Verified by 
The ostrich syndrome
In the 1960s, Noam Chomsky, arguably the world’s most famous linguist, proposed that children acquire their mother tongue because they were born with an inbuilt advantage. Unlike other species, humans, us, have language. And as no one really knew, then, how it was that a child, by around the age of four, could in a relatively short span of years become a linguistic genius. Scientists remained puzzled. Chomsky’s brilliant idea was that they must be born that way: you and I, and every other cognitively normal person on the planet comes to acquire at least one language. And many acquire more than one: bi- and multilingualism is the norm, world-wide, beyond the anglo-centric shores of the English-speaking peoples. But how human infants acquired their mother tongue remained a mystery.
Chomsky argued for what he dubbed ‘Universal Grammar’. The idea was that humans are predisposed to learn language, any language, because they are born with an innately-prescribed set of grammatical principles, that takes much of the challenge out of language learning. English, French, Japanese, Swahili are all underpinned by the same biologically-constrained principles of grammar, Chomsky contends. But as the research has come in, this idea has been steadily challenged. And today, the thesis of Universal Grammar is on life support, clinging on by a mere thread.
Yet, an ardent group of disciples still believe in Chomsky, and his intellectual sleight of hand—language was there all along—which waves away, by magic, the hard problem of how to account for its appearance in the first place. But the science is now overwhelmingly stacked against Chomsky’s ideas. Yet the fervent, likened to a religious cult by objective commentators, still believe. And this, to my mind, is much like the ostrich; according to popular belief, it buries its head in the sand, in the face of the blindingly obvious.
My recent book, The Language Myth, explores the science, and debunks the myth of a ‘language instinct’. But Chomsky’s disciples, remain head-deep in sand. And they have attacked the book with an uncommon, and somewhat irrational ferocity. Rather than attacking the arguments, they, often, appear to have attacked the author, and have spent an incredible amount of time and effort nitpicking over relatively trivial aspects of the book’s presentation, focusing on alleged misuse of quotations and misunderstanding of proposals they seem to be unable to express with clarity. Surely, surely, reputable scientists should face up to the evidence; but they dare not; and one suspects, they have no response.
In a recent exchange in the linguistic journal Lingua, Professor David Adger, perhaps himself suffering from the ostrich syndrome, has been crusading against the alleged injustice of the portrayal of this language instinct in The Language Myth: see his rant, Mythical Myths. My colleague, Dr. Christina Behme, and I replied: Leaving the Myth Behind. And he replied to our reply: More Misrepresentation. Below, Dr. Behme and I given our verdict on Adger’s final reply.
One may wonder by now why we still would bother to respond. It matters for the following two reasons. First, the language myth is wrong, and yet through institutionalised retellings it is still peddled as established fact by the credulous: and there remain many of these, even today, in the Academy. And second, it brings into daylight, for some commentators, the intellectually dishonest practices of some who claim to be seeking ‘the truth’. And as such, it all constitutes a fascinating object lesson in the current paradigm shift taking place in the academic study of language, as I reported on in an earlier blog post: The Structure of Scientific Revolutions. The thesis of Universal Grammar has more than passed its sell by date. Like the dodo, that paradigmatic example of the extinct, Universal Grammar has entered that liminal space of the foregttable, albeit, at least for some, not forgotten. Universal Grammar has had its day in the sun. But seriously: the idea, once brilliant, just turns out to be wrong. In fact, it couldn't be more wrong. The science just isn't there: it's time to move on.
A final reply to David Adger’s ‘final response’, by Christina Behme and Vyvyan Evans
Adger (2015b) replied to Behme&Evans (2015) and this reply requires some clarifying response. But any author would be ill advised to descend to the same level of gratuitous personal attacks and double standards[1]. Hence, this addendum will only correct a few of Adger’s severe distortions. An independent paper (Evans & Behme, forthcoming) reiterates some of the most important points made in TLM, and argues that the kind of generativist paradigm as defended by Adger (2014) does not contribute to a modern science of language.
