New Wave Deflationism Nic Damnjanovic Abstract: Traditional deflationary theories of truth have evolved into a position I call new wave deflationism. This position consists of two core claims: that truth is a logical notion/property and that it satisfies the doctrine of Revelation. I argue that new wave deflationism is immune to many of the standard objections to deflationary theories of truth and so the standard debates between inflationists and deflationists are on the wrong track. In fact, the debate between these views may be of far less importance than is usually thought.

1. Introduction For many, the paradigm of a deflationary theory of truth is the redundancy theory, which is typically taken to consist of two claims: namely (i) that sentences containing the truth predicate are synonymous with sentences not containing the truth predicate (and so the truth predicate is redundant) and (ii) that there is no property of truth.1 The redundancy theory is not an attractive theory of truth since neither of its claims is particularly plausible on its own, and the combination of the two claims is, if not actually inconsistent, at least uncomfortable.2 Very few deflationists nowadays endorse either part of the theory. After abandoning redundancy, deflationists diversified into a bewildering variety. As a result, it is probably not possible to delineate a position that all deflationists share. Nevertheless, in recent years a degree of unity has emerged amongst a number of contemporary deflationists about truth including Hartry Field, Christopher Hill, Paul Horwich, Wolfgang Künne, Scott Soames and Michael Williams.3 Even Colin McGinn, who denies he is a deflationist, has endorsed this position.4 I call all these philosophers new wave deflationists. While new wave deflationists disagree about much, there is a core position they share which deserves careful attention: or so I will argue.5 Here’s what I have planned. First, I’ll outline what I think is the best way to 1

The closest to a real proponent of the redundancy theory is Ayer (1936: 88). The problem I have in mind is this. To capture familiar ‘blind’ ascriptions of truth, such as ‘Truth is sometimes stranger than fiction’, without using the truth predicate, thereby maintaining the claim that it is redundant, the redundancy theorist needs to appeal to some form of higher-order quantification as in ‘There are values of p and q such that p and q and p is more surprising than q.’ But now it looks as if we can define a property of truth like so: 2

For all objects x, (x is true iff there is a value of p such that ((x = the proposition that p) and p)) So, if the truth predicate is redundant, then there is a property of truth we can define. See Soames (1999: 48) for more discussion. 3 See Field (1994), (2001); Hill (2002); Horwich (1998a), (1998b); Künne (2003); Soames (1999), (2003); Williams (1986), (1999). The position Eklund (this volume) calls ‘rejectionism’ has much in common with these positions too. 4 McGinn says he is not a deflationist because i) his view is neither philosophically uncontentious or tame (2000: 107) and ii) he denies that ‘p’ and ‘p is true’ express the same proposition (2000: 92-5). The first claim misses the point, as deflationism is often regarded as radical and highly controversial. As I’ve already mentioned, contemporary deflationists typically reject the second claim. 5 I don’t count prosententialists as new wave deflationists, since they apparently do not view ‘true’ as a predicate. But see Lance (1997) for a different view.

1

understand new wave deflationism, so that we can clearly distinguish the view from inflationary theories of truth. Once we have this formulation of the position, it will become clear, I hope, that a number of the current debates between deflationists and inflationists are really not to the point. The upshot of this discussion is that when we separate the core position of the new wave deflationists from some of the extreme claims added to the position by its main proponents, we can better appreciate the value of the new deflationist position. Finally, I briefly describe the direction in which I think the real debate between inflationists and deflationists lies, and come to a suitably deflationary conclusion about the importance of the debate. (Before beginning, a point of terminology. For ease of expression, I will focus largely on theories of truth for propositions. However, the two key theses that define a position as a version of new wave deflationism can be applied to any truth bearer.) 2. New Wave Deflationism What all deflationists have in common, and what is responsible for them being called ‘deflationists’, is the view that there really is no philosophical problem of truth. This is why deflationists think the traditional debates about truth should be deflated. But why do deflationists think there is no philosophical problem of truth? Compare conjunction. Why does no one think that there is a philosophical mystery of conjunction? Our concept of conjunction is a simple logical concept, and the nature of conjunction (i.e. the truth-function) is fully revealed to us when we grasp that concept. So there is no further need, in thinking about conjunction, for philosophical analysis or inquiry. The whole truth about the nature of conjunction lies open to view to anyone who grasps our concept of conjunction. The new wave deflationists, I suggest, should be understood as saying much the same about truth: the concept of truth is a logical concept such that the nature of the property of truth is fully revealed to us when we grasp that concept. It is probably uncontroversial that most new wave deflationists are committed to these two claims. What is likely to be more controversial is that these two claims are sufficient to express the core of the new wave deflationist position and that, so understood, new wave deflationism is immune from many of the standard objections to it. The rest of this section is devoted to convincing you of these things. 2.1 Truth as a Logical Notion As is now familiar, new wave deflationists typically hold that the truth-predicate is merely a device of generalization. That is, the role and function of the truth predicate in natural language is exhausted by its capacity to allow us to formulate certain generalizations, such as ‘Everything Kripke says is true’. By committing to this thesis about the truth predicate, deflationists commit themselves to the idea that it is a logical predicate, and the concept of truth is a logical concept. Indeed, Field and Horwich have characterized their views with the statement that ‘the truth predicate exists solely for the sake of a certain logical need’6 and is ‘a “logical predicate” like “exists” or “equals”’.7 They also endorse the idea that truth is a logical property.8 6

