« Saturday Night at the Oldies: Baez Sings Dylan | Main | Summer's Here »

Sunday, March 13, 2011


Feed You can follow this conversation by subscribing to the comment feed for this post.

Hi Bill,

Thanks for the post. I am sympathetic to the argument, and in fact I think that if is sound, it would prove to be one of the most important arguments of your book, for reasons I will cover at the end of this comment; but one thing is unclear to me.

I suppose you are arguing against a "groundless" theory of facts (GT) as follows. The GT states (a) There is no ontological ground of the unity of a fact's constituents, and because the GT-ist by hypothesis grants the difference between a fact and the sum of its constituents, he accepts (3). But somehow, some way, the GT-ist must accept (2), which contradicts (3); so GT is false.

I am not clear on why the GT-ist must accept (2). Is there an attempt to infer (2) from (a)?

It is true that the unity of the fact will not be "discoverable" by analyzing the fact, for familiar Bradleyan reasons. It won't be interior to the fact. And we won't discover it by external means such as sense perception. This is all just a consequence of the unity being brute!

But does this show that (a) entails (2)? The GT-ist could, it seems to me, simply say: Yes, the unity of a fact is incapable of discovery by analysis, nor is it empirically detectable. When you analyse a fact, all you get are constituents, not them unified. But nonetheless, the fact is something more than those constituents, and nothing accounts for this "something more."

My problem, therefore, is understanding the inference from (a) to (2).

Or, are you not trying to infer (2) from (a), but rather are trying to argue that the only way to have reason to reject (2) is if we admit of an ontological ground for the unity of a fact's constituents? Again, I would have to say it isn't clear to me why this is so!

I admit that I have framed the discussion in a different way than you did in your post, and this is perhaps the cause of the lack of understanding on my part.

- - -

I said above that this argument would prove to be quite important if sound. Note the following interesting consequence of this discussion, apropos to something else we were discussing Friday. You and I both believe that if we admit a ground of unity of facts (or more broadly, of ontological constituents of concrete particulars), we must admit an external unifier who is God. You and I think that there is a tight logical relation between fact or constituent ontology and theism, if we admit there is a ground of unity. But the problem for us is that it is not clear there must be a ground of unity, as much as the belief that there is a ground seems eminently plausible to us. One could remain a fact ontologist or a constituent ontologist and an atheist, if there need not be a ground of unity. But if there does have to be a ground, as your above argument attempts to show, then the constituent ontologist cannot also be an atheist. And if someone is an atheist, therefore, they must be a non-constituent ontologist.

But I remember you argue non-constituent ontology is incoherent. Then it follows that atheism is incoherent, pace our good friend Peter Lupu. Would you, if we manage to get the above argument cleared up (at least to my mind!), mind going over some of your arguments against non-constituent ontology?


Thanks for the prompt response. You ask why the GT-ist must accept (2), i.e., 'A fact is its constituents taken collectively.' Well, because when we analyze a fact all we find are its constituents. It's like King Milinda and his chariot. The Buddhist monk Nagasena demanded of the good king that he show the monk the chariot, when all the monk could find were wheels and hubs, etc.

What I am saying is that we and the GT-ist have good reason to accept (3) but also good reason to accept (2) -- despite their being logical contradictories.

I don't know what you mean by '(a).'

You make a good point: if the unity of a fact's constituents is a brute unity, then of course there will nothing internal to the fact that accounts for the unity.

Again, what does your '(a)' refer to? My 'a' is a logically proper name of an individual. It does not denote a proposition. One cannot infer anything from an individual.

So I'll stop here. I can't proceed until you tell me what '(a)' refers to.

Sorry, Bill, I should have been more clear. (a) is:

The GT states (a) There is no ontological ground of the unity of a fact's constituents...,

(a) is: There is no ontological ground of the unity of a fact's constituents.

Oh, I thought you were referring to an 'a' in my post. You are asking whether I am inferring
2. A fact is its constituents taken collectively
a. There is no ontological ground of the unity of a fact's constituents.

No I am not. And it wouldn't follow. For one could hold that there is a unifier and it is a constituent of the fact.

My point is that we have as good reason to accept (2) as to accept (3) despite their being logical contradictories. I take it you accept (3). But don't you feel at least some inclination to accept (2)?

More later.

Okay, but now I'm having a hard time understanding what the argument against (a), i.e., the groundless theory is!

You're not, as you say, inferring the false (2) from (a). As you say, this inference wouldn't follow: "For one could hold that there is a unifier and it is a constituent of the fact." [But I think that's mistaken; after all, if the unifier was a constituent of the fact, there would then be an ontological ground of unity, and hence (a) would be false.]

