Category: Facts

Naturalism, Ultimate Explanation, and Brute Facts and Laws

Malcolm Pollack solicited my comments on an article by Tomas Bogardus that appeared in Religious Studies under the title, If naturalism is true, then scientific explanation
is impossible.

Malcolm summarizes:

I’ve just read a brief and remarkably persuasive philosophical paper by Tomas Bogardus, a professor of philosophy at Pepperdine University. In it, he argues that, if we are to have confidence in the explanatory power of science (and he believes we should), then the naturalistic worldview must be false.

Here is the abstract:

I begin by retracing an argument from Aristotle for final causes in science. Then, I advance this ancient thought, and defend an argument for a stronger conclusion: that no scientific explanation can succeed, if Naturalism is true. The argument goes like this: (1) Any scientific explanation can be successful only if it crucially involves a natural regularity. Next, I argue that (2) any explanation can be successful only if it crucially involves no element that calls out for explanation but lacks one. From (1) and (2) it follows that (3) a scientific explanation can be successful only if it crucially involves a natural regularity, and this regularity does not call out for explanation while lacking
one. I then argue that (4) if Naturalism is true, then every natural regularity calls out for explanation but lacks one. From (3) and (4) it follows that (5) if Naturalism is true, then no scientific explanation can be successful. If you believe that scientific explanation can be (indeed, often has been) successful, as I do, then this is a reason to reject Naturalism.

Keywords: philosophy of religion; philosophy of science; scientific explanation; naturalism; supernaturalism; theism; atheism

The gist of the argument is that science, which is in the business of explaining observable phenomena, must offer for every explanandum (i.e., that which is to be explained) some explanans (that which explains). But if the explanans itself requires explaining, the explanation is incomplete, and must rest upon some deeper explanans.

Bogardus’s paper explores the varieties of possible explanatory regression. Either a) we bottom out on a “brute fact”, or b) we encounter an infinite stack of explanations (“turtles all the way down”), or c) our explanations loop back on themselves (so that at some point every explanandum also becomes an explanans), or d) we come at last to some explanans that breaks the chain, by requiring no further explanation.

Bogardus argues that of all brute facts, infinite regressions, and circular explanations explain nothing; the only kind of thing that will serve is (d). But the “laws of nature” do not meet this requirement, because they do not (and cannot) explain themselves.

The heart of Bogardus’s argument, then, is that only some sort of necessary truth, some teleological principle that stands outside of the chain of scientific explanation, can serve as the anchor to which that chain must be fastened. And because Naturalism admits of no such entity, then if scientific explanations are to be considered valid, Naturalism must be false.

My Evaluation

It is given that nature is regular. She exhibits all sorts of regularities. Some of them are codified in scientific law statements. Coulomb's Law, for example, states that particles of like charge repel and particles of unlike charge attract. Another regularity we are all familiar with is that if a gas is heated it expands. This is why I do not store my can of WD-40 in the garage in the Arizona summer. The regularity is codified is Gay Lussac's law: the pressure of a given amount of gas held at constant volume is directly proportional to the Kelvin temperature.  Now why should that be the case? What explains the law? The kinetic theory of gases. If you heat a gas you give the molecules more energy so they move faster. This means more impacts on the walls of the container and an increase in the pressure. Conversely if you cool the molecules down they will slow down and the pressure will be decreased. The temperature of the gas varies with the kinetic energy of the gas molecules. 

But invoking the kinetic theory of gases is not an ultimate explanation.  What about those molecules and the laws that govern them?

So here is a question for Malcolm: Is Bogardus assuming that a genuine explanation must be or involve an ultimate explanation? And if he is making that assumption, is the assumption true?

Here is another example. Farmer John's crops have failed. Why? Because of the drought. The drought in turn is explained in terms of atmospheric conditions, which have their explanations, and so on. Question is: have I not explained the crop failure by just saying that that drought caused it?

Must I explain everything to explain anything? Is no proximate explanation a genuine explanation?

But we are philosophers in quest of the ultimate. That's just the kind of people we are. So we want ultimate explanations. And let us suppose, with Bogardus, that such explanations cannot be non-terminating, that is, they cannot be infinitely regressive or 'loopy,' i.e. coherentist.  Ultimate explanations must end somewhere.  Bogardus:

. . . I believe many Naturalists subscribe to scientific explanation in the pattern of Brute Foundationalism, either of the Simple or Extended variety, depending on the regularity. Here’s Carroll’s (2012, 193) impression of the state of the field: ‘Granted, it is always nice to be able to provide reasons why something is the case. Most scientists, however, suspect that the search for ultimate explanations eventually terminates in some final theory of the world, along with the phrase “and that’s just how it is”.’31

My question to Malcolm (and anyone): Why can't scientific explanations end with brute laws and brute facts? Has Bogardus given us an argument against brute laws? I don't see that he has. Or did I miss the argument for (2) below in Bogardus's main argument:

 

1) Any scientific explanation can be successful only if it crucially involves a natural
regularity.
2) Any explanation can be successful only if it crucially involves no element that calls
out for explanation but lacks one.
3) So, a scientific explanation can be successful only if it crucially involves a natural
regularity, and this regularity does not call out for explanation while lacking one.

4) If Naturalism is true, then every natural regularity calls out for explanation but
lacks one.
5) So, if Naturalism is true, then no scientific explanation can be successful.

Bogardus tries to argue for (2), but I don't see that he succeeds in giving us a non-question-begging reason to accept (2).

I myself reject naturalism and brute facts. My point is that Bogardus has failed to refute it and them. He has merely opposed it and them.  As I use 'refute,' it is a verb of success.  To oppose me is not to refute me. I will oppose you right back.

There is another question that I will address in a separate post:  Can it be demonstrated that there is a Necessary Explainer? Pace the presuppositionalists, the demonstration cannot be circular. A circular demonstration is no demonstration at all.  You cannot prove a proposition by presupposing it. You are of course free to presuppose anything you like. You can even presuppose naturalism and then 'argue': it is true because it is true, and then try to account for everything is naturalistic terms.

Semirealism about Facts: An Exchange with Butchvarov

Jack WebbFacts are the logical objects corresponding to whole declarative sentences, or rather to some of them. When it comes to facts, Panayot Butchvarov appreciates the strengths and weaknesses of both realism and anti-realism. For the realist, there are facts. For the anti-realist, there are no facts. Let us briefly review why both positions are attractive yet problematic. We will then turn to semirealism as to a via media between Scylla and Charybdis.

Take some such contingently true affirmative singular sentence as 'Al is fat.' Surely with respect to such sentences there is more to truth than the sentences that are true. There must be something external to the sentence that contributes to its being true, and this external something is not plausibly taken to be another sentence or the say-so of some person, or anything like that. 'Al is fat' is true because there is something in extralinguistic and extramental reality that 'makes' it true. There is this short man, Al, and the guy weighs 250 lbs. There is nothing linguistic or mental about the man or his weight. Here is the sound core of correspondence theories of truth. Our sample sentence is not just true; it is true because of the way the world outside the mind and outside the sentence is configured. The 'because' is not a causal 'because.' The question is not the empirical-causal one as to why Al is fat. He is fat because he eats too much. The question concerns the ontological ground of the truth of the sentential representation, 'Al is fat.' Since it is obvious that the sentence cannot just be true — given that it is not true in virtue either of its logical form or ex vi terminorum – we must posit something external to the sentence that 'makes' it true. I myself, a realist, don't see how this can be avoided even though I admit that 'makes true' is not perfectly clear.

Now what is the nature of this external truth-maker? It can't be Al by himself, and it can't be fatness by itself. Nor can it be the pair of the two. For it could be that Al exists and fatness exists, but the first does not instantiate the second. What's needed, apparently, is the fact of Al's being fat. So it seems we must add the category of fact to our ontology, to our categorial inventory. Veritas sequitur esse is not enough. It is not enough that 'Al' and 'fat' have worldly referents; the sentence as a whole needs a worldly referent. Truth-makers cannot be 'things' or collections of same, but must be entities of a different categorial sort. (Or at least this is so for the simple predications we are now considering.)

The argument I have just sketched, the truth-maker argument for facts, is very powerful, but it gives rises to puzzles and protests. There is the Strawsonian protest that facts are merely hypostatized sentences, shadows genuine sentences cast upon the world. Butchvarov quotes P. F. Strawson's seminal 1950 discussion: “If you prise the sentences off the world, you prise the facts off it too. . . .” (Anthropocentrism in Philosophy, 174)  Strawson again: “The only plausible candidate for what (in the world) makes a sentence true is the fact it states; but the fact it states is not something in the world.” (174)

Why aren't facts in the world? Consider the putative fact of my table's being two inches from the wall. Obviously, this fact is not itself two inches from the wall or in any spatial position. The table and the wall are in space; the fact is not. One can drive a nail into the table or into the wall, but not into the fact, etc. Considerations such as these suggest to the anti-realist that facts are not in the world and that they are but sentences reified. After all, to distinguish a fact from a non-fact (whether a particular or a universal) we must have recourse to a sentence in the indicative mood: a fact is introduced as the worldly correlate of a true sentence. If there is no access to facts except via sentences, as the correlates of true sentences, then this will suggest to those of an anti-realist bent that facts are hypostatizations of true declarative sentences.

One might also cite the unperceivability of facts as a reason to deny their existence. I see the table, and I see the wall. It may also be granted that I see that the desk is about two inches from the wall. But does it follow that I see a relational fact? Not obviously. If I see a relational fact, then presumably I see the relation two inches from. But I don't see this relation. And so, Butchvarov argues (175) that one does not see the relational fact either. The invisibility of relations and facts is a strike against them. Another of the puzzles about facts concerns how a fact is related to its constituents. Obviously a fact is not identical to its constituents. This is because the constituents can exist without the fact existing. Nor can a fact be an entity in addition to its constituents, something over and above them, for the simple reason that it is composed of them. We can put this by saying that no fact is wholly distinct from its constituents. The fact is more than its constituents, but apart from them it is nothing. A third possibility is that a fact is the togetherness of its constituents, where this togetherness is grounded in a a special unifying constituent. Thus the fact of a's being F consists of a, F-ness, and a nexus of exemplification. But this leads to Bradley's regress.

A fact is not something over and above its constituents but their contingent unity. This unity, however, cannot be explained by positing a special unifying constituent, on pain of Bradley's regress. So if a fact has a unifier, that unifier must be external to the fact. But what in the world could that be? Presumably nothing in the world. It would have to be something outside the (phenomenal) world. It would have to be something like Kant's transcendental unity of apperception. I push this notion in an onto-theological direction in my book, A Paradigm Theory of Existence: Onto-Theology Vindicated. But by taking this line, I move away from the realism that the positing of facts was supposed to secure. Facts are supposed to be ontological grounds, extramental and extralinguistic. If mind or Mind is brought in in any form to secure the unity of a truth-making fact, then we end up with some form of idealism, whether pschological or transcendental or onto-theological. 

So we are in an aporetic pickle. We have good reason to be realists and we have good reason to be anti-realists. (The arguments above on both sides were mere sketches; they are stronger than they might appear. ) Since we cannot be both realists and anti-realists, we might try to mediate the positions and achieve a synthesis. My book was one attempt at a synthesis. Butchvarov's semi-realism is another. I am having a hard time, though, understanding how exactly Butchvarov's semi-realism achieves the desired synthesis. Butchvarov:

Semirealism regarding facts differs from realism regarding facts by denying that true sentences stand for special entities, additional to and categorially different from the entities mentioned in the sentences, that can be referred to, described, and analyzed independently of the sentences. [. . .] But semirealism regarding facts also differs from antirealism regarding facts by acknowledging that there is more to truth than the sentences . . . that are true. (180)

In terms of my simple example, semirealism about facts holds that there is no special entity that the sentence 'Al is fat' stands for that is distinct from what 'Al and 'fat' each stand for. In reality, what we have at the very most are Al and fatness, but not Al's being fat. Semirealism about facts also holds, however, that a sentence like 'Al is fat' cannot just be true: if it is true there must be something that 'makes' it true, where this truth-maker cannot be another sentence (proposition, belief, judgment, etc.) or somebody's say-so, or something merely cultural or institutional or otherwise conventional. And let's not forget: the truth-maker cannot be Al by himself or fatness by itself or even the pair of the two. For that pair (ordered pair, set, mereological sum . . .) could exist even if Al is not fat. (Suppose Al exists and fatness exists in virtue of being instantiated by Harry but not by Al.)

How can semirealism avoid the contradiction: There are facts and there are no facts? If the realist says that there are facts, and that anti-realist says that there aren't, the semi-realist maintains that 'There are facts' is an “improper proposition” (178) so that both asserting it and denying it are improper. In explaining the impropriety, Butchvarov relies crucially on Wittgenstein's distinction between formal and material concepts and his related distinction between saying and showing. Obscurum per obscurius? Let's see.

The idea seems to be that while one can show that there are facts by using declarative sentences, one cannot say or state that there are facts by using declarative sentences, or refer to any particular fact by using a declarative sentence. If there are facts, then we should be able to give an example of one. 'This page is white is a fact,' won't do because it is ill-formed. (179) We can of course say, in correct English, 'That this page is white is a fact.' But 'that this page is white' is not a sentence, but a noun phrase. Not being a sentence, it cannot be either true or false. And since it cannot be either true or false, it cannot refer to a proposition-like item that either obtains or does not obtain. So 'that this page is white' does not refer to a fact. We cannot use this noun phrase to refer to the fact because what we end up referring to is an object, not a fact. Though a fact is not a sentence or a proposition, it is proposition-like: it has a structure that mirrors the structure of a proposition. No object, however, is proposition-like. To express the fact we must use the sentence. Using the sentence, we show what cannot be said.

On one reading, Butchvarov's semirealism about facts is the claim that there are facts but they cannot be named. They cannot be named because the only device that could name them would be a sentence and sentences are not names. On this reading, Butchvarov is close to Frege. Frege held that there are concepts, but they cannot be named. Only objects can be named, and concepts are not objects. If you try to name a concept, you will not succeed, for what is characteristic of concepts, and indeed all functions, is that they are unsaturated (ungesaettigt). And so we cannot say either

The concept horse is a concept

or

The concept horse is not a concept.

The first, though it looks like a tautology, is actually false because 'The concept horse' picks out an object. The second, though it looks like a contradiction, is actually true for the same reason. Similarly, we cannot say either

The fact that snow is white is a fact

or

The fact that snow is white is not a fact.

The first, though it looks like a tautology, is actually false because 'The fact that snow is white' picks out an object. The second, though it looks like a contradiction, is actually true for the same reason.

It is the unsaturatedness of Fregean concepts that makes them unnameable, and it is the proposition-like character of facts that makes them unnameable.

Semirealism about facts, then, seems to be the view that there are facts, but that we cannot say that there are: they have a nature which prevents us from referring to them without distorting them. But then the position is realistic, and 'semirealism' is not a good name for it: the 'semirealism' is more epistemological/referential than ontological.

Other things Butchvarov says suggest that he has something else in mind with 'semirealism about facts.' If he agrees with Strawson that facts are hypostatized declarative sentences, and argues against them on the ground of their unperceivability, then he cannot be saying that there are facts but we cannot say that there are. He must be denying that there are facts. But then why isn't he a flat-out antirealist?

Can you help me, Butch? What am I not understanding? What exactly do you mean by 'semirealism about facts'?

BUTCHVAROV RESPONDS:

Your grasp of the issue is excellent, Bill. “[T]he 'semirealism' is more epistemological/referential than ontological” seems to me right; this is why it is semirealism, not realism. But it is logical semirealism: “Logical semirealism differs from both logical antirealism and logical realism much as Kant’s position on causality differed from both antirealism and realism regarding causality, and Wittgenstein’s position on other people’s sensations differed from both antirealism and realism regarding
“other minds” (page 166).

The reason facts are only “semireal” in my view is that they have a logical structure. As you say in your book A Paradigm Theory of Existence: Onto-Theology Vindicated, “facts could be truth-making only if they are “proposition-like,” “structured in a proposition-like way” – only if “a fact has a structure that can mirror the structure of a proposition.” The structure of a proposition is its logical structure. In Part Two of Anthropocentrism in Philosophy I argue against realism regarding logical structure, but I also reject the simplistic antirealism regarding logical structure that says “there is only language.” Surely there are no ands, ors, or iffs in the world. It’s not just that logical objects and structures cannot be perceived or even “said.” Surely words like “and,” “or,” and “if” do not stand for anything physical, mental, or other-worldly. Yet no less surely they are not merely words.

Since facts necessarily, indeed essentially, possess a logical structure, my argument against logical realism applies also to realism regarding facts but, again, I reject the simplistic antirealism regarding facts that says “there is only language.” I wrote: “[T]here is a third way of understanding facts, which is neither realist nor antirealist. It is semirealist. In general, if a proposition is in dispute between realism and antirealism, with the realist asserting and the antirealist denying it, the semirealist would differ from both by holding that it is an improper proposition, perhaps even that there is no such proposition, and thus that both asserting and denying it are improper. There is an analogy here with sophisticated agnosticism. The theist asserts the proposition “God exists” and the atheist denies it, but the sophisticated agnostic questions, for varying reasons we need not consider here, its propriety” (pages 178-9).

I would share your discomfort if a philosopher said “There are facts and there are no facts”( I can’t find the sentence in Anthropocentrism in Philosophy). But I would understand it, just as I understand Frege’s “The concept horse is not a concept,” Meinong’s “there are things of which it is true that there are no such things,” and Wittgenstein’s “some things cannot be said but show themselves.” All four are puzzling. Sometimes we have to content ourselves with truths that puzzle us, make us wonder. But philosophy begins in wonder. We could, of course, invent new terms, perhaps saying that while facts do not exist they subsist, but I doubt that this would lead to better understanding.

Butch

Posted by: Panayot Butchvarov | Thursday, October 29, 2015 at 10:53 AM

Butch,

Thanks for the response. You never say "There are facts and there are no facts." But it seems to me that you give good arguments for both limbs of this (apparent) contradiction. Because the arguments on both sides are impressive, we have a very interesting, and vexing, problem on our hands, especially if you hold, as I think you do, that there are no true contradictions.

I was under the impression that the doctrine of semirealism (about facts) was supposed to eliminate the contradiction and show it to be merely apparent. It seems to me that if we distinguish between existence and subsistence as two different modes of Being, then we could say that while facts do not exist in the way their constituents do, they are not nothing either — they don't exist but subsist. This would seem to be a way between the horns of the dilemma. In your brilliant formulation, "There is more to the truth of sentences than the sentences that are true." On the other hand, there are the Strawsonian and other arguments against facts. On this way of looking at things, semirealism comes down to a doctrine of modes of Being.

What I don't understand however, is how this is supposed to square with Wittgenstein's say vs. show distinction.

On p. 76 of Anthropocentrism you refer to the existence-subsistence distinction. But then on p. 77 you say that this distinction is not the same as W's distinction between what can be said and what can only be shown — though it resembles it in motivation.

So here is my criticism: you are not using 'semirealism' univocally. If a subsistent, a number say, is semireal, then that is clear to me since I myself advocate (against most contemp. anal. phils.) distinctions between modes of Being. But if you say that numbers are semireal in the sense that 'There are numbers' cannot be *said,* that the existence of numbers can only be *shown* by the use of numerals, then that is a quite different use of 'semireal.'

Why? Because one could take the say-show line while holding that there are no modes of Being, and vice versa.

So I have two problems. One is that you seem to equivocate on 'semireal.' The other is that W's say-show distinction is not clear to me. So if you explain semirealism in terms of the latter, then we have a case of *obscurum per obscurius.*

Posted by: BV | Thursday, October 29, 2015 at 03:03 PM

Butch,

Here is another concern of mine.

>>Frege’s “The concept horse is not a concept,” Meinong’s “there are things of which it is true that there are no such things,”<<

You assimilate these to each other. But I see a crucial difference. Meinong employs a paradoxical formulation for literary effect, a formulation that expresses a proposition that is in no way contradictory. All he is saying is that some items are beingless which you will agree is non-contradictory. In other words, the proposition, the thought, that Meinong is expressing by his clever formulation is non-contradictory despite the fact that the verbal formulation he employs is either contradictory (assuming that 'there are' is used univocally) or equivocal.

But what is going on in "The concept *horse* is not a concept" is quite different. What Frege is saying in effect is that we cannot refer to concepts in a way that preserves their predicative function, their unsaturatedness. 'The concept *horse*' is a name, and only objects can be named. So when we try to say anything about a concept we must fail inasmuch as a reference to a concept transforms it into an object thereby destroying its predicative function.

Frege anticipates Wittgenstein in this. I can say '7 is prime' but not 'Primeness is instantiated by 7.'

This is similar to the problem we have with propositions and facts (which have a proposition-like structure).

As you point out, 'Snow is white is true' is ill-formed. But 'That snow is white is true' is false inasmuch as 'That snow is white' is a nominal phrase that picks out an object, and no object can be true.

At his point someone might propose a disquotational-type theory according to which 'true' in 'That snow is white is true' does not express a property of something but merely serves to transform the nominal phrase back into the sentence 'Snow is white.'

What refutes this is your point that "There is more to the truth of sentences than the sentences that are true."

Posted by: BV | Thursday, October 29, 2015 at 03:40 PM

I am answering two posts. As to the one about Anscombe and GeacH, I agree completely, Bill. I’ve always marveled that philosophers like Anscombe and Geach were so easily influenced by Russell’s attacks on Meinong. Russell of course did know what Meinong meant and initially even agreed with him but then invented his theory of definite descriptions that allowed him to “analyze away” Meinong’s examples.

Now I come to the other post. I am not sure there is genuine disagreement between us. Regarding existence and subsistence, we might look at [Gustav]Bergmann. He “renounced his earlier distinction between existence and subsistence, subscribing now to the seeming paradox that ‘whatever is thinkable exists.’ Yet he acknowledged that ‘the differences among some of the several existents…are very great indeed…momentous, or enormous,’ thus acknowledging the rationale for the distinction” (page 142 of Anthropocentrism in Philosophy). Whether we “distinguish between existence and subsistence as two different modes of Being” or say that everything (‘thinkable’) exists though the differences among some existents are enormous seems to me a matter of words. I am uneasy about using the phrase “modes of Being” because it has had numerous other applications, e.g., matter and mind, universals and particulars, infinite and fine, and so on.
As to the meaning of “semireal,” let me begin with a quotation from the Introduction: “[I]n the case of metaphysical antirealism, numerous qualifications, distinctions, and explanations are needed. No metaphysical antirealist denies the reality of everything, just as no metaphysical realist asserts the reality of everything, including, say, the Easter Bunny. The solipsist says, ‘Only I exist,’ not ‘Nothing exists.’ Berkeley denied that there are material objects, he called them ‘stupid material substances,’ but he insisted on the existence of minds and their ideas. According to Kant…. material objects are ‘transcendentally ideal,’ dependent on our cognitive faculties, but they are nonetheless ‘empirically real,’ not mere fancy. Bertrand Russell distinguished between existence
and subsistence: some things do not exist, yet they are not nothing – they subsist; for example, material objects exist but universals only subsist. According to Wittgenstein’s Tractatus, some things cannot be “said,” i. e., represented in language,but they “show” themselves in what can be said. Among them, he held, are those that matter most in logic, ethics, and religion” (page 15).

I try to explain Wittgenstein’s distinction between saying and showing as follows: “The distinction has a straightforward, noncontroversial application even to ordinary pictures, say, paintings and photographs, indeed to representations generally. And the [associated] picture theory [of meaning] is merely a subtler version of the traditional theory of meaning and thought, which was unabashedly representational, ‘pictorial’: thought involves ‘ideas,’ often explicitly understood as mental images or pictures, and the meaning of an expression is what it stands for” (page 67).”

“In a painting, much is shown that is not and cannot be pictured by the painting or any part of it. For example, the painting may represent a tree next to a barn, each represented by a part of the painting, and the spatial relation between the parts of the painting that represent the tree and the barn would represent their relation of being next to each other. But nothing in the painting represents that relation’s being a relation, nothing ‘says’ that their being next to each other is a relation (rather than, say, a shape or color). Yet the painting shows this, indeed must show it in order to represent what it does represent. What it shows cannot be denied as one might deny, for example, that the painting is a portrait of Churchill. The absence from the painting of what it only shows would not be like Churchill’s absence. Of course, paintings do not consist of words, and sentences are only ‘logical’ pictures. But like all pictures, physical or mental, paintings are logical pictures, though not all logical pictures are paintings” (page 68).

You write, “What I don't understand however, is how this [the distinction between real and semireal] is supposed to square with Wittgenstein's say vs. show distinction.” My concern is with logical semirealism, and Wittgenstein applied his distinction mainly to logical expressions. If some things cannot be said but show themselves, neither calling them real nor calling them unreal would be quite right. So I opted for calling them semireal. Of course, nothing of philosophical importance hangs on what word is chosen.

You write, “if you say that numbers are semireal in the sense that 'There are numbers' cannot be *said,* that the existence of numbers can only be *shown* by the use of numerals, then that is a quite different use of 'semireal.'” I have offered no view about numbers, though Wittgenstein did include “number” in his list of formal concepts: “’Object,’ ‘complex,’ ‘fact,’ ‘function,’ ‘number’ signify formal concepts, represented in logical notation by variables, for example, the pseudo-concept object by the variable ‘x’ (Tractatus 4.1272). The properties they appear to stand for are formal, internal, such that it is unthinkable that what they are attributed to should not possess them (4.123). For this reason it would be just as nonsensical to assert that something has a formal property as to deny it (4.124).” But Wittgenstein was aware that the status of numbers is far too complicated an issue to be resolved by just saying that “There are numbers” cannot be “said.” Much later he wrote his Remarks on the Foundations of Mathematics.

Please forgive me for resorting to such lengthy quotations. I tried to avoid them but found what I was writing inferior to what I had already written.

Posted by: Panayot Butchvarov | Sunday, November 01, 2015 at 01:31 AM

On J. P. Moreland’s Theory of Existence

A re-post from February of 2016 with corrections and addenda.

 

Moreland J. P.What follows is largely a summary and restatement of points I make in "The Moreland-Willard-Lotze Thesis on Being," Philosophia Christi, vol. 6, no. 1, 2004, pp. 27-58.  It is a 'popular' or 'bloggity-blog' version of a part of that lengthy technical article.  First I summarize my agreements with J. P. Moreland.   Then I explain and raise two objections to his theory.  The Moreland text I have under my logical microscope is pp. 134-139 of his 2001 Universals (McGill-Queen's University Press). 

 

Common Ground with Moreland on Existence

We agree on the following five points (which is not to say that Moreland will agree with every detail of my explanation of these five points):

1) Existence is attributable to individuals.  The cat that just jumped into my lap exists.  This very cat, Manny, exists.  Existence belongs to it and is meaningfully attributable to it.  Pace Frege and Russell, 'Manny exists' is a meaningful sentence, and it is meaningful as it stands, as predicating existence of an individual.  It is nothing like 'Manny is numerous.' To argue that since cats are numerous, and Manny is a cat, that therefore Manny is numerous is to commit the fallacy of division.  Russell held that the same fallacy is committed by someone who thinks that since cats exist, and Manny is a cat, that therefore Manny exists.  But Russell was mistaken: there is no fallacy of division; there is an equivocation on 'exists.'  It has a general or second-level use and a singular or first-level  use.

There are admissible first-level uses of '. . .exist(s).'  It is not the case that only second-level uses are admissible. And it is only because Manny, or some other individual cat, exists that the concept cat is instantiated.  The existence of an individual cannot be reduced to the being-instantiated of a property or concept.  If you like, you can say that the existence of a concept is its being instantiated.  We sometimes speak like that.  A typical utterance of 'Beauty exists,' say, is not intended to convey that Beauty itself exists, but is intended to convey that Beauty is exemplified, that there are beautiful things.  But then one is speaking of general existence, not of singular existence. 

Clearly, general existence presupposes singular existence in the following sense:  if a first-level concept or property is instantiated, then it is instantiated by an individual, and this individual must exist in order to stand in the instantiation nexus to a concept or property.  From here on out, by 'existence' I mean 'singular existence.'  There is really no need for 'general existence' inasmuch as we can speak of instantiation or of someness, as when we say that cats exist if and only something is a cat. The fundamental error of what Peter van Inwagen calls the 'thin theory' of existence is to imagine that existence can be reduced to the purely logical notion of someness.  That would be to suppose, falsely, that singular existence can be dispensed with in favor of general existence.  Existence is not a merely logical topic ; existence is a metaphysical topic.

