The author defends Conciliationism as a response to peer-disagreement in ethics against a prominent objection: if in cases of peer-disagreement we have to move our credences towards those of our dissenting peers, then we have to adopt scepticism in fields where disagreement between peers abounds. For this objection, the case of ethics is particularly worrisome. The author argues that the objection from scepticism is based on a highly idealised notion of an epistemic peer. In cases of disagreement about ethical issues, it is often unknown to us what another person counts as her evidence, since one’s notions of what counts as evidence and what weight to attach to different forms of evidence are impacted by one’s global outlook. Being aware of what an agent considers as evidence requires familiarity with that agent’s global outlook. This introduces two constraints on epistemic peerhood in cases of disagreement about ethics: an epistemic constraint (I might not be sufficiently aware of what someone counts as evidence, and hence not consider that person a peer), and a factual constraint (we might disregard each other’s evidence, and hence not consider each other peers).
In the last decade, a heated philosophical debate has emerged about the question of how one should react when faced with people who are as well informed and as rational as oneself and who disagree with one’s views: Is the rational reaction to align one’s views with those of one’s peers or rather to remain steadfast? This debate is of particular interest to us as philosophers, since we work in a field that seems ripe with this kind of disagreement. The question about how to rationally respond to disagreement is therefore also a question about how philosophers ought to conduct themselves when doing philosophy. In the present paper, I will show how and why attribution of epistemic peerhood works differently in cases concerning ethical questions than it does in other cases, such as splitting a restaurant bill. Ethical questions are different, firstly, in the sense that they tend to be more complex, e.g. they potentially involve a number of different considerations of very different kinds. Secondly, and more importantly, ethical questions are linked to what I call an agent’s “global outlook”, her general perspective on the world and what matters to her. The global outlook influences what the agent considers as relevant to answering complex questions. Whilst the global outlook is not exclusively relevant to ethical questions, it has a particular bearing on these issues. The function of the global outlook for our thinking about ethical cases makes the criteria for epistemic peerhood more difficult to satisfy, and it makes it more challenging to recognize someone as an epistemic peer. Hence, we have fewer epistemic peers in this area than we might have in other areas as well as fewer agents we recognize as an epistemic peer.
I will begin by explaining peer disagreement and introducing Conciliationism, the view that we should adjust our credence in a belief in cases of peer disagreement (1). I will then spell out the objection that Conciliationism might lead to scepticism in ethics (2), and critically discuss Adam Elga’s defence of Conciliationism against this objection (3). I will show that, despite my criticism of Elga, we can make use of his main idea that peer disagreement, even in controversial fields, is not as widespread as we think, at least in the case of disagreement about ethics. I will explain why this is the case and argue that disagreement about ethics concerns not only how we react to evidence, but also what we consider to be evidence in the first place. This is a version of what I label the “Enmeshment Reply” (4). Finally, I will discuss some qualifications of my view (5). The contribution of my paper is to show that and how an agent’s global outlook impacts the attribution of epistemic peerhood, and how this helps to defend Conciliationism against the objection that it might lead to a form of scepticism.
I take it that every plausible philosophical position holds that I should lower my credence in a belief, when a person I disagree with provides a good argument against this belief. The question I am interested in is whether the mere fact that there remains disagreement even after all the arguments are on the table provides reasons to adjust credence.
Three remarks are in order here. Firstly, by “having the same evidence” in P1, I mean that agents are not only aware of the same considerations, but also consider them as relevant or reason-giving for what is at issue. I will say more about this in section 4. My argument in this paper also holds for agents who have almost the same evidence. This is important since on some notions of how evidence is individuated different agents can never have the (numerically) same evidence.3 For instance, our respective sense perceptions might be of the same event, but they are different sense perceptions for each agent. If we think that the sense perception (not the event) is evidence, agents do not share evidence in the strict sense of the term. The same might hold for intuitions. Different Agents’ intuitions might have the same content but they are different occurrences of this content. This, however, should not rule out that agents can satisfy P1. There is a danger that if conditions of peerhood are interpreted too restrictive, the occurrence of peer disagreement would become a puzzle. The differences in global outlook I will discuss below are supposed to capture cases with significant divergence in conceptions of evidence and in what evidence agents have, hence they should be clear cases of not sharing evidence.4
(P1)have the same evidence, 2
(P2)are equally likely to react reasonably to evidence, and
(P3)there are no reasons, independently from the fact that they disagree, for those involved in the disagreement to discard the other person as her peer.
Secondly, it is important that the reasons mentioned in P3 are independent from the fact that we disagree. If they were not, I could simply disregard someone as a peer on the grounds that I think my belief is correct and her belief is dissenting from mine and hence wrong, and that this proves that she is not as good as I am in assessing evidence. This illicit move is sometimes called “bootstrapping” (see Elga 2007, sec. 10, Kelly 2010, sec. 5.4). A valid independent reason to discard someone as a peer could be, for instance, that a person shows signs of insanity, drunkenness, or seems to pay no attention to the question.5 In section 4, I will argue that we do discard agents as peers due to differences in global outlook and that this is rational. This discarding of an agent as a peer on grounds other than disagreement about a specific issue satisfies the independence condition.
Thirdly, it is not always clear whether the conditions of peerhood should be understood objectively or subjectively, i.e., whether you have to satisfy criteria P1, P2 and P3 in fact or whether I only have to believe that you do so in order to count you as my peer. I will follow Elga (2007, fn. 14 and fn. 21), whom I will critically discuss in section 3, and accept the subjective interpretation.6 On the objective interpretation it would be difficult to make sense of P3, which directly concerns what it is reasonable for me to believe. The question of pd is how I am to rationally react to disagreement in the light of certain things I believe about the person(s) who disagree with me, namely, that we share the relevant evidence, that we are equally likely to respond correctly to evidence and that we are currently in a position to judge rationally. It is therefore not important what evidence my interlocutor in fact has, but only what evidence I believe her to have. This leaves open that two agents are in fact epistemic peers but that they also have no reason to adjust their credence in cases of disagreement with each other since they are unaware of this status (see my epistemic constraint in sec. 4).
Of course, if my interlocutor does not have the same evidence as I do, I can still falsely believe that we have the same evidence. I assume that I am more likely to acquire and retain the true belief that someone is my peer than I am to acquire and retain the false belief that someone is my peer. If someone is not my peer the differences between us can become apparent at any time once we enter into a dialogue with each other, I read her papers, see her give talks, etc. There is always the possibility that I notice relevant epistemic differences between me and my non-peers. In addition, as I will argue in section 4, in cases in which a question is related to my global outlook, I am likely to suspect that those who disagree with me do not share my standards of evidence. In these cases, I will suspect others of not being my epistemic peers until I become convinced that we do share the same evidence.
I will call disagreements between agents which satisfy P1–3 Peer Disagreement (pd).7 My argument in section 4 will draw on the very demanding conditions for epistemic peerhood and for acknowledging another agent as a peer, and on the great amount of idealization peerhood requires. Due to this idealization the notion of an epistemic peer and of peer disagreement has come under attack lately. Most notably, Nathan King (2012) argues that in the real world pd, which the parties involved acknowledge as pd, is not as wide-spread as the debate assumes. This is due to the highly idealized nature of peer-disagreement.8 My approach is congenial to King’s, though my argument is not based on the notion of an epistemic peer as such but on the nature of ethical questions.9
It is uncontroversial how one should react to pd in a host of cases. Take Christensen’s (2007, 192f.) Restaurant Case: After having dinner in a restaurant we split the bill. I know that you are as good as I am at simple arithmetic and you do not seem to be inebriated or distracted. My conclusion is that each of us owes £23 and you come to the conclusion that we have to pay £25 each. Clearly my confidence in my initial conclusion should be reduced.10 The view that we should adjust our credence in our beliefs if they are subject to pd is called Conciliationism.11 The main competing view is Steadfastness. The idea behind Steadfastness is that it cannot be rational to lower one’s credence in a true belief (see Kelly 2005, sec. 4), and that proponents of Conciliationism ignore the asymmetry between views that are better supported by evidence and those that are not.12 Conciliationism leaves open the question of how much one should change one’s credence in cases of disagreement. The Equal Weight View, for instance, suggests that I give equal weight to my belief and the belief(s) of my peer(s) (see Elga 2007). It is also possible to have a Conciliationism, which requires attaching relatively little weight to your peers’ beliefs and more weight to your own – such a conception is called “Extra Weight View” (see Elga 2007, 485).13 In what follows, I will only discuss whether we should change our credence in cases of pd at all, and, more specifically, discuss an objection against the idea that we should adjust our credence.
The spineless sceptic does not doubt the existence of the external world or of morality – after all, most epistemic peers would disagree with such a sceptical view. “Spinelessness” expresses the worry that a Conciliationist has to reduce her credence in controversial beliefs up to the point where she suspends these beliefs. It is important that Spinelessness is commonly seen as being aggravated by, or as resulting from, the high number of epistemic peers in a debate (Elga 2007, 484):
S1:There are fields such as philosophy, ethics and politics where disagreement is very wide-spread, i.e., many well informed people disagree about many different issues.
S2:Many people in these fields are my epistemic peers.
S3:Therefore, there is a lot of pd in these fields.
S4:A Conciliationist response to S3 is to drastically lower one’s credence in any of the controversial beliefs one holds in these fields. Drastically lowering credence in a belief can lead to suspense of judgement regarding whether or not this belief is true.
S5:This would amount to suspense of judgements concerning many questions, or to a form of scepticism one might call “spinelessness”.
Even if one does not believe that we should give equal weight to peers, it might always be rational, according to Conciliationism, to suspend judgment concerning a controversial issue, if disagreement amongst many epistemic peers abounds concerning this issue.15
Consider an issue on which you count many of your associates as epistemic peers. If the issue is at all tricky, your peers undoubtedly take a wide spectrum of stances on it. (This is especially true if your peers are philosophers.) The equal weight view then requires you to weigh each stance equally, along with your own. But that requires you to think, of each stance, that it is very unlikely to be right. Typically, it will follow that you ought to suspend judgment on the issue.
There is an easy reply to Spinelessness: Biting the bullet and embracing scepticism concerning controversial cases. In fact, some authors such as Christensen (2007, 215) suggest that it would be for the best of philosophy if philosophers were not to hold their positions “with the sort of confidence that philosophers often seem to have about so many controversial issues”. Christensen (ibid. 214) takes this to be an opportunity “for epistemic improvement” in the contested subject areas.
In addition, there are fields in which embracing scepticism seems to be a particularly unattractive option. One of these fields is ethics. For the sake of simplicity, I will in this paper focus on disagreement about concrete moral cases. A spineless sceptic will not be a sceptic about all ethically relevant cases, but only about the contested ones. These are, however, the majority of cases that philosophers deal with, since these cases afford ways of clearly differentiating one’s own theory from competitors.
If there’s some hope of making progress in philosophy, it’s not implausible that we need people vigorously defending positions in a way that people wouldn’t if they didn’t have more confidence in their positions than is rationally warranted. There may be some good—even epistemic good—in our epistemic hubris.
The extensive philosophical debate about how to refute the moral sceptic shows that many philosophers would not accept a response to pd, which tells them to embrace scepticism regarding many controversial ethical cases.16 In fact, many philosophers would consider the suggestion to bite the bullet rather a reductio of Conciliationism. On top of this, it also seems that in some cases embracing scepticism is a problem for the Conciliationist given her aim to reconcile her views with those of her peers. Take as an example a medical doctor who has to make a controversial and irreversible decision concerning the treatment of one of her patients. Say, the patient’s life is in acute danger due to a high-risk pregnancy. An immediate abortion is the medically responsible choice, but the doctor and her patient live in a country in which abortion is generally considered immoral and illegal even in such dire circumstances. The doctor is faced with the decision of saving her patient or obeying the law. Imagine that the doctor knows that many people she considers her epistemic peers (for instance other doctors) advocate that doctors in such a situation stick to the law. The doctor originally favoured performing the operation but she is a Conciliationist and lowers her credence in her conviction to such an extent that she suspends belief about whether or not she should perform the operation and remains inactive.17 In this situation it is not the case that the doctor was neutral or that her decision was in between saving her patient and obeying the law. She did not save 50% of her patient. Rather, she embarked on a specific course of action with fatal consequences to her patient. What the doctor did can be morally evaluated. Depending on a person’s views the doctor’s action was morally blame- or praiseworthy, did or did not violate professional norms of conduct, etc.18 There is an asymmetry in many ethical scenarios in the sense that inaction already favours a specific option and that hence a reaction to disagreement which draws an agent to inaction19 also draws the agent to one of the contentious options that she, as a Conciliationist, cannot embrace.20
Given that most people are not spineless about controversial ethical questions (otherwise these questions would not be controversial), and that scepticism can result in or lead to21 ethically relevant behaviour or actions, a Conciliationist cannot embrace scepticism in ethics for all cases. In many situations inaction is not a compromise but already controversial and can be subject-matter of disagreement. This is an important difference between ethical and other philosophical questions. In ethics there is often no neutral position to retreat to, even if the agent wants to be conciliatory and embrace Spinelessness. Biting the bullet is not as easy as a Conciliationist might think in cases that pertain to ethically relevant actions – unless there is a consensus among one’s epistemic peers that scepticism regarding these ethical questions is the way to go. Conciliationism therefore in some cases fails on its own terms. Instead of taking up a conciliatory position the spineless sceptic who remains inactive might have adopted a very controversial position that will be subject-matter to fierce disagreement.22
In what follows, I will show that Conciliationists can avoid Spinelessness or scepticism regarding many controversial ethical cases.
If Conciliationists want to avoid Spinelesness, they have to show that one of the steps of the argument in the previous section is infelicitous. S2 might be the weakest part of the argument. We saw in section 1 that the peer-relation is quite demanding: The members of an epistemic peer-group must have the same evidence (P1), usually react with equal reasonableness to evidence (P2), and an agent can be (temporarily) discarded as a peer for independent reasons (P3). It is thus not obvious that we have many peers. Furthermore, it is only rational to adjust my credence in the light of pd, if I believe that someone is my peer. In this section, I will introduce the “Enmeshment Reply”23 to Spinelessness and criticise one version of it. In the next section, I will present an enhanced version of this reply
Adam Elga (2007, sec. 12), in a paper in which he seeks to defend the Equal Weight View, argues that in cases of disagreement about ethics or politics “one tends not to count one’s dissenting associates – however smart and well-informed – as epistemic peers” (ibid. 492). Elga (ibid. 482–3) believes that “[o]nly in highly idealized circumstances is it reasonable to defer to someone’s opinion absolutely whatever that opinion might be”. Elga (ibid. 493) argues that people who disagree with me about abortion usually also disagree with me about closely related matters such as social justice, conceptions of personhood, etc. In such a case, however, I would not regard someone as a peer, because I would think these people wrong in many other cases as well, i.e., P2 of the peer relation would not obtain. This reply on behalf of Conciliationism draws on the fact that many issues are closely related to other possible points of controversy, i.e., that issues are enmeshed rather than free-standing. If the Enmeshment Reply is convincing then the transition from S1 and S2 to S3 is blocked: There would be disagreement in ethics but it would not be pd. Therefore, the Conciliationist would not be required to suspend her beliefs.24
Elga (ibid.sec.13) himself addresses a potential objection against the Enmeshment Reply. It might be too quick to postulate that people disagreeing with me about abortion (a, for short), will also disagree with me about different but abortion related issues (ari, for short). Instead, one should rather say that I disagree with my peers on a bigger compound of questions which is composed of (a) and (ari). In that case I could not discard someone as a peer. Our disagreement would simply be a disagreement about the single (complex) subject-matter a+ari. Elga (ibid. 496) replies that the compound a+ari includes such a wide range of issues that setting aside our beliefs about all of these issues in order to evaluate whether someone is a peer or not “is to set aside a large and central chunk” of our ethical and political outlook. Without this chunk, however, there would simply not be a sufficient amount of cases left to determine whether my opponent got it right as often as I did in the relevant field. The entire relevant field would be bracketed.
It is difficult to determine whether it is true that disagreement about abortion (and other issues) is usually or always accompanied by disagreement concerning so many other questions that it is impossible to assess whether a person who disagrees with me is a reliable judge in ethics. Most likely the answer is that some people who disagree with me about abortion will disagree with me on almost everything, while many will disagree with me on some issues and agree on others, and some might even agree with me on almost all related issues. The question of how many people in the field are my peers – i.e. whether S2 is true or not – would require an empirical investigation into how tight abortion and surrounding issues are related for the majority of people.
A reason to think that this investigation would not support Elga’s Enmeshment Reply is provided by Sarah McGrath (2007, sec. 7) who points out that we should not overemphasize the differences between people disagreeing about abortion. Even if these people disagree about abortion and related issues, most likely they will still have more in common with each other than with an Ancient Greek slaveholder or a Nazi. If we have this “background of agreement” (ibid. 106) even with someone who disagrees with us about abortion and related issues, it seems possible to tell whether this person is in general a reliable judge of ethical matters. McGrath even suggests that in most cases this person is a reliable judge, since the person repudiates slavery and at least certain racist prejudices, etc., just as we do.25 I think that whilst Elga is right to point out that the notion of epistemic peerhood is highly idealized and hard to meet, it is insufficient to wonder only about how we assess whether or not another agent is able to correctly respond to evidence. We rather have to question the assumption that there is one shared set of evidence available to me and my alleged peers, and that I can easily know what facts and evidence are available to others and what they consider as evidence.
The Enmeshment Reply in general and Elga’s main idea to attack S2, i.e., the notion that we have many epistemic peers in controversial areas, is plausible, I believe. We saw, however, in the previous section that attacking P2 might not get us very far. I therefore suggest taking a closer look at P1, i.e., the requirement that my peer and I have the same evidence. By “having the same evidence” I mean that my peer and I are aware of the same relevant considerations and that we count the same considerations as relevant for the debate and as providing reasons for (and against) beliefs. It is not sufficient for establishing epistemic peerhood that we simply share awareness of the same considerations. We would not think of an agent as an epistemic peer if she is aware of something that I believe is of paramount importance for an issue, but she does not see it as evidence at all. In this situation, we would rather consider this a strong reason to doubt that an agent is able to judge about an issue as reliably as we do, i.e., that P2 obtains. It is important that in this case the reason why P2 fails to obtain is not because the other agent is less competent to react to what she considers evidence, but because we do not consider the same considerations as evidence. In an important sense, we do not share evidence. The violation of P2 is rooted in P1.26
An alternative interpretation would be that, instead of disagreeing about what constitutes evidence, we merely disagree about the clout or weight we should assign to the facts that we agree count as evidence. This would mean that we do satisfy P1 but not P2. This, however, strikes me as an unnatural diagnosis for many cases. It would, for instance, be strange to say that I consider the fact that the Pope is opposed to abortion evidence that abortion is immoral, but that the bearing this evidence has for me is 0. It is more natural to say that I do not consider what the Pope says about abortion as evidence.
In order to show that P1 rarely obtains in ethical cases, I will develop the Enmeshment Reply that controversial issues in ethics, such as abortion, are not free-standing but linked to other issues, and that it is often unclear what issues these are. Elga postulates this, and I think he is right, but he does not explain and substantiate it sufficiently, and he does not discuss it in the context of P1.
The phenomenon that underlies the Enmeshment Reply and that I want to spell out further is what I will call the “global outlook”.27 An agent has a global outlook if she does not make her normative judgements in a void but based on similar and related cases and on notions of what she values, what matters to her, what she thinks is morally (im)permissible, etc. The most general issues ethical questions are linked to are ones such as “What is a just society?”, “How ought I to live?”, “What kind of person do I want to be?”. The often semi-conscious answers to these questions constitute the core of an agent’s global outlook. I take it to be an uncontroversial empirical claim that all rational agents have a global outlook, since all agents (explicitly or implicitly) assume certain answers to the above questions and these answers matter for her responses to more concrete ethical questions. Furthermore, I do not make any assumptions about whether there are elements that are stable throughout all global outlooks. Some global outlooks could be very exotic and radically different from all others.28 No particular moral theory is implied by the mere having of a global outlook, and agents with a global outlook might be staunch egoists or nihilists.
One’s global outlook contains a conception of who one is, a notion of what matters in life, ideas of how the social world should be structured and how important it is that this state of affairs be reached or maintained. The global outlook is the basis of an agent’s evaluations of his actions, as well as of the actions of other agents, and of the state of the world in general, and it also influences what an agent thinks matters (provides evidence) for particular questions. Whilst I think that the global outlook is particularly important for ethical questions, it can of course have a bearing on other questions as well. It depends on an agent’s global outlook which questions these are. It would be difficult to imagine a global outlook that does not touch upon any ethical issues, whilst a global outlook that mainly concerns ethical issues is probably the norm.
Besides being relevant for our notion of what counts as evidence or what matters for moral questions, the global outlook also influences what we do not count as evidence, or what we think we can safely ignore, and whom we should not pay attention to. The global outlook thus also limits the set of facts that we take to be reason-giving or evidence for our ethical views, and it can be the source of epistemic phenomena such as motivated beliefs and that agents sometimes do not want to know certain things.29 An agent’s global outlook reflects cultural norms acquired as part of education, but it also contains idiosyncratic elements that result from personal experience. It is important that not all the convictions that form the global outlook are explicit. Many convictions might guide the agent’s evaluations tacitly.30
To clarify my notion of a global outlook let me quickly contrast it with two other notions: intellectual temperament and intuitions.
An intellectual temperament, as it was recently introduced by Ruth Weintraub (2013), determines “the correct attitude to a proposition which is not entailed by the evidence” (ibid. 753). Her intellectual temperament determines how suspicious a person is, whether she rather withholds judgment or risks false beliefs, and how she balances certain desiderata of a theory, such as simplicity and explanatory strength, against each other (ibid. 754). Intellectual temperament becomes relevant only once it is clear what the evidence is, and it influences how the agent deals with insufficient evidence. It therefore presupposes the global outlook, which influences, and to a certain extend constitutes, our notion of what counts as evidence. Note, I am not assuming that global outlook completely determines an agent’s notion of evidence. There can be considerations so clearly relevant that they are evidence no matter what the specifics of an agent’s global outlook are. It would be epistemically irrational for anyone to ignore or discard them (see below).
The question of how global outlook relates to the notion of an intuition is important, since some forms of peer disagreement might be described as two equally rational agents having different pre-theoretical intuitions about a specific case.31 I believe that often, when we talk about intuitions, in particular intuitions pertaining to concrete cases, we mean certain beliefs that are unconsciously inferred from elements of our global outlook. These beliefs are still pre-theoretical in an important sense since our global outlook is not a theory, but they are not immediate since they are not independent of everything else the agent holds to be true. Other intuitions, in particular those pertaining to general principles or to the question of what matters for ethical assessment, might be non-inferred from our global outlook and are rather central elements of this outlook themselves. In any case, I think we should resist a picture that views peer disagreement as mainly a clash of intuitions. This certainly does not do justice to the Restaurant Case, and it also misrepresents the debate about abortion and other ethical issues. Whilst intuitions certainly do play a role here, agents frequently appeal to other evidence than intuitions and, in cases of disagreement, they will see a need to justify or support their intuitions further, for instance by trying to find common ground and argue from there.32
My suggestion is that controversial issues in ethics always touch on an agent’s global outlook. This is to a large extent the reason why they are controversial and why they are often subject of public debates, which involve large parts of society. Public debates concern questions most members of society care about, because they are related to their self-conception, self-worth, and to what they believe is owed to others as well as to themselves. Furthermore, members of a society who are not professional philosophers are usually much more invested personally in these debates than in debates about the reference of singular terms or in specific conceptions of the persistence of objects through time. The latter kinds of debates are less closely linked to the (non-philosopher’s) global outlook – if at all.33 I should mention that my idea that ethical questions are of particular concern to agents is inspired by Immanuel Kant (1900ff. v:86) who claims that living up to moral standards is an agent’s “highest vocation” and the capacity to be moral distinguishes humans from other animals and “infinitely elevates” our “worth as an intelligence” (v:162). Kant here articulates the important insight that ethical issues are bound up with an agent’s self-esteem and self-understanding in a way nothing else is.34
The global outlook systematizes an agent’s ethical convictions into a picture of the world. This does not mean that a global outlook always consists of a set of mutually coherent abstract principles and beliefs. I take it that the global outlooks of many agents contain principles and beliefs of various degrees of abstraction, which stand in tension with or might even contradict each other. Often these tensions will not be apparent to the agent herself. If they become apparent, agents are of course rationally required to sort out the tension by dropping or modifying one or more of their beliefs. My credence in my belief that abortion is morally permissible will presumably increase, if I believe that current practices and laws should, amongst other things, aim to correct the results of the political marginalization of certain groups (in this case women), and if there is evidence that current (legal and social) restrictions on abortion are the result of such marginalization. If, by contrast, I discover that my views about abortion clash with beliefs that have a very firm foundation in my global outlook, I am likely to change my mind about abortion. Alternatively, I could instead adjust my global outlook in this case. What I need for my argument is only that the global outlook of an agent is relevant when it comes to ethical controversies, and that it is relevant even on the level of what agents count as evidence. This does not presuppose any specific conception of how the global outlook and an agent’s view about concrete cases are interrelated in concrete cases.
If controversial ethical questions touch upon our global outlooks, I cannot simply suppose that interlocutors with whom I disagree have the same evidence as I do. What an agent counts as her body of evidence depends at least to a large extent on her global outlook. The issue of the moral permissibility or impermissibility of abortion, as already discussed for one specific case in sec. 2, is a good example of how the global outlook impacts what an agent counts as evidence and thus of my version of the Enmeshment Reply. There are three levels we need to take into account here. Firstly, there are certain considerations that are known to everyone who has some idea about a debate. In the case of abortion these will presumably be whether abortion is legal or illegal in a society (and maybe some views as to when it is legal and when illegal), and basic physiological facts, such as how long a pregnancy usually lasts, etc. We can assume that these considerations are widely shared and that they will be relevant for everyone, albeit some agents will attribute more weight to some of these considerations than other agents. We think that everyone has to give some weight to them or, if someone does not, she must tell a story as to why others are wrong to treat those as evidence. If someone simply disregards these considerations, they are rationally criticisable, no matter the specifics of their global outlook.35
Secondly, there are considerations that might be widely known though their status as evidence is controversial. I might count facts pertaining to the social and political role and representation of women (and lack thereof) as central evidence for the debate, whereas someone with a different outlook might disregard these as evidence and instead regard the content of certain scriptures, the verdicts of authorities, wide-spread gut feelings, etc., as evidence.
Thirdly, there are also considerations that are known to some and which count as evidence for these people, but which are unknown to others. An example would be considerations concerning the financial, social and personal hardships that (unwanted) pregnancies involve or the psychological damage they can do. Depending on ones global outlook one may actively seek to learn about this, since one is aware that hardships and psychological damage impact the morality of abortion. Agents might also completely ignore these issues, and even actively do so, if their global outlook disposes them to consider them to be a distraction (or a rhetorical strategy) and to be missing the main point of the debate (such as the sanctity of life, for instance). Knowledge of some basic facts pertaining to ethically relevant issues is presumably independent of our global outlook and everyone should consider these facts to be relevant. Other considerations we only came to know because we actively sought to learn about them, since for a person with our global outlook knowledge of these considerations matters.
To give an example: There is a cartoon entitled “The moment abortion becomes universally legal”. It shows a man saying to a woman: “Honey, I think, I am pregnant”. Depending on your global outlook this will either appear to you as a witty case for abortion or not even as funny at all. If you think that it matters that decisions about the legal status of abortion have traditionally not been made by those who are disproportionately affected by it, the cartoon serves to make this asymmetry and the underlying power structures apparent and shows a disconnect between how decisions about the legality of abortion were made and whom they primarily affect. If you, however, believe that the legal status of abortion should be determined entirely by scripture, the sanctity of life, abstract moral principles, etc., you will believe that it matters little how we came to have the views and social sanctions we have.
My discussion reveals two important constraints on epistemic peerhood in cases of disagreement that touch on the global outlook: one that I call “factual” and one that I call “epistemic”.
The factual constraint: Other agents might count different considerations to be evidence than I do. If I do count the right to bodily self-determination as evidence for the moral permissibility of abortion, and you do not, we can hardly consider each other epistemic peers regarding abortion, even if we know the same considerations, and even if we both react equally reliably to what we both acknowledge as evidence. In the sense that matters for counting someone as my peer, you do not satisfy P1, since you regard as irrelevant for the question under scrutiny something I consider reason-giving and important. This makes you someone who is not in the same situation as I am regarding the evidence at hand: From my perspective, you fail to see important things as evidence (or you see as evidence what I consider irrelevant).
I might, of course, mistakenly believe that someone is my peer, but as pointed out in section 1, this belief will be less stable than the correct belief that someone is my peer, since it should become obvious after some interaction that we take different things to be salient. Depending on how different our views about evidence are, it would take a grave misjudgement on my behalf to count you as my peer and to maintain this belief after I am acquainted with your view. In addition, that we disagree will presumably make me prima facie suspicious of your views, and I will actively try to find out whether you know the same considerations and consider the same considerations as evidence. This is particularly so in cases that I am personally invested in.
Barry Lam (2011, 235) entertains the idea that epistemic peerhood might be “a warranted default assumption about someone when we lack evidence about his or her epistemic reliability”. This might be the case in some of the situations Lam has in mind, such as weather forecasting. In ethics, I believe, the assumption of peerhood is not a default when we encounter disagreement. The objects of moral disagreement matter greatly to our self-understanding. Agents are therefore more reluctant to attribute peerhood, and this is not irrational as long as they are not discarding another agent as a peer merely due to the fact that they disagree about a specific issue. When it comes to cases that greatly matter to agents, they are entitled to be reluctant to endorse a dissenting agent as a peer, but agents must always be open to the possibility that dissenting others, after examination of their evidence and epistemic competences, might turn out to be peers.
The epistemic constraint: Knowing what my interlocutor counts as her body of evidence presupposes familiarity with her global outlook. We need to be aware for instance what considerations she takes to be relevant for an issue and what weight she attaches to them. If we are unaware of this, a person might be our epistemic peer, even though we would not be rationally required to adjust our credence in the case of disagreement. It would be as if we had a peer on the dark side of the moon that we know nothing about.36 Being sufficiently familiar with the global outlook of an agent presupposes a familiarity we do not have with many interlocutors and certainly not with people we only know briefly, from one or a few discussion(s) about abortion, or from reading an academic paper, or from listening to one of her talks, etc. It should be noted that matters might be different when it comes to discarding someone as a peer. This can happen quickly and without being very familiar with the other person, for instance when the other exhibits commitments to beliefs that strike one as obviously false, and when it is plausible to assume that these commitments are central for this person.
My version of the Enmeshment Reply differs from Elga’s view elaborated on in the previous section in two ways. Firstly, Elga believes that abortion and other ethical questions are different not because they are practical questions or connected to our global outlook, but because they are “messy real-world cases […] tangled in clusters of controversy” (Elga 2007, 493). I agree with Elga that they are, but this is just a description of how these cases appear to us. My notion of a global outlook provides a rationale for why these cases are more messy and controversial than Restaurant Cases: They are linked to a bigger picture of what the agent considers valuable, obligatory, relevant, good and bad, etc., and this picture differs among agents and it frames the way an agent thinks about these issues. Secondly, and more importantly, Elga (ibid. 487) assumes that even non-peers have the “same batch of evidence”. I contest this. In cases concerning our global outlook, the nature of evidence itself is controversial in the sense that it is unclear which facts are to be sought out and which considerations count as evidence. I normally do, and can rationally do so, discount an agent as a peer for two reasons: Firstly, she counts different considerations as evidence than I do and I am aware of that. Secondly, I do not sufficiently know what considerations she counts as evidence in the first place. Disagreement can go deeper than Elga acknowledges.
Even if we think that a good part of what it means to react reasonably to evidence is the ability to pick out those considerations that function as evidence for a question (i.e., if some of what I take to fall under P1 falls under P2), my view would still substantially differ from Elga’s. According to my view, issues touching on global outlook are more complicated than other issues, amongst other things because depending on one individual’s global outlook this individual will see different facts, arguments, data, etc., as salient and important to enquire into. Hence different agents presumably often fail to share knowledge or awareness of the same considerations.
Finally, I should mention that I agree with Elga that it can be the case that we do share the same evidence with someone and that we know this, but still do not consider the other person as our peer, since our global outlook not only determines what we count as evidence, but also how to respond to this evidence. We can discard someone as a peer who has a history of reacting inappropriately to things that we both consider as evidence. My point is, however, that the typical case of disagreement about ethics (and the reason why these cases are messy) is that it is contested what to count as evidence and what facts to seek out, not merely how to react to shared evidence. The latter is a tamer version of the usual form of disagreement about ethical cases.
The view presented in this paper can be summarized thusly: In genuine cases of peer disagreement it is rational to adjust credence. However, it is also rational for agents to be careful as to whom they admit as their peers in areas that touch upon their self-understanding, self-worth and worldview. Conciliation is required only when one can be reasonably certain that someone else is in an epistemically equally good position regarding an issue, and this certainty is difficult to acquire for many controversial ethical issues. I argued that, due to ethical questions being linked to our global outlooks, being aware of the same considerations is not sufficient to establish epistemic peerhood. We need to, at least approximately, count the same considerations as evidence. Furthermore, knowing what someone counts as her body of evidence presupposes a familiarity with this person we do not have with most people. Therefore, most people are not our peers in a sense that would require the Conciliationist to adjust her credence. S2 is not true for many issues linked to our global outlook, and these issues are primarily ethical ones.
I would now like to add four qualifications to this version of the Enmeshment Reply. Firstly, of course I do not deny that there is much disagreement and heated debate in ethics. My argument is only supposed to show that disagreement in ethics often is not the kind of disagreement that qua peer disagreement gives reasons to lower credence in one’s contentious beliefs. Secondly, my argument does not aim to establish that agents with different global outlooks can never be epistemic peers. My point is that people with sufficiently different conceptions of evidence regarding an issue are not epistemic peers regarding this issue, and that what we consider evidence is influenced by our global outlook. It is possible that two agents with different global outlooks consider the same considerations to be evidence for a question. Thirdly, I do not deny that it can be rational to suspend one’s judgment in complex and controversial cases in which it is unclear what counts as evidence. There are many reasons why this might be rational, for instance, because we believe that the case is too complex for us to figure out. My argument only shows that in complex and contentious cases it is often not clear that there is peer disagreement and thus not clear that there is specific pressure on the Concilaitionist to suspend judgment.37
Fourthly, even on my view it is of course possible that we do know someone sufficiently to accept her as an epistemic peer and disagree with her. In this case, it is rational to lower the credence in one’s belief. Whether this will lead to suspension of belief depends on how much weight we believe is rational to attach to the dissenting peer’s beliefs, on how strong the original credence in our belief was, and on the strength of the credence the other agent has in hers. Even if we attach little weight to each of our peers, the two constraints I have discussed rule out that the sheer quantity of people disagreeing with us will necessitate suspense of judgement. In addition, they rule out that, due to entrenched controversies in certain fields of ethics, agents ought to suspend their judgements simply because there is a large camp of people advocating one view and other large camps advocating positions incompatible with the first. Furthermore, it is important to bear in mind that an agent whom we believe to be aware of roughly the same considerations about an issue as we are, to count the same considerations as evidence, and to be equally good at responding to evidence, will presumably very often agree with us. Such an agent usually does not pose a threat to the credence we have in our beliefs. However, if we find that such an agent does disagree with us, there is very good reason for us to lower our credence. After all, there is no reason why in this situation we should be prima facie more likely to get it right than she is.
Finally, let me point out a further strength of my account and one problem.
Strength: Not only does the notion of a global outlook provide a response to the objection from scepticism in the face of peer disagreement, it also supports Conciliationism when it comes to what might be labeled the problem of peer agreement. This problem is that, if I have very many peers who agree with me, it seems that, for a Conciliationist, the overwhelming reasons for my view would be that my peers agree with me. The evidence I have for my view independently of peer agreement, even if it is strong, could be dwarfed by the sheer amount of peer agreement. It would be very odd, however, if I claim that we should not torture babies for fun, chiefly because all of my many epistemic peers think that this action is wrong. I would look like someone who has not really understood what it means to torture babies for fun and who just parrots others’ views. If we cannot simply assume that others are our peers in ethical questions then agreement with others also provides less (though certainly some) support for our views, and this is a good thing.
The problem: My argument might solve one specific kind of worry that Conciliationists who want to avoid scepticism have: Spinelessness due to wide-spread peer disagreement concerning concrete moral cases. There might be another kind of scepticism, though, one that could be even more worrying: Scepticism about the correctness or adequacy of one’s global outlook. After all, my argument establishes only that agents are often in a position to discard each other as epistemic peers. This leaves open the possibility that agents may doubt that their own global outlook puts them in a stronger position than others in acquiring true beliefs. If it is not reasonable for me to think that my global outlook is more likely to facilitate the search for true beliefs than the outlook of those who disagree with me, the sceptical pressure seems to have come back.
In response, let me first point out that doubts about the epistemic adequacy of my own global outlook in comparison to other peoples’ global outlooks fall outside the scope of the classical peer-disagreement debate, since here we are not disagreeing about propositions or principles but about our understanding of ourselves and the world as a whole. Here we face a general sceptical scenario: Maybe my whole take on the world is off and thus everything I belief is an unreliable indicator of truth. I admit that there is nothing we can do to establish beyond doubt that this is not the case. In the present paper, I was interested in a very specific route to scepticism regarding certain important and concrete ethical issues, a route according to which there would have been considerable rational pull to suspend judgement concerning these issues. I argued that there is no such pull or at least much less of a pull than is often assumed by critics of Conciliationism. This, of course, does not refute general sceptical scenarios.
Furthermore, it might be the case that a global outlook is the proverbial bedrock that turns the spate, i.e., that an agent cannot consistently question her own global outlook in comparison to other outlooks. After all, what standard were she to use? If I believe that my evidence is genuine evidence, I will also think that my global outlook is to be preferred over the outlooks of those who have a different conception of evidence. Of course, this should not be understood as releasing me from the task of gathering more evidence or revising my standards of evidence. Unless I am a dogmatist, my global outlook will also specify when to adjust my conception of evidence and what would make me distrust something I formerly accepted as evidence.
Last but not least, whilst I have said nothing about disagreement with epistemic superiors and inferiors, I should mention that my argument also entails that we have fewer people in the sphere of ethical reasoning that we take to be our superiors than might be initially assumed. After all, if we disagree with someone about an ethical issue and we realize that this is so because we have different ideas of what counts as evidence in this case, it would exhibit an extraordinary lack of confidence, if we then simply regarded the other person as epistemically superior. It could be that, as a result of a discussion with the dissenting person, we realize that we have ignored something important and adjust our conception of evidence accordingly. We would then become the other person’s epistemic peer, assuming this was the only difference between us. It would seem exceedingly strange if we admitted that the other person has a better conception of evidence, which would make her epistemically superior, but we did not adjust our conception of evidence in turn. Furthermore, it is not clear to me whether we normally should regard someone who has a different conception of evidence as epistemically inferior. It could be that we simply conclude that he is coming from a very different epistemic background, one we cannot (currently) sufficiently understand and assess. In this case, we might need a fourth category: an agent who is neither our peer, nor epistemically superior nor inferior.
Underpinning my Enmeshment Reply is the idea that ethical questions require a different approach than many other questions. Ethical questions are intimately connected to our self-understanding and self-worth, to responsibility, and to notions of punishment and reward. My discussion of pd was an application of this more general view. Restaurant Cases offer too simplistic a model for understanding more complicated cases involving controversy on the level of what counts as evidence. Ethical cases, due to their link to an agent’s global outlook, are paradigms of complex cases that make the obtaining of a peer-relationship much more challenging. The notion of an epistemic peer is highly idealized, but the weak spot of this notion is not that peers have to be equally good in assessing evidence, but rather the more fundamental points of whether someone takes to be evidence the same considerations as I do, whether someone is aware of the same considerations as I am, and whether I am in a position to know or at least believe these things. I argued that due to ethical controversies touching on the global outlook of an agent, it is unlikely that I am very often warranted to regard someone I disagree with as my peer in ethical controversies.
This is, however, not a full-blown rejection of Spinelessness, the argument against Conciliationism from Scepticism. An opponent of Conciliationism could concede that we do not have many epistemic peers in ethics, whereas in other fields we do. It therefore remains a task for the Conciliationist to show that these other fields either do not exhibit the same widespread disagreement as ethics does, or that controversial issues in these fields are also often linked to the global outlook. If neither of these strategies is successful, the Conciliationist might have to bite the bullet and endorse Scepticism in these other fields. The crucial issue for the Conciliationist, if she wants to pursue the strategy suggested in this paper, is whether a case of disagreement concerns a question for which it is controversial what counts as evidence for the correct answer. In the Restaurant Case, this seems uncontroversial, in ethics it is often controversial. Is this a specific feature of ethical questions or does it generalize to many other fields, such as the sciences, the good life, abstract philosophical questions, etc.?
It might well be that the point I made in this paper is not one about ethics in particular, and that ethics is rather an apt illustration of a broader phenomenon. This phenomenon is such that many of the interesting cases of disagreement are those that touch on a multitude of relevant and related issues, and such that an agent’s views on these issues influence what she thinks matters for the case in question. Furthermore, in many of the interesting cases of disagreement something is at stake for those involved in the debate and they will not consider someone a peer as easily as they would in questions of simple arithmetic.
I am grateful to Jessica Brown, Philip Ebert, Steven Hall, Yann Allard-Tremblay, Jens Timmermann, Martha Alicia Trevino-Tarango, Renaud-Philippe Garner, Bren Markey, James Camien McGuiggan, Mario Brandhorst, Kerrin Jacobs, Jörg Schroth, Bruno Haas, Joe Saunders, Susanne Boshammer, Nikola Kompa, André Grahle, Thomas Schmidt, Felix Pinkart, Maike Albertzart, Jörg Winter, Thomas Grundman, Klemens Kappel, Mikkel Gerken, Thomas Raleigh, David Christensen, John Matheson, Marco Tiozzo, Marian David and Lara O’Muirithe for discussion and feedback. Earlier versions of this paper were presented at the universities of St Andrews, Göttingen and Hamburg as well as at the Conference of the German Society for Analytic Philosophy in Osnabrück, the Workshop on the Epistemology of Disagreement in Copenhagen and the Summer School Rationality, Objectivity, Disagreement in Cologne.
Christensen David 2014. “Disagreement and Public Controversy.” In: Essays in Collective Epistemology, edited by Jennifer Lackey, , Oxford: Oxford University Press, 142–167.
Elga Adam 2010. “How to Disagree about how to Disagree.” In: Disagreement, edited by Richard Feldman & Ted Warfield , Oxford: Oxford University Press, 175–187.
Elgin Catherine 2010. “Persistent Disagreement.” In: Disagreement, edited by Richard Feldman & Ted Warfield , Oxford: Oxford University Press, 53–68.
Enoch David 2011. “Not Just a Truthometer: Taking Oneself Seriously (but not Too Seriously) in Cases of Peer Disagreement.” Mind 119, 953–997.
Feldman Richard 2006. “Epistemological Puzzles About Disagreement.” In: Epistemology Futures, edited by Stephen Hetherington , New York: Oxford University Press, 216–236.
Feldman Richard 2007. “Reasonable Religious Disagreements.” In: Philosophers Without Gods: Meditations on Atheism and the Secular Life, edited by Louise Antony , Oxford: Oxford University Press, 194–214.
Kant Immanuel 1900ff. Kants gesammelte Schriften, edited by the Königlich Preußische Akademie der Wissenschaften, Georg Reimer: Berlin.
Kelly Thomas 2005. “The Epistemic Significance of Disagreement.” In: Oxford Studies in Epistemology , edited by Tamar Gendler & John Hawthorne , Oxford: Oxford University Press, 167–196.
Kelly Thomas 2010. “Peer Disagreement and Higher Order Evidence.” In: Disagreement, edited by Richard Feldman & Ted Warfield , Oxford: Oxford University Press, 112–175.
Kelly Thomas 2013. “Disagreement and the Burden of Judgement.” In: The Epistemology of Disagreement: New Essays , edited by David Christensen & Jennifer Lackey , Oxford: Oxford University Press, 31–54.
King Nathan 2012. “Disagreement: What’s the Problem? or A Good Peer is Hard to Find.” Philosophy and Phenomenological Research 85, 249–272.
Kornblith Hilary 2010. “Belief in the Face of Controversy.” In: Disagreement, edited by Richard Feldman & Ted Warfield , Oxford: Oxford University Press, 29–52.
Matheson Jonathan 2009. “Conciliatory Views of Disagreement and Higher-Order Evidence.” Episteme: A Journal of Social Epistemology 6, 269–279.
McGrath Sarah 2007. “Moral Disagreement and Moral Expertise.” In: Oxford Studies in Metaethics, edited by Russ Shafer-Landau , Oxford: Oxford University Press, 87–107.
Spelman Elizabeth 2007. “Managing Ignorance.” In: Race and Epistemologies of Ignorance, edited by Shannon Sullivan & Nancy Tuana , Albany: Suny Press, 119–131.
Van Inwagen Peter 1996. “Is it Wrong, Everywhere, Always, and for Anyone, to Believe Anything on Insufficient Evidence?” In: Faith, Freedom, and Rationality: Philosophy of Religion Today, edited by Jeff Jordan & Daniel Howard-Snyder , London: Rowman and Littlefield, 137–153.
- Search Google Scholar
- Export Citation
Van Inwagen Peter 1996. “ Is it Wrong, Everywhere, Always, and for Anyone, to Believe Anything on Insufficient Evidence?” In: Faith, Freedom, and Rationality: Philosophy of Religion Today, edited by , Jeff Jordan Daniel Howard-Snyder London: Rowman and Littlefield, 137– 153.
Van Inwagen Peter 2010. “We’re Right. They’re Wrong.” In: Disagreement, edited by Richard Feldman & Ted Warfield , Oxford: Oxford University Press, 10–29.
Wedgwood Ralph 2006. “The Moral Evil Demons.” In: Disagreement, edited by Richard Feldman & Ted Warfield , Oxford: Oxford University Press, 216–247.
White Roger 2014. “Evidence Cannot Be Permissive.” In: Contemporary Debates in Epistemology, edited by Matthias Steup & Ernest Sosa , Oxford: Blackwell, 312–24.
The technical term “peer” was originally introduced by Gutting (1982, 83). For further characterization see Christensen (2007, 188f.), Kelly (2005, sec. 2.3), Matheson (2009, 270).
Christensen (2009, 756–7), for instance, speaks of equality in “exposure to the evidence” and Kelly (2005, 174–5) of equal “familiarity with the evidence”, Elgin (2010, 53) stipulates that epistemic peers “have the same evidence”, Wietmarschen (2013, 397) that you have to be “justified to believe that S has the same evidence about p as you do”.
I want to avoid commitment to any specific theory of evidence. Some things that certainly count as evidence are facts, sense perceptions, a priori (mathematical) truth, certain forms of testimony, statistical data, arguments. This broad and non-committal conception of evidence is necessitated by my aim to discuss both relatively simple cases concerning restaurant bills as well as complex ethical cases. For this discussion it would be dangerous to be too restrictive on what counts as evidence.
One way of approaching the distinction I have in mind is the difference between types of evidence and tokens. If I believe that others’ testimony is generally a good way of obtaining information and you, by contrast, never believe anything other people say, it seems that we have radically different notions of evidence. However, even if we share the same conception of what counts as good types of evidence, we might differ so much in the tokens of evidence we have that we do still not satisfy P1.
See Elga (2007, sec. 11) for discussion of the rational responses to special circumstances, such as obvious mistakes, insanity, etc. The debate about peer-disagreement has lately focused on the definition, scope and role of the independence condition (see for instance Christensen 2011, 1). Worsnip (2014, sec. 2–4 and 6) discusses in detail whether it is circular to discard agents as epistemic peers based on independent circumstances and whether the notion of independent circumstances might be misleading. See also Kelly (2013) for problems with the notion of independent circumstances.
See Killoren (2010, sec. 4) for further discussion of whether the peer-relation has to be satisfied objectively or subjectively or both. King (2012, 252–3) explicitly stipulates as a condition for peer disagreement to constitute a reason to change one’s doxastic attitudes that the peers must have good reasons to believe that the conditions of peer disagreement are satisfied. He calls this the “acknowledgement condition” (ibid. 253). Likewise, according to Enoch (2011, sec. 5), someone is a peer whom I justifiably believe to satisfy the conditions of peerhood (see also Wietmarschen 2013, 397). There are thus four options to interpret the requirements for peerhood in the context of peer disagreement: (i) I believe someone satisfies the criteria (subjective interpretation), (ii) I believe with good reasons/justifiably that someone satisfies the criteria (King, Enoch), (iii) the person does in fact satisfy the criteria (objective interpretation), and (iv) the hybrid version that I believe someone justifies the criteria and she in fact does so (subjective and objective interpretation, Killoren). Of these the first (I believe her to justify the criteria) is the least demanding. Yet, I will show that even on this first interpretation there are in certain areas much fewer peers than we think. Any of the other interpretations would make my case even stronger.
I make two further stipulations on pd: Firstly, pd is not merely verbal or due to peers talking past each other. Secondly, pd concerns questions that have one maximally rational answer or only one position that is adequately supported by the evidence. This latter stipulation is sometimes called “Uniqueness” (see Feldman 2007). For a defence of Uniqueness see Lehrer (1976), and for criticism see Rosen (2001), White (2005, 445f.). White (2014, 312) proposes a weaker version, according to which there is one maximally rational position for everyone with the same set of evidence. Kelly (2010) doubts Uniqueness, but, for the sake of his argument, proceeds as if it was true. This seems to me the most prudent way if one wants to accept the standard framework of the peer-disagreement debate and avoid substantial discussion of Uniqueness.
See also Audi (2015, 68) and Worsnip (2014, 5).
See also Kelly (2010, fn. 2) who correctly points out that the track record “that bears most directly on questions of comparative reliability [of agents] will be much easier to come by in some domains than in others”. In the present paper, I will show why the track record is so difficult to establish in ethics.
According to Christensen (2007, 206f.), this is rational even if I am right. However, in his more recent writings he acknowledges that if I am right, it is not necessarily the overall rational option to drastically alter my credence in the face of pd (see Christensen 2011).
For a more detailed characterization of this view see Christensen (2009, 756), Matheson (2009, 269).
For a discussion of this asymmetry see Kelly (2010, sec. 5.2). Besides Kelly (2005) in his earlier writings, Van Inwagen (2010) is a proponent of Steadfastness. Weintraub (2013) has recently defended the view that steadfastness can be rational for both parties of a disagreement (not only for the one who got it right).
This leaves open that there are options in between maximal conciliation and complete steadfastness. Kelly (2010), for instance, proposes a mitigating standpoint with his Total Evidence View. Kelly himself does not consider this a brand of Conciliationism, since he takes Conciliationism to require giving equal weight to disagreeing peers (see Kelly 2013). By contrast, Christensen, the main proponent of Conciliationism, clearly distinguishes between giving equal weight and Conciliationism, with the former being one extreme version of the latter (see Christensen 2011, 2). Worsnip (2014) has developed what he calls a “boringly moderate account”, which encompasses weakened versions of Conciliationism and Steadfastness, according to which the level of adjustment of doxastic attitudes is relative to the particular case of disagreement. Enoch’s (2011) Common Sense View also stresses that the question of what the status of peer-disagreement should be for rational belief revision is to be answered context-sensitive and that there is no general strategy that works alike in all cases of disagreement. That the correct response to disagreement is case-sensitive is already anticipated by Kelly (2010, sec.3). My paper pursues this line of thought and shows why disagreement about ethical cases requires different responses than other cases of disagreement.
Versions of this argument can be found, for instance, in van Inwagen (1996, 275), Plantinga (2000, sec. 13), Pettit (2006), Feldman (2006), Kornblith (2010). See Elga (2007, sec. 4–5), and Kelly (2005, sec. 6) for some other objections against Conciliationism, and Elga (2010, sec. 8), and Christensen (2009, sec. 4) for a defence of Conciliationism.
Lam (2011, 207–8) remarks that scepticism is in particular, but not exclusively, a problem for versions of Concialitionism that require giving equal weight.
Of course, we can make ethical decisions under uncertainty. I assume, however, that there comes a point at which we have so little credence in our relevant beliefs that it is rational to refrain from acting. When this point comes might depend on what is at stake.
In this particular case many people will rather blame society or certain interest groups, not the doctor herself. After all, the problem here is a structural one. My point is not that the doctor’s omission was right or wrong, but that it can be ethically evaluated just as an action can.
I here assume that suspending one’s belief that something ought to be done can lead to inaction. I take it that this is true on every plausible model of an action. I do not claim that we cannot act without a belief, but only that what we believe about the morality of an action has an influence on whether or not we act.
The problem of Spinelessness might become even more apparent if the doctor originally favored not performing the abortion but her peers believe that she should perform it and the doctor suspends her belief as a result and remains inactive. The doctor abandoned her original position, yet what she does is indistinguishable from what she would have done without consulting her peers. This might be an instantiation of the more general problem that actions only have two modalities: do or do not, whereas credence comes in degrees, or we can at least believe a proposition, not believe it and suspend judgment. Actions can be inadequate to reflect the complexity of our propositional attitudes (see Frances 2014, ch. 9).
Admittedly, it is a contentious issue how we should understand the notion of scepticism resulting in or leading to behaviour and actions: Is this a causal connection, or does scepticism provide reasons against actions, or does scepticism undermine motivation? I shall remain neutral regarding this question.
To be precise: not the doing or omission of the Conciliationist will be subject to disagreement, as I talk about disagreement, but her belief that this is the right thing in this situation will be. This problem for Conciliationism is a special case of the more general problem that, if I believe something with a credence of 0.2 and you with a credence of 0.8, we would have to adjust our credence to 0.5 (on the equal weight view), even if everyone thinks 0.5 is wrong. This is an even bigger problem in ethics, since here we can be blamed for inaction. 0.5 is not safer than other options, whereas in other areas suspending belief is usually a safer option.
This, however, does not mean that this non-peer disagreement is absolutely no reason to reconsider one’s beliefs. Elga is also interested in epistemically asymmetric situations.
I abstract from the complication that people might agree for different reasons. One person might be against slavery because it violates her notion of justice, the other might regard slavery as an inefficient economic institution. According to my section 4, both agents here presumably have a very different global outlook and would probably not count each other as epistemic peers concerning the question of slavery and other ethical and social issues.
King (2012, sec.1.2) shows in detail that the requirement to share the same evidence is very demanding. For this purpose he discusses different notions of what counts as evidence, the most restrictive being the “dialectical conception of evidence”, according to which all evidence is discursive and shareable through articulation. Even on this restrictive conception, according to which fewer things can constitute evidence than on other conceptions and we might thus expect more overlap between agents, sharing the same evidence is quite demanding. This is aggravated when we also admit sense perception, intuitions, and background beliefs as evidence and thus create more possibilities for divergence between agents. I believe that sense perception and intuitions can be evidence, though background beliefs should be excluded from evidence. Background beliefs are part of what I call the “global outlook”. Discarding someone as my epistemic peer simply because we have different background beliefs would make reasoning parochial. Background beliefs, however, influence what I count as evidence. It is possible that two people with different background beliefs consider the same things or facts as evidence. It seems to me that these two would be good candidates for epistemic peerhood, if they are also equally reliable in their evaluation of evidence. King (ibid. 260 fn.18) makes basically the same point concerning background beliefs in a footnote.
By “global outlook” I have roughly in mind what the German “Weltbild” and “Weltanschauung” mean. I take the term “global outlook” from Feldman (2006, sec.6.b).
A possible counterexample could be an extreme particularist who makes every decision on a case-by-case basis and without any appeal to general principles or general convictions and previous experience. I am not sure that such an agent would fit the description “rational agent”. If he does, it seems to me that he does have a global outlook which consists in the general conviction that every decision is to be made on a case-by-case basis and without appeal to general principles.
As an example for how the global outlook can create areas of ignorance see for instance Spelman’s (2007) discussion of the difference between the ordinary white American, who is (somewhat intentionally) unaware of the history of racial domination in the u.s., versus the white supremacist, who knows this history but does not take it as evidence for the claim that the grievances of black Americans are valid.
A difficult question is whether the global outlook makes our ethical reasoning irrational. On the one hand, the global outlook seems to contain certain unquestioned assumptions and biases. On the other hand, it is not clear that we could evaluate ethical issues without any global outlook at all. We do not need to discuss this issue here. For my argument, it is sufficient that rational agents have a global outlook and that this outlook matters for our ethical evaluations on the level of what we consider evidence.
See for instance Wedgwood (2006, sec.1) and Killoren (2010, esp. 31). The latter stresses the role of intuitions and propositions about intuitions as evidence that can lead to forms of disagreement in which both parties are rationally justified.
McGrath (2007, 102) specifically emphasizes that “many of us take ourselves to have compelling arguments for our controversial moral convictions” they are not “non-inferential, brute judgments”.
To my knowledge there has never been an angry crowd gathering in front of the St Andrews Arché Research Centre for Analytic Philosophy with signs proclaiming: “Contextualism about epistemic modals is Sin!” or “Jesus hates non-classical logic”. However, to be fair to these issues: Some of them are relevant to questions many people do care about, and maybe there would be angry mobs threatening to torch Arché if angry mobs had more insight into these issues and their implications. It is, I think, fair to say that seemingly far off and theoretical issues will not inspire public outcry, protest, debate, etc. unless the public realizes their relevance for ethical questions.
Kalderon (2005, 8–36) holds a congenial view: Agents are more intransigent in the face of moral disagreement than in the face of other disagreement. Wedgwood (2006) argues against this.
My argument would be more straightforward if what counted as evidence was entirely dependent on an agent’s global outlook, but I believe that this claim would be too strong. I can be rationally criticised for discarding obvious things, even if I have a very exotic global outlook that makes me think that I can in fact ignore it.
My epistemic constraint is congenial to King’s (2012, 261–2) acknowledgement condition: “peer disagreement is a threat to rational belief only if the relevant subjects have reason to think it obtains. The mere fact of peer disagreement carries no significance of its own – at least as far as epistemic rationality is concerned”.