Topoi https://doi.org/10.1007/s11245-018-9568-y

Deep Disagreement and the Problem of the Criterion Scott F. Aikin1

© Springer Science+Business Media B.V., part of Springer Nature 2018

Abstract My objective in this paper is to compare two philosophical problems, the problem of the criterion and the problem of deep disagreement, and note a core similarity which explains why many proposed solutions to these problems seem to fail along similar lines. From this observation, I propose a kind of skeptical solution to the problem of deep disagreement, and this skeptical program has consequences for the problem as it manifests in political epistemology and metaphilosophy. Keywords  Deep disagreement · Problem of the criterion · Skepticism · Metaphilosophy

1 The Problem of the Criterion The problem of the criterion is at its core a regress problem for justification.1 Sextus Empiricus, in his statement of it in his Outlines of Pyrrhonism, explicitly argues that what he calls the ‘Two Modes’ suffers from regress. He holds that things are apprehended either through themselves or through something else. Were things apprehended through themselves, we would have no controversy. Since there is controversy, we must turn to apprehending things through some second thing. However, given controversy about that by which we apprehend things, these second things must not be apprehended by themselves, either. And so, there must be a third thing through which we apprehend our second thing, by way of which we apprehend our first thing. Sextus draws the conclusion by noting: If that through which an object is apprehended must always be apprehended through some other thing, one is involved in circular reasoning or in regress ad infinitum. (PH 1.179) This result, Sextus takes as a reductio of the notion of apprehension overall. Sextus later states the problem explicitly as one of criteria: [I]n order to decide the dispute which has arisen about the criterion, we must possess an accepted criterion * Scott F. Aikin [email protected] 1



Department of Philosophy, Vanderbilt University, 111 Furman Hall, Nashville, TN 37240, USA

by which we should be able to judge the dispute; and in order to possess an accepted criterion, the dispute about the criterion must first be decided. (PH 2.20) Notice that Sextus’ program begins with the thought that there is a dispute about some object of judgment, so the program is primarily dialectical.2 However, he also states a version of the problem that is a purely epistemically procedural matter: If he is to pass judgment on the impressions, he must certainly judge them by a criterion; this criterion, then, he will declare true […] If he shall declare it true, he will be stating that the criterion is true either without proof or with. But if without proof, he will be discredited; and if with proof, it will certainly be necessary for the proof to also be true […] For the proof always 1

 See Hankinson (1995), Cling (1997 and forthcoming), and Thorsrud (2009) for accounts of the problem of the criterion as an instance of the regress problem of justification. The connection between the problem of the criterion and the five modes of Agrippa, then, is that the five modes can be applied to the question of what the proper criterion is. 2  For accounts of the Pyrrhonist program as primarily dialectical, see Vazquez (2009), Lammenranta (2008, 2011), and Hazlett (2014). This is a question of how essentially dialectical the skeptical challenge is here, and how central the dialectical considerations are to the procedural epistemic version of the problem of the criterion. See Michael Williams (1999, 2003) for the case that the skeptical challenges arise from a misreading of dialectical norms. Michael Rescorla (2009) holds that Sextus is right about the dialectical norms, but is wrong about whether they provide an adequate model for epistemic structure. For a defense of Pyrrhonian dialectical norms as models for epistemic justification, see Aikin (2011 and 2014b).

13

Vol.:(0123456789)

S. F. Aikin

requires a criterion to confirm it, and the criterion is also a proof to demonstrate its truth […] (PH 1.114-6) We need a criterion of truth in order to judge between true and false impressions, but those criteria must be shown to be true or reliable, and so we must have a criterion for judging the true criteria from the false. And so a vicious regress or circular reasoning ensues.3 Contemporary versions of the problem of the criterion take on an epistemically procedural mode of presentation. Roderick Chisholm’s first-order version of the problem is procedural in this sense, as he notes (paraphrasing Montaigne): To know whether things really are as they seem to be, we must have a procedure for distinguishing appearances that are true from appearances that are false. But to know whether our procedure is a good procedure, we have to know whether it really succeeds […] And we cannot know whether it really succeeds unless we already know which appearances are true and which ones are false. And so we are caught in a circle. (1973, p. 3) Chisholm’s proceduralist version of the problem runs on two principles of backing for our commitments: 1. Subject S is justified in sorting the true from false appearances only if S has a criterion C for sorting and S is justified in holding that C is a good criterion. 2. S is justified in holding that C is a good criterion only if S is justified in sorting true from false appearances. Chisholm and Sextus both capture this notion of epistemic procedural priority by using temporal terms like ‘already’ and ‘first,’ but the point is about epistemic priority—one is a necessary condition for the other.4 A good criterion for proper sorting is epistemically prior to and necessary for properly sorting true appearances from false, but properly sorting truths and falsities is epistemically prior to identifying whether a criterion has done its job of sorting well. And so, if two matters are epistemically prior to and necessary conditions for each other, then they must fail to obtain. Two assumptions must be in place for the skeptical conclusion to follow. First, that circular justifications are not justification affording. Second, that no infinite series of justifying reasons could satisfy the requirement of justifying a

criterion.5 And so the following would be disallowed for S, because of its circularity: (i.) P is held true on the basis of it being endorsed by criterion C.   Q: On what basis do we hold that C is a good criterion? (ii.) C is held as a good criterion on the basis of its successfully sorting truths from falsities—it sorts P (as true) from not-P (false).   Q: How is P known to be a truth? (iii.) P is held true on the basis of its being endorsed by criterion C.   Q: How is P known to be a truth? Alternately, the following would be disallowed by the view that infinite series of reasons are not justification-affording: (iv.) P is held true on the basis of its being endorsed by criterion C.   Q: On what basis is it held that C is a good criterion? (v.) C is endorsed by another criterion D.   Q: On what basis is D a good criterion? (vi.) D is endorsed by criterion E.   Q: On what basis is E a good criterion? … So long as criteria and the truths they endorse do not bootstrap their way to justification or that a series of nonrepeating criterial endorsements can only transmit, but not produce, justification, then no belief is justified, and by extension, no disputes are rationally resolved. That is the problem of the criterion.

2 Deep Disagreements The problem of deep disagreement arises first from an observation about what may be called ‘normal’ disagreements. If, say, two people disagreed about when their favorite team last won the championship, they may argue it out, tracing their reasons to some shared set of commitments in order to determine who is right. And so, both will remember that the Hornblowers won the cup the year before the Magpies did, and the Magpies won the cup just days before that big political event in 2015. So it must have been 2014. Alternately,

3

  That the Agrippan five modes are relevant here is clear, since both depend on a similar backing requirement for the judgments at issue (see Sextus PH 1.164-177). 4  Amico (1988, 1993) and Cling (forthcoming) have made it clear that the idiom of temporalizing epistemic procedure is a way of indexing epistemic priority.

13

5

 These considerations are surveyed as background principles for running most versions of the regress problem in epistemology. See Aikin (2011) and Cling (2009) for how these principles are supposed to be justified.

Deep Disagreement and the Problem of the Criterion

they could agree that they could consult the team’s or the league’s website. In these cases, there are identifiable shared premises or warranting rules from which to derive content which resolves the disagreement or there is an agreed-upon source or procedure for producing the resolving content. These are normal disagreements. Deep disagreement arises when there are no uncontroversial premises for deriving resolving content and no uncontested sources or procedures for producing it, either. Under these conditions, it seems that arguments, and even the procedures we use to assess and critique arguments, cannot work. As noted by Robert Fogelin in “The Logic of Deep Disagreement,” “the possibility of a genuine argumentative exchange depends […] on the fact that together we accept many things” (1985, p. 6). Consequently, deep disagreement cases “cannot be resolved through the use of argument, for they undercut the conditions essential to arguing” (1985, p. 8). At its core, the problem of resolving deep disagreements takes the form of a regress problem, one that has a structural overlap with what was revealed to be the core mutually excluding necessary conditions driving the problem of the criterion. One of the principles behind the challenge of deep disagreement is a procedural commitment about dialectical priority. If we assume that the ‘sharing’ of commitments and procedures Fogelin invokes is between arguers who take turns giving arguments to each other, we may identify them as alternating between roles of speaker (S) and hearer (H). What arises, given this turn-taking element of argument, is what I call the Dialecticality Requirement for speakers: DR1 Arguments are good only if they are dialectically appropriate for their audiences. DR2 Speaker S’s argument (A) is dialectically appropriate for H only if: a. A’s premises are accepted or acceptable to H, b. The support that A’s premises provide the conclusion is accepted by, or acceptable to, H, and c. S has, in proposing A, addressed any of H’s subsidiary objections or concerns.6 Given DR1 and 2, we have a way of showing that, in deep disagreement cases, we have either regresses or circularity. To show this, imagine the following scenario. S proposes proposition P, and H rejects it. S then produces an argument 6  The Dialecticality Requirement is recognized widely within the informal logic and argumentation theory literature as a default rule for arguments. See as exemplary: Johnson (2000), Pinto (2009), Tindale (1999), Walton (1989), and van Eemeren and Grootendorst (2004).

