The Correspondence theory of truth Frank Hofmann

1. draft, July 2003 The Correspondence theory of truth Frank Hofmann 1 Introduction Ever since the works of Alfred Tarski and Frank Ramsey, two views on truth have seemed very attractive to many people. On the one hand, the correspondence theory of truth seemed to be quite promising, mostly, perhaps, for its ability to accomodate a realistic attitude towards truth. On the other hand, a minimalist conception seemed appropriate since it made things so simple and unmysterious. So even though there are many more theories of truth around - the identity theory, the prosentential theory, etc. -, it is fair to say that these two views have acquired the status of the main contenders in the field. Most recently, John Searle and David Lewis have taken sides on the issue. Searle defends a new version of the correspondence theory which takes facts to be the correspondents of true statements, whereas Lewis wants to hold on to a minimalist conception of truth - allowing for an appendix of purely ontological claims that he extracts from the correspondence theory. What is new is that both have tried to see the correspondence theory not as a rival to the minimalist conception, but rather as essentially compatible with it. Still, in the end, Searle arrives at his version of the correspondence theory, and Lewis votes for minimalism. Both philosophers have sympathy with the spirit of the correspondence theory, to say the least. So one wonders why it is that they come to so different conclusions. Both start from the spirit of the correspondence theory, and both see it as essentially compatible with the heart of the minimalist conceptions, and still is there no agreement on what theory is to be accepted in the end. It is my goal in this paper to analyze the arguments that Searle and Lewis have presented and to evaluate their validity. Both positions contain valuable insights. Yet I see mistakes in the arguments of both, and I would like to present my own view which will be a version of the correspondence theory. Taking up the idea of Searle's and Lewis's that the correspondence theory and the minimalist theory are not really in conflict with one another, I will try to show why the correspondence theory prevails in the end. In effect, the correspondence theory can do so by swallowing up the core of the minimalist theory, as it were, as a condition of adequacy on the correspondence relation. So the relation between the correspondence theory 1

and the minimalist theory is simply that of the whole to a proper part of it. And that explains well, in my view, why so many have found the minimalist theory quite attractive. 2 Searle's correspondence theory of truth Let us begin by looking at Searle's argument. John Searle has recently tried to defend the correspondence theory of truth. He wants to show that the minimalist theory can serve as a kind of supplementation to the correspondence theory of truth. The correspondence theory of truth he takes to be saying the following: (C) A statement is true if and only if it corresponds to a fact. 1 Here, Searle allows us to speak of a correspondence to the facts instead of the singular, as in (C). But it is clearly mandatory and essential for Searle that the correspondence theory be stated by means of the notion of a fact. The correspondence theory is, according to Searle, a theory that essentially refers to facts. And Lewis's statement of the correspondence theory makes equally use of the notion of a fact. 2 This is, I believe, a mistake. There is nothing in the idea of correspondence as it occurs in the intuitions standing behind the correspondence theory of truth that forces us to speak of facts. And indeed, to state the correspondence theory by reference to facts from the very beginning, leads into serious trouble, as we will see soon. But let us first follow Searle in his argument. The second starting point for Searle is the disquotation criterion (or principle) for truth. This derives from Tarski's famous T sentences that are supposed to constrain a definition of the truth predicate. It has inspired Ramsey's redundancy theory of truth, and equally the deflationary and minimalist theories of truth. 3 It can be put as follows: (D) For any statement S, S is true if and only if p, where for 'S' we have to substitute a quotation or other specification of the statement, and for 'p' we have to substitute the statement itself. 4 1 Cp. Searle (1998), p. 387. 2 Cp. Lewis (2001b), p. 275, for example. 3 Cp. Searle (1995), p. 209. 4 Cp. Searle 1998, p. 385. In Searle (1995), p. 201, essentially the same principle is stated, only with talk of sentences instead of talk of statements. I will suppose that this difference is irrelevant for our purposes here. 2

