For Humility
Why you should be less confident in many of your beliefs
(1/4) Suppose Alice and Bob disagree.1 Each is roughly as smart, thoughtful, and informed on the issue as the other. Who’s right?
Given the symmetric setup, you are likely to shrug your shoulders.
And if, say, Bob were to insist that he is much more likely to be right than Alice because his arguments and evidence are better than Alice’s, you might politely note to him that Alice has just the opposite appraisal of the quality of their arguments and evidence. And, you might add, if Bob isn’t justified in believing his appraisal ability is superior to Alice’s — and given the setup, there is no reason to think it is — this symmetry should be enough to get Bob to join you in thinking he and Alice are roughly equally likely to be right.
(2/4) Now, as it turns out, Alice and you disagree on some other issue. And each of you — just like her and Bob — are roughly as smart, thoughtful, and informed on the issue as the other. Who’s right?
Despite the symmetric setup, you are not inclined to shrug your shoulders. The correct answer is obvious — you. Sure, Alice has some evidence and arguments in support of her view, but her stuff just isn’t as good as yours.
But then Bob reminds you of what you told him. Alice has just the opposite appraisal of the quality of your arguments and evidence. And, he might add, if you aren’t justified in believing your appraisal ability is superior to Alice’s — and given the setup, there is no reason to think it is — this symmetry should be enough to get you to join Bob in thinking you and Alice are roughly equally likely to be right.2
(3/4) Of course, disagreements aren’t always one-on-one affairs. But adding more people to each side shouldn’t change one’s attitudes, so long as the relevant symmetry is preserved.
Suppose instead of it being just Alice vs. You, it was Alice and Alfred vs. You and Yosef. If each is still as smart, thoughtful, and informed on the issue as all the others, does this change now justify believing one side is much more likely to be right than the other? It seems not. And, if we add a third, fourth, or however many equally qualified people to both sides, this also wouldn’t seem to matter.
And even if we introduce heterogeneity to each side such that each has people of varying smarts, thoughtfulness, and informedness, this shouldn't sway one either way to any significant degree.3
So, we end up with the following conclusion: if two groups of roughly equal size and roughly equal distribution of smarts, thoughtfulness, and informedness4 are on different sides of an issue, one should regard the sides as roughly equally likely to be right. Call this the Rough Symmetry Principle.
(4/4) Still, one might wonder how frequently such cases of rough symmetry arise — and whether we fail to be humble when they do.
Notably, there is often no such symmetry concerning scientific issues. You, like I, are presumably quite confident that the earth is round. Accepting the Rough Symmetry Principle shouldn’t decrease your confidence in this belief, for the flat-earthers are not nearly as numerous nor typically as smart, thoughtful, and informed as us globers.5
And when such symmetry arises, we do sometimes have the appropriate level of humility. Suppose you go out to dinner with a friend, and it’s time to pay the check. You and your friend agree to split the cost evenly and decide a 15% tip would be appropriate. You calculate the 15% tip to be $17.53, while your friend calculates it to be $18.23. You, like many others, I presume, would usually become much less confident in your belief upon hearing your friend’s. And, assuming you don’t believe you are notably better at this sort of calculation than your friend, it wouldn’t be all that unusual for you to regard your friend as just as likely to be right as you.
But there are many cases of rough symmetry where we routinely fail to have the appropriate level of humility.
There is perhaps no better example than politics — where people tend to have very strong opinions about very many issues. All despite these issues often bisecting the population, with smarts, thoughtfulness, and informedness similarly distributed on each side. Thus, politics — where we perhaps most often show great confidence — is where the Rough Symmetry Principle perhaps most often recommends great humility.
And our conflicts with the Rough Symmetry Principle are not limited to politics but extend to sports, work, and household affairs. Hence, it seems either the Rough Symmetry Principle is wrong or we ought to be far less confident in many of our beliefs across a wide range of domains. Absent good arguments for the former, one may be inclined to think the latter.
This is a rewrite of a now-removed post.
One might object that you do have reason to believe your ability to appraise arguments and evidence is superior to Alice’s. For, you believe she is wrong about the issue and therefore she — unlike you — misappraised the arguments and evidence in this case. Below are two responses to this objection.
First, even if you should believe your appraisal ability is superior to Alice’s on this basis, it seems that you still shouldn’t believe it is much better. For, the only new evidence against her would be this one case of misappraisal. So, even if the objection runs, it still seems plausible you should judge Alice to be roughly as likely as you to be right.
Second, your apparent reason to believe appraisal ability is superior to Alice’s is based on an unjustified belief. Namely, your belief that Alice is wrong. Before disagreeing you already believed Alice was similarly smart, thoughtful, and informed on the issue as you. So, when you later find out you two disagree, you are not justified in forming the belief that Alice is wrong in the first place. (To be fair, one can imagine cases where someone finds out they disagree with some other before forming a belief about the other’s smarts, thoughtfulness, and informedness. I think the objection has more force in these cases, but they are not cases I intend to discuss here.)
Lastly, if the objection runs, then it also implies Alice would be justified in believing her appraisal ability is superior to yours. And, thus, justified in believing she is more likely to be right than you. A further implication is that each subsequent time you two disagree, your assessments of each other would justifiably become more polarized.
To be clear, the heterogeneity introduced here is intra-group, not inter-group. Just as before, both sides have roughly the same distribution of relevant qualities. For example, if one side has three members around the 90th percentile of thoughtfulness, then the other side has three similarly thoughtful members. The introduced heterogeneity just allows for both groups to have three members around the 90th percentile of thoughtfulness as well as, say, two members around the 40th percentile.
Smarts, thoughtfulness, and informedness are, admittedly, not the only relevant traits. Notably, the independence with which members of each group form beliefs is relevant. But, for ease, I only talk of the first three. This, I claim, is not overly convenient. When smarts, thoughtfulness, and informedness are roughly equally distributed, it’s a decent bet the other relevant traits are too. For, the relevant traits — i.e., the epistemic virtues — tend to be positively correlated with one another.
To be clear, even if the average flat earther was as smart, thoughtful, and informed as the average glober, the Rough Symmetry Principle might still not apply. For, the Rough Symmetry Principle requires rough symmetry in the distribution of relevant traits.
Why impose this restriction? Well, suppose you have four friends debating whether Paris was founded in an even or odd year. Two think even; two think odd. One of your friends who thinks even is extremely smart, thoughtful, and informed on the issue. In fact, he is probably the most qualified person in the world to answer this question. Your bumbling fool of a friend thoughtlessly agrees with him despite not even knowing what “Paris” is. Your two friends on the other side are ordinarily smart, thoughtful, and informed on the issue. Plausibly, both sides have roughly the same average smarts, thoughtfulness, and informedness — despite having very different distributions of those qualities. Because we ought not to conclude that both sides are roughly equally likely to be right in this case, we ought to conclude that the average is not what matters.