1 Introduction

1.1 Wright-Price expressivism?

I’ve long regarded Crispin Wright as an ally, on some topics. Even so, I was surprised recently—re-opening my copy of Truth and Objectivity after many years—to find the following comment on the last page of Chapter 1: ‘CW as my kind of global expressivist!’ The comparison may seem implausible to some readers, but what surprised me was less the claim itself than the discovery that I had already expressed it, as bluntly as this, on some forgotten previous reading .

Lest the remark escape me a second time, I’m nailing it to the mast in this piece, as a hook for a comparison between my views and Wright’s on a range of topics in which we have common interests—among them truth, pluralism, and debates between metaphysical realism and its rivals. In other words, I’ll treat it in the spirit of an examination question: ‘Crispin Wright is a Pricean expressivist’—Discuss.

Some differences will emerge, of course, and I’ll pay special attention to those. As we’ll see, one major and central difference is that I am an expressivist about truth itself, in a way in which Wright is not.Footnote 1 I’ll explain how this difference makes a difference to our understanding of the nature of ordinary disagreements about many other matters. And at the end of the paper I’ll contrast my brand of expressivism and Wright’s alethic pluralism, as strategies for making sense of what might neutrally be called linguistic or discourse pluralism. As I’ll argue, however, all these differences sit on top of some deeper similarities—among them, an attraction to some such pluralism in the first place, and a suspicion that popular forms of realism provide an unnecessarily extravagant framework within which to explore it.

I can’t recall when I left this remark in my copy of Truth and Objectivity, but I presume it dates from the present century. I didn’t use the label ‘global expressivism’ before then, so far as I can recall.Footnote 2 But like the view for which I now use the term, my sense that Wright and I are on a similar page goes back a lot further—it predates Truth and Objectivity, in fact. In ‘Metaphysical pluralism’ (Price, 1992) I cite a typescript copy of Wright’s ‘Realism: The Contemporary Debate—W(h)ither Now?’ (Wright, 1993), claiming that my own conclusions in that piece—a close ancestor of what I later came to call global expressivism—are comparable to Wright’s anti-realism.Footnote 3

I’ll begin below (Sect. 1.2) with an account of what I took myself to be doing in ‘Metaphysical pluralism’, and why I saw Wright as an ally. Because that piece also mentions Blackburn’s quasi-realism, this will take us quickly to the reason why my remark about Wright is likely to seem implausible. Wright was a prominent critic of quasi-realism, and yet I present my own view, in ‘Metaphysical pluralism’ and many later pieces, as a sympathetic advance from quasi-realism.Footnote 4 How could I be close both to Wright and to Blackburn?

To answer this question I’ll need to explain what I take expressivism to be. I’ll do that (Sect. 2) by explaining my response to the charge that semantic minimalism is incompatible with expressivism (a charge of which Wright’s argument in Truth and Objectivity is an influential example, of course), as the response developed in the 1990s and after. In other words, I’ll be outlining my take on what Jamie Dreier (2004) was to dub the problem of Creeping Minimalism. As I’ll explain, the upshot is that in those discourses that Wright regards as operating with merely his minimal notion of truth, he is an expressivist, in my sense. So far, then, good news for my marginal remark.

Differences begin to emerge when we dig more deeply into our approaches to truth itself. Here I can explain the differences by beginning (Sect. 3.1) with a later engagement with Wright’s views, this time with Truth and Objectivity itself. A few years after ‘Metaphysical pluralism’, I wrote about Wright’s objection to Paul Horwich’s deflationism about truth. In Truth and Objectivity Wright argues that Horwich’s deflationism cannot account for the normativity of truth—this despite the fact, Wright claims, that such normativity can be shown to be a straightforward consequence of the disquotational schema at the heart of Horwich’s account. In a piece written in the mid-1990s (Price, 1998) I argued that although Horwich has the better of the immediate battle, Wright wins the war. Truth is normative, in a way that Horwich’s view does not explain, though for reasons that Wright himself does not correctly identify.

Happily, I can explain the core argument of (Price, 1998) using a framework that Wright himself puts on the table in one of his most recent papers (Wright, 2013). In this paper Wright amends the view defended in Truth and Objectivity to allow for so-called ‘faultless disagreement’ in domains that are merely, as Wright puts it, ‘minimally truth-apt.’ His examples include judgements of taste, such as ‘Sushi is delicious.’ He argues among other things that his approach gives a better understanding of disagreements about such matters than that of recent relativists, such as John MacFarlane.

Several aspects of this recent piece provide very helpful points of comparison and contrast to my own views on similar matters. So I’ll outline this paper in Sect. 3.2 (mostly in Wright’s own words), and then explain in Sect. 4 how it relates to my own account of truth, developed in Facts and the Function of Truth (Price, 1988; hereafter FFT), as well as several of my earlier and later pieces. Along the way (Sect. 4.2), I’ll compare my own proposal about the coordinative role of the truth norm to proposals by MacFarlane, as well as Wright (2021) himself.

This will take us two directions. First, it will lead (Sect. 5) to what I think is a substantial disagreement about what Wright regards as the non-minimal cases—about those, as I’ll explain, my view about truth itself is resolutely expressivist, in a way in which Wright’s is not. At crucial points, I think, the option isn’t even on his radar. Second, it will lead (Sect. 6) to a disagreement about the remaining cases, those that Wright regards as merely minimally truth-apt. Wright (2021) argues that his view offers a better account of faultless disagreement than recent relativist approaches, including that of MacFarlane. I’ll explain why I disagree with both Wright and MacFarlane, and explain my alternative account of these cases.Footnote 5

In Sect. 7.1 I turn to a methodology that has been an enduring feature of Wright’s alethic pluralism: namely, an approach to the metaphysics of truth via a platitudes-based ‘analytical theory’, as Wright calls it (2001, p. 760). This is a proposal in which I have had a particular interest. It is essentially what John Hawthorne and I (1996) dubbed ‘the Canberra Plan.’ In that form I have raised objections to it in a number of places. I relate those objections to Wright’s use of the approach, and explain why I think an expressivist should be suspicious of it (especially so in the light of the normativity of truth, on which Wright and I agree). In Sect. 7.2, however, I note that there is an alternative kind of alethic pluralism in the offing, one that an expressivist can happily embrace.

In conclusion, in Sect. 8, I compare the motivations that led Wright and me to our respective views. Not surprisingly, perhaps, there is a comparable mix of similarities and differences to that between our views themselves. One of the things we share, as I have already noted, is an attraction to some sort of linguistic or discourse pluralism. The last issue I discuss is the difference between my kind of expressivism and Wright’s alethic pluralism as ways of making sense of such pluralism. I close with the suggestion that as presently formulated, Wright’s view is a kind of transit station, where the alethic pluralist needs to make a choice between metaphysical and expressivist foundations for the plurality claimed at the level of truth itself.

1.2 Metaphysical pluralism

My targets in ‘Metaphysical pluralism’ (Price, 1992) were philosophers I called ‘lapsed Humeans’. I characterise the Humean ideal, from which these philosophers lapse, as the view that

the pinnacle of metaphysical virtue [is] a world in which the only facts are the mundane first-order physical facts about how things actually are—“a vast mosaic of local matters of particular fact, just one little thing and then another,” as David Lewis puts it.Footnote 6 (Price, 1992, p. 34)

Why do metaphysicians who are attracted to this ideal fail to live up to it? They do so, I said, because the

Hume world has frequently been judged too cramped. Some lapsed Humeans extend the bare Humean structure to fit in modal facts; some extend it to fit in subjective experience or intentional mental states; others to fit in moral facts; and so on. The common theme is that a respectable metaphysics simply cannot survive on the bare regime that Hume prescribes. (34)

I contrast two responses to the comparative profligacy of these lapsed Humeans (their failure to live according to their own spare conception of metaphysical virtue). One criticises them from the Humean side, by arguing that these extensions to the bare Hume world are actually unnecessary. This kind of response is very familiar. Think of many objections to Lewis’s own modal realism, for example, or Lewis’s own approach to many other matters.

The second response, which was mine, was much less familiar. It was to criticise the Humean conception of metaphysical virtue, by ‘deflating’ the pluralism of the lapsed Humeans’ views. I characterised the argument of the paper like this.

Roughly, its effect is to undercut the distinction between various non-Humean forms of metaphysical realism and something akin to a Wittgensteinian linguistic pluralism. To the extent the distinction can be drawn, moreover, the latter is the default position. So, not only is this form of pluralism an important and widely neglected option in a range of contemporary metaphysical debates; it actually has claim to be the pre-eminent option. In a sense I shall explain, it is the philosophical geodesic, the course from which no one is entitled to depart without good reason. The paper thus presents a challenge to the lapsed Humeans of contemporary metaphysics: embrace worldly pluralism, or return to the pure faith, for there is no virtuous middle way. (35)

I point out that despite its profligacy by the standards of orthodox Humeans, the lapsed Humean position is intended as a form of monism. It is intended as a claim about a single world or reality, even though its thesis is that that single world contains more than orthodox Humeans would like to find there. Thus it is an ‘additive monism’, as I call it, compared to the orthodox Humean view.

I then characterise ‘the central issue of the paper’ as the question ‘as to whether there is actually a tenable distinction between this monist position and our target [Wittgensteinian] pluralism.’ The paper goes on to argue

that despite the avowedly ontological nature of their concerns, these lapsed Humeans cannot avoid reliance on a certain semantic distinction – essentially, the distinction between descriptive and nondescriptive uses of language – and points out that this puts them at at least a prima facie disadvantage compared to the pluralists, who need no such distinction. [It] outlines a case for thinking that the required distinction cannot be drawn, and draws attention to some connections between the resulting sort of pluralism and certain other recent approaches to the same metaphysical topics. (35)

One of these ‘other recent approaches’ was that of Wright, another that of Blackburn—more on those comparisons in a moment.

In its focus on ‘the distinction between descriptive and nondescriptive uses of language’, the paper draws heavily on my conclusions in FFT, published a couple of years previously. One of the main concerns of FFT is with this distinction, and with the various guises in which it appears—e.g., that between ‘factual’ and ‘nonfactual’ uses of language. The main conclusion of the first part of the book is that all these formulations are ill-grounded—a fact we miss because, as in other areas of philosophy, the different formulations tend to take in each other’s washing. The three main themes of FFT were (i) rejection of this assumption concerning a division in language, and of the notion of strict or genuine factuality on which it depends; (ii) a defence of a kind of global non-factualism; and (iii) an account of truth and falsity (and their limits) in terms of their normative role in disagreement. I’ll say more below about what I meant by these conclusions; the last of them, in particular, relates to some of the most interesting differences between myself and Wright.

For the moment, I want to explain how it was that when I appealed to FFT in ‘Metaphysical pluralism’, it seemed so natural to enlist Wright as a potential ally—as I proceed to do, as soon as I have described my conclusions in FFT:

This approach has affinities with recent work of Crispin Wright’s (1993) on what might be called the “fine structure” of truth. Wright has distinguished a number of different components that may, but need not all, be characteristic of the use of truth in association with a particular area of discourse. In effect, he suggests that we may classify a subject matter according to which of the set of these characteristics we take its notion of truth to involve. As I understand it, his concern is mainly descriptive and taxonomic. I think there is some prospect, however, that the structure thus discerned will turn out to be explicable as sketched above, in terms of some general account of the function of truth in language. Whether the best general account will be in the terms I outlined earlier remains to be seen. But if it is to be in keeping with Wright’s program, I think it will have to be like my account in being explanatory rather than analytic: its focus will be not on the question “What is truth?” but on the question “Why do ordinary speakers have such a notion as truth?”

Interestingly, Wright too takes [his] stance on truth to support a noncommittal metaphysics, at least as the natural fallback position. He characterizes this position as an antirealist one, in Michael Dummett’s sense, on the grounds that the minimal notion of truth it requires can be thought of as derived from assertibility. ... I take the following remark of Wright’s to be in much the same vein as my emphasis on the economical advantages of discourse pluralism: “Anti-realism thus becomes the natural, initial position in any debate. It is the position from which we have to be shown that we ought to move. All the onus, everywhere, is on the realist” (1993, p. 69). Substitute ’pluralist’ for ‘antirealist’ and ‘nonpluralist’ for ‘realist’, and these are my sentiments exactly. (Price, 1992, p. 48)

Not everything in this comparison now seems to me to be accurate. Whether or not I misinterpreted (Wright, 1993), Wright’s later work makes it clear that we are not on the same page with respect to the question I characterised as the difference between explanatory and analytic approaches to truth. Wright finds a great deal more interest than I do in the question ‘What is truth?’ The distinction between these two approaches, and advocacy of the former at the expense of the latter—including, as I would now put it, advocacy of expressivism about truth itself—were major themes of the second part of FFT. As I’ve already noted, I now think that this marks a central difference between my views and Wright’s. Much more on this below.

Relatedly, my appropriation of the term ‘pluralism’ in the context of this comparison to Wright now seems a little quick. Wright has his own strategy for pluralism, going via pluralism about truth. Again, more on this below. But the parallel I draw concerning the claimed ‘economical advantages’ of our respective views still seems to me to be a good one.Footnote 7

But let’s turn to a second comparison I draw in ‘Metaphysical pluralism’. The paragraph about Wright is immediately followed by this one:

The project also has affinities with Blackburn’s quasi-realism. The quasi-realist about moral discourse (for example) wants to argue that, although moral judgments are not really factual, they are entitled to the trappings of factuality, including a respectable notion of truth. So, although there is only one domain of genuine facts, it is quite proper on this view for ordinary usage (and truth in particular) to work as if there were many. As we noted earlier, what separates this from discourse pluralism is the quasi-realist’s assumption that his project is a limited one, bounded in its application by the availability of a substantial distinction between factual and nonfactual uses of language. If that distinction lapses, quasi-realism is no longer a distinct alternative to discourse pluralism. Discourse pluralism is again the default position. But so long as a more limited quasi-realism remains a live project, the quasi-realist’s interest in explaining why certain discourses should usefully employ an “artificial” notion of truth will apparently coincide with the project described above. (1992, p. 49)

As I noted above, readers familiar with Wright’s criticisms of Blackburn (in Chapter 1 of Truth and Objectivity, among other places) may find it puzzling how I could think of myself as close to both. The explanation rests on two points. First, my claimed affinity is not to quasi-realism as such, but to what it becomes when ‘globalised’, as I was later to say—when the distinction between ‘real’ and ‘quasi’ is lost. In FFT I had already drawn attention to the tendency, as I saw it, for Blackburn’s ‘local’ version of the view to be too successful for its own good—for the account it offers of truth and other ‘apparently factual’ aspects of language in the so-called quasi cases to work equally well in all cases. This was a tendency I applauded, of course.Footnote 8

The second and more major point concerns the consequences of the deflation of the distinction between ‘real’ and ‘quasi’ cases. I think that this was not very well understood at the time of Truth and Objectivity, on either side of the argument—i.e., either by Wright and others making similar criticisms of expressivism, or by their expressivist opponents. Much of the fault here lies on the expressivist side, and the position has been greatly clarified since then, in part in response to Wright’s challenge and others like it. Some of my own attempts to clarify what I meant by expressivism were responses to challenges of this kind, and in one case to what I saw as a flawed response to them by others. By explaining some of this history I can explain how from my present perspective it still seems appropriate to regard Wright’s view in Truth and Objectivity as close to expressivism, in my sense (except that he baulks at endorsing the global version of the view).

