David Lewis holds that morality is indexical to individuals. That is, each individual is an extreme moral egoist. This position is forced upon him if he wishes to avoid criticisms such as those raised by Robert Adams and others to the effect that, given counterpart relations across worlds, there is no reason to avoid evil action.[i] The net balance of good and evil is unaffected no matter what my moral decision may be. Lewis’ response is sufficient to save his theory from Adams, but there is a particularly devastating critique that can be leveled against him in return. If we assume that, contrary to what Lewis holds, there are transworld counterfactuals, we can demonstrate that counterpart theory is incompatible with our held moral intuitions and thereby extreme modal realism (EMR) is false.
For every way the actual world could be there is a concrete possible world that is that way. This, together, importantly, with the doctrine of counterpart relations, is the EMR picture of modality. One immediate worry with this picture, raised by Adams, is that the evils occurring in other worlds are every bit as real as those in the actual world; why not then go ahead and commit evils?[ii] For example, in a nearby possible world there are Guantanamo-Bay-counterparts who are brutally tortured for useful information. Well, if we do not torture the actual inmates, these counterparts will be. So why not torture them here and save the others the worry? At worst, this leads to moral indifference. Lewis’ response is moral indexicality: our moral aims are egocentric.[iii] For the virtuous agent moral concern is strictly limited to their worldmates and, even more so, to themselves (presumably given that what is under an agents control is what they causally bring about). This assumes that causation cannot occur between worlds, that worlds are isolated to the extent that moral considerations need not extend from one to another.
Lewis holds that, given a counterfactual analysis of causality, transworld causation “comes out as nonsense”.[iv] So, transworld counterfactuals ought to be nonsense.[v] However, take the following slight adaptation of an argument from Alexander Pruss, to the effect that this nonsense is possible.[vi] Imagine a world where there is only one free action possible (say, the Oracle of Delphi can speak to Socrates or not) and all other actions at this world, before and after, are determined (given the laws of nature and initial conditions). Further, assume that in this actual world I am the Oracle of Delphi. At time t I am able to decide whether to freely speak to Socrates or to freely refrain from speaking to Socrates. Hence, there are two worlds involved here that match the actual world. Let W1 be the world in which my counterpart speaks to Socrates and W2 be the world in which my counterpart refrains from speaking to Socrates. Given the Lewisian account, the following is true at W1:
How would we evaluate (T)? Looking at W1, the closest world to W1 where the antecedent of (T) holds is W2. However, at W2 it is true that “my counterpart in the other world that matches the actual world would speak to Socrates” given that it is true at W2 that “the other matching world is W1” and it just so happens that, in W1, my counterpart speaks to Socrates. Thus, we have a transworld counterfactual.
With this in place several arguments can go forward, including versions of Mark Heller’s criticism that have even more bite than his original.[viii] Heller speaks of an inevitability relation, which is a logical relation, holding between worlds wherein an agent contemplates saving a drowning child or not. The problem is one of moral obligation or moral excuse, given counterpart relations and genuine EMR. We can give the situation in terms of a transworld counterfactual: if agent Xwas to save drowning child Y, then in some world W a counterpart of X, X`, will let some child Y` drown. If, as EMR assumes, X` and Y` and all such are genuine persons, real people, then we have a situation where more people than we would otherwise expect must enter into our moral calculations.[ix] But, if more genuine persons enter into our moral calculations, then X can be indifferent to the drowning of Y in the actual world. That is, given that the transworld counterfactual (T) holds, it will inevitably be the case that some child somewhere drowns, thus making it the case that X is under no obligation, all things being equal, to save the drowning Y.[x] EMR allows X to be morally indifferent when our ethical intuitions scream the contrary, and to that extent EMR is false.[xi] Ordinary morality dictates that if a trivial action of mine can save the life of an actual person, then it is morally obligatory; EMR allows me to take the easy way out and let the child drown (and have an untroubled conscience!).
Consider, further, the case of self-torture. Let us assume that, in the argument laid out above for a transworld counterfactual, all the conditions are exactly the same except that instead of the one freedom to speak to Socrates or to refrain I am free to brand myself with a red-hot poker or to refrain. According to (T), then, if I were to refrain from branding myself with a red-hot poker my counterpart in the other world that matches the actual world would brand himself. This is a matter of causality, indeed, of logical necessity. Extend the case some: I and a complete stranger are in a Medieval Ottoman Turkish dungeon facing torture, except that the Sultan declares that the person to brand himself will spare the other. It would be a morally praiseworthy, indeed heroically virtuous, deed for me to brand myself and spare the stranger.[xii] Hence, if Lewis is correct, then I right now, sitting at my desk, would be heroic if I spared some stranger in another world the horror of self-torture by branding myself. But this is not morally praiseworthy at all; rather it is morally insane. Even if there were more free choices available rather than just this one would not matter. One possible response, for Lewis, might be to admit that the overall balance of good and evil in reality is not affected; go ahead and let the other person brand himself, after all, it makes no ultimate difference.[xiii] This, however, would be to straightforward cede the previous case of the drowning child. There would be no point for X to attempt to save Y, sub specie aeterinitatis. Ordinary morality dictates otherwise.
