Religious diversity in the contemporary world is characterized by its pervasiveness. Perhaps more than any other time in history, people are aware of the various conflicting religious opinions that exist, and that for any religious belief they hold, there is likely someone who disagrees with them. The result is that actual (and not merely hypothetical) cases of interreligious disagreement—i.e., disagreement between at least two persons, at least one of whom considers him- or herself religiously distinct from the other—are part of the lived reality of religious believers. The pervasiveness of interreligious disagreement naturally leads to questions about what it means to respond constructively in such cases. Constructive responses to religious diversity, whatever else might be said of them, should hold several competing priorities at once: the epistemic challenges of religious diversity are not ignored, and superiority or exclusivistic claims, while not forbidden, are subject to rational scrutiny. Yet, importantly, the outcome of any particular constructive response is not necessarily to eliminate distinctness. Rather, constructive responses to religious disagreements allow interlocutors to engage with one another robustly, opening the possibility for a range of epistemic outcomes while acknowledging the possibility of persistent and pervasive religious diversity.

Though not typically understood under the broader heading of constructive, many of the responses articulated as part of theories of interreligious dialogue might be understood as such, including interreligious learning, respect, accommodation, and appropriation, to name just a few (Cornille, 2020). But, importantly, these responses differ in the epistemic demands they place on participants in a dialogue. For instance, appropriation is more demanding than learning, learning more than respect, and respect more than mere accommodation. To learn from a religious other requires that one’s beliefs undergo some kind of transformation, be it a kind of conceptual enrichment or change in one’s understanding of how his or her beliefs relate to those of the other (Clooney, 2010; Cornille, 2020). To respect the differing beliefs of another does not require such epistemic accommodation. I can respect another’s beliefs by recognising that the other’s beliefs are “live options,” even if I am not obligated to, entitled to, or even justified in holding those same beliefs (Grube 2015, 2020).

Given the various epistemic demands that constructive interaction places on parties to interreligious dialogue, what philosophical and epistemological considerations must be developed to support a broad theory of constructive interreligious interaction? In this paper and in the following paper by Dirk-Martin Grube, we take up various issues central to the development of the foundations for a broad theory of constructive interreligious interaction in response to religious diversity. We both find promise in the notion of modal justification, which I develop in this paper, as a way to break the impasse that results from disagreement over the alethic status of a proposition. Yet, we deploy this notion to underwrite distinct understandings of the central tasks and goals of interreligious dialogue. For Grube, the purpose of interreligious dialogue is not necessarily to resolve disagreement through convergence. Rather, it is to come to understand the potentially enlightening difference that arises from robust interaction with the views of another in order to become better versions of ourselves religiously. I, on the other hand, see convergence as the ideal terminal goal for interreligious dialogue, so that constructive interaction with other beliefs should serve this goal. This means, in part, that a crucial component of an account of constructiveness is some theory of how robust engagement with differing beliefs is possible even when one deems those beliefs to be false. As such, Grube and I differ on what makes interreligious dialogue constructive: Grube locates a crucial constructive-making feature of dialogue in understanding difference, and I locate it in the possibility of convergence on truth.

The plan of this paper is as follows. I begin by offering an account of positive and negative epistemic status that depends to some degree on the relation between the alethic and epistemic status of a proposition. Even on a strong view of that relation (one which stipulates that the epistemic status of a belief is determined by its alethic status), there are nonetheless a variety of ways a proposition can have positive epistemic status, even when one thinks that the other’s belief is false. I then note how theories of interreligious dialogue are implicitly built around the question of how much positive epistemic status one is rationally permitted to assign to the differing beliefs of other religious traditions, with pluralists believing that positive epistemic status is somewhat easy to come by and exclusivists rejecting this. After this, I begin to develop my account of modal justification as an imaginative engagement with the disagreeing beliefs of another that can rationalise the ascription of positive epistemic status to the other’s belief. I then make reference to epistemological theories of disagreement, addressing in particular the role that epistemic peerhood has played in those discussions, and noting how reliance on the notion of peerhood may actually hinder the possibility of constructive interaction. Next, I consider an important objection to my proposal, namely, that the programme of ascribing positive epistemic status to the other’s belief may result in cognitive overload for the individual and that this overload makes my proposal implausible, or, at least, impractical. I defend my proposal against two versions of this objection, finding both to be somewhat overblown. Finally, I conclude by offering some reflections on the differences between my and Grube’s application of modal justification for the goal of constructive interreligious dialogue.

