Functionalism and mental boundaries

I’ve decided to dust off some of the papers from a themed issue that I co-edited five years ago since I happen to be very much in “extended mind” mode just now. First up is Larry Shapiro – below is his into; here is his abstract.

1. Introduction

Where are minds? For most people the answer to this question is obvious: the mind is in the head. The tough questions about minds typically concern how the physical stuff in the head produces minds. Surprisingly, however, there is growing controversy among psychologists and philosophers over how to answer the first question I asked. Traditional cognitive scientists (henceforth cognitivists) continue to defend the obvious answer. Researchers in the area of extended cognition (henceforth extended cognitivists) have urged a different answer. According to extended cognitivists, the mind’s location is only partly in the head. In addition, extended cognitivists have argued, the mind is located in parts of the world outside the body.

Clearly there is much at stake in this dispute. If extended cognitivists are right, there is much about psychology that is wrong. Cognitive neuroscience, for instance, would have been grounded in the false belief that all cognitive processes emerge somehow from neural processes. Computational psychologists would have to look beyond the brain to specify in full the implementation of algorithmic processes that previously had been thought to occur only in the head. Studies of psychopathologies could not limit themselves to an investigation of brain disorders.

Moreover, the possibility of extended cognition suggests new lines of research within the domain of social cognition. If minds extend, the boundaries that define the units of social interaction become less certain. Perhaps minds overlap. If, as some extended cognitivists believe, features of the environment comprise parts of a cognitive system, then a single piece of the world might constitute a piece of distinct cognitive systems. More dramatically, perhaps parts of a mind of one individual may be located within the mind of another. Insofar as extended cognition can make such possibilities plausible, social psychologists will need to re-interpret the nature of social interaction, will need to re-examine how the motivations and emotions of a single agent can influence an extended cognitive system, and so on.

Perhaps more seriously, if minds are extended then our ordinary ways of describing and thinking about human beings must undergo dramatic revision. We might have to learn to make sense of claims like “Welch accidentally left his memory on the bus,” or “Dixon stubbed his mind on his way to work this morning.” To traditionalist ears, both these claims sound like category mistakes, as in this example from Gilbert Ryle: “she came home in a flood of tears and a sedan-chair” (1949, p. 22). Just as floods of tears and sedan-chairs belong to distinct logical categories whose combination is jarring, the re-conception of minds that extended cognition promotes is likely to strike many as at least unnerving and quite possibly incoherent.

Of course, the possibility of extended minds must rest on a theory of mind. By this I mean that talk of extended minds can make sense only given various assumptions about what minds are. For instance, if one thought that minds are identical to brains – that mental properties are identical to neural properties – then the claims of extended cognition could be rejected outright. Grounding extended cognition must be a theory of mind that is consistent with the possibility of extended minds. This point may make one wonder whether the dispute between cognitivists and extended cognitivists is in fact a dispute over theories of mind. If so, this would be disappointing. The controversy is interesting only insofar as its participants share a view about what minds are but disagree over how to draw the mind’s boundaries.

Fortunately, many involved in the dispute seem committed to a common theory of mind, viz. functionalism. From the perspective of functionalism, mental states are identical to particular functional roles. Agreement about this lets the controversy over extended cognition take place at the appropriate level: the dispute can now focus on where minds are given a common assumption about what minds are.

Unfortunately, or so I shall argue, functionalism is the wrong perspective from which to judge the merits of the extended cognition program. Indeed, commitment to functionalism makes arguments for or against extended cognition too easy. Consequently, the decision regarding the mind’s extent must take place against the backdrop of non-functionalist considerations.

In the first part of this chapter I show how functionalism has been used to support a case for extended cognition. I then consider an argument that tries to drive a wedge between functionalism and extended mind. Although this argument is compelling, I next present what I take to be a more significant barrier to those who use functionalism to motivate extended cognition. More specifically, I argue that functionalism is ill-equipped to answer a boundary problem that confronts decisions about the extent of a property’s realization. Because functionalism cannot solve the boundary problem, I conclude that any principled assessment of extended cognition must rest in part on non-functionalist grounds. Before starting on these tasks, however, I must say something about the content of functionalism.