Friday, February 19, 2010

Holism and Metaphysical Grounds

[Some thoughts inspired by Shamik's recent talk 'On the Plurality of Grounds'...]

Suppose comparativism about mass is true, i.e. "the fundamental facts about mass are facts about the mass-relations between things". Absolute mass claims, such as that my laptop is 2 kgs, must ultimately be grounded in mass-relational facts, such as that my laptop is twice as massive as the standard kilogram weight in Paris. Of course, no single such relation suffices: the standard kilogram might have been twice as heavy as it actually is (in comparative terms: mass-related to the vast majority of other objects at double its actual ratios), in which case the aforementioned relation would imply that my laptop was 4 kgs rather than just 2. So the full ground of my laptop's being 2 kgs must instead be massively plural, perhaps depending on the entire universe of comparative mass facts.

At this point, Shamik worries that the universe of comparative facts can't ground the single absolute fact about my laptop's mass, because it includes facts that are (intuitively) explanatorily irrelevant. For example, the mass relation between my laptop and some particle on Alpha Centauri seems irrelevant to my laptop's being 2 kgs. Shamik's solution is to introduce plural grounding: no single absolute mass fact has a ground, but the set of all such facts is plurally grounded in the set of all relational mass facts. Neat enough, but I wonder whether we should really accept the intuition of explanatory irrelevance that prevents us from simply grounding each individual absolute mass fact in the set of all relational mass facts.

Granted, the mass relation between my laptop and some Alpha Centauri particle clearly doesn't play an especially central role in the explanation of my laptop's being 2 kgs. But if we accept comparativism, I don't see the basis for insisting that this relation must play strictly no role at all. After all, it's pretty natural to interpret comparativism as a radically holistic view. For example, it's not enough even to look at all the intra-Earthly mass relations -- if those all stayed the same, as did strictly extra-terrestrial relations, but extra-terrestrial objects were all doubly mass-related to terrestrial objects, I take it that this is most naturally described as everything on Earth shrinking, rather than everything else in the universe doubling in mass. (Not that these are really two distinct possibilities, according to comparativism. But the point is that all the local mass relations, that in the actual world partially ground my laptop's being 2 kgs, would in this scenario instead ground my laptop's being 1 kg.) Given such an holistic picture, it seems natural to say that every part of the whole -- every mass-relation in the universal set -- is at least slightly relevant. After all, if the mass-relation between my laptop and the Alpha Centauri particle, along with a whole lot of other laptop-ET mass relations, had been halved, then my laptop would weigh 1 kg rather than 2. So that fact that all of those laptop-ET mass relations aren't halved is explanatorily relevant to my laptop's weighing 2 kg after all.

As an analogy, consider semantic holism: the view that the meaning of a word is fixed by the totality of our dispositions regarding its use (rather than by any privileged subset of 'analytic' inferential dispositions). One could go Shamik's way and say that this is a case of plural grounding, whereby one's entire vocabulary set (as a collective) is grounded in one's total set of linguistic dispositions. But has anyone ever thought this necessary? (I'm totally unfamiliar with the relevant literature, so that's a genuine question.) I would've thought that most people would be happy to allow the semantic holist to say that each word's meaning is (individually) grounded in the totality of their linguistic dispositions. And if that's right, presumably it's legitimate for the comparativist to say the same sort of thing about the holistic grounding of absolute mass claims in comparative mass relations.


  1. [Shamik responds:]

    Thanks Richard, these are really interesting points. And sorry for taking so long to respond, I got caught up in meeting a deadline.

    Anyway, I didn't do a lot to argue against the view that my laptop's being 2 kgs is grounded in the totality of all mass relations it stands in, other than say "surely lots of those relations (e.g. to electrons in Alpha Centauri) are irrelevant". So you're absolutely right to press me on this. But I don’t think that it’s being 2 kgs is grounded in all those mass relations, so let me say why.

    First, a pedantic point. You argue like this:

    If X had been different, Y wouldn't obtain
    Therefore, X is part of Y's ground

    where X are facts about the mass-relations between my laptop and extra-terrestrials, and Y is my laptop's being 2 kgs. I just want to point out that this form of inference isn't valid: counterfactuals can pick up on all sorts of dependency relations including (famously) causal dependency, but X can cause Y without grounding Y. So, in the case that X causes Y, it might be that the premise is true but the conclusion false.

    But I don't want to rest anything on this, since I can see a reading of the counterfactual you use that does appear to pick up on grounding relations rather than causal relations (and I take it you have that reading in mind). And I agree that it is tempting to take that counterfactual to motivate the view you outline, i.e. that my laptop's being 2 kgs is grounded in the totality of all its mass relations. But I think that my view predicts and explains the truth of the counterfactual, so its truth doesn't really tell in favor of the view you outline. Also, I think that the view you outline is unattractive for other reasons. I’ll say why below. None of this will be at all definitive (I don't think there are "knock down" arguments to be had in this area of metaphysics), but it might serve to show that, all things considered, there's something to be said in favor of the "pluralistic" view I'm in favor of over the view that kilogram facts are individually grounded in the totality of all mass relations.