What does it mean for a belief to have positive or negative epistemic status? Positive and negative are blanket terms, ascribed as the result of the relevant kinds of appraisal. While there are various ways to distinguish between positive and negative epistemic status (e.g., the difference between having rational insight into a proposition or lacking it, the difference between being justified in believing a proposition or lacking justification, etc.), one common way to think of this difference is that a belief’s epistemic status is related to the alethic status of the target proposition. This begins by following the traditional analysis of propositions in terms of alethic bivalence, according to which a proposition is either true or false.¹ For a proposition to be true means that it is the case or that it is actual in the world in which we live. False means that the state of affairs expressed by the proposition is not the case, that it fails to obtain in our world. In this sense, the alethic value “true” is ascribed to a proposition as a result of a positive appraisal of a proposition, and, hence, we can say that the proposition has positive alethic status. Similarly, if a proposition is evaluated as “false,” it has negative alethic status.

While true and false are the most basic kinds of alethic statuses, standard alethic bivalence can be interpreted modally. A proposition that is actually true may have the even stronger modal-alethic status of necessary truth, in which case the relevant proposition could not fail to be true in our or any other world. In the same way, a proposition that is actually false may have the unfortunate negative alethic status of necessarily false, in which case it could not possibly be true in our or any other world.

In general, epistemologists recognise some connection between the alethic status of a proposition, P, and the epistemic status of a belief that has P as its object, though there is disagreement over the precise nature of that relation. One way to make the connection between alethic and epistemic status is via the relationship between the notions of truth, knowledge, and justification. One such interpretation of this relationship (i.e., infallibilism) holds that if P is false, one cannot know it, and thus there can be no justification for believing it. This position reflects a strong view of the relationship between alethic and epistemic status, and it is, truth be told, a controversial view. There are other ways of understanding this relationship, such that knowledge that P is compatible with P’s being false. On this interpretation, it may be that one is justified in believing that P, or that one may know that P, even if P is false. For example, one might suppose that even if one cannot know that 2+2=5, there is a sense in which one might be justified in believing it to be true. Perhaps one has grown up in a community that accepts 2+2=5 as a pillar of orthodoxy, and despite misgivings, when asked to solve the equation 2+2=, one answers consistently and without hesitation “5.” This reflects a more moderate position on the relationship between alethic and epistemic status of a proposition.

Given the plausibility of moderate positions, I expect that there may be an impulse to reject the strong view on its own merits. One might argue that the epistemic outlook it promotes is implausible and deficient, since it makes the question of one’s rationality entirely dependent on whether one believes what is true, and not whether one has met the relevant demands of rationality. For, surely, we all hold a number of false beliefs about various matters, but the simple fact that we hold false beliefs does not in itself constitute decisive information about our rationality. It is true that though the epistemically judicious individual is more likely to have more true beliefs and fewer false beliefs, nonetheless, epistemic judiciousness does not guarantee true beliefs. These considerations suggest that we should reject the strong view.

Though it is important to wonder whether the strong view is plausible, I will abstain from further judgement on this matter, since what is important to me at this point is not whether we ought to take the strong view of the relationship between alethic and epistemic status. Rather, my interest is this. I will argue in the next section that the strong view helps to make sense of some versions of religious exclusivism. What we should want to know is whether these commitments affect the potential for religious exclusivism to guide constructive interaction, or, to put it somewhat differently, whether it is possible to interact constructively with beliefs that one holds to be false. In what follows, I hope to show that this is, in fact, possible.

