Plain alethic pluralism: the human face of truth

Abstract

Within the secular and still ongoing discussion about truth, what is by now known as alethic pluralism has proved to be one of the most interesting proposals advanced in the field. Indeed, from the last decade of the last century onwards, the field of theories about truth has been enriched with the idea that truth is not only one but also many—many alethically potent properties that differently characterise different domains of discourse. Moreover, thirty or so years of discussions have also shown that alethic pluralism itself is many rather than one—a family of views with different metaphysical commitments. As it may happen in all good families, though, some members are neglected or overshadowed. This is the case with the view called “simple alethic pluralism”, and this paper seeks to remedy this. Accordingly, a new kind of simple alethic pluralism (termed “Plain Alethic Pluralism”) is advanced. It highlights the features of the meaning of the word “true”, in particular its focal meaning, while reaffirming the dependence of truth on the world. In drawing attention to the way speakers use the truth-predicate, making it acquire the meaning it has, it intends to qualify as a conception of truth with a human face.

1 Introduction

Within the secular and still ongoing discussion about truth, what is by now known as alethic pluralism has proved to be one of the most interesting proposals recently advanced in the field. Indeed, from the last decade of the last century onwards the field of theories about truth has been enriched with the idea that truth is not only one, but also many—many alethically potent properties that differently characterise different domains of discourse. Moreover, thirty or so years of discussion have also shown that “alethic pluralism itself is many rather than one” (Pedersen, 2006, p. 112)—a family of views with different metaphysical commitments (cf. Pedersen & Lynch, 2018).

As it may happen in all good families, though, some members are neglected or overshadowed. This is the case with the view called “simple alethic pluralism”, and this paper seeks to remedy this.

Accordingly, in §2 a definition of the view is provided, and four different versions of it are described. In §3 a fifth proposal is put forward, whose central element is what will be termed the focal meaning of the word “true”, which is explained in §4. With this in hand an account of the sense of “true” is offered in §5. §6 stresses the need for a proper account of the world in matters of truth. Finally, §7 rounds off the paper pointing out the advantages of the new proposal.

Just a clarification before starting. One of the questions which one has to state in talking about truth is the question of truth-bearers, i.e., the task of specifying which kind of things truth should be attributed to. I must confess that I do not have a definite stance on this issue, as I am inclined to think that what I want to say about simple alethic pluralism has little bearing on this. However, I will take propositions as truth-bearers just because statements, utterances, sentences, judgments, or the like can well be considered to express propositions.

2 Simple alethic pluralism

Let’s stipulate that a Simple Alethic Pluralism is any pluralistic conception about truth according to which the word “true” has multiple meanings. Usually a word that has multiple meanings is considered ambiguous, but I think it need not always be so. Basically, there are two cases in which we can speak of a word having multiple meanings: homonymous words and polysemous words. The former includes words whose meanings are unrelated, like “swallow”, “mole”, “tick”, “bat”, “rip”; the latter includes words whose meanings are related, like “cane”, “chair”, “car”, “tram”, “bar”. I propose to consider only the former as genuine cases of ambiguity. With this in mind, let us see how many simple alethic pluralisms there are currently on the scene. To my knowledge there are four: I will call them SAP₁, SAP₂, SAP₃, and SAP₄.

SAP₁ is the view envisaged (but not endorsed) by Michael Lynch, to the effect that the word “true” “conveys different concepts or meanings in differing contexts. Sometimes ‘truth’ means ‘correspondence’ (in the teleofunctional sense, say); sometimes it means ‘superwarrant’” (Lynch, 2009, p. 54). This renders “true” semantically ambiguous on a par with words like “step” or “bank”, which as mentioned possess completely different and unrelated meanings depending on the context in which they appear. On the authority of SAP₁, a clear sign that we have here different unrelated meanings is the fact that we must learn the meanings in question separately, on a case-by-case basis, without the slightest possibility of exploiting the semantical content of the very same word occurring in a different context. This is why, according to SAP₁, there are as many concepts of truth as many domains of discourse there are in which “true” occurs. Moreover, since the many occurrences of “true” in different domains can be taken to be occurrences of the ‘same’ word only by outward appearance, SAP₁ actually contemplates a plurality of truth-predicates which just happen to be homonymous, each one picking out a different property of truth.¹ That we have here different words that only happen to possess an identical form is reflected by the convention adopted by several dictionaries with regard to ambiguous words, i.e., of rigorously separating the explanation of the different meanings by numbers, not just alphabetical letters—or even by different entries.

Lynch himself underlines a bunch of problems that an alethic pluralism such as SAP₁ fatally encounters, chief among them the problem that in the end “it isn’t even a pluralist view of truth at all. It is a pluralist view of the meaning of the word ‘true’. Compare: we don’t say that there is more than one way to be a bank. We say there are different meanings to the word ‘bank’” (Lynch, 2009, p. 58; cf. Lynch, 2013, p. 27). Seen in this light, SAP₁ appears to be “a disguised form of truth nihilism” (Lynch, 2009, p. 59), given that it proves to be well attuned with claims such as “there are many kinds of ‘truths’, and therefore there is no truth” (Nietzsche, 2017, p. 313).²

So much for SAP₁. Another simple alethic pluralism on the scene, SAP₂, is the view envisaged (but not endorsed) by Crispin Wright that is very much like the preceding one in claiming that “true” has no single meaning, but different in a very important respect: it doesn’t presuppose that “true” is semantically ambiguous. So, the truth-predicate isn’t a member of the category “bank” or “step” belong to: it is rather like “‘fixing’, or ‘dispensing’, or ‘poor’, as in ‘Fixing dinner, fixing a race, fixing the car’; ‘Dispensing a prescription, dispensing justice, dispensing sweeties to the class’; ‘Poor relation, poor performance, poor wee thing’. These aren’t ambiguities exactly. You don’t have to learn each type of use separately” (Wright, 2013, p. 126).

