Newsgroups: sci.lang,rec.arts.books,comp.ai.philosophy
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news3.near.net!paperboy.wellfleet.com!news-feed-1.peachnet.edu!gatech!ncar!uchinews!internet.spss.com!markrose
From: markrose@spss.com (Mark Rosenfelder)
Subject: Re: Chomsky on Consciousness and Dennett
Message-ID: <D9EzC4.ILF@spss.com>
Sender: news@spss.com
Organization: SPSS Inc
References: <D94Fo0.9CL@cup.hp.com> <JMC.95May25104241@sail.stanford.edu> <3q2put$p07@Mercury.mcs.com> <JMC.95May25182319@sail.stanford.edu>
Date: Tue, 30 May 1995 23:16:51 GMT
Lines: 47
Xref: glinda.oz.cs.cmu.edu sci.lang:39621 comp.ai.philosophy:28471

In article <JMC.95May25182319@sail.stanford.edu>,
John McCarthy <jmc@cs.Stanford.EDU> wrote:
>   The point, though, is that "promise" will not be the best word to use,
>   if it leads one to expect features that the computer omits.  And if
>   it *doesn't* lead one to expect those features, then those features
>   don't really belong in the definition...
>
>Webster offers:
>    1a: a declaration that one will do or refrain from doing something 
>    specified
>    1b: a legally binding declaration that gives the person to whom it 
>    is made a right to expect or to claim the performance or forbearance 
>    of a specified act
>    2: reason to expect something <little promise of relief>; esp
>    :ground for expectation of success, improvement, or excellence <shows 
>    considerable promise> 
>    3: something that is promised
>
>The items in a definition are alternatives.

I doubt that many linguists would accept dictionary definitions as more
than rough hints about the meaning of a word.

I remember an article by James McCawley criticizing another linguist's
analysis of the meaning of a word-- "accuse", I think it was.  The first
linguist had half a dozen conditions for felicitous use of the word; 
McCawley added several more.  I amused myself by coming up with four or five
additional aspects that both linguists had missed.

>Speculating about what people expect is very squishy, and Searle
>offers no evidence that people expect the promisser to expect benefit
>to the promissee.  If you were making a definition of promise as a
>technical term, would you include it?

One should attempt to offer at least informal evidence for one's 
definitions.  For instance, let's take your Webster citation for "promise":
>    1a: a declaration that one will do or refrain from doing something 
>    specified

So let's take some declarations that one will do or not do something:
     I'm going to the bathroom.
     I'm not voting for Gramm.
     I'll kill that little pest.
     We're going to the Bahamas for our vacation.
     I'll have a Big Mac.

You would maintain that those are promises?
