Online research in philosophy

Representation is a central part of models in cognitive science, but recently this idea has come under attack. Researchers advocating perceptual symbol systems, situated action, embodied cognition, and dynamical systems have argued against central assumptions of the classical representational approach to mind. We review the core assumptions of the dominant view of representation and the four suggested alternatives. We argue that representation should remain a core part of cognitive science, but that the insights from these alternative approaches must be incorporated into models of cognitive processing.

In Connectionism and the Philosophy of Psychology, Horgan and Tienson (1996) argue that cognitive processes, pace classicism, are not governed by exceptionless, “representation-level” rules; they are instead the work of defeasible cognitive tendencies subserved by the non-linear dynamics of the brain’s neural networks. Many theorists are sympathetic with the dynamical characterisation of connectionism and the general (re)conception of cognition that it affords. But in all the excitement surrounding the connectionist revolution in cognitive science, it has largely gone unnoticed that connectionism adds to the traditional focus on computational processes, a new focus – one on the vehicles of mental representation, on the entities that carry content through the mind. Indeed, if Horgan and Tienson’s dynamical characterisation of connectionism is on the right track, then so intimate is the relationship between computational processes and representational vehicles, that connectionist cognitive science is committed to a resemblance theory of mental content.

In Connectionism and the Philosophy of Psychology, Horgan and Tienson (1996) argue that cognitive processes, pace classicism, are not governed by exceptionless, “representation-level” rules; they are instead the work of defeasible cognitive tendencies subserved by the non-linear dynamics of the brain’s neural networks. Many theorists are sympathetic with the dynamical characterisation of connectionism and the general (re)conception of cognition that it affords. But in all the excitement surrounding the connectionist revolution in cognitive science, it has largely gone unnoticed that connectionism adds to the traditional focus on computational processes, a new focus – one on the vehicles of mental representation, on the entities that carry content through the mind. Indeed, if Horgan and Tienson’s dynamical characterisation of connectionism is on the right track, then so intimate is the relationship between computational processes and representational vehicles, that connectionist cognitive science is committed to a resemblance theory of mental content.

I explain why, within the nonclassical framework for cognitive science we describe in the book, cognitive-state transitions can fail to be tractably computable even if they are subserved by a discrete dynamical system whose mathematical-state transitions are tractably computable. I distinguish two ways that cognitive processing might conform to programmable rules in which all operations that apply to representation-level structure are primitive, and two corresponding constraints on models of cognition. Although Litch is correct in maintaining that classical cognitive science is not committed to the first constraint, it is committed to the second. This fact constitutes an illuminating gloss on our claim that one foundational assumption of classicism is that human cognition conforms to programmable, representation-level, rules.

Connectionism provides hope for unifying work in neuroscience, computer science, and cognitive psychology. This promise has met with some resistance from Classical Computionalists, which may have inspired Connectionists to retaliate with bold, inflationary claims on behalf of Connectionist models. This paper demonstrates, by examining three intimately connected issues, that these inflationary claims made on behalf of Connectionism are wrong. This should not be construed as an attack on Connectionism, however, since the inflated claims made on its behalf have the look of cures for which there are no ailments. There is nothing wrong with Connectionism for its failure to solve illusory problems.

Several accounts of representation in cognitive systems have recently been proposed. These look for a theory that will establish how a representation comes to have a certain content, and how these representations are used by cognitive systems. Covariation accounts are unsatisfactory, as they make intelligent reasoning and cognition impossible. Cummins' interpretation-based account cannot explain the distinction between cognitive and non-cognitive systems, nor how certain cognitive representations appear to have intrinsic meaning. Cognitive systems can be defined as model-constructers, or systems that use information from interpreted models as arguments in the functions they execute. An account based on this definition solves many of the problems raised by the earlier proposals.

