Disagreement as Evidence: The Epistemology of Controversy

How much should your confidence in your beliefs be shaken when you learn that others – perhaps ‘epistemic peers’ who seem as well-qualified as you are – hold beliefs contrary to yours? This article describes motivations that push different philosophers towards opposite answers to this question. It identifies a key theoretical principle that divides current writers on the epistemology of disagreement. It then examines arguments bearing on that principle, and on the wider issue. It ends by describing some outstanding questions that thinking about this issue raises. If you’d like to make a professional philosopher uncomfortable, try asking for clear examples of our discipline’s achievements in settling the questions we study: Katie Couric: I’m just going to ask you one more time – not to belabor the point. Specific examples, in the last 2600 years, of important philosophical questions settled... Philosopher (visibly straining to look upbeat): I’ll ... try to find ya some, and I’ll bring ’em to ya! I suspect that many philosophers have a bit of a guilty conscience about this issue. Of course, the worry is not about any dearth of philosophers with firm opinions on the great questions. It is about how few of these opinions have, over the years, achieved anything like consensus. Lack of consensus might well – at least in certain conditions – be taken as evidence that the parties to the dispute lack good reason for confidence in their positions. Nevertheless, the issue has not attracted much serious attention in mainstream epistemology until recently. Clearly, the question of how to react to disagreement by apparently well-qualified others is not one whose import is restricted to philosophy. On many factual questions – examples from politics or economics are easy to think of – widespread disagreement, even among experts, is the norm. Contemporary responses to this issue may be roughly arrayed along a spectrum. At one end are views on which the disagreement of others should typically cause one to be much less confident in one’s belief than one would be otherwise – at least when those others seem just as intelligent, well-informed, honest, free from bias, etc. as oneself. Following Elga (forthcoming), I’ll label this the ‘Conciliatory’ end of the spectrum. At the other end are views on which one may typically, or at least not infrequently, maintain one’s confidence in the face of others who believe otherwise, even if those others seem one’s equals in terms of the sorts of qualifications listed above. Let us call this the ‘Steadfast’ end of the spectrum. Much of the recent discussion has centered on the special case in which one forms some opinion on P, then discovers that another person has formed an opposite opinion, where one has good reason to believe that the other person is one’s (at least approximate) equal in terms of exposure to the evidence, intelligence, freedom from bias, etc. (Such a Philosophy Compass 4/5 (2009): 756–767, 10.1111/j.1747-9991.2009.00237.x a 2009 The Author Journal Compilation a 2009 Blackwell Publishing Ltd person is often referred to as one’s ‘epistemic peer’.) In addition, discussions typically assume that one has no special reason – such as knowing that the other is drunk or tired – to discount her opinion. Of course, the interest of this sort of special case lies largely in the light it can shine on the more general question of how one’s beliefs should be affected by one’s knowing the opinions of others. 1. Motivations for Conciliationism The main motivation for Conciliatory positions on disagreement can be illustrated simply in a case where one comes to believe P on the basis of certain evidence, and learns that one’s apparent epistemic peer has reached the opposite conclusion on the basis of the same evidence. It begins with two thoughts: that the peer’s disagreement gives one evidence that one has made a mistake in interpreting the original evidence, and that such evidence should diminish one’s confidence in P. This does not yet speak to how conciliatory one should be, but the following sort of example supports the claim that (at least in certain cases), one should be highly conciliatory, suspending belief (or, in graded terms, adopting a middling credence) in response to disagreement. Mental Math: You and your friend have been going out to dinner together regularly for many years. You always tip 20% and split the check (with each person’s share rounded up to the nearest dollar), and you each do the requisite calculation in your head upon receiving the check. Most of the time you have agreed, but in the instances when you have not, you have taken out a calculator to check; over the years, you and your friend have been right in these situations equally often. Tonight, you figure out that your shares are $43, and become quite confident of this. But then your friend announces that she is quite confident that your shares are $45. Neither of you has had more wine or coffee, and you do not feel (nor does your friend appear) especially tired or especially perky. How confident should you now be that your shares are $43? Many people agree that in this sort of case, strong conciliation