¹ By “self-interest” I mean to include, roughly, one's welfare, what is good for one, what is beneficial to one insofar as one is concerned with one's own well-being. “Self-regard” has broader uses. For example, a self-regarding duty may be simply to treat oneself in a certain way (e.g., protecting one's health) even if the justification has nothing to do with one's personal interest in one's own welfare but stems instead, say, from one's promise to a spouse. To act out of “self-regard,” however, seems properly understood as the same as acting out of self-interest. My understanding here, I hope, simply follows ordinary usage.

² According to the coherence-and-efficiency model, one can also appeal to empirical facts about our other-regarding desires, because this model does not assume that the ends one wants to fulfill are all benefits for oneself.

³ These tasks are obviously too much to take on here. Defending what I call “the common-sense view” would by itself be a major philosophical project, and establishing its credentials as very widely held among many otherwise diverse peoples would be a very challenging sociological /historical task. Neither is necessary for my purposes, though admittedly my conclusions will be of less interest to those who deny the plausibility or the breadth of appeal of the view I attribute to common sense. The main point is that this view, sketched here in familiar, nontechnical language, seems to have more affinity with the Kantian model, despite the austere Kantian terminology, than with the self-interest model, the coherenceand-efficiency model, or the consequentialist model. Readers, I expect, will find this claim more striking and important the more they agree with my conjecture that what I call “the common-sense view” is plausible and widely held; but the extent to which they agree on this is a matter for each to judge.

⁴ In my discussion I will deliberately deviate from the philosophical practice of distinguishing “rational” from “reasonable,” e.g., as explicitly done in Rawls, John, Political Liberalism (New York: Columbia University Press, 1993), pp. 48–54.Google Scholar Ordinary language does not honor a sharp distinction here, though there are subtle differences in the various “reason” expressions we use. The Kantian view of reason, I shall suggest, is closer to a common-sense idea of the reasonable than the thinner and more technical idea of the rational in contemporary philosophy, but I continue to use “rational” in a broader sense (including the reasonable) as a reminder that (as I believe) philosophers have kidnapped that ordinary term for their own purposes. This was a well-meaning offense, no doubt, but it is a constant source of confusion at the intersection of technical and nontechnical discourse.

⁵ This idea fits well with Rudiger Bittner's suggestion that prudential reason does not “demand” but merely “advises.” See his What Reason Demands, trans. Talbot, Theodore (New York: Cambridge University Press, 1989), esp. chs. 6 and 7.Google Scholar The idea is, I believe, congenial with the position developed by Slote, Michael in Beyond Optimizing: A Study of Rational Choice (Cambridge, MA: Harvard University Press, 1989).CrossRefGoogle Scholar

⁶ I pass over several matters of detail here, partly because I suspect that they may not be settled features of the overlapping views I call “common sense.” For example, is it ever reasonable to sacrifice one's sight, or life, to complete some personal project that one has come to treasure above all, assuming (artificially) that others' interests are not involved? For our purposes, it is enough to say that common sense holds that reasonable people place a very high priority on life and basic human capacities. Questions can also arise, of course, about what the basic capacities are and what degree of maintenance is rationally required. That one should try to preserve one's sight and limbs is uncontroversial, but does reason demand that we train our bodies to the peaks of athletic fitness? It is generally conceded that to “fry one's brains” on drugs is foolish, regardless of the pleasures one might gain; but to what extent does reason demand that we develop our intellectual acuity, memory, etc.? More-theoretical issues that I pass over here concern how the “unreasonableness” of self-destructive behavior is to be explained. Some hold that it is inherently irrational; but one might argue that it merely reflects means-ends incoherence or cognitive confusion, given that the human desire to survive and thrive as human is in fact deep and virtually universal in us.

