Abstract
This article reconsiders Kripke’s (1977, in: French, Uehling & Wettstein (eds) Contemporary perspectives in the philosophy of language, University of Minnesota Press, Minneapolis) pragmatic, univocal account of the attributive-referential distinction in terms of a metalinguistic apparatus consisting of semantic reference and speaker reference. It is argued that Kripke’s strongest methodological argument supporting the pragmatic account, the parallel applicability of the apparatus to both names and definite descriptions, is successful only if descriptions are treated as designators in both attributive and referential uses. It is not successful if descriptions are treated à la Russell, contrary to what is often assumed. Thus a third theoretical option for the semantic analysis of definite descriptions arises, neglected by both supporters and opponents of Russell: a univocal, referentialist analysis of descriptions in conjunction with a pragmatic account of the attributive-referential distinction. Contrary to Kripke, and to much of the literature, it is noted that not all so-called referential uses involve implicatures. In the course of the argument Kripke’s innovative apparatus is subjected to improvements and fine-tunings. Also, some general critical comments are made about analogical reasoning, on which Kripke’s argument is partly based. This leads to a clarification of the fundamental notion of speaker reference. The paper concludes with reflections on the challenge to and need of systematic empirical evidence in this field, a desideratum noted by Kripke and still not met.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
As opposed to a same-category ambiguous word like ‘bank’, which disambiguates into different noun-phrases. An uncontroversial example of cross-category ambiguity is ‘leaves’.
This was anticipated in Grice (1969), an article curiously not mentioned by Kripke.
Such intuitions are also problematic for more specific reasons. (a) Expectation versus surprise that a word can ‘be disambiguated by separate and unrelated words in some other languages’ is to count as a kind of a priori data, ‘independently of any empirical investigations’ (Kripke 1977, p. 19). However, from the fact that Saul says he would be surprised that p, while Mark does not, nothing decisive follows whether or not p. The ambiguist theorist won’t expect to be surprised, and neither will be all those many speakers with no opinions about or interest in our topic. Semantics is to be kept free of such psychologism. In addition, the failure to discover a language in which AttRef is captured by different words could be accommodated by auxiliary hypotheses. Thus it is not obvious that ambiguity has to be an accidental feature of a language (Szabó 2000, p. 54); it may well be that there is a systematic connection between attributive ‘the’ and referential ‘the’, such as the possibility that the attributive use grew out of the referential use (demonstrative reference). Also, even if we discovered some other language with two words for ‘the’, this would be of semantic relevance only if we could establish on independent grounds how the meaning of those two words relate, i.e. disambiguate them in our language. Hence, there is no need for a trans-linguistic detour. Ambiguity is an intra-linguistic property, independent of the contingent existence of other languages; see also Kripke himself on this (1977, p. 26, fn. 29). (b) Kripke considers the following exchange: ‘A: Her husband is kind to her. B: No, he isn’t. The man you’re referring to isn’t her husband’, where the man referred to in context is her kind lover, and her real husband is cruel (Kripke 1977, p. 21). Kripke claims the quantifier theory can and the equivocal theory can’t explain this dialogue. For Donnellan must claim that ‘he’ links back to ‘her husband’, which is used referentially, and hence ‘he’ must refer to the lover, by definition, since a description used referentially has its own meaning, distinct from the attributive one. But clearly, ‘he’ picks out the husband, hence only the Russellian theory, which is not restricted by the ambiguous meaning of the definite description, can explain the dialogue. But this is not so. Quite generally, there is no reason why the use of an anaphoric pronoun must match the use or meaning of a preceding designator. Compare: ‘Fog lay on the city. And so it did on his soul.’ Here ‘it’ connects back to ‘fog’, but is used in the metaphorical, not the initial literal sense of the former (see also Ramachandran (1996, Sect. 9) on this). (c) Kripke considers the judgements, or rather ‘intuitions’, about the truth-value of statements containing descriptions whose semantic referent diverges from the speaker referent (‘Her husband is kind to her’), or empty descriptions (‘The king of France is bald’). He is inclined to assign Russellian truth conditions to such statements. But the debate between referentialism and the quantifier theory cannot be thus decided. For a version of referentialism is compatible with Russellian truth conditions, as discussed below.
See Kripke (1977, pp. 14f., 18). The pairs ‘what is said’/‘what is meant’ and ‘semantic reference’/‘speaker reference’ are not synonymous, but functionally equivalent. The first pair applies to the propositional level, the second to the sub-propositional level. In what follows they will be sometimes used interchangeably.
