From Haecceities To God
Haecceities, for the record, are a philosophical concept, not a kind of spell
(In a previous draft, I talked about de se facts, rather than haecceities, but that was a mistake. However, switching to talk about haecceities completely wrecked my ability to make hilarious puns, including titling the article “De Se There’s A God.” This is very sad.)
Philosophers often believe in what are known as irreducible haecceities. Haecceity facts are irreducible thisness facts—facts about something being this one that doesn’t depend on its properties. Sam Cowling writing for the Stanford Encyclopedia of Philosophy explains:
Imagine the following alternative history of the world: Things are qualitatively just as they actually are. There is no difference in anything like the shape, size, or mass of objects. There is no difference in the number of entities. Even so, there is a non-qualitative difference and it concerns you in particular. According to this alternative history, you fail to exist. In your place, there is a distinct individual, Double. Double has all the qualitative properties, whether mental or physical, you actually have, but, despite all these similarities, you and Double are distinct individuals. So, according to this alternative history, you do not exist.
Is this alternative history of the world a possible one? And what should we make of other alleged qualitatively indiscernible possibilities? For example, is it possible for a pair of siblings to “swap” their actual qualitative roles—i.e., where the actually firstborn twin is born second and vice versa—but where no qualitative features of the world are altered? In a similar vein, suppose, following Black (1952), that there could be a world containing only two qualitatively indiscernible iron spheres. Are there also other possible worlds where these spheres are “replaced” by numerically distinct yet qualitatively indiscernible doppelgängers?
An affirmative answer to these questions entails haecceitism, according to which the world could differ non-qualitatively without differing qualitatively. So, if the alternative history described above where Double replaces you is a genuine possibility, it is a maximal possibility—i.e., a total way the world could be—that differs haecceitistically from actuality. Similarly, if it is possible for twins to swap their birth orders while leaving all qualitative matters unchanged, that maximal possibility also differs haecceitistically from actuality.
It’s a tricky concept to get your head around, so let me give some examples to explain it:
Imagine there are two identical red balls on opposite ends of the universe. Then God snaps his fingers and switches them out. This seems possible. But crucially, no fact about the universe has changed—there are still two identical red balls with exactly the same properties. It seems the thing that has changed is that, though there are now balls with identical properties, each ball is in a different position from the one it was on.
There are two disembodied minds—minds without bodies. Both of them have the same experience. This scenario seems possible. But what could distinguish them? They have the same properties (no physical properties, possessing exactly the same experience). So in order to distinguish them, there must be irreducible thisness facts—facts about which one is which that don’t reduce to any qualitative facts. In other words, there are facts of the form “this is this one,” that aren’t about the qualities of the thing.
Imagine that there are two worlds, created at the same time, identical in all the things that happen in them. This seems coherent. Yet for them to be different things, there must be something that differentiates them, something that makes them different. It can’t be any qualitative properties, because they have no differentiable qualitative properties. Thus, the thing that differentiates them would have to involve irreducible haecceities—one is this one, the other is that one! Additionally, it seems that, if one of them stops existing at some time, this would be different from the other one stopping to exist at some time—but because they have the same non-haecceistic properties, on alternative views, such an event would be impossible. As Cowling writing for the SEP says:
Consider, again, a possible world w1w1, in which there are two qualitatively indiscernible globes; call them Castor and Pollux. Being indiscernible, they have of course the same duration; in w1w1 both of them have always existed and always will exist. But it seems perfectly possible, logically and metaphysically, that either or both of them cease to exist. Let w2w2, then, be a possible world just like w1w1 up to a certain time tt at which in w2w2 Castor ceases to exist while Pollux goes on forever; and let w3w3 be a possible world just like w2w2 except that in w3w3 it is Pollux that ceases to exist at tt while Castor goes on forever. That the difference between w2w2 and w3w3 is real, and could be important, becomes vividly clear if we consider that, from the point of view of a person living on Castor before tt in w1w1 and having (of course) an indiscernible twin on Pollux, it can be seen as the difference between being annihilated and somebody else being annihilated instead. But there is no qualitative difference between w2w2 and w3w3.
I think these arguments are very decisive in showing irreducible haecceities. If this is right, then there’s a strong argument for the existence of God.
