LINGUIST List 6.189

Sat 11 Feb 1995

Disc: Innateness/ Language & Species

Editor for this issue: <>

Directory

1. Sherman Wilcox, Re: 6.132 Innateness
2. , 6.136 Language and Species
3. benji wald, lg & species

Message 1: Re: 6.132 Innateness

Don Churma writes:

"This elephant is too big for any one "blind man" to figure out alone!"

Probably true, especially for this blind man, but we probably should keep
our hands firmly in contact with the beast anyway. For me, this means
taking to heart warnings such as Michael Studdert-Kennedy's:

"Characteristic motor systems have evolved for locomotion, predation,
consumption, mating. Matching perceptual systems have evolved to guide the
animan in these activities. The selection pressure shaping each species'
perceptuomotor capacities have come, in the first instance, from physical
properties of the world. By contrast, these perceptuomotor capacities
themselves must have played a crucial role in the form of a social species'
communication system. ... Certainly, specialized neuroanatomical signaling
devices have often evolved, but they have typically done so by modifying
pre-existing structures just enough for them to perform their new function
without appreciable loss of their old. ... Language has evolved within the
constraints of pre-existing perceptual and motor systems. We surrender much
of our power to understand that evolution if we disregard the properties of
those systems."

And...

"If there is indeed a universal set of linguistic features that owes
nothing to the nonlinguistic capacities of talkers and listeners, their
biological origin must be due to some quantal evolutionary jump, a
structure producing mutation. While modern biologists may look for
favorably on evolutionary discontinuities that did Darwin, we are not
justified in accepting discontinuity until we have ruled continity out.
This has not been done. On the contrary, the primacy of linguistic form has
been a cardinal, untested assumption of modern phonology -- with the result
that phonology is sustained in grand isolation from its surrounding
disciplines."

Sherman Wilcox
Dept. of Linguistics
University of New Mexico

Just a brief remark a propos of Bemji Wald's comment on Chomsky's
comment on whether apes can be shown to command 'reflexivization'; in
particular the question of whether recognizing your image in a mirror
as 'self' would be a sign of that ability.

If it's of interest, cats have a very peculiar relationship with
mirrors. IN my experience, most cats do not recognize images in
mirrors, or TV, etc. as three-dimensional at all, and simply
disregard them. But there are smart cats who do recogtnize their own
images AS CATS in mirrors: but almost invariably as 'non-self'. The
typical reaction of a cat seeing itself in a mirror, if it's a
'recognizer', is to bristle and hiss and go into defense-mode, or
sometimes attack-mode.

This of course raises what I like to call the Dr Doolittle Problem:
since we can't talk to animals we have to anthropomorphise and try to
guess by analogy what they might be doing, but have no sense of what
it FEELS like to be doing whatever. But in any case reflexivity is a
bad example, because in general most animals do not have, in nature,
any opportunity to see themselves; animals that do confront mirror-
like objects a lot (say surface predators that hunt under water like
herons, some cats, raccoons) probably must deliberately as it were
disregard the image they see, because they have to concentrate on
refraction and what's below the surface.

Roger Lass
Department of Linguistics
University of Cape Town

 The language & species discussion has not recurred this week, but I
 had already prepared the following, which I think would be of general
 enough interest to the recurrent misunderstandings about innateness
 and human language to publish on the list.

Since my last posting on the language & species
discussion I have received some interesting comments
and checked on the reference that I had in mind
(since I own a copy). The reference is a volume
called "Speaking of Apes: A Critical Anthology of
Two-Way Communication with Man". eds. T.A. Sebeok
 & Jean Umiker-Sebeok. NY: Plenum Press, 1980.
In it is Chomsky's article: "human language and other
semiotic systems." pp. 429-40. Chomsky's article is
pretty much as I remembered it, but with a great many
other observations on the nature of human language,
characteristic of his concept of human language, which
contrast/s with what the animal psychologists of the time
had taught apes (and perhaps even thought of teaching them),
e.g., potentially infinite embedding of phrases within other
phrases. Beyond that, he challenges the notion that human
language is (merely/primarily) a system for (social)
communication, which serves his notion of the
inappropriateness of comparing human language with animal
systems of communication for drawing conclusions about
the evolution of human language (and may also imply that
it is a strictly human device for interpreting external
stimuli and "thinking").

