Please do not quote or share without the author‟s permission
Emotional Expressions of Moral Value
In reading the Nicomachean ethics one lesson to be learned is this: how people feel reveal what they value. When a person is pained at parting with money in order to repay a loan, this reveals that he overvalues money. And when a person gives away too much money with pleasure,
1this reveals that he undervalues money. The rash man feels no fear in battle and this reveals that
he does not properly value his life, whereas the courageous man does feel fear thereby showing
2that he does value his life. Aristotle goes so far as to integrate an agent‟s feelings into his
description of whether an action is involuntary; an agent who acts by force or because of
3ignorance and who is pained, regretful, and distressed by what comes about acts involuntarily.
In “Moral Luck” Bernard Williams revisits these themes of involuntary action, feelings
4and values. Williams describes a lorry driver who, through no fault of his own, runs over a child. The example, as I understand it, is meant to be one in which the agent neither intentionally nor negligently kills the child. Williams does not elaborate on the details of the case, but doing so will be helpful to our discussion. Imagine that the street is commercial—a place where children
do not live or play. The driver kept within the speed limit, was focused on his surroundings, and so on. The child, let‟s suppose, ran out from behind a parked car, which is why the driver didn‟t see the child until it was too late; he hit the child before there was even a chance to apply the
1 Nicomachean Ethics, Book IV, chapter one. 2 Nicomachean Ethics, Book III, chapter six through nine. 3 Nichomachean Ethics Book III, chapter one. See also Rosalind Hursthouse‟s article “Acting and Feeling in
Character: Nicomachean Ethics 3.i,” Phronesis (1984). 4 “Moral Luck” in Moral Luck (Cambridge University Press, 1981). All the references to and quotes by Willaims are drawn from this article.
breaks. Williams claims that the driver “will feel differently from any spectator, even a spectator next to him in the cab” (p.28). He labels this feeling “agent-regret.”
The goal of this paper is to explore the values revealed or expressed in the driver‟s
feeling agent-regret. There are really two kinds of values or evaluation at issue: the driver‟s
feelings reveal a negative evaluation of his action—he regrets his killing of the child (and in a
way that is different from how others might regret it)—and that act evaluation is itself grounded
on something else the agent values. To illustrate these two values and their relation, think back to the repayment example. An agent who repays a loan, but is pained by doing so, negatively evaluates his repayment and he has this negative evaluation since the repayment is judged as contrary to the value he places in money. The question is what sort of negative evaluation does the driver make of his action and what value does he see his action as contrary to or in conflict with.
I want to defend a particular hunch, namely that the driver‟s agent-regret is a moral
feeling, by which I mean that the values that give rise to or find expression in the feeling are moral. The challenge this paper undertakes is explain what sort of moral evaluation the driver makes of his action and what moral value grounds this act evaluation.
In section one I point out that not all forms of agent-regret are moral and I discuss some initial difficulties with explaining why the driver‟s agent-regret is moral. In the second section I
aim to show that appeals to moral obligation cannot explain why the driver‟s feeling of agent-
regret is a moral feeling. I describe what I see as the two most plausible views of moral obligation, one deliberatively oriented and the other performance oriented, and show that neither can make sense of the driver‟s agent-regret. The goal of section three is to show that there is a
morally evaluating an action without reference to moral obligations. I suggest that there are at
least three conditions on the success of an action: first, the action must result from deliberations that reflect what the agent values; second, the action must realize the ends reached in deliberation; and third, the agent must have the right values. The driver‟s action meets the first and third conditions but it does not meet the second condition, and so his action is a failure. Moreover, it is a moral failure because the ends he fails to realize are moral ends, that is, they reflect his moral values.
Section One: Moral and Nonmoral Agent-Regret
Notice that nothing Williams says in elucidating the concept of agent-regret obviously
5 of agent-regret is “how much indicates that it is a moral feeling. The “constitutive thought”
better if I had acted otherwise” (27). But there is nothing about this thought that is obviously moral. I can imagine having this thought upon discovering that seven is the winning number rather than the number I betted on. Nothing in my thought, “how much better if I had bet on 7”
reveals anything about my moral values. It does tell you that I value winning or making money, but there is not anything obviously moral about that value. So, on its face, there are certainly nonmoral examples of agent-regret. And yet the driver‟s regret strikes me as moral and it is this
intuition that I want to explain.
Perhaps the driver‟s thought “how much better if I had acted otherwise” indicates his
belief that his act lacks justification? Kant thought there was a close connection between moral value and justification. But it strikes me that moral value is just one element of justification and
5 I plan to put aside the issue of how thoughts and feelings are related, that is, whether thoughts constitute, cause or are caused by feelings.
