Jacques Maritain Center: Ethics Without God?In her article titled "The Reasons We Share" Christine Korsgaard appears to endorse the claim that in the well-known case of Jim and the Indians, Jim may be justified in shooting one Indian to save the rest.{1} This conclusion is a surprising one for a professed Kantian. I want to suggest that at least part of the reason she ends up taking this position is because of an objection she presents against a view she attributes to Kant. Ultimately, I think her objection fails as an objection to Kant, as does the view that she herself advocates in response to the objection. The reasons why, however, reveal a lot about the conceptual foundations of different ethical theories, and the role that belief in God must play in them. Ultimately, the conclusion I hope to suggest is that without belief in God, there is deep problem that exerts an unavoidable pressure to accept some form of consequentialism. This pressure, then, can explain why Korsgaard stops short of endorsing universal moral rules against lying, suicide, and the direct killing of the innocent.
The objection that Korsgaard presses against Kant stems from the consideration of particular cases. She considers, for example, Bernard Williams' familiar case of Jim and the Indians, where Jim is given the choice of killing one Indian or watching soldiers kill twenty. A second, also familiar, case involves a murderer approaching one's door and asking if the person he intends to be his next victim is at home. The choice in this case is meant to be between telling the murderer the truth, that the person he seeks is at home next door (which will lead to the victim's death) and lying to the murderer in order to save the intended victim's life.
First, I want to consider a general problem that such cases present for non-consequentialist ethical views. It is this general problem which is the basis for the specific objection Korsgaard presents. Then I will consider Korsgaard's objection to Kant in some detail.
What these cases have in common is that in each the course of action normally considered to be morally forbidden (killing an innocent, lying) is the only way to secure the outcome that a good person would most like to realize given the possibilities available in each set of circumstances (e.g., saving the most Indians possible, saving the potential murder victim).{2} In short, these cases illustrate that it is possible for the ends we would like to seek, the ends that we believe a moral person, a virtuous person, would seek, to pull apart from the actions that are normally thought to be moral or virtuous. This generates a problem in securing the intelligibility of the project of morality, of moral action as allowing us to achieve certain goods. If moral action cannot secure certain ends, ends that appear to be the best outcome of a given situation, then what is it that moral action aims at? What is moral action meant to secure for us if it cannot guarantee our fulfillment in this important respect? This thought, perhaps rightly, can lead to a rejection of the moral project as incoherent.
To illustrate consider one of several claims Robert Adams thinks it is important to believe if we are to act morally. He thinks it is important to believe that a moral life is "better for the world." {3} We must believe that acting as we should does not, on balance, make the world a worse place. As Adams asks, "how else can we care about morality as morality itself requires?"{4} If morality can, and sometimes does, lead to disaster, how can we understand a universal call to be moral? What is important, then, about the cases Korsgaard considers is that in each, by acting morally, more evil seems to result, more death and more harm. Each makes it seems as if the course of action we are morally required to take is not consistent with our belief that morality is good for the world. The problem is not one of providing a grounding for or justifying a moral requirement. The problem is maintaining the coherence of action in accordance with this moral requirement as seeking some genuine good.
Of course, one obvious solution to this problem is the consequentialist one. Consequentialism avoids the problem by defining it out of existence. By making consequences definitive of right action there can be no separation of the two, and hence no lack of fit to threaten the intelligibility of moral action seeking good ends. What is moral is always good for the world, by definition. Korsgaard, then, hopes to find a position that can justify actions that violate certain general moral rules, without falling into a simple consequentialism.
The objection to Kant that I want to consider can be found in Korsgaard's article titled: "The Right to Lie: Kant on dealing with evil."{5} It is more specific than the general problem I just mentioned, but ultimately stems from the same source. She objects to Kant because he holds strong to certain moral rules, such as never kill an innocent or never lie, even if horrible consequences result. Korsgaard appears to believe that his justification for doing this has to do with his claims about when agents can be held responsible for the outcomes that follow from actions.{6} She suggests this claim about responsibility follows from his acceptance of a particular kind of ethical view, what she calls a single-level ethical view.{7} She distinguishes single-level ethical theories from two-level theories asserting that a single-level view does not take into account both ideal and non-ideal cases, while a two-level theory does.
