1 The lay of the land: pluralism and deflationism

It has been 30 years since the publication of Crispin Wright’s Truth and Objectivity (1992), in which Wright argued for what is now known as alethic pluralism, i.e., pluralism about truth. Though the view has gained quite some traction, and many different versions of alethic pluralism have been developed in the last decades, it seems that there is still a lot of resistance to the idea that truth could be plural.

In his book, Wright identifies the deflationist as the pluralist’s greatest rival. He devotes the entire first chapter to an argument against deflationism. Deflationism has been seen as the most dangerous threat to pluralism by other pluralists as well, such as Michael Lynch and Douglas Edwards. Lynch also devotes a chapter of his book Truth as One and Many (2009) to the development of an argument against deflationism, and so does Edwards in The Metaphysics of Truth (2018). Edwards says that the main aim of his book, besides the development of his pluralist theory, is to legitimize inquiry into the nature of truth and to thereby counter “deflationist and primitivist movements”. (Edwards, 2018, p. 3) I shall leave primitivism aside in this paper, and I focus solely on the debate between deflationists and pluralists about truth.

The deflationist scoffs at the pluralist’s attempt to divide and conquer when it comes to answering questions about the nature of truth. They urge them to leave those questions be, and to rest content with a description of the way we use truth in our linguistic practices; we can find out what truth is like by exploring what use we make of it, without having to inquire into its metaphysical constitution. An early example of this approach is P. F. Strawson’s non-descriptive theory of truth:

[T]he phrase ‘is true’ is not descriptive at all. If we persist that it describes (is about) something, while denying that it describes (is about) sentences, we shall be left with the old, general questions about the nature of, and tests for, truth, about the nature of the entities related by the truth-relation, and so on. Better than asking “What is the criterion of truth?” is to ask: “What are the grounds for agreement?” — for those we see to be not less various than the subjects on which an agreed opinion can be reached. And this will perhaps also discourage us from seeking to mark the difference between one kind of utterance and another by saying, for example, “Ethical utterances are not true or false”. It is correct to say that utterances of any kind are true or false, if it is correct usage to signify agreement or disagreement with such utterances by means of the expressions ‘true’ or ‘false’. (Strawson, 1949, p. 94)

Strawson argues that we primarily use a phrase like ‘is true’ to express our agreement, for instance when we say “What Strawson said is true”. Other deflationists (e.g., Horwich, 1998) agree with Strawson’s analysis of truth but they do not reject the existence of a truth property. What they reject instead is that an appeal to the metaphysics of this property could explain, or is needed to explain, how we use truth in the way we do. They believe that truth is metaphysically uninteresting, and that its nature is fully captured by the following platitude, the equivalence schema: < p > is true if and only if p.Footnote 1 They thus agree with Strawson that the inquiry into truth’s nature is futile.

Pluralists would of course disagree. While they might accept the deflationist’s starting point of taking the use of the concept of truth, and phrases like ‘is true’, as their primary explanatory objective, they would reject the claim that an inquiry into truth’s nature would be futile for trying to figure out why we use truth in the way we do.

For example, in “Truth: A Traditional Debate Reviewed” (1998), Wright argues, in agreement with the deflationist, that the question “What is truth?” cannot be answered directly by answering the question “What does truth consist in?” Instead, he articulates platitudes about truth, such as the platitude that “a proposition may be true without being justified, and vice-versa”, which help identify the target concept (Wright, 1998, p. 60) This allows us to give a network analysis of the concept of truth. Lynch (2009) gives a functionalist twist to this story: the platitudes characterize the role of truth. They show how truth functions in our discursive practices. This kind of analysis of truth is very congenial to deflationism.

However, pluralists like Lynch argue that different truth properties are needed in different domains of discourse for truth to play its role. This is because they believe that truth does not merely function as a device for expressing agreement, it functions as a standard of correctness and as the goal of our inquiry, and for it to serve this purpose, truth needs to be more than what is conveyed by the equivalence schema. For example, the correctness of moral propositions might be epistemically constrained, so that truth could consist in some kind of indefeasible justification, whereas the correctness of scientific propositions might be evidence-transcendent, in which case truth could be better thought of as consisting in some relation of correspondence between a proposition and mind-independent reality.

Edwards’s determination pluralism, for instance, takes truth to be analogous with winning a game. Yet, what determines when a game is won differs per game and, likewise, what determines whether a proposition is true differs per domain of discourse—per language game—such that a scientific proposition may be true because of its correspondence to reality, whereas a mathematical proposition would be true because of its coherence with a set of mathematical axioms. The determination pluralist’s claim is that a plurality of truth-determining properties is called for to explain truth for different language games. (Edwards, 2013).

This is not the only objection to deflationism from the pluralists, but it is nonetheless a prominent objection. The objection is that we cannot account for the role of truth in our discursive practices without having to get into a substantive metaphysical inquiry into truth’s nature. The reason for this is as follows. First of all, the role of truth is normative since it is used as a standard of correctness and as a goal of inquiry. Secondly, given the variability of our discourse—the plurality of language games that make up assertoric discourse—there must be different truth properties (or truth-determining properties) for truth to play this normative role as standard and as goal of inquiry. Hence, we need to reject deflationism and accept pluralism about truth.Footnote 2

One could reimagine this argument as saying that truth needs to be plural at the metaphysical level to sustain its unity at the conceptual level; if we want truth to play its role—e.g., as the goal of inquiry—for all domains of discourse, there must be multiple domain-specific truth properties. I believe that versions of this kind of alethic pluralism, that pair conceptual unity with a metaphysical plurality, are the legacy of Wright’s Truth and Objectivity.

This article offers a new kind of pluralism that breaks with this tradition. It offers a kind of pluralism not at the metaphysical level, but at the level of use. It argues that truth does not have a single uniform role in our linguistic practices, it has two distinct roles. Nevertheless, this dual role of truth does not call for two different truth properties. As such, both the deflationist and the pluralist could get behind my theory, though neither could do so without making concessions. I thereby offer this theory as an opportunity for reconciliation, and as a way forward for both the pluralist and the deflationist.

My article owes a great debt to Huw Price’s recently published “Global Expressivism and Alethic Pluralism” (2022).Footnote 3 Price’s article, in combination with his other writings, offered somewhat of a roadmap for the development of my theory. I show, in Sect. 2, how Price tries to account for the normativity of truth without negating deflationism. He does so by arguing that the normativity of truth isn’t explained by its metaphysics but by the practical utility of taking truth as our aim. However, I argue in Sect. 3 that this argument doesn’t work for matters on taste. This means either that the argument is faulty or that truth isn’t normative for discourse on taste. I explore the latter option. Section 4 shows that this is indeed a viable option, one that has recently been endorsed by pluralists such as Ferrari and Moruzzi (2020), and by Wright (2021) himself. Section 5 clears the way for my pluralism at the level of use by arguing for a pragmatist meta-alethic framework that can accommodate it. Finally, Sect. 6 explores the pluralist credentials of my theory.

