Monday, September 24, 2007

From Ockham to nihilism?

Some thoughts I've been having as a result of a conversation with Daniel Nolan:

Does Ockham’s razor give us reason to be mereological nihilists? Ockham’s razor tells us not to multiply entities beyond necessity. A popular argument for nihilism is that mereologically complex entities don’t do anything simples arranged a certain way wouldn’t do on their own. That’s why Merricks, for example, believes that the only complex entities are conscious ones: he thinks that consciousness is a property that can’t be had collectively by a plurality of simples, so there needs to be a conscious mereologically complex object; but unconscious complex entities like tables and chairs wouldn’t do anything simples-arranged table/chair-wise on their own wouldn’t do, and so it would be a violation of Ockham’s razor to admit their existence.

But is it true that tables don’t do anything that collections of simples arranged table-wise wouldn’t do on their own? One thing I think my table does is stop my glass – which is, I think, sitting on it – from falling to the floor. If there were no table would the collection of simples arranged table-wise do this on their own? You might think not, because the closest possible world in which the simples arranged table-wise exist but the table doesn’t is one in which the glass also doesn’t exist, and only the simples arranged glass-wise are being kept from falling to the floor (or, rather, the simples arranged floor-wise).

An assumption here in the above is that a nihilistic world is closer to our world (assuming nihilism is in fact false) than a world where some of the collections that would compose in our world compose but some don’t. Why believe that? Well if the simples arranged table-wise don’t compose anything but the simples arranged glass-wise do then it would seem to be an entirely arbitrary matter whether a collection of simples compose something. We’d like to be able to explain why some collection composes or fails to compose some thing by saying something like ‘they compose something because every collection composes something’, or ‘they don’t compose anything, but no plurality of things ever composes some thing’, or ‘they compose something because they’re close enough together’, or some such thing. There doesn’t seem to be any natural condition that the parts of the glass meet but the simples arranged table-wise don’t meet, however; so in the world in question, composition seems to be an arbitrary affair. The nihilistic world – despite being unlike our world in many ways – is like ours (or like we hope ours is) in that composition occurs or fails to occur systematically: there is intra-world supervenience of the composition facts on the non-compositional facts. Perhaps that’s enough to give some reason for thinking that the nihilist world is closer to ours than the ‘mixed-world’. In any case, let us grant the assumption for the sake of argument.

Does it follow from that assumption that there is no Ockham’s razor argument for nihilism? The nihilist might object that the argument is question-begging, since it assumes that some of the work that is to be accounted for is the work of keeping the glass from falling to the ground. We should, the nihilist might counter, start from a neutral ground, in which case the datum to be accounted for can only be described as that the table-like simples are keeping the glass-like simples from falling to the ground-like simples.

But wouldn’t that be equally question begging on the part of the nihilist? Why shouldn’t I appeal to the table’s ability to stop the glass itself – not just the simples arranged glass-wise – from falling to the ground as something that the table itself – not just the simples arranged table-wise – does? After all, I think it’s true that there is a glass on a table and that it would fall to the ground were the table not there. Isn’t the nihilist’s insistence that I only admit objects to explain facts that the nihilist accepts as true simply stacking the deck in favour of nihilism?

The nihilist might question my reason for believing that to be true. Fair enough – obviously I should have a reason for taking as true the truths I want my ontology to ground. Here’s my reason then: I can see the glass on the table, and by induction I know that were the table not there the glass would fall to the floor. The nihilist denies that this is what I see, of course; but why let her set the debate by accepting her description of the phenomena rather that mine?

Anti-nihilism, one might think, is the default view. We have to be argued away from a belief in tables in chairs towards a nihilist ontology; we don’t have to be argued from a nihilist ontology towards a belief in tables and chairs. The burden of proof is on the nihilist, not the compositionalist. If that’s right then it’s perfectly appropriate for me to take as a datum that the glass is held up by the table and to try to explain it. I grant that if one can explain that datum with a nihilist ontology then there’s an Ockham’s razor argument for nihilism over compositionalism; but the fact that one could explain a nihilistic paraphrase of that datum with a nihilist ontology is neither here nor there. Since it is doubtful that a nihilist ontology can explain why my glass is held up by the table, it’s not obvious that there’s even a pro tanto reason for nihilism by way of Ockham’s razor.

6 comments:

Alex said...

