Against Possible Worlds

Against Possible Worlds

The phrase “possible world” has become ubiquitous in philosophy, but it is a meaningless phrase, a linguistic monster.[1] It derives from perfectly good uses of the word “world” while departing significantly from those uses. That word can mean planets, regions, cultures, habitats, professions, hobbies, religions, and more; but it can’t mean the totality of reality or maximal sets of propositions or some such technicality. Nor is there such a thing as “the actual world”, because this phrase is entirely unspecific about the kind of world intended. The problem isn’t that the doctrine of possible worlds is ontologically extreme—the objection that these things don’t really exist—it is that the concept is faulty (meaningless, empty). A “possible world” isn’t really a world in any intelligible sense. When someone says they live in the real world, they don’t mean the whole of reality (whatever that may be); they mean something like the world of money or hard labor or the daily grind. No one ever refers to what the philosopher purports to refer to; his meaning is far removed from the everyday meaning of the word. It is pure invention. When some intellectual speaks of the “best of all possible worlds” they don’t mean what the contemporary metaphysician means by that word; they mean something like the best human society or moral order (“the moral world”). The metaphysician’s use of the word is unprecedented and unexplained, a kind of pseudo-word living off another word’s credentials. There are no “possible worlds”; that word is not a genuine sortal.

To be sure, there are possible dogs and lifestyles and hairdos, because these are sortals in good standing; but there are no possible worlds, construed as real individuated entities out there. For how can they be counted, identified, named, tracked, sorted? Have you ever had one (and only one) in mind, or remembered it, or attended to it and not others? You can picture possibilities, but not entire worlds of them—unless you mean possibilities of a specific type of world, such as a planet or country or culture or profession. You have been tricked into thinking you mean something when you say that necessity is truth in all possible worlds, picturing some nebulous giant chunk of reality. When Wittgenstein speaks of “everything that is the case” he is not denoting an entity at all; that is not really a singular term denoting a maximal object of some sort. When someone speaks of a “brave new world” he is not speaking of one of these spectral super-beings coming to exist; he simply means (say) a new political system. These putative worlds are not like places and times that can be located, ordered, measured, and experienced; they are a theoretical construct created to enable modal words to be fitted into standard quantificational logic. They have no other ontological standing. They are just the postulated values of variables in logical notation (“for all worlds w”); they aren’t denizens of reality in any other sense. That’s why attempts to explain their meaning invariably fall back on the vernacular use of “world”, which is entirely legitimate in itself. We don’t quantify over possible worlds because they exist; they exist because we quantify over them—or pretend to. They are a pseudo-ontology. And don’t say they are merely formal devices equivalent to “set of propositions or sentences”; such things are not worlds at all but descriptions of worlds. The whole story is a put-up job, a kind of con. The philosophy profession has been conned into accepting a bunch of nonsense, literally. Not a bunch of exotic non-existent objects, mark you, but a nonsensical concept. You might as well use the word “wardrobe” to refer to the objects in question, as in “in all possible wardrobes, p”. To repeat, a world (vernacular) is not a “world” (technical): we would never say “in all possible worlds, p” meaning “on all planets, in all professions, in all cultures and habitats, p”.  A wardrobe is a wardrobe, as a world is a world—not some sort of super-duper-world (super-duper wardrobe). This is a misleading metaphor at best. As I say, a con job, not an interesting ontological discovery; a kind of metaphysical shell game (now you see it, now you don’t).

Possible worlds metaphysics and possible worlds semantics are therefore a horrible hot mess—confusion piled on solecism, a pyramid of nonsense. And isn’t it quite preposterous to suppose that when an ordinary person uses the word “necessary” or “must”, they are referring to a vast multiverse of these rebarbative pseudo-entities? It is just not credible that this is a good analysis of these ordinary words.  Possible worlds must have been resorted to for want of anything better. Before them we had only the theory that necessity is analytic truth, which is far too narrow. So, we are left with no theory at all. We have only a bonkers theory or a lame theory. We can, of course, take our modal vocabulary as primitive, using it to formulate our favorite doctrines—the idea that names are rigid designators, essentialist claims, and rejections of materialism. But this is hardly satisfactory; so, we hanker after the illusory charms of “possible worlds” (the quotation marks are compulsory now). Better that than nothing, you might say. But it’s worse than nothing; it’s a fantasy, a piece of nonsense on stilts dressed up as rigorous logic. I think people will look back on it as one of the more embarrassing episodes in the history of analytical philosophy.[2]

[1] See my “Counting Worlds”.

[2] Look, I know I am being excessively harsh here, but I am criticizing my earlier self too (and it is kinda fun). Seriously, it’s hard not to see the fashion for possible worlds as an infatuation with formal systems and the prestige attaching to them. Someone has to point the trembling finger at the naked emperor. It was a fun game while it lasted, but let’s move on. (Note that I name none of the main culprits.)