The idea is familiar enough. The truth-predicate serves as a means for semantic ascent. We may have certain quarrels with the way this idea is formulated in (D), but this should not deter us from finding the idea itself very appealing. From the disquotation criterion we extract the minimalist theory of truth which will serve well our purposes for the ensuing discussion. The heart of what is standing behind the disquotation criterion can be put well by making use of the notion of a proposition. Assume that statements involve propositions, and that a statement is true if and only if its proposition is true. Then we arrive at the following sentences: (M) The proposition that the cat is on the mat is true iff the cat is on the mat; the proposition that pigs fly is true iff pigs fly;... Here we have what Lewis has aptly called the redundancy biconditionals. 5 And the minimalist theory of truth can be simply taken as the infinite conjunction of all these (nonparadoxical) redundancy biconditionals. It has certain advantages, when compared with the disquotation formulation (D). The relation between a sentence, or a speech act ('statement'), and its propositional content has been filtered out, by speaking of the propositional content only. So no quotation is needed any more. And we no longer have any universal quantification at the beginning, but work with an infinite conjunction instead. This all helps to bring out the essentials more clearly, and to avoid any distractions. Therefore, I will use the minimalist theory (M) in what follows. But please note that all of what I will argue in the following can also be argued by strictly sticking with Searle's disquotation criterion (D) literally. Now, Searle observes that (C) suffers from a certain shortage. For (C) states that a statement is true if and only if it corresponds to a fact (or the facts), but it does not tell us which fact this is supposed to b. And here Searle sees help in the disquotation criterion. The disquotation criterion is by no means incompatible with the correspondence theory, argues Searle. Rather, it can help out the correspondence theory in order to arrive at a more 'complete' correspondence theory of truth. This can be seen as follows. From (C) and (M) we can derive 5 Cp. Lewis (2001a), pp. 602-3; Lewis (2001b), p. 275. (M) can also be found in Paul Horwich's minimalist 3

(1) The statement that p corresponds to a fact iff p. (Here I have filled in (C) by mentioning explicitly the propositional content of the statement.) 6 And now, Searle recognizes, we have almost arrived at a characterization of what fact it is to which the statement that p is supposed to correspond if it is to be true. For which fact else could be this searched for fact if not simply the fact that p?! Thus Searle concludes (2) The statement that p is true iff it corresponds to the fact that p. In this way, Searle has arrived at his 'full' correspondence theory of truth: the statement that p is true if and only if it is a fact that p, and the fact to which the statement corresponds (if and only if it is true) is simply the fact that p. Correspondence theory and disquotation principle/minimalist theory go hand in hand, there is no conflict or incompatibility between the two. So far, so good. However, the trouble with this version of the correspondence theory is that it leads into the classical problems with true negative predications, true negative existentials, and the like. If the statement that there is no unicorn is true, what fact is the fact that there is no unicorn? Do we really want to assume that there is such a fact, that the fact that there is no unicorn exists as an entity? Equally, when the cat is not on the mat, do we really want to assume that there is the fact that the cat is not on the mat? To many (including myself) this is too much to be swallowed. Probably, there are facts. But many true statements do not correspond to facts, they correspond to other elements of reality, or simply to reality as a whole. Searle acknowledges the problem. He realizes that we are inclined to reject the negative fact that the cat is not on the mat as the fact which makes the true statement 'The cat is not on the mat' true. Rather, we would like to say that it is some other positive fact - for example, the fact that the cat is standing in the kitchen - which makes the statement true. But for Searle, this is a 'metaphysical urge' that we should question. And Searle offers a diagnosis of what is wrong with it: our specification of the truthmaker as 'the fact that the cat is not on the mat' leaves it open in which way exactly the world makes the statement true. But this, he argues, is exactly the same in the case of a true positive statement: theory of truth (cp. Horwich 1990, ch. 1 and 2) and Scott Soames's deflationary view (cp. Soames 1999, ch. 8). 6 If we use Searle's disquotation criterion literally, we arrive at (1') The statement S corresponds to a fact iff p. with S and p as above in (D). The argument continues just as well. 'The statement S' takes the place of 'the statement that p'. 4

"To put this point rather grandly, it is in the nature of language that it allows a slack in the determination of truth conditions by statements, but it is in the nature of reality that it allows no slack in anything. Strictly speaking this is as much a problem for affirmative as for negative statements. Just as there are an indefinite number of ways for the cat to be not on the mat, so there are an indefinite number of ways for it to be on the mat." (Searle 1998, p. 398, emphasis in original) So the positive and the negative true statement are companions in guilt, if they are guilty at all. All - or at least many - statements, positive or negative ones, are somewhat indeterminate in what features of the world would make them true. It is typically true of any statement that "the statement allows a certain slack in what determinate features of the real world would constitute its truth condition. This, I believe, is the underlying reason why we feel there is a problem about negative facts or hypothetical facts." (Searle 1998, 398) The conclusion that Searle draws from this is that there is a deep and as yet unresolved tension in our concept of a fact. "We are, in short, sometimes asking the word 'fact' to do an impossible job and this is the deeper source of our philosophical confusions." (Searle 1998, p. 399) It is not easy to state exactly what tension Searle has in mind here. To me it seems to be the following one. Sometimes we use the term 'fact' according to a disquotational rule. In this sense, the statement that the cat is not on the mat is made true by the fact that the cat is not on the mat. Here the expression 'the fact that...' describes the condition that the world has to satisfy, and we arrive at negative facts. But sometimes we use the term 'fact' in order to designate the determinate way the world is that satisfies the statement's truth condition. Then we cannot simply use the disquotational rule, but have to find out what the determinate way of the world is in order to make the appropriate substitution for 'p' in 'the fact that p'. The problem with Searle's diagnosis here is the following. There may be such a tension in the ordinary notion of a fact. There may be even two notions of a fact, a 'deflationary' one and a more 'substantial' one. Other philosophers have made essentially the same observation, 5

