1. The problem of the unity of truth

1Is truth one or many? In other words, is there a single and unique property of being true which applies to all truths whatsoever, or are there different kinds of truth? Our intuitions go both ways. On the one hand if truth is truth it should not be truth-P for P-things or P- discourse, truth-R for R-things, etc. On the other hand we want to account for the fact that there are different kinds of things and properties to which truth applies. As Blackburn [Blackburn 1984, 230] remarks, we know individually what makes ‘[it is true]’ applicable to judgments or sentences of an understood language. For instance “Penguins waddle” is true if and only if penguins waddle, “snow is white” is true if and only if snow is white. The reason the first sentence deserves the predicate is that penguins waddle. The reason the judgment that snow is white deserves the predicate is that snow is white. But these reasons are entirely different. There is no single account in virtue of which each deserves the predicate, for deciding whether penguins waddle has nothing much in common with deciding whether snow is white. There are as many different things to do to decide that the predicate applies, as there are judgments to make. So how can there be a unified “common” account of the property which these different decision procedures supposedly determine? What is common to truths? What is common to moral, physical, mathematical, normative truths? And about aesthetic matters, about the comic? etc.

2In general there are four types of solutions to this problem. a) A first solution is to deny that all discourses are truth apt (e. g. expressivism about moral discourse, or fictionalism about mathematical truths). Let us call this truth chauvinism. b) Another solution is to accept that they are all true alike. Let us call this truth monism. c) A third solution is to say that not all truths are true alike. This is called alethic pluralism. d) There is a fourth solution, which is associated with a particular view of truth. According to the deflationist conception of truth, all truth are alike—and in this sense this conception resembles truth monism, but they are not alike in a substantial way. Truth is not a “heavy weight property”: it is not a property such as being square, or being made of wood, or even like being red. It is a “thin” or “light” property. In another sense, however, the deflationist accepts the disunity of truth. The deflationist bites the bullet. Truth is not a common property that our sentences or judgments have. It is only a syntactical device by which we can form an infinite disjunction of sentences of the form

“P” is true iff P
or
it is true that P iff P.

3So if you talk of penguins you will say that it’s true that they waddle iff they waddle, if you talk of cabbages it’s true that they are tasty iff they are tasty, if you talk of kings it’s true that they are polite iff they are polite, etc. [Blackburn & Simons 2000]. So a form of pluralism is the immediate consequence of deflationism: there are as many truths as there are kinds of discourses and kinds of things to talk about. Truth is equally applicable to each discourse: truth is fiction, in mathematics, in ethics, etc. No problem with that, for truth is purely an expressive device, lightweight and with no ontological consequence. Another consequence of deflationism is quietism: realism disputes vanish. In so far as realism is signaled by the applicability of the predicate “true” to a given domain, realism is applicable everywhere. But this is just to say that it is applicable nowhere, for if everything is truth-valuable, hence real, then nothing is.

2. Wright’s minimalism

4Wright’s minimalism [Wright 1993] seems to illustrate these commitments of deflationism. Although Wright claims that he is not a deflationist about truth, the problem of the unity and plurality of truth is particularly salient for him. First Wright defends syntacticalism about truth, the thesis that “true” is a predicate, although a “lightweight” one, which satisfies only formal or syntactic features: it obeys the equivalence or the disquotational schema, is such that statements which are apt for truth have negations which are likewise, that they can be embedded in conditional and propositional attitudes constructions. The minimalist in Wright’s sense also claims that truth satisfies a set of platitudes:

- To assert a statement is to present it as true (transparency)

- Truth is correspondence to the facts (correspondence)

- A statement may be justified without being true and vice versa (contrast)

- Truth is absolute and has no degrees (absoluteness)

- Truth is timeless (timelessness)

- Truth is objective, and implies convergence (convergence).

6An important consequence of Wright’s minimalism is that the issue of the nature of truth and the issue of realism are not systematically linked. A T-predicate, satisfying the syntactic and formal constraints, applies to a number of discourses: ethical, mathematical, etc. It will not be uniform across all discourses. There is no reason to expect that truth about the physical world will be the same as truth about mathematical entities, and that both will be the same as truth about moral matters. On Wright’s view, it may well be that truth could be a form of correspondence with independent states of affairs about the physical world, whereas it is closer to superassertibility in mathematical or moral matters. It may also be different in the domain of the comic. Some discourses may be more apt for truth than others, even though truth applies to a variety of discourses. But how are we to frame the realist-anti-realist disputes? Wright proposes the familiar Euthyphro contrast: that an act is pious if and only if it is loved by the gods, can be understood a) in the anti-realist or response dependent sense: pious acts are such because they are loved by the gods, or b) in the realist sense: it is because some acts are pious that the gods love them. Generalising, there are certain discourses which are true because P's best opinion has it that P, whereas others are such that best opinion has it that P because P. How do we know that we are in one case or the other? There are some a priori requirements for a discourse to be truth-apt or not. One is cognitive command, the other is width of cosmological role. Cognitive command is the feature instantiated when observers disagree on a feature of X, this can be traced to some shortcoming of their cognitive apparatus, because the feature is reliably tracked, in the usual cases, by this apparatus. Width of cosmological role is the feature instantiated when a subject matter can best be explained by independent states of affairs.

