Monday, September 05, 2005

Truthmaking

I am a bit puzzled by the modal claims often made concerning the relation between a proposition's being true and its truthmaker existing. Here's one of the sources of my puzzlement.

Suppose you think facts (actual worldly states of affairs) make propositions true. In particular, suppose you think that the fact that Mu and Marks purr is what makes true the proposition All Carrie's cats purr.

Why should you think that it is impossible for the same fact to obtain without making proposition true? Consider a world where I have two more cats besides Mu and Marks, and one of these other cats does not purr. In this world, the fact that Mu and Marks purr still obtains, but does not make true All Carrie's cats purr.

I guess it could be said that the truthmaking fact in the actual world is not the fact that Mu and Marks purr but the fact that all Carrie's cats purr. But I wonder whether this sort of fine-grained individuation of facts is in keeping with the conception of them as worldly states of affairs (which, presumably, are supposed to have some degree of independence of the different ways in which we can represent them).

Why do we want a modal tie between the existence of the truthmaker and the truth of the proposition, anyway? We want truthmakers to have some special relationship to the propositions they make true, of course, but why this sort of special relationship? Wouldn't it be better to drop the claim that a proposition p's truthmaker is something that could not exist without making p true, and say instead that p's truthmaker is that thing in virtue of whose existence p is true? (NB: I intend these all as genuine, i.e. not merely rhetorical, questions!)

24 comments:

Ross said...

I agree that whatever makes it true that all your cats purr could exist in a world where you had an extra cat who did not purr. (See Cameron (2005) 'Truthmaker Necessitarianism and Maximalism')

Armstrong has an argument that we need the relationship between the TM and the proposition to be necessitation, but it presupposes the claim that every truth has a truthmaker, which should be rejected.

I like the 'in virtue of' suggestion; in my paper I argued for a restricted version of necessitation, but if the in-virtue-of theory went through I'd be happy.

Carrie Jenkins said...

As I read your paper, it says that Armstrong argues that necessitation plus maximalism requires the existence of various higher-order facts, but you can avoid this by rejecting maximalism. An actual Armstrongian *argument* for necessitation seems though to be present in the little passage you quote at the end of your paper.

The sort of thing I had in mind could be respected by saying, not that Mu and Marks are all Carrie’s cats has no truthmaker, but that whatever fact makes it true is modally independent of the fact which at the actual world makes true All Carrie‘s cats purr (so that the latter fact could obtain at a world where the former did not - and indeed where no other fact made true Mu and Marks are all Carrie’s cats either). (I’m not particularly keen to defend truthmaker maximalism or anything, just exploring options.)

If I understand it, your response to the Armstrong passage that you quote at the end of your paper is that the kind of ‘full’ (necessitating) truthmaker Armstrong wants is not available in many cases. It would be even more satisfying, though, if in addition to saying why we can’t always provide what A wants, we could also say why it would be a mistake to want that sort of thing even supposing it were available. (This is my intuition.)

The universally quantified examples seem to me to provide a good prima facie reason not to hanker after ‘full’ (necessitating) truthmakers. So all we need to do is say what’s wrong with the argument in favour of them. In the little passage you quote, A says that a contingent truthmaker is incomplete: because it only operates in contingently obtaining circumstances, those circumstances must be added in to give the full truthmaker. But why should we think that you have to add in the relevant contingent circumstances in order to give the ‘full’ truthmaker? That would obviously follow if we already accept TM necessitarianism - we haven't given the full truthmaker if its existence doesn't necessitate the proposition - but this was meant to be an argument for necessitarianism …

I guess non-necessitarians do face a challenge, which A may be trying to avoid: that of saying what the difference is between worlds where F makes p true and worlds where it doesn't. But that difference seems to have to do with the obtaining of other facts, and with the way propositions relate to the world. In my example, what makes the difference is that the target proposition contains a universal quantifier and in the one case Mu and Marks are all my cats and in the other they're not.

Andrew said...

