1 Introduction

Martin Peterson argues for two interesting and appealing claims: multi-dimensionalism and degrees of rightness. Multi-dimensionalism is the view that more than one factor determines whether an act is right. For example, total utility could be one factor, equality another, promise-keeping a third, and so on. According to Peterson’s multi-dimensionalism, these factors are not simply ways of achieving some greater aggregate good. Degrees of rightness is the view that some actions are more wrong or less right than others without being entirely wrong. It is of course, compatible with this, that some actions are right or wrong to a maximal degree, or entirely right or wrong.

Here’s how Peterson wants to combine these two claims:

An act a is

  1. i)

    entirely right if and only if the agent has a verdictive reason to perform it and no verdictive reason not to perform it.

  2. ii)

    right to some non-extreme degree if and only if the agent has a verdictive reason to perform a and a verdictive reason not to perform a.Footnote 1

  3. iii)

    entirely wrong iff the agent has a verdictive reason not to perform a and no verdictive reason to perform a.

  4. iv)

    morally neutral if and only if the agent has neither a verdictive reason to perform a nor any verdictive reason not to perform a. (Peterson 2013, p. 33)

Multi-dimensionalism and degrees are taken to be intertwined. On Peterson’s view, if there were only one dimension, we would not need degrees; where only one dimension applies, an act is entirely right or entirely wrong. (Peterson 2013, p.11) To see this, imagine that there were only one dimension, say utility. In that case, any action would either maximize utility or fail to do so. If it maximized utility, there would be a verdictive reason to perform it, and no verdictive reason not to perform it. So, it would be entirely right according to (i) above. On the other hand, if it failed to maximize utility, there would be a verdictive reason not to perform it and no verdictive reason to perform it. So, it would be entirely wrong, according to (iii) above. Either way, it would be entirely right or entirely wrong, with no room for degrees. If there were two or more dimensions, say, utility and equality, in play, and they pulled in different directions, then the agent would have a verdictive reason to do each action and a verdictive reason not to do it. So, that action would be right to some degree and wrong to some degree, but would not be entirely right, according to (ii) above. In sum, Peterson claims degrees of rightness or wrongness only arise because there are multi-dimensions, and that an act cannot be entirely right if it is wrong on some dimension. I shall argue against both of these claims.

2 Degrees Without Dimensions

In this section, I will argue that degrees of rightness and wrongness are compatible with the denial of multi-dimensionalism.Footnote 2

In support of this claim, let us consider the analysis of “right” and “wrong”. It is tempting to think that these concepts are so fundamental that no more fundamental, non-tendentious, but illuminating, analysis is possible. An indirect approach to this issue focuses on the functional roles of these concepts. Here are some of the roles that the concepts of right and wrong play in our thought and life:

  1. 1.

    Advice: insofar as we think that an action is wrong, we are inclined to advise (or insist) that someone not do it.

  2. 2.

    Punishment, blame, or disapproval: if someone has done nothing wrong, there is a strong prima facie case against blaming or punishing her; if someone has done wrong, there is some (again overridable) reason to blame her.

  3. 3.

    Regret, shame, guilt: if someone has done nothing wrong, it seems inappropriate for her to feel guilt. Some regret may be appropriate if she is innocently involved in some horrendous event, however.

I shall argue that each of these roles fits well with the notion that right and wrong are matters of degree even if there is only one relevant dimension of rightness.

2.1 Advice and Degrees

When thinking of advice, a degree notion makes sense. Consider an example:

Vegan Relative

Imagine an avid animal rights advocate, who is advising (or urging) his family to become totally vegan. Suppose they see his point about the rights of animals, but say that he is going too far and refuse to do what he says. He could simply give up, or he could offer them a second best option. For example, “Well, if you are not ready to go completely vegan, how about cutting out all meat except fish?” or “…have a meatless day or two every week.”?

That is, he might give them a ranked list of best, second best, third best, etc. options, and encourage them to get as high up on the list as possible. A natural corollary of this is that if they choose the second (out of say, 10) options, their actions are more right than they previously were, but not as right as they might be. Here’s a second example:

Anxious Mother

I have two young teenage sons. They do not go to teen parties where alcohol is served yet, but I am sure this will come up. My preferences (and my beliefs about what ought to happen) are ranked like this: 1. Do not drink at all. 2. If you drink (or drink more than a tiny bit) call me and I’ll fetch you rather than driving home drunk. 3. Drive home drunk.

So, I might say to them, “Do not drink at all. But if you do drink, call me, no matter how late, and I’ll fetch you. Whatever you do, do not drive home drunk.”