7Now, given that these a priori requirements are not uniform, pluralism about truth follows:

One who takes a minimalist line about truth, that the concept is fixed by a number of platitudes about it, ought to be open to the possibility…that it may have a variety of models…This potential plurality reflects the distinctions that are relevant to the realism/antirealism debate: that the justification of a broadly realist or antirealist view turns on the character of the local realist-antirealist debate. [Wright, C. 1996, PPR 4, 923].

8Officially, Wright’s minimalism is neither a realism nor an anti-realism about truth, since truth is not uniform across discourses. But Wright’s actual view may be called a minimal anti-realism, because “the general rule is that realism has to be earned” and that “anti-realism is the default position” [Wright 1992, 149]. One starts from an anti-realist predicate, where truth is response dependent, and one sees from there whether the domain is more or less shaped by a stronger notion of truth, such as superassertibility and correspondence. In most cases, superassertibility is the proper model of truth (in particular ethics), in other domains (in particular the comic) it is not.

9Wright’s view, however, raises the problem of unity again. As David Wiggins [Wiggins 2002] puts it, any predicate, say φ, is a truth predicate if, for every sentence S of language, or any proposition expressible in a given language, we have: φ (S) iff S (where “(S)” is the name of a sentence of the language and the rightmost “S” is the sentence itself). The question then is: how can we fix the meaning of φ without presupposing that it denotes truth itself?

10The main difficulty for truth pluralism has been raised by Williamson [Williamson 1994], see also [Tappolet 1997]. It is the problem of mixed inferences. If one infers (3) from (1) and (2):

(1) Waddling penguins are amusing.

(2) That penguin waddles.

(3) Therefore that penguin is amusing.

11The inference cannot go through, since (1) is a comic truth, whereas (2) is a truth about penguins. Because the word “true” does not have the same meaning is (2) and in (1) and (3) respectively, the inference is not valid. The same problem affects generalizations about truth, such as:

(4) “Everything you said is true”.

12But if you said things about ethics, mathematics, law, the comic and the physical world, then we can’t generalize with (4). What then can the criterion of truth-aptitude be? We encounter a dilemma. On the one hand, if the criterion of truth-aptness is but the capacity of a predicate for each domain to behave “syntactically” or formally as a truth predicate, then in this sense all discourses all equal, and truth-aptness is a well-shared property. If, on the other hand, truth aptitude is a further issue which can only be settled locally, how can it make for different “kinds of truths” in each respective domain? For if the different kinds are kinds of truths, how can truth be thought of as uniform in the first place? The answer given by Wright to the problem of mixed inferences shows the proximity of minimalism to deflationism [Wright 1994]: in the penguin inference (1)-(3), the premises need only satisfy the platitudes. So the predicate “true” is not ambiguous, and it is lightweight. The problem with this answer, as Gerald Vision [Vision 2004, 102] has remarked, is that either the platitudes yield the concept of truth we are after, or they don’t. If they do, this amounts to deflationism. If they don’t this does not solve the mixed inference difficulty.

3. Truth functionalism

13A solution, functionalism about truth, has been proposed by Pettit [Pettit 1996] and Lynch [Lynch 2001]. It seems to have been partially endorsed by Wright himself [Wright 1996]. Just as, according to functionalism about mental states, our use of “belief”, “desire” and other mental terms, can be individuated in terms of the role that they play, together with other states, in mediating between inputs and outputs, we could say that the predicate “true” is a place mark for a certain role marked by the usual platitudes: asserting statements that one believes, which correspond to reality, on which people can converge, etc. But the issue of the nature of the properties which “realize” these roles is left open. Truth is a property which can be variably realized, just as a functional property can be so. Summarizing the idea, Wright says:

The concept of truth admits a uniform characterisation wherever it is applied — the characterisation given by the minimal platitudes, which determine what is essential to truth… The form of pluralism for which space is allowed by this overarching uniformity is variable realization. What constitutes the existence of a number is different from what constitutes the existence of a material object. [Wright 1996]

14In other words truth is a second-order property of our statements, which has to be realized in various ways in first order properties which will underlie this role. In each case we shall give arguments to secure that the realist story is correct or not. The idea is close to one which had been advanced by David Wiggins in his theory of identity [Wiggins 1980]. Identity is the one and only relation captured by the principles of reflexivity, symmetry, transitivity, substitutivity and the identity of indiscernibles, but there are varying conditions of identity for material objects, persons, events, etc. Similarly truth is a multiply realizable role-property, which we can “ramsify” with a Ramsey sentence:

x has a property that plays the truth-role if and only if
∃t₁[A(t₁ …O₁ …O_n) & x has t₁],
where “A” is the conjunction of our intuitive beliefs about truth.

