Dialetheism in action: a new strategy for solving the equal validity paradox?

This paper starts from the Equal Validity Paradox, a paradoxical argument connected to the so-called phenomenon of faultless disagreement. It is argued that there are at least six strategies for solving the paradox. After presenting the ﬁrst ﬁve strategies and their main problems, the paper focuses on the sixth strategy which rejects the assumption that every proposition cannot be both true a false. Dialetheism is the natural candidate for developing strategy six. After presenting strategy six in detail, we formulate a normative problem for the dialetheist and oﬀer a tentative solution to it. We then elaborate further considerations connecting strategy six to pluralism about truth and logic. Even if strategy six is a hard path to take, its scrutiny highlights some important points on truth, logic and the norms for acceptance and rejection. 1

in the same world, the proposition that ginger is tasty and the proposition that ginger is not tasty are both true. By LNC these two latter propositions cannot both be true in the same world, hence, by reductio, we conclude that, despite the equal validity intuition, yes-no utterances cannot both be true. Thus yes-no utterances must differ in truth-value. Conclusion: the equal validity intuition cannot find a coherent formulation in terms of equality of truth of the utterances involved in a dispute of inclination.

Six Ways Out of the Paradox
A way to object to the paradox is to criticise the reasoning on the basis that some of inferences are invalid. We will set aside this solution since we do not think that objecting to modus ponens or conjunction introduction in relation to the inferences involved in the paradox is an helpful strategy for explaining of the nature of the paradox.
Given the structure of the paradox there are at least six solutions to the Equal Validity Paradox: (i) to deny Equal Validity; (ii) to deny Alethic Equal Validity; (iii) to deny Propositionality; (iv) to deny Contradiction; (v) to deny Truth Equivalences; and, finally, (vi) to deny LNC. In this section we will sketch these strategies.

Strategy i: Revisionary Vs descriptive approaches
Strategy (i) -i.e. the denial Equal Validity-consists in a revisionary approach to the problem. 8 According to the revisionary approach disputes of inclination depend on a misguided view on the subject matter: when confronted with disputes of taste, for example, we wrongly think that both views are legitimate, whereas there is no sense of entitlement for these views capable of making sense of both views as rationally responsible attitudes -the only rational attitude is that of abstaining from believing and disputing. Such a radical stance on the problem denies the phenomenon we wish to account for -i.e. the intuition that subjective discourse has its own specificity. Of course, such a radical departure from our practice needs an articulated motivation, for the challenge for any revisionary approach is the possibility of elaborating a theory that offers a descriptive account of the phenomenon while at the same time solving the Equal Validity Paradox. Strategies (ii)-(vi) can then be seen as attempts to solve the paradox and to offer at the same time a descriptive account of faultless Semantic strategies include at least three strategies -i.e. rejection of either theses 2, or 3 or 4-which are well known in the literature. Yet for each of these routes there are well-founded doubts that they can actually avoid a revisionary approach to faultless disagreement. In this section we will sketch these strategies and will rehearse some worries on the feasibility of these options. 10 3. determine what the correct answer is. 12 More importantly: if truth is so remotely connected to our practices of subjective judgements, it is difficult to describe these practices as rationally sustainable when it comes to the activity of disputing.
when the relativisation of propositional truth is in place: it becomes unclear, how according to non-indexical contextualism, two utterances, which are true at different circumstances, involve any incompatibility of their respective assertions; whereas truth-relativism does not seem to have a non-metalinguistic way to express the idea of faultlessness since from each perspective each subject is right is criticising the opponent's view. 13 Another implementation of strategy (ii) is to argue that disputes of inclination operate in areas of discourse where there is no fact of the matter as to whether an utterance or a proposition is true. Call this strategy indeterminism.
The main gist of indeterminism is that there is an "absence of mistake-makers " when yes-no utterances are involved (Beall 2006, 64), and this absence is due to the absence of any fact of the matter that would otherwise determine one of the utterances as incorrect. A natural way to implement indeterminism is to argue that yes-no utterances have a third semantic status different -and incompatible-with truth. A well known way to model this option is to adopt a strong Kleene semantics and hold that yes-no utterances have a gappy semantic status as they are neither true nor false.
The problem with indeterminism lies in the tension between the thesis of semantic under-determination as a third semantic status other than truth and falsity and the thesis of equal validity.
If we adopt the rather weak truth-norm: Norm it is permitted to assert p only if p is true yes-no utterances (and their respective beliefs) would count as incorrect. 14 Could we reasonably modify the norm for allowing that a gappy proposition is assertable? Norm* it is permitted to assert p only if p is true or gappy.
Being a gappy proposition would thus par with truth in being a semantic value compatible with the correctness of an assertion and a belief.
However Norm* cannot capture the normative role of a gappy proposition.

Expressivism, Second-order indeterminacy
Whereas the latter three positions agree on the idea that in disputes of inclination the discourse involves genuine propositions and hence it is truth-apt, strategy (iii) rejects the idea that yes-no utterances express genuine propositions; one way to make sense of this option is to go expressivist. 15 Expressivism has been historically the preferred route for escaping the problem: given that no genuine proposition is involved, there is no point in making sense of the truth-conditions of yes-no utterances. However, well known problems afflict the expressivist route, first and foremost the so-called Frege-Geach problem. 16 Whether or not this daunting problem is genuine is beyond the scope of this paper. Suffice here to say that Schroeder (2008) has thoroughly elaborated the expressivist position concluding that the Frege-Geach problem is answerable only by incurring in heavy costs that compromise compositional semantics.
Another way to implement strategy (iii) is to hold that a yes-no utterance expresses one proposition but that it is indeterminate which one it actually expresses. Norm and this indeterminacy thesis jointly entail that it is indeterminate whether yes-no utterances are permitted. Note that this indeterminacy strategy is different from the gappy indeterminism mentioned in relation to strategy (ii). Whereas gappy indeterminism claims an indeterminacy regarding the truth-value of the proposition expressed (first-order indeterminacy), this indeterminacy thesis concerns which proposition is expressed by a yes-no utterance. Borrowing the terminology of Matti Eklund 17 , we can call this indeterminacy second-level indeterminacy and the indeterminacy of gappy indeterminism first-order indeterminacy.
The basic worry for this second-order indeterminacy strategy is that the indeterminacy in the content expressed casts doubt on the existence of a genuine disagreement: if Mary's utterance does not determinately express that ginger is tasty, then Jane's utterance does not determinately contradict what Mary has said. A further worry is related to the appropriate stance related to secondorder indeterminacy. If Mary fails to express a proposition determinately, it seems that the appropriate stance to have in this case is to withhold belief from any of the propositions that her utterance indeterminately expresses. Agnosticism seems thus the mandated attitude in disputes of inclination, and thus the proposal seems to fall back onto a form of revisionism.

Strategy (iv): the denial of Incompatibility -Indexical Contextualism
As to the strategy (iv) -the denial of Incompatibility-indexical contextualism is the case in point: indexical contextualists deny that the same proposition is involved in disputes of inclination; according to them the semantic content of evaluative expressions such as "tasty" is sensitive to the context of use. Whether or not the contextualist semantic story is credible, the most pressing problem for this strategy is the so-called lost disagreement problem: the difficulty of making sense of the idea of disagreement if yes-no utterances do not express incompatible contents. 18 None of the previous considerations is of course intended to provide a knockdown argument against these proposals, nor it is the aim of this work to do so.
What matters here is rather to show that there is a well-grounded suspicion that these solutions to the Equal Validity Paradox cannot but end up in being revisionary with respect to the phenomenon of faultless disagreement.

Strategy (v): the denial of Utterance-Proposition
Truth Schema According tho this strategy a yes-no utterance expresses a proposition but the truth of the utterance does not entail the truth of the proposition expressed: <s,c> expresses <p> and <s,c> is true but <p> is not true in circ(<s,c>) Utterance-Proposition Truth Schema would thus fail. It is hard to find a principled motivation for this strategy, nor we do know of any proposal in the literature implementing this strategy. Few remarks will suffice for the purposes of this work. Given that yes-no utterances are true they are correct as speech acts. On the other hand, given that the propositions expressed by these utterances are not true in the relevant circumstances, believing these propositions would be incorrect. 19 It would follow that the correctness conditions belief and assertion potentially diverge. When Mary utters "Ginger is tasty" her linguistic move is correct, while her internal representation of the taste of ginger is mistaken. One way to make sense of this option is to hold a pragmatic story explaining why assertions about taste such as Mary's are correct, while at the same holding an error theory about taste-properties: there is no genuine taste property, but only physiological properties that play a different role from the evaluative role that in our ordinary talk we attribute to taste-vocabulary.
The basic problem with this way of implementing strategy (v) is that it is clearly committed to a revisionary stance: we are systematically mistaken in having beliefs about what is tasty. 20 19 Notice that in order to avoid violation of LNC, it should claimed that either a) one proposition is true and the other is false; or b) the propositions expressed are not contradictory; or c) that <p> is untrue (but not false) and that <not-p> is false. As for option a), it seems tantamount to realism, so option a) collapses into strategy (ii). Analogously, option b) seems just indexical contextualism, so we have a collapse into strategy (iv). Finally, option c) requires to adopt some version of threevalued semantics with a value (call it "untruth only") that is excluded by truth of a negation. This latter option seems the most promising option for avoiding the collapse into other strategies. 20 Assuming that propositions are objects of belief.

Strategy vi: the denial of LNC -Dialetheism
The last option is to deny LNC. This can be done by claiming that yes-no utterances express propositions that are both true and false. From a logical point of view, being true and false is here treated as a the third semantic status which is designated. This third status is the dual status of a truth-value gap: a truth-value glut. The most popular logic for modeling this view is LP -the Logic of Paradox-which is exactly like strong Kleene semantics except that the third value is designated. 21 This brings us to see option vi a more promising solution than the gappy theory: dialetheism.
According to dialetheism, LNC fails 22 since a yes-no utterance expresses a proposition that is a dialetheia -i.e. a proposition that is both true and false.
This latter solution seems more promising: since a dialetheia is both true and false, it is also false but its falsity does not imply its untruth and thus Norm* does not evaluate yes-no utterances as incorrect.
Whether or not a contradiction is acceptable is, for a dialetheist, a question to decide case by case. According to Priest, there seems to be no general criterion for deciding whether or not a contradiction is rationally acceptable. 23 The 21 The idea of using dialetheism for accounting for the phenomenon of faultless disagreement is sketched in Beall (2006) where it is claimed that analetheism -which interprets gappy propositions as designated-is a better option. See Beall and Ripley (2004) for a presentation of analetheism. Beall (2006) prefers this strategy to dialetheism in relation to taste on the grounds that analetheism makes sense of the idea that there is no fact of the matter in matters of sense (taste propositions are neither true nor false). We agree with Armour-Garb and Priest (2005) in thinking that analetheism does not do justice to the truth-norm for the same problems we have raised before in relation to the indeterminist gappy theory (see supra 3.1). Insofar as truth is the norm for assertion, we do not think that this model of indeterminacy is sufficient for accounting for the rational sustainability of a dispute. We will thus focus on dialetheism. 22 Recall that we are using a semantic reading of LNC. In LP the formula ¬(p ∧ ¬p) is never false only since, when p is dialetheia, the formula itself is a dialetheia. 23 "I am frequently asked for a criterion as to when contradictions are acceptable and when they are not. It would be nice if there were a substantial answer to this question-or even if one could give a partial answer, in the form of some algorithm to demonstrate that an area of discourse is contradiction phenomenon of equal validity could then be a reason to accept a contradiction as true.
This dialetheist solution to the Equal Validity Paradox allows one to say that in a dispute where Mary asserts that ginger is tasty and where Jane denies it, Mary's and Jane's opposite judgements are actually contradictory, so no hidden or extra parameters are invoked for relativizing yes-no utterances' truth (as opposed to contextualism or truth-relativism); moreover it allows one to maintain that Mary's and Jane's judgements are equally valid because of they are value gluts.
Note that this solution can be seen as very similar to a relativistic solution in the following sense. Though it is difficult to state precisely what a fully general relativistic doctrine is, we can try to gesture at least at the spirit that has motivated relativistic proposals. Relativism finds its inspiration in the basic and rough idea that there are areas of discourse where we cannot describe reality in absolute terms without falling into contradiction. This thought can be developed in two ways: i) in a metaphysical way by claiming that reality is somehow richer than what is described by any individual consistent description; ii) alternatively, it can be developed in terms of representation by claiming that our ways of describing reality are expressively incapable of providing an absolute and coherent description of it. 24 Now, historically this basic thought has been developed by introducing some mechanism of relativisation because the idea of embracing a contradiction was abhorred. 25 However, if this rough description of the basic relativistic thought is right, it is perfectly consistent with the relativistic spirit to go dialetheist by claiming that there is no consistent free. But I doubt that this is possible. Nor is this a matter for surprise. Few would now seriously suppose that one can give an algorithm-or any other informative criterion-to determine when it is rational to accept something. There is no reason why the fact that something has a certain syntactic form-be it p ∧ ¬p or anything else-should change this. One can determine the acceptability of any given contradiction, as of anything else, only on its individual merits." (Priest 1998b, 423 absolute description of reality because in some areas of discourse contradictions are true. In analogy with the metaphysical and representational version of the basic relativistic thought, dialetheist relativism can mirror these two versions by adopting, respectively, either metaphysical dialetheism (e.g. there are inconsistent facts in the subjective domain) or semantic dialetheism (e.g. natural languages are intrinsically inconsistent when it comes to areas of subjective discourse; see Mares (2004) for the distinction).
To sum up, dialetheism has a way of dealing with the contradiction involved in a dispute that gives rise to faultless disagreement. It also seems capable of avoiding the problems that contextualist proposals usually face -namely the disappearance of disagreement. At the same time it distances itself from truthrelativism in making sense of equal validity, while truth-relativists have no way to make sense of the idea that relatively to the same parameter, to which truth is relativized, both opinions are equally valid.
In the following sections strategy (vi) will be scrutinised in detail in order to assess its stability.

