August 25, 2015

Regarding Harman and Greenberg’s response to Fodor and Lepore

In the philosophy of concepts, Conceptual Role Semantics is the view that the content of any of the concepts we use in order to think about the world (and, therefore, to categorise the perceptual input into meaningful categories) is determined by “at least” some of its inferential connections to other concepts. Let’s focus on the “at least” conception of Conceptual Role Semantics.

Fodor and Lepore (1992) (also important are Fodor 1990, 1998, ch.4, and 2000; Fodor and Pylyshyn 2015, among others) have argued that,

a) because no two people are committed to exactly the same inferences and
b) because there is no way of determining how the extent to which the addition of a new belief to one’s stock of standing beliefs directly or indirectly alters the inferences to which one is committed and
c) because there are no analytic beliefs,

the view that the content of concepts can be determined by a limited number of inferences (or conceptual roles) must be wrong. By contrast, Fodor’s view is that the content of a concept isn’t determined by its relation to other concepts, but, instead, by its relation to the world. This latter theory of content goes by the name of Informational Semantics.

Alternatively, Harman and Greenberg (2006) defend Conceptual Role Semantics (CRS). In particular, they have argued for an ecumenical version of CRS that includes Fodor’s Informational Semantics as a special case of CRS. Their central contention is that the content of mental states (and, therefore, concepts) is determined by “any part of their role and use in thought” (p. 296). According to H-G, Fodor and Lepore’s objections to CRS have failed to make the case that content-determination by means of conceptual role is fatally flawed. What H-G specifically claim is that F-L are really objecting a very limited conception of CRS, namely, one in which the inferential relations that are determinants of content are also constitutive of content. By contrast, H-G claims that CRS does not need to be committed to that assumption. In their own words,

“[…] from the claim that a given aspect of conceptual role, a certain belief for example, is part of what determines that a symbol has a given meaning, it does not follow that someone without the belief cannot have a symbol with the same meaning.”

These authors admit that a viable CRS cannot be exhausted by purely internal aspects of conceptual role, which is why they argue for a CRS that is not either wholly internalist or wholly externalist in scope. In their ecumenical view of CRS, as H-G claims, no theory of CRS can escape the internal aspects of use that help determine the content of concepts. This view is normally committed to the idea that content similarity, instead of content identity, is sufficient to guarantee the stability of content that is required for different agents (or the same agent at different time slices) in order to share the same concepts.

However, we are not told exactly how concepts can be shared in virtue of pure content similarity, and the claim that the conceptual roles that determine the content of concepts are not constitutive of content is controversial. According to H-G, it is perfectly possible that people with different beliefs can have a symbol with the same meaning (see previous quote). However, this is exactly what Informational Semantics predicts. So, it is perfectly possible—as it might be—that the sameness of meaning that H-G think their ecumenical conception of CRS can explain is really something that no special case of CRS can explain without this CRS being Informational Semantics. According to Informational Semantics (e.g. Fodor 1990, 1994), people with different belief systems can very well have concepts with exactly the same content, given that (according to this theory) it is the extension of a concept that determines its content.

H-G’s conception of CRS wants to have it both ways, but it can’t because, from the claim that people can have concepts with the same content (or symbols with the same meaning, for that matter), it does not follow that similar conceptual roles are part of what determines that a concept has a given content (or, for that matter, that a symbol has a given meaning).Alternatively, the claim that people have similar conceptual roles associated to a given concept does not preclude the possibility that similarity of meaning is explained by content identity without appealing to the idea that conceptual roles are constitutive of content.


Greenberg, M. and Harman, G. (2006). “Conceptual Role Semantics.” In Ernest LePore & Barry Smith (eds.), Oxford Handbook of Philosophy of Language, pp. 295-322. Oxford: Oxford University Press.
Fodor, J., (1990). A Theory of Content and Other Essays. Cambridge, MA: MIT Press.
_______. (1994). The Elm and the Expert: Mentalese and Its Semantics. Cambridge, MA: MIT Press.
_______. (1998). Concepts: Where Cognitive Science Went Wrong. Oxford: Oxford University Press.
_______. (2000). The Mind Doesn’t Work That Way: The Scope and Limits of Computational Psychology. Cambridge, MA: MIT Press.
Fodor, J. and Lepore, E. (1992). Holism: A Shoppers' Guide. Oxford: Blackwell.
Fodor, J. and Pylyshyn, Z. (2015). Minds without Meanings: An Essay on the Content of Concepts, MIT Press, 2015

No comments: