Home  |  Introduction |  Legal Puzzlers   | Law Reviews  |  Online Forum   | Paper Topics  |  E-mail  me


Spring 2000


ico4.gif (285 bytes)
5. Henry's Fatal Desert Crossing
Henry plans to go hiking in the desert. John, intending to kill Henry, puts poison in his canteen. Alice, who has no idea what John is up to, also intends to kill Henry. Alice punctures Henry's canteen. Henry sets out on his trip. All the water in his canteen leaks out and he dies of thirst. Who caused Henry's death? Was it Alice? But she only deprived Henry of some posioned water that would have killed him more quickly than thirst? Is Alice responsible for Henry's death?



Commentary. . If doubts have not already been raised in your mind about the value of the sine qua non test, this scenario should help to give it a permanent black mark. If we apply the test without qualification to Henry's desert crossing, we would have to say that neither John nor Alice caused Henry's death. We would be inclined to say that both did it or one or the other of them (probably Alice) did it, but not that neither of them did it. But if we apply the test and ask ourselves whether Alice's act was a necessary condition of Henry's death, the answer would have to be "no." And if we ask whether John's act was necessary to Henry's death, the answer would also have to be "no." The answer is "no" in both instances because in either instance what Alice or John did was not necessary since Henry would have died anyway at the hand of the other. If not for Alice, Henry would have died because of what John did. And if not for John, Henry would have died because of what Alice did. So in this case at least the sine qua non test produces a counter-intuitive result. But consider the following:

John's and Alice's Reckless Cycling is a Cause for Alarm: John and Alice have just bought brand new Harley-Davidson motorcycles and they are out for a spin. They spot Henry on his horse, sauntering along a dirt road to his neighbor's farm. John and Alice zoom by, each to either side of Henry and his horse. The sound of the motorcycles and the dirt they kick up so frighten Henry and his horse, that the horse bolts, tipping Henry out of the saddle. Henry's foot catches in a stirrup and he is dragged along, bumpity-bump, into a potato field where he eventually becomes disentangled. He suffers a dislocated shoulder, a badly bruised hip and a broken ankle. Who caused Henry's injuries?

Assume that Henry's accident would have occurred (let's say) even if either Alice alone or John alone passed by on one side or the other of Henry's horse. Now apply the sine qua non test. According to the test, neither John nor Alice caused the accident. Alice's action did not cause Henry's injury because Henry's accident would have occurred anyway; it would have happened even if Alice had not driven by on her motorcycle because John's action was sufficent to cause Henry's horse to bolt. And John's action (according to the test) did not cause Henry's bruised hip and dislocated shoulder because Henry would have been injured even if John had not zoomed by since Alice's action would have been enough to cause the damage. Neither John's action nor Alice's action is necessary for the occurrence of Henry's accident, yet both are causes. Clearly, something's amiss with the sine qua non theory. This scenario and the puzzler it is intended to illuminate suggest that the sine qua non test fails in situations where an event is "over-determined," where two independent causes are sufficient to bring about the same result.

This observation may be the beginning of the end for the sine qua non test. The more we rely on it, the less confident we are likely to be with its results. Indeed, the sine qua non test seems to lead to the absurd conclusion that any causal factor, any pre-condition, necessary for the occurrence of some consequence, is its cause. So, for example, if John and Alice had not been born, they woud not have been alive to buy motorcycles or to frighten Henry and his horse, and Henry would not have dislocated his shoulder. But for John's and Alice's having been born, Henry would not have been injured. Indeed for any accident, the accident would not have occurred if one of the parties to the accident had not been born. If Henry had not been born, he would not have been injured. So, according to the sine qua non test, Henry caused his own accident. Or let Henry be born: if he had not gone out riding, he would not have bumped into John and Alice on their motorcycles, and he would not have been injured. His decision to take his horse out on that day at that time was a necessary condition for the accident. Had Henry left his horse in the barn, there would have been no accident. Henry should have been more prescient. He should have kept his horse in the barn and himself at home. This, of course, is absurd. The sine qua non test has to go. Whatever test we muster with regard to legal causation, the test shall need, it would seem, to be able to distinguish between cause and causal condition and between the cause and all the other causal factors, between the primary or immediate cause and other causes. The sine qua non test cannot make these distinctions.

What makes something the cause of a particular result? And what distinguishes it from all other possible causal candidates? How about a view that holds that an action A causes some event B in those cases where A is both necessary and sufficient for the occurrence of B? This rescues us from the absurd conclusion that all accidents are caused by the births of the parties involved. However much we may wish that John and Alice had never been born, their births were not sufficient for the occurrence of Henry's accident. But this alone will not do. If we want to claim, for instance, as we did above that John's zooming by Henry's horse on his motorcycle was sufficient to cause Henry's horse to bolt and that, therefore, what John did was one of the causes of the accident, we would need to make one very important qualification. We would have to say, something to the effect, that John's action caused Henry's accident, "all other things being exactly as they were." The accident would not have occurred if, for instance, Henry had been a better rider, or if someone else had been standing nearby to grab the reins of Henry's horse as it started to bolt. John's action is sufficient only in a given context. Only in that context is John's action sufficient to cause Henry's accident. Is this then a good theory? Should we be happy with it?