2) Existence cannot be an ordinary property of individuals.  While existence is attributable to individuals, it is no ordinary property of them, and if one were to define properties as instantiable entities,  it is no property at all.   There are several reasons for this, but I will mention only one:  you cannot add to a thing's description by saying of it that it exists. Nothing is added to the description of a tomato if one adds 'exists' to its descriptors: 'red,' round,' ripe,' etc.   As Kant famously observed, "Being is not a real predicate," i.e., being or existence adds nothing to the realitas or whatness of a thing. Contrary to popular scholarly opinion, Kant did not anticipate the Frege-Russell theory.  He does not deny that 'exist(s)' is an admissible first-level predicate.  (See my "Existence: Two Dogmas of Analysis" in Novotny and Novak, eds. Neo-Aristotelian Perspectives in Metaphysics, Routledge 2014, pp. 45-75, esp. 48-50.)

3) Existence is not a classificatory concept or property.  The reason is simple: there is no logically prior domain of items classifiable as either existent or nonexistent.  Pace Meinong, everything exists.  There are no nonexistent items.  On Meinong's view, some items actually have properties despite having no Being at all.

4) Existence makes a real difference to a thing that exists.  In one sense existence adds nothing to a thing.  It adds nothing quidditative.  In another sense it adds everything:  if a thing does not exist, it is nothing at all! To be or not to be — not just a question, but the most 'abysmal' difference conceivable.    In this connection, Moreland rightly speaks of a "real difference between existence and non-existence." (137)

5) Existence itself exists.  This is not the trivial claim that existing things exist.  It is the momentous claim that that in virtue of which existing things exist itself exists.  It is a logical consequence of (4) in conjunction with (3). As Moreland puts it, "[i]f existence itself does not exist, then nothing else could exist in virtue of having existence." (135)

The above five points are criteria of adequacy for a theory of existence: any adequate theory must include or entail each of these points.   Most philosophers nowadays will not agree, but  I think Moreland will.  So he and I stand on common ground.  I should think that the only fruitful disputes are those that play out over a large chunk of common ground. 

But these criteria of adequacy also pose a problem:  How can existence belong to individuals without being a property of them?  Existence belongs to individuals as it would not belong to them if it were a property of properties or concepts; but it is not a property of individuals.  How can we uphold both of these insights?

Moreland's Theory

Moreland's theory gets off to a good start:  "existence is not a property which belongs, but is the belonging of a property." (137)  This insight nicely accommodates points (1) and (2) above:  existence is attributable to individuals without being an ordinary property of them.  Indeed, it is not a property at all. I infer from this that existence is not the property of having properties. It is rather the mutual belongingness of a thing and its properties.  Moreland continues:

Existence is the entering into the exemplification nexus . . . . In the case of Tony the tiger, the fact [that] the property of being a tiger belongs to something and that something has this property belonging to it is what confers existence. (137)

I take this to mean that existence is the mutual belonging together of individual and property.   It is 'between' a thing and its properties as that which unifies them, thereby tying them into a concrete fact or state of affairs.  The existence of Tony is not one of his properties; nor is the existence of Tony identical to  Tony.   And of course the existence of Tony is not the being-exemplified of some such haecceity property as identity-with-Tony.  Rather, the existence of Tony, of that very individual, is his exemplifying of his properties. The existence of a (thick) individual in general is then the exemplification relation itself insofar as this relation actually relates (thin) individual and properties.  Existence is not a first-level property, or a second-level property, but a very special relation, the exemplification relation.

Moreland implies as much.  In answer to the question how existence itself exists, he explains that "The belonging-to (exemplification, predication) relation is itself exemplified . . ." (137)  Thus the asymmetrical exemplification relation x exemplifies P is exemplified by Tony and the property of being a tiger (in that order). Existence itself exists because existence itself is the universal exemplification relation which is itself exemplified.  It exists in that it is exemplified by a and F-ness, a and G-ness, a and H-ness, b and F-ness, b and G-ness, b and H-ness, and so on.  An individual existent exists in that its ontological constituents (thin particular and properties) exemplify the dyadic exemplification relation which is existence itself.

The basic idea is this.  The existence of a thick particular such as Tony, that is, a particular taken together with all its monadic properties, is the unity of its ontological constituents.  (This is not just any old kind of unity, of course, but a type of unity that ties items that are not facts into a fact.)  This unity is brought about by the dyadic exemplification relation within the thick particular.   The terms of this relation are the thin particular on the one hand and the properties on the other. 

Moreland's theory accommodates all five of the desiderata listed above which in my book is a strong point in its favor.

A Bradleyan Difficulty

A sentence such as 'Al is fat' is not a list of its constituent words.  The sentence is either true or false, but neither the corresponding list, nor any item on the list, is either true or false.  So there is something more to a declarative sentence than its constituent words.  Something very similar holds for the fact that makes the sentence true, if it is true.  I mean the extralinguistic fact of Al's being fat.  The primary constituents of this fact, Al and fatness, can exist without the fact existing. The fact, therefore, cannot be identified with its primary constituents, taken either singly, or collectively.    A fact is more than its primary constituents.  But how are we to account for this 'more'? 

On Moreland's theory, as I understand it, this problem is solved by adding a secondary constituent, the exemplification relation, call it EX, whose task is to connect the primary constituents.  This relation ties the primary constituents into a fact.  It is what makes a fact more than its primary constituents.  Unfortunately, this proposal leads to Bradley's Regress. For if Al + fatness do not add up to the fact of Al's being fat, then Al + fatness + EX won't either.  If Al and fatness can exist without forming the fact of Al's being fat, then Al and fatness and EX can all exist without forming the fact in question.  How can adding a constituent to the primary constituents bring about the fact-constituting unity of all constituents?  EX has not only to connect a and F-ness, but also to connect itself to a and to F-ness.  How can it do the latter?  The answer to this, presumably, will be that EX is a relation and the business of a relation is to relate. EX, relating itself to a and to F-ness, relates them to each other.  EX is an active ingredient in the fact, not an inert ingredient.  It is a relating relation, and not just one more constituent that needs relating to the others by something distinct from itself.  For this reason, Bradley's regress can't get started.  This is a stock response one can find in Brand Blanshard and others, but it is not a good one.

The problem is that EX can exist without relating the relata that it happens to relate in a given case.  This is because EX is a universal.  If it were a relation-instance as on D. W. Mertz's theory, then it would be a particular, an unrepeatable, and could not exist apart from the very items it relates.   Bradley's regress could not then arise.  But if EX is a universal, then it can exist without relating any specific relata that it does relate, even though, as an immanent universal, it must relate some relata or other.  This implies that a relation's relating what it relates is contingent to its being the relation it is.  For example, x loves y contingently relates Al and Barbara, which implies that the relation is distinct from its relating.  The same goes for EX: it is distinct from its relating.  It is more than just a constituent of any fact into which it enters; it is a constituent that does something to the other constituents, and in so doing does something to itself, namely, connect itself to the other constituents.  Relating relations are active ingredients in facts, not inert ingredients.  Or we could say that a relating relation is ontologically participial in addition to its being ontologically substantival.  And since the relating is contingent in any given case, the relating in any given case requires a ground.  What could this ground be? 

My claim is that it cannot be any relation, including the relation, Exemplification.  No relation, by itself, is ontologically participial. More generally, no constituent of a fact can serve as ontological ground of the unity of a fact's constituents.  For any such putatively unifying constituent will either need a further really unifying constituent to connect it to what it connects, in which case Bradley's regress is up and running, or the unifying constituent will have to be ascribed a 'magical' power, a power no abstract object could possess, namely, the power to unify itself with what it unifies.  Such an item would be a self-grounding ground: a ground of unity that grounds its unity with that which it unifies.  The synthetic unity at the heart of each contingent fact needs to be grounded in an act of synthesis that cannot be brought about by any constituent of a fact, or by the fact itself. 

My first objection to Moreland's theory may be put as follows.  The existence of a thick particular (which we are assaying as a concrete fact along the lines of Gustav Bergmann and David Armstrong) cannot be the fact's constituents' standing in the exemplification relation.  And existence itself, existence in its difference from existents, cannot be identified with the exemplification relation.

Can Existence Exist Without Being Uniquely Self-Existent?

I agree with Moreland that existence itself exists.  One reason was supplied by Reinhardt Grossmann: "If existence did not exist, then nothing would exist." (Categorial Structure of the World, 405) But I have trouble with the notion that existence itself is the exemplification relation.  Existence as that which is common to all that exists, and as that in virtue of which everything exists cannot be just one more thing that exists.  Existence cannot be a member of an extant category that admits of multiple membership, such as the category of relations.  For reasons like these such penetrating minds as Martin Heidegger, Roman Ingarden, and Panayot Butchvarov have denied that existence itself exists. 

In my 2002 existence book I proposed a synthesis of these competing theses:  Existence exists as a paradigm existent, one whose mode of existence is radically different from the mode of existence of the beings ontologically dependent on it.  From this point of view, Moreland has a genuine insight, but he has not taken it far enough: he stops short at the dubious view that existence is the relation of exemplification.  But if you drive all the way down the road with me you end up at Divine Simplicity, which Moreland has  good reasons for rejecting.

Is There Such a Thing as Metaphysical Explanation?

M. L. writes,
 
I've been enjoying your critique of [Peter] van Inwagen. [The reader is presumably referring to  my "Van Inwagen on Fiction, Existence, Properties, Particulars, and Method" in Studia Neoaristotelica: A Journal of Analytical Scholasticism, 2015, vol. 12, no. 2, 99-125]  I was initially astonished at his claim that metaphysics/ontology doesn't explain, but it also got me curious about where the explanation is going on in ontological accounts (especially of properties, however construed).
 
I'm doing a Ph.D. in metaontology and I'm contrasting neo-Quinean (van Inwagen) and neo-Aristotelian (Lowe) approaches. 
 
Can you direct me to where you might have written about, if indeed you have, how it is ontology/metaphysics explains?
 
Well, I haven't discussed the issue head-on in a separate publication, but I have discussed it en passant in various contexts. Below is a re-do of a 2012 weblog entry that addresses the question and may spark discussion. Combox open.
 
………………………
 

Let 'Tom' name a particular tomato.  Let us agree that if a predicate applies to a particular, then the predicate is true of the particular.  Predicates are linguistic items.  Tomatoes are not. If Tom is red, then 'red' is true of Tom, and if 'red' is true of Tom, then Tom is red. This yields the material biconditional

1. Tom is red iff 'red' is true of Tom.

Now it seems to me that the following question is intelligible:  Is Tom red because 'red' is true of Tom, or is 'red' true of Tom because Tom is red?  'Because' here does not have a causal sense.  So the question is not whether Tom's being red causes 'red' to be true of Tom, or vice versa.  So I won't speak of causation in this context.  I will speak of metaphysical/ontological grounding.  The question then is what grounds what, not what causes what.   Does Tom's being red ground the application (the being-applied)  of 'red' to Tom, or does the application (the being-applied) of 'red' to Tom ground Tom's being red?

I am not primarily concerned with the correct answer to this question, but with meaningfulness/intelligibility of the question itself.

Grounding is asymmetrical: if x grounds y, then y does not ground x.  (It is also irreflexive and transitive.)  Now if there is such a relation as grounding, then there will be a distinctive form of explanation we can call metaphysical/ontological explanation.  (Grounding, even though it is not causation, is analogous to causation, and metaphysical explanation, even though distinct from causal explanation, is analogous to causal explanation.)

Explaining is something we do: in worlds without minds there is no explaining and there are no explanations, including metaphysical explanations.  But I assume that, if there are any metaphysical grounding relations, then  in every world metaphysical grounding relations obtain.  (Of course, there is no grounding of the application of predicates in a world without languages and predicates, but there are other grounding relations. For example, if propositions are abstract objects that necessarily exist, and some of the true ones need truth-makers, then truth-making, which is a grounding relation, exists in worlds in which there are no minds and no languages and hence no sentences.)

Grounding is not causation. It is not a relation between event tokens such as Jack's touching a live wire and Jack's death by electrocution.  Grounding is also not a relation between propositions.  It is not a logical relation that connects propositions to propositions.  It is not the relation of material implication, nor is it entailment (the necessitation of material implication), nor any other logical relation wholly situated at the level of propositions.  Propositions, let us assume, are the primary truth-bearers. 

In our example, grounding is not a relation between propositions — it is not a logical relation — since neither Tom nor 'red' are propositions. 

I want to say the following.  Tom's being red grounds the correctness of the application of 'red' to Tom.  'Red' is true of Tom because (metaphysically, not causally or logically) Tom is red, and not vice versa.  'Red' is true of Tom in virtue of  Tom's being red.  Tom's being red is metaphysically prior to the truth of 'Tom is red' where this metaphysical priority cannot be reduced to some ordinary type of priority, whether logical, causal, temporal, or what have you.  Tom's being red metaphysically accounts for the truth of 'Tom is red.' Tom's being red makes it the case the 'red' is true of Tom.  Tom's being red makes 'Tom is red' true.  

I conclude that there is at least one type of metaphysical grounding relation, and at least one form of irreducibly metaphysical explanation. 

We can ask similar questions with respect to normative properties.  Suppose Jesus commands us to love one another.  We distinguish among the commander, the act of commanding, the content of the command, and the normative property of the commanded content, in this case the obligatoriness of loving one another.  If Jesus is God, then whatever he commands is morally obligatory. Nevertheless, we can intelligibly ask whether the content is obligatory because Jesus/God commands it, or whether he (rightly) commands it because it is obligatory.  The 'because' here is neither causal not logical.  It is metaphysical/ontological.
 
This of course a variation on the old Euthyphro Dilemma in the eponymous Platonic dialog.
 
I freely admit that there is something obscure about a grounding relation that is neither causal nor logical. But of course logical and causal relations too are problematic when subjected to squinty-eyed scrutiny. 
 
I conclude with a dogmatic slogan. Metaphysics without metaphysical explanation is not metaphysics at all.  

Penultimate Draft: Meinertsen Review for Metaphysica

REVIEW ARTICLE

Bo R. Meinertsen, Metaphysics of States of Affairs: Truthmaking, Universals, and a Farewell to Bradley's Regress, Springer 2018, 174 + xviii pp.

Summary

Professor Meinertsen's detailed treatment of states of affairs agrees with the spirit and much of the letter of David M. Armstrong's middle period as represented in his A World of States of Affairs (Cambridge UP, 1997). States of affairs in this acceptation are not abstract objects, as they are for some philosophers, but concrete denizens of the natural world of space-time. They are “unified complexes that are instantiations of properties or relations by particulars.” (1) Unlike Armstrong, however, Meinertsen is not concerned to argue for their existence (3, 13), or to show their utility in different philosophical areas. His focus is on states of affairs themselves, their main theoretical role, the nature of their constituents, and the problem of their unity.

