Page 52 - Encyclopedia of Philosophy of Language

Language, Metaphysics, and Ontology
implications. Crudely, it recommends that one should construe prepositional attitude sentences opaquely, by ignoring the referential semantic properties of their embedded 'that' clauses.
1. Defining Methodological Solipsism(MS)
Though the term is found in Kant, methodological solipsism in its late-twentieth-century sense was denned by Putnam (1975:220) as 'the assumption that no psychological state, properly so called, presupposes the existence of any individual other than the subject to whom that state is ascribed.' Davidson (1987), cri- ticizing this definition, pointed out that describers or descriptions presuppose things; psychological states do not. To sidestep this, it would be possible to sub- stitute 'constitutively requires' for 'presupposes.' Fodor (1987: 42) defines MSas 'the doctrine that psychological states are individuated without respect to their semantic evaluation.'
2. Identifying Psychological States
Mary's knowing that Paris isfull of tourists is not a pure psychological state because she cannot be in that state unless Paris is indeed full of tourists, which requires that Paris exists and a lot of tourists exist. 'Know,' and many other psychological verbs, are fac- tives.ButMary'sbelievingthatthereisabeautiful city
full of tourists, on the other hand, is a state that could obtain even if there were no city and no tourists. It appears, then, to be a psychological state properly so called.
It seems commonsensical to ignore whether an agent's beliefs are true or false when explaining behavior. The majority of philosophers and psy- chologists have always accepted this. Suppose that two 'green' individuals, A and B, share many attitudes and aspirations. Each believes that his local atmo- sphere is dangerously polluted. Because of that belief, each decides not to buy a car. A lives in MexicoCity, which is dangerously polluted, while B lives in Copen- hagen, which is not. The fact that A's belief is true while B's is false is irrelevant to the fact that their beliefs have the same cognitive role and the same effect. Psychology should try to build upon gen- eralizations of this sort, abstracting from the fact that people inhabit numerically distinct local environ- ments.
3. PsychologicalStatesandtheExplanationofAction
that the act being explained is a success whose achieve- ment depended upon a belief's being true. Defenders of MSreply that such explanations are hybrids in which contextual truths and relationships are 'woven in' to the description of the agent's mental states and actions. To maximize one's chances of picking up significant generalizations, one should prise off the explanation's external component and then try to describe the agent's mind and behavior in a pre- suppositionless way.
Some rationales for MS rely on strong philosophical assumptions about the nature of the mind. Stich (1983) defends a principle of the autonomy of psycho- logical states: 'the states and processes that ought to be of concern to the psychologist are those that supervene on the current, internal, physical state of the organism.' Any scientific inquiry into how a sys- tem works must define the boundary between the sys- tem and what lies outside it. Psychology is concerned to discover how the mind works, and the mind is a system that depends on the brain. So, if two people had physically and functionally identical brains, their psychological states ought to be counted as the same. The fact that they have different histories and are in different environments is irrelevant, unless and until such differences cause the two people's current inner states todiverge.
4. Inner States and Cognitive Processes
Fodor (1980) assumes that cognitive processes are computational manipulations of internal symbolic representations. Such operations, in both animals and machines, are purely formal or syntactic. 'Formal operations,' says Fodor, 'are the ones that are speci- fied without reference to such semantic properties of representations as, for example, truth, reference, and meaning.' But, as Stich (1983) is quick to point out, the formality assumption threatens all content-based classifications of psychological states, because content per se is a kind of meaning. Having a sense is a sem- antic property. If all semantic properties are explan- atorily irrelevant, then cognitive science must become entirely syntactic (as Stich recommends).
The assumptions made by Stich and Fodor entail conclusionsthatarestrongerthanthedoctrineofMS defined by Putnam. It is one thing to ignore a belief's truth-value, quite another to refuse to look at the conditions under which it would be true. If MSis inter- preted in the latter way, as enjoining that psycho- logical states should not be individuated by their truth-conditional contents, then psychological states will be far removed from mental states as normally conceived. Even then, though, MS does not undercut all talk of content in cognitive science. Two-factor' theorists such as Schiffer (1981), McGinn (1982), and Block (1986), hold that a part of a belief's content is
Not all explanations abstract from the agent's embed-
dedness in a particular environment, nor is the truth
of a belief always irrelevant to the explanation. It
depends on how the behavior is described, and on
the type of explanation. Suppose that A and B are
motivated to sell their old cars. If A's action is
described as 'selling A's car,' then B's act is not of that
type. This merely shows that actions, too, should be
individuated according to the MS principle. Suppose its conceptual role, and that this component is
30