Isn't the idea that P's truthmaker explains the truth of P? Then the intuition is that answering

Why is "All Carrie's cats purr" true?

with

"Because Mu purrs and Marks purrs"

isn't to give a full explanation, since it couild easily be the case that Mu and Marks purr without the general proposition being true, as you note. (The explanation seems enthymematic.) Now obviously, this presupposes a certain conception of what metaphysical explanation amounts to. But I think this thought is what leads Necessitarianism to seem attractive. Now obviously the 'in virtue of' locution is intended to supply a form of explanation too, so I guess the way to proceed is to get clearer on what's different about that form, and whether it supplies all of what we wanted in the way of metaphysical explanation of the truth of a proposition. (A natural distinction is between what actually makes a thing true - in the sense of what grounds its truth - and what it is for that thing to be true - what it's truth essentially consists in, if you like. The latter looks closer to requiring necessitation, although it's finer grained. I'm working up a paper on this stuff at the moment).

Carrie Jenkins said...

Hi Andrew,
That's interesting stuff - I'd be keen to see your paper if you're willing to send me a draft. I'm very interested in the ways necessitation, explanation, essence and IVO claims interact in general. I guess I'm tempted to think that what's required for a good explanation (metaphysical or otherwise) depends a lot on what the audience already knows and what they want from the explanation, but that most explanations needn't necessitate their explananda. One might think that's because they're elliptical, and that the 'full' explanations *do* necessitate the explananda. But for one thing, doesn't that lead to weird results (a lot of inexplicable facts) if the laws of nature are chancy?

Andrew said...

Hi Carrie,

I think that it's a more ontological sense of explanation that these people have in mind. The truthmaker itself is the explanans, not some statement involving our concept, theory or characterization of it. Explanation in the latter, more subjective sense may be partial, interest and context dependent, and relative to prior audience knowledge, etc, but I don't have a good grasp on how explanation in the former sense could be (partly because it's unclear what that kind of explanation could be, but also because the former, objective type of explanation needn't e.g. involve any audiences, interests, and so on).

I'm not sure I understand your laws of nature point in this context. It could certainly be chancy whether e.g. a truthmaker exists (or obtains, etc), but given that it does exist, it will be available to be the relata of the objective explanation relation. Is the idea that explanans in general needn't e.g. causally necessitate their explananda? That seems right, but these people are going to say (with some plausibility) that the case is quite different in the case of e.g. constitutive necessity. If our aim is to say what it is for a general proposition to be true - where that is heard in an ontic rather than conceptual way - then what would play the role analogous to that of indeterministic laws in the causal case? Chancy essential/constitutive relations? If we thought those could obtain, we would have to commit ourselves to claims like:

What it is to be an F is just, constitutively, to be a G and yet there's a chance C that something could be an F and not a G, or vice versa.

That seems to me not to capture what these people are after when they talk about e.g. constitutive relations. It's a bit like saying that G is the real definition of F, but that there's a chance that something could be G and not F. Seems to me that that just wouldn't deserve to be called a real definition relation.

Anyway, lots of difficult issues. I'll send you a copy of the paper when it's tidied up a bit.

enjoy the blog, btw!

A

Carrie Jenkins said...

Andrew,

I'm not sure I know what explanations are unless they're explanations for some audience. It sounds like metaphysical *grounding* is really what these people have in mind.

The point about chancy laws was much more simplistic than the things you mentioned (!) - it was just meant to illustrate that you could give all the information about the state of the world just before event e, and all the information about the laws governing the world, without giving information which necessitates e. So either e is inexplicable or explanations don't need to necessitate their explananda.

And then the thought would be: if some explanations do not necessitate their explananda, why should truthmaking explanations have to?

Andrew said...

