1. I assume that a rational defense of the criminal law is possible. For a challenge to this assumption, and an argument that many of the distinctions central to the criminal law are incapable of rational reconstruction, see Kelman, Mark, A Guide to Critical Legal Studies (Cambridge: Harvard University Press, 1987)Google Scholar and, more recently, Norrie, Alan, Crime, Reason and History (London: Weidenfeld and Nicolson. 1993).Google Scholar

2. Perhaps the most fundamental problem is to justify the central place of culpability in the criminal law. Applications of methodologies that have been useful in understanding other areas of the law, such as economic analysis, are notoriously unable to provide an account of culpability in the criminal law. See Coleman, Jules, Crime, Kickers, and Transaction Structures, in Pennock, Ronald and Chapman, John, eds.: Nomos XXVII: Criminal Justice (New York: New York University Press. 1985), at 313.Google Scholar

3. For useful criticisms of several proposed accounts, see Horden, JeremyCriminal Culpability: The Possibility of a General Theory, 12 Law and Philosophy 193 (1993).CrossRefGoogle Scholar

4. Many such factors are included in sentencing guidelines. For example, the Federal Sentencing Guidelines punish defendants more severely if their victims are especially vulnerable due to age or physical condition. See Sentencing Guidelines and Policy Statements (U.S. Government Printing Office. 1987).Google Scholar Sec. 3A 1.1.

5. See note 53 infra.

6. Many theorists are attracted to structure in criminal law defense theory. For a critical discussion, see Husak, Douglas. The Serial View of Criminal Law Defenses, 3 Criminal Law Forum (1992) at 369.CrossRefGoogle Scholar

7. See Simons, Kenneth. Rethinking Mental States, 72 Boston University Law Review 463, 469 (1992)Google Scholar

Many other commentators apparently endorse the sequential principle, although they do not explicitly mention whether the same ordering of culpable states applies to each and every criminal act Consider, for example: “I accept, of course, the culpability rankings of the Model Penal Code … Purposeful behavior is more culpable than knowing, knowing is more culpable than reckless and reckless is more culpable than negligent.” Low, Peter, The Model Penal Code, the Common Law, and Mistates of Fact: Recklessness, Negligence, or Strict Liability?, 19 Rutgers Law Journal 539, 552 (1988).Google Scholar

8. Arguably, the culpability structure of the Model Penal Code may be a good deal more complicated Perhaps the Code surreptitiously makes motive relevant to criminal liability. See Husak, Douglas, Motive and Criminal Liability, 8:2Criminal Justice Ethics (1989), at 3.CrossRefGoogle Scholar

9. These respective definitions of purpose, knowledge, recklessness, and negligence (with respect to results) are found in Model Penal Code, Sec. 2.02(2) (a), (b), (c), and (d).

10. The details of a desert theory of sentencing are described and defended in a series of works by Andrew von Hirsch. See his Doing Justice (New York: Hill and Wang. 1976)Google Scholar; Past or Future Crimes? (New Brunswick: Rutgers University Press, 1985)Google Scholar; and Censure and Sanction (Oxford: Clarendon Press, 1993).Google Scholar

11. I suspect that disagreement about the number of distinct culpable states contributes to the greater skepticism among British commentators about the prospects or providing a theory of culpability. See Horder, , op. cit. note 3.Google Scholar

12. See Robinson, Paul. A Brief History of Distinctions in Criminal Culpability, 31 Hastings Law Journal 815 (1980).Google Scholar

13. There may or may not have been a time when all criminal liability was “strict” See Robinson, supra note 12, at 823.Google Scholar

14. See Wootton, Barbara, Crime and the Criminal Law (London: Stevens & Sons, 1963).Google Scholar Wootton argues that distinctions among culpable states are important only to provide a basis for deciding how offenders should be treated in order to minimize the probability of recidivism.

15. See Sayre, Frances. The Present Significance of Mens Rea in the Criminal Law, in Campbell, Morton et al. , eds.: Harvard Legal Essays (Cambridge: Harvard University Press, 1934), at 399, 404Google Scholar

16. See Acimovic, Mihajlo, Conceptions of Culpability in Contemporary American Crimnal Law, 26 Louisiana Law Review 28, 51 (1965).Google Scholar

17. Law Reform Commission of Canada, Report No. 30, Recodifying Criminal Law (1986), at 18–21.Google Scholar

18. Law Commission No. 143. Codification of the Criminal Law: A Report to the Law Commission (Proposed Draft, 1985), at 183–184.Google Scholar

19. Model Penal Code, Commentaries to Sec. 2.02. at 230, n.3.Google Scholar

20. Simons, , supra note 7.Google Scholar

21. Id., at 464.

22. Simons argues that yet. “a third hierarchy, of conduct, which does not essentially involve mental states” must also be countenanced (id., at 464). But this third hierarchy plays a much smaller role in his theory.

23. Id., at 496.

24. Id., at 477 (emphasis In original).

25. See Robinson, , supra note 12, at 816.Google Scholar

26. Model Penal Code. Commentaries to Sec. 2.02. at 229.Google Scholar

27. Id., at 230.

28. More precisely, different culpable stales might attach to the material elements (such as a taking) that compose the actus rem or the criminal act of theft. I make this point because some persons may regard the concept of a “negligent theft” to be incoherent. See note 29 infra.

29. This generalization is subject to many exceptions. Some criminal acts could not be described meaningfully with each of the four different culpable states. The nonconsummate offenses of attempt, solicitation, and conspiracy are not the only such exceptions. Some statutes use verbs, such as assault, which require intentionality. These offenses could not be committed negligently. A defendant might swing his fists in the air and injure someone, but his act would not be described as an assault unless he intended his blows to land. See Husak, Douglas, Philosophy of Criminal Law (Totowa: Rowman & Littlefield, 1987), at 122–128.Google Scholar

30. The Model Penal Code does not create separate offenses of homicide depending on whether a defendant kills purposely or knowingly. Both are murder. Moreover, it allows that a reckless killing can be murder, if the degree of recklessness is sufficiently extreme. See Sec. 210.2(1)(b).

31. Model Penal Code, Sec. 250.5.

32. This result follows by application of Section 2.02(3), which specifies that recklessness suffices as the level of culpability when “the culpability sufficient to establish a material element of an offense is not prescribed by law.”

33. For some reservations about leaving these determinations to sentencing authorities, see Robinson, Paul and Darley, John. Justice. Liability, and Blame (Boulder: Westview Press, 1995), at 208–209.Google Scholar

34. For impressive empirical evidence that laypersons would not be confused by creating separate grades of the same offense, see Robinson, and Darley, , supra note 33, at 91.Google Scholar

35. Simons accepts this interpretation. He cites this section as evidence that “the Model Penal Code explicitly orders the mental state categories in a hierarchy.” See op.cit. note 7, at 479.Google Scholar

36. This interpretation has ample support in the commentaries. The explanatory note indicates that Sec. 2.02(5) “makes it unnecessary to state in the definition of an offense that the defendant can be convicted if it is proved that he was more culpable than the definition of the offense requires” (emphasis in original). The commentaries add that “it is only necessary to articulate the minimal basis of liability in drafting specific offenses for the more serious bases to be implied.”

37. Clearly the application of the “greater includes the lesser” rule is less than obvious in circumstances in which the greater does not entail the lesser. To cite just one problematic application, suppose it is true that a state has the power to abolish a given defense altogether. This “greater” power might be thought to include the “lesser” power to allocate the burden of proof with respect to that defense in anyway the state might choose. For some difficulties with this application of “greater includes the lesser” rule, see Underwood, Barbara. The Thumb on the Scales of Justice: Burdens of Persuasion in Criminal Cases, 86 Yale Law Journal 1299 (1977).CrossRefGoogle Scholar

38. In other words, the existence of a justification precludes (by definition) that an agent acts recklessly, but does not preclude that he acts knowingly. Of course, the doctor who knowingly wounds her patient will almost certainly escape liability because of some defense or another (such as necessity or consent). By contrast, a reckless battery lacks a justification by definition.

