Moral constructivism is a metaethical view about the nature of moral truth and moral facts (and properties), so called because the intuitive idea behind the view is that such truths and facts are human constructs rather than objects of discovery. More precisely, constructivism involves both a semantic thesis about moral sentences and a two-part metaphysical thesis about the existence and nature of moral facts and properties. According to the semantic thesis, ordinary moral sentences purport to be fact-stating sentences and thus purport to be genuinely true or false. And, according to the metaphysical thesis, there are moral facts whose existence and nature are in some sense dependent upon human attitudes, agreements, conventions, and the like. Thus, constructivism represents a metaethical view in partial agreement with versions of moral realism. Like the realist, the constructivist is a so-called cognitivist (descriptivist)—moral sentences have descriptive content and thus purport to be genuinely fact stating. Again, like the realist, the constructivist accepts the view that there are moral facts that serve as the truth makers of true moral sentences. But unlike the realist, the constructivist rejects the idea that there are moral facts (and properties) that are independent of human attitudes, conventions, and the like.
It is useful to distinguish between simple and sophisticated versions of constructivism as well as between nonrelativist and relativist versions. Simple versions of constructivism are represented by certain views that would construe moral truth in terms of the actual attitudes of individuals or actual agreements within cultures about matters of moral concern. More sophisticated versions of constructivism construe moral truths (and associated moral facts and properties) in terms of the hypothetical attitudes of individuals or perhaps hypothetical agreements among members of a group reached under suitably constrained circumstances. Nonrelativist versions of constructivism maintain that all individuals and groups whose attitudes, agreements, and so forth provide the basis for moral truths and facts do or would accept the same set of basic moral norms with the result that there is a single set of moral truths and facts. Usually, such views are wedded to some version or other of sophisticated constructivism.
Thus, a version of the ideal-observer view of moral truth—according to which basic moral truths are represented by the moral norms that would be accepted by an ideal observer, where the notion of an ideal observer is so characterized that all ideal observers will agree on the same set of basic moral norms—is a version of sophisticated nonrelativist constructivism. Relativist versions of constructivism allow that there may be more than one individual or group with differing attitudes and agreements that serve as the basis for different (and conflicting) sets of basic moral norms. Versions of moral relativism, according to which moral truths and facts are a matter of what basic moral norms a culture in fact accepts, represent versions of simple, relativistic constructivism; versions of relativism, according to which moral truths and facts are a matter of what would be accepted under conditions that are ideal for choosing such norms, represent sophisticated relativistic versions of constructivism. Versions of the ideal-observer view are relativistic if they allow that there can be ideal observers who would accept different (and conflicting) sets of moral norms. So-called Kantian constructivism of the sort elaborated and defended by John Rawls, which appeals to choices made by hypothetical individuals behind a veil of ignorance (a version of contractarianism), is yet another sophisticated and apparently nonrelativistic constructivist view.
Constructivism, at least in its sophisticated versions, is supposed to capture what is plausible about moral realism, leaving behind what is problematic about realist views. Thus, constructivism can accommodate quite well certain "objective pretensions" of commonsense moral thinking. Some of these pretensions have to do with the form and content of moral discourse. A good many moral sentences are in the declarative mood (e.g., "Abortion, except in cases of rape and incest, is wrong") and are thus naturally interpreted as genuinely fact-stating sentences. Moreover, some such sentences appear to make references to (putative) moral facts and properties (e.g., "The evil of American slavery was partly responsible for its demise as an institution"). Other objective pretensions have to do with such activities as moral deliberation, debate, and argument. These critical practices are seemingly aimed at arriving nonarbitrarily at true or correct moral views, ones that would ideally resolve intrapersonal and interpersonal conflict and uncertainty about moral issues. Like realism, constructivism is attractive in apparently being able to accommodate such objective pretensions of ordinary moral discourse. Moreover, it attempts to accommodate these features without endorsing the sorts of metaphysical commitments to independently existing moral properties and facts countenanced by the realist. In short, at least certain versions of constructivism boast a robust notion of moral objectivity without problematic metaphysical commitments.
One serious challenge to constructivism is represented by the argument from moral error. According to constructivism, moral truths and associated facts are to be understood in terms of the attitudes and agreements of individuals and groups. However, if we take ordinary moral discourse and argument seriously, then since such discourse and argument presuppose that there are right answers to moral questions whose correctness outstrips any actual or even ideal set of attitudes or agreements, the constructivist view cannot be correct. To understand this objection more clearly, it will be useful to distinguish between basic and nonbasic moral truths and facts. Basic moral truths and facts are of a quite general sort, properly expressed by moral principles, and are the direct objects of choice by those under ideal conditions of moral thought and deliberation. Nonbasic moral truths and facts are those truths and facts that, in some sense, follow from the basic ones together with nonmoral information.
