A Linguistic Prizefight


Why Files

Learning language

Inventing writing

Birds' mating call

Dog talk

 If your language only has words for "one," "two" and "many," how would that change your view of the world?



Much to the relief of forest animals, that's Peter Gordon holding the bow.

Subjects saw a line of batteries, and were asked whether the second line had the same amount of batteries.

Graph data from "Numerical Cognition..." (see bottom of page)

F: When the original set of objects was shown for only eight seconds, performance plunged.

D: When matching a set of uneven lines, performance improved at higher numbers. Were the subjects "chunking" the objects into smaller groups that they could count?

Pirahã man with his catch, in dugout canoe.

Man sits at table with hands on surface.How do you think about things you can't talk about? Can you think about them at all? Linguists and anthropologists love to tussle over the issue. In this corner, weighing 238 pounds and wearing blue trunks, are the "culture-first" crowd, who argue that human culture determines language. In that corner, tipping the scales at 245 pounds and wearing purple trunks, are the Whorfians, who argue that language molds how we think.

Pirahã man in a test of his numerical abilities.

Benjamin Lee Whorf was an insurance man who became an influential linguist in the first half of the 20th century.

Now comes evidence for the strong Whorfian hypothesis, that we cannot think about some things because our language does not allow it. Imagine knowing three number words: "one," "two," and "many." That's roughly the situation with the Pirahã tribe in the Brazilian Amazon. Resistant to assimilation, the tribe's 200-odd members hunt, fish and occasionally farm in the soggy lowlands.

map shows Terra Indigena Piraha in North Central Brazil

Peter Gordon, who studies language development at Columbia University, became fascinated with the tribe several years ago. Gordon says that when traders reach the villages, the Pirahã often get ripped off. The reason? They seem to have a very faint feeling for numbers, because they don't have much to count.

Man with bow and two others standing in mud. As Gordon notes, "You need numbers for exact computation. We have shown that they can estimate numbers pretty well, and a lot can be done by estimation. How do you build a hut? You put up enough sticks until they hold a roof. Their life is not built around the necessity of using numbers."


Sure, this sounds logical, but the logic is disturbingly circular. They don't use numbers because they don't have numbers. They don't have numbers because they don't need numbers. And they don't need numbers because they don't use numbers.

It's a far cry from the industrialized, westernized existence. And after waking up at 7:15, hearing that it's 78 degrees outside, tuning the TV to channel 11, counting your pocket change, putting $25.75 worth of gas in the guzzler, driving 67 in a 55 zone on the freeway, and getting to work at 9:05, five minutes late, aren't you curious to get inside the head of people who devote three words to the numerical realm?

line graph going straight then fluctuating Gordon was. An associate professor of biobehavioral sciences, he studies how infants and kids learn language. He saw the Pirahã as a chance to confirm or refute Whorf.

Specifically, did the impoverished lingo for numbers affect how the Pirahã think and work in the numerical realm? Gordon teamed up with linguists Daniel Everett and Keren Everett, who had already been studying the Pirahã for 15 years. On his second visit to Pirahã villages, Gordon put some Pirahã men (women were reclusive) through some tests.

He couldn't ask the men to count, since they don't count. Instead, he asked them to match two sets of objects. If I put out four sticks, can you put out four sticks? That sort of thing.

For two or three objects, the men did pretty well. But their performance usually plunged at higher numbers, especially as the mental tasks got more complex.

line graph starts high, declines rapidlyline graph starts up high, declines steeply, then goes back up

Adding it up
So what does the research mean?

Aside from problems with traders, the paucity of number words seems to affect craft work. Like many tribal people, they make necklaces, Gordon says. "When you look at tribes that count, their necklaces tend to be symmetrical... but the Pirahã may put four objects on side and two on the other."

Gordon calls the experiment "an existence proof" for the strong Whorf hypothesis. "It's not a general claim about language and thought. ... There are lots of cases where language influences thought, where people might have a preference to construe something in a particular way."

Only rarely, he says, does a case arise where "you would not be able to get your mind around something because you do not have the words for it."

man with paddle and two dead fish (on bottom of boat) in carved dugout canoe.

And one case does not prove that the strong hypothesis governs every aspect of our mentality. "I think there is a huge amount of universality in the way we think about the world. I think language can tweak it at the edges, and the weak sense of Whorf [that language influences how we think], in some ways, has to be true."

-- David Tenenbaum

All photos this page courtesy Peter Gordon

Numerical Cognition Without Words: Evidence from Amazonia, Peter Gordon, Science Express, published online Aug. 19, 2004. 10.1126/science.1094492.


The Why FilesCredits | Feedback | Search