ABSTRACT
Gareth Evans (1985) and Sven Rosenkranz (2008) have respectively formulated two objections to truth relativism that would show that this view does not cohere with our practice of asserting. I argue that the relativist should answer such objections by appealing to the notion of assessment sensitivity. Since the relativist accounts for this notion by means of a technical truth predicate relating propositions to contexts of assessment, the task left to her turns out to be to make sense of assessment sensitivity by making sense of this predicate (i.e. by showing that it expresses a truth notion).
Keywords
Assertion; truth relativism; assessment sensitivity; monadic truth; non-monadic truth
RESUMO
Gareth Evans (1985) e Sven Rosenkranz (2008) formularam, respectivamente, duas objeções ao relativismo da verdade que mostrariam que esta visão não é coerente com nossa prática de afirmar. Defendo que o relativista deve responder a tais objeções apelando à noção de sensibilidade da avaliação. Tendo em mente que o relativista considera essa noção por meio de um predicado de verdade técnica, a tarefa deixada para ele calha de ser dar sentido à sensibilidade da avaliação mediante dar sentido a esse predicado (quer dizer, mostrar que ele expressa uma noção de verdade).
Palavras-chave
Afirmação; relativismo da verdade; sensibilidade da avaliação; verdade monástica; verdade não monádica
Truth relativism, as it is understood here, holds that both the truth of certain propositions and the accuracy (i.e. correctness in a truth-related sense) of their assertions1 can vary with the value taken by some non-traditional parameters in the circumstances of evaluation (e.g. a standard of taste, a standard of knowledge, an information state, a judge or a perspective).2 In John MacFarlane’s (2005, 2007, 2011, 2014) framework, such truth and accuracy relativizations boil down to a relativization of truth to a context of assessment fixing certain parametric values.
Some authors have argued that a theory that relativizes the accuracy of assertions does not cohere with our practice of making assertions (Evans 1985, Rosenkranz 2008).3 In this essay I argue that the relativist should deal with these objections by appealing to the notion of assessment sensitivity introduced by MacFarlane. Be that as it may, as we shall see, the relativist still needs to make proper sense of this notion.
In the first section, I explain the main features of MacFarlane’s framework. In the second section, I address an objection to truth relativism that was raised by Sven Rosenkranz (2008) and calls our attention to the relation between the notion of propositional truth and the notion of accuracy as applied to assertions. I argue that in order to answer that objection the relativist has to appeal to the notion of assessment sensitive truth, which allows her to make justice to the traditional picture of how truth and accuracy are related. To make my case, I criticize Max Kölbel’s (2008c, 2009) alternative way of conceiving of this relation, which he takes as compatible with the relativization of accuracy and does not tread on the notion of assessment sensitivity. In the third section, I address the objection famously raised by Gareth Evans (1985) to the relativization of the accuracy of assertions. In this section, I argue that the most promising response to this objection also rests on the notion of assessment sensitivity. In the fourth section I show that the relativist still needs to make proper sense of this notion, which amounts to making sense of the notion of truth relative to a context of assessment (i.e. to show that this notion is a truth notion), and briefly take notice of two possible approaches to accomplish this task. The moral of this section is that the proposed answers to the objections leveled by Rosenkranz and Evans are dependent on whether the relativist can actually make sense of assessment sensitivity, which is a question this paper leaves open. Be that as it may, the paper attempts to show that, as they stand, these objections fail, as long as they do not take into account the notion of assessment sensitivity and so the question of whether we can make sense of it. In the fifth and last section, I give a brief summary of the paper.
1. Truth relativism as an account of assessment sensitivity
Truth relativist views are meant to account for our use of some expressions (e.g. predicates of personal taste, the verb ‘know’ or epistemic modals) in declarative sentences. Relativists (Lasersohn, 2005; MacFarlane, 2007, 2014; Kölbel, 2009; Richard, 2008) argue that only a relativist treatment of such expressions can vindicate certain disagreements and retractions.4 Consider an alleged disagreement where John sincerely and assertively utters sentence (1) and Ann does the same with sentence (2):
-
(1) Mutton is tasty.
-
(2) Mutton is not tasty.
According to the relativist, we have at the same time the impression that John and Ann disagree with each other and the impression that both of them are perfectly entitled to make their assertions and have the beliefs their assertions express - what else but their own taste could license their assertions and beliefs?5 Max Kölbel (2003, 2009) introduced the category of faultless disagreement to describe such cases: according to him, these cases involve a genuine disagreement between two people that are both free of fault.
