Paul Schweizer

The Truly Total Turing Test

The paper examines the nature of the behavioral evidence underlying attributions of intelligence in the case of human beings, and how this might be extended to other kinds of cognitive system, in the spirit of the original Turing Test (TT). I consider Harnads Total Turing Test (TTT), which involves successful performance of both linguistic and robotic behavior, and which is often thought to incorporate the very same range of empirical data that is available in the human case. However, I argue that the TTT is still too weak, because it only tests the capabilities of particular tokens within a preexisting context of intelligent behavior. What is needed is a test of the cognitive type, as manifested through a number of exemplary tokens, in order to confirm that the cognitive type is able to produce the context of intelligent behavior presupposed by tests such as the TT and TTT.

A syntactical approach to modality

ConclusionThe systems TN and TM show that necessity can be consistently construed as a predicate of syntactical objects, if the expressive/deductive power of the system is deliberately engineered to reflect the power of the original object language operator. The system TN relies on salient limitations on the expressive power of the language LN through the construction of a quotational hierarchy, while the system TMrelies on limiting the scope of the modal axioms schemas to the sublanguage LinfM+, which corresponds exactly with the restrictive hierarchy of LN. The fact that LinfM+ is identical to the image of the metalinguistic mapping C+ from the normal operator system into LM reveals that iterated operator modality is implicitly hierarchical, and that inconsistency is produced by applying the principles of the modal logic to formulas which have no natural analogues in the operator development. Thus the contradiction discovered by Montague can be diagnosed as the result of instantiating the axiom schemas with modally ungrounded formulas, and thereby adding radically new modal axioms to the predicate system.The predicate treatment of necessity differs significantly from that of the operator in that the cumulative models for the predicate system are strictly first-order. Possible worlds are not used as model-theoretic primitives, but rather alternate models are appealed to in order to specify the extension of N, which is semantically construed as a first-order predicate. In this manner, the intensional aspects of modality are built into the mode of specifying the particular set of objects which the denotation function assigns to N, rather than in the specification of the basic truth conditions for modal formulas. Intensional phenomena are thereby localised to the special requirements for determining the extension of a particular predicate, and this does not constitute a structural modification of the first-order models, but rather limits the relevant class of models to those which possess an appropriate denotation function.

The Externalist Foundations of a Truly Total Turing Test

The paper begins by examining the original Turing Test (2T) and Searle’s antithetical Chinese Room Argument, which is intended to refute the 2T in particular, as well as any formal or abstract procedural theory of the mind in general. In the ensuing dispute between Searle and his own critics, I argue that Searle’s ‘internalist’ strategy is unable to deflect Dennett’s combined robotic-systems reply and the allied Total Turing Test (3T). Many would hold that the 3T marks the culmination of the dialectic and, in principle, constitutes a fully adequate empirical standard for judging that an artifact is intelligent on a par with human beings. However, the paper carries the debate forward by arguing that the sociolinguistic factors highlighted in externalist views in the philosophy of language indicate the need for a fundamental shift in perspective in a Truly Total Turing Test (4T). It’s not enough to focus on Dennett’s individual robot viewed as a system; instead, we need to focus on an ongoing systemof such artifacts. Hence a 4T should evaluate the general category of cognitive organization under investigation, rather than the performance of single specimens. From this comprehensive standpoint, the question is not whether an individual instance could simulate intelligent behavior within the context of a pre-existing sociolinguistic culture developed by the human cognitive type. Instead the key issue is whether the artificial cognitive type itself is capable of producing a comparable sociolinguistic medium.

Intentionality, Qualia, and Mind/Brain Identity

The paper examines the status of conscious presentation with regard to mental content and intentional states. I argue that conscious presentation of mental content should be viewed on the model of a secondary quality, as a subjectiveeffect of the microstructure of an underlying brain state. The brain state is in turn viewed as the instantiation of an abstract computational state, with the result that introspectively accessible content is interpreted as a presentation of the associated computational state realized by the brain. However, if the relation between consciousness and representational content is construed in this manner, then conscious presentation does not provide an adequate foundation for the claim that human mental states areintrinsically intentional. On this model, I argue that functionalism is able to account for (non-intrinsic) intentionality, but not for consciousness, which has implications for the computational paradigm, as well as for Searles Chinese room thought experiment.

