The epistemic dimension of reasonableness

I Justificatory dilemma

Political liberalism faces an intrinsic tension within its theoretical framework. The attempt to provide a strictly political account of the justificatory framework for liberal theories of justice gives rise to a dilemma between two aspects of the justificatory procedure. On the one hand, the philosophical side demands that the procedure of justification provides a sound argument granting the normativity of the whole structure; on the other hand, practical constraints involving a major focus on the actual circumstances of justice and on the fact of pluralism call for a less idealized account of public justificatory procedures. In order to develop a procedure of justification consistent with both commitments, Rawls (1993: 10–12) claims that the philosophical argument for justifying principles of justice should be free-standing; i.e. completely detached from any specific comprehensive doctrine. Yet the arguments that inform the public – and free-standing – justification should be consistent with different theories of the good life, so that the resulting political principles might obtain support from different moral perspectives. It follows from this tension that any attempt, by a strictly political account of liberalism, to provide free-standing arguments for justifying the political system will end up facing a dilemma between stability for the right reasons and pluralism. ¹ The more the theory respects the fact of pluralism, employing strictly free-standing arguments, the more the theory loses philosophical depth and consequently does not grant the stability for the right reasons. Thus, it might look as though political liberalism has reached a dead end. Either political liberalism accepts the need to face the actual circumstances of justice here and now and therefore becomes inclined to focus its attention more on the issue of the actual consent than on the quest for a philosophical normative argument, or it tries to overcome the fact of pluralism by reaffirming the validity of liberalism as a true – and philosophically justified – theory of justice.

Many authors (Eberle, 2002; Gaus, 2011; Maffettone, 2010; Quong, 2011; Talisse, 2009) have recognized this dilemmatic outcome of the justificatory paradigm of political liberalism and have worked on normative solutions that might avert it. Most commonly, such authors have focused on the public reason debate; in this article instead my goal is to address this justificatory dilemma from the perspective of moral epistemology. In order to do so, let us reframe the justificatory dilemma in a way well suited for an epistemic analysis. Examining the dualistic aspects of liberal theories of justification means, at the epistemic level, wondering how a justification of political principles that is free-standing might anyway achieve the necessary robustness (see D’Agostino, 1996; Gaus, 1996). Robustness is an epistemic notion according to which a ‘theory T1 is robust vis-à-vis T2 to the extent that changes in T2 – including the total rejection of T2 in favor of some competing theory T2’ – do not weaken the justification of T1. Robustness is to be contrasted with sensitivity; to the extent that the justification of T1 is affected by changes in T2 T1 is sensitive to T2’ (ibid.: 6). The regulative ideal that political liberalism tries to respect is that of providing a public justification of political principles that demonstrates how these principles can achieve robustness in relation to extremely different comprehensive doctrines. If a public justification fulfills the task of making liberal tenets robust vis-à-vis a wide variety of comprehensive views, these theories say, then political liberalism will be justified through a procedure that shows: (1) it is philosophically adequate; (2) it respects the fact of pluralism and the variety of comprehensive doctrines held by citizens. However, the attempt to grant political theories the adequate robustness implies different drawbacks that, again, call for the dilemmatic outcomes of political liberalism. On the one hand, if the procedure of justification has to be free-standing vis-à-vis a wide menu of comprehensive views, then the array of philosophical arguments that can be employed in the justification is extremely reduced. An actually free-standing justification of political theories of justice would result in being almost ‘empty’, given the necessity to avert the reference to comprehensive ideals and values (Dworkin, 2000; Galston, 2002; Habermas, 1995). On the other hand, a free-standing justification of political theories leaves room for questions concerning the necessary motivational force that this justification should guarantee in order to face the unconstrained set of comprehensive beliefs held by citizens. In fact, if the public justification of political principles cannot refer to the beliefs that citizens hold more strongly, since such beliefs are connected with different comprehensive doctrines (such as T2 and T2’), then it seems that the motivational force attached to the resulting political principles is substantially reduced. In this regard, political liberalism seems to become counter-intuitive, as it requires citizens to avert any appeal to most of the beliefs they hold more strongly, when they are engaged in a public deliberation over political matters. ²