At the heart of responses to religious diversity is the question of how much, if any, positive epistemic status it is rational for one to ascribe to the contradictory religious beliefs of others. The results of this task are essential to making constructive interreligious interaction possible. Approaches to religious difference situate themselves relative to two poles of a spectrum: classical pluralist and exclusivist.² The classical pluralist response to religious diversity seeks to give an account of how multiple, even apparently contradictory, religious claims can enjoy the same degree of positive epistemic status.³ Hick (1989, 2001) defends the pluralistic hypothesis, by which he argues that differing, even contradictory religious traditions, insofar as they constitute genuine responses to a transcendent and ineffable Real, are equally valid, i.e., have the same degree of positive epistemic status. Following a distinct pluralistic tack, Schmidt-Leukel (2009, 2017) argues for a programme of synthesis founded upon what he calls a fractal theory of religious diversity. On this view, different religious traditions are parts of the same whole whose identity is, in turn, formed as a result of the constructive interaction between the parts. Importantly, both Hick and Schmidt-Leukel understand themselves (as well as others) as engaging the pluralistic programme from the perspective of their respective faith traditions. They claim that their programme cannot be carried out from a disengaged perspective; rather, one must be affiliated with some religious tradition or perspective. In doing so, they aim to defend religious difference, allowing for, and even encouraging, religious diversity. Of course, not all approaches to religious diversity understand the task of ascribing positive epistemic status as grounded in the manner suggested by pluralists. In a recent contribution to the Christian theology of religions, Daniel Strange (2014) takes up the Reformed theology of religions of Jonathan Edwards and J. H. Bavinck. Strange’s brand of exclusivism argues that non-Christian religions are “antithetically against yet parasitically dependent upon the truth of the Christian worldview, [and that] non-Christian religions are ‘subversively fulfilled’ in the gospel of Jesus Christ” (Strange, 2014, p. 42). Consistent with his understanding of the Christian Reformed theological tradition, the Bible is the normative epistemological standard for knowledge. Beliefs are justified (i.e., have positive epistemic status) only to the extent that they are consistent with the affirmations of Scripture. If a belief contradicts an affirmation of Scripture, that belief has negative epistemic status. As such, he argues that Christians cannot learn directly from other religions if “by learning we mean the reception of new or neglected truth in the answers or whatness the religious others might give to the perennial questions of existence, the ‘magnetic points’ concerning God, humanity and the world.” (Strange, 2014, p. 332) This is because, in his view, especially with regard to non-believers whose sin inhibits the proper function of their truth-seeking abilities, accurate epistemic judgements should be seen as “a God-given ‘inconsistency’ or, to use Van Til’s famous phrase, must be shown to be the ‘stolen capital’ of the Christian worldview.” (Strange, 2014, p. 91)

Despite the rather sharp contrast between the pluralist agenda, on the one hand, and, in this case, the exclusivism of the Reformed evangelical theology of religions view, on the other, at issue for both is the question of the degree of positive epistemic status one can and ought to grant to other religious traditions. Pluralists seek to grant equal epistemic status to more than one religious tradition. For Hick, the great religions of the world have very high epistemic status; Schmidt-Leukel sees the religions of the world as moderately justified by themselves, but as strengthening their justification through interaction with one another. In sharp contrast to these views, Strange argues against according positive epistemic status to any religious tradition that rejects the Christian Scriptures as the ultimate epistemological criterion.

Differences aside, they share a crucial methodological assumption: if a belief has only negative epistemic status, constructive responses—learning, respect, et al.—are not possible. Conversely, if there is reason to think that a belief has some positive epistemic status, constructive interaction with that belief is an obligation. Recognizing this assumption helps to explain the pluralist impulse to avoid, sometimes at all costs, assigning negative epistemic status to differing or contradictory beliefs. It also explains the reactive tendencies underlying exclusivism: differing or contradictory beliefs have only negative epistemic status, and, hence, there is nothing to be gained from them.

To be sure, we should find nothing wrong with this assumption (i.e., that beliefs with only negative epistemic status do not merit constructive interaction). Teachers of mathematics do not learn from, respect, or even tolerate the belief that 2+2=5, nor should we learn from, respect or tolerate racist or sexist ideologies, and for good reason: those beliefs have only negative epistemic status. Such beliefs are the results of epistemic errors, e.g., faulty reasoning, carelessness, or vice. As such, there is nothing to be gained from directly interacting with such views.