The Normative Problem for Strategy (vi)
Consider again our yes-utterances: Mary utters "Ginger is tasty" and Jane utters "Ginger is not tasty". Suppose that basic taste discourse exhibits faultless disagreement and that dialetheism should be applied to it. Incompatibility between the utterances would fail: both parties should admit that also their opponent is right, not just from her own point of view, but tout court. This consequence is itself a virtue of this strategy since it satisfies what Crispin Wright has dubbed Parity: [...]it is the requirement that faultlessness be appreciable, and endorsable, from the point of view not just of neutrals but of the committed parties in a dispute (Wright 2012, 439) In addition to Parity, Wright (2012) lists three other features of a dispute involving faultless disagreement. Two of these features are mentioned in the Equal Validity Paradox: Contradiction (equivalent to the principle Incompatibility) and Faultlessness (equivalent to Equal Validity). To these features Wright adds Sustainability: that is perfectly rational for the disputants to hold on disputing.
In the dialetheist setting Parity seems to jeopardise the sustainability of the dispute: if Mary thinks that Jane is not faulty in thinking that ginger is not tasty because Mary recognises that it is true that ginger is not tasty, and if Jane recognises that Mary is not faulty in thinking that ginger is tasty, then given they respective beliefs, they seem to have no real disagreement insofar as the latter requires a belief in the incompatibility concerning the truth of the contents believed. They should therefore cease to quarrel since they should both recognise that also the other party is right. The normative significance of holding contradictory contents evaporates since it does not involve any genuinely Assume the previous norm of assertion is in place: Norm it is permitted to assert p only if p is true The latter considerations can then be presented as a normative elaboration of a familiar problem for dialetheism in making sense of an exclusive notion of negation (Parsons 1990, Batens 1990, Berto 2008): ARGUMENT 1 : 1. Suppose p is a dialetheia and that A and B have evidence for this;