It seems to me that there may well be e.g. causal explanations of cetain physical events that we will never be able to actually offer, and maybe no idealized inquirer would be able to either (perhaps the explanations are infinitely complicated, and infinite inquirers are conceptually or nomically impossible, or maybe the explanations are nomically such that getting into a position to identify one of their aspects prevents us from being able to discover another. These causal explanations won't be for audiences. But it still seems quite natural to say that there is e.g. a physical explanation of why the events occur - just one that we don't know. I don't think this kind of claim - that explanation is an objective, non-subject-relative relation - is peculiar to the metaphysical case, and it certainly doesn't seem to be reducible to a grounding relation in this type of nomic case.

Thanks for the elucidation on the chancy laws point. I guess the answer I would offer to your question if I was a truthmaker theorist would run as follows. We don't require necessitation in the case of causal explanation, because it is the natural laws that are the source of the relevant modal force, and we can make sense of indeterministic natural laws. But we require it in the case of truthmaking because truthmaking is about what-it-is for a proposition to be true, where it is a constitutive relation that is the source of the relevant modal force, and we can't make sense of indeterministic constitutive relations.

Carrie Jenkins said...

Hi Andrew,

"It seems to me that there may well be e.g. causal explanations of cetain physical events that we will never be able to actually offer ... These causal explanations won't be for audiences."

I suspect that to the extent I am tempted to think of them as (potential) explanations that's because of the truth of some counterfactual that does involve an audience. Perhaps it's because these explanations are the kinds of things that would help us understand why those events took place if we could only appreciate them.

I guess I think that once the necessitarian starts talking about essence - what-it-is-for (WIIF) claims - she's probably appealing to something a bit different from explanation. I can see how a WIIF view of truthmaking could lead to necessitarianism (though even then some people seem to think that essence claims and corresponding modal claims are independent). The question is: why assume at the outset that looking for the *ontological grounds* of truth (truthmakers) amounts to looking for the *essence* of truth? The identification of ontological grounds with essence might be an interesting conclusion to some argument, but it would be a weird thing to assume straight off.

Andrew said...

So you think that in those cases

P explains Q because were it to be the case that we offered a description of P, and idealizing conditions held, (perhaps per impossible), our audience would understand why Q.

This can't be quite right, since obviously we could describe P in unhelpful terms. So it has to be something like

P explains Q because were it to be the case that we offered a description of P under the 'P' mode of presentation, and idealizing conditions held, our audience would understand why Q.

I think these people would say that that gets things the wrong way round. It's (partly) because P objectively explains Q that were it the case that we offered a description of P under the P mode of presentation, in ideal conditions, our audience would understand why Q. If P didn't objectively explain Q, then the audience might e.g. have an illusion of understanding, but they wouldn't really understand why Q.

Wrt to your second point, I don't think that that accurately characterizes the state of the dialectic, at least from these people's point of view. It isn't that you start off looking for the ontological grounds of truth and then assume that this amounts to looking for the essence of truth. Rather, the line of thought seems to be something like this. You take the goal of a theory of truth as being the provision of a satisfying account of what the truth of a truthmaker consists in (what it essentially is). This seems quite plausible to me. You then claim that it consists in the obtaining of a certain type of relation between a truthbearer and a part or aspect of the world. You take yourself to have to say more about the nature of the relation and the parts/aspects in question. You say, a bit unhelpfully, that R is called the truthmaking relation. In the case of

(C) All Carrie's cats purr

you're reluctant to say that it consists in the relation R holding between (C) and Mu's purring and Marks' purring, because that doesn't seem to be what its truth *consists* in, since (i) (C) could be true even though Mu and Marks aren't purring and (ii) (C) could be untrue even though Mu and Marks are purring. So you take yourself to have to say more about what it does consist in. You don't want cases like (i) or (ii) happening, so you look for a relata that necessitates the truth of (C).

That line of thought may be misguided, but I think that's what's going on. It's not that they assume straight off that ontological grounds = essence.

Carrie Jenkins said...

"I think these people would say that that gets things the wrong way round."
Agreed. Unsurprisingly, I think the same about their view! (To be clear, though, it's not that I think *all* the facts about whether something is a good explanation are determined by facts about the audience, just that it's always the case that some of them are.)