A′ in support of P, which has a premise Q. H rejects premise Q, so S then produces another argument A″, with a premise R. H rejects R, so S proposes a third argument A′′′, with a premise S, and so on. Alternately, we could do this with H rejecting the support relation between the premises and conclusion, or bringing out defeaters (so, invoking DR2-b, and DR2-c respectively). On the assumption that the disagreement is deep, any finite, non-repeating series of arguments will fail the Dialecticality Requirement. And so, the only two options structurally left are a series of arguments with premises that repeat between arguments (and so arguments that are structurally circular), or a non-repeating, non-finite series (and so, a regress ad infinitum). Assuming that circular arguments are dialectical failures and infinitely long argument series defeat the purpose of timely argument resolution, it seems that, in deep disagreement cases, there can be no dialectically acceptable resolution of the disagreement. In the same way that the problem of the criterion has mutually-cancelling necessary conditions, so does the argumentative problem of deep disagreement. (1) S has resolved a disagreement (about the acceptability of P) with H only if S has provided dialectically satisfying arguments for H that P. (2) S has provided dialectically satisfying arguments for H that P only if S has resolved a disagreement with H (about the acceptability of C, as a criterion for the acceptable resolution of P). In this regard, the phenomenon of deep disagreement is best seen as an instance of what I’d termed earlier Sextus’s dialectical versions of the problem of the criterion—insofar as there is persistent controversy, argument cannot resolve what is at issue.

3 Parallel Challenges Deep disagreement cases are special instances of the dialectical form of the problem of the criterion. Consequently, the logical form of the problem of the criterion, that of mutually denying necessary conditions, explains why deep disagreement cases are challenging in the same fashion the problem of the criterion is. As a consequence, purported solutions to the two problems will have structural similarities, and so will problems that dog these solutions. One proposed solution to the problem of the criterion is to suspend one of the backing requirements for justification. On Chisholm’s model, there are two forms for this strategy: what he calls Particularism and Methodism. The Particularist holds that we begin with a set of truths or a number of unproblematic cases of knowledge, and from these, we

13

S. F. Aikin

“work out” what kinds of criteria and procedures are reliable in terms of these (1973, p. 20). Alternately, the Methodist begins with a source or procedure from which information is derived, and then the variety of truths we endorse may be produced. And so, one may “begin with the criterion,” and start sorting the truths from falsities (as does the Methodist), or one may start with the given truths and falsities and from these work up to rules (as does the Particularist). Either way, we have suspended one form of the backing requirement, because we simply start with one set of commitments without any further reason. The same goes for deep disagreement cases, as one may start with a set of commitments beyond which one will not see another as a target of rational persuasion, or one may take some sources of information or specific arguments as sacrosanct. Just as Methodists and Particularists break with backing, this strategy breaks with the Dialecticality Requirement for reasons and arguments. The quality of an argument, on this view, is not dependent on the audience’s acceptances or commitments. And so it overcomes what Michael Huemer, in the context of moral disagreement, calls “the idiot’s veto”: “Someone’s mere failure to accept the conclusions of an argument does not constitute a flaw in the argument” (2005, p. 131). Facing a similar challenge, Richard Rorty claims that we must “grasp the ethnocentric horn of this dilemma. We should say that we must, in practice, privilege our own group, even though there can be no non-circular justification for doing so” (1991, p. 29). And Jeremy Barris holds that “various standard fallacies are legitimate elements in the negotiation between globally different frameworks” (2015, p. 70).7 The trouble many see with suspending the backing requirement for the problem of the criterion extends to the suspension of the dialecticality requirement for deep disagreement. First, suspending backing for our commitments and beliefs seems to steal whatever resolution there is in assent. That is, if you were to assent to something, that means that you’ve committed to something, and not something else. Doing so explicitly without evidence is less like commitment, and more like saying “whatever.” Imagine yourself saying this of any of your beliefs:

I believe all the things told to me by S, but I have no reasons to believe that S is a good source. Such statements seem to sap our assents of the commitment that make them assents. The same goes for belief. Our beliefs, when seen from this perspective, seem more to be things that have befallen us, instead of things that we’ve accepted rationally. Such states, perhaps, can be explained, but explanations are not justifications.8 The same, as I see it, goes for the Dialecticality Requirement. Return to the problem of the idiot’s veto from before. In suspending the Dialecticality Requirement, we must think that statements of the following form are fine: My arguments for P are good, but those with whom I disagree are not moved by them. Or take Richard Rorty’s ethnocentric orientation of overt reasoning: My arguments for P are good, but they move only people in my particular culture. Or take Jeremy Barris’s suspension of the prohibition on fallacious arguments: My arguments for P are good, but I must argue fallaciously when posing them to those with whom I deeply disagree. Again, I’m inclined to say that when we say these sorts of things to ourselves, we should feel uncomfortable with our assessments of our arguments.9 This is for two reasons. For one, the idiots should be easier to answer, not harder. This is to say that insofar as one’s dialectical opponent is missing things that are progressively more and more obvious and clear, it should be within one’s power to make it explicit what has been missed—one masters the obvious in these exchanges, and this project, though time-consuming, is a matter more of patience than intellectual difficulty. The second reason to be uncomfortable with the dismissive program is that it seems that on this form of reasoning, the wider the intellectual gap between two discussants, the less relevant their disagreement can be for the other.10 And so, proceeding in this fashion with deep disagreement makes

I believe that P, but I have no evidence supporting it. Or

7   Others who see the deep disagreement cases as temporary hold that there can be various non-argumentative means of persuasion used so that those on the other side can be brought around with: emotional appeal (Freeman 2005), non-doxastic showing (Davson-Galle 1992), educational programs (Turner and Wright 2005; Duran 2016), or polemical programs of confrontation (Kraus 2010).

13

8

  See Adler (2003), Talisse (2009) and Aikin (2014a) for versions of this thought that Moore’s Paradox provides support for evidentialism. 9  See Aikin 2014b for a reply to Huemer on the idiot’s veto with moral reasons. 10  The moral hazards of arguing in deep disagreements have been surveyed by Campolo (2005, 2013). The concern is that we develop bad argumentative habits in such circumstances. I agree with Campolo that there are great temptations, and in some ways, I believe that Barris’s endorsement of fallacies under these conditions is evidence that Campolo is right. However, the question is whether not arguing also has (perhaps greater) moral hazards.

Deep Disagreement and the Problem of the Criterion

it so that one may only really argue with those with whom one already agrees, and those with whom one disagrees significantly, one may simply ignore. Call this rule the norm of dialectical neglect. Notice that this norm of neglect is symmetric—it is permissive for both sides of the disagreement. And so, though we may have the comfort of saying our arguments are just fine, even though the idiots are not moved by them, the idiots think the same of their own arguments (and they likely call us the idiots). And if they follow the other advice from Rorty, they will feel no compulsion at all to address any arguments to us, and if they do, they won’t be worried about them taking the form of fallacies. Surely we can appreciate what’s objectionable about the norm of neglect when we are the ones on the receiving end of the dialectical neglect it endorses. What arises is a lesson about identifying acceptable norms for argumentative exchange—they must not only be norms one will agree to as an arguer, but also as a hearer. Argumentative exchanges are turn-taking games, and so when we make rules for these games, we must remember that we are not just making these rules for ourselves when we imagine ourselves in the role of arguer or speaker, but we must also think of what norms one must think are acceptable for those who are addressed by those arguments. This is because we will, for every bit of dialectic, be on the receiving end of arguments in addition to being on the giving end. And so, the norm of dialectical neglect does not pass this test of symmetry—that we would make a rule of argument that we could imagine ourselves as both speaker and hearer endorsing. This, I believe, is the basis for the Dialecticality Requirement. Arguments of the type I’ve just given are arguments from cognitive hygiene for both the backing norm and the Dialecticality Requirement. They are, like hygiene, just good rules to follow in maintaining intellectual health, similar to physical health. We should eat right, exercise, and get plenty of sleep. By analogy, we should be able to survey supporting reasons for our beliefs, know what our critics say of our views, know how those with whom we disagree justify theirs, and find out more about what other alternatives there are. But there is another argument that intersects with the cognitive hygiene argument to support the dialecticality requirement in particular. Call it the argument from moral recognition. In essence, the thought behind the recognition argument is that argument must be dialectical if we are to recognize the moral considerability of our audiences and the blameworthiness of the things they say or do as a consequence of our disagreement. If we blame someone for doing something, we are not only holding that what they have done is bad, but that they had reasons not to do it. Moreover, we must hold that those reasons are accessible to them. So

