« Best to be Neither Poor nor Rich | Main | Saturday Night at the Oldies: Marie/Maria/Mary »

Saturday, November 05, 2011


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

It's hard for me to say this intelligibly, but:

Perhaps there is a difference between meta-facts (facts about facts) which CAN be said intelligibly, and the way that a fact about the world corresponds or represents (if true) the world itself, which cannot be fully contained in the descriptions of it in our words. But our talk, our meta-facts about facts, since they are words about words, can contain their abstractions.


You have outlined three arguments against facts:

1) The Strawsonian argument which claims that facts are mere “hypostatized sentences”. You summarize this argument as follows: “After all, to distinguish a fact from a non-fact (whether a particular or a universal) we must have recourse to a sentence: 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.”

I find this argument unsatisfactory. We cannot *distinguish* objects either except by recourse to some form of language. Does the anti-realist about facts wishes to extend this obvious point to all objects and maintain that they too are merely “hypostatized phrases”? We cannot *describe* the world without using language. Does it follow that the world is a linguistic entity or a “hypostatization” of the use of true descriptions? Pursuing this line, the anti-realist about facts might just as well maintain that only languages exist. But, of course, these conclusions are absurd.

This anti-realist argument seems to me to be based on confusion and a systematic misunderstanding of the realist project. The realist argues that the world exists independently of any language or users thereof; do anti-realists about facts deny this? The realist also maintains that some uses of language are true about this world; do anti-realists about facts deny this? Then the realist asks: What do we need in order to explicate the fact that some declarative sentences are true about this language-independent world? The realist then attempts to give an answer by using, what else, language.

The anti-realist’s objection seems to be that since the realist must use language in order to state his theory of this relationship, he is “trapped” in the web of language and can never escape it. But this argument is confusing the use of language to talk about the world and the use of language to theorize about language and its relationship to the world. This confusion is at the bottom of the anti-realist objection. Once the confusion is exposed, the force of the anti-realist objection dissolves; at least so I claim.

2) The Unperceivability Argument: “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.”

Merely because we do not, and perhaps even cannot, perceive certain kinds of entities it hardly follows that they do not exist. I won’t recite the obvious counter-examples to such outlandish claims. Instead let us focus on whether we do “see” relational facts or not. Consider this: I judge that Tom is heavier than John based on a casual observation of the two men. Suppose I have no clue as to the weight of each. I simply look at them and “see” that one is heavier (fatter, taller, shorter, etc.,) than the other. According to the anti-realist’s unperceivability-argument this is impossible, for I cannot see a relational fact. But surely my judgment is not based upon seeing first the weight of one, the other, and then comparing the two “seeings” in my mind. Why not? Because what is it that I am comparing according to this account? It cannot be some measure of the weight of each represented by some numerical value since ex-hypothesis I do not know any such value. Moreover, if I have to represent the weight of each in my mind independently from the other and then go through a process of comparison, then the question arises how do I perform this mental feat, if I cannot perform the same feat regarding my perception of the men themselves? If my judgment is not based upon perceiving that one is heavier than the other, then my judgment cannot be based upon a comparison of their mental representations through introspection either. For why should introspection achieve something that perception cannot already achieve?

3) The Unity Argument: “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 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.”

In some sense the unity argument may be the only argument that I find to have a real bite. But, upon reflection I must ask whether it is ultimately conclusive. You seem to suggest that the problem of the unity of facts either leads to a vicious infinite regress or to idealism. Neither is palatable. The former because of obvious reasons, whereas the later because it defies the very motivation to posit facts; namely, realism. If that were correct, then the realist about facts would be in real trouble. But is he?

Consider the second option you list above and the argument against it: i.e., the option that a fact is “an entity in addition to its constituents.” But why can’t a fact be an entity over and above its constituents and at the same time depend upon them so that “apart from them it (i.e., the fact; PL) is nothing.”? Let me explain. A house is made out of bricks; so apart from the bricks, *this* house is nothing. But at the same time this house is not numerically identical to a pile of the bricks it is made out of: hence, it is something over and above them. Thus, a house is an emergent entity that depends upon the bricks and the manner in which they are organized.