    First, how does my view predict the truth of the relevant counterfactual? The relevant counterfactual is something like this: "if the mass-ratios between all extra-terrestrials and my laptop were doubled, but the mass-ratios between all terrestrials and my laptop were the same, then my laptop would be 1 kg (rather than 2 kgs)". To evaluate this, we have to think about how to compare masses, and in particular kilograms, across possible worlds. Now, as you yourself point out, *any* comparative view (whether pluralist or not) will say that (strictly speaking) mass-relations do not hold between objects in different possible worlds (actually, a comparativist who's also a modal realist in Lewis' sense might disagree, but I'll put this complication aside here). Moreover, since on the pluralist view I defend there is no relational matter of fact that grounds my laptop's being 2 kgs, it follows that given any non-actual possible world, there is (strictly speaking) no fact of the matter whether my laptop is 2 kgs.

    So how do we evaluate the counterfactual? With a "mass-counterpart" theory, which runs analogously to normal counterpart theory. In normal counterpart theory, we use a "counterpart relation" which is not identity but which stands in for it when evaluating de re modal claims. In mass-counterpart theory, we use a "mass-counterpart" relation which is not the same-mass-as relation but stands in for it when evaluate claims about how the masses of things could have differed and (in particular) how their mass-in-kilograms could have differed. And, analogously to normal counterpart theory, two objects in different possible world are mass-counterparts when they are resemble one another with regards their mass-relations.

    [to be continued...]

  2. [Shamik continues...]

    In the possible world W you describe, the mass-counterpart relation is naturally taken to hold between the extra-terrestrials in the actual world and the extra-terrestrials in W. The last piece of the puzzle is something like the following principle: that if x actually is r kgs and has a mass-counterpart y in W, then y is (or, perhaps better, is to be described as being) r kgs too. It follows that all the extra-terrestrials in W have (or, perhaps better, are to be described as having) the same mass-in-kilograms as they actually have, and so it follows from the mass-relations in W that my laptop is 1 kg in W. In this way, my view about plural grounds predicts and explains the data you appealed to in order to motivate the view you outlined.

    So that’s why I think the counterfactual you appeal to doesn’t settle things. Of course, the view you outline predicts that counterfactual too… but I think it is unattractive for other reasons. Why? Well, the view was that my laptop's being 2 kgs is grounded in the totality of all its mass relations. But now consider a possible world W in which the mass facts are ever so slightly different---perhaps a single particle in Alpha Centauri is a little more massive than it actually is. Intuitively, my laptop is still 2 kgs in W, but what is this grounded in? On the view we're supposing, my guess is that the natural thing to say is that it's grounded in the totality of all my laptop's mass relations in W. So it's having 2 kgs has one ground in the actual world, and a different ground in W.

    I think this is an odd consequence. It's not that what grounds my laptop's being 2 kgs is a contingent matter, for I think there are plenty of examples of contingent grounds. For example, the disjunctive fact that snow is white or grass is purple is grounded in snow's being white, but if snow were black and grass were purple it would be grounded in grass's being purple instead. But in this case, we have a systematic explanation of why the grounds should be contingent in this way, namely the nature of disjunction. I’m not sure whether we have nearly as convincing an explanation of why the grounds of my laptop’s being 2 kgs should be so contingent in the way that the view you outline implies. My view avoids this problem: for on my view the primary question is what grounds the plurality of all facts about mass-in-kilograms, and of course any relational difference between worlds will give rise to a difference in some part of that plurality. So we are never in a position where one and the same plurality of kilogram facts has a different mass-relational ground.

    One last point that’s worth keeping in mind here: the idea that my laptop's being 2 kgs had anything to do with mass-relations is, I think, the result of theory. I think that the *intuitive* thing to say is that my laptop's being 2 kgs is grounded in its having a certain intrinsic mass. Speaking for myself, my pre-theoretical intuitions say that its being 2 kgs has *absolutely nothing* to do with particles in Alpha Centauri at all! Being 2 kgs is, intuitively at least, an intrinsic property of my laptop that it has in virtue of the nature of the substances that compose it. Now, of course, as a comparativist I cannot say this, but I think that pluralism gives the comparativist the best way to preserve that intuition: it preserves the intuition that my laptop's being 2 kgs has nothing to do with its mass relations to other things. That's true on my pluralist view, but it's true not because its being 2 kgs has to do with its having an intrinsic mass; rather, it's true because its being 2 kgs doesn't have any kind of ground at all! On the view you outline, one has to bite the bullet and give up the intuition.

    [end quote]


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)