15This is an appealing view. But truth functionalism is, like Wright’s minimalism, unstable [Wright 2005], [Lynch 2006]. First, on truth functionalism, is truth a role property or is it a realizer property: is truth the higher order property in which various realizer properties are realized (truth about ethics, comedy, maths, etc. ) or is it the realizer property itself? If truth is the realizer property, it is not a general property at all. Second, is truth a higher order property or a concept? If truth is a higher order property of properties multiply realized how can it be the common property of all these different properties? The alternative would be to say that the concept of truth (captured by the platitudes) is common, but that the properties are differently realized. The problem then is that the analogy with functionalism breaks down. Either way alethic functionalism ressembles deflationism, and we are back to the difficult already encountered with Wright’s minimalism. If there is no common property of truth which various domains realize, but only a concept associated with various platitudes, then the view comes close to deflationism.

4. Minimal realism

16Some writers have proposed a view related to truth functionalism. Alston [Alston 1996] defends what he calls alethic realism, and Lynch [Lynch 2001] what he calls alethic pluralism. I have myself defended a related view which I called minimal realism [Engel 2002]. Minimal realism incorporates four minimalist commitments:

(1) it agrees with minimalism on the fact that truth is a “thin” notion satisfying the discipline of syntax and the associated platitudes about assertion, correspondence, convergence, etc;

(2) it rejects, however, the thesis that truth is a mere logical device of assertion or of disquotation ; truth registers a distinctive norm;

(3) it takes truth-bearers to be propositions, or the contents of beliefs, and assumes that we need to have an independent account of these contents;

(4) it is not, however, pluralistic, since it does not take the truth predicate to be ambiguous with respect to different domains ; truth has a uniform core-meaning defined by its role but which is realized in different ways from domain to domain.

17Minimal realism is also a form of realism because it says that:

(5) the uniformity of the truth-predicate does not neutralize the issues about realism and anti-realism which arise from domain to domain ; a minimalism about truth does not imply a minimalism about truth-aptness;

(6) in each domain, truth-aptness is to be judged after the realist criterion of the independence of a domain from our responses, and of verification transcendence: our best conceptions might be false;

(7) in each domain, realistic truth, in the sense of (6) is the norm of our inquiries.

18Minimal realism, however, has the same problem as truth functionalism, since (4) leads to the same problem as truth functionalism.

5. The norm of truth and extended realism

19None of the versions of functionalism about truth is satisfactory. The problem does not come simply from the ambiguous status of truth as both a higher-order and a realizer property, but it comes also from the conception of the norm of truth that is presupposed by truth functionalism. On Wright’s picture, the content of this norm is a special form of warranted assertibility: superassertibility. A discourse obeys the norm of truth if it carries with it the requirements of syntactic discipline and if it aims at superassertability as a model for truth. But, as we saw above, the discourse does not need, on his view, to carry over to realist truth, although it can, if realism happens to be justified within a specific domain. Anti-realism is the default option. But what if instead we took realism as the default option? What if, instead of truth, we took the norm of assertion to be knowledge? The result would be, in my view, a much more unified conception of the truth property. But that would also lead us to a rejection of pluralism about truth. For Wright superassertibility is at least a necessary condition for knowledge:

(KS) If p is knowable, then p is superassertible.

20Note that the converse is not true: superassertibility is not a sufficient condition for knowledge [Wright 1993, 58]. What KS involves is that assertion implies warranted assertibility in a strong enough sense to imply its objectivity, although in a weak enough sense to imply that it is not knowledge (superassertibility). Let us agree with Wright that truth is the norm of belief, which is stronger than warranted assertibility. If so, its being a norm of assertion in a general, substantial, property of truth itself. But where the minimalist, the pluralist and the functionalist about truth go wrong is that they do not acknowledge that standards of assertion carry more weight than truth pluralism allows. Although I shall not defend this view here, and take it for granted, there are good reasons to hold [Williamson 2000], [Engel 2002, 129] the knowledge account of assertion according to which assertion is governed by the norm of knowledge:

KAA Assert p only if you know that p.

21On this view, an assertion is warranted only if the speaker knows that P. Therefore the criterion of truth aptitude is the capacity to know a proposition in the respective domains. Let us apply this to the issue of pluralism. We shall say that:

If a discourse permits warranted assertions then it permits the expression of knowledge

22This applies to mathematical truth, to moral truth, to aesthetic truth, etc. This extends to the comic: if I do not know that Chaplin is funny then I cannot assert that Chaplin is funny.

23But does KAA entail realism? No. To say that a discourse permits the expression of knowledge is not immediately to vindicate realism. A minimal anti-realist notion of truth can be associated to a minimal anti-realist notion of knowledge. But knowledge entails realism. To know about a subject matter is to know about a domain to which realist truth applies. If this is so, we can reformulate Wright’s notion of cognitive command, in a stronger way. The general criterion of knowledge is safety, in the sense that if one knows that P, one is reliable about P and one cannot easily be wrong about P [Sosa 1999], [Williamson 2000], [Engel 2007].

24The condition of warrant to assert P applies to a range of cases only if the truth condition of P applies to a wider range of cases. (This is an application of Williamson’s Margins or error principle). Such a condition of knowledge on assertion is stronger than cognitive command.