Evidence and argumentation: A reply to Everett (2009)

This discussion note is a brief response to Everett (2009, E09), which was a reply to our assessment (Nevins, Pesetsky, & Rodrigues 2009, NP&R) of Everett's (2005, CA) earlier claims about the Amazonian language Pirahã. An important (and somewhat hidden) feature of E09 is a set of new empirical assertions presented in defense of CA's hypothesis that Pirahã lacks embedding. We argue that these new claims have not been supported by appropriate evidence; but if they are correct nonetheless, they WEAKEN rather than strengthen the case against embedding. We conclude with a discussion of comparable questions of argumentation and evidence that arise elsewhere in E09, and the relevance of these issues to the public discussion of CA's claims.

1. Goals Of This Reply

One of our reasons for undertaking the project reported in NP&R was the extraordinary attention that Everett's claims about Pirahã have received in the popular press—and the equally extraordinary significance that has been attributed to them. The New Scientist (March 18, 2006) suggested, for example, that Pirahã might constitute 'the final nail in the coffin for Noam Chomsky's hugely influential theory of universal grammar'; and the Chicago Tribune (June 10, 2007), under the headline 'Shaking language to the core', reported that Everett had 'fired a volley straight at the theory when he reported that the Brazilian tribe he was studying didn't use recursives [sic]'. More recently, the Times of London (October 24, 2008) has characterized Everett's claim that 'Pirahã lack the grammatical principle of recursion' as an 'astonishing find'.

If the conclusions in NP&R are correct, of course, Pirahã presents us with no nail, no coffin, no volley, and no astonishing find. Much of NP&R was devoted to an examination of the claims and conclusions about Pirahã language and culture that were put forward in CA, the article that triggered the publicity.¹ We argued that many of CA's claims were not coherent as stated; that the logic by which factual evidence was said to support them was flawed; and also that CA failed to discuss a wealth of published counterevidence (indeed, failed to acknowledge even the existence of this counter evidence). Finally, we argued, the linguistic evidence presented in CA not only fails to support claims of Pirahã exceptionality, but actually suggests THE OPPOSITE: that Pirahã fits straightforwardly into the known typology of human languages. In the field of linguistics, as in other fields, evidence does matter—as does the logic of the argumentation that links evidence to hypothesis. It is precisely in this domain that linguists can contribute to the Pirahã discussion in ways that reporters cannot. Since very few of the press reports on Pirahã had even raised the key questions of evidence and argumentation, we considered it important to try to put these issues on the agenda at large, a task rendered particularly timely by the continuing public interest.

Similar considerations prompt us to offer this brief response to Everett's reply (E09) that accompanied the publication of NP&R. At the same time, we do not believe that Language readers would welcome a point-by-point discussion of all the claims and counterclaims taken up in E09. For example, although E09 mischaracterizes many arguments and claims from NP&R, these errors can be easily found by simply comparing E09's citations of NP&R with what we actually wrote, so we see little reason to correct them here.² We also believe that the most effective replies to E09's comments on such issues as the proper meaning of 'universal grammar' can already be found in the sections of NP&R to which E09's comments respond—so once again, we do not rehearse these issues a second time.

Instead, we focus in this commentary on one aspect of E09 that is less obvious, but of some significance nonetheless, precisely because of its direct bearing on questions of evidence and argumentation. Everett represents E09 as demonstrating that CA 'is essentially correct in its description of the facts and that it corrects errors...