Suppose A asserts p;
3. Suppose B's typical of way of disputing with A is by asserting not-p; 4. B's assertion of not-p does not prevent her from also accepting p (since p is a dialetheia Norm is respected); 5. Incompatibility between A and B's assertions is lost; 6. The dispute between A and B on p is not rationally sustainable.
In a nutshell, the problem stems from the nature of the dialetheia: if a proposition is a dialetheia, the mutual correctness of accepting the proposition and its negation does not seem to leave space for any substantial disagreement: both accepting p and rejecting p -intended as accepting not-p 27 -are correct.
So opposite attitudes over p do not seem to motivate any rational dispute.
A possible way out of this problem is to distinguish four notions in play here (Priest 2005, 96-99) (Priest 2006, 103): Acceptance of p = mental state of believing p; Assertion of p = the speech act of expressing the act of accepting p; Rejection of p = the mental state of refusing to believe p; 28 Denial of p = the speech act of expressing the act of rejecting p.
Here is how Priest explains these distinctions: Someone who rejects A cannot simultaneously accept it any more than a person can simultaneously catch a bus and miss it, or win a game of chess and lose it. If a person is asked whether or not A, he can of course say 'Yes and no'. However this does not show that he both accepts and rejects A. It means that he accepts both A and its negation. Moreover a person can alternate between accepting and rejecting a claim. He can also be undecided as to which to do. But do both he can not. (Priest 1989, 618) 27 The equivalence between acceptance of not-P and rejection (or denial) of P is known as "denial equivalence" in Ripley (2015) and as "classical denial" in Murzi and Carrara (2014). Also gappy theorists reject this equivalence -see Parsons (1984). 28 We can interpret acceptance of p and rejection of p as "yes-or-no questions, respectively reading Given this distinction, Priest (2006, 110) states three different normative principles for rationality involving acceptance and rejection: (Accept-p) One ought rationally to accept p if there is good evidence for the truth of p.
(Accept-Not-p) One ought rationally to accept not-p if there is good evidence for the falsity of p.
(Reject-p) One ought rationally to reject p if there is good evidence for the untruth of p.
First, these principles are epistemic norms and not alethic norms -principles 29 See Grim (2004), Berto (2008). Berto (2008, 174-8) states some worries regarding the idea that acceptance and rejection are primitively exclusive mental states. Berto argues that it is better to explain this incompatibility between attitudes through a notion of content exclusion. We agree with Berto's point, however this issue will not be crucial for the following discussion. 30 In other words the exclusivity relation between acceptance and rejection involve that accepting about what is rational to believe on the basis of evidence for the truth of a proposition and not principles about what is permitted to believe in relation to the truth-value of a proposition. Second, note the Reject-p is not an exclusive norm for rejection since accepting p and rejecting p can be jointly correct attitudes.
An exclusive norm would be (Reject-p-Ex) One ought to reject p if p is untrue.
However, Murzi and Carrara (2014) have argued that an exclusive denial norm -plus the assumption that denial is expressible in the language-gives rise to a paradox once very minimal assumptions are in place. 31 Contrary to the norms for exclusive denial, Reject-p does not make assertion and rejection normatively incompatible. However, according to Priest, these acts cannot be performed at the same time and hence, though they are not normatively incompatible, they are nonetheless metaphysically incompatible: given any proposition p a subject cannot perform both the act of accepting p and the act of rejecting p. Murzi and Carrara object against the metaphysical impossibility (Murzi and Carrara 2014, PAGE).
In order to recapture the normative incompatibility between acceptance and rejection, we could modify the rejection rule as follows: (Reject-p*) One is permitted to reject P if there is good evidence for P s untruth, unless there is also good evidence for its truth. 32 According to Reject-P* one may deny P if one has good reasons for thinking that P is untrue only.
Carrara and Murzi criticise this norm on the basis of its asymmetry with the acceptance norm: Unlike assertion, any denial may later turn out to be incorrect, since any false sentence can in principle be discovered to be a glut. Thus, you can disagree with my assertion that 0 = 0, and thus deny 0 = 0. But, even if you can prove 0 = 0, and hence disprove 0 = 0, you can never be fully confident that your denial is correct: a proof of 0 = 0 may always turn up. By contrast, if you have proved 0 = 0 and thereby assert it, you can be fully confident that your assertion is correct.We find this asymmetry problematic: nothing in our practice of asserting and denying things, it seems to us, suggests that assertion can be indefeasible in a way that denial is not. (Murzi and Carrara 2014) In the quoted passage it seems crucial that evidence for accepting 0 = 0 is a proof, and thus an indefeasible evidence. At the same time the evidence for the rejection p sufficient for rejecting p is defeasible: once it is discovered that p is a dialetheia its rejection becomes unpermitted for the subject. But this argument assumes that assertion always requires indefeasible evidence. However in many other (non mathematical) cases the evidence for a proposition is feasible: I believe that this fish is a mackerel but I am wrong due to bad lighting conditions. 33 34 If we bracket this issue we nevertheless have the following problem: ARGUMENT 2 : 1. Suppose p is a dialetheia and that A and B have evidence for this.

