LINGUIST List 6.304

Sun 26 Feb 1995

Sum: Discussion of human and non-human language

Editor for this issue: <>


Directory

  1. Steven Schaufele, Sum: More discussion of human and non-human Language

Message 1: Sum: More discussion of human and non-human Language

Date: Sat, 25 Feb 1995 11:17:24 Sum: More discussion of human and non-human Language
From: Steven Schaufele <fcoswsprairienet.org>
Subject: Sum: More discussion of human and non-human Language

Content-Length: 18462

There's been a fair amount of discussion about language among non-humans,
both publicly on LINGUIST and privately with me, since my previous summa-
ry in LINGUIST 6-28. I have already posted to LINGUIST a bibliographical
list, broadcast in LINGUIST 6-195; here follows a summary of some of the
main points that have come up in discussion.

First of all, i'd like to thank the following people who posted messages
to me personally:

Dan Alford, a.k.a. Moonhawk (dalfords1.csuhayward.edu)
Celso Alvarez-Caccamo (lxalvarzudc.es)
Rachel Lagunoff (ihw1009mvs.oac.ucla.edu)
Dorothy Hinshaw Patent (doropatentaol.com)
John E. Limber (jelchrista.unh.edu)
Alex Schwartz (Alex_Schwartzsagepub.com)
Andrew Spencer (spenaessex.ac.uk)
Mike Tomasello (tomasfs1.psy.emory.edu)
Bill Turkel (billhivnet.ubc.ca)
George Williams (gwwnavisoft.com)

In addition to these individuals, the following people have posted items
on the subject of language among nonhumans to LINGUIST since 6-28. I'm
going to focus less on these in the following summary, on the assumption
that most of you have already read what they had to say.

Lloyd Anderson (ecolingapplelink.apple.com)
Sherri L. Condon (slc6859usl.edu)
Jane A. Edwards (edwardscogsci.berkeley.edu)
Larry Gorbet (lgorbetmail.unm.edu)
Jacques Guy (j.guytrl.oz.au)
Gilbert Harman (ghhprinceton.edu)
Marion Kee (Marion.Keea.nl.cs.cmu.edu)
Bob Krovetz (krovetzcs.umass.edu)
Roger Lass (rogerbeattie.uct.ac.za)
David Pesetsky (pesetskmit.edu)
Benji Wald (ibeneawjmvs.oac.ucla.edu)
Sherman Wilcox (wilcoxalcor.unm.edu)

One major issue that came up in discussion was the methodological and
ethical justification of evaluating the linguistic ability of non-humans
by confronting them with the task of mastering a language belonging to
Homo sapiens, instead of by investigating the means by which they inter-
act amongst themselves.