II Epistemic circumstances of justice

The concept of robustness, when imposed as a regulative constraint to the procedure for publicly justifying a political theory or principle, stresses the fact that reasonable pluralism should be accepted as a stable outcome of our public debate over political matters. Indeed, in a justificatory domain in which the fact of pluralism is taken for granted, the requirement for political theory to be robust vis-à-vis different comprehensive doctrines is justified through the epistemic idea that disagreement cannot be solved and, therefore, it is necessary to distinguish between public reasons that are employed within the free-standing justification of the political domain and private reasons by which citizens might support the very same political module. Different private beliefs, even when inconsistent with each other, might overlap in the justification of the same political module for extremely different comprehensive reasons. Consequently, we can define reasonable pluralism as a sensible outcome of a justificatory framework in which it might be the case that two individuals are both justified in holding beliefs that are simultaneously justified – by private reasons – and yet incompatible with each other. This epistemic circumstance stresses the fact that it is fundamental to distinguish between: (1) private and public justification; (2) the doxastic and non-doxastic aspects of the justificatory processes.

Public justification – and the epistemic and normative constraints that apply to this category – is suitable only for a specific range of matters and beliefs. In order to claim that a belief (or principle) is publicly justified, such belief must as a result be robust enough to be justified regardless of the other comprehensive beliefs that different persons hold in their sets of beliefs. This means that a political principle is publicly justified when the arguments employed for such justification might result in being consistent with private reasons that are incompatible with each other. This definition is consistent with a common interpretation of overlapping consensus as an intermediate justificatory stage that is antecedent to that of public justification. According to this approach, overlapping consensus is that justificatory stage in which each citizen, thanks to the appeal to her or his own private reasons, justifies the political module that has been already justified by free-standing arguments (Quong, 2011; Rawls, 1995). This means that in a stable overlapping consensus, a political principle p might result in being justified for Julia thanks to reason α, while the same principle p would be justified for Mark thanks to reason β. Julia and Mark agree on the validity of principle p, but they disagree on the private reasons that justify such principle p. The role played by overlapping consensus within the justificatory picture outlined by Rawls is exactly that of establishing whether the free-standing arguments in favor of a political principle might be consistent with the different comprehensive system of beliefs held by citizens. In this regard, the overlapping consensus plays a fundamental motivational role, being the justificatory stage in which each citizen looks for a possible consistency between his or her comprehensive set of beliefs and the political module he or she is required to support. It does not matter that Julia and Mark do not agree on the very same private reason in favor of p (i.e. α or β), rather it is important that they overlap, thanks to their different private reasons, on the validity of p.

Once the fact of pluralism has been acknowledged as a given fact of our political life, and consequently the distinction between private and public justification has been cashed out, then it should be clear that the justificatory procedure could not overlook the epistemic role played by citizens within the deliberation. In order to understand the epistemic relevance of the personal standpoint, we can recall a quite common epistemic distinction (Scanlon, 2003: 140):

to justify a principle or judgement is ‘to say that it is supported by good and sufficient reasons’;

According to this distinction, the epistemic role of justification is not exhausted by the introduction of a set of reasons R that provides a propositional justification for p. Rather, if the epistemic value of a justification hinges in part on the deliberative role played by moral agents, then a full account of the justificatory processes should involve a doxastic analysis that refers to the actual possibility that an agent S might believe, for sound reasons, that p. Consequently, the provision of a set of good reasons R is not a sufficient condition for the agent S to believe correctly that p. Agent S might believe that p for different, and unjustified, reasons or, even, might not accept the set of reasons R as acceptable within her doxastic system. A doxastic justification is usually defined as a justification in which a non-doxastic justification is coupled with a basing requirement, namely the fact that agent S bases her belief on the reasons that propositionally justify it. ³ The doxastic presupposition claims that, for a correct epistemic analysis of internal deliberation, it is not sufficient to assess the set of reasons R that is available to an agent S, as we should as well evaluate the epistemic performance carried out by agent S. ⁴ The doxastic presupposition implies that, from the epistemic perspective, it is important to distinguish between the two following concepts (see De Paul, 1993; Skorupski, 2000; Wedgwood, 2007):

Judgement, i.e. activity that pertains to the epistemic appraisal of the judge; this is a subjective epistemological notion governed by the rational connections among different agents’ internal mental states. Even if subjective, this epistemic activity of producing judgements is ruled by normative criteria (e.g. being correct or incorrect; meeting rationality constraints) and therefore epistemic justified beliefs are those beliefs which it is rational to hold.

