RE: "Semantics" vs. "No Semantics"

-----Original Message-----
From: public-rif-wg-request@w3.org
[mailto:public-rif-wg-request@w3.org] On Behalf Of Peter F.
Patel-Schneider
Sent: Monday, May 08, 2006 10:30 PM
To: Obrst, Leo J.
Cc: jos.deroo@agfa.com; public-rif-wg@w3.org
Subject: Re: "Semantics" vs. "No Semantics"


From: "Obrst, Leo J." <lobrst@mitre.org>
Subject: RE: "Semantics" vs. "No Semantics"
Date: Mon, 8 May 2006 18:21:13 -0400

> There are three ways to define the semantics, two formal, one
informal:
> 1) proof theory: axioms; this is typically harder than (2)
                                                         ^^^ I'm not so
                                                             sure of
this
> 2) model theory
> 3) a natural language description describing the operational
semantics.
>
> Of course a 4th way is some combination of these.
>
> What else is there?

Well, there are various abstract machine semantics, including the ones
for
Prolog and PDP-11s, to pick two quite different examples.  One can make
operational semantics as formal as one wants, including using a very
formal
theory of computation such as Turing machines or the lambda or pi
calculi
and a very formal mapping from syntax into this target.  Many modern
programming languages have a formal semantics in this style.

LEO: Oh, yes, sorry, you are quite right: formal machine semantics is a
5th. 

> Thanks,
> Leo

peter