I suggest that facts are emergent entities from their constituents and the manner in which these constituents are organized? But now you will inevitably ask: what does the organizing? Don’t we need something that organizes the constituents in just the manner they are arranged so as to make out this particular fact rather than some other fact? My answer is: sure we do? But we already have this organizing something in the relational universals that enter into the fact and without which no facts can exist. I suggest that relational universals are the sort of entities that essentially unite constituents that instantiate them. A fact just is the result of the unification by a universal of those objects that instantiate it. Thus, in addition to other reasons, universals are needed in order to unify constituents into a whole that is a fact. This solution stops the threatening vicious regress and at the same time avoids the horrendous pitfalls of idealism.


Thanks for the detailed and perceptive comments. As you know, I am not quite endorsing the arguments against facts since I accept the latter myself. But I don't want to paper over the difficulties that the positing of facts involves. I have time now only for a comment on the first argument.

You raise the important question whether anti-realism with respect to facts can be 'contained' or whether it must inevitably by its own logic spread to other ontological categories. This is a question that we should put to the Londonistas, Ed and David.I have already expressed my suspicion that Ed will end up a lingusitic idealist.

A. To have a theory about facts we must nominalize sentences. When asked what I mean by a truth-making fact, I say an entity like Tom's-being-seated. But this involves transforming 'Tom is seated' into 'Tom's being seated.' The latter, however, is not a sentence, but a noun phrase. And that suggests to Strawson and the boys an illicit hypostatization.

B. To have a theory about properties we must nominalize predicates. When asked what I mean by a property, I say an entity like being-seated. But this involves transforming 'is seated' into 'being-seated' or 'seatedness.' The latter, however, is not a predicate but a name.

I take it that what you are getting at is that if the (A) consideration is a reason to deny facts, then the (B) consideration is a reason to deny properties (whether these be construed as universals or as particulars, e.g., tropes).

And so we push our nominalist pals into denying properties. But then what are they saying? That the world in itself is just a bunch of naked particulars (I avoid 'bare particular' which has a technical meaning)?

If we can force them to deny properties, then we can force them to deny ordinary particulars like tomatoes. For things like tomatoes are obviously propertied, not 'naked.'


I of course know very well that you do not endorse these arguments, although I think you do suspect that at least some of them have sufficient weight to render the problem of facts a real philosophical problem. Hence, the metaphilosophical position at the end of your post.

Your response on behalf of the anti-realists is part of what I meant in my post; but I thought that the anti-realists had a broader argument against facts.

The objection you address that appeals to nominalization of sentence or predicates raises in my mind the following question: What nominalization has to do with the (non)existence of facts or properties? What justifies the inference from the observation that *one* way of referring to facts is by means of nominalized sentences to the conclusion that facts are mere hypostatized and shadowy entities? I simply fail to see the basis for such an inference.

Secondly, nominalization is merely one grammatical construction that enables us to refer to facts. Another is the construction 'It is a fact that...' where in the blank we insert a true sentence such as 'John is sitting'. Moreover, I am very skeptical of nominalization as a reliable criterion of the ontological commitments of a theory. We also have Quine's well-known criterion of ontological commitment which does not rely upon nominalization at all. And there are other criteria better suited.

However, the broader argument which anti-realists frequently invoke is that facts and properties are language dependent because we cannot access them except by means of language. But, of course, we cannot access, talk about, and describe individuals either except by means of language. I think we agree that this is a bad argument, for it entails that merely talking about or describing something turns it into a language dependent entity or no entity at all.

>>And that suggests to Strawson and the boys an illicit hypostatization

I assume "the boys" are the Londonistas.

That's not quite right. Rather, Londonistas (some of them anyway) think that nominalisation throws away a tiny part of the reality. We don't necessarily object to 'Tom's being seated" referring to some sort of thing. Not in this part of the fight, anyway. Rather, if it were a thing, it would have left something out, namely the part contributed by the verb. A bit of 'active ingredient' if you like.

And they say that what you can't talk about, you must pass over in silence. For you can never get to that 'thing'. If you could, you would have to refer to it by a noun phrase, and thus miss out again on the 'active ingredient' of reality.