A relativist treatment of ‘tasty’ would vindicate these appearances by (i) including a novel parameter NP in the circumstances of evaluation (e.g. a standard of taste parameter, a judge parameter or a perspective parameter); (ii) taking the content of ‘tasty’ as well as the content (i.e. a proposition) of a simple sentence containing this expression as neutral to this parameter (i.e. as not conveying information about a particular taste standard, judge or perspective); (iii) taking the extension of ‘tasty’ and so the truth-value of the propositions expressed by simple sentences containing it as possibly varying with the value taken by this parameter; (iv) and taking the accuracy of the assertions of (beliefs in) these propositions as relative to NP, just as truth is. Thus, John and Ann would be fault free in the sense that their respective assertions and beliefs would be accurate relative to their own standard of taste or perspective, and they would genuinely disagree with each other insofar as, from any single standard of taste or perspective, the accuracy of John’s assertion and belief precludes the accuracy of Ann’s assertion and belief and vice versa. In other words, a disagreement would consist in the phenomenon MacFarlane (2014, pp. 125-128) classes as preclusion of joint accuracy. 6,7
Insofar as the existence of propositions is conceded, the accuracy of an assertion is traditionally seen as deriving from the truth of the asserted proposition, or the truth of this proposition at some relevant context(s). That is, an assertion is supposed to be accurate in virtue of the asserted proposition being true or true at some relevant context(s). MacFarlane’s relativist framework vindicates this way of conceiving of the relation between truth and accuracy by making the assessment sensitivity of accuracy a result of the assessment sensitivity of truth.
MacFarlane (2005, 2007, 2011, 2014) makes use of two notions of propositional truth: a monadic assessment sensitive one allegedly expressed in English by ‘true,’ and a triadic one expressed by a metalinguistic predicate from the relativist theory.8 The first notion is monadic insofar as its extension is a set of propositions and not a set of ordered n-tuples, whereas it is assessment sensitive because it has an extension relative to parametric values in the circumstances of evaluation (e.g. a standard of taste, a standard of knowledge or an information state) that are determined by the context of the assessor of an assertion or belief (i.e. the context of assessment) instead of by the context of the assertion or belief (i.e. the context of use). The triadic truth notion, in turn, relates propositions to contexts of use and contexts of assessment and is meant to account for the assessment-sensitivity (and use-sensitivity) of the monadic notion.9 On this framework, the accuracy of assertions, acceptances and rejections of assessment-sensitive propositions is itself assessment-sensitive. Insofar as we need a context of assessment in addition to a context of use to assess for truth an assessment-sensitive proposition (i.e. a proposition whose truth-value can vary with the context of assessment), we also need it to assess for accuracy an assertion, acceptance or rejection of such a proposition.
The just described notion of accuracy is a monadic one that would be responsible for our disagreement and retraction intuitions, and so it should not be seen as a technical notion (despite the fact that the use of term ‘accuracy’ to express it may be a technical use). Nevertheless, its assessment-sensitivity is accounted for by means of a dyadic and theoretical notion of accuracy relating assertions, acceptances or rejections of propositions to contexts of assessment. And it is in terms of this latter notion that MacFarlane could make sense of the faultlessness of a range of disagreements. Returning to our working example of disagreement, MacFarlane takes ‘tasty’ as invariantly expressing a monadic assessment-sensitive property, i.e. a monadic property whose extension (the set of tasty things) can vary with the context of assessment. As a result, John’s and Ann’s respective sincere assertive utterances of (1) and (2) would express propositions that contradict each other (i.e. that cannot be jointly true at a circumstance) and are true or untrue depending on the context of assessment, and so the accuracy of these assertions would also depend on this context. Thus, John and Ann would be fault free in the sense that their assertions would be accurate relative to their own respective contexts of assessment, and they would genuinely disagree with each other insofar as their assertions could not be jointly accurate at a single context of assessment.
To be sure, the difference between MacFarlane’s proposal and other relativist views that are not formulated in terms of the notion of assessment-sensitivity may be thought to be non-substantial. Most relativists hold that the English predicate ‘true’ is a monadic predicate that is correctly applied to a proposition just in case this proposition is true relative to the relevant circumstance of evaluation. Thus, insofar as these views relativize the accuracy of assertions, this predicate could be seen as expressing a monadic assessment-sensitive truth notion. Be that as it may, Kölbel (2008c, 2009) has advanced an alternative view on the relation between propositional truth and accuracy that makes no room for assessment-sensitivity by making no room for the assessment-sensitivity of accuracy, while at the same time claiming that the correctness (in a truthrelated sense) of assertions about personal taste matters is relative to some non-standard parametric value. In the next section, I argue that MacFarlane’s view, unlike Kölbel’s, allows the relativist to give an answer to an objection raised by Rosenkranz (2008) to truth relativism.
2. Rosenkranz’s objection and the relation between truth and accuracy
Rosenkranz’s (2008) objection questions the possibility of faultless disagreement, assuming that this possibility goes hand in hand with the intelligibility of truth relativism. The objection is based on the idea that in asserting a proposition one thereby presents it as true or as true relative to something,10 and on a simple rationale for deriving accuracy from propositional truth that follows from this idea.
Before explaining this rationale and how Rosenkranz’s objection is based on it, it is worth pointing out two things. First, we are not defining assertion as a presentation of a proposition as true or as true relative to something, but just taking assertion to necessarily involve such a presentation. Second, this idea is hardly rejected by someone who accepts the legitimacy of an ordinary propositional truth notion. As MacFarlane (2014) observes, it is a standing convention that one aims to assert truths (i.e. true propositions) so that “in making assertions one represents oneself as aiming to put forward truths” (MacFarlane, 2014, p. 101). And this way of representing oneself, in turn, would be inseparable from representing or presenting the asserted proposition as true.