There are two kinds of single-level view. They either (1) focus only on what a perfectly just society would be like (what is normally thought to be morally required of us in terms of the discussion above, never kill an innocent, never lie), or (2) focus only on the realities of injustice in the world that we are in (the outcomes given the nature of our world).{8} Korsgaard claims utilitarian views are single-level in the second sense because they only focus on how to act in the particular circumstances we confront. Such views do not independently consider what we should do apart from outcomes; outcomes determine what is right.{9} Kant's view, then, is single level in the first sense because, in her view, it focuses only on the ideal (never lie, never kill) and does not address the often less-than-ideal particulars of real cases (and because of this, there can be the lack of fit between what we should do, and what results, mentioned above). She asserts that "The standard of conduct [Kant] sets for us is designed for an ideal state of affairs, we are to always act as if we were living in a Kingdom of Ends, regardless of the possible disastrous results."{10} It is Kant's perceived inability to adequately respond to the reality of these disastrous results that leads Korsgaard to argue in favor of a two level view.
By focusing only on an ideal state of affairs, she argues, Kant "defin[es] a determinate ideal of conduct to live up to rather than setting a goal of action to strive for."{11} As such, he gives the individual a "definite sphere of responsibility" such that "if you act as you ought, bad outcomes are not your responsibility."{12} Korsgaard considers this suggestion in relation to Williams' famous case of Jim and the Indians, and concludes that it is "grotesque" to argue that Jim is justified in not killing a single Indian when this results in the death of 20 by simply saying that "[Jim has] done [his] part . . . and the bad results are not [his] responsibility" (as she suggests Kant does).{13} It is this grotesqueness objection, then, that is at the heart of her rejection of a Kant-style single level view and hence helps to explain her acceptance of exceptions to general prohibitions against lying, suicide, and killing of the innocent.
In response to this objection, I first want to consider whether the two level theory that Korsgaard defends is up to meeting the general challenge of making moral action intelligible as seeking some good without falling into consequentialism. Then I will turn to consider whether she is right to think that this claim about responsibility plays the role she identifies in Kant's ethical view. I will argue that she is not right in thinking this, and that there is a very different set of considerations that are key to understanding Kant's response.
One danger that the proponent of a two level view must face is that of lapsing into the other kind of single level view, lapsing into consequentialism. Korsgaard needs to provide an answer to the question of when it is okay to deviate from the strictures of the moral ideal in a way that either does not make reference to consequences or limits the role of consequences to a specific range of cases. She recognizes this danger observing that the "common sense" answer to the question of when we can deviate is when the consequences of not doing so "would be 'very bad.'"{14} She rightly observes that this answer is both too vague and leads down a slippery slope to consequentialism (at any threshold, we can ask about cases just over the line and wonder why they are ruled out and others are not).{15} She claims that her own approach avoids both these problems.
Korsgaard suggests that she avoids consequentialism by providing a more clear, identifiable threshold at which consequences can become relevant.{16} Drawing on an example from Rawls, Korsgaard suggests that the point at which it becomes morally acceptable to trade gains in efficiency for reductions of liberty (for example) is the point at which inefficiency itself becomes a threat to liberty.{17} The suggestion seems to be that there is something special about this point, the point at which the consequences of acting in accord with a rule embodied in the goal we seek become so bad that they make the achievement of that goal impossible.
This seems like a promising suggestion, but it does not solve the problem for it is hard to see how it generalizes to individual moral choices. Consider the case of Jim and the Indians. Jim is supposedly, in his specific circumstances, morally permitted to kill one Indian in order to save twenty. When we deviate from the ideal, Korsgaard argues, the ideal is not simply cast aside, but comes to define "the goal towards which we are working."{18} This is, then, how she can avoid the problem associated with maintaining the intelligibility of moral action with respect to ends. But how could it be that killing one Indian to save twenty aims at this ideal? How could we see this action as aiming at securing a world in which this rule is universal? In fact, if we accept this threshold, we seem to be incapable of ever violating the ideal, for how can an action contrary to the moral rules contained in the ideal (goal) be a means for achieving a world in which such moral rules are always followed? The goal, then, cannot actually be securing the ideal if we are justified in deviating. But, if the goal must be something other than achieving the ideal, then the supposed threshold cannot be sustained. What makes this a unique threshold is that it marks where following the ideal causes problems with respect to the ideal. If we grant that causing problems with respect to some other goal is relevant, then we have started back down the slippery slope to a robust consequentialism. It becomes hard to see why any important goal is not relevant, why any sufficiently bad outcome cannot justify a different course of action.