⁷ Here I am also supposing that common sense differs from the position of Richard Brandt and others who maintain, roughly, that the rational thing to do is to efficiently pursue maximum satisfaction of the desires one would have if fully informed, subjected to “cognitive psychotherapy,” etc. See Brandt, Richard B., A Theory of the Good and the Right (Oxford: Clarendon Press, 1979), esp. pp. 110–29;Google Scholar and Brandt, , “The Concept of Rational Action,” Social Theory and Practice, vol. 9 (1983), pp. 143–64.CrossRefGoogle Scholar For criticism of theories of this type, see, for example, Gibbard, Allan, “A Noncognitivist Analysis of Rationality in Action,” Social Theory and Practice, vol. 9 (1983), pp. 199–221;CrossRefGoogle ScholarLoeb, Don, “Full Information Theories of the Good,” Social Theory and Practice, vol. 21, (1995);CrossRefGoogle Scholar and Rosati, Connie S., “Persons, Perspectives, and Full Information Accounts of the Good,” Ethics, vol. 105, no. 2 (1995), pp. 296–325.CrossRefGoogle Scholar Some desires and preferences that common sense recognizes as quite foolish or crazy may be revealed as such by exposure to more information in cognitive psychotherapy, but it seems doubtful that common sense is committed to the general thesis that choices based on desires that would be extinguished under such a process are thereby irrational. (Brandt, I should note, introduces his definitions as stipulative and makes no claims that they conform to common sense.)

⁸ I explain and try, informally, to make this view persuasive in my Autonomy and Self-Respect (Cambridge: Cambridge University Press, 1991), ch. 12.

⁹ The same point, I believe, applies to the hypothetical inclinations that we would have with full information and cognitive psychotherapy, but this is more controversial. In any case, this more sophisticated controversy–whether members of a special class of hypothetical filtered, informed desires always give us reasons–involves technical complications about which we should not expect “common sense” to have an opinion.

¹⁰ Another possible example is the choice of a less expensive brand of food, judged equally as tasty and nourishing as another brand which has a more attractive package. If the cost difference “means” relatively little to one, it seems not unreasonable to choose the slightly more expensive brand even though one realizes that, on reflection, the pretty package “means” even less and thus one has slightly better reason to choose the less expensive brand. This point that it is not always unreasonable not to act on the best reasons has been more thoroughly discussed by others and is not a major thesis on which I want to insist. I note, however, that common sense's apparent adoption of this idea may be subject to several different explanations. One may be that, in the case in question, the agent, on reflection, endorses the end of having the pretty package now, thus making it his or her “preference,” despite realizing that he or she would not endorse this ranking as a general policy. That current preference, as a reflective endorsement, is itself a good reason for acting, absent compelling counter-reasons, and thus tips the scale of reasons in favor of now buying the more expensive and attractive package. We can say that the opposite choice also would not have been unreasonable, for, had the agent made it, he or she would no doubt have endorsed the end of saving a little money now over having the pretty package, and thus again would have been acting on the stronger reasons.

¹¹ I am referring here to the content of the ends that a person may reasonably be expected to have, as opposed to (further) constraints on the adoption of ends, which I count as procedural, having to do, for example, with ascertaining that they are attainable, compatible with one's other ends, internally consistent, etc.

¹² These controversial cases are treated more fully in my discussions of hypothetical imperatives, in Dignity and Practical Reason in Kant's Moral Theory (Ithaca, NY: Cornell University Press, 1992), chs. 1 and 7.

¹³ Most people, no doubt, think that it matters how many others are as well-situated to help, whether the victims are “remote” in distance and affiliation, etc. I am not supposing, then, that most people accept Peter Singer's demanding standards for sharing resources, but only that most would grant that, other things equal, one has “some reason” to help if one can help effectively and at very little cost. (See Singer, Peter, Practical Ethics, 2d ed. [Cambridge: Cambridge University Press, 1993], ch. 8, esp. pp. 229–46Google Scholar).

¹⁴ We should note, however, that one's own preference, or reflectively endorsed inclination, not to do a particular favor for someone is also a reason that one may often take into account. When I really prefer not to do a favor, then, I am not ignoring the preferences of others “for no reason.” This is not to say that my preference not to do favors is always, or even generally, a sufficient reason to make it reasonable to refuse the favor; but it is a reason, often relevant.

