Dealing With Disagreement : Distinguishing Two Types of Epistemic Peers

Epistemologists have recently debated how we should respond to apparent cases of rational disagreement. Is it possible for two people to disagree and have both people still be rational? Those involved in this debate make use of the idea of epistemic peers. Two people are epistemic peers if they share the same knowledge of a given topic and have similar epistemic virtues. My paper argues that we have different kinds of epistemic peers; close peers who think similarly to ourselves, and remote peers who think very differently. I argue that remote peers are genuine peers, but that we should respond to disagreement from remote peers differently than disagreement from close peers.

Epistemologists have recently debated how we should respond to apparent cases of rational disagreement.Is it possible for two people to disagree and have both people still be rational?Those involved in this debate make use of the idea of epistemic peers.Two people are epistemic peers if they share the same knowledge of a given topic and have similar epistemic virtues.My paper argues that we have different kinds of epistemic peers; close peers who think similarly to ourselves, and remote peers who think very differently.I argue that remote peers are genuine peers, but that we should respond to disagreement from remote peers differently than disagreement from close peers.
Disagreement is a constant feature of our epistemic experience.Often this disagreement is explained by the fact that one of the parties to the dispute lacks some piece of data, or else by the fact that one of the disputants has greater cognitive ability or experience in making judgements of the kind at issue.In other cases, however, there is continuing disagreement even between equally well-informed and intelligent investigators, often referred to as epistemic peers.It is tempting in such instances to hold that this is a case of "rational disagreement" in which both disputants can be rational in continuing to hold their conflicting positions.However, several prominent epistemologists have denied that cases of rational disagreement are possible, holding that in such disputes the only rational response is for both parties to suspend judgement on the issue.This position is often called the "equal weight view."I wish to introduce a new viewpoint into this debate by distinguishing between two different kinds of epistemic peers.Roughly speaking, some peers reason about evidence in similar ways, considering the same sorts of evidence relevant and the same kinds of arguments valid.Other peers reason in very different ways from each other and yet still appear to be equally well informed and intelligent.An example of peers of this second sort might be a religious ethicist and an atheist ethicist discussing some moral problem,

Wald
Dealing With Disagreement or a string theorist and a critic of string theory considering the desiderata of theory construction.I wish to defend the claims that there can in fact be peers of this second kind and that we can identify people who count as such peers for us.I argue that disagreement from an epistemic peer carries different epistemic consequences depending on which sort of peer it is from.While the equal weight view is convincing for peers of the first kind, I will argue that we are rationally justified in maintaining our position while disagreeing with a peer of the second type.On the other hand, intellectual engagement with this second type of peer can potentially have much greater long-term effects: it can give us an opportunity to reflect on our epistemic outlook as a whole and possibly adapt or amend it.
To begin with, we should consider what definition of epistemic peers is at work in the existing literature on rational disagreement.Most of the philosophers who have written on this topic seem to agree on a general definition of what makes someone an epistemic peer.Thomas Kelly provides a clear statement of this definition, defining someone as an epistemic peer with respect to some given issue if and only if they satisfy the following two conditions: (i) they are equals with respect to their familiarity with the evidence and arguments which bear on the question, and (ii) they are equals with respect to general epistemic virtues such as intelligence, thoughtfulness, and freedom from bias.(Kelly 2005, 174-75) While something like this definition is used in most of the discussions on epistemic peers, philosophers disagree on whether people with substantial disagreements can still meet these requirements.Richard Feldman, for example, holds that "cases of seemingly rational disagreement are all around us," (Feldman 2006, 217) and proceeds to provide examples from law, philosophy, politics, and religion.Feldman sees these cases as potential instances of rational disagreement because he holds that these areas "involve intelligent, serious, and thoughtful people with access to the same information who come to different and incompatible conclusions."(Feldman 2006, 219) According to Feldman, all of these areas of inquiry, despite widespread disagreement, allow for epistemic peers. 1 Adam Elga, on the other hand, holds that we should consider someone an epistemic peer only if we think they are as likely as we are to arrive at the correct conclusion.This judgement will be influenced by what we think about the potential peer's other beliefs and past track record (Elga 2007, 493).If I am an atheist and I know that my friend is a devout Catholic, I may judge him less likely to make a correct judgement on the morality of abortion.In the context of a debate on abortion, therefore, Elga would claim that I should not treat this friend as a peer.While people may have many peers in such clear-cut fields as mathematical calculation, "in messy real-world cases... though agents in those examples may count their associates as thoughtful, well-informed, quick-witted, and so on, they often do not count those associates as peers."(Elga 2007, 493) For Elga, only those who agree with one's general outlook and opinions will count as epistemic peers.
To clarify what is at issue in this disagreement, I think it is helpful to make use of the distinction Mark Vorobej introduces between close peers and distant peers (Vorobej, unpublished manuscript).Vorobej defines close peers by stating that two individuals are close, or perfect peers, with respect to a body of evidence E and a topic T, just in case, in addition to being peers in Kelly's sense, both (i) [they] assess (or reason about) the probative value of E in a very similar manner, with each individual following a method that is transparently intelligible to the other; and (ii) [they] each have good reason to believe that [they] have comparably good [or bad] track records of forming true beliefs from bodies of evidence pertaining to T. (Vorobej, unpublished manuscript) A distant peer is an epistemic peer who fails to meet at least one of these conditions, and a distant peer who satisfies neither of the two conditions specified above is a remote peer (Vorobej, unpublished manuscript).This framework can be used to understand debate over whether or not two people can be epistemic peers despite widespread disagreement about the subject at issue.Feldman would likely agree that the kind of people we have referred to as remote peers are, in at least some instances, genuine peers.He believes the disputants in philosophy are often epistemic peers, for example, but a utilitarian and a Kantian assess and reason about evidence in ethics in very different ways.Elga's stance, on the other hand, implies that he would reject the idea that remote peers are actually our peers, at least in cases where remote peers disagree with us.According to Elga, we should treat someone as a peer only if we believe beforehand that he or she is as likely to arrive at the correct conclusion as we are.In philosophy, a utilitarian would probably judge a

Dealing With Disagreement
Kantian less likely to arrive at the correct conclusion to an ethical dilemma.A utilitarian thinks that the utilitarian methods of assessing evidence are the most effective at reaching truth in ethics; otherwise, he or she would not use them.Since a Kantian uses different assessments of evidence, he or she is less likely to arrive at truth and thus would not be an epistemic peer.The Kantian would follow a similar chain of reasoning and conclude that the utilitarian is not his or her epistemic peer either.
I wish to argue that remote peers are genuine peers but that we are rationally justified in treating disagreement from remote peers differently than disagreement from close peers.The equal weight view, which holds that when we disagree with an epistemic peer we should both suspend judgement, seems convincing for close peers.After all, in these cases we both accept the same standards of evidence, and so one of us must be misapplying these standards in this case.Given that close peers share equivalent intellectual virtues, neither has a reason to assume that the error is more likely to have been made by their interlocutor, so the rational response is for both parties to suspend judgement.2However, when we consider remote peers, I think a different conclusion is warranted.
In order to determine what response is rationally justified when confronted with disagreement from remote peers, we must first have some idea of what comprises rationality.In dealing with remote peers, it seems inappropriate to adopt any particular substantive theory about what kinds of inferences and evidence make a given belief rational.After all, by definition, remote peers reason in very different ways from one another and so will endorse different substantive theories of rationality.Since our remote peers are as intellectually virtuous and knowledgeable as ourselves, it seems as unwarranted to assume that our own theory of rationality is more likely to be true as it would be to assume that our judgement in the specific case is more likely to be true.As such, it seems appropriate to make use of a content-neutral, or procedural, theory of rationality.
A useful version of such a theory is provided by Richard Foley, who provides what he refers to as a subjective foundationalist theory of epistemic rationality (Foley 1987).For Foley, a belief is rational for an individual if it accords with his or her deepest epistemic standards.These epistemic standards are in turn defined by what the individual would believe if he or she were sufficiently reflective.The theory gives a sophisticated account of epistemic standards, but for our purposes, the main points are that different people have different reflective styles of reasoning and that there is at least one sense of rationality in which it is rational for people to form beliefs in accordance with this personal style.Such a theory of rationality allows us to criticize sloppy or unreflective reasoning but is neutral as to the specific types of arguments or evidence of which it is rational to make use.As such, this theory provides a sense of rationality that is ideally suited to addressing the problem posed by remote peers.We can further refine our account of distant peers using Foley's notion of epistemic standards.A close peer would be someone who meets Kelly's two conditions for being an epistemic peer: who shares our own deepest epistemic standards and who has a similarly strong track record with regard to questions of the type at issue.A distant peer is one who fulfills Kelly's conditions and either lacks a track record or makes use of different epistemic standards, while a remote peer is a distant peer who both lacks a comparable track record and uses different epistemic standards.
This definition might seem unhelpful in practice.To identify someone as a remote peer, we need to judge that they fulfill both of Kelly's two conditions for being an epistemic peer, despite the facts that they do not have a track record on beliefs of this type and that they make use of different epistemic standards.Kelly's first condition, that a peer be equally well informed on the topic in question, is relatively easy to verify.A pair of prospective peers need only to compare their respective knowledge of key arguments and pieces of evidence pertinent to the debate.However, verifying Kelly's second condition, which demands that the pair of prospective peers have equivalent "epistemic virtues," or as I prefer, "intellectual virtues," is more difficult.Is it even possible to identify someone as intellectually virtuous when they embrace entirely different standards of knowledge?I believe that it is, and to see how this is so, it will be helpful to clarify what exactly we mean by an "intellectual virtue."Kelly himself gives us little guidance on what he means by this condition.However, he uses as examples of intellectual virtue intelligence and freedom from bias.This suggests that Kelly sees intellectual virtues as character traits of the epistemic agent, which points us in the direction of some version of virtue responsibilism, rather than to a reliabilist account of intellectual virtue whereby virtues are reliable, belief-producing faculties. 3Such an account seems most appropriate for our purposes here.In cases of disagreement between potential remote peers, the reliability of the methods being used is one of the key points of contention.Given that remote peers do not have a track record to compare, declaring one's own methods of arriving at beliefs to be more reliable seems to beg the question against the potential peer.Furthermore, what we are trying to capture in the notion of a remote peer is someone whose general intellectual character one respects as comparable to one's own, despite differences in outlook and methodology.This suggests that it is the potential peers' epistemic character traits and habits which are at issue.
A useful definition of intellectual virtue is offered by Linda Zagzebski, who defines a virtue as "a deep and enduring acquired excellence of a person, involving a characteristic motivation to produce a certain desired end and reliable success in bringing about that end" (Zagzebski 1997, 10, emphasis in original).At first glance, it seems that reliability has re-entered our definition here, but the crucial difference is that this is reliability in achieving the aims of the motivational component of the virtue, not reliability in achieving truth.For instance, Zagzebski describes the motivational component of the virtue of open-mindedness as a disposition to "restrain the natural impulse to ignore or be unreceptive to views contrary to one's own," (Zagzebski 1997, 10) and the reliability component would be success at this goal rather than success at the more general goal of arriving at truth.Intellectual virtues are those virtues whose motivational component is derived from the desire for knowledge.I would add to the motivation to attain knowledge the complementary motivation to avoid error.The weighting of these two elements may differ between individuals, with some people being very concerned with attaining truth even at the risk of error and others willing to miss out on true beliefs as long as errors are avoided.Nonetheless, only virtues whose motivational elements can be traced back to one of these two goals should count as epistemic, as opposed to moral or prudential, virtues.
In order to identify intellectual virtue in a prospective remote peer, we need both to identify the appropriate motivations and to judge the individual's success in bringing about the goal of these motivations.This will be more difficult in the case of a remote peer than a close peer, and in some cases, it may prove impossible.Identifying the motivational element of the virtues will not be that much harder for remote peers than for close peers.However, it will often be more difficult to judge the reliability with which an individual achieves the goals of these motivations when he or she is a potential remote peer.A remote peer will often come to conclusions with which we disagree, and it might be difficult to determine whether this difference is due to a failure on the individual's part to achieve the aim of the motivational component of intellectual virtue.For instance, if Joe is a proponent of evidence-based medicine and Judy champions the value of clinical expertise, Joe may ascribe this difference to Judy's failure to be impartial when evaluating her own ability to predict clinical outcomes.However, this difficulty can be overcome.When we examine the overall trends in an individual's belief formation, intellectual virtues and vices should become apparent as trends.If someone always refuses to alter opinions once formed, it becomes likely that he or she is not open-minded, whereas if he or she has a history of adopting new ideas from time to time then open-mindedness is plausible.This implies that we can identify our remote peers at least some of the time.
The question now becomes what our response should be when we encounter disagreement with remote peers.Foley can aid us here as well.In his 2001 book Intellectual Trust in Oneself and Others, Foley discusses the issue of when we should trust the judgements of others.Particularly germane to the present discussion is his point that, given that we must trust ourselves if we are to avoid skepticism, we must also trust our means of arriving at beliefs.If I am confident that most of my beliefs based on sense perception are true, then I must also be confident that sense perception is a generally reliable way to form beliefs.However, the perceptual and cognitive faculties of all humans are strikingly similar.Thus, if I believe that perception is reliable, and that nearly all humans have similar powers of perception, I must, if I am to be consistent, believe that other people are generally reliable (Foley 2001, 102-103).The same argument holds for knowledge gained through our cognitive ability, Kantian categories of experience, or any other source of beliefs that is widely shared.
We can use Foley's account of our trust in others to argue that the disagreement of even remote peers should have an influence on our belief, although this influence should fall short of full suspension of judgement.We all start out with similar mental capacities and perceptual equipment.However, from this same starting point, people in different cultures and in different situations arrive at very different epistemic standards.This is so despite the fact that these epistemic standards are all designed to allow us to fulfill the epistemic goal of now having true beliefs and now avoiding false beliefs (Foley 1987, 8).How can we explain this diversity?
In the case of a remote peer, if we have stipulated that the individual has equal knowledge and intellectual virtues to ourselves while also accepting that such an individual began with similar capacities to our own, then it appears that we have no reason to assume our own epistemic standards are more conducive to believing the truth or avoiding falsehoods than those of our remote peers.We cannot assume that we are using the correct standards and everyone else is in error insofar as they depart from them.After all, we acknowledge that others who are as intelligent and knowledgeable as ourselves arrived at different standards.This might be due to the coincidence of upbringing or some other factor, but whatever the source, we have no grounds to suppose it is due to some particular merit of our own epistemic standards.We are not forced to suppose that all such standards are equally effective, since they may contradict one another.However, we should accept that these alternate standards have as strong a claim to being effective as our own.
The difficulty is that even if we accept that these standards are as likely to be correct as our own, we must still choose some determinate set of standards in order to pursue our goals of believing the truth and avoiding falsehoods.Since these epistemic standards contradict each other, we cannot reconcile them, so we are stuck choosing between them.We could of course become skeptics and disbelieve or suspend belief on almost everything, but this too would be the adoption of an epistemic standard, and would be just as well or poorly warranted as any other.Given this dilemma, it still makes sense for us to commit ourselves to our own deepest epistemic standards, since they are what strike us most forcefully as convincing.While the specific level of confidence with which we hold any individual belief will not be affected by the disagreement of distant peers, widespread disagreement from distant peers in a given area of discourse should cause us to recognize the uncertainty of beliefs in that area and hold the possibility of error more firmly in mind.Thus, we should be more aware of the possibility of error in fields such as philosophy, where disagreement from remote peers is commonplace, than in mathematics, where it is relatively rare.
Although the mere fact of disagreement from a remote peer may not cause us to give up a belief entirely, it may result in a change in our long-term epistemic outlook.In the course of ordinary deliberation, our epistemic standards tend to recede from view.Evidence and arguments are interpreted using these standards, but the standards themselves remain in the background.Disagreement between close peers, who share epistemic standards, will also assume the correctness of these standards.Given that close peers will agree on the standards to be used in evaluating disputes, these standards will not themselves be called into question and will instead serve as the (generally unexamined) framework for the disagreement as a whole.The debate will focus on the application of the mutually accepted standards to a particular case.It is primarily in disagreements with those who make use of different standards that our own modes of thought are explicitly revealed and the possibility of changing them considered.The disagreement of remote peers thus provides an opportunity to inspect more closely the standards we use and to examine the reasons we find them convincing.It also provides a certain impetus to do so.Since remote peers are those who are equally knowledgeable and intellectually virtuous, when we encounter one who disagrees with us, it demonstrates that there are other possible epistemic standards endorsed by individuals whose judgement we should respect.

Dealing With Disagreement
This effect is heuristic; the mere fact of disagreement from remote peers does not in itself generate any rational requirement for us to alter our epistemic standards.Given the number of remote peers, and the variety of viewpoints they hold, we cannot expect to find any standards that will avoid disagreement altogether.However, epistemic standards are based on what we would come to believe if we were sufficiently reflective.Usually we do not have the time or interest to actually reflect on our deepest epistemic standards.Disagreement with remote peers can foster this form of reflection because it challenges us to understand why we hold our own position in the face of this disagreement.Occasionally, we may discover that the principles under which we have been tacitly operating do not stand up to reflection, and this can result in a readjustment of our beliefs.Thus, disagreements with remote peers can have a greater effect on our beliefs over the long term than disagreements with close peers, since the former cause us to reconsider our core epistemic standards in a way that the latter do not.
Remote peers are genuine peers, and their disagreement should be taken into account in evaluating our confidence in our beliefs.It will be harder to identify the necessary epistemic virtues in remote peers when they are applying these virtues according to epistemic standards we find unconvincing.Nonetheless, there will be cases when we can indeed identify someone as a remote peer.This implies that the members of different epistemic communities can count as remote peers to one another.The disagreements between such communities need not be due to any failure of rationality.Thus, the epistemic boundaries embodied in the divergent practices of these different communities can be, at least in some cases, rationally justified.However, this does not imply that discussion and dialogue between such communities is fruitless.The disagreement of these remote peers can influence the development of our epistemic identity, by encouraging us to critically consider our epistemic standards.To ignore the disagreement of remote peers would be to unfairly devalue the influence of the intelligent, knowledgeable people who happen not to share our own standards and to forsake a key opportunity to evaluate our own standards.