On Moral Luck: A Defense of Gauguin

In the paper “Moral Luck” Bernard Williams makes the striking claim that the justifiability of an act can depend on its outcome. To illustrate this, Williams appeals to a fictionalized story of the French painter Gauguin. In the story, Gauguin decides that he cannot become a great painter unless he abandons his family in France and moves to Tahiti. Williams claims that Gauguin’s choice to abandon his family has some justification if he succeeds and becomes a great painter, but not if he fails. If he succeeds, he has at least something to say for himself. If he fails, he does not. He is a talentless hack who left his family for nothing.

At this point, most seem to agree that whatever insights the paper contains, Williams’ central claim cannot be correct. I’m not so sure that Williams’ is wrong, however, so I want to take the time here to (somewhat begrudgingly) defend Gauguin.

The first complaint many have is that Gauguin’s choice just doesn’t seem justified in any sense regardless of the outcome. Either way, he abandoned his family. That’s a terrible thing to do. I’m sympathetic to this point, but it sidesteps the problem. What is important about Gauguin’s choice is that he decided to infringe the claims of others on the assumption that this would enable him to do something that would help justify the infringement. The problem with this response is that it doesn't work against structurally similar cases where future action could more plausibly justify infringing someone's claims.

For example, suppose a politician in a state with rampant corruption needs to engage in some corruption to win an election (maybe they need to unfairly award contracts to mafia backed construction companies). They are committed to political reform, but they cannot help reform the government without winning the election. Now, imagine that they give the mafia boss what they want, but afterwards succeed in reforming the government. Here it seems like they could view their choice as unfortunate in some sense, but as ultimately the right thing to do. By contrast, imagine that they are pulled down by the system and become just as corrupt as everyone else. Here it seems like they no longer have much to say for themselves, and they should basically regret their choice. Good intentions alone don’t seem to offer much of a defense. This seems to get at what Williams had in mind (in fact he compares the case of Gauguin to political dirty hands cases at Williams, 1981: 37).

Another common objection is that Williams’ position is either trivial or incoherent (Heuer, 2022). Williams could be talking about “fact-relative” or about “evidence-relative” justification. If Williams is talking about “fact-relative” justification, he is correct, but his point is hardly worth making. Everyone can grant that fact-relative justification can depend on luck. Obviously, when we lack information, the best gamble relative to our evidence might not be the gamble we would be advised to make by someone who had all the facts. On the other hand, if Williams is talking about “evidence-relative” justification, his claim is incoherent. The justification of an action depends on the reasons that there were for the agent to perform the action. The reasons Gauguin had relative to his evidence, however, definitely cannot depend on the outcome.

It isn’t clear, though, that all justifications really need to cite reasons for actions. If I make a terrible chess move for example, I can justify the move to you (in some sense) by claiming that the rules of chess allow me to make that move. The fact that the rules of chess allow me to make a move, however, is not a reason to make it. Williams can plausibly claim, then, that the sort of justification he is interested in is not justification in terms of reasons either in an evidence or fact-relative sense. This seems to be what Williams had in mind anyways (see Williams, 1981: 35 ¶ 3).

A final set of objections claim that Williams’ argument fails because regret doesn’t have the connection to justification Williams thinks it does (ex. Wallace, 2012; Dan-Cohen, 2008). For example, Jay Wallace argues that, in some cases, we cannot rationally regret unjustified choices (Wallace gives the example of someone conceiving a child they know they cannot take care of). As a result, even if the rationality of Gauguin’s regret can depend on the outcome, that does not show that the justifiability of their action can depend on the outcome.

While Williams does focus almost exclusively on regret, it isn’t clear that they need to do this. In the political case offered above, for example, it seems like the politician’s ability to respond to the objections of their citizens can depend on whether or not they succumb to corruption. If they successfully reform the government, they can tell their concerned citizens that they only engaged in corruption because it was necessary to combat corruption. By contrast, if they are themselves corrupted by the political process they can’t seriously make this defense. It is harder to see, though, how our ability to defend ourselves against the objections of others could have nothing to do with justification.

There’s obviously more to say. In summary, though, I think that Williams’ central intuition cannot be easily explained away as a simple confusion. Ultimately, I think we may have to either concede to Williams that there is some sense of justification that is sensitive to luck, or else we may have to put our foot down and say that Williams was just flat wrong here. I’m not sure that Williams was just flat wrong, however, so I think Williams’ central argument from “Moral Luck” presents a more serious challenge than it is often given credit for.