Wright stresses that it is a fact about our ordinary linguistic usage that we sometimes stretch the use of some words, employing them in different contexts from the ones they were originally supposed to have their place. The verb “to stretch” is well-chosen—it gives the image of a widening of the use of a word to cover further cases, meanwhile keeping something of the use in the original case. This shows that that word’s meaning is elastic, i.e., extendable while capable of retaining what is necessary to make one recognize the word’s physiognomy in passing from context to context. Indeed, cases in which this physiognomy ceases to be recognized are a clear sign of the elastic’s rupture. Moreover, the fact that it is a question of elasticity tells us that stretching is not coining a metaphor. It is true that metaphors usually exploit the elasticity that is intrinsic in meanings in virtue of the manipulations we can do of an expression in the course of our talking; however, I am inclined to see metaphors not just as involving elasticity, but as what happens as the result of the elastic’s rupture—roughly, as the creation of a new meaning through the use of an expression, where recognition of the pertinence of this use is not immediate and at first glance appears incongruous. In contrast, nothing incongruous appears in the stretched use of an expression. This is how I see Wright’s account of SAP₂. SAP₂ is then “the thesis that there is a relevant species of semantic elasticity in the truth-predicate, and that it’s manifest in its various applications when you look carefully” (Wright, 2013, p. 127). In contrast to SAP₁, for SAP₂ there is just one concept of truth and one truth-predicate, which picks out a different property of truth depending on the context.

Although Wright declares to be somewhat sceptical about the potential of this reading of alethic pluralism, he thinks it is worth a careful analysis and suggests giving it a try. I am happy to follow his advice and present a kind of simple alethic pluralism along the lines he traced.³ Since I am not sure it is thoroughly compatible with SAP₂, I decided to call it differently. But, before describing it in some detail, let’s see what SAP₃ and SAP₄ amount to.

SAP₃ is the view put forward by Max Kölbel to the effect that “‘true’ as used in natural language is ambiguous” (Kölbel, 2008, p. 361; he doesn’t give a name to his view, so I call it SAP₃ just for the sake of the present explanatory list). The idea is that “even if de iure the truth-predicate is just a denominalizer, [it turns out] de facto to express an interesting property” (Kölbel, 2008, p. 363). The role of the truth-predicate as a denominalizer is the central tenet of (moderate) alethic deflationism, according to which truth expresses no metaphysical property attributable to a truth-bearer by the truth-predicate, so that the latter has just an expressive import. Therefore, all there is to it is the infinitely many non-contradictory instances of the Equivalence Schema “It is true that p if, and only if, p”—a schema that some authors prefer to call Denominalization Schema “since the sentence nominalized in the left branch of the biconditional is denominalized in the right branch” (Künne, 2003, p. 18; cf. Horwich, 1998, p. 5). Kölbel accepts this but disagrees with the deflationary claim that this exhausts the meaning of “true”, and thinks that, next to this deflationary meaning, speakers reveal that “true” has a more substantial meaning: it expresses at least one metaphysically significant property, e.g. that true propositions represent objective reality correctly. This combination, he maintains, renders the truth-predicate ambiguous: “on some occasions of use, ‘true’ expresses a deflationary concept [truthD] and on other occasions of use it expresses some substantial concept of truth [truthS]” (Kölbel, 2008, p. 368).

The idea that the truth-predicate is ambiguous makes SAP₃ close to SAP₁, and it is natural to inquire how close. Kölbel offers a tripartite understanding of the ambiguity in question: syntactic, semantic, and pragmatic ambiguity. The first kind comprises lexically ambiguous words such as “coach” or “bank”, where at the level of deep syntax we have distinct syntactic elements—“trainer” and “bus” in the former case, “river-bank” and “financial institution”, in the latter one. This is the case of homonymy typical of SAP₁. In the second kind of ambiguity we have only one syntactic element at the level of deep syntax, and its interpretation is context-sensitive: “gradeable adjectives like ‘rich’ or ‘tall’ have this feature [because they] supposedly express different properties in different contexts of use” (Kölbel, 2008, p. 368). This case is usually referred to as a case of polysemy—words that have distinct but related meanings. The third kind of ambiguity is located at the pragmatic level in that it has to do with those expressions that, while “syntactically unambiguous and semantically context-insensitive, [are] nevertheless not always used to express the same concept” (Kölbel, 2008, p. 369). This is the case of metaphors, for instance. Indeed, speakers sometimes use an expression to communicate a concept which is different from the one semantically encoded in that expression. In brief, according to Kölbel, if the standard case of ambiguity is represented by the syntactic case, “semantic ambiguity is usually called ‘indexicality’, […] and pragmatic ambiguity usually goes under the label ‘conversational implicature’” (Kölbel, 2008, p. 369).

Now, as far as the word “true” is concerned there is an initial oscillation in Kölbel’s thought. In Kölbel 2008 he does not really make a choice between the three cases, for he thinks that the available evidence (he provides some) doesn’t help decide which of the three kinds of ambiguity applies to “true”. Nevertheless, he clearly inclines toward syntactic ambiguity, provided that certain further methodological assumptions are made, and since the latter ambiguity is the case of “bank” and “coach”, we might conclude that there is a sort of overlap between SAP₃ and SAP₁. However, in clarifying what is the relationship between the two meanings expressed by “true”—truthD and truthS—he contradicts this conclusion.

He compares “true” to words like “dog” and “duck”, which have a neutral meaning (general dog and general duck) and a specific meaning (male dog and female duck)—the latter arising by way of distinction with “bitch” and “drake”, respectively. This kind of double meaning is different from the one possessed by “coach”, because in the case of “coach” the alternative meanings are mutually exclusive: “nothing is both a coach in the sense of bus and also a coach in the sense of trainer” (Kölbel, 2008, p. 375). On this basis Kölbel contends that the deflationary and substantialist meanings expressed by “true” are related in the same way in which the meanings in “dog” are related, so that the ambiguity of “true” is on a par with the ambiguity of “dog”: “truthS is a species of truthD” (Kölbel, 2008, p. 375; my italics), because “the concept truthS [is] subsumed under the concept truthD” (Kölbel, 2013, p. 288; my italics).⁴ Since this idea is confirmed in Kölbel 2013, we can conclude that this is what he actually thinks, stopping the oscillation initially detectable in his thought. (Accordingly, Pedersen and Lynch give his position the label polysemy pluralism: cf. Pedersen & Lynch, 2018, p. 554.)

Notice, however, that Kölbel’s conclusion leads us to think that “dog” and “duck” aren’t ambiguous at all. Every competent speaker of English knows quite well that “dog” can be used in all the cases in which “bitch” is used (barring the necessity of using “bitch” to refer to female dogs in specific cases), but not vice versa. And the same goes for “duck”. Moreover, quite interestingly, in order to evaluate Kölbel’s proposal Stefano Boscolo and Giulia Pravato submitted the truth-predicate to three well-established tests for detecting ambiguities, and the result of their analysis is that “‘true’ is not ambiguous in Kölbel’s sense” (Boscolo & Pravato, 2016, p. 50). Therefore, despite Kölbel’s insistence on the label “ambiguity” to refer to his stance, based on the very argument he deploys “true” isn’t ambiguous at all, and this makes it appear a certain closeness between SAP₃ and SAP₂—a closeness regarding just the idea that the truth-predicate isn’t ambiguous, however clear in Kölbel’s mind.

Be this as it may, we have that in contrast to SAP₁ and SAP₂, for SAP₃ there are at least two concepts of truth expressed by the truth-predicate, which therefore picks out a different property of truth depending on the context.

Finally, SAP₄ is the view that one can gather from Andy Yu’s thesis according to which the word “true” as a predicate of sentences is vague, context-sensitive, and ambiguous (Yu, 2021). This is the lesson we can take from the semantic paradoxes, according to Yu, since a possible solution to them resides in taking the concept of truth as indefinitely extensible thanks to subtle shifts in the meaning of “true”: “‘true’ shifts from less inclusive meanings to more inclusive ones. In my view, this is because the underlying concept that ‘true’ expresses, truth, itself changes” (Yu, 2021, p. 557). Given that these shifts form “a hierarchy of concepts of truth and a corresponding hierarchy of truth predicates in formal languages” (564), each with its distinct meaning, at first glance it would seem that SAP₄ is just like SAP₁, but it would be a wrong impression.

First, Yu himself thinks that the concepts generated by the indefinite extensibility of truth are subsumed under a unified concept, which in a sense guarantees the semantical stability of the truth-predicate. The concepts subsumed are labelled “subconcepts” and the idea is that “the unified concept of truth is coarser grained, while the various subconcepts are finer grained. Still, each of these concepts deserves the honorific ‘concept of truth’” (561). The presence of a single general concept of truth thus ensures the constancy of the meaning of the truth-predicate across time and contexts. Second, he believes that the multiple meanings the truth-predicate acquires in the hierarchy are related and not completely disconnected from each other—he uses the label polysemy to characterise the semantical nature of “true”.⁵ Third, although Yu is less interested in analysing truth in natural languages than in providing a solution to semantic paradoxes, and although he is fully aware that “we cannot make straightforward inferences about what a given theory says about natural language from claims it makes about formal languages” (564), he believes that ultimately his main thesis has a bearing on the truth-predicate used in a natural language—hence, his belief that the polysemous character of the predicate “in English is represented by means of distinct expressions corresponding to distinct (formal) languages” (564).

For all this I take it that SAP₄ is, mutatis mutandis, like SAP₂. Next up, my proposal.

