From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!ogicse!psgrain!m2xenix!quagga!undeed!levy Tue Nov 19 11:09:28 EST 1991
Article 1247 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!ogicse!psgrain!m2xenix!quagga!undeed!levy
>From: levy@daisy.ee.und.ac.za (David Levy)
Newsgroups: comp.ai.philosophy
Subject: Re: Animal Intelligence vs Human Intelligence
Message-ID: <1991Nov06.101715.13520@daisy.ee.und.ac.za>
Date: 6 Nov 91 10:17:15 GMT
References: <1991Oct30.091241.9820@cs.bham.ac.uk> <1991Nov1.060336.28229@nuscc.nus.sg> <1991Nov1.151043.14789@news.larc.nasa.gov> <1991Nov2.073022.9468@nuscc.nus.sg>
Organization: Univ. Natal, Durban, S. Africa
Lines: 69

In <1991Nov2.073022.9468@nuscc.nus.sg> smoliar@maclane.iss.nus.sg (stephen smoliar) writes:

>I am delighted to see an opportunity for the discussion of the Whorf-Sapir
>hypothesis. If anyone knows of specific experimental evidence which may be
>taken as grounds for refuting it (or even treating it with skepticism), I
>would be very interested in accounts of that evidence.  I was always under
>the impression that the jury was still out.

>I think it is an open question whether or not Scott's hypothetical garage
>mechanic would actually acquire the same CONCEPTS Scott had for computer
>architecture.  The problem is that we can make observations about the words
>people use;  but it is much more difficult to draw conclusions about how (or
>even if) those words correspond to nodes in some vast semantic net internalized
>in the brain.  The other problem is that, whatever we may have in our heads, we
>are extremely flexible in the way we use out words, constantly adjusting their
>connotations as a result of verbal "negotiations" which take place implicitly
>during conversation.

>I would argue that a computer architect and a garage mechanic are too
>"conceptually close," so to speak, to constitute much evidence one way
>or the other (even recognizing the limited extent to which evidence may
>be gathered). 

Excuse a simple electronic engineer for intruding on this fascinating
discussion, but I think the following comment, made in a completely 
different context, is interesting and pertinent. In:

Newsgroup: comp.arch
Subject: Re: A Critique of the CM-5
Date: 5 Nov 91 17:44:54 GMT
Reply-To: bobg@phx.mcd.mot.com (Bob Greiner)

Bob Greiner writes:

> In article <1991Nov5.023131.2989@morrow.stanford.edu> dow@nova1.stanford.edu (Keith Dow) writes:
> >...  He should not use ideas like Minkowski-space-time though unless he 
> >plans to whirl his computer around at speeds near light. 
> 
> I have found that Minkowski-space-time is a quite useful concept in 
> hardware design.  Not in mathematical detail, but in the idea that 
> "simultaneous" is undefined.  
> .
> .
> .
> What is needed is good conceptual models and languages that allow 
> calculations to naturally follow this paradigm.  Data flow and 
> functional languages are one feasible approach; I am sure there are 
> many others. 

It is the last paragraph that is relevent here.
It is very common that experiment, experience and discovery result
in new concepts and even entire conceptual frameworks.  The new model
requires a new vocabulary for discourse, and soon discourse in the new
field is impossible without the new vocabulary.

It would seem likely, therefore, that different languages should incorporate
different concepts, with associated vocabulary, and that this
should have the effect of limiting communication until the would-be
communicants have learnt not only the vocabulary but also the associated 
concepts.

I would imagine that the most difficult issue here is how to measure the
effect reliably.

-- 

David C Levy, Dept of Electronic Eng, University of Natal, King George V Ave,
Durban 4001, South Africa,  levy@ee.und.ac.za, levy%ee.und.ac.za@rain.psg.com
Tel: (+27)-(31)-816-2727/8  Fax: ---2111   (Decnet): psi%(6550)13601353::levy


