Newsgroups: sci.lang
Path: cantaloupe.srv.cs.cmu.edu!das-news.harvard.edu!news2.near.net!MathWorks.Com!europa.eng.gtefsd.com!howland.reston.ans.net!EU.net!uknet!bcc.ac.uk!link-1.ts.bcc.ac.uk!skgtjt1
From: skgtjt1@ucl.ac.uk (Jim Tyson)
Subject: Re: language and memory
Sender: news@ucl.ac.uk (Usenet News System)
Message-ID: <1994Sep28.141211.49011@ucl.ac.uk>
Date: Wed, 28 Sep 1994 14:12:11 GMT
References: <1994Sep20.100935.1@ahab> <3644ba$8v5@newsbf01.news.aol.com>
Zone: if this offends you...too bad.  There is also at least one
Organization: Bloomsbury Computing Consortium
Lines: 80

In article <3644ba$8v5@newsbf01.news.aol.com> hafthors@aol.com (HafthorS) writes:
>In article <1994Sep20.100935.1@ahab>, rpayne@kean.ucs.mun.ca writes:
>
>Being bilingual, I think I qualify to make a response.
>

Eh, why?  I mean I am human but don't ask me to explain, eg, metabolic
processes that requires more than a human it requires a biologist.

>The brain (mine, anyway) stores information in a non-language specific
>manner. Pictures and concepts are built during conversation and the
>language is tacked on as a last step. 

How on earth do you know anything about inofrmation storage in the
brain?  Your subjective experience is of your *mind* and there is
no particular reason to suppose that they are constrained by some
representational homomorphism.  I mean maybe an appeal to common
sense views will gain you some support but...so what this is
*sci*.lang we are looking for rather more.  At least some
information must be stored in a language specific manner to account
for bilingualism and cross-linguistic variation.  For example,
some verbs of motion in Spanish and english involve quite different
argument structure realisations of conceptually (virtually)
identical situations.  Compare:

He swam into the river
El intro nadando el rio

(Forgive any egregious errors in my Spanish I am too lazy to
check that it is correct: the linguistic point should be
unaffected).

If we take the view that these sentences are equivalent in some
way then a Spanish speakers lexicon and its links to conceptual
representations is linguistically and specifically different to
that of an English speakers.  Bilingual English/Spanish speakers
can presumably unconsciously make the adjustment between these two
different realisations.  At least one popular model of speech
production (the Garrett model) proposes a serial account of
mappings from a high level representations (a message level) to
a phonetic realisation, but this is in no way tacking language
on at the end to pictures, or even pictures and concepts and
what you say can only apply to utterance production, comprehension
involves something like the reverse.  Further, many people
(not me I hasten to add but people whose views I am forced
to respect...) find serial models such implausible and favour
distributed/spreading activation models.

>FYI: When learning new languages,
>one should not try to match words of one language to another. Not just
>because of grammer differences and the occasional lack of direct word
>equivalance. It makes it simpler to remember the new words and easier to
>speak in the language because of the lack of the translation step.

This could be true but do you have any evidence?  Can 2L learners
possibly avoid assuming some lexical equivalence?  Competent
2L speakers report that they reach a stage where there appears to
be no level of "translation" involved in speaking in their 2L
(I feel this myself about some languages) but can learners simply
assume this state?

>Ideally, you should associate the new words with the actual objects or
>concepts they represent. Professional translators amaze me since they have

Which would mean you could only learn them by ostension...

>to first be able to speak while listening, have a buffer for remembering
>past contexts, and convert one language to an abstract and back again to
>another language.

Please don't think me rude, but is this any more than a folk theory of
translation.  This maybe what goes on, I don't know and neither does
Newmark nor Bell nor...(insert name of your favourite translation
theorist) and if this is what they do, is it very different from
what we all do in natural language interactions?  Sounds quite
straightforward to me.  Wonder why I can't do it (even in my
cradle languages)...

Jim Tyson