¹⁵ I should note that I am thinking of cases where we only consider that someone has a personal preference, setting aside special relationships. For example, if I am legally appointed to be impartial executor of funds for underage heirs, then I might be explicitly obliged to satisfy more of their permissible preferences, other things equal, rather than fewer, in fulfilling my task; but that is a very special context. I might also note that the common-sense permissibility of not acting on the “best” other-regarding reasons stemming from others' preferences might be explained in terms of an overriding self-regarding permission to choose for oneself, without counting the numbers, where to invest one's “charity” efforts (in promoting the preferred ends and activities of others). But these matters go beyond my main concerns here.

¹⁶ I leave aside, for example, the familiar view that the fact that an act would cause an animal extreme pain is a reason for anyone to refrain from it, whether they “prefer” to or not. See note 55.

¹⁷ See Slote, Michael, From Morality to Virtue (New York and Oxford: Oxford University Press, 1992).Google Scholar Slote gives an extensive description of “common-sense morality” as he sees it, noting where it diverges from utilitarianism (pp. 31–84). He argues that common-sense morality is incoherent in various ways, but that common-sense views on virtue are more plausible. Regarding virtue, he suggests, common sense seems to give roughly equal regard to self and to all others:

I am saying that our ordinary thinking about the virtues treats the category of trait-possessor and the category of “other people” (i.e. people other than the trait possessor) as of roughly equal importance, and this latter suggests (perhaps it does more than suggest, but I don't at this point want to claim any more than it suggests) an ideal of character and action that, for any given choice of agent/possessor, exemplifies roughly equal concern for the agent/possessor and for others treated as a class or category to which everyone other than the agent/possessor belongs, (p. 98)

¹⁸ It is common now for philosophers to insist that the “shoulds” and “reasons” requiring concern for others be labeled “moral” in contrast to the “nonmoral” or “rational” “shoulds” and “reasons” pertaining to self-interest and efficient pursuit of one's ends; but common sense, I think, does not draw such a sharp line, and to insist on it here would beg questions that are at issue (e.g., what makes reasons “moral”?).

¹⁹ The view I sketch here is akin to what is often called “ethical egoism,” i.e., the view that what one morally ought to choose is just what maximizes one's self-interest. (Another version says that it is always morally permissible to do what maximizes one's self-interest.) By contrast, the view I sketch is about what it is reasonable to choose. The latter view, I suspect, is more common, but some philosophers apparently hold both the view I sketch and some form of ethical egoism. Uncontroversial historical examples are hard to find, but surely among the best candidates to illustrate the self-interest model of reasonable choice would include Epicurus, Thomas Hobbes, and Henry Sidgwick. See Long, A. A. and Sedley, D. N., The Hellenistic Philosophers (Cambridge: Cambridge University Press, 1987), pp. 102–57;Google ScholarHobbes, Thomas, Leviathan, ed. Macpherson, C. B. (Baltimore: Penguin Books, 1968), part I, esp. pp. 110–222;Google ScholarSidgwick, Henry, The Methods of Ethics, 7th ed. (Chicago: University of Chicago Press, 1962), pp. 119–95.CrossRefGoogle Scholar See also Derek Parfit, Reasons and Persons, esp. chs. 1 and 6.

²⁰ The latter is proposed in Kavka, Gregory, Hobbesian Moral and Political Theory (Princeton: Princeton University Press, 1986), p. 42.Google Scholar

²¹ Another variation worth noting is a view that while the “objectively rational thing to do” is what best promotes self-interest, because self-interest is best served by not pursuing it directly it is not rational for us, most of the time, actually to engage in deliberation about what will best promote self-interest. The view is a response to the “paradox of egoism,” the alleged fact that it is not most in one's self-interest deliberately to pursue what is most in one's self-interest.