2 Minimalism and expressivism

2.1 Defeat or global victory?

Let’s begin with Wright’s argument from Chapter 1 of Truth and Objectivity. The last section of that chapter is entitled ‘Minimalism’. As Wright says, his object there is

to commend ... a kind of minimalism about truth—a species of deflationism, if you will, but unencumbered by the classical deflationist’s claim that truth is not a substantial property. The minimalist view is that when a predicate has been shown to have the relevant features, and to have them for the right reasons, there is no further question about the propriety of regarding it as a truth predicate. Minimalism is thus at least in principle open to the possibility of a pluralist view of truth: there may be a variety of notions, operative within distinct discourses, which pass this test. (Wright, 1992, pp. 24–25)

A couple of pages later, Wright asks the question that interests us here.

[N]eed any of this provoke a proponent ... of expressivism, or “quasi-realism”, ... to disagree? In effect, I have argued that there is a notion of truth aptitude which is carried in train by possession of assertoric content ... But I have so far ventured nothing about possession of assertoric content as such; in particular, nothing at variance with the suggestion, integral to the expressivist tradition, that possession of genuine assertoric content is a relatively deep feature of the sentences of a discourse, which its overt syntax can serve to mask, or merely to simulate. The final ingredient in any conception of truth and truth aptitude which can be common ground between realist and anti-realist is that this is not so.

An analogy may help. Elsewhere I have argued that Frege’s platonism about number is best interpreted as based on the view that an expression’s candidacy to refer to an object is a matter of its syntax: that once it has been settled that a class of expressions function as singular terms by syntactic criteria, there can be no further question about whether they succeed in objectual reference which can be raised by someone who is prepared to allow that appropriate contexts in which they do so feature are true. There is, that is to say, no deep notion of singular reference such that an expression which has all the surface syntactic features of a Fregean proper name, and features in, say, true contexts of (by surface syntactic criteria) predication and identity, may nevertheless fail to be in the market for genuine—“deep”—reference. So too, in the present context, the claim must be there is no notion of genuine—deep—assertoric content, such that a discourse which exhibits whatever degree of discipline (there are firmly acknowledged standards of proper and improper use of its ingredient sentences) and which has all the overt syntactic trappings of assertoric content (resources for—apparent— conditionalisation, negation, embedding within propositional attitudes, and so on)—no notion of genuine assertion such that a discourse with all this may nevertheless fail to be in the business of expressing genuine assertions. Rather, if things are in all these surface respects as if assertions are being made, then so they are.Footnote 9

On reflection, it is possible to be a little less combative. It is not necessary to insist that there is no suitable notion of deep assertoric content. It suffices that there is, at any rate, at least a more superficial one, carried by surface syntactic features; and that a minimal truth predicate is definable on any surface-assertoric discourse. If we can go on to explain what, after they have agreed that claims about what is funny are apt for truth and falsity in the minimalist sense, could still be at issue between realists and anti-realists about the comic, perhaps that will supply a sense in which minimally truth-apt claims can yet fail to be deeply assertoric. But I don’t think that will be the happiest way of expressing the significance of the points of debate which will later concern us. (1992, pp. 28–29)

Wright is here elaborating what was becoming a familiar objection to expressivism. If truth is ‘minimal’, or ‘deflated’, then it seems easy to be truth-apt or truth-conditional, and implausible to claim that utterances that appear to make factual claims (e.g., moral utterances) nevertheless fail to do so. Similar arguments had recently appeared in Boghossian (1990) and Humberstone (1991), though the point goes back further. In FFT, I put a very similar interpretation on a remark from McDowell (1981). After describing a thin approach to truth based on the equivalence schema, McDowell adds this footnote:

It is a philosophical issue whether there are respectable purposes for which a stronger notion of truth is required. A familiar sort of non-cognitivist about values, for instance, making play with the idea that real truth is correspondence to objective reality, will not be content with the application of my platitude to, say, ethical assertions. ... I am inclined to suppose that this is a matter not so much of an alternative notion of truth as of a characteristically philosophical misconception of the only notion of truth we really have: one which the platitude in fact suffices to determine. (1981, p. 229, n. 9)

When I cite this remark in FFT (Price, 1988, p. 40), I am interested in the question whether one can appeal to truth to draw the distinction I mentioned earlier between ‘factual’ and ‘non-factual’ uses of languages—e.g., in the latter category, questions and commands. The observation that ‘thin’ notions of truth are unable to do so goes back at least to Bernard Williams (1966/1973), who proposes the term ‘substantial theory of truth’ for an account thick enough to draw this kind of distinction. Williams offers Strawson’s early ‘performative’ theory of truth (Strawson, 1949) as an example of an insubstantial account.Footnote 10 As I put it in FFT:

The fact that Strawson’s theory is insubstantial in this way is noted by Bernard Williams, in the paper in which he introduces the idea of a substantial theory of truth. Williams says that he doesn’t see ‘how on such a theory it could be more than an accident of language that “is true” signified agreement with assertions rather than agreement with anything else’ [Williams 1966/1973, p. 203]. (Price 1988, p. 26)

This is not quite the same as the argument that a thin notion of truth is inadequate for expressivism, but I think it deserves to be seen as a close ancestor. After all, traditional non-cognitivists seemed to be assuming the same factual/non-factual distinction, but with a view to placing it somewhere within the class of indicative utterances.Footnote 11

At any rate, the versions of the argument from Boghossian, Humberstone, and Wright in the early 1990s soon generated a response from Jackson et al. (1994), who proposed a defence of the coherence of expressivism in the face of minimalism. They argued that minimalism about truth need not imply minimalism about truth-aptness, and that the latter will do for the expressivist’s purposes. Moreover, they suggest that the latter can be explained in psychological terms. Truth-apt claims are those that express beliefs, rather than some other sort of psychological attitude.Footnote 12

In FFT I had considered the attempt to rest a factual/non-factual distinction on a belief/non-belief distinction, and concluded that it wouldn’t bear the weight. Under pressure, in fact, it turns out to be natural to define beliefs in terms of their semantic properties, rather than vice versa:

Like Hume himself, the non-factualist may therefore feel inclined to ... say that desires are distinguished from beliefs in virtue of the fact that it is ‘impossible ... they can be pronounced either true or false’ [Hume, 1978, p. 458]. (Price, 1988, p. 93)

Thus the argument of (Jackson et al., 1994) left me unconvinced. Moreover, I felt it missed a much more interesting avenue for defending the compatibility of expressivism and ‘thin’ theories of truth. In a response written with John Hawthorne in 1994, I offered what I took to be a better way of combining expressivism with semantic minimalism. It rested on two observations. First, what is essential to expressivism (or, as we then said, ‘non-cognitivism’) is the idea that the disputed vocabularies have distinctive functions, and that other philosophies go wrong in missing this point.

Most basically, we suggest, what these positions have in common is that they characterise a linguistic function, or category, in terms of which the non-cognitivist may claim that the disputed sentences serve a different function from, or belong to a different category to, other parts of language (and in particular, to paradigmatic causal–explanatory parts of language). (Hawthorne & Price, 1996, p. 114)

We note that it is a ‘distinction of this sort that underpins the philosophical impact of non-cognitivism, particularly in opposition to reductionist or eliminativist moves’:

Non-cognitivists argue that these programs, and the philosophical concerns from which they arise, rest on a distinctive kind of mistake about language—on a misidentification of the linguistic category within which particular families of concepts have their home. (Hawthorne & Price 1996, p. 114)

Our second observation was that the functions in question need not be defined in terms of notions such as truth and belief, thus rendering the view immune to deflation of these notions by minimalism. We give three examples: Brandom (1984) on reference, Gibbard (1990) on norms, and Horwich (1990) on truth. We point out that all these views subscribe to the ‘metaphysics rests on a linguistic category mistake’ idea, and yet at the same time are thoroughly comfortable with the application of minimal notions of truth and belief, in the vocabularies in question.

In a slightly later piece I characterised the upshot of these points like this:

Noncognitivists were right in thinking that the notion of linguistic function provides a naturalistic solution to metaphysical concerns, but wrong to try to characterise the functions concerned in the terms normally used. To get things right, we need to retain the insight that different bits of language may serve different functions in a way which isn’t obvious at first sight, but set aside the usual attempts to characterise the functions concerned in terms of truth, factuality, belief, and the like. (Price, 1997, p. 135)

In that piece I call the approach I am recommending ‘functional pluralism’ (noting that it is what Hawthorne and I had referred to as ‘a version of noncognitivism’).

Later again [e.g., in Macarthur and Price (2007) and Price (2009)], it seemed to me more helpful to say that traditional expressivism already had the functional stories it needed, but simply made the mistake of bundling them with something it didn’t need, and which minimalism undermined. Put this way, traditional expressivism is interpreted as making two claims: a negative claim, to the effect that the utterances in question are not truth-apt (or something of that semantic kind); and a positive claim, to the effect that the utterances in question have such-and-such other (not semantically-characterised) function.

As Macarthur and I (2007) point out, semantic minimalism undermines the negative claim, which is incompatible with the deflationist’s thesis that semantic terms do no substantial theoretical work; but it leaves the positive claim intact. This means that so long as expressivism puts all its weight on the positive, functional story, it is completely untouched by the deflation of the negative claim. Certainly some of the ways that expressivists traditionally characterised their view have to be abandoned, but this clarifies and strengthens the view.Footnote 13

Indeed, the upshot is that semantic minimalism is bad news for the expressivist’s opponents, because it deprives them of the semantic vocabulary in which they find it natural to state their own view—i.e., a vocabulary that does put theoretical weight on notions such as truth and reference. That may be a problem for local expressivists, but because it recommends global expressivism,Footnote 14 not because it recommends global factualism. In a deflationary sense it does indeed recommend global factualism, but this is no help to the opponent who wishes to disagree with the expressivist’s claim about the functions of particular vocabularies.

As I have said, I take Wright’s comments in Truth and Objectivity about the consequences of minimalism to be pushing in the same direction. Contrary to his own interpretation, the team who actually come out ahead, where truth is merely minimal, are the expressivists. For it is the would-be non-expressivists who find their theoretical vocabulary too deflated to be useful. (Once again, it helps in making this point to distinguish the positive claim, which is essential to expressivism, from the negative claim, which is not.)Footnote 15

2.2 Creeping minimalism

Pursuing the theme of the relationship between expressivism and semantic minimalism into the present century, let me briefly mention Jamie Dreier’s influential paper, ‘Meta-ethics and the problem of Creeping Minimalism’ (Dreier, 2004). Dreier characterises Creeping Minimalism as follows.

Minimalism sucks the substance out of heavy-duty metaphysical concepts. If successful, it can help Expressivism recapture the ordinary realist language of ethics. But in so doing it also threatens to make irrealism indistinguishable from realism. That is the problem of Creeping Minimalism. (Dreier, 2004, p. 26)

Dreier doesn’t raise this as a problem for one side or the other, but for the field as a whole:

[T]he problem is not a problem for realists or for irrealists, but more a problem in meta-meta-ethics. It’s not as if one side had better be able to come up with something clever to say about how to distinguish realism from irrealism or else the other side wins. It’s rather that those of us who feel confident that there is some difference between the two meta-ethical camps should be concerned that we don’t know how to say what that difference is. (2004, p. 31)

The solution Dreier proposes—drawing, as he says, on (Hawthorne & Price, 1996), as well as work of Kit Fine (2001) and Allan Gibbard (2003)—rests on the idea that what distinguishes expressivism is a distinctive explanatory thesis:

[T]o explain what it is to make a moral judgment, we need not mention any normative properties. So Hawthorne and Price’s characterization of expressivism turns out to mesh with Fine’s criterion for distinguishing irrealism, and also with Gibbard’s understanding of what distinguishes his own sophisticated expressivism from realism. I’ll call the integrated account of the distinction, the “explanation” explanation. (2004, p. 39)

Dreier concludes that the ‘divide between realism and irrealism, at least in meta-ethics, rests on the substance of questions about metaphysical explanation.’ (2004, p. 42). Returning to these issues in (Dreier, 2018), Dreier calls this the explanationist proposal.

It would take us too far afield to explore my response to Dreier in detail, but I want to note that Wright and I both seem to disagree with Dreier in one important respect, in the same direction. Wright and I are both inclined to see minimalism as imposing a greater burden on realism than on its rivals, with respect to the task of making a case that it has something distinctive to say. ‘All the onus, everywhere, is on the realist’, to repeat the words from Wright (1993) that I borrowed in ‘Metaphysical pluralism’.

In other words, I think that both Wright and I agree with Dreier that ‘[m]inimalism sucks the substance out of heavy-duty metaphysical concepts’, and that this ‘can help [what I would call] Expressivism [to] recapture the ordinary realist language of ethics’ (Dreier, 2004, p. 26). But, contra Dreier, we think that this is not ‘a problem in meta-meta-ethics’, but merely a discovery in meta-ethics. It is the discovery that anything worth calling realism, in a sense intended to be incompatible with expressivism, needs to reject semantic minimalism. (I think Wright would also agree that this is not exclusively an issue in meta-ethics. The issues at stake here extend well beyond the ethical realm.)

To reinforce the sense that Wright are on the same page on this point, let me note a short remark from one of Wright’s most recent papers.

[L]et us be mindful that, where truth is deflated and so registers no independent norm operative over the acceptance and assertion of statements, an account of meaning—of correct linguistic practice—has to proceed in other, non-truth-conditional, broadly use-theoretic terms. So we need to focus directly on the use of the signature statements in [the] discourses [in question]. (Wright, 2021, p. 441)

It is common ground among expressivists that one of the central distinguishing marks of the view is that it explains the meaning of a target vocabulary in terms of use conditions, rather than truth conditions. While this may not be all it takes to be an expressivist, it is an important and major step in that direction – one that Wright here takes to be a direct consequence of a discourse’s being merely ‘minimally truth-apt’, as he puts it (2021, p. 439).