Now the constructivist can allow for certain sorts of errors in moral judgment. For instance, simple moral relativism can allow that individuals and groups can be mistaken about particular moral judgments owing to misinformation about particular cases or perhaps to faulty reasoning from basic moral principles to concrete cases. This kind of moral relativism, however, cannot allow for error at the level of actual agreements, since such agreements constitute basic moral truths. The sophisticated constructivist can allow for error at the level of communal agreement, since it is possible on such views that the actual agreements of actual groups are at odds with those hypothetical choices constitutive of moral truth on this sort of view. However, the sophisticated constructivist cannot allow for error at the level of choice made under ideal conditions—call this "deep moral error." After all, the constructivist construes such choice as constitutive and not just evidence of basic moral truths and facts. But, so the objection goes, given our critical practices, we can sensibly raise questions about the truth of those moral principles and norms that are chosen under ideal circumstances. This indicates that moral truth is one thing and the norms and principles chosen even under the most ideal of circumstances is another. Hence, constructivism, in both its simple and sophisticated versions, is not acceptable.
In response, the constructivist can perhaps block the argument from moral error in the following way. First, the constructivist can note that it is dubious that our critical practices presuppose that deep moral error—error at the level of choice under ideal conditions—is possible. After all, our commonsense critical practices are not finely tuned to subtle differences in metaethical positions, and, in particular, common sense does not (so the constructivist might plead) make any distinction between the sort of realist objectivity that presupposes the possibility of deep moral error and a kind of constructivist objectivity that denies this possibility. Can we, for instance, really make sense of the idea that we might be mistaken about such basic moral principles as one that prohibits torture for fun? Furthermore, the constructivist can question the basic move featured in the argument from moral error—that is, the move from (1) it is quite sensible to raise questions about choices that purport to be made under ideal conditions to (2) an explanation of this phenomenon requires moral realism. Granted, the supposed gap between the truth of moral principles on the one hand and choice of such principles under ideal conditions on the other is one way to explain how we can sensibly raise questions about the truth of moral judgments made under ideal conditions, but this is not the only way to make sense of such critical stances.
The constructivist can note that in the context of everyday discussion where we have to judge whether or not to accept the moral judgments of others, one can sensibly raise questions about some judgment by raising questions about the judger herself. After all, whatever is involved according to the constructivist in being ideally well situated for choosing basic moral principles, it is not likely to involve features of the judger and her situation that are easy to detect. For example, part of being ideally well situated would seem to require having all sorts of factual information, being free from certain forms of bias, and properly weighing the interests of parties affected by the choice of principles. But it is difficult to determine that someone has satisfied these and other relevant desiderata for being well situated. So, even if it is not possible for someone who really is well situated to be mistaken in moral judgment, it is possible for critics who acknowledge that such error is not possible to raise sensible questions about the truth of a person's moral judgment. Hence, although the constructivist cannot allow for the possibility of deep moral error, she can plausibly argue that our commonsense critical practices do not presuppose that deep moral error is possible. Moreover, she can go on to accommodate the idea that it makes sense to criticize those who are ideally situated. The constructivist, it would appear, can plausibly respond to the argument from moral error.
Brink, D. O. Moral Realism and the Foundations of Ethics. Cambridge, U.K.: Cambridge University Press, 1989. In chapter 2 Brink uses the argument from moral error against constructivism. Appendix 4 is a critical discussion of Rawlsian constructivism.
Firth, R. "Ethical Absolutism and the Ideal Observer." Philosophy and Phenomenological Research 12 (1952): 317–345. A classic statement of the ideal-observer version of constructivism.
Milo, R. "Skepticism and Moral Justification." Monist 76 (1993): 379–393. Milo defends a contractarian version of constructivism.
O'Neill, O. "Constructivisms in Ethics." In Constructions of Reason. Cambridge, U.K.: Cambridge University Press, 1990. O'Neill criticizes Rawls's version of constructivism and sketches what she takes to be a more plausible version of the view inspired by Immanuel Kant's writings.
Rawls, J. "Kantian Constructivism in Moral Theory." Journal of Philosophy 77 (1980): 515–572. An elaboration of a constructivist view that centrally involves a Kantian conception of persons.
Timmons, M. "Irrealism and Error in Ethics." Philosophia 22 (1993): 373–406. A critical discussion of the argument from moral error.
Wong, D. Moral Relativity. Berkeley: University of California Press, 1984. Wong both criticizes various versions of moral relativism and defends his own version.
Mark Timmons (1996)