But, importantly, believing that some proposition is false does not oblige one to ascribe only negative epistemic status to the relevant belief. For, though we all hold opinions about the falsity of some beliefs, we likely think that at least some of these views are reasonable. Given that reasonableness is a kind of positive epistemic status, the fact that we can affirm that a belief is reasonable though false suggests that merely believing a proposition to be false does not guarantee that one will ascribe only negative epistemic status to the relevant belief. This means that even versions of exclusivism that adhere to a strong view of the relationship between truth, knowledge, and justification may be able to foster constructive interaction. What is unclear is how exactly such theories may be able to do so. In the following section, I will propose a mechanism for grounding the ascription of positive epistemic status, which, if adopted by exclusivists and pluralists, can serve as a philosophically responsible way of guiding constructive interreligious interaction.

The notion of what I will call modal justification provides a way for individuals to ascribe a minimal degree of positive epistemic status to the other’s belief even though they think that that belief is false. This makes it possible for a believer to maintain that the other is mistaken yet still view the other’s belief as having some positive epistemic status. Constructive interactions based on modal justification can be analysed according to the following form: If A should recognise that P is possibly justified, A should interact constructively with B’s belief that P.

Modal justification requires a few preliminary moves to get it off the ground. To begin, I return to the alethic statuses mentioned earlier (necessarily true, actually true, actually false, necessarily false). The statuses of actually true and actually (but not necessarily) false can be analysed as possibly true, since if P is actually true or false, at minimum it must be possibly true.⁴ Hence, though an actually false proposition fails to obtain, it might have obtained. The Empire State Building in New York City opened in 1931, but, given a few construction delays, it could have opened in 1932. The alethic status of possibly true has implications for a belief’s epistemic status, for if a proposition might have been true, that means it might have been justified.

While these modal considerations have bearing on matters of epistemic evaluation more generally, my goal here is much more modest.⁵ Consider that a great number of religious beliefs—from the contents of divine revelation to prophetic authority, even down to robust doctrines such as the Incarnation—fall squarely into the category of possibly true propositions. That is to say that the states of affairs expressed by those propositions, even if they are actually true, could have easily been false. God could have chosen not to reveal certain aspects of his character, not to have called this or that prophet, or to have refrained from taking on human nature. That these propositions could have easily been false constitutes a reason for their possibly being justified as a kind of epistemic status. Even if they turn out to be actually false, their status as possibly justified counts as a reason to engage in constructive dialogue about their ultimate epistemic status.⁶

To recognise that a proposition is possibly justified is a way of recognising that there are some rational considerations in its favour. Yet, to commit to a belief as possibly justified does not commit one to affirming its truth. This means that the exclusivist who adheres to the strong view of the relationship between alethic and epistemic status can continue to deny that P is true while ascribing it some positive epistemic status. And this recognition of the possible justification of P is enough (or so I will continue to argue) to allow for constructive engagement about P-related matters.

What reasons are there for thinking that a false belief could have positive epistemic status? For most propositions (including religious propositions), it is often not the case for an actually false but possibly true proposition that none of the justifications for that belief fail to obtain. Let’s say that P is the belief that the Qur’an is divine revelation. The (exclusivist) Christian denies that this belief is true, and, hence, that it is actually justified. Though there are myriad apologetic and doctrinal questions that need resolution, it would seem to me that the Christian can recognise that P is possibly true. If P is possibly true, it has propositional justification, some of which (very likely) obtains in our world. For instance, a world in which the Qur’an is divine revelation must be a world in which the Qur’an speaks of God as compassionate and merciful, since these are, I take it, essential attributes of God. If one can recognise that a belief one considers to be actually false as possibly true and that some of the propositions that would justify that belief obtain in our world, then this is enough to license the ascription of positive epistemic status (i.e., possibly justified) to that belief. Hence, one can interact constructively with that belief: he or she can disagree that it is actually true, but nonetheless take action consistent with constructiveness.