I am going to discuss this recent paper of Wright’s in detail below (Sect. 3.2), and we’ll see that it offers further confirmation of this reading. Where truth is merely minimal, in Wright’s sense, that’s good news for what I call expressivism, and bad news for any opposing view that seeks to call itself realism.Footnote 16

2.3 What is expressivism?

So what do I mean by expressivism? I now think of it as a recipe with about five main ingredients.Footnote 17 One important ingredient, as just mentioned, is a use-first approach to meaning. Expressivism focusses on how words are used, rather than what they are about. I have a rather broader conception of the factors the relevant accounts of use are allowed to involve than many expressivists. I think it is unhelpful to restrict them to psychological states, as opposed to more general aspects of speakers’ circumstances.Footnote 18

The second ingredient is a programme that presents itself as an alternative to metaphysics, or ontology. It may be motivated in the same way by so-called ‘placement problems’—that is, in their typical form, questions about the ‘place’ of some seemingly problematic subject matter (e.g., morality, modality, meaning, or the mental), in the kind of world revealed to us by science. But expressivism combines an insistence that these be regarded as primarily linguistic or psychological issues—Why do we talk or think this way?—with a renunciation of the ‘representational’ moves that lead from there back to metaphysics (e.g., that of seeking ‘referents’, or ‘truthmakers’, in some non-deflationary sense).

The third ingredient, closely linked to the first and second, is an explanatory programme. It aims, roughly speaking, to account for the existence and practical relevance of the vocabularies in question; typically the former in terms of the latter, in some way. Why do creatures like us employ these terms and concepts? And why do these terms and concepts exhibit distinctive links to various aspects of our practical lives? I call the latter question the Practical Relevance Constraint (Price & Weslake, 2010; Price, 2023, ch. 11), and argue that it is often a great advantage of expressivism over various rivals that it meets it so easily. I’ll illustrate this point below, in making a distinction between my views and Wright’s about truth itself.

The fourth ingredient, closely linked again to the third and first, rests on identification of features of speakers—typically features of practical or ‘pragmatic’ significance—that play characteristic roles in expressivist accounts of particular vocabularies. I have called these features the pragmatic grounds of the vocabularies in question (Price, 2019, p. 146).

The fifth ingredient, finally, is a kind of perspectivalism, with the pragmatic grounds of a vocabulary playing the role of the perspective from which the users of that vocabulary speak. I link this ingredient to the Copernican metaphor familiar from Kant, noting how well it characterises the sense in which expressivism provides an alternative to metaphysics. What we took to be in need of metaphysical investigation is instead explained as a perspectival matter, in which features of our own situation carry the main explanatory burden.

2.4 An alternative to metaphysics

I have noted that Wright takes something that I would count as expressivism to be a consequence of a discourse’s being merely minimally truth-apt, in his sense—more on this below. In some contexts, however—including, crucially, the case of truth itself—Wright seems by my lights to be strangely blind to the expressivist option. In particular, he seems blind to the second ingredient in the recipe above, that of a non-metaphysical route to an understanding of a philosophical topic. In order to have this possibility clearly in view, let me describe it in a little more detail, and contrast it to some methodological remarks from Wright himself.

In the disagreements with rivals, expressivists need to be clear that they take the relevant explananda to be aspects of human linguistic and cognitive behaviour, rather than anything particularly ontological or metaphysical. In my own work I have often distinguished the expressivist’s explanatory project from metaphysics by calling it ‘anthropology’, or ‘psychology’. In FFT, as I noted above, I distinguished between analysis and explanation, and emphasised that concerning truth itself, what I was recommending was the latter. In recent work, I have noted that Ramsey is one early figure who sees the importance of this distinction.

[W]hen Ramsey comes this way he speaks of psychology, not anthropology. Here he is in ‘General Propositions and Causality’ ..., reflecting on a possible response to the account of causation he has just sketched – an account that we might now call expressivist, or pragmatist.

What we have said is, I think, a sufficient outline of the answers to the relevant problems of analysis, but it is apt to leave us muddled and unsatisfied as to what seems the main question—a question not of psychological analysis but of metaphysics which is ‘Is causation a reality or a fiction; and, if a fiction, is it useful or misleading, arbitrary or indispensable?’ (Ramsey, 1929, p. 141)

Ramsey doesn’t address this concern directly, but I think it is clear that his view is that metaphysics is the wrong mode of enquiry, in this case. The illuminating enquiry is what he calls ‘psychological analysis’—an investigation into how we come to think and talk in causal terms, conducted in a manner that we do not presuppose that the helpful answer will lead us back to the objects. (In other words, we do not presuppose that the answer will be ‘We talk this way because we are keeping track of the causal facts’, or anything of that kind.) (Price, 2017, p. 151)

A theorist who does say that we talk this way because we are keeping track of the causal facts, or its equivalent in the ethical case, is someone who has acknowledged the expressivist’s questions about language, but taken them to lead back to ontological issues. That’s the possibility that Dreier has in mind when he says that in contrast, the expressivist’s view is that ‘to explain what it is to make a moral judgment, we need not mention any normative properties’ (2004, p. 39).Footnote 19

For present purposes, the crucial point is that for an expressivist, the linguistic or psychological questions are where we start. Metaphysics often starts somewhere else, such as with questions about the nature and reality of causation, moral properties, or whatever. (As Ramsey puts it, ‘Is causation a reality or a fiction; and, if a fiction, is it useful or misleading, arbitrary or indispensable?’) The difference between expressivism and its rivals often shows up as a difference in which questions are taken to be worth asking.

I want to contrast this aspect of expressivism with a methodological perspective outlined by Wright himself, in a paper in which he is defending, as I would call it, the project of a metaphysics of truth. Arguing that such a project need not depend on conceptual analysis in the traditional sense, Wright notes the generality of what he has in mind:

[S]uch issues arise for any putative characteristic, \({\upphi }\). Should we (ontologists) take \({\upphi }\) seriously at all, or is some sort of error-theoretic or deflationary view appropriate? If we do take it seriously, should we think of the situation of an item’s being \({{\upphi }}\) as purely a matter of how it is intrinsically with that item, or are we rather dealing with some form of relation? Is an item’s being \({{\upphi }}\) an objective matter (and what does it mean to say so)? These are analytic-philosophical issues par excellence, but their resolution need not await – and might not be settled by – the provision of a correct conceptual analysis. (Wright, 1998, p. 35)

My point, and the point I take to be illustrated by Ramsey’s remarks about causation, is that there is an alternative to what Wright here calls ontology, as a path to a philosophically illuminating understanding of a notion we find in common use. The expressivist simply doesn’t ask these questions (at least as starting points, though some of them may turn out to have expressivist readings). Nevertheless, like Ramsey about causation, she claims to show us something interesting, important, and central about the notion \({{\upphi }}\) in question.

Wright certainly has this expressivist perspective on his radar, in some cases. Indeed, as we’ve seen, he appears to take it to be mandatory for any domain in which truth is sufficiently minimal, in his sense. But he does not seem to countenance it, or at least in my view insufficiently so, in the case of truth itself. This will emerge as one of the biggest differences between us.

3 Wright on truth and objectivity

3.1 The normativity of truth

I noted above that ‘Metaphysical pluralism’ (Price, 1992) predates Wright’s Truth and Objectivity. When I first wrote about Truth and Objectivity, a few years later, it was from a different angle. I was interested in Wright’s ‘Inflationary Argument’ against Paul Horwich’s version of a deflationary theory of truth. The debate between Horwich and Wright sets the scene in my piece ‘Three norms of assertibility, or how the MOA became extinct’ (Price, 1998). This is how I describe my own viewpoint.

My own view, in a nutshell, is that Horwich wins the battle but Wright wins the war. I think that truth is normative, in a way not explained by [Horwich’s] deflationary theory; but that Wright has not given us a good argument for this conclusion. In this paper I want to reinforce Horwich’s objections to Wright’s argument, but then to offer an alternative argument to the same conclusion. As I’ll explain, however, this conclusion does not require that we abandon ... the claim that truth is not a substantial property. It simply requires that our explanation of the folk use of the concept of truth should not be grounded solely on [the disquotational schema], but rather needs to appeal to the utility within a speech community of the distinctive kind of norm that truth provides. (Price, 1998, pp. 241–242)

I won’t rehash my objection to Wright’s Inflationary Argument.Footnote 20 Rather, I want to focus on the alternative case I make, in (Price, 1998) and elsewhere, for thinking that truth is normative, in a way not captured by Horwich’s view.

Happily, I can now describe this alternative case using materials that Wright himself provides in one of his most recent pieces. In ‘Alethic pluralism, deflationism, and faultless disagreement’ (Wright, 2021), Wright puts on the table many of the considerations on which my argument rests (including, though in an indirect sense, the central point). This piece is so useful for our present purposes that I’ll devote the next subsection to a summary of the key points, mainly in Wright’s own words. I’ll add a few flags from my side, to explain how Wright’s points connect with various of my own interests (and return to these points later in the paper).

3.2 Alethic pluralism, deflationism, and faultless disagreement

Wright puts the question he wants to address like this:

One of the most important “folk” anti-realist thoughts about certain areas of our thought and discourse—basic taste, for instance, or comedy—is that their lack of objectivity crystallises in the possibility of “faultless disagreements”: situations where one party accepts P, another rejects P, and neither is guilty of any kind of mistake of substance or shortcoming of cognitive process. On close inspection, however, it proves challenging to make coherent sense of this idea, and a majority of theorists have come to reject it as incoherent. (Wright, 2021, p. 432)

He notes that there are ‘two significant exceptions’ to this consensus ‘in the contemporary literature’, with both of which he takes issue:

[R]elativists often hold it up as something of a coup for their view that it can make straightforward sense of faultless disagreement; and the author of this paper has argued (Wright, 2006) that making judicious intuitionistic revisions to classical logic can provide resources that suffice to stabilise the notion. The present paper argues that neither relativism nor intuitionism in fact provides a satisfactory account and indicates how an alethic pluralist framework enables us to do better. (Wright, 2021, p. 432)

Flag 1

The topics that Wright introduces here have been interesting to me at several levels, and at several periods. In some of my earliest work in the 1980s, ‘no fault disagreements’, as I called them in FFT, played a central role. Four decades later, one of my most recent pieces (Price, 2022) compares and contrasts my way of dealing with such phenomena with John MacFarlane’s relativism. In the latter case Wright and I turn out, once again, to share a common opponent – though there are differences, as I’ll explain, in what we offer as an alternative to relativism.

Wright describes the linguistic phenomena in question like this:

Does this idea [of faultless disagreement] really make sense, however? The idea that it does is, I think, entrenched in folk-philosophical thought about some of our most basic values—notably values of culinary taste, comedy, and some values of social propriety, aesthetics, and (for some folk thinkers) ethics. The value judgements concerned are cases where we feel parties can genuinely disagree—that is, affirm mutually incompatible opinions—without anyone ... being mistaken about the matter at hand. ...

Call this kind of thinking the Folk View of (some) basic evaluations. ... My question is whether it makes sense at all, and if so, what kind of framework is needed, or able, to safeguard its coherence. (pp. 433–434)

He offers this as a familiar example:

Some people find sushi delicious. Others are repelled by the characteristic odour of raw fish and find sushi highly disagreeable. The proposition (P) that sushi is delicious is a good example, if anything is a good example, of the kind of evaluation that is apt to give rise to faultless disagreements of the kind sanctioned by the Folk View. (p. 434)

Wright distinguishes a number of requirements that a satisfactory vindication of the Folk View would need to meet. One of them, in particular, he regards as problematic for his own earlier intuitionistic proposal for accounting for the Folk View:

Call this ... requirement Parity. In effect, it is the requirement that faultlessness be appreciable, and endorseable, from the point of view not just of neutrals but also of the committed parties in the relevant kind of dispute.

How are we to explain Parity? Here Wright describes the crucial move, according to his new view:

Our problem now is Parity: in opining that sushi is delicious, you opine that it is false that it is disagreeable, and hence that my opinion is false. So how can you regard that opinion as no worse than your own?

Well, regarding my opinion as false compromises its parity with yours only if “false” carries its normal normative punch. It is here that I wish to invoke aspects of the treatment of truth and truth aptitude that I proposed in Truth and Objectivity (1992) ... (p. 438)

He relates this to a central claim of Truth and Objectivity.

A central contention of Truth and Objectivity was that—at least over merely minimally truth-apt discourses—truth need carry no payload of accurate substantial representation. When merely minimally truth-apt claims are at stake, to regard a statement as false need not be to attribute any representational fault to someone’s acceptance of it. So if there need be no other kind of fault, the way is open for the idea that, in such a case, to describe an opponent’s view as “false” is, in effect, to go no further than to record one’s disagreement with it, with no implication of any further deficiency. There would be an imputation of fault, and hence a compromise of Parity, only when “true” demands some kind of richer interpretation. But where merely minimally truth-apt discourses are concerned, is there any good reason to suppose that this has to be so in general? (p. 439)

Flag 2

To foreshadow the use I want to make of this material below, let me note that in (Price, 1998) I argue that, in different ways, both Horwich and Wright miss or misunderstand what Wright here calls the ‘normal normative punch’ of truth and falsity. Here Wright is arguing (i) that in the case of minimally truth-apt claims we can do without this normative punch altogether; and (ii) that in other cases we can take it to be grounded on a norm of ‘accurate substantial representation’. As I’ll explain, I disagree on both points.

Wright claims that his ‘minimalist’ proposal makes better sense than relativism of the Folk View.

When you affirm the truth of your view [that sushi is delicious], you are not to be interpreted, as author relativism will interpret you, as committed merely to its satisfaction of a truth condition that simply has no role in my assessment and the satisfaction of which I do not dispute. Rather just as, on the surface, it appears you are committed to an appraisal of sushi that I do indeed reject. But, again, your commitment to the truth of your appraisal need involve no imputation of fault to me. You are indeed committed to the falsity of my opinion. But since this is merely minimal falsity, and tagging my opinion as “false” is, where basic taste is concerned, simply another way of expressing your disagreement with it, with no implication of cognitive-procedural or alethic shortcoming on my part, my view can be none the worse for that.

The suggestion, in summary, is that locally minimal—fully deflated—notions of truth and falsity are available to allow one to describe a contested opinion as false without thereby doing more than recording one’s disagreement with it, in particular without imputing any kind of cognitive or other shortcoming to its author. “Locally” is of course important. A global deflationist who took this line would have to explain what makes the difference when, in imputing falsity to a view, one precisely does intend to impute fault. That does not mean that global deflationism cannot avail itself of the present suggestion. But it does mean that an account of the distinction between disputes that are liable to betoken faultless disagreement and disputes of more substantial matters, where shortcoming of some kind is essentially involved, will have to proceed in other terms. That someone thinks something false, merely, will underdetermine the issue. By contrast, within the alethic pluralist framework of Truth and Objectivity, fully deflationary conceptions of truth and falsity can be reserved for the problematic subject matters with which we are currently concerned, without any commitment to so conceiving of truth across the board. (Wright, 2021, pp. 339–440; here and in the remainder of the paper I use underlining for my own added emphasis)

We then get the remarks that I quoted in (Sect. 2.2):

[L]et us be mindful that, where truth is deflated and so registers no independent norm operative over the acceptance and assertion of statements, an account of meaning—of correct linguistic practice—has to proceed in other, non-truth-conditional, broadly use-theoretic terms. So we need to focus directly on the use of the signature statements in discourses where the Folk View seems intuitively apt. (p. 441)

Wright then introduces a distinction very similar to one on which, as we’ll see, I relied in Price (1998):

What we then find is that there is, in these discourses, a distinction in use between what I shall here term the objectifying idiom exemplified by “Sushi is delicious” and “The party is going to be fun” and the corresponding subjectiverelational [S–R] reports: “I find sushi delicious” or “Sushi tastes delicious to me,” and “I’m going to enjoy the party” or “The party will be fun for me.” Moreover, it appears that our practice is to treat the objectifying claims as in general somehow stronger: witness that, in a wide class of contexts, a subjective-relational claim provides a fallback when an objectifying statement runs into difficulty. Filippo asserts, “The sushi is delicious” but then finds that all his dining companions are expressing regret at ordering it and falls back to “Well, I am enjoying it at any rate.” (p. 441)

After that, we get more music for my expressivist ears, in two senses. Wright detaches assertoric content from truth-conditions, and then proceeds to introduce exactly the explanatory stance on linguistic practice that the expressivist requires.

