On Christmas Day of 2022, Marcellin Siliai got into an argument with Charlotte Tyrell, the mother of most of his children, which culminated in his accidentally running over and killing her. Here’s how the New Zealand Herald describes the incident:
As Siliai started the car and began to accelerate, Tyrrell ran to the front of the vehicle to try and make him stop.
Tyrrell accused him of planning to head back to his ex-partner’s house and the couple got into a shouting match. Tyrrell put her hands through the top of the driver’s side window to try and force it down. Siliai drove the car forward to dislodge her but stopped shortly after.
She reached for the keys in the ignition and Siliai drove forward again.
When he accelerated she fell onto the road, under the vehicle and Tyrrell was run over by the rear driver’s-side tyre.Siliai stopped and tried to help her but she died on the road shortly after.
The article goes on to note that Siliai was initially sentenced to two years in prison for dangerous driving causing death, but that his punishment was reduced to six months home detention on appeal. Here is a second story, this time from my own personal experience. Recently, I was riding my bike home from university in the bike lane of a busy road, when I noticed a car approaching from a side street. Vehicles on the side street are required to give way to traffic on the main street. But this car did not give way. Instead, it continued to drive out onto the road. To avoid hitting the car, I slammed on the brakes of my bike, causing my back wheel to skid out. I just managed to keep the bike balanced and slow down enough to avoid hitting the car. If I’d been paying a bit less attention, I’d probably have slammed straight into the car. Would I have died? Who knows, but I do know that whatever happened wouldn’t have been pleasant.
These stories have one thing in common: they both feature someone who committed the moral failing of dangerous driving. But there’s also a key difference between the stories: the consequences of this dangerous driving. Marcellin Siliai killed someone, whereas the anonymous driver caused no negative consequences (apart from irritating me). This is (in part, at least) why Marcellin Siliai is serving a six-month home detention sentence, whereas even if the anonymous driver had been apprehended they would likely have gotten away with a small fine.
Many people have the intuition that this is appropriate. After all, Siliai killed somebody, and killing somebody is worse than merely irritating somebody. And so, intuitively, it seems that each perpetrator received the punishment they deserved. Let’s call this intuition Intuition 1.
But is this intuition correct? When we step back from our immediate intuitive response, another, more abstract intuition emerges which calls this first intuition into question. After all, when we morally evaluate an agent’s actions, what we are interested in is what is under her control. Evaluating an agent more negatively for factors that she could not control seems wrong. Morality, it seems, evaluates an agent for what she can control. But if this is true, then it is unclear how we can morally evaluate the three drivers above any differently. Plausibly, each evinced a similar degree of recklessness in his driving; and the fact that the consequences of the recklessness were different is a factor that is outside the agents’ control. Intuitively, then, it seems as if both drivers are equally blameworthy, and should have received the same punishment. Let us call this Intuition 2.
What we have, then, is a clash between two intuitions: one holding that results can affect blameworthiness, and one holding that they can’t. This clash presents a problem that was the focus of two famous papers from the 1970s by Bernard Williams and Thomas Nagel. Williams called it the ‘problem of moral luck’—a name that has stuck.
Solving the Problem
How can we reconcile these conflicting intuitions? If we cannot reconcile them, which should we discard? Intuition 2 seems fairly secure; it simply seems correct that the degree to which an agent is blameworthy for an action cannot depend on factors outside of her control. So let’s examine intuition 1 more closely. Is it actually the case that we consider an agent who does harm more blameworthy than an agent who is equally reckless but does no harm? I do not think it is.
Imagine that, after Marcellin Siliai runs over his girlfriend, he is greatly contrite, and blames himself deeply. In fact, he considers himself little better than a murderer. How would we respond to this? I submit that we would attempt to convince him that he is being too harsh on himself. Yes, he was reckless, we would concede, but it was merely bad luck that this recklessness led to the death of his girlfriend. He didn’t mean to kill her.
Note that, in convincing him to regard himself as less blameworthy than he evaluates himself as being, we tell him that the consequences of his recklessness were not under his control. This means that, in a situation in which both intuitions should be asserting themselves, we see intuition 2 as more plausible and authoritative. If this were not the case—if intuition 1 seemed more plausible—we would not mention lack of control as a reason for Siliai to stop blaming himself so severely. So it seems as if intuition 2 overrides intuition 1.
What, then, explains the fact that we have intuition 1? I think it is that, when one is responsible for a morally bad outcome, we expect them to take responsibility for it, even if they are not (entirely) morally responsible for the outcome. As Susan Wolf noted two decades ago in a paper called “The Moral of Moral Luck,” we expect people to take responsibility for what they are merely causally responsible for. If, for instance, I am in a store and I knock a vase over, I am required to take steps to remedy the bad consequences of my action—that the vase is broken—even if it was not my fault (perhaps the vase was poorly placed, for instance.) Wolf calls the virtue grounding this requirement the “nameless virtue.”
If Wolf is right, and there is indeed a virtue that grounds such a requirement, then it is this that grounds intuition 1. What intuition 1 is really getting at, then, is that a person like Siliai, whose negligence had tragic consequences, faces onerous moral requirements stemming from those tragic consequences. Intuition tells us that Siliai is in a different moral situation from somebody whose negligence did not result in tragedy. But this does not mean he has done greater wrong. His level of culpability, and therefore wrongdoing, is the same as that of the driver who nearly ploughed into me.
Radically Revising Our Punishment Practices
I’ve argued here that outcomes are irrelevant to blameworthiness. If this is true, then it has some pretty serious consequences, both for our interpersonal blaming practices and for the criminal justice system. Let me set the former aside and focus on the latter. What implications do Wolf's ideas have for criminal justice? What is the moral of “the moral of moral luck”?
First, it seems as if the criminal justice system needs to eliminate discrepancies in punishments for crimes that are merely attempted and crimes that are successfully carried out. After all, if one attempts to commit a murder but fails, one is just as culpable as one who does not fail. The success of the attempt is outside the agent’s control.
Second, in cases of negligence, sentences ought to be based upon the degree of negligence an agent demonstrates in performing a certain action, and not on the severity of the consequences of the negligent action. In other words, it is plausible that Marcellin Siliai and the driver who almost hit me on my bike are deserving of the same degree of blame, and therefore of the same punishment.
Third, our justice system, with its impersonal nature, is ill-equipped to deal with the fact that moral requirements arise when we are causally responsible for bad outcomes but not (entirely) responsible for these outcomes. For instance, when a driver negligently hits a pedestrian in his car, he is typically advised not to visit the victim in the hospital to offer apologies, for it might expose him to further litigation. We aim to deal with perpetrators objectively; but the moral requirements that arise in situations of legal liability are inseparable from the harm that has been inflicted. The demands of morality require interaction between the harmer and the harmed. Without such interaction—without providing the harmer with the opportunity to take responsibility for the harm they have caused—the nameless virtue cannot be satisfied.