2 Price and the normativity of truth

What is it for truth to be normative? In the above section I have said that part of truth’s normativity is that it is a standard of correctness and a goal of inquiry. However, these claims seem rather platitudinous. Who would deny that it is correct to believe < p > if and only if < p > is true? And who would deny that we aim to gain beliefs that are true and avoid those that are false? Even the staunchest deflationist would accept these platitudes.Footnote 4 So why is it that the pluralist believes that arguing for the normativity of truth is an effective strategy for arguing against deflationism?

The answer is that the debate isn’t about whether truth is normative per se but about where this normativity comes from. Deflationists like Rorty (1995) or Horwich (2006) believe that the normativity of truth is derived from the normativity of assertibility. Pluralists, such as Lynch (2004), tend to argue for a reversed order of explanation.

Price’s view is an outlier within this debate since he is a deflationist who nevertheless agrees with the pluralists that truth offers a separate norm (Price, 1998, p. 241). In this current section I shall briefly outline his view.

Price argues that the aim for truth gives us a common goal. If we merely aimed for justification we would not care about differences of opinion, since we could both be individually justified in having these different opinions. Yet, we recognize that our opinions, insofar as they conflict, can’t both be true, so if truth is our aim, we would have a reason to engage with each other and to try to resolve the conflict, to seek agreement through argumentation. This is how Price puts it:

Without the [truth] norm, differences of opinion would simply slide past one another. Differences of opinion would seem as inconsequential as differences of preference. […] The [truth] norm makes what would otherwise be no-fault disagreements into unstable social situations, whose instability is only resolved by argument and consequent agreement. […] If reasoned argument is generally beneficial — beneficial in some long-run sense — then a community […] who adopt this practice will tend to prosper, compared to a community who do not. (Price, 2003, pp. 180–181)

Price calls the truth norm a coordination device: it provides the pressure needed for us to try to resolve our disagreement and to coordinate our opinions. It is this use of truth that is overlooked by the kind of deflationist who believes that truth is merely a device for expressing agreement.

So, what kind of deflationist is Price? Why doesn’t he feel compelled to inflate truth based on his finding that the truth norm is a distinct norm? Price argues that truth is conceptually important, but he resists the pluralist’s argument that truth’s conceptual role is satisfied by multiple distinct properties. A comparison between Price’s formulation of the truth norm and Wright’s notion of Cognitive Command shows where the difference between Price’s position and that of the pluralists lies.

According to Wright, a domain of discourse displays Cognitive Command if and only if “it is a priori that differences in opinion […] will involve something which may properly be regarded as a cognitive shortcoming” (Wright, 1992, p. 144). Roughly, we can say that a discourse displays Cognitive Command whenever we assume that any disagreement must be indicative of a fault. There could be many faults at the root of a disagreement, such as a misevaluation of evidence, some kind of cognitive processing error (like making a miscalculation), or perhaps the fault of basing one’s judgment on insufficient evidence. Yet it is important to see that Cognitive Command entails an a priori commitment to there being a fault, and therefore we must assume that all disagreements imply some kind of fault, even if we are unable to identify this fault. The fault we take the disagreement to imply is an alethic fault. We know that at least one of us must be at fault because it is not possible that both of our claims are true; one of us is at fault because of endorsing a claim that is false.

Compare this to what Price says about the truth norm:

[The truth norm] is a norm which speakers immediately assume to be breached by someone with whom they disagree, independently of any diagnosis of the source of the disagreement. Indeed, this is the very essence of the norm of truth, in my view. (Price, 2003, p. 164)

Price’s interpretation of the fault of breaching the truth norm is the same as Wright’s interpretation. However, Wright argues that Cognitive Command is a local feature, that it is a feature of realist discourse, and that it thereby indicates that truth for such discourse is correspondence. While we could both be justified in our respective beliefs, these beliefs couldn’t both correspond to the way things actually are. Thus, if having our beliefs correspond to reality is our aim, we know that at least one of us must be at fault.

Price, conversely, believes that his truth norm ranges over all assertoric discourse. It is, he says, this kind of normativity that makes discourse assertoric in the first place. (Price, 1983) Hence, the a priori alethic fault is not just present when we disagree about “realist” matters such as matters of science, it is present in all instances of disagreement; whenever we disagree, we must assume that only one of our beliefs could be true and that, therefore, at least one of us is mistaken.

By severing the connection between this conception of truth and the realist domains of discourse, Price hopes to suck the metaphysical life out of it. While he agrees with the idea of truth as correspondence, he argues that it offers nothing but a mere platitude. It neither explains nor justifies how truth is used (Price, 2003, p. 180).

3 The practical value of the truth norm

If it isn’t some metaphysical fact about the nature of truth that explains its normativity, then what is it instead? Price is clear on this: it is the practical utility of the truth norm that explains why it is instituted by our assertoric practices. He does not claim that the institution of the truth norm was an intentional achievement by some past generation that was aware of its practical advantages, as if it is some policy we adopted after lengthy deliberation. Instead, the story is that our assertoric practices have evolved to include it because of its advantages, regardless of whether we were or are aware of such advantages (Price, 1998, p. 251).

The truth norm is useful because it aids disagreement, and what (ideally) happens in disagreement is that the better view wins out. Given Price’s pragmatist framework, this would be the view that is most advantageous. Sometimes, “two heads are better than one” (Price, 1983, p. 356). He puts it as follows:

In [disagreements], rival, incompatible views are exposed to common scrutiny. Ideally the more well-justified prevails, and one speaker recants, accepting the view of the other. Plausibly, there is enough of a general advantage in such dispute behaviour to explain the existence of a powerful linguistic device to facilitate it (i.e., the use of ‘true’ and ‘false’). This advantage will be explained in terms of the behavioural consequences of particular views, and the consequent benefits of basing one’s views on as wide a body of experience as possible. [...] This explains why language has developed a general means of indicating such agreements and disagreements, in the application of the terms ‘true’ and ‘false’ to the associated utterances. Utterances such as questions, commands and requests, on the other hand, characteristically result from states of mind for which no such reason for unanimity exists. Different speakers can reasonably hold conflicting such states of mind (conflicting in the sense that no one person could hold them concurrently), even if fully acquainted with each other’s viewpoint. Appropriately, ordinary usage does not apply ‘true’ and ‘false’ to the types of utterance which express, or result from, these states of mind. (1983, pp. 356–357)

Price argues that the Strawsonian deflationist, who thinks of truth merely as a device for expressing agreement, misses this crucial point. If truth were merely a way of saying ‘ditto’ then why can’t a question or a command be true? Why would it be inappropriate to reply with “It’s true” when our companion at a restaurant says, “I would like to have the lentil soup”, and we want to convey that we would like to have the soup as well? In other words, why can phrases such as “It’s true” only be applied to assertions? The Strawsonian deflationist has no answer.