You're right, Ed. The charge is not merely one of hypostatization, but that the entities introduced -- facts as truthmakers -- cannot do the job they are supposed to do. What we need in the fact is something that corresponds to the verb -- or the verbal element of the verb -- in a sentence in use. For example, in 'Tom is sitting' there must be something in the fact that corresponds to the 'is.' Call this the copulative element. But it must be a COPULATING copula! (Shades of the Russell-Bradley debate and relating relations as opposed to relations as inert ingredients.) The 'active ingredient',' however, does not survive nominalization/objectification.

But it doesn't follow that we must pass over the
'active ingredient' is silence.


As my exchange with Ed shows, the main isseue is not hypostatization but the issue of whether introducing further entities -- facts -- over and above individuals and properties really gets us anywhere. See above.

>>Another is the construction 'It is a fact that...' where in the blank we insert a true sentence such as 'John is sitting'.<<

But surely this does not involve reference to facts in the relevant sense of 'fact.' See http://maverickphilosopher.typepad.com/maverick_philosopher/2010/01/three-senses-of-fact.html

We should distinguish between facts-that and facts-of. A fact-that is just a true proposition or sentence. You will have noticed that when I am speaking of truth-making facts I am always careful to write: the fact of Tom's being tired, not the fact that Tom is tired. The fact that Tom is tired cannot be a truth-maker; it is itself rather in need of a truth-maker.

'It is a fact that John is sitting' could be accepted by Ed with equanimity. It says no more than 'It is true that John is sitting.'

Ed no doubt grants that there are true sentences; what he does not grant is that any true sentence has or needs a truth-maker.

At this point I show Ed the instruments of torture and demand an answer to this question: Is there anything of a nonlinguistic nature that is involved in the truth of 'John is sitting'? If yes, then what? If not, then don't you admit your position is crazy?

>>At this point I show Ed the instruments of torture and demand an answer to this question: Is there anything of a nonlinguistic nature that is involved in the truth of 'John is sitting'? If yes, then what? If not, then don't you admit your position is crazy?

Ouch. The problem is your 'anything'. If it were such 'things', we would be able to quantify over them, which is precisely what the nominalist denies. When we substitute into a quantified term, we substitute a referring phrase. This is not a verb phrase, hence cannot express the reality that we want to get hold of.

How about this: a proper sentence like 'Caesar stands' expresses more than than the verbal noun 'Caesar-standing'. It says that Caesar-standing is a fact, or that it exists in reality. So we can express this 'more' by means of a verb, and I absolutely agree that what is expressed or signified is non-linguistic. What could be more non-linguistic than saying that something is a fact, that it obtains in reality? No linguistic idealism at all.

What I do deny is that we can name or refer to this extra component using a noun phrase or referring phrase. Only a verb can do it.

Thank you, Bill. I am very well.

This is a fascinating discussion. The more I think about it the more it seems to make contact with other topics you have covered.

Here is a 'sorites' or 'topological' argument against truthmakers as presently conceived: Intuitively the truth value of 'Al is fat' depends continuously on Al's weight, say. Suppose 'Al is fat' is made true by Al's fatness. The existence of Al's fatness is an all-or-nothing business---it's either fully present or completely absent---and hence cannot be continuously dependent on Al's weight, since there can be no continuous function from the reals onto {0,1}. The truth of 'Al is fat' cannot depend on the existence of some discrete thing.

It would seem that 'Al is fat' is an extremely coarse approximation to reality---it conveys one bit of information---but it is not entirely disconnected from reality. Rather it's a judgement we make which may (presumably is) useful in certain circumstances. A less coarse approximation might be to describe the distribution of Al's adipose tissue---so many pounds here, so many there. But any rule which derived the truth of 'Al is fat' from this distribution must, by the same argument, have some arbitrary or conventional element. And in making this move we have changed the subject. The next move (to molecules perhaps) will change the subject again. Can the process converge?

It rather looks as if this Londinista denies some properties, or at least relocates them elsewhere. Shades of the reductionsism/eliminativism discussion.

I'm not sure I follow the latest turn though. The suggestion is that a nominalisation necessarily loses some verbal element. What if we nominalise to a process rather than to an object? A blink, say, or a fattening, or even a life?


