[Wittrs] Re: An Issue Worth Focusing On
- From: "SWM" <SWMirsky@xxxxxxx>
- To: wittrsamr@xxxxxxxxxxxxx
- Date: Sun, 09 May 2010 18:37:48 -0000
--- In Wittrs@xxxxxxxxxxxxxxx, Joseph Polanik <jPolanik@...> wrote:
> SWM wrote:
>
> >Joseph Polanik wrote:
>
> >>SWM wrote:
>
> >>>My point is that "constitution" can be read as either a causal or an
> >>>identity claim.
>
> >>my point is that I'm not writing 'constitution' when I mean identity;
> >>so, if you are reading it that way, you are seeing the bat you put in
> >>the inkblot.
>
> >My point is that this is a discussion and debate about what Searle was
> >doing, ...
>
> as I see it, the issue is whether the third axiom is true; and, we are
> not limited to considering whether Searle wrote what you read him as
> saying. any evidence/argument that the third axiom is or is not true
> is available for consideration.
>
Yes, we can consider the truth or falsity of the third premise independent of
the CR but then the argument is no longer Searle's CRA (which is what we have
been discussing). I have already agreed that computational processes running on
computers (Searle's "syntax" in the CR) might not be adequate to produce
understanding. We might, by implementing and testing appropriate models,
discover that what a theory like Dennett's predicts would simply fail. But that
is not the issue here because Searle is giving us an argument that he says
demonstrates, via logic, that Dennett's thesis could not succeed. Searle is
arguing against the possibility of "strong AI" while Dennett's thesis is only
an argument in favor of the possibility. So Dennett's is open to empirical
consideration while Searle's is not.
You have elsewhere suggested there are other mathematical or quantum physics
related arguments for believing the third premise is true. I agree there could
be other, better logical arguments than the CRA though I have yet to see any.
However, this dispute is about whether the CRA succeeds and whether Searle made
a mistake in his formulation of it. I have argued that it doesn't succeed and
that Searle did make such a mistake (a number of them, in fact).
> >... not what you stipulate you mean.
>
> my commentary is correctly understood only when interpreted according to
> the definitions I give you for terms in dispute.
>
My comments are solely in relation to whether Searle's CRA succeeds. If you
have other, different arguments that take us to the same conclusions and which
do so in a more effective way than Searle's, we can, of course, consider them.
But even if one such argument exists, that, in itself, is not an argument that
Searle did not err in his formulation of the CRA which must stand or fall on
its own logical underpinnings.
> >Thus what's at issue is what Searle means (or seems to mean) by the
> >statements he incorporates into his Chinese Room argument (CRA).
>
> the CRT is the conceptual analysis on the basis of which it is said that
> the third axiom is conceptually true. the third axiom contains two
> negative statements; and, each is a denial of a positive claim that
> (allegedly) the CRT shows to be false.
>
As I have noted, the CR does not show that semantics qua understanding
cannot be understood/explained on a system level and, if it can be so
understood, then the third premise does not show that "syntax does not
constitute and is not sufficient for semantics".
> so, the question is whether the third axiom only contains denials of
> claims that are legitimately rejected; although, for the moment we are
> only considering claims that may be rejected on the basis of the CRT.
>
The CR is the basis of the CRA. The latter is formulated entirely on claims
derived from the former. If those claims are misguided or not supported, then
the CRA fails. The equivocal statement Searle formulates as his third premise
serves to mask the problem by making a non-identity claim seem to support a
non-causation claim.
> >1) His claim that the premise is "conceptually true" (which is only the
> >case for the denial of identity);
>
> according to you, the claim of identity is legitimately denied on the
> basis of the CRT; and, I agree. that is scenario 1.
>
Actually I don't think the CR is the basis for that denial but it certainly
does exemplify that and so serves a useful conceptual purpose.
> however, I do not take your word as evidence that no other false claims
> are rejectable on the basis of the CRT. I will take my own inventory.
>
I wouldn't want or expect you to!
> scenario 2 considers the claim of constitution.
>
> in this scenario, I hypothesize that the relation between syntax and
> semantic is one of constitution (where 'to constitute' means 'to count
> as').
"Count as" is to include something in a particular class. "Constitutes" doesn't
mean that exclusively. Why should we take Searle to have meant "count as"
rather than some other use when he said "does not constitute"?
> hypothesizing that syntax counts as semantics leads to the
> expectation that there will be semantic understanding in the CR; but,
> this is not the case. hence the hypothesized relation between syntax and
> semantics is rejected; and, I conclude that (given that sense of
> 'constitute') syntax does not constitute semantic understanding.
>
If we do take that use as the right one, the outcome still wouldn't change
because saying 'a car doesn't count as the class of motor vehicles' is no
different than saying the car is not the same as its class (a claim of
non-identity). Thus, "syntax doesn't constitute semantics" read as "syntax
doesn't count as semantics" is, finally, tantamount to saying "syntax is not
the same as semantics".
> [NB: one can also repeat scenario 2 using some other definition of 'to
> constitute'.]
>
> now, clearly the logic by which a claim of constitution is rejected in
> scenario 2 is the same as that by which a claim of identity is rejected
> in scenario 1. only the content of the hypothesis is different.
>
> so, Stuart, the question remains: what is the basis for claiming that
> the CRT does not support rejecting the hypothesis that syntax
> constitutes semantics?
>
Your version of translating "doesn't constitute" into "doesn't count as" is
still a non-identity claim. Of course, the CR does show the non-identity of
syntax and semantics as I've already said. The problem is that we still have to
get to a non-causal claim as in brains cause minds but computers can't! And the
CR does not demonstrate that what we call minds are not just an outcome of a
certain arrangement of syntactical constituents which is a legitimate way in
which we could say that causation occurs.
> your previous responses have been evasive and/or irrelevant; for
> example, from #5562 on 2010-05-06:
>
> >>defining 'constitutes' informally as 'counts as' and avoiding any
> >>conflation of constitution with either identity or causation, do you
> >>admit or deny that the hypothesis [2] is refuted?
>
> >Since your question is built on a series of mistakes, including an
> >insufficient understanding of how language works, a failure to grasp
> >that this is about Searle's argument, not your stipulations, and, your
> >inaccurate use of a word that you have made critical to your own claims
> >("constitutes"), any reply risks inadvertently sustaining these errors.
>
> taking inventory of what may be legitimately rejected as false on the
> basis of the CRT is *prior to* the construction of the third axiom from
> statements said to be true based on the CRT; meaning, that all of your
> speculations concerning the interpretation of the third axiom
> *presuppose* that such an inventory has been taken.
>
> let's just say I'm reviewing your work; and, I find that your inventory
> is ... incomplete.
>
> you correctly note that a claim of identity can be rejected on the basis
> of the CRT; but, you've failed to note that the claim of constitution
> can be rejected by the same reason line of reasoning. hypothesizing
> that a relation of constitution holds between syntax and semantics such
> that syntax counts as semantics yields a prediction that is falsified.
>
> Joe
The problem is that the claim of constitution as "counts as" that you have
offered unpacks to a variant of a claim of identity. While "constitutes" has a
range of uses in English, the only one that's relevant here looks like it has
to do with identity. Noting that a thing is not the same as the class in which
it is placed, that while it counts as a member of that class it is not the same
as that class, is still a claim of non-identity. And such a claim still does
not imply non-causality.
SWM
=========================================
Need Something? Check here: http://ludwig.squarespace.com/wittrslinks/
Other related posts:
