Editing David Lewis (philosopher) (section)

== Counterfactuals and modal realism ==
<!-- Missing image removed: [[Image:DavidLewis.jpg|thumb|David Lewis]] because is politician -->

Lewis went on to publish <cite>''Counterfactuals''</cite> (1973), which gives a modal analysis of the truth conditions of [[counterfactual conditionals]] in possible world semantics and the governing logic for such statements. According to Lewis, the counterfactual "If kangaroos had no tails they would topple over" is true if in all worlds most similar to the actual world where the [[Antecedent (logic)|antecedent]] "if kangaroos had no tails" is true, the [[consequent]] that kangaroos in fact topple over is also true. Lewis introduced the now standard "would" conditional operator ◻→ to capture these conditionals' logic. A sentence of the form A ◻→ C is true on Lewis's account for the same reasons given above. If there is a world maximally similar to ours where kangaroos lack tails but do not topple over, the counterfactual is false. The notion of similarity plays a crucial role in the analysis of the conditional. Intuitively, given the importance in our world of tails to kangaroos remaining upright, in the most similar worlds to ours where they have no tails they presumably topple over more frequently and so the counterfactual comes out true. This treatment of counterfactuals is closely related to an independently discovered account of conditionals by [[Robert Stalnaker]], and so this kind of analysis is called [[Counterfactual conditional#Possible worlds accounts|Stalnaker-Lewis theory]]. The crucial areas of dispute between Stalnaker's account and Lewis's are whether these conditionals quantify over constant or variable domains (strict analysis vs. variable-domain analysis) and whether the Limit assumption should be included in the accompanying logic. Linguist [[Angelika Kratzer]] has developed a competing theory for counterfactual or [[Subjunctive mood|subjunctive]] conditionals, "premise semantics", which aims to give a better heuristic for determining the truth of such statements in light of their often [[Vagueness|vague]] and [[Context-sensitive language|context-sensitive]] meanings. Kratzer's premise semantics does not diverge from Lewis's for counterfactuals but aims to spread the analysis between context and similarity to give more accurate and concrete predictions for counterfactual truth conditions.<ref>{{Cite book|last=Kratzer|first=Angelika|title=Modals and Conditionals: New and Revised Perspectives|publisher=Oxford University press|year=2012|isbn=9780199234691|pages=chapter 3}}</ref>

===Realism about possible worlds===
What made Lewis's views about counterfactuals controversial is that whereas Stalnaker treated possible worlds as imaginary entities, "made up" for the sake of theoretical convenience, Lewis adopted a position his formal account of counterfactuals did not commit him to, namely [[modal realism]]. On Lewis's formulation, when we speak of a world where I made the shot that in this world I missed, we are speaking of a world just as real as this one, and although we say that in that world I made the shot, more precisely it is not I but a ''[[counterpart theory|counterpart]]'' of mine who was successful.

Lewis had already proposed this view in some of his earlier papers: "Counterpart Theory and Quantified Modal Logic" (1968), "Anselm and Actuality" (1970), and "Counterparts of Persons and their Bodies" (1971). The theory was widely considered implausible, but Lewis urged that it be taken seriously. Most often the idea that there exist infinitely many causally isolated universes, each as real as our own but different from it in some way, and that alluding to objects in this universe as necessary to explain what makes certain counterfactual statements true but not others, meets with what Lewis calls the "incredulous stare" (Lewis, ''On the Plurality of Worlds'', 2005, pp.&nbsp;135–137). He defends and elaborates his theory of extreme modal realism, while insisting that there is nothing extreme about it, in ''[[On the Plurality of Worlds]]'' (1986). Lewis acknowledges that his theory is contrary to common sense, but believes its advantages far outweigh this disadvantage, and that therefore we should not be hesitant to pay this price.

According to Lewis, "actual" is merely an indexical label we give a world when we are in it. Things are [[logical truth|necessarily true]] when they are true in all possible worlds. (Lewis is not the first to speak of possible worlds in this context. [[Gottfried Wilhelm Leibniz]] and [[C.I. Lewis]], for example, both speak of possible worlds as a way of thinking about possibility and necessity, and some of [[David Kaplan (philosopher)|David Kaplan]]'s early work is on the counterpart theory. Lewis's original suggestion was that all possible worlds are equally concrete, and the world in which we find ourselves is no realer than any other possible world.)

=== Criticisms ===
This theory has faced a number of criticisms. In particular, it is not clear how we could know what goes on in other worlds. After all, they are causally disconnected from ours; we can't look into them to see what is going on there.<ref>[[Robert Stalnaker]], ''Inquiry'', MIT Press, 1984, p. 49: "But if other possible worlds are causally disconnected from us, how do we know anything about them?"</ref> A related objection is that, while people are concerned with what they could have done, they are not concerned with what people in other worlds, no matter how similar to them, do. As [[Saul Kripke]] once put it, a presidential candidate could not care less whether someone else, in another world, wins an election, but does care whether he himself could have won it (Kripke 1980, p.&nbsp;45).{{citation needed|date=May 2021}}

Another criticism of the realist approach to possible worlds is that it has an inflated [[ontology]]—by extending the property of concreteness to more than the singular actual world it multiplies theoretical entities beyond what should be necessary to its explanatory aims, thereby violating the principle of parsimony, [[Occam's razor]]. But the opposite position could be taken on the view that the modal realist reduces the categories of possible worlds by eliminating the special case of the actual world as the exception to possible worlds as simple abstractions.

Possible worlds are employed in the work of Kripke<ref>"[[Naming and Necessity]]". In ''Semantics of Natural Language'', edited by D. Davidson and G. Harman. Reidel, 1980 (1972), pp. 253–355.</ref> and many others, but not in the concrete sense Lewis propounded. While none of these alternative approaches has found anything near universal acceptance, very few philosophers accept Lewis's brand of modal realism.

=== Influence ===
At Princeton, Lewis was a mentor of young philosophers and trained dozens of successful figures in the field, including several current Princeton faculty members, as well as people now teaching at a number of the leading philosophy departments in the U.S. Among his prominent students were [[Robert Brandom]], [[L. A. Paul]], [[J. David Velleman]], [[Peter Railton]], [[Phillip Bricker]], [[Cian Dorr]], and [[Joshua Greene (psychologist)|Joshua Greene]]. His direct and indirect influence is evident in the work of many prominent philosophers of the current generation.