Incidentally, my rereading of the article changed my earlier
impression that Chomsky lacked "grace" in not explicitly
acknowledging the accomplishments and discoveries of
animal psychologists, to a perception that such explicitness
would be irrelevant and distracting to the points he wanted to
make. Of course he couldn't help activating involuntary
visceral hostility in some researchers when he accused them
of lack of logic in their arguments, but that is another matter.

Most interesting was his final position that regardless of
what apes may prove capable of learning, he saw evidence for
a qualitative distinction between the human and ape natural
intellectual endowment in the fact that humans acquire most
(syntactic) properties of language without (even the possibility
 of) explicit teaching, while apes obviously do not, despite the
 "evolutionary advantages" (C's phrase) that it would bestow
on them.

I particularly liked the last sentence of the following passage:

"Now it is difficult to imagine that children learning English
receive specific instruction about these matters, or even that
they are provided with relevant experience. In fact, we find
that while children make many errors in language learning,
they never make such mistakes as these: they never assume,
until corrected, that "the candidates wanted me to vote for
each other" means that each candidate wanted me to vote for
the other. In fact, relevant experience is never presented for
most speakers of English, *just as no pedagogic grammar
would ever point out these facts.*" p.432

To tell the truth, I don't get the error in the example (maybe
because I've never been corrected?), but I get the point. It's
the point about anaphoric reference that I mentioned in the
last posting (though I think the passage is trying to rely on
 some point about syntactic embedding of anaphora of the type
common at the time among generativists, cf. the parallel
reflexive "the candidate wanted me to vote for ?him-/herself
[unstressed]").

In any case, one might argue (I would not) that there is an
anthropocentric bias inherent in Chomsky's perspective on
the "evolutionary advantage" of human language -- I would
suppose stemming from what I think is the evolutionary tenet
that whatever promotes indefinite increase of the population
of a species is an evolutionary advantage, since that is supposed
to maximise the chance that at least some of the members of
the species will survive to continue the reproduction of the
species.

I guess an objection might be that in some sense apes "know"
something that we don't know that makes them shy away from
retaining or developing something like human language, e.g.,
that the technological advances allowed by human social
organization and motivation facilitated by language will
eventually lead to our extinction, a notion that would probably
 have evoked more rhetorical sympathy in the mid 1980s when
fear of nuclear holocaust peaked (or more persistently but less
clearly the Malthusian notion that uncontrolled human population
increase puts dangerous pressure on the ecological support system).

I doubt such an objection has any chance of being taken seriously
(in the form just given at least) by the scientific spirit. Imagine
the unimaginable that some human society (ANY human society) came
to this conclusion and rejected human language as ultimately
threatening to the species. Even so, I would guess that the
innateness hypothesis would predict that humans would still not
be able to "help" learning and manifesting language
(manifesting --) learning by future generations), and if that would
 contribute to eventual extinction, too bad. Nothing in
evolutionary theory prevents "defective" aberrations from arising.
The species with them would simply arise and then disappear
(relatively quickly?). However, I'm sure humans are constitutionally
 incapable of seeing the human language faculty as such an injurious
aberration -- I can't. At worst I can only see it as a possible means
of salvation from the jeopardy that some of our more sinister
instincts may have placed us in. Forgive me for even inventing what
I consider an idle and repulsive speculation -- but I think it throws
in relief what might be inferred in assessing Chomsky's ultimate
argument as I understand it. The simple summation of Chomsky's
argument is: if apes are capable of learning "human" language, why
don't they (do it naturally -- like people do)?