is called for: you should become much less confident in $43 – indeed, you should be about as confident in $45 as in $43. There are, of course, hard questions about how far the conciliatory response that seems correct here will generalize. The Conciliationist’s thought is that it will generalize widely: the claim is that, in many controversies, participants on each side have good reason to think that they are as likely as those on the other side to have gone wrong; thus they should become much less confident in their opinions. A related route to motivating Conciliationism in fields like philosophy, where disagreement is rife, derives from the point that reliable methods of inquiry must tend to produce agreement. Thus the persistence of the degree of disagreement on important issues we see in fields like philosophy indicates that, in general, practitioners in the field do not form beliefs reliably. If one is a practitioner in such a field, then, absent some reason to think oneself special, one should not have confident opinions on the field’s controversial questions. 2. Motivations for Steadfastness The most obvious motivation for Steadfast views on disagreement flows from the degree of skepticism that Conciliationism would seem to entail. There must be something wrong, the thought goes, with a view that would counsel such widespread withholding of belief. If you have an opinion on, for example, compatibilism about free will, scientific realism, or contextualism about knowledge, you must be aware that there are very intelliThe Epistemology of Controversy 757 a 2009 The Author Philosophy Compass 4/5 (2009): 756–767, 10.1111/j.1747-9991.2009.00237.x Journal Compilation a 2009 Blackwell Publishing Ltd gent and well-informed people on the other side. Yet many are quite averse to thinking that they should be agnostic about all such matters. The aversion may be even stronger when we focus on our opinions about politics, economics, or religion. The thought that steadfastness is appropriate in certain particular sorts of cases is even stronger. Consider: Careful Checking: The situation is as in Mental Math, but this time you do not do the arithmetic in your head. You do it carefully on paper, and check your results. Then you do it in a different way. Then you take out a well-tested calculator and use it to do and check the problem a few different ways. Each time you get $43, so you become extremely confident in this answer. But then your friend, who was also writing down numbers and pushing calculator buttons, announces that she has consistently gotten $45! In this sort of case, many feel that very little, if any, conciliation is called for. Of course, here, too, questions arise about how the intuitive response to this sort of case will generalize. Steadfast theorists will see disagreements in fields like philosophy as more analogous to Careful Checking than to Mental Math. Another motivation for Steadfastness derives from the thought that many evidential situations leave room for more than one completely reasonable doxastic response. If that is so, then when my (equally informed, intelligent, and so on) friend disagrees with me, we might both be fully rational in our beliefs. Some Conciliationists have indeed argued for their position by supporting Uniqueness principles, which hold that only one doxastic response will be (maximally) rational in a given evidential situation. And Kelly (forthcoming) argues that a certain sort of strong Conciliatory view carries a commitment to Uniqueness. But many find Uniqueness principles highly implausible. So to the extent that Conciliationism’s plausibility is tied to that of Uniqueness, one may be motivated to take a more Steadfast view. 3. A Principle Dividing the Camps If examples can be found which seem to favor both steadfast and conciliatory responses, the debate will turn in part on what principles may explain these responses in a way that can be extended to cases where initial intuitions are less clear. There is at least a rough sort of principle that has been put forth by some Conciliationists, and attacked by their opponents, which may prove useful to consider here. The general idea emerges as a diagnosis of what would be wrong with the following sort of intuitively bad response to the suggestion that I should be conciliatory in the Mental Math case: Sure, my friend is generally as good at mental math as I am, and has been right as often as I when we’ve disagreed in the past. But I have special reason to think that I’m right this time: our shares of the bill are $43, and she thinks they’re $45. So she’s made a mistake, and I needn’t reduce my confidence in my initial answer. This response crudely begs the question in favor of the agent’s initial belief. In response, it might seem natural to impose something like the following sort of condition on rational responses to disagreement: Independence: In evaluating the epistemic credentials of another person’s belief about P, to determine how (if at all) to modify one’s own belief about P, one should do so in a way that is independent of the reasoning behind one’s own initial belief about P. Conciliationism will result from combining this sort of principle with the thought that, to