¹ Chambers's Twentieth Century Dictionary (London: Chambers, 1959)Google Scholar. There is very little new that can be said on the topic of this essay—every idea has its antecedents, and an essay that attempted to cite them all would be impossible to write, to read—and to print. I have done my best to reference key items, taking comfort from Alfred North Whitehead's aphorism “Everything of importance has already been said by someone who did not discover it.”

² Some talk of “responsibility to oneself,” but this is metaphorical, usually employed to disguise selfishness (“I'm doing it for me”) as altruism (“I'm doing it for someone else”).

³ Ziman, J. M., Public Knowledge: An Essay Concerning the Social Dimension of Science (Cambridge: Cambridge University Press, 1968).Google Scholar

⁴ For simplicity and on the basis of historical precedent—and because most criminals are male—I use the masculine generic.

⁵ Menninger, Karl, The Crime of Punishment (New York: Viking Press, 1968)Google Scholar; Skinner, B. F., Beyond Freedom and Dignity (New York: Bantam Books, 1971).Google Scholar

⁶ See Dershowitz, Alan, The Abuse Excuse (Boston: Little, Brown, 1994)Google Scholar, for an entertaining description of the creative flights of the legal profession in this direction. See Horowitz, D. L., “Justification and Excuse in the Program of the Criminal Law,” Law and Contemporary Problems 49 (1986): 109–26CrossRefGoogle Scholar, for an account of some of the legal aspects.

⁷ This is Skinner's term—but see note 27.

⁸ Note that the contrary proposition, that we are, or could be, aware when our behavior is determined, leads to endless regress: because we are not then aware that our awareness may itself be determined. Note also that failure to predict does not imply absence of determinism. The behavior of a chaotic system like the logistic map becomes less and less predictable further and further into the future, even though it is perfectly deterministic, for example.

⁹ Skinner, , Beyond Freedom and Dignity, 17.Google Scholar

¹⁰ Dennett, Daniel, Elbow Room (Cambridge, MA: MIT Press, 1984)Google Scholar; Levin, Michael, Why Race Matters (New York: Praeger, 1997), 172 and 318Google Scholaret seq.; Staddon, John, “On Responsibility and Punishment,” The Atlantic Monthly, 02 1995, 88–94Google Scholar. I describe the concept of contingency in more detail below.

¹¹ He may feel remorse for past misdeeds, but he is responsible to no one for his actions on the desert island.

¹² In fact, for many practical matters, chaos theory is not relevant. Space shuttles and geostationary satellites achieve their orbits through Laplacean calculations. The “physics of the macroworld,” as it is called, is pretty much predictable. And we have no solid evidence that human behavior, at the gross level that is important for much of public policy, is unpredictable either.

¹³ See, for example, Penrose, Roger, The Emperor's New Mind (Oxford: Oxford University Press, 1990).Google Scholar

¹⁴ In 1843, Daniel M'Naghten murdered Edward Drummond, secretary to Sir Robert Peel, the British Prime Minister who was M'Naghten's intended victim. M'Naghten contended that his mission was guided by the “voice of God.” This trial yielded the M'Naghten rule: insanity is proved if the defendant was “labouring under such a defect of reason, from disease of the mind, as not to know the nature and quality of the act he was doing; or if he did know it, that he did not know he was doing what was wrong.” See, for example, Livermore, L. and Meehl, P., “The Virtues of M'Naghten,” Minnesota Law Review 51 (1967): 789–856.Google Scholar

¹⁵ I term this component social, even though it represents a capacity of individuals, because it depends on the uniquely human ability to learn about consequences through signals (signs, language) rather than only from direct experience.

¹⁶ See, for example, Staddon, J. E. R., “Operant Behavior as Adaptation to Constraint,” Journal of Experimental Psychology: General 108 (1979): 48–67CrossRefGoogle Scholar, and other papers in that issue, for many examples of the relations between psychological and economic concepts.

¹⁷ Or, at least, neoclassical economists. There are other varieties that take a broader view (e.g., Armen Alchian, Brian Arthur, Ludwig von Mises).