blame attaches not just to bad actions, but to actions we see as rationally negligent. Consider: those who had no reason available to them not to perform some action may have done something bad, but they are not to blame for that action. For example, a driver who couldn’t see a dog running across the street isn’t to be blamed for hitting the dog. Or take the person who leaves hedge clippers in the front yard when taking a break from gardening—only if she had reason to worry that the toddlers visiting next door could get loose and use them is she blameworthy for leaving them out. It is certainly the case that these agents could be negligent for not taking the proper care to acquire information or reflect on their respective situations, and so could be blameworthy for that negligence. But if they did not have access to those reasons in the first place, then it is hard to see why blame is appropriate—we can recognize the badness of the actions or their consequences and take steps to inform them of that badness. But blame seems appropriate only if those reasons were accessible to the agent performing the action. Consequently, if we are finding someone blameworthy for their actions or beliefs, we must hold that some reasons are accessible for them, and so arguments are possible. The possibility of dialectically successful argument, then, is a precondition for having standing to blame. The point, then, is that insofar as we can see those with whom we disagree as being blameworthy for their beliefs and actions, we must see them as having accessible reasons not to believe and act the way they do. And so it is to those reasons that we must appeal when we argue with them.11 The consequence is that we have two convergent arguments for keeping the dialecticality requirement, and from them we see the costs of rejecting it. Here they are in short form: Cognitive Hygiene Costs: If we reject the Dialecticality Requirement, we undercut the rational credence constitutive of assent and belief, and we allow those with whom we deeply disagree to feel no compulsion to address us in ways that would seek our rational credence. Moral Recognition Costs: If we reject the Dialecticality Requirement, we forego the necessary condition of moral recognition for blaming those with whom we deeply disagree for the bad things they do or believe. I think that in the process of thinking through solutions to the problem of deep disagreement, a desideratum of a solution has emerged. It is that any solution must not only be

11

  This requirement is a form of ought implies can for blaming, but in this case, it takes the form of ought implies can have reasons to. See McKiernan (2016) for a full-dress version of this reasons-accessibility requirement for blame.

13

S. F. Aikin

something we can be comfortable in taking on for ourselves, but it must also be something we can be comfortable with if those with whom we disagree follow it. That is, solutions to the deep disagreement problem must be symmetric; they are solutions not just for us, but also for those with whom we disagree. One reason, I believe, the temptations to suspend the symmetry of the dialecticality requirements are so tempting is the thought that the situation is actually asymmetric. The asymmetry thought begins with the fact of commitment— one has one’s commitments, and being committed means one thinks those commitments are substantively true. If one is substantively right about the issues, then the temptation is to think the procedural rules that get in the way of what is correct must be objectionable burdens. The trouble is that, not only is being right not enough to know, being right is not enough to have a good argument. The requirement of reasons in these cases is one that not only picks out truths, but it makes it clear why we pick out these truths, and not others. Criteria and arguments are not just about their target truths, but they are about the sorting or discerning role they play in targeting those truths. And the discerning role of argument in particular is that of sorting many views and considerations for and against them. Arguments are necessary and arise because of identified issues of dispute, and issues of dispute are identified when conversants hold contrary views. Argument must not only make it clear why one view ought to be favored over another, but it must do so on the basis of reasons which those not committed to the favored view can appreciate as reasons, else argument is not sorting in a way that actually resolves the disagreement—the disagreement would just migrate from the issue up to the purportedly resolving argument. And so, ultimately, the cognitive hygiene and moral recognition arguments converge with a version of a pragmatic take on the dialecticality requirement—if argument is to be effective at rationally resolving disagreement, it must be dialectical.