in some way or other, and I agree. 7 But for a philosophical theory of truth that relies on facts, it is necessary to have an unambiguous and clear notion of a fact. So Searle has to decide which notion of a fact he wants to rely on in his correspondence theory of truth. However, there is a dilemma for him. On the one hand, if he wants to choose the deflationary notion, which is subject to the disquotational rule, then he will have to admit negative facts, universal facts, hypothetical facts, and so on. If, on the other hand, he chooses the more substantial notion of a fact, then his whole correspondence theory of truth breaks down, since he can no longer use the disquotational criterion. Given that the disquotational criterion was one of his starting points, the latter alternative is excluded, and so Searle is forced into the former horn of the dilemma, the acceptance of negative facts and the like. It is of no help to Searle that there is another notion of fact according to which there are no such negative facts, universal facts, conditional facts, and so on. According to his notion of a fact there are these negative facts and the like, and that is a highly undesirable result. 8 Searle might attempt to show that the result is not so undesirable after all. For, he might argue, the term 'the fact that p' is simply a term that picks out, in the context of use, a certain entity, or certain entities, that make the statement that p true on that occasion. For example, in the case of our statement 'The cat is not on the mat', the term 'the fact that p' might pick out the cat which is standing in the kitchen on one occasion, and the cat which is lying on the grass on another occasion, and so on. In one word, the term 'the fact that p' simply picks out the truthmaker(s) of the statement that p. And which entity or entities are the truthmakers may vary from case to case. This maneuver, however, is not successful, since it will make Searle's correspondence theory of truth empty. It was Searle's main idea to claim that the correspondent in the world for the true statement that p is the fact that p. If we are now told that the fact that p is simply the truthmaker for the statement that p, then Searle's claim turns into the claim that the correspondent in the world for the true statement that p is the truthmaker for the statement that p. This, however, is empty as long as we are not told what the relation of truthmaking is. As long as no truthmaker account is supplied, the number three could be the truthmaker of all true statements, their correspondent, for all we know. Because Searle nowhere produces an account of truthmaking, his proposal remains empty. 9 7 Compare, for example, Hugh Mellor's distinction between facts and facta (Mellor 1995, ch. 13.4). See also Armstrong's distinction between facts and states of affairs (Armstrong 1997, p. 19). 8 According to Searle, facts are conditions. But this is also of no help to avoid the undesirable conclusion just mentioned. As long as it is claimed that these conditions exist, that they are entities in the world, the conclusion follows. And Searle does say so at several places. Cp., for example, Searle (1995), p. 211; Searle (1998), p. 396: "And the word 'fact' has evolved precisely so that we can specify those conditions that exist in the world in virtue of which true statements are true." (my emphasis) 9 I suspect that this maneuver is still the best way for Searle to go. For it might turn out that it leads to a theory 6