Assuming that propositions are neutral at most with respect to the world, this feature of assertion gives us a straightforward explanation of why if an asserted proposition is true the assertion is accurate (i.e. correct in a truth-related sense), whereas if this proposition is untrue the assertion is inaccurate. We derive the accuracy of an (actual, not merely possible) assertion of a proposition p from p being true because the assertion presents this proposition as true.11 That is, the accuracy of the assertion consists in its presenting the proposition as it actually is. Accordingly, in case the asserted proposition were not true, the assertion would be inaccurate, as long as the proposition would not be as it is presented by the assertion.
According to Rosenkranz (2008), in a framework that countenances propositions that are neutral with respect to perspectives or whatever other non-standard parametric values (e.g. taste standards), we need to assume that in asserting such a proposition a speaker thereby presents it as true relative to a particular such value (e.g. a perspective or taste standard). Assuming a framework that relativizes truth to perspectives, a speaker s who asserts a perspective-neutral proposition p can be taken to present p as true relative to s’s perspective. Rosenkranz’s objection, then, simply runs as follows. Insofar as an assertion is accurate if and only if it presents a proposition as being as it actually is, we would have to deem the above-mentioned assertion accurate just in case p is true relative to s’ perspective, otherwise we would have to judge it as inaccurate.12 Hence, there is only one perspective (or set thereof) that is always relevant to evaluate an assertion of a perspective-neutral proposition for accuracy, namely the one relative to which the assertion presents the proposition as true (in the just considered case, s’s perspective). And this means that there is no room for relativizing accuracy. With regard to the purported cases of faultless disagreement, the outcome is that they turn out to be non-genuine disagreements (in case each party respectively presented a proposition as true relative to a different perspective), non-faultless (if at least one of the parties asserted a proposition that is not true relative to the perspective she presents the proposition as true) or both. To be sure, this objection can also be stated if we assume that assertions of p present p as true relative to a context or contexts, or as true relative to other parametric values than perspectives.
Rosenkranz’s (2008) objection assumes that the relativist needs to take an asserter to present a proposition as having a relational assessment insensitive truth property (e.g. true relative to s’s perspective). But the relativist need not do this. The relativist can avoid the just stated difficulty in relativizing accuracy by claiming that the truth property an asserter presents a proposition as having, is assessment-sensitive.
Insofar as the relativist takes propositions to be use-sensitive, at most, with respect to the world, he can claim that an assertion presents a proposition simply as true, while conceiving of ordinary monadic truth (i.e. the notion commonly expressed by the English predicate ‘true’) as assessment-sensitive (i.e. a notion or property whose extension is relative to the context of assessment).13 Monadic accuracy would be derived from monadic truth using a simple pattern of inference (i.e. an assertion is accurate if and only if the asserted proposition is true), and would come out as assessment-sensitive just as ordinary truth is. To be sure, non-monadic truth and accuracy notions from the relativist theory would account for the assessment-sensitivity of the monadic notions of truth and accuracy.14 According to this line of reasoning, it makes sense to treat accuracy as relative to a parameter in the circumstance of evaluation as long as this treatment amounts to accept a dyadic accuracy notion -i.e. a notion relating assertions/beliefs to contexts of assessment- in terms of which we account for the assessment sensitivity of an ordinary notion of accuracy. Having said this, we must note that the just presented answer is not unproblematic since, as we shall explain, the relativist still needs to make sense of the notion of truth relative to a context of assessment that is meant to account for the assessment sensitivity of truth and leads to the relativization of accuracy to contexts of assessment.
I would like to address now a possible complaint. It could be claimed that our answer to Rosenkranz’s objection, as well as the objection itself, depends upon conceiving of accuracy as derived from or grounded on truth, but Max Kölbel (2008c) has argued for a different way of understanding the relation between truth and accuracy that may allow us to face Rosenkranz’s objection without relying on the notion of assessment sensitivity. I will next contend that Kölbel’s understanding of this relation does not constitute a satisfactory alternative for a relativist.
Kölbel (2008c) argues that there is no conceptual link between ordinary monadic truth and his technical truth notion relating propositions to worlds and standards of taste. But this, according to him, does not mean that his view is devoid of practical and philosophical significance. On his view, there is an independent notion of belief correctness (and, derivatively, assertion correctness) that indirectly confers significance on his non-monadic truth notion. This notion of correctness has to do with our competence in acquiring beliefs and applying concepts. Being competent with a concept requires applying it only under certain conditions and in the case of certain concepts, like the concept of tastiness, these conditions would be sensitive to some features of the believer. Accordingly, the same object could be correctly judged as tasty by one believer and correctly judged not to be tasty by another believer, despite both having access to the same evidence. Other concepts, which Kölbel considers to be objective as opposed to non-objective ones, would not allow for any variation in their correct application: if one believer applies the concept to an object and another thinks the concept is not applicable to the object, at least one of them is making a mistake. This notion of correctness could be extended to assertions as follows:
An assertion (or utterance of an assertoric sentence) is correct to the extent to which it is (or would be) correct for the utterer to believe the proposition asserted. (Kölbel, 2008c, p. 250)
Kölbel (2008c) notices that correctness so understood coincides with ordinary monadic truth in the objective range but not in the non-objective range. For instance, if I judge that someone correctly asserts that the sun is a star, I must - on pain of incoherence - judge that the asserted proposition is true, but I can coherently judge that someone correctly asserts that mutton is tasty (because it is correct for the utterer to believe the asserted non-objective proposition) despite the asserted proposition not being true (because it is incorrect for me to believe or assert this proposition and correct for me to believe or assert its negation, and the ordinary truth predicate is subject to the schema ‘p is true iff p’).15 According to this, we cannot identify correctness with ordinary truth.