A common idea when it comes to simplicity is that simplicity has to do with how long it would take to fully describe the thing that’s going on. Of course, we’re not talking about description in English—you can have one English word that describes something very complicated—but description in some ideal language that doesn’t use a single word to express a very complicated idea (I don’t think this should just be about how easy it would be to write a computer program expressing the idea, because that would imply that if consciousness doesn’t reduce to physical stuff, for instance, it’s infinitely complicated on account of being impossible to spell out with a computer program, but I think the basic idea is right).
Suppose you grant this. Well then atheism will have to have an extremely complicated account of what ultimate reality looks like. On such a picture, it’s literally impossible to specify the atheist’s model of fundamental reality, because it involves irreducibly de se facts. If there are an infinite number of atoms, differentiated only by their irreducible haecceities (their irreducible thisness), then there’s no way to write a program to pick out one particular one. This would mean that in an ideal language, it would be literally impossible to pick out a particular atom.
Now, you might object: if simplicity is about how long it would take to specify some object in an ideal language, couldn’t the ideal language simply brutely pick out particular atoms? If objects have irreducible haecceities, couldn’t the ideal language involve some unique operator to brutely pick out a particular atom—perhaps giving each atom its own unique symbol?
I’m not sure if this is right (it might be impossible to devise such a language with uncountably infinite symbols) but even if it is, it doesn’t help the problem much. It seems like on naturalism, it’s very unlikely that there will be just a single fundamental entity. Probably fundamental reality will involve a bunch of different things—e.g. a bunch of strings. But if you need a unique symbol to pick out every single individual string, then your picture of ultimate reality becomes monstrously complicated. You’ll need at least 10^89 symbols, it looks like, to pick out each of the specific strings that exist if string theory is true (I might be wrong about this, so if I am, correct me if you know about string theory). Theories of fundamental physics tend to require lots of different fundamental entities.
Now, you might object and deny that the right way to measure simplicity is by ease of specifying something in a formal language. My argument has so far assumed that, but if you deny it, then it may seem the argument doesn’t work. I think this doesn’t work though.
I think that the account of simplicity is very plausible. It’s hard to come up with another account that can, in principle, analyze the relative complexity of different properties.
If there is another such account, then I think atheism will turn out extremely complicated. On atheism, there are likely tons of different fundamental entities (e.g. huge numbers of different strings). Now, someone who accepts my preferred account of simplicity can explain why this doesn’t balloon complexity—it’s easy to specify many things with the same property—but on alternative accounts, this would be very complicated because it involves lots and lots of different things.
The basic point about atheism ballooning simplicity isn’t specific to my account. Imagine if you thought that every fundamental particle had a different kind of conscious state with no unifying explanation of them. Such a theory would be very complicated. But if there is no God, then fundamental reality has similar complexity—every fundamental particle has some irreducible property. This should be complicated for the same reason that it would be complicated for all fundamental particles to have different conscious states.
You might object: doesn’t this apply to God as well? Wouldn’t you need some way to pick out God rather than a qualitative duplicate of God? But this is not so obvious:
If God is necessary and there’s only one, then he won’t have qualitative duplicates. So if you think God exists necessarily, this isn’t a problem.
It’s not obvious that God is the sort of thing that could have qualitative duplicates. It’s not obvious that God has irreducible haecceities or that if he does, he’s the sorts of thing that could have a qualitative duplicate. If God is either pure act, perfection, or a limitless mind, it’s not at all obvious that this admits of irreducible haecceities, that there are other possible beings that are exactly like God but are irreducibly different people.
Even if there are, the theist will then have to posit only a tiny bit of extra complexity because they believe in just one fundamental thing. While it might be a cost to pick out which of the qualitative duplicates of God that they believe in, it’s a much less great cost than that the atheist must bear when they posit huge numbers of irreducible haecceities. While this might slightly add complexity to theism, makes atheism so much more complicated that it’s enough to single-handedly rule out the overwhelming majority of atheist views of ultimate reality.
The only objection I can think of—other than, of course, denying haecceitism—is that simplicity purely has to do with the description length of the qualitative elements in a theory. In other words, while irreducibly haecceities may be real, they don’t detract from simplicity.
But I think this is objectionably arbitrary. Why wouldn’t the irreducible haecceity facts count towards simplicity the way everything else does? To me, this seems like an arbitrary carve-out, an arbitrary epicycle, a bit analogous to a person who thinks that all fundamental particles have their own unique, irreducible mental state simply declaring that this doesn’t detract from simplicity. Such a theory would be implausible and bizarre! But how does it differ from the theory according to which irreducible haecceities exist but don’t detract from simplicity?