²² See, for example, Butler, Joseph, Five Sermons, ed. Darwall, Stephen (Indianapolis: Hackett Publishing Co., 1983).Google Scholar For a rather different strategy of argument, see Olson, Robert G., The Morality of Self-interest (New York and Chicago: Harcourt, Brace, and World, Inc., 1965).Google Scholar

²³ Even the indirect self-interest theory mentioned above (in note 21) has implausible consequences, though less blatant ones. On that view, anyone who deliberately pursued the greater good of others at a slight sacrifice of his or her own good would be doing something “objectively irrational”; and any pursuit of the good of others as an end at all would be irrational unless grounded somehow (e.g., by prior policy decisions) in one's aim for one's own best self-interest. Although the theory is consistent, it involves the bizarre claim that people who deliberate and act well are in most cases doing so in order to pursue ends quite different from the egoistic end that, according to the theory, really justifies their acting as they do. See Stocker, Michael, “The Schizophrenia of Modern Moral Theories,” Journal of Philosophy, vol. 73 (August 12, 1976), pp. 453–66.CrossRefGoogle Scholar

²⁴ Psychological egoism comes in many forms, but for now it may suffice to characterize it as the view that it is an unalterable law of human nature that every human being always aims ultimately for what he or she believes is in his/her best self-interest, and for nothing else. A somewhat weaker version, i.e., “each person is so constituted that he will look out only for his own interests,” is well discussed in James Rachels, The Elements of Moral Philosophy (New York: Random House, 1986), pp. 53–64.

²⁵ See Butler, Five Sermons; Hume, David, Enquiries Concerning Human Understanding and Concerning the Principles of Morals, 3d ed., ed. Selby-Bigge, L. A. and Nidditch, P. H. (Oxford: Clarendon Press, 1978), pp. 212–32;Google ScholarFeinberg, Joel, “Psychological Egoism,” in Feinberg, Reason and Responsibility, 4th ed. (Encino, CA: Dickenson Publishing Co., 1978)Google Scholar, reprinted in Moral Philosophy, ed. George Sher (San Diego, New York, and Chicago: Harcourt Brace Jovanovich, 1987), pp. 1–15.

²⁶ When I to “virtually cost-free pleasures,” I do not mean to deny that there may always be some costs in opportunities, time, effort, etc. What I have in mind (more strictly) are significant pleasures (not very-minor or barely discernable ones) that can be obtained at very small, insignificant costs so that, on balance, the costs are negligible for practical purposes. It is regrettable, I think, but not contrary to demands of practical reason that people often let the quality of their lives be somewhat diminished by minor but persistent habits of self-denial of pleasures of this sort. Perhaps fuller understanding of why they do this would help to free them from such habits, but the appropriate response, I suggest, is not a charge of irrationality of choice but sympathetic encouragement to “lighten up.”

²⁷ A principle of efficiency, as I understand this, would direct us, given fixed goals, to take means that are least costly in terms of resources, or, given fixed resources, would direct us toward a higher degree of satisfaction of our ends. If all our values are commensurable and expressible in terms of some common denominator, then it might be possible (and many would say desirable) to dispense with “coherence” and reduce rational choice to efficiency, seen as choosing so as to bring about the greatest possible amount of that value (which might be, for example, “preference-satisfaction”). The model I call “coherence and efficiency” does not assume that all values are commensurable and thus introduces standards of informed reflection and coherence among means and ends, distinct from and beyond efficiency. For example, if certain means are necessary to an end, one must choose the means or else give up the end; to hold on to an end while refusing to take the necessary steps to achieve it is a form of practical incoherence. (We might be incapable of such incoherence if we were immune to self-deception.) Similarly, it is generally a mark of incoherent (though possible) practical thinking to pursue goals that undermine one's other goals or to employ means that violate the values that were the basis for choosing one's goals. Finding himself in such a situation, a reasonable person will make some adjustment in his set of goals to make them more “coherent,” even if there is no quantitative measure of value to indicate which ends among his incoherent set of ends should be revised or abandoned.