What are we to do then? Can we rescue the correspondence theory from this problem? It seems to me, Yes, we can. All we need to do is to eliminate all reference to facts from the statement of the correspondence theory of truth. It is no part of the correspondence theory as such that there are facts. This seems to me the basic mistake in the formulation (C) of the correspondence theory, as given by Searle and equally by Lewis. Neither Searle nor Lewis provides any argument for the claim that the correspondence theory contains reference to facts. And there is no need for this. All of the intuition, or intuitions, standing behind the correspondence theory of truth can be captured without speaking of facts. It is an unnecessary burdening of the correspondence theory to include facts in its formulation. For we can equally express the realistic attitude that truth depends on reality by simply speaking of correspondence to reality. Thus we can state the correspondence theory as: (C') A statement is true if and only if it corresponds to reality. No mentioning of facts is necessary here. We can leave entirely open the question of what reality consists of when we state the correspondence theory. It is no part of the correspondence theory to tell us what entities reality consists of. This is the job of a comprehensive ontological theory. The correspondence theory is no such theory, but only a theory of what makes a statement true. Clearly, (C') captures the spirit behind the correspondence theory, namely, that the truth value of a statement depends on what reality is like. No more commitments have to be included in the correspondence theory, and for reasons of transparency, they should not. Furthermore, mention of facts misleads us into Searle's line of thought. Once we mention facts in the statement of the correspondence theory, as Searle does, we are immediately lead by the very formulation to bring up the question 'what fact does the statement that p correspond to (if and only if it is true)?' and to answer it in Searle's way. But no such thing happens when we state the correspondence theory as (C'), without reference to facts. Then we can derive, from (C') and (M), simply the following conclusion: (3) The statement that p corresponds to reality iff p. that is ontologically more or less the same as the proposal that I will present, except that the account of the relation of truthmaking is missing. The other remaining difference is merely semantical: Searle wants to call all truthmakers 'facts' - so for example the cat Bruce would be called a fact, in virtue of being a truthmaker for the statement 'There are cats' -, whereas I would like to reserve the term 'fact' for a special category of entities (something like Armstrong's states of affairs). 7

Thus we have simply used reality as a whole as that to which a statement corresponds when it is true, and we have not introduced any problematic facts, and are not even inclined to do so. We run into the Searle's problem of negative facts etc. only if facts are mentioned in the statement of the correspondence theory at the very beginning. 10 3 Lewis's minimalism Lewis has also presented the view that the correspondence theory of truth is not really incompatible with the minimalist theory, but rather a kind of ontological rider on the true - minimalist - theory of truth (Lewis 2001a, 2001b). In his view, the classical formulation of the correspondence theory is something like this: (CB) The proposition that the cat is on the mat is true iff it corresponds to fact; the proposition that pigs fly is true iff it corresponds to fact;... (cp. Lewis 2001b, p. 275) (This formulation suffers from the same defect as Searle's, since it makes reference to facts. But let us ignore this for the moment.) The minimalist theory of truth, on the other hand, consists of the following infinite conjunction of propositions: (M) The proposition that the cat is on the mat is true iff the cat is on the mat; the proposition that pigs fly is true iff pigs fly;... Thus, we are able to derive biconditionals of the following form: (4) The proposition that the cat is on the mat corresponds to a fact iff the cat is on the mat. Or, schematically, 10 This is not to say that there is no real problem of finding a truthmaker for negative predications, negatives existentials, and so on. But this truthmaker need not be a fact, in contrast to what Searle has to provide, given his version of the correspondence theory. 8

(1) The statement that p corresponds to a fact iff p. (For the sake of simplicity, I do not distinguish between the statement that p and the proposition that p here. 'Correspondence to fact' and 'correspondence to a fact' I take to be synonymous.) But now we have arrived at a set of statements that are not about truth at all, since truth is not mentioned at all in (4) and (1) and any other instance of (1). Thus, Lewis concludes, (1) and (4) and the like represent the real content of the correspondence theory of truth - only that this theory is misnamed as a theory of truth. The theory is not about truth but rather "about the existential ground of all manner of other things: the flying of pigs, or whathave-you" (Lewis 2001b, 279). The theory is a purely ontological theory. And this theory we should think of as the true correspondence theory of truth. 11 But is Lewis's argument really plausible? Lewis is right in thinking that the correspondence theory adds something to the minimalist theory. What it adds is an identification, namely, the claim that the property of being true is the property of standing in the correspondence relation to reality. (From now on, I switch from 'correspondence to fact' to 'correspondence to reality', following out earlier insight.) 12 The property of being true is such that the minimalist's biconditionals (M) The proposition that the cat is on the mat is true iff the cat is on the mat; the proposition that pigs fly is true iff pigs fly;... hold. This claim - the claim of the minimalist theory - is somehow part of the correspondence theory. But looking at things in this way opens up a new way of conceiving of the correspondence theory of truth, namely, as follows: 11 In this context, Lewis spells out 'correspondence to (a) fact' by means of necessitation, as it is usually done in truthmaker theories (cp. Lewis 2001b, p. 278-9). But it is irrelevant to his argument that the correspondence relation be spelled out in this way (or any other way, at least, as long as the notion of truth does not occur in the spelling out). Therefore, I have stayed with the general correspondence formulation. I will come back to truthmakers in the next section. 12 Indeed, one of the two complaints that Lewis raises against the correspondence theory of truth, when spelled out as a truthmaker theory, is that not all truthmakers are facts. This complaint evaporates once the correspondence theory is taken as speaking of correspondence to reality. Cp. Lewis 2001b, p. 278. His second complaint is the one that is discussed in the main text, the misnomer complaint. 9