Price does have an answer, which is that truth does more than express endorsement; it facilitates disagreement by giving us a shared norm. Assertions could be thought of as, using Robert Brandom’s phrase, moves in “games of giving and asking for reasons”. (Brandom, 1994, p. xviii) Disagreements are a central part of this game. Since the truth norm facilitates disagreements, it facilitates games of giving and asking for reasons. Through these games we (ideally) converge toward the beliefs that work best for everyone. According to Price, these games wouldn’t exist without the truth norm, which is why it is an essential characteristic of an assertion as opposed to any other kind of utterance. The assertion is the kind of utterance for which it is useful to engage with each other’s views by subjecting them to common scrutiny.

I shall evaluate Price’s claim about the advantages of aiming for truth in more detail below and argue that these advantages go missing in the case of matters of taste. In later sections, I suggest that Price’s argument might still work, and that truth simply isn’t normative for discourse on taste. I then draw on the work of Shapiro (2021) to show that truth has another use for this domain of discourse. Hence, truth is a dual-purpose tool.

3.1 The same boat

In Facts and the Function of Truth, Price talks about the “Same Boat Property” (SBP), which he describes as follows:

An utterance type is thus correlated with a class of mental states that share what I shall call the Same Boat Property (SBP). A class of mental states have the SBP if their typical behavioural consequences are such that their behavioural appropriateness, or utility, is predominately similar across a speech community. If a mental state has the SBP, then if it is appropriate for any one of us, it is appropriate for all – we are all in the same boat. (Price, 1988, p. 152)

Price argues that (1) it is only useful to aim for truth if the same boat property is satisfied and (2) the assertion is the only utterance type that possesses this property. If we make conflicting assertions, it is likely that one of us possesses a correlating mental state that is behaviorally disadvantageous. This explains why the truth norm only ranges over assertions.

There are two potential problems with this view though. The first is that it can become very difficult to track the behavioral consequences of certain kinds of assertions. For example, assertions that belong to the most theoretical branches of mathematics might be found lacking in any behavioral consequences whatsoever. Price would presumably treat such assertions as outliers. But this brings us to our second problem. Even if the behavioral consequences of an assertion are clear, the question of whether such consequences are advantageous depends in a large part on the idiosyncrasies of an individual’s background beliefs and desires.

Price’s discussion of the SBP seems to say that beliefs are such that it pays to pool our resources regardless of our goals. For example, we should join forces when it comes to our mathematical beliefs, regardless of whether your goal is to build a bridge or demolish one. Mathematics is a shared system of beliefs from which we can, generally, all profit. But not all beliefs are like this. Take, for instance, beliefs about gustatory taste. We can presume that almost everyone has the goal to satisfy their own tastes, even if this goal might at times be counteracted by other goals, such as the goal to pursue a healthy diet. But the goal to satisfy your tastes is not in any way advanced by there being a shared system of beliefs about gustatory taste.

There might be some advantages to the coordination of taste, such as that it would be easier to settle on which restaurants to go to. However, it would mostly be counterproductive to try to coordinate our beliefs through argumentation. Price argues that the aim for truth leads to coordination by adding friction to our disagreements. It makes us keen to try to resolve our disagreements. The attitude that would get in the way of this resolution is one of steadfastness. Hence, if the aim for truth is to have the practical effects Price thinks it does, it must outlaw the steadfast response to disagreement (or to peer disagreement at least). When we encounter disagreement, we are to suspend our judgment until the disagreement is resolved. It is this suspension of judgment that creates the tension to resolve our disagreements; if I were allowed to hold onto my beliefs in the face of disagreement, disagreement per se wouldn’t bother me.

Apply this thought to the matter of taste, and you can see disaster looming. If we wouldn’t be allowed to have firm beliefs about taste until the relevant disagreements are resolved, we would, for the most part, not be allowed to have any beliefs about taste at all. This attitude would create an excess of doubt that would impede the satisfaction of our goals. Better to remain steadfast, to agree to disagree, instead of seeking endlessly for a resolution.

It seems, then, that the utility of the normativity of truth does not apply to assertions about matters of taste. Moreover, if the aim for truth has the kind of practical effects Price takes it to have, it would be disadvantageous to aim for truth when it comes to such matters. I have also shown that Price appeals to the utility of the truth norm to argue for the normativity of truth. Hence, his argument seems not to apply to discourse on taste.

4 How to do without the truth norm

We have now hit upon a crossroads. I have shown that Price attempts to explain the normativity of truth by appealing to the utility of aiming for truth, instead of by looking into the nature of truth itself. However, I have also argued that this utility goes missing in the case of matters of taste. The case of taste might thus be interpreted as a counterexample to show why Price’s argument fails and why we need to look into the nature of truth after all. However, might it not be possible that Price’s argument does work, but that truth just isn’t normative for discourse on taste?

In the pluralist spirit, I want to explore this option. After all, if truth could be more than one thing, wouldn’t it be conceivable that it is normative for some domains of discourse and not for others? This current section shall argue that it is. First by responding to Price’s claims that the truth norm is part and parcel of the act of asserting, and then by looking into some of the pluralist literature that has long contended that truth must be normative across the board but that now seems to turn away from this claim.

4.1 Why disagree about taste?

Price argues that without the truth norm we have no reason to criticize another’s beliefs: “Differences of opinion would seem as inconsequential as differences of preference” (Price, 2003, p. 180). Therefore, without the truth norm we would not even enter into games of giving and asking for reasons. We would only be inclined to do this when we see the need to change each other’s beliefs. And, as I have shown, this need depends on the utility of coordinating our beliefs through games of giving and asking for reasons.

Insofar as discourse on taste is assertoric, it must be possible to engage in games of giving and asking for reasons, but why even give or ask for reasons at all if we’re not trying to resolve our differences and coordinate our beliefs? I argue that, in the case of taste, games of giving and asking for reasons can be useful even when we do not expect them to result in a coordination of belief. When we have to defend our taste judgments, we gain the opportunity to explore the reasons behind these judgments and to discover structures that interlink our individual judgments, which in turn enables us to extrapolate our tastes into undiscovered areas, to seek out the things we enjoy—and avoid things we wouldn’t enjoy—more effectively. We become better at predicting our own tastes once we have explored the reasons behind our taste judgments. When we aren’t triggered to defend these judgments by our interlocutors, we are at risk of being guided by prejudice instead of reason, of assuming that there are good reasons for our opinions where there are none.

