Abstract
We show that, contrary to conventional wisdom, Frege’s distinction between sense and reference does not reconcile a classical logic of identity with apparent counterexamples to it involving proper names embedded under propositional attitude verbs.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Our 2 plays the role of Kaplan’s unnecessarily confusing (11) involving two uses of “F.D.R.”. The sort of “restrictions [...] attendant on consideration of a language containing ambiguous expressions” that Kaplan has in mind are not on which sentences to count as instances of schemas, but rather on which sentence uses are such that their falsity threatens the validity of schemas of which the used sentence is an instance. He would not deny, for example, that 2 is an instance of the schema “If \(\Phi\), then \(\Phi\)” (as his distinction between ambiguity-based and what he calls “mono-denotationalist” ways of thinking about of such sentences makes clear). He writes: “The natural analysis of (11) involves pointing out that the name ‘F.D.R.’ is ambiguous, and that in the second clause it denotes a television show rather than a man. Substitutions or any other logical operations based on the assumption that the name has here its usual denotation are pointless and demonstrate nothing [..., unlike] transformations based on a correct analysis of the name’s denotation in this context [...]” (Kaplan 1968, p. 183, emphasis his).
The idea that Fregeans should hold that sentences like 1 are valid is not an idiosyncratic suggestion of Kaplan’s. On the contrary (and to our surprise), it has been a common reaction, both in talks and in conversation, whenever we have claimed that the validity of such sentences marks an important contrast between Fregean and Millian treatments of the semantics of names.
This is not to say that, according to Frege, “Hesperus” and “Phosphorus” refer only to their ordinary senses in all of their occurrences in 4. Frege held that the complement clauses of factive attitude verbs need “to be taken twice over, with different referents, of which one is a thought, the other a truth value” (Frege 1892, p. 228). In other words, expressions in the complement clauses of ascriptions involving factive attitude verbs (i.e., those which create a presupposition of the truth of their complement clauses) refer to their ordinary senses for the purpose of determining those ascriptions’ truth-values and refer to their ordinary referents for the purpose of determining those ascriptions’ presuppositions.
It is worth pausing on this point because a popular approach to theorizing about validity in languages with vocabulary that create presuppositions has the consequence that validities are not closed under classical consequence. Call a sentence Strawson-valid (von Fintel 1999) if it is guaranteed to be true if its presuppositions are satisfied. Now consider “The king of France is a king”, “If the king of France is a king, then France has a king” and “France has a king”: the first two sentences are Strawson-valid, and they classically imply the third sentence, but the third sentence is not Strawson-valid. However, notice that any classical consequence of some Strawson-valid sentences will itself be Strawson-valid provided that the sentence’s presuppositions are at least as strong as the presuppositions of the sentences that classically imply it. Since the presuppositions of 4 are at least as strong as the presuppositions of 1 and 3 (given standard assumptions about presupposition projection) the pattern of presuppositions of sentences in our argument does not raise trouble for our appeal to the closure of validities under classical logic. We could also sidestep this issue by replacing 3 in our argument with “If the thought that Hesperus is Phosphorus is a true thought, then Hesperus is Phosphorus”, since “is a true thought”, unlike “know”, does not generate presuppositions. (Frege’s own views about truth as a predicate of thoughts are complex, but in respects that do not threaten the modified version of our argument; see Heck and May (2018).)
One might resist this argument by claiming that, despite the fact that every word in 4 would have the same reference in each of its occurrences if 4 were used to speak falsely, such uses would nevertheless involve a kind of equivocation because they would involve a mid-sentence shift in context. Compare “If today is Christmas, then tomorrow is boxing day”: if the clock struck midnight on December 26th at the moment “then” was uttered during a use of this sentence, then the falsity of that use would involve a kind of equivocation (since it would turn on the fact that “today” and “tomorrow” were uttered in different contexts on different days) despite the fact that every word in the sentence would have the same reference in each of its occurrences. Here is one way in which this contextualist proposal might be fleshed out. Suppose that context supplies a function from entities to senses, and that, when a name is used embedded under a single propositional attitude verb (as in the consequents of 1 and 4) it refers to the sense assigned by context to its ordinary referent. On this view, when 1 and 4 are used to speak falsely, the occurrences of “Hesperus” and “Phosphorus” in “Hammurabi knew that Hesperus is Phosphorus” must be in different contexts, since they refer to different senses despite having the same ordinary referent.