I maintain that for a political theory of justice, in order to respect and deal with the fact of pluralism and the doxastic aspects of our moral knowledge, it is important to adhere to a moderate epistemology and to sustain a fallibilist account of knowledge. According to fallibilism, knowledge, when viewed as strictly connected with our doxastic evidence for it, is compatible with the possibility of error. In this regard, a distinction should be made between two meanings of ‘knowing’:

If S knows p, then S is not mistaken about p.

The first goal of this article has been to show that, once the doxastic aspects of political deliberation are recognized as a true epistemic fact of our deliberative life together, then the most adequate epistemic framework for dealing with political justification is a moderate one. I define this framework as moderate, since it implies an epistemic modesty according to which we have to look for a justificatory framework that, although aiming at objectivity in moral and political deliberation, acknowledges the impossibility, for political agents, to bridge the epistemic gap between justification and warrant. Once it is demonstrated that this epistemic gap can never be filled, then the justificatory framework should face the fact that, even though the doxastic justification is still a justification for believing true a certain proposition p, the whole epistemic procedure of justification cannot be defined as able to guarantee truth in an uncontroversial way. ⁸ A political theory could be objective to the extent that its justified principles are at least fallibly (and less than infallibly) justified. Since there are no valid moral and political deliberations in which a sound justificatory argument (propositional perspective) can be introduced as completely detached from the fact that a moral agent is actually justified in holding this specific political view (doxastic perspective), then a moderate epistemic framework acknowledges the fact that we have to deal carefully with the epistemic difference between robustness and uncontroversiality. Therefore, a moderate epistemic account of justification replies to the skeptical challenges defending the possibility that a political principle could be defined as robust, since it has been publicly justified through sound reasons. Yet this very same political principle cannot be defined – and justified – as infallibly true, because the agents involved in the justificatory procedure lack the epistemic ability to fill uncontroversially the gap between the doxastic and non-doxastic aspects of knowledge.

III Epistemic peerhood

In the previous two sections of this article I discussed the dilemmatic outcomes of a justificatory structure underpinned by strictly political arguments and I introduced some epistemic categories that can help in figuring out what does not work correctly within such justificatory framework. My line of argument has been focused on the fact that some of the dilemmatic outcomes might be avoided – or at least reinterpreted in a less problematic way – once an epistemic analysis of such justificatory procedure is correctly outlined. In this regard, I agree with Gaus (1996: 4) when he claims: ‘If public justification is the core of liberalism (as I believe it is), and because there is no such thing as an uncontentious theory of justification, an adequately articulated liberalism must clarify and defend its conception of justified belief – its epistemology.’

I have already defended a moderate epistemology that upholds a fallibilist perspective and that highlights the fundamental epistemic role played by agents’ doxastic states. These two theoretical findings are fundamental to develop the next arguments of this article, because of the fact that fallibilism – when coupled with a doxastic presupposition – provides us with a specific account of disagreement. In fact, given the fallibilist account of knowledge, it follows that we have good epistemic reasons for claiming that disagreement must be acknowledged as an intrinsic feature of contemporary democracies, since it does not necessarily depend upon the fact that one of the two disagreeing parties has been irrational or that one party’s cognitive processes have been defective. If reasonable disagreement among fellow citizens is an actual epistemic possibility, then this fact affects both the political account of public deliberation practices and the definition of the fundamental concept of reasonableness. From this, I argue, it follows that the very same notion of legitimacy must be modified, once the fact of disagreement is acknowledged as an intrinsic feature of contemporary democracies. When a political principle is legitimated if and only if it can be justified in the light of the different perspectives of citizens to whom it is addressed, then the fact that citizens hold different, often contradictory, doxastic sets of beliefs is a fundamental matter. On the one hand, such an outcome modifies the way in which political deliberation must be framed and carried out. On the other hand, a reasonable citizen should be prompted to recognize that she has not the political or epistemic authority to impose the comprehensive beliefs that are justified within her doxastic set as ultimate truths to other citizens. Naturally, this reasonable attitude by one citizen calls for a symmetric attitude by the other citizen. In this regard, I believe that the political reasons that have been provided by theorists in order to prompt citizens to be mutually reasonable with regard to both meeting the constraints of public reason and respecting other citizens’ comprehensive views lack motivational force, because the epistemic aspects of such dynamics have been completely dismissed. We need epistemic arguments in order to explain to citizens that, within a political debate, it is possible for an agent S to be justified both in believing that p and in believing that Z is justified in believing that ˜p. ⁹ How is it so? In order to grasp the political meaning of such an epistemic claim I need to introduce a new concept, namely that of epistemic peers.

