Edward Gibson, a professor at MIT, mainly studies memory and how it affects language. He recently returned from studying the PirahÃ£ and examining other more controversial issues surrounding the tribe.
The linguistic features of the PirahÃ£ people, an Amazonian tribe of hunter-gatherers, have been the subject of heated controversy among academics. Professor Edward Gibson presented his recent field work and research with the PirahÃ£, and his own thoughts about the debate.
At the heart of the issue is Daniel Everett, a linguist who first translated and analyzed the PirahÃ£ language. He has made strong claims that the PirahÃ£ lack sentence recursion, syntactic embedding, color terms, numbers, and relative tenses.
“Theoretical claims on why this is,” Gibson explained, “is something called the ‘Immediacy of Experience’ principle, which says that the PirahÃ£ culture constrains conversation to non-abstract subjects which fall in the immediate experience of the interlocutor.”
Gibson and many other linguists were skeptical of this blanketing theory—linguists, in particular, about the claims that refuted renowned linguist Noam Chomsky’s theory that all human languages have recursion.
“All we mean by recursion here is some kind of embedding being able to repeat the same category over and over again…so sentences can embed one within the other,” Gibson said.
He gave an example: “The Red Sox beat the Yankees. Mary said that the Red Sox beat the Yankees. John thinks that Mary said the Red Sox beat the Yankees.” As the later sentences show, by using additional verb phrases, sentences can be infinitely nested within each other.
The alternative in PirahÃ£ would use discrete units which infer relation—“Mary said something. The Red Sox beat the Yankees.”
Gibson made sure to note that this lack of recursion does not preclude their ability to conceive of recursive relations—clearly the PirahÃ£ alternative can mean the same thing as the English sentence above—they simply do not say it the same way as other languages do. The difference is not semantic, but syntactic.
Despite the great interest in proving or disproving Everett’s claim with more data, Gibson did not reach any conclusive results. His tests, which involved showing adults four pictures and asking how they were related to one another—hoping for a description that the last picture was “Ko’oi’s spouse’s parent’s dog”—yielded a lack of recursion.
“So we didn’t falsify the recursion claim, but we didn’t provide evidence for the lack of recursion, either” Gibson said. The provided test could have been cognitively taxing, involving extensive use of short-term memory for a task the PirahÃ£ were probably uninterested in.
This is one of the difficulties of testing the PirahÃ£: Gibson, from his interactions with the people and discussions with Everett, thinks that “they are not very interested in outside cultures.” For example, Everett “did translate sections of the Bible, which they found entertaining. They never found it very compelling.”
His conclusive result from his trip to the Amazon in the winter of 2007, however, was with regards to PirahÃ£ numbers. Gibson and his colleagues performed several simple cognitive tests on the PirahÃ£ and found consistent results: they only had three words for number which seem to signify few, some, and many.
In an “orthogonal matching task,” for example, they placed spools of thread in a vertical line and asked participants to lay down, horizontally, as many balloons as there were spools of thread. The orthogonality made it difficult to perform the task without number words.
“The response is exactly the same in all the trials,” Gibson said. Once the number of spools reached about 10, accuracy fell off drastically. Some would overestimate balloons up to 12, others would underestimate from about 7. All of their other tasks supported the idea that though they were able to estimate quantities well, their lack of number words hindered their ability to succeed in the more complex tasks.
Previous studies done by Frank Gordon showed that the PirahÃ£ could not perform simple matching tests that did not require number words—putting down one line for the other, for example. Their hypothesis was that the lack of number words completely precluded their ability to perform such tasks—that their language strongly affected their cognition.
Gibson performed the same matching tests and found the PirahÃ£ “did perfectly fine.”
“What we think is going on here is that number is a cognitive technology and is an example of a way to encode and maintain relative distinctions over time,” Gibson concluded. He hypothesized that perhaps the ability to count exact numbers is not needed in a hunter-gatherer society, and estimates of quantity suffice instead.
In addition to their recursion and number tests, they performed other cognitive tests on visual perception and short term memory. These too did not have clear-cut results or implications for Everett’s controversial Immediacy of Experience principle.
In the end, though, Gibson aligns himself with Nevins, Pessetsky, and Rodrigues in their 2007 critique of Everett’s observations.
“Some of Piraha’s supposed ‘inexplicable gaps’ (both linguistic and cultural) will be argued to be illusory or non-existent,” they wrote. In particular, Gibson is highly skeptical of Everett’s claims about recursion.
However, he finds the entire endeavor worthwhile and fascinating.
“Even if it is extremely controversial, I think it is very useful to raise these interesting research issues…[recursion] has become quite an interesting research issue because of this debate. Before it had just been assumed and asserted that all languages are the same,” he said.
“It’s also very interesting to investigate the role of culture in that…also the investigations generally of the relationships between culture and the cognition of language.”
Due to the limitations of the College’s dashboard, we have intentionally excluded the tilde from atop the h in “Piraha”.
Did you like this article? Consider joining the DG! Open staff meetings are every Monday at 6:30 p.m. in Kohlberg; or email us at firstname.lastname@example.org.