This proposal is Kaplanian to the extent that it upholds the validity of 1 and 4 by convicting their false uses of a kind of equivocation, and Fregean to the extent that it deploys a sense/reference distinction in the semantics of propositional attitude ascriptions. But it does not invoke the sort of equivocation Kaplan had in mind, nor does it fit with Frege’s way of thinking about the connection between sense and reference. Regarding Kaplan, note that the equivocation is not between two differently embedded occurrences of the same name (one of which refers to an ordinary object and the other to a sense) but rather between two occurrences of different names embedded under the same attitude verb. Regarding Frege, he did not think that the senses of proper names are context-sensitive in the way the proposal requires—he would not think that, as long as context is held fixed, there is a backward road from reference to sense. (Dorr (2014) argues that proponents of a Fregean semantics of attitude ascriptions should, pace Frege and most of his followers, accept something like the above contextualist proposal.)
How should Fregeans characterize the validity-relevant notion of non-equivocation? The simplest proposal would be to count a reading of a sentence not containing quotation marks as equivocation-free just in case, for every elementary expression in the sentence, there is a sense such that every occurrence of the expression embedded under an attitude verb (or in some other ‘indirect context’) refers to that sense and every occurrence of the expression not embedded under an attitude verb refers to the referent determined by that sense. A different proposal (for those who think that iterated attitude ascriptions require a hierarchy of senses) would be to say instead that, for every elementary expression in the sentence, there is a hierarchy of senses \(s_0,s_1,\dots\) such that each \(s_n\) is the referent determined by \(s_{n+1}\) and every occurrence of the expression in the scope of n attitude verbs refers to the referent determined by \(s_n\).
One salient dissimilarity here is that the pattern of distinct referents of “Phosphorus” in 1 is conventionalized, obviating the need for the sort of explicit indication of the intended reading that we just made above regarding 2; see Kripke (2011 [2008], p. 262, f.31).
For some recent work in this direction, see Bacon and Russell (2017) and Caie et al.(forthcoming).
References
Bacon, A., & Russell, J. S. (2017). The logic of opacity. Philosophy and Phenomenological Research
Caie, M., Goodman, J., Lederman, H. (forthcoming). Classical opacity. Philosophy and Phenomenological Research.
Carnap, R. (1947). Meaning and necessity: A study in semantics and modal logic. Chicago: University of Chicago Press.
Dorr, C. (2014). Transparency and the context-sensitivity of attitude reports. In M. Garcia-Carpintero & G. Martí (Eds.), Empty representations: Reference and non-existence (pp. 25–66). Oxford: Oxford University Press.
Frege, G. (1892). Über Sinn und Bedeutung. Zeitschrift für Philosophie und philosophische Kritik, 100, 25–50. Translated as “On sense and reference” by Max Black in Philosophical Review, 57(3), 209–230 (1948).
Heck, R. K., & May, R. (2018). Truth in Frege. In M. Glanzberg (Ed.), The Oxford handbook of truth (pp. 193–217). Oxford: Oxford University Press.
Kaplan, D. (1968). Quantifying in. Synthese, 19, 178–214.
Kripke, S. A. (2011). Frege’s theory of sense and reference: Some exegetical notes. In Philosophical troubles (pp. 254–291). Oxford University Press, 2011 [2008]. First published in Theoria, 74(3), 181–218.
von Fintel, K. (1999). NPI licensing, Strawson entailment, and context dependency. Journal of Semantics, 16(2), 97–148.
Acknowledgements
Thanks to Dave Chalmers, Cian Dorr, Kyle Landrum, Robbie Williams, Jack Woods and especially an anonymous reviewer for helpful comments on earlier drafts of this paper.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Goodman, J., Lederman, H. Sense, reference and substitution. Philos Stud 177, 947–952 (2020). https://doi.org/10.1007/s11098-018-1214-4
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11098-018-1214-4