Modal Realism with Modal Tense 1

¹Various shorter versions of this paper were presented at: Keio University, Tokyo, and Kanazawa University, Kanazawa, Japan, October 2004; Southern California Philosophy Conference, Los Angeles, October 2005; the American Philosophical Association Pacific Division Annual Meeting, Portland, OR, March 2006. I thank the commentator on my Portland presentation, Uriah Kriegel, the Editor of this journal, and an anonymous referee for their helpful suggestions, which improved the paper.

Abstract

Modal realists should fashion their theory by postulating and taking seriously the modal equivalent of tense, or modal tense. This will give them a uniform way to respond to five different objections, one each by Skyrms, Quine, and Peacocke, and two by van Inwagen, and suggest a non-Lewisian path to modal realism.

Notes

¹Various shorter versions of this paper were presented at: Keio University, Tokyo, and Kanazawa University, Kanazawa, Japan, October 2004; Southern California Philosophy Conference, Los Angeles, October 2005; the American Philosophical Association Pacific Division Annual Meeting, Portland, OR, March 2006. I thank the commentator on my Portland presentation, Uriah Kriegel, the Editor of this journal, and an anonymous referee for their helpful suggestions, which improved the paper.

²It is customary to characterize modal realism along the following lines: (1) there are merely possible worlds, as well as the actual world, (2) there are possible individuals existing at those worlds, and (3) the possible individuals existing at merely possible worlds are as real as those existing at the actual world. I avoid such a characterization because I do not accept (1) and (2) unconditionally. I find the use of the existence predicate ‘there are’ in them problematic. I do not agree with David Lewis's claim, with which many others concur, that the existence predicates (‘there are’, ‘exist’, and ‘there exist’) have a unique absolutely unrestricted reading. (See Section V.D and Section VI.) My view is that no absolutely all-inclusive domain is available to legitimize such a reading and therefore every occurrence of an existence predicate needs to be understood relative to a restricted domain. Each of (1) and (2) is true relative to some restricted domains and false relative to others, and ontologically serious discussion of those domains will already involve live issues between modal realists and their rivals.

The only widely known and widely discussed modal realist theory is David Lewis's but my characterization of modal realism allows other versions of modal realism. See Section VII.

³This leaves it open whether some important modal facts are neither modal-tensed facts nor modal-tenseless facts. Some such facts might be able to be designated or quantified over adequately both in modal-tensed terms and in modal-tenseless terms, while others might be able to be designated or quantified over adequately neither in modal-tensed terms nor in modal-tenseless terms; the latter facts would defy expressive completeness of our modal language, with or without modal tense.

⁴Bertrand Russell says, ‘The occurrence of tense in verbs is an exceedingly annoying vulgarity due to our preoccupation with practical affairs’ 1956: 248].

⁵I am inclined to think that the modal analogues of some such complex tenses can be made intelligible, but any serious discussion of this topic would take us too far afield.

⁶In general w* will almost always contain many members, whereas in the temporal case t* may well contain just one member, e.g., the year 1964. Hardly any humanly possible context of utterance is so detailed as to determine a single possible world as the member of the salient set w*. Almost all humanly possible contextual determinations of w* proceed propositionally. No matter how much information the propositions in question may carry, they will almost never carry enough information to determine a unique possible world.

⁷There is much to be said about the use of the actuality tense in ‘is _a true’, ‘is _a evaluated’, and ‘is _a (not) the actual world’, but space does not permit elaboration. For some related but limited discussion, see the end of Section VI.

⁸This way of looking at my proposal was suggested to me by Uriah Kriegel. Also, see Haslanger 1989: 7 – 8].

⁹Colin McGinn 2000: 74 – 83] offers a similar idea under the label ‘the copula modifier theory’. He contrasts the copula modifier theory with the predicate modifier theory David Wiggins 1976 proposes and says, ‘Thus, according to the copula modifier theory, we do not work with an ontology of modal properties, rather, we take the stock of non-modal properties and think of them as possessed in different modes’ 2000: 77]. There is much in McGinn's proposal that I agree with, but there are some important points of difference. McGinn says, ‘If I just say, “Socrates is a man”, I do not commit myself to the mode of instantiation involved—the copula is modally neutral here—but the instantiation itself is always either necessary or contingent’[80 – 1]. I differ. If I just say, ‘Socrates is a man’, I usually commit myself to the actuality-tensed understanding of the copula. I also do not understand what it means to say of ‘instantiation itself’ that it is ‘either necessary or contingent’. Another difference between us is that I work within the possible-worlds framework, whereas McGinn rejects the framework: ‘modality belongs to a special ontological category: it consists neither in objects (unlike the possible worlds theory) nor in properties (unlike the idea of modal properties that goes with the predicate modifier view), but rather in items I have called modes’[ibid.: 83]. It is, however, unclear how deep this last difference is, as I allow possible worlds to constitute special ontological dimensions; see Section VII.

¹⁰M. J. Cresswell 2006 argues that presentism and four-dimensionalism (which he calls ‘eternalism’) are logically inter-translatable both at the object-language level and at the meta-language level. If Cresswell is right, some might interpret his result as showing that the disagreement between presentists and four-dimensionalists is illusory, hence the difference between actualism and modal realism is illusory. That would be a mistake. Cresswell's result, assuming it is correct, shows no such thing and has no such implication. As Cresswell himself repeatedly says, it instead shows that the dispute between presentists and four-dimensionalists cannot be solved by logic alone. The implication of this for us is at best that the difference between actualism and modal realism goes beyond logic.

¹¹The predicate ‘is a talking donkey’ is a surface form of ‘is a donkey and talks’. The noun phrase ‘talking donkeys’ should be understood as shorthand for ‘individuals that are donkeys and that talk’. Thus, the modal-tensed phrase ‘talking _p donkeys _p ’ should be understood as shorthand for ‘individuals that are _p donkeys and that talk _p ’. The mere possibility-tense should not be understood as a modifier of nouns.

¹²Further paraphrasing would reveal more occurrences of ‘are’. Among them would be one in ‘individuals who are possible’. This would raise an interesting issue of modal-tensing the possibility predicate, but we will not pursue it here.

¹³Modal-tensing is not incompatible with counterpart theory, but we shall not take up this issue here.

¹⁴A modal-tenseless sense is at best an artificial product manufactured out of modal-tensed senses by means of disjunction in a way analogous to the way the four-dimensionalist tenseless sense of ‘is’ may be construed by a presentist to be the disjunction of ‘is’ (in the present tense), ‘was’, and ‘will be’.