Probably more interesting and arguable to the list discussion is
Chomsky's point of distinguishing "human language" and "language".
Chomsky sees human language as a subject for scientific inquiry
with properties which are quite specific, including the syntactic
 properties of reference, embedding, etc. that we are all familiar
with as current linguists. In contrast, my understanding of the
article is that he sees "language" as a non-scientific concept,
something vague and not even promising as a potential scientific
field of inquiry. In this vein, he concedes that apes and many other animals
may -- in fact, he does not doubt -- make use of symbolic
systems (semiotic systems) apparently comparable in principle to
the lexical component of language in some way, though less
extensive, and, if I understand, less discrete (in the linguistic
 sense of "discrete"). And he supposes that such symbolic
systems in other animals may be related to shared intellectual
capacities of humans and these other animals, but that with humans
they interact with distinctive linguistic capacities (among the
latter I suppose the way lexicon fills in more abstract linguistic
categories in grammatical derivations). In an illustrative passage
(p.437) he objects to the Gardners' characterisation of teaching
apes to use Ameslan lexical signs as teaching them Ameslan
as a (human) language.

By the way, he considers acquisition of the sign for "and" as
trivial, with respect to comparison with human language. I
don't suppose that that is meant to detract from recognition of
the apes' ability to grasp this "logical operation" (by human
definition) -- but that such recognition is irrelevant to an
appreciation of what is distinctive about human language (well,
at least we now know that the "logical" concept "and" is NOT
distinctive to humans -- the concept "plural", then, probably
isn't either -- more problematic is the concept "dual" as far as
I know -- have apes been taught to count? Hey! last time I
looked I had THREE identical rubber ducks, now I only have TWO!).

In sum, then, Chomsky has a very specific and single-minded
notion of HUMAN language which allows him to immediately
"see through" claims about animal manifestations of "language",
just as it had earlier allowed him to criticise (and condemn)
Skinner's notions about the "nature" of human language. I think
the usual difficulty in seeing his point is not so much in the
persistence of linguistic debate about whether and to what extent
"autonomous syntax" is a valid notion (let's not get into that here),
but in the intuitive notion among linguists, as well as everybody
else, that lexicon is a major "part" of "human language". It is
certainly not the part that Chomsky associates with the distinctive
innate human faculty of language. Rather, the innate faculty is
somewhere in the systems which organise combinations of signs,
that "somewhere" being crucial to whether or not there is an
evolutionary discontinuity between human language and animal
potentials for "language".

Of course, another source of resistance to the idea of an innate
human language faculty is a generalised sneaking suspicion that
anything that proposes to set humans apart from other animals in a
fundamental way is self-deluding anthropocentric self-
aggrandising propaganda, cf. the discredited (I think) argument
against the heliocentric theory of the solar system that humans
are the "center" of creation and therefore their location MUST be
at the pivot of the material universe. While sneaking suspicions are
certainly appropriate issues to bring up for something as informal as
the ling.list discussion, it is not clear to me how it fits in to more
formal scientific argument.

Misguided as a source of resistance would be the idea floated in the
Kant/innate discussion that a theory that something is "innate" is a
killer to further attempts at "explanation". The killer to explanation
is the "just" in "that's JUST the way things are". On the contrary,
take out the "just" and there would be nothing to explain if there were
no "that's the way things are". If we get confused about this, it's
because, as scientists we DON'T know how things are, and our
"explanations" are hypotheses to test if things are the way
we THINK they are. I forgot what the context of Kant's discussion
of "explanation" was, but in the context of "pure" reason it would
have to be the "just". In "practical" reason I suppose whatever
aids remembering the "facts" is sufficient "explanation".

Finally, a message from Massimo Piattelli-Palmarini, director of the
dept of cognitive science at the Istituto San Raffaele in Milan, alerted
 me that among the discussion of comparisons between human and ape
cognition relevant to language capacity, "The truly definitive piece is
by Mark Seidenberg and Laura Petitto in Cognition, Vol 7, 1979,
pp.177-215." Although this article was published before the volume
I referred to above, it was too recent for most of the papers published
in that volume to fully discuss, so that there are only glancing
references to it in some of the papers. I still haven't read it yet,
 not that I've even read most of the articles in the Sebeok volume.
Incidentally, Massimo reminded me that apes were indeed found to
be able to recognise their reflections in mirrors, monkeys not (and I
think I read that in Roger Fout's popular book about teaching apes to
communicate with humans which came out in the late 70s). From what
I gather, animal psychologists etc. are not (or no longer) hostile to the
idea of a discontinuity between human language and what animals are
capable of, but remain (why not?) interested in discovering of what
animals are capable of, and what that might suggest about
human evolution. Benji