3 Plain alethic pluralism

The kind of simple alethic pluralism I am going to put forward lies in the wake of SAP₂, and therefore inherits its central tenets. It’s just like SAP₂ in spirit, if not in letter—I’ll call it “plain alethic pluralism” (henceforth PAP). It is the view according to which the many occurrences of the word “true” are not occurrences of different homonymous words like the occurrences of “coach”, “bank”, and “case”—one single phonological form and distinct unrelated meanings—but they are occurrences of the same polysemous word, as in the case with words like “face” and “foot”—one single phonological form and distinct related meanings. The central idea of PAP is that the word “true” is not univocal—it does not have the very same meaning in each context of usage—nor equivocal—it does not have a completely different meaning in each different context where it is correctly used. Roughly, the two occurrences of the word “true” in “It is true that some drops of wine spilt on the table” and “It is true that Chaplin’s tramp is a funny character” do not mean the same thing because in the former case “true” conveys the sense of correspondence to an empirical state of affairs, while in the latter case it conveys the sense attributed to it by the set of principles that in a certain cultural context fix what counts as humour. Yet, these two uses of “true” are not entirely unrelated: if not by way of intuition, we see that there is a relation between them and understand the second occurrence without the need to learn an utterly different meaning. If, as I think, all this reflects the actual use of the word “true” we do in the many circumstances of our daily life, then the word is not equivocal or semantically ambiguous. On the contrary, the fact that the many uses of the word “true” across different domains of discourse convey “not always absolutely the same meaning, nor several absolutely different meanings” (Austin, 1961b, p. 42), reveals that there is an analogy between them. How does PAP account for this analogy? Let’s start by seeing what the just quoted philosopher has said about the analogy among the uses of a same word.

First of all, Austin warns us to handle questions of analogy and similarity with due care, given the risk of hastily saying that certain things are similar, disregarding the respects in which they are different: “If we say that [pleasures] are all called ‘pleasures’ ‘because they are similar’, we shall overlook [the fact that they] also differ precisely in the way in which they are pleasant” (Austin, 1961b, p. 41). This warning can be of some help given the pluralist assumption of our analysis: just as it would be a mistake to think “that pleasure is always a single similar feeling, somehow isolable from the various activities which ‘give rise’ to it” (Austin, 1961b, p. 41), so it would be a mistake to think that truth amounts to a single similar thing, somehow isolable from the various propositions which we justifiably call true.

Secondly, Austin claims that it is no simple task to say what the analogy between the different uses of a same word amounts to, in that there can be different reasons for which we call different things by the same name, and these reasons are often of the wrong type. In particular, “it is not in the least true that all the things which I ‘call by the same (general) name’ are in general ‘similar’, in any ordinary sense of that much abused word” (Austin, 1961b, p. 37). As a result, he goes on to say, talk of similarity and analogy can be quite void and misleading. Take for instance the phrases “the foot of a mountain” and “the foot of a list”: here “there is a good reason for calling the things both ‘feet’ but are we to say they are ‘similar’?” (Austin, 1961b, p. 40). Or take one of Aristotle’s favourite examples (you can come across it e.g. at the beginning of Metaphysics Γ): the adjective “healthy” when applied to a body, a complexion, or an exercise. Austin remarks: “are we to be content to say that the exercise, the complexion, and the body are all called ‘healthy’ ‘because they are similar’? Such a remark cannot fail to be misleading. Why make it? And why not direct attention to the important and actual facts?” (Austin, 1961b, p. 39).

The latter recommendation leads to the real interesting feature of his analysis, i.e., the proposal to break the sense of a word such as “healthy” down into components, and to call the sense in which it is used of a healthy body the primary nuclear sense of “healthy”. He explains: “I call this nuclear because it is ‘contained as a part’ in the other two senses which may be set out as ‘productive of healthy bodies’ and ‘resulting from a healthy body’” (Austin, 1961b, p. 39). I find this interesting because it gets us closer to the characterization of the analogy between the different uses of “true” that is at the centre of PAP. Closer, but just up to a certain point, though: as appealing as it might sound, Austin’s proposal isn’t entirely satisfactory.

The reason lies in the fact that the adjective “nuclear”—deriving from “nucleus”—gives the idea of an essence owned by the meaning of “true” and “truth”—an essence contained as a part in every use of these words. But, whatever the meaning of an expression is taken to be, it hardly appears plausible to suppose that it is centred in an essence. Moreover, given the relationship between the meaning of an expression and its reference, it would be quite odd—to say the least—to suppose that all the items falling into the extension share a given essence. A criticism Enrico Berti addressed to Austin is appropriate here.

Referring to Aristotle’s example, Berti maintains that “we cannot say that health is the essence of the healthy medicine, or the healthy climate, or the healthy complexion. We must say that health is the product of the healthy medicine, or the thing preserved by the healthy climate, or finally the thing manifested by the healthy complexion. In all these cases there is surely a relation between the single healthy thing and health. This relation is nevertheless different in each case” (Berti, 2001, p. 195; my italics). Berti’s point is that, far from trying to envisage a purported essence, we should pay attention to the web of relations that surrounds our object of study—being it something such as health or truth. (And also—I would add—to the web of relations around the relevant word, given the correspondence between the web surrounding the thing and the web surrounding the word.) As to truth, we could then say that the word “true” has a primary meaning, and that the meanings the word acquires in different contexts of usage bear a different relation to it. In a nutshell, there is a primary meaning, on the one hand, and a constellation of different meanings related to it, on the other hand. Thus conceived, the primary meaning “is only a term of reference, i.e., that to which the others stand in relation […], and it is common to all just for this reason” (Berti, 2001, p. 196), not because it is or represents an essence. Berti himself finds it more convenient to characterize the notion of primary meaning by employing an expression used by G. E. L. Owen, i.e., focal meaning. It is an expression meant “to describe the first of the meanings of a word said in many senses, but all relative to one of them” (Berti, 2001, p. 198; cf. Owen, 1986, p. 184), and I happily borrow it. Indeed, this expression “does not presuppose that the first meaning is a part of the others, but indicates that it is only the focus, i.e., the term of reference, of the others” (Berti, 2001, p. 198).

