From: "Wallace Polsom" <wallace-AT-raggedclaws.com>
To: <bhaskar-AT-jefferson.village.Virginia.EDU>
Subject: BHA: Various things (with the addition of Devitt & Sterelny)
Date: Mon, 29 Sep 1997 11:25:20 -0600


Thanks to Colin and Tobin for their swift responses to my first message to
the list and to Howard for his warm welcome this morning. Couldn't ask for
more.

Here's a section from _Language and Reality_ that relates to the "BHA: Re:
Various Things" thread (though it does not address postmodernism--that comes
later in the book) and will perhaps give those who are unfamiliar with
Devitt and Sterelny's work a taste of what awaits them (the humour contained
in the Twin-Earth fantasy is characteristic of D and S's style):

(Italics are indicated with opening and closing underlines, _like so_. My
additions are in square brackets. I apologize in advance for any typos.)
__________

<begin quotation>

3.5 Rejecting Description Theories

We think that description theories of names are wrong not merely in details
but in fundamentals. The whole programme is mistaken. Of course, we do not
claim to have _demonstrated_ this: there are few knockdown arguments in
philosophy. So, moves can still be made to continue the programme; for
example, by denying the evidence ("The ignorant do _not_ designate
Einstein"). We won't follow these lines of thought further. Instead, we aim
to undermine the description theories in two further respects. First, we
shall emphasize their essential limitations in the rest of the chapter.
Second, we shall develop an alternative [the causal theory of meaning] in
the next few chapters.

Even were we to set aside the fact that a description theory requires
speakers to have more knowledge than they actually have, the theory is
essentially incomplete. A description theory of names explains the
referential properties of one category of term, names, by appeal to those of
another, definite descriptions: on the classical theory, '_a_'  designates
_x_ in virtue of being associated with 'the _F_' which denotes _x_;
description is explained in terms of denotation. The referential properties
of descriptions are explained, in turn, by appeal to those of general terms:
'the _F' denotes _x_ in virtue of the fact that '_F_' applies to _x_and
nothing else; denotation is explained in terms of application (3.1). What
account of general terms is on offer? In virtue of what does '_F_' apply to
_Fs_? Perhaps a description theory of _some_ general terms would be
satisfactory: words like 'bachelor', 'judge' and 'murderer' may seem
definable. This process cannot, however, go on forever: there must be some
terms whose referential properties are not parasitic on others. Otherwise,
language as a whole is cut loose from the world. Description theories, which
explain one part of language in terms of another, can give no clues as to
how, ultimately, language is referentially linked to reality. These theories
pass the referential buck. But the buck must stop somewhere.

Strangely enough the essential incompleteness of description theories of
reference was not noticed until recently. It is still ignored by most
writers on reference.

A famous science-fiction fantasy, invented by Hilary Putnam to refute
description theories of natural kind terms ([_Mind, Language and Reality_,]
1975: 223-7), helps to bring out this incompleteness (and shows again that
identifying knowledge is not necessary). Imagine that somewhere in the
universe there is a planet, Twin Earth. Twin Earth, as its name suggests, is
very like earth. In particular, each Earthling has a _doppleganger_ on Twin
Earth who is molecule for molecule the same as the Earthling. Consequently,
many Twin Earthlings speak a language that seems like English. Indeed it is
phonologically and syntactically the same as English. Is it semantically the
same too? It cannot be if we deem it to include the proper names that Twin
Earthings use, because it is not referentially the same. When an Earthling
uses a name in English, he refers to an object on Earth. When his
_doppleganger_ uses what is apparently the same name, he refers to an object
on Twn Earth. Friend Oscar declares his voting intentions: "I'll vote for
Reagan; we need a dangerous president for a dangerous world". He is talking
about our local Earthly Reagan. Twin Oscar produces an utterance that sounds
the same, but he is not referring to Reagan; he has never heard of Reagan
nor of any other Earthling. He has his own problems with Twin Reagan in Twin
USA.

What does this show? It shows that reference, and hence meaning, does not
depend solely on the association of some words with other words, for all
those associations are the same in Twin English as in English. Further,
there is no internal state of the speaker that determines the reference, and
hence the meaning, of his words, for Oscar and his twin are alike
internally. No association of descriptions or mental images will make
Oscar's words refer to Reagan rather than Twin Reagan. No state of Oscar's
brain will do the job. As Putnam put it, "'meanings' just ain't in the
_head_" (1975: 227). We must look for some relation that language and mind
have to things outside themselves to explain meanings.

Indeed, _how could_ meanings be in the head? Meaning depends on reference
and reference relates a person and his words to something outside him. To
suppose that something internal to an objet is sufficient to determine its
relation to a particular object external to it is to suppose that it has a
truly magical power. A person no more has this power over the relation of
reference than he has it over the relation of kicking, teaching, being
taller than, or being the father of.

(Twin-Earth fantasies appeal to philosophers, but not to many others, as
Stephen Stich has emphasized; 1983: 62n. If you are one of the others, try
to invent a less outlandish example that makes the same point.)

For all this, there are some important truths in the description theory. The
idea that we must know something aboutan entity in order to refer to it by
name is appealing. We shall return to it (4.4). The idea of reference
borrowing [an idea put forward by a number of advocates of description
theories of reference, including Strawson] is especially important. It
highlights what Hilary Putnam calls "the linguistic division of labour"
(1975: 27-8). Our ability to use language is, in part, a social capacity. It
depends on our interactions with others in a community of varying interests,
capacities and expertise. Putnam made the point about general terms: our
capacity to talk about chromosomes, microchips and curved spacetime, despite
our ignorance of these things, is a result of our social links to others
whose acquaintance with them is more intimate. The same is true of names: a
person often succeeds in designating an object only in virtue of being at
the end of reference chains running through her linguistic community to the
object. However, our view of the natur3e of those links is very different
from that of the description theorist."

<end quotation>

__________

Thus endeth the lesson.

Almost.

Because, as an afterthought, here's a passage from the introduction that
will warm the heart of every supporter of depth realism.

__________

<begin quotation>

Our theory of language must . . . be physicalistic. Any linguistic facts
there are must be, ultimately, physical. Semantic notions like meaning,
truth and reference can be used only if they can be explained in
non-linguistic terms; they are not "primitive", not theoretically
fundamental. Biologists were not satisfied to leave the notion of the gene
as primitive, they wanted to understand the mechanism by which inheritable
characteristics are encoded in a cell. Their search led to the discovery of
the structure of DN. Similarly, we seek deeper explanation of semantic
notions. We might, for example, hope to explain them in psychological terms;
then, hope to explain the psychological in neuro-anatomical and biochemical
terms; then, explain those in physical and chemical terms . . . . if notions
like meaning and truth cannot be explained in non-semantic, naturalistic
terms, we should do without them in our theory of language.

<end quotation>

__________

Obviously, I have WAY too much time on my hands.

Wallace

Another Reading Recommendation:
Dieter Freundlieb, "Semiotic Idealism," _Poetics Today_ 9 (1988). This
article contains, in embryo, nearly all of the major objections to
poststructuralist theories of language. Freundlieb's examples are taken from
literary theory, but his arguments are clear enough that familiarity with
the original texts should not be necessary. (Freundlieb's main point is that
poststructuralist theories of language are generally best analyzed as forms
of "relativized Kantianism," and he is convincing!) As far as I can tell,
Freundlieb was (is?) a realist.





     --- from list bhaskar-AT-lists.village.virginia.edu ---