What Are Possibilities? Part 3A: Extreme Modal Realism

David Kellogg Lewis was an American philosopher who taught briefly at UCLA and then at Princeton from 1970 until his death
David Kellogg Lewis was an American philosopher who taught briefly at UCLA and then at Princeton from 1970 until his death in 2001.


Extreme Modal Realism (EMR) was a theory of the metaphysics of modality championed by David Lewis (1941–2001). Lewis maintained that possible worlds are collections of concretely existing things which are spatio-temporally related to each other but to nothing else.[1] He argued that propositions are sets of worlds in which they are true, and that properties are sets of individual entities across worlds which have certain features in common. For example, the property of being circular is nothing more than the set of all circular entities in this world and/or others; there is no abstract property of ‘circularity’ that exists independently of those entities.

A key feature of Lewis’ view is the attempt to explain the modal in terms of the non-modal. In Pruss’ (2011) words, “everything modal is analyzed in terms of the relations between concrete and unproblematic entities. [So] the mystery of modality is removed” (p.81). For instance, the modal proposition “I can jump” is true only if there is concretely existing world in which I do jump. Likewise, I have the property of “being able to jump” only if there are other worlds in which I jump. In this way, facts about what is possible are reduced to facts about worlds and the individuals existing in them.

Lewis’ claim that I exist in various worlds can be interpreted in two main ways: either the individuals in other worlds are me (the identity theory) or they are similar to me (the counterpart theory). Therefore, “if we accept identity theory, then the same individual ends up existing in multiple worlds….If we accept counterpart theory, then each individual exists in exactly one world but may have counterparts in some others” (Pruss, 1996, p.82). As we shall see, Lewis’ view faces sizable objections on either interpretation.


First, Lewis claims that actuality is indexical: a world is actual if the real objects within it are spatio-temporally related to me, whereas a world is non-actual if its objects are not related to me in that way. Because entities in other worlds are just a real as those in the world I inhabit, it follows that actual entities are only a small subset of all the real ones. Now at first glance, this view runs contrary to the common sense notion that all real entities are actual, but perhaps such a revision of common sense will be warranted by the other merits of EMR.[2]

Second, EMR implies that I have counterparts in other worlds, but it is unclear how these counterparts are related. Lewis cannot define the counterpart relation in terms of accidental and essential properties because the ability to distinguish these properties (say, by asking “could I have been a doctor?”) already presupposes modal notions, such as “could.” But if Lewis presupposes modality to explain who my counterparts are, then his attempt to reduce modal notions to the non-modal fails. What if my counterparts are related in terms of similarity to me, not by the essential properties I share with them? In my view, this is Lewis’ best bet, but using the concept of similarity to explain the counterpart relation may not be so helpful after all. Why? Because relations of similarity are complex ones – they are often context-relative and amenable to degrees. Thus, one may wonder if Lewis’ appeal to similarity is any less mysterious than the concept of modality he is attempting to explain.

Third, EMR is implausible when it comes to the truth-makers for statements of ability. Consider the ability-statement “I can jump.”  What makes this statement true? Most of us would say it is true by virtue of facts about me and my abilities in the actual world, because those facts play a causal role in the possibility of my jumping. For instance, the fact that my shoes are not stapled to the floor is causally relevant to whether I can jump.

What grounds the truth of ability-statements such as "I can jump"?
What grounds the truth of ability-statements such as “I can jump”?

By contrast, EMR suggests that causally irrelevant facts are involved making ability-statements true. It has the unfortunate result that “entities other than I, my properties and my surroundings are…involved in the truthmaker of the proposition that I can jump” (Pruss, 1996, p.89).[3] The other entities just are my counterparts who jump in worlds causally isolated from me.[4] But Lewis’ analysis does not really explain why it would still be possible for me to jump even if this world was the only real world; even if there were no other real worlds containing counterparts who jump.[5]

Lewis’ view implies that statements about the whole of reality have no modal status

Fourth, Pruss thinks Lewis’ analysis of modality is too restrictive because it can’t account for the possibility that the sum of all concrete worlds might have been different in certain respects – say, by containing fewer worlds. Let’s call that possibility P. Now, Pruss argues that P is conceivable, but Lewis is forced to deny P as a genuine possibility. Why is this so?

Recall that Lewis’ semantics analyzes modality in terms of what is true in particular worlds: something is possible if it is true in a world; impossible if it is not true in any world; and necessary if it is true in each world.  Unfortunately, his analysis can’t handle P because it makes a modal claim about the sum total of ALL concrete worlds. And that leaves Lewis without any remaining worlds for P to be true in!

What if Lewis argued instead that P is true at a world within the sum total? Again, this will not help, but for more complicated reasons: if there could have been fewer worlds, as P suggests, then Lewis’ semantics requires that there is a world W at which there ARE fewer! But this is something Lewis cannot accept. Why? Because he takes each concrete world to be part of the same sum total of reality. If each is part of the same sum total, then that sum remains the same in each world, including W. But then Lewis is forced to embrace a logical contradiction: that is, the sum total of reality is both the same and different at W, which is absurd! So it seems that Pruss is correct about the semantics of EMR. It is too restrictive.

Stay tuned for my next week’s post, where I briefly examine three more objections to Extreme Modal Realism.

[1] This entails that if two worlds are spatio-temporally connected, then strictly speaking they just are two parts of the same world. They are not distinct worlds.

[2] “For Lewis, ontologically all worlds are on par….Our world is actual, but this only says it is home…a world’s being actual is nothing but an indexical claim” (Pruss, 1996, p.80).

[3] The traditional Platonist faces a similar oddity insofar as the fact that I can jump is made true by an abstractly existing (and causally effete!) proposition having the primitive property of possibility.

[4] Pruss argues that by making attributions of ability dependent on what my counterparts do in other worlds, Lewis is committed to the truth of transworld counterfactuals, e.g. “if I refrain from A, my counterpart in another world will perform A.” But transworld counterfactuals are prohibited by Lewis because if they are true, then they violate the principle that worlds are causally insulated from each other.  I will not pursue Pruss’ objection here, but I see no good reason to think that such counterfactuals would  have to be causal counterfactuals.

[5] One way around this objection is to reject counterpart theory and favour the identity theory by claiming that it is I (not some counterpart) who is doing the jumping in these other worlds. But even if the identity theory solves the problem of causally irrelevant truthmakers, it leads to serious ethical problems (to be discussed later).

Leave a Reply

Your email address will not be published. Required fields are marked *