We can find such an analysis of focus in virtually every perspective according to which “the word ‘sense’ […] is much more flexible than the word ‘meaning’” (Putnam, 1992, p. 375), as questions like “In what sense do you mean that?” reveal. The idea is that the meaning of an expression E constrains the use of E, without always determining its sense, so that the latter is a sort of enrichment of the former that can be different on a context-by-context basis. Central to this picture is “a conception of linguistic meaning as least common denominator [where] the meaning of an expression is the minimal content that must be associated with it by a rational agent […] in order to communicate with other members of the linguistic community” (Soames, 2010, p. 172). And this is the case of the predicate “true”, which has a minimal meaning that, functioning as ‘least common denominator’, constrains its use in every context—making it recognizable as that predicate—and then gets stretched to acquire an enlarged sense in virtue of the many uses in different domains of discourse.

From a linguistic point of view, PAP is then the view according to which the analogy between the occurrences of the truth-predicate in different contexts consists in “true” having a focal meaning shared as a term of reference by all its occurrences. But, now, what could be said about the focal meaning of “true”?

4 The focal meaning of “true”

I think that it consists of (at least) three elements: the Equivalence Feature, the Minimal Correspondence Feature, and the Normative Feature. Let’s look at them one by one.

According to the Equivalence Feature, every use of “true” answers to the aforementioned Equivalence Schema “It is true that p if, and only if, p”, for every proposition p. This is a purely logico-linguistic feature, has a ‘formal’ character, and allows “true” to fulfil important logical needs: to begin with, the need to express a proposition that for some reason we cannot explicitly identify (e.g., “What Isabella said is true”), or the need to concisely express an infinite number of propositions (e.g., “Every proposition of the form ‘p is F or not F’ is true”), or the need to show agreement with a proposition someone expressed (e.g., when during a conversation we utter “That’s true”, performatively corroborating what an interlocutor formerly stated; the same goes for “That’s false” and disagreement). These are all uses of “true” permitted by the Equivalence Schema, compliance with which shows that we do have a predisposition to state one of the horns of the equivalence when we state the other, and vice versa.⁶

The second semantic feature of the truth-predicate is the Minimal Correspondence Feature—if a proposition is true, it is because something makes it so. This is an example of philosophical awareness that we have possessed at least since the time of Aristotle, who gave voice to nothing but a piece of common sense which characterizes a basic trait of our thinking. As is well-known, it was Aristotle one of the first to make explicit that “it is not because of our truly thinking you to be pale that you are pale, but it is rather because you are pale that we who say this speak the truth” (Metaphysics Θ, 10, 1051^b8-9). Centuries later Michael Dummett has maintained that “If a statement is true, there must be something in virtue of which it is true” (Dummett, 1976, repr. 52), deeming this a ‘regulative principle’ concerning truth. This feature conveys the immediate and natural insight that, if a proposition is true, it is so by virtue of ‘something else’—something different from it. I call it ‘minimal’ because the tie it envisages between a true proposition and what makes it true doesn’t need any commitment to a metaphysically heavy understanding of truth—typically, a correspondence theory. Rather, the latter should be seen as a sort of elevation of this feature to a general metaphysical level, whereas the feature itself expresses nothing but a pre-philosophical intuition of an explanatory character regarding how a proposition comes to be true. In this respect, both Crispin Wright’s Correspondence Platitude and Michael Lynch’s Objectivity Truism seem to make the same point (cf. Wright, 2001, p. 760; Lynch, 2009, p. 8). Finally, it should be clear that the Minimal Correspondence Feature is intended to give the general idea that what is true is always objective, hence also in cases in which truth and falsity are stipulated by us, as it happens in epistemic situations such as the Rules of the Road and the Penal or Civil Code.⁷

The third element constituting the focal meaning of the word “true” is the Normative Feature. Virtually nobody would object to claims such as “to regard an assertion or a belief or a thought as true or false is to regard it as being right or wrong” (Putnam, 1999, p. 69). It appears a piece of common philosophical wisdom to think that truth exerts a normative constraint on assertions—wisdom that comes from observing the linguistic practice of human beings. For instance, one of the harshest criticisms we can make of an assertion made by someone is simply “it’s not true”. Failure to stick to what is the case immediately proves an assertion to be defective. All this then makes truth appear as a normative concept. It is possible to discern different dimensions along which the normativity of truth can be analysed: attribution of truth to an assertion may signal that one deems the assertion correct, or good to entertain, or the goal one was aiming at, or that everybody ought to entertain it. Truth may thus function as a criterion to sort the correct assertions from the incorrect ones, or as a value-conferrer, or as the aim of inquiry, or as the sign of the obligation to entertain a given assertion. Following Filippo Ferrari (2016, 2020, 2022), we may distinguish four dimensions of the normativity expressed by truth: criterial, axiological, teleological, and deontological. Now, since we are considering what we intuitively deem the normative import of truth to be, I surmise that the last quotation emphasises that at a fundamental level we feel that the normative nature possessed by truth has to do with the ‘correctness’ of what we think or say, and points out that the only normative dimension we naturally attribute to truth is the criterial one. In light of this we can claim that the Normative Feature of the meaning of “true” is to the effect that truth represents a criterial norm—it expresses the correctness of the assertion it is applied to. This prompts two remarks.

