From Practical Normativity to Moral Normativity

So far, I have been discussing reasons for action in general, and, in the last section, I suggested that we should make a distinction between explanatory reasons and justificatory reasons. Indeed, our experience of normative constraints indicates that not all ways of responding to a situation by an agent are acceptable: an agent can doubt that her ways of responding are correct even in the face of full information. This presupposes that there are wrong and right ways of being for agents, that is, wrong and right ways in which their appetitive and cognitive faculties can be structured.

The idea that an explanatory reason might not be justificatory, though, does not yet account for moral reasons.¹⁴ Sometimes I have used the word “moral,” but that was not for the sake of clarifying that term: it was only to offer examples of normativity, without any concern for what is special about morality. What I have been suggesting so far holds for the reasons we have to buy a nice painting, to play a nice piece of music, to go for a walk, or to choose a career. It is true that I have introduced the notion of normative constraints, but I did not consider whether its domain is co-extensive with that of morality. For what I have said, the notion of normative reasons that I used might hold for all that we do, in general. Moral actions and moral reasons must be a subset of those that we have so far considered. I want now to suggest that moral reasons are a subset of normative reasons. I will try to define what the characteristic feature which picks this subset of normative reasons is. The upshot will be that they are not as special as it is sometimes claimed.

I would like to start by suggesting that one of our everyday uses of the word “moral” - one that I take to be central - suggests that the main difference between moral and non-moral normative practical reasons is that the moral imply an obligation that the non-moral lack. In a sense, non-moral normative reasons also indicate what one ought to do, but this is a weak sense of “ought:” the indications of non-moral normativereasons can be ignored simply by giving up assumptions on which those reasons rest. As Kant put it, those reasons imply imperatives that are hypothetical, not categorical. The assumptions might concern a result that the agent may want to attain or the kind of person that she may want to be. “If you want to play that passage on the violin well, you must study it in separate strings.” That statement might give Toni a normative reason to study the passage in separate strings. Toni could also decide not to study the piece that contains that passage at all, however. “If you want to be the kind of person who is always invited to parties, you ought to wear brand-name clothing and always smile.” Good, but if I do not want to be that kind of person, I can avoid doing those things.

In the case of moral normative reasons, by contrast, an obligation is stronger (cf. Audi 2001, 221-2). “If you want to help that person who is starving you can give her a piece of your sandwich.” If Toni said that he has no interest in helping that person - provided that he is the only one who can do it, and all other necessary amendments - we would think that he is not as justified as he was when he decided not to practice a certain passage on the violin in separate strings. “If you do not want to be a liar, you must avoid telling even small lies.” The answer “I do not care about not being a liar” does not seem as acceptable as the answer “I do not care about being a party type.”

Moral normative reasons are problematic since they seem to request a stronger normative constraint than is required by non-moral normative reasons. Where does the stricter constraint come from? Now I want to address this question.

Let us consider our previous example: “if you want to play that passage you must study it in separate strings.” Let us imagine that this is a genuine justificatory reason: it is true that the given order (i.e. the technical level of the player, the characteristics of the piece and so on) calls for a completion of a certain kind (studying in separate strings). That order, however, can also be completed in other ways: for example by studying an easier but equally pleasant piece. The existing order calls for a response from the agent, but there is more than one way to complete that order and the subject can choose any of those ways. Hence, there is a set of viable responses offered by the situation in which the agent is placed, which are allowed. On the other hand, there will usually also be ways of actualising viable possibilities that would not count as completions of the existing order at all: for example, one could (in a physiological sense, not in a normative sense) play badly or destroy the violin. Those options, though, would not make the situation better off. These possible courses of action fit in the set of possible courses of action which are not allowed by the situation in which the agent finds herself.

When we face non-moral, normative reasons, like the case that we are considering, it seems that an open possibility or a legitimate option for the agent is just to renounce to a presupposition, which divides the

Action, Human Nature and Morality 119 possible courses of action offered by the situation into what is allowed and not allowed. Indeed, suppose that the agent decides to give up a career in music: that decision reshapes the distinction between allowable and unallowable courses of action. (Of course, some courses of action could remain unallowable, like playing badly while someone is hearing, or breaking a valuable violin, which could be resold if one gives up a musical career.) That means that, once the presupposition that one wants a career in music is given up, some facts, which constituted reasons before, do not constitute reasons any longer.