Can Lewis’ response to Adams withstand the counterfactual cases presented? It seems not. Why, after all, should we care about worldmates over and above the fate of counterparts in other worlds, if all are equally real? Lewis says the virtuous person will care about his own character, but it is in fact his character that is the cause of the death, or torture, of his counterparts. Hence, on EMR, the virtuous person ought to be a super-saint and care about all his counterparts and possible actions in other worlds. But this is surely contrary to ordinary morality.
The point can be made, paradoxically enough, with an appeal to Humphrey-style cases. Kripke offers the following critique to Lewis’ counterpart theory:
The counterpart of something in another possible world is never identical with the thing itself. Thus if we say ‘Humphrey might have won the election…’ we are not talking about something that might have happened to Humphrey but to someone else, a ‘counterpart’. Probably, however, Humphrey could not care less whether someone else, no matter how much resembling him, would have been victorious in another possible world. Thus Lewis’s view seems to me even more bizarre than the usual notions of transworld identification it replaces.[xiv]
This is an objection concerning the modal attitudes Humphrey takes to certain modal facts. It is a pre-theoretical virtue that we ought to remain faithful to our modal concerns. As Gideon Rosen is apt to point out, one measure of an ontology is whether it remains faithful to our pre-theoretical concerns and modality is no different; we respond emotionally, morally, and pragmatically to our modal thoughts.[xv] Kripke’s objection is that EMR fails this constraint; Humphrey simply does not care about his counterpart. More substantially, this is the objection that accepting a modal theory should not force a modification of one’s modal concerns and Humphrey seems forced to give up his modal concerns. Lewis’ response could very well be to bite the bullet and say, no, we only need to revise our concern for perfect (other-worldly) strangers.[xvi] Humphrey, it turns out, ought to care about the facts concerning his counterparts given that he cares about the modal facts concerning him and these just are the facts about his counterparts. It should be obvious, however, that in light of the preceding transworld counterfactual argument, if Lewis were to take this obvious route he would ipso facto lose the remaining plausibility of his modal ethical egoism thesis. Hence, Lewis’ response to Adams fails to cover the transworld counterfactual cases, and EMR fails to hold up to moral considerations.[xvii]
Adams, Robert Merrihew. “Theories of Actuality,” in: Loux, Michael. The Possible and the Actual: Readings in the Metaphysics of Modality. Ithica, New York: Cornell University Press, 1979.
Divers, John. Possible Worlds. London: Routledge, 2002.
Heller, Mark. “The Immorality of Modal Realism, or: How I Learned to Stop Worrying and let the Children Drown,” Philosophical Studies 114: pp. 1-22, 2003.
Kripke, Saul. Naming and Necessity. Cambridge, Mass.: Harvard University Press, 1980.
Lewis, David. On the Plurality of Worlds. Oxford: Blackwell Publishing, 1986.
Pruss, Alexander. Actuality, Possibility, and Worlds. New York: The Continuum International Publishing Group, 2011.
Rosen, Gideon. “Modal Fictionalism,” Mind, New Series, Vol. 99, No. 395, pp. 327-354, July 1990.
[i] D.C. Williams allegedly raised a similar objection in a lecture at the University of Notre Dame, 1974, in which it seems strange to have one strive to rid the world of evil, if they just reappear in other worlds. See Lewis, David. On the Plurality of World. Oxford: Blackwell Publishing, 1986, (page 123, note 6).
[ii] See Adams, Robert Merrihew. “Theories of Actuality,” in: Loux, Michael. The Possible and the Actual: Readings in the Metaphysics of Modality. Ithica, New York: Cornell University Press, 1979, (page 195). Adams asks, “ ‘What is wrong with actualizing evils, since they will occur in some possible world anyway if they don’t occur in this one?’”
[iii] Lewis, David. On the Plurality of World. Oxford: Blackwell Publishing, 1986, (127). To quote, “If you actualise evils, you will be an evil-doer, a causal source of evil. That is something which, if you are virtuous, you do not want to be. Otherworldly evils are neither here nor there. They aren’t your evils. Your virtuous desire to do good and not evil has nothing to do with the sum total of good and evil throughout reality.”
[v] Lewis’ counterfactual analysis can be described as: “We have a world W where event C causes event E. Both these events occur at W, and they are distinct events, and it is the case at W that if C had not occurred, E would not have occurred either. The counterfactual means that the closes worlds to W at which C does not occur, E does not occur either.” Ibid.
[x] After all, should X decide not to save Y, her counterpart in another matching world, X`, is such that she will save Y`. So X may reason that, no matter what, one child is saved and, if it is inevitable that her actions result in the living and dying of some child, she is hardly under moral obligation to save Y in lieu of Y`.