On the state of the dialectic with the truthmaker necessitarian, couldn't one think that although there *are* truthmakers - ontological grounds of truth - it's not the case that some story about truthmakers constitutes one's *theory of truth* i.e. reveals what the essence of truth is? If so, then, even if (as a matter of fact) the extant necessitarians do think the correct theory of truth is some truthmaker story, and even if in order to play that sort of role truthmakers must necessitate the propositions they make true, the necessitarian won't have achieved her stated aim of arguing that something *cannot be a truthmaker* for p unless its existence necessitates p. She'll only have shown that, *if the correct theory of truth is a truthmaking theory* something cannot be a truthmaker for p unless its existence necessitates p.

Andrew said...

Somebody could adopt the position you suggest. But then the issue shifts to broader issues, it seems to me. Your original question was

Why do we want a modal tie between the existence of the truthmaker and the truth of the proposition, anyway?

I suggested that people wanted such a tie because (i) they wanted to say what the truth of a truthbearer consists in, and (ii) they want to give a correspondence-theoretic account of this, which seem to jointly pressure them towards necessitation. If that's right, then your question becomes

Why do we want to say what the truth of a truthbearer consists in? Why do we want to give a correspondence-theoretic account of truth?

To the first, I just want to say: because we are philosophers interested in the metaphysics of truth. The second is too general and complicated to answer very straightforwardly.

It's perhaps worth noting that if all you want is ontological grounding, then it's not clear that truthmaking can do all the work people have wanted from it. Say you are suspicious of the presentist claim that only while the present exists, there are past and future directed truths. It doesn't seem all that difficult just to metaphysically ground these truths - what makes it true that Socrates was snub-nosed is the present's (primitively) having had a past in which Socrates was snub nosed, and what makes it true that the sun will rise on the 9th September 2005 is the present's (primitively) having a future in which the sun rises on the 9th September 2005. So if the rules of the game only demand that we ground truths in being, presentism isn't ruled out, in the way that e.g. Lewis suggested. (Joe Melia has a paper on this somewhere - maybe his 'Continuants and Occurrents', Procedings of the Aristotelian Society, supplementary volume, 2000?)

Now, the 'in virtue of' locution seems richer than grounding. Intuitively, that the sun will rise on the 9th September 2005 isn't true in virtue of the way the present is, but in virtue of the movement of the sun and earth on that date. So there is a kind of failure of explanation - the putative explanans doesn't stand in the relevant relation to the explandum. But mere appeal to ontological grounding doesn't seem sufficient to manufacture a problem for the presentist who's content to appeal to brute temporal ways the present is. Something stronger and more explanation-like is needed.

A related point is that the presentist can straightforwardly make many of the same moves that you did in the final paragraph of your response to Ross. (What make it the case that "Will (Sun rises tomorrow" is true at this presentist world, but not at an intrinsic duplicate of it? Well, it's that the proposition contains a future-operator, and in the one case, the world has a future where the sun rises, and in the other one it doesn't. This sort of explanation is fun, but not very satisfying, especially to the non-presentist. I think the real suspicion that the non-presentist has isn't being addressed: that there ought to be a robust story about what having a temporal property consists in. Ditto with brute dispositions, etc.)

Carrie Jenkins said...

Andrew,

You write:
"your question becomes
Why do we want to say what the truth of a truthbearer consists in? Why do we want to give a correspondence-theoretic account of truth?"

I don't think that's quite it - my question is meant to be more like this:
Granted that we're interested *both* in thinking about truthmakers and in saying what the truth of a truthbearer consists in, and granted if you like that we want a correspondence account of truth, why do we think our account of what truthmakers are will answer the question of what truth consists in?

I'm not convinced that "the 'in virtue of' locution seems richer than grounding." I agree that "Intuitively, that the sun will rise on the 9th September 2005 isn't true in virtue of the way the present is, but in virtue of the movement of the sun and earth on that date." But that's because I'm not a presentist and I don't think the fact the sun will rise at that time is ontologically grounded in present facts. I'm dubious about whether a presentist who thinks there are present truthmakers for facts about the future will share this intuition. So truthmaking intuitions and IVO intuitions seem to go along together, as far as this case goes.