forward is to ask whether there are less vicious ways to beg the question than others. Perhaps, in the same way that reasons reach a reflective equilibrium between our substantive commitments and our procedures, so our arguments may demonstrate the coherence of a viewpoint. Consequently, it is not in any particular act of overt reasoning that one displays the means for resolving a dispute, but it is in the responsible deployment of those coordinations over time that we show the rationality of a view and the viewpoint it manifests.13 Insofar as we are fallibilists, we must be able to say that it is possible to be rational and wrong. So we must take that same forgiving attitude to the acts of reasoning others display, and in doing so, we must focus on what, internal to the reasons they see as relevant, makes their reason-coordination rational.14 Focusing on particular deep disagreement cases, one thing that is revealed is that the reason why prohibitions on suspending the Dialecticality Requirement are symmetric is because both sides wish to be treated as rational agents— the kind of beings that are to be moved by reasons only, to move another otherwise is coercion or oppression. We, then, argue with each other because we see each other as rationally self-regulating beings. With argument, then, we recognize each other as morally considerable and rational. And it is in recognizing each other’s rational self-regulation that the Dialecticality Requirement is rooted—we respect intellectual autonomy. The lesson is that when we face deep disagreement problems for argument, we are facing conditions where there is not sufficient agreement to resolve the debate, but there must (insofar as it is an argumentative situation) be some significant overlap. This can come out as a transcendental argument:

4 Dialecticality and Symmetry

And so, though deep disagreements are those wherein we disagree about many, many things, insofar as they are

If I am right, there are no obviously non-question-begging or regressive ways out of the problem of the criterion or the problem of deep disagreement.12 One thought about ways

12   I have, admittedly, run past a variety of alternatives here, ranging from forms of infinitism as a response to the problem of the criterion (exemplified by the Hegelian line, as outlined in Aikin 2010) and forms of reflective equilibrium with respect to criteria and judgments (as exemplified by DePaul 1986 and; McCain and Rowley 2014). Both of these forms, on my reading, are more skeptical or management approaches to the problem of the criterion than solutions to the problem.

13

The condition for the possibility of two sides addressing arguments to each other is that both sides see each other as rational and thereby both resolve to move each other only by reasons they themselves see as reasons.

13

  Claudio Duran has posed a program of identifying just how deep disagreements are, with the hope of showing that some are only temporarily intractable (2016), and Teresita Matienzo has proposed a strategy of literary bridges between competing traditions in deep disagreements (2014). Godden and Brenner (2010) provide a model for these transitions as making a case for and using a set of (sometimes new) concepts. 14   One consequence may be that a set of meta-argumentative commitments may arise from this observation, as noted by Finocchiaro (2013). Godden (2013) holds that such a program is likely to lead to “more of the same;” however, when these arguments are posed as internal strategies, there may be room for rapprochement.

Deep Disagreement and the Problem of the Criterion

argumentative contexts, we must nevertheless share something—a respect for paths of reason that the other can travel. The ironic turnaround is that if we can mutually recognize a deep disagreement, we must not share enough to rationally resolve the issue, but we nevertheless share enough to see ourselves both bound by the Dialecticality Requirement (and the ethic of symmetry of the argumentative norms). That’s not much, but that’s not nothing.

5 Consequences Insofar as we can appreciate the skeptical upshot of both the problem of the criterion and the problem of deep disagreements, we can see the importance of the backing requirement for knowledge and its cousin, the dialecticality requirement for argument. Further, if we appreciate the dialecticality requirement as symmetric, then we must see the respect for rational autonomy as a binding norm for argumentatively interacting with others. There are, then, political epistemic consequences of this arrangement. One political upshot is a case for familiar liberal constraints on public reason. When we deliberate, we are directing our arguments and reasons toward determining the common good and the means to it.15 But these reasons must be public in the sense that they must be shareable by all the relevant discussants, since to propose and enforce a good for a citizen, even if it truly is a good, that the citizen does not see as a good is oppression. John Rawls terms this the duty of civility, and it is pursuant to the thought that state force is legitimate only if it is in accord with ideals that people can endorse.16 Democratic deliberation, then, requires that we argue from a shared public perspective, and refusing to abide by that norm is a failure to treat other citizens as equals. For Rawlsians, comprehensive doctrines must be off the table for public deliberations, because reasonable people reject them.17 The problem of political polarization strikes not only at the heart of our capacity to access reasons from (and give accessible reasons to) the other side, but as we see these reasons less and less as reasons, we see the others themselves as less and less recognizably reasonable. That is, polarization is not just a phenomenon of the two sides of a disagreement growing further and further apart in their respective views, but it has further consequences of not accessing reasons the other side gives as reasons. The other side overwhelmingly sounds in one’s ears less as communication and more as

