Death in the Desert

The desert traveler’s water bottle maliciously tampered with by one malefactor and then stolen by another has bedeviled the jurisprudence of criminal causation since 1929.

Well, the bedeviltry worked by this problem has been only occasional, not because there has ever been consensus as to its correct resolution, but because the fact pattern doesn't arise so very often in the real world.The inadequacies in our standard formulation of causation brought into the spotlight by the hypothetical have been of much interest only to those of a pathologically theoretical bent. Without pretending to exempt myself from the latter category, I would claim that seeing the water bottle problem in the right way can contribute to an understanding of criminal causation – an understanding that, could it become widespread, might save judges and commentators from some of the false, and sometimes silly, things that they say when addressing the causal leg of criminal liability.

I turn for a second time to the problem also because I have thought for some time that I got it wrong in A Retributive Theory of Criminal Causation, 1994 J. Contemp. Legal Issues 65.

In McLaughlin’s original version, “Proximate Cause” Harvard L. Rev. 39 149-199 (1929), the first malefactor puts salt in the water keg. Later, an entirely independent evil-doer steals said keg from said hiker. The hiker dies of dehydration in the desert. McLaughlin ends his 3 line footnote discussion of the case with: “Who killed him?” Hart & Honore, in a belief that they were preserving the essentials of the case while make it crisper, have the first bad guy poisoning the water in the keg. The second, again, steals it. Causation in the Law, (1985) 239-241.

Water Poisoner. Let us do the easier issues first. In the Hart & Honore hypothetical, it seems quite impossible to hold the first malefactor liable for anything more than attempt. He tried to poison the hiker. The hiker was not poisoned. When the poison went down the drain, his attempt was foiled as completely as an attempt ever is. 

Salt Thief. Only slightly less obvious is that the thief in McLaughlin’s version didn’t kill the hiker. Yes he stole a water keg, but it contained only salt. The theft is as irrelevant to the hiker’s death as if the thief had purloined a spare sock from the hiker’s pack. Salt is of no more help in warding off dehydration than is hosiery. 

Thief of Slow Acting Poison. If dissolved in the the keg's water was a poison so slow acting that the hiker would have lived longer had he had it, rather than nothing, to drink, then there is little room for doubt that the thief caused his death. It was the lack of water that cut short his life. That the water of which the theft deprived him, being poisoned, would not have gotten him very far makes no difference. If you put a bullet through the brain of someone trapped in a burning building, it is you who have killed him, not the arsonist.

Thief of Fast Acting Poison. This seems to be the case that Hart & Honore had in mind. The thief was responsible for the hiker's death as he died and when he died in that the hiker would have died earlier if the thief had not intervened. Having acted with malign intent, however, the thief is not automatically exonerated by the clock. (See my 8/29 post for a little more on this and Retributive Theory 75-76.) Yet, as you will shortly see, I now think I was wrong in the article in concluding that the theft was a legal cause in this fast poison case.

Yes, the hiker's death was exactly of the sort the thief had in mind, and he was unquestionably responsible for the fact that there was no keg in the hiker's rucksack. This all tempts us into thinking that the thief, in effect,  took the liability onto himself when he substituted his own plan for the (unknown) plan of the poisoner. 

We should, however, resist that temptation. The thief is not responsible for the lack of potable water in the rucksack. That circumstance was already established before the thief got his hands on the keg. He stole nothing that could help the hiker survive. His theft was a mistake, his plan having become, on closer inspection, impossible. If one is trying to kill by dehydration stealing a keg of fast poison is as much a failure as stealing a keg of stove fuel. What killed the hiker is that he left for his hike without potable water, and the thief contributed to that only in certain possible worlds. The actual world is the one for which we punish, and in that world the thief's actions were causally irrelevant.

Defense counsel might fairly go so far as to argue that, were we to look to non-actual worlds, in many of them the thief's action would have saved the hiker altogether. A slightly more conscientious hiker would double check for water before setting out into the desert. A poisoned or salted keg would likely pass muster. A missing one would not. This is no consolation to this particular hiker, of course, but it is of at least rhetorical force in illuminating the tenuous connection of the theft to his death.

 

For the same reason that one cannot kill a dead person, one cannot kill by theft of water one who had no water.

Keg Salter. I contended in “Retributive Theory” that salt substitution was a cause of the hiker’s death – arguing that the hiker's doom was sealed at that moment. The hiker died of dehydration in this world, and generally in any nearby possible world. In the very nearest by of worlds, those most like reality, the keg is successfully stolen. Slightly more distant are the worlds in which the thief fails to get his hands on the keg. In some of these the hiker may have chanced to handle the un-stolen keg, and, salt being roughly the density of water, failed to notice that something was different. In all these, the hiker dies of thirst. 