In other words, I give them ranked advice. Insofar as these activities – drinking and driving, eating meat, are morally significant, my moral advice mirrors (my view of) the rightness or wrongness of these various choices.

These examples suggest that there are degrees of wrongness. It could be argued, however, in each case, there is only one right action, just as perhaps, there is only one way to be completely flat, but lots of degrees of bumpiness; just one way for a container to be completely empty, but lots of degrees of fullness. In support of this, we might note that we do say, “that’s very wrong” but it seems odd to say, “that’s very right.”

In some cases, however, particularly where the best is very demanding or heroic or saintly, and where there is a smooth continuum from the very best to the very worst, the analogy with flatness seems weak.

Consequentialists are often criticized for being too demanding in insisting that agents do the very best they can do. Some consequentialists respond by shifting to a satisficing view, according to which reaching a threshold less than the best is enough for one’s action to count as right. And everything below that counts as wrong. But if we embrace such a view, if one can satisfice (i.e., do right) by giving 10 % of one’s income to charity, surely it is better, if not required, to give even more? So, you might say that there are barely right actions, and even righter ones, just as a man of, say, 5′11 might be barely tall, and a man of 6′3″ is even taller.

If we treat height, or tallness, as the model, then we can have multiple degrees of wrong and multiple degrees of right. However, this is still a (perhaps fuzzy) line between right and wrong. At least in some cases, the notion of a line (even a fuzzy line) between right (in all its many degrees) and wrong (in all its many degrees) may seem entirely arbitrary. Some advocates of an ordinal approach will be content with a fuzzy line between right and wrong; others will perhaps push for a completely scalar notion where actions can be compared for rightness and wrongness, but no action (except perhaps the best) is right simpliciter. I see virtues and vices for each of these approaches, but shall not insist on either here. What is important for my purposes here is that it is reasonable to speak about degrees of rightness in advice – even when only one dimension is in play.

2.2 Blame and Punishment and Degrees

There are obviously degrees of blame and punishment, number of days in prison, for example, number of dollars for the fine, and so on. More serious punishments tend to be a response to more serious crimes. On a judicial level, one has to pay a steeper fine for a higher speeds or for speeding in a more dangerous areas; possession of larger amounts of drugs, or more serious drugs, generally comes with a longer prison term; on a personal level, we react with greater degrees of hostility to friends who have injured us more severely, or done so with weaker excuses.

If we link “wrong” even loosely with “punishable” or “blameworthy,” and we recognize that some actions are more blameworthy or more punishable than others, it follows that we should think that some actions are more wrong than others. I.e., that wrongness comes in degrees. The more wrong the behavior the more severe the appropriate penalty.

We could do something similar for degrees of right (or perhaps good) action. Actions that are more right call for higher degrees of praise and reward.Footnote 3

2.3 Regret, Shame, Guilt Also Come in Degrees and be Mingled with More Positive Reactions of Pride etc.

So, in the same way, insofar one of the important functional roles of the concept of wrongness is linked to regret, shame and guilt, we have another reason for treating wrongness as a matter of degree. So, these three aspects of “wrong” and “right” suggest that a degree notion of these might be quite sensible.

This may be so even if there is only a single dimension at stake. It might be difficult to imagine such a scenario, since, for all we know, it is a necessary truth that there are many different dimensions of rightness and wrongness. However, perhaps we could try to put ourselves into the mindset of an agent or an advisor or a critic who recognizes only one dimension.

I suggest that the above functional roles of right and wrong encourage such a person to consider a degree notion. The agent who does her best on the one dimension has done something perfect or entirely right. The agent who has done the second best of her options has done something slightly less right.

Peterson may object that the cases I have offered as compelling examples of a degree notion of right and wrong are intuitively appealing only because other values are at stake. For example, in thinking about Vegan Relative, it makes sense that the family member who chooses less than the option of perfect veganism, say, who eats no meat except seafood, has acted in a way that is partially right, only because she has a strong self-interested reason, in terms of pleasure or health, for eating meat. If there were no other value at stake, it would seem perverse and in fact, entirely wrong, to eat fish.

