Humanities desk
< February 3	<< Jan \| February \| Mar >>	February 5 >

Welcome to the Wikipedia Humanities Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

February 4

Thought experiments, ethics, and cause

I heard a sort of thought experiment recently that I had not heard before. It's perhaps vaguely related to a trolley problem, but it's not so much about what to do as much as what has been done, and how to assign responsibility and causes. It goes like so:

Alice and Bob are on top of a very tall building. Alice pushes Bob off, a surely fatal move. Meanwhile, Carol, who is on top of another building, has the falling Bob in the sights of her sniper rifle, and cleanly shoots him through the head before he can hit the ground.

The question is, who caused, or is responsible for, Bob's death? Alice, Carol, both, or other? Alice would hold that clearly Carol killed Bob, as she fired a bullet through his head. Carol would retort that Bob was already as good as dead, and her bullet did not have any effect with regard to the outcome of the situation. Bob is of course unavailable for comment.

Now, I can puzzle over this all I like, and maybe it seems simple or silly to some, but my question for the ref desks is: Is this just a re-phrasing of some other well-known thought experiment? Or can you recommend any other similar puzzles that relate to the issue of how to assign cause? Thanks, SemanticMantis (talk) 17:40, 4 February 2016 (UTC)[reply]

Causation (law)#Independent sufficient causes may cover this. -- BenRG (talk) 18:36, 4 February 2016 (UTC)[reply]

Thanks! That seems to cover the legal situation for negligence in the USA, still interested in other legal or philosophical/ethical precedents. Actually, upon re-reading, I'm having a hard time reconciling our coverage with this discussion [1] of the case. Our blurb makes it sound like A and C would both be held liable in my example. But the casebriefs link seems to imply that if B fell off the building, then C would not be liable. If I'm interpreting that correctly, then that's interesting; sort of a violation of independence of irrelevant alternatives, because the liability of C seems to rest on the liability of A, even though C's actions were sufficient to cause damage! SemanticMantis (talk) 18:56, 4 February 2016 (UTC)[reply]

There's a section further down with some discussion of English criminal law. And the rest of the article covers both US and English/non-US common law positions to some degree, but a basic torts and criminal law text book (respectively) should have more details and references to cases. In terms of criminal liability in the example you cited, you would probably get all of the people responsible, either under an actual offence or under an inchoate offence. --PalaceGuard008 (Talk) 19:29, 4 February 2016 (UTC)[reply]

For the position in English law, see Murder_in_English_law#Causation_and_foreseeability, which mentions the concept of novus actus interveniens or "breaking the chain", although that article only refers to civil cases. This article describes a fight between two soldiers, one of whom was stabbed and then dropped twice by those carrying him on the stretcher and finally mis-diagnosed by the medical officer resulting in his death. The soldier that did the stabbing was convicted of murder but this was overturned on appeal (Regina v Smith, 1959). Not the same as your scenario, but I suspect that such a case has never appeared before the courts and would result in much wig-scratching (this is not legal advice). Alansplodge (talk) 19:52, 4 February 2016 (UTC)[reply]

I think you misread that source. In R v Smith (Thomas Joseph) the conviction was upheld because the stab wound was the "operating and substantial cause" of death. In R v Jordan the conviction was overturned because the stab wound was largely healed but the victim was given antibiotics which he was allergic which would probably have been fine were it not for the fact another doctor prescribed the same antibiotics the next day after they had been stopped due to the allergy. This is not legal advice but while the fall may not be the "operating and substantial cause" of death it would have been the cause of death had the victim not already been dead when hitting the ground so it seems closer to Smith in so much as you can learn anything from them. (It's not like Jordan where court considered the defendent had basically survived the stabbing.) If the victim was fine after somehow hitting an anti-suicide airbag except the sniper then shot the airbag and caused an explosion because it had accidentally been filled with oxygen, then you have someting more akin to Jordan. Nil Einne (talk) 06:26, 5 February 2016 (UTC)[reply]

Quite right, I must have done (good job I'm not a barrister). This article about the same case summarises: 'To operate as a novus actus interveniens, the intervening event must be so potent and independent of the defendant’s actions as to render those actions “insignificant”.' Alansplodge (talk) 14:37, 5 February 2016 (UTC)[reply]

SemanticMantis' ~~exact~~ scenario is described in the introduction to the film Magnolia, incidentally, the name of the suicide (or murder) victim being Sydney Barringer. Tevildo (talk) 20:52, 4 February 2016 (UTC)[reply]