You argue: (P1) a thoroughly nihilistic world is closer to our world (so long as we assume that nihilism is false at our world) than a world where some of the collections that would compose in our world compose but some don’t, and (P2) if so, then tables have effects that collections of simples arranged table-wise wouldn’t on their own.

I'm confused by how you support (P2). To show that (P2) is true, you need to show that if the thoroughly nihilistic world were considered as actual, then the simples arranged table-wise would fail to prevent the glass from falling. But by hypothesis, there is no glass in such a world: so how could this be a world in which the simples fail to prevent the glass from falling? Moreover, you do not claim that they would fail to prevent the simples arranged glass-wise from falling. So I don't see what effects they would fail to have in such a world that the table would have in the actual world. So I don't see why I should think (P2) is true.

(I'm also confused by how you support (P1), but students are knocking down my door to talk about their papers...So I'll save it for another post.)

Ross Cameron said...

I was thinking it was enough to show that (P2) is true that there is some effect that wouldn't obtain were the table not to exist. That effect is: the glass not being held up. There are two ways for it to fail to be the case that the glass isn't held up: the glass could exist and fall, or it could not exist. In all the nearby worlds in which the table doesn't exist neither does the glass (grant (P1) for the sake of argument here), so were the table not to exist that effect wouldn't obtain. Hence there's something that the table does that wouldn't be done were it not to exist.

Mike said...

If there were no table would the collection of simples arranged table-wise do this on their own? You might think not, because the closest possible world in which the simples arranged table-wise exist but the table doesn’t is one in which the glass also doesn’t exist, and only the simples arranged glass-wise are being kept from falling to the floor

But isn't the Ockhamist claiming that in the closest worlds where there is both a glass and simples arranged table-wise, the arranged simples would do just what the actual table does? It doesn't matter to the Ockhamist's argument that this mixed world is farther from us than the closest nihilistic world. The mixed world shows that the arranged simples can do all the work of the tables. Or they can do the glass-holding work anyway.

Ross Cameron said...

I think it does matter Mike: to show that y is doing genuine work I just need to show that were y not to exist, the work wouldn't be done. That's the case in my example, if the mixed-world is indeed further away than the nihilistic world.

Now it's true that y isn't *needed* to do the work, in that it could be done in worlds without y. But the fact that y isn't needed to do the work doesn't mean that it's not in fact doing it. It *is* doing it, because without y it wouldn't be done.

I guess it comes down to what we take Ockham's razor to be. I was thinking that the compositionalist only violates it if the complex objects don't in fact do anything. On this understanding, I don't think there's a violation. You might have a stronger reading that said not to believe in a thing if the existence of that thing isn't necessary to do the work that needs to be done. Then I think the compositionalist does violate it. But I also think that's far too strong a principle.

Maybe there's something in between in strength. Did you have something like that in mind, or am I misrepresenting you?

Mike said...

I think it does come down to how we take the Ockhamist principle. I simply assumed it was some variant on the modal claim, "entities should not be multiplied unnecessarily". I take it that anything like this version you think too strong. I might be misreading you, but I'm a little confused. You say,

I was thinking that the compositionalist only violates it if the complex objects don't in fact do anything.

If this is right, then your counterfactual situations produce misleading results. It can be true that e and e' together actually explain D, but the closest ~e world be one in which D holds too. So you have both e and e' doing actual work, despite the fact that, in the closest ~e, worlds the work gets done anyway. Certainly the effort of Smith and Jones in moving the chair explain the actual movement. But it could easily be that in the closest worlds where Smith fails to exist, Jones moves the chair anyway. You don't want to say, I'm sure, that the actual explanation for the chair moving event does not include both Smith and Jones.

Ross Cameron said...

Hmmm . . . I actually think I recant. I was thinking that the table did the work but that the simples arranged table-wise didn't, so that it wasn't a case of over-determination. While I still think that the table does do the work if it exist, I think I was to think that the simples only do it if the table doesn't exist. If the test is 'would the work be done were the thing(s) not to exist?' then the simples plausibly pass. Because were the simples not to exist the table wouldn't either. I think there is a possible world where the simples don't exist but the table does (the table is an extended simple in that world), and in that world the table still holds the glass up. But I think (for reasons analogous to those above) that that world is further away than the world where neither the table nor the simples exist. So it's true that were the simples not to exist the glass wouldn't be held up. So they're both doing the work if the table exists, so it is a case of over-determination, so I do think Ockham's razor gives us some reason not to believe in the table.

Now I'm finding it hard to think back into my original mindset, though, so I may be missing something.