Consider another controversial belief in Christian-Muslim dialogue: the affirmation or denial of the crucifixion of Jesus of Nazareth. The Christian believes that the crucifixion occurred, and having investigated the matter, believes that he or she is highly justified. However, following theologians like Thomas Aquinas, the Christian may recognise that events like the Incarnation (and, hence, the crucifixion) need not have happened; God could have brought about salvation through alternative means. The resources exist for the Christian to recognise that although he or she has good reason to believe that he or she is right about the crucifixion, it is possible that it might not have occurred and that beliefs that entail the denial of the crucifixion are possibly justified.

Of course, this all depends on possibly justified counting as a kind of positive epistemic status. Should it be considered as such? My answer is yes, and I will make this argument by considering the counterarguments. Along these lines, there are, so it seems, some reasons for thinking that modal justification should not count as a positive epistemic status. One important reason against allowing possibly justified to count as a positive epistemic status is that it is too generous. Allowing this to stand will result in a great number of false beliefs attaining positive epistemic status. However, I do not see this as a problem. Remember that the goal here is to ascribe enough positive epistemic status to enable constructive interaction. Interlocutors who are part of constructive interactions may begin and end the dialogue believing that the other is mistaken, since constructiveness does not presume that interlocutors will have to converge in every single instance of dialogue. But this minimal positive epistemic status at least allows them to engage one another in a robust, non-trivial manner.

Another reason against possibly justified as a positive epistemic status is that it treats matters of truth and falsity too casually. It ought to mean something that the exclusivist holds the other’s belief to be false and appealing to possible justification for the other’s belief does not respect the exclusivist’s assessment of its falsity. Yet, by my lights, modal justification does precisely that, viz., it respects assessments about the falsity of the other’s belief. The trick is that it also allows the exclusivist to assign positive epistemic status to a belief that he or she deems to be false. Much of the hesitation expressed in this worry can be mitigated by keeping an important constraint in mind. To consider a belief to have positive epistemic status does not imply that the belief is true. Rather, it is simply to say that it has some rational-making consideration in its favour. In this way, rational belief remains distinct from true belief.

A natural question about my proposal is whether it says anything that goes beyond the already vast and growing literature on the epistemology of disagreement. For it would seem that the basic premise on which constructive interaction—based, as it is, on modal justification—is built (i.e., the need to ascribe positive epistemic status to the beliefs of the other) is rendered irrelevant by the theories of disagreement on offer (i.e., Conciliationism and Steadfastness). According to this line of thought, there is reason to think that Conciliationism and Steadfastness are inherently constructive. Since advocates of Conciliationism and Steadfastness counsel parties to a disagreement to engage robustly with the other’s evidence and to take seriously the other’s cognitive virtue, however we go about developing an account of what constructiveness is, it will likely show that the most plausible versions of Conciliationism and Steadfastness inherently possess the resources for parties to a disagreement to engage one another constructively.

However, there is one feature of the epistemology of disagreement that suggests that these theories may not be inherently constructive. Especially in the early literature on the topic but present throughout subsequent discussions, disagreement between epistemic peers—i.e., interlocutors who possess roughly (or, for some, exactly) the same evidence and are roughly (or exactly) cognitive equals—is what makes disagreement philosophically interesting (Kelly, 2005, 2010). As I hinted at in the previous paragraph, robust engagement with the other’s evidence and appropriate deference to the other’s cognitive abilities may be crucial components of constructive interaction. If this is so, then the constructiveness of Conciliationism and Steadfastness consists of (perhaps to a significant degree) interlocutors’ assessments about whether the other is a peer.

Yet, despite the potential importance of peerhood for the constructiveness of theories of disagreement, it is my contention that epistemic peerhood is not necessary for, and, in fact, may hinder the possibility of constructive interaction. This is because much of the literature on the epistemology of disagreement has not considered cases in which one encounters an epistemic superior or inferior to be philosophically interesting. The prevailing assumption is that when one encounters an epistemic superior—one who is more intellectually virtuous and better informed evidentially relative to P—with whom one disagrees, one will have to revise his or her belief to some degree. Alternatively, if one has reason to believe that one is epistemically superior to one’s interlocutor, he or she not only has reason to remain steadfast in his or her belief relative to P but one should expect one’s interlocutor to significantly revise his or her belief.