Clearly there are advantages to subjugating one’s taste judgments to games of giving and asking for reasons, but what would trigger our interlocutor to challenge our judgments? Price would say that they are inclined to challenge our views because they believe our views are wrong and they want to convince us of this. But they would only want to do this if the aim is truth rather than mere justification.

However, there is another reason why we might want to ask others to provide reasons for their taste judgments. If you didn’t like the pasta I made you, I would love to know why you didn’t, not because I hope to change your mind but because I can avoid making you food that isn’t to your taste in the future. More generally, there are distinct advantages to me understanding your taste judgments that have nothing to do with me wanting to convince you that your judgments are wrong. By understanding how these judgments are structured, I can better understand your dispositions to act, and when you deviate from the expected course of action, I can hold you accountable, e.g., ‘I thought you said you hated cheese?’ Hence, we have reason to critically engage with each other’s opinions even in the absence of the same boat property and, therefore, even in the absence of the truth norm.

4.2 Asserting without the truth norm

As mentioned above, Price is set against the possibility that assertoric discourse might lack the truth norm. One of his arguments turns on the difference between a seemingly objective assertion, e.g., ‘Pistachio ice cream is tasty’, and its overtly subjective counterpart, ‘I like pistachio ice cream’. Price believes that this difference cannot be explained without taking the former claim to be guided by the truth norm. He argues that the latter claim is an “escape hatch” that cancels out this truth norm. He then claims that these escape hatches are more densely dispersed over discourse of basic taste because the utility of coordination through argumentation is more defeasible when it comes to matters of taste. According to Price, the ease with which we can switch from the objective claim to a subjective one—‘Well, I like it at any rate’—explains why we think of the discourse as (relatively) subjective (Price, 2022, p. 43).

However, this explanation of the overtly subjective claims as escape hatches for the seemingly objective statements can still be endorsed without committing to truth being normative for such statements. In that case, it isn’t the normativity of truth we try to escape from, it is our justificatory burdens that we want to avoid. When we say ‘Pistachio ice cream is tasty’ we allow others to demand reasons for this claim and to scrutinize the reasons given. But what often happens when it comes to such basic matters of taste is that we feel that no more reasons can be given. Our interlocutor won’t let us get away with asserting that pistachio ice cream is tasty if we’re unable to provide reasons for this assertion, but they would let us get away with saying that we just happen to like pistachio ice cream. We use the escape hatch to get out of the disagreement, to relieve ourselves of the justificatory burdens born by our initial assertion, when we realize that there are no more good reasons at hand.

One might understandably be wary of this talk of reasons and justification, in the absence of truth, given that such notions are standardly defined in relation to the aim for truth. The pragmatist could, of course, argue that such notions are to be understood practically, which would undermine the distinction between epistemic and practical reasons or justification. But, of course, not everyone who takes truth to be normatively deflated has such a radical stance on these epistemic notions. For example, Brandom thinks that reasons are provided by the rules of the relevant language game, and that by grasping the conceptual content of the terms in play, we understand how these rules are to be applied. For example, our concept of spiciness prohibits the inference from ‘This burrito is spicy’ to ‘This burrito is bland’, while it licenses the inference to ‘This burrito has a strong taste.’ Brandom understands justification as an entitlement to a belief, which is granted by these inferential rules (Brandom, 1995, p. 903). I agree with Brandom that it is possible to understand assertions as moves in games of giving and asking reasons, and that such games can in turn be made sense of without appealing to a truth norm.

Price, of course, disagrees. He believes that all assertoric discourse is guided by the truth norm and that the presence of this norm is what makes an assertion an assertion. He argues that if truth isn’t normative, we could not explain why it applies only to assertions and not also to a host of other speech acts.

However, now that we’ve seen that we can still engage in games of giving and asking for reasons without the truth norm, there is an alternative explanation for why truth only applies to assertions even if it isn’t normative. This is Shapiro’s (2021) deflationist account of truth. Shapiro also thinks of assertions as moves in games of giving and asking for reasons, but he conceives of truth not as a guiding norm but as performing a different role in these games. According to Shapiro, we make claims such as ‘What S said is true’ as a way of transferring our justificatory burdens. In one dialectical context, someone could challenge my assertion of p and I could defend this assertion by saying ‘S said that p and what S says is true’. This is a pretty standard appeal to authority. I thus use truth to transfer my justificatory burden onto S. In another dialectical context, in which we see that S is being challenged and struggling to defend their assertion, I can say ‘What S said is true’, thereby stepping in to shoulder S’s justificatory burden. All that truth needs to be to play this role in games of giving and asking for reasons is a device for expressing agreement. Yet, it is only within these games that truth can play this role and that is why truth only applies to assertoric discourse.

4.3 Pluralists on the normativity of truth

As mentioned above, many pluralists have argued or presumed that truth is distinctively normative. This tradition goes back to Wright’s Truth & Objectivity (1992) in which he argued that though truth coincides in positive normative force with warranted assertibility, these norms are distinct. Warranted assertibility and truth diverge in extension whenever there are neutral states of information, i.e., states that would neither warrant the assertion of p nor of its negation (Wright, 1992, p. 20). This means that, while aiming for warranted assertibility and aiming for truth might be the same activity, their satisfaction-conditions can come apart, e.g., when one asserts a proposition for which one lacks warrant yet which is nevertheless true. Wright shows that the equivalence schema alone entails this result and hence that the commitment to the equivalence schema alone already shows that the truth norm is distinct from the norm of assertibility, and that truth must be more than a mere device for expressing endorsement or for generalizing.

However, Ferrari and Moruzzi (2020) have argued that if one adopts a subjectivist and relativist metaphysics and epistemology, this argument will lose its force. For in that case, neutral states of information would either not occur at all or they would be cases in which the proposition in question lacks a truth-value. Either way, there couldn’t be the difference in extension between truth and assertibility upon which Wright’s argument relies. They argue furthermore that the domain of taste is one for which such metaphysics and epistemology might be most likely to be correct. Hence, they show that the case of taste can serve as a local counterexample to Wright’s inflationary argument.

Wright himself has actually recanted his argument in a recent article. He argues that assertoric discourse is guided by more than just the norm of assertibility. There are, for instance, also norms of restraint, retraction, or denial. While Wright maintains that the normativity of truth cannot be derived solely from the normativity of assertibility, he argues that it might still be derived from the normativity of our assertoric practices more generally, as seen as including all of these norms. He says that “Any discourse controlled by such a complex pattern of norms will, once “true” is introduced via the [equivalence schema], throw up the contrasts between “true” and “assertible” that drive the Inflationary Argument” (Wright, 2021, p. 447). There is a myriad of norms that guide our assertoric discourse, and the deflationist might still maintain that the normativity of truth is derived from these norms, instead of it being a further distinguishable assertoric norm.