When we deal with moral reasons, by contrast, it seems that the distinction between courses of action that are allowed and those which are not allowed cannot be reshaped at will by the agent. A choice among one of the ways of a completing order is compulsory for a suitably situated agent. There might be a variety of ways in which a certain situation can be made better off, but the stronger force of moral obligations seems to depend on the fact that in these examples the prospective agent can choose a course of action among a set which is fixed and she cannot change at will. I can deal with the starvation of a beggar by giving him money, by offering him food, by finding him a job, etc. However, giving him rotten vegetables can never be an option, unlike in the case of the mediocre violinist who burns her cheap instrument in her fireplace, after devoting herself to something else. In the case of non-moral normative reasons, there is a range of open possibilities for the completion of order, which may be reshaped at the will of the agent. In the case of moral reasons, by contrast, it seems that the set of allowed viable possibilities must be accepted by whoever happens to be in a certain situation, apart from her will, desires, or choices. We can be put it in this way. Non-moral normative reasons are escapable, but moral reasons are not.¹⁵ On what does this difference depend?

Let us note that, in a moral case, the fact that a response from a set of viable possibilities is obligatory means that a response is requested no matter what desires, interests or inclinations the agent has, that is, the response is requested from everyone. In those cases, a response is requested regardless of the subjective characteristics of the putative agent. No one should lie to one’s friends, at least other things being equal, no matter whether one desires doing so, or whether one has an interest in doing so, or whether one did not make one’s point in life being truthful and trustworthy. By contrast, one might decide not to play music or not to go to parties, if that best satisfies one’s desires, or one has an interest in doing so, or one has other specific characteristics, for example, one is not very talented in music or does not find a life of small talk and gin and tonics very fulfilling. I think that one can generalise this point by pointing out that, in non-moral normative reasons, the obligation is conditional upon the assumption that the agent has an intention to be a certain kind of agent or to choose a certain kind of end. In the case of moral normativereasons, the obligation holds apart from the kind of person that an agent intends to be and the kind of ends that the agents want to attain.

Before proceeding with the analysis of moral reasons, I would like to address an objection that could be raised against the point that I have just made. A reader could counter that an agent can change her moral obligation at will, just as the musician I imagine can change the set of allowed viable courses of action by deciding to give up music. Here is an example. The musician has a moral obligation towards her parents to practice three hours per day, as her teacher requires, since her parents pay for her tuition. When she decides to give up music, she does not have that moral obligation towards her parents any longer.

It seems to me that this objection can be answered. In the non-moral example, the will of the agent changes the dividing line between the courses of action which are allowed and those which are not allowed in the situation. In the moral case, by contrast, the will of the agent changes the situation itself, and this, in turn, changes the facts that furnish reasons to the agent. Thus, in the moral case, the change of the will of the agent purports that she does not take violin lessons any longer, and that relieves her from the obligations towards her parents, who were paying her tuition.

Let us now go back to the main line of argument. We have seen above, while discussing non-moral normative reasons in the last section, that an agent can always ask herself whether the reasons that seem normative to her are really such. We have also seen that that question presupposes a belief about what an ideal agent - i.e. a person whom the agent trusts and respects - would do in a circumstance similar to hers. It is natural to expect that the chosen person may be a concrete or an imagined person, who in a certain way paradigmatically exemplifies the kind of person that the agent wants to be or who fully realises certain ends that she has chosen.

In the case of moral reasons, the criterion can no longer be a (concrete or ideal) person of a kind to which the agent belongs or intends to belong: the question then is how any human being should respond in the given circumstances, that is, anyone who can be fully counted among humans. I do not think here of an abstract “point of view of humanity.” It is rather the point of view of a concrete human being who has all and only the qualities that make him or her a well-flourishing human, that is, someone who chooses not on the ground of belonging to or intending to belong to a certain specific category, but only under the generic assumption that she or he is human.

At this point, I think that I need to rest again and consider another possible objection. This will allow me to clarify my proposal better. The objection has been raised against views that account for reasons for action in terms of conditional thoughts of the agent. A typical view of this sort is John McDowell’s thesis that a right action is the action that an agent would choose in her situation if she were fully rational (McDowell 1995). Williams has rightly objected that there are reasons that agents have precisely in force of the fact that they lack full virtue (Williams 1995): Toni might have a reason not to go to parties since he knows that when he is at parties he cannot restrain himself and drinks too much. If he were fully virtuous, though, he would not have that reason at all. In order to overcome this line of objection, Michael Smith has distinguished an advice model of an agent’s counterfactual thinking about reasons from the more traditional example model of McDowell (Smith 1994). According to Smith, an agent grasps her reason for action not by thinking what she would do in the situation if she were fully virtuous, but by thinking what her fully virtuous counterpart would want her actual self to do in the actual situation. Although this move overcomes Williams’s objection to conditional thinking accounts of reasons, it has still the burden of explaining how to make sense of the idea of a fully virtuous counterpart having desires about an actual weaker self (Bedke 2010).