But if you're not inferring (2) from (a), I don't understand how any of this is an argument against the GT-ist. After all, what have you said that a GT-ist couldn't say? Why couldn't the GT-ist say: "Ah, yes, there is more to a fact than its constituents: it is the constituents unified. But nothing accounts for or grounds this unity."

Are you trying to argue that the only way we can reject (2) is if we accept that there be a ground of unity, and hence (a) is false?

My problem is understanding the connection of the whole argument to (a), the groundless theory.


You didn't answer my question: Do you or do you not feel inclined to accept (2)? My point is that both (2) and (3) are reasonably accepted. Do you agree? If not, my argument will leave you cold.

How do you know that (2) is false?

Answer these questions and then we can proceed.

In response to your questions:

I think I must not be getting something, because I don't find myself at all compelled to accept (2). You're right, you don't find the unity of the fact just upon reflection of its constituent parts, but that hardly seems to me to be reason to accept that a fact is just the parts together; you don't find the unity of a chariot by reflecting upon its various parts, but if you know what a chariot is, you know it is not just a bunch of parts taken in collection.

Or, if that is some kind of reason to accept (2), it is only prima facie reason to believe (2); for upon further reflection one will find defeaters for (2). So belief in (2) may be reasonable, but only prima facie. Upon further reflection, it can easily be seen (2) is false.

I know (2) is false because there is clearly a difference between a mereological sum of some fact's constituents and the fact itself: the unity of the fact being what the difference is.

Thanks for the nice, clean, direct response. You say you know that the following is false: a fact F = F's constituents. Let F be the relational fact Rab, e.g., Al's loving Betty. Then an instance of what you claim know is that Rab is not identical to R + a + b. (X + y is the mereological sum of x and y.) But some philosophers accept Extensionality of Composition, the principle according to which, if x and y are composed of the same things, then x = y. Now Rab and (R + a + b) are composed of the same things, hence fact and sum are identical.

This is not plausible as it stands, but is more plausible if you add the thesis of the Relata-Specificity of Relations: a relation is relata-specific iff it is in its nature to relate specific relata. If R in our example is relata-specific, then, as soon as it exists, it relates a and b. Consequently, R could not have existed and failed to relate a and b.

An objection has been made to my view along theses lines in "Relata-Specific Relations: A Response to Vallicella," Dialectica 62, 4 (December 2008) p. 509-524

So if one combines Extensionality of Composition with Relata-Specificity of Relations, then there is an argument for (2), and this would undermine your claim to know that (2) is false.

I have a few objections to such a view.

1. The argument from Extensionality of Composition conjoined with Relata-Specific Relations to (2) is invalid. Even if R could not exist and fail to relate a & b does not show there is no difference between the mereological sum R + a + b, and the fact Rab. Rab is still a structured entity, the mereological sum is not.

2. If R could not exist and fail to relate a + b, then Rab exists just as necessarily or contingently as R does. But an abstract critter like R would exist necessarily if at all, whereas Rab obtains contingently. Hence there must be a distinction between R's existing and R's obtaining, or relating, and thus there is room for Bradley-style worries about R operating as a unifier of a fact's constituents, even if R could only operate upon a & b.

3. A single ball is round, a certain weight, a certain color. But upon heating it, it loses the roundness and perhaps its weight while keeping its color. Heat it some more, then it loses its color as well. If relations are determined by their nature to relate the relata they do, then there must have been, in a single ball, 3 relations: one to relate the particular and some particular color, some particular weight, and some particular shape. This is uneconomical, given that that the ball has many more properties than just those three; we'd need who-knows-how-many relations to exist for any one of the particulars that exist in the actual world, not to mention that there are lots of particulars in the actual world. Positing a single relation which is a universal, i.e. not determined by its nature to relate any particular relata, is more economical.

4. There are a number of relations which appear to me to not be relata-specific in the above way suggested.

I exemplify whiteness, and so did my grandfather and Socrates (let's say). If numerically the same relation relates all of us to our whiteness, then this means no one else besides the persons that actually exemplified whiteness could have exemplified whiteness. But there are possible white persons.

Another example. If God exists, then everything stands in the relation "being created by" to God. But because of the relata-specificity of this relation, nothing else besides what God actually created could have been created by God. But this is contrary to divine omnipotence.

The comments to this entry are closed.

My Photo
Blog powered by Typepad
Member since 10/2008



February 2024

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29    
Blog powered by Typepad