There’s no tort of “attempted negligence.” In order for ultimate liability to accrue, a defendant must not only breach a duty owed to a harmed plaintiff, but that breach must be the cause-in-fact of the harm. Someone speeding through the pedestrian stop light in Harvard Square isn’t held responsible for harm done a few minutes later when a carefully driven car just so happens to independently get into an accident with a pedestrian, even though it was mere fortuity that the careless driver didn’t hurt anyone. If I carelessly spill oil on the floor and wander away, and you fall as you turn the corner, causation is pretty clear. But what if you were running so quickly over the area that I can show you would have taken a spill irrespective of the oil? That’s not so easy to prove, but in theory it could defeat causation, just as a negligently maintained dam collapsing in a big storm could be thought of as mere “accident” if the storm were so large that even a well maintained dam would have buckled. Causation can be tricky to prove in other contexts, such as establishing that a particular chemical causes cancer. To what level of certainty must a causal link be determined? Again, the usual answer is “more likely than not,” but other formulations might better capture the moral instinct that requires causation, and even small changes in phrasing can affect a jury’s deliberations. Once we understand that causation is a required element, there are still questions about what exactly it means to say that X caused Y. Should the presence of other causes matter? One common formulation, good enough for most purposes, is “but-for” causation: without the defendant’s negligence, the harm would not have arisen. But what happens when there are two negligent parties, with each party’s act sufficient to cause the harm? Can each point the finger at the other, since one alone is sufficient to have harmed, and therefore neither is a “but for” cause? Not to bury the lede: the answer there tends to be “no”; we can’t allow a surfeit of negligence across multiple parties to perversely result in no liability. When else might “but-for” causation be loosened? When there is more wrongdoing than there is harm to go around, the law runs into the issue of how to allocate liability among defendants. One simple route is “joint and several liability,” where each wrongdoer can be found liable for up to 100% of the harm, and the plaintiff may elect from whom to collect what, so long as no overcollection takes place. Should this be the rule, for example, when intentional and negligent acts combine, such as an assailant hurting someone thanks to a hotel’s negligently maintained lock?