form

Frederick Newmeyer fjn at U.WASHINGTON.EDU
Thu Jan 23 19:53:30 UTC 1997


The Vargha-Khadem, et al. paper that Liz refers to is nothing less than
scandalous. This 3  1/2 page (!) paper, which Liz calls 'thorough', was
published in 1995, yet refers to no work by Myrna Gopnik and her group on
the K family that was published after 1991. In that period, they carried
out dozens of tests on the family that directly contradict the claims of
Vargha-Khadem, et al. The slipperiest thing that V-K do is to imply that
nongrammatical problems manifested by one (or some) of the affected family
members are manifested by *all* of the affected family members, giving the
illusion that there is a broad syndrome of problems associated with the
inability to handle inflectional morphology. In fact, there is none. The
low IQ scores for the affected members reported by V-K not only contradict
the scores reported by Gopnik and her associates, but also contradict the
scores published by *Varga-Khadem's own research group*. There is no
explanation for this discrepancy; in fact, there is no evidence that the
affected members of the family have statistically significantly lower IQs
than the nonaffected members.

The 'intelligibility' problems reported by V-K and repeated by Liz appear
to be almost entirely a function of the testing situation. The V-K group
brought the (uneducated working-class) family members into a laboratory
and pointed bright lights and video cameras at them. Relaxed settings
(party-like atmosphere in the subjects' homes) revealed vastly improved
articulatory abilities and few of the other problems reported by V-K. Bill
Labov taught us linguists decades ago about the importance of a
nonthreatening environment if one wants to assess natural speech. Few
psychologists, it would seem, have learned the lesson.

Implications that an auditory processing deficit is responsible for the
dysphasia cannot be correct. Affected members of the K family perform
excellently on phoneme-recognition tasks, and, moreover, have no
difficulty perceiving unstressed word-final segments that mimic the form
of inflectional suffixes (e.g. the final alveolar in words like 'wand').
Furthermore, whatever deficit the affected family members might have in
articulation, it could hardly explain why they make errors with suppletive
past tenses ('was', 'went') and with irregular pasts, regardless of the
sound that happens to occur in final position ('took', 'drove', 'got',
'swam', etc.). And, and Goad and Gopnik have pointed out, 'it is very hard
to see how articulatory problems could prevent them from making correct
grammaticality judgments or ratings which require them to just nod yes or
no or to circle a number'.

I could write much much more, but instead will refer you to an upcoming
special issue of Journal of Neurolinguistics, in which this issue will be
discussed in detail.

--fritz


On Thu, 23 Jan 1997, Elizabeth Bates wrote:

> But shall we assume (and I promise NOT to say "Dear Fritz...") that you
> DID endorse the claim that the genetic basis of grammatical morphology
> has been discovered?  That claim has made the rounds for several years,
> it is all based on a premature report about the K Family in London, and
> the report is stunningly wrong.  Faraneh Vargha-Khadem and her colleagues,
> who have studied this family for years and were in no way responsible
> for the original report (a letter to Nature by Myrna Gopnik) published
> a thorough study of the family in 1995 in the Proceedings of the National
> Academy of Sciences, showing that there is absolutely no dissociation
> between regular and irregular morphology, or (for that matter) between
> grammar and other aspects of language, because the affected members of
> the family are significantly worse than the unaffected members on a host
> of different languages tests and on a number of non-linguistic measures
> as well.  They also have a serious form of buccal-facial apraxia, i.e.
> they have a hard time with complex movements of the mouth, so severe
> that some members of the family supplement their speech at home with
> a home signing system.  A separate paper (not in the Proceedings) shows
> that they also have a hard time (relative to familial controls) with
> a finger-tapping task!
>
> Imagine the following tabloid headline: "Elton John and Lady Diana
> spent hot night together in Paris."  You buy the paper, open up to
> page 17, and discover that they spent the night together with 400
> other people at a party.  Kind of changes the interpretation, no?
> In the same vein, the grammatical deficits displayed by the affected
> members of this family are part of a huge complex of deficits, in no
> way specific to grammar much less to specific aspects of morphology.
> But the rumor continues to be passed around....-liz
>



More information about the Funknet mailing list