Abstract
One of the most important objections to information-based semantic theories is that they are incapable of explaining Frege cases. The worry is that if a concept’s intentional content is a function of its informational content, as such theories propose, then it would appear that coreferring expressions have to be synonymous, and if this is true, it’s difficult to see how an agent could believe that a is F without believing that b is F whenever a and b are identical. I argue that this appearance is deceptive. If we heed the distinction between the analog and digital contents of a signal, it is actually possible to reconstruct something akin to Frege’s sense/reference distinction in purely information-theoretic terms. This allows informational semanticists to treat coreferring expressions as semantically distinct and to solve Frege cases in the same way that Frege did—namely, by appealing to the different contents of coreferring expressions.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Following Schneider (2005), I’m treating ‘Mother’ as a proper name in Oedipus’ idiolect used for referring to his mother.
I’m using words in all caps to refer to the corresponding mental representations and words in italics to refer to the meanings of these mental representations.
This, in fact, is a bit too crude—even as an account of the crudest sort of IBSTs—for it is thoughts rather than concepts that can be properly said to have informational contents. Thus, it’s not the concept dog that depends upon the presence of dogs, but the thought that’s a dog or Lo, a dog! Strictly speaking, what informational semanticicsts should say is that dog means dog because the thought Lo, a dog! depends upon the presence of dogs—or something to that effect. We’ll have more to say on this below. For now, we need simply note that the apparent problem with fine-grainedness persists even if we adopt this stricter way of speaking.
Of course, not all solutions to the disjunction problem neatly fit into the ‘circumstances C’ framework. Fodor’s (1990a, b) account, for example, distinguishes the meaning-constitutive from the non-meaning-constitutive triggers of a concept in terms of the asymmetrical dependence of the latter on the former. The point is that, on anyone’s account, some further condition has to be added to the crude informational account in order to screen off the so-called ‘wild’ tokenings of a concept.
See, e.g., Dretske’s (2000, pp. 211-213). Incidentally, this feature of informational content is one of the reasons why an informational account of conceptual content is superior to a purely causal account, for while informational relations are intensional, causal relations are extensional. Note, moreover, that even so-called ‘causal covariation’ versions of informational semantics are not purely causal. On Fodor’s account, for example, dog means dog not because dog is caused by dogs, but because it’s a law (a counterfactual supporting generalization) that dog is caused by dogs (in circumstances C).
Note, however, that even here a potential difficulty arises. The complex concepts ab and cd can be distinguished in terms of their different primitive constituents only if we have some way of distinguishing these primitives. If they can be semantically distinguished, all is well. But suppose that these primitives also corefer. Now we can’t distinguish the complex concepts in terms of their syntactic properties without distinguishing the primitives in terms of their syntactic properties. But what sort of syntactic properties do primitive expressions have?
What’s the connection between being simple and lacking a definition? Well, if some name n were a complex concept composed of the simple concepts a and b, we’d be able to define n as a + b. Thus, if names lack definitions, they must be simple rather than complex (see Davis 2006, pp. 326–331).
See Aydede (1999) for a more complete overview of these and other objections. Schneider (2011) has made a valiant attempt to respond to these criticisms, but even she admits (vide p. 247) that a theory that allows for the sharing of narrowly individuated thoughts would be preferable to the syntactic approach, which does not.
The nature of this inheritance relation is clear enough in the case of referential content. To a first approximation, Hesperus refers to Venus, because thoughts containing Hesperus as a constituent counterfactually depend upon states of affairs containing Venus as a constituent. A bit more precisely: thoughts of the form Hesperus is F counterfactually depend upon states of affairs of the form Venus is F. Or, if we follow the strategy described below of making recognitional thoughts semantically basic, Hesperus refers to Venus because tokens of Lo, Hesperus! depend upon the presence of Venus in circumstances C. Nonreferential conceptual content is a bit more complicated, but the same general account will apply. On the view I’m proposing, the thought Hesperus is shining counterfactually depends not simply on the state of affairs in which Venus is shining, but on the state of affairs in which Venus is shining in the evening sky. Thus, what Hesperus contributes to the thoughts of which it is a constituent is not merely a referent, but a referent under a specific mode of presentation. Much more on this below.
Of course, this is true only on the assumption that our cartographer is not aware that Gaurishankar and Everest corefer. If the cartographer is aware of this coreference, it’s possible that her Gaurishankar concept will be tokened by the Everest mode of presentation or that her Everest concept will be tokened by the Gaurishankar mode of presentation. Are Gaurishankar and Everest now synonymous? No, for this is just another variation on the disjunction problem. How can Everest mean Everest, even though it is sometimes tokened by Gaurishankar? Well, pick your favorite solution to the disjunction problem. If we follow Prinz’s (2002) incipient cause condition, for example, we can say that Everest continues to mean Everest rather than Everest or Gaurishankar because it was encounters with the Everest mode of presentation rather than the Gaurishankar mode of presentation that originally gave birth to the concept.