Their main role is to serve as truthmakers. Suppose it is contingently true that Tom is red, where 'Tom' denotes a tomato of our acquaintance. (The use and justification of such “toy examples” is nicely explained on p. 5) Intuitively, such a truth is not just true; it needs an ontological ground of its truth. What might that be? Rejecting both tropes (Chapter 3) and D. W. Mertz's relation instances (Chapter 4) as truthmakers, Meinertsen argues that states of affairs do the job. In this example, the truthmaker is Tom's being red. On Meinertsen's use of terms, all and only states of affairs are truthmakers (84-85).

A state of affairs is a complex, and complexes are composed of distinct constituents. The composition of a state of affairs, however, is non-mereological. Mereological complexes are governed by the unrestricted composition axiom of classical mereology. (8) What the axiom states is that any plurality of items composes something: the existence of some items entails the existence of the sum of those items. The constituents of a state of affairs, however, can exist without the state of affairs existing. For example, Tom's being red entails the existence of the sum, Tom + instantiation + the universal redness. But the existence of the sum does not entail the existence of the state of affairs. A state of affairs, then, is a non-mereological complex. We will return to this important point when we come to the problem of the unity of a state of affairs.

First-order states of affairs have as their constituents particulars, properties or relations, and instantiation. The particulars are bare or thin (Chapter 5). What makes them bare is not that they lack properties, but the way they have them. The bareness of a bare particular consists in its instantiating, as opposed to including, its properties. (73) The properties that enter into states of affairs are sparse as opposed to abundant: not every predicate picks out such a property. In addition, the properties in states of affairs are universals, and thus multiply instantiable. If an immanent (transcendent) universal is one that cannot (can) exist uninstantiated, then Meinertsen's universals are immanent. Immanence so defined admits of abstractness. Meinertsen's universals, however, are concrete. (Chapter 8) The concrete is that which is “spatially and/or temporally located.” (119) Given naturalism, which Armstrong endorses and to which Meinertsen “inclines” (119), every existent is concrete and therefore located, including universals. The locatedness of universals, which is unlike that of particulars, has three implications. The first is that a universal is “wholly located in many places at the same time.” (120) The second implication is that “the region occupied by any such universal is not a mereological part of the region occupied by the whole thick particular.” (121) The third implication is that “more than one universal can have the same spatiotemporal location.” (121)

I note in passing that the banishing of so-called abstract objects demanded by uncompromising Armstrongian naturalism exacts a high price. The price is paid in the coin of the three implications just listed. The abstract-concrete distinction is replaced by a distinction between two categories of concreta, particulars and universals. This replacement requires that one accept the view that universals are ones-in-many (as opposed to ones-over-many) not merely in the sense that a universal cannot exist uninstantiated, but also in the sense that, if it exists, it is wholly present in each of its many spatiotemporal instances without prejudice to its being one and the same universal. This is a highly counter-intuitive consequence, as philosophers from Plato to R. Grossmann have appreciated, but it must be accepted by a states-of-affairs ontologist who is both a naturalist and an upholder of universals. (121)

Chapter 7 is devoted to relations, but in the interests of brevity I will not report on this chapter but advance to Chapters 9 and 10 which treat the problem of unity and Bradley's Regress respectively. This is the most exciting and original part of the book.

Meinertsen and I agree that the problem of the unity of a state of affairs is the central problem for a states of affairs ontology. The problem arises because states of affairs have “non-mereological existence conditions” (7): the existence of the constituents does not entail the existence of the state of affairs. What then accounts for there being one state of affairs having several distinct constitutents? What makes a one out of the many? A state of affairs is not just its constituents; it is these constituents unified. Something more is needed to make of the constituents a state of affairs. “I believe that it is the relating of a unique relation that is needed.” (8)

We can call this 'something more' the unifier. On Meinertsen's approach the unifier is internal to the state of affairs: “the unifier of a state of affairs is a [proper] constituent of it.” (135, emphasis in original) I added 'proper' in brackets to underscore that Meinertsen is not maintaining that states of affairs are self-unifying either in the positive sense that they unify themselves or in the privative sense that they are not unified by another. They are truth-making unities, but not as a matter of brute fact: they need a unifier to account for their unity. The unifier U is a special sort of relation, indeed it is a unique relation as I have just quoted him as saying. It relates the material constituents in the state of affairs, but it does so by being related to them. It is not just a relator of what it relates; it is a relator of what it relates by being related to what it relates. So if U relates the constituents of R(a, b), U does this by being related to each of them, including the relation R. This implies, of course, that U is not identical to R. Some say it is the business of a material relation to relate; not so on Meinertsen's view: it is the business of the formal relation U, and it alone, to relate. We also note that a consequence of U's being related to what it relates, and not merely a relator of what it relates, is that U enters as a constituent into every state of affairs. On an externalist view, by contrast, U unifies the constituents of a state of affairs S without entering into S as a constituent.

Now U is either related by another to what it relates, or it is related by itself to what it relates. If the former, then Bradley's regress is up and running, a regress both infinite and vicious. (Chapter 10) To avoid it, Meinertsen posits that “The U-relation is related to its relata by itself.” (143). This is what makes it unique: it is the only relation that has this “ability,” a word Meinertsen employs. This view, which he dubs “self-relating internalism,” has not been maintained before as far as I know. “To emphasize this unique self-relating ability of U on self-relating unternalism, I shall call it the 'U*-relation.'” (143) Because U* is a constituent of every state of affairs whose constituents it unifies, the monadic case of a's being F may be depicted as follows: U*(U*, F, a). The occurrence of the sign 'U*' both outside and inside of the parentheses indicates that the concrete universal U* is both the bringer of unity and one of the items brought into unity. It is a constituent of every state of affairs without which there would be no states of affairs.

Is U* the same as the instantiation relation? Meinertsen waxes coy: he is “inclined” to say that it is, but this would be an “extrinsic thesis.” What he means, presumably, is that a full assay of R(a,b) might list the following constituents: U*, dyadic instantiation, R, a, and b. Or it might list the foregoing items except instantiation. In the latter case, U* is instantiation. For example, “Edinburgh's being north of London is unified if and only if the U*-relation relates itself to being north of, Edinburgh and London.”(143). Either way, it would seem that U* must be a multi-grade relation, one that can be had by a variable number of items, and which therefore has different 'adicities.' For example, if U* is the instantiation relation, then U* is tetradic in U*(U*, R, a, b) but triadic in U*(U*, F, a). If U* is distinct from the instantiation relation I, then U* is pentadic in U*(U*, I, R, a, b) and tetradic in U*(U*, I, F, a). Meinertsen is aware of all this, and of the apparent problems that arise, but he thinks that they can be adequately dealt with. (157-159) The reasoning is intricate and obscure and to save space I will not comment on it.

The main point is that U* is the master concrete universal without which no state of affairs could exist. A state of affairs exists if and only its constituents are unified, and no plurality of constituents is unified in the state-of-affairs way as a matter of brute fact; ergo, unity demands a unifier as its ground. Being a universal, the unifier U* is multiply instantiable. Being concrete implies that U* cannot exist uninstantiated. It also implies that U*, if multiply instantiated, is multiply located and 'at work' in every state of affairs as that which ties its constituents into a state of affairs. As a self-relating relation, it does its work without igniting Bradley's vicious regress. (Chapter 10) To cop a line from Armstrong, “Nice work if you can get it.”

                                                                            Critique: The Problem of Unity

I will focus my critical remarks on Meinertsen's fascinating and original internalist theory of the unifier U*. What struck me about his theory is its structural similarity to the externalist suggestion I made in a number of my writings. (I thank Meinertsen for his close attention to them.) The points of similarity are the following. Meinertsen and my earlier self both accept that there are middle-Armstrongian states of affairs; that their main role is to serve as truthmakers; that they are complexes composed of distinct constituents; that the composition of these complexes is non-mereological; that their material constituents are particulars and universals; that the unity of a state of affairs, and therewith its difference from the mere plurality of its constituents, needs accounting in terms of a unifier; and above all, that there is a very special, indeed a unique, entity that serves as unifier. The main difference is that Meinertsen's unifier is a constituent of states of affairs while mine is external to states of affairs. Not only is there a similarity, but the two theories, as different as they are, are open to some of the same objections. But before discussing these objections, I want to state my objections to Meinertsen's account of unity, and how my theory avoids them.

First Objection

If there is a constituent of a state of affairs that explains its unity, this constituent must have a unique feature: it must be self-relating. But 'self-relating' has two senses, and this duality of senses give rise to a dilemma. Either (1) U* is self-relating only in the privative sense that it is not related by another to what it relates, supposing it is actually related to what it relates, or (2) U* is self-relating in the positive sense that it actually relates itself to what it relates. If (1), then U* blocks Bradley's regress, but fails to ground unity. It fails to ground the difference between the state of affairs, which is one entity, and the corresponding plurality of its constituents, which is a mere manifold of entities. If (2), then U* is an active as opposed to an inert ingredient in the state of affairs. It is a unity-maker, if you will. It plays a synthesizing role. It brings together the constituents, including itself, which otherwise would be a mere plurality, into a truthmaking unity.

But analysis cannot render this synthesizing intelligible, and therein lies the rub. All ontological analysis can do is to enumerate the constituents of a state of affairs, or, more generally, the parts of a whole. Analytic understanding proceeds by resolving a given whole into its parts, and ultimately into simple parts. But there is more to a (non-mereological) whole than its parts. There is the unity in virtue of which the parts are parts of a whole. The whole is one entity; the parts are many entities. Now if we try to understand this 'more' analytically we can do so only by positing a further part, a unifying part. I say 'posit,' not 'find.' In Fa, one can reasonably be said to find a particular and a character, but not a distinct copulative entity that grounds the truthmaking unity of the constituents. And so Meinertsen posits a unity-grounding entity. But the attempt to understand synthesis analytically is doomed to failure. First of all, no proper part of a whole is its unity, and this for the simple reason that the unity is the unity of all the parts. What one could say, though, is that the unity of the parts, which is distinct from any part, and from all of them, is brought about by a special part, the unifier. But then that special part, without ceasing to be a proper part, would have to exercise a synthesizing function. This synthesizing is what eludes analytic understanding. Simply to posit that the unifier U* has the ability to synthesize is make a kind of deus ex machina move. Leaving God out of it, Meinertsen's U* is a principium ex machina. I will come back to this later in connection with Meinertsen's talk of “inference to the best explanation.” (144) My present point is that even if there is some occult constituent internal to states of affairs that grounds and thus explains their contingent unity, its existence and its operation must remain a mystery and cannot be rendered perspicuous by the analytic method of constituent ontology. Let me explain further.

Does Meinertsen's U* exist? If there are states of affairs as Meinertsen conceives them, then U* has to exist. But if U* exists, then it is (a) a distinct entity independent of us and our synthetic activities, and (b) a distinct item that we can single out in thought if not in perception. If I see that a book is on a table, then I see a book, a table, and possibly also the relation referred to by 'on.' What I don't see, however, is the referent of 'is': the being of the book's being on the table. Since I don't see the being of the book's being on the table, I do not see U*. I cannot single it out in perception. Can I single it out in thought? To do so I would have to be able to distinguish U* from S, the state of affairs the unity of whose constituents U* grounds. There is a problem here. The ordinary (material) constituents in a state of affairs S are weakly separable: each such constituent could exist apart from every other one in S and apart from S itself, but not apart from every other entity. For example, let S = Fa. If Fa is a Meinertsenian state of affairs, then a can exist without instantiating F, and F can exist without being instantiated by a, and each can exist without being constituents of S. (The separability is said to be weak because a cannot exist without properties, and F cannot exist uninstantiated.) Now the immanent universal relation U* can exist apart from a and apart from F provided it is instantiated elsewhere, but not if it is the actual unifier of a and F. As the latter, as the active ingredient in S, it is inseparable from a, from F, and from S. But then U* is quite unlike the material constituents in S, which are all inert, and it is unintelligible in what exact sense U* is a constituent of S. The analytic assay lays out the constituents of a state of affairs, but it can do this only because of the logically antecedent unity of the constituents in virtue of which there is a state of affairs to assay. To understand this unity analytically by positing a special unifying constituent would make sense only if said constituent were inert like the material constituents. But of course it cannot be inert if its is to be a unity-grounder.

Another way of appreciating the problem is by asking what the difference is between U* as an active ingredient in S, and S. Clearly, S cannot exist without U*. But it is also true that U*, as the active ingredient in S that unifies precisely a and F, cannot exist without S. This is because U* is a unifying unifier only when instantiated/located in a state of affairs with determinate material constituents. In every state of affairs S in which the in rebus immanent universal U* exists, it unifies precisely the constituents of S, and cannot do otherwise. So U* and S are mutually inseparable. It follows that U* both is and is not weakly separable from S. As a constituent of S, U* is weakly separable from S. As an active ingredient and unity-maker, however, U* is not weakly separable from S. We ought to conclude that it is unintelligible how a (proper) constituent of a state of affairs could serve as its unifier. As a constituent, U* must be inert in S; as unifier, U* must be active. But it can't be both because it cannot be both weakly separable from S and not weakly separable from S.

An analogy may help clarify my criticism. The existence of two boards and some glue does not entail the existence of two boards glued together. That is obvious. It is also obvious that there would be no need for super-glue to glue the glue to the boards should someone glue the boards together. If there were a need for super-glue, then one would need super-duper-glue to glue the super-glue to the glue and to the boards, and so on. We can express this by saying that ordinary glue glues itself to what it glues; it is not glued by another to what it glues. In this sense, ordinary glue is self-gluing. This is in analogy to Meinertsen's claim that U* is self-relating. But note that 'self-gluing' can only be taken in a privative, not a positive, sense. The same goes for 'self-relating.' By 'privative' I simply mean that the self-gluing glue is not glued by another. If the glue and the relation U* were self-gluing and self-relating in a positive sense, then they would be agents of an action. They would be active as opposed to passive or inert. But surely self-gluing glue does not do anything: it does not apply itself to the boards or bring it about that the two boards are glued together; self-gluing glue is merely such that if the two boards are glued together by a genuine agent, no further glue would be needed to glue the glue to what it glues. Likewise, self-relating U* does not do anything: it does not bring it about that U*, a, and F are 'cemented' into a state of affairs; it is merely such as to insure that if U*, a, and F are brought together to form a state of affairs, no further formal U-type relations are needed to do the job.

Meinertsen credits me with appreciating that the problem of regress-avoidance and the problem of unity are two and not one. “As Vallicella (2004, p. 163) . . . eloquently puts it: 'A regress-blocker is not eo ipso a unity-grounder, pace Russell, Alexander, Blanshard, Grossmann, et al.'” If I am right, however, Meinersten has not really taken this insight on board. My point against him is that his U* can do only the regress-blocking job but not the unifying job. The problem is that no constituent of a state of affairs can do the unifying job. A fortiori, no relational constituent can do the job. By my lights, Meinerten fails to appreciate this, and it may be that he fails to appreciate it because he illicitly slides from the privative sense to the positive sense of 'self-relating.'