(CTT) There is a relation C - the correspondence relation - that can hold between a proposition and reality; and the property of being true is the property of standing in the relation C to reality; and C is such that (M'). where (M') The proposition that the cat is on the mat stands in C to reality iff the cat is on the mat; the proposition that pigs fly stands in C to reality iff pigs fly;... Thus, the minimalist's biconditionals have been reformulated as biconditionals about the correspondence relation C that is postulated by the correspondence theorist. C gives the nature of the property of being true - as it has always been the intention of the correspondence theorist. And the minimalist biconditionals (M) turn out to be a collective condition on this relation C. This collective condition is a close relative of what Tarski called the convention T. 13 It all comes in one package. But the package is a package about truth, through and through. (M) is only an extraction, a part of the package, the part which leaves open the nature of the property of being true. This part is by no means trivial. For there are other relations between propositions and reality that do not conform to the redundancy biconditionals. For example, the relation of being thought by some subject does not conform to the redundancy scheme. Call this relation 'R'. Then many of the claims (5) The proposition that the cat is on the mat stands in R to reality iff the cat is on the mat; the proposition that pigs fly stands in C to reality iff pigs fly;... are clearly false, since there are many propositions which are the object of some thought but are not true, and many propositions which are true but not the object of some thought. So satisfaction of the schema expressed by (M') is by no means trivial. Still, satisfaction of this scheme, as it is expressed by (M), is only a proper part of the correspondence theory of truth. 13 Cp. Tarski (1956), pp. 187-188. 10

Whether the adequacy condition is formulated in the form (M) or (M') does not matter from the point of view of the correspondence theorist. For the correspondence theorist the two amount to the same, since being true and having C to reality are one and the same. Only for those who reject this identification the difference matters. Thus, the minimalist speaks of something the nature of which he leaves open. By now, I have turned Lewis's view upside down. For Lewis, the minimalist theory is the theory of truth, and the correspondence theory is a purely ontological theory. Combining both leads to a set of claims that are usually, traditionally called 'the correspondence theory of truth'. In my view, both theories are about truth. The correspondence theory (CTT) is a claim about the nature of truth, and the minimalist theory is a claim about truth. The minimalist theory (M) can be seen as the outcome of combining the correspondence theorist's identification claim (identification of truth with correspondence to reality) with her claim (M') about the correspondence relation C. Lewis has not provided any argument to the effect that we have to see things in his way. Because the way of looking at things that I have presented respects the traditional view that the correspondence theory really is a theory about truth, and even about the nature of truth, it is preferable to Lewis's view. So I recommend (CTT) as the proper formulation of the correspondence theory of truth. 14 4 The correspondence theory further developed The correspondence theory can be de developed further in certain ways. The redundancy biconditionals (M') state only the first set of conditions that the correspondence relation is supposed to satisfy. But further things can be said - and should be said - about the correspondence relation. We can take what (CTT) says as part of our ordinary concept of truth. Then, (CTT) has something like the status of a conceptual truth. Any further claim about correspondence is a theoretical claim. And there is indeed a lot of room for further theory here, since (CTT) leaves 14 This puts new light on the issue of whether Aristotle and Tarski can be seen as correspondence theorists. They have explicitly stated (a part of) something like the minimalist theory of truth, but Aristotle perhaps wanted, and Tarski explicitly said that he wanted, to express the view that truth is some kind of correspondence. Cp. Aristotle, Categories, 4b 8-10; 14b 13-22. At these passages, Aristotle even seems to articulate the idea of a truthmaker as the correspondent of a truth (which he calls a 'ground' - 'aitios' - of the truth). Cp. Tarski (1956), pp. 154-156, where Tarski also refers to Aristotle. If I am right, they were on the right track, since there is nothing that could have prevented them from taking the minimalist theory as a proper part of the correspondence theory. 11