Suppose A asserts p.
33 It is also doubtful that in mathematics the evidence for an assertion must always be a proof. 34 The other option Carrara and Murzi consider is Beall (2013)'s proposal according to which we can add the so-called shriek-rules to express that a sentence is true-only. For each sentence A which is not glutty it is added a non-logical rule of the form: A, ¬A ⊥, where '⊥' is an operator that implies triviality. By means of the shriek-rules we can then express that a theory is consistent by stating shriek-rules for all the predicates -a shrieked theory. A shrieked-theory is thus either consistent or trivial. Notice, however, that this proposal will not work if the dialetheist position for the Equal Validity Paradox entails local triviality as argued below -see infra §6.

Suppose B's way of disputing with A is rejecting p.
4. B's rejection of p prevents her from also by accepting p (Incompatibility satisfied).
5. But since p is a dialetheia and B has evidence of it, B is not permitted to reject p if B has evidence for the truth of p (Reject-p*).
6. B ought not rationally to reject p (ditto for A wrt not-p).
7. The dispute between A and B on p is neither faultless nor rationally sustainable.
The conclusion of Argument 2 is thus worse than the conclusion of argument of 1 for the dialetheist solution to the Equal Validity Paradox: assuming that both subjects have evidence that p is a dialetheia, not only can we make no sense of the rationality of the dispute, but B's rejection of p is also incorrect. Hence,faultlessness is lost. Therefore, a dispute on matters of inclination cannot be presented as involving the norm of rejection. However, the mere acceptance of a proposition and of its negation does not justify any sense of genuine disagreement between the parties to the disputes.
In principle there are several reasons why a subject could start disputing over a proposition that has been recognised a dialetheia. She could, for example, be interested in persuading the other subject to endorse some course of action instead of a different one. The point of ARGUMENT 2 is that there are no epistemic reasons (reasons for believing) that could motivate a dispute once a dialetheia is recognised. We think that precisely for this reason a dialetheist strategy (vi) must go revisionary in the end.
In conclusion, arguments 1 and 2 set up a dilemma for a dialetheist solution to the Equal Validity Paradox: either a dispute between subjects A and B involves A's acceptance of a proposition and B's mere acceptance of its negation; or else, it involves A's acceptance of a proposition opposed to B's rejection of it; if the former, incompatibility between the correctness of subjects' attitude is lost and hence, if the dispute is rational, ignorance of the fact that the proposition is a dialetheia must be then imputed to subjects (they mistakenly take acceptance of the negation of the proposition to be equivalent to the rejection of it). If, on the other hand, the dispute between subjects A and B involves A's acceptance and B's rejection of a proposition, B's rejection could be rational only insofar as she ignores that the proposition is a dialetheia, for recognition that the proposition p is a dialetheia involves recognition that both p and not-p are true and hence that rejection of each is incorrect. In both cases the idea that the dispute is rationally conducted and sustainable is at odds with the assumption that subjects have correct beliefs about the dialetheic status of the proposition.
Hence strategy (vi) is committed to a form of revisionism and it no longer seems a promising strategy for making sense of a descriptive solution to the Equal Validity Paradox.