²⁸ This point, incidentally, goes beyond Hume, as I understand him. For Hume, acts, choices, and preferences cannot be “contrary to reason,” except in the derivative sense mentioned. If they involve wasting resources, taking unnecessary risks of failure, etc., they are foolish and imprudent but not, strictly speaking, irrational. But the coherence-and-efficiency model treats such inappropriate selection of means to fixed ends as a paradigm of practical irrationality.

²⁹ Rawls, John, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971), pp. 407–16.Google Scholar

³⁰ I discuss this principle at some length in Dignity and Practical Reason in Kant's Moral Theory, chs. 1 and 7.

³¹ According to the coherence-and-efficiency standard, it is not in itself irrational, for instance, to risk one's life for trifles, but if one has that attitude, it is unwise to invest much in long-term projects.

³² Hume, David, A Treatise on Human Nature, ed. Selby-Bigge, L. A. (1888; reprint, Oxford: Clarendon Press, 1995), p. 416.Google Scholar

³³ There are many important distinctions that have emerged from the vast literature on utilitarianism and consequentialism, but I hope no harm is done by my oversimplifying here. For other purposes, for example, one might note that “consequences,” as usually understood, are not all that counts in many versions, e.g., G. E. Moore's; for acts themselves, as distinct from their consequences, could have intrinsic value. See Moore, G. E., Principia Ethica (Cambridge: Cambridge University Press, 1902), pp. 1–21.Google Scholar The many subtle variations on “rule-utilitarianism,” as well as “motive-utilitarianism,” are also highly significant in many discussions, but not, I think, here.

³⁴ Some argue that encouraging people to use the utilitarian principle as a deliberative guide would result in less than the most possible intrinsic value; and for this reason they recommend that, for the most part, we (or common folk) rely on familiar specific moral rules as our deliberative guides. See Hare, R. M., Moral Thinking (Oxford: Clarendon Press, 1981), pp. 44–64.CrossRefGoogle Scholar

³⁵ Here I am thinking that “the most intrinsic value, on balance” is the result when negative intrinsic values are “subtracted” from positive ones. (That we can really perform such calculations is, of course, a fiction.)

³⁶ We need, in addition to empirical premises, also some account of intrinsic value, such as Mill's, which allows us to identify such value as happiness, or high-quality happiness, or Aristotelian thriving, or something similar. See Mill, J. S., Utilitarianism (Indianapolis: Hackett Publishing Co., 1988), pp. 7–11.Google Scholar

³⁷ See, for example, Brandt, Richard B., “Toward a Credible Form of Utilitarianism,” in Morality and the Language of Conduct, ed. Castaneda, Hector-Neri and Nakhnikian, George (Detroit: Wayne State University Press, 1965), pp. 107–43;Google Scholar and Brandt, A Theory of the Good and the Right (supra note 7), pp. 163–99. For the distinction between “act” and “rule “ varieties of utilitarianism and consequentialism, see note 41 below.

³⁸ Many introductory texts and anthologies on ethics, as well as professional books and articles, offer such “counterexamples” to utilitarianism together with discussions of rule-utilitarian, or other consequentialist, devices to circumvent these problems. See, for example, Sir Ross, David, The Right and the Good (Oxford: Clarendon Press, 1930), pp. 16–47;Google ScholarRachels, James, The Elements of Moral Philosophy (New York: Random House, 1986), pp. 90–113;Google ScholarFrankena, William K., Ethics (Englewood Cliffs, NJ: Prentice Hall, 1973), pp. 34–43;Google ScholarDonagan, Alan, The Theory of Morality (Chicago: University of Chicago Press, 1977), pp. 172–99;Google ScholarDonagan, , “Is There a Credible Form of Utilitarianism?” in Contemporary Utilitarianism, ed. Bayles, Michael D. (New York: Doubleday Anchor Books, 1968);Google ScholarSmart, J. J. C. and Williams, Bernard, Utilitarianism: For and Against (Cambridge: Cambridge University Press, 1973);CrossRefGoogle Scholar and Lyons, David, Forms and Limits of Utilitarianism (Oxford: Clarendon Press, 1965).CrossRefGoogle Scholar

³⁹ On this point, the self-interest standard and the coherence-and-efficiency standard seem quite correct; the Kantian disagreement with them lies elsewhere, as should become evident later.