However, given that in the objective range the correctness of an assertion coincides with the ordinary truth of the asserted proposition, Kölbel (2008c) thinks that we can, for semantic purposes, link this notion of correctness to a technical truth notion relating sentences to contexts, worlds and taste standards by means of the following principle (where ‘TrueS’ stands for Kölbel’s sentential truth relation and ‘@’ stands for the actual world):
An utterance of a sentence s in a context c is correct iff TrueS (s, c, <@, the standard determined by c>). (Kölbel, 2008c, p. 250)
In turn, the principle linking correctness to propositional truth would run as follows (where ‘TrueP’ stands for the propositional truth relation):
In other words, Kölbel claims that we can, for strictly semantic purposes, allow ourselves to use technical truth predicates like ‘TrueS’ or ‘TrueP’ that are connected to correctness both in the objective and non-objective range via the principles just stated. For Kölbel, then, his non-monadic truth notions derive its significance from an independent pre-theoretical notion of correctness that is responsible for our faultless disagreement intuitions (for this reason, we could call it an accuracy notion).
Leaving aside the merits or demerits of this view for now, what is clear is that this notion of correctness is not relative but absolute. The standard relevant for assessing an assertion is the one the asserter is subject to, that is the one fixed by the context of use. Nevertheless, Kölbel (2008b, 2009) also claims that the correctness of assertive utterances of simple sentences containing predicates of personal taste is relative. What he has in mind may be made clear by his claim that there is not just one notion of correctness but lots of them:
There are many ways of evaluating utterances: when Anna says that whale meat is tasty, I can evaluate what she says against her own standard of taste, against my own standard, or against some other person’s. Thus, her utterance can be correct on her standard, my standard, someone else’s standard. Thus, [...] we should take into account that there is not just one notion of correctness for utterances of sentences expressing propositions concerning matters of taste, but there are many. (Kölbel, 2009, p. 387)
Each one of these notions of correctness is absolute: there is just one legitimate answer to the question of whether an assertion is correct relative to a particular standard. However, each of these notions can be seen as the result of fixing a particular standard for a dyadic correctness notion relating assertions and standards of taste.17 In particular, the monadic notion of correctness having to do with our competence in applying concepts can be seen as the result of introducing a descriptive condition -i.e. the standard of the speaker (believer)-on the standard relatum of this dyadic notion. The dyadic notion, in turn, can be taken as the result of generalizing the monadic (but structurally complex) notion of correctness according to the standard of the speaker (believer).
There is a problem with this account of relative accuracy that points to the need of positing a monadic assessment sensitive truth notion or property in order to relativize accuracy. This way of conceiving of relative accuracy prevents us from vindicating the appearances of genuine disagreement adduced in support of a relativist treatment of a domain of discourse. If B’s assertion is inaccurate relative to A’s perspective, A’s assertion is inaccurate relative to B’s perspective, and our non-monadic notion of accuracy does not account for the assessment sensitivity of a single monadic notion, we have no clear reason to see the previous assertions as being in conflict. A and B could - without retracting one or the other assertion - perfectly agree on how their assertions must be assessed in terms of these two monadic but structurally complex notions of accuracy. That is, insofar as these assessments are not incompatible, we have no clear reason to the see the assessed assertions as expressing a disagreement. But being able to take such apparent disagreements as genuine disagreements involving preclusion of joint accuracy is supposed to be the main practical difference that truth relativism has with non-indexical contextualism, and so the distinctive trait of truth relativism.18 We can hope to avoid this problem by accepting MacFarlane’s framework, which makes use of technical non-monadic truth and accuracy predicates to account for the alleged assessment sensitivity of our monadic notions of truth and accuracy.
In sum, truth relativism, as we understand it here, is a view that relativizes accuracy as a way to vindicate some purported intuitions of faultless disagreement. And according to what we said in this section, MacFarlane’s framework, unlike Kölbel’s, promises to relativize accuracy in a way that accounts for the alleged phenomenon described as faultless disagreement. As a matter of fact, Kölbel’s view is arguably a form of pluralism, not a form of relativism, about accuracy, since it countenances several monadic accuracy notions that, as long as they are different notions, do not result in incompatible assessments of assertions and beliefs.