The epistemology of dialetheias
A natural answer to the normative problem raised so far is to deny the assumption that both subjects have evidence of the dialetheic status of p. Perhaps in this way the rationality of the dispute could be regained.
The assumption that both subjects have evidence that the relevant proposition is a dialetheia needs more careful scrutiny. The crucial question is: how can subjects acquire evidence that a proposition on matters of taste is a dialetheia? Consider again the proposition that ginger is tasty. Mary tastes ginger finding it tasty, whereas Jane (tasting the same piece of ginger) finds it disgusting. At the time of tasting neither of them has first-order evidencei.e. evidence coming from their senses-that ginger is different from how their respective tastes presently inform them it is. After expressing their own views, they have nonetheless higher-order evidence -evidence coming from what the other believes-that ginger is different from how each of them finds it. Suppose they are open to the view that the proposition that ginger is tasty is a dialetheia: each subject would both have (first-order) evidence for the proposition she has asserted and (second-order) evidence for the opposite proposition.
Assuming that Mary and Jane are epistemic peers with respect to tasting gin-ger, they would thus have evidence that the proposition that ginger is tasty is a dialetheia. Would they be irrational in disputing given this epistemic situation?
The standard problem in the epistemology of disagreement is how it can be rational to disagree with an epistemic peer: the root of the problem comes from the intuition that if evidence is knowingly shared by subjects with the same skills and if it is agreed that only one view can be right, then the responsible attitude to take, once disagreement comes into light, is agnosticism. If I believe a proposition based on some evidence, further evidence that a peer takes a different and incompatible view is in its turn evidence against that proposition.
Such evidence undermines the rationality of my belief. However, once we drop the belief that the propositions in play are incompatible, the problem vanishes: the evidence for the opposite view does not automatically speak against my own view. So the question of the rationality of disagreeing between epistemic peers in the case of a dialetheia cannot be framed as the standard problem of the epistemology of disagreement.
The problem for the dialetheist strategy to the Equal Validity Paradox is in fact different, it is the problem of making sense of why Mary and Jane, say, should take one view and not both if they don't believe that their views are incompatible. To put the problem in a nutshell: if they have evidence that the proposition that ginger is tasty is a dialetheia, why should not they believe this proposition and its negation? But if they believed both propositions, then any sense of there being a rationally sustainable dispute would vanish on the assumption that both subjects have evidence that the proposition is a dialetheia.
Here is one way to answer this epistemological worry. Notice that each subject has access to different pieces of evidence for the two views: one piece is first-order (coming from the senses) for each view, the other is second-order (coming from the acknowledgement of the view of the other) for the opposite view. Following a suggestion made by Kelly (2005), there is a problem in the role that second-order evidence plays for deciding what to believe when this evidence is grounded in a belief based on the same body of evidence as the one on which one's belief is based. More specifically: If Mary credits the second-order evidence that Jane believes that ginger is not tasty as a further evidence relevant to deciding what to believe, she then gives additional weight to Jane's belief that ginger is not tasty is true when she has already considered (to the best of her ability) all first-order elements on which Jane's belief is based. Kelly's point is that this additional weight is suspicious.
Kelly's point can be developed in a slightly different direction: the secondorder evidence coming from the opponent's view cannot be as strong as my