¹⁸ It is, of course, irrelevant to the economic argument whether support payments actually reward childbearing or simply remove a disincentive. Charles Murray has provided a pathbreaking analysis of this and related issues: see Murray, , Losing Ground: American Social Policy: 1950–1980 (New York: Basic Books, 1984).Google Scholar

¹⁹ Breland, K. and Breland, M., “The Misbehavior of Organisms,” American Psychologist 16 (1961): 661–64CrossRefGoogle Scholar. See Staddon, J. E. R. and Simmelhag, V., “The ‘Superstition’ Experiment: A Reexamination of Its Implications for the Principles of Adaptive Behavior,” Psychological Review 78 (1971): 3–43CrossRefGoogle Scholar, for a theoretical account of the phenomena described in the text.

²⁰ See Honig, Werner K. and Staddon, J. E. R., eds., Handbook of Operant Behavior (Englewood Cliffs, NJ: Prentice-Hall, 1977)Google Scholar, for many of these examples.

²¹ Hughes, Robert, The Fatal Shore (London: Collins, Harvill, 1987), 481.Google Scholar

²² Frayne, Laurence, “Memoir on Norfolk Island,” NSW Colonial Secretary Papers, vol. 1 (re. NSW 1799–1830), Ms. 681/1, ML, Sydney, 25–26.Google Scholar

²³ Literature provides many vivid examples in which the threat of severe punishment fails to achieve the desired behavior. For instance, in the seventeenth-century play Surgeon of Honor, by Spanish writer Calderon de Barca, the clown-servant Coquin is given the ultimatum that he must make the King laugh—or else have all his teeth pulled out. This particular reinforcement contingency is more likely to leave Coquin toothless than the King in fits of mirth.

²⁴ Hughes, , The Fatal Shore, 499.Google Scholar

²⁵ Negative reinforcement is the removal of an aversive stimulus (for example, reducing the convict's time of servitude); positive reinforcement was provided in Maconochie's scheme in the form of goods (such as food, tobacco, and clothes) that could be bought with “marks.” “Time off for good behavior” is an impoverished modern version of Maconochie's “marks.”

²⁶ Ibid., 500.

²⁷ “Control” in Skinner's misleadingly forceful terminology. The effects of contingencies are always limited by the means organisms have for adapting to them; hence, “predict” is a more accurate term than “control.”

²⁸ More precisely, such explanation is teleonomic, in C. S. Pittendrigh's sense (see Pittendrigh, , “Adaptation, Natural Selection, and Behavior,” in Behavior and Evolution, ed. Roe, Anne and Simpson, George G. [New Haven: Yale University Press, 1958]).Google Scholar

²⁹ There are extensive discussions of these issues in the biological literature. See, for example, Smith, J. Maynard, “Optimization Theory in Evolution,” Annual Review of Ecology and Systematics, 1978: 31–56CrossRefGoogle Scholar; Oster, J. F. and Wilson, E. O., “A Critique of Optimization Theory in Evolution,”Google Scholar in Oster, and Wilson, , Caste and Ecology in the Social Insects (Princeton, NJ: Princeton University Press, 1978), 292–315Google ScholarPubMed; Staddon, J. E. R., Adaptive Behavior and Learning (New York: Cambridge University Press, 1983)Google Scholar; Staddon, J. E. R., “Optimality Theory and Behavior,” in Dupré, John, ed., The Latest on the Best: Essays on Evolution and Optimality (Cambridge, MA: Bradford/MIT Press, 1987), 179–98Google Scholar; Staddon, John, Behaviorism: Mind, Mechanism, and Society (London: Duckworth, 1993)Google Scholar; and Staddon, J. E. R. and Hinson, J. M., “Optimization: A Result or a Mechanism?” Science 221 (1983): 976–77.CrossRefGoogle ScholarPubMed

³⁰ Simon, H. A., “Rational Decision Making in Business Organizations,” in Advances in Behavioral Economics, vol. 1, ed. Green, L. and Kagel, J. H. (Norwood, NJ: Ablex Publishing, 1987)Google Scholar. (This was originally Simon's 1978 Nobel lecture.)

³¹ Staddon, , Behaviorism.Google Scholar

³² These direct effects need not be immediate. Instinctive drift follows the laws of Pavlovian (rather than operant) conditioning and takes some time to develop. The point is that what I am calling “direct effects” are effects of antecedent, not consequent, events. Psychologists sometimes use the term elicited for these effects.

³³ This actually gives too much credit to the economists, who were entirely uninterested in dynamics, process, etc.