Wright goes on to also argue that truth isn’t normative in the case of taste. Instead of talking about the metaphysics and epistemology of the domain of taste, Wright talks about the phenomenon of faultless disagreement to support his claim.

The puzzle of faultless disagreement is as follows. When we disagree, we take the other’s belief to be false. Hence, we take the other’s belief to be worse than ours; ours is true, whereas theirs is false. So how could any disagreement be faultless?

The contextualist would say that our assertions of taste are implicitly indexicalized. What I mean by saying that pistachio ice cream is tasty is that it is tasty for me. What you mean when you say that it isn’t is that it isn’t tasty for you. That’s why both our assertions can be true at once and why the disagreement is faultless. However, as John MacFarlane argues, if this is the case then we aren’t really disagreeing at all, we are simply talking past each other (MacFarlane, 2014, p. 8).

A relativist like Kölbel (2004) or MacFarlane (2014) would say that the truth of a claim like ‘Pistachio ice cream is tasty’ is relative to whomever assesses the claim. Hence, one and the same proposition could be true when assessed by you and false when assessed by me, which is how we could come to disagree without either of us being alethically at fault.

Yet, Wright argues that instead of fracturing the truth norm in this way, so that everyone has their own standard, it would be simpler to do away with the normativity of truth entirely for matters of taste:

[Faultless disagreement is] a hopeless idea if the discourse is thought of as answerable to a single norm of truth with which no statement and its negation can simultaneously comply. So if faultless disagreement is to be a possibility, there must be no such single alethic norm. That leaves two options. One is, in one way or another, to — as it were — fracture the norm, multiply the ways of being true, and spread the pieces around, so that conflicting opinions can each alight on a shard. [Relativism] attempts a particular implementation of that option. The other option is to suction out the substance of the alethic norm, leaving only the formal shell. (Wright, 2021, p. 445)

If truth doesn’t pack its “normal normative punch” (Wright, 2021, p. 438), then we can disagree with someone, thereby taking their belief to be false, without automatically taking them to be at fault for having this belief.

Both Ferrari and Moruzzi as well as Wright have hereby shown that it is conceivable, if not plausible, that truth isn’t normative for matters of taste. Hence, it is still possible that Price was right about his pragmatic vindication of the truth norm; maybe the practical advantages of aiming for truth do explain why truth is normative. However, in that case, the practical disadvantages of aiming for truth would likewise explain why truth isn’t normative for discourse on taste.

5 Meta-alethic pragmatism

The above suggestion—that truth might be normative for those, but only those, domains of discourse for which the aim for truth is practically advantageous—can only be made to work if we accept a particular pragmatist framework. We must see truth as a tool that we use for a particular purpose. The idea is, then, that truth is a tool that can be used in different ways. In discourse on matters of taste we just use it to transfer justificatory burdens, whereas in other, more “objective” areas of discourse we use it as a standard, a norm, and thus as a coordination device. This is a pluralism about truth at the level of use.

It is exactly the kind of pluralism that Price anticipates in his recent paper. He shows that, back in 1988, he still believed in faultless disagreements, and he thought of the mark of objectivity or factuality of a discourse in terms of its resistance to faultless disagreements. Reflecting back, he says that this view implied a kind of “internal-to-expressivism alethic pluralism”, “a story told at the level of use about how truth ‘amounts to subtly different things’ in different domains” (Price, 2022, pp. 48–49).

To make the move from how truth is used to what it is like, a meta-alethic perspective is needed that brings these two closer together. As it turns out, it is exactly the meta-alethic views of the deflationists that do the trick.

Some deflationists hold rather extreme meta-alethic views, such as Strawson, as mentioned in the first section, who believes that truth-talk simply isn’t descriptive of reality (Strawson, 1949, p. 94). On such a view, it would make no sense to say that there might be any difference between how truth is used and what it is really like. William Gamester calls this kind of thesis “Meta-Alethic Expressivism”, which states that alethic judgments are about how we ought to live rather than about what the world is like.Footnote 5

This meta-alethic view is clearly congenial to my position, but I expect that most contemporary deflationists would reject it. However, there is a lot of theoretical space between saying that alethic discourse does not describe reality at all and saying that the kind of reality it describes is completely insulated from any practical concerns. Call this latter view meta-alethic realism. It is the view that says that the matter of what truth is like simply has nothing to do with how we are to use the concept of truth or the word ‘true’ in practice and is purely a matter of metaphysics, i.e., of figuring out what truth really is independently of any of our concerns with it.

Meta-alethic realism is antithetical to deflationism. The deflationist says that we must analyze truth by discovering what use we make of it in practice. We look at the role truth plays in our language games, and this tells us all we need to know about what truth is like. Compare this, for instance, to the question of what a chess piece such as a rook is like. All we need to do to answer this question is discover how the piece is used in the game of chess. It matters not what material it is made of or what size it is. Yet, for the meta-alethic realist it is the material that matters. The realist is like the gemologist who says of a piece of jewelry that has been used as if it is made of diamonds, that it is really made of glass. For the meta-alethic realist, the question of what truth is like can come apart from the question of how truth is used.

This meta-alethic stance would go against the dialectic of this article. I want to argue that deflationists should, and can, accept a novel kind of pluralism, at the level of use. To show this, I play by the deflationist’s rules. Yet, the deflationist won’t be a meta-alethic realist. Instead, they seem committed to some form of meta-alethic pragmatism, and the theory I am presenting is best described as a combination of alethic pluralism and meta-alethic pragmatism. I have defended this combination before in “Alethic Pluralism for Pragmatists” (2022), in which I argued that truth is a tool we have constructed for practical purposes, and in which I also attempted to construct a pluralism on practical grounds. Meta-alethic pragmatism entails that truth is an entity that can fully be described by the role(s) it plays in our language games, like a rook in chess. There is no need for a further inquiry into what kind of stuff truth is made out of to understand what it is like. The pragmatist view says that truth is a tool and that what it is like is determined by how it is best used:

(Meta-Alethic Pragmatism) What truth is like is determined by the practical purpose(s) it serves.

5.1 Which kind of truth for meta-alethic discourse?

In my previous paper, I recommended an epistemic, pragmatist, notion of truth for alethic discourse (Kaspers, 2022, p. 13). However, if truth-attributions belong to alethic discourse, then truth for such discourse could not be epistemic, as shown by Chase Wrenn. Wrenn starts his argument with an example of a proposition that he assumes, for the sake of the argument, to be true but unknowable:

Expand The universe expands and contracts eternally, without beginning or end.