Andrew said...

So is the idea that you have truthmakers and truthmaking, AND correspondence relata and the relation of correspondence?

I think these people might say that they moved to truthmakers + truthmaking because we were making so little advance with the correspondence relata + relation? The former was meant to be a way of elucidating what was right/intuitive about the latter, or reducing the latter to the former.

So we think that our account of what truthmakers are will (partly) answer the question of what truth consists in because we are correspondence theorists and think that truthmakers are the relata of the correspondence relation.

Now, you might have a different, original, idea of what correspondence to the facts amounts to, in which case, great! Or you might think that truthmakers can't do the job that correspondence theorists wanted them to, and neither can anything else that we can identify, and accordingly have less faith in the correspondence theory. Then you're going to be less inclined to necessitation. But if someone wants truthmakers as correspondence relata, and share the vision of the aims of a theory of truth, and what I argued above was roughly right, they're going to be pushed towards looking for necessitating truthmakers.

(They'll elucidate/reduce as follows:

Facts = Truthmakers
Correspondence = Necessitation

So "truth is standing a correspondence relation to the facts" = "truth is standing in the necessitation relation to a truthmaker". Then they'll complain that the position that you sketch in your last comment doesn't offer a elucidation/reduction, and has a bloated ontology+ideology, without obvious explanatory payoff.)

That was the answer I was trying to give to the first two non-rhetorical questions of your original post.

The best concessive move for you at this point wrt to the third question seems to be the Lewis move - claim that truthmaking isn't really about truth, and that the most eligible realizer for the 'correspondence' relation doesn't always involve necessitation by difference-makers.

Carrie Jenkins said...

There's a third option between saying truthmakers aren't really about truth and saying that they supply a full account of truth.

You say:
"So we think that our account of what truthmakers are will (partly) answer the question of what truth consists in because we are correspondence theorists and think that truthmakers are the relata of the correspondence relation."

The parenthetical "partly" is key. Someone who held the view I'm toying with could say that truthmakers will have *some* role to play in explaining what truth consists in, but they can't do it all by themselves.

Andrew said...

I guess everybody is going to agree that truthmakers only have a partial role in explaining what the truth of a truthbearer consists in, if only because they'll accept that we'll have to hear a lot more about e.g. what truthbearers and necessitation are. But even if we restrict ourselves to the 'worldly' aspect of truth, there's still an issue about the relationship between what makes P true in the actual world, and the currently unspecified added ontology/ideology that completes the story of what truth consists in.

After all, if we have a general account of what it is for a truthmaker to be true - what a possible world has to be like in order for some proposition to be true there (and it looked like you were granting, at least pro tem, that this was an intelligible and desirable goal) - doesn't it seem as if this account, together with a sense of what's around in the actual world, and which propositions are true here, should enable us just to deduce what the 'local' grounds are - the truthmakers, in your sense? So it might seem that rather than playing a fundamental albeit partial role in the explanation of what truth is, such local truthmakers will be derivative elements of the theory, provided via theorems derived from the more general account, conjointly with other scientific and metaphysical theories of the nature of the actual world?

I think that perhaps a better idea for you would be to follow Barry Smith and Jonathan Simon in (1) denying that truthmaker theorists have any business trying to give a real or nominal definition of truth and (2) claiming that truthmaker explanations (or the metaphysical correlates of truthmaker explanations, if you want to reserve the term 'explanation' for the acts of explaining, rather than the worldly relations picked out by successful such acts) admit of degrees, so that some explanation would still have been given by the citing of local truthmakers like Mu and Marks, even if there was still more to be said.

Carrie Jenkins said...