Second Objection

On Meinertsen's internalist theory, the unifier U* is a constituent of every state of affairs. Now corresponding to every state of affairs there is the sum of its constituents. So, corresponding to a's being F, there is the sum a + U* + F. Clearly, the particular a in the state of affairs is numerically the same as the particular a in the sum, and the universal F in the state of affairs is numerically the same as the universal F in the sum. The state of affairs and the sum share these material constituents and do not differ in respect of them. But what about the concrete universal U*? Is it numerically the same entity in the state of affairs and in the sum? If yes, then trouble, and if no, then trouble.

States of affairs are contingent. The contingency of a state of affairs derives from the contingent unity of its constituents. So it must be possible that the same constituents exist either unified or not unified. Thus the state of affairs and the sum must have the same constituents. Now U* is a constituent. It follows that U* must be be numerically the same in both the state of affairs and the corresponding sum. Two items, x, y, are numerically the same just in case thay have all the same properties. So U* must be either inert in both state of affairs and sum, or active in both. Now if U* is inert in both, then no state of affairs is constituted. If, on the other hand, U* is active in both, then the unity of the state of affairs is necessary. (For if U* is active in both, then there is no difference between the state of affairs and the sum. ) Either way, no contingent state of affairs is constituted. Therefore, U* cannot be numerically the same in both state of affairs and corresponding sum.

If, on the other hand, U* is active in the state of affairs, but inert in the sum, we get the same problem. A state of affairs is contingent just in case its constituents can exist without forming a state of affairs. It must be possible for the same constituents to be either unified into a state of affairs or not so unified. But active U* is not the same as inert U*. It follows that the state of affairs and the sum do not have the same constituents, which implies that the state of affairs is not contingent, but necessary. We ought to conclude that the unifier of a state of affairs cannot be a constituent thereof.

Third Objection

The first objection focused on the existence conditions of states of affairs; the third focuses on the existence conditions of concrete universals, in particular, the existence conditions of U*. What I will try to show is that Meinertsen's theory is involved in an explanatory circulus vitiosus. Roughly, he attempts to explain the unity, and thus the existence, of a state of affairs by positing a special unifying constituent when that very constituent can exist only in a state of affairs. Here is my argument:

a) A state of affairs exists if and only if its constituents form a unity.
b) U* is a constituent of states of affairs that explains their unity.
Therefore
c) U* is a constituent of states of affairs that explains their existence. (from a, b)
d) U* cannot exercise its explanatory function unless it exists.
Therefore
e) The existence of U* explains the existence of states of affairs.
But
f) U* cannot exist except in a state of affairs.
Therefore
g) The existence of states of affairs explains the existence of U*
h) Given the asymmetry of explanation, (e) and (g) are contradictory, and Meinertsen's explanation of the existence of states of affairs in terms of U* is viciously circular.

The above argument rests on the following assumptions. First, there is such a procedure as metaphysical explanation. Second, it is asymmetrical: if x explains y, then it is not the case that y explains x. Third, a circular explanation, violating as it does the asymmetry of explanation, is not an explanation, or is not a successful explanation. Fourth, the unity/existence of states of affairs, being modally contingent, needs explanation, i.e., it cannot be a factum brutum. Meinertsen is committed to all four assumptions. He is committed to the first since he accepts truthmaking. The truthmaker metaphysically (not logically and not causally) explains the truth of the truth-bearer. He is obviously committed to the second and third. He is committed to the fourth because he takes seriously the problem of unity, which is the problem of explaining the difference between a state of affairs and the mere plurality of its constituents.

An External Unifier Avoids the Above Objections

I admit that the theory of my earlier self is not much better than Meinertsen's: in the final analysis they are both unsatisfactory, although for different reasons. But my theory does avoid the above objections. Meinertsen gets into trouble by making his unifier U* a constituent of states of affairs. This exposes him to the first objection because no constituent of a state of affairs could be an active, unity-grounding ingredient. Or at least it is unintelligible how anything like that could exercise a synthesizing function. A state of affairs is a synthetic unity the synthetic character of which cannot be understood by ontological analysis. Analytical understanding here reaches one of its limits. An ontological assay is merely a list of constituents. But the unity of these constituents is not a further item on the list. Nor can adding a special constituent to explain this unity avail anything. For either this further constituent is inert or it is active. If the former, no progress as been made in accounting for unity. If the latter, then the further constituent must be ascribed a special synthesizing power that nothing else has, and that nothing that analysis could reveal could have. How could analysis reveal such an occult power?

U*'s being a constituent opens Meinertsen to the second objection because a state of affairs is contingent only if the same constituents can exist either unified or not. But this sameness is impossible if U* is both a constituent and a unifier. U*'s being a constituent also exposes him to the third objection because no constituent can exist without being a constituent of some state of affairs or other. So if the unifier is a constituent, then it cannot exist unless states of affairs exist. This however gives rise to the explanatory circle. We ought to conclude that if there is a unifier, then it cannot be internal.

My external unifier unifies but without thereby entering into the states of affairs whose unity it brings about. It thereby evades all three of the objections lately listed. Kant's transcendental unity of apperception provides a model of an external unifier. That which brings about the synthesis of representations in the unity of one consciousness, thereby constituting an object of experience, is not itself a part of the object so constituted. One obvious objection from a realist, naturalist, and empiricist point of view to an external unifier, whether developed along transcendental lines or, as in my 2002, along onto-theological lines, is that it leads us away from realism to idealism. It brings mind into the picture as the synthesizing factor. But if (irreducible) mind is brought in, then naturalism is abandoned for some sort of 'spiritualism.' Empiricism too is abandoned if one invokes an external unifier along either transcendental or onto-theological lines.

The brings me to the deus ex machina objection that has been lodged against my proposal. Roughly, I put God to work to solve the problem of the unity of states of affairs. (“God has his uses,” my teacher J. N. Findlay once said.) Curiously, Meinertsen is open to a similar objection, call it principium ex machina. He does not call upon God, but upon a sui generis entity, U*, which is unique among concrete universals due to its synthesizing power. Well, what exactly is wrong with these ex machina moves? Meinertsen and I will be told that the moves are objectionably ad hoc. Meinertsen is sensitive to the criticism:

The U*-relation is of course an 'ad hoc' entity in the sense that it is only introduced to solve a problem, viz. the problem of unity. Some authors, such as Betti (2015), would consider that a big drawback of self-relating internalism. However, one man's 'ad hoc' – solution is another man's inference to the best explanation. (143)

If any of my three objections above are sound, however, Meinertsen's inference to the best explanation is an inference to an explanatory entity that cannot exist or at least cannot be intelligibly posited.

Meinertsen on the Merely Apparent Existence of Thick Particulars

Meinertsen's bookThis is the second in a series on Bo Meinertsen's 2018 book. It is part of a 'warm-up' for a review article to appear in Metaphysica.  Here is the first installment.

A thick particular in the parlance of David Armstrong is an ordinary particular taken together with its non-relational properties. But an ordinary particular is distinct from each and from all of its properties: it is that which has these properties. If we consider an ordinary particular in abstraction from its properties, what we have before our minds is the particular qua particular. From here it is but a short step to the much maligned and hotly contested bare or thin particular. Meinertsen ably defends bare or thin particulars as constituents of states of affairs in Chapter 5.

A tomato will serve as an example. Call it 'Tom.' There are any number of contingent truths about Tom. Tom is red; Tom is ripe; Tom is round; etc.  Meinertsen and I agree that these truths need truthmakers. As I would put it, they can't just be true. What in the world makes them true? For Meinertsen, states of affairs (STOAs) play the truthmaker role.  A (first-order) state of affairs is a unified complex consisting of the instantiation of a property by a thin particular, or the instantiation of a relation by two or more thin particulars. Instantiation is an asymmetrical external relation that, in the monadic case, connects a thin particular to a property thereby forming a state of affairs.  The truth that Tom is red is thus made true by the state of affairs, Tom's being red, where the subject constituent is a thin particular, thin-Tom if you will, and not thick-Tom, Tom together with his intrinsic properties. And the same goes for the truth that Tom is ripe, and the truth that Tom is round.  For each truth there is a truthmaking state of affairs, a thin state of affairs we can call it since it includes only one of thick-Tom's properties.  

Now take the conjunction of all of Tom's intrinsic properties. The result is a conjunctive property.  Call it the nature N of Tom. The instantiation of this nature by a thin particular is a state of affairs.  This is because N is a bona fide property, and the instantiation of any property by a thin particular is a state of affairs. This state of affairs is a thick state of affairs, and is identical to the thick particular, Tom. So the following comment (in the earlier thread) by Meinertsen comes as a bit of a surprise:

As to (4), well, in my view, thick particulars aren’t real STOAs, merely apparent ones. It’s true that I assay a thick particular as the instantiation of N, the conjunctive property that is the conjunction of its intrinsic properties. But I also argue that conjunctive properties are truthmaking reducible (TM-reducible) – i.e. only existing at the level of truths, not at the level of truthmakers – and that the instantiation (‘instantiation’) of a TM-reducible property isn’t a real STOA.

This is puzzling because the dialectic started with a really existent thick particular, Tom together with his properties, but seems to end with the elimination of the starting point and the demotion of the thick particular to a mere appearance.

The reasoning seems to proceed as follows. The contingent truth that a is F needs a truthmaker, and so does the contingent truth that a is G. But the conjunction of the two truths — which is 'automatically' true given the truth of the conjuncts — does not need its own truthmaker.  So these three truths need only two truthmakers. There is no need for a third truthmaker because the truth of the conjunctive proposition supervenes on the truth of its conjuncts. It's an aletheiological 'free lunch.'

Now consider the conjunction C of all the truths about a, or about Tom in our example. What makes this conjunction true are the 'thin' states of affairs corresponding to and grounding each of the truths  in the conjunction. The 'thin' states of affairs do all the truthmaking work: there is no need for a separate 'thick' state of affairs to serve as truthmaker for the conjunction itself.  But if there is no need for 'thick' states of affairs, then there is no need to posit thick particulars in reality. (A thick particular just is a 'thick' state of affairs.)  So thick particulars are best regarded as merely apparent.

That is the argument as far as I can tell. Did I get it right, Bo?

Critique

But if there is no thick particular in reality, then what makes it the case that each of the thin particulars in each of the thin states of affairs is the same thin particular? Meinertsen speaks above of "the conjunctive property that is the conjunction of its intrinsic properties." (emphasis added) What is the antecedent of the pronoun 'its'?  That would have to be Tom in our example, thick-Tom, Tom together with all its properties. So the very identity of C — its being the conjunction it is and not some other conjunction — presupposes the reality of thick-Tom, Tom together with his intrinsic properties.  For C to exist and to be true, thick-Tom must exist. 

I conclude that one cannot take thick particulars to be merely apparent.  Their reality is presupposed if the STOA style of ontology is to get off the ground in the first place.

Now the tomato example is what Meinertsen rightly calls a "toy example." (5).  We philosophers employ such examples for convenience ignoring the fact, if it is fact, that tomatoes and other meso-particulars are not ontologically fundamental. So it may make sense to say that thick-Tom and his colleagues do not really exist. But surely the micro-entities of physics do exist and are thick particulars and thus 'thick' states of affairs. There have to be some thick particulars somewhere.

On p. 70, Meinertsen tells that at the level of truthmakers, there are no such things as molecules. Presumably he will say the same about their constituent atoms.  But what about sub-atomic particles?  Could he be telling us that, no matter how far down we go, we will never encounter anything fundamentally real?

Non-Substantial Change, Trope Bundle Theory, and States of Affairs

MeinertsenI am presently writing a review article for Metaphysica about Bo R. Meinertsen's Metaphysics of States of Affairs: Truthmaking, Universals, and a Farewell to Bradley's Regress (Springer 2018). Since I will probably incorporate the following critical remarks into my review, I want to give Bo a chance to respond. 

Substantial and Non-Substantial Change

One way a thing can change is by coming into being or passing away. This is called substantial change.  We could also call it existential change. The other way could be called alterational change. This occurs when a thing, persisting for a time, alters in respect of its intrinsic properties during that time.  Consider the ripening of a tomato. This typically involves the tomato's going from green to red.  This change in respect of color is an alterational, or accidental, or non-substantial change. One and the same entity (substance) persists through a non-zero interval of time and instantiates different properties (accidents) at different times.  As I would put it, there is no alterational change without existential unchange: numerically the same tomato is green, hard, inedible, etc. at time t and red, soft, and edible at later time t*. Bo and I are both assuming that things in time persist by enduring, not by perduring. 

The Problem of Non-Substantial Change of Continuants

This is 

. . . the problem of how to ground the fact that continuants 'persist through change'. For instance, a tomato's changing from red to green [sic] is a case of non-substantial change, and how do we ground the fact that the tomato that has changed exists both before and after the change? The bundles of basic trope theory essentially have the members they actually have and are therefore incompatible with such change. (Meinertsen 2018, 49)

The problem is that we want to say that one and the same tomato goes from being green to being red. We want to be able to uphold the diachronic identity of the tomato as it alters property-wise.  But this is impossible on basic bundle-theoretic trope theory because trope bundles have their members essentially.  This means that if bundle B has trope t as a member, then it is impossible that B exist without having t as a member. The counterintuitive upshot is that a green tomato assayed as a bundle of tropes ceases to exist when it ceases to be green.  This implies that our tomato when so assayed cannot undergo alterational, or accidental, or non-substantial change when it goes from green to red, hard to soft, etc.  It implies that every change is a substantial change. I agree with Meinertsen that this is a powerful objection to the basic bundle-of-tropes assay of ordinary particulars.

Does a State of Affairs Ontology Face the Same Problem?

Meinertsen says that it does not:

State of affairs ontology has no problem in dealing with the problem of non-substantial change. None of the properties of a particular in a state of affairs — which as we shall see in Chap. 5 is a bare particular — is included in it, as opposed to instantiated by it. Hence, it changes non-substantially if and only it ceases to instantiate at least one of these properties or whenever it instantiates a new property. (49)

It seems to me, though, that states of affairs (STOA) ontology faces, if not the very same problem, then a closely related one.

Critique

It is true that a bare particular does not include its properties: the bare or thin particular stands to its properties in the asymmetrical external relation of instantiation.  So what Meinertsen is telling us is that it is the bare particular that remains numerically the same over time while some of its properties are replaced by others. This is what grounds the diachronic numerical identity of the continuant.  The substratum of change is the bare particular 'in' the tomato, not the tomato as a whole.