To illustrate this, consider:

Stingy Button Pusher

An agent is offered a series of buttons, numbered 1 through 10. If she pushes 1, one person will be saved; if she pushes 2, 2 people will be saved etc. up to 10. If she pushes no button, all will die. Now, suppose that it does not cost her anything to push any button. Would not it seem extremely perverse, even reprehensible, to push 9? Do we really want to say that the agent who does so is almost completely right? On the other hand, if there’s a price to the agent for pushing these buttons, a price that increases with the size of the number, say $1000 per person saved, then we might well say that the agent who pushed 9 was good, even somewhat right, but not perfectly right? Does not that confirm Peterson’s view that degrees only come in when there are several dimensions at stake? I do not think so. The self-interested reason not to push button 10 (that doing so will be very costly) does not count as a verdictive reason not to push button 10. If it were, then Peterson would have to say that the person who pushed button 10 had acted somewhat wrongly. But this seems wrong. Saving all ten in this situation is supererogatory, but not partially wrong. The cost of saving ten is an excuse for failing to save all ten, but not a moral reason not to do so.

3 Doing One’s Best is Entirely Right

The second aspect of Peterson’s position I disagree with is his view that “an act is entirely right if and only if the agent has a verdictive reason to perform it and no verdictive reason not to perform it.” (Peterson 2013, p. 33.)

Contrary to this, I believe that if an agent does the best she can do, then she has acted entirely rightly, even if she has a verdictive reason not to perform it. That is, I believe there is good reason to reserve the notion of less-than-perfect degrees of rightness for agents who fail to do their best. Now, it may be objected that this is a terminological issue and that Peterson is entitled to use the expression “entirely right” in whatever way he chooses. However, the expression does have currency in our language outside Peterson’s work, and it seems legitimate to argue that he cannot simply stipulate his preferred usage, or that we have no reason to follow him when he does so.

Let us consider some cases where different dimensions of rightness are in play.

Sometimes when these dimensions pull in different directions, the pulls are of equal or (roughly equal) strengths and we have what is sometimes called a dilemma. Some describe these as cases where everything an agent does is wrong. Peterson might describe these as cases where there are two options that are equally right and equally wrong, perhaps both equipoised between right and wrong. Let us set these sorts of cases aside and focus instead on cases where it is clear that one moral consideration is more morally important than the other. For example, imagine that you have power over the fate of two babies. You can make one, Tom, live for 50 years and the other, Tim, live for 51 years; or you can make both live for exactly two years. The first option creates more utility, the second more equality. The first is clearly the better option and hence is more right. Even if we agree with Peterson in saying that it is not entirely right, we should acknowledge that it is more right.

I claim that, in this case, the first option is entirely right, that is to say, it is the thing to do. It is what I would advise; and if it were done, it would seem inappropriate to blame the agent for choosing it, and the agent herself should feel no guilt for having done so. I am assuming here at an action cannot be obligatory if it is not entirely right. Peterson seems to agree with the latter point. In cases where there is more than one dimension pulling in different directions, he claims that one action can be more right than its alternatives without being entirely right. In such a case he says that each of the distribution options is somewhat right, one being more right than the other. This is all the moral advice the agent will get. As for rational advice, Peterson suggests that she choose by some sort of randomizing advice that gives more weight to the action that is more right. (Peterson 2013, p. 117).

Here’s a case that supports my view against Peterson’s here.

Ross’s Driver: imagine our agent, Joe, has made a promise to meet a friend for lunch. On the way to the restaurant, Joe hits a child who runs into the road. Let us suppose that Joe has not acted wrongly in hitting the child as he could not have foreseen that the child would run into the road. Nevertheless, the child is seriously injured and needs medical attention to stay alive. (We might add the Joe has no working cell phone.) Now, Joe has the choice between driving on to the restaurant, on the one hand, and staying and helping the child, on the other.Footnote 4

This is a case of conflicting duties or conflicting dimensions of rightness. Joe should keep his promise and he should help the child. It is important to the case that the one dimension is dwarfed by the other.

It seems to me completely obvious that Joe ought to stay and help the child. Peterson’s view seems to be that helping the child would be much more right than keeping the promise but not entirely right; going to the restaurant is partially right; the rational thing for Joe to do would be to toss a very weighted coin and then act on the result. This is a case where moral intuitions oppose Peterson’s view.

Who is right? Let us return to my suggestion from section I above about how the functional roles of rightness and wrongness can shed some light on disputes regarding those concepts. Let us consider advice-giving, sanctions and guilt and regret.

If Joe turned to you and said, “What should I do?” Or “What next?” Surely you would say “Help the kid.” You would not say, “Well, there’s a lot to be said in favor of helping the kid; but then again, you did make a promise” You would not say, “Toss a weighted coin and if it lands heads, go to your lunch date.” No, that would be crazy advice. What you would say is: “Help the kid!”

