To trim their sail to every wind that blows

by Carl Pierer


The standard case of epistemic peer disagreement has two people going regularly to a café, where they regularly drink and eat. Sometimes A has to pay more, sometimes B. Often, they disagree about who has to pay how much – sometimes A gets it right, sometimes B. This time, A calculates that she has to pay £14, while B thinks A has to pay £12. Both are equally smart, they have access to the same information and because of the stipulated regularity, neither of them is particularly likely to get it right on this occasion. They are so called epistemic peers on this question. What should A do now, in light of the further information that B disagrees with her sum of £14?

Suggestions for what ought to be done when disagreement arises between epistemic peers (EPs) fall either under conformism (backing off the initial confidence) or non-conformism (remaining unmoved). I am not aware that anybody suggested opportunism (losing confidence altogether and taking the opponent's view). All three positions work with the following definition:

Epistemic Peers: A&B are EPs iff:

(i) A&B are cognitive equals (equally smart, diligent, thorough, etc.)

(ii) A&B have access to the same evidence. Whether or not this just means same amount of evidence or exactly the same evidence is disputed, but irrelevant for the present discussion.

(iii) There is no reason to suppose that one of them is particularly apt to figure out the right response in D.

Call such a disagreement between EPs an epistemic-peer disagreement (EPD).

In this essay I will argue that, without a telos in relation to which we hold and modify our views, opportunism cannot be ruled out. First, I outline the arguments for conformism and suggest these cannot exclude opportunism. Second, I present the non-conformist argument, thereby illustrating that an analogous argument supports opportunism. Third, I suggest that only with an agreed telos for our views can opportunism be rejected.


According to (Lackey 2010), conformists hold that if two EPs disagree, neither can continue to rationally sustain their view unless the other party's view can be independently discounted. The conformist relies on the following principle, which we grant:

Rational Uniqueness (RU): there exists “(…) a unique maximally epistemically rational response to any given evidential situation.” (Christensen 2007, p.210) Call such a response MERR.

Note that RU has an immediate corollary: in a disagreement at most one party has the MERR. The conformist argument (CA) can then be formulated as:

1. The disagreement at hand is an EPD.

2. By the noted corollary, at most one of the EPs' views is the MERR.

3. By (iii) both are equally likely to have the MERR.

∴ Both A&B ought to lower their respective confidence that their view is the MERR.

Here, two points shall be noted: firstly, the very idea of modifying one's confidence due to an EPD renders opportunism possible. If you have to change confidence in your view, then – to rule out opportunism – there needs to be a principled criterion for why you shouldn't lose it altogether. Secondly, the argument is not valid as it stands. The premises describe the situation of the disagreement. The conclusion is normative, it states what ought to be done. What is needed is a principle to bridge the is-ought gap in the argument.

The strongest conformist reply to the first issue is given by (Elga 2007). Firstly, he acknowledges that sometimes – albeit rarely – it is “(…) reasonable to defer to someone's opinion absolutely whatever that opinion might be.” (Elga 2007, p.483) However, those are not cases of EPD, hence irrelevant to the discussion. If, Elga continues, the two parties are EPs, then equal weight should be given to both positions. Elga formulates a refined version of the Equal Weight View, which can roughly be characterised as: the amount of confidence you should place in your own view, upon disagreeing with an EP, should equal the amount of confidence you would have in this view prior to the disagreement and prior to having thought through the matter yourself (cf. Elga 2007, p.490).

Elga argues that any non-equal assignment of confidence is subject to the “bootstrapping” objection (Elga 2007, p.488): if it were permissible to give your own reasoning more weight than your EP's, through continuous disagreements of the same sort, you would arrive at the conclusion that you are superior to your initial EP. But the mere fact that there are a large number of disagreements between you and your EP cannot support the claim that you are an epistemic superior – or, conversely, an epistemic inferior.

This reasoning motivates the independence condition (cf. Christensen 2009), which requires reasoning independent of the disagreement in question to evaluate the EP's “epistemic credentials”. Combined with a desire to avoid “bootstrapping”, this informs a supplement to CA, CA2:

1. By CA, both A&B ought to lower their respective confidence that their view is the MERR.

2. If A&B do not assign equal confidence to their views being MERR and they do not have a reason satisfying the independence condition, they face the “bootstrapping” objection.

3. The “bootstrapping” objection must be avoided.

4. By (iii) they do not have a reason satisfying the independence condition.

∴ Both A&B ought to assign equal confidence to their views being the MERR.

If the conclusion of CA is rejected, the present one has to go as well. This entails that CA2 is valid only if CA is. But, as argued earlier, CA is not valid, so neither is CA2. (Elga 2007) ignores the problematic leap from is to ought. This leap then allows for the re-introduction of opportunism:

8372897720_be05554aa7_oFor the sake of argument, accept CA and CA2. Hence, A&B ought to assign equal confidence to their views being the MERR. However, if neither A nor B cares about MERR, they can sustain their initial view without lowering confidence. Equally justified, A can adopt B's view, or vice-versa. This is because CA&CA2 do not show that A&B ought to lower confidence in their views unconditionally in case of an EPD. Rather, they illustrate how A&B's confidence levels in their views have to change to track MERR. What is needed, then, is an argument for why their views should align with MERR – a telos, in other words. Even if CA&CA2 are accepted, opportunism remains an option.

Until the “is-ought” gap is bridged, opportunism can creep in. Any such bridge would be equivalent to a conditional. Because of the normative nature, any justification for such a conditional (or equivalent) will

appeal to one telos or other. Hence, without explicitly and independently formulated telos, conformism cannot reject opportunism.


The popular alternative is to hold that “(…) there can be reasonable disagreement among epistemic peers.” (Lackey 2010, p.2). According to the non-conformist, a disagreement between EPs does not require revision of the initial beliefs. As the most sophisticated non-conformist position is developed by (Kelly 2005), this section aims to show that even Kelly's view cannot rule out opportunism without appeal to some telos.

The correct reasoning view (CRW) suggests that the symmetry between EPs can be broken by the extra weight provided by each party's respective belief that their view is in fact best supported by the evidence. The non-conformist argument can then be formulated as:

1. As in CA.

2. If there is a disagreement between EPs, A&B can take the other to fail condition (iii) of EPs.

3. If somebody fails a condition of EPs, they are not an EP.

∴ A&B can take each other not to be EPs

4. If A&B take each other not to be EPs, they do not have to revise their initial beliefs.

∴ A&B do not have to revise their initial beliefs.

Kelly tries to establish that it is no threat to reasonability for both A&B to hold onto their views and not to modify their confidence. However, an analogous argument shows that it is not a threat to reasonability to be systematically suspicious of the correctness of one's own reasoning. It is easy to see that if A&B take themselves to fail (iii), rather than the other, a similar argument re-introduces opportunism.

Even on a very charitable reading, Kelly does not address the issue of opportunism. At best, he can be understood to think that such systematic self-doubt is unreasonable. Thus, a certain concept of reasonability would be invoked and posited as a telos for our views. This, however, is not a problem for the argument that precisely such a telos is lacking to rule out opportunism. Furthermore, sans an explicitly and independently formulated telos, there is no reason to comply with Kelly's appeal to an intuitive grasp of reasonableness.


The foregoing argument only shows that the arguments supporting conformism and non-conformism do not suffice to rule out opportunism. This, however, cannot establish opportunism as a viable position. Extra reasons to reject opportunism, which are not found in the arguments for conformism or non-conformism, can easily be given. For example, if in an EPD both parties were opportunists, they would just keep flipping views.

What is left to show, then, is that any such extra reason must be derived from some accepted telos. Looking at similar debates in ethics, it seems reasonable enough to suppose that this can be achieved. Unfortunately, this lies beyond the scope of this essay.

As a final remark, it should be noted that opportunism is not a very plausible position to defend. It runs counter our most basic intuitions of what knowledge should be, and the virtues we associate with it. However, much like the sceptic, it serves as a hypothetical opponent, against which theories ought to be tested.


Christensen, D. (2007). Epistemology of Disagreement. Philosophical Review, 187-217.

Christensen, D. (2009). Disagreement as Evidence: The Epistemology of Controversy. Philosophy Compass, 756-767.

Elga, A. (2007). Reflection and Disagreement. Noûs, 478-502.

Kelly, T. (2005). The Epistemic Significance of Disagreement. In J. Hawthorne, & T. Gendler Szabo, Oxford Studies in Epistemology (Vol. 1, pp. 167-196). Oxford: Oxford University Press. Retrieved February 2014, 2015, from

Lackey, J. (2010). A Justificationist View of Disagreement's Epistemic Significance. In A. Haddock, A. Miller, & D. Pritchard, Social Epistemology. Oxford University Press. doi:10.1093/acprof:oso/9780199577477.003.0015

Like what you're reading? Don't keep it to yourself!
Share on Facebook
Tweet about this on Twitter
Share on Reddit
Share on LinkedIn
Email this to someone