Horwich (1998b: 2). Field (1992: 322), attributes this view to Horwich, but also seems to accept it. 8 See also McGinn (2000: chapter 5); Künne (2003: 91, 338). 7

2

I take it that the distinction between a logical concept or property and a non-logical concept or property is fairly intuitive, however hard it might be to specify a precise definition of ‘logical concept’ or ‘logical property’. And while there may be borderline cases of logical concepts, there are at least uncontroversial examples of both the logical and non-logical. For example, I assume we would all agree that on the logical side of this distinction we have expressions like ‘and’, ‘every’, ‘or’, ‘not’, and on the non-logical side we have expressions like ‘blue’, ‘tree’, ‘magnetic’, ‘justice’ and ‘conscious’. Corresponding to these expressions, we have logical and non-logical properties and relations.9 Because there are these uncontroversial examples of logical notions, it is also pretty straightforward, in most cases, to determine whether a theory of truth entails that truth is a logical notion. For example, one type of deflationary theory entails that the concept of truth is a merely logical concept because it defines truth solely in terms of logical concepts. For example, one could define truth using (1). 1. For all x, x is true iff ∃p((x = ‘p’) and p). As (1) appeals just to universal quantification, the biconditional, existential substitutional quantification, quotation marks, identity and conjunction, it entails that truth is a logical notion—assuming that (1) is a complete definition of truth. Similarly, Hill has offered a version of deflationism, called substitutionalism, that defines truth using (2). 2. For all x, x is true iff ∃p((x = the proposition that p) and p). As Hill asserts, definitions like this one ‘reduce semantic concepts to substitutional quantification’10 and so his position also implies that truth is a logical concept.11 Alternatively, some theories describe the meaning or function of the truth predicate in ways that, if correct, make the truth predicate much more like other logical expressions than non-logical expressions. For example, those who argue that the truth predicate is merely a device of generalization, or merely a device of disquotation, clearly think that the truth predicate is more like logical expressions such as ‘every’ and ‘some’, or devices of quotation, than expressions such as ‘justice’ or ‘magnetic’. Similarly, Tim Maudlin has argued that giving the semantics for ‘true’ is just like giving the semantics for ‘and’—it is simply a matter of pointing to a certain truthfunction. That is, he suggests that the truth predicate expresses the ‘identity-map’ truth function so that attributions of truth to a proposition have the same truth-value as the proposition to which truth is being attributed. On this account, ‘true’ is on all fours

9

Because any n-ary function can be thought of as an n+1-ary relation, I will speak of functions, like the truth-function of conjunction, as relations, or relational properties. 10 Hill (2002: 23). Hill assumes it is legitimate to use substitutional quantification when dealing with propositions rather than sentences, though this is not a question I can answer here. 11 It should be noted, though, that both (1) and (2) make use of name-forming devices—quotation marks in (1), and ‘the proposition that …’ in (2). Although I think such name-forming devices deserve to be called ‘logical’ devices, others might find them to be borderline cases of logical notions. If so, perhaps truth is best characterized as quasi-logical or ‘broadly logical’ (Künne 2003: 338).

3

with other logical expressions.12 Thus, while there may be other accounts of the truth predicate that make it a borderline case of a logical expression, I think the distinction between logical and non-logical expressions is tolerably clear, and clear enough to be used in the formulation of a theory of truth.13 It is up to particular new wave deflationists, of course, to convince us that truth is in fact a logical property. 2.2. Revelation As well as believing that truth is a logical property, new wave deflationists believe that truth has no hidden essence. For example, some new wave deflationists maintain that grasping the instances of (PT) tells us all there is to know about the nature of truth (for example Horwich 1998b). PT. The proposition that p is true if and only if p. And deflationists typically hold that believing the instances of (PT) is a possession condition for the concept of truth. Taken together these ideas entail that by grasping the concept of truth we are thereby in a position to grasp the full nature of truth. But what, in general, does it mean to say that truth has no hidden essence? While the claim is often mentioned,14 it is rare to find an attempt at explanation. But one thing the ‘no hidden essence’ idea does not have to imply is that there is no essence or nature of truth at all: it only implies that the nature is not hidden. And there is an idea discussed in the philosophies of colour and mind that we can employ to explain what it is for a property’s nature to not be hidden—the doctrine of Revelation. We can define a revelatory concept as follows: C is a revelatory concept if and only if any subject S who grasps the concept C is in a position to know the full nature of the property C without further empirical investigation or a priori argumentation. If a concept is revelatory, I say that the concept, or corresponding property, satisfies 12