In LINGUIST, Benji Wald pointed out that Chomsky himself, in a paper pub-
lished in the Sebeok/Sebeok vol. mentioned in my bibliography in LINGUIST
6-195, carefully drew the distinction between 'human language', the cog-
nitive system built into the *human* brain that enables us to master the
various human languages that are the raw data of our studies, and 'Lan-
guage' (which i shall in this posting capitalize -- not to be confused
with LSA's journal), the set of all possible similar systems, in whatever
species (or machine) they may manifest themselves. This phase of the dis-
cussion began with the question of the relation between an ability to re-
cognize one's image in a mirror and an ability to handle pronouns of vari-
ous sorts, including reflexives. (Briefly, dogs and cats are able to re-
cognize their reflections as members of their own species but consistently
treat them as 'others', behaving toward them as they would to strangers
of their own species; Roger Lass has suggested that the ability to re-
cognize the reflection as 'self' may not be evolutionarily advantageous
to such predatory animals. The anthropoid apes, however, manage after a
few minutes to realize that the reflection is 'self', at which point they
switch from the 'social' behaviours they would normally use to make the
acquaintance of strangers of their own species to using the mirror as a
tool for self-inspection. But there has been no evidence presented, to
my knowledge or that of anyone else on the List, that chimps, gorillas,
etc. have any ability to handle anaphoric reference the way human apes
do. Sherri Condon referred us to the work of her colleague at the Univer-
sity of Southwestern Louisiana, Daniel J. Povinelli, on this subject.)
But ultimately this discussion came back to a fundamental issue raised in
my previous summary: Even were it possible, through ingenious programme
design and/or arduous effort, to teach a real human language to a chimp
or a gorilla, the mere fact that the system in question would have to be
'force-fed', as it were, is evidence in itself that there is a serious
difference between the cognitive systems of the two species (human and
non-human) in question, since humans pick the stuff up with almost no
training whatsoever. Since much of our research programme is dedicated,
more or less directly, to investigating the ability of humans to learn
individual human languages 'naturally' (i.e., with a minimum of effort)
-- at least in childhood -- this unredoubted fact places the linguistic
abilities of non-humans at a significant distance from our discipline's
focus area. Whether the field of linguistics ought to broaden its focus
area to include non-human quasi-linguistic semiotic/communicative systems
is another question, of course, that was raised in private discussion
with me. Speaking for myself, i am very sympathetic to the notion of
embedding (to use a very professional word) linguistics as we now under-
stand it within a broader programme of research into Language, including
the semiotic/communicative systems of non-human animals.

Supposing that such animals could accomplish the task of mastering a hu-
man language (which has so far not happened), this would certainly prove
that the difference in species is not relevant to the nature of the lan-
guage in question, much less Language in general. But failure to accom-
plish the task is not as probative, since it can be interpreted in either
of at least two ways:

(1) the animals in question are no more than 'protolinguistic' (to use
Bickerton's term), unable to master a system of the complexity of human
language -- a quantitative evaluation.

(2) the animals in question are already in possession of an equivalent
system of comparable complexity but of incompatible organization (which
ipso facto interferes catastrophically with the proposed task) -- a qua-
litative evaluation.

Which of these alternatives is correct will require a completely diffe-
rent set of experiments, carefully studying the communicative behaviours
natural to these animals as we linguists study the communicative beha-
viours natural to humans.

One of my correspondents said, 'Why is all this effort devoted towards
teaching apes a modified version of English? A complex and highly social
species such as this has an extremely sophisticated natural communication
system of its own, yet we know very little about this ... Surely, before
embarking on experimental meddling of dubious methodological validity, it
would be more prudent to construct a research program aimed at mapping
out the cognitive abilities of the various species and more particularly,
their home-grown communication systems. Presumably, it's only against the
background of how apes communicate with each other in the wild that we
will really be able to interpret the kinds of artificial experiments pio-
neered by the Gardners and others ... When you provide the reference list
it would be extremely interesting (though a lot of hard work) to compare
the amount of research effort expended on teaching apes 2-year old
English, compared to the number of studies devoted to ethological study
of ape communication.'

A further, related issue was the epistemological question of how to reco-
gnize behaviour in another species that is equivalent to language in hu-
mans. I remarked myself at one point, 'Human language is not purely a
communication system, and its structure is not based solely on its commu-
nicative function. I use language for a variety of functions ... It does
not strike me as logically self-evident that all of these functions
should be served by the same system. I can imagine a species that en-
gaged in all of these behaviours but used a radically different system
for each. Would we be able to recognize all these different systems? and
would all of them properly fall into the field of inquiry we call 'lin-
guistics'?

Larry Gorbet, in his LINGUIST posting, touched very well on a further as-
pect of this problem, addressing in particular one of the criticisms that
has been made of the results achieved by the Gardners et al. 'Any formal
element whose meaning includes displacement (or more generally, *any*
fairly abstract meaning) will be intrinsically difficult to recognize as
such, simply because the "search space" (for the observable behavior that
points to meaning) is so large. The result is an *intrinsic* sample bias
in semantic analysis: concrete meanings are more likely to be *disco-
vered* than are more abstract ones, relative to their actual frequency of
occurrence. In addition to abstract meanings in the more obvious sense,
meanings which concern internal states (e.g. "emotions") can be difficult
to discern if one doesn't experience those states or similar ones.'

