Path: utzoo!censor!geac!torsqnt!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!mcsun!unido!uklirb!powers
From: powers@uklirb.informatik.uni-kl.de (David Powers )
Newsgroups: comp.ai
Subject: Re: Metaphors, Information and War
Keywords: metaphor, analogy
Message-ID: <7455@uklirb.informatik.uni-kl.de>
Date: 15 Jan 91 14:17:52 GMT
References: <shahar.663768754@shum> <26303@uflorida.cis.ufl.EDU> <28045@cs.yale.edu> <2012@oravax.UUCP>
Organization: University of Kaiserslautern, W-Germany
Lines: 47

daryl@oravax.UUCP (Steven Daryl McCullough) writes:

>In article <28045@cs.yale.edu>, mcdermott-drew@cs.yale.edu (Drew McDermott) writes:
>> Let me begin by saying that I agree completely with George Lakoff's
>> political conclusions in his recent postings.
>> ...
>> Having said that, I would like to argue that Lakoff's argument from
>> his "theory of metaphor" is just silly.

>Hear, hear! Too often in political debates, people are willing to
>swallow any cockamamie theory if it leads to politically correct
>conclusions. It is nice to see an illustration that it is possible to
>agree with someone and still think that their reasoning is
>wrong-headed.

>And (the eternal question) does any of this have anything to do with AI?

Whilst I cannot say I am in complete agreement with everything
Lakoff says OR concludes, he has helped ME (prompted ME) to look
past the metaphor filled rhetoric.  It is necessary at times to
strip away the clutter of mutually inconsistent frames.

In summary,

1. Metaphor is an essential aspect of language and learning;
2. Multiple metaphorical frames can be applied to the same facts;
3. Different metaphorical frames are somewhere mutually inconsistent;
	cf.  For every proverb there is one with the opposite force.
	     Our everyday language contradicts our scientific beliefs.
	     The parable is not intended to elucidate every detail.
4. It is hard work to understand complex situations without metaphors;
5. Our whole past experience is a metaphor for our future expectations.
	cf.  Life goes on as usual.
	     A good theory just hasn't been invalidated yet.

As far as relevance to AI is concerned,

1. Our understanding and recognition of the ubiquitousness of metaphor
	has been considerably enhanced by our attempts to formalize
	language and learning into computational models;
2. Our understanding of our understanding can be enhanced by
	understanding and stripping away metaphor (cf ontology);
3. AI systems are probably already part of the "war machine" (met. &
	lit.) being employed in this "crisis" (cf etymology);
4. The language of tomorrow is moulded by the metaphors of today.

David Powers