17.1320, Disc: Problems with Recent Proposals on Recursion & FLN

Sat Apr 29 15:00:37 UTC 2006

LINGUIST List: Vol-17-1320. Sat Apr 29 2006. ISSN: 1068 - 4875.

Subject: 17.1320, Disc: Problems with Recent Proposals on Recursion & FLN

Moderators: Anthony Aristar, Wayne State U <aristar at linguistlist.org>
            Helen Aristar-Dry, Eastern Michigan U <hdry at linguistlist.org>
Reviews (reviews at linguistlist.org) 
        Sheila Dooley, U of Arizona  
        Terry Langendoen, U of Arizona  

Homepage: http://linguistlist.org/

The LINGUIST List is funded by Eastern Michigan University, Wayne
State University, and donations from subscribers and publishers.

Editor for this issue: Ann Sawyer <sawyer at linguistlist.org>

To post to LINGUIST, use our convenient web form at


Date: 28-Apr-2006
From: Daniel Everett < dan.everett at manchester.ac.uk >
Subject: Problems with Recent Proposals on Recursion & FLN 

-------------------------Message 1 ---------------------------------- 
Date: Sat, 29 Apr 2006 10:59:23
From: Daniel Everett < dan.everett at manchester.ac.uk >
Subject: Problems with Recent Proposals on Recursion & FLN 

The recent article in Nature purporting to show that birds can recognize
recursive structures doesn't seem very convincing. On this I agree with
Chomsky's recent esponse that the results are more likely related to
short-term memory.

On the other hand, I am not particularly impressed with Chomsky's own
handling of the notion of recursion, nor of any of the things said about it
in the different articles by Hauser, Fitch, and Chomsky. (There are some
extremely useful articles showing how the reasoning about formal grammars
used to interpret results of primate experiments cited in these HF&C papers
is deeply flawed.)

They say, for example, that recursion is perhaps the only thing in the FLN
(Narrow Faculty of Language). And yet, when presented with the possibility
that there are languages without recursion, they say (see their recent
reply to Pinker and Jackendoff) that the only thing that really matters is
whether people can learn recursion, so that if a particular language lacks
it, this is not a problem. But think about this. If *a* language can lack
recursion with no damage to the Hauser, Fitch, Chomsky idea of FLN, then
*all* languages could lack recursion. That is, what they claim to be the
single essential characteristic of human language could be lacking in all
human languages. This claim is, well, contentless.

But then what of the claim that that is OK to think of recursion as the
only significant component of FLN, even in a world where all languages lack
recursion, so long as people can learn recursion? I think that this also
makes little sense. If people can learn recursion all that teaches us is
that it is part of the brain. Anyone surprised? FLN proponents will then
respond that this means that recursion is either part of FLN or part of FLB
(Broad Language Faculty). We have already seen that it is rather silly to
think in this case that recursion is part of FLN. But why would we even
need to believe that the learnability of recursion is part of FLB rather
than simply deriving from general properties of the brain, properties also
responsible for music, for culture in general, for gourmet cooking, and the
like? Another likely support for recursion as a general property of the
brain, as opposed to something as arcane/specific as FLN/FLB, comes from
the work of Herbert Simon. Simon has argued that thinking and handling of
information by human cognitive-based systems requires recursion, i.e. that
recursion in this sense comes 'for free' without any need to appeal to
grammar-in-biology (other than human brains in general). A Nobel
prize-winning economist, Simon makes this point about recursion in his
'Architecture of Complexity' article in 1962, a paper ignored by most

It seems more likely that brains have recursion and this can, but need not,
be reflected in human languages, given its usefulness (Simon 1962) in the
processing of complex information. 

That is, the whole issue of recursion as part of some FLN (and, really, I
think that FLN and FLB are unnecessary obfuscations in an already murky set
of proposals) seems deeply misguided. 

I defy any reader of this list to cite a single crucial experiment in the
history of psycholinguistics showing evidence for FLN/FLB or UG. 

Linguistic Field(s): Philosophy of Language

LINGUIST List: Vol-17-1320	


More information about the Linguist mailing list