Sunday, November 13, 2011


I've been following the debate between Maverick Philosopher and our friend Ocham on whether there are "truthmakers" with great interest. The most recent entries are here and here, respectively. The most interesting thing for me is largely the fact that Ocham, who by all indications is a filthy nominalist and so ought to be anathema to us Scotists, seems to make more sense than Vallicella, with whom as a realist-leaning semi-Platonist I ought to agree more. Not because Vallicella comes across as wrong, exactly, so much as that his approach seems to highlight everything I dislike about contemporary anglophone analytic philosophy. I just can't figure out for the life of me what is the use of all the talk about facts and truthmakers. (By "use" I mean "helpfulness in making sense of the world".) It seems riddled with ambiguity and equivocation and a preoccupation with the way we talk about things at the expense of an inattention to the way things are. I've been wondering whether and how to put in my two cents for a while now, but luckily Vallicella's latest post provides a very helpful summary of the status quaestionis, and I will use it as a platform for my own comments.

Let us confine ourselves to true affirmative contingent nonrelational predications. If you deny that there is any extralinguistic fact or state of affairs that makes it true that Tom is smoking [another oft-used example is Vallicella's is "Al is fat"], then what is your positive theory? Here are some possible views, 'possible' in the sense that they are possibly such as to be held by someone whether fool or sage or someone in between.

So Vallicella will lay out what he sees as the possible alternatives.

1. A contingently true sentence like 'Tom is smoking' is just true; there is nothing external to the sentence, nothing at all, that plays any role in making it true. There is no more to a true sentence than the sentence. Thus no part of the sentence has a worldly correlate, not even the subject term. On this view there is no extralinguistic reality -- or at least no extralinguistic reality that bears upon the truth or falsity of our sentences -- and thus no ontological ground of any kind for the truth of true contingent representations, whether declarative sentences, propositions, judgments, beliefs, whatever the truth-bearers are taken to be.

This alternative is plainly unacceptable. On my view a linguistic expression, such as an sentence, is a (complex) sign. A sign is defined as something that brings to mind something else. A sentence can either be true or false if it asserts something, truth being understood as a sign-relation such that the sign accurately signifies the signified, and falsity as a sign-relation such that the sign does not accurately signify the signified. On this account, then, a sign "all by itself" is neither true nor false. A sign signifies something; if it signifies truly it signifies its significate as it is.

2. A rather less crazy view is that our sample sentence does have something corresponding to it in reality, and that that item is Tom, but nothing else. On this view 'Tom is smoking' has a truth-maker, but the truth-maker is just Tom. On this view the truth-maker role is a legitimate one, and something plays it, but there are no facts, and so no fact is a truth-maker. Note carefully that the question whether there are facts is not the same as the question whether there are truth-makers. It could be that the truth-making riole is played by non-facts, and it itr could be that there are facts but they have no role to play in truth-making.

This must be wrong. "Tom", supposing that "Tom" signifies such-and-such an existing man, does not signify "Tom is smoking", but just "Tom", i.e. "this man". It seems to me that before we can figure out the relationship between the complex sign "Tom is smoking" and the truth it signifies (assuming Tom is in fact smoking) we should understand the relationship between the simple sign "Tom" (the name, vocal expression, written characters, thought) and the existing man Tom. But no one in the debate has tried to do that, as far as I've seen.

Note that in my view Tom cannot be the "truthmaker" for "Tom is smoking" or for any other assertion, because Tom is not signified by "Tom is smoking" in my thought or speech or other true expression, but by the name "Tom", and the name is neither true nor false, being non-assertoric. "Hamlet" (the name of a non-existing fictional character) signifies just as well as "Tom", and is also neither true nor false.

3. On a variant of (2) it is admitted that besides Tom there is also an entity corresponding to the predicate, and the truth-maker of 'Tom is smoking' is the set or the mereological sum, or the ordered pair consting of Tom and the entity corresponding to the predicate.

This is where Vallicella starts to lose me. But the reasons are hard to explain given the way he sets out the alternatives, which is partially why he loses me. Let me present the fourth alternative as well before saying what my problem is.