We can, and must, drop the idea that assertoric content has to go hand in hand with  truth-conditional  content.  To be sure,  assertoric content does go hand in hand  with  amenability  to  a  disquotational  truth  predicate,  but  it is a further step  to  take  this to be amenability to representation or some other substantial notion of truth. ...

How else, then? This is not the occasion to embark on a full development of the minimalist alternative. But in barest outline, in the second stage of a fuller development, the question focused on will not be: what kind of fact must O-statements [i.e., objectifying statements] be taken to describe if both their assertibility on the basis of an appropriate S–R response and their conditions of defeasibility adumbrated above are to be explained, but rather: what   point   would   the   institution   of   such   assertions   serve—why   would   it be   worthwhile   having   a   practice  in  which  such  statements  were  treated  as assertible   on   the   basis   of  S–R  responses  but  defeasible  under  the  kinds  of conditions reviewed? (p. 443)

Flag 3

We’ll see below that I asked a very similar question to this one, but with a much broader scope. Taking my cue from Dummett, I asked it about assertions in general.

Wright outlines a possible answer to this question, and notes how close it takes him to expressivism.

Focusing now on the positive S–R responses, a reminder may be apt of a range of mundane and contingent but very important facts about them. First, in a wide class of cases, our enjoyment of values of taste, the intensity of the associated S–R responses, is characteristically enhanced by sharing and socialisation: the ride is more fun when others are with you and enjoying it too; we like to eat together; we—most of us, at least—prefer to go to the theatre with friends. Second, we do naturally share many of these responses. Third, they are also in many cases to a high degree tractable—one can acquire and refine patterns of response of these kinds by experience and education. Fourth, many of these responses have a rich causal provenance in their objects, which is receptive to study, technique, and manufacture—to the arts of cuisine, comedy, musical composition and performance, dance, and drama. Fifth, we do regard these responses as subject to conditions of appropriateness in the light of other of our social and personal values.

All of these factors combine to create a situation where we have an interest in having an idiom that enables us, more than merely reporting a response we personally have, to project it as a possible point of coordination, something that may be shared and thereby enhanced, is dependable rather than ephemeral, something that is a reaction of our normal, healthy selves, and free of taints of spite, schadenfreude, cruelty, and other morally reprehensible features, and whose causal prompts it may be worthwhile to understand with a view to developing an associated art.

I am not of course suggesting that ordinary speakers characteristically have such considerations in mind in making O-statements. Rather, even in this whistle-stop overview, the beginnings can be seen of how an account might run of the social utility of an objectifying idiom of taste, or comedy, which both assigns the importance it had better assign to grounding in personal responses and explains the broad range of defeaters we have noted without any need to reconceive the content of O-statements along contextualist or relativist lines, or to imbue them with substantial truth conditions, or to query appearances of disagreement where ordinary speakers take it to occur. This  minimalist  approach  shares with expressivism a rejection of the idea that in making such statements, we are normally in the business of trying to “report the  facts”;  but its expressivism is advanced as a thesis of pragmatics—a thesis concerning   what   participants  in   a   discourse   are  characteristically  doing  in endorsing its distinctive kind nof evaluations—not a claim about the semantics of the statements in question. And minimalism agrees with relativism both in accepting that basic disagreements about taste are just that—disagreements focused on exactly the shared propositional content that they seem to concern—and in rejecting the idea that in asserting or denying such a content, one purports to represent an objective fact. This anti-realism, however, is now accomplished without any need for relativistic manoeuvrings with the truth predicate. (pp. 443–444)

Flag 4

In a deflated sense my expressivist is happy to say that we are ‘in the business of trying to “report the facts”.’ The important thing is that no explanatory weight gets carried by the facts in question. On the contrary, the talk of facts has to be seen as ‘downstream’ of the use-theoretic description of the function of the discourse in question—downstream of the ‘pragmatics’, as Wright puts it. My expressivist can say the same thing about purporting to represent an objective fact, in fact, though here my difference with Wright turns on my different treatment of the no fault cases—more on that below.

Finally, Wright summarises the advantages of his approach to faultless disagreement compared to that of relativism, as he sees them.

[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 to subserve the contrasts whose contours are exploited by the Inflationary Argument. I have argued that the first option will not deliver what is wanted, and that the second is the way to go. I regard it as a strength of the combination of minimalism about truth aptitude and alethic pluralism defended in Truth and Objectivity that it provides a natural setting for the elaboration of the second direction. (p. 445)

Flag 5

As we’ll see, I think that there is a third alternative, one that both Wright and MacFarlane miss. It is that truth provides a flexible norm, ‘positive-presumptive’, to borrow another Wrightean phrase (Wright, 1989, p. 251), but cancellable. The ordinary business of ascribing fault to speakers with whom we disagree comes with ‘escape hatches’—ways of stepping back from a dispute, and cancelling an accusation of fault. What is distinctive about examples such as predicates of taste (or, for different reasons, ascriptions of probability) is that these escape hatches often lie close to the surface. But no topic is entirely immune from them, in my view—that’s why what I offered in FFT was a global view.

Enough flag-waving, I’m getting ahead of myself. But as the flags indicate, my views connect with several elements in this recent piece of Wright’s, though the package I offer is in some ways very different. To explain where the elements originate in my case, I need to reach a long way back—all the way back to my Cambridge PhD thesis (Price, 1981).

4 Explaining objectivity and its limits

4.1 The origins of Pricean expressivism

My PhD thesis was a defence of what I would now call an expressivist view of claims of the form ‘It is probable that P.’ In other words, it proposed a version of what Yalcin (2012) calls ‘credal’ expressivism. Influenced by Dummett, I put the central claim in terms of the Fregean sense–force distinction. I claimed that ‘It is probable that ...’ and similar probabilistic expressions modify the force of an utterance, not its sense. I described utterances of this form as partial assertions, extending the familiar terminology of partial belief.

At the time—the minimalist alternative being far in the future—I took it that this conclusion required an argument that such claims don’t have genuine truth conditions. One of the arguments I offered was that disputes about such claims have a strange feature. What seems like a straightforward ‘factual’ disagreement can ‘evaporate’ (as I put it later in FFT), when it turns out that the parties are relying on different evidence. At least in some such cases, neither party feels that the other is at fault, once both understand the source of the disagreement. I argued that genuinely truth-conditional claims would not exhibit such behaviour. Thus I already had in play a kind of faultless disagreement, linked to what I would later call a kind of expressivism.Footnote 21

In the years immediately after my PhD I had a postdoc at ANU, Canberra. This gave me time to write up this central argument from my thesis, which appeared as ‘Does “Probably?” modify sense?’ (Price, 1983a)—my first published paper, if I recall correctly. It also gave me time to dig more deeply into Dummett’s work, which became a major influence and focus, central to two other papers from the same year (Price, 1983b, c). By describing the use I made of some of Dummett’s famous remarks on truth, I can explain why some of the passages I quoted from Wright’s recent paper seem so congenial to me, as well as the respects in which I take similar thoughts in different directions.

One of my Dummett-inspired papers from that period is a piece called ‘Could a Question be True? Assent and the Basis of Meaning’.(Price, 1983c). This piece stemmed from my interests in negation and disagreement, and the sense–force distinction, and the connections between the two. By this point, my concern with what I saw as ‘non-standard’ ascriptions of truth and falsity—e.g., ascriptions to probability judgements, which I took to lack truth conditions—had led me to Dummett’s question: Why do we call each other’s utterances ‘true’ and ‘false’ in the first place? What is the point of doing so, as Dummett puts it?

At one time it was usual to say that we do not call ethical statements ‘true’ or ‘false’, and from this many consequences for ethics were held to flow. But the question is not whether these words are in practice applied to ethical statements, but whether, if they were so applied, the point of doing so would be the same as the point of applying them to statements of other kinds, and, if not, in what ways it would be different. (Dummett, 1959, p. 3)

In (Price, 1983c) I proposed an answer. It was the beginnings of an account of the role of truth and falsity that was to play a central role in FFT, and in later work, such as (Price, 1998, 2003). But it is this early paper that ties the proposal most directly to Dummett.

I begin by asking ‘why it is not appropriate to assent to or dissent from utterances such as questions, commands and requests’ (1983c, 355). I note that ‘[o]ne function of the terms “true” and “false” is to provide a uniform means of endorsing or rejecting a statement made by a previous speaker’, and that this is in effect Strawson’s (1949) account. ‘However’, I say,

if this were all that ’true’ and ’false’ did, it would be unclear why they shouldn’t be applied to utterances other than assertions – to questions, for example. There are many circumstances in which it would be useful to have a simple device for putting one’s weight behind a question asked by someone else; for indicating, in effect, that one is asking the same question oneself. Similarly, there are situations in which it would be useful to be able to indicate in a simple way that a question did not have one’s endorsement. ... If a major function of the terms ’true’ and ’false’ is to provide a simple means of so endorsing or rejecting a previous utterance, why should these terms be applied only to assertions, and not to utterances of other force-types?Footnote 22 (1983c, p. 355)

I note that another possible answer is ‘that questions (commands, requests, etc.) don’t have truth-conditions, and are therefore not properly described as true or false’ (1983c, p. 355). But I argue that this looks ‘suspiciously circular’, especially so if ‘holding true’ and ‘holding false’ are supposed to provide the evidential basis for a theory of meaning, as in the Davidsonian framework I had in the background. I suggest that a more promising approach

lies in the claim that, as Dummett puts it, ‘the roots of the notions of truth and falsity lie in the distinction between a speaker’s being, objectively, right or wrong in what he says when he makes an assertion’ (1978, p. xvii). That is, to call an utterance ‘true’ is to say that it (or its speaker) is correct, or right; to call it ‘false’, to say that it (or the speaker) is incorrect, or wrong. (1983c, p. 355)

However, I say, ‘[t]his is a step in the right direction but not a solution’:

For it has not been explained how to call an utterance ‘correct’ is to do more than simply endorse it (in the sense seen to be equally appropriate for questions); and to call it ‘incorrect’, to do more than to reject or decline to endorse it. It is no use appealing to further ways of saying that an utterance is incorrect, or wrong (that it is ‘at fault’, say, or ‘mistaken’). Rather we need an account of the function of this collection of descriptions, and some understanding of their consequences in typical conversations. (1983c, pp. 355–356)

In effect, then, I was suggesting that Dummett does not push his own question about truth far enough. We need to ask about the point of having some such norm in the first place. I then offer my own proposal, which, in four paragraphs, is the core of the account of the functions of the norms of truth and falsity that I was to develop in FFT and later work.

It seems to me that the primary significance of these forms of criticism lies in the fact that they constitute a challenge to a speaker to justify an utterance, and an indication of readiness on the part of the critic to engage in a dispute. In such a dispute, 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.

For the deliberately vague term ‘views’ here, it is natural to read ‘beliefs’. There would seem too little point in such a “dispute” with respect to utterances expressing, say, desires, which could reasonably vary from speaker to speaker, even in the face of the same evidence as to matters of fact. In other words, the suggestion is this: some utterances (call them ‘assertions’) characteristically express states of mind (‘beliefs’) with respect to which there is reason to seek agreement between speakers. For these states of mind, “two heads are better than one”; there is a general advantage in exhibiting differences between speakers in this respect, so that less well-justified beliefs may be replaced by more well-justified ones. 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.

The claim is thus that there would be a certain value in a general device for merely, in effect, repeating or declining to repeat a previous utterance; this value would lie in brevity and convenience. But the process thereby facilitated would be one of less significance than that which forms a part of a dispute procedure, whereby conflicting beliefs are brought under common scrutiny, to the general advantage of the speakers concerned. So the facilitating of this latter process is the more important task of the terms ‘true’ and ‘false’, and therefore the one which takes priority, in determining their extension to utterances of a particular kind. (1983c, pp. 356–357)

4.2 Comparisons with Wright and MacFarlane

Let me pause to compare this proposal to the one offered by Wright (2021), to explain why we give voice to value judgements in what he calls the objectifying idiom. As Wright puts it,

we have an interest in having an idiom that enables us, more than merely reporting a response we personally have, to project it as a possible point of coordination. (2021, p. 443)

My suggestion is in one sense very much like Wright’s. But there are two big differences. First, I’m applying the idea much more generally, to all ‘states of mind ... with respect to which there is reason to seek agreement between speakers’ (1983c, p. 356). Second, I’m suggesting that the normative dimension of truth and falsity, the very thing Wright is proposing to abandon in the case of the minimal truth, is a crucial element. It is what provides pressure for coordination, what distinguishes truth as we have it from a mere device for saying ‘Ditto’.

By way of comparison, think of a group of diners, asked by a waiter for their drink orders. Their responses are more than mere reports of preference, of course. (Imagine the reaction of the waiter, if one of them says, ‘But I didn’t order this coffee—when I said “I would like a coffee” I was simply describing my state of mind, nothing more.’) So the diners are doing more than ‘merely reporting a response [they] personally have’, as Wright puts it. Moreover, their choices certainly provide ‘a possible point of coordination’, as they may find out if the waiter restricts their options—‘Sorry, guv, the kitchen can’t do a pot of tea just for one.’ And simple verbal devices such as ‘Ditto’, ‘Same again’ or ‘Not me’, for piggybacking on utterances by other diners, can be useful in these contexts, too. But these devices provide no pressure for coordination, in normal circumstances, because they are not normative.Footnote 23

Since Wright has relativists in his sights, let me also compare my proposal to one that MacFarlane offers, where he asks about the point of disagreement in ‘subjective’ cases, such as matters of taste. MacFarlane asks:

[W]hat is the point of fostering controversy in “subjective” domains, if there is no (nonrelative) truth on which both parties can converge? Why shouldn’t we just talk about our own tastes, rather than ascribing subjective properties to the objects?

