Re: Human language acquisition

in article [EMAIL PROTECTED], David Longley at
[EMAIL PROTECTED] wrote on 12/1/03 11:37 AM:

> In message <[EMAIL PROTECTED]>, Joe Legris
> <[EMAIL PROTECTED]> writes
>> Glen M. Sizemore wrote:
>> 
>>> 
>>> JL: But this suggests that the difference between a dog barking in
>> response
>>> to his name being called and a child responding correctly to "Will the
>>> boy the birthday party is for please stand up!" is simply a matter of
>>> degree.
>>> 
>>> GS: Is the dogıs barking an operant under stimulus control of its
>> name being
>>> called? If so, then the two behaviors mentioned have something in
>> common -
>>> assuming the childıs behavior is operant behavior under stimulus
>> control. I
>>> donıt have any problem with this.
>>> 
>> 
>> The point is that children's acquisition of language has not been
>> demonstrated to be operant behaviour under stimulus control. If it is
>> operating then why don't words that are physically similar (i.e.
>> pronunciation, spelling, position in sentence) generally have similar
>> grammatical properties or meanings? Where does it apply?
>> 
>> [snip]
>> 
>> 
> 
> Hold on.
> 
> I've read and re-read your objections (and I know the Chomsky line), but
> I don't see what you are objecting to. First of all, there are others
> who have taken a very firm Skinnerian line *and* directly responded to
> Chomsky's odd attacks on empiricism (and I'm not referring to K.M).
> 
> I'm not referring to any of your specific points in particular, but
> given that language is our most sophisticated individual behaviour, why
> do you find the absence of a complete and detailed theory to be
> something which in any way vitiates Skinner's thesis that it is an
> operantly shaped set of behaviours? And more importantly, why do you
> find the cognitivist "alternative" in any way helpful or persuasive?
> Surely one could turn the same counter-argument on Chomsky viz the
> failure of machine translation etc?
> 
> The overall evidence (and I'm not going to repeat why *I* think there is
> good reason to accept the Radical Behaviorist approach, I've done that
> elsewhere at length and in detail) in favour of the approach would seem
> to be to be rational grounds enough. There is more to this than
> *psychology*.
> All that cognitivism has done over the past 40 years or so is distract a
> generation or two of psychologists in my view.


| Certain words and phrases inspire passionate responses (Carlin, 1978),
| even from scientists. Bayes' rule provides an example of this phenomenon.
| In modern language, the mathematics of Bayes' rule is elementary -- it
| follows from the definition of conditional probability distributions.
| Nonetheless, mention of Bayes' rule can still trigger heated discussion,
| with partisans displaying a nearly religious zeal. For a review of the
| history and current status of the controversies see Earman (1992). The
| problem (we think) lies with the claim that all prior expectations about
| the world can be encapsulated in a probability distribution. To give an
| example from the history of physics (Weinberg 1982), Pauli postulated
| the existence of an elementary particle that we now call the neutrino,
| a particle that would be very difficult to detect, as a way of explaining
| the apparent non-conservation of energy in radioactive decay processes.
| He viewed the neutrino as a distasteful hypothesis, but the idea that
| energy could be created or destroyed seemed even more objectionable.
| Clearly the "strength" of his belief in conservation of energy exceeded
| his concerns about nearly unobservable particles. But would he have been
| willing to state a probability that energy is not conserved? In the
| present discussion we try to be quite explicit in saying that sensory
| signals are drawn from a probability distribution, so there is, in
| principle, no ambiguity. This "in principle" is translated into practice,
| since we as experimenters choose the stimuli and we can design the
| experiment so that stimuli are in fact chosen at random from a
| distribution P[s(t)] that we have constructed. In a natural setting there
| is a deeper question: Is there a well defined probability distribution
| from which natural signals are drawn? Some efforts at specifying this
| distribution are collected in section 5.2.

-- Rieke, Warland, de Ruyter van Steveninck, Bialek, (1999) "Spikes:
Exploring the Neural Code", MIT Press.

> In the end, there are many alternative conceptual schemes which might be
> selected, however, which ones are picked for which domains and how well
> they survive will be a function of how good their explanatory/predictive
> utility is, how simple they are, how they fit in with the rest of
> science without causing unnecessary disruption elsewhere.