As far as I can see, there is just one domain in which the focal meaning of “true” gets anaesthetised, so to speak, i.e., both the Minimal Correspondence Feature and the Normative Feature are ineffective, and elastically shrinks covering the meaning ‘it’s all subjective’: the domain of basic taste. When we say something like “It is true that eels are disgusting” or “that Neapolitan pizza is the most delectable food in the world is true”, the use of the truth-predicate doesn’t show its usual Minimal Correspondence and Normative Features, but is governed only by the Equivalent Feature—by its expressive and performative aspects. The truth-predicate is used in an empty way, as we might say. In this domain truth doesn’t appear with its metaphysical and epistemological load: no conformity with an objective standard or with an intersubjective norm is in question here, but merely an expression of agreement with a given proposition—something analogous to “I agree that eels are disgusting”, accompanied by the deep awareness that nothing objective is being conveyed. This habit is so entrenched in our practice that from Medieval times a maxim is virtually universally shared to the effect that De gustibus non est disputandum. Accordingly, I advance the following Dormancy Conjecture: in the domain of basic taste the metaphysical and epistemological aspects of truth are dormant and truth enters only with its logico-linguistic property.⁸

Despite the maxim just mentioned, the Dormancy Conjecture may strike us as unfaithful to our practice, in that there is an obvious link between the concept of belief and the concept of truth: believing the proposition p is equivalent to believing that p is true, and vice versa—and, one may argue, this link obtains also in the case of beliefs about basic taste. But notice that this link has merely a logico-linguistic character, it is inherently formal, and nothing metaphysically substantial is involved by it. Indeed, it suffices the Equivalence Feature to license it. Hence, there is no alethic, normative or other kinds of fault here: domains like this are devoid of any objective standards—and subjectivity prevails.⁹

The second remark is that—being the taste domain practically the only one that “anaesthetizes” the concept of truth—the sense of the word “true” carries with it a normative load in all the remaining domains of discourse. Every use of this word has a normative connotation, in the sense of the criterial dimension: it underlines the correctness of what is being said. This also applies to usages such as “true friend” or “true work of art”, through which we state the conformity of the name’s referent to a certain standard. Therefore, a true friend is a person who fully fulfils the qualities embedded in the meaning of the word “friend”, and a true work of art is an object that meets all the requirements (however nuanced) set by the current idea of a work of art (for a brief discussion of this point, cf. Künne, 2003, p. 105).

Thus, three are the features of the focal meaning of “true”. They enable truth to function as a passe-partout across the many domains of discourse, averting the risk of the predicate’s semantical ambiguity and ensuring a unity to the concept itself: a unity throughout the plurality of the different meanings it acquires within the domains in which it is correctly applied—a plurality that makes up the sense of “true”. What plurality am I talking about?

5 The sense of “true”

What happens is that the focal meaning combines with the meaning the predicate gets when occurring within a specific domain of discourse—which we may call the domain-related meaning of the predicate. This is a meaning couched in terms of correspondence, superassertibility, coherence, etc. depending on the domain in question, which brings to light the differences the usage of the predicate is elastically put to. For instance, in the empirical domain we use “true” to mean correspondence, in the mathematical domain coherence, in the comedic domain coherence again (with a given set of standards for comedy possibly different from culture to culture), in the ethical domain superassertibility, etc. All these uses are interrelated in virtue of the focal meaning of “true”: as soon as the word “true” is used within a given domain, the focal meaning enlarges, as it were, to cover the use the word is put to in this domain. It’s a case of stretching, to use Crispin Wright’s well-chosen expression, not conversational implicature or metaphor—not something “over and above what one asserts” (Soames, 2010, p. 145). The focal meaning embodies then the general semantic features of “true”, while the domain-related meanings may be regarded as the specific semantic features of the truth-predicate, each tied to the property that the given domain specifies. Accordingly, the latter account for the plurality inherent in truth talk, while the former accounts for its unity.

So far, the explication of what PAP is has focused mainly on the sense of the word “true”, and this may give the mistaken impression that PAP is nothing more than a truth pluralism at the level of sense, dealing with a word and its uses regardless of the world. We have indeed been mentioning properties—something typically taken as a part of the world—but the impression might arise that we have a variety of truth properties because we have a variety of uses of the predicate “true”, and that therefore properties are involved into the picture just because of the many different applications we make of “true” to truth-bearers—which in a sense represents the metaphysically minimal commitment to properties a speaker can reveal. A bit of clarification is in order here.

6 In the beginning was the world

The fact that PAP stresses the importance of linguistic usage in the understanding of truth doesn’t mean that it is a “language-oriented” view of truth or an idealistically-oriented conception. On the contrary, PAP acknowledges and underlines the centrality of the world in matters of truth—truth depends on the world. Only, according to PAP the world is multifaceted: it possesses human and non-human features that differently impinge on our life and language, constraining our verbal and non-verbal behaviour.

Roughly, the human and non-human features of the world constitute different “fields of states of affairs”, each enabling a determinate domain of discourse and specifying a property that the use of the predicate “true” captures and attributes to whatever proposition p of that domain is entitled to it. This entitlement comes from the assessment of p’s truth-value, which consists of gathering evidence in favour of p within the field of states of affairs in question, while the property that the predicate captures is the property relevant to truth in that field. This occurs because the predicate naturally captures whatever property satisfies the features of the focal meaning of “true” (more on this below). Thus, a state of affairs represents a fact of the matter which is responsible for p’s truth or falsity—equivalently, for the attribution (or non-attribution) of the field-related property through the application of the truth-predicate to p. As mentioned, all this applies also to features that human beings create, as happens, for example, in socio-institutional settings. In cases like these, the truths and falsehoods that belong to a domain of discourse are of our own making—we are the ones who determine when a proposition can be said to have a certain truth property. Nevertheless, these truths and falsehoods have a kind of mind-independence and don’t lose their objective validity—they are truths and falsehoods in their own right.¹⁰