open very much about what the nature of the correspondence relation C itself is. 15 We should therefore go on and theorize about C. One very promising route here is to spell out correspondence in terms of a truthmaker relation. Many people have found a truthmaker principle both plausible and congenial to the correspondence theory of truth. And some have proposed that we should understand the truthmaker principle, if suitably accompanied by an account of the nature of truthmaking, as the true correspondence theory of truth. Thus, for example, David Armstrong says: "Are truthmakers for truths just the 'correspondents' envisaged by the Correspondence theory of truth? I think the answer to this is a qualified 'Yes'. Truthmaker theory is a correspondence theory, but it does not burden itself with the dogma that the correspondence has to be one-one. This point was not, I think, ever explicitly stated by Russell, but he understood it well. The freeing-up that results from jettisoning the one-one dogma breathes new life into the Correspondence theory." (Armstrong 2001, p. 30) "Anybody who is attracted to the Correspondence theory of truth should be drawn to the truthmaker. Correspondence demands a correspondent, and a correspondent for a truth is a truthmaker." (Armstrong 1997, p. 14) 16 But what then is truthmaking? - There are many open questions about what exactly is the right definition of 'truthmaking'. But the starting point of all investigations is clear enough: a truthmaker is something in the world that makes a truth true; and truthmaking is something like necessitation. More exactly, we can say preliminarily that an entity x makes the proposition that p true if x's existence (metaphysically, or logically in the broad sense) entails that p. Thus, for example, John Bigelow says: "I urge that what the Truthmaker axiom is really saying is this: Whenever something is true, there must be something whose existence entails that it is true. The 'making' in 'making true' is essentially logical entailment." (Bigelow 1988, p. 125) 17 15 Compare Alston's distinction between the concept of truth and the property of truth (Alston 2001). 16 Other philosophers who have taken the truthmaker theory as the 'true' correspondence theory are J. Bigelow and J. Dodd. Cp. Bigelow (19988), p. 122; Dodd (2000), ch. 1. Dodd identifies the correspondence theory and the truthmaker theory from the very beginning. 17 Cp. Lewis (2001a), p. 605; Lewis (2001b), pp. 277-279; Armstrong (1997), p. 115. It is fair to say that in the meantime entailment has become the standard way of explaining the truthmaking relation. 12

Then Bigelow goes on and notes that this account of truthmaking may not be exactly right. Perhaps, he suggests, some cases of entailment are not cases of truthmaking, but all cases of truthmaking are cases of entailment. So Bigelow proposes, as a preliminary refinement: "Whenever something is true, there must be something whose existence entails in an appropriate way that it is true." (Bigelow 1988, p. 126) Here we can indeed stop and leave it to further investigations to find out what the 'appropriate way' is. 18 The main point that I wanted to make is simply: there is a whole body of ideas centering around necessitation that can be made use of in the attempt at spelling out the nature of truthmaking. So the identification of the correspondence relation C with the truthmaker relation is both promising and an open project that requires further theoretical work. Some truthmaker account seems well-suited for the job of further theoretical illumination of the correspondence relation C. Along this way, we can hopefully find a mature correspondence theory of truth. 19 References: Alston, W. (2001), "Truth: concept and property", in: R. Schantz (ed.), What is Truth?, de Gruyter, Berlin, 2001, 11-26. Armstrong, D. (1997), A World of States of Affairs, CUP. Gregory, D. (2001), "Smith on Truthmakers", AJP 79 (2001), 422-427. Horwich, P. (1990), Truth, Blackwell, Oxford. Lewis, D. (2001a), "Truthmaking and difference-making", Nous 35:4 (2001), 602-615. Lewis, D. (2001b), "Forget about the 'correspondence theory of truth'", Analysis 61:4 (2001), 275-280. Mellor, D.H. (1995), The Facts of Causation, Routledge, London. Searle, J. (1998), "Truth: a reconsideration of Strawson's views", in: L.E. Hahn (ed.), The Philosophy of P.F. Strawson, Open Court, Chicago and Lasalle, Illinois, 385-401. 18 For example, Barry Smith has tried to give an account of the 'appropriate way' (Smith 1999). For a critique of his account see Dominic Gregory (2001). Smith (2002) contains Smith's reply to Gregory. 19 Acknowledgment. 13

Searle, J. (1995), The Construction of Social Reality, The Free Press, New York. Smith, B. (2002), "Truthmaker realism: response to Gregory", AJP 80 (2002), 231-234. Smith, B. (1999), "Truthmaker realism", AJP 77 (1999), 274-291. Soames, S. (1999), Understanding Truth, OUP, Oxford. Tarski, A. (1956), The concept of truth in formalized languages, in: A. Tarski, Logic, Semantics, Metamathematics, Clarendon, Oxford, 152-278. 14