Except that there was no suicide in SM's scenario. -- Jack of Oz ^{[pleasantries]} 20:56, 4 February 2016 (UTC)[reply]

True, "exact" is not correct - the shooting in the film was also accidental rather than deliberate. The more traditional formulation of the paradox involves the victim intending to cross the desert - one murderer poisons his canteen, the second murderer makes a hole in it. Which of them is responsible for his death? Tevildo (talk) 21:06, 4 February 2016 (UTC)[reply]

There are still alternative causes of death in all these scenarios too. --Scicurious (talk) 21:32, 4 February 2016 (UTC)[reply]

It is similar to the Ronald Opus urban legend, which is the inspiration of Magnolia, the film. Snopes has an analysis of it.

The situation you describe seems to be a peculiar form of involuntary euthanasia, of a person who had only a couple of seconds left. Scicurious (talk) 21:44, 4 February 2016 (UTC)[reply]

SM, according to this posts message: A and C. By Law C if C gets caught, otherwise A if you happened to catch A; still A will get out of it cause it was a sniper. - This is just a thought, not a legal advice... -- Apostle (talk) 21:38, 4 February 2016 (UTC)[reply]

Thanks all, very interesting stuff. @Tevildo: has come the closest to what I was hoping for, the desert scenario is definitely isomorphic for me. I may have even heard it before. While I have some interest in the legal ramifications, I'm much more interested in the philosophical/ethical/ontological considerations. Ideally, I'd like to read a scholarly essay on this "desert scenario" or one like I described, or a close variant with suicide, etc. The key I think is that either action is sufficient cause on its own, but it's challenging to assess cause unequivocally when both actions are done. So if anyone can point me to some primary philosophical documents on the matter or something analogous to our article on the trolley problem, or even help find more canonical names for the puzzle, I'd be grateful. SemanticMantis (talk) 23:45, 4 February 2016 (UTC)[reply]

An interesting real-life example is the court case resulting from the death of Sammy Yatim in Toronto, where the man who undisputedly killed him was acquitted of murder, but convicted of attempted murder. --76.69.45.64 (talk) 06:38, 5 February 2016 (UTC)[reply]

There are many cases of people having fallen from an airplane without a functioning parachute and surviving (Juliane Koepcke for example) - so it's certainly plausible (albeit unlikely) that B might have lived had C not gone and shot him. Similarly, there are plenty of cases of people surviving a gunshot to the head (Gabrielle Giffords, for example) - and A couldn't be certain that C wouldn't miss.

So A cannot argue that she knew that C would kill B before he hit the ground - because B might have survived the gunshot. C cannot argue that A had already killed B - because he might have survived the fall had she not shot him. So both committed "attempted murder" (at the very least) - and ethically, both are 100% at clearly fault because neither of them could know for 100% sure that their actions wouldn't be the cause of B's death.

The only question that seems tricky is whether either of them gets charged with full-on murder. That brings up the medical question of whether the gunshot caused B to die instantly - or whether he was still alive when he hit the ground. That would clearly determine who committed murder and who committed attempted murder. If (for example) there is security camera footage of B's head breaking up into a gazillion pieces and a more or less headless corpse hitting the ground - then it's decided. But if forensic evidence can't tell us which was the cause of death - then it's a different matter.

There are plenty of cases of someone being stabbed or shot repeatedly by two or more people - with it being impossible to determine which stab or bullet wound finally caused the death and which person inflicted that specific wound. In such cases, all of the people who stabbed or shot are guilty of murder...and this hypothetical case doesn't seem any different. On that basis, you'd have to say that both A and C murdered B.

But the outcome of this kind of event isn't going to be clearly described in law. In the end, judges and juries will wind up deciding. Maybe C claims she committed a "mercy killing" - maybe A claims that she saw C's laser gunsight highlighting B's forehead and felt that pushing B off the roof was the only way to give him at least a tiny chance at survival? Neither seems a particularly plausible explanation - so it's down to judges and juries to determine motives.

SteveBaker (talk) 15:07, 5 February 2016 (UTC)[reply]

You are right about the chance of surviving such events (even if it's slim). However, including this information in the analysis would imply jumping the hypothetical limits set by SemanticsMantis. Such scenarios are simply not part of the question. — Preceding unsigned comment added by Scicurious (talk • contribs) 15:33, 6 February 2016 (UTC)[reply]