6 Moreover, Williams warns us so an action can be unjustified without being morally deficient.
away from understanding the constitutive thought of agent-regret as a thought about justification. He claims “we should not entirely assimilate agent-regret and the wish, all things taken together,
to have acted otherwise.” (31). At least in some cases he thinks that an agent can think his action
justified and yet still feel agent-regret.
Does Williams provide any other clues to how we might understand the driver‟s agent-
regret as a moral feeling? Williams points out that agent-regret motivates what might seem to be moral actions. Williams points out that the driver will, if he can, make reparations or restitution or engage in self-punitive actions (28). I wonder whether self-punitive actions are necessarily moral actions. I might well punish myself (slap my thigh) after losing a point in a tennis game, even when I played my best. This is a response to not measuring up to my ideals or goals, and perhaps even a method of getting myself to measure up. But none of this is about moral ideals or moral goals, but rather about measuring up to standards, standards which are not moral but rather
7about good tennis playing.
The descriptions “restitution” and “reparation” however, do have a moral ring. Notice, however, that for an action to bear these descriptions the action must be done with a particular intention. That is, for an act to be an act of restitution the agent must do the act with the intention of making up for what he has done. If I bring you flowers, but not in order to make up for a missed appointment, but so that I have something nice to smell while I‟m in your dreary office,
my bringing you flowers is not an act of restitution for the missed appointment. It is necessary
6 I argue for this view in Moral Action and Moral Motivation (Ph.D dissertation). Michael Smith also argues for this view, although in a different way in The Moral Problem, p.185 and in “In Defence of The Moral Problem: A Reply
to Brink, Copp, and Sayre-McCord” (especially pp.280-291) in Ethics and the A Priori. 7 The cause of my not measuring up the standards of good tennis playing needn‟t be a moral failure. Perhaps the
other player is simply better and out hit me. It needn‟t be that I‟m lazy or careless. And are laziness and carelessness necessarily moral failings anyway? Or are they moral failings only when they get in the way of successful moral activity?
that I act with the intention of making up for my missed appointment. That is, I must see some previous action of mine as lacking. Moreover, for restitution to be a moral action, I must think of my previous action as morally lacking. And so we are back to the question of how the driver
evaluates his action; in what way does he see it as morally lacking and what are his reasons for viewing his action in this way? Without such an account in hand, we cannot explain why the driver‟s feeling of agent-regret is a moral feeling.
Section Two: Unsuccessful Accounts of Agent-Regret
In this section I aim to show that appeals to moral obligation cannot explain why the driver‟s feeling of agent-regret is a moral feeling. I will describe two views of moral obligation that I find appealing. The first is deliberatively oriented and the second is performance, and hence outcome oriented. Neither view, when coupled with what I will call the Evaluative Thesis—that an action can only be morally evaluated by reference to the agent‟s moral
obligations—can make sense of the driver‟s feeling of agent-regret as a moral feeling.
According to the deliberative view of moral obligation, our sole moral obligation is to
8 When deliberate in a particular way. There is no moral obligation to perform a specific action.such a view held in combination with the Evaluative Thesis, it follows that all actions are morally evaluated solely on the basis of the agent‟s deliberations that led to the action.
To make the view more concrete consider a Kantian version of this deliberative view of
9moral obligation. According to the Kantian, the mark of a moral value is its connection to
8 Of course, one can also speak of deliberation as an activity, but doing so makes it difficult to describe the view I have in mind. 9 As I‟ll show later, one needn‟t be a Kantian to accept a deliberative view of moral obligation.
10 To value rational nature. Rational nature, according to Kant is unconditionally valuable.rational nature properly is a matter of the agent‟s practical deliberation having certain features. In
particular, to value rational nature unconditionally just is to adopt and pursue the end of respecting rational nature and to give this end its proper status with respect to one‟s other ends.
