Bennett, Jonathan - Whatever The Consequences

10/15/2010

Ethics, Thomson & Dworkin Eds, Harper & Row 1968

This paper seeks to dispute the so-called principle of double effect, which posits a moral difference between actively doing x and passively allowing x to happen (when you could have stopped it). The example that author uses throughout the entire paper is one of a woman who is in labor but has reached complications and will die unless a late-term abortion is performed. However, it is also possible to deliver the baby alive, but the mother will suffer complications and eventually die. Author wants to attack those who believe that there is a moral difference between killing the baby and 'letting the mother die' by delivering the baby. The principle under attack is: "It would always be wrong to kill an innocent human, whatever the consequences of not doing so", though at the outset (pg212) author acknowledges that it isn't possible to refute that principle directly. Instead the strategy will be to show that the structure of the principle, something like 'always don't do x, whatever the consequences' is morally vacuous.

To start, author does not seek discussion with interlocutors who take religious authority or divine command as the basis for the principle; author wants to talk with those who believe there is an independent way to ascertain the validity of the principle. (pg213) Those who follow the principle are provisionally stipulated as 'conservative'. Author argues the conservative makes the following claims: (pg214)
-It is wrong to kill an innocent, whatever the consequences
-operating on the baby will kill it
-not operating (and delivering the baby) is not killing the mother, since her death is only a consequence, not an act of killing
-operating is wrong, not operating is permissible
What is going on here is an Action/Consequence distinction, which author tries to unpack in most of the rest of the paper. An action is considered motion in the physical environment, but will also typically include some of the 'upshots', that is what happened as a result of the physical motion. But not all the 'upshots' or consequences will be included in an action-- some could be considered to fall on the side of mere consequences. Author believes this is a valid distinction, but wonders if there is moral salience to it, especially in the case of the woman in labor. For author, there are 6 ways to approach the distinction between a human action and the consequences that come from it: (pg216-7 & pg222-3)
a) The actor was highly confident that the upshot (consequences) would follow from the action
b) There was a high degree of certainty or high likelihood that the upshot would follow from the action, even if the actor didn't know it
c) The aim or intention with which the action is performed does not include the upshot (consequence), even though such a consequence may inevitably follow
d) An impartial observer who assesses the entire situation would assign a particular upshot to be part of an action
e) There is a high degree of immediacy, 'simplicity of causal connections', short time-lag, etc. for actions but not for consequences
f) Actions can only be activities of a person, while not-doing something, whatever the consequences, is not an action.

Author argues there is moral salience to a-c but they do not apply to the 'obstetrics case', and d-f do apply to the obstetrics case but aren't morally significant. (pg216) To start, a and b both are stipulated in the case that the doctor knows either the woman or the child will die depending on the doctor's actions. So the obstetrics case does not have these elements. And with c, the doctor doesn't kill the baby for the sake of killing the baby-- the doctor does so to save the mother. And vice-versa with saving the baby but killing the mother-- in neither case is the action performed for the sake of killing the innocent human. (pg217)

Author then takes an elaborate detour into the general philosophy behind the conservative principle. The conservative might want to argue that there is a higher degree of immediacy or certainty that the child will die compared to the dying mother. However, author believes this is the wrong kind of argument the conservative would want to make to justify her position. If that argument underwrites the distinction between an action and its consequence, then it might be possible to concoct a case where performing some intentional voluntary motion would create a small chance of killing an innocent but have a large immediate upshot of saving more innocents. This kind of argument is not a 'whatever the consequences' argument. Author discusses how a conservative like Anscombe first denies this kind of argument as an underwriter, but then helps herself to it in her footnotes. (pg218-221) In general, this is a discussion about the permissibility of using 'fantastical' examples in moral theory.

The second group of ways to support an Action/Consequence distinction d-f might have differences in the obstetrics case, but author claims that no one should consider them morally relevant. Author claims that d would actually argue against the conservative position and moves to e, the case of a closer causal connection or a higher degree of immediacy in the killing than in the letting die. Author believes this element is loosely connected with a and b. But e has a difference in immediacy in the obstetrics case, (pg224) where there is no asymmetry in cases a and b (author claims). Author ties e and f together as basically using a distinction between actions and refraining-from-acting to do its work. Though there is a difference, author claims the difference isn't morally relevant. The difference between acting so that x happens and not-acting and allowing x to happen are laid out by author: (pg225-7)
Similarities:
-bodily movements
-the happening of x (e.g. someone dying)
Differences:
-there is a limited set of activities that could be performed in order to kill
-there is a wide set of activities that could be performed when letting-die

These differences are merely of degree, which may not be comforting to the conservative anyway, since it is possible to construct a circumstance where actions are so limited to destroy the difference in set size. More immediately, your moral judgment on whether someone is responsible for a death does not lessen once you learn that A could have killed B a whole variety of ways, not just the way A actually killed B. (pg228-229) And this is the fatal difference-- the set size of the ways in which an impermissible action might have occurred is metaphysically significant but not morally so.

The last portion of the paper discusses the reasons to perhaps stick with such a 'whatever the consequences' rule like the one. A conservative might object to the example, claiming it is too fantastic or fanciful and that it unfairly tests a good rule. Author counters that the example used in the paper is of a kind: there will be other times where this kind of case might come up. A conservative who pushes further might use two different arguments: (pg231-2)
1) You missed something when you analyzed the elements of the Action/Consequence distinction-- something that is morally salient but not nameable. Reply: this is a serious issue; do the tough work to elucidate the moral difference. If you can't, then drop the reply.
2) We must have some rules, or else we would never be able to decide anything on time. A host of particular situations are too complex to try to resolve one-by-one-- we need rules to do it efficiently. Reply: This is a focus on the particular, not the specific. You can make two rules: one that specifies what to do in certain kinds of situations, another that specifies what to do in other kinds of situation. The objection is about the particular, not the specific. Further, this argument is consequentialist: 'if we don't have rules to follow, look at the consequences that might occur!', which doesn't seem to be appealing to the conservative anyway. (pg234)