³⁴ The work of Sarnoff Mednick, who tracked the careers of criminals with and without a history of consistent punishment for their crimes, is a notable exception. For these individuals, punishment was evidently a strong deterrent; see, e.g., Brennan, P. A. and Mednick, S. A., “Learning Theory Approach to the Deterrence of Criminal Recidivism,” Journal of Abnormal Psychology 103 (1994): 430–40.CrossRefGoogle ScholarPubMed

³⁵ Bradshaw, John, Lear's Magazine, 01 1993, 42.Google Scholar

³⁶ Seventeenth-century diarist Samuel Pepys was willing to be “cut of the stone” (probably a gallstone), sans anesthetic, because without the operation, death was certain. Depressives will tolerate the pain of electroconvulsive shock because the treatment, crude as it is, often provides relief. People will accept pain from a therapist if their condition is dire and they believe relief is probable.

³⁷ A critical discussion of psychotherapy, particularly psychoanalysis, that is both knowl edgeable and entertaining is experimental psychologist Sutherland, Stuart's description of his own treatment for mental-health problems: Breakdown: A Personal Crisis and a Medical Dilemma (London: Weidenfeld and Nicolson, 1976)Google Scholar. A more recent critique is to be found in Campbell, Terence W., Beware the Talking Cure: Psychotherapy May Be Hazardous to Your Mental Health (Boca Raton, FL: Upton Books, 1994)Google Scholar. It is not just capitalist insensitivity that makes the bean-counters who run health maintenance organizations reluctant to pay for mentalhealth benefits. The cost-benefit calculations are not favorable.

³⁸ As induction into some religious orders—a sort of psychotherapy, one might suppose—used to do.

³⁹ This myopia seems to be characteristic of all the “caring” professions. Some readers may recall the objections of physician Dr. Marcia Angell, editor of the New England Journal of Medicine, to a government-funded study that proposed to compare very low doses of anti-HIV drugs with a no-dose control group, as part of an effort to find an inexpensive treatment for endemic AIDS in sub-Saharan Africa; see Angell, Marcia, “Tuskegee Revisited,” Wall Street Journal, 10 28, 1997Google Scholar. Her objection: that some experimental subjects received no treatment. The fact that no subjects were injured, that the value of the “treatment” given to experimental subjects was conjectural, and that the results of the study might benefit millions, carried no weight. Her attention was entirely focused on the fact that the control subjects (under the “care” of the experimenters, in her view) received no treatment. (See John Staddon, letter to the editor, Wall Street Journal, 11 11, 1997Google Scholar, for a contrary view.)

⁴⁰ In the animal laboratory, the most persistent behavior is generated by shock-postponement schedules. These work by delivering brief, painful electric shocks at fixed intervals. The next shock can be postponed for a fixed time by pressing a lever or making some other operant response. Well-trained animals respond frequently enough to avoid all shocks, and may continue to do so indefinitely, long after the shock generator has been disconnected. No schedule of positive reinforcement produces such persistent effects. (It is an interesting sidelight on the nexus between policy advice and laboratory science, that the inventor of this procedure is nevertheless a passionate opponent of aversive control in public policy; see Sidman, Murray, Coercion and Its Fallout [Boston, MA: Authors Cooperative, 1989]).Google Scholar

⁴¹ In Aubrey, John's The Life of Mr. Thomas Hobbes of Malmesburie, in English Biography in the Seventeenth Century, ed. de Sola Pinto, Vivian (London: Harrap, 1951), 188.Google Scholar

⁴² The commonest form of punishment, incarceration, serves a third purpose: prevention of further offenses by the criminal during his prison time. Incarceration of repeat offenders can be very effective in reducing crime rates (see, e.g., Methvin, E. H., “Mugged by Reality,” Policy Review, 07 1997, 32–38)Google Scholar. But the control here is physical, not behavioral.

⁴³ Many legal theorists have argued for the contrary view. Most recently, Moore, Michael (see Placing Blame: A General Theory of the Criminal Law [Oxford: The Clarendon Press, 1997])Google Scholar has argued persuasively and at length that “[o]f the possible functions for criminal law, only the achievement of retributive justice is its actual function” (78–79)Google Scholar. I have no space here to contest this position. Suffice it to say that: (a) Achieving consensus on what is retributively just is likely to be even more difficult than getting agreement on what minimizes general suffering, (b) The near-universal agreement that punishment for attempted murder should be less than for successful murder strongly supports the idea that judicial punishment should have a retributive component and not be based on deterrence only. But (c) punishment certainly does deter, so it seems unwise to formulate criminal-justice policy without taking this element into account.