But this answer is less than satisfactory. What changes over time is not a thin particular, but a thick particular. It is the green tomato with all its properties that loses one or more of them and becomes a red tomato.  This is supported by the fact that we do not see or otherwise perceive the thin particular; we do, however, see and otherwise perceive thick particulars.  What we have before us is a tomato that we see to be green and feel to be  hard, etc.,  and that we then later see to be red and feel  to be soft, etc.  

Arguably, then, it is the thick particular that is the substratum of non-substantial change, not the thin particular. If so, then a problem arises similar to the problem that arose for the bundle-of-tropes theory. How?

Well, the green tomato is a STOA whose nature is N1, where N1 is a conjunctive property the conjuncts of which are all the intrinsic properties of the green tomato. The red tomato is a STOA whose nature is N2, where N2 is a conjunctive property the conjuncts of which are all the intrinsic properties of the red tomato.  These STOAs differ numerically for they differ in one or more constituents.  The first has greenness as a constituent, the second does not. A STOA is a complex, and two complexes are the same iff they have all the same constituents. 

So what's the problem? The problem is that any non-substantial change in the green tomato assayed as a STOA destroys its identity just as surely as any non-substantial change in the green tomato assayed as a bundle of tropes destroys its identity. On either account, there is no adequate explanation of non-substantial change. This is because there is no numerically self-same substratum of change that endures through the change in properties. The thin particular is not plausibly regarded as the substratum. I note en passant that Gustav Bergmann regarded bare particulars as momentary entities, not as persisting entities.

The problem set forth as an aporetic sextad:

  1. There is no change in intrinsic properties of an ordinary particular over time without a numerically self-same substratum of change. (endurantist assumption)
  2. The green tomato changes to red. (pre-theoretical datum)
  3. The green tomato that changes to red is a thick particular. (pre-theoretical datum)
  4. Thick particulars are STOAs. (theoretical claim)
  5. STOAs are complexes. (true by definition)
  6. Two complexes are the same iff they share all constituents. (theoretical claim)

These six propositions are collectively inconsistent. My question to Meinertsen: which of these propositions will you reject? Presumably, he will have to reject (3) and say that 'the green tomato' refers to an invisible thin particular, and it is this item that changes from green to red and that serves as the substratum of change.

What do I say?  For now I say merely that, pace Bo, on the issue before us, STOA ontology is no better than the bundle-of-tropes theory.

Indexicality and Omniscience

Patrick Grim gives something like the following argument for the impossibility of divine omniscience. What I know when I know that

1. I am making a mess

is an indexical fact that no one else can know. At most, what someone else can know is that

2. BV is making a mess

or perhaps, pointing to BV, that

3. He is making a mess.

Just as no one except BV can refer to BV by tokening the first-person singular pronoun, no one except BV has access to the indexical fact that, as BV would put it to himself, I am BV. Only BV is privy to this fact; only BV knows himself in the first-person way. Now an omniscient being knows everything that can be known. Although I am not omniscient, there is at least one proposition that I know — namely (1) — that is not known by any other knower, including an omniscient knower. So an omnisicent being is impossible: by its very definition it must know every fact that can be known, but there are indexical facts that it cannot know. God can know that BV is making a mess but he cannot know what I know when I know that I am making a mess. For any subject S distinct from God, the first-person facts appertinent to S are inaccessible to every mind distinct from S, including God's mind. That is what I take to be Grim's argument.

I suppose one could counter the argument by denying that there are indexical facts.  But since I hold that there are both indexical propositions and indexical facts, that response route is not available to me.  Let me see if I can respond by making a distinction between two senses of 'omniscience.'

A. X is omniscient1=df X knows every fact knowable by some subject or other.

B. X is omniscient2 =df X knows every fact knowable by some one subject.

What indexical facts show is that no being is or can be omniscient in the first sense. No being knows every indexical and non-indexical fact. But a failure to know what cannot be known does not count against a being's being omniscient in a defensible sense of this term any more than a failure to do what cannot be done counts against a being's being omnipotent. A defensible sense of 'omniscience' is supplied by (B). In this second sense, God is omniscient: he knows every fact that one subject can know, namely, every non-indexical fact, plus all facts pertaining to the divine subjectivity. What more could one want?

Since no being could possibly satisfy (A), (A) is not the appropriate sense of 'omniscience.' Compare omnipotence. An omnipotent being cannot be one who can do just anything, since there are both logical and non-logical limits on what any agent can do.  Logical: God cannot actualize (create) an internally contradictory state of affairs. Non-logical: God cannot restore a virgin.  So from the fact that it is impossible for God to know what is impossible for any one being to know, it does not follow that God is not omniscient.

To sum up. There are irreducible first-personal facts that show that no being can be omniscient in the (A)-sense: Patrick Grim's argument is sound. But the existence of irreducible first-personal facts is consistent  with the truth of standard theism since the latter is committed only to a being omniscient in the (B)-sense of 'omniscience.'

Is Sin a Fact? A Passage from Chesterton Examined

A correspondent asked me my opinion of the following passage from G. K. Chesterton:

Modern masters of science are much impressed with the need of beginning all inquiry with a fact. The ancient masters of religion were quite equally impressed with that necessity. They began with the fact of sin — a fact as practical as potatoes. Whether or no man could be washed in miraculous waters, there was no doubt at any rate that he wanted washing. But certain religious leaders in London, not mere materialists, have begun in our day not to deny the highly disputable water, but to deny the indisputable dirt. Certain new theologians dispute original sin, which is the only part of Christian theology which can really be proved. Some followers of the Reverend R. J. Campbell, in their almost too fastidious spirituality, admit divine sinlessness, which they cannot see even in their dreams. But they essentially deny human sin, which they can see in the street. The strongest saints and the strongest sceptics alike took positive evil as the starting-point of their argument.

What Chesterton is saying is that sin is a fact, an indisputable fact, whether or not there is any cure for it. Not only is sin a fact, original sin is a fact, an observable fact one can "see in the street." Chesterton also appears to be equating sin with positive moral evil.

Is moral evil the same as sin? If yes, then the factuality of moral evil entails the factuality of sin. But it seems to me that moral evil is not the same as sin. It is no doubt true – analytically true as we say in the trade — that sins are morally evil; but the converse is by no means self-evident. It is by no means self-evident that every moral evil is a sin.  Let me explain.

An Exchange on the Metaphysics of Truthmaking

Dan M:

Discussing a puzzle about divine simplicity has led us to the metaphysics of truthmaking; I'll just focus on the latter for now – but the broader dialectic is this: I was thinking that a particular view about truthmaking can help us with that puzzle about simplicity. [Cf. first related article below.]

Take your sentence 'Al is fat', and suppose it's true. I agree it must be somehow *made* true, and I agree it can't be made true by Al, or fatness, or the sum or set of the two.

I suspect that we disagree about the following question: Must the sentence be made true by an item (entity, etc.)? If we answer "yes", then the natural proposal is to posit an entity with, as you say, a proposition-like structure, such as a state of affairs of Al's being fat. But suppose we answer "no": though 'Al is fat' must, if true, be made true, it needn't be made true by an item. How could it be made true without being made true by an item? Well suppose we express its being made true as follows:

(*) The sentence 'Al is fat' is true because Al is fat.

That is, the sentence (a linguistic item) is true because Al (a man) is fat. The sentence to the right of 'because' in (*) expresses what it is about the world in virtue of which the sentence 'Al is fat' is true. But (*) nowhere refers to an *item* of Al's being fat. The only referring term appearing to the right of 'because' is 'Al'.

Bill V:

Dan grants that some truthbearers need truthmakers, but thinks that truthmakers needn't be entities. Right here I must lodge an objection. A truthmaker is an entity by definition.  That truthmakers are entities is built into the theory. If the true sentence 'Al is fat' (or the proposition expressed by a thoughtful utterance of this sentence) needs a truthmaker, then this sentence/proposition cannot just be true: there must be, external to the sentence/proposition, an entity that 'makes' it true.  But of course this entity cannot itself be a truthbearer, whether a declarative sentence, a Fregean proposition, an Aristotelian proposition, a judgment, a statement, a belief, or any cognate item.  This point is crucial, so forgive me for belaboring it a bit.

Suppose we have a valid deductive argument all of the premises of which are true.  Then, from Logic 101, we know that the conclusion must also be true.  To put it precisely, and taking care not to confuse the necessitas consequentiae with the necessitas consequentiis: Necessarily, if the premises are all true, then the conclusion is true.  In this precise sense the truth of the premises necessitates the truth of the conclusion.

Could one say that the conjunction of the premises 'makes true' the conclusion, that the conjunction of premises is the truthmaker  of the conclusion?  One could say this, but this is not what truthmaker theorists mean when that say that a truthmaker makes true a truthbearer, or that a truthbearer needs a truthmaker.   What they mean is that some if not all truthbearers need truthmakers that are not truthbearers.

As I use 'truthmaker,' no truthmaker is a truthbearer.  (I ignore some recherché counterexamples.) So the proposition Tom is tall is not the truthmaker of the proposition Someone is tall. And this despite the fact that the first proposition entails the second. Does the second proposition have a truthmaker? Yes. In fact it has more than one. Tom's being tall is one, Bill's being tall is another.  But these are not propositions, but ontological grounds of true propositions.

So if 'Al is fat' has a truthmaker, then there exists an entity external to this sentence and to every sentence (proposition, etc.) that makes the sentence (proposition, etc.)  true.  If entailment is a logical relation, then truthmaking is not a logical relation.  Logical relations connect propositions to propositions; truthmaking, however, connects a non-propositional chunk of external reality to a proposition (or cognate item).  Al's being fat, for example, is not a proposition.  It is a state of affairs or concrete fact. Propositions are either true or false, but it is neither; it either exists or it does not. If it exists, then it it can serve as the truthmaker of 'Al is fat.'  Concrete (Armstrongian) states of affairs are not bipolar or bivalent items. In this respect they are not like Chisholmian-Plantingian abstract states of affairs.

What Dan should say is there is no need for truthmakers, not that truthmakers needn't be entities. 

 Dan offers

(*) The sentence 'Al is fat' is true because Al is fat

to show that a truthmaker need not be an entity.

It seems to me, though, that Dan is confusing a truthmaker with a truth condition.  A truthmaker is concrete chunk of extralinguistic and extramental reality whereas a truth condition is just another sentence, proposition, or cognate item.  Our old friend Alan Rhoda in an old blog post does a good job of explaining the distinction:

. . .truthmakers are parcels of reality . . . .

Not so with truth conditions. Truth conditions are semantic explications of the meaning of statements. They tell us in very precise terms what has to be true for a particular statement to be true. For example, a B-theorist like Nathan Oaklander will say that the truth conditions of the sentence "The 2006 Winter Olympics are over" is given by the sentence "The 2006 Winter Olympics end earlier than the date of this utterance". Thus truth conditions are meaning entities like statements that are used to spell out or analyze the meaning of other statements.
 

Dan's (*) merely sets forth a truth condition. It doesn't get us off the level of propositions and down to the level of truthmakers.

Another important point has to do with the asymmetry of truthmaking: if T makes true p, it does not follow that p makes true T.  It's an asymmetry of explanation. If one thing explains another, it does not follow that the other explains the one. The truthmaker theorist takes seriously the project of metaphysical explanation. Truthmakers explain why true truthbearers are true.  Dan's (*), however, entails the following non-explanatory biconditional:

(**) The sentence 'Al is fat' is true iff Al is fat.

But (**) has nothing to do with truthmaking; it is but an instance of Quine's disquotational schema according to which the truth predicate is but a device of disquotation. We remain on the level of sentences (propositions, etc.)

In sum, I see no merit in Dan's suggestion that there are truthmakers but they needn't be entities. That shows a failure to grasp the notion of a truthmaker. What Dan should say is that there is no need for truthmakers.  He might also try arguing that the truthmaking relation is bogus or unintelligible since it is neither a logical relation nor a causal one.  

Related articles
Mental Act Nominalism with an Application to Divine Simplicity
Is New Jersey an Artifact? And Everything Else Too?
In a Philosophical Conversation, Three's a Crowd
Facts, Opinions, and Common Core
Does the Atheist Deny What the Theist Affirms?
Carnap and Clarity
Objective Truth as a Condition of Intelligibility

Mental Act Nominalism with an Application to Divine Simplicity

This entry continues a discussion with Dan M. begun here.  

Before we get to the main event, a terminological quibble.  A view that denies some category of entity I would call eliminativist, not nominalist. I say this because one can be a nominalist about properties without denying their existence. Tom is a tomato of my acquaintance. Tom is red and ripe and juicy and other things besides.  It is a Moorean fact, I would say, that Tom has properties, and that, in general, things have properties.  After all, Tom is red and ripe, etc. It's a datum, a given, a starting point.  A sensible question is not whether there are properties, but what they are. Of course there are properties. What is controversial is whether they are universals or particulars, mind-dependent or mind-independent, immanent or transcendent, constituents or not of the things that have them, etc.

Still, there are those parsimonious souls who deny that there are properties. They accept predicates such as 'red' and 'ripe' but deny that in extralinguistic reality there are properties corresponding to these or to any predicates.  These people are called  extreme nominalists.  It's a lunatic position in my view valuable only as a foil for the development of a saner view.  But moderate nominalism is not a lunatic view. This is the view that there are properties all right; it's just that properties are not universals, but particulars, trope theory being one way of cashing out this view.  My Trope category goes into more detail on this.

The present point, however, is simply this: a moderate nominalist about properties does not deny the existence of properties.  So my suggestion is that if you are out to deny some category K of entity (i.e., deny of a putative category that it has members) then you should label your position as eliminativist about Ks, not nominalist about Ks.  Dan is an eliminativist about mental acts, not a nominalist about them.

But this is a merely terminological point.  Having made it, I will now irenically acquiesce in Dan's terminology for the space of this post.  Dan writes with admirable clarity:

As you explain my proposal (I'll call it "Mental Act Nominalism" or "MAN"), an ontological assay of propositional attitudes will only turn up two entities, the agent and the proposition. The agent's having the relevant attitude (e.g., belief, doubt) to the proposition is not itself construed as an additional entity. You say that this view is committed to "a denial of mental acts and thereby a denial of the act-content distinction."

[. . .]

Turning to your concern. You suggest that "such a parsimonious scheme cannot account for the differences among" various propositional attitudes (belief, doubt, etc.). And after discussing some examples, you say they provide "phenomenological evidence that we cannot eke by with just the subject and the object/content but also need to posit mental acts." And you add: "The differences among [various attitudes] will then be act-differences, differences in the type of mental acts."

