T. M. Scanlon, portrait from Steve Pyke’s Philosophers series
Contractualist theories attempt to base morality (or some part thereof) on a hypothetical social contract arrived at through reasonable agreement. T. M. Scanlon’s contractualism identifies moral principles with principles for the general regulation of behaviour that no one can reasonably reject. The basic moral motivation is to conduct oneself in ways that can be justified to each person, insofar as they are reasonable.
One popular objection to Scanlon’s version of contractualism is that it has difficulty explaining our obligations with respect to future people, and especially that it runs into trouble with the nonidentity problem formulated by Derek Parfit.
The nonidentity problem arises cases where different courses of action will result in different people coming into existence. Suppose, for example, I recklessly dump chemical waste in a stream. As a result, Mary is born some years later, and suffers severe birth defects (though not so severe as to make her life not worth living). If I had not dumped the chemical waste, a healthy baby — Jerry — would have been born instead, and Mary would never have existed at all. Intuitively, by dumping the chemical waste, I’ve done something wrong. Most importantly, though, I’ve done something wrong to Mary, even though my actions have made her no worse off than she would otherwise be. But if she is no worse off because of my actions, it’s not clear how she can object on her own account to what I’ve done. This (allegedly) leaves contractualists at a loss to explain how I have wronged Mary at all.
Scanlon actually has a pretty compelling response at the ready, I think. In What We Owe to Each Other, he distinguishes between actual individuals to whom justification is owed, and the more abstract standpoints from which we consider whether principles could reasonably be rejected.* When I am considering whether it would be wrong to dump chemical waste in the stream, the question is not “What grounds does Mary have to reject a principle that permits me to dump this waste?” Instead, rejectability is to be assessed from the generic standpoint of “persons who live near the stream.” Although Mary would, if she existed, certainly be a person who lives near the stream, Mary’s standpoint is not identical to the generic standpoint of persons who live near the stream. Perhaps Mary does not have grounds to reject a principle that would permit me to dump the chemical waste, but it seems to me that the abstract person who lives by the stream does. This means it is wrong for me to dump the waste, and I do wrong to Mary because I’ve done something to her that I cannot justify to her — something that is disallowed by a principle that no one can reasonably reject. On the contractualist view of moral justification, “You wouldn’t exist otherwise” is not much of a justification at all; to justify myself to Mary I would have to show that no one could reasonably reject a principle that permitted me to dump the waste, and that’s a tall order.
In the second volume of On What Matters, Parfit acknowledges and offers a critique of this defence against objections based on the nonidentity problem (pages 235-236). He refers to abstract personal standpoints like “persons who live near the stream” as general persons, contrasting them with individuals. General persons are groups of possible individuals; in the example above, the general person who lives by the stream includes the possible individuals Jerry and Mary.
Parfit claims there are two reasons why Scanlonian contractualism cannot assess principles from the standpoint of general persons. First, doing so denies individuals moral standing, instead treating them as “merely parts of a general person.” This ignores the moral significance of the separateness of persons, Parfit claims, committing the same error as utilitarians for whom individuals are only morally considerable insofar as they contribute to aggregate utility. Second, Parfit thinks that the using the standpoints of general persons is inconsistent with Scanlon’s purpose of providing an account of “the particular form of concern that we owe to other individuals.” General persons are not individuals, so assessing principles from the standpoints of general persons does not tell us what we owe to other individuals.
Both of these arguments are invalid. When assessing principles, we are to consider them from generic standpoints, but we are using these generic standpoints to find out what we owe to individuals. General persons don’t have moral status, on Scanlon’s view. The reasons attached to these standpoints are only of interest insofar as they help actual individuals find principles they can justify to one another. Because Scanlon’s contractualism grants only individuals and not general persons moral status, the complaint that contractualism fails to respect the separateness of persons misses the mark. And because the only reason why we care about general persons because they’re a useful tool for understanding “the particular form of concern that we owe to other individuals”, the argument that focusing on the rejectability of principles from the standpoint of general persons doesn’t tell us about what we owe to other individuals is incomplete.
Parfit’s confusion on this point parallels the way he (and other critics of Scanlon’s contractualism) conflate the forms of reasoning that contractualism employs with the content of the principles that contractualist reasoning generates. Some critics complain that by ruling out the aggregative reasoning that counterintuitively justifies imposing severe burdens on a small number of people in exchange for tiny benefits to some much larger number of people, contractualism also rules out benign cases of aggregative reasoning (as in decisions where we must choose between saving a larger or smaller number of people from some serious harm). As Scanlon points out, however, contractualism rules out aggregative reasoning, not aggregative principles.** Aggregative principles are only ruled out if such principles can only be arrived at by aggregative reasoning. But contractualists argue (convincingly, I think) that some aggregative principles are principles that no one can reasonably reject.
Similarly, Parfit conflates contractualist reasoning about general persons with the object of that reasoning. I consider principles from the standpoints of general persons in order to be able to justify myself to actual individuals, not so I can justify myself to general persons. Critics who want to press objections based on the nonidentity problem need to show that consideration of generic reasons is of no use in justifying ourselves to actual individuals. But this is a core commitment of Scanlon’s contractualism. If its critics can show that such a core commitment is false, objections based on the nonidentity problem would be superfluous.
*From What We Owe to Each Other, page 202: “According to contractualism, our concern with right and wrong is based on a concern that our actions be justifiable to others on grounds that they could not reasonably reject insofar as they share this concern. ‘Others’ figure twice in this schema: as those to whom justification is owed, and as those who might or might not be able reasonably to reject certain principles. When we think of those to whom justification is owed, we naturally think first of the specific individuals who are affected by specific actions. But when we are deciding whether a given principle is one that could reasonably be rejected we must take a broader and more abstract perspective.”
And from page 204: “As this discussion of the points of view that must be considered in deciding whether a principle could reasonably be rejected brings out, an assessment of the rejectability of a principle must take into account the consequences of its acceptance in general, not merely in a particular case that we may be concerned with. Since we cannot know, when we are making this assessment, which particular individuals will be affected by it in which ways (who will be affected as an agent required to act a certain way, who as a potential victim, who as a bystander, and so on), our assessment cannot be based on the particular aims, preferences, and other characteristics of specific individuals. We must rely instead on commonly available information about what people have reason to want. I will refer to this as information about generic reasons.”
And from “How I am not a Kantian”, in On What Matters, Volume II, page 131, footnote 86): “Parfit and I may take different views about the correct characterization of the ‘individuals’ whose reasons are to be considered. Although he does not say so explicitly, some of what he does say suggests that he has in mind actual persons affected by the action, or by the acceptance of the principle. In my case what we consider are not the reasons of actual persons but the ‘generic’ reasons that someone would have in virtue of occupying a certain role in regard to the principle in question, such as being the person who has relied on the assurance of others, or a person in need of help, or a person called upon to give it.”
**”Many people may be drawn to consequentialism because they see that there are some situations in which it the morally correct way to decide what to do is to figure out what would produce the best consequences overall. Decisions by public officials about what kind of hospitals to build may be a good example. Because producing the best consequences seems so obviously to be the right standard in these cases, people then infer that this idea is always morally basic. This seems to me to be a mistake: producing the best consequences might be the correct standard in these cases not because it is the basis of morality but because it is what is owed to people in situations of that kind, by agents who stand in a certain relation to them.” (from “How I am not a Kantian”, in On What Matters, Volume II, page 139)