What do you think?
Rate this book
Cartesian Psychology and Physical Minds: Individualism and the Science of the Mind
This book offers an extensive critique of individualism in psychology, a view that has been the subject of debate between philosophers such as Jerry Fodor and Tyler Burge for many years. Rob Wilson approaches individualism as an issue in the philosophy of science, and by discussing issues such as computationalism and the mind's modularity, he opens the subject up for non-philosophers in psychology and computer science. Because the topic is so central to the philosophy of mind, an area generating enormous research and debate at present, the book has implications for a very broad range of philosophical issues .
288 pages, Paperback
First published May 26, 1995
13 people want to read
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 of 1 review
November 7, 2025
The long methodological beating of a dead horse senseless. Some notes:
Fodor’s 1987 argument from causal powers for individualism is invalid: you don’t need causal powers to individuate what you can scientifically talk about, only causal properties. Causal properties single out ‘victims of the Hiroshima bombing’, but their causal powers do not (p. 34). It’s also not true that the relational properties of things can be recapitulated in their causal powers (as in Fodor’s 1991 updated version of the argument). As Wilson writes later in the book, “The basic problem is that an entity’s relational properties do not supervene on that entity’s intrinsic physical properties.” (p. 125)
In ch. 2, discussion of computationalism: against the argument from computationalism for individualism. The idea is that wide computationalism, viz. treating more than just the individual, but the individual plus its embedding environment, as a computational system, is plausible. Based on Gallistel’s work, it seems like the computations carried out by insects only get them what they want (i.e. have content) because of an isomorphism relation that they have with formal relations obtaining in the environment. (Note: predictive processing is a form of wide computationalism: see Friston 2002; 2005 and Williams 2018; see also Ryder 2004).
Distinctions between computational vs physical and semantic vs syntactic ‘levels’ of content:
• The correlation thesis (Stich 1983): either you cannot have a difference in semantic content without a difference in syntactic form (false: Twin-Earth cases) or you cannot have a difference in syntactic form without a difference in semantic content (false: bats and dogs can be related to the same things, either via echolocation or via perception, so via different computational states, albeit with the hypothetically same content; see Sterelny 1990)
• Both the computational and the semantic levels of content only gain causal ‘power’ by piggybacking on their physical implementation
The individualist says that wide properties are Cambridge properties (Geach 1969): properties like being 12,000 miles south of London. They have no causal power. This argument, Wilson says, turns on an implicit identification of ‘genuine’ (i.e. not mere Cambridge-) properties with intrinsic properties, but that is false of most scientific taxonomies, which routinely trade in relational properties (p. 123). In fact, I think there are two stands one can take on this issue: (i) either show that (a) some relational properties, e.g. being a mother, being unemployed, being in a magnetic field, are causally efficacious, (b) and even that some essentially relational ones, like being a highly specialized species or not, are causally efficacious (highly specialized species tend to extinction in times of ecological catastrophe, and being a highly specialized species is relative to an environment and an evolutionary history) (Wilson’s strategy), (ii) or say that the individualist’s claim pushes us into a purely metaphysical/ontological debate about the status of ‘properties’ that is just too boring to bother about (‘Is being a mother a causally efficacious property?’) (my strategy).
It is not the same thing to say that what psychology is about, mental states, is internal (in the head), which is something both individualists (like Fodor) and non-individualists (like Burge) agree on, and to say that what is in the head must be intrinsically (and not extrinsically, i.e. relationally) individuated (p. 152).
Is narrow content oxymoronic? Consider Jackson & Pettit (1993): a case for a truth-conditional, non-individualistic concept of narrow content. First, folk-psychology is predicated on predictive content: conditionals associating circumstances to behaviour. Content so ascribed may be indeterminate (Jackson & Pettit, pp. 269, 275). It is truth-conditional, because it relates to patterns of behaviour in the actual world, and it is narrow, because doppelgängers would have the same. The problem, Wilson argues (p. 179), is that at least some of the conditionals’ antecedents will refer to mental states, whose contents are, by Jackson & Pettit’s own admission, wide.
Fodor’s 1987 argument from causal powers for individualism is invalid: you don’t need causal powers to individuate what you can scientifically talk about, only causal properties. Causal properties single out ‘victims of the Hiroshima bombing’, but their causal powers do not (p. 34). It’s also not true that the relational properties of things can be recapitulated in their causal powers (as in Fodor’s 1991 updated version of the argument). As Wilson writes later in the book, “The basic problem is that an entity’s relational properties do not supervene on that entity’s intrinsic physical properties.” (p. 125)
In ch. 2, discussion of computationalism: against the argument from computationalism for individualism. The idea is that wide computationalism, viz. treating more than just the individual, but the individual plus its embedding environment, as a computational system, is plausible. Based on Gallistel’s work, it seems like the computations carried out by insects only get them what they want (i.e. have content) because of an isomorphism relation that they have with formal relations obtaining in the environment. (Note: predictive processing is a form of wide computationalism: see Friston 2002; 2005 and Williams 2018; see also Ryder 2004).
Distinctions between computational vs physical and semantic vs syntactic ‘levels’ of content:
• The correlation thesis (Stich 1983): either you cannot have a difference in semantic content without a difference in syntactic form (false: Twin-Earth cases) or you cannot have a difference in syntactic form without a difference in semantic content (false: bats and dogs can be related to the same things, either via echolocation or via perception, so via different computational states, albeit with the hypothetically same content; see Sterelny 1990)
• Both the computational and the semantic levels of content only gain causal ‘power’ by piggybacking on their physical implementation
The individualist says that wide properties are Cambridge properties (Geach 1969): properties like being 12,000 miles south of London. They have no causal power. This argument, Wilson says, turns on an implicit identification of ‘genuine’ (i.e. not mere Cambridge-) properties with intrinsic properties, but that is false of most scientific taxonomies, which routinely trade in relational properties (p. 123). In fact, I think there are two stands one can take on this issue: (i) either show that (a) some relational properties, e.g. being a mother, being unemployed, being in a magnetic field, are causally efficacious, (b) and even that some essentially relational ones, like being a highly specialized species or not, are causally efficacious (highly specialized species tend to extinction in times of ecological catastrophe, and being a highly specialized species is relative to an environment and an evolutionary history) (Wilson’s strategy), (ii) or say that the individualist’s claim pushes us into a purely metaphysical/ontological debate about the status of ‘properties’ that is just too boring to bother about (‘Is being a mother a causally efficacious property?’) (my strategy).
It is not the same thing to say that what psychology is about, mental states, is internal (in the head), which is something both individualists (like Fodor) and non-individualists (like Burge) agree on, and to say that what is in the head must be intrinsically (and not extrinsically, i.e. relationally) individuated (p. 152).
Is narrow content oxymoronic? Consider Jackson & Pettit (1993): a case for a truth-conditional, non-individualistic concept of narrow content. First, folk-psychology is predicated on predictive content: conditionals associating circumstances to behaviour. Content so ascribed may be indeterminate (Jackson & Pettit, pp. 269, 275). It is truth-conditional, because it relates to patterns of behaviour in the actual world, and it is narrow, because doppelgängers would have the same. The problem, Wilson argues (p. 179), is that at least some of the conditionals’ antecedents will refer to mental states, whose contents are, by Jackson & Pettit’s own admission, wide.
Displaying 1 of 1 review