This argument has quite a few moving parts to it, and so I’m not sure exactly how convincing it is. Still, if there are irreducible haecceities, then it seems extremely strong.
Leaving aside questions of personal identity for a moment, quantum mechanics seems to very clearly exclude a notion of persistent identity for particles.
First, in the simple case where the amount and type of particles doesn't change, we can set up experiments where two multi-particle states are arrived at with opposite phase, and see if they interfere with each other. When the states differ only in the identity of the particles, we see that they do, so for example, "electron A in position X, electron B in position Y" can destructively interfere with "electron B in position X, electron A in position Y". Or it can interfere constructively, in which case each assignment of labels is equally valid.
Secondly, in the more fundamental theory of quantum fields, individual particles are just waves in those fields. Individual waves can have a meaningful kind of identity, if they don't interact too much with other waves, but in the general case waves can be created, destroyed, combine or split apart, or enter complicated interactions where you can't really pick out any waves any more at all, let alone identify "which" input wave they "really are". The situation in the simpler case arises because states are not distinguished by how they arose, only by the field value at each point, so "swapping" two identical particles is just the same state you started with.
This notion of fundamental identities also has a serious epistemological issue: if these identities don't effect any feature of the physical world, how could you possibly have learned of them? For example, you probably have an intuition that normally identities are persistent over time, but this intuition was not caused by you observing any physical event that demonstrates persistent identity because identity doesn't cause any physical events. So you would retain it even if, in truth, the identity of all trees on Earth were actually being shuffled every 6 hours, unless perhaps you also hold that God intervened to give you correct intuitions.
I think this sense that things have identities that persist over time independently of their physical state is just a cognitive shortcut humans use, which is approximately valid in our ordinary experience, but has no fundamental reality. Our experience involves lots of physically contiguous objects that don't change much over time, and don't often merge with, or split from, other objects of interest, and two previously distinct objects are almost never rendered physically identical in every way, so it's useful to think of objects as having a unique identity and associated history, but if you leave the ordinary experience this can quickly break down.
I definitely don't think that, whatever your theory of complexity/simplicity is, hypotheses should be penalized for failing to specify haecceities, as opposed to just positing the existence of objects which bear certain relations between them but are otherwise "anonymous." I don't exactly even know what it would mean to specify a haecceity, other than to point ostensively to things around you and being like "that specific thing over there." And I don't think anyone grades the simplicity of a theory of physics based on pointing to anything, as opposed to its abstract mathematical character which would be the same even if you woke up in a black room with no memories and couldn't point to anything at all.
And I still don't really see how God is helping you here. First because it's not clear how God decides to create this electron instead of a different, indistinguishable electron, and second because your theory of what God does is going to have to build in all of his countless decisions to build this and that haecceity-specific electron in exactly the same way you accuse the atheist of needing to do. So the complex-due-to-haecceity theory of physics has been derived from something else, yes, but it's an equally complex theory of divine action.
I think most of your examples motivating haecceitism seem like they'd be easy to bite the bullet on for an anti-haecceitist - indeed, they are precisely sort of the paradigmatic cases meant to illustrate to novices what anti-haecceitism even means! It actually seems sort of common in metaphysics in general to argue that the existence of a symmetry implies the illusoriness of some intuitive metaphysical thing. For instance, a lot of people would be motivated to say that God couldn't have created the whole universe shifted five meters to the left of where it actually is (perhaps we need to pretend we live in a Newtonian universe for this to make sense, IDK), because what would be the real difference?
Finally, I've privately wondered if haecceity-centric viewpoints undercut SIA, or at least lead it to drastically different conclusions than it's usually taken to lead to. For example, in any of the standard red-vs.-blue jacket thought experiments, we're called to ask what the probability is of God's coin landing heads conditional on our observing ourselves having a red jacket - SIA says the answer is going to be based on the expected number of people who have such observations depending on coin outcomes. But if we instead ask what the probability is of God flipping heads conditional on ourselves observing *this specific* experience of a red jacket, then there's never going to be more than one of those in any world, no matter how many qualitatively identical observers there are. In other words, if SIA is based on computing the number of possible people who have my evidence, then no more than one person truly has *my* evidence in this sense.