An epistemic peer is someone whom you take to be equally likely to make a mistake (Elga, 2007; Peter, 2013b). There are other definitions of epistemic peer-hood that refer to the degree of equality with regard to epistemic virtues such as intelligence, coherence, attentiveness, intellectual honesty, etc. ¹⁰ Since I have already provided sound reasons in favor of fallibilism as the most adequate epistemic paradigm in the field of moral and political reasoning, then I believe that it makes sense to accept the former definition of epistemic peers as suitable for our discussion. Two agents that disagree, but that mutually recognize each other as epistemic peers, are able to acknowledge that, for some issues, it is extremely difficult to establish who has the epistemic authority to make claims that give others sufficient reasons for believing. ¹¹ This is especially true in a deliberative domain as the political one. Indeed, when dealing with the interpretation of factual beliefs and in many circumstances of our daily life it is very often possible to find an agreement about who are the epistemic authorities on determinate matters (e.g. a doctor for establishing how to treat a patient; an engineer for building a non-collapsing bridge; a lawyer in a courtroom). On the contrary, when dealing with disagreements about ethics, politics, religion and so on, it is extremely difficult to solve our disagreement by referring to the opinion of an expert. These are fields in which the disagreement is pervasive and in which the ability to interchange reasons publicly becomes salient, as the epistemic authority does not stem from a major expertise or specific ability, rather it is established through the public confrontation among disagreeing agents. This conclusion is compatible with the doxastic presupposition I discussed previously. Indeed, when a disagreement cannot be solved thanks to the appeal to an expert whom opinion counts as a third-person epistemic authority, then the deliberation is tied down to the reference to second-person perspectives and to the assessment of the epistemic performances provided by deliberating agents.

Might the concept of epistemic peer-hood help us in clarifying the epistemic and deliberative possibility both that agent S might be justified in believing that p and that agent Z, who had access to the same piece of evidence, is justified in believing that ˜p? This epistemic circumstance pictures a situation in which two epistemic peers, S and Z, after having acknowledged their disagreement, decide to stick to their own guns (i.e. S keeps believing that p and Z keeps believing that ˜p) and yet, each mutually recognizes the other as a valid source of justification given her and his personal doxastic attitude. This is not the standard description of the doxastic response to disagreement by epistemic peers that is provided in literature, but following Fabienne Peter (2013a, 2013b) I hold that this is the most fitting one for addressing political disagreement.