Perhaps the point is to bring about agreement by leading our interlocutors into relevantly different contexts of assessment. If you say “skiing is fun” and I contradict you, it is not because I think that the proposition you asserted is false as assessed by you in your current situation, with the affective attitudes you now have, but because I hope to change these attitudes. Perhaps, then, the point of using controversy inducing assessment-sensitive vocabulary is to foster coordination of contexts. We have an interest in sharing standards of taste, senses of humor, and epistemic states with those around us. The reasons are different in each case. In the case of humor, we want people to appreciate our jokes, and we want them to tell jokes we appreciate. In the case of epistemic states, it is manifestly in our interest to share a picture of the world, and to learn from others when they know things that we do not. (MacFarlane, 2007, p. 30)

Like Wright, MacFarlane is proposing this answer for the subjective cases, taking for granted, apparently, that some other account will work elsewhere.Footnote 24 My view, in contrast, is that in order to answer Dummett’s question about the point of truth, we need to be more ambitious. We should ask this question about the point of ‘fostering controversy’ as a question about assertoric language in general, rather than about these special cases. In case it is unclear how the answer can be anything but obvious in the general case, I note two things.

First, it is easy to find ways of cancelling controversy, for any sort of assertion whatsoever. It isn’t just in these evaluative cases that we can fall back, as Wright notes, to a subjective idiom. On the contrary, we do this kind of thing all the time when we want to reduce the temperature of a disagreement. Instead of simply saying ‘P’, I say ‘Well, my own view on this is that P.’ It is worth asking why this isn’t the default. In other words, why doesn’t saying ‘P’ have this more gentle, controversy-reducing feel from the start? More on this thought below—it was a central point in (Price, 1998).

Second, any seemingly obvious answer in the general case is likely to appeal to our intuitions about truth and falsity, e.g., by pointing out that where two speakers disagree about a non-subjective case, one of them believes something false. But that won’t do if our interest is in getting at the ‘point’ of truth and falsity themselves, in Dummett’s sense.

4.3 Challenging the bifurcation thesis

After introducing the proposal above, (Price, 1983c) goes on to discuss some different ways in which two speakers can disagree, based on differences of evidence and other things. It defends a criterion for being a genuine assertion, or a genuinely truth-conditional claim, on which I had relied in my thesis and in (Price, 1983a). In the language I was to use later in FFT, the test is a matter of not permitting no fault disagreements (NFDs).Footnote 25

A few years later, however, I had come to doubt whether the NFD criterion could mark a sharp line. One crucial influence was exposure to the so-called rule-following considerations, in Kripke’s (1982) book, Wittgenstein on Rules and Private Language. I had had the opportunity to discuss this book at ANU in 1983, in a reading group with Philip Pettit, Peter Menzies, and others. I eventually became convinced that the ineliminable possibility that different speakers might diverge from one another, in virtue of possessing different dispositions to ‘go on in the same way’, was a universal potential source of NFDs. If so, NFDs couldn’t possibly mark a boundary between genuinely truth-conditional uses of declarative language and all the rest (except in the trivial sense in which there is a boundary between a null class and everything else).

Thus, my reliance on NFDs as a mark of the kind of non-factuality I had wanted to attribute to probabilistic claims, had pushed me to the conclusion that no uses of assertoric language were fully factual, in the assumed sense. And this seemed to me an appealing option to have on the table, not least in Kripke’s own case. I felt that the assessment of Kripke’s ‘sceptical solution’, by Kripke himself amongst others, was insufficiently sensitive to the broader implications of the arguments. To put it crudely, it was less shocking that there were no genuine facts about meaning, if one had in any case had to renounce the conception of genuinely factual language, against which the contrast was supposedly being drawn.Footnote 26

By the time I wrote FFT, I had thus become sceptical of an assumption on which I had relied in Price (1983a) and Price (1983c), and which was, and still is, very widely taken for granted. This assumption is the so-called Bifurcation Thesis—the view that there is a division within the class of indicative utterances between genuinely factual claims, on one side, and utterances with some other role or status, on the other.Footnote 27 I had become convinced of the attractions of a kind of global non-factualism, a view that treated the familiar expressivism about ethics and probability as some of the tips of an all-encompassing iceberg. And I had come to this position, in part, by thinking about the role of responses such as ‘That’s true’ or ‘That’s false’ in dialogue—both in general, and in what I had previously seen as special cases, such as probability judgements. Together, these conclusions correspond to the main themes of FFT: rejection of the Bifurcation Thesis and of the notion of strict or genuine factuality on which it depends; a defence of a kind of global non-factualism; and an account of truth and falsity, and their limits, in terms in terms of their normative role in disagreement.

4.4 Truth as convenient friction

In the years after FFT, I developed this ‘dialectical’ approach to truth in two main papers. The first of them (Price, 1998) is the one inspired by Wright’s disagreement with Horwich. It proceeds by distinguishing the conversational norm I take to be associated with truth and falsity from two other norms, those of sincerity and warranted assertibility. Truth is thus my ‘third norm’, and the focus of the paper is on what this third norm adds to our conversational practice—on the question how things would be different if we didn’t have it.Footnote 28

As an aid to answering this question, I imagine a speech community who don’t have the third norm, but who nevertheless use speech acts to give voice to their opinions. I call such speech acts ‘merely opinionated assertions’, or MOAs, and imagine them used by a community called the Mo’ans. As I point out, and noted above, it is easy to get to something like MOAs in our own linguistic practices, by using devices to cancel the third norm: ‘My own opinion is that P’; ‘Mine is that not-P.’ In Wright’s (2021) terms, then, the Mo’ans are speakers who don’t, or no longer, have the device of putting their opinions into a public realm.Footnote 29

I cover similar ground in ‘Truth as convenient friction’ (Price, 2003). This paper, too, has a connection to Wright. It began as a response to a piece in which Rorty criticises Wright’s conclusions in Truth and Objectivity. Rorty’s main target is Wright’s defence of truth as a norm distinct from warranted assertibility. Once again, I argue that Wright has the better of this contest, though for reasons he himself does not clearly identify. The paper begins like this.

Rorty begins by telling us why pragmatists such as himself are inclined to identify truth with justification:

Pragmatists think that if something makes no difference to practice, it should make no difference to philosophy. This conviction makes them suspicious of the distinction between justification and truth, for that distinction makes no difference to my decisions about what to do. (Rorty, 1995, p. 19)

Rorty goes on to discuss the claim, defended by Wright, that truth is a normative constraint on assertion. He argues that this claim runs foul of this principle of no difference without a practical difference:

The need to justify our beliefs to ourselves and our fellow agents subjects us to norms, and obedience to these norms produces a behavioural pattern that we must detect in others before confidently attributing beliefs to them. But there seems to be no occasion to look for obedience to an additional norm—the commandment to seek the truth. For—to return to the pragmatist doubt with which I began—obedience to that commandment will produce no behaviour not produced by the need to offer justification. (1995, p. 26)

Again, then, Rorty appeals to the claim that a commitment to a norm of truth rather than a norm of justification makes no behavioural difference. (Price, 2003, p. 163)

I argue, on the contrary, that the third norm makes a huge behavioural difference:

I want to maintain that in order to account for a core part of ordinary conversational practice, we must allow that speakers take themselves and their fellows to be governed by a norm stronger than that of justification. ... [I]t 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. It gives disagreement its immediate normative character, a character on which dialogue depends, and a character which no lesser norm could provide.

This fact about truth has been overlooked, I think, because the norm in question is so familiar, so much a given of ordinary linguistic practice, that it is very hard to see. Ordinarily we look through it, rather than at it. In order to make it visible, we need a sense of how things would be different without it. Hence, in part, my reason for beginning with Rorty. Though I disagree with Rorty about the behavioural consequences of a commitment to ‘a distinction between justification and truth’, I think that the issue of the behavioural consequences of such a commitment embodies precisely the perspective we need, in order to bring into focus this fundamental aspect of the normative structure of dialogue. (2003, p. 164)

Invoking the MOA again, I go on to argue that ‘some of the basic functions of assertoric discourse could be fulfilled in an analogous way, by a practice which lacked the third norm.’

But it will be clear, I hope, that that practice would not support dialogue as we know it. What is missing – what the third norm provides – is the automatic and quite unconscious sense of engagement in common purpose that distinguishes assertoric dialogue from a mere roll call of individual opinion. Truth is the grit that makes our individual opinions engage with one another. Truth puts the cogs in cognition, at least in its public manifestations. (2003, p. 165)

Again, let me note the similarity between this view and the ones that Wright (2021) and MacFarlane (2007) suggest for the special cases. The similarity is even more apparent in this passage:

The third norm doesn’t just hold open the conceptual space for the idea of improvement. It positively encourages such improvement, by motivating speakers who disagree to try to resolve their disagreement. Without the third norm, differences of opinion would simply slide past one another. Differences of opinion would seem as inconsequential as differences of preference. With the third norm, however, disagreement automatically becomes normatively loaded. The third norm makes what would otherwise be no-fault disagreements into unstable social situations, whose instability is only resolved by argument and consequent agreement – and it provides an immediate incentive for argument, in that it holds out to the successful arguer the reward consisting in her community’s positive evaluation of her dialectical position. If reasoned argument is generally beneficial – beneficial in some long-run sense – then a community of Mo’ans who adopt this practice will tend to prosper, compared to a community who do not. (2003, pp. 174–175)

What Wright and MacFarlane both miss, in my view, is the need for an explanation of this kind in the (so-called) non-subjective cases.

4.5 Blackburn’s missing bull’s-eye

To be fair, Wright and MacFarlane are in excellent company. This deeper issue about disagreement is on almost nobody’s radar. In a later piece (Price, 2006), I make a similar point against Blackburn’s Ramseyan version of minimalism about truth. I use a different analogy to highlight the question posed by the Mo’ans. Since the point is so crucial to the senses in which I both agree and disagree with Wright, I want to explain this alternative analogy as well.

I begin by distinguishing two kinds of examination. In the first (‘autological’) case, the aim is simply to be sincere, to say what you actually believe. If the question is ‘Is Aristotle a Belgian?’ then the right answer is ‘Yes’ if you believe that he is, and ‘No’ if you believe that he isn’t. In the second (‘heterological’) case, sincerity isn’t sufficient, or indeed necessary. You simply get marked right or wrong, depending on whether the examiners agree or disagree with what you say. I then use this distinction to make a parallel point about assertion.

We can draw a parallel distinction between two kinds of assertion, or conversation. In one kind (‘autological conversation’), the aim is simply to give voice to what one actually believes, as accurately as possible. In the other kind (‘heterological conversation’), the aim is as if one’s assertions were answers in a heterological examination, with one’s interlocutors taking the role of the examiners. In the second case but not the first, in other words, it is as if one’s utterances are being held to be accountable to an objective standard—a standard that mere sincerity doesn’t guarantee.

As in the case of heterological exams, our interlocutors don’t have to be gods, to subject our assertions to this kind of ‘objective’ standard. It is sufficient that they apply the standard on the basis of their own beliefs: if we say ‘Aristotle was a Greek’ and they believe that he was actually Belgian, they are entitled to say, ‘That’s not correct—your claim is mistaken.’ Indeed, the main difference between this case and that of the corresponding exams is that conversation is symmetric. Both sides are playing the same game, and each is entitled to act as ‘examiner’ with respect to the other. (Heterological conversation is the heterological examination of all against all, as it were.)

I hope it is clear that as a matter of fact, our conversational practices are substantially those of heterological conversation rather than autological conversation. In my view, this fact is of great significance for [an understanding of truth]—though its significance has usually been overlooked, I think, probably because the relevant features of conversational practice are so familiar as to seem unremarkable. ...

[T]o all intents and purposes, the question ‘What is truth?’ is the same question as ‘What is this correctness and incorrectness, rightness and wrongness, that we claim for various moves in our conversational games?’ The point of the case of autological conversation [as of the MOAs] is to throw the significance of these questions into relief, by pointing out that we can imagine a conversational game that lacks this particular kind of normative standard. And the bare redundancy theory of truth doesn’t address the issue, because the truth predicate it offers us is so thin that it works equally well in either game.

Like most minimalists, Blackburn misses this point. But what he misses is not the fact that conversation has such a normative standard. On the contrary, here he is calling loud attention to it, by way of pointing out where some forms of relativism go wrong:

[T]here is no one place from which it is right to look at the Eiffel tower, and indeed no place that is better than another, except for one purpose or another. But when it comes to our commitments, we cannot think this. If I believe that O. J. Simpson murdered his wife, then I cannot at the same time hold that the point of view that he did not is equally good. It follows from my belief that anyone who holds he did not murder his wife is wrong. They may be excusable, but they are out of touch or misled or thinking wishfully or badly placed to judge. I have hit a bull’s-eye, which they have missed. (Blackburn, 2005, pp. 65–66, my italics)

Rather, what Blackburn misses is what I’ve highlighted by distinguishing two kinds of conversational games: the fact that there’s something important that needs explaining here, viz., that in the game as we actually play it, there is a norm, or a bull’s-eye, of precisely this kind. At another point he says that ‘[t]o make an assertion at all is to put a view into the public space, up for acceptance and rejection.’ (68) Again, he’s right, but he misses the question: why are our assertions treated like this? Why aren’t a chap’s beliefs treated as entirely his own affair, as it were—as they are, by default, in the game I’ve called autological conversation? ...

It follows that there’s a residual form of Pilate’s question [‘What is truth?’], absolutely untouched by the suggestion that we deflate truth in Ramsey’s way: What are these norms of ‘rightness’ and ‘wrongness’, ‘correctness’ and ‘incorrectness’—this bull’s-eye that distinguishes ordinary heterological conversation from the autological alternative? It might seem tempting to make the notions of rightness and correctness follow truth itself, in walking Ramsey’s deflationary plank—a temptation encouraged by the fact that for many purposes, ‘true’, ‘right’ and ‘correct’ are interchangeable. But this would simply deprive us of Blackburn’s distinction between ways of looking at the Eiffel tower and ways of taking a view about whether O. J. Simpson murdered his wife. For better or worse, our ordinary conversational practice does admit such a distinction, and we can’t sweep it under the carpet simply by forcing all the predicates we use to express it into the minimalist box—the right lesson, rather, is that we were too hasty in forcing truth itself into that box. (Price, 2006, pp. 608–610)

5 Contra Wright: the non-minimal cases

5.1 Do Wright and I agree about the objective cases?

I expect much of what I have just said to seem highly congenial, from Wright’s point of view. I am agreeing with him, against Horwich, Rorty and Blackburn, that there is something essentially normative to truth—something missed both by the pragmatist’s identification of truth with warranted assertibility, and by what Rorty calls ‘Tarski’s breezy disquotationalism’ (Rorty, 1995, p. 21).