Maudlin (2004: 30 and 68–70). Maudlin nowhere calls himself a deflationist, but his claims that truth is a logical notion and that once we grasp the truth-function associated with it ‘there is nothing more about it to be known, just as there is nothing more to know about conjunction once one has the truth table ...’ (2004: 70) strongly suggest he is. 13 In particular, I don’t think we should object to using the concept of a logical notion to formulate deflationism on the grounds that there is a controversy over how the logical expressions in a formal language should be defined. The notion of logical expression (or constant) that these theories are trying to define is a technical notion which has a role in formal theories of logical consequence. While this technical notion is obviously related to our pre-theoretical notion, the two are distinct, and it may well turn out that because of the theoretical burden placed on the technical notion, it cannot fully line up with our pre-theoretical notion. But it is the pre-theoretical notion I am relying on, and which Field and Horwich are relying on above. Tarski (1936) also originally relied on the pre-theoretical notion and later tried to give a formal account of it. He said: “Underlying our whole construction is the division of all terms of the language discussed into logical and extra-logical. This division is certainly not quite arbitrary. If, for example, we were to include among the extra-logical sign the implication sign or the universal quantifier, then our definition of the concept of consequence would lead to results which obviously contradict ordinary usage.” (1936: 412) 14 For example, Blackburn and Simmons (1999: 3) and Lynch (2001: 4).

4

Revelation. It should be kept in mind, however, that a property can’t satisfy Revelation simpliciter, but only relative to a particular concept. In other words, there can be both revelatory and non-revelatory concepts for the same property. When a property is revealed to us by our concept of it, there is no more for us to learn about its nature. There is no more to learn either through empirical investigation or a priori reasoning. But that is not to say that grasping a revelatory concept is sufficient to know the nature of the property. It may be that some reflection on C is required, where such reflection falls short of a priori argument or reasoning. This is the point of saying that grasp of the concept only puts us in a position to know the nature of the property.15 According to the ‘no hidden essence’ idea of truth, then, our concept of truth, the everyday concept, is a revelatory concept and so truth satisfies Revelation. If truth is a revelatory concept, this puts it in the company of relatively few other concepts, including, probably, our concepts of the part-whole relation and of conjunction.16 Perhaps, if certain notorious claims are true, then our colour concepts and phenomenal concepts might also be revelatory.17 Before moving on, I should note that my definition of a revelatory concept is stronger than others that could reasonably be offered. One could allow, for example, that a concept is revelatory even if a priori argumentation is required to discover the full nature of the property. On this approach, concepts like table, philosopher, or even justice may be revelatory. For our purposes, however, we need the stronger notion. As I explain below, new wave deflationists require that the nature of truth is not hidden from view in any way, including being hidden at the end of a chain of a priori argument. 2.3. Logical Notions and Revelation New wave deflationists believe that truth is a logical notion that satisfies Revelation. But are both these claims required to delimit the position? If so, are the two claims sufficient to delimit the position? Let’s take the first question first. Revelation alone is clearly not enough to distinguish new wave deflationists from inflationists. On its own, Revelation only tells us that the nature of truth is revealed to us upon reflection on our, everyday, concept of truth. This claim is clearly consistent with the nature of truth being as ‘substantial’ as we like. For although revelation tells us something about the relation between the concept and the property of truth, it tells us nothing about the nature of the property itself until we know the content of the concept, or what it is we know in grasping the concept. That is, revelation is not itself a metaphysical claim about the nature of truth and so is consistent with any claim about the nature of truth. 15

By ‘being in a position to know that p’ I mean roughly what Williamson (2000: 95) means; namely that ‘no obstacle must block one’s path to knowing p. If one is in a position to know p, and one has done what one is in a position to do to decide whether p is true, then one does know p.’ 16 For the claim about the part-whole relation see Lewis (1997: 338 fn 21). For the claim about conjunction see Maudlin (cited in n. 12). 17 Russell (1912: 47) endorsed revelation for colors. The name for the principle is due to Johnston (1992).