⁴⁴ Philosophical economists have pointed out that utility cannot be compared between individuals: even President Clinton cannot really “feel my pain.” Nevertheless, such comparisons are essential to any utilitarian analysis of social effects. In practice, most writers seem to assume a highly nonlinear scale that effectively rules out extreme (“cruel and unusual”) punishments. Rawls, John's influential A Theory of Justice (Cambridge, MA: Harvard University Press, 1971)Google Scholar also argues for this view.

⁴⁵ See Beccaria, Cesare, An Essay on Crimes and Punishments (1764; Stanford, CA: Academic Reprints, 1953)Google Scholar, for what seems to be the earliest statement of this view.

⁴⁶ “Why worry about whether the degree of suffering imposed on the criminal is appropriate to the crime?” some might ask. “Why not just minimize the suffering of victims?” One reason is that just retribution probably cannot be excluded as a component in criminal justice, even if it is probably not the only component. Another is that without including the suffering of the criminal in the utilitarian equation, we can raise no principled objection to a colleague of mine who, only half-joking, proposed the death penalty for littering.

⁴⁷ What about the individual who is sensitive to punishment in the normal way, but is willing to disobey the law anyway because he disagrees with it? Mohandas Gandhi and Nelson Mandela, but also Oliver North and Leon Trotsky, come to mind. Such individuals will naturally fail the deterrence test. Consistency implies that they should therefore be treated as irresponsible, but dangerous, like a mentally ill person. And indeed, political prisoners are often treated in this fashion, restrained indefinitely but not necessarily punished in any other way (cf. Mandela's long-term incarceration, the indefinite house-arrest of the Burmese opposition leader Aung San Suu Kyi, etc.).

⁴⁸ A skeptic might ask: “Well, if he would have heeded the angel, why did he shoot?” The obvious answer is: “Because he thought he wouldn't be caught—or, he thought that if he were caught, he could get off.” The belief that a crime will go undetected or unpunished is, of course, a major “root cause” of crime.

⁴⁹ Ms. Landers, now in her eighties, has been writing an advice column syndicated in hundreds of U.S. newspapers for more than forty years.

⁵⁰ Menninger, , The Crime of Punishment (supra note 5).Google Scholar

⁵¹ Of course, prohibition is one solution to this dilemma. But again, nonaddicts are pun ished needlessly by being deprived of a harmless (for them) enjoyment. Is the greater good served? What about personal freedom, which is also a good? Again, these questions go beyond practical science.

⁵² See, for example, Kutchins, Herb and Kirk, Stuart A.'s recent exposé Making Us Crazy: DSM—The Psychiatric Bible and the Creation of Mental Disorders (New York: The Free Press, 1997)Google Scholar; and Davis, L. J., “The Encyclopedia of Insanity—A Psychiatric Handbook Lists a Madness for Everyone,” Harpers Magazine, 02 1997.Google Scholar

⁵³ MacFarquhar, Larissa, “Diagnosis: Totally Sane—The DSM Isn't Crazy in the Slightest, A Review of Kutchins and Kirk,” Slate (internet magazine), 11 12, 1997.Google Scholar

⁵⁴ There is a more technical meaning for self-control, as capacity to delay gratification (i.e., impulse control), which is treated at length in an elegant experimental literature. (See, for example, Ainslie, George, Picoeconomics [Cambridge: Cambridge University Press, 1992]Google Scholar; and Rachlin, Howard and Green, Leonard, “Commitment, Choice, and Self-Control,” Journal of the Experimental Analysis of Behavior 17 [1972]: 15–22.)CrossRefGoogle ScholarPubMed I intend the simpler meaning—susceptibility to deterrence—here.

⁵⁵ There is a category of diminished responsibility that is the mirror image of the repeat offender. The repeat offender is, by definition, insufficiently deterred by punishment. The complementary category is people who are excessively encouraged by reward. It may seem odd to see such people as less responsible than average, but this is apparently the view of a colleague of mine who recently termed payments to a public lottery an “implicit tax.” Granted that a tax is “a contribution exacted by the state: anything imposed, exacted or burdensome,” to call voluntary and presumably pleasurable participation in a lottery a tax is, on its face, absurd—unless you feel that people who play a lottery cannot control themselves—are not fully responsible. This category is not yet fully accepted in the legal arena (“compulsive gambling” and “sex addiction” as mental illnesses are contenders, I suppose), but its time will probably come.