The view that I support in this chapter rests also on the idea that an agent grasps her reasons via conditional thoughts. I think, however, that it is not open to the line of criticism that I just mentioned. Let us see how I can resist Williams’s point against McDowell. My proposal is different from McDowell’s in two respects. First, my view does not involve counterfactual thinking by the agent about her fully virtuous counterpart and her reasons. Rather, it involves conditional thinking about another fully rational, virtuous, ideal agent, who can be a real person whom the agent sees as exemplar. Second, my proposal does not require that the agent identifies her reasons in a situation with the reasons that the ideal agent would have in that very situation. By contrast, it requires the agent to consider what reasons the ideal agent would have and reconstruct her own reasons accordingly. Hence, unlike McDowell’s agent, my agent does not think “What would I do in this situation if I were fully virtuous,” but “How would a fully virtuous human being respond to this situation and what could I do - given how I am - to get close to that?” For example, let us imagine that Toni thinks about an ideal agent, who has no reason not to go to parties since that agent has full control of her drinking. Toni could still have a reason not to go to parties if he sees doing that as a way to get closer to the ways of responding to situations (e.g. not getting drunk) which are characteristic of his ideal agent. In this way, my account is closer to advice than to exemplar versions of the conditional thinking conception of reasons for action.

Let us now turn to Bedke’s further criticism against advice-conceptions. Unlike those conceptions, my proposal does not purport that the virtuous, ideal agent identifies with rhe actual agent to find out what his own desires would be, if he were in her position. According ro my view, the actual agent thinks about what desires the ideal agent would have and tries to find out what courses of action she should follow to approach those of the ideal agent. This thought delivers her reasons for action. The counterfactual thinking, therefore, does not involve considerations about the attitudes and thoughts of the ideal agent about the actual agent, as the advice versions of the conditional thinking accounts of reasons do.

I need to make a last remark about the distinction between general normative reasons and moral reasons. We can observe that there are different degrees of strength within the domain of reasons that we would consider moral.¹⁶ A reason not to lie might be a moral one, and therefore it can be much stronger than a reason to practice a piece of music in a certain way. However, it might be still weaker than a reason not to kill someone, for example. A reason to leave a seat to an elderly person in a bus can be even weaker than a reason not to lie: there are certainly moral reasons to do that, but failing to respond to them would not necessarily be a moral failure. In the case of supererogatory actions, there might even be moral reasons to which one is not obliged to respond at all. I cannot offer an explanation of this aspect of our moral experience here, but I need at least to show that the difference between levels of strengths of moral reasons does not jeopardise my account of the distinction between moral and non-moral reasons.

One worry could be that my way of distinguishing moral from non-moral reasons rests on the claim that the former are inescapable and the latter are escapable. Does not the fact that there are weak moral reasons show that moral reasons are also escapable? I would like to claim that it does not. The point of the distinction between inescapable and escapable reasons was that the latter merely fade away when the will of the agent changes. When the musician decides to give up music, the facts entrenched in the situation - although they remain unchanged - stop delivering her reasons. She just does not see the situation as calling for perfecting interventions form her. Moral reasons, by contrast, do not fade away just for a change in the will of the agent, if the situation remains unchanged. (As we have seen, the change in the will of the agent can change the situation but that is a different case.) This is the sense in which moral reasons are inescapable (Hampton 1998, 104-5). And this is true of weak moral reasons as well: they persist even when one is not responding to them and that lack of a response is not a moral failure. When I have a reason to offer my seat on the bus and I decide not to do it, that decision does not make the reason fade away. Similarly in the case of supererogatory actions: the fact that I decide not to do what I have reason to do, does not trump my reasons, which keep calling for my response.

The fact that the strength of moral reasons comes in degrees is significant also on the side of strong reasons. Moral reasons that are very strong may become even legal obligations: their violation cannot be tolerated and legitimate authorities can punish the one who violates them: “do not kill,” “do not steal,” etc. I cannot add more about this

Action, Human Nature and Morality 123 distinction at this point and discussion of it will have to wait until the third part of this work.

In this section, I have argued that, from the point of view of an agent who chooses, asking whether those that seem morally normative reasons to one are really such, means asking what is human in the concrete given circumstances and this involves understanding what being human means. That conclusion allows me to note that moral reasons limit the range of possibilities open to agents more than non-morally normative reasons because of the second role played by human nature in agency, as discussed in the last section. The limitations of moral reasons depend on a criterion that requires the recognition of what is human and the understanding of how humans “become” in different conditions, in which their lives can be realised. Let us remember that so far I have been reasoning from the point of view of the agent, that is, of a subject who chooses what to do and how to do it. When an agent asks herself whether the reasons that seem normative to her are really such, she cannot reach an absolute point of view (Audi 2001, 179-80). Even if she decided to follow some moral authority, that choice would still depend on reasons that she sees as such and thus the point of view that she could reach would not be absolute.