5

That said, most views that combined Revelation with the idea that truth is a nonlogical notion would be implausible, since we don’t seem to be in a position to know any substantial theses about truth of this sort merely upon grasping the concept of truth. Nevertheless, even if we rule out these implausible positions, the field remains open to primitivism about truth, as espoused in the early twentieth century by Moore (1899) and Russell (1904). Moore and Russell held that truth (and goodness and the colours) was a primitive property that satisfied Revelation. In fact, according to Russell at this time, propositions are true in the same way he said roses are coloured—they simply are and there is no further account we can give of what makes that the case.18 And this is clearly not a deflationary view. Nothing, neither the traditional debate nor the property of truth, has been deflated: truth has simply been declared beyond analysis. Can new wave deflationism be delimited merely by the claim that truth is a logical notion? Colin McGinn has argued that truth is a logical property, and his discussion strongly suggests that all logical properties ‘are transparent in the sense that their nature is given in the concepts we have of them’ (2000: 84). So perhaps the talk of revelatory concepts is superfluous. I’m uneasy about this idea. It is not clear to me that all logical concepts are revelatory in the way McGinn suggests. Let P be a formal system satisfying the relevant conditions that make a Gödel-style proof of its incompleteness possible. The property of provability in P is a logical property, and yet it took a genius like Gödel to show us that there are sentences of P that are true but not provable within P. This fact is arguably part of the nature of provability within P and yet it is not a fact that we grasp when we grasp the concept of provability within P. So it is plausible to say that provability in P had a hidden nature revealed by Gödel, and so that revelation is not true of this logical property. This general worry can be made more specific. Suppose we discover, after drawn-out and difficult a priori reasoning, that truth is a logical property. Would the resulting theory count as deflationary? A toy example will help. Suppose that after centuries of philosophical analysis we finally realize that truth is a property of sets of propositions and is in fact the same thing as logical consistency. This is implausible, of course, but suppose nevertheless that someone uncovers a knock-down argument for the view (and neatly turns away all prima facie objections). This consistency theory of truth is clearly not a deflationary theory of truth. Yet it would be classed as one if deflationism were defined simply with the claim that truth is a logical notion. Neither the claim that truth is a logical notion, nor the claim that it satisfies revelation is on its own enough to distinguish new wave deflationism from inflationary theories of truth. But how about both together? Obviously the consistency theory just mentioned would not count as a version of new wave deflationism, since it violates revelation. Likewise, more plausible, traditional theories of truth such as the correspondence theory, coherence theory, pragmatic theory and so on will also be ruled out by one or other of the two new wave theses. Functionalist views are also not deflationist, since they hold that the properties that fill the functional role of truth are not given to us along with our grasp of the concept of truth but are discovered, 18

Russell (1904: 75).

6

presumably by a priori reasoning.19 Moore and Russell’s primitivism is also not a version of ‘new wave’ deflationism. Their primitivism was consistent with the idea that we could say something about how truth supervened on other states of affairs, but inconsistent with the idea that the supervenience facts are part of the nature of truth. This idea is no part of the new deflationist view, however, as many deflationists hold that the nature of truth is given by the instances of (PT). PT. The proposition that p is true if and only if p. And the instances of (PT) tell us how the truth of each proposition supervenes on the state of the world. Similarly, a primitivist like Davidson (1996) is not a new wave deflationist, since he denies that truth is a logical notion. (That said, a primitivism only committed to truth’s unanalyzability in non-circular terms, like Patterson (this volume) is consistent with new wave deflationism.) 3. What the Debate is not about Now we’ve tightened up our understanding of new wave deflationism, we can see that a number of debates between deflationists and inflationists are not really to the point. 3.1. Is truth a causal-explanatory property It was once quite common for deflationism to be described as the view that truth is not a property. More recent commentators have characterized deflationism as the view that truth is not a genuine property, or not a property with a nature.20 Neither of these characterizations is very helpful, however, since they would automatically count property nominalists as deflationists about truth. Once we formulate deflationism with the help of Revelation, we can see why debates about whether truth is a property or has a nature are neither here nor there. While this point is sometimes conceded, it is often replaced with the idea that deflationists must hold that truth is not a causal-explanatory property.21 This view stems from Field (1972) in which he argued that if truth plays some causalexplanatory role then it must be susceptible to physicalistic reduction. One place truth may have such a role is in explanations of our successes in getting what we want. That is, often it is because we have some true belief about how to get what we want that we succeed in satisfying our desires. Another important theoretical, and perhaps therefore causal-explanatory, role for truth is in theories of meaning and content, and in particular, truth-conditional theories of meaning and content. As a result, a large portion of the debate about deflationism has turned on these two issues—whether truth plays a causal role in bringing about the satisfaction of our desires, and whether truth plays an essential role in the correct theory of meaning. However, whether or not truth plays a causal-explanatory role does not seem to be a 19