No one could reasonably deny that linguistic research requires collaboration between researchers from different theoretical frameworks. But it does not follow that any and all paradigms need to be included in such collaborative efforts. Evans (2014a) argued persuasively that the Chomskyan paradigm Adger subscribes to has outlived its usefulness and should be rejected for that reason. Adger’s misleadingly claims “that not everyone shares [his] perspective on theoretical pluralism, as Evans 2014a (The Language Myth, hence TLM) makes clear“. Yet Evans (2014a) only rules out one of the many current theoretical approaches. To date Adger has not addressed the bulk of the substantive challenges in Evans (2014a) nor has he provided an unambiguous formulation of the generativist commitment he subscribes to[2].
It is curious that Adger would comment that “paragraphs [in Behme&Evans (2015)] are packed with citations as though that counted as an argument” without providing specific examples of instances where citations were used as arguments and not as supporting evidence for arguments. He would have been hard pressed to find any. It is equally curious that Adger would deny that there was any ‘minimalist fury’, given that he contributed to blogposts in which Norbert Hornstein proposed it would be part of the public service of generativists to combat the junk produced by Evans (2014a,b), suggested the possibility that “Language hates 2/3 of the field” even called on fellow linguists to leave the LSA when Language would proceed with a planned book-review of TLM. It is equally notable Adger would accuse Behme & Evans (2015) of “insinuations about motivation” (p. 3) but remain silent when Hornstein (2015) implied not just gross misunderstanding but even malice and dishonesty: “… trying to figure out whether the gross misunderstanding is a product of ignorance or malice [is difficult]. This is a particularly difficult decision to make in the case of Vyvyan Evans’ oeuvre as the work is both deeply misinformed … and, apparently, very dishonest” (Hornstein, 2015). One can, of course, express intellectual disagreement in strong language. But the kind of quasi-religious rhetoric used by Hornstein and never objected to be Adger should have no place in academic discourse. Finally the inconsistencies in Adger 2015b are quite revealing. On the one hand he claims that “TLM and BE15 fail to maintain [basic standards of scholarship]” (p.3). On the other hand he accuses Behme & Evans (2015) of misrepresenting him as “criticizing the whole book [TLM]” (p. 3, fn1). Failing to maintain basic standards of scholarship is of course a severe criticism. Finally he claims “…one of the positive points about TLM is that it is fairly focussed on ideas” (p. 7). There is no need to speculate about motives, the facts speak for themselves: depending on Adger’s agenda the same book “fails to maintain basic standards of scholarship” or “is focussed on ideas” [a basic standard of scholarship].
Adger claims repeatedly that Evans (2014a) is misrepresenting and caricaturizing the generativist commitment. One would expect, such allegations to be carefully substantiated and the generativist commitment to be clearly stated. Adger (2015b) does neither. From TLM he cites out of context “Chomsky famously proposed a Universal Grammar, which he dubs ‘a general principle of linguistic structure on the basis of observation of a single language’: English”. (p. 2) and proposes that this is “the opposite of what [Chomsky] says”. First, it is unclear what Adger means by ‘the opposite’: did Chomsky say the principles of UG could be proposed based on two languages [as the example of Japanese and English provided by Adger (2015b) might suggest] or can they be proposed only based on the study of all languages [which would be impossible for actual empirical research]? In any event, Adger’s quibbles seem rather pointless because one data analysis can suggest universals and one can propose a universal based on the study of one language. This proposal could be right or wrong, testable or untestable, clear or vague. What matters is whether the proposal makes testable predictions of any interest. Like Pesetsky, Adger uses selective quotation to misrepresent what Chomsky (quite reasonably, if one accepts the generativist biolinguistic commitments) proposed: “…we may plausibly postulate that P is a property of universal grammar on the basis of the investigation of a single language. There is no paradox here. The argument rests on the alleged fact that something is known without relevant experience so that knowledge must be attributed to the language faculty itself, a faculty common to the species. Deep analysis of a single language may provide the most effective means for discovering nontrivial properties of universal grammar” (Chomsky, 1977: 65).
In fact, Adger’s (2015b) arguments seem to suggest that contemporary Chomskyan generativists either (i) have no consistent notion of Universal Grammar or (ii) are unable to account for language acquisition. If the Chomskyan claim is the opposite of what TLM claims, then it is not clear what is ‘universal’ in universal grammar for the following reasons.