Still, I want to press on two sources of disagreement. One will involve the no fault or ‘subjective’ cases, but let’s leave those aside for a moment. The second is a point about the appropriate order of explanation for the normativity on which we agree. In the recent piece we have been examining, Wright appears to say that the fault we attribute in the objective or non-minimal cases, to a speaker with whom we disagree, is to be explained in terms of a (perceived) failure of a norm of ‘accurate substantial representation.’ He explains the space he takes his view to leave for the no fault cases in terms of a contrast with some such substantial norm:

A central contention of Truth and Objectivity was that—at least over merely minimally truth-apt discourses—truth need carry no payload of accurate substantial representation. When merely minimally truth-apt claims are at stake, to regard a statement as false need not be to attribute any representational fault to someone’s acceptance of it. ... There would be an imputation of fault, and hence a compromise of Parity, only when “true” demands some kind of richer interpretation. (2021, p. 439)

Focussing on what Wright regards as the non-minimal cases, let us ask the following questions: Can we explain the normative ‘friction’ of disagreement in terms of some such ‘payload of accurate substantial representation’? Can the friction be a matter of the two parties each thinking that the other’s claim is inaccurate, in some such sense?Footnote 30

In my view, no. For suppose the two parties in question are Mo’ans. Each thinks that they have achieved an accurate substantial representation and that the other has not. Why should that feel any more friction, any more need to engage, about that matter than about their original disagreement? A claim about accurate substantial representation becomes simply one more expression of a purely subjective opinion. That’s the point about the Mo’ans. All their ‘claims’ are like that, autological rather than heterological.

I made this point in (Price, 2003):

If the Mo’ans don’t already care about disagreements, why should they care about disagreements about normative matters? Suppose that we two are Mo’ans, that you assert that p, and that I assert that not-p. If this initial disagreement doesn’t bother me, why should it bother me when – trying to implement the third norm – you go on to assert that I am ‘at fault’, or ‘incorrect’? Again, I simply disagree; and if the former disagreement doesn’t bite then nor will the latter. And if what was needed to motivate me to resolve our disagreement was my acceptance that I am ‘at fault’, then motivation would always come too late. If I accept this at all, it is only after the fact – after the disagreement has been resolved in your favour.

To get the sequence right, then, I must be motivated by your disapproval itself. This is an important point. It shows that if there could be an assertoric practice which lacked the third norm, we couldn’t add that norm simply by adding a normative predicate. In so far—so very far, in my view—as terms such as true and false carry this normative force in natural languages, they must be giving voice to something more basic: a fundamental practice of expressions of attitudes of approval and disapproval, in response to perceptions of agreement and disagreement between expressed commitments. I’ll return to this point, for it is the basis of an important objection to certain other accounts of truth. (2003, pp. 173–174)

Later in the piece, with Rorty’s proposed identification of truth with warranted assertibility in mind, I say this:

I noted above that the same point applies to the normative predicates themselves. If we weren’t already disposed to take disagreement to matter, we couldn’t do so simply by adding normative predicates, for disagreement about the application of those predicates would be as frictionless as disagreement about anything else. My claim is thus that the notions of truth and falsity give voice to more primitive implicit norms, which themselves underpin the very possibility of ‘giving voice’ at all [in the heterological sense]. In effect, the above argument rests on the observation that this genealogy cannot be reversed: if we start with a predicate – warrantedly assertible or any other – then we have started too late. (2003, p. 179, n. 22)

In other words, if we don’t make it constitutive that assertion is a move in a cooperative project—a game in which players care about normative assessments by other players—adding a norm of substantial representation doesn’t get us there. It just provides another possible topic for frictionless disagreement. But if we start with the third norm in such a practice, and explain truth and falsity as ways of giving voice to that norm, we don’t have to put it in at any later stage.

5.2 Truth, Euthyphro, and the Practical Relevance Constraint

To sum up, I have called attention to the role in normal conversation of a practice of approving and disapproving of speakers with whom one agrees or disagrees. Without these attitudes, linguistic disagreement simply wouldn’t matter to us in the way that it does. In that case, no addition of normative labels could make it matter, because disagreement about these normative matters would be as frictionless as disagreement about anything else.

Once we have this fundamental practice of treating disagreement in this way, labels that we apply to give voice to these dialogical attitudes of approval and disapproval are effectively ‘true’ and ‘false’, or the special senses of ‘correct’ and ‘incorrect’, applicable to assertions, to which Dummett’s questions direct us. And we have an immediate explanation for the fact (in the sense that it is a fact) that truth is what speakers are aiming for, in making their assertions. What they are actually aiming for, unconsciously as it were, is gaining the approval that comes with having their assertions agreed to by others, and avoiding the disapproval of the opposite case. But this looks from the inside like aiming for truth, once ‘true’ and ‘false’ are understood as expressions of the relevant sort of approval and disapproval.

One way to understand the disagreement on this broad point is in terms of a Euthyphro question.Footnote 31 When I agree with a claim made by a fellow speaker, do I endorse what she said because I take it to be true? Or should we rather understand what it is to take a claim to be true in terms of a practice of endorsing the claims of others? I have interpreted Wright, and elsewhere MacFarlane (Price, 2022), as taking the former option. Whereas my money, as I have said, is on the latter.Footnote 32 The former option cannot explain the practical relevance of taking the claims of another to be true or false—its role in our highly non-Mo’an game of giving and asking for reasons, and our sense that there is a single common bull’s-eye that all participants in the game strive to hit.

This is an example of what in Sect. 2.3 I called a Practical Relevance Constraint. In many instances, an advantage of expressivist accounts of philosophically interesting notions is that by beginning with the practical role of the notion in question, they avoid the problem of having to explain it later. A famous case is that of probability, where writers as diverse as Ramsey, William Kneale, D H Mellor and David Lewis have argued that the connection to betting and partial beliefs needs to be built into one’s theory from the beginning (see Price 2023, ch. 12, for discussion of this case). Another well-known case involves what Smith (1994) calls the moral problem—the challenge of explaining how evaluative beliefs can be relevant to action, in the way in which we take them to be. Smith himself calls this feature of evaluative beliefs the practicality requirement.Footnote 33 Indeed, we could think of the point I have just made about truth as a close cousin of the moral case.

This is what makes my view an expressivist account of truth. It begins with the practical relevance of truth and falsity—in other words, in Dummett’s terms, that of having interpersonal norms of correctness and incorrectness in this kind of linguistic activity. As I point out in (Price, 2003) and elsewhere, this also makes it a pragmatist theory of truth, in one very natural sense—though it is very different from what is often regarded as the pragmatist approach to truth (the identification of truth with success, or some such).Footnote 34

5.3 Why does Wright miss the expressivist option for truth?

I’m not sure why Wright misses this expressivist alternative for truth itself. I find it especially puzzling in papers such as Wright (1998) and Wright (2001), in which Wright presents his Inflationary Argument, and then carefully considers the deflationist’s options for responding to it. For these purposes, deflationism amounts to the denial that truth need be regarded as a ‘substantial’ property—in other words, as something that contrasts with the ‘thin’ property of truth that, as Wright notes, deflationists such as Horwich are happy to allow. As I have said, I take my own view of truth to be deflationist in this sense, even though it disagrees with Horwich about the essential function of truth. It takes ‘true’ to be first and foremost an expression of the crucial third norm of heterological conversation, rather than a mere device of disquotation.

Let’s pick up Wright’s challenge to deflationism in Wright (2001), after he has reviewed his argument that truth is a norm distinct from warranted assertibility. Wright then says this:

Minimalism [i.e., Wright’s own view] now claims that these facts about assertoric practices stand in need of explanation. In particular, it maintains that it needs to be explained what this further norm of correctness amounts to in such a way that it becomes clear how it and warranted assertibility, although potentially divergent in extension, coincide in normative force: how it can be that warrant is essentially warrant to think that this other norm is satisfied when there is no guarantee that they are always co-satisfied. And such an explanation, it is contended, while it will have to do much more than this, must at least begin by finding something for the truth of a proposition to consist in, a property that it can intelligibly have although there may currently be no reason to suppose that it has it, or may intelligibly lack even though there is reason to think that it has it. Warrant can then be required to be whatever gives a (defeasible) reason to think that a proposition has that property. (Wright, 2001, p. 757)

Wright thus identifies an explanatory challenge, one that he thinks any deflationist view will be unable to meet.

For my part, I have identified a different explanatory task, that of accounting for the friction-generating role of the third norm. I have argued that Wright’s apparent order of explanationFootnote 35—i.e., an approach that begins by identifying a property (‘something for the truth of a proposition to consist in’), and then seeks to show that the property in question does the normative work required—cannot possibly do the full job. In the absence of the third norm, disagreements about the property in question would be as frictionless as anything else.

But what about the explanatory task that Wright has in mind? Here I think he misses the possibility of an expressivist approach to the challenge. Suppose we ask what usage rules would be required, to generate an assertoric practice of ascribing norms with the character of truth and falsity—in particular, with the relation to warranted assertibility that Wright here describes. The answer looks clear, at least for the initial and main steps. Speakers simply need to ascribe truth and falsity to claims by others with which they agree or disagree, respectively. Crucially, these ascriptions need to carry a positive or negative normative load. In other words, they need to express (a special sort of) approval or disapproval. This immediately makes the practice heterological. It renders each speaker’s normative status hostage to the judgement of her peers, so that it is not guaranteed by warranted assertibility from her own point of view. Yet at the same time it is obvious why for any individual speaker, warrant is, as Wright says, a defeasible reason to think that a claim is true. That falls straight out of the usage rule.

One thing that makes it puzzling why this option is not on Wright’s list is that it seems by my lights so close to options he takes elsewhere. We saw that in Wright (2021) he is perfectly happy with this pragmatic, use-theoretic approach to other topics—indeed, he takes it to be obligatory, wherever truth-aptness is minimal in the sense of that paper. And elsewhere, famously, he defends approaches that look very much like expressivism, in my sense. Let me quote again from the paragraph from Truth and Objectivity that has my marginal remark.

Elsewhere I have argued that Frege’s platonism about number is best interpreted as based on the view that an expression’s candidacy to refer to an object is a matter of its syntax: that once it has been settled that a class of expressions function as singular terms by syntactic criteria, there can be no further question about whether they succeed in objectual reference which can be raised by someone who is prepared to allow that appropriate contexts in which they do so feature are true. There is, that is to say, no deep notion of singular reference such that an expression which has all the surface syntactic features of a Fregean proper name, and features in, say, true contexts of (by surface syntactic criteria) predication and identity, may nevertheless fail to be in the market for genuine—“deep”—reference. (Wright, 1992, pp. 28–29)

The message here is that once a certain pattern of use is in place, there are no further deep questions of metaphysics, or deep questions of the semantic kind that do indirect duty for metaphysics – no further deep question about whether there are numbers, for example, or whether number terms ‘really’ refer to anything. I agree wholeheartedly with this message. And as I’ve explained, and as my comment in the margin says, the resulting view counts as expressivism, by my lights.

Hence my puzzle. Why doesn’t Wright explore the possibility of saying the same kind of thing about truth? Perhaps he thinks that the result would be the option he describes elsewhere as ‘Frege’s indefinabilism’. The following passage is from Wright (1998), and corresponds to a similar point in Wright’s dialectic to the paragraph quoted above from Wright (2001). Wright has laid out the Inflationary Argument, and is considering possible deflationist responses.

Can the deflationist regroup? What the foregoing forces is an admission that, for each particular proposition, we have the concept of a norm which is distinct from warrant and is flagged by the word “true.” And once it’s allowed that the role of “true” is to mark a particular kind of achievement, or failing, on the part of a proposition, contrasting with its being warranted or not, there will have to be decent sense in the question, what does such an achievement, or failing, amount to? To be sure, that is a question which may turn out to admit of no very illuminating or non-trivial answer – but if so, that would tend to be a point in favour of Frege’s indefinabilism, rather than deflationism. If a term registers a distinctive norm over a practice, the presumption ought to be that there will be something in which a move’s compliance or non-compliance with that norm will consist. And whichever status it has, that will then be a real characteristic of the move. So what room does deflationism have for manoeuvre? (1998, p. 45)

But imagine a version of Frege’s indefinabilism that parallels Wright’s own neo-Fregean view about number. Such a view of truth could help itself to Wright’s own insistence that neo-Fregean platonism rejects ‘deep’ metaphysical and semantic notions, thereby blocking his attempts to embarrass the deflationist with questions that seem in need of substantial answers. Such questions would be the equivalent of questions like this, addressed to the neo-Fregean platonist: ‘What are these things—“numbers”—to which you take number terms to refer?’

With these metaphysical challenges blocked, how could Wright respond to the proposal that Fregean indefinabilism, construed by analogy with his own neo-Fregean platonism, might be an option for deflationism? I have two suggestions, though neither gets very far, by my lights. First, against a deflationist such as Horwich, Wright can point out that deflationism doesn’t get normativity for free, so it is still missing something. I agree, but this cuts no ice against my version of deflationism, which builds normativity into the use rules from the very beginning (arguing that only by starting at that point can we meet the Practical Relevance Constraint). Second, Wright could claim that this neo-Fregean indefinabilism doesn’t do the explanatory work described in Wright (2001). But I have already responded to that challenge—that work is done by the use rules.

Elsewhere in Wright (1998), Wright puts his challenge to the deflationist like this:

[W]hat the deflationist clearly cannot allow is that “true,” when used to endorse, has the function of commending a proposition for its satisfaction of some distinctive norm which contrasts with epistemic justification and which only “true” and equivalents serve to mark. For if there were a distinctive such norm, it could hardly fail to be reckoned a genuine property of a proposition that it did, or did not, comply with it. And if the norm in question were uniquely associated with “true” and its cognates, that would be as much as to allow that there was a special property of truth – at which point the deflationary game would have been given away. (1998, p. 42)

My answer is that there is indeed a genuine property, in the same sense in which there are genuine properties in any of the other domains in which it is expressivism that offers the best path to philosophical illumination. But as in these other cases, it is not a ‘substantial’ property, if by that we mean (something like) a property whose nature is an interesting matter for theoretical investigation. In that sense, the deflationary game has not been given away. On the contrary, with the normative use of the predicate now firmly in its sights, the deflationary game has found its feet.

There is one other aspect of Wright’s work that seems relevant to this puzzle, namely his attraction to a platitudes-based approach to ontological questions that he attributes to Michael Smith and Frank Jackson (Wright, 2013, p. 128). I suspect that Wright’s sympathies for this approach contributes to his sense that there is more work for a non-trivial metaphysics of truth than my expressivism allows. I will come back below to my reservations about this programme. Before that, let’s turn to the subjective or no fault cases, where I claim to offer an alternative both to Wright and to relativism.

6 Contra Wright: the minimal cases

6.1 Does Wright’s account work for the subjective cases?

I can introduce my main concern about Wright’s (2021) proposal by way of a comparison that Lionel Shapiro (2014) draws between the goals of relativism and those of some of my work. Here is Shapiro’s description of the key linguistic phenomena.

The goal of advocates of relativist semantics has been to explain the apparently distinctive functioning of a wide-ranging class of expressions. As my examples, I will take epistemic modals such as ‘it’s possible that ...’ and so-called predicates of personal taste such as ‘tasty’ and ‘fun’. At issue is the way their use appears to have a subjective or perspectival aspect. ...