Rawls' example, dealing with broader social/political concerns, seems to be intuitively different from cases of individual moral choices. There is at least some plausibility to the claim that temporary injustice or restriction of liberty is necessary to set up the social conditions needed to secure greater liberty or equality in the long term.{19} However, it is not at all clear how killing an innocent man can be aimed at setting the conditions needed to make killing always immoral. How does killing one Indian serve to prevent another Jim at another time from finding himself in the exact same circumstances? How can committing suicide be aimed at making the world such that no one encounters the kind of profound suffering Korsgaard thinks can make suicide morally acceptable (more on suicide in a moment)?
A second problem with Korsgaard's approach is that it does not provide a coherent reason for deviation from the ideal other than the recognition that the consequences of not deviating are "very bad." If there are no relevant factors other than consequences justifying deviation, there is no clear threshold to be drawn. To see this, consider how she constructs the two levels of her account. She relies on the claim that in some cases the Formula of Humanity of the Categorical Imperative is stricter than the Formula of Universal Law. This allows these two to define the two levels of the view; the former gives us the ideal, while the latter is responsive to the world.{20} Importantly, however, the arguments under the Formula of Humanity are perfectly general, they do not appear to admit the possibility of exceptions, so how can there be a reason for setting aside these arguments other than through the recognition that outcomes can trump any such arguments.
Consider, for example, the case of suicide. Korsgaard argues that Kant's argument against suicide under the Formula of Universal Law does not work,{21} but that "Under the Formula of Humanity we can give a clear and compelling argument against suicide."{22} That argument is, simply, that "Nothing is of any value unless the human person is so, and it is a great crime, as well as a kind of incoherence, to act in a way that denies and eradicates this source of value."{23} She then claims that "it might be possible to say that suicide is wrong from an ideal point of view, though justifiable in circumstances of very great natural or moral evil."{24} In The Sources of Normativity Korsgaard further discusses the circumstances under which suicide becomes morally licit. It is when "The ravages of severe illness, disability, and pain can shatter your identity by destroying your physical basis, obliterating memory or making self-command impossible."{25} The question, then, is how these considerations can provide a reason to deviate from the ideal that does not lead to outright consequentialism.
Notice that none of these considerations undermine or are even responsive to the argument given under the Formula of Humanity. It simply does not follow from the fact that nature can rob us of our identity that we are then justified in acting to destroy it. If it is indeed a "great crime" and incoherent to commit suicide in ordinary circumstances, to deliberately "eradicate the source of all value", then it still seems to be so in these special circumstances. To put the point another way, the fact that nature makes the achievement of a goal that we see as important impossible (in this case the preservation of our identity) simply does not make it coherent for us to act in a way that similarly makes the achievement of that goal impossible (committing suicide). The argument against suicide given under the Formula of Humanity does not admit to exceptions in such circumstances because such circumstances are, ultimately, irrelevant to the argument that is the basis for the rule.
Her suggestion, then, that sometimes circumstances can make suicide morally licit cannot be grounded in anything other than the claim that sometimes the badness of an outcome can override a moral rule. Since there is no threshold contained within the arguments for the rules themselves, the threshold must be found in the relationship of the outcomes to the rule. As we saw in the first argument above, however, Korsgaard's attempt to bootstrap a threshold from those considerations fails to make violating the rule coherent. All Korsgaard is left with, then, is the claim that the consequences are really bad and no principled dividing line. Korsgaard thinks she can defend a threshold by holding to the Formula of Universal Law, but if really bad consequences allow us to violate the moral ideal (the Formula of Humanity), why can't the same consequences, or even worse consequences, allow us to violate the requirements of the Formula of Universal Law? Korsgaard insists that the Formula of Universal Law provides the threshold at which "morality become uncompromising" but she cannot give a principled defense of this claim.{26} All she is left with, then, is a long, slippery slope to consequentialism.