Adger cites Chomsky claiming that one would wish to confirm the proposed principles through the study of other languages. But that is an entirely different matter. Chomsky has claimed fairly consistently that language acquisition is made possible by UG because UG allows children to learn what could not possibly be learned just from the available language input (poverty of the stimulus arguments). One can of course study the (alleged) poverty of the stimulus by investigating just one language: For any language L, if x is not in the input and cannot be generalized to based on the input, yet x has been learned by children acquiring L, then x must be supplied by/be part of UG. If, on the other hand there are no principles that occur universally in all actual and possible human languages (as seemingly argued in Adger 2015b and Hornstein & Boeckx, 2009), then it is not clear how UG could account for the fact that a human infant can learn any language s/he is exposed to. These conflicting claims have been discussed in the literature for decades, yet Adger offers no explanation as to how they could be reconciled into one coherent framework.
Adger’s replies to other ‘substantial issues’ are equally unsatisfactory[3]. He reiterates the accusation that Evans misunderstands the generativists’ commitments regarding the localization or distribution of the hypothesized language organ. He cites irrelevant literature referring to other bodily organs but does not offer even a hint of the biological structure of the postulated language organ. According to Chomsky, biolinguistic research has been going on for over 60 years. Yet Adger is unable to offer any description of, or evidence for, the hypothesized biological organ. Given that Adger is one of the foremost experts on the Chomskyan framework one could conclude that currently no empirical evidence supporting the biolinguistic claims is available. Compare this state of affairs to the tremendous progress that has been made in genetics since the discovery of the double helix at roughly the same time as Chomsky’s first biolinguistic proposals. Actual work in biology has produced results (some of which are relevant to language and cited in TLM), while the framework defended by Adger has not. All exclusively language-specific biologic structures remain purely hypothetical and by now there is good empirical evidence that no such structures exist.
In his published replies to Evans’ work, Adger has not provided any defence against the substantial challenges to the generativist paradigm provided by Evans (2014 a,b). Instead, he has focused on alleged misrepresentations of a framework he seems unable to express with clarity. Overall, Adger’s arguments are akin to those of someone who defends phlogiston theory based on the fact that the critics have not described phlogiston correctly.
References:
Adger, D., 2015. Mythical myths: Comments on Vyvyan Evans’ ‘‘The Language Myth’’. Lingua 158: 76 - 80.
Adger, D. 2015b. More misrepresentation: A response to Behme and Evans 2015. Lingua
Adger, D., 2014. Syntax. Wiley Interdisciplinary Reviews in Cognitive Science. http://dx.doi.org/10.1002/wcs.1332
Behme, C. & V. Evans . 2015. Leaving the myth behind: A reply to Adger (2015)Lingua, Volume 162, 149-159.
Chomsky, N. 1977. Essays on Form and interpretation. New York: North Holland.Studies in Linguistic Analysis.
Evans, V. 2014a. The language myth: Why language is not an instinct. Cambridge: Cambridge University Press.
Evans V. 2014b. Real talk: There is no language instinct. Aeon: http://aeon.co/magazine/culture/there- is-no-language-instinct
Hornstein, N. 2015. Quotational dyslexia: Thank you Masked Man. On-line: http://facultyoflanguage.blogspot.ca/2015/01/quotational-dyslexia-thank…
Hornstein, Norbert. & Cedric Boeckx. 2009. Approaching Universals from below: I- Universals in light of a minimalist program for linguistic theory. In: Morten H. Christiansen, Chris Collins & Shimon Edelman (Eds.) Language Universals. (79 - 98). Oxford: Oxford University Press.
Notes:
[1] Adger objects to Behme & Evans (2015) alleged “insinuations about motivation” (p. 3). But he speculates about motivation throughout his reply: “What is criticized is not the generative position, it’s a chimaera invented to allow TLM to make an attack” (p. 2), “BE15 seems to assume that readers will not attend closely enough to the quotations to detect the misrepresentation here” (p. 4), etc.
[2] Adger claims to represent a wide range of generativists. This is not the case. Over the years generativists have diverged greatly in their commitments. In recent debates between Hauser, Chomsky and Fitch on the one side and Jackendoff and Pinker on the other it became clear that even generativists formerly very closely aligned with Chomsky no longer share at least some of his views. Jackendoff’s support of Behme&Evans (2015) shows that he also does not share the views defended by Adger (2015).
[3] The confusions and mischaracterizations about recursion and metaphysical issues are so profound that they cannot be dealt with here.