This paper explores the difference between Connectionist proposals for cognitive a r c h i t e c t u r e a n d t h e s o r t s o f m o d e l s t hat have traditionally been assum e d i n c o g n i t i v e s c i e n c e . W e c l a i m t h a t t h e m a j o r d i s t i n c t i o n i s t h a t , w h i l e b o t h Connectionist and Classical architectures postulate representational mental states, the latter but not the former are committed to a symbol-level of representation, or to a ‘language of thought’: i.e., to representational states that have combinatorial syntactic and semantic structure. Several arguments for combinatorial structure in mental representations are then reviewed. These include arguments based on the ‘systematicity’ of mental representation: i.e., on the fact that cognitive capacities always exhibit certain symmetries, so that the ability to entertain a given thought implies the ability to entertain thoughts with semantically related contents. We claim that such arguments make a powerful case that mind/brain architecture is not Connectionist at the cognitive level. We then consider the possibility that Connectionism may provide an account of the neural (or ‘abstract neurological’) structures in which Classical cognitive architecture is implemented. We survey a n u m b e r o f t h e s t a n d a r d a r g u m e n t s t h a t h a v e b e e n o f f e r e d i n f a v o r o f Connectionism, and conclude that they are coherent only on this interpretation.

William Ramsey’s Representation Reconsidered is a superb, insightful analysis of the notion of mental representation in cognitive science. The book presents an original argument for a bold conclusion: partial eliminativism about mental representation in scientific psychology. According to Ramsey, once we examine the conditions that need to be satisfied for something to qualify as a representation, we can see those conditions are not fulfilled by the ‘representations’ posited by much of modern psychology. Cognitive science—or at least large swathes of it—has no warrant for positing representations. The structure of Ramsey’s argument repeats a familiar eliminativist strategy (c.f. Churchland (1981); Stich (1983)).1 First step: argue that in order for something to be an X, it must satisfy a certain description D (say, beliefs must satisfy the description given in folk psychology). Second step: argue that to the best of our knowledge, nothing satisfies description D (e.g. folk psychology is false). Third step: conclude that since nothing satisfies description D, there are no Xs (no beliefs). Here is how the strategy is played out in the book. First, Ramsey argues for certain minimal conditions that a representation must satisfy (what he calls the ‘job description’). Second (this takes the bulk of the book), he considers the ways in which our best psychological theories use the notion of representation. Ramsey argues that none of these uses satisfy the job description associated with a genuine representation. (A wrinkle is that some representations—those posited by the classical computational theory of cognition—do qualify as true representations. But, Ramsey claims, classical theories are in a minority in cognitive science, and their hold on the field is shrinking.) Therefore, Ramsey concludes, in most of cognitive science, there are no mental representations.

According to Ramsey (Representation reconsidered, Cambridge University Press, New York, 2007 ), only classical cognitive science, with the related notions of input–output and structural representations, meets the job description challenge (the challenge to show that a certain structure or process serves a representational role at the subpersonal level). By contrast, connectionism and other nonclassical models, insofar as they exploit receptor and tacit notions of representation, are not genuinely representational. As a result, Ramsey submits, cognitive science is taking a U-turn from representationalism back to behaviourism, thus presupposing that (1) the emergence of cognitivism capitalized on the concept of representation, and that (2) the materialization of nonclassical cognitive science involves a return to some form of pre-cognitivist behaviourism. We argue against both (1) and (2), by questioning Ramsey’s divide between classical and representational, versus nonclassical and nonrepresentational, cognitive models. For, firstly, connectionist and other nonclassical accounts have the resources to exploit the notion of a structural isomorphism, like classical accounts (the beefing-up strategy); and, secondly, insofar as input–output and structural representations refer to a cognitive agent, classical explanations fail to meet the job description challenge (the deflationary strategy). Both strategies work independently of each other: if the deflationary strategy succeeds, contra (1), cognitivism has failed to capitalize on the relevant concept of representation; if the beefing-up strategy is sound, contra (2), the return to a pre-cognitivist era cancels out.