the extent that one’s dispute-independent evaluation gives one strong reason to think 758 The Epistemology of Controversy a 2009 The Author Philosophy Compass 4/5 (2009): 756–767, 10.1111/j.1747-9991.2009.00237.x Journal Compilation a 2009 Blackwell Publishing Ltd that the other person is equally likely to have evaluated the evidence correctly, one should (in the case where one is quite confident that P, and the other person is equally confident that !P) suspend belief (or adopt a credence close to 0.5) in P. It is worth pausing to note that while Conciliationism may recommend something close to ‘splitting the difference’ doxastically in cases like Mental Math, it should not be construed as the view that in disagreements with an apparent peer, one should always split the difference. The difference-splitting view is initially attractive, and seems a natural way of putting into practice the Conciliatory idea of giving the peer’s opinion equal weight. But in Elga’s (2007) – which is the source of the label ‘Equal Weight View’ popularly used for strong Conciliationism – there is no mandate for uniform difference-splitting. This is a good thing, for two reasons. First, uniform difference-splitting generates technical difficulties (see Shogenji (2007) and Jehle and Fitelson (forthcoming)). Second, such mechanical difference-splitting with peers is not in the end true to the motivation for Conciliationism. Suppose, for example, that I am a doctor determining what dosage of a drug to give my patient. I’m initially inclined to be very confident in my conclusion, but knowing my own fallibility in calculation, I pull back a bit, say, to 0.97. I also decide to ask my equally qualified colleague for an independent opinion. I do so in the Conciliatory spirit of using her reasoning as a check on my own. Now suppose I find out that she has arrived – presumably in a way that also takes into account her fallibility – at 0.96 credence in the same dosage. Here, it seems that the rational thing to do is for me to increase my confidence that this is the correct dosage, not decrease it as difference-splitting would require. But this is not inconsistent with giving equal weight to my colleague’s opinion, or with the Independence principle. Commitment to some sort of Independence principle may also lie behind Conciliatory arguments based on the thought that widespread intractable disagreement in a field indicates that practitioners are not generally reliable. Absent some such principle, a practitioner would seem to be able to use her own views in the field to support the claim that while most practitioners are indeed quite unreliable, she was an exception, as evidenced by her having figured out so many issues correctly. Not surprisingly, some criticisms of Conciliationism have centered around Independence. Consider cases in which I begin with extremely high rational confidence in my belief, such as Careful Checking. Suppose that it is correct to say that I should not be very conciliatory toward my friend in that case. Intuitively, the reason that I resist full conciliation in that case is that I think that something screwy had gone on with my friend. But my only reason for suspecting that is that she announced getting $45. My only reason for taking that as a sign of trouble seems to spring from my confidence that $43 is the right answer – confidence, that is, that depends on my own initial reasoning on the disputed matter. This suggests that I must, after all, evaluate the epistemic credentials of my friend’s assertion in a way that violates Independence. It is not clear that the Conciliationist should concede that a steadfast response to Careful Checking requires giving up Independence. It can be argued that one’s reason for thinking something screwy went on with one’s friend does not, properly understood, rest on the reasoning behind one’s belief that the answer is $43. One might instead argue as follows: It would be incredibly unlikely for two people to clear-headedly go through the sort of extensive checking described and come up with different answers. Thus if two people who ostensibly went through this process announce different answers, it’s very likely that something screwy – either some sort of severe cognitive impairment, or a case of insincere assertion (lying, joking, The Epistemology of Controversy 759 a 2009 The Author Philosophy Compass 4/5 (2009): 756–767, 10.1111/j.1747-9991.2009.00237.x Journal Compilation a 2009 Blackwell Publishing Ltd etc.) – is involved. As I can eliminate many such possibilities much more decisively for myself than I can for my friend, I should think that it’s more likely that the answer she announced is incorrect. Note that this reasoning does not depend on my calculations, or on $43 being correct. The success of this defense of Independence will depend on making it plausible that similar responses are available in related cases, and that these responses are really independent of the arguer’s initial reasoning on the matter under dispute. One may also argue that Independence does not have the conciliatory punch it might seem to have for controversial issues, and thus that the worries about excessive skepticism can be mitigated. Elga (2007) holds