Andrew,
You write:
"if we have a general account of what it is for a truthmaker [truthbearer?] to be true ... doesn't it seem as if this account, together with a sense of what's around in the actual world, and which propositions are true here, should enable us just to deduce what the 'local' grounds are - the truthmakers, in your sense?"

Presumably (especially since some of the true propositions will be true propositions about what the truthmakers are ...). But if (as I was imagining) the general account was a truthmaking-plus-X account, that won't mean that "truthmakers will be derivative elements of the theory".

Andrew said...

So the X takes care of what every other possible world has to be like for (C) to be true, and the local truthmakers take care of what this world has to be like for (C) to be true? I find it prima facie very unlikely that a suitable substantive X is going to be found which meets this constraint. I think the best you can do is just to get a big disjunctive X:


If this is world 1 (insert description of how the world is intrinsically) then Mu and Marks make (C) true &
If this is world 2 then Mu, Marks and Mike make (C) true &
If this is world 3...etc

This together with

This is world 1

will enable us to derive

Mu and Marks make (C) true.

This is what I meant by saying that facts about the local truthmakers were derivative. But I can also see how you might think that this theory puts local truthmakers first (after all, it's just a big list of local truthmakers).

Problem 1: It isn't clear that this is the right kind of shape for a story of what being true consists in. (Compare a constitutive account of redness that just constructs a set of all the actual and possible red objects, then e.g. pairs it with a RED-representation.)

Problem 2: It's not enough just to give the intrinsic descriptions, since e.g. w2 may have a part which is an intrinsic duplicate of w1. We need something like a '...and that's all' clause. But if we can already make metaphysical sense of "and that's all" clauses, what's the worry about necessitating truthmakers for general statements?

Carrie Jenkins said...

"So the X takes care of what every other possible world has to be like for (C) to be true, and the local truthmakers take care of what this world has to be like for (C) to be true?"

I wasn't quite thinking of it like that. Let p be a proposition and let 't' range over the facts (or whatever) which are p's truthmakers at various worlds. I was thinking that, at each world, what p's truth-at-that-world-w consists in is the existence of some t at w plus the obtaining at w of other facts X which make it the case that t a truthmaker for p at w.

But now I'm wondering why the following simpler thing wouldn't be a way to combine an *entirely* truthmakerish, correspondence account of truth with the denial of necessitarianism:

What it is for a proposition p to be true at a world is for p to have a truthmaker at that world.

Andrew said...

So your explanation is of the form:

(W)What is is for p to be true at w is for p to have an M at w.

Now, when does p have a M at w? Presumably, on your suggested account, when it stands in the local grounding or 'in virtue of relation' to M. So

(W*) What it is for p to be true at w is for p to stand in the 'in virtue of' relation to an M at w.

Now, when does (C) stand in the IVO relation to Mu and Marks? What does the world have to be like for that to happen? It's not enough for us to say: the world has to have certain intrinsic properties i1, i2, i3, etc, since a part of another world could have all those properties and yet (C) not stand in the IVO relation to Mu and Marks there. What has to be the case is for the world to have the intrinsic properties in question and no other relevant properties. That is, it has to be the case not only that Mu and Marks purr, but that they are all Carrie's cats - the world has to be such that all Carrie's cats purr. That's what it is for the IVO to hold between the purrers Mu and Marks and (C). So we have:

(1) What it is for (C) to be true at the actual world is for (C) to stand in the IVO relation to the purrers Mu and Marks at the actual world. (from W*, on the mooted hypothesis that Mu and Marks' respective purring the local truthmaker for (C))

(2) What it it for (C) to stand in the IVO relation to the the purrers Mu and Marks at the actual world is for the following three facts to obtain:

(a) (C) is the proposition that all Carrie's cats purr
(b) The actual world is intrinsically such that Mu and Marks purr
(c) Mu and Marks are all Carrie's cats at the actual world

Now make the plausible assumption that 'what it is for' claims are 'transitive' in the following sense:

WIIF(P) = Q and WIIF(Q)=R =>
WWIF(P)=R

then we have

(3) What it is for (C) to be true at the actual world is for the following three facts to obtain

(a) (C) is the proposition that all Carrie's cats purr
(b) The actual world is intrinsically such that Mu and Marks purr
(c) Mu and Marks are all Carrie's cats at the actual world

But this just appeals to the kind of fine-grained, general fact that you originally objected to. If you have (c), why not just have a necessitating 'All Carrie's cats purr' fact and be done with it?