⁴⁰ The need for the qualification “inappropriately” here should be evident; for some attention to our actual circumstances, including our particular loves and hates, is often relevant to the reasonable application of general moral principles. What is needed now is more constructive effort to work out standards of appropriateness of particular concerns in various contexts and less rhetoric about the evils of impartialism (or the opposite).

⁴¹ “Act-consequentialism” generally refers to any theory of normative ethics that affirms, perhaps with minor modifications, that one acts in the right (or best) way if and only if one does what will (or probably will) have the best results, compared to one's options, in the long run, considering all persons (or all sentient beings). “Rule-consequentialism” generally refers to any version of normative ethics that affirms, perhaps with minor modifications, that one does what is right (or best), among one's options, if one acts as directed by an ideal (or actual) moral code, or set of rules, such that general acceptance of that code (which is not the same as perfect conformity to it) by most people in the relevant community would have the best results in the long run, considering all persons (or sentient beings). Consequential-ists differ among themselves as to what makes results “good” and whether there need always be a quantifiable “best.” “Utilitarians” are often considered consequentialists who reduce “good results” to pleasure and pain, or at least benefits and setbacks to the welfare of individuals (but this usage does not fit G. E. Moore's “ideal utilitarianism”). See the citations in note 38.

⁴² Endorsing this idea, a colleague of mine says that we should be utilitarians but think like Kantians.

⁴³ Strategies for showing the rationality of adopting rule-utilitarianism are discussed in Hare, Moral Thinking, ch. 12, and Brandt, A Theory of the Good and the Right, chs. XI and XVII. Advocates of either of our first two models of rational deliberation would need to argue from empirical evidence that it is, derivatively, rational to adopt rule-consequentialism as a moral philosophy. Also, I suspect confusion about “intrinsic value” sometimes gives a false appearance of supporting the thesis that trying to maximize intrinsic value, directly or by rules, is rationally necessary. If one defines “intrinsically valuable” as “having properties that provide good reasons for anyone to favor the thing,” or something like this, then the move to justify maximizing intrinsic value could at least get started. But traditionally, from classic utilitarians to Moore and his successors, this has not been what is meant. Bentham, for example, identified intrinsic value with certain sensations, and Moore treated it as a metaphysical, simple, nonnatural, supervenient quality. In both cases, intrinsic value is conceptually independent of what we have reason to choose.

⁴⁴ My aim here is merely to reconstruct and summarize a Kantian position rather than to show how it is drawn from Kant's texts; but for purposes of comparing my sketch with Kant's writings, the main texts to consider are the following: Kant, Immanuel, Groundwork of the Meta-physic of Morals, trans. Paton, H. J. (New York: Harper and Row, 1964);Google ScholarCritique of Practical Reason, trans. Beck, Lewis White (New York: Macmillan Library of Liberal Arts, 1985);Google ScholarReligion within the Limits of Reason Alone, trans. Greene, T. M. and Hudson, H. H. (New York: Harper and Row, 1990);Google Scholar and The Metaphysics of Morals, trans. Gregor, Mary (Cambridge: Cambridge University Press, 1991).Google Scholar Page numbers in brackets following citations of Kant's texts refer to the standard Prussian Academy edition numbers, included in most translated editions.

⁴⁵ This is not to suggest the silly view that we actually go through all these steps consciously every time we make significant choices. The elements are meant to reflect aspects of what is involved, often as background beliefs and commitments, when, as we ordinarily say, “his reason for doing that was …,” and the like.

⁴⁶ Other instrumental principles might be, for example, those that Rawls (in A Theory of Justice, pp. 407–16) calls “counting principles.”

⁴⁷ I develop this view of the Hypothetical Imperative in Dignity and Practical Reason in Kant's Moral Theory, chs. 2 and 7.