The gist of my reply is that we can perhaps account for the differences you speak of without committing ourselves to the existence of the relevant mental acts/states.

Consider these two situations:

(A) Dan wonders whether Bill owns cats.

(B) Dan believes that Bill owns cats.

(We may suppose there was a time lapse between them.) What should the ontological assays of (A) and (B) include? As you described MAN, its ontological assays of propositional attitudes deliver just two entities, the relevant agent and proposition. So on this approach, we get these two assays:

(A Assay 1) Dan, the proposition Bill owns cats.

(B Assay 1) Dan, the proposition Bill owns cats.

These assays fail to differentiate situations A and B. However, it's not clear to me that MAN has to be implemented in this way. Consider these alternative assays:

(A Assay 2) Dan, the relation wondering whether, the proposition Bill owns cats.

(B Assay 2) Dan, the relation believing that, the proposition Bill owns cats.

These assays do differentiate A and B, by virtue of the different relations. I think MAN is prima facie compatible with these assays, since the main aim of MAN is not to deny the existence of propositional attitude relations per se, but to deny the existence of mental acts or states consisting in the agent's having the relevant attitude. So, MAN must reject, for example, these assays:

(A Assay 3) Dan, the proposition Bill owns cats, the state Dan's wondering whether Bill owns cats.

(B Assay 3) Dan, the proposition Bill owns cats, the state Dan's believing that Bill owns cats.

So perhaps we can be realists about propositional attitude relations, but nominalists about propositional attitude states (of affairs). The former would give us a robust basis to differentiate different kinds of propositional attitudes, while the latter would preserve MAN.

BV: The issue is now one of deciding which tripartite assay to accept, mine, or Dan's.  Where I have mental acts or states, he has relations. Mental acts are datable particulars, where a particular is an unrepeatable item.  Dan's relations are, I take it, universals, where a universal is a repeatable item. 

Suppose that Dan, who has not seen his elderly neighbor Sam come out of his house in a week, fears that he is dead. What does the world have to contain for 'Dan fears that Sam is dead' to be true?  Suppose that it contains Dan, the relation fears that, and the proposition Sam is dead, but not the mental act, state, or event of Dan's fearing that Sam is dead.  Then I will point out that Dan, the relation fears that, and the proposition Sam is dead can all three exist without it being the case that Dan fears that Sam is dead.  The collection of these three items does not suffice as truthmaker for the sentence in question.

This is the case even if the relation in question is an immanent universal, that is, one that cannot exist instantiated. It could be that Dan exists, the proposition Sam is dead exists, and the relation fears that exists in virtue of being instantiated by the pair (Pam, the proposition Hillary is sad.)  It is possible that all three of these items exist and 'Dan fears that Sam is dead' is false.

We need something to tie together the three items in question.  On my tripartite analysis it is the mental act that ties them together. So I am arguing that we cannot get by without positing something like the particular Dan's fearing that Sam is dead.

How can a simple God know contingent truths, such as Bill owns cats? On the version of MAN that accepts bona fide relations, we say: God bears the relation believing that to the proposition Bill owns cats. There are just three entities to which this situation commits us: God, the relation, and the proposition. There is no state (construed as a bona fide entity) of God's believing that Bill owns cats.

BV: But if S bears R to p, this implies that R is instantiated by the ordered pair (S, p), and that this relation-instantiation is a state or state of affairs or event.  It is clearly something in addition to its constituents inasmuch as it is their truthmaking togetherness. And this bring us back to our original difficulty of explaining how a simple God can know contingent truths.

Related articles
Again on Divine Simplicity and God's Knowledge of Contingent Truths

Against Ostrich Nominalism

As magnificent a subject as philosophy is, grappling as it does with the ultimate concerns of human existence, and thus surpassing in nobility any other human pursuit, it is also miserable in that nothing goes uncontested, and nothing ever gets established to the satisfaction of all competent practitioners.  (This is true of other disciplines as well, but in philosophy it is true in excelsis.) Suppose I say, as I have in various places:

That things have properties and stand in relations I take to be a plain Moorean fact beyond the reach of reasonable controversy. After all, my cat is black and he is sleeping next to my blue coffee cup.  ‘Black’ picks out a property, an extralinguistic feature of my cat.

Is that obvious?  Not to some.  Not to the ornery and recalcitrant critter known as the ostrich nominalist.  My cat, Max Black, is black.  That, surely, is a Moorean fact. Now consider the following biconditional and consider whether it too is a Moorean fact:

1. Max is black iff Max has the property of being black.

As I see it, there are three main ways of construing a biconditional such as (1):

A.  Ostrich Nominalism.  The right-hand side (RHS) says exactly what the left-hand side (LHS) says, but in a verbose and high-falutin' and dispensable way.  Thus the use of 'property' on the RHS does not commit one ontologically to properties beyond predicates.  (By definition, predicates are linguistic items while properties are extralinguistic and extramental.)  Predication is primitive and in need of no philosophical explanation.  On this approach, (1) is trivially true.  One needn't posit properties, and in consequence one needn't worry about the nature of property-possession. (Is Max related to his blackness, or does Max have his blackness quasi-mereologically  by having it as an ontological constituent of him?)

B. Ostrich Realism.  The RHS commits one ontologically to properties, but in no sense does the RHS serve to ground or explain the LHS.  On this approach, (1) is false if there are no properties.  For the ostrich realist, (1) is true, indeed necessarily true, but it is not the case that the LHS is true because the RHS is true.  Such notions as metaphysical grounding and philosophical explanation are foreign to the ostrich realist, but not in virtue of his being a realist, but  in virtue of his being an ostrich.

C. Non-Ostrich Realism.  On this approach, the RHS both commits one to properties, but also proffers a metaphysical ground of the truth of the LHS: the LHS is true because (ontologically or metaphysically speaking)  the concrete particular Max has the property of being black, and not vice versa.

Note 1: Explanation is asymmetrical; biconditionality is symmetrical.

Note 2: Properties needn't be universals.  They might be (abstract) particulars (unrepeatables) such as the tropes of D. C. Williams and Keith Campbell.  Properties must, however, be extralinguistic and extramental,  by definition.

Note 3: Property-possession needn't be understood in terms of instantiation or exemplification or Fregean 'falling-under'; it might be construed quasi-mereologically as constituency: a thing has a property by having it as a proper ontological part.

Against Ostrich Nominalism

OstrichOn (A) there are neither properties, nor do properties enter into any explanation of predication.  Predication is primitive and in need of no explanation.  In virtue of what does 'black' correctly apply to Max? In virtue of nothing.  It just applies to him and does so correctly.  Max is black, but there is no feature of reality that explains why 'black' is true of Max, or why 'Max is black' is true.  It is just true!  There is nothing in reality that serves as the ontological ground of this contingent truth.  Nothing 'makes' it true.  There are no truth-makers and no need for any.

I find ostrich nominalism preposterous.  'Black' is true of Max, 'white' is not, but there is no feature of reality, nothing in or at or about Max that explains why the one predicate is true of him and the other is not!?  This is not really an argument but more an expression of incomprehension or incredulity, an autobiographical comment, if you will.  I may just be petering out, pace Professor van Inwagen.

Can I do better than peter?  'Black' is a predicate of English.  Schwarz is a predicate of German.  If there are no properties,  then Max is black relative to English, schwarz relative to German, noir relative to French, and no one color.  But this is absurd.  Max is not three different colors, but one color, the color we use 'black' to pick out, and the Germans use schwarz to pick out. When Karl, Pierre, and I look at Max we see the same color.  So there is one color we both see — which would not be the case if there were no properties beyond predicates.  It is not as if I see the color black while Karl sees the color schwarz.  We see the same color.  And we see it at the cat.  This is not a visio intellectualis whereby we peer into some Platonic topos ouranos.  Therefore, there is something in, at, or about the cat, something extralinguistic, that grounds the correctness of the application of the predicate to the cat.

A related argument.  I say, 'Max is black.'  Karl says, Max ist schwarz.  'Is' and ist are token-distinct and type-distinct words of different languages.  If there is nothing in reality (no relation whether of instantiation or of constituency, non-relational tie, Bergmannian nexus, etc.) that the copula picks out, then it is only relative to German that Max ist schwarz, and only relative to English that Max is black.  But this is absurd.  There are not two different facts here but one.  Max is the same color for Karl and me, and his being black is the same fact for Karl and me.

Finally, 'Max is black' is true.  Is it true ex vi terminorum?  Of course not.  It is contingently true.  Is it just contingently true?  Of course not.  It is true because of the way extralinguistic reality is arranged. It is modally contingent, but also contingent upon the way the world is.  There's this cat that exists whether or not any language exists, and it is black whether or not any language exists.

Therefore, I say that for a predicate to be contingently true of an individual, (i) there must be individuals independently of language; (ii) there must be properties independently of language; and there must be facts or truth-making states of affairs independently of language.  Otherwise, you end up with (i) total linguistic idealism, which is absurd; or (ii) linguistic idealism about properties which is absurd; or (iii) a chaos, a world of disconnected particulars and properties.

The above is a shoot-from-the hip, bloggity-blog exposition of ideas that can be put more rigorously, but it seems to to me to show that ostrich nominalism and ostrich realism for that matter are untenable — and this despite the fact that a positive theory invoking facts has its own very serious problems.

Metaphilosophical Coda: If a theory has insurmountable problems, these problems are not removed by the fact that every other theory has problems.  For it might be that no theory is tenable,while the problem itself is genuine.

Correspondence and Truthmaking

I posed the following problem:

A. Some sentences are true in virtue of their correspondence with extralinguistic reality.

B. If so, then reality must have a sentence-like structure.

C. Reality does not have a sentence-like structure.

London Ed solves it by rejecting (A).  But let me first say why I accept (A).

Consider a true contingent sentence such as 'Tom is sad,' or the proposition expressed by an assertive utterance in appropriate circumstances of such a sentence.  I maintain that the sentence or proposition cannot just be true:  if true it is true in virtue of something external to the sentence. The external something cannot be another sentence, or, more generally, another truthbearer.  Nor can it be someone's say-so. So the external something has to be something 'in the world,' i.e., in the realm of primary reference, as opposed to the realm of sense. The basic idea here is that some truths need ontological grounds:  there is a deep connection between truth and being.  There is more to a true sentence than the sentence that is true.  There is that in the world which makes it true.  Call it the truthmaker of the truth.  Some truthbearers need truthmakers.  As far as I am concerned, this is about as clear as it gets in philosophy.  Which type of entity is best suited to play the truthmaker role, however, is a further question.

Please note three things.  First, the direction of the truthmaking relation is from the world to language.  More broadly: from external concrete reality to the realm of representations, where Fregean propositions count as representations, despite their not being tied to specific languages, and despite their independence of minds.   Second, correspondence is an umbrella notion that covers two quite different relations, naming, and making-true.  Naming is a word –> world relation, whereas truthmaking goes in the opposite direction.  I am tempted to say that truthmaking is the converse of naming.  Third, I unpack 'correspondence' as it occurs in (A) in terms of truthmaking, not naming.

Here is what Ed says in rejection of (A):

The exam question is my argument against (A), namely that some sentences are true in virtue of their correspondence with extralinguistic reality. I shall also be taking on why my reasons are properly nominalistic, given that your version of nominalism is not mine.

1. Starting with nominalism. Classic nominalism is formulated by Ockham in Summa Part I, 51. “the root [of the error of the Realists] is to multiply entities according to the multiplicity of terms, and [to suppose] that to every term [or expression] whatsoever there corresponds a thing [quid rei].”

2. My target is a formulation of the correspondence theory that violates classic nominalism, as I have defined it. There may be other formulations of the theory that are OK.

3. My formulation of the correspondence theory is that an assertoric sentence is true in virtue of naming or referring to or signifying a fact. Let that naming relation be R. Then the correspondence theory says that a sentence S (e.g. ‘Socrates is sitting’) is true iff S stands in the relation R to some fact F (e.g. ‘that Socrates is sitting’).

4. Suppose ‘Socrates is sitting’ names the fact that Socrates is sitting, and assume that it always so names. Then that fact must always exist, assuming the name is always names the fact. So ‘Socrates is sitting’ must always be true, i.e. ‘Socrates is sitting’ always stands in the relation R to the fact that Socrates is sitting. But it is not always true, clearly.

5. We might get out of this in two ways. First, by supposing that ‘Socrates is sitting’ fails to be meaningful, namely when the fact it purports to names ceases to exist, such as when Socrates stands up, or runs. This is absurd, however. The purpose of a sentence is always to mean something.

6. The other way is to suppose that the sentence sometimes names a fact, and sometimes does not. I.e. it actually names something else – a proposition – and the proposition is a fact when the sentence is true, otherwise not a fact. However we have now failed to explain the ‘correspondence’. The sentence ‘Socrates is sitting’ always bears the naming relation R to the proposition that Socrates is sitting, even when Socrates is not sitting.

7. What we really need to name is not the proposition (which may be true or false), but the reality that corresponds when the proposition is a fact. Perhaps ‘the proposition that Socrates is sitting being a fact’ or ‘the actuality of Socrates’s sitting’ or something like that. But there we have the same problem. Either the name ceases to be meaningful when Socrates is not sitting, or it continues to name something. But the former we agreed was absurd, and the latter means that we have not fully captured the relation we want.

8. The problem in general is that if the object of the relation R is something we can talk about i.e. name at all, then we have to deal with the problem of the fixity of reference. The purpose of a name is always to name what it names. But reality is not thus fixed. Whatever supposedly corresponds to the truth of ‘Socrates is sitting’ comes into existence when Socrates sits and goes out of existence when he stands up. But if ‘Socrates is sitting’ is true in virtue of naming this thing, either the sentence becomes meaningless when Socrates stands up, which is absurd, or it names something that does not go out of existence, and so does not name what the correspondence theory purports to name.

9. Bringing this back to nominalism. The problem above arises from the supposition that ‘Socrates is sitting’ is the name of some fact, and thus from supposing that every expression (‘Socrates is sitting’) has a name or referent or whatever.

Ed does two things above.  He confronts the truthmaker theorist with a certain (supposedly insoluble) problem, and then he explains how this problem arises by way of a false assumption.  First, the problem.  I will summarize it as I understand it.

