Derek Bickerton is one of the most striking critics of the search for language origins. He does not mince words when he thinks you’re wrong. Now it is time for him to look in the other direction and tell us what is right. He sums up the thesis of his new book, Adam’s Tongue in its subtitle: How Humans Made Language, How Language Made Humans. In other words, humans and language made each other, or to use a fancier term, they co-evolved. Surprised? Neither am I. It turns out that Bickerton is a more exciting critic than proponent.
It was a little over a year ago that this blog reported after the Barcelona conference that linguistic co-evolution was the new orthodoxy (see: Co-Evolution Idea Won Big in Barcelona), but Bickerton is pushing the idea over a broader range than usual. The focus at Barcelona concerned the co-evolution of grammatical speech, and while Bickerton takes that point seriously his position is more general: language and humanity co-evolved by shaping the peculiar human niche, culture.
The book’s strongest parts are in its complaints, summarized in the introduction (so don’t skip that chapter). Bickerton fiercely rejects the notion—which he calls primate-centric—that there is no serious discontinuity between humans and other apes. The rival position in favor of continuity is found in Christine Kenneally’s The First Word and I am on Bickerton’s side in this quarrel, for mostly the same reason. All humans do things that can only be understood in terms of a person’s culture. Apes have differences that can be called ‘cultural,’ in the philistine sense that they are learned rather than genetic, but they cannot be called ‘cultural’ in the more profound sense. That is to say, ape cultural behavior can be explained in terms of the natural environment. One group of chimpanzees may make tools from stones and another makes them from sticks. It is a difference, to be sure, but in both cases the point of the tool is to obtain food, a purely natural function. Meanwhile, humans make artifacts like necklaces or paint pots or prayer beads whose functions promote nothing natural and serve a need that can only be explained in terms of other elements of the culture. Chimpanzee culture is an adaptation to nature; human culture is an adaptation to culture itself. If we must call what chimpanzees do "culture," then lets call what humans do "culture squared."
Bickerton also rejects what he calls homo-centric bias. That’s the belief that humans constitute the apex of a pyramid of communication in which only humans have achieved the top rung. This approach looks for precursors of speech in other species, and is thus related to the primate-centric search for continuity. Bickerton rejects this approach more by fiat than by evidence, and I think it hurts him. If he looked, for example, at the precursors of human communication with the close study that somebody like Michael Tomasello has brought to the work, the nature of the break between chimpanzee society and human communities would be much clearer. (None of Tomasello's research is in the bibliography.)
Third, Bickerton explicitly rejects the notion that language has been the target of natural selection. It is a disastrous mistake and contributes to the book’s breezy, superficial tone. You can get a taste of the price this rejection charges by looking at Bickerton’s brief and cavalier dismissal of the FOXP2 gene. He insists FOXP2 is too remote from language and too involved in other tasks to be a language gene. And plainly it is not the language gene, but he cannot simply wave away the fact that people with mutated FOXP2 genes have serious language problems. Over a year ago this blog reported research led by Constance Schariff establishing the importance of the mutated FOXP2 gene in enabling young zebra finches to learn their group’s song. (See: Birds Also Use FoxP2) True, such observations can only be suggestive about humans, but I thought it was a most exciting finding and one that should only encourage investigators like Bickerton because of the way it straddles the great nature-nuture conundrum. The gene does not provide the bird with its song, only the capacity to learn it through imitation, so even if we imagine a perfectly matching relationship between humans and birds we would see that language itself is still entirely a cultural product while the capacity to learn to speak is natural and genetically supported.
Bickerton’s rejection of the relevance of natural selection to language origins does not mean he rejects evolution. In the end he does not even reject natural selection so much as he rejects using it as a tool for thinking about how something might happen. Part of this stance seems to be a reaction against the extreme, genetic-level only approach that has dominated evolutionary thinking for the past 40 years and has probably done as much as the Piltdown Forgery to slow down inquiry into human evolution.
Every inquiry into the mechanisms of language origins has had to deal with this road bump. One approach is to embrace selfish-gene theory, and Jean Louis Dessalles probably took the orthodoxy as far as it can go. Another approach is to look for a way around the barrier, as Terrence Deacon did when he promoted the Baldwin Effect. Still another approach is to reject the legitimacy of the barricade altogether in favor of a multi-level selection. Michael Tomasello and this blog have both rejected the selfish gene in favor of the rise of a cooperative species. Bickerton has another workaround, “niche construction theory.”. It is like Deacon’s effort to outflank the problem, but lacks the rigor that Deacon seeks. (I’m not saying niche construction theory isn’t rigorous, only that Bickerton’s use of it is not rigorous.)
The heart of niche construction, as summarized by Bickerton, is that
animals themselves modify the environments they live in, and then these modified environments, in turn, select for further genetic variation in the animal. So a feedback process begins…. [p. 99]
I have no quarrel with this point, and confess to being a bit puzzled at the suggestion that it is not perfectly obvious. I believe I have taken it for granted from the start of this blog that speech has produced changes in the human environment which in turn forced changes in people. Some examples previously cited on this blog include the adaptation of the human ear to hear and distinguish speech sounds, the adaptation of the brain to handle more complex and rapid utterances, the rise of imaginative speech, the rise of metaphorical speech. In short, I am not bowled over by the book’s thesis that the evolution of speech proceeded through a feedback process in which a little bit of language changed the human lineage so that it could handle increasingly more language. What does astonish me is the idea that this process can be understood without thinking hard about the natural selection processes going on.
Bickerton's most important previous contribution to the study of language origins has been his proposal of a protolanguage, and the book's strongest point is its defense of that idea against Noam Chomsky who denies that protolanguage is even possible. Bickerton uses his familiarity with pidgins to great advantage and I will discuss protolanguage in next week's post.
Most disappointing the book's lack of new insights into either the nature of language or what it means to be a human. In short, I’m not very startled by this book, and that’s a great letdown because I have been looking forward to its release. Bickerton is a lively iconoclast who brings zest to his style and he moves right along. His books is fifty times the breezier read than Michael Tomasello’s Origins of Human Communication, but Tomasello’s book is fifty times more important.
At the book’s outset Bickerton makes an amazingly cocky statement:
Fortunately for the publisher, it doesn’t seem to be a money-back guarantee.
I was equally underwhelmed by Bickerton's book. I really can't understand how he can claim that the communicative capacities of other great apes are an insufficient platform for language evolution without mentioning any of the important work done by Tomasello and others.
I was also quite baffled by his handwaving confounding of language and (compositional) thought. This even more so given that he mentions Penn et al.'s Relational Reinterpretation-Paper int the Behavioral and Brain Sciences but fails to adress any of their arguments or any of the arguments for a Language of Thought, for that matter.
Posted by: Michael | April 20, 2009 at 08:03 AM
"Third, Bickerton explicitly rejects the notion that language has been the target of natural selection."
Note 1:
Bickerton merely says that it is not possible to find a convincing answer to the question "Language evolved, so language had to be selected for. Well then, what selected for it?"
The problem is that language is a very highly developed feature and before you can have language you must first have more primitive components that eventually combine to make language. Rather than asking about what selected for the full-blown language that we use every day (and whose development left few, if any, traces in the fossil record), start at an earlier time and instead ask the question "What might be the precursors of language? What changes could have been selected for that produced the precursors of language in our ancestors?"
Note 2:
If the FOXP2 gene is responsible for language then where is my talking finch, gorilla, and ferret? Would perhaps language somehow not provide an advantage to those animals as it does to us humans?
Posted by: tndal | June 23, 2009 at 10:47 AM