Interacting Poisons and Criminal Causation

A recent case note: “United States v. Smith, District Court Denies Oxycodone Distributor's Post-trial Motions in Penalty-Enhancement Case.” 129 Harv. L. Rev. 2297 (June 10, 2016) makes the point that understanding such statutory language as “result” and “causes” in terms of “but for” causation runs into a problem with concurrent sufficient cause cases, what philosophers call “overdetermination.” I want to take a look at some hypothetical poisoning cases, with independently sufficient causes, to pursue some issues of interaction, timing, and the general way to approach hard questions of criminal causation.

With intent to kill, and knowledge of efficaciousness, Badone surreptitiously mixes substance A into Vic’s green tea. Without knowledge of this, Badtoo, with similar intent and knowledge, puts substance B into the cream that Vic used with his tea.

Half Hour. Suppose now that Vic died half an hour after drinking his tea with cream. A doctor on board declared Vic dead, and said that what he saw was consistent with poisoning. As it happens that is all we know on that score however. No autopsy could be performed because Vic’s body was washed off the deck and lost at sea. 

Both Badone and Badtoo, are clearly guilty of attempted murder. More difficult questions arise for charges of murder. 

Let us imagine a joint trial of the two on the murder charge. In the real world there would almost certainly be a trial of one of the two, and then, perhaps the other, but for our purposes, it is useful to imagine that the liability question for each is before the same judge and jury.

There is uncontradicted expert testimony that the dose of A in Vic’s tea would, acting alone, have killed him in about an hour. B alone would have had the same effect in 3 hours.

Defense counsel will argue, properly, that Vic, though young and having recently passed a physical with flying colors, might have fallen to a stroke, heart attack, arrhythmia, pulmonary embolism, or aneurysm. In their motion at the end of the prosecution case, defense emphasizes that, absent an autopsy, there is no evidence directly inconsistent with death unrelated to the poisons. The judge, however, given the timing and expert testimony, will let that question go to the jury.

If the jury concludes that Vic was poisoned, it would, on the present facts, have to convict both Badone and Badtoo. Because Vic died before either poison acting alone would have killed him, it must be that it was the combination of A and B that did him in. It is strongly analogous to the classic bleeding from two wounds fact pattern, and both malefactors would be liable, even though they were not acting together. 

Two Hours. What now, if Vic had died two hours after drinking the tea. Here it seems at first glance as if Badone but not Badtoo could properly be found liable. The most natural inference is that B must have acted as a partial antidote to A, somehow slowing up A’s action, perhaps by neutralizing some fraction of it. At least, in the absence of autopsy, it would initially seem hard to convict Badtoo on these facts alone. 

That is the way I think things would stand in most such cases. Consider, however, an extraordinary case. A chemist testifies that A and B mixed in those proportions would form a new compound, C, a poison intermediate in speed between that of A and B. It could well lead to death in two hours. 

 

The prosecution contends that this brings Badtoo’s liability right back, he being equally responsible for the specific poison that in fact killed Vic. It is no different, she argues, than if Badone and Badtoo had each procured Vic’s ingestion of one half the total dose of C. Badtoo’s substance B was unquestionably a but for cause of Vic’s actual specific death – when he died and of what he died. Causation is a matter of science, here physics, chemistry, and physiology, and they show that the ingestion of B was a direct cause of Vic’s death.

Badtoo's counsel concedes that Badtoo was part of the causal chain leading to Vic’s death from the point of view of science, but then so was the person who brewed the green tea, and all those who designed, manufactured, sold, transported, or washed, however many times, the tea cup. With only a little extravagance we might say that everything within the light cone of Vic’s death was its cause. Legal cause, in order to do the job we ask of it, must be a good deal narrower than the branching lines of causation known to science. Perhaps this very inclusive chain of causation will do for a first “cause in fact” phase of the causation inquiry, but when we get to “legal cause” we use norms to draw lines across the chains of scientific causality, excluding what is too remote, too incidental, too little foreseeable, or otherwise normatively inappropriate for criminal liability.

Badtoo, his counsel continues, cannot possibly be considered the cause of Vic’s death, as Vic would have died a half hour earlier had B had not been in the cream. Slowing someone’s death is nothing like administering a second wound that speeds the draining of the victim’s lifeblood. Had Vic been brought to an emergency room, a doctor with only B relevantly at hand, might have given Vic a shot of B to prolong his life until the stat call brought safer antidotes for A and C from across town. Something that could be used to save Vic’s life, cannot be the legal cause of his death. Yes Badtoo was the cause of Vic’s actual death, at the moment he died, but so was the chest pound that kept Vic alive an extra minute. For Badtoo’s action, Badone’s poisoning of Vic meets the legal definition of a “superseding cause.” It was no different than if Vic had been shot before lifting the tea cup with its poisoned cream to his lips. Indeed, Badtoo’s relation to the superseding cause is even better than that, because he interfered with poison A, slowing its fatal action. 

Given a chance to rebut, the prosecution argues that Badtoo, with his intent to kill, is a world away from an ER doc who might give a patient the same dose of B in an attempt to save a patient from A poisoning.

The defense responds that Badtoo will be properly punished by his bad intent by the attempt charge, to which his client has always been willing to plead in exchange for dismissal of the top count. Intent cannot be made to do double service as the key element of attempt and also a surrogate for causation.