Carrie Jenkins said...

I was entertaining (not advocating, you understand, although I do find it appealing) the following proposal:

WIIF p to be true at w is for p to have an M at w (which is close enough to your W*).

This isn't meant to imply that for a particular world w1 where p is true IVO M1, WIIF p to be true at w1 is for p to stand in the IVO relation to M1 at w1. We shouldn't lose the quantified nature of the WIIF claim.

Similarly, WIIF a person P to be a parent is for him to have a child. It would be wrong to say that, in cases where P1's child is C1, WIIF P1 to have a child is for P1 to have C1 as his child.

But we could still get a version of your point, along the lines that I'd be saying:

What it is for (C) to be true at the actual world is for the following three facts to obtain:

(a) (C) is the proposition that all Carrie's cats purr
(b) The actual world is such that some fact F exists at @ and
(c) The actual world is such that F is a truthmaker at @ for all Carrie's cats purr.

We have (b) as the quantified version of the requirement that Mu and Marks's purr, and (c) corresponding to the requirement that Mu and Marks are all Carrie's cats.

We could then say, as you do:
"But this just appeals to the kind of fine-grained, general fact that you originally objected to. If you have (c), why not just have a necessitating 'All Carrie's cats purr' fact and be done with it?"

The way I'd be tempted to respond if I was defending this view would be by saying that the (c) clause isn't to do with ontological grounding of truth (what we're calling truthmaking).

What the (b) clause and the (c) clause are supposed to make perspicuous is that a proposition p's truth is a matter of certain ontological grounds for its truth existing *and* other facts obtaining which make these things *be* ontological grounds for p's truth. We can't theorize some class of useful, fine-grained, necessitating facts into existence which will single-handedly play the role of both clauses, if we don't believe in those sorts of facts.

Andrew said...

I think I think that this looks sound, granting the idea of local truthmakers, and (1):

(1) WIIF p to be true at w is for p to have an M at w
(2) WIIF p to be true at @ is for p to have an M at @ (from 1, UI)
(3) WIIF p to have an M at @ is for p to stand in the IVO relation to M1 at @ (assumption)
So by 'WIIF transitivity' as above
(4) WIIF p to be true at @ is for p to stand in the IVO relation to M1 at @

But as you say, I don't really need that to make the point (would you reject (3)? To my ears that is to treat WIIF claims more like conceptual claims than metaphysical ones - in terms of reality, what it is for me to have my unique dog at @ is for me to stand in the ownership relation to Fido, I think. That's obviously different from what it is for me to have a unique dog simpliciter, which isn't related to Fido in the same way).

I'm not sure I understand the latter part of your reply. it looked like

What it is for (C) to be true at the actual world is for the following three facts to obtain:

(a) (C) is the proposition that all Carrie's cats purr
(b) The actual world is such that some fact F exists at @ and
(c) The actual world is such that F is a truthmaker at @ for all Carrie's cats purr.

was fairly explicit - what it is for (C) to be true was for three entities to be located in the world - namely, the three obtaining facts, some of which were gneeral facts. It didn't look like the claim could be read as having (b) and (c) act as proxies for other, unidentified facts. Is the idea that e.g. WIIF (b) and (c) to obtain is for these other, reducing facts to obtain? But won't you still seemingly need some general claim, (or at least to show that no generality is appealed to)? E.g. isn't that in danger of just repeating the pattern exemplified in moving from

(C) All Carrie's cats purr

to

(b*) The following fact G obtains: at @, some fact F exists such that TM(F,(C),@)

In particular

(a) In the case of (b) and (c), what right have we to be confident that F is not a general fact?