The two traditional positions with regard to peer disagreement are the Conciliatory View and the Steadfast View. According to the Conciliatory View (Christensen, 2009; Elga, 2007), the fact that an epistemic peer disagrees with me with regard to the evaluation of the same piece of evidence is a good reason to ‘bite the bullet’ and revise my belief in the attempt to find a middle ground between our opposite beliefs. By contrast, the Steadfast View (Kelly, 2010) claims that, given the fact that there is not a way to appeal to a third-person epistemic authority for solving the disagreement, it is a better doxastic response to ‘stick to my own guns’ and demote the other party epistemic position given the fact that we have good epistemic reason to trust our beliefs as long as they are ours. ¹² At the end of the day, both these two approaches do not acknowledge a genuine epistemic role for reasonable disagreement. On the one hand, the conciliatory view makes disagreement a ‘solvable’ problem, since it describes peers as willing to ‘split the difference’ (Gelfert, 2011: 508) and looking for a middle ground. On the other hand, the steadfast view describes belief formation as a unilateral process in which the fact that I disagree with someone I take to be an epistemic peer has no influence whatsoever on my doxastic justification. The conciliatory view stresses the social aspects of the belief formation processes, but puts the cart before the horse requiring that any epistemic peer will be willing to meet the other peer at the middle ground. By contrast, the steadfast view is more consistent with moral phenomenology, recognizing the extreme difficulty in solving the actual disagreement arising among agents deliberating between each other. Still, the steadfast view completely overlooks the social aspects of the deliberation, since it does not attach any relevant epistemic significance to the fact that someone I take as a peer disagrees with me. To conclude, both the views, for different reasons, reject the possibility of genuine disagreement and provide two different accounts of how epistemic peers may react to the fact that they hold different interpretations of the same piece of evidence.

According to Peter (2013a and forthcoming) what is wrong with the conciliatory view and the steadfast view is that they look at the ways in which reasonable disagreement can be addressed starting from epistemic circumstances in which both the peers have a perfect – and equal – access to the full evidence. This critique, highlighting the fact that we have to inquire the epistemic possibility of having a full access to evidence, is perfectly consistent with the fallibilist account that I defended in the previous section. Indeed, claiming that justification and warrant might contingently coincide, but not necessarily so, implies describing our doxastic access to the evidence as always mediated and partly ‘opaque’. ¹³ Therefore, I claim, at least in the political domain, we have to discuss the fact of reasonable disagreement starting from a realistic depiction of the epistemic circumstances of justice. Non-idealized circumstances of justice provide us with an epistemic context in which it is extremely difficult, if not impossible, that moral agents can reach a full disclosure of the evidence at stake. This means that moral agents have epistemic good reasons to deliberate together in order to reach a better disclosure of the evidence. Moreover, since our epistemic abilities are limited and our access to the available evidence is opaque, then looking at agents we disagree with as epistemic peers does not sound as irrational or unjustified. According to this view, reasonable disagreement might be a genuine outcome even in a case in which both the agents have been reasoning correctly and in a rational way while addressing the very same piece of evidence. The idea is that there is the possibility both for the disagreement to be solved (scenario 1) and to be reasserted (scenario 2). In scenario 1, it could be the case that a more specific examination of the evidence or a ‘purification’ from incoherence or irrationality thanks to a good deliberative procedure will grant a better disclosure of the body of evidence, providing therefore some good reasons for determining which of the two peers has the right to claim epistemic authority on such specific case and to require an agreement on her doxastic perspective (that is now supported by a more adequate propositional set of reasons). ¹⁴ In scenario 2, instead, the disclosure of evidence remains uncertain even after the deliberation – this outcome might be either related to a specific difficulty in accessing a single piece of evidence and evaluating it or stemming from the fact that it is impossible to find an agreement between the two different interpretations that the peers hold in the light of the same piece of evidence that they share (again, I do believe this is especially true in such a context as the political domain, in which the appeal to a third-person epistemic authority is very often unjustified). In scenario 2, two peers keep disagreeing even after the deliberation and, given the opacity thesis, there is no possibility to appeal to an external epistemic authority – conjointly acknowledged by both the parties ¹⁵ – to solve the disagreement. Consequently, the two parties that disagree can still be justified in ‘sticking to their own guns’ and yet the deliberation – that plays an epistemic and normative role – provides both the parties with good epistemic reasons for, at least, diminishing their confidence in their own belief and for not dismissing the other party’s belief as totally unjustified.

Once the genuine epistemic possibility for reasonable disagreement has been granted, we can wonder which deliberative framework better fits with the opacity view regarding our access to evidence. The discussion over the most adequate deliberative account brings us back to the epistemic circumstance that has prompted this section of the article, namely if it is possible that, after public deliberation on political matter, an agent S might be justified in both acknowledging agent Z as an epistemic peer and believing that while she is doxastically justified in believing that p, Z is justified in believing that ˜p according to his set of beliefs. ¹⁶ I claim that this epistemic circumstance is the perfect expression of what it would mean to be reasonable among epistemic peers. In the last section of this article, therefore, I shall connect the aspects of the epistemology of disagreement that I have outlined, with my personal account of the epistemic features of reasonableness.