Lynch (2009). Lynch (2001: 4) and Vision (2004: 8) portray deflationists as denying truth has a nature and Engel (2002: 12) says deflationists deny that truth is a genuine property. 21 For the most prominent admission that deflationism is consistent with truth being a property see Horwich (1998b: 36–8); for a recent reaffirmation by a critic, see Lynch (2009: 106). Field (2001: 29), Horwich (1998b: 38) and Devitt (2001) argue that deflationism is committed to denying that truth is a causal-explanatory property. 20

7

basic commitment of deflationists about truth, however commonly deflationists endorse the thesis. Indeed, as I’ve argued elsewhere (Damnjanovic 2005), whether or not truth plays a causal-explanatory role depends heavily on basic, and controversial, issues in the theory of causation, since on some views even logical properties are capable of being ‘causally relevant’. In short, only if it is combined with controversial views about causal explanation does deflationism imply that truth can’t play a causalexplanatory role. The more fundamental debate is about whether truth is a logical notion (which will require, inter alia, providing a more detailed account of what logical notions are). 3.2. Instances of (PT) have a special status So far I have intentionally downplayed the importance of various equivalence schemata, such as (PT), to the deflationist position.22 I have done this because I think that new wave deflationism is less wedded to these schemata than its ancestors may have been, and that overlooking this leads to a distorted view of what divides deflationist from inflationist. Sometimes, for example, it is said that deflationists grant instances of (PT) some sort of special epistemic or other status, such as that the instances of (PT) are analytic, concept-constituting, a priori, necessary and/or trivial. Usually when these claims are made it is implied that inflationists do not grant (PT) such a special status. But in fact none of these claims on its own mark out a deflationist position. To see this, consider the strongest (albeit implausible) such claim, namely that the two sides of (PT) are synonymous. Now an inflationist may claim to have discovered another equivalence involving truth, which gives some further insight into the nature of truth. One such example might be (3). 3. The proposition that p is true if and only if the proposition that p corresponds to some fact. Now, however plausible such a claim is, it is open to an inflationist to maintain that the two sides of (3) (or of their preferred alternative analysis) are synonymous. And the thesis that instances of (PT) also express claims of synonymy doesn’t rule this out, but only implies that the right hand sides of (PT) and (3) are also synonymous. So, even the very strong view that the two sides of instances of (PT) are synonymous is consistent with inflationism. And since the strongest version of the thesis that instances have a ‘special status’ leaves room for inflationism, the weaker, and more plausible, versions do too. So it is hopeless to try to characterize deflationism as the view that (PT) has some sort of special status.23 3.3 All Instances of (PT) are True While inflationists might grant (PT) a special status, a typical worry about deflationism is that it is committed to (PT) having some sort of special status, while 22

The other schemata I have ignored are those applying to other truth bearers, including (ST). ST. The sentence ‘s’ is true if and only if s. 23 For this reason we should not accept Soames’s recent formulation of deflationism as a commitment to the views ‘i) that p and the proposition that p is true are trivial, necessary and a priori consequences of one another, and ii) that any warrant for asserting, believing … (or taking any of a variety of related attitudes) toward one of those propositions is a warrant for asserting, believing … (or taking the relevant related attitude) toward the other’ (2003: 372).

8

inflationists need not be. The worry takes different forms depending on what view is being attributed to the deflationist, but one common attribution is the idea that they are thereby committed to the truth of all instances of (PT). For example, this assumption is a central plank in the argument that deflationists must deny the existence of bivalence failures, and reject all forms of nonfactualism.24 Yet some deflationists explicitly deny that all instances of (PT) are true, particularly paradoxical and pathological instances.25 Nor can a general commitment to the truth of all instances of (PT) be immediately inferred from the claim that instances of (PT) are analytic, necessary, concept-constituting, etc.26 After all, it is possible to say that only some specified sub-set of the instances of (PT) are analytic, necessary, conceptconstituting … . And on the surface, there is nothing about truth being a logical notion which satisfies revelation that is inconsistent with this. If there is a problem here, it must be that the deflationist does not have the theoretical resources to distinguish those instances of (PT) that are (logically) true from those that aren’t. In other words, the worry must be that if a deflationist tries to specify the relevant instances of (PT) they will inevitably appeal to resources that inflate their theory of truth. But it is hard to see why this should be so. Why can’t a deflationist distinguish the appropriate instances as those which are genuinely about propositions—and providing an account of what it takes for a sentence to express a proposition—or about sentences with a certain sort of meaning? Is it because deflationism is incompatible with any theory of meaning rich enough to make these distinctions? If so, this needs to be argued. More importantly, it is clear that the fundamental question concerns what theories of meaning deflationism is compatible with. I return to this question in the final section. It is also worth pointing out that whether or not deflationists are committed to the truth of all instances of (PT), they are not committed, qua new wave deflationist, to the view that those who grasp the concept of truth are disposed to accept all instances of (PT) (as Horwich (1998b: 121) claims). In particular, it might be thought that Revelation somehow implies this, since if the instances of (PT) tell us part of the nature of truth, then grasping the concept of truth must put us in touch with all those facts. However, it seems plain that many of us who possess the concept of truth don’t believe all instances of (PT) for the simple reason that we are finite creatures and there will be propositions that are too complex or too complicated for us to grasp, or which employ concepts we don’t as a matter of fact possess.27 Nor can it be responded that even if we don’t understand certain instances of (PT), we are nevertheless disposed to accept them. I might not accept (4), for example, for fear that it is 24