The key phenomenon is pointed to in Huw Price’s ... Facts and the Function of Truth (1988), which deserves to be recognized as a close precursor to the current discussion of relativism. This is the seeming possibility of what Price calls “evaporative disputes,” which he says “populate the margins of factual discourse” (1988: 159ff). Price himself focuses on disputes over whether something is probable, but his observations apply equally to disputes involving epistemic possibility and predicates of personal taste. Here are two examples:

  • Two physicians, Dr. Adams and Dr. Brown, are discussing a mutual patient. Adams says: ‘It’s possible the patient has Lyme disease’. Brown protests: ‘no, that’s not possible! You must be misinterpreting the test results.

  • Two friends, Alice and Ben, are discussing which cheeses to buy for a party. Alice says: ‘Limburger is a tasty cheese. My sister told me so.’ Ben replies: ‘Did she really? I’ve tried it, and it’s not tasty at all!’

To all appearances, both examples satisfy the following schematic description:

  1. i.

    one party has affirmed, and the other has denied, that P,

  2. ii.

    both parties recognize (i), and

  3. iii.

    each takes (i) to mean that there is something mistaken about the other’s speech act.

Suppose the appearances are correct and (i)–(iii) are satisfied. In a familiar sense, this suffices for the parties to be engaged in a dispute about whether P. Thus Adams and Brown are engaged in a dispute about whether it’s possible the patient has Lyme disease, and Alice and Ben are engaged in a dispute about whether Limburger cheese is tasty.

Imagine, now, that Adams and Brown discover that only Brown has seen the most recent test results ruling out their patient’s having Lyme disease. According to Price, (ii) will continue to obtain. In particular, Brown will continue to regard Adams as having affirmed that it is possible that the patient has Lyme disease. Yet (iii) now seems to fail: each will cease to regard the other’s assertion about the patient as mistaken in any way. In particular, Brown will not merely excuse Adams’s mistake as blameless given Adams’s uninformed perspective (Price, 1988, pp. 162–163). Rather, in view of that perspective, Brown will cease taking Adams to have made any mistake, even though Brown takes Adams to have described something as possible that is not in fact possible. What was formerly a dispute about whether it’s possible that the patient has Lyme disease has evaporated—or so it has seemed to Price and his relativist heirs. (Shapiro, 2014, pp. 141–142)

Shapiro develops his own proposal for dealing with these cases, within the framework of Brandom’s account of the norms of assertion; more on that later. For now I want to focus on what we might call the dynamic character of these cases, nicely captured in Shapiro’s example. By this, for the moment,Footnote 36 I mean our sense that we do have a substantial disagreement at the beginning of the exchange. This must be so if we want to allow—as seems obviously true in such cases—that ‘evaporation’ may never occur. Many disputes about such matters take a normal course, with no relaxation of fault.

In my view, Wright’s account fails to explain this dynamic aspect of faultless disagreements. It withdraws fault too easily, and too early. I have made a very similar objection to MacFarlane’s account, so let me begin there.

6.2 MacFarlane and Wright on the phenomenology of disagreement

Here is a passage from MacFarlane (2007), giving his sense of what disagreement looks like in what he thinks of as relativistic cases—disputes about matters of taste, for example.

This [relativist] account captures the distinctive phenomenology of disagreement about matters whose truth is relative. The challenger thinks (rightly) that he has absolutely compelling grounds for thinking that the assertion [e.g., ‘Liquorice is tasty’] was not accurate. But the original asserter thinks (also rightly, from her point of view) that the challenger’s grounds do nothing to call in question the accuracy of the assertion. The asserter’s vindication will seem to the challenger not to show that the assertion was accurate, and the challenger will continue to press his claim. (Until the game gets boring.) Thus we have all the normative trappings of real disagreement, but without the possibility of resolution except by a relevant change in one or both parties’ contexts of assessment. (2007, p. 29)

Is MacFarlane right that his version of relativism ‘captures the distinctive phenomenology of disagreement’ about matters of taste, and that it offers us ‘all the normative trappings of real disagreement’? In my view, no, on both points.Footnote 37 On the contrary, I think it makes a mystery of the phenomenology of disagreement, because it does such a poor job of making sense of the normative trappings. Why, given relativism, should I, or a fellow speaker, care that I think I have compelling grounds for thinking that their assertion is inaccurate? As MacFarlane himself puts it earlier in the same piece:

How can there be disagreement between Abe and Ben, on the relativist’s view, if the proposition Abe asserts and Ben denies is true relative to Abe’s standard of taste and false relative to Ben’s? Aren’t they just talking past each other, in some sense? (2007, p. 21)

It is true, as MacFarlane notes, that the need or desire for coordination may provide a reason to care about the disagreement. But such a reason is not there by default. We see this easily in expression of intention cases, such as my restaurant example (Sect. 4.2). Imagine the following scene. The waiter approaches a table, and asks, ‘Would anyone care for coffee?’ We can imagine an alternating sequence of responses from two diners: ‘Yes, please’; ‘No, thank you’; ‘As I said, yes!’, ‘And as I said, no!’; ... This becomes boring, too, and for the same reason. There is nothing more than a rather shallow trick of the language to make the two diners think there is any issue between them.

The trick of the language might be more convincing in the tasty case, because the labels ‘true’ and ‘false’ (or ‘accurate’ and ‘inaccurate’, to use MacFarlane’s own terms) are permitted, and the two parties apply them differently. This would indeed explain the phenomenology of disagreement, under the supposition that the disputants don’t realise that the case is subject to context-relative standards of accuracy. But this is no help to MacFarlane, who wants to get the phenomenology right for folk like him, who do know the story. (He wants the relativist account of truth and accuracy to explain the phenomenology.) Whereas it seems to me that we only ’get the normative trappings’ if we can ignore what we know about the perspectival character of the accuracy conditions—if we pretend (or mistakenly believe) that there is an objective matter at issue, as it were.

In my view these objections apply equally to Wright’s account. No fault means no friction, and hence no pressure for coordination. Again, ill-informed speakers might be confused about this, but it seems fair to apply the same test to Wright that we applied to MacFarlane. His account should work for folk like him, who are assumed to know the full story.

This challenge to Wright becomes particularly pressing once we realise that no fault disputes can be dynamic, in the sense I described earlier. We need to explain why properly informed speakers do take such disputes to involve fault, before the point—never reached, in many cases—at which the disputes ‘evaporate’. MacFarlane perhaps has some resources to account for this, in terms of shifting contexts of assessment. But for Wright, without these resources, it is hard to see how the dynamic character of no fault disagreements could be anything other than a mystery.

6.3 FFT’s view of evaporative disputes

In different ways, Wright and MacFarlane both think of the subjective cases as fundamentally distinct from the objective cases—cases in which they take there to be no option for no fault disagreement. My view, in contrast, maintains that the subjective cases involve exactly the same normatively-loaded game of challenge as disputes about, say, the age of the Earth (one of MacFarlane’s examples). I take ordinary heterological conversation in general—i.e., the game of giving and asking for reasons (GGAR), in Brandom’s phrase—to be the kind of coordination device that Wright and MacFarlane both propose for the subjective cases.

Crucially for our current concerns, this game of challenge is well adapted to variability with respect to the ‘need’ and value of coordination. It is part of the normal functioning of the game that there are exit moves—ways of cancelling the norms, in particular disagreements.Footnote 38 The so-called subjective cases are those in which there is often and obviously a rationale for using these in-built escape hatches. The phenomenology of subjectivity arises from the fact that use of the escape hatches cancels the normal objectivity of the bull’s-eye—the normal presumption that there is a common goal—but we only see this properly for what it is when we understand the source of the latter. Proper play of the game, required for it to serve its coordinative function, requires that the escape hatches initially be ignored. Fault is positive-presumptive, as I put it earlier.

In my view, as I have said, the escape hatches are never absent altogether, but I don’t need to insist on that point here. Even if I allowed cases in which there was no possibility of evaporative disputes—no possible role for the escape hatches—the story I would tell about the source of the phenomenology of disagreement would be no different from the one I tell about the so-called subjective cases. In all cases, it takes the source of the phenomenology to be the third norm, at the heart of the GGAR. This works for me, in a way in which it can’t for Wright or MacFarlane, because of the pragmatist direction of explanation, in Brandom’s sense—in effect, because truth and factuality are regarded as downstream rather than upstream of the GGAR.Footnote 39

By analogy, imagine a play-fighting game, say paintball, in which players compete to shoot each other with small paint-filled capsules. Imagine there is a designated escape phrase, say ‘Time out!’, that immediately pauses the game when uttered or displayed. For better or worse, participants who wear ‘Time out!’ on their hats never get to experience the competitive phenomenology of the game. In different ways, MacFarlane and Wright accomplish something similar. MacFarlane puts the players in different rooms, and Wright takes away their weapons. My view puts them in the same room, with their weapons, and relies on their knowledge of the escape options provided in normal play of the game to cancel the conflict, when that seems appropriate.

Certainly, there is a good question why expressions of taste, and other obviously subjective cases, should ever invoke the machinery of such a game in the first place. But as we saw, MacFarlane and Wright propose a similar answer: it has engineering advantages, in communities of social creatures like us. In general, it is an advantage of an expressivist conception of the functional role of commitments of various kinds that it stands back far enough back to consider the question of the merits of ‘making those commitments explicit’ in the framework of the GGAR. As I argued in FFT, we should expect that different communities will draw the line in different places, and should not be surprised to find both practices in use in a single community—as we do, of course, with pairs such as ‘Yum!’ and ‘That’s tasty!’

6.4 Shapiro’s pragmatic relativism

There are nice questions about how we should model a linguistic game of this kind, for various formal purposes. Here I think Shapiro’s (2014) suggestion is attractive. As I noted earlier, Shapiro points out that the linguistic phenomena on which MacFarlane and other relativists have focussed have a substantial overlap with the NFDs of FFT. However, Shapiro feels that these relativists are addressing the issues posed by such cases in the wrong key.

The last decade has seen an explosion of interest in relativism in the philosophy of language. Relativist accounts have been proposed to explain discourse about knowledge, epistemic possibility, matters of taste, and contingent future events. In this context, relativism is usually taken to be, or to presuppose, a semantic thesis. According to relativists, understanding how some discourses function requires recognizing that speakers express propositions whose truth or falsity must be evaluated relative to parameters in addition to a possible world ...

In this paper, I propose a different way to think about how the discourses in question motivate relativism in the philosophy of language. The central thrust of relativism, I argue, can and should be understood independently of any semantic framework of relativized truth. Instead, relativism should be understood in pragmatic terms, as corresponding to a particular understanding of assertoric force. The idea starts with Robert Brandom’s analysis of “fact-stating discourse” (1994, p. 607) as a “game of giving and asking for reasons” whose basic move is asserting. Brandom’s account of assertion makes no appeal to truth. My proposed revision of his analysis makes room for a broader conception of fact-stating discourse, by allowing assertoric force to depend on speakers’ perspectives. What is distinctive and plausible about relativism, I will argue, is best captured by the resulting liberalized version of Brandom’s game of giving and asking for reasons. (Shapiro, 2014, pp. 139–140)

Specifically, Shapiro proposes an amendment to Brandom’s account of the ‘authority’ claimed by someone who makes an assertion. According to Brandom, as Shapiro puts it: ‘the authority claimed in asserting a proposition is to license others to assert the same proposition and use it as a premise in their theoretical and practical reasoning.’ (2014, p. 153) Shapiro suggests that to capture ‘what is distinctive and plausible about relativism’, we should restrict this claimed authority:

Stated generally, the idea is that assertoric force has a perspectival aspect. On this view, an assertion may carry its licensing potential not vis-à-vis its entire potential audience, but rather vis–à-vis a target audience restricted to those who share a particular perspective. (2014, p. 154)

As I say, this seems to me a very promising approach. It seems particularly well-suited to cope with the dynamic character of the relevant linguistic phenomena. Speakers simply update their sense of who belongs in the relevant target audience.

I have not yet mentioned another aspect of the dynamics of these phenomena that I discuss elsewhere (Price, 2023, ch. 11). We seem to need a distinction between divergent and convergent cases. In Shapiro’s terms, this amounts to the distinction between cases in which speakers split or amalgamate their target audiences, when a dispute evaporates. There are cases of both kinds, in my view, and both may involve a withdrawal of fault. For example, judgements of taste are often divergent; epistemic cases, such as probability judgements, are typically convergent. As I say, Shapiro’s framework seems to handle these distinctions very easily.

7 Alethic pluralisms

7.1 Avoiding the road to Canberra

Let’s now come back to Wright’s attraction to a platitudes-based approach to metaphysics, an approach he associates with the work of Smith (1994) and Jackson (1998). Wright describes the idea like this:

Let us call an account based on the accumulation and theoretical organization of a set of ... platitudes concerning a particular concept an analytical theory of the concept in question. Then the provision of an analytical theory of truth in particular opens up possibilities for a principled pluralism in the following specific way: in different regions of thought and discourse, the theory may hold good a priori of – may be satisfied by – different properties. If this is so, then always provided the network of platitudes integrated into the theory is sufficiently comprehensive, we should not scruple to say that truth may consist in different things in different such areas: in the possession of one property in one area, and in that of a different property in another. For there will be nothing in the idea of truth that is not accommodated by the analytical theory, and thus no more to a concept’s presenting a truth property than its validating the ingredient platitudes. In brief, the unity in the concept of truth will be supplied by the analytical theory, and the pluralism will be underwritten by the fact that the principles composing that theory admit of variable collective realization. (2001, pp. 760–761)

For my part, I associate this methodology with what Hawthorne and I (1996, p. 130) called the Canberra Plan (CP). CP begins with location or placement problems: Where do normativity, meaning, mentality, and other puzzling domains ‘fit’ in the kind of world described by science? It proposes to answer questions of this kind with a generalisation of the Ramsey-Carnap-Lewis approach to the meaning of theoretical terms.Footnote 40

The proposed solution comes in two steps. At Step 1 we collect the core truths or platitudes about the target entity or property—the entity or property Target, let us say—and conjoin them to form the Ramsey sentence, R(Target). At Step 2 we ask what in the world satisfies or makes true the sentence R(Target)—or to what the term ‘Target’ refers. As Haukioja (2009) puts it, Step 1 is a matter of ‘a priori analysis of our philosophically interesting everyday concepts and folk theories’; Step 2 of ‘consult[ing] the best scientific (typically, physical) theories to see whether ... referents [for the terms so analysed] are to be found in reality.’ Typically, as here, this is understood to mean natural reality, the world described by natural science, but this isn’t essential to the method. A non-naturalist could also frame her investigations in these terms.

So far as I can see, an expressivist need have no distinctive objection to Step 1 (except the obvious objection, from her own point of view, that this platitude-marshalling does little or nothing to address what she sees as the interesting explanatory questions). Other objections may be raised to Step 1—for example, that it pays insufficient attention either to the analytic—synthetic distinction, or to the grey zone that results from taking seriously Quinean objections to such a distinction. But if anything such objections are likely to trouble expressivism less than they do CP, I think, because they threaten Step 2, which is where expressivism and CP really differ.