So, according to the metaphysics of PAP there are many properties relevant to truth couched in terms of correspondence, coherence etc., each related to a specific field of states of affairs, and a general truth property whose nature is grasped simply by understanding the meaning of the predicate “true”. This happens because the property of truth enjoys the features that the focal primary aspect of the meaning of the predicate “true” enjoys—Equivalence, Minimal correspondence, and Normativity. Thus, possession of this property by a proposition p makes p minimally correspond (in the sense specified in §4) to the state of affairs it deals with, makes it correct to say, and equivalent to an application of the truth-predicate to it. So conceived, the general property of truth is irreducible, in that to understand it no reduction to a more fundamental concept is needed, being it fully described by the meaning of the truth-predicate.

On the other hand, the field-related relevant-to-truth properties are so relevant precisely because they too satisfy the features of the concept of truth as expressed by the focal meaning of “true”. As we noted above, these features are intuitive and pre-theoretical: our intuition guides the use of the truth-predicate within a given domain of discourse, enabling it to capture in the corresponding field of states of affairs exactly that property which possesses the same characteristics as truth in that field.

Douglas Edwards has interpreted the relationship between these two kinds of property as an ‘entailment’ relationship and has exemplified it with the aid of conditionals, each relating to a specific domain of discourse: in domain x, if proposition p has the property relevant to truth specified by the field of states of affairs that enables x, then p has the general truth property (my wording: for Edwards’, cf. Edwards, 2011, p. 43, 2018, p. 126). Edwards calls his theory determination pluralism precisely because the property mentioned in the antecedent of the conditional ‘determines’ the property of truth mentioned in the consequent. Determination pluralism is the best illustration of the metaphysics underlying PAP, in that it requires no special metaphysical mechanism to explain how the two kinds of property are combined and thus helps keep PAP simple.¹¹ It also helps PAP differentiate itself from two proposals that have acquired centre stage in the debate on alethic pluralism: Crispin Wright’s (1992, 1999) minimalism and Michael Lynch’s (2009) manifestation-functionalism. According to the former, once realized that the concept of truth is whatever concept that satisfies a set of platitudes concerning truth (cf. e.g. Wright, 1999, p. 227), one also realizes that “truth may consist […] in the instantiation of one concept in one area, and in that of a different concept in another” (Wright, 1999, p. 228, my italics; in the 2003 reprint (p. 273) “concept” is replaced by “property”). The properties relevant to truth that each field of states of affairs differently expresses play the role of truth in their respective field: they “serve as truth by dint of satisfying those [platitudes]” (Wright, 2001, p. 752). Now, although Wright seems to argue that the idea just seen does not amount to an identification between concept and properties, the very idea of instantiation, or of serving as truth, goes very close to an identification—the property functions as a proxy of the concept, it stands in for the concept. Therefore, since all in all the field-related properties identify with truth,¹² we have as many truth concepts as there are fields that allow truth talk—and no general truth concept that remains constant across fields.¹³

Lynch also starts by listing some platitudes about truth, which he calls “truisms”, and these “tell us that true propositions are those that have a property that has a certain function in our cognitive economy” (Lynch, 2009, p. 71). According to his theory, there is a general truth property, and this is manifested in the properties relevant to truth, or equivalently immanent in them. In this case the general truth property is somewhat contained in the properties relevant to truth as a part, and this mixture makes it difficult to keep the two kinds of properties distinct. Again, the possibility of characterising truth independently of any field is compromised, if not lost, and the result is that there is no clear general field-free truth property—contrary to Lynch’s thought.

This causes both Wright’s and Lynch’s theory suffer from a serious drawback that several critics noted shortly after the appearance of pluralist alethic theories. The drawback has to do with so-called mixed propositions, i.e., compound propositions consisting of two or more propositions belonging to different domains of discourse. Suppose we are analysing a true mixed proposition, such as the conjunction “Chaplin’s tramp is funny and 3 is a prime number”, trying to make out why it is true. If the two conjuncts are true in virtue of different properties—each conjunct acquiring a property relevant to truth from a specific field of states of affairs—then the following question imposes itself: in virtue of which property is the conjunction true?¹⁴ By failing to give a distinctive characterization either of the general truth property or the concept of truth, on the one hand, and the field-related relevant to truth properties, on the other, it seems that both Wright’s and Lynch’s theories are unable to provide a satisfactory answer to the latter question—in both cases truth does not seem able to free itself from the ties that bind truth-relevant properties to fields of states of affairs, and therefore in both cases it is difficult to say what the general truth of the mixed conjunction is per se.¹⁵ On the contrary, for both determination pluralism and PAP the two kinds of property are clearly distinct, and this is a definite plus.

I said above that—on the level of language—any domain-related meaning of the truth-predicate is tied to its focal meaning: now I can say that this is because—on the metaphysical level—having a property relevant to truth entails being true, i.e., having the truth-property.¹⁶ In light of this I contend that PAP is not only a truth pluralism at the level of sense or concept, but also at the level of properties.