Perhaps the first person to raise the problem was Dummett (1959). See also Boghossian (1990). Horwich (1998b) makes the restriction to non-pathological instances, and Field (1994) restricts the claim to those instances we understand. 26 Contra McGee: ‘For the disquotationalist repudiating bivalence is not an option, for the principle is built into the very meaning of the word “true.” [(T)-sentences are] … on the disquotational account, not merely true, but, if I may use the word, analytic’ (2005: 77-8). 27 As Anil Gupta points out, if we did need to believe all instances of (PT) in order to grasp the concept of truth, we couldn’t grasp the concept of truth until we had grasped all other possible concepts. Gupta (1993, p. 297-8). See also Kunne (2003: 324). 25

9

nonsense. 4. The proposition that a quantum of quarks is worth two in the bush is true iff a quantum of quarks is worth two in the bush. Regardless of whether this instance is meaningful or not, I may not accept it merely because I am not sure that it is, and so not sure that it is a legitimate instance of (PT). Finally, as Mark Richard (1997) has stressed, non-factualists about moral statements, for example, surely possess the concept of truth, yet they may refuse to accept certain instances of (PT), such as (5) 5. The proposition that gluttony is wrong is true if and only if gluttony is wrong. since they deny that assertions of gluttony is wrong are capable of being true or false. While these problems are real, they are all particular versions of the problems facing any dispositional account of concept possession. For most concepts, it is difficult to find any specific statements and/or inferences such that being disposed to accept them is necessary and sufficient to qualify as a possessor of the concept. Indeed Kripke (1982) has given powerful arguments for the conclusion that any sort of dispositional account of concept possession is bound to fail. So, while the problems I have raised are difficult, they are only problems for those, like Horwich, who hold a dispositional account of the conditions for possessing the concept of truth. That is, Revelation only implies that grasp of the concept of truth entails a disposition to accept all instances of (PT) if i) all instances of (PT) are true and ii) a dispositional account of concept possession is correct. 3.4. The Instances of (PT) Tell Us Everything About Truth If deflationists aren’t committed to the truth of all instances of (PT), then perhaps they are at least committed to the idea that (some sub-set of) instances of (PT) exhaust the true theory of truth. One way this idea has been developed is to hold that the instances of (PT) exhaust the content of the concept of truth (Collins 2002). This idea should be treated with some scepticism, however, since presumably some part of the content of the concept of truth explains why some instances of (PT) are paradoxical or pathological.28 And this suggests that there is more to our grasp of the concept than those instances themselves tell us. Moreover, the claims that truth is a logical property that satisfies revelation does not imply that the content of the concept of truth is exhausted by instances of (PT). In particular, while no deflationist I know of has taken this option, new wave deflationism allows for accounts of the content of the concept of truth that go beyond the instances of (PT)—though presumably the instances of (PT) will play a central role in any correct account of truth. Perhaps, then, it is not the concept which is exhausted by the instances of (PT), but the nature of the property (Horwich (1998b); Williams (2002: 153)). But while some new wave deflationists have said this, it is not clear that they must. True, if the concept were so exhausted, then Revelation would imply that the nature of the property is likewise exhausted. But as we have just seen, the new wave deflationist 28

See Soames (1999: 247), and Gupta (2006: 135-9)

