Culture, Not Biology, Shapes Language
There's no language gene.
There's no innate language organ or module in the human brain dedicated to the production of grammatical language.
There are no meaningful human universals when it comes to how people construct sentences to communicate with each other. Across the languages of the world (estimated to number 6,000-8,000), nouns, verbs, and objects are arranged in sentences in different ways as people express their thoughts. The powerful force behind this variability is culture.
So goes the argument in Language: The Cultural Tool, the new book I'm reading by Daniel Everett. Next week, I'll have more to say about the book itself; this week, I want to explore how Everett's years of living among the Pirahã Indians of Amazonian Brazil helped shape his conclusions — and why those conclusions matter.
The Pirahã are hunter-gatherers who live along the Maici River in Brazil's Amazon region. They fish, gather manioc and hunt in the forest. As is true with any human society, Pirahã communities are socially complex.
Everett first showed up among the Pirahãs as a missionary associated with the Summer Institute for Linguistics (SIL), with the goal of converting the natives to Christianity by translating the Bible into the local language. He left many years later as an atheist, knowing that the Pirahãs "were not in the market for a new worldview."
In between, Everett found that the Pirahãs have no words for "please," "thank you," "you're welcome" or "I'm sorry." They have no color words, but instead deploy phrases such as "it is temporarily being immature" for green. They have a limited kinship term system, one that does not distinguish between parent and grandparent or brother and sister. And their sentences lack recursion. This means there are no embedded clauses, as in the English sentence "Bring me the fish that Mary caught."
In his previous book Don't Sleep, There Are Snakes, Everett explained that the Pirahã culture drives the Pirahã linguistic system. For example, the language lacks recursion because of what he calls the principle of immediate experience. The Pirahãs don't discuss events that are not experienced by the speaker or by someone alive during the speaker's lifetime. (They do believe in spirits, but to them, the spirits are real and thus directly experienced.)
The Pirahãs would not say "Bring me the fish that Mary caught." They would say "Bring me that fish. Mary caught that fish." Two independent sentences are needed in Pirahã rather than the single English recursive one.
In corresponding with me earlier this week, Everett explained this for my non-linguist's brain (and it's explained near the end of Language, too): The Pirahã language has what's called "an evidence requirement." Each utterance is marked, by means of a suffix on certain words, as how the information contained in that statement came to be known. Was it witnessed directly by the speaker, heard from a third party or deduced from available evidence? In our example, that the fish should be brought and that the fish was caught by Mary are statements that each must be marked on its own; the two cannot be combined recursively because each has its own evidence requirement.
If you're familiar with Noam Chomsky's theorizing on language, you by now will have intuited that Everett's data directly challenge it. I refer both to Chomsky's insistence on universal grammar as an inborn set of rules in every human brain that allows a child to learn grammar, and also to his more recent work with Marc Hauser and Tecumseh Fitch that recursion is the defining feature of all human languages.
Language is learned, on Everett's account. And the Pirahãs do just fine, he says, without recursion in their language. Everett couldn't be more forceful in claiming that Chomsky is wrong.
This is no polite academic disagreement. Everett told me, "Over the years, first because of my ties to SIL and then because of my 'traitorous' turn against Chomsky, I have been accused of racism, of mining uranium, of stealing the Pirahãs' teeth, of fathering children with Pirahã women, of 'stealing their language' and on and on and on."
These charges Everett finds — as do I, to the degree that I can judge such things — absurd. The racism charge is plainly baseless; in his books Everett portrays the Pirahãs as clever people. In Language he takes on this issue directly:
"People seem to worry that if we say a given language lacks grammatical devices that are found in other languages, then that this is tantamount to claiming that the speakers of one language are somehow inferior to the speakers of the other. But nothing could be further from the truth. ... Languages are tools that fit their cultural niche."
So why does all this matter? For one thing, it challenges the seductive, heavily biologized discourse that I've complained about before. Of course, as a biological anthropologist, I know that, in an important sense, culture is part of our biology, that the two shouldn't be split. But keep in mind, Everett is challenging a dominant discourse (Chomsky's); in that context, he's right to harp on culture.
For another thing, it shows how getting out into the field to live among speakers of human language makes a difference. This is no Chomsky-esque armchair theorizing; it's immersive anthropology at work. Another fine example of this approach comes from Nicholas Evans and Stephen Levinson in their article on the "myth" of human linguistic universals.
I'm finding this material fascinating. I'll be back next week with more.
You can keep up with more of what Barbara is thinking on Twitter.