1 Some of the basic ideas in this paper are presented in a section of my paper ‘Reference, Generality, and Anaphora,’ to appear in a volume in honor of Keith Donnellan, edited by Paolo Leonardi. However, the version in the present paper extensively emends and extends those ideas, and I believe that the presentation is significantly clearer. It certainly contains a more detailed investigation of some of the phenomena in question.

2 Evans, Gareth ‘Pronouns, Quantifiers, and Relative Clauses (I),’ Collected Papers (New York: Oxford University Press 1985), 114-5Google Scholar. Evans uses the example “Just one man drank champagne, and he was ill.“

3 Neale, StephenDescriptions (Cambridge, MA: MIT Press 1990). See especially chapters 5 and 6.Google Scholar

4 A node A of a phrase structure tree c-commands a node B iff (i) A does not dominate B, (ii) B does not dominate A, and (iii) the first branching node that dominates A also dominates B. So, if S is a sentence containing an anaphor and its antecedent, the antecedent c-commands the anaphor iff the node of the phrase structure tree for S that dominates the antecedent c-commands the node that dominates the anaphor. (There are alternative conceptions of c-command, but their differences would not affect our discussion.) See, for example, Haegman, LilianeIntroduction to Government and Binding (Oxford: Basil Blackwell 1991), 198.Google Scholar

5 In speaking of ‘classical’ DRT, I am following the (admittedly loose) usage in Chierchia, GennaroDynamics of Meaning: Anaphora, Presupposition, and the Theory of Grammar (Chicago: University of Chicago Press 1995)CrossRefGoogle Scholar. The paradigms of the work he is referring to with this label are found in Kamp, H. ‘A Theory of Truth and Discourse Representation’ in Groenendijk, J.Janssen, T. and Stokhof, M. eds. Formal Methods in the Study of Language (Amsterdam: Mathematical Centre 1981), 277–322Google Scholar, and Heim, I.The Semantics of Definite and Indefinite Noun Phrases (Ph.D. dissertation, the University of Massachusetts, Amherst, 1982)Google Scholar. The dissertation was published in 1989 by Garland Press, New York.

6 In classical DRT, a rule of Existential Closure was widely adopted that assigned existential force to indefinites (and other variables) that are not otherwise bound by a quantifier in the discourse.

7 For a discussion and defense of this kind of conception, see, for instance, Fiengo, Robert and May, RobertIndices and Identity (Cambridge, MA: MIT Press 1994)Google Scholar, especially chapter 1. Also, I should note that my simple notation of underlining becomes immediately inadequate as soon as one considers any example involving two or more such anaphoric connections. More adequate and more commonly employed is some system of indexing the noun phrases and anaphors, as they appear in the relevant chains. An extensive discussion of the theory of indices is presented in Fiengo and May, and their discussion brings out the conceptual and other complexities that are implicated in a full theory of this topic. I have used underlining only because it seemed much easier to read and follow, given the examples studied in this paper.

8 For example, Reinhart, Tanya in Anaphora and Semantic Interpretation (London: Croon Heim 1983)Google Scholar rejects relations of obligatory co-reference induced by the linguistic meanings of the relevant sentences or discourse fragments. She attempts to give a ‘pragmatic’ account of the character of the antecedent-anaphor linkages that I am here viewing as semantically induced.

9 Evans,96

10 For a very clear explication of restricted quantifiers, see Neale, 38-47.

11 Geach, PeterReference and Generality (Ithaca: Cornell University Press 1962), 128Google Scholar

12 Essentially the same diagnosis is given by Soames, Scott ‘Direct Reference, Propositional Attitudes, and Semantic Content,’ Philosophical Topics 15 (1987) 47–87CrossRefGoogle Scholar.

13 Obviously, the concept of ‘a case’ employed here and in the rest of the paper would require much more careful discussion. In particular, if one moves beyond the relatively simple examples I examine, the structure of a case can become quite complicated, but I do not believe that those complications change the basic picture I am attempting to draw.

14 For an excellent discussion of the concept of ‘impliciture’ and its applications,see Bach, Kent ‘Conversational Impliciture’, Mind and Language 2 (1987) 124-62.Google Scholar

15 In the second part of her dissertation (cited in n. 5), Irene Heim introduced a kind of ‘file change’ semantics that has been developed by various writers and that continues to be influential. Although I will not explore the similarities and differences here, her ‘files’ and my ‘cases’ are introduced to play a similar role, i.e., to register content that has been introduced earlier in a discourse and stored among the presuppositions of the discourse- content that may then be exploited in helping to determine the truth conditions of utterances that occur in the subsequent discourse. See also the discussion of this and related questions in Chierchia's Dynamics of Meaning.

16 Neale, 241-51

17 Neale, 234-5

18 These remarks are meant to do no more than give a suggestive picture of the role of the disjuncts in the antecedents of the conditionals under study. Clearly, the relevant notion of ‘a supposition’ and of ‘truth under a supposition’ would require extensive treatment. (Although I believe that my remarks are compatible with a number of possible approaches to these matters.) This would force us to enter, seriously and at length, into the daunting territory of the theory of conditionals and the enormous literature it has generated.

19 I have profited, in thinking about these materials, from suggestions from Kent Bach, Paolo Leonardi, Peter Ludlow, and Ernesto Napoli. I am especially indebted to Jeff King for extensive comments on an earlier version and to many discussions, over the years, with Mark Wilson.