In the next section, I show that the notion of assessment-sensitivity also allows the relativist to provide an answer to Evans’s (1985) objection to the relativization of accuracy. The task left to the relativist, as we explain in the fourth section, will be to make sense of the notion of truth relative to a context of assessment.
3. Evans’s objection
Before the contemporary debate over truth relativism took place, Evans (1985) identified and criticized a possible treatment of tense that amounts to a truth relativist proposal, since it countenances time-neutral (tensed) propositions and relativizes the accuracy (in Evans’s words, the correctness) of their assertions. According to this view, sentences like ‘Peter is smiling’ or ‘Peter is not smiling,’ which do not make explicit reference to a particular time, are normally used to assert time-neutral propositions that can have different truth-values at different times, and as a result the accuracy of such assertions is time-relative. The core of Evans’s objection to this proposal is briefly stated in the following passage, where ‘correct’ is meant to express the accuracy notion from a truth relativist theory about tense:
If a theory of reference permits a subject to deduce merely that a particular utterance is now correct, but later incorrect, it cannot assist the subject in deciding what to say, nor in interpreting the remarks made by others. What would he aim at, or take the others to be aiming at? Maximum correctness? But of course, if he knew the answer to this question, it would necessarily generate a once-and-for-all assessment of utterances, according to whether or not they meet whatever condition the answer gave (Evans, 1985, pp. 349-350).
The objection can be analyzed in two related parts. First, Evans contends that the proposal gives rise to an unacceptable view on communication, since according to it a subject could not know what to assert and how to interpret others’ assertive utterances if accuracy were relative. An assertive utterance of ‘Peter is smiling,’ for instance, could be accurate at the time when it is made (in case Peter is smiling at that time) but inaccurate as assessed from a later time (in case Peter is not smiling at this later time). Accordingly, the asserter would have no control over how her assertion should be assessed later, and so would have no reason to assert this tensed proposition instead of its negation. In other words, the theory would fail to provide a norm of assertion that specifies what asserters should aim at. Second, Evans claims that if the defender of the proposal attempted to specify such a norm, the specified norm would provide a basis for an absolute assessment of assertions as accurate or inaccurate. For instance, if the relativist said that the asserter should aim at truth and accuracy at the time of the context of assertion, this would provide the basis for absolutely assessing the assertion for accuracy taking into consideration the time of the context of the assertion. According to this line of argument, truth relativism is not compatible with taking accuracy as something asserters should aim at.
It is worth noting that a relativist view on tense may be particularly implausible insofar as it may strike us as obvious that it clashes with our linguistic practice. For instance, if John assertively utters ‘Peter is smiling’ in the morning and Ann assertively utters ‘Peter is not smiling’ at night (where ‘Peter’ is used by John and Ann to refer to the same person), it seems clear that they are not (faultlessly) disagreeing with each other in any significant sense. On the other hand, it has been argued that relativism about other expressions (e.g. personal taste, moral and aesthetic predicates, knowledge ascriptions or epistemic modals) is supported by strong linguistic evidence consisting in the apparent existence of a range of disagreements and retractions. Be that as it may, Evans’s objection purports to provide the basis for a general objection to any proposal that relativizes the accuracy of the assertions of a range of propositions. More precisely, the objection intends to show that any view that relativizes the accuracy of assertions fails to make sense of the practice of asserting.
Now, a relativist like MacFarlane has a simple answer to Evans’s objection at her disposal. She can face this objection by claiming that an asserter should aim at accuracy and that this is an assessment-sensitive notion or property. The assessment-sensitivity of this notion of accuracy, in turn, would be accounted for by means of a metalinguistic dyadic accuracy notion relating assertions to contexts of assessment. This answer does not give rise to an absolute assessment of assertions and respects the intuitive idea that the proposed truth and accuracy norms stating what asserters should aim at determines how the assertion is to be assessed for accuracy. It would happen that these norms involve assessmentsensitive notions or properties. Thus, the relativist should endorse the following norm stating the accuracy aim of assertion:
Assessment-sensitive accuracy: An asserter should aim at accuracy.
To be sure, as it happens with the relativist answer given in the last section to Rosenkranz’s objection, this answer to Evans’s objection is not unproblematic since, as we shall explain, the relativist still needs to make sense of the notion of truth relative to a context of assessment that is meant to account for the assessment-sensitivity of truth and leads to the relativization of accuracy to contexts of assessment.19
At this point, it may be objected that that Evans’s objection can be taken as showing that we cannot make sense of a notion like truth relative to a context of assessment by showing that we cannot make sense of our practice of asserting in terms of a relativist non-monadic notion of accuracy (recall that relativizing truth to contexts of assessment leads us to relativize accuracy to such contexts). I think that this observation has some force, but that nonetheless the previous answer to Evans’s objection still shows that this objection, as it stands, is unsuccessful for two reasons. First, Evans simply does not consider a relativist framework that treads on the notion of assessment-sensitivity and could make room for a norm such as assessment-sensitive accuracy. He considers a relativist framework where the non-monadic truth and accuracy notions are not taken to respectively account for the assessment-sensitivity of monadic truth and accuracy. And second, as we shall see in the next section, it has been a matter of debate whether we should make sense of a non-monadic truth notion, i.e., a notion like the one MacFarlane uses to account for the notion of assessment-sensitivity, by means of an account of our assertion practices or by means of a definition in terms of our ordinary monadic truth notion. Thus, it is not uncontroversial whether we should read Evans’s objection as putting into question, besides the possibility of truth relativism to explain linguistic communication, the possibility of making sense of a non-monadic truth notion like MacFarlane’s (i.e. of showing that this notion is a truth notion). According to all this, if we wanted to use Evans’s objection to question the possibility of making sense of assessment sensitivity, we would have to complement it with other considerations that would turn it into a different, more complex objection.
