Re: Feasible Learnability as a Universal Constraint on All, even Alien, Languages

Hi Tom,

On Tue, 1 Nov 2005 Tom Chappell wrote:
>
> --- In conlang@yahoogroups.com, Yahya Abdal-Aziz
> <yahya@m...> wrote:
> >
> > Hi all,
> >
> > On Mon, 31 Oct 2005, Tom Chappell wrote:
> >
> > > It's a fact that every human natlang must be learnable by an infant
> > > who knows no language at all.
> > ... [snip]
> > > This fact actually puts non-trivial constraints on the class of all
> > > possible human natlangs.  Knowing these constraints, in turn, aids
> > > the learner in learning the language from the sparse data available.
> > > The same is true when an adult learns an L2.
 ... [snip]

> > ... With regard to the constraints you
> > have listed, I have two questions:
> >
> > (1)  Did you arrive at them a priori, or do you have
> > objective evidence for them?
>
> Neither.  I found them in some academic papers on the Web.  I'll try to
give some URLs.
>
> http://arxiv.org/PS_cache/cmp-lg/pdf/9611/9611002.pdf
>
> http://acl.ldc.upenn.edu/C/C88/C88-1001.pdf
>
> http://acl.ldc.upenn.edu/P/P88/P88-1028.pdf
>
> http://www.bic.kyoto-u.ac.jp/pathway/mami/pubs/ismb94.pdf
>
>
http://64.233.167.104/search?q=cache:NAG9r37tE4wJ:www.genome.ad.jp/manuscrip
ts/GIW94/Oral/GIW94O03.ps+beta-sheet+inside-outside+algorithm&hl=en

Thanks for the links.  I've downloaded them for study.


> > (2)  Given that some alien lifeform may have a system
> > of communication ("language") that uses different
> > senses than human natlangs do. to what extent might
> > this invalidate your constraints, or require their
> > extension?
>
>
> Dang good question.  I thought of the question, but don't have an answer,
so I didn't put the question on the list.  However, now that you have, I'll
go ahead and speculate.
>
> The fact that computational biochemists, and other specialists, have been
able to profitably use, and extend, the results the computational linguists
came up with, and their extensions have then been profitably applied by the
computational linguists, tends to make me think that these results can be
generalized to languages using different organs of speech and of
speech-perception, and even languages using different senses, than human
natlangs.  Don't forget that the various national sign languages all count
as human natlangs.
>
> That the constraints might have to be somehow modified or extended to
accomodate these different senses, seems possible, or even probable, to
me -- I am aware of my layman's ignorance here.

You and me too! :-)  Tom, we don't even have to leave the planet to find a
whole group of "alien species" that is apparently very good at communicating
using another sense.  I refer, of course, to the squids and cuttlefish that
communicate using rapidly changing patterns of colour displayed on their
bodies.  (I don't think this is true of all cephalopod species.)  From the
little I've gleaned so far, marine biologists consider these inverterbrates
to be among the most intelligent sea creatures, but have a long way to go to
decipher their languages - if indeed they deserve the name.  IIRC their
codes for communication outnumber those of most birds and primates, some of
whom do have quite large repertoires of coded calls.  A little Google
research should provide some links.

The only obvious constraints on which senses might be used in language is
that they must be expressive and informative; that is they must make
different configurations or patterns - in short, forms - available for other
sentient beings to detect.  On both counts, the human sense of taste is not
generally informative, since it requires contact with the tasted substance
in order to sense it, but does not itself produce those substances.  The
lateral-line sense in many fish is another nonhuman sense which is not
expressive.

Regards,
Yahya

--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.362 / Virus Database: 267.12.7/154 - Release Date: 1/11/05