Adopting and pursuing the end of respecting rational nature generates a variety of further sub-
11ends. For instance in some cases respecting rational nature will be a matter of sustaining rational nature, in other cases a matter of promoting it. Moreover, in giving the end of respecting rational nature its proper status relative to other ends, the agent will take there to be a deliberative presumption against negatively impacting rational nature, even if doing so would
12advance some end that he has. This presumption can be rebutted, but only in very specific ways. So, for example, there is a presumption against killing another person, since this negatively impacts their rational nature. However, this presumption against killing is rebutted in cases in which we say a person‟s right to life can be overridden (trolley case), forfeited (self-defense), or
13waived (euthanasia). In addition, the more specific ends (sustaining and promoting rational nature) and constraints (not negatively impacting rational nature) generate yet further sub-ends, such as being aware of whether one‟s action will sustain, promote or negatively impact rational
10 Groundwork of the Metaphysics of Morals, section one. 11 The relation between the end and sub-ends is a rational connection. If one adopts the end of respecting rational nature then one must, on pain of irrationality, adopt the relevant sub-ends. 12 I use the formulation “negatively impact rational nature” rather than “undermine rational nature” or “disrespect rational nature” since not all killings are instances of undermining or disrespecting rational nature. Insofar as Kantians wish to capture the intuition that killings are presumptively morally problematic, it is important to think of killing as problematic because they negatively impact rational nature. 13 In cases of abortion, the fetus does not yet have a rational nature and so abortion cannot be considered to negatively impact rational nature. Kantians might attempt to capture the intuition that abortion is morally problematic, though not necessarily impermissible, by appealing to the fact that the fetus potentially has rational nature and so has a special moral status. There are, however, difficulties with this line of argument. See, for example, Joel Feinberg‟s article “Abortion” in (ed. Regan) Matters of Life and Death (Temple University Press, 1980). 14 For a more nuanced Kantian account along these lines, see, for example, Barbara Herman‟s book The Practice of
Moral Judgment (Harvard University Press, 1993).
According to the deliberative-Kantian view of moral obligation, our sole moral obligation is to adopt and pursue the ends, sub-ends and deliberative constraints related to respecting
15rational nature. This deliberative-Kantian view, combined with the Evaluative Thesis (an action can only be morally evaluated by reference to the agent‟s moral obligations), implies that we are
to morally evaluate actions in the following way. If an agent does not have or pursue the end of respecting rational nature, or does not give it the status it should have relative to other ends, or if the agent does not adopt and pursue the relevant sub-ends, then the resulting action is morally wrong. For instance, if an agent knowingly kills another without rebutting the deliberative presumption against doing so, his action is morally wrong. And if an agent unknowingly kills another while ignoring whether his action would have this result, his action is morally wrong. In both cases the agent‟s deliberation does not meet the criteria described above and this explains why the action is evaluated as morally wrong. However, if an agent‟s action is the result of
deliberation that does have the features described above, then the action has moral worth. So the moral evaluation of action depends solely on whether the agent‟s deliberations reflect the fact
that the agent values rational nature unconditionally.
With these materials in hand, can we explain why the lorry driver‟s agent-regret is a
moral feeling? The driver, as I mentioned at the outset, neither knowingly nor negligently kills. He adopts and pursues (I‟m supposing) the end and sub-ends related to respecting rational nature.
He pursues his end (doing his job/driving down this street) on the condition that and with the
16belief that his action will not kill anyone (any rational natures). Moreover, he has and pursues
15 Notice that the deliberative view of moral obligation makes use of the ordinary (and conceptual) connection between valuing and the content and structure of one‟s deliberations. Consider a person who values money. Such a person will adopt the end of making money, sustaining the money they have, and there will be a deliberative presumption against losing money. Moreover such a person will be on the lookout for which of his actions make, sustain, or lose money. 16 The child is not so young as to lack a rational nature.
the end of being aware of whether his action of driving will kill anyone. The trouble is that he arrives at the false belief that driving down this street will not kill anyone. His belief is false but reasonable. He meets all the reasonable requirements for justification in forming this belief. If actions are morally evaluated solely on the basis of an agent‟s deliberation, then the driver‟s action of killing the child is not morally deficient, since the action results from deliberations that have the features that amount to valuing rational nature unconditionally. And so the deliberative-Kantian view cannot account for the driver‟s negative moral evaluation of his action as revealed
by his feeling agent-regret.
As I mentioned earlier, there are other versions of the deliberative view of moral obligation apart from the Kantian one described above. Some philosophers, such as Aristotelians,
17 On a deliberative Aristotelian view of moral think that living well is what is morally valuable.
obligation, one is only morally obligated to adopt the end and sub-ends related to living well
18(eudaimonia). Living well is the chief human good and valuing living well is a matter of
adopting the end of living well and so also the sub-ends of living justly, generously, courageously, and so on. There is a deliberative presumption against killing, since it is contrary
19to generosity and justice unless shown otherwise. However, notice that an agent can deliberate
well with respect to the end and sub-ends related to living well while nevertheless blamelessly arriving at a false belief. If actions are morally evaluated solely on the basis of an agent‟s deliberation, then the driver‟s action of killing the child is not morally deficient, since the action results from deliberations that have the features that amount to valuing living well as the chief human good.
