What Is Peer Disagreement? What to Do When Someone Just as Reasonable Disagrees

Two people are epistemic peers on a question when they have the same evidence, comparable reasoning ability and comparable care and honesty. If they reach opposite conclusions, at least one has made a mistake, and neither has any reason to think it was the other rather than themselves, since by hypothesis they are equally reliable. That symmetry is what makes the case difficult. Simply maintaining your view requires believing that you got it right and they got it wrong, which the assumption of equal reliability does not support, and abandoning your view entirely means conceding to someone equally likely to be mistaken. The question is what a rational person does with the discovery, and it arises constantly in ordinary life without being noticed.

Philosophers have taken several stances and each has costs:

• •Conciliationism, holding that you should move substantially towards the other view, in the strongest form splitting the difference
• •Steadfastness, holding that you may retain your view, since your own reasoning is available to you in a way theirs is not
• •The total evidence view, holding that the disagreement is one piece of evidence to be weighed alongside everything else
• •The right reasons view, holding that whoever actually reasoned correctly may retain their view, which is true and useless in practice
• •Justificationism, distinguishing what you are justified in believing from what is true
• •Context sensitivity, where the right response differs by domain, which most people apply informally

The strongest conciliatory position implies that widespread disagreement should leave a person with almost no confident views on contested matters, since disagreement on politics, religion, ethics and much else is ubiquitous and involves people who are plainly reasonable. That is called the spinelessness objection, and it is either a reductio or an uncomfortable truth depending on the reader. The steadfast position faces the opposite difficulty, since it appears to license everyone retaining whatever they started with, which makes disagreement evidentially inert and is hard to distinguish from stubbornness. It also generalises badly, since if each party may retain their view on the grounds of their own reasoning, both are entitled to do so and the disagreement is permanent by construction. The intermediate positions are more plausible and less precise.

A related and more common case concerns disagreement between a non-expert and an expert, which is not a peer situation and has its own structure. Deferring to expertise is generally rational, since the expert has evidence and training the non-expert lacks, and the difficulty is identifying genuine expertise from outside the field, which requires judgement the non-expert also lacks. Useful indirect markers exist, including whether a claim reflects a consensus among people who would gain from overturning it, whether the person has a track record that can be checked, whether their reasoning is available for inspection and whether they acknowledge uncertainty appropriately. Disagreement among experts is the harder case, since it removes the option of deferring to the field, and it is the situation that most public controversies actually present.

The practical value lies in a small number of questions the framework sharpens. Whether the other party really is a peer is the first and is generally answered too quickly in one's own favour, since demoting a disagreeing party to non-peer status resolves the problem conveniently and is exactly what a person motivated to keep their view would do. Whether the disagreement is about evidence, about interpretation or about values matters, since those call for different responses. Whether the other party has information you lack is frequently the actual situation rather than a genuine peer case. And the number of people disagreeing matters, though less than it appears, since many people sharing a view may reflect one source rather than independent judgement. The honest application is uncomfortable, which is the sign it is being applied.

Two people with the same evidence and comparable judgement reaching opposite conclusions means one is wrong with no basis for saying which, which is what makes the case hard. Strong conciliation implies suspending judgement on most contested matters, and steadfastness makes disagreement evidentially inert. Demoting the other party to non-peer status is the move a motivated reasoner makes first.