We both reject linguistic idealism. We both agree that there is an extralinguistic reality and that a sentence such as 'Caesar stands,' if true, is true because of this extralingusitic reality. But if you agree with all this, then why not say that 'Caesar stands' has a truth-maker?

What exactly is our dispute? I think it is as follows.

I say that some sentences have truth-makers and that these truth-makers, though not themselves propositions, must be proposition-like: they must be facts. You say that no sentence has a truth-maker.

What if I said that there are truth-making facts, but they cannot be named; they can only be referred to by using a sentence. Would you buy that? So I cannot access a fact by means of the noun phrase 'Peter's being seated'; but I can access it via 'Peter is seated.'


Your argument seems to be this:

1. If there are facts, then the fact of Al's being fat either exists or does not exist: it does not admit of degrees.
2. But surely Al is more or less fat which we can express by saying that Al's being fat does admit of degrees.
3. There are no facts.

But consider 'Al is a native Californian.' If true, this is contingently true. 'Native Califonian' is not a vague term nor does it admit of degress. Either a person was born in California or he was not.

So at best your argument disposes only of truth-makers for sentences like 'Al is fat.' It does not dispose of truthmakers for all true contingent affirmative singular sentences.

But also: why couldn't I say that there are continuum-many truth-makers, one for each degree of fatness?

David writes,

>>I'm not sure I follow the latest turn though. The suggestion is that a nominalisation necessarily loses some verbal element. What if we nominalise to a process rather than to an object? A blink, say, or a fattening, or even a life?<<

Right. Ed seems to think that whatever we refer to using a noun phrase must be some static object or 'thing' in his jargon. 'Tom's running,' though it is not a sentence but a name, does seem to pick out an activity or a process.


Doesn't Frege quantify over unsaturated entities without thereby transformning them into objects (saturated entities)?

And doesn't Frege implicitly if not explicitly distinguish between naming and referring? Predicates refer to concepts (which are unsaturated) but surely they do not name concepts. Only objects can be named.

We agree that sentences are not names. So no sentence names a fact. But if we distinguish between naming and referring we can say that sentences refer to facts. Why not?

As usual Bill puts my argument better than I can myself!

I guess I could reply that a birth is a process distributed in time and space with vague boundaries. It could take place around midnight and about the state line, idealising the latter conventional legal term as a geometric line. Behind all this is the thought that words are necessarily Procrustean---the detail in the world must be squeezed and chopped and coerced to fit the limited carrying capacity of sentences. Hence 'approximation', though this doesn't convey quite what I am trying to say. Can we find a CASS that does not suffer from this effect?

We could indeed say that there is a truth-maker for each continuum-many degree of fatness, but this may be going to the opposite pole of finding more information in the world than there is. Standard measure theory would give each of these possibilities a probability of zero. And where does this leave the property 'fat'? It still contains an element of arbitrariness.


There are degrees of being born given that the neonate's emerging from its mother admits of degrees. But consider the property of having been born. Surely this property does not admit of degrees. This makes no sense: 'Sam and Dave were both born, but Sam is more born than Dave.' Compare: 'Sam and Dave both are bald, but Sam is more bald than Dave.' The latter does make sense.

So consider this CASS: 'Tom was born.' One can imagine a situation in which that would actually be said. 'Tom, unlike his older brother Tim who was aborted, was born.'


David suggests that it is the process that makes the sentence true. I reply, the sentence is true only if the process actually takes place.

David may then suggest that it is the taking place of the process that makes the sentence true. I reply, the sentence is true only if the taking place of the process actually happens.

David may then suggest that it is the happening of the taking place of the process that makes the sentence true. I reply, that the sentence is true only if the happening of the taking place of the process is a reality.

David may then suggest that it is the being a reality of the happening of the taking place of the process that makes the sentence true. I reply, that the sentence is true only if the being a reality of the happening of the taking place of the process actually exists.

Is it the running out of words that ends this regress?