that for many controversial issues, disagreements will involve large knots of interconnected claims. For example, Ann and Beth, who disagree about the morality of abortion, will probably disagree about a great many related moral, psychological, theological, and ordinary factual issues. But if Ann attempts to evaluate the epistemic credentials of Beth’s beliefs independent of all these, she will fail: to the extent that Ann abstracts from disputed considerations, there will be no fact of the matter concerning her opinion of Beth’s credentials. So Ann need not take a conciliatory attitude toward Beth’s belief. It is not clear how effective this line will be in assuaging worries about skepticism on controversial issues. One might object that if Ann takes Beth to be honest, intelligent, familiar with the arguments, and so on, and if she thinks that these are the attributes primarily responsible for figuring out the correct answers to difficult questions, she might well have enough dispute-independent information on Beth to arrive at some dispute-independent opinion about Beth’s likelihood of getting the right answer on abortion. In fact, it seems to me that a related line of thought leads to a serious problem for Independence that has not been developed so far in the literature (although it has been pressed in conversation). Suppose that someone earnestly challenges virtually all of my belief, including the beliefs that my general cognitive processes are reasonably reliable. Independence now enjoins me to evaluate the epistemic credentials of his beliefs in a way that abstracts even from my taking myself to be a reliable thinker. So I cannot dismiss him as I dismissed my friend in the Careful Checking case. It might well seem that I cannot cite any dispute-independent reasons for thinking that my beliefs are more likely to be correct, if only because, given the breadth of the territory under dispute, I cannot cite dispute-independent reasons for much of anything at all. Does Conciliationism now force me to near-global skepticism? The answer will depend on how exactly the dispute-neutral assessment demanded by Independence is to figure in one’s final beliefs. The most natural thought is that, absent independent reason to think the other person more likely to be mistaken, one must be conciliatory. But this natural thought seems to lead to skepticism quickly in the present case, as one lacks dispute-neutral reasons for one’s judgments. It turns out that the natural thought is too much like the thought that one must be able to provide a non-questionbegging answer to the global skeptic. This observation suggests a different approach to taking account of the dispute-neutral assessment demanded by Independence. One might hold that conciliation is required only to the extent that the dispute-neutral assessment provides one with strong positive reason for thinking the other person likely to have gotten things right. On this sort of view, one is not forced to be conciliatory in cases where the scope of the dispute precludes strong dispute-neutral reasons for making judgments about the other person’s epistemic credentials. Of course, this is just a sketch 760 The Epistemology of Controversy a 2009 The Author Philosophy Compass 4/5 (2009): 756–767, 10.1111/j.1747-9991.2009.00237.x Journal Compilation a 2009 Blackwell Publishing Ltd of an approach to the problem. At this point in the discussion, I think it is not clear how serious a problem it will turn out to be. 4. More Arguments on the Wider Issue 4.1. DOES CONCILIATIONISM THROW AWAY EVIDENCE? A prominent theme in Kelly (2005, forthcoming) is the worry that Conciliationism’s prescription for doxastic compromise is not appropriately sensitive to the possibility that one of the parties to the disagreement may have initially reasoned well, while the other did not. Kelly defends the view that one who has initially reasoned well should not be required to compromise doxastically to the same extent as one who has not. To see why Conciliationism might seem deficient in this respect, consider a case (adapted from Kelly (forthcoming)) in which two equally reliable thinkers, who have extensive evidence of each other’s equal reliability, independently study the same evidence E. In this case, Rita reaches the right credence of 0.8 – the level of confidence supported by E – while Wayne reaches the wrong credence 0.2. Then they learn about each other’s credences. Supposing that this is not a case in which disagreement would be strong evidence that something screwy had occurred, Conciliationism will hold that the correct response for both Rita and Wayne to the evidence provided by the other’s disagreement will be to compromise epistemically with the other. Suppose they do this, and each reaches about 0.5 credence. Kelly argues that it is implausible to hold that Rita and Wayne should make equally extensive revisions in their initial beliefs. That would ignore a clear epistemic asymmetry between them: Rita evaluated the original evidence correctly, while Wayne blew it. Kelly argues that the Conciliationist prescription would seem to amount to disregarding the original evidence E (after all, the 0.5 compromise was dictated by Rita’s and