In my estimation, the ways that epistemic superiors and inferiors should relate to one another misses something of the essence of constructive interaction. For it would seem to me that there are many cases in which I should interact constructively with someone’s belief even if he or she fails to be my epistemic peer. Suppose a Christian encounters a Muslim who did not finish high school but insists that Jesus was not crucified. It may be clear from interaction with the Muslim that the latter cannot give a good account of the counterevidence presented, nor does it appear that the Muslim has given much thought to the question at all. The Christian should not be faulted for leaving the encounter with at least the same degree of justified confidence in the crucifixion as before the encounter. But, more generally, how should the Christian think about those who deny the crucifixion? Of course, in this particular instance, the Christian did not encounter a peer. If this were one’s only encounter with a disagreeing interlocutor on this matter, one may tend to think that any Muslim who denies the crucifixion is not a peer and should thus revise his or her belief. But this way of handling the situation does not fully account for the possibility that there may still be something for the epistemic superior to learn from the encounter. Perhaps the Muslim presents an objection that the Christian has never before considered, so that, though the Muslim is not an epistemic peer, the Christian can gain something epistemically.

Taking these matters to heart, in the place of peerhood, I suggest that an account of recognition can do the necessary work of grounding modal justification for constructive interaction. ⁷ Recall the articulation of modal justification stated above: If A should recognise that P is possibly justified, A should interact constructively with B’s belief that P. Recognition, then, plays a foundational role in the success of modal justification. In a disagreement between A and B, the obligation for A to interact constructively with B’s differing belief relative to P is grounded in A’s recognition that it is possible that B’s belief relative to P has some positive epistemic status. This acknowledgement can be more basically analysed in terms of two logical moments: (i) A conceives of possible justification for P and (ii) A believes that B’s epistemic circumstances make it likely that B has some justification for P.⁸

Regarding the first logical moment, A’s ability to conceive of possible justification for P is a part of what constitutes a rational response to disagreement. Conceivability is grounded in the simple fact that human beings are imaginative creatures, able to cognise a great number of possibilities. We can imagine that for any proposition that could be true, that proposition has a number of possible justifications. Hence, the imaginative capacity of human persons is such that persons can consider what conditions would need to obtain in order for a belief to be true.⁹

The second logical moment relies on a distinction between our world and the possible worlds that are closest to ours. Merely being able to conceive of the justification for P does not mean that there is in our world a way of knowing such that one can appropriate those justifications for belief that P. Recognition goes beyond mere conceivability, requiring that some way of knowing that P in a possible world in which P is true is present in the actual world. More likely than not, if B believes that P and A can conceive of possible justifications for P, it is probably the case that B has some actual justification for P. However, given the circumstances that led to their disagreement (A has a different belief about P), A may understand B’s justification as based on less than fully conclusive reasons. This means that in most cases, barring bizarre (though possible) counterinstances, B’s belief will have some actual justification. Importantly, this does not mean that A must conclude that B is justified, because, in addition to the possibility that a bizarre counterinstance obtains, for various reasons it may be that A cannot fully analyse B’s justifications. A only need recognise that it is very likely that B’s belief has some degree of justification, and A can do this without knowing precisely of what that justification consists and without acknowledging that B is in fact justified in the contested belief.

These dual considerations—the ability of A to conceive of propositional justifications for P and the acknowledgment that it is likely to be the case that B will believe that P only if B has some justification for P—are what ground A’s recognition of B’s belief that P as possibly justified.

To this point, I have argued that constructive interreligious interaction is desirable since it encourages interlocutors to engage with religious difference in a robust manner. I suggested that theoretical approaches to dialogue assume that interlocutors’ ability to engage constructively with disagreeing religious views depends implicitly on the degree to which they can assign positive epistemic status to the other’s view. I then proposed the notion of modal justification as a philosophically responsible way for interlocutors to assign positive epistemic status to the beliefs of others, thus making constructive interaction possible.