On sentences 'referring' to facts, there’s another argument of Frege’s this brings to mind, namely when he talks about the reference of a ‘that’-clause, e.g. ‘that the sun is shining’ and the reference of a whole sentence e.g. ‘it is true that the sun is shining’ or just ‘the sun is shining’. He says we can grasp the reference of the that-clause, and make a question out of it, e.g. ask whether the sun is shining. I.e. we can grasp the reference of the ‘that’ clause without grasping whether the corresponding sentence is true. But he says that the sentence itself cannot have a reference, for then its nature would include the predicate ‘true’ or ‘false’, and then, simply by knowing what was referred to, you would know whether the sentence was true or false.

In other words, whatever a truthmaker was, we could never know one pro isto statu.

Hi Ed,
I thought you would say that, and I agree. The verb-like component in a verbal noun is not assertive, right? But isn't the assertive component we need built in to our working truthmaker concept back at the beginning? Bill says that a truthmaker for p is a thing t such that 't exists' implies p. The verb-like, assertive element ('exists') is added explicitly to the noun-like element in the definition. That's how a truthmaker can be noun-like. Nobody is saying that a bare noun functions as a proposition, which, if you'll forgive me, seems to be the position you are opposing, eg, here

Hi Bill,
In a way your example illustrates what I am trying to say. The reality grounding 'Tom was born' is surely biochemical and physiological interactions of enormous complexity proceeding in the matter comprising the bodies of Tom and his mother. That is what a birth is. And yet all this is squeezed down to just three words. On the other hand, little of this is directly accessible to an observer who sees and feels fleshy surfaces, hears cries and gurgles. Perhaps these sensory phenomena are better candidates for truthmakers. Either way a lossy compression or many-one map is taking place. The move from 'Al is fat'/'Tom was born' to 'Al's fatness'/'Tom's birth', where the latter pairings are seen as elements of reality, is analogous to reversing a lossy compression or inverting a many-one map, and is consequently ill-defined. Perhaps this is just another way of putting Strawson's objection.

>>Bill says that a truthmaker for p is a thing t such that 't exists' implies p.

Yes but then we are right back to the question of what is the truthmaker of 't exists'. Is it t? Or the existence of t?

Ed writes,

>>But he says that the sentence itself cannot have a reference, for then its nature would include the predicate ‘true’ or ‘false’, and then, simply by knowing what was referred to, you would know whether the sentence was true or false.<<

You are confusing sense and reference. Frege, as you must know, maintains that sentences have reference: they refer to the True and the False. His point is that truth and falsity cannot be properties included in the sense of sentences.


I agree with you as against Ed. The truth-maker has a built-in assertive/unifying component. So we could say that a truth-maker is both noun-like and verb-like. It is an entity or thing in Ed's plain English. But it doesn't just 'sit there': it does something, namely, grounds the truth of the sentential representation. So it has a verb-like character. The unity of the sentence whereby the sentence is more than a list is grounded in the unity of the truth-making fact.

Now in my book I argue that the truth-making fact cannot unify its constituents by a further internal unifying constituent, but needs an external unifier -- but this takes us beyond our present discussion


The T-maker move, I would say, is from 'Al is fat' to Al's being fat not to Al's fatness. For 'Al's fatness' could be taken to pick out a trope or particularized property which is something like an Aristotelian accident. A trope is not a fact. Both are particulars in that they are not instantiable, but tropes are simples while facts are complexes. The fact of Al's being fat has Al, fatness, and an exemplification nexus as constituents. But the trope Al's fatness does not contain Al or any exemplification nexus.

A trope could be a constituent of a fact. Suppose, to appease Ed, we deny universals. Suppose properties are particulars, tropes. The fact of Al's being fat would then have as constituents: Al (all 200 lbs of him), a fatness trope, and an exemplification nexus, at a bare minimum.

My point is in part terminological. As against Peter above I said that we shouldn't speak of the fact THAT Al is fat because there is no difference between a fact that and a true proposition. Similalrly we shouldn't use 'Al's fatness' because that doesn't tell us whether we are speaking of a trope or a fact.

David sez: >>Bill says that a truthmaker for p is a thing t such that 't exists' implies p.<<

Ed replies: >Yes but then we are right back to the question of what is the truthmaker of 't exists'. Is it t? Or the existence of t?<

I thought I answered that question, Ed. The truthmaker of 't exists' is t. Thus the truth-maker of 'Al's-being-fat exists' is Al's-being-fat. Tell me what exactly is wrong with this answer!