Dialetheism, Trivialism and Pluralism
Call a domain of discourse a set of propositions that concern a certain subject matter. 35 It is not easy to define what a subject matter is, but we have an intuitive grasp of this notion when we say that the domain of morality is made of propositions that concern what is good and bad, that the domain of mathematics is the set of propositions that concern abstract properties such as being a set There are no logical reasons against trivialism. It has been argued that trivialism cannot be easily dismissed as logically absurd (Priest 2000, Priest 2005, ch.3 and Kabay 2010. Some have argued against it on the basis of its theoretical uselessness (Beall 2009, 34), others (Plebani 2015 If we conceptualise a judgement directed towards an atomic proposition as a judgement attributing properties or relations to objects -i.e. as representing that certain things are related in such and such way-then the thought of global atomic equal validity is that it is always equally correct to hold that objects are related so and so and to hold the contrary opinion. Global atomic equal validity is a difficult doctrine to accept. Can this doctrine be stabilised in 36 "LP has its so-called trivial model: if all atomic sentences of the relevant language are both true and false (or both true and false at the base world where truth is evaluated, if we have a worlds semantics for LP plus arrow), then all sentences are true and false." (Berto 2008, 170 One way to argue from empirical evidence to trivialism is to hold that optical illusions related to colours are evidence for the more general thesis that all states of affairs are inconsistent: How do I know that the room does not go red and green all over when I put on the glasses? Again, the facts that only I see it in this way, that the colouring comes and goes with the glasses, etc., clearly make the hypothesis that the colouring is subjective a superior one. (Priest 2005, §3.5) But this line of argument is not effective against an empirically based evidence related to the phenomenon of faultless disagreement in relation to basic taste. If we take seriously the idea of disagreement, then faultless disagreement in the discourse of basic taste might be taken as empirical evidence for the existence of dialetheias.
Priest's second argument against trivialism is centred on the idea that assuming that the acquisition of meaning requires contrastive applications (correct and incorrect applications of expressions), trivialism implies that language is meaningless since there are correct applications only. Priest's himself recognises that this argument is not effective against trivialism: "They may not be able to assert anything, but that does not stop them [from] uttering things. They may take themselves to be just babbling-to use 39 For alethic pluralism see Lynch (2009) and the collection Pedersen and Wright (2013).
a helpful phrase of Peter Unger. But we don't; and we still have to take into account the force of what they say, or at least, of what we take to be that force." (Priest 2005, §3.8) Moreover the objection is less forceful against local trivialism, since the local trivialist might argue that the acquisition of the language of the domain in question -e.g. taste predicates-is secondary to the acquisition of the language of other non trivial domains.
The third argument against trivialism is a sort of "transcendental deduction from certain features of consciousness to the impossibility of being a trivialist" (Priest 2005, §3.9): the argument is based on the idea that choosing is intending to bring about a state of affairs, but "One cannot intend to act in such a way as to bring about some state of affairs, s, if one believes s already to hold".
The trivialist believes that everything already holds, hence the trivialist cannot choose. But is it part of our phenomenological life that we (including the trivialist?) ought to choose. Thus, "We cannot, therefore believe everything to be the case about this world." Hence, we have a sort of transcendental argument against trivialism.
However suppose you have trivialism for the domain of basic taste. What is intending to act for basic taste propositions? Does Mary intend to act to bring about that ginger is tasty? Mary intends to act for tasting ginger, but the latter is not a basic taste proposition. Mary tastes ginger and then she judges that ginger is tasty. Her purpose was to taste ginger, once the purpose has been reached she has deliberated that ginger is tasty. However Mary did not act to bring about that ginger is tasty, nor to bring about her deliberation that ginger is tasty. It could be replied that if May intended to eat ginger she would be unable to choose what to eat in the future, assuming, as seems plausible, that in general we choose to eat tasty food and to avoid food we don't consider tasty. The objection would be effective if Mary's evidence for thinking that ginger is tasty were the same as her evidence for thinking that ginger is not tasty. However, crucially, Mary's evidence for thinking that ginger is tasty is stronger than her evidence for thinking the contrary as we have argued before (see supra 5.1). Hence, Mary can rationally choose what to eat in the future even if basic taste propositions are dialetheias.
Priest's objections to global trivialism thus do not apply to local trivialism.
Whether or not local trivialism is acceptable is not easy to answer. On the one hand, it offers a dialetheist solution to the Equal Validity Paradox leaving open the possibility that truth is realised by different properties in other domains.
On the other hand, it endangers to destabilize the intelligibility of the targeted domains. Thus, whether or not a dialetheist solution to the Equal Validity Paradox should embrace local trivialism depends on the reasons we have for being pluralist. If the explanatory advantages gained with pluralism overcome the disadvantages of local trivialism, then local trivialism can be overall acceptable.
Otherwise, better to pursue strategy (vi) in a monist framework.

Conclusions
Of the five descriptive strategies to solve the Equal Validity Paradox, the dialetheist implementation of strategy (vi) is in danger of not doing justice to the rational sustainability of disputes related to the phenomenon of faultless disagreement. Strategy (vi) thus risks to fall back onto strategy one -the revisionary strategy-, a danger shared by all other descriptive strategies. In this work it has been argued that strategy (vi) has some space for resisting the revisionary challenge and that pursuing strategy six might have interesting consequences for pluralism about truth. In fact, if strategy (vi) is applied to some domains only -like the domain of basic taste-, it provides new motivations for claiming that in these domains truth is inconsistent. If in other domains truth turns out to be realised by truth-properties such as coherence, correspondence or super-assertability that are not inconsistent, we would have a new case for alethic pluralism.
Whether or not strategy (vi) amounts to the most stable solution to the Equal Validity Paradox is still an open question, but a question that is worth enquiring.