17 Some Aristotelian‟s eschew the talk of morality all together, but others aim to give a virtue theoretic account of
moral value. 18 Nicomachean Ethics, Book I. 19 See for example, Philippa Foot‟s discussion of the relation between killing, justice, and generosity in “Euthanasia” in Virtues and Vices (University of California Press, 1978).
20 so long as moral Regardless of which substantive moral theory one appeals to,
obligations are deliberative, and one embraces the Evaluative Thesis, one will not be able to account for the moral aspect of the driver‟s agent-regret. Insofar as one is committed to the
deliberative view of moral obligation and the evaluative thesis, one will be tempted to think that
21the driver‟s feeling of agent-regret is irrational if it is in fact a moral feeling. Notice that when
Williams first introduces agent-regret, he states that the driver will feel agent-regret (p.28) but
that doesn‟t speak to whether the driver is rational in feeling agent-regret. Williams does go on to
claim that “it would be a kind of insanity never to experience sentiments of this kind towards anyone, and it would be an insane concept of rationality which insisted that a rational person never would” (p.29). But some might find this mere assertion unsatisfying. Moreover, Williams claims that others will rightly (p.28) attempt to get the driver to stop feeling agent-regret. If it is right for them to move him away from such feelings, how can it also be right for the driver to have the feeling in the first place?
The goal of this paper is to show that the driver is rational in feeling agent-regret, where this feeling is understood to be a moral feeling. As I see it, a feeling is rational only if the feeling has the appropriate object, that is, if the agent conceives of the object in the right way and the
22object really does have those features. So, for example, fear is rational if the agent conceives of
20 The deliberative view of moral obligation is by no means restricted to just two moral theories (Kantianism and Aristotelianism. 21 There is a difference between a feeling‟s being typical, understandable, and rational. It might be typical for enraged drivers in Los Angeles to shoot when cut off in traffic, but it is neither understandable nor rational. It is understandable that a parent would become excessively angry when his child is only slightly harmed, but it is not rational. A feeling is rational if the agent feels what he should, when he should, at the object he should, in the degree that he should, and so on. 22 Moreover a feeling is rational only if it has the appropriate intensity, duration, and so on. For purposes of this paper, I will focus only on the object or content of the feeling.
23 What I plan to show is that the what he fears as dangerous and if the object really is dangerous.
driver‟s feelings of agent-regret are rational by showing that he is right to think of his action as
Would an alternative view of moral obligation, coupled with the Evaluative Thesis, fare better at showing agent-regret to be both rational and a moral feeling? I do not believe so. Consider a performance account of moral obligation. According to the performance view of moral obligation, agents are morally obligated to perform and refrain from certain sorts of actions. For many actions, to be that type of action, the action must have a certain outcome. And so this view of moral obligation is somewhat outcome oriented. For example, whether one‟s
action is a helping or merely an attempted helping depends, at least partly, on whether the person in need is helped. And whether one‟s action is a killing depends partly on whether the person
dies as a result of what one does. If one is morally obligated to help and not to kill, then actions that fail to help or that are killings will be evaluated as morally lacking. While such a view of moral obligation, in combination with the Evaluative Thesis, seems like a promising avenue for explaining the rational and moral aspect of the driver‟s feeling of agent-regret, ultimately I do
not think the view can offer us the right account of why the driver‟s feelings are rational and moral.
23 See for example Aristotle‟s discussion of fear in Book III, chapters six through nine, and his discussion of anger in Book IV, chapter five. See also Justin D‟Arms and Daniel Jacobson paper “The Moralistic Fallacy: On the „appropriateness‟ of Emotions,” Philosophy and Phenomenological Research (2000). 24 I do not plan on arguing that the driver‟s feelings are required (rationally or morally). It seems to me that a feeling can be rational without being required. Consider the example of fear again. Perhaps there is a certain risk of being bitten by a dog. In such a scenario, feeling fear is rational but not required. If a feeling is required, then not having the feeling is not rational. Some would think ill of the driver if he didn‟t feel bad for killing the child and so
conclude that feeling agent-regret is required. But I think if the driver could convince us that he really did value the child and disvalue his action in the way that I go on to describe, then we would not think ill of him for not feeling agent-regret. This shows, I believe, that feeling of agent-regret in such situations is not required but merely rational.