In general, we choose to implicate ‘q’ precisely by saying ‘p’, and this is one complex speech act, the deliberate result of one intention, not the coincidence of two intentions. Otherwise, ‘Gettier’-like coincidences between two intentions could be constructed, causing havoc to any simple and intuitive definition of implicature.
Kripke mentions implicatures only cursorily, but his references to Grice’s writings are unambiguous.
The notation ‘ϕ(d)’, the form of an elementary subject-predicate (or argument-function) sentence, by which the two distinctions are introduced, is Kripke’s and a case in point. See Kripke (1977, p. 15).
Recanati (1993, pp. 244ff.).
Kripke comes close to realising that at least the Smith-Jones case does not involve implicatures. He writes that this is a case of misidentification of two people. Given that misidentifications do not trigger, by default, implicatures, there is a small step here to establishing that the Smith-Jones case does not involve implicatures. But this is not a conclusion Kripke draws. Instead, he points out an asymmetry between names and descriptions, contrasting misidentifications arising through complex uses of names with corresponding referential uses of descriptions, where no such misidentification need arise (Kripke 1977, p. 25, fn. 26). However, what matters here is the more general phenomenon of misapplication of a term, whether name or description. Since misapplications certainly do not involve implicatures by default, my point is not affected by Kripke’s observation about the said asymmetry.
To illustrate an infelicitous case in the example above: the real Stalin has eliminated Professor Quirk and masquerades as him at the department meeting, hence SpRef(‘d’) = SemRef(‘d’) (to make this a more realistic case, think yourself a member of Moscow university during the Great Purge of 1936–1938). The utterance simply turns out to be literally true, if unintended, with no implicature arising.
To illustrate the infelicitous case here: imagine I am under the misapprehension that the man occupying the throne is the usurper; in fact, he is the real king, having replaced the usurper long ago, but since they look very similar, I never noticed the difference. I see the man occupying the throne in his countinghouse and say to the guards: ‘The king is expecting me’. Hence, SpRef(‘d’) = SemRef(‘d’). From the guards’ perspective my utterance is felicitous, but not from my own, since I had intended my utterance to be about the usurper, and thus to let me through to him. But instead, it is taking me to the wrong person, the real king.
But even if they didn’t, no implicature would necessarily be present. People could be accustomed to refer to the usurper as ‘the king’, or even ‘the real king’, out of cynicism, conformism, fear etc.
See Reid (1853, p. 237) on this particular question.
See Kripke (1977, p. 16). This sounds paradoxical, as both banning X from Y and using Z instead of X still involves reference to X.
Bach (2004) has recently claimed that only a quantificationalist interpretation of definite descriptions can explain their referential uses. But nothing he adduces in favour of this refutes the possibility of referentialism. See my forthcoming “Bach on Definite Descriptions”.
As opposed to cross-category ambiguity, which the Russellian usually imputes to his opponent. The details of Sainsbury’s account do not concern us here. See e.g. Sainsbury (2004, p. 379f.).
However, as established in Sect. 4, there can also be implicature-free complex cases modelled by Def 4* . One such case for definite descriptions would be Donnellan’s aforementioned example about the king in his countinghouse.
A similar example would be ‘Look, our kindest colleague is back’. A quite different case might be one in which ‘our kindest colleague’ is used ‘echoically’ (see Sperber and Wilson (1986, p. 238f.), e.g. in a situation in which some newcomers misled by their first impression of Quirk describe him as the kindest colleague. Somebody knowing Quirk’s true character could echo this ironically with ‘You will soon be delighted by our kindest colleague’. Such a case seems to require a mixed analysis, in terms of both Def4 and Def4*. For the irony is meant to echo a previous infelicitous referential use modelled straightforwardly by Def4. Nevertheless, we would also need Def4* for this case, since ‘our kindest colleague’ appears to be still employed with a referential intention and in the belief that SpRef(‘d’) ≠ SemRef(‘d’). However, this all depends on whether the description is indeed employed with a referential intention, or rather only with a metalinguistic commenting intention. If it is the latter, we might well have to exclude such uses from the apparatus of Def1–4/4*, since that contains referential intentions as a necessary ingredient. The resolution of this matter must be left for another occasion.
An infelicitous case can be constructed on the analogy of the case in footnote 14.
