NetNews Usenet Archive 1992 #27/NN_1992_27.iso/spool/bit/listserv/csgl/1583

Text File | 1992-11-20 | 7.0 KB | 147 lines

Comments: Gated by [e-mail]
Path: sparky!uunet!paladin.american.edu!auvm!BEN.DCIEM.DND.CA!MMT
Message-ID: <[e-mail]>
Newsgroups: bit.listserv.csg-l
Date:         Fri, 20 Nov 1992 18:04:40 EST
Sender:       "Control Systems Group Network (CSGnet)" <[e-mail]>
From:         [e-mail]
Subject:      Re: code metaphor
Lines: 136

[Martin Taylor 921120 18:00]
(Bruce Nevin 921118 10:35:06)

I hope that this will provide enough fun to compensate for the use
of net bandwidth on what is not PCT.  A view over the fence.

> Our discussions over the past
>almost two years have not made it clear to me where you stand on
>the literal merit of the metaphor.
>
>To see that this metaphor is wrong you have to recognize that a
>code has no information structure in it other than that imparted
>to it by the process of encoding; that is, on this view an
>utterance can exhibit only such information structure as was
>previously present in the message or belief held to be "in" the
>sender.
>
>There is no warrant for claiming that perceptions associated with
>an utterance (as the meanings of the utterance) are structured in
>the way that the utterance is, and there is much to suggest that
>they are not.

I partially answered this (921118 13:30), but I think you would be amused
by some of the argument currently going on about the meaning of "meaning"
on the Usenet group sci.cognitive.  If you think I lean toward "encodingism"
look at these (unattributed to save reputations) quotes that I read today...
------------------
1.
Consider a theory analyzing meaning-preserving tratsformations.  Let the
said transformations be regarded as argument-valued functions (in the
general sense, rather than taken in extension) on arguments, i.e.
self-contained units of discourse.  (An elementary argument may be
construed as a statement, i.e. a sentence in a context.)  In order to get
a reasonable theory of such beasts, you will want them to be broken down
to the elementary level, where each transformation will correspond in
extension to a one-one function t:A->A.  Of the said functions, one would
clearly be the identity I_A; moreover, if an elementary meaning-preserving
transformation is characterized by a function t:A->A, then it is to be
expected that there exists another such transformation corresponding to
its inverse, t^{-1}:A->A.  Finally, and most controversially from a
natural language standpoint, the philosopher will expect that, given two
meaning-preserving functions t,s:A->A, their composition, ts:A->A will
likewise be meaning-preserving.  Given these conditions, which obviously
likewise be meaning-preserving.  Given these conditions, which obviously
can be motivated in a way not depending on any reference to generative
processes, all arguments will be partitioned by the equivalence relation
induced by the class of all meaning-preserving transformations, into
equivalence classes of synonymous arguments.  Once your theory starts
referring to the said equivalence classes by including them in the ranges
of its existential quantifiers, you will thereby commit yourself to an
ontology of abstract propositions and concepts.

If you still cannot understand what I mean, you will have a problem in any
analytic discipline.  Go study logic.

(Name deleted)

P.S.  Is anybody out there objecting to transitivity of synonymy?

P.P.S.  As for the person who intoned that meaning-preservation was a folk
concept: if you really mean it, kindly go to a far-off place, and die by
the Death of a Thousand Cuts.  I can't mean that, can I?  But to resolve
any residual doubts, ask a Chinese friend to explain.
---------------
2.
>>
>>If you don't feel that being bound by
>>logic is a *moral* obligation, would you still have any basis for a
>>reasonable expectation that your tooting your car's horn at a bunch of
>>Samoan Hell's Angels does not *mean* your informed consent for them to
>>peel off your skin and rip out your heart?  Do you have a glimmer of
>>understanding that your human rights are not worth the market value of
>>the chemicals that comprise your anatomy, *if* there be no such things
>>as matters of fact about meaning?
>
>It seems to me that this statement mixes several immiscible issues.
>Let us remember that human rights are based on a *committment* to
>logic.  We shouldn't believe that logic is fundamentally 'right' or
>something.

If you do not believe that it is fundamentally right, you are not
committed to it.

----------------
3.
>Moreover, if understanding talk about "meaning" requires the adoption of
>some theory of meaning, then it would follow that philosophers can seldom
>understand each other when they talk about this subject, since there have
>been many and various theories of meaning. Thus, philosophers who hold
>different theories must be talking about different things when they discuss
>"meaning"; they only think they are talking to each other about the same
>thing.

I may be getting the wrong impression, but I do believe that philosophers
more often than not do not understand each other when they talk about meaning.
And then, after a while, you see them start talking only to those people
who do understand them, and their theory.

>>I did NOT insist that there are reasons to formulate theories of meaning - I
>>took that for granted. ...

>I thought it might be interesting to ask you _why_ you take this for
>I thought it might be interesting to ask you _why_ you take this for
>granted. Is there some innate joy in formulating theories? Are they perhaps
>aesthetic objects that we admire for their beauty? Or does this theory have
>some _use_? If the answer is, "well, if you have to ask, then you can't
>understand", then I will, of course, have to desist from my questioning.
>Formulating theories of meaning is, perhaps, a natural urge that some
>people have and that can't be explained further. Those who don't get the
>urge can't expect to participate.

The answer is: the theory has to be there in order to explicate what we are
really talking about. Some people may find some "innate joy" in formulating
theories or contemplating their aesthetic value. I think however there is
a more practical issue here.

--------------
4.
> P.S.  Is anybody out there objecting to transitivity of synonymy?

I would object to the assumption that synonymy is a property of
human language. Trivially, if we believe that two expressions
"mean the same thing" then we have some meaningful criteria to
distinguish them and some reason for doing so and, unless you want
to multiply abstract entities endlessly, they therefore do not "mean
the same thing". More concretely, for any pair of human language
expressions that you claim to be synonymous in a given context, I
claim that I can produce another context in which they are not
synonymous. Finally, while genuine synonymes *might* exist as
transient phenomena in human language, surely one would quickly
be driven out by the other for simple reasons of efficiency (would
you want to keep two identical copies of the same procedure in
a program?).
---------------

Martin