⁴⁸ The distinction between “substantive” and “nonsubstantive” is perhaps relative to context, just as the line between “specific” and “nonspecific” is. Thus, I am not claiming that the Kantian basic norms of rational deliberation are so “formal” as to have no action-guiding or deliberation-constraining content at all, but am only trying to contrast them with more specific and controversial substantive principles such as “Maximize your wealth and power,” “Follow God's commands, “ “Choose the most pleasant life,” and “Promote the greatest happiness of the greatest number.”

⁴⁹ For example, moral attitude is one that values humanity in each person and con-strains an agent's choices by principles determined by trying to deliberate according to the procedural standards of the Kantian legislative perspective, to be sketched below.

⁵⁰ See, for example, Moore, Principia Ethica, pp. 1–21; Frankena, William K., “The Natu ralistic Fallacy,” Mind, vol. 48 (1939), pp. 464–77;CrossRefGoogle Scholar and Hare, R. M., The Language of Morals (Oxford: Clarendon Press, 1952), pp. 79–93.Google Scholar

⁵¹ See Perry, Ralph Barton, Realms of Value (Cambridge: Harvard University Press, 1954), pp. 3, 107, 109;CrossRefGoogle Scholar and Brandt, A Theory of the Good and the Right (supra note 7), ch. 6.

⁵² By “fundamentally valuable” here I mean only moral and nonmoral values as defined or characterized in the most general terms, not the more specific values that we might draw from such general characterizations in the light of empirical facts.

That persons have dignity for Kant was morally fundamental; that adultery is wrong was derivative. Similarly, that each person's “happiness” consistent with duty is good for that person would be a basic point about personal value, whereas that playing games is good for everyone, or any particular person, would be derivative (if true). The naturalists' error, from the Kantian perspective, was not in their (correct) recognition that empirical facts about what we want, find satisfying, etc., are crucially relevant to most of our value judgments; it was rather that they identified the value judgments with judgments about natural properties.

⁵³ Kant, Groundwork, pp. 88–92 [420–25].

⁵⁴ Kant, Groundwork, pp. 95–98 [427–30].

⁵⁵ I distinguish the “thin” from the “thick” interpretation of the humanity formula of the Categorical Imperative in “Donagan's Kant,” Ethics, vol. 104, no. 1 (1993), pp. 22–52. Neither interpretation, I should note, makes explicit any grounds for the decent treatment of animals, although neither denies that there are such grounds or implies that the ground is merely Kant's unsatisfactory argument that cruelty to animals creates habits of cruelty to people. A theory that, in the end, cannot convincingly articulate good grounds for the decent treatment of animals is woefully incomplete, as Kant's critics have often pointed out.

⁵⁶ Kant's aim is similar in some respects to Hobbes's, when Hobbes, after noting that in a state of nature we each call the objects of our diverse desires “good,” insists that we need a common authoritative standard to determine a “good” that all will acknowledge. See Hobbes, Leviathan, pp. 120–21. For Hobbes, the standard, of course, was the Sovereign's voice, once the state was established–whereas Kant's standard, the voice of “practical reason” is a construct from ordinary reasonable deliberation, namely, an ideal of duly constrained joint deliberations of persons presumed to have certain general dispositions (traditionally associated with our “rational nature”).

⁵⁷ This is meant as an example, with obvious gaps, not a general criterion of treating someone as a mere means. Sometimes, of course, coercion is justified and a person's failure to consent is not a barrier; to sort cases, we need, among other things, to appeal to the idea of hypothetical consent or justifiability to the person under specified ideal conditions.

⁵⁸ Kant, Groundwork, pp. 98–102 [431–34].

⁵⁹ Note that among these concerns would be the general concern with one's own happiness: that is, wanting, within the bounds of reason, to realize some large set of jointly possible ends that one endorses upon reviewing the many diverse things one feels inclined toward. This concern, of course, will be important, though not the only factor, in applying the abstract idea of reasonable deliberation to actual conditions.

⁶⁰ Kant, Groundwork, pp. 100–102 [433–35].

