From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!csd.unb.ca!morgan.ucs.mun.ca!nstn.ns.ca!bonnie.concordia.ca!garrot.DMI.USherb.CA!uxa.ecn.bgu.edu!mp.cs.niu.edu!rickert Thu Apr 16 11:33:42 EDT 1992
Article 5014 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!csd.unb.ca!morgan.ucs.mun.ca!nstn.ns.ca!bonnie.concordia.ca!garrot.DMI.USherb.CA!uxa.ecn.bgu.edu!mp.cs.niu.edu!rickert
>From: rickert@mp.cs.niu.edu (Neil Rickert)
Newsgroups: comp.ai.philosophy
Subject: Re: syntax and semantics
Message-ID: <1992Apr9.191840.29417@mp.cs.niu.edu>
Date: 9 Apr 92 19:18:40 GMT
References: <92099.194744JPE1@psuvm.psu.edu> <1992Apr9.011549.15678@mp.cs.niu.edu> <92100.123228JPE1@psuvm.psu.edu>
Organization: Northern Illinois University
Lines: 57

In article <92100.123228JPE1@psuvm.psu.edu> <JPE1@psuvm.psu.edu> writes:
>In article <1992Apr9.011549.15678@mp.cs.niu.edu>, rickert@mp.cs.niu.edu (Neil
>[Emmer:]
>      I think there is some confusion here over what is meant by 'precise'.
>When someone else first used this term in this discussion, I took it to be
>just another way to state the fact that computers are formal systems.  You
>you seem to be saying that the formal systems instantiated by various
>computers only approximate the task of performing floating point computations,
>and that even this approximation differs from machine/system to machine/system.
>In other words, the computer's "precise" formal manipulations are used by us
>as an "imprecise" approximation of some task we wish to achieve.  I'm sorry
>if I am responsible for this equivocation.  I hope this clears it up.

  What I am saying is that computers are not formal systems.  They are
versatile mechanical simulation tools.  They can be used to simulate formal
systems, but this is only one use of them.  When running a language compiler
they are certainly simulating a formal system.  But when a computer is
doing massive numeric computations on floating point data, interpreting the
action as a formal system gives a very distorted and unnatural view of
what the computer does.  You might just as well say that the universe
is merely a formal system manipulating the various elementary particles
in accordance with some specifications.

>[Rickert:]
>>  Whether there is actually semantics in the floating point unit is a
>>different question.  I don't suggest there is - the semantics would be in
>>the data.  But I am not currently making a case that computers are capable
>>of semantics.  I am making the case that the reasoning you use to convince
>>yourself that they are not is seriously flawed.
>
>[Emmer:]
>     This I don't understand.  In my earlier post, I suggested that, if
>semantics has a 'where' it's in the pragmatic context.  When you say
>semantics is not "in the floating point unit" but perhaps "in the data",
>what are you sugsuggesting?  I don't know what you intend to refer tloatinge "f

  A human's semantics surely has something to do with the memory traces
(whatever they may be) accumulated through experience and interaction.
In contrast most commonly used computer program start out fresh and act on
new data, without making any use of a repository of memory traces built
through prior experience.  If semantics depends on accumulated experience,
then you would not expect much of it in a computer program which does not
attempt to accumulate experience.  But there is nothing inherent in computers
which requires you to limit them to programs of this type.  A program which
accumulated some kind of information from prior experience, and used that
accumulated data to modify the way it processes new information quite
possibly might have some semantic abilities.  But of course, once the
computer acts this way, there is no reasonable interpretation of it as
a formal system, since the behavior of a formal system is always exactly
repeatable whereas the actions of such a computer system are dynamic
based on experience and thus not exactly repeatable.

-- 
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
  Neil W. Rickert, Computer Science               <rickert@cs.niu.edu>
  Northern Illinois Univ.
  DeKalb, IL 60115                                   +1-815-753-6940


