The more I think about it, the more unhappy I am about the whole Pirahã counting discussion.
I'm not unhappy because people are getting Whorf wrong, though it's true that they are, as Kerim Friedman at Keywords explains.
I'm not unhappy because people are leaving Edward Sapir out of it, though they're doing that too. The Sapir-Whorf hypothesis is sort of like GNU-Linux systems. People often leave a crucial piece out of the name, and also out of their thinking. Sapir -- from whom Whorf got the idea in the first place -- balanced his discussion of the ways in which "the language habits of our community predispose certain choices of interpretation" with a claim about the "formal completeness of language". By this he meant that "a language is so constructed that no matter what any speaker of it may desire to communicate ... the language is prepared to do his work".
No, what bothers me about the Pirahã counting discussion is that I'm not convinced that language is relevant at all, in the sense of playing a causal role in the Pirahã's lack of counting ability.
Here's an analogy. Suppose that there's an isolated group -- call them the Nerdahã -- who just aren't interested in throwing things. They don't stone people, they don't have snowball fights, they don't play ball games, they don't skip flat rocks on the surface of ponds, they don't pitch pennies, nothing. There's no religious or moral prohibition against throwing, they just think it's boring and a bit stupid, when they bother to think about it at all, which is rarely.
As a result, Nerdahã kids don't ever practice throwing. Not for speed, not for accuracy, not for fancy effects like curving through the air, or coming down vertically through a hoop placed at a distance, or bouncing off in funny directions after landing. Therefore, when they grow up, they're hopelessly bad at throwing. In the parlance of pre-Title 9 America, they "throw like girls", an expression used not because women can't in principle throw very well indeed, but because girls traditionally didn't practice throwing skills.
Because of their complete lack of interest in throwing, the Nerdahã language is completely lacking in throwing vocabulary. They have no words for pitch, fling, chuck, toss, sidearm, slider, curveball, bouncepass, and so on. They have a verb [ˈpʊʃ] that they can use for propelling something a short distance through the air, as in [ˈpʊʃ.ɪɾˈo.vɚˈɦiɹ] "throw that to me", but the same verb can be used for any sort of ballistic transfer, including sliding over a flat surface, and even for propulsion in which the agent remains in contact with the item moved throughout its motion. In fact, this verb can even be used for trying to shift an object that remains immobile. And they have another verb [ˈsɛnd ] that can be used for propelling something a longer distance through the air, but the same verb can also be use for any transfer that is more indirect.
Now an American psycholinguist comes to visit the Nerdahã. Her field linguist guide, who has been working among the Nerdahã for decades, mentions their lack of ball sports and their lack of throwing vocabulary, and the psycholinguist realizes that this is a marvelous opportunity to evaluate not only the Sapir-Whorf hypothesis, but also the idea that throwing is a human instinct, an innate module which may even have played a key causal role in the evolution of the hominid line. So she tests the throwing skills of the Nerdahã.
Well, you can figure out the rest for yourself. At a target distance of one or two feet, the Nerdahã do okay. As the target gets further away, projectiles start flying off in random directions at variable but generally low rates of speed. "Sheesh", says the psycholinguist to herself, "Whorf was right! Language does determine cognitive capacity. These people have only two words for throwing, and as a result, they can't throw for spit!"
There's something wrong with this story, don't you agree?
Posted by Mark Liberman at August 22, 2004 12:12 PM