As for punishment, if Peterson allows any conceptual link between wrongdoing and punishment, he will have to say that the promise constituted some sort of mitigating factor for not helping the child.

Moreover, on Peterson’s view (supplemented by a link to punishment) if Joe does not go on his lunch date, but helps the child, he deserves some sort of punishment for breaking his promise. This seems wrong. If I were the friend to whom the promise had been broken, I might be annoyed while being stood up, but I would later come to withdraw my feelings of annoyance with Joe when I understood the circumstances, supposing that none of this was in any way Joe’s fault, for example, he did not cause the accident, it’s not his fault that his cell phone does not work etc. My reaction would be similar to the person who is stood up for a lunch date because the other person (entirely innocently) is knocked down by a car, and so is simply unable to keep the promise. When I am ignorant of the circumstances, I am annoyed; when I learn what happened, I am no longer annoyed. In light of full information, it seems that sanctions are inappropriate.

I do think a little regret might be appropriate in Joe for breaking his promise. Ross agrees with this. He speaks of a residue of regret left by the prima facie duty of promise-keeping. This seems to me right. It is regrettable that a prima facie moral obligation was left unfulfilled, just as it is regrettable that the friend was left looking silly and feeling disappointed at the restaurant, but I do not think the best move is to translate that regret into the claim that the agent has (in some verdictive or substantive sense) acted wrongly. Note that guilt, unlike regret, does not seem appropriate. Note also that something like regret is present in cases where some force outside of Joe’s control prevents him from making the meeting. E.g., a huge storm causes flooding that blocks his path to the restaurant, a huge traffic jam delays traffic by four hours and traps him on the road (again, with no cell phone). The fact that regret is somewhat appropriate is not reason to suppose that Joe has acted somewhat wrongly.

In case the promise involved in the last example seems too trivial, let us turn to another example.

Lifesaver sees four people drowning after a shipwreck. Three are grouped together on a rock; and another one is on a different rock. The two rocks are equidistant from Lifesaver’s starting point. Lifesaver sets out in a lifeboat, rescues the three first and then tries to reach the other one. Long before she reaches him, the fourth victim is lost.

I think that Lifesaver has acted in an exemplary way, in fact, in a perfect way.

Peterson, if I understand him correctly, will say that Lifesaver has acted not entirely rightly, but partially wrongly, because she had a moral reason to save the three, and a different weaker reason to save the one; and she was choosing between those options. She failed one of the dimensions of rightness.

If you say that someone’s action is partially wrong, you owe the agent or the listener some account of what the agent should have done. For example, a professor should be able to tell the student who earned a B- how she could have earned an A. On the moral front, at the pearly gates, St. Peter might tell me that my life was pretty good but not perfect. When I protest that I did my best, he might point that I should have yelled at my kids less; I should have done more to help the starving and so on, and that these were things I could have done. To say that someone’s behavior is unsatisfactory invites the question of how could it have been entirely satisfactory.

In the case of Lifesaver, what would Peterson say about how she could have done better? Given his views about when to randomize, I can think of only two possible answers.

  1. 1.

    You should have tossed a weighted coin to indicate the separateness of persons and then acted in accordance with the outcome of the coin toss.

  2. 2.

    You should have saved all four people.

If he says (1) I say “Fine, but suppose my agent does that and then saves three. My understanding of Peterson’s view is that this is still partially wrong because the agent has failed to save the other one. If the agent tosses the coin and saves only one, we seem to get a paradox because Peterson has already granted that there is more reason to save the three than the one. Either way the behavior is unsatisfactory by Peterson’s lights and so it has not answered the challenge to show how the agent could have been perfect.

If Peterson answers the challenge by saying (2) Agent should have saved all four victims, I say “Violation of Ought implies Can!” and hence, unacceptable.Footnote 5

To conclude this section, I have argued that to say that an agent has acted somewhat wrongly implies that there is some way in which her behavior could have been better. If, by hypothesis, the agent has done her best, there is no way her behavior could have been better. So, if an agent has done her best, or the best, then she has acted entirely rightly, at least in the standard sense of that expression.

4 Conclusion

In conclusion, while I agree with some version of Peterson’s multi-dimensionalism and while I am sympathetic to the idea of degrees of rightness and wrongness, I disagree with two of the ways he wants to combine these ideas. In particular, I have argued that one can coherently and plausibly embrace degrees of rightness and wrongness without more than one dimension of rightness and wrongness; and that an act can be entirely right even if it is wrong on some dimension.