Note that what we’ve said about innate concepts applies, mutatis mutandis, to innate knowledge. Knowledge—at least the sort of knowledge that IBSTs are interested in—involves propositional attitudes and propositional attitudes are composed of concepts. If S innately knows that xs are F, then experience of xs that are F may not be involved for the acquisition of this belief. However, if x’s being F is a contingent fact about the world, then the thought that x is F gets its content from its informational relation to a contingent feature of the external world, and, once again, such informational relations are possible only through the mediation of perception. Moreover, even if innate knowledge or innate concepts could be sustained by other than perceptual means, it’s unlikely that this would be relevant to our specific concerns. Few would be willing to maintain that we possess proper names innately or that we possess innate knowledge of the individual referents of these names.
I’m grateful to an anonymous referee for this articulation of the objection.
See Fodor and Lepore (1992) for a similar sort of objection. Note as well that a parallel problem faces any attempt to develop a notion of syntactic similarity—at least, that is, if syntax is understood to be a matter of a symbol’s computational role. Simply read ‘computational liaison’ in place of ‘inferential liaison’.
I am indebted to an anonymous reviewer for bringing this objection to my attention.
Note that as we add the necessary detection mechanisms to Bob’s repertoire, we’re able to identify distinct modes of presentation for his concepts. Suppose Bob remembers that he overheard one token of ‘Chomsky’ from Sam and another from Sally. He may now be capable of establishing an informational link to Chomsky via Sam or Sally. But now Sam’s and Sally’s testimonies come to function as distinct modes of presentations. If Bob’s Chomsky 1 depends upon Sam’s testimony, it will carry information about Chomsky, but this information will be nested within the more specific information associated with the fact that it was Sam rather than Sally who identified Chomsky.
References
Aydede, M. (1999). On the type/token relation of mental representations. Facta Philosophica, 7(2), 23–50.
Chisholm, R. (1957). Perceiving: A philosophical study. Ithaca: Cornell University Press.
Cohen, J., & Meskin, A. (2006). An objective counterfactual theory of information. Australasian Journal of Philosophy, 84(3), 333–352.
Davis, W. (2006). Nondescriptive meaning and reference. New York: Oxford University Press.
Dretske, F. (1981). Knowledge and the flow of information. Cambridge, MA: MIT Press.
Dretske, F. (2000). If you can’t make one, you don’t know how it works. In Perception, knowledge, and belief (pp. 208–226). New York: Cambridge University Press.
Fodor, J. (1990a). A theory of content, I. In A theory of content and other essays (pp. 51–88). Cambridge, MA: MIT Press.
Fodor, J. (1990b). A theory of content, II. In A theory of content and other essays (pp. 89–136). Cambridge, MA: MIT Press.
Fodor, J. (1990c). Substitution arguments and the individuation of belief. In A theory of content and other essays (pp. 161–176). Cambridge, MA: MIT Press.
Fodor, J. (1994). The elm and the expert. Cambridge, MA: MIT Press.
Fodor, J. (1998). Concepts. New York: Oxford University Press.
Fodor, J., & Lepore, E. (1992). Holism: A shopper’s guide. Cambridge, MA: Blackwell.
Frege, G. (1948). Sense and reference. Philosophical Review, 57(3), 209–230.
Kripke, S. (1980). Naming and necessity. Cambridge, MA: Harvard University Press.
Kulvicki, J. (2004). Isomorphism in information-carrying systems. Pacific Philosophical Quarterly, 85(4), 380–395.
Margolis, E. (1998). How to acquire a concept. Mind & Language, 13(3), 347–369.
Margolis, E., & Lawrence, S. (2007). The ontology of concepts: abstract objects or mental representations? Noûs, 41(4), 561–593.
Prinz, J. (2002). Furnishing the mind. Cambridge, MA: MIT Press.
Prinz, J. (2005). The return of concept empiricism. In H. Cohen & C. Lefebvre (Eds.), The handbook of categorization in cognitive science (pp. 679–694). San Diego: Elsevier.
Putnam, H. (1975). The meaning of ‘meaning’. Midwestern Studies in the Philosophy of Science, 7, 131–193.
Schneider, S. (2005). Direct reference, psychological explanation, and Frege cases. Mind & Language, 20(4), 423–447.
Schneider, S. (2009a). LOT, CTM, and the elephant in the room. Synthese, 170(2), 235–250.
Schneider, S. (2009b). The nature of symbols in the language of thought. Mind & Language, 24(5), 523–553.
Schneider, S. (2011). The language of thought. Cambridge, MA: MIT.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Rellihan, M. Informational Semantics and Frege Cases. Acta Anal 28, 267–294 (2013). https://doi.org/10.1007/s12136-012-0183-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12136-012-0183-z