Consider this truth-attribution:

T-Expand Expand is true.

Because T-Expand entails Expand, any warrant for T-Expand would supply warrant for Expand. If you could be indefeasibly warranted in believing or asserting T-Expand, then you could be indefeasibly warranted in believing or asserting Expand. If T-Expand were knowable, Expand would be too. So, since Expand is unknowable, T-Expand must be unknowable too.

Assume for reductio that aletheiological truth is epistemic: <<p> is true> is true only if <<p> is true> is knowable. Since T-Expand is unknowable, it follows that T-Expand is not true. (That doesn’t mean T-Expand is false; epistemic truth may tolerate truth-value gaps.) By hypothesis, Expand is unknowably true, so:

Expand is true, but T-Expand is not true.

Substitution yields:

(*) Expand is true, but <Expand is true> is not true.

which has the form:

p, but <p> is not true.

This is inconsistent with ES [the equivalence schema] and the non-negotiable logical principle of Capture: p ⊢ <p> is true. So, by reductio, if truth is non-epistemic in any discourse, it must be non-epistemic for truth-attributions. (Wrenn, 2020, p. 313)

I think that this argument is correct and that we must therefore accept its conclusion: truth for truth-attributions cannot be epistemic (at least not for all truth-attributions). Therefore, to maintain that truth for alethic discourse is epistemic, one must argue that truth-attributions do not belong to the alethic domain, that they aren’t statements about truth. Could this be done?

When we classify which sentences belong to which domains of discourse, it is customary to look at the predicate. For example, when I say ‘Shoplifting is a source of revenue loss’ my statement belongs to the domain of finance, but when I say ‘Shoplifting is wrong’ my statement belongs to morality. One can question whether ‘is true’ is a predicate in the same way as ‘is wrong’ is. Some deflationists would argue that it isn’t because it is a purely syntactic device, a logical expression, and that it thus doesn’t belong to any domain at all. In that case, truth-attributions wouldn’t be statements about truth. However, if truth is more than a syntactic device, if it is a real standard we aim to meet, it seems likelier that ‘is true’ is a predicate that belongs to the alethic domain of discourse, just like ‘is wrong’ is a predicate that belongs to the moral domain. Therefore, for those who believe that truth for alethic discourse is an epistemic property, it is difficult to maintain that truth-attributions don’t belong to the alethic domain.

What does this leave us with? If we explore the options offered by Wright (2021) and we rule out epistemic truth properties, the choice ends up being between a realist truth property and a deflationary truth property. The former option would lead us back to meta-alethic realism. Hence, the only option left is the deflationary truth property.

5.2 Is truth normative for alethic discourse?

Yet, there is still the question as to which deflationism applies to truth for alethic discourse: Price’s normative deflationism or Wright’s understanding of deflationary truth as truth “without its normal normative punch”? (Wright, 2021, p. 438) I argue that it has to be the latter kind of deflationism.

Suppose that truth-attributions belong to alethic discourse. A truth-attribution is equivalent to its corresponding object-level proposition. Hence, a disagreement about < p > is true wouldn’t be different in any way from the disagreement about p. Now, suppose the proposition in question is one about a matter of taste. In this case, truth wouldn’t be normative and there would be no alethic fault. Yet, if truth is normative for alethic discourse, then a disagreement about < p > is true would always be indicative of an alethic fault, no matter which discourse p belongs to. Given that < p > is true and p are equivalent, this would imply that disagreement about p would always be indicative of an alethic fault, no matter which discourse p belongs to. The consequence of taking truth to be normative for alethic discourse is thus that truth would always be normative, for all domains of discourse. This is contrary to what we have established in the previous section.Footnote 6

But there is a reason to believe that truth about alethic discourse isn’t normative even if alethic discourse doesn’t include truth-attributions. This reason has to do with the possibility of faultless disagreement. I have maintained Price’s characterization of the truth norm as the a priori commitment to taking all disagreements to contain faults. Hence, to show that truth isn’t normative for alethic discourse, it would suffice to show that some of our alethic disagreements are faultless.

I think that the meta-alethic pragmatist would concur. For example, Price makes use of the thought experiment of a community of “Mo’ans”, who are Merely Opiniated Asserters, and who use truth merely as a device for recording agreement (Price, 1998, p. 247). The point of his thought example is that what they call “truth” is nothing like our concept of truth. Of course, the point is also that their concept of truth is defective, that it is wrong because it does not serve them well (in some discourses at least). Their conception of truth is at fault in virtue of being practically inadequate. However, this need not always be the case. We could recognize that a society’s or community’s practical needs might be different enough to warrant a different conception of truth. Wyatt (2018) argues that empirical data suggests that our folk concept of truth differs across communities and cultures, and he consequently argues for a conceptual pluralism about truth. It would be presumptuous to believe that other communities and cultures have impractical or otherwise defective folk concepts of truth, for it could very well be the case that different practical contexts (whatever these may be) call for a different concept of truth. In this case we might recognize that the disagreement is faultless.Footnote 7 The possibility of such a faultless disagreement about truth shows that truth isn’t normative for alethic discourse. If it were, all our disagreements about truth would contain an alethic fault and none of them could possibly be faultless.

6 Is it pluralism?

My reader is probably wondering at this point whether my theory is really a kind of pluralism about truth and, if so, what this pluralism would look like. If truth is normative for some but not all domains of discourse, it follows that there are at least two truth properties: a normative one and a nonnormative property. In this respect, my kind of pluralism would be very similar to the “ecumenical alethic pluralism” of Ferrari and Moruzzi (2019). Ferrari and Moruzzi use their proposal of a subjectivist and relativist metaphysics and epistemology for the domain of taste, as mentioned above, to argue that the normativity of truth is locally deflated, and that there must thus be a normatively deflated truth property in addition to one or more normative truth properties. Yet their concern isn’t with practical considerations about how truth is best used but with, for instance, metaphysical considerations about the status of discourse on taste. This marks a difference in their approach to pluralism as well. Their pluralism follows the classical model: the concept of truth is characterized by a network of platitudes and truth is plural only at the metaphysical level. My pluralism operates primarily at the level of use. The question is whether the plurality at this practical level carries over to the conceptual and metaphysical level.

6.1 Function and use

I shall first discuss the conceptual level. As mentioned earlier on, the network analysis of Wright (1992) and Lynch (2009) specifies the function of truth in terms of platitudes that relate the concept of truth to that of, for instance, justification, belief, and inquiry. I also mentioned that among these platitudes are those that express the normativity of truth, such as the platitude that truth is the goal of inquiry. Ferrari and Moruzzi cannot maintain that a truth property must satisfy all these platitudes, for their nonnormative truth property couldn’t satisfy these platitudes that express truth’s normativity. They suggest that the network analysis imposes a flexible instead of a strict requirement on truth properties, meaning that a property need not satisfy every single one of the core platitudes for it to count as truth (Ferrari & Morruzi 2019, p. 387). This is their way of maintaining the conceptual unity of truth.

