Newsgroups: comp.ai.nat-lang
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!gatech!news.cse.psu.edu!news.math.psu.edu!news.cac.psu.edu!howland.reston.ans.net!spool.mu.edu!usenet.eel.ufl.edu!bofh.dot!warwick!bham!sunserver1.aston.ac.uk!usenet
From: rupraibs <rupraibs@aston.ac.uk>
Subject: Re: Transition Network for English Grammar. Possible or not?
X-Nntp-Posting-Host: denmark.aston.ac.uk
Content-Type: text/plain; charset=us-ascii
Message-ID: <DrCxzn.G8D@aston.ac.uk>
To: Shez@sv.span.com
Sender: usenet@aston.ac.uk
Content-Transfer-Encoding: 7bit
Organization: Aston University
References: <DqsH03.IL3@aston.ac.uk> <831782552snx@sv.span.com>
Mime-Version: 1.0
Date: Mon, 13 May 1996 18:57:22 GMT
X-Mailer: Mozilla 1.1N (X11; I; SunOS 5.4 sun4c)
X-Url: news:831782552snx@sv.span.com
Lines: 36

Shez <Shez@sv.span.com> wrote:
>rupraibs <rupraibs@aston.ac.uk> writes:
>>
>> If it were sufficiently big and complex enough, could a Transition Network be
>> set up to parse any sentence in the whole of the English Language?
>
>Such a network would have to embody a lot of semantic information, as
>properly parsing natural languages requires a lot of knowledge. Also,
>natural languages evolve continually, and differ from one locality to
>another, so the network would have to be updated on the fly from new
>meanings disclosed during communication with others.

I assume you are talking about 'real-world' knowledge?
Forgive me if I'm wrong but I thought that if you just wanted to parse 
a language using a grammar, you don't have to worry about semantics. If you
just wanted to check that the arrangement of the words were in a valid order, 
then the meaning of the sentence is irrelevent, isn't it? I would think this
is definitely true if you were only considering grammatically correct
sentences that conformed to some old-fashioned school teacher's strict
 definition of 'correct' usage of grammer. I agree that trying to find out what
 these sentences mean is another matter entirely. Could anyone help me here?

I originally put this posting up as I suspect the question:
"If its network were made sufficiently large and complex, could a transition
network parser process all English sentences? Justify your answer."
may come up in my AI exam. 
I always assumed that if you had a written piece of English that the
above hypothetical school master would mark as grammatically correct (ie no
errors or slang), then a sufficiently large transition network would have no
 problem in parsing it.
However, because the question had come up in a previous exam paper and needed
justification, I began to lose faith in my conviction that the answer was too
obvious to even ask the question!
This is why I wanted to hear other people's views on the issue. I apologise if
my original post was too vague in trying to convey what I was after.

