Elise Woodard (MIT), "Epistemic Atonement"

Forthcoming, Oxford Studies in Metaethics

By Elise Woodard

Joe Biden used to be a supporter of the Hyde Amendment, which bars the use of federal funds to pay for abortion in most cases. On June 6, 2019, he reversed his position and denounced the Hyde Amendment, ostensibly after intense criticism from fellow Democrats and people on the Left. You might expect that this change would be welcomed, even celebrated, by liberals demanding the change. Perhaps surprisingly, the opposite occurred.

After the change, Biden was subject to intense criticism, particularly by those who thought he made the right change. Some of the criticism leveraged at Biden appeared epistemic in nature, focused—for example—on whether he had good reasons for changing his mind or for his initial view in the first place. For example, Li Zhao questioned the flawed arguments he offered for changing his mind. Similarly, others criticized his inadequate justifications for changing his mind now, rather than sooner. To justify his reversal, Biden observed that “circumstances had changed” and women’s rights were under attack these days. This led one commentator, Ana Marie Cox, to balk: “so different from all the other days,” suggesting that his reasoning for changing now was insufficient. Jon Lovett echoed this criticism, remarking that “the justification for the reversal doesn’t make a ton of sense.” Relatedly, Guy Branum criticized him for being on “autopilot” and “stuck in 1987,” failing to be attuned to the changing world around him. The fact that Biden changed his mind after intense public pressure did not help his case.

Initially, I found these reactions puzzling. After all, Biden did what they wanted them to do all along! What else did they want? It felt like yelling at someone to clean the dishes but then ridiculing them for having done so.

From one perspective, these criticisms are both natural and appropriate. Biden failed to offer adequate justification for either his previous or his new view. Similarly, he demonstrated a lack of understanding of why he ought to have changed his mind sooner, offering “no apologies for the last position.” From another perspective, though, these criticisms are perplexing. After all, Biden came to believe, or at least publicly support, what those criticizing him thought he should believe. As Cox admitted, “I’d rather he change his mind than not change his mind!” What more did Biden have to do? What more could he do?

Biden’s reversal and the ensuing discourse are what first got me thinking about the phenomenon that I call epistemic atonement. (An alternative name may have been ‘epistemic repair.’) The basic idea is this: just as we need to make up for our moral mistakes, we also need to make up for our epistemic mistakes. A central problem with Biden is that he failed to make up for—much less acknowledge—his previous epistemic failures, including his lack of responsiveness to epistemic considerations and inability to adequately justify his views. The problem wasn’t simply that he failed to respond to moral reasons—though that may also be true. Rather, the worry was in part that insofar as his stated reason for changing his mind was a good one, it was an epistemic reason for him to change his mind far sooner: the attack on women’s rights did not begin in June 2019. The concept of epistemic atonement allows us to interpret these criticisms of Biden—which otherwise may seem perplexing—as both intelligible and potentially fitting. Had Biden offered a good explanation for why he changed his mind, apologized for his past errors, and committing to avoiding similar mistakes in the future, then such criticisms would not have seemed appropriate, had they even been leveled.

Given that we’re fallible, many of us will find ourselves in a position like Biden’s. We will have made mistakes and need to change our minds. How can we do so responsibly? On a more practical level, how can we do that without being subject to further ridicule? As Russ Roberts observes, “if I ever admit I was wrong or ever admit that I had an imperfect view of the other side, or I've come to believe something different, I get savaged.” For both epistemic and practical reasons, agents need to be able to explain to others’ why they made the mistakes and assure them they won’t make them in a future. In short, they need to know how to epistemically atone.

Why is epistemic atonement important? A central reason is this: when an agent makes an epistemic mistake, this often indicates that they are less than fully trustworthy. Hence, others cannot rely on them for their beliefs. (How untrustworthy they will be judged depends on the egregiousness of the mistake; the more egregious, the more work repair work is needed.)  Agents thus need to restore trust that was lost. I offer an epistemological self-help guide for how to achieve this. In particular, I show that agents can restore trust by employing strategies familiar from the moral domain, such as taking responsibility, offering narrative explanations, committing to improve, and engaging in what I call ‘epistemic community service.’ Drawing on philosophical and empirical literature on apologies, I demonstrate how each of these strategies can function to restore trust and trustworthiness.

Importantly, there are several ways in which demands for epistemic atonement can go wrong. First, sometimes agents are perceived as making an epistemic mistake when they haven’t. For example, sometimes we criticizes politicians or scientists for changing their mind, when this change reflects solely a change in evidence rather than any previous epistemic failing. Still, insofar as trust-restoration is important, it can still be valuable for such individuals to engage in atonement strategies. (Hence, for example, the public health institutions ought to have done more to restore trust in light of perceived epistemic failings during the pandemic, even if they were just following their evidence.) Second, it may be imprudent to demand atonement of people who have since changed their mind, even when such demands would be fitting. Leveraging such criticisms may disincentive future changes of mind, both in the target and hearers. Especially given that changing one’s mind is often difficult, we ought not make it harder by pillorying agents who admit their mistakes. On the contrary, there may even be practical reasons to celebrate changes of mind.

“Epistemic Atonement” aims to show that those who criticize public figures for their past judgment often get something right: politicians, scientists, and laypersons alike often incur an obligation to restore trust in light of their epistemic mistakes. By clarifying obligations to atone and why they are important, we can better understand when such obligations count as fulfilled. At the same time, for prudential reasons, we sometimes ought to be careful about demanding atonement. Otherwise, we risk disincentivizing changes of mind, leaving us both epistemically and politically worse off. Thus, articulating when such demands are not only fitting but also prudent is an important project not just in epistemology but also in social and political philosophy.