You have not made good on your regress threat, and I don't see that Frege can help you. You must know that the mature Frege eschews t-making facts. On the side of Bedeutung he has das Wahre und das Falsche. No Sachverhalten.


Spell this out for me, please:

>>The move from 'Al is fat'/'Tom was born' to 'Al's fatness'/'Tom's birth', where the latter pairings are seen as elements of reality, is analogous to reversing a lossy compression or inverting a many-one map, and is consequently ill-defined. Perhaps this is just another way of putting Strawson's objection.<<

>>I thought I answered that question, Ed. The truthmaker of 't exists' is t. Thus the truth-maker of 'Al's-being-fat exists' is Al's-being-fat. Tell me what exactly is wrong with this answer!

Well you did answer it, but with no convincing arguments against my 'symmetry' argument. And it looks very suspicious to me. I don't have any further arguments on this today, except to note that when I something looks suspicious to me, it generally is suspicious.

Please also note my latest blog on Frege and the (impossible) thought whose being consists in being true. And note the brief remark I made in an earlier post, and which I shall develop later, about truthmakers for future tense statements. Briefly, is the truthmaker for 'it will rain tomorrow' something that exists today, or tomorrow, assuming that it does rain tomorrow? If it only exists tomorrow, we have to explain what makes it true that it comes into existence. It can't be the truthmaker on its own. But if it exists today, the future seems to be determined already.

>>when I something looks suspicious to me, it generally is suspicious.<<

So you have a power the rest of us lack?

You say that there are no truth-makers at all. I say that there are some truth-makers. That is the fundamental issue between us, is it not?

Whether or not future-tensed sentences have T-makers is not germane to the fundamental dispute. I am not a truth-maker maximalist. See http://maverickphilosopher.typepad.com/maverick_philosopher/2010/03/truthmaker-maximalism-questioned.html

Butchvarov on Semi-Realism about Facts

The condition for Al's fatness to be the truthmaker of 'Al is fat' is that for every way the world could be if Al's fatness exists then Al is fat. What can we mean by 'Al's fatness exists'? Presumably it means that there is some local feature in the distribution of Al's matter---he has a pot belly say. The noun is suggestive of something we might point to. Now suppose that the adjectival 'Al is fat' rests on some global aspect of Al---his body-mass index exceeding some threshold, say, which we can't point to. It seems quite plausible that of all the ways the world could be there is at least one in which Al's fatness exists but he is not fat (maybe he is rather tall). So Al's fatness cannot be the required truthmaker. The key here, which I've been groping towards in earlier comments, is that if 'having fatness' is not 'locked' to 'being fat' by virtue of being a Strawsonian shadow, but rather has some independent meaning, then the two meanings can come apart sufficiently to defeat truthmaking as we are currently thinking of it. Connected with this is that different sentences throw away different information---they are grossly many-one in different ways. If we could usefully say Al is exactly thus: and give some vast exact description, the problem would not arise.

Hello Bill,

I have thought some more about your 04:55 reply to me. I agree that 'Al's fatness' is not equivalent to 'Al's being fat'. The former has an object-like character that the latter lacks. We could joke that Al's fatness always precedes him but not that his being fat precedes him. However, I contend that 'that Al is fat' is equivalent to 'Al's being fat'. I can't find a context in which replacing one with the other alters the sense. Examples:

a) I see that Al is fat = I see Al's being fat
b) I know that Al is fat = I know (of) Al's being fat
c) that Al is fat worries his wife = Al's being fat worries his wife
d) that Al is fat exacerbates his heart condition = Al's being fat exacerbates his heart condition
e) that Al is fat is obvious = Al's being fat is obvious
f) that Al is fat is a fact = Al's being fat is a fact
g) it is a fact that Al is fat = it is a fact Al's being fat?

The RHS of (b) is perhaps a little awkward. I conclude that 'Al's being fat obtains/holds/exists' <---> 'that Al is fat is true' <---> 'Al is fat' and Strawson's objection kicks in. Furthermore, we can move between these by superficial syntactic transformations. The truth of one implies the truth of the other. We don't need to penetrate even to the Tarskian depth of showing that someone called 'Al' exists for which the property called 'fat' applies. It's just the way that proper names, monadic concept words, 'true', and the verb 'to be' are used together.