⁶¹ Kant, Groundwork, pp. 102–3 [435–36]. See also my “Donagan's Kant.”

⁶² My efforts to work out this legislative model (and to identify the problems with it) so far include mainly chapters 2, 3, 4, 10, and 11 in my Dignity and Practical Reason in Kant's Moral Theory; “A Kantian Perspective on Moral Rules,” Philosophical Perspectives, vol. 6 (1992), pp. 285–304; and “Donagan's Kant.” An application of the idea, independent of Kant, can be found in my Autonomy and Self-Respect, ch. 6.

⁶³ The supreme moral principle, what is supposedly expressed in the various forms of the Categorical Imperative, is “categorical” (indeed supposedly the only “categorical” imperative) in a further, stronger sense, implying that to show its rational necessity, unlike that of derivative principles, one does not need empirical premises. (Here I report the Kantian view without endorsing it.)

⁶⁴ The idea that reasonable Kantian legislators would reach agreement is more plausible, though perhaps still not guaranteed, when the principles are quite general and leave some room for possible exceptions–e.g., “Everyone should make some efforts to contribute to the happiness of others.” Even specific, unqualified principles may be reasonably presumed to be agreeable to all who take up the legislative perspective (with its stipulated attitude and constraints)–e.g., “One should not torture and kill human beings solely for the amusement of oneself or others.” Those who insist on the inevitability of disagreement are usually focusing on borderline “hard cases,” are not heeding the Kantian constraints on moral deliberation, or both.

⁶⁵ It should be clear that this ideal construction of a legislative perspective makes no pretense at being describable, or defendable, entirely in “neutral,” nonevaluative terms–although how various “evaluative” terms are to be analyzed remains a matter of controversy.

⁶⁶ There is a trivial sense, of course, in which one always acts, if one acts intentionally, on one's own judgment; but what I have in mind is something else. There is a possibility of judging (even after consultation with others) that one thing is best, but then doing instead what “most people” judge best simply because they say so (rather than because they have convincing reasons). It is this latter kind of “acting against one's judgment” that goes against the conscience of a responsible agent.

⁶⁷ Alternatively, if they cannot agree on the specific treatment, can they justify to each other some general principles governing institutions that might acceptably arbitrate residual disagreements on specific issues?

⁶⁸ Skeptics about the necessary rationality of morality should not feel cheated by this stipulation, for the stipulations are aspects of an analysis of deliberation from a moral point of view and not covert attempts to allay or circumvent skeptical doubts. Skeptical worries can still be raised at a different point: why accept the results of what Kantians call “reasonable deliberation” as authoritatively binding on ourselves? At this point, one might argue, Kant's arguments seem to run out and he must appeal to “the fact of reason,” i.e., the supposedly inescapable sense of moral consciousness that we have some genuine moral duties. What Kant in the end defends is that, by analysis, we see that, if you accept the fact of reason, you must accept that rational deliberation goes beyond instrumental reason in the ways I have sketched. But if you really reject the “fact of reason,” not just by saying that it is intellectually unproven but by freeing yourself (like a sociopath?) from the moral attitude and purging your dispositions to judge and act according to it, then probably neither Kant nor anyone else has an argument that will change you.

⁶⁹ The “etc.” here refers to the rest of the various perfect and imperfect duties that Kant argues for in the Groundwork, The Metaphysics of Morals, and various political writings. For the latter, see Kant: Political Writings, ed. Reiss, Hans (Cambridge: Cambridge University Press, 1992).Google Scholar The list includes more duties than I explicitly mentioned, but the main point is that both moral and nonmoral reason leave considerable leeway for individual choice of way of life.

⁷⁰ Kant, The Metaphysics of Morals, p. 209 [409].

⁷¹ See Kant, The Metaphysics of Morals, pp. 192–94 [387–91] and 243–54 [448–61]. Interpretation here is somewhat controversial. My interpretation and evidence can be found in Dignity and Practical Reason in Kant's Moral Theory, ch. 8.

⁷² See Hare, Moral Thinking, pp. 44–64.