4. A more radical view is that the truth-maker role is not a legitimate role, hence does not need filling by the members of any category of entity. On this view there are no truth-makers becsuae the very notion of a truth-maker is incoherent. One who takes this line could even admit that there are facts, but he would deny that they play a truth-making role.

The presumption seems to be that "truthmaker" will be a sort of being, a categorial entity, and determining what a truthmaker is means determining what category it falls under. But this presumption is indeed incoherent. A truthmaker could only be a category of entity if every entity about which there were truths fell under the same category; but they don't. I would make the exact same objection to the notion of "facts" as a kind of entity. I admit that there "are" facts and truths, in the sense that many thoughts and sentences are true, and that they are true insofar as they correspond to the facts; but I deny that facts and truthmakers are a kind of category of being.

Let's look at the examples. What makes "Tom is smoking" true? The "fact" that Tom is smoking. But what is this fact? Is it a being? I'm not at all sure that this question is going to get us anywhere. Instead I would ask, what is Tom? and, what is smoking? Now Tom is a man, which means he is a kind of substance. And smoking is an action. So to say that Tom is smoking is to say that a certain substance is performing a certain action. Which is to say that a certain accident belongs to a certain substance. So what are the entities involved in the truth that Tom is smoking? Only, Tom's substance and one of his accidents. The "fact" that Tom is smoking is not something other than these. But note that the "truthmaker" is not the substance Tom nor the accident the action of smoking, but the inherence of the accident in the substance, the "ens per accidens" that is Tom as smoking.

So I'm saying that a "truthmaker" is an accident? No, because here what makes "Tom is smoking" or "Al is fat" true is not the accident "smoking" or "fat" but the inherence of the accidents in their subjects. So I'm saying that a "truthmaker" is a mereological sum? No! First, I deny that a substance and its accidents can be understood as parts of a whole. What makes "Tom is smoking" true is not {Tom}+{smoking}, but the fact that Tom is smoking, the statement of fact being not an assertion of summation but an assertion of inherence. Since Tom is a substance and smoking is an accident, they belong to different categories, and I deny that entities of different categories can be added together as parts of a whole, except on an equivocal understanding of "whole".

So a "truthmaker" is a fact constituted by the inherence of an accident in its subject, like fatness in Al or smoking in Tom? No! Because a truthmaker is defined as the entity that makes its truthbearer true. Now simply stated like this there must be truthmakers, if their denial means that true statements are just true for no reason. That would be crazy, as Vallicella rightly notes. But we're asking what kind of entity a truthmaker is, and it seems clear to me that it can be no kind of being. Because if what makes "Tom is smoking" true is the inherence of an accident in its subject, then it seems that truthmakers are accidental inherences or facts about accidental inherences. But this can't be right, because there are many kinds of truths other than truths about accidental inherences, for instance, the truth that Tom is a man. What is the "truthmaker" for "Tom is a man"? It can't be the inherence of any accident in Tom, because humanity is essential, not accidental, to Tom. "Humanity" is not something that inheres accidentally in Tom but is part of what it is to be Tom. And the statement of fact that Tom is a man is not an assertion of inherence but an assertion that Tom has a certain nature. But essences and accidents are not the same kinds of being, i.e. an action and an essence belong to different categories. So if when I call Tom a man and when I say that Tom is smoking I say that these are two facts about Tom, I cannot be saying that there is one kind of entity, facthood, to which "Tom's-being-a-man" and "Tom's-smoking" belong. Rather there are (here) two kinds of entity, essences and accidents, and Tom has a certain essence and Tom has a certain accident.

In other contexts, however, we might want to state truths about entities that are non-categorial entirely, e.g. God. God exists in no genus, and when I say "God exists" or "God is wise" I'm not predicating a species of a genus or an accident of a subject. Nevertheless I want to say that my assertions are (at least possibly) true. The truthmaker then for such an assertion can't be a substance or an accident or an inherence or anything of the sort.