-- David Longley (Re Differential Cognition).

o The structure and physiology of the human brain set limits on what and how
humans can learn.

o The structure and physiology of the human brain are subject to both
genetic and environmental influences.

o Genomes are subject to the influence of natural selection.

The question is not *whether* natural selection is responsible for the
ability of humans to learn and use language, but rather *how*.  Is human
"verbal behavior" a reflection of some general purpose problem-solving
ability of the brain, or is it mediated via language-specific neuronal
circuitry? How much is innate, how much is learnt? The hypothesis that there
are specific neuronal circuits mediating human linguistic abilities makes
testable predictions: it should be possible to observe genetic defects of
linguistic ability; it should be possible to observe language-specific
impairment as the result of disease of, or injury to, the brain.

A few more quotes:

-----------------------

To the Editors:

It is of course gratifying to find my work on inherited language impairment
repeatedly cited by such distinguished scholars as Philip Lieberman and Noam
Chomsky, as has happened in a recent and extended exchange in your columns
[NYR, Letters, October 10 and December 19, 1991]. They clearly appreciate
that this inherited disorder is important because it allows us to
investigate the innateness of language in new ways. I am bemused, however‹as
I am sure any of your readers who have followed this exchange must also
be‹to see my findings used to support all sides of the argument. It is
important, therefore, for me to make clear the significance of my research.

Zuckerman, responding to Chomsky, raises as an ongoing question an issue
that I believe can be clearly settled by looking at inherited language
impairment over several generations. In The New York Review he claims that
whether "man's syntactical abilities [are] due to one set of interacting
genes or more than one [is] anyone's guess." While this question may have
been "anyone's guess" in the past, there is now converging evidence from
several studies that provide a clear answer: though certain cases of
developmental language impairment are associated with a single autosomally
dominant gene, these impairments affect only part of language‹the ability to
construct general agreement rules for such grammatical features as tense and
singular/plural‹and leave all other aspects of language, such as word order
and the acquisition of lexical items, unaffected. These facts answer
Zuckerman's question: Language must be the result of several sets of
interacting genes that code for different aspects of language rather than a
single set of interacting genes. In fact it is misleading to think of
"language" or "grammar" as unitary phenomena. Inherited language impairment
shows that different parts of grammar are controlled by different underlying
genes.

Lieberman cites my work to support the argument that there is genetic
variability in language. In fact, my work shows that there are two different
genotypes, one which results in the normal development of morphological
rules in language and another that results in an inability to learn this
specific aspect of language in the normal way. This means, obviously, that
language involves different genotypes; but variability across genotypes is
quite a different issue than variability within a genotype. Language
impaired subjects are not merely less good at certain aspects of language
than normals‹the way they learn these aspects of language is profoundly
different than the way the rest of us do. The kind of variability within
normal language users that Lieberman is searching for cannot really be
extrapolated from the populations I have studied.

Chomsky is sanguine that my results can be welcomed because they are
"interesting evidence for what had been assumed." In fact, while this work
does support his basic hypothesis that some aspects of language are innate,
it raises serious questions about precisely which aspects of language they
are. For example, it can be shown that a wide range of phenomena, from
noun-verb agreement to the sequence of tenses in narratives, are impaired in
similar ways. Though these properties are all "agreement relationships" in
some sense, there is no present theory of grammar, including Chomsky's, in
which they are treated in the same way.

The arguments about innateness have been based on inferences from
theoretical models of language. We now have a natural experiment that
provides an entirely new source of information. These new facts do not
merely confirm what has been assumed, but rather force us to revise widely
held assumptions about the fundamental structure of language.

M. Gopnik
Department of Linguistics
McGill University
Montreal, Canada

-------------------------------

A forkhead-domain gene is mutated in a severe speech and language disorder.
Lai CS, Fisher SE, Hurst JA, Vargha-Khadem F, Monaco AP.
Wellcome Trust Centre for Human Genetics, University of Oxford, Roosevelt
Drive, Oxford OX3 7BN, UK.
Individuals affected with developmental disorders of speech and language
have substantial difficulty acquiring expressive and/or receptive language
in the absence of any profound sensory or neurological impairment and
despite adequate intelligence and opportunity. Although studies of twins
consistently indicate that a significant genetic component is involved, most
families segregating speech and language deficits show complex patterns of
inheritance, and a gene that predisposes individuals to such disorders has
not been identified. We have studied a unique three-generation pedigree, KE,
in which a severe speech and language disorder is transmitted as an
autosomal-dominant monogenic trait. Our previous work mapped the locus
responsible, SPCH1, to a 5.6-cM interval of region 7q31 on chromosome 7
(ref. 5). We also identified an unrelated individual, CS, in whom speech and
language impairment is associated with a chromosomal translocation involving
the SPCH1 interval. Here we show that the gene FOXP2, which encodes a
putative transcription factor containing a polyglutamine tract and a
forkhead DNA-binding domain, is directly disrupted by the translocation
breakpoint in CS. In addition, we identify a point mutation in affected
members of the KE family that alters an invariant amino-acid residue in the
forkhead domain. Our findings suggest that FOXP2 is involved in the
developmental process that culminates in speech and language.