IV Epistemic reasonableness

I began this article addressing the issue of public justification within the framework of political liberalism. Then, in order to analyse from the epistemic perspective the dilemmatic outcomes of the public justification enterprise, I introduced and defended a specific epistemic view regarding the doxastic perspective of moral agents and their deliberative relation as epistemic peers. Now, I want to reconnect this technical analysis with the strictly political one through to the analysis of the Rawlsian concept of reasonableness. When distinguishing between rational and reasonable, Rawls (1993: 48–58) claims that rational agents are able to pursue a conception of the good thanks to a self-centered perspective that very often (but not necessarily always) employs a means–ends way of reasoning. By contrast, a reasonable person is aware of the normative constraint of reciprocity ¹⁷ and, moreover, she acknowledges among the epistemic circumstances of justice the burdens of judgement. Recognizing the value of fairness within a cooperative context means that reasonable agents have a sense of justice, whereas being aware of the burdens of judgement implies that we are ready to accept the public reason constraints.

The fundamental line of distinction between rationality and reasonableness is that an agent might be rational even in the case of her being the only person alive on the earth, whereas for being reasonable, she needs at least a second person to be in cooperation with. ¹⁸ The ‘reciprocity constraint’ is defined by Rawls as depending upon the moral appraisal of the sense of justice. In my opinion, this is not incorrect, but in order to solve the dilemma of public justification, I claim that the principle of moral reciprocity is not enough, as alone it does not have the motivational force for prompting citizens in accepting free-standing argument as normatively binding against their own personal comprehensive beliefs. Therefore, I argue, the reciprocity constraints can be enhanced through the reference, in a context of deep political disagreement, to the normative value of epistemic reciprocity. The willingness to recognize the burdens of judgement can be reframed in a more technical way, namely claiming that those who recognize the burdens of judgement are ready to acknowledge both the validity of a fallibilist account of knowledge and the relevance of doxastic aspects in our appraisal of evidence. This means that reasonable citizens share an epistemic attitude that grants a justificatory framework that can explain the arising of reasonable disagreement, instead of simply taking it as a given fact of our political societies. Furthermore, this framework provides sound epistemic reasons in favor of the appeal to the use of public reason in certain deliberative contexts. It has been claimed that reasonable agents abide by the constraints of public reason because their sense of justice compels them to do so. Aiming at something more ambitious, I believe that it is also possible to provide strong epistemic reasons for showing to agents that, in order to be reasonable, when it comes to fundamental political questions on which the appeal to an established epistemic authority is unavailable, then they should tolerate the perspective of their fellow citizens and the beliefs that happen to be justified in those citizens’ specific doxastic set of beliefs. ¹⁹

According to an epistemic account of reasonableness, the recognition of pervasive disagreement as a stable feature of our political life together has implications both at the epistemic and the normative level. I have investigated the epistemic side of this issue in the previous section. Now, it is time to address the normative upshots of such an account. As I said before, the recognition of disagreement as a genuine possibility, even in a case in which both the parties have reasoned in a flawless way, imposes on us as epistemic agents a ‘modest’ attitude regarding the possibility of our fully accessing evidence. Indeed, if we accept fallibilism as a correct account for our epistemic capacities, then, when challenged by the disagreement with an epistemic peer, we are required – for epistemic reasons – to diminish the confidence in our own beliefs and to be tolerant of the view of the other. This means that reasonable agents would be able to abide by a sort of democratic epistemic ethos that requires them to regard someone as a political peer even in the circumstances in which there are deep epistemic inequalities and consequently epistemic performances are different. ²⁰

