Re: Standardizing FOL
Message-id: <>
Date: Thu, 23 Sep 1993 15:02:13 +0000
Subject: Re: Standardizing FOL
Hi John-

I know I said last word, but Ive got to explain what you still don't seem
to understand. 

A semantics isnt (usually) the kind of thing that fits inside a computer:
that's a category mistake, like asking the weight of the letter 'J'. A
semantics is an account of how expressions inside computers might relate to
ways the world might be. Giving a semantics for a logic isnt a process of
programming. Its a process of explaining rigorously (mathematically, if you
like) what the expressions of the logic are supposed to mean. KIF and/or
CG's won't provide me with tools for doing this, and my doing it couldnt
*possibly* have anything other than intellectual effects on anyone else. If
I make a mess of my semantics, nobody but me will ever know unless I try to
publish it. Offering to provide a standard shell to develop semantics
within, like an operating system for software, seems crazy. I know you
aren't, so we must not be understanding one another.


PS  I didnt mean 'hack' to connote bad or careless programming: in the
culture I grew up in, hackers were regarded with awe. I meant only to
distinguish programming or describing a syntactic form (maybe in a
metalanguage) and defining a concept in a logical language. A definition
would, for example, enable you to draw conclusions from assertions using
that syntax.
The reason you can't define lambda-conversion in FOL is that its definition
essentially involves a schema which is not equivalent to any finite set of
first-order expressions (in contrast, for example, with the schema for

PPS, I PROMISE no more to Internet from me on this. Honestly. 

Beckman Institute                                    (217)244 1616 office
405 North Mathews Avenue        	   (217)328 3947 or (415)855 9043 home
Urbana, IL. 61801                                    (217)244 8371 fax  or