Newsgroups: comp.ai.philosophy
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!fas-news.harvard.edu!newspump.wustl.edu!news.ecn.bgu.edu!usenet.ins.cwru.edu!lerc.nasa.gov!magnus.acs.ohio-state.edu!math.ohio-state.edu!howland.reston.ans.net!news.sprintlink.net!news.bluesky.net!solaris.cc.vt.edu!insosf1.infonet.net!internet.spss.com!markrose
From: markrose@spss.com (Mark Rosenfelder)
Subject: Re: What's innate? (Was Re: Artificial Neural Networks and Cognition
Message-ID: <D3snLG.IBo@spss.com>
Sender: news@spss.com
Organization: SPSS Inc
References: <D38qGn.H6L@gpu.utcc.utoronto.ca> <D3LGHJ.LpM@gpu.utcc.utoronto.ca> <D3nIwA.5CA@spss.com> <D3r847.EuB@gpu.utcc.utoronto.ca>
Date: Fri, 10 Feb 1995 17:08:04 GMT
Lines: 92

[I've rearranged some paragraphs]

In article <D3r847.EuB@gpu.utcc.utoronto.ca>,
Andrzej Pindor <pindor@gpu.utcc.utoronto.ca> wrote:
>In article <D3nIwA.5CA@spss.com>, Mark Rosenfelder <markrose@spss.com> wrote:
>>* They can express a wide variety of concepts, by no means limited to 
>>ones they have heard before.  The major purpose of language, after all, is
>
>Do you mean that three year olds invent rutinely new concepts? There are
>of course geniuses, but in general I am very sceptical.

I was thinking of the concept expressed by an entire sentence.  By a new
concept I don't mean one of genius, just one that's new to the speaker.

I think it's also worth emphasizing that the child has to do some very
complex theorizing to learn language and to learn how to live in the world;
the insights and concepts acquired are of course commonplaces, but they 
are new to the child.  

>>* They've grasped the conceptual and semantic features underlying all that, 
>>including developing a system of categorization (cf. _Women, Fire, and 
>>Dangerous Things_), an understanding of time, causation, and possession, 
>>understanding of reasons and goals, acquaintance with various speech acts and 
>>their effect, and of course an enormous amount of real-world knowledge.
>>
>Like in the list above, I think that these statements are far too strong 
>and tainted by a theory one wants to prove by these 'facts'. Note that many 
>adults have problems with 'understanding of reasons and goals' except in
>very simple situations. 

I listed that based on sample sentences such as
  When it's got a flat tire it's need a go to the station.
  Why you put the pacifier in his mouth?
  I'm going to mail this so the letter can't come off.

To me, these sentences suggest a simple but firm idea that things are done
for reasons.  Naturally, the sort of reason that satisfies a child may not
be of the same quality that an adult would demand: when I asked my niece
why her dog barked so much, her answer was, "Because he's a doggie!"

>Ascribing all this to children is in the same category
>as insisting that Koko the Gorilla (or other ape) understands the sign
>language (or whatever else). It is to a large extend a matter of interpretation
>or theory used to organize the observations. 

Well, this counter-claim seems too strong to me.  I quite agree that we must
be very, very careful when investigating ape or child language.  (Or AI.) 
But we shouldn't let skepticism become a kneejerk naysaying, either.  Seeing
some sample sentences is not, of course, enough to base a firm conclusion on;
but we could easily design good experiments to test how well children 
understand reasons, time, possession, etc.  And in many cases these
experiments have been done already.

>For instance children make a lot
>of mistakes in language use (as you have observed yourself). At which stage 
>are we going to decide that they know the correct use of prepositions, for
>instance, if sometimes they use them right and sometimes wrong? 

How about counting how often they get them right?  Researchers have checked
large corpora of three-year-olds' utterances, and found that any given rule
(proper inflections, proper use of auxiliaries, insertion of function words,
etc.) is followed 90% or more of the time.  There are simply so many things
going on in language that even with this level of mastery, very many sentences 
contain some error.

>>Considering all this, most linguists don't find an explanation in terms of
>>copying what is heard very compelling.  

>It seems to me that part of the disagreement comes from different 
>interpretations of the of the word "copying". I did say aboove that children
>combine the elements of the language they heard in various ways, like
>they experiment with toy blocks, etc. My main point is that they experiment
>with material which comes from listening (in case of language) to the adult's
>speech. Even inventing new words (or their private speech, which sometimes
>happens) can be considered as copying what adults do - after all they hear
>adults making sounds which do not mean anything to the kids.

>Of course the copying is only a part of the process, but it provides the
>material on which learning is build and it seems to me that importance of
>this input is not sufficiently recognized. 

This is a nice illustration of the importance of one's point of view, I think.
Linguists and psychologists tend to have the opposite impression: they 
believe that the usual preconception is that language learning is *just*
copying what the parents say, and they fight against this preconception.

>As you probably know, children
>brought up in orphanages from the very early age (as babies) have a lot
>of developmental problems, for instance they do not speak very well. 

This is no doubt why the experimental psychologist Newt Gingrich wishes
to do more work in this area.
