sees the brain as ready to host language as a kind of parasite that adapts ever more precisely to the brain’s operations. Deacon’s co-evolution involves a brain adapting to language and vice versa, whereas Haider focuses almost exclusively on the language adapting to the brain part. The difference between the older notions and the newer one is ultimately as stark as the difference between Louis Pasteur and the naturalist doctors who preceded him.
The naturalist sees language as a thing generated by the body, and thus, as Chomsky has directly stated, is a kind of organ that grows and becomes part of us. Like any other bodily organ, it can be studied by itself and found to possess all kinds of properties and structures. Meanwhile, the Pasteurite sees language as a kind of germ that enters the body and takes over some part of it. The evolutionary history of naturalist language studies the evolution of the organ. The Pasteurite’s evolutionary history explores how the germ changes in order to survive and grow.
One might object that Pasteur’s germs are real and have an existence of their own, while language has no separate existence, but is that so? Language is truly shared by contemporaries and through generations. Language can be preserved in writing and even recovered, as the Mayan, Egyptian and Linear B languages were recovered. This argument may seem more confounding than persuasive. Darn it; we surely know that language has no separate existence apart from the speakers, listeners, and readers who use it. Talk of germs merely muddies the water.
Perhaps we can gain a little ground by specifying what kind of germ language is. It is a virus and like any virus it can only come to life when it enters a host. Once stirred, to life it goes about its own tasks and has an identity separate from its host. We are used to thinking of viruses as bad things, but that is a prejudice. I would not be surprised to learn that there are helpful viruses, and it seems certain that the future will bring doctors who can cure disease through the introduction of artificial viruses built for medical purposes.
So let’s say that—language is an artificial virus introduced by people for their own communicative needs, but once introduced it took on a life of its own and over many thousands of years has evolved quite a bit.
One benefit of this approach is that it immediately frees the theorist from having to find selective advantages for humans who introduce a change in language. What is the selective advantage for people who add adverbs to speech? There seems no explanation more convincing than a just-so story. But flip it around. From the virus's point of view and adverb may just be a mutation that adapts itself to the human cognitive system and survives. Just-so stories go out the window.
So the virus concept could actually be helpful in thinking about language origins and development. In future posts I will explore this idea to see how far it can take us before collapsing under the weight of its own metaphor.