blog.kennypearce.net

Hi Kenny,

I just read your paper. It's an interesting view! I wondered if you could clarify something about it for me:

Normally we say that there are composite objects nobody's thought about, like individual boulders on Mars or whatever. We've thought about boulders, but not all the particular boulders. As I understand your view in the paper, these boulders won't exist, because nobody's unified them in their mind. You could have a view - mereological phenomenalism, maybe - on which the undiscovered simples arranged boulderwise do compose because they're arranged in accordance with our boulder concept. Was mereological phenomenalism the view you had in mind?

I think there might be a dilemma lurking around the choice between the two views, too. Mereological phenomenalism is, I guess, the one that does more justice to our normal judgements about what there is. But it also seems susceptible to the vagueness objection in a way that full-on mereological idealism isn't. The defence on page 8 - "in order to wonder whether *that* is a cloud, I must unify *that* in thought under some concept" - doesn't seem (to me) to work for mereological phenomenalism. It *is* vague whether some undiscovered bunch of simples fit any of our composite object concepts.

I'd be interested to know which of the views you were after, and what you think about the dilemma. Thanks!

Mike

Hi Michael,

Thanks for your comments! This is an issue several people have raised in response to this draft, so I'll definitely have to address it directly in the revised version.

I do want to say there are boulders on Mars, and I want to say that it's an open, empirical question whether there are (now) clouds on any planets that might be circling Alpha Centauri. But I don't want to allow vague existence, so (as you point out) I can't say that they exist if the molecules are arranged in accord with our concept.

I think I want to make this move instead: when we think about boulders on Mars, we create boulders on Mars. We may not be thinking of particular boulders, but we're still regarding the Martian landscape in a certain way. It's only later, of course, that we find out whether our thoughts were thoughts of/about any real stuff. But we've already created the unified objects before we find out they're real.

Now here are some questions you might have about this.

How many boulders are there on Mars? Of course, if we mean how many real boulders, that'll be indeterminate. But how many boulders of any kind, whether real or not? That depends on how many Martian boulders are ever thought of. Generally speaking, whenever we ask, 'are there more than n?' we're going to be thinking of boulder number n+1, and thereby bringing about its existence, so this is one of those questions we won't ever be able to answer precisely because of the way our investigation affects the facts.

Were there boulders on Mars before anybody thought about them? Yes, but they depended for their existence on the fact that people were going to think about them in the future. (I'm an eternalist about time.) Similarly, there were clouds on earth before anybody had the concept cloud. Our thoughts about those clouds now suffices for their existence insofar as our thoughts really truly are about the molecules that were floating around earth's atmosphere back then.

Where on Mars is the boulder I'm thinking of right now located? What particles is it made of? I'm happy to allow the answers to these questions to be indeterminate. But of course it's determinately true that my boulder is on Mars, and it's determinately true that my boulder is made of rock molecules. Will this particular boulder turn out to be real? Probably not because I probably don't succeed in making cognitive contact with some particular individual observable quantity of stuff. But when I think vaguely of boulders like this, then I carve up the Martian landscape in a certain way and create the boulders.

I plan to spell this out in the next draft, so I'd be interested in any pressing questions or objections to it!

Kenny

Kenny,

Thanks for your reply; I'm glad to see other people have had the same thought as me and that it's something you've thought about.

I guess there's a sort of objection to your response which is big but probably not dialectically appropriate, which is that it relies quite heavily on the Schaffer/Thomasson/Brentano(?)/etc view that existence is easy, and referential failure is correspondingly difficult. I've done quite a lot of work on fiction and referential failure and things and I don't really like that view myself, but if you do like it then putting it to work here seems fair enough.

I think there are two questions I'd like answered when you're spelling it out though.

First, are there real boulders on Mars? If nobody's making cognitive contact with any particular groups of appropriately arranged simples on Mars, then it sounds like the answer is no, and this violates our ordinary judgements about what there is just as much as the view that there are no boulders on Mars would. (It's kind of reminiscent of the objection to Meinongianism that it has as much trouble with the existent golden mountain as everyone else has with the golden mountain. Meinongians sometimes appeal to a watered-down property of existence which the existent golden mountain can have without existing, but this isn't pretty.)

Second, I wonder if you'll have indeterminate identity as well as indeterminate parthood, and whether this will be the benign kind resulting from indeterminacy in what's being referred to or thought about, or the scary kind resulting from things out there on Mars with indeterminate identity relations between them. I was thinking about this in particular because I thought that if I'm thinking about a boulder on Mars and then someone goes to Mars and finds a boulder it might be indeterminate whether the one they found was the one I was thinking about, especially if the simples composing the astronaut's boulder weren't determinately not the simples composing mine. But even if we conceded this it'd still leave open both the benign and the scary options, as far as I can see.

Mike

There are certainly real boulders on Mars, since the way I'm thinking about the Martian landscape when I think about the boulders there is an appropriate way of carving up the landscape - i.e., the landscape is really like that. One might think that there being real boulders on Mars implied that there be some particular Martian boulder that was determinately real, but I'm not sure that's true. There are some issues about the logic of vagueness here that I need to think (and read) about. If it does imply that, I might be in trouble.