I now think, however, that I was wrong in the article about the salter too. Whether or not it is right that in all the appropriately close possible worlds the hiker died dehydrated, the salt substitution did not kill the hiker. The keg was not there when the hiker died. That salt was or was not in it at the crucial moment had nothing to do with the death.

Paradox: Killed by Human Action, but Killed by No One. So of the six evil doers in the various versions of the desert dehydration, only one is guilty of murder because only one, the thief of the water spiked with a slow acting poison, causes the hiker's death. In McLauglin’s case and in the fast poison case, no one is guilty of more than attempt. How is this possible when the hiker was unquestionably done in?

We would have no trouble in finding both guilty if the malefactors had been acting in concert. “I will try to put salt (or fast poison) in his water keg, but as backup when you later get your chance, you try to swipe it. One of us should succeed, and it’s no problem if we both do.”

Is it not a plausible postulate of criminal law theory that if the actions of two persons operating together are the legal cause of a death, then the actions of at least one of the two must cause the death even if they are operating independently? If this seems plausible to you, it shouldn’t. Acting in concert in effect creates a single actor, a fictional, but legally salient, individual. The actions of the actual co-conspiring persons are brought together as those of the fictional person and treated as a single cause. The analogy of this fictional person to co-conspirators is close enough to assimilate criminal liability of the fiction to the actual malefactors; for independent perpetrators it just isn't.

If one thief turns out the lights, and he and a second unrelated miscreant takes advantage to grab purses and a watch or two from the cocktail partyers, the prosecution cannot add the value of what each had when apprehended to make out grand larceny. If the two had been co-conspirators, that addition would be perfectly proper. 

Co-conspirators who set fires in two separate parts of the hotel, which ultimately combine to burn it down, are liable for the entire damage even though neither embryonic blaze by itself would have burned much of the hotel. (Either alone was within the capacity of the sprinkler system.) Independent fire setters who did exactly the same thing would not be liable for the whole, at least criminally, because neither would be responsible for the other’s essential contribution to the result.

You may want to say that justice is better served in that case if both are held criminally responsible for the whole damage. Consider, however, that the part of the structure near where A started his fire – the room that was fully destroyed before the two blazes even joined. B cannot be made liable for that damage without some very strong principle of the strict liability ("at risk") family that overrides all considerations of causation. 

The bleeding from two wounds principle in the strong form applying to independent inflictors of wounds, neither of which would be fatal alone, is such a principle. Its generalization could get us to the fire case, and its aggressive generalization to these desert hiker cases. These are generalizations that I think we should abjure. 

There is not really any paradox in the proposition that something caused by human beings, all of whom have criminal intent, is not caused, for criminal law purposes, by any one of them. This happens when the causal chain is, for example, too indirect or unforeseeable, e.g. the voodoo pin sticker who, being surreptitiously observed by the intended victim, causes such victim, not previously known to be superstitious, to die of a hear attack. In the hiker cases the partially conflicting vectors of the malefactors' actions add up to cause death in a way not foreseeable by either. Our common sense tell us that together they caused the death, but, as the evil doers were not acting in concert, a close examination shows that neither can be held criminally liable for that result.

Still, an aroma of paradox, and the so prominent role of moral luck in my solution to the hiker problems will persuade some that the right answer really is liability for all. Hart and Honore thought that our causation concepts do not apply well in these conditions, and certainly do not apply to make poisoning or theft causes in the usual senses. They suggest, however, the possibility of holding such malefactors guilty of murder as a matter of policy. 

When you poison, salt, or steal water bottles, and that is part of the whole story that leads to the death of your intended victim, you should get, some argue, no legal cover from the fact that other interventions undermined yours. To go to a theory of criminal causation strong enough to reach this conclusion, however, would seriously deflate the role of resulting harm in the criminal law. The badness of the malefactor is normatively important but so is the harm, where there is harm, and whether the malefactor can, with justice, be said to have caused it. 

Yes, there is a recurring, and popular, argument that nothing beyond the wickedness of the actor or that of her immediate intended act should be salient for a rational system of criminal liability and punishment. This argument would relegate such language as “causes the death of such person or of a third person” to the bad old days in which a superstitious society barely sublimated its primitive passion for revenge. 

My own view is that the resulting harm component of the law has always been there and will continue to be there for a very good reason. A death really is worse than a close call. For Macbeth and Hamlet we honor Shakespeare, not because Marlowe or Bacon lacked the genius, insight, or energy to have written these plays, but because of the happenstance that the quill that penned them was in the hand of Shakespeare.Who actually caused a result makes a difference to such creatures as we.