A couple of references to the science fiction literature are relevant
here, science fiction being often very useful for such 'thought-experi-
ments' in the Einsteinian sense. H. Beam Piper, who while definitely not
PC was brilliant, wrote a story called 'Naudsonce' (pp. 57-112 in the col-
lection Federation, published 1981 by Ace. Piper suicided in 1964, and i
have no idea whether 'Naudsonce' was published anywhere during his life-
time), in which an exploration party from Earth discovers an extrasolar
planet inhabited by an obviously sapient species which the explorers la-
bel 'Svants'. These people are living at roughly a Neolithic cultural/
technological level, and there's no question that they communicate by
means of deliberately modulated sound waves transmitted through the air,
so the humans immediately jump to the conclusion that they have something
akin to what we call 'language', and the professional linguist they've
brought along for just such eventualities gets to work trying to analyse
the Svants' 'language' and establish inter-species communication. No go.
Turns out that although both Svants and humans communicate by means of
deliberately modulated, etc., the resemblance ends there. Our auditory
systems (by which i mean both the auditory centers of our brains and the
neurosensory systems in our ears that serve as their input) are designed
to convert auditory impulses into a distinct sensory experience that we
call 'hearing'. The Svants' neurosensory systems are designed to convert
such impulses into something more closely approximating the sensory expe-
rience that tells us about our internal states (e.g., feelings of hunger,
thirst, satiation, heartburn, etc.). Thus, when a human reports a feel-
ing of pain or pleasure, another human apprehends the message intellectu-
ally; when a Svant reports such a feeling, an equivalent feeling is di-
rectly induced in the body of another Svant. In order to have any hope
of inter-species communication, the human explorers have to rely on two
peculiar individuals: a young male Svant who is as a result of a birth
defect 'deaf', but is otherwise quite intelligent and shows promise of
being able to master an ideographic writing system, and his mother, who
is 'normal' but sticks with her son out of maternal affection, and can
interpret for him to the rest of the Svant community. My point in giving
this rather lengthy summary is that recognition of the fundamental diffe-
rence between the human and Svant communicative systems constitutes a ma-
jor, and very challenging, cognitive leap for the human explorers; can we
be confident that a similar cognitive leap would not be necessary for us
in evaluating the cognitive behaviours of non-human apes, cetaceans, etc.?
(Piper, speaking through some of his characters, expresses scepticism
that a species whose principal communication system so thoroughly bypas-
ses the 'higher/rational' cognitive centers can develop very far in the
way of civilization. I personally don't share his scepticism. Several
of us during discussion have considered that cetaceans use sound both as
their principal means of exploring the world around them, via sonar, and
apparently to communicate. This suggests that a cetacean mode of commu-
nication might involve 'projecting' a sonar 'image' of what one is talk-
ing about into the listener's brain. Such projection may involve vocali-
zation skill levels outside the abilities of cetaceans; but assuming they
could do it, i don't think such 'direct' communicative methods preclude a
high level of civilization. I'm not sure they're all that different from
the ability of the Chinese, or of fluent signers, to communicate highly
abstract concepts by means of what is essentially visual imagery.)

Those of you who have read Suzette Hayden Elgin's intriguing though dif-
fuse novel Native Tongue may remember that in that novel, humans, having
over the course of several decades established contact with a variety of
alien species and managed to negotiate profitable trade deals with seve-
ral of them in spite of the obvious language barriers (the only reason
the generally despised profession of 'linguist' is allowed to survive in
this crypto-fascist state), encounters a species that is obviously sapi-
ent and obviously endowed with something in the way of a language (if i
remember correctly, it's not made clear exactly how this is recognized),
but that 'language' is so radically different from human language in its
structural organization that no human seems to be able to get the hang of
it (several people die trying).