My objection is to the form of the truthmakers currently being discussed. They simply lack the explanatory value we seek. So my feeling is the move you make in para (2) from 'Al is fat' to 'the fact of Al's being fat' leaves us in the same place, 'the fact of _' behaving analogously to '_ is true'. But truthmakers of a different structure might be viable.

In your para (1) you say

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. 
This is surely right. I assume 'way' is understood just as it is in 'there are six ways of ordering a red, a white, and a blue token'. This may look like an existential assertion but these ways cannot be entities in the world. Rather they are possibilities of how the world might be. What makes p true is that the actual world is arranged in a certain way. Another possible world may be so arranged that p is not true. So this idea would make contact with modality but place truthmakers, understood as 'ways the world might be', outside the world.

Very good, David.

Your main point seems to be that there is nothing explanatory in saying that 'Al is fat' is true because the fact of Al's-being-fat obtains. And this is presumably because the fact cannot be picked out except by using a phrase -- 'Al's being fat' -- that is equivalent to 'that Al is fat.'

Could your point be summarized by saying that the purported explanation is worthless because circular? That it really amounts to saying that 'Al is fat' is true because Al is fat?

Yet you concede that there must be something external to language, that 'Al is fat' cannot just be true. 'True' implies a relation to something language-transcendent.

The question is what this extralinguistic entity is. I say it cannot just be the concrete particular, Al. It has to be something with a proposition-like structure, a fact. You seem to be denying this when you say that truthmakers are outside the world. But then where are they? Are the ways, the structurings, added by the mind or added by language? But then you lose the notion that truth of a sentence is determined by something in the world.

I am tempted to say 'so what?' to the Strawsonian-Butchvarovian point that we cannot access facts except by using the very sentences that the facts are supposed to make true. This by itself does not show that we are not getting outside language. Or does it?

Your view, and perhaps Ed's too, issues not in full-blown linguistic idealism, but a sort of 'Kantian' idealism according to which particulars are in the world beyond language, but we supply the structuring. So in reality you have a and F-ness, but a's being F is our doing.

But then the notion of a truthmaker is shot to hell . . .

Hello Bill,
Suppose we had a phone application that could photograph the pieces on a chessboard and emit a description of the position in the standard format one finds in newspapers and chess books. How might such an app work? One obvious approach would be (1) to apply some kind of pattern recognition algorithm to the photo image to derive a description of the board position in an 8x8 array of (colour, piece) pairs. For example, at the start of the game the pair (white,king) would appear at element [1,5] of the array. Then (2) we could transform this description into the conventional format.

Let's take our phone at some point in a game and have it snapshot the state of play. The resulting output is a list of symbols which we can think of as a sentence. It makes sense to talk about its truth or falsity, and if it's true we can ask what 'makes' it true. I suggest that there are two aspects to this corresponding to the two steps outlined above. Firstly there is the detection of some pattern in the way the pieces are arranged on the board. The array of pairs in the phone memory after step (1) represents this pattern. Secondly there is an encoding of the detected pattern into a linear sequence of symbols. The first step requires that the phone be in causal contact with the chessboard. The second step, though mediated by causal processes, is governed by rules imposed by convention. The final output is true in so far as (1) the pattern is correctly detected, and (2) the detected pattern is correctly encoded according to the rules.

The nearest we get to a truthmaking 'thing' in this account is what I've been calling a 'pattern'. I appreciate that this term carries a heavy load. Patterns are 'in' the world in the sense that we say that there is a pattern in a piece of wallpaper, or a pattern exists in the paper. But I'd rather say that the wallpaper bears an instance of the pattern, which itself is an abstract thing. And it's difficult to find in a pattern anything of a propositional character, I think. This is only to be found at the linguistic end of my putative process. Lastly, though I think causation plays a large part in the process of getting from a state of the world to one containing a true statement about it, I'm not claiming that truthmaking is causation in the sense that you inveigh against. Rather I say that a symbol string is true of a chessboard position iff the phone would emit the string were the board presented to the phone.

The comments to this entry are closed.

My Photo
Blog powered by Typepad
Member since 10/2008



June 2024

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