Let us consider, then, Kant's own account, and see if it cannot do better. Korsgaard clearly thinks that the only solution which Kant offers is the simple assertion that the bad consequences that follow from moral actions cannot be seen as the responsibility of the person who has acted morally. I believe, however, that there is much more to his answer. Either the claim about responsibility is not meant to provide a candidate end, in which case it cannot solve the problem, or it does propose a candidate end that is clearly inadequate given Kant's own view. If we take the avoidance of responsibility to be the end that is sought, then the action is no longer consistent with the requirements of the moral law. If Jim's action aims at avoiding the dirtying of his own hands, then he is using the Indians as a means, a means to secure his own good conscience. Aiming at this end in these circumstances is, in fact, incompatible with the requirements of the moral law so it cannot serve as the end that is sought when we act from the moral law. Kant can rightly say that acting in this way is grotesque, not because of the consequences that result, but because it is immoral to do so. If this is Jim's end, then his action cannot be motivated out of respect for the moral law.
Korsgaard is right to think that the solution of the problem requires identifying some goal, some good, that moral action seeks. Kant's claims about responsibility, then, cannot solve the problem because they do not provide an appropriate positive goal for moral action. Not just any end will serve as a solution to this problem of ends. The concern is that in cases such as those under consideration here, there appears to be no connection between moral action and an appropriate end. In such circumstances, remembering Adams' phrase, morality does not seem to be good for the world. Limiting responsibility does not solve the problem because aiming at this cannot be seen as aiming at a genuine good for the world.
What, then, is Kant's answer? The only way to solve the problem is to provide an end, something that can rightly be viewed as a good that is sought through moral action even if other, bad consequences also result. In this way, acting morally can still intelligibly be understood as seeking some good, even in such difficult circumstances. The place where Kant talks about the end of moral action in most detail is in the third Critique under the rubric of the highest good. There he claims that there is an end that we seek when we act morally. He calls this end the highest good and suggests it is constituted by both the requirements of the moral law and happiness, where the former is the condition for the achievement of the latter.{27} Since he does not believe happiness and morality are connected by definition, since the concept of morality does not imply happiness, the only possible way these two could be connected is for them to be causally connected.{28} For Kant, then, when we act morally we seek the highest good; we seek a world in which moral action is connected to the achievement of our ends as a causal condition for that achievement.
Importantly, then, Kant suggest that the moral rule is constitutive of the good that is sought in such a way that this good cannot be sought by any means other than moral means. We fall victim to irrationality if we seek a world where moral action is a condition for the achievement our goals by acting immorally. In such a case we would be willing, at the same time, that we achieve our goal and that it be impossible for us to achieve our goal by the means we have selected.{29} It is, of course, important that Kant thinks this is the highest good. The end that is sought must be the greatest good, the complete, unconditioned end so that this end is always capable of securing the intelligibility of moral action in terms if ends. If some end could be higher than this, then the problem could be reconstituted in cases where an immoral act is the only way to secure this higher good.
The problem, however, is that simply asserting that moral action must have this relationship with the good that we seek does not completely solve the problem. As the cases Korsgaard considers suggest, there are some circumstances where moral action seems unable to achieve the goals we seek.
More generally, there is no guarantee that living a moral life will lead to a good life.{30} People who always act morally suffer calamities just like everyone else. Why should someone who has faced one calamity after another when acting morally continue to act morally when he can achieve other good ends by not doing so? Why should Jim act morally when he knows that he could save the lives of so many Indians if, in this case, he sets aside his moral commitments?
It is to this very real problem that Kant responds by introducing the practical postulates. We must, according to Kant, believe in God (or at least a god capable of securing the connection between moral action and the ends we seek) and the immortality of the soul (so that there is time for this connection to be realized).{31} In this way, we can see the possibility of moral action connecting up with the ends that we seek, and the intelligibility of the project of morality in terms of ends can be secured. Rather than trying to make clear the reasoning here in purely theoretical terms, let me instead consider a further example.
The case I have in mind is that of the Maltese conjoined twins treated in England in the year 2000 and referred to in the press by the names Mary and Jodie. There are several reasons for selecting this case, but the most important are the fact that it is a real world case, not a philosopher's thought-experiment, and that it is comes about naturally, that is it does not rely on a person acting immorally in order to generate the problem (unlike the other cases mentioned so far). The twins presented a moral dilemma because they were joined in such a way that even though Mary had all her own vital organs, her heart and lungs were not capable of circulating and oxygenating her blood. An artery connecting Mary's circulatory system to Jodie's heart was all that kept Mary alive. Unfortunately, Jodie's heart, while strong enough to support her own life, was not strong enough to support Mary's as well. Doctors believed that Jodie's heart would only continue to function for a matter of months before it gave out and both twins died.{32} The choice, then, was a difficult one: separate the twins and kill Mary, or do nothing and watch both Mary and Jodie die.{33} I do not want this discussion to get bogged down in questions of double effect, which are interesting in their own right, but have no bearing on the point I want to make here today. Let us stipulate, then, for the sake of this discussion that compelling arguments can be given that separating Mary and Jody would count as a case of direct killing.