1 0

need not, and should not, hold that there is no more to the content of the concept of truth than is told to us by (PT). 4. What the Debate Might Be About If I am right so far, then many of the familiar skirmishes between inflationists and deflationists are, at best, only indirectly relevant to the real debate. The one issue which really seems pressing is the question concerning what the deflationist can say about the nature of meaning. I’d like to finish by saying a few (all-too-few) words about this issue, and suggest that here, too, the differences between inflationists and deflationists might be more apparent than real. It is sometimes said that deflationism implies that truth-conditional theories of meaning are false. One reason for thinking that deflationists cannot endorse truthconditional theories of meaning is that doing so would imply that truth is a causalexplanatory property. As I argued, however, this on its own is not a crucial concern for the new wave deflationist (or at least should not be). But there are other reasons to worry about the claim too. Soames (1989) has argued that we should separate semantic theories from theories of semantic competence. Yet if semantic theories are not required to explain language users’ abilities, then what reason is there to think of them as causal-explanatory theories at all? Also, Williams (1999) and Kölbel (2001) have argued that we should separate the expressive role the truth predicate may play in semantic theories from any explanatory role it may play in such theories. 29 That is, it may be that ‘true’ plays merely its familiar, logical role when it appears in truthconditional theories of meaning. A different reason for thinking that deflationism is incompatible with truthconditional theories of meaning is that the combination would be circular. This worry primarily applies to sentential deflationism. Consider instances of (ST): ST. The sentence ‘s’ is true iff s. As Horwich (1998b: 68) says, if instances of (ST) are both definitional of sentential truth and used to give the meanings of the sentences in the language, then they are like ‘equations with two unknowns’ and so unsolvable. In other words, we are trying to define meaning in terms of truth, and also truth in terms of meaning. This argument is hard to evaluate because it is unclear in what way the sentential deflationist is being said to take instances of (ST) to be ‘definitional’. Even the most radical of sentential deflationists, Hartry Field (2001: 152), holds that the basic definitional claim about sentential truth is something like (STT): STT. If ‘s’ is translatable as ‘p’, then necessarily ‘s’ is true iff p. Of course, this makes the instances of (ST) trivial, but then why is there a special problem for the deflationist here? Won’t the inflationist likewise accept that instances of (ST) are trivial? 30 A similar worry arises concerning the idea that deflationists cannot say anything substantial about meaning at all. To see this, consider the following, trivial, ‘meaningto-truth’ conditional 29

But see Horisk (2007). For discussion see Bar-On et al (2005). For criticism of the circularity charge see Gupta (1993); (Lance 1997); Williams (1999); Horisk (2008). 30

1 1

If sentence ‘s’ means that p, then ‘s’ is true if and only if p. It may seem that anything substantial one says about meaning will transfer over this conditional to inflate the theory of truth. In short, it may seem that a substantial account of meaning implies that whether or not a sentence has truth-conditions is a substantial matter, and so inconsistent with deflationism.31 But, again, it is far from clear that this is so. Suppose we invent a predicate which is a purely disquotational device. Call this the ‘struth’ predicate. There seems to be a valid meaning-to-struth conditional too. (i.e. If ‘s’ means that p, then ‘s’ is strue if and only if p). But surely the fact that there is such a disquotational predicate shouldn’t have serious consequences for the theory of meaning. Why then should the deflationist claims that truth is a logical notion that satisfies revelation have such implications? It is unclear, to say the least, what the real commitments of deflationism are in these respects. 5. Conclusion Even when it comes to their commitments in the theory of meaning, the differences between inflationists and deflationists may turn out to be minimal. Perhaps more interesting, however, is simply that the main debate between inflationists and deflationists seems largely to reduce to a debate about the nature of meaning. This is interesting, since one would think that the real debate would be focused on the nature of truth. Instead, it seems that, at best, there is fundamental debate about meaning, and the debate about truth has become strangely epiphenomenal. If deflationism is the view that traditional debates between inflationists about truth should be deflated, then perhaps the real conclusion might be metadeflationism: the view that the contemporary debate between deflationists and inflationists should also be deflated.32 References Ayer, A.J. (1936/1946). Language, Truth and Logic, second edition. London: Gollancz. Bar-On, D., Horisk, C. and Lycan, W. (2005). ‘Deflationism, Meaning and TruthConditions’ and ‘Postscript to ‘Deflationism, Meaning and Truth-Conditions’ in B. Armour-Garb and J.C. Beall (eds.) Deflationary Truth (321–352), Chicago and La Salle: Open Court Press. Blackburn, S. and Simmons, K. (eds) (1999). Truth. Oxford and New York: Oxford University Press. Boghossian, P.A. (1990). The Status of Content. Philosophical Review, 99: 157–184. Collins, J. (2002). On the Proposed Exhaustion of Truth. Dialogue: Canadian Philosophical Review, 41 (4): 653–679. Deviit, M. (2001). The Metaphysics of Truth. In M. Lynch (ed.), The Nature of Truth (579–611). Cambridge, MA and London: MIT Press. 31

Eklund (this volume) distinguishes the question of the compatibility of deflationism with certain theories of meaning from the question concerning the compatibility of rejectionism with those theories, where rejectionism is the view that ‘true’ is merely an expressive device. I think new wave deflationism is closer to rejectionism than deflationism as Eklund understands it. 32 For helpful discussions and/or written comments on various versions of this paper I am very grateful to Stewart Candlish, Martin Davies, Michael Devitt, Matti Eklund, Michael Lynch, Dan Marshall, Greg Restall, Stewart Saunders, and Daniel Stoljar. Special thanks are also due to Nikolaj Pederson and Cory Wright for helpful comments and for bringing this volume into existence.