Since Socrates is  a past individual, but nothing in this discussion has to do with time, I will change the example to 'Tom is red.'  Tom is a tomato of my present acquaintance.  We assume that the sentence is true.  And of course, if true, then contingently true.  My type of TM-theorist holds that contingent true predications such as 'Tom is red' have worldly correspondents called facts.  These concrete facts are the truthmakers of contingent predications. Note that the fact corresponds to the sentence as a whole.  So not only does 'Tom' have a worldly correspondent, and presumably also the predicate 'red'; the sentence has a worldly correspondent as well.

Note also that the sentence is not about the fact; it is about Tom, or, if you insist, it is about Tom and the property of being red.  Still, there is some relation R that connects the sentence and/or the proposition it expresses and the fact.  Notice, I wrote 'and the fact,' not 'to the fact.'  'To the fact' suggests a direction from language to world, and not vice versa, whereas 'and the fact' leaves the directionality open.  Is the truthmaking relation R naming?  Ed thinks it is, but this is not clear.  Indeed, I will argue in a moment that the truthmaking relation is not the naming relation.  It is clear that 'Tom' names Tom.  It is not clear that 'Tom is red' names anything.  Suppose it doesn't.  This doesn't exclude the possibility that the sentence has a truthmaker.  Maybe it has a truthmaker, but that truthmakers cannot be named.  Note also that what Ed says above is nothing like what any TM-theorist has ever said.  Truthmaking is a relation that runs from the world to representations, whereas naming and referring and 'signifying' run from representations to the world.  Truthmaking is more like the converse of the naming relation.  We shall see.

But let us suppose arguendo that the truthmaking relation R is naming.  On this supposition, Ed sets up a clever little dilemma.  It is based on three plausible theses.  

T1.  If N is a name, then N cannot be vacuous: it must have a nominatum or referent.

T2. If N is a name, then it has an existing referent.  That is, there is no naming of nonexistent objects, pace Meinong.

T3. If a name N names an object O, then at every time at which N names something, it names O.  So the following is impossible: at some times at which 'Kripke' is in use as a name, it names Kripke, at other times Shkripke.  I think this is what Ed means by "the fixity of reference."

The Dilemma.  Either sentence S names fact F or it doesn't.  On either alternative, trouble.  Remember, Ed is assuming that the truthmaking relation is a naming relation and that declarative sentences name facts.

Horn One.  If S names F, then, by the conjunction of the three plausible theses,  F exists at every time at which S exists, which is plainly false.  Clearly, 'Tom is red' both as type and as token can exist at times at which the fact of Tom's being red does not exist.  (I might assertively utter 'Tom is red' while Tom is green, or after Tom has been dunked into molten chocolate.)  If you say instead that S is meaningless when the fact does not exist, then truthmaking is not naming  (by T1), which is all it can be on Ed's (mis)understanding of truthmaking.

Horn Two.  If S does not name F, then there is no truthmaking.  For truthmaking is a naming relation.

Critique

It is clear that Ed does not understand truthmaker theory.  The key idea is not that sentences name facts, but that facts make sentences true.  That truthmaking is different from naming is clear from the different directions of the relations, but also because truthmaking is a many-many relation whereas naming is a many-one relation.  That truthmaking is many-many can be seen as follows.  One and the same truth can have many different  truthmakers.  For example, 'Something is red' is made true by a's being red, b's being red, c's being red, etc.  And one and the same truthmaker can make true many truths.  For example, Tom's being red makes true 'Tom is red,' 'Tom is red or Shlomo is sad,' etc.  (Cf. Armstrong 1997, pp. 129-130.)

Nominalism

Ed has an understanding of nominalism which contemporary analytic philosophers will find idiosyncratic and vacuous to boot.  No philosopher today thinks that for every bit of language there is a corresponding bit of reality.  So we are all nominalists in this vacuous sense. And no one is a realist if  “the root [of the error of the Realists] is to multiply entities according to the multiplicity of terms, and [to suppose] that to every term [or expression] whatsoever there corresponds a thing [quid rei].”  And surely it is a bad joke to claim or suggest that TM-theorists straightaway infer the existence of facts from the existence of declarative sentences. 

Related articles
Tropes as Truth-Makers? Or Do We Need Facts?
Does Reality Have a Sentence-Like Structure?
The Truthmakers of Truths About Truths
Plantinga Reviews Philip Kitcher, Life After Faith: The Case for Secular Humanism
Working Draft: The Case Against Facts
Peter van Inwagen, "A Theory of Properties," Exposition and Critique

The Truthmakers of Truths About Truths

Josh writes,

I would be interested to see how you respond to the following dilemma (from Peter Geach, "Truth and God," Proceedings of the Aristotelian Society, [1982]: 84).

Say proposition P1 is true because it corresponds to fact F. Does the proposition "Proposition P1 is true" (call it proposition P2) have a truthmaker? It seems that it should. Not only that, it seems that the truthmaker of P2 should be the same as P1 (i.e. F). But it's not obvious how F could make P2 true, since it is not obvious that F shares P2's "propositional" or "language-like structure," as you put it.

You've already said that some propositions do not have truthmakers, so perhaps you could just deny that P2 has a truthmaker. Or perhaps there is a way that F could do the job of truthmaking with respect to P2? Or perhaps P2 could be analyzed in a way that shows it is not really different from P1?

Thanks for your high-quality blogging!

You're very welcome!  Interesting puzzle. It seems obvious that P2 has a truthmaker and that it has the same truthmaker as P1.  Note also that if P1 is contingent, then P2 will also be contingent.  For example, 

Tom is sad

and 

'Tom is sad' is true

are both contingently true and have the same truthmaker, namely,  the contingent fact of

Tom's being sad.

And the same holds for all further iterations such as 

"'Tom is sad' is true" is true.

Iteration of the truth predicate preserves the modal status of the base proposition. The regress here is infinite but benign.  Whatever makes the base proposition true makes true every member of the infinite series of truth predications.

Now the problem you raise is that, while there is a clear isomorphism between 'Tom is sad' and Tom's being sad, there is not the same isomorphism between "'Tom is sad' is true" and Tom's being sad.  The predicate in P2 is the predicate 'true', not the predicate 'sad.'  P1 is about a man and says of him that he is sad; P2 is about a proposition and says of it that it is true.  You are making an assumption, perhaps this:

A. If two or more propositions have the same truthmaker, then they must predicate the same properties of the same subjects.

The truthmaker theorist, however, is not committed to (A).  The singular 'Tom is sad' and the existentially general 'Someone is sad' have the same truthmaker, namely, Tom's being sad, but the two propositions differ in logical form, and the second is not about what the first is about. The singular proposition is about Tom while the general proposition is not.

My point, then, is that the puzzle arises only if we assume (A).  But (A) is no part of truthmaker theory.  Truthmaking is not a 1-1 correspondence.  'Someone is sad' has many different truthmakers, and Tom's being sad makes true many different propositions, indeed, infinitely many. 

Related articles
In a Philosophical Conversation, Three's a Crowd
The Delight of the 'Find'
Does Reality Have a Sentence-Like Structure?
Working Draft: The Case Against Facts
Incarnation Approached Subjectively: The Mystical Birth of God in the Soul
Theism Meets Metaphysical Naturalism
Wittgenstein: On the Fritz

Does Reality Have a Sentence-Like Structure?

 Our problem may be formulated as an antilogism, or aporetic triad:

A. Some sentences are true in virtue of their correspondence with extralinguistic reality.

B. If so, then reality must have a sentence-like structure.

C. Reality does not have a sentence-like structure.

This trio of propositions is inconsistent. And yet one can make a plausible case for each member of the trio.

Ad (A).  Consider a true contingent sentence such as 'Tom is sad,' or the proposition expressed by an assertive utterance in appropriate circumstances of such a sentence.  Surely, or rather arguably, the sentence or proposition cannot just be true:  if true it is true in virtue of something external to the sentence. I should say that I reject all deflationary theories of truth, including  Ramsey's redundancy theory, Quine's disquotationalism, and Paul Horwich's minimalism. The external something cannot be another sentence, or, more generally, another truthbearer.  Nor can it be someone's say-so: no truth by fiat unless your name is YHWH. So the external something has to be something 'in the world,' i.e., in the realm of primary reference, as opposed to the realm of sense, to invoke a Fregean distinction. The basic idea here is that some truths need ontological grounds:  there is a deep connection between truth and being.  There is more to a true sentence than the sentence that is true.  There is that in the world which makes it true.  Call it the truthmaker of the truth.  Some truthbearers need truthmakers.  As far as I am concerned, this is about as clear as it gets in philosophy.  Which type of entity is best suited to play the truthmaker role, however, is a further question.

Ad (B).  At a bare minimum, external reality must include Tom, the subject of our sentence.  Part of what must exist for 'Tom is sad' to be true is Tom himself.  But Tom alone does not suffice since the sentence says, and says truly, that Tom is sad.  So it would seem that external reality must also include properties including the property of being sad.  How could something be F if there is no F-ness in the world?  There are of course extreme nominalists who deny that there are properties.  I consign these extremists to the outer darkness where there is much wailing and the gnashing of teeth.  Theirs is a lunatic position barely worth discussing.  It is a datum that there are properties. One cannot reasonably ask whether they are; the only reasonable question is what they are.   Moderate nominalism, however, is a respectable position.  The moderate nominalist admits properties, but denies that they are universals.  In contemporary jargon, the moderate nominalist holds that properties are tropes.  A trope is a property assayed as a particular, as an unrepeatable item. Accordingly, the sadness in Tom is not repeated elsewhere: it is unique to him. Nor is it transferable: it cannot migrate to some other concrete particular.  I'll 'turn' back to tropes in a 'moment.'  (Get the double pun?)

For now suppose properties are immanent universals and that reality includes Tom and the property of being sad.  Could the sum Tom + sadness suffice as the ontological ground of the truth of 'Tom is sad'?  I will argue that it cannot.  A universal is a repeatable entity.  Universals are either transcendent or immanent. An immanent universal is one that cannot exist unless instantiated.  A transcendent universal is one that can.  Suppose sadness is an immanent universal instantiated by Shlomo.  Then sadness exists and Tom exists.  But the mere(ological) sum of the two does not suffice to make true 'Tom is sad.'  For if the property and the particular each exist, it does not follow that the particular has the property.  A tertium quid is required: something that ties the property to the particular, sadness to Tom.  

What this suggests is that the truthmaker of a contingent predication of the form a is F must be something that corresponds to the sentence or proposition as a whole. It cannot be a by itself, or F-ness by itself; it must be a's being F.  It is the BEING F of Tom that needs accounting.  You could call this the problem of copulative Being.

Enter facts or states of affairs.  (These are roughly the states of affairs of Armstrong's middle period.) We now have the concrete particular Tom, the property sadness, and the fact of Tom's being sad.  This third thing brings together the concrete particular and the property to form a truthmaking fact.  Now this fact, though not a proposition or a sentence, is obviously proposition-like or sentence-like.  Although it is a truthmaker, not a truthbearer,  it is isomorphic with the truthbearer it makes true.  Its structure is mirrored in the proposition.  It is a unity of constituents that is not a mere mereological sum of parts any more than  a sentence-in-use or a proposition is a mere mereological sum of parts.  Plato was already in possession of the insight that a declarative sentence is not a list of words.  'Tom is sad' is not the list: 'Tom,' 'sad,' or the list: 'Tom,' 'is,' 'sad.'

This  argument to facts as worldly items in addition to their constituents requires the assumption that properties are universals.  For this assumption is what makes it possible for the sum Tom + sadness to exist without Tom being sad.  To resist this argument for the sentence-like structure of external reality, therefore, one might try insisting that properties are not universals.  And here we come to Arianna Betti's proposal which I have discussed in painful detail  in a draft the final version of which will soon appear in the journal METAPHYSICA.  She suggests that properties are bearer-specific and that relations are relata-specific.  

Well, suppose sadness is bearer-specific, or more precisely, bearer-individuated.  This means that it cannot exist unless its bearer, Tom, exists.  We can depict the property as follows:  ____(tom)Sadness.  Tom can exist without this property because it is contingent that Tom is sad.  But the property cannot exist or be instantiated without Tom.   On this scheme there cannot be a difference between the sum Tom + ___(tom)Sadness and the fact of Tom's being sad.  Given the particular and the property, the fact 'automatically' exists.  Betti takes this to show that some mereological sums can serve as truthmakers.  But, as she notes, the bearer-specific property by itself can serve as truthmaker.  For if ___(tom)Sadness exists, it follows that 'Tom is sad' is true.  This is because it cannot exist without being insdtantiated, and because it is the "nature" (Betti's word) of this property to be of Tom and Tom alone.  So if it exists, then it is instantiated by Tom, by Tom alone, and without the services of a tertium quid.

Now the point I want to make is that whether we take properties to be universals or tropes, it seems we have to grant that reality has a proposition-like structure.  Either way it has a proposition-like structure.  We saw how this works if properties are universals.  The mereological sum Tom + the universal sadness does not suffice as truthmaker for 'Tom is sad.'  So we need the fact of Tom's being sad.  But this fact has a proposition-like structure.  To avoid Armstrongian facts, Betti suggests that we construe properties as monadic tropes.  But these too have a proposition-like structure. Even if Betti has shown a way to avoid Armstrong's middle period facts or states of affairs, she has not shown that the world is just a collection of things bare of proposition-like or sentence-like structure.

How so?  Well, ___(tom)Sadness obviously in some sense involves Tom, if not as a constituent, then in some other way.  There has to be something about this property that makes it such that if it is instantiated, it is instantiated by Tom and Tom alone. It is very much like a Fregean proposition about Tom.  Such a proposition does not have Tom himself, with skin and hair, as a constituent, but some appropriately abstract representative of him, his individual essence, say, or his Plantingian haecceity.  

Ad (C).  According to the third limb of our triad reality does not have a sentence-like structure.  This will strike many as obvious.  Are worldly items syntactically related to one another?  Do this make any sense at all?  Arianna Betti, Against Facts, MIT Press, 2015, p. 26, italics in original:

Only linguistic entities . . . can strictly speaking have syntax.  Facts are neither linguistic nor languagelike, because they are that of which the world is made, and the world is not made of linguistic or languagelike entities at the lowest level of reference.  Thus the articulation of a fact cannot be logical in the sense of being syntactical.  It is a categorical mismatch to say that there is a syntactical articulation between a lizard and light green or an alto sax and its price. 

So how do we solve this bad boy?  I say we reject (C).

In the beginning was the Logos and the Logos was with God, and the Logos was God, and  the Logos ex-pressed itself LOG-ically as the world. 

 

 
 
 
Nothing is Written in Stone
Language and Reality
Working Draft: The Case Against Facts
Visual and Propositional Contents of That-Clauses: An Aporetic Hexad
Tropes as Truth-Makers? Or Do We Need Facts?
Related articles