noise. Consequently, the deeper the disagreements get, the harder they are to even recognize as disagreements, and the more they tend to appear simply as clashes. The great irony is that deep disagreements and the polarization that yields them are particularly endemic to liberal democracies, since with the protections of freedom of conscience, we yield further and further proliferation of views and discourses. And so disagreements are made wider and deeper. Further, deep disagreements will be more prevalent and discoverable in democracies, since with speech freedoms, we can express and develop further a breadth of views and uncover more and more widespread items of controversy. And finally, deep disagreement is most problematic for liberal democracies, since if we are to proceed according to the rule of public reason, then deep disagreements threaten to make resolution impossible. But let me close with a hopeful note. One thing that regularly arises from deep disagreements is a series of progressively philosophical questions. That is, as disagreements get deeper, the matter regularly becomes one of the proper semantics of a term—justice, truth, God, knowledge, fairness, perhaps. Philosophy is, it seems, a place where we face the depth and interminability of many, many disagreements, and with philosophical training, it seems we have many ways to identify, manage, and make progress on thinking things through. Often that progress isn’t in the form of getting closer to a resolution, but rather it is identifying the connections between commitments on the various sides, the way the reasons in the debates hang together. And so, if you’ve accepted the skeptical solutions to these problems, like Sextus, it doesn’t follow that you give up on philosophical argument just because the arguments don’t seem to ever yield resolution. Rather, we pursue philosophical argument as a kind of end in itself. Sextus cast his Pyrrhonian skepticism as a zetetic or searching program of philosophy—one that concedes that the path to answers is not readily given, but that does not guarantee that there are no answers. We can still argue in deep disagreement cases, as is shown by the longstanding well-run debates in philosophy. The path forward is making sure we manage those debates well in philosophy, set good examples for our students and the many onlookers to our dialogues, pause to make sure our arguments reach the audiences they purport to address, and make sure we hear their arguments and objections, too.18

Compliance with Ethical Standards  Conflict of interest  I do not have any Conflicts of Interest, Funding Sources, or testing ethics statements to submit beyond the following:

15

  Rawls (1971, pp. 360-1 and 1997, p. 96) and Cohen (1997, p. 68).   Rawls (1997, p. 96). 17   This view is defended as an argumentative rule of public reason in Aikin and Talisse 2014. 16

18

  Thanks to John Casey, Andrew Cling, David Godden, Robert Talisse, the anonymous reviewers at Topoi, and the audience at Michigan State University for comments on earlier drafts of this paper.

13

S. F. Aikin Ethical Approval  This article does not contain any studies with human participants performed by any of the authors.

References Adler J (2003) Belief’s own ethics. MIT Press, Cambridge Aikin S (2010) The problem of the criterion and Hegel’s model for epistemic infinitism. Hist Philos Q 27:379–388 Aikin S (2011) Epistemology and the regress problem. Routledge, New York Aikin S (2014a) Evidentialism and the will to believe. Bloomsbury, London Aikin S (2014b) Prospects for moral epistemic infinitism. Metaphilosophy 45:172–181 Aikin S, Talisse R (2014) Why we argue (and how we should). Routledge, New York Amico R (1988) Roderick Chisholm and the problem of the criterion. Philos Pap 17:217–229 Amico R (1993) The problem of the criterion. Rowman and Littlefield, Lanham Barris J (2015) Metaphysics, deep pluralism, and paradoxes of informal logic. Int J Philos Stud 23:59–84 Campolo C (2005) Treacherous ascents: on seeking common ground for conflict resolution. Informal Log 25:37–50 Campolo C (2013) Argumentative virtue and deep disagreement. OSSA 10 conference archive, paper 32 Chisholm R (1973) The problem of the criterion. Marquette Univerisity Press, Milwaukee Cling A (1997) Epistemic levels and the problem of the criterion. Philos Stud 88:109–140 Cling A (2009) Reasons, regresses, and tragedy. Am Philos Q 46:333–346 Cling A (Forthcoming) Procedural reasons and the problem of the criterion. In: McCain K, Poston T (eds) The mystery of skepticism. Brill, Leiden Cohen J (1997) Deliberation and democratic legitimacy. In: Bohman J, Rheg W (eds) Deliberative democracy. MIT Press, Cambridge Davson-Galle P (1992) Arguing, arguments, and deep disagreements. Informal Log 14:147–156 DePaul M (1986) Reflective equilibrium and foundationalism. Am Philos Q 23:59–69 Duran C (2016) Levels of depth in deep disagreement. In: Bondy P, Benacquista L (eds) Argumentation, objectivity, and bias: OSSA 11 proceedings. OSSA archives Finocchiaro M (2013) Deep disagreements: a meta-argumentative approach. In: Zenker F (ed) Argumentation: cognition and community. Proceedings of the 9th OSSA Fogelin R (1985) The logic of deep disagreement. Informal Log 7:1–8 Freeman F (2005) Emotional backing and the feeling of deep disagreement. Informal Log 25:51–63 Godden D (2013) Commentary on Finocchiaro. In: Zenker F (ed) Argumentation: cognition and community. Proceedings of the 9th OSSA Godden D, Bronner W (2010) Wittgenstein and the logic of deep disagreement. Cogency 2:41–80