(b) If we get lucky and F is a particular claim, what right have we to be confident that the other facts sufficient to make it into a truthmaker for (C) aren't explicitly or implicitly generality-involving?

I didn't mean to suggest that we could theorize anything into existence. We can see what our theories commit us to, however. At the moment, it looks like asserting the WIIF claim involving (a)-(c) commits one to the obtaining of general facts. If one is already so committed, what harm can there be in appealing to them elsewhere? They may as well earn their bread...

Carrie Jenkins said...

I would reject 3. It seems wrong to me, for the same reason it seems wrong to say that WIIF P1 to be a parent at @ is for him to have C1 as a child at @, or that WIIF me to have a dog at @ is to stand in some relation to Fido at @. We could have a whole separate discussion about your appeal to the conceptual/metaphysical distinction (things are complicated here by my view that the concepts we're justified in using are the ones that mirror metaphysically real aspects of the world). But why would it be such a bad thing to treat this WIIF claim as a conceptual rather than metaphysical matter, assuming that is what I'm doing? What this WIIF claim is trying to supply is a theory of truth - in the tradition of conceptual analyses.

I guess the answers to your questions (a) and (b) would be that if one doesn't believe in general facts one will be confident that they will appear neither as truthmakers nor as the facts that turn things into truthmakers. I can see some prima facie motivation for worrying that my clause (c) appeals to general facts. But couldn't it be that what it is about @ that makes F into a truthmaker for C is the *non*-existence at @ of any other facts of the form X is Carrie's cat, rather than the existence of a fact of the form Nothing else is Carrie's cat?

Andrew said...

The conceptual/metaphysical issue is complicated when it comes to WIIF claims. My feeling is that as metaphysicians we are primarily concerned with what makes the conceptual claim:

(CON) In the conceptual sense, WIIF p to be true at @ is for there to be something in @ that makes p true at @

true, rather than in putting forward the claim. In the same way, we may be interested in WIIF one thing to be identical with another (e.g. ask: is strict identity a genuine relation or a pseudo-relation? What kind of entity is a pseudo-relation?) even though we can define identity pretty easily in the conceptual sense. I think this probably relates to the stuff about explanation earlier, but doesn't hang on it. Basically, we're pursuing real definitions, not nominal ones. We want to know what p's being true IS. But I don't have a lot of arguments for this.

I don't agree with your methodology vis a vis general facts. Compare:

"I can be confident that my adherence to physical theory won't implicitly commit me to the existence of numbers. After all, I don't believe that numbers exist!"

The point is that it's not clear that you can coherently make the claims you want to make (consider, etc) and yet not commit yourself to things you don't want to be committed to. If you make your move, the question just becomes: what reason, independent of our thinking that (C) is true and that general facts don't exist, do we have to believe that a non-general fact G obtains? If it's

(C) is true
There are no general facts

that are one's only support for the claim that G obtains, one can hardly then appeal to the obtaining of G to explain the absence of those problematic general facts. Independent grounds would be needed.

How do lacks or gaps stand in making relations? Again, we may be able to explain what it is for something to be a truthmaker in conceptual terms by citing gaps. But that wasn't really the challenge. The challenge is to say: what is the making relation, what are its relata, and what makes that relation hold between those relata? I don't see how the absence of something can be the ontological relata of the type of relation we're interested in.

(Of course the fact that there are no other facts of the right form could be a potential relata. But that's a general fact.)

Andrew said...

Just occurred to me that maybe Crispin's stuff about warrant transmission would apply or be relevant to the general fact case. If

I have hands
If I have hands, then the real world exists
So, the real world exists

fails to transmit warrant because the conclusion is presupposed by the first premise, the same would be true of

G exists
If G exists, then general facts are not required to explain general truths
So, general facts are not required to explain general truths

I.e. it would fail to transmit warrant for the same reason (even though e.g. all the premises might be true, and the argument valid).

Does Crispin discuss this kind of case?