How do we see our way through such a case, to a description of choosing not to separate the twins such that it can be understood as aiming at an appropriate end? Notice first that this case does fit the general problem we have been considering. Acting in the way normally thought to be moral does seem to lead to more suffering, more death. It does, then, seem that moral action, in this case, is not consistent with believing morality is better for the world. In response, Kant suggests that this apparent problem is just that, an appearance. If we understand what morality requires of us properly, we see that moral action not only aims at some good, but also is the only way to achieve this good.
We have also seen, however, that just any good will not do, there are certain requirements that the end must meet. As pointed out above, for example, the goal of avoiding responsibility will not serve because it is not a genuine good for the world and is not compatible with the moral law. Further, if we note that the prohibition against killing is founded on the requirement that we recognize the value of other beings like ourselves, we see that this end cannot serve since this end is not an end for the twins.{34} It is not properly part of the project of their lives so taking this end does not seem to be consistent with aiming at a good for them. Since either course of action clearly causes harm to one or both, the fact that this end is not an end for them makes it impossible to see the action as consistent with the belief that their lives are valuable. Following the rule for this reason, then, would be to act contrary to the very basis for the rule.
The only way to avoid this problem is for the end to be an end for them as well, something that can genuinely viewed as a good for them. We have, then, four requirements that the end must meet: 1) right action must be constitutive of it, 2) it must be a possible end for those affected by the action, 3) it must rightly be seen as good for the world, and 4) it must be plausibly seen as the highest, or most important, good. Kant takes what he calls the highest good, a world in which moral action always results in achieving the goods sought, to be one such end. Other possibilities include union with God or human flourishing. In each case, it is plausible (or necessary) to claim that the end can only be sought by right action, the good that is sought is a good for all human beings, and its achievement is good for the world. Each of these ends, then, is a candidate for what we seek when we act morally.{35}
The fact that such action results in the deaths of both Mary and Jodie, however, must be reckoned with. Since it results in their deaths, it seems impossible to see the moral action (or failing to act) as seeking these goals for them. Even when the end sought could be, in general, an end for them, the fact that it is not just harm that immediately results, but death, makes it hard to see how it can be an actual end for them in this case. But notice that this is only true if what we are concerned with is life in this world.{36} Kant argues that we must believe the soul to be immortal so that there is an infinite period of time in which persons can move toward the highest good.{37} Even with an infinite amount of time, however, the world seems to be such that we will be frustrated. The goal itself does not always seem to be achievable, given the world that we experience, because of the way moral action and the ends we seek can pull apart. Nature simply is not such that we can avoid such frustration even with infinite time. Believing in the possibility of achieving such a goal, then, similarly depends on believing in the possibility of a God capable of arranging the world such that the goal is possible.{38} For Kant the end in question is something that we just do seek, in so far as we are moral beings, so there is no separation between our seeing this end as important for Mary and Jodie, and their own seeking of it. Both stem from the same source. In so far as we are morally obligated to encourage moral action on the part of others, then, we are morally obligated to see this as a possible end for them, a part of their project.{39} It is also important that for Kant following the moral rule is itself necessary for actually making progress towards achieving such a moral, transcendent end because the end is itself partly constituted by the moral law itself. This precludes the possibility of achieving the end in question by any means other than moral means.{40}
If we admit a transcendent dimension to human life, then, we begin to see the possibility of acting in a way consistent with seeing the lives of Mary and Jodie as valuable, their projects as in some sense our own, valuable to us, even when our failure to perform the surgery results in the death of both infants.{41} The special relationship between the rule and the ends in question, as illustrated in the case of Kant, explains why this end can only be sought by following the rule, and underscores the necessity of following the rule even when we focus attention on the ends of action, and not on the rule or the action itself. Since not performing the surgery results in the death of both Mary and Jodie, the only way to sustain the intelligibility of the project of morality is to see their projects, their lives, as not limited to the material world as we experience it. In short it is to believe in the possibility of God's existence and of the immortality of the soul.