In my interpretation, the reciprocity constraint attached to the recognition of the burdens of judgement inherits its normativity from the fact that fallibilism gives us good reasons for recognizing other fellow citizens as epistemic peers; even if this recognition imposes a peer-hood in a very minimal sense – and naturally after that a minimal epistemic acceptability threshold has been drawn. ²¹ Once an epistemic threshold of acceptability has been drawn in order to establish which beliefs should be publicly cast out as unjustifiable, notwithstanding the fact that they might be embedded in some citizens’ sets of beliefs, then any belief that is above such epistemic threshold cannot be ruled out as totally unjustifiable. The advantage of a fallibilist characterization of peer-hood is to recognize that the equal status is not given by a potentially contestable standard of assessment of those virtues, but rather by the conditions of deliberation themselves. Two agents that disagree in a context of deliberation will at least admit that they are on a par as disagreers. This consideration has a fundamental impact on democracy, because it provides agents with good epistemic reasons to recognize other participants as potential epistemic authorities. This specific account of democracy requires citizens to address each other with reasons in the attempt to achieve an agreement about the decision to make. The interpersonal exchange among participants has an epistemic value, regardless of the fact of making or not making correct decisions. In fact, taking each other as peers implies that we should respect a symmetric credibility constraint and that therefore, in a deliberative process, we must attribute some epistemic credibility to other parties’ perspective and recognize the fact that we owe them some sort of response. ²² This assumption of credibility has an epistemic value; however, such value is still procedural, not outcome-oriented.

The epistemology of disagreement provides us with a sound epistemic argument in favor of deliberation as an adequate decision-making practice. Starting from the fact that people, given the nature of the evidence and the epistemic circumstances of belief formation processes, very often (almost every time) find themselves in an epistemic situation where there is no way to be certain about which is the correct point of view from which disagreement could be resolved, it follows that the only correct way – both from a normative and an epistemic perspective – for dealing with such disagreement is to accept the deliberative side of the epistemic processes for assessing the evidence at stake. ²³ In fact, two agents that disagree would at least have good reasons for mutually acknowledging the other party as a source of valid epistemic claims. The moderate epistemic account, therefore, stresses the fact that the epistemic authority should be shared in every circumstance in which no one can claim direct access to a third-person point of view of the evidence at stake. The willingness to share the epistemic authority expresses the fact that reasonable agents have recognized their fallibility as epistemic agents. Moreover, their accepting that they share the epistemic authority implies that agents are keen to engage themselves in a public deliberation over political matters that are undecided and on which disagreement is wide.

Historically, the normative constraints for the practice of exchanging reasons within a deliberative political context have been justified in referring to moral concepts, such as fairness (Rawls, 1993), equal respect (Larmore, 2008), acceptability constraint (Habermas, 1996; Scanlon, 1998), right to justification (Forst, 2012; Gaus, 2011), etc. One of the goals of this article is to show that we can appeal to the epistemology of disagreement in order to provide sound epistemic reasons for favoring deliberation as the most adequate political decision-making practice within a context of deep and very often insoluble disagreement among peers. If the analysis about the dynamics of mutual accountability among epistemic peers is correct, then we can provide agents with epistemic reasons for understanding why each side of the disagreement deserves some weight for its claim. Moreover, in cases in which the deliberation does not solve the disagreement – as there is no way to establish who had the right to claim epistemic authority in dealing with a specific piece of evidence – the epistemology of disagreement shows us that the adequate doxastic response is to recognize that I do not have an ultimate justification for my belief that can be recognized in any doxastic set of beliefs and that therefore can be claimed to be public and acceptable by everybody. I might be justified in keeping my own belief, but not in dismissing other parties’ beliefs as nonsensical. Consequently, tolerance might be defined as a fundamental normative upshot of the fact that epistemic peers happen to disagree strongly on political matters. ²⁴