Wayne’s initial takes on E, and not influenced at all by what level of credence was really supported by E). To put the point another way, to call Wayne’s 0.5 credence rational would make rational belief too easy to come by. This can be made more vivid by thinking of two people who reach, say, credences 0.2 and 0.4 after considering E, and then compromise at 0.3. Remembering that E actually strongly supports P, we should not say that the compromisers’ 0.3 is rational. What should the Conciliationist say here? Kelly’s sort of examples show that when the Conciliationist says, for example, that Rita and Wayne each should respond to the evidence provided by the other’s disagreement by moving to credence 0.5, we must not understand this as meaning that the resulting credences will be fully rational. Conciliationism should instead be seen as a view just about the bearing of one particular kind of evidence. Clearly, taking proper account of one bit of evidence cannot be expected to erase previous epistemic blunders. Thus the Conciliationist should agree that Wayne’s belief in the example is not fully rational. She should agree that Wayne does, indeed, have more reason for doxastic revision than Rita does – the reason provided by the original evidence, which supports P. And the Conciliationist should take a similar view of the two unfortunates who both screwed up before compromising. But this is all consistent with the Conciliationist’s view of the evidential import of disagreement. So Kelly’s examples reveal something important about how Conciliationism must be understood. But on this understanding, it is not clear that Conciliationism makes the original evidence irrelevant, or makes rational belief too easy to come by. The Epistemology of Controversy 761 a 2009 The Author Philosophy Compass 4/5 (2009): 756–767, 10.1111/j.1747-9991.2009.00237.x Journal Compilation a 2009 Blackwell Publishing Ltd 4.2. DOES STEADFASTNESS BETTER ACCOUNT FOR THE FIRST-PERSON PERSPECTIVE? Conciliationists are fond of examples involving our epistemic reliance on devices such as watches, which are external to us. After all, if my watch says 4:20, and my friend’s says 4:30, then (absent independent reason to favor the verdict of one of the watches), it seems that I should accord both possibilities equal weight. But is the comparison apt? Foley (2001) and Kelly (2005) have argued that it is a mistake to view the problem of resolving disagreements with others as analogous to third-person cases involving arbitration between two external sources. After all, in arbitrating the dispute between oneself and another, one has no choice but to do so from one’s own perspective. Clearly, arbitrating the dispute from one’s own perspective need not entail disregarding evidence that one might be wrong – that would mean ignoring disagreement even from one’s epistemic superiors (Christensen 2007). But this leaves open the possibility that some aspect of arbitrating from the first-person perspective will result in a measure of steadfastness greater than that permitted by Independence-based Conciliationism. One such possibility has been advanced by Wedgwood (forthcoming). Wedgwood notes that my own beliefs, experiences, and intuitions can guide me directly in a way that the beliefs, experiences, and intuitions of others cannot (even if I know about them). He then suggests that this asymmetry may ground the rationality of somewhat Steadfast responses to disagreement, even when one has no independent reason for thinking oneself to be more likely to be correct. Filling in this picture more fully will require explaining why this undoubted asymmetry should license Steadfastness. The fact that I am guided directly by some beliefs, and only indirectly by others, does not require Steadfastness; so it would appear to be an open question whether the directness of guidance should mitigate the rational pressure toward Conciliation. Still, to the extent that one finds Steadfast responses intuitively rational, this asymmetry provides a theoretical framework from which to pursue justifying such responses. 4.3. IS CONCILIATIONISM DEFECTIVE BECAUSE IT IS SELF-UNDERMINING? Several people have noted that, at least given the current state of epistemological opinion, there is a sense in which Conciliationism is self-undermining. For example, I, as a conciliationist, know full well that several excellent philosophers oppose my view; in fact, it seems to me that opinion on Conciliationism is presently divided roughly evenly. By my own lights, then, I should not be highly confident of Conciliationism. So in a sense, my Conciliationism is self-undermining. One might distinguish here between principles which automatically self-undermine, and principles which do so only potentially – that is, they self-undermine only under particular evidential circumstances. Clearly, Conciliationism belongs to the latter category. There are several different aspects to this problem. One is that even describing how I should react to the news of other philosophers’ rejection of Conciliationism is more complex than it might seem at first. Suppose I take account of others’ views, and reduce my confidence in Conciliationism. That, it seems, should affect the way