⁵⁶ See, for example, Rapport, M. D., “Attention-Deficit/Hyperactivity Disorder,” in Hand book of Psychological Treatment Protocols for Children and Adolescents, ed. Van Hasselt, V. B. and Hersen, Michel (Mahwah, NJ: Lawrence Erlbaum, 1998), 65–107Google Scholar, for the current view on this fluid category.

⁵⁷ There are, of course, many more dimensions to the problem than simply punishment and reward. The child is sensitive to all sorts of aspects that cannot easily be quantified: Is the punishment just? Is the teacher to be respected? To be believed? What is right and wrong? Will my parents support me or my teacher? How about my friends? And so forth. The effects of these aspects, which may deflect the expected effects of reward and punishment, come under the heading of the direct effects of reinforcement contingencies, discussed earlier.

⁵⁸ See Murray, , Losing GroundGoogle Scholar, for a thoughtful discussion of this problem.

⁵⁹ “Repeated misbehavior” is a valid measure of insensitivity to discipline, of course, and those who failed this test used to be segregated in “reform schools,” but this practice has fallen out of favor. I am arguing that some fraction of ADHD children probably belong in the “reform school” category.

⁶⁰ The traditional self-serving male chauvinist view is, of course, the opposite.

⁶¹ Mamet, David's play OleanaGoogle Scholar presents a vivid picture of this scenario.

⁶² See Young, Cathy, “Domestic Violations,” Reason, 02 1998, 23–31.Google Scholar

⁶³ The entertainment media have certainly picked up on this theme. Two examples: When asked by a television reporter about his simultaneous affairs with two sisters and their mother, English political charmer, aristocrat, and self-confessed rake Alan Clark recently responded: “Whatever happened to free will?” The answer seems to be that for women it has been, to some degree, suspended. And in the popular movie As Good as It Gets, the misanthrope romance novelist played by Jack Nicholson, when asked how he writes believable female characters, responds: “I think of a man. Then I take away reason and accountability.” Plus ça change …

⁶⁴ Women tend to lead men in a few crimes, such as poisoning and embezzlement; otherwise, men are well ahead, particularly in violent crime.

⁶⁵ The bald fact of lower crime rates by women leaves open the question of how this comes about. The Darwinian metaphor offers two possibilities that are not mutually exclusive: women are less likely to initiate crimes (i.e., show less behavioral variation of this type); or they are more responsive to punishment. The two might be distinguished by comparing the recidivism rate for men and women comparably punished for comparable crimes. (Of course, women might first-offend less often than men because they are more readily deterred by the prospect of punishment—i.e., they may exhibit more of the social aspect of responsibility.)

⁶⁶ For a survey, see Gladwell, Malcolm, “Crime and Science: Damaged,” The New Yorker, 02 24, 1997, 132–47.Google Scholar

⁶⁷ Ibid., 134–35.

⁶⁸ Ibid., 138.

⁶⁹ Ibid., 134.

⁷⁰ C. S. Lewis, in a prophetic essay, points out the dangers of medicalizing crime—and argues for a retributive theory of punishment; see Lewis, , “The Humanitarian Theory of Punishment,”Google Scholar in Lewis, , God in the Dock—Essays on Theology and Ethics, ed. Hooper, Walter (Grand Rapids: Eerdmans, 1970)Google Scholar; Lewis, 's essay was originally published in 20th Century: An Australian Quarterly Review 3, no. 3 (1948).Google Scholar

⁷¹ O'Brien, Patrick, The Mauritius Command (New York: W. W. Norton, 1977), 210.Google Scholar

⁷² Wilson, James Q., Moral Judgment: Does the Abuse Excuse Threaten Our Legal System? (New York: Basic Books, 1997), 36–37.Google Scholar

⁷³ Bazelon, David L., “Veils, Values, and Social Responsibility,” American Psychologist 37 (1982): 115–21.CrossRefGoogle Scholar

⁷⁴ Wilson, , Moral Judgment, 37.Google Scholar

⁷⁵ Ibid., 112.

⁷⁶ Ibid., 36.

⁷⁷ This view is, apparently, mistaken, since hate crimes are neither common nor increasing in frequency; see Jacobs, James B. and Potter, Kimberly, Hate Crimes: Criminal Law and Identity Politics (New York: Oxford University Press, 1998).Google Scholar

⁷⁸ I do not add “legal,” because I assume that law is a product of science, politics, and morality, not an independent entity. But this view may not be generally held.