1 This paper is a completely rewritten version of an unpublished paper, ‘The Varied Sorrows of Modality, Part II.’ I am indebted to several colleagues for information used in writing that paper, and for advice given on it once written, and I would like to thank them all - Gil Harman, Dick Jeffrey, David Lewis - even if the portions of the paper with which some of them were most helpful have disappeared from the final version. But I would especially like to thank Scott Soames, who was most helpful with the portions that have not disappeared.

2 Hintikka, J. ‘Is Alethic Modal Logic Possible?’, Acta Philosophica Fennica 35 (1982) 89–105Google Scholar, opening paragraph. In context it is clear this is a description, not an endorsement, of a widespread impression.

3 The most important of Quine's, presentations is ‘Reference and Modality,’ in From a Logical Point of View (Cambridge: Harvard University Press, three editions, 1953, 1961, 1980)Google Scholar. Citations of this twice-revised work here will be by internal section and paragraph divisions, the same from edition to edition. This work supersedes the earlier ‘The Problem of Interpreting Modal Logic,’ Journal of Symbolic Logic 12 (1947) 43-8. For commentary see Linsky, editor's introduction to Reference and Modality (Oxford: University Press 1971)Google Scholar, and ‘Reference, Essentialism, and Modality,’ therein 88-100. See also Føllesdal, D. ‘Quine on Modality,’ in Davidson, D. and Hintikka, J. eds., Words and Objections: Essays on the Work of W. V Quine (Dordrecht: Reidel 1969), 175-85CrossRefGoogle Scholar; and ‘Essentialism and Reference,’ in Hahn, L. E. and Schilpp, P. A. eds., The Philosophy of W. V Quine (La Salle: Open Court 1986), 97–113.Google Scholar

4 A theme in his reviews in the Journal of Symbolic Logic 11 (1946) 96-7 and 12 (1947) 95-6

5 See the last paragraph of the third section of ‘Reference and Modality,’ ending: ” … for if we do not propose to quantify across the necessity operator, the use of that operator ceases to have any clear advantage over merely quoting a sentence and saying that it is analytic.“

6 Contrast the opening section of ‘Reference and Modality,’ on knowledge and belief contexts, with the antepenultimate paragraph of the paper, beginning: “What has been said in these pages relates only to strict modality ….“

7 For a contemporary account deploring such tendencies, see Kneale, W. and Kneale, M.The Development of Logic (Oxford: Clarendon Press 1962), 628ffGoogle Scholar. Such tendencies are exemplified by the usage of all the participants in the exchange discussed in section II below.

8 For a less rough formulation, see Parsons, ‘Essentialism and Quantified Modal Logic,’ The Philosophical Review 78 35–52CrossRefGoogle Scholar, reprinted in Linsky, 73-87.

9 It may be worth digressing to mention that Quine's one and only contribution to the formal side of modal logic occurred in connection with this law, though the history does not always emerge clearly from textbook presentations. The earliest derivations of the law took an old-fashioned approach on which identity is a defined second-order notion, and on such an approach the derivation was anything but straightforward, and only went through for systems at least as strong as the second-strongest Lewis system S4. Quine was one of the first to note that on a modem approach with identity a primitive first-order notion, the derivation becomes trivial, and goes through for all systems at least as strong as the minimal normal system K. This is alluded to in passing in the penultimate paragraph of the third section of ‘Reference and Modality.’ For the original presentation see Barcan, R. (Marcus), ‘Identity of Individuals in a Strict Functional Calculus of Second Order,’ Journal of Symbolic Logic 12 (1947) 3–23CrossRefGoogle Scholar. For a modern textbook presentation see Hughes, G. E. and Creswell, M. J.An Introduction to Modal Logic (London: Methuen 1968), 190Google Scholar.

10 In the original paper where (17) was derived there were no singular terms but variables, and nothing was said about application to natural language. For an idea of the range of options formally available, see the taxonomy in Garson, J. ‘Quantification in Modal Logic,’ in Gabbay, D. and Guenthner, F. eds., Handbook of Philosophical Logic II (Dordrecht: Reidel 1984) 249–308.CrossRefGoogle Scholar

11 See Ackerman, ‘De Re Propositional Attitudes Toward Integers,’ in Shahan, R. W. and Swoyer, C. eds., Essays on the Philosophy of W. V. Quine (Norman: University of Oklahoma Press 1979)Google Scholar. Lectures of Kripke have brought this formerly underappreciated paper to the attention of a wider audience. See also Shapiro, S. ed., Intensional Mathematics (Amsterdam: North Holland 1985)Google Scholar; and especially Boolos, G.The Logic of Provability (Cambridge: Cambridge University Press 1993), xxxiv and 226Google Scholar.

12 Workers in the cited fields have in effect suggested that something like indices can serve as canonical terms for more fine-grained intensional analogues of recursive sets and functions. But these too would be very special objects. The best discussion of these matters known to me is in some papers of Leon Horsten still at the time of this writing ‘forthcoming.’

13 Whose published proceedings make up one issue of Acta Philosophical Fennica 16 (1963), and include not only Kripke's, ‘Semantical Considerations on Modal Logic,’ 83–94Google Scholar, but also Hintikka's, ‘Modes of Modality,’ 65–82Google Scholar.

14 ‘Naming and Necessity: Lectures Given to the Princeton University Philosophy Colloquium, January, 1970,’ in Davidson, D. & Harman, G. eds., Semantics of Natural Language (Dordrecht: Reidel 1972), 253–355 and 763-9CrossRefGoogle Scholar; reprinted with a new preface (Cambridge: Harvard University Press 1980).

15 ‘Modalities and Quantification,’ Journal of Symbolic Logic 11 (1946) 33-64; Meaning and Necessity: A Study in Semantics and Modal Logic (Chicago: University of Chicago Press 1947).

16 Journal of Symbolic Logic 12 (1947) 139-41; with elaboration in his paper ‘Modality and Description,’ Journal of Symbolic Logic13 (1948) 31-7. Smullyan's priority for his particular response to Quine has been recognized by all competent and responsible commentators. See note 15 in Linsky, ‘Reference, Essentialism, and Modality,’ and Føllesdal, ‘Quine on Modality,’ 183.Google Scholar

17 Thus the items are: (i) the compendium, ‘Modalities and Intensional Languages’; (ii) the ‘Comments’ later retitled ‘Reply to Professor Marcus’; and (iii) the edited ‘Discussion.’ They appear together in the official proceedings volume, Wartofsky, M. W. ed., Proceedings of the Boston Colloquium for the Philosophy of Science 1961/1962 (Dordrecht: Reidel 1963), 77–96CrossRefGoogle Scholar (compendium), 97-104 (commentary), 105-16 (edited discussion). The same publisher had printed them in 1962 in Synthese in a version that is textually virtually identical down to the placement of page breaks, (i) and (ii) in a belated issue of the volume for 1961, and (iii) in an issue of the volume for 1962. (There have been several later, separate reprintings of the different items, but these incorporate revisions, often substantial.) Two of the present editors of Synthese, J. Fetzer and P. Humphreys, have proposed publishing the unedited, verbatim transcript of the discussion, with a view to shedding light on some disputed issues of interpretation; but according to their account, one of the participants, Professor Marcus, has objected to circulation of copies of the transcript or the tape.

18 These are the closing words of Kripke, ‘Is There a Problem about Substitutional Quantification?’ in Evans, G. and McDowell, J. eds., Essays in Semantics (Oxford: Oxford University Press 1976)Google Scholar. The fallacy recurs again and again in other contexts in the literature. See Copeland, B. J. ‘On When a Semantics Is Not a Semantics: Some Reasons for Disliking the Routley-Meyer Semantics for Relevance Logic,’ Journal of Philosophical Logic 8 (1979) 399–413.CrossRefGoogle Scholar

19 Notably the Barcan or Carnap-Barcan formulas, which give formal expression to F. P. Ramsey's odd idea that whatever possibly exists actually exists, and whatever actually exists necessarily exists. (The ‘Barcan’ label is the more customary, the ‘Carnap-Barcan’ label the more historically accurate according to Cocchiarella, N. ‘Philosophical Perspectives on Quantification in Tense and Modal Logic,’ in Gabbay, D. and Guenthner, F. eds., Handbook of Philosophical Logic II (Dordrecht: Reidel 1984) 309-53CrossRefGoogle Scholar, which also explains the connection with Ramsey.) If these formulas are rejected, one must distinguish a thing's having a property necessarily (for every possible world it exists there and has the property there) from its having the property essentially (for every possible world, if it exists there, then it has the property there). I have slurred over this distinction so far, and will for the most part continue to do so.

20 As shown by examples in the opening section of ‘Reference and Modality.’ This point seems to be conceded even by some who otherwise take an uncritically positive view of the compendium, as in the review by Forbes, G.Notre Dame Journal of Formal Logic 36 (1995) 336-9Google Scholar. The last sections of ‘Is There a Problem about Substitutional Quantification?’ in effect point out that the claim that the ordinary language ‘there is’ in its typical uses is a ‘substitutional quantifier’ devoid of ‘ontological commitment’ is absurd, since ‘ontological commitment’ is by definition whatever it is that the ordinary language ‘there is’ in its typical uses conveys.

21 “What I've been talking about is quantification, in a quantificational sense of quantification, into modal contexts, in a modal sense of modality,” Wartofsky, 116

22 ‘Essentialism in Modal Logic,’ Noûs 1 (1967) 90-6. And about the same time we find even the usually acute Linsky (editorial introduction, 9) writing: “Terence Parsons bases his search for the essentialist commitments of modal logic on Kripke's semantics, and he comes up (happily) empty-handed …. He finds modal logic uncontaminated.” The continuation of this passage better agrees with Parsons’ own account of his work and its bearing on Quine's critique.

23 See Wartofsky, 90-2. It is just conceivable that this is deliberate exaggeration for effect, a rhetorical flourish rather than a serious exegetical hypothesis. ‘Essentialism in Modal Logic’ cites some other authors who have written in a similar vein about the example.

24 And “I did not say that it could ever be deduced in the S-systems or any systems I've ever seen,” Wartofsky, 113. Despite these forceful remarks, the understanding of Quine's views has not much improved in the later ‘Essentialism in Modal Logic.’

25 An earlier paper by the author of the compendium, ‘Extensionality,’ Mind 69 (1960) 55-62, reprinted in Linsky, ed., gives a more concise statement of the response in its last paragraph, where a footnote acknowledges the author's teacher Fitch, Frederic The latter, in his ‘The Problem of the Morning Star and the Evening Star,’ Philosophy of Science 16 (1949) 137-41CrossRefGoogle Scholar, and ‘Attribute and Class,’ in Farber, M. ed., Philosophic Thought in France and the United States (Buffalo: University Press 1950), 640-7Google Scholar, acknowledges Smullyan. (See footnote 4 in the former, footnote 12 in the latter, and the text to which they are attached.)

26 The major one being Weiss, P. ed., The Encyclopedia of Philosophy, six volumes, (New York: MacMillan 1967)Google Scholar, and the minor one the collection of survey articles, Klibansky, R ed., Contemporary Philosophy, four volumes (Firenze: Editrice Nuova Italia 1968)Google Scholar. The former contains Prior, ‘Logic, Modal,’ V 5-12; while the latter contains Marcus, ‘Modal Logic,’ I 87-101. The conference talks are to be found in the previously cited proceedings, Marcus’ ‘Classes and Attributes in Extended Modal Systems,’ 123-36, and Prior's ‘Is the Concept of Referential Opacity Really Necessary?,’ 189-99. Another advocate of closely related ideas has been J. Myhill.

27 Let me not fail to cite chapter and verse myself. For the most relevant pages of the most recently reprinted work, see The Philosophy of Logical Atomism, (La Salle: Open Court 1985), 113-5.

28 Reply to Sellars, in Davidson and Hintikka, 338. This 1969 formulation is the earliest adequate one known to me, the rebuttal even in the 1961 version of ‘Reference and Modality’ being inadequate.

29 As was pointed out in Kripke's last few remarks in the discussion at the colloquium. Quine seems to accept the observation in his last remark. Marcus had apparently ceased to follow by this point.

30 Fitch, ‘The Problem of the Morning Star and the Evening Star,’ explicitly claims that Quine's contention is “clearly” false if the key expression are taken to be names.

31 Prior, ‘Is the Concept of Referential Opacity Really Necessary?,’ 194-5Google Scholar. Prior was from Balliol, and I have heard it asserted- though I cannot confirm it from my own knowledge - that there was a tradition of setting examples of this kind on undergraduate examinations at Oxford in the 1960s.

32 ‘Classes and Attributes in Extended Modal Systems,’ 132. Note the characteristically Carnapian expression “meaning postulates.“

33 For the published version, too familiar to bear quoting again, see Wartofsky, 115. This is one of the parts of the discussion where comparison with the verbatim transcript could be most illuminating. It is a shame that the scholarly public should be denied access to so significant an historical document.

34 Wartofsky, 83-4. This passage has sometimes been misleadingly cited in the later literature as if it were unambiguously about ordinary names in ordinary language.

35 Wartofsky, 101. Quine surely means that (12a’) is not just a linguistic empirical discovery but a properly astronomical empirical discovery. By contrast, Marcus in Wartofsky, 115, distinguishes “such linguistic” inquiry as leads to discoveries like (12a’) from “properly empirical” methods such as lead to discoveries about orbits.

36 The quotation from Quine is from ‘Reference and Modality,’ antepenultimate paragraph. The work of Føllesdal where it is quoted is ‘Quantification into Causal Contexts,’ in Cohen, R.S. and Wartofsky, M.W. eds., Boston Studies in the Philosophy of Science, II (New York: Humanities Press 1965), 263-74Google Scholar; reprinted in Linsky, ed., 52-62. Føllesdal's final footnote suggests that “causal essentialism” is better off than “logical essentialism,” and that Quine's own proposal to treat dispositions as inhering structural traits of objects is a form of “causal essentialism.“

37 Klibansky, 91ff. This echoes Fitch, ‘Attribute and Class,’ where it is said (553) that: “Smullyan has shown that there is no real difficulty if the phrase [sic] ‘the Morning Star’ and ‘the Evening Star’ are regarded either as proper names or as descriptive phrases in Russell's sense.” The syntactic ambiguity in this last formulation as to whether “in Russell's sense” is supposed to modify “proper names” as well as “descriptive phrases” matches the ambiguity in the formulation quoted earlier as to whether “Russellian” is supposed to modify “the distinction between proper names and descriptions” as well as “theory of descriptions.” The ambiguity is appropriate, since the theory of names in question is neo-Russellian.

38 Though this may not yet have been made clear at the time the encyclopedia article was written, since the formulation of the rebuttal I have quoted dates from two years later.

39 See Føllesdal, §17, 96ff. of Referential Opacity and Modal Logic, doctoral dissertation, Harvard, 1961; reprinted as Filosofiske Problemer, 32 (Oslo: Universitetsforlaget 1966). Church, review in the Journal of Symbolic Logic 15 (1950) 63. Both address Smullyan and Fitch.

40 For work on difficulties with the Fregean theory in the 1950s and early 1960s, see the discussion in ‘Naming and Necessity,’ and J. Searle's article on ‘Proper Names and Descriptions’ in Weiss, VI 487-91. The doctrines in ‘Naming and Necessity’ were first presented in seminars in 1963-64, and whereas that work apologizes for being spotty in its coverage of the literature of the succeeding years, it is pretty thorough in its discussion of the relevant literature (work of P. Geach, P. Strawson, P. Ziff, and others) from the immediately preceding years. (Searle discusses work of yet another contributor, Elizabeth Anscombe.)

41 In, Realism, Mathematics and Modality (Oxford: Basil Blackwell 1989), chapter 3. Field also cites several expressions of the same or related views from the earlier literature, and such citations could in a sense be carried all the way back to the ‘principle of predication’ in Wright, G. H. vonAn Essay in Modal Logic (Amsterdam: North Holland 1951)Google Scholar.

42 In Entities and Indices (Dordrecht: Kluwer 1990). Cresswell also cites several expressions of the same or related views from the earlier literature, and such citations could in a sense be carried all the way back to Lewis, D. K. ‘Anselm and Actuality,’ Noûs 4 (1970) 175-88CrossRefGoogle Scholar. This is the earliest relevant publication known to me, but its author has suggested that there was very early unpublished work on the topic by A. P. Hazen and by D. Kaplan. The parallel phenomenon for tense in place of mood was noted even earlier by P. Geach.

43 See Prior, Past, Present, and Future (Oxford: Clarendon Press 1967)CrossRefGoogle Scholar, chapter VII, and among later work Thomason, R. H. ‘Combinations of Tense and Modality,’ in Gabbay and Guenthner, 135-65CrossRefGoogle Scholar. The purely modal part is also S5 for virtually all the workers there cited, as well as later ones like A. Zanardo.

44 ‘A Theory of Conditionals,’ in Rescher, N. ed., Studies in Logical Theory (Oxford: Basil Blackwell 1968), 98–112Google Scholar. This feature becomes even more prominent in later work on the same topic by D. K. Lewis and others.

45 Unfortunately this comes in the form of a review of a book by a third party, and is subject to the limitations of such a form. The third party is Linsky, ; the book is his Names and Descriptions (Chicago: University of Chicago Press 1977)Google Scholar; the review is by Marcus, Philosophical Review 87 497–504CrossRefGoogle Scholar. The three quotations to follow come from 498, 501, and 502-3.

46 In this connection mention may be made of one serious historical inaccuracy of a kind extremely common when authors quote themselves from memory decades after the fact— to be found in the book review, where it is said that the compendium maintained “that unlike different but coreferential descriptions, two proper names of the same object were intersubstitutable in modal contexts“ (502). In actual fact, in the compendium it is repeatedly asserted that two proper names of the same object are intersubstitutable in all contexts.

47 Frege's Puzzle (Cambridge, MA: MIT Press 1986). While the early Marcus followed Smullyan, the later Marcus has developed in response to Kripke an idiosyncratic theory that may be described as intermediate in degree of Russellianism between Salmon's and Smullyan's. See her ‘Some Revisionary Proposals about Belief and Believing,’ Philosophy and Phenomenological Research 50 (Supplement) 133-53.

48 For Kripke's rejection of this view, see the closing paragraphs of the preface to the second edition of ‘Naming and Necessity.’

49 Bull, R. A. and Segerberg, K. ‘Basic Modal Logic,’ in Gabbay and Guenthner, 1–88CrossRefGoogle Scholar. Other articles in the same work, some of which I have already cited, do recognize the importance of the distinction.

50 It would be out of place to enter into technicalities here. See Burgess, J. and Rosen, G.A Subject with No Object (Oxford: Oxford University Press 1997)Google Scholar.

51 In actual fact, on Kripke's theory, for instance, a name can be given to any object that can be described, not excluding mathematical objects. But again see Burgess and Rosen. (The theory of P. Geach probably deserves and the theory of M. Devitt certainly deserves the label ‘causal,’ and does have nominalistic implications.)

52 For comparatively moderate instance see the review by Lavine, S.British Journal for the Philosophy of Science 46 (1995) 267-74Google Scholar.

53 For an extreme instance see Hintikka, J. and Sandu, G. ‘The Fallacies of the New Theory of Reference,’ Synthese 104 (1995) 245-83CrossRefGoogle Scholar. This work acknowledges no important differences among: (i) the neo-Russellian theory of Smullyan as expounded by the early Marcus (which incidentally is erroneously attributed to Marcus as something original, ignoring the real authors Smullyan and Russell); (ii) theories adopted in reaction to Kripke by the later Marcus; and (iii) the theory Of Kripke.

54 In context, what is said to be right is specifically the rebuttal to Smullyanism on names quoted earlier. See ‘Naming and Necessity,’ 305.