This might be more difficult for me to do given my meta-alethic pragmatism and the fact that truth has multiple uses. Does this not mean that truth has multiple functions—that there is no such thing as the function of truth—and that, thus, truth must be plural at the conceptual level as well? I do not think that it does. Let me explain why.

As Amy Thomasson notes, there is a difference between an entity’s function and the way in which it is used (Thomasson, 2020, p. 53). This distinction is subtle. For instance, a hammer can serve many purposes, such as cracking nuts, but its function is to drive nails. I don’t think there is a principled distinction between functions and uses that holds for all kinds of entities. However, there is, I believe, an intuitive difference in the case of truth. The deflationist argues that truth is primarily a linguistic device, and that it has a linguistic function. Remember Strawson’s (1949) claim that truth is a linguistic device for the expression of agreement. As I have shown, Shapiro (2021) takes this conception on board but asks why we only use this device for assertions. He argues that a device for the expression of agreement can be used to transfer justificatory burdens. Thus, truth is a tool that functions as a means of expressing agreement, but this tool can be used to transfer justificatory burdens.

I think this is basically right, but I want to add to it that truth also has a second use. If truth functions as a linguistic device for expressing endorsement, and there is a domain of discourse for which it pays to try to coordinate our beliefs through argumentation, then it pays to have truth be our collective aim for this domain of discourse. Hence, truth can be used as a norm for such a discourse. By attaching a positive normative force to our endorsements—that is, to our truth-attributions—we also attach a fault to our attributions of falsity. This is why our disagreements—in which we say of the other’s assertion that it is false—come to contain alethic faults, which is exactly what is supposed to happen, according to Price, if truth is to be used as a means for reaching convergence via argumentation. Yet, this story is consistent with a run-of-the-mill deflationism about the linguistic function of the truth predicate. This is what provides the conceptual unity of truth; truth is a single tool that can be used in two different ways for two different purposes.

6.2 How many truth properties?

Ferrari and Moruzzi are open to accepting multiple normative truth properties, such as a causal interpretation of correspondence for astronomical discourse and coherence for arithmetic. The nonnormative truth property is, of course, a deflationary truth property (Ferrari & Moruzzi, 2019, p. 389). I am less inclined to accept a plurality of normative truth properties. The acceptance of different normative truth properties is again driven primarily by metaphysical or epistemological considerations, which seem out of place in my pragmatist pluralism.

In his recently published book, Ferrari seems to have changed his mind a bit. His “normative alethic pluralism”Footnote 8 highlights the normative variability of truth, but it doesn’t entail a metaphysical variability. He has come to accept minimalism about truth by rejecting that the nature of truth is metaphysically substantive (Ferrari 2022, p. 172). I am very sympathetic to this move, as it is an excellent way of combining pluralism with deflationism at different levels of analysis. In fact, I want to do something similar here.

Many deflationists accept the platitude that truth is correspondence. They see it as a strength that deflationism is compatible with this intuition. Take, for example, Paul Horwich’s following claim:

The common-sense notion that truth is a kind of ‘correspondence with the facts’ has never been worked out to anyone’s satisfaction. Even its advocates would concede that it remains little more than a vague, guiding intuition. But the traditional alternatives — equations of truth with ‘membership in a coherent system of beliefs’, or ‘what would be verified in ideal conditions’, or ‘suitability as a basis for action’ — have always looked unlikely to work, precisely because they don’t accommodate the ‘correspondence’ intuition[.] (Horwich, 1998, p. 1)

Horwich is an example of a deflationist who believes that truth is correspondence to the facts without believing that truth offers a norm that is separate from the other norms of assertibility. Such a deflationist could say that it is precisely because truth doesn’t imbue inquiry with its own brand of normativity that the idea that truth is correspondence is platitudinous. Since, as long as truth is not a distinct norm, one can accept this idea of correspondence without thereby committing oneself to taking on the metaphysical burdens that are imposed by the existence of an external, evidence-transcendent, standard.

Yet, the deflationist on the other side of the aisle, Price, also accepts the idea that truth is correspondence, as a platitude. Just like Horwich, he believes that it is a great weakness of the traditional alternatives to correspondence that they can’t account for the correspondence intuition:

It is often felt that [coherence] theories fail to make sense of the intuition that truth has something to do with answerability to some external standard – the intuition, to put it crudely, that we can agree until we are blue in the face, yet still be wrong. Correspondence theories try to make sense of this world-imposed ‘rightness’, as the basis of an analysis of truth. (Price, 1988, p. 89)

Price accepts that truth is an external standard, a world-imposed standard, which adds a sense of objectivity to our inquiry. Yet, he doesn’t accept that this can serve as the basis of an analysis of truth. It is, instead, the analysandum. For Price it is the utility of the a priori belief that disagreements are indicative of a fault that explains why we take truth to be normative and why we believe that our inquiry is tracking objective facts that our true beliefs correspond to.

It seems, then, that one can take truth to be correspondence regardless of whether one takes truth to be normative. Does this mean that there is only one truth property? Or are there two distinct correspondence properties, a normative and a nonnormative one?

When truth is normative in Price’s sense, it seems to become an “external” and “world-imposed” standard, with some or maybe even most of the realist trimmings. As I have mentioned in an earlier section, the a priori alethic fault thus created makes that the discourse in question possesses the property of Cognitive Command. Wright argues that it is the possession of Cognitive Command that is indicative of a metaphysically inflated truth property:

Showing that a discourse exerts Cognitive Command thus has the effect of “beefing up” the Correspondence Platitude in just the kind of realism-relevant way I advertised. One shows precisely that the idea of representation featured therein has a characteristic which minimal truth aptitude does not impose, but which it had better have if there is to be real substance to the idea that, in using the discourse in ways which respect the standards of assertoric warrant by which it is informed, we function as representational systems, responsive to states of affairs which, when we are successful, our beliefs and statements serve to map. (Wright, 1992, p. 147)

I think one could agree with this while maintaining that there is only one truth property across the board. This is because the metaphysical weight added when we use truth as a standard does not pertain directly to the nature of truth itself but to the nature of the discourse in question. Wright says a lot about correspondence in his book, but he refrains from talking about the metaphysics of the correspondence relation. Is it a causal relation, or an isomorphic relation? Wright does not say. He seems to agree with Horwich that it “has never been worked out to anyone’s satisfaction”, and therefore his claim that truth is correspondence “remains little more than a vague, guiding intuition” (Horwich, 1998, p. 1).