-------------------------------------

Putting language genes in perspective

Dorothy V. M. Bishop

Dept of Experimental Psychology, University of Oxford, South Parks Road,
Oxford, UK OX1 3UD 

 
Abstract
A point mutation has recently been found in a gene from affected members of
a family with an autosomal dominant pattern of inheritance for specific
speech and language impairment. However, this does not mean we have
localized the `gene for language'. The phenotype is complex, and the
affected gene, which is concerned with regulating activity of other genes,
is common to human and mouse. The discovery is nevertheless important,
because it will help us to identify target genes that play a role in
development of the neural circuitry involved in language.

--------------------------------------

Compounding and Inflection in Language Impairment:
Evidence from Williams Syndrome (and SLI)
Harald Clahsen & Mayella Almazan
Department of Linguistics
University of Essex

Abstract

We have examined potential dissociations between lexical and grammatical
knowledge in language impairment by investigating noun plurals and plural
formation inside compounds in Williams Syndrome (WS) subjects. For
comparison, we also report results from studies investigating the same
linguistic phenomena in children with Specific Language Impairment (SLI). We
found that nouns taking regular plurals were always correctly inflected by
the WS subjects, whereas they performed much worse on nouns that take
irregular plurals. The WS subjects heavily overregularized the regular -s
plural, and they even used the plural -s as non-head elements inside
compounds, i.e. in circumstances in which unimpaired children (and adults)
would typically not use regular plurals. We argue that the excessive use of
the regular -s plural in WS results from an impairment of the lexical system
and/or its access mechanisms. The dissociation between lexical and
grammatical phenomena in WS supports the theoretical distinction between
listed lexical entries and a (rule-based) computational system for language.

-------------------------------------------------

Auditory Verb Recognition in Developmental Language Impairment

Phaedra Royle and Gonia Jarema

Universite´ de Montre´al and Centre de Recherche de lıInstitut Universitaire
de Ge´riatrie de Montre´al, Montreal, Quebec, Canada

and

Eva Kehayia

McGill University, Jewish Rehabilitation Hospital, and Centre de Recherche
Interdisciplinaire en Re´adaptation, Montreal, Quebec, Canada

Published online January 22, 2002

This research addresses the issues of frequency and regularity in auditory
verb recognition and their importance for the organization of the mental
lexicon in DLI francophones. A primed auditory lexical decision task probes
reaction times and response accuracy in DLI and control participants on
frequent and infrequent, and regular and irregular, inflected forms of
verbs. Results show no priming for inflectionally related forms as well as
strong frequency effects in DLI participants. These findings lend additional
support to the hypothesis that developmental language impairment is a
deficit in lexical representation.

--------------------------------------------

Natural selection of the critical period for language acquisition

Natalia L. Komarova 1,2 and Martin A. Nowak 1

1 Institute for Advanced Study, Einstein Drive, Princeton, NJ 08540, USA
2 Department of Mathematics, University of Leeds, Leeds LS2 9JT, UK

The language acquisition period in humans lasts about 13 years. After
puberty it becomes increasingly dificult to learn a language. We explain
this phenomenon by using an evolutionary framework. We present a dynamical
system describing competition between language acquisition devices, which
differ in the length of the learning period. There are two selective forces
that play a role in determining the critical learning period: (i) having a
longer learning period increases the accuracy of language acquisition; (ii)
learning is associated with certain costs that affect fitness. As a result,
there exists a limited learning period which is evolutionarily stable. This
result is obtained analytically by means of a Nash equilibrium analysis of
language acquisition devices. Interestingly, the evolutionarily stable
learning period does not maximize the average fitness of the population.

------------------------------------------------

Regards,
-- Michael