A Connectionist Defence of the Inscrutability Thesis and the Elimination of the Mental

Calvo Garzon, Francisco Jose (1999) A Connectionist Defence of the Inscrutability Thesis and the Elimination of the Mental. PhD thesis, University of Glasgow.

Full text available as:
[img]
Preview
PDF
Download (9MB) | Preview

Abstract

This work consists of two parts. In Part I (chapters 1-5), I shall produce a Connectionist Defence of Quine's Thesis of the Inscrutability of Reference, according to which there is no objective fact of the matter as to what the ontological commitments of the speakers of a language are. I shall start by reviewing Quine's project in his original behaviouristic setting. Chapters 1, and 2 will be devoted to addressing several criticisms that Gareth Evans, and Crispin Wright, have put forward on behalf of the friend of semantic realism. Evans (1981) and, more recendy, Wright (1997) have argued on different grounds that, under certain conditions, structural simplicity may become alethic-i.e., truth-conducive-for semantic theories. Being structurally more complex than the standard semantic theory, Quine's perverse semantic route (see chapter 1) is an easy prey for Evans' and Wright's considerations. I shall argue that both Evans' and Wright's criticisms are unmotivated, and do not jeopardize Quine's overall enterprise. I shall then propose a perverse theory of reference (chapter 3) which differs substantially from the ones advanced in the previous literature on the issue. The motivation for pursuing a different perverse semantic proposal resides in the fact that the route I shall be offering is as simple, structurally speaking, as our sanctioned theory of reference is meant to be. Thanks to this feature, my strategy is not subject to certain criticisms which may put perverse proposals a la Quine in jeopardy, thereby becoming an overall better candidate for the Quinean to fulfill her goal. In chapter 4, I shall introduce and develop a criterion recently produced by Wright (1997) in terms of 'psychological simplicity' which threatens the perverse semantic proposal I offered in chapter 3. I shall argue that a Language-of-Thought (LOT)-model of human cognition could motivate Wright's criterion. I shall then introduce the reader to some basic aspects of connectionist theory, and elaborate on a particularly promising neurocomputational approach to language processing put forward by Jeff Elman (1992; 1998). I shall argue that if instead of endorsing a LOT hypothesis, we model human cognition by a recurrent neural network a la Elman, then Wright's criterion is unmotivated. In particular, I shall argue that considerations regarding 'psychological simplicity' are neutral, favouring neither a standard theory of reference, nor a perverse one. In the remainder of Part I, I shall focus upon certain problems for the defender of the Inscrutability Thesis highlighted by the friend of connectionist theory. In chapter 5 I shall introduce a mathematical technique for measuring conceptual similarity across networks that Aarre Laakso and Gary Cottrell (1998; 2000) have recently developed. I shall show how Paul Churchland makes use of Laakso and Cottrell's results to argue that connectionism can furnish us with all we need to construct a robust theory of semantics, and a robust theory of translation-robustness that may potentially be exploited by a connectionist foe of Quine to argue against the Inscrutability Thesis. The bulk of the chapter will be devoted to showing that the notion of conceptual similarity available to the connectionist leaves room for a "connectionist Quinean" to kick in with a or\Q-io-many translational mapping across networks. In Part II (chapters 6, and 7), I shall produce a Connectionist Defence of the Thesis of Eliminative Materialism, according to which propositional attitudes don't exist (see chapter 7). I shall start by rejoining to two arguments that Stephen Stich has recently put forward against the thesis of eliminative materialism. In a nutshell, Stich (1990; 1991) argues that (i) the thesis of eliminative materialism, is neither true nor false, and that (ii) even if it were true, that would be philosophically uninteresting. To support (i) and (ii) Stich relies on two premises: (a) that the job of a theory of reference is to make explicit the tacit theory of reference which underlies our intuitions about the notion of reference itself; and (b) that our intuitive notion of reference is a highly idiosyncratic one. In chapter 6 I shall address Stich's anti-eliminativist claims (i) and (ii). I shall argue that even if we agreed with premises (a) and (b), that would lend no support whatsoever for (i) and (ii). Finally, in chapter 7, I shall introduce a connectionist-inspired conditional argument for the elimination of the posits of folk psychology put forward by William Ramsey, Stephen Stich, and Joseph Garon. I shall consider an objection to the eliminativist argument raised by Andy Clark. I shall then review a counter that Stephen Stich and Ted Warfield produce on behalf of the eliminativist. The discussion in chapter 5 on 'state space semantics and conceptual similarity' will be used to show that Clark's argument is not threatened by Stich and Warfield's considerations. Then, in the remainder of Part II, I shall offer a different line of argument to counter to Clark. A line that focuses on the notion of causal efficacy. I hope to show that the thesis of eliminativist materialism is correct. Conclusions, and directions for future research will follow.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Additional Information: Adviser: Jim Edwards
Keywords: Philosophy
Date of Award: 1999
Depositing User: Enlighten Team
Unique ID: glathesis:1999-76121
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 19 Nov 2019 16:37
Last Modified: 19 Nov 2019 16:37
URI: http://theses.gla.ac.uk/id/eprint/76121

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year