August 23, 2004

Pirahá and the art of mathematics

Via Language Log and Crooked Timber, I take note of a Reuters dispatch over here and a brief in the Science section of the WaPo here.

Having just read the relevant paper from here, which makes even more remarkable claims for the Pirahá tongue. I am with Mark Liberman in disregarding the vulgar Whorfian conclusion, which I note that Dan Everett, the author of the paper cited above, also disregards.

In a series of videotaped psychological experiments, Gordon and K. Everett (ca. 1993) collect data to investigate the claim that the very concept of counting is foreign to the Pirahás. Gordon (2003) develops this theme in more detail, with impressive statistical interpretation of this and additional experimental results. Although I disagree with his assertion that Pirahá has a one-two-many system of counting (there are no such numbers, as has been seen above) and although I find his conclusion that the Pirahá facts offer support for Whorfianism unconvincing, I nevertheless agree completely with the principal conclusion he draws from his experimental results, namely, that Pirahá people neither count nor understand the concept of counting. I want to briefly review his results here.

Everett goes on to summarise Gordon's findings and a number of other claims about Pirahá. They are quite remarkable. Nothing in the way of a perfect tense, no numbers beyond a many/few distinction, no colours (and a pointer to what is wrong with the whole "universals of colour naming" thesis - namely that if you look for colour words, you'll find some whether they really are words for colours or not), a pronoun system that they appear to have borrowed from an adjacent tribe, and a lack of embedded clauses that shreds any claims about the universal recursivity of natural language. They also lack any strong conception of historical experience and have no founding myths or creation stories. They appear to lack any curiosity about the outside world and speak no Portuguese despite two centuries of contact with Brazilian culture.

Pretty hefty stuff.

What Everett is claiming, however, is the opposite of vulgar Whorfism, and exactly what I said over here:

The idea that language reflects the thoughts of speakers is very important. Speakers' categories of thought reflect the way they interact with the world and how you interact with the world may vary a lot depending on who you are and what context you live in. Language is not an arbitrary protocol, it is the concrete manifestation of the worldview of the culture that supports it. It is an expression of the reality of its speakers' lives.

Everett claims that the material simplicity of Piraná life, its lack of complex kinship relations, narratives or larger worldview, is the cause of its lack of many elements we take for granted as present in all languages. Without any need to express counts, numbers are missing. Without any stories to tell, there is no perfective tense. The reality of Piraná life is missing these elements and thus so is their language.

Thus, it should hardly be a suprise - and even less grounds for universal claims about the acquisition of numbers - when children prove much better at learning to count when introduced to the concept than elders do.

Now, I am not an expert on this tribe and can only summarise the claims made by others. It's possible that Everett's account is fundamentally flawed. But still, this seems to me to be a better explanation than Whorifism, and it certainly poses challenges for a universalist culture-free conception of language.

Posted 2004/08/23 10:59 (Mon) | TrackBack

I have started a series of notes on the issue in my blog. Hope it may help readers.

Posted by: Tony Marmo at September 2, 2004 21:18
Post a comment

Remember personal info?