Though the end that Kant selects (the highest good) is particular to his own ethical view, the central move can be generalized. I have already mentioned two other candidate ends (flourishing, relationship with God). In each, moral action must be constitutive of the ends that we seek, to make it impossible to achieve what is thought to be good by any means other than moral means. That good must, of course, be general, be good for those affected by the action and good for the world in general, given the arguments we have seen here. Nonetheless, the heart of the solution is the assertion of this strong connection between what is right, and the good we seek. The problem that leads to the need for belief in God, then, can also be generalized.{42} We simply have no reason to believe that such a connection is born out as we act in this world. Right actions pull apart from the achievement of good ends and threaten the intelligibility of moral action as seeking such ends. The only solution is to look beyond this world, to look beyond to something capable of transforming the world and securing the connection that is needed.
As we have seen, then, there is a general problem in making moral action intelligible as action for ends. It is important that what we believe to be right connects up with the ends that we seek. We have also seen that there are several possible ways of thinking about this connection.
First, we could think that what achieves the ends that are important to us defines what is right; this is the consequentialist solution and avoids the problem by defining it out of existence.
Second, we could think that what is right is of necessity a condition for our achieving the ends that we seek. For Kant, right action is what makes us eligible for, deserving of, achieving our ends. For most modern virtue theorists, human flourishing cannot be achieved through immoral action. In the world we inhabit, however, we can come to have reason to doubt this connection can be realized, that we can indeed achieve our ends through moral action. No matter how often we do the right thing, it is always possible that the ends we seek will not be fulfilled; suffering can always result. Sometimes, as in the cases important here, suffering appears to be the only possible result. Since this connection does not seem to be hold in the world we experience, but belief in its possibility is crucial to maintaining the intelligibility of morality in terms of ends, taking this solution requires the belief that the world we experience is not how the world has to be, it requires the belief that the connection is at least possible. Since we cannot secure it on our own, since we cannot change the very nature of the world, it requires the possibility of God to secure it for us.
Third there is Korsgaard's approach, which attempts to maintain that sometimes moral action is important to the fulfillment of the ends we seek, but sometimes the fulfillment of our ends should allow us to deviate from what is ordinarily considered right. The challenge for this view, as we have seen, is to provide some form of principled dividing line. Korsgaard's own attempt, as we have seen, fails. It is worth wondering if any such account could succeed. On what principled basis could consequences define what is right only sometimes? If we have a principle defense of a moral rule, how can we make consequentialist exceptions to it without admitting that consequences are playing the defining role? How can we hold fast to any rule, if no rule connects up with ends in the way required?
Such questions are not a conclusive argument, but they do, along with the other arguments I have presented today, suggest a provocative conclusion. It may be that we are faced with a fundamental choice if we want to be moral: either accept some form of consequentialism, or hold onto hope that there is a god who can secure the connection between good ends and moral actions.
{1} Christine Korsgaard, "The Reasons We Share" in Creating the Kingdom of Ends (Cambridge, Cambridge University Press, 1996), page 296.
{2} This assumes, as is necessary given Williams intention in presenting the case, that there is no course of action that can save all the Indians.
{3} Robert Adams, "Moral Faith" Journal of Philosophy, vol. 92 no. 2 (February 1995) page 80.
{4} Adams, page 80.
{5} Christine Korsgaard, "The Right to Lie: Kant on dealing with evil" in Creating the Kingdom of Ends (Cambridge, Cambridge University Press, 1996), pages 133-158. Hereafter TRL.
{6} See TRL, page 150.
{7} TRL, page 147ff.
{8} TRL, page 147.
{9} TRL, page 149.
{10} TRL, page 149.
{11} TRL, page 150, original emphasis.
{12} TRL, page 150.
{13} TRL, page 150. I do not deny that Kant makes claims about responsibility that suggest this conclusion, what I object to is the thought that these passages can be taken in isolation, apart from arguments Kant gives elsewhere concerning the end of moral action. See below, page 11ff.
{14} TRL, page 150.
{15} TRL, page 150.
{16} TRL, page 150-151.
{17} TRL, page 150.
{18} TRL, page 151.
{19} The claim here is tentative because it is not at all clear to me that Rawls' own examples don't fall victim to the same problem. The problem is, however, at the very least much more obvious in the case of individual morality Korsgaard discusses.
{20} TRL, page 151.