Now, Ramiro Caso (2014) offers an answer to Evans’s objection different from the one we gave. As we shall see, we have reason to reject this alternative answer. Caso claims that the relativist can simply say that an assertion is governed by the norm that the asserter should aim at truth relative to her perspective or context of assessment. The resulting norm could run as follows:
Caso’s principle 1: An asserter ought to assert only propositions that are true as assessed from the world and perspective (or, alternatively, from the context) she occupies.20
It could be claimed that this answer determines an absolute assessment of assertions, and so yields an accuracy norm that runs exactly like the norm Assessment sensitive accuracy but deploys an absolute (non-relative) notion of accuracy. More precisely, it seems that according to Caso’s principle 1, in case the asserted proposition is true as assessed from the world and perspective of the asserter, the assertion is absolutely accurate. More precisely, it seems that according to Caso’s principle 1, in case the asserted proposition is true as assessed from the world and perspective of the asserter, the assertion is absolutely accurate, and so that the word ‘ought’ occurring in this principle deserves an absolute reading According to this, Caso’s principle 1 would lead to the following principle endorsed by Kölbel, which, as we saw, is understood by this author as deploying an absolute monadic notion of accuracy:
An assertive utterance of a proposition p made in c is correct iff TrueP (p, <@, the standard determined by c>).
In spite of all this, Caso (2014) -based on some distinctions first introduced by MacFarlane (2005, 2014)- contends that a principle like Caso’s principle 1 provides a satisfactory relativist answer to Evans’s objection, that is an answer that is genuinely relativist and does not prevent us from vindicating our alleged faultless disagreement intuitions.
Caso argues that the relativist should claim that the practice of assertion has two sub-practices: the practice of making assertions and the practice of assessing them. These sub-practices, in Caso’s opinion, are guided by different principles, i.e. norms that relativize truth to different contexts. Caso’s principle 1would be the norm that guides the making of assertions but not their assessments, which would be guided by the following principle (where wc1 is the world of context c1 and Pc2 is the perspective relevant at context c2):
Caso’s principle 2: An assertion of proposition pc1 is accurate as assessed from context c2 iff p is true at < wc1, Pc2>.
The resulting picture, in Caso’s words, is the following:
In making assertions, the asserter aims to correctness according to her own circumstance. In interpreting the utterances of others, the interpreter presupposes compliance with the norm of assertion (or, at any rate, the intention of complying with it). However, in assessing the assertions of others (or hers at a later time), an assessor does so according to her own circumstance, not the asserter’s (or hers at the time of utterance) (Caso, 2014, pp. 1315).
According to Caso (2014, pp. 1318-1321), the norm stating what asserters should aim at (i.e. Caso’s principle 1) does not determine how we ought to assess the assertion for accuracy. In spite of this, he claims that the accuracy notion (in Caso’s words, the correctness notion) implicit in Caso’s principle 1 is not ultimately different from the one used in Caso’s principle 2. Rather, Caso’s principle 1 would be insufficient to determine a particular notion of accuracy, since it would only concern the production of assertions. But Caso’s principle 1 together with Caso’s principle 2 would determine a relative notion of accuracy. The impression that Caso’s principle 1 gives rise to an absolute accuracy notion would stem from the fact that there is a privilege perspective for the purpose of making an assertion.21 At this point, it is worth stressing that Caso (2014, pp. 1318-1321) contends that there is a single relativist notion of accuracy at stake in his answer to Evans’s objection (this notion would be a dyadic one relating assertions/beliefs to contexts of assessment), and so that he intends his first principle (i.e. Caso’s principle 1) to be different Kölbel’s seemingly equivalent principle. As I explain below, I believe that Caso’s answer does not succeed in doing this.
I have the following objection to this line of response to Evans’s objection. I believe that, unless we read Caso’s principle 1 and Caso’s principle 2 as metalinguistic principles that jointly account for the assessment sensitive norm Assessment sensitive accuracy, there cannot be one single notion of accuracy in play in Caso’s proposal. After all, taken at face value, the word ‘ought’ in Caso’s principle 1 does presuppose a monadic notion of accuracy in terms of which assertions can be absolutely assessed, whereas Caso’s principle 2 does deploy a dyadic notion of accuracy relating assertions to contexts. Thus, if we do not take them as two principles jointly accounting for the norm Assessment sensitive accuracy, a view that endorses both of them countenances two unrelated notions of accuracy. Whereas the first principle would presuppose a non-relative notion of accuracy, the second one would deploy a dyadic one that - as we saw in the previous section - the relativist has reason to reject.22 Accordingly, the relativist should accept a norm that simply states that in asserting one ought to aim at accuracy and take accuracy as an assessment-sensitive property or notion. Non-monadic truth and accuracy notions from the relativist theory would, in turn, account for this assessment-sensitivity. To be sure, as we suggested, we can read the principles presented by Caso as metalinguistic principles that account for the assessment-sensitive and object language principle assessment-sensitive accuracy. But that move presupposes that the proper answer to Evans’s objection consists in claiming that speakers ought to aim at accuracy, and that this is an assessment-sensitive notion.