So let's say that facts are the ways things are, and that truthmakers are what make propositions etc true. If so, then facts and truthmakers cannot belong to a certain category. Because among the things that are, are things belonging to many different categories; and not everything that is belongs to any category. So, to ask "what kind of an entity is a truthmaker?" is in a certain sense a category mistake. A truth maker is not a kind of entity, but an entity of whatever kind. Is this what Vallicella's fifth alternative suggests?:

5. On a still more radical view, there is an extralinguistic reality, but we cannot say what categories of entity it contains.

This is not what I'm saying, since I'm saying that there is an extralinguistic reality, and it contains many categories of things, and there is also some being which is in no category.

On this view one abandons the notion that language mirrors reality, that there is any correspondence or matching between parts of speech and categories of entity. Thus one would abandon the notion that truth is correspondence, that the 'Al is fat' is true just in case the referent of 'Al' exemplifies the property denoted by 'fat.' One would be abandoning the notion that language is any guide at all to ontology.

I do deny that language mirrors reality, in the sense that reality has to have (as Vallicella calls it in several places) a "proposition-like structure" in order for propositions about reality to be true. "God is wise", "Tom is a man", "Tom is smoking", "Al is fat", all have the same linguistic structure, but the "facts" to which they refer, that is, the beings about which they are true, have very different entitative structures.

It occurs to me to ask why we should think that, just because our speech can express what is, language ought to be "any guide at all to ontology". I suppose this is me doubting the usefulness of the whole "linguistic turn". Why should an examination of how we say things tell us anything special about the way things are? Shouldn't we examine the way things are and then ask if the way we talk about them is true or false? However, I certainly agree that Al's being fat is the reason it is true to say "Al if fat", and that Al's and Tom's being human is the reason it is true to say "Al and Tom are men".

So perhaps it seems that I'm suggesting that there's an alternative Vallicella left out. Between the notion that a truthmaker must be some category of entity and the notion that no category of entity can be a truthmaker, perhaps we should say that any categorial or supracategorial being can serve as a truthmaker? In this case, however, why should I ever talk about "truthmakers" or "facts"? Why shouldn't I just talk about being? It's totally clear and obvious that there are beings, and it also seems obvious to me that metaphysics ought to study what being is like: what are the categories, what falls under them, what belongs to no category, etc. But if "truthmakers" and "facts" are just beings, or rather, ways of talking about being in relation to our assertions, and in such a way as to render it enormously confusing whether they are supposed to be some special sort of being or not, what is the "value added" of dragging these terms into metaphysics?

Ocham seems to agree that if it is true to say "Tom is smoking" or "Al is fat", this is because in reality Tom is smoking, or Al is fat. That is to say, our speech is true because of the way being is. But the beings in question are Tom and his activities and Al and his qualities and quantities. Do we need to drag in beings other than this? Or are the "truthmakers" of our sentences and the "facts" supporting them just these beings, but considered under a certain aspect, as related to our speech? If this is Ocham's confusion, then I share it, along with his dubious attitude towards the need to talk about truthmakers and facts as distinct from the beings our assertions are about at all.


awatkins909 said...

Wonderful post! Your argument is convincing. I find myself agreeing more and more these days with the idea which you state, that we shouldn't assume the way we talk about things implies anything about the way things are. This leads to many errors in analytic philosophy. Off the top of my head, the immensely popular Quinean view of ontology might be one of them. Since analytic philosophy makes little distinction between categories of being, with some more fundamental than others, just by saying "there are ____" you've likely generated a new category of things. I'd suggest at least some of it has to do with its Kantian heritage (the name 'analytic philosophy' itself bears the name of the great thinker).

Michael Sullivan said...

Thanks, awatkins.

By the way, Ocham (if you read this), the "filthy" in "filthy nominalist" is meant in fun, as pertaining to school rivalry - no slight intended!

E.R. Bourne said...

This was truly an excellent post. It is most gratifying to see a discussion of a very unnecessarily confusing contemporary topic from a scholastic perspective. More often than not, today's philosophy takes on an opaqueness that is the direct result of a refusal to begin at the beginning. Without a thorough elucidation of first principles all we are left with is a muddleheaded attempt to turn intuitions and modern biases into real theories of reality and language. My suspicion is that much more contemporary philosophy can use a dose of this.