{21} She suggests that Kant's argument relies on a certain teleological claim, that our instinct to improve our lives "cannot universally be used to destroy our life without contradiction" (TRL, page 158n20). She does not believe such teleological claims have any place in the test associated with the Formula of Universal Law, and hence does not find his argument convincing. Further, she does not see any way, given her understanding of this test, to argue suicide is morally illicit (TRL, page 158n20). Though I will not dispute her view here, I do not find her arguments convincing.
{22} TRL, page 152.
{23} TRL, page 152.
{24} TRL, page 152.
{25} Christine Korsgaard, The Sources of Normativity (Cambridge, Cambridge University Press, 1996), page 162.
{26} TRL, page 154. Further, it is worth noting that the general problem of securing the intelligibility of an action with respect to ends is not entirely solved simply by moving to a two level view (even if it could be defended). The moral ideal, remember, is the goal when we are justified in deviating from the strictures of the goal, but what is the positive goal that we seek when the bad consequences of an action in accord with the ideal fall just short of allowing us to deviate? Might there not be some cases where following the moral ideal leads to bad, but not quite bad enough, consequences? How do we secure the intelligibility of action in accord with the ideal in such circumstances? Say Jim must choose between executing one Indian and watching twenty be tortured (but not killed)? Or between killing one and watching the twenty be separated from their families and children with no hope of ever returning? What is the positive good we seek in such cases? Notice, importantly, that the goal cannot be the ideal itself, for if the ideal is the goal when we act according to the ideal (not just when deviating), then Korsgaard cannot generate her grotesqueness objection in the first place. There would no longer be a need simply to deny responsibility for the bad outcome, because there would be an overriding a positive good that one is seeking. This thought is, I hope to argue, the heart of Kant's own solution to the problem.
{27} Immanuel Kant, Critique of Practical Reason (Cambridge, Cambridge University Press, 1997), 5:110-111. Page references to Kant will be to the standard German Academy edition pagination.
{28} Kant, 5:111ff.
{29} The reason Korsgaard's view falls victim to the first objection I mentioned above is because she appears to endorse this same view of the relationship of the goal and moral rule in cases where deviation from the ideal is acceptable. In such cases, the moral ideal becomes the goal. She then tries to suggest that this goal can be sought be means inconsistent with the moral rules that are part of the goal, but this, I argued above, cannot be made intelligible.
{30} Obviously, "a good life" is taken to mean something more than doing what is morally required; it also implies happiness or fulfillment.
{31} Kant, Critique of Practical Reason, 5:122ff.
{32} The British Court of Appeal nicely summarizes the details of the case in the introduction, summary and conclusions of the decision in the case. The case number was B1/2000/2969, and the decision was released on September 22, 2000.
{33} I will assume without argument, though I think there are compelling arguments that can be given, that the separating of the infants would be a case of directly killing Mary to save Jodie.
{34} Whatever reason we have for valuing the lives of other human beings, whether it is because the are rational, created in God's image or whatever, the requirement to act in a way that recognizes that value still remains.
{35} The fact that each of these ends are ends for Mary and Jodie is important, because only in this way can we see seeking this end as not using them to achieve an end. Since in each of these cases, the end sought is also the ultimate end for them, we are furthering their projects as much as are own, we can see this end as consistent with seeking their fulfillment. As will be argued in a moment, however, more needs to be said before this claim can be fully made good.
{36} Notice that it is not the value of their lives that needs to be secured, it is the need to recognize or, perhaps better, acknowledge the value of their lives through our action. This does not mean that life in this world is not important, not good, or cannot be seen as such without admitting the possibility of an afterlife. All it means is that in this case, we need to see the possibility of there being a further good in order to make our action intelligible as seeking a good for them given what results from the moral course of action.
{37} See Kant, The Critique of Practical Reason, 5:122-124.
{38} See Kant, The Critique of Practical Reason, 5:124-132.
{39} In fact, more than just a part of their own project, but the ultimate end of their life.
{40} These arguments are here phrased in terms of Kant's own theory, but I believe they apply equally well to other non-consequentialist ethical views that can meet the challenge discussed here.
{41} The problem is not finding value in their lives, but acting in a way consistent with recognizing that value.
{42} Of course, in the case of Aquinas, the end explicitly acknowledges the need for God, so no further argument is required. In his case, this argument only highlights why reference to God is a necessary part of that end.