The second objection is indeed pretty tricky. One way to go would be to appeal to a different notion of sameness, on which two boulders count as the same if they are made real by the same underlying stuff. So the boulder I'm thinking of and the boulder you're thinking of are not numerically identical, but they're the same in the sense that matters. Then this sameness relation could be vague due to the vagueness about what stuff is making each boulder real.

I'm not totally happy with this line because I think we can both think about the same fictional character, and since I think fictional characters exist I think they should have well-defined identity conditions. An analogous move, with a weaker notion of sameness, could be made here. (Maybe it has been made here?) But it certainly would be nicer to stick with straight Leibnizian identity and not appeal to anything else.

I think I'm getting it now.

Your logic of vagueness definitely doesn't have to say that if it's determinate that there are real Martian boulders then there is a Martian boulder that's determinately real. There could be two things, and it's determinate that one of them is a real Martian boulder, but indeterminate which one. So I don't think you need to worry about that.

I want to say that you can get everything you want with just vanilla Leibnizian identity too. If I'm thinking generically about a boulder on Mars then it's indeterminate which Martian simples compose it, and maybe it's even indeterminate if any do (and so indeterminate whether it's real). And if you go to Mars and find a boulder, it's indeterminate whether it's the one I was thinking about or not. But it's still determinate that boulders are identical iff they're composed of the same parts.

Also, if two people's thoughts about Martian boulders are appropriately related to exhibit intentional identity (in Geach's terminology), or if they've got the same Martian-boulder-mode-of-presentation as a constituent, then we could say that this is sufficient for us to be determinately thinking about the same boulder, even if it's indeterminate which boulder we're thinking about (or even if it's a real one).

That kind of phenomenon needn't be different in kind from how it can be indeterminate what height is the cutoff for "tall", but determinate that it's the same for your tallness thoughts and my tallness thoughts.

Yes, I see what you mean. Thanks, this is very helpful!

On the logic of vagueness point, I just wasn't sure whether standard logics for vagueness allowed that sort of thing. Are you saying that they do? I need to read more on this before writing the final version.

I'm still a little worried that since on my view your thinking of the boulder makes the boulder exist we might get the kind of vagueness that gunks up the logic: it's determinately true that your boulder exists, and it's determinately true that my boulder exists, but it's indeterminate whether they are identical. Isn't this enough to raise problems via Leibniz's Law? For instance, it's determinately true that your boulder is thought of by you but indeterminate whether my boulder is thought of by you. By Leibniz's law, the boulder will have both of the properties being determinately thought of by you and being such that it is indeterminate whether it is thought of by you, which looks like a contradiction. I guess, though, that you don't get this effect unless you are treating being determinately though of by you as a property that's in the scope of Leibniz's Law. If we insist that determinately/indeterminately can function only as sentence operators, we might not be able to run the problem. On the other hand, if we have abstraction in our language we won't be able to enforce that restriction.

On logics of vagueness: yes, that is what I'm saying. A popular way to go is supervaluational: you have a set of sharp (ie not vague) models, and something is determinately true iff it's true at all models in the set. A different thing could be the real Martian boulder at each sharp model, which gets the result you want. There might be logics of vagueness where things don't go as smoothly, but the supervaluational stuff is pretty mainstream.

I think you're right to still be a little worried - it does change things that thinking about the boulder makes it exist. The way you set up the problem is a lot like Evans's argument against vague objects, but I think even people who think semantically indeterminate a=b sentences are sometimes OK could still worry in this case. The David Lewis of "Vague Identity: Evans Misunderstood" would probably still worry.

Put it this way: Suppose only two people ever think about boulders on Mars, each thinks about one, and it's indeterminate whether it's the same one. Then it's indeterminate whether there's one boulder on Mars or two. I know plenty of people who'd be fine with that, but I do live in Leeds.

Hello again,

I had an idea for how to solve the problem we were just talking about, and while you might not like it I thought I'd share it in case you did.

If you didn't want it to be indeterminate what exists, you could get on the Composition As Identity bandwagon and say that the wholes that we unify when we think about some things as one thing just are the parts. It isn't indeterminate what things there are; it's just indeterminate which groups of them are a whole.

This would be a version of the restrictivist composition as identity suggested by Ross Cameron in his "Composition as identity doesn't settle the special composition question".

I guess someone might say that this is still a case of indeterminacy in what exists, because it's still indeterminate whether the wholes exist. I don't really think that's right, as the things the wholes would be do determinately exist - they're the pluralities - it's just indeterminate whether they're eligible values of singular variables. We're already used to it being indeterminate whether things are eligible values of personal variables, and you could say this was more of the same.

You might not like it because you think composition as identity is ridiculous, though. Lots of people do. (Not me, although I prefer the universalist brand.)

But composition as identity requires many-one identity, and I'm trying to avoid being revisionary about the logic of identity. I think vague identity would be a smaller revision than many-one identity.

I don't really think many-one identity demands any more of a revision of the logic of identity than mental-physical identity, but I know the issue's delicate at best. I thought it was worth mentioning, anyway :)