4. The task left to the relativist
I have argued that two significant objections that purport to show that truth relativism cannot make sense of the practice of asserting could be answered by appealing to the notion of assessment-sensitivity. But it is contentious whether we understand our talk of assessment-sensitive truth and, relatedly, whether our ordinary truth notion is assessment-sensitive. As MacFarlane (2014, p. 97) observes, in order to see truth as assessment-sensitive we need to make sure that we understand what is meant by the metalinguistic predicate ‘true as used from c1 and assessed from c2’ that would account for the assessment-sensitivity of the ordinary truth notion.23 But, as MacFarlane (2005, 2014) points out, it is not clear that the notion of truth admits of a relativization to assessors. MacFarlane, following Jack Meiland (1977), states the question as a dilemma:
If ‘true’ as it occurs in ‘true for X’ is just the ordinary, nonrelative truth predicate, then it is unclear what ‘for X’ adds.24 On the other hand, if the occurrence of ‘true’ in ‘true for X’ is like the ‘cat’ in ‘cattle’ -an orthographic, not a semantic, part- then the relativist needs to explain what ‘true-for-X’ means and what it has to do with truth, as ordinarily conceived (MacFarlane, 2005, p. 312; 2014, p. 97).
Hence, the proposed relativist answer to the objections we considered faces the challenge to make sense of the non-monadic truth notion relating propositions to contexts of assessment (i.e. to show that this notion is a truth notion).
MacFarlane (2005, 2014) proposes to illuminate this notion by showing how it is connected with our practice of making assertions. More precisely, he tries to make sense of this notion by showing how it can be brought to bear in a truth norm account (MacFarlane, 2014) and in a truth commitment account of assertion (MacFarlane, 2005) that are supported by evidence from our linguistic practice. In turn, other authors (Cappelen and Hawthorne, 2009, 2011; Montminy, 2009; Soames, 2011) contend that any theory that introduces a technical non-monadic truth notion has to make sense of it by defining or characterizing in terms of ordinary monadic truth (a notion, it can be assumed, we already grasp to some extent). In particular, MacFarlane’s relativism would have to make sense of the notion of truth relative to a context of assessment in such a way, which in turn would allow the view to make sense of (ordinary) truth as assessment-sensitive.25
How the task of making sense of this relativist non-monadic truth notion should be carried out and whether this task can be successfully accomplished are issues that I do not address in this essay. Thus, insofar as the proposed answers to the objections leveled by Rosenkranz and Evans depend on whether the relativist can actually make sense of assessment-sensitivity, this paper leaves open the question of whether these answers can ultimately be made to work. Be that as it may, I hope to have shown that, as they stand, the just mentioned objections fail, as long as they do not take into account the notion of assessment-sensitivity and so the question of whether we can make sense of it. Moreover, I also attempted to show that this is the only genuinely relativist line of answer to these objections, since the alternative answers lead to forms of pluralism about monadic accuracy.
5. A final summary
I argued that the notion of assessment-sensitivity allows the relativist to give a simple answer to Rosenkranz’s (2008) and Evans’s (1985) objections, both of which purport to show that we cannot make sense of the accuracy of assertions being relative. In the first section, I introduced the notion of assessment-sensitivity by briefly explaining MacFarlane’s framework. In the second section, I argued that this notion allows the relativist to face Rosenkranz’s objection by keeping the Fregean pattern for deriving accuracy from truth without compromising the relativity of accuracy. I also showed in this section that Kölbel’s (2008c) non-Fregean understanding of the relation between truth and correctness (in a truth-related sense) does not provide a way out to the difficulty posed by Rosenkranz, since it prevents truth relativism for vindicating our purported faultless disagreement intuitions by giving rise to a pluralist view about monadic accuracy. In the third section, I argued that the relativist could face Evans’s objection by claiming that an asserter should aim at accuracy understood as an assessment-sensitive notion. In this section, I also criticized Caso’s (2014) alternative reply to this objection by arguing that Caso’s two principles should be seen as metalinguistic principles jointly accounting for the assessmentsensitive principle assessment-sensitive accuracy, and this presupposes that the proper answer to Evans’s objection is the one we gave in terms of this latter norm. Finally, in the fourth section, I showed that the relativist still needs to make sense of the notion of truth relative to a context of assessment (i.e. to show that it is a truth notion) and took notice of two possible strategies to do this. Insofar as the proposed answers to the above-mentioned objections depend on whether the relativist can accomplish this last task, I left open the question of whether these answers can ultimately be made to work. Be that as it may, I hope to have shown that, as they stand, the objections fail, as long as they do not take into account the notion of assessment-sensitivity. I also sought to show that an answer to these objections that do not tread on the notion of assessmentsensitivity gives rise to a pluralist view about monadic accuracy, which cannot possibly vindicate the purported phenomenon of faultless disagreement. In sum, then, I pretend to have shown that the notion of assessment-sensitivity, and in particular the question of whether we can make sense of it, should be at the center of the debate over truth relativism.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Reference
- CAPPELEN, H., HAWTHORNE, J. “Relativism and Monadic Truth”. 1st ed. Oxford: Oxford University Press, 2009.