This does not mean that all definite descriptions in subject position need be designators. Think of obvious counterexamples like ‘The clever linguist is better than any philosopher’ or ‘The average man in Britain suffers of depression’. But the apparatus of Defs1–4* can serve as a filter for the considerable number of possible uses of descriptions in subject position. Only contexts which can be modelled by the apparatus will count as cases in which the description is a designator in the sense of univocal referentialism.
The apparatus of speaker reference and semantic reference, and of simple and complex uses of designators […] is applicable to all languages (Kripke 1977, pp. 18, 21).
At least not of the relevant kind. This qualification might be needed, for some semantic analyses of quantifiers can be construed as assigning semantic referents to them, namely concepts of concepts (Frege) or sets of sets. But then why not resolve the debate by saying that definite descriptions are referring expressions after all, except that they don’t refer to what they are ordinarily taken to refer to (e.g. persons), but rather to abstract second-order entities? This interesting line of thought will not be pursued here.
Do we even know what an attributive use of a quantifier (phrase) is?
I owe this example to Mike Inwood.
Displayed in categorial grammar, in ‘ϕ(Q)’ ‘ϕ’ is an expression of the type S/N (first-order predicate), needing a name as an argument to yield a sentence, while ‘Q’ is an expression of the type S/(S/N) (second-order predicate), needing itself an S/N expression as an argument to yield a sentence.
See, symptomatically, Bach: ‘The speaker thinks of a certain object, takes that object to be the F, and uses “the F” to refer to it’ (Bach 2004, p. 203).
In the Fregean tradition. Term logic has no reservations in assigning first-order semantic values to quantifiers.
An additional difficulty: On a Fregean view of sentential reference, we would have to admit a strange duality, with sentences having truth-values as semantic referents and some other entities (Professor Quirk, etc.) as speaker referents. This would preclude that sentences can have speaker referents ontologically on a par with their semantic referents (truth-values), unlike other designators, e.g. proper names. This anomaly would threaten a uniform semantic/speaker reference distinction.
And we could add: the speaker chooses to utter ‘Q ϕ(x)’ precisely because ‘Q’ has no, neither semantic nor speaker, reference. For it is this choice that achieves the roundabout, insinuating effect of the speech act.
This is not meant to be only a point about the meta-language. Accounting for the fact that S is referring to e by implicating something about e falls itself, like every speech-act involving implicatures, under the availability principle. The speaker herself must be able to give such an account. It is therefore questionable whether a language allowing only for speaker reference, but no semantic reference, is possible, at least if speaker reference is understood as in the present article.
It does not attach to the implicature, since an implicature is not an expression, but an inference arising from the contrast between principal utterance and implicatum. It does not attach to the implicatum either, since the implicatum is not expressed in the speech act. We could, maybe, say that roundabout speaker reference attaches to the expression of the implicatum, if it were uttered. But this complicates the framework beyond reasonable necessity, and it brings it no closer to the original apparatus.
As Kripke describes the application of the notion of speaker referent to cases ‘where existential quantification rather than designation is involved’ (Kripke 1977, pp. 15, 17).
These points must be duplicated to account for appropriate pragmatic derivations on the audience’s part.
But it is even disputed whether scalar inferences, the original phenomenon, involve implicatures at all. See Noveck and Sperber (2007).
I am grateful to the referee of this journal for pointing this out, and the following comment:
‘[p]resumably any cross-linguistic descriptions are going to presuppose a particular analysis, and since these are done by linguists these would by and large assume a referentialist account. And corpus studies that relied on data that has already been syntactically and semantically tagged would also mean that a particular analysis was assumed’.
For a paradigmatic example see the elaborate schematic models of such derivations hypothesised in Neale (1990, p. 89f.).
References
Bach, K. (1987). Thought and reference. Oxford: Clarendon Press.
Bach, K. (2004). Descriptions: Points of reference. In M. Reimer & A. Bezuidenhout (Eds.), Descriptions and beyond. Oxford: Oxford University Press.
Bach, K. (2006). The top 10 misconceptions about implicature. In B. Birner & G. Ward (Eds.), Drawing the boundaries of meaning: Neo-Gricean studies in pragmatics and semantics in honour of Lawrence R. Horn. Amsterdam: John Benjamins.
Bach, K. (2007a). Referentially used descriptions: A reply to Devitt. European Journal of Analytic Philosophy, 3, 2.
Bach, K. (2007b). The main bone of contention. European Journal of Analytic Philosophy, 3, 2.