Physical Instantiation and the Propositional Attitudes

The paper addresses a standard line of criticism of the computational theory of mind, based on the claim that the notion of realizing a computational formalism is overly liberal to the point of vacuity. I argue that even for interesting and powerful cases, realization is essentially a matter of approximation and degree, and interpreting a physical device as performing a computation is always relative to our purposes and potential epistemic gains. However, while this may fatally undermine a computational explanation of conscious experience, I contend that, contra Putnam and Searle, it does not rule out the possibility of a scientifically defensible account of propositional attitude states in computational terms.

Consciousness and Computation

In his review of Mind versus Computer (Gams et al., 1997), Charles Dunlop (Dunlop, 2000) supplies a brief and quite dismissive critique of my article ‘Computation and the Science of Mind’. Dunlop’s basic verdict is that my argument is so trivial that it doesn’t even make a claim worth refuting. He summarizes by stating that my “ . . . core idea is that computational descriptions are abstract, but abstractions cannot in and of themselves provide the causal foundation for brain-based consciousness. Well, of course, but who would disagree?” (p. 293). I’m glad that Dunlop finds my major premise so obviously correct. The only problem is that he stops there, and neglects to follow the discussion through to its conclusion, which is normally deemed to be a rather important component of most arguments. My conclusion is that the functionalist/computationalist approach is unable to provide an adequate theory of consciousness, precisely because we are not conscious in virtue of computational structure. Hence I reject the standard functionalist line that any physical system realizing the same computational arrangement that’s implemented in my brain would perforce have the same conscious experiences as me. I argue that, unlike computational structures, conscious states are inherently non-abstract: they are actual, occurrent phenomena extended in physical time. Furthermore, they are by nature qualitative, while abstract structures are devoid of any qualitative dimension. Like numbers, sets and mathematical functions, systems of rule governed symbol manipulation (as well as connectionist networks formally conceived) are colorless and tasteless abstractions which exist neither in time nor in space, and which lack causal efficacy as well as qualitative aspect. Now, perhaps this last bit is uncontentious and, as Dunlop says, “Contemporary (symbolic) computationalists are not averse to allowing that mentality arises only when an appropriate Turing machine is physically realized . . . ” (p. 293). It might well be difficult (though not impossible) to find someone who does not accept that physical realization is a necessary condition for mentality. However, my own claim is much stronger, because I hold that the realization of an “appropriate Turing machine” in a non-neural substrate does not provide a sufficient condition for consciousness. The computational camp makes a critical error by espousing multiple realizability as a hallmark of their theory, while at the same time contending that identical

Physicalism, functionalism and conscious thought

In this paper, I provide further elaboration of my theory of conscious experience, in response to the criticisms made by David Cole, and I directly address a number of the issues he raises. In particular, I examine Coles claim that functionalism rather than neurophysiology is the theoretical key to consciousness. I argue that weak type-physicalism provides an analysis which is more fine grained, makes weaker assumptions, and allows more scope for empirical methods.

Realization, Reduction And Psychological Autonomy

It is often thought that the computational paradigm provides a supporting case for the theoretical autonomy of the science of mind. However, I argue that computation is in fact incompatible with this alleged aspect of intentional explanation, and hence the foundational assumptions of orthodox cognitive science are mutually unstable. The most plausible way to relieve these foundational tensions is to relinquish the idea that the psychological level enjoys some special form of theoretical sovereignty. So, in contrast to well known antireductionist views based on multiple realizability, I argue that the primary goal of a computational approach to the mind should be to facilitate a translation of the psychological to the neurophysiological.

Blind grasping and Fregean senses

ConclusionThe foregoing considerations have shown that on the Fregean model, no descriptive rendition of the meaning of a word, and no feature of the subjects psychological state, will be sufficient to answer the question of how reference takes place. Reference is determined by an independent semantical object, and the mind is limited by its perceptual access to this external semantical realm. The psychological and epistemic states of the language user will be causally influenced by this perceptual contact, and such causal influences will enable the subject to evince the appropriate behavioral signs of comprehension. But the internal effects produced by the perception of meaning are not in principle sufficient to individuate their semantical causes, and hence some of the prominent criticisms of the traditional theory are simply not applicable to Freges system.

In What Sense Does the Brain Compute

I analyse the notion of computation in the physical world and argue against the widely held view that merely implementing the ‘right’ sort of computational procedure is sufficient to transform a given configuration of matter and energy into a genuinely mental system. Instead, I advocate a more scientifically plausible version of the Computational Theory of Mind, wherein the interpretation of the brain as a computational device should (i) provide the theoretical bridge between high level intentional states and causally efficacious physical structure, and (ii) supply the integrated key for predicting both future brain states viewed as implementations of abstract computational states, and output behaviour viewed in cognitive terms.