Given the problems with Possibilism, it might seem that we should abandon possibilism and embrace actualism. Actualism certainly captures the right kind of inequality. Actual things are real, and there simply are no non-actual things. Absolutely everything is actual!
At the same time, there could have been things that do not exist. Perhaps there could have been flying amphibians, even though in fact there are none. This is consistent with actualism, so long as we don’t include any flying amphibians in our domain of quantification. We can say that there are no flying amphibians, even though there could have been some. Other possible worlds have their own domains of quantification, and some are (or, more precisely, would be) quite different from the actual domain of quantification. We’re tempted to say that there are things in those other domains of quantification of some of those other worlds that are not contained in the actual domain—but we can’t say that! All we can say is that there would be things that would belong to those domains if those other worlds were actual, in place of the one actual world.
There are two principal difficulties with this strict actualism. First, it must treat all reference to non-actual individuals and species in an abstract, de dicto fashion. We can’t name non-actual individuals because there is literally nothing there to be named. There can be no de re facts about non-actual things, although actualists do permit de re modal facts about actual individuals.[1] This introduces a very implausible asymmetry into our modal universe, forcing us to treat the actual world as intrinsically unique, which makes it hard to make sense of the claim that any other possible world could really have been actual.
This deficit in de re modalities is most clear when we consider possible worlds that would contain only non-actual individuals in some perfectly symmetrical arrangement. Consider, for example, Max Black’s thought experiment of a possible world containing two perfectly spherical and homogeneous iron spheres eternally rotating around their common center of gravity. Since we’re supposing that the world contains two distinct but indiscernible spheres, we should be able to label one ‘Castor’ and the other ‘Pollux’. Now, if the Max-Black world were actual, there would be distinct de re possibilities involving each of the spheres. For example, there should be a possible world in which only Castor and not Pollux exist, and a second, distinct possible world in which it is Pollux and not Castor that exists. Let’s call these worlds C and P. However, there is, from the perspective of our actual world, no way to distinguish C from P. They are abstractly and qualitatively exactly the same, and they differ only in respect of which non-actual thing they contain. But (according to actualism) there are no non-actual things, so the resources of reality (on the strict actualist view) are insufficient to ground any difference between the two worlds.[2]
This problem is what led both Duns Scotus and Alvin Plantinga to posit bare haecceities, “thisnesses”, which exist necessarily, one haecceity for each possible non-actual entity. With haecceities, we can suppose that the actual world contains two distinct and uninstantiated haecceities, Castorness and Polluxity, and these two properties can be the basis for the distinctness of worlds C and P. These haecceities lead to their own metaphysical problems (see Koons and Pickavance 2017, 345-7). They do, however, solve the problem of de re modal facts about non-actual things.
The second difficulty with actualism is its inconsistency with the ancient formula, Nihil fit ex nihilo—nothing comes from nothing. Parmenides used this principle to demonstrate the impossibility of change, since all real change involves some substance or accident making the transition from non-being to being. Since non-beings don’t exist, it isn’t possible so much as to think or speak about them (de re). On this point, the strict actualists are fully in agreement with Parmenides. Consequently, it should be impossible for any actual thing to have the power to produce some non-actual substance or accident.
Aristotle’s answer to this Parmenidean challenge depended on his introducing the category of merely potential being. Change involves the transition, not from absolute non-being to being, but from potential to actual being. This Aristotelian solution requires some deviation from the strict possibilist framework.
But, once again, an actualist who embraces haecceities has an alternative solution to the Parmenidean problem. Haecceitists can theorize change as involving a transition from an uninstantiated to an instantiated haecceity.
As we shall argue, haecceity theory is subject to a number of compelling objections. If we do reject haecceity theory, then (as Aristotelians) we will have to reject actualism altogether. As we mentioned above, Lewis’s egalitarian possibilism is also incompatible with Aristotelianism, since Aristotelians are committed to an absolute distinction between the actual and the merely potential.
How, then, can possibilists account for actuality as an absolute feature of some possible entities and not others? As we saw, we cannot treat actual existence as another accident or accident-like entity, since such entities would have to be included in all possible worlds, whether actual or not. Instead, the factor that explains something’s actuality will have to be external to that thing as a possibility. There are at least four ways to do this. Before we can lay out these four possibilist options, we must first discuss the alternative metaphysical theories of predication, which I will do in my next post.
[1] A de re fact is a fact about an individual as such. We predicate a property to an individual de re just in case we can pick out the individual without relying on some other description the individual satisfies. Actualists cannot make de re predications of non-actual individuals, and they cannot countenance de re facts about such things, since there are simply no such things (in our actual domain of quantification). They can only say that there could have been individuals satisfying certain descriptions (where these descriptions must be constructible using only the ontological resources of the actual world).
[2] See Christopher Menzel (1991) for the best actualist attempt to solve this problem. Menzel proposes using abstract objects (like real numbers) to stand in for the non-actual entities that exist according to some other possible worlds. He fails, however, to explain how an actual fact about a real number could be the truth-maker for a subjunctive truth about a possible concrete object.
The Analytic Thomist 
One thought on “Two Problems with Actualism”