- CAPPELEN, H., HAWTHORNE, J. “Reply to Lasersohn, MacFarlane and Richard”. Philosophical studies, Vol. 156, Nr. 3, pp. 449-466, 2011.
- CASO, R. “Assertion and relative truth”. Synthese, Vol. 191, Nr. 6, pp. 1309-1325, 2014.
- EVANS, G. “Does tense logic rests upon a mistake?”. (pp. 343-363). In: G. EVANS (ed.). Collected Papers Oxford: Oxford University Press, 1985.
- FREGE, G. “On sense and reference”. (pp. 56-78). In: P. GEACH, M. BLACK (eds. and trans.), Translations from the Philosophical Writings of Gottlob Frege Oxford: Blackwell, 1980.
- IACONA, A. “Faultless or disagreement.” (pp. 287-296). In: M. GARCÍA-CARPINTERO and M. KÖLBEL (eds.). Relative Truth Oxford: Oxford University Press , 2008.
- KÖLBEL, M. “Faultless disagreement”. Proceedings of the Aristotelian Society, Vol. 104, pp. 53-73, 2003.
- KÖLBEL, M. “‘True’ as ambiguous”. Philosophy and Phenomenological Research, Vol. 77, Nr. 2, pp. 359-384, 2008a.
- KÖLBEL, M. “The evidence for relativism”. Synthese, Vol. 166, pp. 375-395, 2009.
- KÖLBEL, M. “Introduction: motivations for relativism”. (pp. 1-38). In: M. GARCIA-CARPINTERO, M. KÖLBEL (eds.). Relative Truth . Oxford: Oxford University Press, 2008b.
- KÖLBEL, M. “Truth in semantics”. Midwest studies in philosophy, Vol. 32, pp. 242257, 2008c.
- KÖLBEL, M. “Should we be pluralists about truth?”. (pp. 278-297). In: N. J. L. L. PEDERSEN, C. D. WRIGHT (eds.). Truth and Pluralism: Current Debates New York: Oxford University Press, 2013.
- LASERSOHN, P. “Context dependence, disagreement and predicates of personal taste”. Linguistics and Philosophy, Vol. 28, pp. 643-686, 2005.
- MACFARLANE, J. “Making sense of relative truth”. Proceedings of the Aristotelian Society, Vol. 105, pp. 321-339, 2005.
- MACFARLANE, J. “Relativism and disagreement”. Philosophical studies, Vol. 132, pp. 17-31, 2007.
- MACFARLANE, J. “Truth in the garden of the forking paths”. (pp. 81-102). In: M. GARCIA-CARPINTERO, M. KÖLBEL (eds.). Relative Truth . Oxford: Oxford University Press , 2008.
- MACFARLANE, J. “Nonindexical contextualism”. Synthese, Vol. 166, Nr. 2, pp. 231-250, 2009.
- MACFARLANE, J. “Simplicity made difficult”. Philosophical studies , Vol. 156, Nr. 3, pp. 441-448, 2011.
-
MACFARLANE, J. “Assessment Sensitivity: Relative Truth and its Applications”. 1st ed. Oxford: Clarendon Press, 2014. https://johnmacfarlane.net/reltruth.pdf
» https://johnmacfarlane.net/reltruth.pdf - MEILAND, J. “Concepts of relative truth”. The Monist, Vol. 60, pp. 568-582, 1977.
- MONTMINY, M. “Contextualism, relativism and ordinary speakers’ judgments”. Philosophical studies , Vol. 143, pp. 341-356, 2009.
- RECANATI, F. “Perspectival Thought: A Plea for (Moderate) Relativism”. Oxford: Oxford University Press , 2007.
- RICHARD, M. “When Truth Gives Out”. Oxford: Oxford University Press, 2008.
- ROSENKRANZ, S. “Frege, relativism and faultless disagreement”. In: M. GARCIA-CARPINTERO, M. KÖLBEL (eds.). Relative Truth . Oxford: Oxford University Press , 2008. pp. 225-237.
- SMITH, B. C. (2010) “Relativism, disagreement and predicates of personal taste.” (pp. 195-225). In: F. RECANATI, I. STOJANOVIC, N. VILLANUEVA (eds.). Context-Dependence, Perspective and Relativity Berlin/New York: De Gruyter Mouton, 2010.
- SOAMES, S. “True at”. Analysis, Vol. 71, Nr. 1, pp. 124-133, 2011.