1 2

Damnjanovic, N. (2005). Deflationism and the Success Argument. Philosophical Quarterly, 55: 53–67. Davidson, D. (1996) The Folly of Trying to Define Truth. Journal of Philosophy, 93: 263–78. Dummett, M. (1959/1978). In his Truth and Other Enigmas (1–24). London: Duckworth. Engel, P. (2002). Truth. Chesham: Acumen. Field, H. (1972). Tarski’s Theory of Truth. Journal of Philosophy, 69: 347–375. Field, H. (1992). Critical notice: Paul Horwich’s Truth. Philosophy of Science, 59: 321–330. Field, H. (1994). Deflationist Views of Meaning and Content. Mind, 103: 249–284. Field, H. (2001). Truth and the Absence of Fact. Oxford: Clarendon Press. Gupta, A. (1993). A Critique of Deflationism. Philosophical Topics, 21: 57–81. Gupta, A. (2006). Do the Paradoxes Pose a Special Problem for Deflationists. In J.C. Beall and B. Armour-Garb (eds.), Deflationism and Paradox (133–147), Oxford: Clarendon Press. Hill, C. (2002). Thought and World: An Austere Portrayal of Truth, Reference, and Semantic Correspondence. Cambridge: Cambridge University Press. Horisk, C. (2007). The Expressive Role of Truth in Truth-Conditional Semantics. Philosophical Quarterly, 57: 535–557. Horisk, C. (2008). Truth, Meaning, and Circularity. Philosophical Studies, 137: 269– 300. Horwich, P. (1998a). Meaning. Oxford: Clarendon Press. Horwich, P. (1998b). Truth. Oxford: Clarendon Press. Johnston, M. (1992). How to Speak of the Colours. Philosophical Studies, 68: 221– 263. Kölbel, M. (2001). Two Dogmas of Davidsonian Semantics. Journal of Philosophy, 98: 613–635 Kripke, S. (1982). Wittgenstein on Rules and Private Language. Oxford: Basil Blackwell. Künne, W. (2003). Conceptions of Truth. Oxford: Clarendon Press. Lance, M. (1997). ‘The Significance of Anaphoric Theories of Truth and Reference’, in E. Villanueva, (ed.), Philosophical Issues 8: Truth (181–198). Atascedero, Calif.: Ridgeview Publishing Company. Lewis, D. (1997). Naming the Colours. Australasian Journal of Philosophy, 75: 325– 342. Lynch, M. (2001). Introduction. In M. Lynch (ed.) The Nature of Truth (1–6). Cambridge, MA and London: MIT Press. Lynch, M. (2009). Truth as One and Many. Oxford: Oxford University Press. Maudlin, T. (2004). Truth and Paradox: Solving the Riddles. Oxford: Clarendon Press. McGee, V. (2005). Two Conceptions of Truth? – Comment’. Philosophical Studies, 124: 71–104. McGinn, C. (2000). Logical Properties: Identity, Existence, Predication, Necessity, Truth. Oxford: Clarendon Press. Moore, G.E. (1899/1993). The Nature of Judgment. In T. Baldwin (ed.), G.E. Moore: Selected Writings (1–19). London: Routledge. Richard, M. (1997). Deflating Truth. In E. Villanueva (ed.), Philosophical Issues, 8, (57–78). Atascedero, Calif.: Ridgeview Publishing Company. Russell, B. (1904/1974). Meinong’s Theory of Complexes and Assumptions. In D.

1 3

Lackey (ed.), Essays in Analysis (21–76). New York; Braziller. Russell, B. (1912/1946). The Problems of Philosophy. London: Oxford University Press. Soames, S. (1989). Semantics and Semantic Competence. Philosophical Perspectives, 3: 575-96. Soames, S. (1999): Understanding Truth, Oxford: Oxford University Press. Soames, S. (2003). Understanding Deflationism. Philosophical Perspectives, 17, Language and Philosophical Linguistics: 369–383. Tarski, A. (1936/1983). On the Concept of Logical Consequence. In his Logic, Semantics, Meta-mathematics (409–420). Indianapolis, Indiana: Hackett Publishing Company. Vision, G. (2004). Veritas: The Correspondence Theory and Its Critics. Cambridge, MA: MIT Press. Williams, M. (1986). Do We (Epistemologists) Need a Theory of Truth?’ Philosophical Topics, 14: 223–242. Williams, M. (1999). Meaning and Deflationary Truth. Journal of Philosophy, 96: 545–564. Williams, M. (2002) On Some Critics of Deflationism. In R. Schantz (ed.), What is Truth? (146–158). New York: Walter de Gruyter. Williamson, T. (2000). Knowledge and Its Limits. Oxford: Oxford University Press.

1 4