But some important questions remain. Here I want to consider two versions of the cognitive overload objection to my proposal.¹⁰ The versions of the objection run like this. On the first version, we should wonder about the kinds of situations that give rise to the demand that we interact constructively with other religious beliefs. The concern is that the wide diversity of opinions on any given religious matter means that the matter of when to interact constructively with other beliefs must be carefully delimited. For, surely, we cannot be expected to interact constructively with every differing or disagreeing belief, since we are limited in terms of our time, our ability to access the relevant arguments for differing religious perspectives (for instance, if they are written in a language that we do not understand), etc. Second, a further worry about my proposal is that the task of ascribing positive epistemic status to the other’s beliefs may place burdensome cognitive demands on interlocutors. In some cases, this may lead to cognitive incongruities. For example, to pick up on the example in the previous section, the epistemically judicious Christian will have to do significant cognitive work in order to ascribe positive epistemic status to the disagreeing beliefs of his or her Muslim interlocutor, who may or may not exert similar cognitive effort. Since I argued that one can do this even in situations when the other fails to be a peer, the question is whether the kind of cognitive work that is required to motivate constructiveness is really worth the effort, especially in cases in which the other is not similarly moved by the disagreement, or if the epistemic benefits of constructive engagement are likely to be marginal.

Regarding the first version of the cognitive overload objection, I have made some suggestions in previous sections (i.e., “Positive Epistemic Status and Religious Diversity”) that point to the potential limits of constructive interaction. In that section, I noted that theorists of religious diversity seem to agree that we are not obliged to interact constructively with beliefs that have only negative epistemic status, and I concurred with this point. Yet, this seems to only reaffirm this worry about cognitive overload, since concluding that a belief has only negative epistemic status implies that one has thoroughly evaluated the relevant belief and found it to be deficient in important ways. Thus, while a belief’s having only negative epistemic status may relieve pressure for one to interact with it constructively, it leaves open the possibility that one will have to engage with a wide range of disagreeing beliefs.

To address this worry, I want to suggest two further limits on the obligation to interact constructively. First, one is not obligated to interact constructively with a belief if one does not have a prima facie reason to think that belief has positive epistemic status. Lacking a prima facie reason to think that a belief has positive epistemic status is, of course, different than believing that that belief has negative epistemic status. The Christian may think that the Muslim’s disagreeing belief about the crucifixion lacks positive epistemic status, even though the Christian is unsure about whether that belief has only negative epistemic status. In such cases, there seems to be no obligation for one to further engage with one’s interlocutor in a constructive manner, since having positive epistemic status is what obliges one to respond in this way. Because constructive interaction is grounded in this way, if it is not clear whether a belief has positive epistemic status, this is enough to relieve the pressure to act constructively. Yet, there are important differences between thinking prima facie that a belief lacks positive epistemic status and thinking that a belief has only negative epistemic status. In the former case, one may refrain from constructive interaction with the belief in question but remain open-minded to the possibility of constructive interaction in the future. In the case of the latter, thinking that a belief has only negative epistemic status may motivate more serious epistemic interventions, such as censoriousness. Nonetheless, for present purposes, it is enough to say that having a reason to think (even prima facie) that a belief has positive epistemic status is necessary for motivating constructive interaction, so that lacking a prima facie reason to think that a belief has positive epistemic status excuses one from having to interact constructively.

Importantly, there are some cases in which extreme beliefs such as conspiracy theories may appear to have some positive epistemic status. Such beliefs often possess a certain psychological allure as they have the ability to offer cogent narratives that explain a host of facts. If such beliefs seem to have positive epistemic status, should we expect to interact constructively with them always? In short, I think not. This points to a second limit on constructive interaction: we are not obliged to interact with a belief if we have reason to think that doing so will cause moral or epistemic harm. In such cases, the conflict between one’s moral and epistemic systems allows one to override any of the direct epistemic considerations in favour of constructive interaction.¹¹ If my interlocutor believes in capital punishment for religious blasphemy, I may be within my rights to refuse to interact constructively with his or her belief on the suspicion of moral or epistemic harm that may be caused by interacting with it in this way. A moral harm is any harm that distorts my moral convictions, for instance, a belief in free speech, while an epistemic harm is a harm that I incur by treating a controversial belief as epistemically respectable, even when my intuitions strongly suggest that it is false. These considerations excuse individuals from interacting constructively with the other’s belief, though they do not demand this.