13

Hankinson RJ (1995) The sceptics. Routledge, New York Hazlett A (2014) A critical introduction to skepticism. Bloomsbury, London Huemer M (2005) Ethical intuitionism. Palgrave, New York Johnson R (2000) Manifest rationality. Lawrence Erlbaum, Mahwah, NJ Kraus M (2010) Cultural diversity, cognitive breaks, and deep disagreement: polemic argument. In: van Eemeren F, Garssen B, Godden D, and Mitchell G (eds) ISSA 2010 proceedings Lammenranta M (2008) The Pyrrhonian problematic. In: Greco J (ed) The Oxford handbook of skepticism. Oxford University Press, New York Lammenranta M (2011) Disagreement, skepticism, and the dialectical conception of justification. Int J Study Skept 1:3–17 Matienzo T (2014) Desacuerdos profundos y persuasion literaria. In: Arroyo G, Matienzo T, Marafioti R, Santibáñez Yáñez C (eds) Esporando el desacuerdo: epistemología, cognición, y sociedad. Universidad Nacional de General Sarmiento, Buenos Aires McCain K, Rowley W (2014) Pick your poison: beg the question or embrace circularity. Int J Study Skept 4:125–140 McKiernan A (2016) Standing conditions and blame. Southwest Philos Rev 32:145–151 Phillips D (2008) Investigating the shared background required for argument: a critique of Fogelin’s thesis on deep disagreement. Informal Log 28:86–101 Pinto R (2009) Argumentation and the force of reasons. Informal Log 29:268–295 Rawls J (1971) A theory of justice. Harvard University Press, Cambridge Rawls J (1997) The idea of public reason. In: Bohman J, Rheg W (eds) Deliberative democracy. MIT Press, Cambridge Rescorla M (2009) Epistemic and dialectical regresses. Australas J Philos 87:43–60 Rorty R (1991) Solidarity or objectivity? Objectivity, relativism, and truth: philosophical papers, vol 1. Cambridge University Press, Cambridge Sextus E (1990) Outlines of pyrrhonism. Bury RG (trans) Prometheus Books, Amherst. Referenced in text as (PH. Book.Page) Talisse R (2009) Democracy and moral conflict. Cambridge University Press, Cambridge Thorsrud H (2009) Ancient scepticism. University of California Press, Berkeley Tindale C (1999) Acts of arguing. SUNY Press, Albany Turner D, Wright L (2005) Revisiting deep disagreement. Informal Log 25:25–35 van Eemeren F, Grootendorst R (2004) A systematic theory of argumentation. Cambridge University Press, Cambridge Vazquez D (2009) Reason in check: the skepticism of sextus empiricus. Hermathena 186:43–57 Walton D (1989) Informal logic. Cambridge University Press, Cambridge Williams M (1999) Skepticism. In: Greco J, Sosa E (eds) The Blackwell guide to epistemology. Blackwell, Malden Williams M (2003) The Agrippan argument and two forms of skepticism. In: Sinnott-Armstrong W (ed) Pyrrhonian skepticism. Oxford University Press, New York