Burge, T. (1974). Truth and singular terms. Noûs, 8, 309–325.
Clark, H. H., et al. (1983). Common ground and the understanding of demonstrative reference. Journal of Verbal Learning and Verbal Behavior, 22, 245–258.
Devitt, M. (2004). The case for referential descriptions. In M. Reimer & A. Bezuidenhout (Eds.), Descriptions and beyond. Oxford: Oxford University Press.
Devitt, M. (2007a). Referential descriptions and conversational implicatures. European Journal of Analytic Philosophy, 3, 2.
Devitt, M. (2007b). Referential descriptions: A note on Bach. European Journal of Analytic Philosophy, 3, 2.
Donnellan, K. (1966). Reference and definite descriptions. Philosophical Review, 75, 281–304.
Evans, G. (1985). Collected papers. Oxford: Oxford University Press.
Grice, H. P. (1969). Vacuous names. In D. Davidson & J. Hintikka (Eds.), Words and objections: Essays on the work of W. V. Quine. Dordrecht: Reidel Publishing.
Kaplan, D. (1969). Quantifying in. In D. Davidson & J. Hintikka (Eds.), Words and objections: Essays on the work of W. V. Quine. Dordrecht: Reidel Publishing.
Kripke, S. (1977). Speaker’s reference and semantic reference. In P. French, T. Uehling, & H. Wettstein (Eds.), Contemporary perspectives in the philosophy of language. Minneapolis: University of Minnesota Press.
McCulloch, G. (1989). The game of the name. Introducing logic, language, and mind. Oxford: Oxford University Press.
Neale, S. (1990). Descriptions. Cambridge, Mass: MIT Press.
Noveck, I. A. (2001). When children are more logical than adults: Investigations of scalar implicature. Cognition, 78, 165–188.
Noveck, I. A., & Reboul, A. (2008). Experimental pragmatics: A Gricean turn in the study of language. Trends in Cognitive Sciences, 12(11), 425–431.
Noveck, I. A., & Sperber, D. (2007). The why and how of experimental pragmatics: The case of ‘Scalar inferences’. In N. Burton-Roberts (Ed.), Pragmatics. Basingstoke: Palgrave.
Papafragou, A., & Musolino, J. (2003). Scalar implicatures: Experiments at the semantics-pragmatics interface. Cognition, 86, 253–282.
Ramachandran, M. (1995). Methodological reflections on two Kripkean strategies. Proceedings of the Aristotelian Society, XCV, 67–81.
Ramachandran, M. (1996). The ambiguity thesis versus Kripke’s defence of Russell. Mind & Language, 11, 371–387.
Recanati, F. (1993). Direct reference: From language to thought. Oxford: Blackwell.
Reid, T. (1853). Of analogy. In W. Hamilton (Ed.), Essays on the intellectual powers of man. Edinburgh: MacLachlan & Stewart.
Reimer, M. (1998). Donnellan’s distinction/Kripke’s test. Analysis, 58, 89–100.
Sainsbury, R. M. (2004). Referring descriptions. In M. Reimer & A. Bezuidenhout (Eds.), Descriptions and beyond. Oxford: Oxford University Press.
Sainsbury, R. M. (2005). Reference without referents. Oxford: Oxford University Press.
Salmon, N. (1991). The pragmatic fallacy. Philosophical Studies, 63, 83–97.
Sperber, D., & Wilson, D. (1986). Relevance: Communication and cognition. Oxford: Blackwell.
Szabó, Z. G. (2000). Descriptions and uniqueness. Philosophical Studies, 101, 29–57.
Wettstein, H. (1981). Demonstrative reference and definite descriptions. Philosophical Studies, 40, 241–257.
Acknowledgements
The author would like to thank for comments and inspiration Hanoch Ben-Yami, Timothy Chan, David Cram, Jonathan Dancy, Anna Dimitríjevics, Dorothy Edgington, Michael and Fiona Ellis, Hanjo Glock, Steven Hall, Mike Inwood, Joel Katzav, John Preston, Murali Ramachandran, Severin Schröder, Stephen G. Williams, Tim Williamson, Zoltán Gendler Szabó, and the anonymous referee of this journal.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Kanterian, E. Kripke’s metalinguistic apparatus and the analysis of definite descriptions. Philos Stud 156, 363–387 (2011). https://doi.org/10.1007/s11098-010-9608-y
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11098-010-9608-y