7 - DEVIANT CAUSAL CHAINS

Summary

In the last thirty years, philosophers have formulated analyses of a number of psychological notions, analyses which share the following feature: they stipulate that the psychological analysandum can be applied only if some psychological state and some non-psychological state stand in a causal relation. I shall concentrate on three phenomena which have received this treatment: perception, memory and action.

Causal accounts of all three notions have become part of the received wisdom about the mind but, at the same time, it is widely recognised that these analyses are inadequate. Simply to require that two states of the relevant type stand in a causal relation is not enough for either perception, memory or action, for there are causal chains which link the states in question but which preclude the application of the relevant psychological concepts. These chains have been branded ‘deviant causal chains’ and conceptual analysts have set about trying to say what they are so that they can be formally excluded.

I believe that the most successful attempts to characterise deviant causal chains all implicitly rely on principles and ideas enunciated in earlier chapters of this book. To eliminate these deviant chains, we must maintain, as I did in chapter 1, that causal explanation is neither transitive nor agglomerative, and we must embrace the idea, broached in chapter 6, that there are distinct and irreducible causal mechanisms to be found on different levels of explanation. In this chapter and the next, I will examine, among other things, Peacocke's treatment of perception and action, and Martin and Deutscher's theory of memory, with a view to demonstrating their tacit reliance on these assumptions.