Concerning Step 2, my view is that an expressivist should simply deny that it leads to non-trivial results, in general. Clearly, Target satisfies R(Target), if anything does. Why should we expect anything else, in general? I have argued (Menzies & Price, 2009; Price, 2004a, 2009) that in practice, this expectation rests heavily on an implicit appeal to non-deflationary readings of semantic terms such as satisfies, makes true, or refers—readings that expressivism rejects, of course. As I have noted elsewhere, Blackburn makes a similar point:

Blackburn [1998b, 78] notes that on Ramsey’s view, the move from ‘P’ to ‘It is true that P’ – “Ramsey’s ladder”, as he calls it – doesn’t take us to a new theoretical level. He remarks that there are “philosophies that take advantage of the horizontal nature of Ramsey’s ladder to climb it, and then announce a better view from the top.” (Price, 2011, p. 15)

I take it that CP is one of the philosophies that Blackburn has in mind. His point is that talk of truthmakers, denotations, and the like adds nothing to the repertoire of metaphysics, unless the semantic notions in question are more robust than those of Ramsey, Quine, and later deflationists. If a proponent of CP tries to embrace this conclusion, saying that their own use of semantic notions is similarly ‘thin’— that in effect, Step 2 simply asks ‘What is the X such that R(X)?’—then the expressivist says again that we already have a trivial answer to that question, but no reason in general to expect a non-trivial one.Footnote 41

The qualification about generality is important. Expressivism may have no need to challenge particular cases, including those of theoretical identification in science. But it will argue that in these cases the semantic characterisation is inessential—the questions can be phrased without it. Some proponents of CP might agree, and argue that the science model is all we need—CP should simply be seen as generalised functionalism. On this view, R(Target) encodes the causal and functional role of Target, and Step 2 simply enjoins us to look for whatever it is that plays this causal role—a question for natural science, in principle. However, as Peter Menzies and I pointed out (Menzies & Price, 2009), this version of CP doesn’t have the generality to which CP aspires—it cannot handle the metaphysics of the causal relation itself, for example.

Deflationism is not the only threat to the attempt to ground a general version of CP on semantic notions. In its general version, the programme will want to encompass the metaphysics of the semantic notions themselves.Footnote 42 But then it becomes worryingly circular, for reasons related to Boghossian’s challenges to irrealism about content and to Putnam’s model-theoretic argument (see Price 1998b, 2004a, 2009). It is an interesting question whether any such charge could be pressed against Wright’s use of the methodology.

But are these objections—the one based on deflationism, or the one based on circularity—relevant in the present context? After all, won’t an expressivist about moral value, or beauty, allow that it makes perfect sense to ask what kinds of things we take to be morally valuable, or beautiful, in particular domains. What does beauty amount to for vintage sports cars, say? Something different, presumably, than for mountain landscapes. So if the expressivist doesn’t challenge the platitudes about moral value, or beauty, isn’t the way open to a pluralism about these notions, analogous to Wright’s pluralism about truth?

In the case of moral value, however, there’s a familiar response. One of the platitudes about moral goodness links it to motivation and hence to behaviour. It is a priori that moral worth has a (positive) motivating character: believing a possible action to be good or right provides a motivation for doing it, all other things being equal.Footnote 43 Notoriously, it is a challenge for any non-expressivist account of the content of moral belief to explain this fact. As I noted earlier, this is Smith’s (1994) practicality requirement, a key element in what he calls the moral problem. As I also noted, it is an instance of the Practical Relevance Constraint, which is a central part of the general expressivist recipe.Footnote 44

So long as this motivational principle is included among our platitudes for good, there is an obstacle to taking any merely naturalistic property P to instantiate those platitudes, even if it is clear (at least within some given domain) that competent speakers apply the term ‘good’ to things with property P. For P won’t satisfy the motivational platitude. It won’t be true a priori that thinking that an action has the property P is a motivation for doing it.

Similarly, I suggest, for true. It may be true that we can find properties that tend to correlate with acriptions of truth by competent speakers, in various domains. In that sense, there may well be something in the alethic pluralist’s thought that truth amounts to different things in different domains. (Again, think of beauty for vintage cars and mountain landscapes.) But if one of our platitudes for true and false links the ascription of these terms to the attitudes of approval and disapproval that generate friction in ordinary heterological conversation, it will be an additional step to take such properties to ‘[validate] the ingredient platitudes’, as Wright puts it. On the face of it, as I argued, no such property can do the work of explaining friction.Footnote 45

7.2 Prospects for alethic pluralism

Where does this leave us? I think it leaves my sort of expressivist with two sorts of reason for resisting any reading of a platitudes-based analytical theory (as Wright calls it) that takes it to be in the business of substantial metaphysics, the way it is intended by proponents of the Canberra Plan. One sort of reason turns on objections to the programme in general, and in particular to two (linked) things: its claim to offer some sort of guarantee of non-trivial answers, and its own reliance on substantial semantic notions. The other sort of reason turns on the Practical Relevance Constraint, pointing out that that in itself provides an obstacle to non-expressivist approaches to truth, once some manifestation of the normativity of truth is included among the platitudes.

On the other hand, more irenically, there’s a different kind of expressivist-friendly alethic pluralism in the near vicinity. Indeed, there may even be a plurality of such possibilities. The possibility already on the table is the analogue of our point about aesthetic judgements. Clearly, even an expressivist should have no trouble making sense of the idea that aesthetic judgements respond to different features in different domains. Plausibly there are commonalities, too, but the differences may well be interesting.

A different possibility is that aesthetic discourse itself might in some ways work differently, in different domains of application. I don’t have a clear example to offer, but, moving back to the alethic case, I think we can see a space here for a kind of internal-to-expressivism alethic pluralism. Indeed, my own view in FFT provides an example. Once I had abandoned the Bifurcation Thesis, as I have explained, I saw my project in FFT as making sense of intuitions about the relative ‘factuality’, or ‘objectivity’, of different subject matters, in terms of variable susceptibility to NFDs. As I put it in (Price, 2023, ch. 11):

We could find No Fault Disagreements, popping up for different reasons in different language games. And we could correlate those differences both with intuitions about comparative ‘factuality’, or lack of it, and with variations in the underlying functional stories.

In effect, this is a story told at the level of use about how truth ‘amounts to subtly different things’ in different domains (where I put the characterisation in quotes to indicate that it should not be read too literally, as a metaphysical story about what truth is).

Neither of these possibilities is Wright’s kind of alethic pluralism, as I have interpreted it, but they both share some common ground with it. It would be interesting to explore this common ground further, in the particular applications of Wright’s own alethic pluralism that he and others have discussed. Indeed, Tom Kaspers has suggested to me that Wright’s own view might be interpreted in something like the latter way, though with an interesting normative twist: Wright might be read as asking the question what kind of truth property we should construct in a given domain.Footnote 46

8 Conclusion

To close, let me step back a little, and compare Wright’s motivations for alethic pluralism with mine for expressivism, both about truth and about other matters. Not surprisingly, perhaps, there is a similar mix of similarities and differences to the one I have described above, in looking at the ways in which our motivations play out, in our respective views.

In a recent piece Wright reports that there were two main motivations for his attraction to what has become known as alethic pluralism.

[The first] was because it looked as though making some sense of different kinds of truth might help to explain why the traditional debate about truth turned out to be sterile and incomplete. Maybe the reason why the correspondence theorists, the coherence theorists, and the pragmatists couldn’t get anywhere was because they were all over-generalizing. ... What really seems wrong with correspondence is that it seems a tendentious way to think about mathematics, for instance, and a bad way to think about the comic: one doesn’t want to be saddled with some metaphysics of ‘out there’ comedic facts to which one’s impressions about comedy may correspond just by being willing to apply ‘true’ to ordinary ascriptions of ‘funny’. (Wright, 2013, p. 123)

For my part, I am attracted to a different reason why the traditional debate about truth got stuck: it was asking the wrong kind of question. I got to where I am by pushing at Dummett’s (1959) question about the point of truth, and adopting the engineering stance.

Before describing his second motivation, Wright adds a wrinkle to the first one:

This idea connected with my desire to resist deflationary accounts of truth, which of course originally drew a large part of their credibility, for those who found them credible, from the failure of the traditional debate. In general, I distrust philosophical accounts of anything that say, ‘There’s not much here, it’s not as interesting as you think it is.’ I don’t want to be told that something isn’t interesting. I want to be told, ‘It is more interesting than you think’—because one has missed certain ramifications and nuances, for example. So, perhaps just as a matter of temperament, I wanted to find a way of avoiding the collapse into deflation, and I saw that collapse as primarily motivated by the sterility of the original debate and a different diagnosis of it: that the truth debate was bad because the antagonists weren’t talking about anything, because ‘the nature of truth’ is not an authentic subject. That’s not the right account of the matter, in my view. (2013, p. 124)

Here Wright and I differ, of course. I am attracted to the view that the nature of truth is not an authentic subject, though I don’t take it to imply ‘there’s not much here.’ What it implies instead, in my version, is something like: ‘You’ve been looking for the wrong kinds of things—there is an interesting story, but it has to be told somewhere else.’ I was predisposed to be attracted to this idea in the case of truth—and hence sympathetic to deflationism, even though I thought that in some versions it gave the wrong story about the function of truth, by missing the normative role—because I had come to these questions about truth from the standpoint of an expressivist approach to other matters.Footnote 47

What about Wright’s second motivation?

The other [motivation] was my long-standing interest in the debates about realism and objectivity, Dummett and Wittgenstein, and all that. Dummett had given us a model of those debates, or some of those debates, where what’s at stake are differing conceptions of the form that statement-meaning takes in the region of discourse in question. And I thought that he was right, up to a point, because if you are a correspondence theorist about truth, you are thinking of meaning as consisting in, so to speak, correspondence conditions. And if you are not a correspondence theorist, you may still say, ‘I am thinking of meaning as truth-conditional’, but you are not thinking of truth-conditions in the same way. So it does look as though there would be implicitly differing conceptions of meaning in play if you conceived of the different disputes in that way. (2013, p. 124)

Wright goes on to describe Dummett’s attempt to do semantics in terms of assertion-conditions, and its problems.

Dummett got into trouble trying to sustain the meaning-theoretic model of the disputes, because he couldn’t actually construct any assertibility-conditional accounts of meaning. ... Generally speaking, the assertibility-conditions of a statement are not purely recursively semantically determined, so of course a proper semantic theory can’t fully characterize assertibility-conditions. So my thought was just that if we’ve got differing notions of truth, or differing conceptions of what truth consists in ... then you didn’t need to engage any of that. You could just allow that truth-conditions are fine across the board; that disquotational—Davidson-style—semantics is fine across the board (as far as it goes, whatever it’s supposed to illuminate exactly). What’s really varying is the way in which the various discourses engage with reality, the kind of truth that applies. So, if disquotational semantics is an adequate basic semantics, then the realist/anti-realist debate is not a semantic debate in the end. (2013, pp. 124–125)

I agree with pretty much all of this, except the sentence underlined. For me, what’s varying is the kind of material that goes into an expressivist account of meaning in terms of use. I agree that this needs to be sharply distinguished from, though is not incompatible with, what Wright calls a disquotational basic semantics; see Price (2004b).

Summing up, then, there are some similarities in our starting points. Dummett plays an important role for both of us, though in different ways. And we are both interested in an account of truth that can handle what we might term the problem of Many Subject Matters (MSM). If anything, the latter became an even stronger motivation for me than it is for Wright. For me, as ‘Metaphysical pluralism’ (Price, 1992) reflects, this motivation strengthened as I engaged with alternative proposals for dealing with this kind of plurality that were popular in my Australian environment. But one thing I seemed to have shared with Wright, in opposition to writers such as Lewis, Armstrong, Jackson, and many others, was a tendency to associate MSM with the linguistic pluralism of (later) Wittgenstein. Still, I think I take this tendency further than Wright does, as demonstrated by his continuing interest in the nature of truth. By my lights, that’s one of the bad old metaphysical questions, and it helps to realise that, if we want truth to help us with the problem of MSM.

Finally, how much of the weight of MSM—that is of the task of describing and explaining the variety of different discourses that seem salient from this loosely Wittgensteinian perspective—can be carried by an alethic pluralism, whether of Wright’s kind or of mine? This question is especially interesting, from my own perspective, because, in its own way, FFT attempted something like this. Roughly speaking, it attempted to squeeze the bulk of our sense that some subject matters are less factual, or more subjective, than others, into the framework provided by NFDs—in other words, under the umbrella provided by FFT’s account of the role of truth in disagreement.

In recent work, I have considered the question of the relation between this approach in FFT and the kind of conception of expressivism that has emerged in my later work. This is my conclusion.

I think it is fair to say that FFT offered would-be expressivists a rather meagre diet, compared to the multigrain recipe [described above]. FFT wasn’t blind to the other ingredients, but it did maintain that a focus on NFDs did essential work in keeping expressivism from the clutches of determined factualist opponents, once the bifurcation thesis had been abandoned. That conclusion was too pessimistic. The positive sides of expressivism – its explanatory programme, its approach to accounts of meaning, and the alternative it often provides to metaphysical rivals struggling with the Practical Relevance Constraint – all these survive the kinds of accommodation with factualism on which the argument of Part I [of FFT] was based. In effect, this is the lesson of Creeping Minimalism. Expressivism has no reason to be embarrassed by looking much like realism, so long as it maintains its explanatory credentials. It is true that local expressivism has a different reason to be concerned about the failure to find an adequate ground for the bifurcation thesis, but that’s an embarrassment it incurs by underreaching, not by overreaching. (Price, 2023, ch. 11)

I could put this by saying that in FFT I looked too much at the surface manifestations of the important underlying differences between discourses, and not enough at their causes or explanatory grounds.

Can a similar charge be laid against Wright’s alethic pluralism, in so far as it seeks to make sense of discourse pluralism? It may be less subject to the charge than my account in FFT, in virtue of appealing to something a bit more substantial than the varying usage-patterns with respect to NFDs. Unlike me, Wright has available the properties he regards as candidates for truth, in different domains.

All the same, explanatory questions will remain. There are two issues. First, there is no guarantee that there will be enough varieties of truth to mark all the lines MSM inclines us to draw; and second, even if there were, they aren’t sufficiently explanatory. What explains the different truth properties in different discourses? We get tugged in one of two directions: to differences in the nature of subject matter (metaphysics), or to differences in the function of language or concepts in question (expressivism). Alethic pluralism thus becomes a kind of transit station, delaying a choice that needs to get made at some point – where does the explanatory basis for its pluralism reside?

I conclude that if alethic pluralism is to have a future at all, it needs to make this choice between metaphysics and expressivism. This gives us three options altogether,Footnote 48 nicely summarised in a form we can borrow from the title of (Wright, 1993): alethic pluralism—w(h)ither now?Footnote 49