I react to other disagreements, making me more Steadfast. Indeed, looking back retrospectively on my compromise vis-à-vis Conciliationism, I may now judge that I went too far! Blog posts by Brian Weatherson and Matthew Weiner discuss the technical question of whether a stable view of Conciliationism is possible in such circumstances. Less technically, one might just worry that there is something intrinsically wrong with an epistemic principle that would sometimes tell you that it is incorrect. Elga (forthcom762 The Epistemology of Controversy a 2009 The Author Philosophy Compass 4/5 (2009): 756–767, 10.1111/j.1747-9991.2009.00237.x Journal Compilation a 2009 Blackwell Publishing Ltd ing) argues that even potential self-undermining is fatal – at least for Conciliationism as a basic epistemic principle – because it amounts to a kind of inconsistency: the principle yields incompatible prescriptions for certain evidential situations. It is important to realize, however, that these worries about potential self-undermining are not restricted to Conciliationism. Consider the following principle. Minimal Humility: If I have thought casually about P for 10 minutes, and have decided it is correct, and then find out that many people, most of them smarter and more familiar with the relevant evidence and arguments than I, have thought long and hard about P, and have independently but unanimously decided that P is false, I should become less confident in P. Clearly, Minimal Humility will self-undermine in certain evidential situations. Given its evident plausibility, we should be cautious before taking potential self-undermining as showing a principle false. In fact, it seems that any general position on disagreement short of a radically Steadfast refusal to take anyone’s contrary opinions into account ever – in other words, any halfway reasonable account of disagreement – will be faced with the fact that it is potentially self-undermining. So it is not clear how seriously Conciliationists in particular should worry about the fact that Conciliationism has this property. The worries explored by Weatherson and Elga appear to be worries for pretty much everyone. However, even if this is right, and Conciliationism being potentially self-undermining does not itself pose a special problem, a difference remains between Conciliationism and, say, Minimal Humility: given the present state of epistemic opinion, one might well remain consistently confident of Minimal Humility. But one cannot say the same for Conciliationism. Indeed, it seems to me those of us who find ourselves strongly drawn toward Conciliationism in these contentious times should not be confident that Conciliationism is correct. (Of course, we may still work hard in producing and disseminating arguments for the view, hoping to hasten thereby the day when epistemic conditions will brighten, consensus will blossom, and all will rationally and whole-heartedly embrace Conciliationism.) 5.Some Additional Outstanding Questions 5.1. HOW CLOSELY IS CONCILIATIONISM TIED TO UNIQUENESS? As noted before, Conciliatory views have been supported by invoking rational Uniqueness principles, according to which only one doxastic response is maximally rational in a given evidential situation. Steadfast views have been supported by invoking Permissiveness, the denial of Uniqueness. But it is not fully clear how closely the two controversies are linked. There is the obvious point that if the leeway granted by a Permissive account is fairly limited, it might not permit Steadfast responses to the sort of sharp disagreement we seem to see running all through fields like philosophy, history, and economics. But there are also more subtle questions to consider. While it might seem clear that, to the extent that one denies Uniqueness, one permits Steadfast responses to disagreement, the relation between these positions is not that simple. Permissiveness entails that the following sort of case can arise: A and B are fully rational, and have exactly the same evidence, but have different levels of credence in P. Now, as Kelly (forthcoming) argues, if each of A and B realize that both levels of credence are fully rational, then it would seem that they can then learn of their disagreement without having any reason for belief-revision, contra strong versions of Conciliationism. The Epistemology of Controversy 763 a 2009 The Author Philosophy Compass 4/5 (2009): 756–767, 10.1111/j.1747-9991.2009.00237.x Journal Compilation a 2009 Blackwell Publishing Ltd There still may, however, be Permissive accounts of rational belief that support very strong Conciliationism. One sort of example would involve Permissiveness motivated by the thought that fully rational belief cannot require god-like epistemic perfection. One might hold that an agent’s credence in P is fully rational if: (a) it is close enough to the Ideal Evidential Probabilities given the first-order evidence, and (b) the agent lacks independent reason for suspecting that her credence is too high (or for suspecting it is too low) relative to the ideal credences. Such an account is compatible with taking peer disagreement to provide agents with reason for suspecting their credences are too high (or too low). So such a view could allow differing responses to the evidence from isolated agents, but require conciliation when the agents become aware of peer disagreement. As Ballantyne and Coffman (ms) point out, this sort of view is premised on allowing that agents may fail to be in a position to know what the most rational response to their evidence is. That might make it implausible, but it might also comport with one of the main motivations for Conciliationism, which is precisely that people cannot always tell directly what their evidence supports. A different type of example might involve a view on which rational belief depends on evidential support plus practical stakes. Such views clearly violate Uniqueness. But on such a view, I may often (e.g., when I have evidence that practical factors are equal) get reason from peer disagreement to think that I have misevaluated evidential support. This could be held to require strong conciliation. Such situations may arise particularly often in fields like philosophy, where differences in practical stakes would frequently seem insufficient to underwrite significant differences in the degree of evidential support required for belief. Other views, which deny Uniqueness for other reasons, may be more hospitable to the Steadfast position. Kelly (ms) points out that Uniqueness is violated by views which allow agents’ weightings of epistemic values (such as believing truths and avoiding belief in falsehoods) to help determine rational belief. Such views would seem amenable to Steadfast reactions to peer disagreements, at least insofar as one was rational in thinking that the disagreements were value-based, rather than based on different assessments of evidential bearing. In sum, it does seem clear that embracing Permissiveness has the general potential to push our view of disagreement toward the Steadfast end of the spectrum. The point of these toy examples is just to show that the details of this relationship are more complex than they might at first seem to be, and to suggest that one’s reasons for denying Uniqueness may well affect the extent to which one’s general account of rational belief is hospitable to Conciliatory or Steadfast responses to disagreement. 5.2. HOW MUCH DISAGREEMENT IN PHILOSOPHY IS GENUINE? Another dimension of the disagreement problem, especially as it applies to disagreement in philosophy, is the degree to which some disagreements may be merely verbal. Sosa (forthcoming) emphasizes this point; indeed he suggests that it accounts for why many of the apparent disagreements we find in philosophy are reasonable on all sides. (This is not to say they are trivial. Sosa distinguishes superficial disputes based on ‘bank’-type homonyms from the sort that more plausibly occupy some philosophers, based on terms whose contents largely overlap.) Clearly, this issue has important implications for the degree to which Independencebased views would require suspension of belief in controversial philosophical theses. But it is also clear, as Sosa notes, that it is difficult to generalize about this issue: approaching it seriously would require detailed work on particular examples. 764 The Epistemology of Controversy a 2009 The Author Philosophy Compass 4/5 (2009): 756–767, 10.1111/j.1747-9991.2009.00237.x Journal Compilation a 2009 Blackwell Publishing Ltd 5.3. HOW DO THE VIEWS APPLY TO REAL-LIFE CONTROVERSIES? The literature on disagreement has concentrated on artificially simple cases designed to support or refute general principles governing the correct response to disagreement; the hope is that this abstract understanding will throw light on how we should react to controversial matters in philosophy and elsewhere. But it may be more difficult than it would seem to take this next step. For example, everyone seems to agree that one’s assessment of the credentials of those with whom one disagrees is an essential determinant of how much, or whether, one should revise belief on the basis of their disagreement. But in real controversies, we lack the sort of track record that provides for robust evaluations in some of the artificial cases. It is not totally clear what sorts of factors one should take as relevant to, say, the likelihood of a person’s arriving at correct opinions on mental causation, or abortion, or the existence of gods, or even the likely effects on poorer Americans of cutting capital gains taxes. Other dimensions of complexity affecting real-life cases are the numbers of people on different sides of the issue, and the types and degrees of causal or epistemic dependence of some people’s opinions on others. If 117 people independently come to disbelieve P, and only 4 of us believe it, then ceteris paribus it is likely that some revision on our part is in order. But if 116 of those people are all unthinking disciples of the 117, the case for revision would seem to be vastly weakened. Presumably, all views on disagreement will take these factors as relevant, but specifying precisely how they play out even in the abstract will not be easy. And to the extent that we do come up with abstract principles covering these factors, applying these principles to actual controversies will no doubt be harder still.