Why pick John's action out from the context and not some other factor within the context: Henry's failure, say, to hold onto the reins? Or recall Alice's guzzling down that bowl of soup and arsenic. If we presume that her drinking the soup caused her death, we would need (again) to make it clear that her drinking the soup was a necessary and sufficient condition, "all other things being exactly as they were." Had the poison been less potent or had she been unable to digest it, she would not have died. These causal factors are necessary for her death, and her drinking the concoction prepared by John would not have been sufficient to bring about her death without the existence of these and other causal factors. On what basis then, according to what criteria, do we "select" the cause in a given context and treat it as significant? How do we distinguish a cause from a causal factor or a cause from a mere condition? R. G. Collingwood suggests the following:

[The] 'selection' is by no means arbitrary. It is made according to a definite principle. If my car 'conks out' on a hill and I wonder what the cause is, I shall not consider my problem solved by a passer-by who tells me that the top of the hill is further away from the earth's center than its bottom, and that consequently more power is needed to take a car uphill than to take her along the level. All this is [of course] quite true; what the passer-by has described is one of the conditions which, together, form the true cause of my car's stopping; and he has 'arbitrarily selected' one of these and called it 'the cause.' But now suppose someone from AAA comes along, opens the bonnet, holds up a loose tension lead, and says "look here, sir, you're running on three cylinders." My problem is now solved. I know the cause of the stoppage. It is the cause. It has not been "arbitrarily selected." It has been correctly identified as the thing that I can put right, after which the car will go properly. If I had been a person who could flatten hills by stomping on them, the passer-by would have been right to call my attention to the hill as the cause of the stoppage; not because the hill is a hill, but because I can flatten it out. (R. G. Collingwood, Proceedings of the Aristotelian Society, 1938)

Collingwood's idea is that the cause of the event as opposed to a mere condition is that aspect of the situation over which we have some degree of control and which we can produce or prevent, "the thing," as he says,"that I can put right." But this way of trying to get a grip on the matter turns out to have almost as many problems as the sine qua non test with which we started. Part of the reason for the problems is similar to the reason there are problems with trying to give a definition of murder or theft or of crime itself. Any general account tends to over-simplify the process whereby we determine in actual, concrete cases what went wrong and who did what to whom. The determinations we make in actual, concrete cases vary in subtle and complex ways. The variations are also not easily captured by a simple and neat formula, but they are not arbitrary either. Why a case is decided one way as opposed to another or why a judge decides in favor of one litigant as opposed to another litigant is always defensible. There is no single, knock-down defense nor is there a perfect set of principles which will settle in each and every case the matter once and for all. Still some defenses are better than others and it is always possible to be clearer, even if it is not always possible to be absolutely clear. This can make the study of law frustrating, but it is also what makes such study challenging and fun.

The first five puzzlers make plain that even a matter as apparently simple as sorting out "whodunit" can can have its own challenges and pitfalls. Perhaps all we can say of value at this point is that abnormal events, such as a bolt of lightning emerging out of a freak thunderstorm and the free, deliberate acts of others can break a causal chain and prompt us to revise our original tracing of the consquences. With the qualification of such phrases as "on the whole," by and large," and "all other thing's being equal," an act ("A") is the cause of some injury or harm ("B") if A was necessary to the occurrence of B and sufficient to produce B without the intervention of the free and deliberate acts of others or an abnormal conjunction of events. Or, put perhaps a bit too simply, if your conduct was both a sine qua non of some harm and sufficient to produce it, then you caused it you did it unless another person voluntarily and deliberately intervened to produce it or some freak combination of events materialized to give rise to it.

Although I do not like to give answers, especially in cases where the answers are debatable, let me say that I think that the negligence of the nurse in the first puzzler, her placing herself in close proximity to patients when she was contagious with scarlet fever, not John, caused Alice's death. John did not kill her, although he did attempt to kill her. Let me say, too, that I think that the lightning in the second puzzler caused the forest to burn down since it was an abnormal occurrence (the result of a freak thunderstorm) and that the free and deliberate act of Alice, her dumping gasoline on the smoldering ash, not John, caused the fire in puzzler number three. Let me say too that I think the man in the South African case did not kill his mother and that Alice killed and John attempted to kill Henry in Henry's fatal desert crossing.

These answers are debatable, but not without reason. You may have different answers, also not without reason. The better answers may depend in the end less on how well they resolve particular cases, but how well they serve to make sense of and resolve other cases like them. With this last reflection in mind, consider the next two puzzlers.

  Back to top


Search by Number



  Back to top

Home  |  Introduction |  Legal Puzzlers   | Law Reviews  |  Online Forum   | Paper Topics  |  E-mail  me

Page last edited: December 18, 1999