Richard Feldman, "Epistemological puzzles about disagreement," in Stephen Hetherington (ed.), Epistemology Futures (Oxford, 2006), 216-236
This paper distinguishes three kinds of questions one can ask about cases of disagreement:
Feldman discusses several argumentative strategies for defending the possibility of "reasonable disagreements" in any of these senses. One strategy appeals to "private evidence," like different sensory experiences or different "intuitions." A second strategy appeals to different "starting points" or "epistemic frameworks," which are either equally rational or not up for rational evaluation. He agrees that these views can motivate reasonable disagreements in sense 1, but not in the other senses.
A third strategy says that each subject has a special reason to trust her own faculties. A fourth strategy says that when evidence is equivocal, you are sometimes permitted to freely choose from a range of beliefs. Feldman doubts whether these can even motivate reasonable disagreements in sense 1.
So Feldman's overall conclusion is that when peer disagreement is disclosed, the only reasonable response will be to become agnostic/suspend judgment:
[C]onsider those cases in which the reasonable thing to think is that another person, every bit as sensible, serious, and careful as oneself, has reviewed the same information as oneself and has come to a contrary conclusion to one’s own... An honest description of the situation acknowledges its symmetry... In those cases, I think, the skeptical conclusion is the reasonable one: it is not the case that both points of view are reasonable, and it is not the case that one’s own point of view is somehow privileged. Rather, suspension of judgement is called for.
(One shortcoming of this paper is that it only considers cases where the subjects start out believing vs disbelieving, rather than believing vs suspending. Also, Feldman only considers cases where the subjects "fully conciliate" to a single shared doxastic response, or don't conciliate at all.)
Feldman observes that his argument has two stages: (i) it's not the case that just one side is justified in her HO beliefs. (Feldman also seems to be assuming here that it's not the case that both are justified in thinking they're the one who assessed the evidence correctly.) So their HO beliefs are both defeated. (ii) Generally, it won't be reasonable for a subject to become akratic, even if this can be reasonable in some special circumstances. So her FO belief should also be dropped.
In discussing the first strategy, Feldman makes the claim that "Evidence of evidence is evidence. More carefully, evidence that there is evidence for P is evidence for P. Knowing that the other has an insight provides each of them with evidence" (p. 424 in my manuscript). This claim later came under criticism, partly for technical reasons (using a framework that's inhospitable to Feldman's idea), and partly for informal reasons. We've looked at some of the informal issues when discussing Lasonen-Aarnio and Williamson's clock; another objection and Feldman's reply are discussed in the summary to Feldman 2009, below. See also §2 of Lasonen-Aarnio 2013.
In discussing the second strategy, Feldman observes that what is needed to get reasonable disagreement after disclosure is not just that the starting point has some initial justification, but that it retains that justification in the face of knowledge that (otherwise equally capable) subjects have different starting points "with as much 'objective' initial credibility as one's own."
In discussing the third strategy Feldman says that "self-trust cuts both ways," in that one has self-trust-originating reasons to rely on the reliable expert you have identified, too. He also discusses a variant of this strategy which says that it's never beliefs but only changes in belief that are justified, when an alternative attitude is better supported. But in Feldman's cases he thinks there is a better-supported alternative, namely suspending.
A couple of times Feldman makes the (I think unfair) move of assuming that if a subject is justified in giving greater weight to her own insight, she must be doing so "because" or on the basis of the fact that it is her own insight (see pp. 424-5, 428, 441). Compare the discussion of what Elga calls the "Extra Weight" View in our summary of Kelly's papers.
If there can't be reasonable disagreements after full disclosure, is that because of what Elga calls a "Right Reasons" View (like Kelly's), which says it's reasonable for only one of the parties to remain steadfast, or is it instead because both parties should suspend? Section 5 of Feldman's paper argues against the first alternative. Feldman admits that a Kelly-like story may be correct in some cases (disputes about astrology). But in more challenging cases, the evidential situation of the parties looks "symmetrical" to a neutral outsider, and we have to "treat like cases alike" (439).
Richard Feldman, "Respecting the evidence," Philosophical Perspectives 19 (2005), 95–119
This paper discusses two main cases, one where a student is (falsely) informed by an epistemic superior that their perceptual evidence doesn't count as good evidence for their belief that there is a tree on the quad, and the other where van Inwagen contemplates how to respond to the fact that an epistemic peer assesses the philosophical arguments differently on matters about which they disagree.
Feldman describes three views one might take about these cases:
The paper's main project is to argue against the level-splitting view, and to rebut defenses of it. If the level-splitting view is correct, the subjects seem to be in a position to justifiably say "P but my total evidence doesn't support P." That's odd. Moreover, this is something she can't know to be true; and which she can know is unreasonable to believe if it is true.
That said, Feldman allows that claims like this might sometimes be reasonable, for instance if the subject has reasonable false beliefs about what it takes for her total evidence to support P. But he thinks such cases would be unusual, and in his own examples he thinks that the justified HO beliefs do have a first-order effect. (This doesn't address the prospect Worsnip was arguing for, that the evidence might have more of a HO effect than it does a first-order effect, and so can in some cases leave the subjects justified in a pessimistic HO attitude, but nonetheless on balance still justified in retaining their FO belief.)
When discussing putative defenses of the level-splitting view, Feldman sympathizes with the claim that you don't need justified beliefs about the merits of your evidence to have justified first-order beliefs. But the level-splitting view says, more strongly, that your first-order justification can survive acquiring good reason to doubt your evidence.
Feldman acknowledges that there's a way in which the HO evidence in his examples is unlike many other kinds of undermining evidence. The latter often leave a generally reliable connection or first-order reason intact, and just give reason why it can't be relied on in some specific case. Whereas the HO evidence testifies to there having been no generally reliable connection or first-order reason in the first place. Despite this difference, Feldman thinks in his cases the HO evidence does still have an undermining effect on the first-order justification.
Against the Titlebaum/Smithies view: why can't authoritative and reliable testimony from an apparent epistemological expert defeat at least your justification for your contrary epistemic belief? Also, Feldman finds it incredible that it should be impossible to have justified false beliefs about what it takes to have epistemic support.
The Fully-Conciliate/Equal Weight View has an odd implication: it makes the skeptic's conclusions true, at least about subjects like the student in our example. But it isn't the unsound skeptical arguments (by themselves) which have that effect, but only in combination with authoritative testimony.
Richard Feldman, "Evidentialism, Higher-Order Evidence, and Disagreement," Episteme 6 (2009), 294-312
Abstract: Evidentialism is the thesis that a person is justified in believing a proposition iff the person's evidence on balance supports that proposition. In discussing epistemological issues associated with disagreements among epistemic peers, some philosophers have endorsed principles that seem to run contrary to evidentialism, specifying how one should revise one's beliefs in light of disagreement. In this paper, I examine the connection between evidentialism and these principles. I argue that the puzzles about disagreement provide no reason to abandon evidentialism and that there are no true general principles about justified responses to disagreement other than the general evidentialist principle. I then argue that the puzzles about disagreement are primarily puzzles about the evidential impact of higher-order evidence – evidence about the significance or existence of ordinary, or first-order, evidence. I conclude by arguing that such higher-order evidence can often have a profound effect on the justification of first-order beliefs.
Feldman thinks that the disclosure of a peer disagreement about P has some negative first-order bearing for a subject who believes P. But he resists stronger claims that attempt to specify when such a subject wouldn't be all-things-considered justified in believing P. He says we need only to subscribe to evidentialism, plus the thesis that information about disgreeing peers evidentially bears on the first-order questions we disagree about.
Also, what we say about peers who share their evidence should be continuous with our account of the relevance of information about the evidence-assessments of non-peers, or of people who aren't assessing exactly the same evidence as ourselves.
Feldman discusses three kinds of scenarios where you have higher-order evidence:
He observes that when you acquire first-order evidence, you'll nearly always also acquire HO evidence, but that the latter doesn't in those cases make you more justified in your first-order belief. (This bears on how his claim that "evidence of evidence is evidence" needs to be understood.)
He also observes that sometimes philosophers who seem to be in different camps in the Disagreement debate are really just weighting the negative bearing of HO evidence somewhat differently.
Feldman's section 3 focuses on scenarios of sort (a). He asks what attitude a theorist should take to the possibility of epistemic akrasia:
As in 2005, Feldman thinks of the level-splitting view as denying that the HO evidence has any first-order impact.
Against the Titlebaum/Smithies View, Feldman briefly voices the same kinds of qualms he aired in 2005.
Feldman's section 4 talks about scenarios of sort (b) and (c). He revisits his 2006 claim that "evidence of evidence is evidence." Hudson objected: what if I lie to someone that P, then I will have evidence that there is some evidence for P (namely my testimony), but I don't thereby get any evidence for P. Feldman prefers to say you do have some evidence for P, but it's defeated by your knowledge that the testimony came from yourself and was a lie. He also points out that acquiring more evidence for P doesn't always make you more justified in believing P, even when the evidence isn't defeated. (Consider our discussion earlier in term about how "redundancy" in one's evidence can pattern formally like undermining does.)
Feldman also discusses some remarks from Ernie Sosa that in cases of interesting disagreements, some of our own reasons are "inscrutable" or "hard to uncover," and "fails to be fully formed and operative in one fell swoop," and so our reasons cannot be fully shared. Feldman responds:
Sosa is almost surely right about this. But it is a mistake to conclude that this somehow protects the justification of our beliefs in these cases.
Suppose that Sosa is right about our inability to report our evidence. I therefore know that you have been unable to tell me your whole story. I also do not know my own entire evidential story. But I also know your strengths. I have evidence that you have evidence for your conclusion. I know that there are two lives, one that led to belief in P, one that led to belief in ∼P. Both are, I have agreed, generally epistemically worthy lives. I am more familiar with one – my own. I do not see how this makes me better justified in believing what the path I have followed has led to than what the path that you have followed has led to. (p. 310)
Another Feldman paper on disagreement, which I'm not going to summarize here, is:
Richard Feldman, "Reasonable religious disagreements," in Antony (2007)