Within the political domain, describing reasonableness thanks to the appeal to epistemic features means that reasonable citizens (i.e. agents that accept both the epistemic and moral constraints of reasonableness), when publicly discussing over political matters, will themselves abstain from appealing to p and ˜p in cases in which there is a reasonable disagreement between p and ˜p. This account of reasonableness is perfectly consistent with the public reason constraints as outlined by the public justification debate (Cohen, 2008; D’Agostino, 1996; Forst, 2012; Gaus, 2011; Habermas, 1995; Quong, 2011; Rawls, 1993). Indeed, the practice of public reason requires that citizens, when discussing publicly, respect the bounds shaped by reasonableness, namely accepting the requirement of not addressing other citizens with reasons that they do not believe, sincerely, that the others can accept or reconcile with their doxastic systems of beliefs (i.e. not to address a fellow citizen with the reasons p, when we are aware that she holds ˜p as true in her doxastic set of beliefs; see Gaus, 1996: 140). According to the democratic deliberation interpretation of public reason, the major task accomplished by public reason is to guarantee that political decisions in a democratic arena are reached following a normative framework of discussion. In this regard, the legitimation of political decisions relies in part on the normative relevance attached to the public reason’s constraint.

To conclude, the justificatory framework that I have been developing in this article grants political inclusiveness partly referring to criteria of epistemic fairness and argues in favor of an account of deliberative democracy that does not dismiss the epistemic aspects of the public practice of exchanging reasons. According to my line of argument, the normative value of reasonableness is not exhausted by the appeal to the sense of justice that liberal citizens should demonstrate having. Rather, being reasonable means as well being aware of the limitations of our doxastic processes for establishing knowledge and, consequently, being ready to recognize other citizens as epistemic peers and therefore as potential epistemic authorities. Naturally, the two aspects of reasonableness (the moral and the epistemic) are consistent and it is sensible to hope that a liberal citizen would be aware of both of them, namely showing herself or himself to have a sense of justice and a willingness to accept the deliberative reciprocity imposed by the epistemology of disagreement. However, the arguments that I developed in this article suggest that, from the motivational perspective, in order to account for the normative constraints of deliberation, it might be less demanding to refer to the mutual recognition among epistemic peers, rather than attempting to ground the reciprocity constraints solely on moral reasons. Furthermore, I claimed that clarifying the epistemic circumstances of justice is a fundamental task for understanding if and how a strictly political account of liberalism can avert dilemmatic outcomes. ²⁵ I have been arguing that it is impossible to overcome the tension between the dualistic aspects of political justification if the epistemic aspects of reasonableness are not investigated properly. The only way to provide a philosophical normative argument for justifying political theories, meanwhile granting the necessary motivational force, is to show that: (1) it is sensible, from an epistemic perspective, that agents might support a political principle for completely different, and very often incompatible with each other, private reasons; (2) the very same agents have good reasons – both moral and epistemic – for respecting other agents’ opinion and for engaging themselves in a deliberative decision-making process which acknowledges the status of epistemic peer to any fellow citizen that is able to meet some minimal epistemic conditions.

V Conclusion

The main goal of this article was to investigate the epistemic dimension of reasonableness and to show that the epistemology of disagreement provides us with an adequate theoretical framework for arguing in favor of the democratic practice of public reasoning. In order to establish public reason as the standard of deliberative procedures, I claimed that it is important to provide an epistemic framework that grants a normative democratic ethos through the analysis of the epistemology of disagreement. ²⁶ Public reason requires that citizens, when discussing publicly, respect the bounds of the normative framework of deliberation and therefore do not address other citizens with reasons that they cannot translate in ‘public reasons’ that can be understood even from a completely different doxastic perspective. This follows from the fallibilist account I have been defending, since accepting reasonable disagreement means that every citizen is required, for being actually reasonable, to recognize that some other citizens may hold undefeated beliefs in their doxastic systems that are instead defeated in her or his doxastic set of beliefs. From this account, it follows that it is correct to claim that political deliberation is a public practice that begins with disagreement, instead of agreement. To conclude, the inquiry concerning the epistemic circumstances of justice has allowed me to highlight the relevance, for political accounts of liberal theories, of the mutual recognition as epistemic peers by citizens that deliberate together on political matters. Moreover, appealing to the epistemic constraints of deliberation allows us to make sense of the epistemic difficulties derived from the public reason requirement. Hence, the epistemology of disagreement provides us with solid arguments for claiming that the normative request of democracy to grant equal respect to any participant in the decision-making process compels agents to address each other as epistemic peers. From this it follows that any decision-making procedure that denies the shared status of epistemic peer to some members of the constituency frustrates fairness and therefore is intrinsically anti-democratic.