After due thought, the judge announces that she will not dismiss the murder charge against Badtoo and that she will charge the jury that a poisoning may be a cause of death if it contributes to the death even though it acts in an unforeseeable way together with another poison, so long as death by poison was reasonably foreseeable. She explains that she thinks that matters of intent can be relevant to legal causation. It makes a difference that Badtoo was not an ER doc trying to prolong Vic’s life. Had B acted only as an antidote to A, Badtoo would not be legally responsible for Vic’s death. Because B was, however, an ingredient of substance C that the jury could conclude killed Vic, Badtoo ought not to be able to take advantage of the incidental circumstance its operation led to Vic’s living a little longer than he would if only the victim of poison A. Badone’s poison was not a superseding cause that would cut off Badtoo’s liability because the two poisons were not independent but combined in their effect.

She might, if she were an exceptionally scholarly judge, go on as follows, but you may skip this unless you are as nerdy as she:

My conclusion is consistent with Lawrence Crocker’s definition of “cause in fact” in “A Retributive Theory of Criminal Causation,” 5 J. Contemp. Legal Issues 65, 69-77 (1994). The victim’s death, when he died, followed by the laws of nature from no set of initial conditions that did not include both Badone’s and Badtoo’s pouring of noxious substances into tea or cream. I recognize that Crocker’s definition also makes it a cause in fact that there was no hole in the tea cup and that it would make Badtoo’s pouring a cause in fact even had B been a wholly benign antidote to A. 

Whether we patch the definition of cause in fact or take care of it at the next phase, that of legal causation, does not seem to me to matter much. If we are to follow tradition in distinguishing these two phases at all, I suppose it is at the second that we say that giving B to extend life is separated out from giving B to kill. Really, however, I think Crocker went altogether too far in endorsing the traditional distinction between cause in fact and legal cause. The “common sense” component that separates cause in fact from purely scientific causation is continuous with the normative elements of the legal cause analysis. There is no neat division between the two, and legal cause type considerations have to be brought in for some problems that would traditionally seemed to raise pure cause in fact issues. I say this without intending to deny that it is good practice to start on a causation problem from common sense notions of causation as deepened by our best science, and then bringing legal norms to bear.

Three or More Hours. Here, again, it is clear enough that A and B are in some way interfering with each other, but still very likely that one or the other or both together would be on the bad side of the story explaining Vic’s death.

Let us put aside, for a moment, the wave that washed Vic’s corpse away and suppose that we can bring the full resources of science to bear on the question of criminal liability. Blood toxicity tests and careful examination of organs might show that B acts only as an antidote to A, unfortunately not in quite large enough a dose to have neutralized all of A. (No B still in blood stream; some A; only A-targeted organs compromised.) Badone should be convicted of murder; Badtoo found guilty only of attempt.

Alternatively, B might be an antidote to A with enough B left over after neutralizing A to cause death itself. (No A still in blood stream; some B; only B-targeted organs compromised.) Badtoo should be convicted of murder: Badone of attempt.

In a middle case, A’s independent effect is nullified but, either be forming a new compound C or by other mechanisms, A and B acting together lead to the demise of Vic. So we are back to the considerations with which the judge had to wrestle in the two hour case. She would reach the same conclusions, and we might have a few qualms –  B here extended Vic’s life yet longer beyond the baseline of A poisoning alone.

Science, or more accurately common sense deepened by science and supplemented in the hard cases by norms, can settle these cases. What if science has one hand tied behind it – our yachting no autopsy case? There is always the possibility that the chemical and physiological interactions of A and B are so well understood by the experts that evidence can be given, even without autopsy, to convict Badone, Badtoo, or both. More likely is that without the corpse we can know with high certainty that one or both caused Vic’s demise without being able to  convict either. The heart of retributive justice is that we not convict the innocent, and the reasonable doubt standard makes it inevitable that there will be cases in which the factually guilty cannot be found legally guilty. The identical twin hypothetical is the poster child here. The outcome might be different in a wrongful death action. There is good sense in placing the financial burden of Vic’s death on Badone and Badtoo not Vic’s family.

In this particular case we can console ourselves with the severity of attempted murder sentencing. 

Moral of the Story. Science is the essential first step for criminal causation. Were criminal jurisprudence only a matter of the wickedness of the accused, we would not have to be so concerned with the actual mechanism by which the prohibited effect came about. The law has, however, long been guided by the principle of retributive justice that those who commit bad acts are not guilty of as serious an offense if those acts do not in fact produce the bad effect aimed at. “In fact produce” is matter of common sense as supplemented by science – the better the science the better.

Science alone settles the question when an act can be shown in the particular circumstances to function only to counteract another act sufficient to cause the prohibited harm (or otherwise sufficient to do so.) When, however, there is any way in which both acts contribute to the harm, even if one or both also partially counteract the other, then science alone will not settle the question of liability. Normative judgments have to be made wherever there is positive interaction. 

Bleeding from both wounds is usually enough for liability for both independent assailants. It is not, enough, however, if one scratched and the other severed the carotid. It may not even be enough if one’s punch to the nose caused a serious nose bleed but the other’s scimitar slashes got carotid, femoral, and brachial. A poison that both delays the effect of another poison and contributes to the ultimate death is usually the legal cause of death, as our judge concluded. Perhaps our responsibility norms however will put it on the no-liability side of the line if the death is very long postponed or if the science can tell us, e.g., that 99 out of 100 parts of the B substance acted as an antidote to A, and only 1 part combined with A towards the death that would have come only slightly less quickly by the remaining un-neutralized A alone. Science is always essential; our responsibility norms sometimes are. 

On occasion determination of criminal cause requires more science than can be brought to bear under the circumstances. In such cases the guilty will properly be found not guilty.

Because the interacting poison cases bear such a close relation to the McLaughlin-Hart-Honore cases of the independent malicious tamperers with a desert hiker’s water bottle, I  plan to reconsider my previous take on those cases next.