Now I would like to emphasize the ‘symmetry’ between the semantics of predication and the metaphysics of properties—once it is made clear that the former depends on the latter. We have seen that there is a correlation between domains of discourse and fields in the world, in that the truth-predicate enters any domain with its (default) focal meaning and picks out the specific property that in the correlative field is relevant to truth, thereby stretching its semantical value by acquiring a domain-related meaning. Now, the relationship between the focal meaning and the domain-related meanings, on the one hand, and the relationship between the general truth property and the field-related properties that are relevant to truth, on the other, reveal an interesting symmetrical configuration: just as the focal meaning is not a part of the domain-related meanings, but only their term of reference, so the general truth property is not a part of the field-related properties that are relevant to truth, but only their term of reference—that to which the field-related properties stand in a relation of entailment—and is common to all just for this reason.

We may summarize all this by claiming that truth inherits its substance from the world. If there is a diversity of uses of “true”, as PAP contends, this depends on the plurality of fields of states of affairs in the world and on the fact that those uses respond to the many properties specified by this metaphysical plurality. This is how PAP secures both conceptual unity and metaphysical plurality: it countenances one concept of truth, one truth property, and one truth-predicate, which captures a different property relevant to truth depending on the context (the property of correspondence, coherence, etc.). It is the view according to which in the course of our practice we reveal a remarkable flexibility in the use of “true” across domains. In some of them, the use stretches so far as to express the sense of correspondence or coherence etc., in others it simply conveys the logico-linguistic feature dear to the alethic deflationist¹⁷: the Equivalent Feature. Moreover, the connection meaning–world that is at the basis of PAP helps to show how PAP manages to ward off the charge that Lynch directs against a simple alethic pluralism—SAP₁, in his mind—i.e., the charge to the effect that it isn’t a view of truth at all but a view of the meaning of “true”. Undeniably, PAP is a (pluralist) view of the meaning of the word “true”, but, for one thing, since we might conceive of this word as the representative of truth on the level of language, we can say that talking about “true” is at the same time talking about truth, and vice versa; and, for another, the possibility to talk about truth in talking about the word “true” is enabled both by the connection language–world and by the focal meaning that plays a unifying role that binds uses of the word in different contexts—being the unrelatedness of the meanings the word takes within different contexts the real problem with SAP₁. Hence, if we believe that the concept of truth is revealed by the sense of the word “true”, that the use of “true” is driven by the concept of truth, and that the concept is driven by the world, we can conclude that PAP qualifies as a pluralistic view of truth on all counts.

7 The advantages of PAP

Since there are currently several pluralist positions, the question arises as to why another proposal should be considered. I therefore conclude the paper by pointing out some advantages of PAP: at least two points should be emphasized.

PAP nicely stands alongside other pluralist positions in addressing the problem I mentioned in the previous section—the one concerning mixed propositions. This is not unique to PAP, but I think it is worth pointing out given the emphasis that the discussion of mixed problems has taken on the analysis of alethic pluralism since its inception. Based on PAP we can claim that, even though there is no single domain a true mixed proposition belongs to, its truth is explained by logic—i.e., by the truth table for the involved connective—and the mechanism of truth-determination mentioned above.¹⁸ For instance, if we have a mixed proposition such as p₁ & p₂, then the truth value of each component is determined by assessing whether it possesses the metaphysical property that in the field of states of affairs in question is relevant to truth. If so, the component is true in virtue of its possession of the general truth property determined by the metaphysical property. Hence, by applying the standard truth table for the conjunction connective, we know that p₁ & p₂ is true if both conjuncts are true. On this basis we can say that our proposition of the form p₁ & p₂ is true.¹⁹

So much for the first point I wanted to make. As for the second, PAP tries to provide an account of truth that is as simple as possible. As we have seen, the general truth property is simple in that to understand what it is there’s no need to reduce it to other (seemingly more fundamental) properties: its nature is entirely captured by the features specified by the meaning of the truth-predicate—Equivalence, Minimal correspondence, and Normativity—revealed in turn by our use of the predicate itself. It is a ‘lightweight’ property because it depends only on the use of the predicate and possesses no metaphysical character—the metaphysical load being borne by the field-related properties. As for how field-related properties confer truth to a given proposition, PAP again makes it as simple as possible, since it is the result of an entailment obtained under a sheer logical move, and no cumbersome metaphysical mechanism of determination is called for.

Moreover, the simplicity of PAP also comes from the fact that it emphasises the role of speakers in truth talk—sometimes left at the margins of the discussion about truth. We are the ones who realize that the truth-predicate can be applied to a proposition belonging to a given domain, we are the ones who see that field-related properties have something in common with the general truth property, we are the ones who make the word “true” acquire the meaning it has. It is thanks to us that platitudes and truisms about truth play a role—a role that is shown by our use of the truth-predicate. The platitudes and truisms listed by Wright and Lynch, respectively, are what they are because that is how they appear to us—they are not ‘out there’, so to speak, they are not principles we receive from above. That is why they are embedded in the focal meaning of “true”.

PAP therefore helps to express an obvious fact concerning truth. Namely, that truth is a human matter: if there were no human beings, there would be no truth. There would probably be mountains and lakes, but no truth about mountains and lakes—not to mention truths of an ethical, aesthetic, social, political kind. Truth gains its objectivity and mind-independence from the world, but it is our use of language that fixes and strengthens the connection between mind and world, showing how truth is parasitic on the latter. If the specific traits of a domain of discourse—degree of closeness to experience, degree of mind-independence, width of the role of argumentation (itself a matter of degree) vis-à-vis poor evidence, kind of truth-relevant property taken from the world, etc.—succeed in conferring a certain meaning on the truth-predicate, this happens because of our role within the entire picture. In stressing this, PAP qualifies as a conception of truth with a human face.