In both cases, we are left with the Big Question: What, exactly, consti-
tutes Language, or a 'language'? And how do we recognize one if it exists?

On the subject of the cognitive organization of the brains of non-human
primates, Steven Pinker, on p. 350 of his recent book The Language Ins-
tinct: How the Mind Creates Language, discusses their equivalents of
Broca's and Wernicke's Areas: 'The neuroanatomists Al Galaburda and
Terrence Deacon have discovered areas in monkey brains that correspond in
location , input-output cabling, and cellular composition to the human
language areas. For example, there are homologues to Wernicke's and
Broca's areas and a band of fibers connecting the two, just as in humans.
The regions are not involved in producing the monkeys' calls, nor are
they involved in producing their gestures. The monkey seems to use the
regions corresponding to Wernicke's area and its neighbors to recognize
sound sequences and to discriminate the calls of other monkeys from its
own calls. The Broca's homologues are involved in control over the mus-
cles of the face, mouth, tongue, and larynx, and various subregions of
these homologues receive inputs from the parts of the brain dedicated to
hearing, the sense of touch in the mouth, tongue, and larynx, and areas
in which streams of information from all the senses converge.' For those
of you who are interested, the relevant references are:

Deacon, T. W. 1988. 'Evolution of Human Language Circuits' in H.
Jerison & I. Jerison, eds., Intelligence and Evolutionary Biology. New
York: Springer.

----. 1989. 'The Neural Circuitry Underlying Primate Calls and Human
Language' Human Evolution 4:367-401.

Galaburda, A. M. & D. N. Pandya. 1982. 'Role of Architectonics and Con-
nections in the Study of Primate Brain Evolution' in E. Armstrong & D.
Falk, eds., Primate Brain Evolution. New York: Plenum.

Beyond this, there is the fact that the integration of even complex so-
cial systems is not enough to account for the complexity of human lan-
guage. A large number of human social situations seem to be manageable
by just a few dozen different utterances; yet we have a natural linguis-
tic ability vastly in excess of that. I suspect that our ancestors may
have developed that ability through some nonce mutation, and then over
the subsequent millenia have gradually developed uses for it. (For some
at least tangentially relevant thoughts on this subject, see Theodore
Zeldin's recently published Intimate History of Humanity, Harper-Collins.)

Before closing this posting, i will quote the following from Dorothy
Hinshaw Patent's message to me, and reiterate my plea in LINGUIST 6-28
for more user-friendliness in our interactions with people who are rea-
sonably intelligent, may be well-educated, may even be fellow scientists,
but just aren't linguists.

'I was VERY frustrated by the critics of the ape language work; they
seemed to be looking for reasons to say the work didn't have meaning, and
I think your piece summarized many of the frustrations of people like me
very well. Would that everyone were able to think so clearly! What I'd
like to see is a linguistic analysis -- using minimal linguistic jargon,
so zoologists and psychologists could understand it, too -- of the work
done with apes, dolphins, sea lions, and parrots, comparing the "accom-
plishments" of the different species to one another and to human chil-
dren. I'd like the person doing it to make the assumption, as you have,
that the work DOES have meaning, and go from there, rather than trying to
find picky reasons for discounting it because of human hubris.'

Let us instead adopt the attitude expressed by Sherri Condon: 'I think we
have much to learn about cognition and communication in all species, and
we are fortunate that careful researchers are on the job.' Of course, as
is always true not *all* researchers 'on the job' are 'careful', by which
i understand 'responsible, not wedded to their hypotheses to such an ex-
tent that they are unable to wrestle appropriately with contrary evidence
or alternative approaches', but the responsible ones definitely deserve
encouragement!

Best,
Steven
---------------------
Dr. Steven Schaufele
712 West Washington
Urbana, IL 61801
217-344-8240
fcoswsprairienet.org

**** O syntagmata linguarum liberemini humanarum! ***
*** Nihil vestris privari nisi obicibus potestis! ***
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue