Deeply social knowledge and epistemic humility

Does social knowledge production require epistemic humility? Epistemic humility is typically conceived of as the mean between epistemic arrogance and servility. But the dispositions appro­priate to cumulative culture seem more servile than humble.

Of course, human beings are clever animals, and we also innovate, invent, and probe. But suc­cessful innovations tend to build on the platform provided by cumulative culture, because this platform embodies knowledge the adaptive value of which it is almost impossible to replicate de novo.The dispositions to copy, to accept testimony, to adopt the local way of doing things seem more servile than humble.

Much of our most adaptive epistemic behavior therefore seems to manifest epistemic servility, not humility.The role for epistemic humility seems more circumscribed than is usually thought. On this basis, we may think that only in our innovations do we abandon servility sufficiently to be in the humility game. In fact, the scope for epistemic humility is more circumscribed still: when we are not servile, we are often — appropriately — arrogant, rather than humble. Perhaps surprisingly, I contend that we display epistemic arrogance not in counterbalance to our deeply social knowledge generation and acquisition, but in the service of social knowledge acquisition.

There is extensive evidence that when we do deliberate (rather than acquire ways of going on by imitation), groups do much better than individuals, at least under a variety of conditions (Mercier and Sperber 2017). Consider our performance at reasoning tasks, like the Wason selec­tion task.Though the task is logically simple, most people do badly at it: around 10% of people select the right cards on the task. But groups of individuals do very much better, and transcripts of exchanges within groups indicate that success is explained by the exchange of reasons (rather than, say, the recognition that one individual is smarter than the others). Indeed, groups of deliberators may exhibit the assembly bonus effect, where the group performs better than the best individual within it. Even deliberation at its best is often social.

But, notoriously, groups have a bad name (think of books with titles like Popular Delusions and the Madness of Crowds).

We are, it seems epistemic individualists who are reliant on social networks and culture for our epistemic success.Why are we epistemic individualists? Group deliberation is powerful, but is subject to characteristic limitations and pathologies. Information cascades can overwhelm the group; powerful individuals can carry disproportionate weight and people may self-silence in the face of prejudice or anxiety. All of these problems can be mitigated if people are epistemi­cally arrogant: if they are disposed to give their private information and their individual delib­eration greater weight than it deserves in group deliberation.

Information cascades (Bikhchandani, Hirshleifer and Welch 1992) occur when people delib­erate sequentially, and information about previous decisions is available to each. In this situation, each may rationally discount their own private information, because the public information outweighs it. A situation may arise in which the publicly available evidence is highly mislead­ing, while many individuals are in possession of private information that would correct the picture. The problem of information cascades can be reduced if individuals are overconfident: if they take their private information to have a weight greater than it ought, relative to public information. It seems intellectually arrogant to think that one's own private information or one's opinion should be given greater weight than the apparent information of several (or many) indi­viduals who are one's epistemic peers, but this kind of intellectual arrogance can be conducive to group deliberation.

If it is intellectually arrogant to place greater weight on one's own opinions than those of one's epistemic peers, it is even more arrogant to prefer one's own opinions to those of epis- temic superiors. But the disposition to do so may be epistemically fruitful, because self-silencing by those who recognize their inferiority may also lead to ‘hidden profiles'; information relevant to deliberation going unshared. In fact, groups may benefit from what looks like the intellectual arrogance of intellectual inferiors, even when the information they insist on turns out to be misleading (Surowiecki 2004). Again, the apparent individual vice of epistemic arrogance may be a virtue at the level of the group. Mercier and Sperber (2017) have argued that some of our individual-level reasoning pathologies may be adaptations for collective deliberation: the con­firmation bias, for example, which leads us to overvalue evidence in favour of hypotheses we are well disposed to and undervalue evidence contrary to these hypotheses, may conduce to the division of epistemic labor. Similarly, I suggest, our epistemic individualism — our disposition to under-weigh the views of others and to think that group deliberation brings few or no benefits over that of the individual — may be an adaptation for collective deliberation (Levy 2019).

I have suggested that we owe our success very significantly to collective and cultural mecha­nisms of knowledge generation. The mechanisms whereby cumulative culture is generated are very heavily dependent on an epistemic deference so thoroughgoing that it often looks more like epistemic servility than humility. Moreover, our collective mechanisms of knowledge gen­eration may require epistemic arrogance. If something like this picture is accurate, then epis­temically well-functioning agents do not exhibit anything like the virtue of epistemic humility; not, at any rate, in much of their most important epistemic activity. Instead, they are disposed to oscillate between different dispositions: dispositions to servility (in blind deference and imita­tion), to arrogance (preferring their own views to those of acknowledged experts and numerous dissenting peers) and — no doubt — to humility, as the situation demands it.⁶

40.3