As for the second version of the cognitive overload objection, assuming that we can properly limit the cases in which constructive interaction is appropriate, the question remains about whether making constructive interaction dependent on ascribing positive epistemic status to the other’s belief is too cognitively demanding. Certainly, I think that an account of constructive interaction should not place too many demands on either participant in a dialogue. Yet, I do not think the demands that are implicit in my account are more burdensome than, say, the demands implicit in the accounts of peerhood discussed in the literature on disagreement. In those discussions, the matter of the extent to which one’s interlocutor is a peer is important to understanding the extent of belief revision. Determining whether one is a peer depends on assessments about the other’s cognitive and evidential credentials. Is the other similarly competent as I am on the matter at hand? Does the other have roughly the same evidence as I have on the matter? Assuming that in most cases in which I become aware that I disagree with another on P, I should make some rough assessment about whether the other is a peer, it is unclear to me that assessing whether the other’s belief has positive epistemic status is any more cognitively demanding than assessing whether he or she is a peer. In fact, it may actually be less cognitively demanding, since, as I have it, one can assess whether the other’s belief has positive epistemic status without assessing the other’s cognitive or evidential credentials. This being the case, there is reason to think that recognising positive epistemic status may not be too cognitively demanding.

Taken together, I do not think that my proposal makes implausible cognitive demands. Constructive interaction is only obligatory when one has reason to think that the other’s belief has some positive epistemic status, and even this can be overridden by prima facie suspicions that interacting with a particular belief will result in moral or epistemic harm. Further, the task of assigning positive epistemic status to the other’s beliefs is no more demanding than determining whether the other is a peer, and it may, in fact, be less cognitively demanding than assessments about peerhood. Thus, the cognitive overload objection may be overblown.

In this paper, I have made the case that the basic task that drives constructive interreligious interaction is the ascription of positive epistemic status to the beliefs of others. One mechanism for completing this task is modal reasoning about the possible justifications for the other’s belief, what I call modal justification. I defended this proposal against the cognitive overload objection, arguing that constructive interaction based on the ascription of positive epistemic status makes sense of our obligations in the pervasively diverse world in which we find ourselves.

To conclude my discussion, I want to offer some reflections on the relationship of my proposal to the suggestions that Dirk-Martin Grube will offer in the subsequent paper. It is important to note that Grube and I agree on the basic framework that I have proposed here. We both agree that constructive interaction is important to the task of interreligious dialogue and that modal justification provides a philosophically responsible way to ground this kind of interaction. As I suggested in the introduction to this paper, where Grube and I differ is the goal of interreligious dialogue and, ultimately, how modal justification and the ascription of positive epistemic status supports that goal. Grube sees the goal of dialogue as coming to an understanding of what he calls “potentially enlightening difference” which can help interlocutors become better versions of themselves religiously. He understands the notion of truth to play an important role in guiding interreligious dialogue, even if converging on the truth is not, properly speaking, the goal of any particular interreligious dialogue.

I agree with Grube that one possible result of interreligious dialogue is that interlocutors come to understand their differences in a new way, and that robust constructive engagement can be epistemically enriching for interlocutors, as they are forced to consider the reasons for their beliefs and embrace their cognitive limits. Yet, I think this can and should be done with full awareness of the fact that in many cases of disagreement over some belief, at most only one of the parties to the disagreement can be correct. Thus, while becoming better versions of ourselves religiously is an important outcome, its value must always be considered in light of its truth-conduciveness. Consequently, it seems to me that we should acknowledge the importance of convergence on truth as the ultimate goal of interreligious interaction, even if specific cases of dialogue do not result in convergence.