Napoleon and the tsar meet at Tilsit to agree on their spheres of interest. (But Tolstoy still got a story to tell)
Talk about a marriage of irreconcilables! PLOS Biology has published an article titled "How Could Language Have Evolved?" and like all PLOS papers it is available online to anybody interested. The topic is absolutely the question I've been asking myself for the past 45 years so it should be up my alley. Regrettably it doesn't seem to say anything that hasn't already been discussed in this blog. Nevertheless, it has a distinguished set of authors who are famous for their contrasting perspectives. I don't suppose it is quite as astonishing an alliance as Churchill and Stalin sitting down as allies, but it is surprising.
Johan Bolhuis and his work has been mentioned a few times on this blog (here and here). Most notably he has compared the neurobiology of bird song and speech, leading this blog to report: The conclusion that we have little positive data about speech to gain from studying chimpanzees, but much to be learned from birds now seems settled. Language in his work is mainly concerned with the biology that supports communication.
Ian Tattersall is an anthropologist and curator at the American Museum of Natural History. I haven't had much to say about his work of late, but I did cover him a couple of times in the blog's early days. Most notably I discussed his article in Scientific American in which he claimed that language was the result of invention and insight. Some great insight led to the use of symbols and changed the whole history of the human lineage. Language in his work is mainly concerned with the cultural role of symbols.
Robert Berwick is at MIT, which is Chomsky territory indeed. At the same time, Berwick is his own man and in 2010 I reported that he has developed a list of core universals of language that differs from the single trait listed in the famous Hauser-Chomsky-Fitch paper. He shares Chomsky's primary focus.
So what do these four agree on? It turns out that their paper is something more familiar to the history of diplomacy than science. It breaks no new ground, but offers a ringing statement of mutually agreeable propositions, each to be interpreted by its signers in a slightly different way.
The paper opens with "It is uncontroversial that language has evolved, just like any other trait of living organisms." Of course there are people who do [or at least did] contest the evolutionary proposition. Tattersall has said that the systems that support language evolved for other purposes. Chomsky and Berwick have doubted that considering language from an evolutionary perspective can be at all helpful. The one who is most strongly associated with this sentence is Bolhuis. Meanwhile the closing sentence--"we argue that the basic principle that underlies language's hierarchical syntactic structure is consistent with a relatively recent evolutionary emergence."—is what we expect from the other three authors.
The paper's first two sections are on language itself and embrace the basic Chomsky/Berwick proposition, "In our view, for the purposes of scientific understanding, language should be understood as a particular computational cognitive system, implemented neutrally." The final bit about neural implementation makes Bolhuis's position clear. I might have expected trouble from Tattersall, thinking he might share my doubts about language being a computational system, if by computation you mean (a) a complete system with rules and definitions built in, or (2) a procedure that leads to a single solution. But Tattersall agreed to the definition.
The authors continue, "Externalized language may be used for communication, but that particular function is largely irrelevant in this context." What context is that? They have given us their definition of language, but it is so abstract that it is deprived of all context. Suppose I said that I had a piece of software that should be understood as a particular computational input/output system, implemented on the Android operating system. I have just told you something, but something stripped of functional context. If I added, "Thus, there is nothing useful about the app's origin to be discovered by considering my need for a magnifying lens." You might say, "Whaa?" Since I've told you nothing about the app's function, you have no idea what is or is not relevant to the origin. Yet, this kind of non-sequitur appears in the paper. The authors continue, "Thus, the origin of the language faculty does not generally seem to be informed by considerations of the evolution of communication." I wonder why both Bolhuis and Tattersall let that go by without at least a little more clarity. Presumably Chomsky and Berwick wouldn't come on board without it.
They also spend time analyzing the merge procedure, which is the most recent of Chomsky's various generative procedures. For purposes of this post I need only say that it is crucial to Chomsky and Berwick and pretty much irrelevant to Bolhuis and Tattersall.
Bolhuis seems most firmly represented in the paper's third section, "The Nature of Evolution." The first paragraph says, "Questions of evolution or function are fundamentally different from those relating to mechanism, so evolution can never 'explain' mechanisms." I have no quarrel here. The merge process at the center of Chomsky's theorizing, of course, is a mechanism. So we are immediately going to concede that there is no evolutionary explanation of merge. This may be what Bolhuis wanted. He can study what he wants to study without alarming Chomsky's crowd.
The next section, Paleoanthropology, is Tattersall's baby. It states the position, frequently criticized on this blog, that language began when non-linguistic symbols first began to appear in the archaeological record.
The final paragraph before the conclusion is where the four contributors come together. It begins, "The Strong Minimalist Thesis (SMT), as discussed above, greatly eases the explanatory burden for evolutionary analysis, since virtually all of the antecedent "machinery" for language is presumed to have been present long before the human species appeared." The first part of the sentence is Chomsky/Berwick's claim and the last part (beginning with since) is Tattersall's, but I wondered at first why Bolhuis agreed. His work on the neurobiology of learning to speak certainly seems relevant to language. Then my penny dropped. According to Chomsky, language precedes any social use of it, thus the stuff Bolhuis studies could have evolved after merge. That process does not allow much time to evolve the complex changes in the brain needed to produce speech, but evolution can happen very quickly when there is strong enough selective pressure.
The resultant paper is something like a peace treaty that declares spheres of interest. Each party has staked out a well-established position and agreed not to challenge the others. On the whole, it looks to me like Chomsky is the biggest winner, as the work of people like Bolhuis was making his position increasingly untenable. I don't know why Bolhuis felt he should go along when his work can fit in many scenarios, but Chomsky can be quite the charmer.