Furthermore, the metaphysical weight is more the result of how we use truth than of what truth is really like. This point holds more generally. Yes, in some sense there must be two truth properties, a normative one and a nonnormative one, but this is only due to us using the same thing in two different ways. Suppose we find a rock on the pavement, and we decide to use it as a rook for our game of chess. After our game is finished, I use the stone once more to scrape some dirt off the sole of my shoe and then I discard of it. When we use it as a rook we instill it with a kind of normativity through the rules of the game of chess, whereas when I use it to scrape the dirt off my sole it is void of normativity. Has the stone become two entities, one normative and one nonnormative? Perhaps some metaphysicians would concur that it has. But on an intuitive level, it is one and the same entity that we happen to have used in two distinct ways. It is in this way that I claim that there is only one truth property, correspondence, even though truth can be used in two distinct ways.

If alethic pluralism is the thesis that there are multiple truth properties, then I suppose my theory wouldn’t count as a pluralist theory about truth. I nevertheless maintain that my theory captures an important aspect of the pluralist spirit. Alethic pluralism in the tradition of Wright’s Truth and Objectivity (1992) has used the inquiry into truth as a stage for debating the metaphysical bearings of different domains of discourse. Wright’s pluralist is a methodological anti-realist, who believes that we mustn’t assume that all our inquiries aim to correspond to an objective reality; this kind of realist talk must be earned (Kellen, 2018). And instead of doing straight up metaphysics to establish a discourse’s realist credentials, Wright looks not at the entities that the discourse refers to but at the features of the discourse itself. One of the relevant features is the way in which we engage with disagreements. If this calls for an understanding of truth as an external standard, then the discourse has earned (some of) its realist stripes. This project can be continued if one accepts my theory: one can look at whether we use truth as an external standard for a domain of discourse by looking at how we engage with disagreements within this domain. If we do use truth as an external standard, then the discourse is objective.

Of course, this stops short of vindicating the kind of realism that Wright and other pluralists are after. For instance, one could argue that mathematical discourse is anti-realist even though it isn’t subjective, and even though mathematical truth is normative. One could argue, for example, that there are no unknowable mathematical truths because the principle of bivalence does not hold for mathematics (this is what Dummett 1991 does) or one could maintain that mathematical entities are generated by mathematical discourse (Edwards, 2018, Chap. 4). This shows that there are many kinds of realism/anti-realism distinctions. Discourse that is realist in the one sense of the word can fail to be realist in another sense. Objectivity is one of many features of realism, and establishing the objectivity of a domain of discourse is one way of marking a metaphysical distinction in language, but it is only one of many. Nevertheless, the fact that my pluralism at the level of use allows us to draw this contrast does represent an important departure from the deflationist tradition, which contends that the inquiry into truth shouldn’t be used as a metaphysical battleground and tries to insulate truth from all metaphysical connotations.

This represents an important payoff of endorsing the theory proposed in this article. The Topical Collection to which this article belongs—Alethic Pluralism and Its Critics—is a retrospection of 30 years of Truth & Objectivity, the book that started it all for the pluralists. But Wright’s book wasn’t just about truth itself, it was about the metaphysical differences between domains of discourse. In some sense, truth is just a vehicle for expressing these differences. Ever since, this has been one of the biggest draws to alethic pluralism. It allows us to draw distinctions between domains of discourse that might otherwise be overlooked. The Topical Collection is also—as the title suggests—a platform for engaging with the critics of the theory. And deflationists have been some of the most prominent critics of alethic pluralism. They get hung up on the baroque metaphysics of pluralism. My theory offers a way to be sensitive to the differences between inquiries—the different language games we can play within our assertoric practices—without having to commit to bold claims about the metaphysical nature of truth.

In a less constructive spirit, one could argue that I’m handing deflationist a way of defeating alethic pluralism by saying ‘No truth itself isn’t plural, we just use it in different ways.’ However, the deflationist’s meta-alethic pragmatism makes it difficult to clearly distinguish between what truth itself is like and how it is used. Hence, I suggest that we be more constructive. For deflationists to accept my proposal, they must be willing to meet the pluralists halfway. I thus proclaim that my pluralism at the level of use represents a resolution between the deflationist and the pluralist, which can, hopefully, advance the conversation.

7 Conclusion

In this article, I have adopted the deflationist methodology of trying to answer the question of what truth is like by exploring how the concept of truth, and the word ‘true’, is used in practice. Within this framework, I have defended a pluralist theory of truth at the level of use, by arguing that there are practical considerations for taking truth to be normative for some but not all domains of discourse. I then endorsed meta-alethic pragmatism, which is the view that truth is a tool in our linguistic practices, and that what it is like is determined by how it is best used by these practices.

I’ve distinguished between truth’s linguistic function, as a device for the expression of agreement and generalizations, and its uses. I have shown that this device can be used to transfer justificatory burdens in any assertoric discourse, but it can also be used as a standard for those discourses for which it pays to aim for the coordination of beliefs through argumentation. By attaching a positive normative force to our endorsements—to our truth-attributions—we also attach a fault to our disagreements—to our attributions of falsity. When we understand truth’s linguistic function as a device for the expression of endorsements, we understand how it can accommodate these dual uses. This is how unity can be preserved at the conceptual level.

What distinguishes my pluralism from traditional forms of alethic pluralism is that it also preserves unity at the metaphysical level. As long as truth isn’t normative, one could accept that truth consists in correspondence without thereby committing to an external standard. When truth is normative, it does provide an external standard, but it can be argued, following Price, that there need be nothing mysterious about this standard, nothing that needs explaining in, say, causal or isomorphic terms. The external standard is simply brought about by our intolerance toward disagreement, it is a consequence of this practical attitude. Hence, we can maintain that truth is correspondence, in a deflationist spirit, regardless of whether truth is normative.

My pluralism about truth at the level of use is a pluralism that operates exclusively at the level of use. It is an open question whether it is pluralistic enough to count as a form of alethic pluralism. I have suggested that one reason why it might do is that the theory allows for the connection between certain features of truth and the objectivity, or lack thereof, of a domain of discourse, which is the kind of employment of alethic pluralism as envisaged by Wright in Truth and Objectivity (1992). This also marks the greatest departure from deflationism. In many other respects, however, my pluralism at the level of use and deflationism, especially the deflationism of Price, are kindred theories. I allow both the pluralist and the deflationist to claim my theory as belonging to their doctrine. I’ll leave it up to them to declare me friend or foe. In any case, I hope that the theory can fill a new theoretical space between the two doctrines and that it can, thereby, push their discussion forward.