Whether knowledge (or purpose), like recklessness, should be defined to include the lack of justification is a different and difficult matter. See Simons, , op. cit. note 7, at 492–493.Google Scholar

39. Sometimes, however, courts relativize the test of what a reasonable person should know in light of the actual expertise of the defendant. Perhaps the test would be whether a “reasonable chemist” would have been aware of the explosion.

40. See Simons, . op.cit. note 7, at 479.Google Scholar

41. This interpretation of “lesser included offense” has played an important role in applying the rule against double jeopardy. See Blockburger v. U.S., 284 U.S. 299 (1932).Google Scholar

42. The “greater includes the lesser” rule is the contrapositive of the “absence of the lesser includes the absence of the greater” rule. Propositions are logically equivalent to their contrapositives.

43. I am (sometimes) inclined to say that a person who kills in self-defense kills knowingly, but lacks the culpability of a person who kills knowingly (but who does not act in self-defense). If this inclination is sensible—and I do not endone it without reservations—it follows that having the culpability of knowledge is not simply a matter of knowingly committing a criminal act.

44. In contending that offenses based largely on a paternalistic rationale provide “hard cases” for defenders of the sequential principle. I reserve judgment about whether the principle is vulnerable to counterexamples in cases in which a criminal law is designed to prevent persons from harming others.

45. Perhaps judgments about persons who injure themselves, or who run risks of injuring themselves, are judgments about the lack of prudence rather than about blame. Or perhaps we judge such persons to be blameworthy because we judge them to be imprudent. In the context of offenses that are based largely on a paternalistic rationale, judgments about blame and the lack of prudence seem inextricably linked.

46. Admittedly, judgments about the permissibility of suicide (or attempted suicide) are most likely to be shared when someone is suffering and terminally ill. Still, I would think that many persons would judge suicide (or attempted suicide) to be permissible whenever someone lacks an interest in remaining alive.

47. For a defense of the claim that the criminal law protects individual rights that are constructed from personal interests, see Feinberg, Joel, Harm to Others (New York: Oxford University Press, 1984).Google Scholar

48. See the useful discussion in Feinberg, , supra note 47, at 115–117.Google Scholar

49. See. for example, the discussion of risk assessment in Viscusi, W. Kip, Smoking: Making the Risky Decision (New York: Oxford University Press, 1992).Google Scholar

50. See the sophisticated discussion of voluntariness in Feinberg, Joel, Harm to Self (New York: Oxford University Press, 1986), at 98–169.Google Scholar

51. See Husak, Douglas. Drugs and Rights (Cambridge: Cambridge University Press, 1992), ch. 2 and 3.CrossRefGoogle Scholar

52. But see the critical discussion in Winick, Charles, Social Behavior, Public Policy, and Nonharmful Drug Use, 69 Milbank Quarterly 437 (1991).CrossRefGoogle ScholarPubMed

53. It might be objected that I have failed to show that a different ordering of culpable states should be applied to paternalistic legislation. Applications of the sequential principle of relative culpability are qualified by the requirement that all other variables that affect the blameworthiness of offenders must be held constant. But, the objection continues, a different ordering of culpable states only appears to pertain to paternalism because I have neglected this important ceteris paribus clause. It might be conceded that persons who lack an interest in living, for example, are not blameworthy for killing themselves, even though persons who have an interest in living should not be permitted to take given kinds of risks. But if the interests of persons are held constant, there is no need to alter the ordering of culpable states from its more familiar sequence. I have stated that, in standard cases, persons have an interest in not being victimized by criminal activity. But what should be said about nonstandard cases? Suppose that Green lacks an interest in remaining alive. How does the culpability of someone who kills Green affect his blameworthiness? How should culpable states be ordered in such a scenario? Is it worse to wilfully kill Green than to do so accidentally?

I have no clear intuitions about how these questions should be answered. Perhaps culpable states should be ordered here in the same way as my conjecture about legislation that is based largely on a paternalistic rationale. If so, a person who negligently or recklessly kills Green is actually mart blameworthy than a person who kills knowingly or purposely. If this answer is correct, the number of counterexamples to the sequential principle may be even more numerous than I have supposed thus far. That is, the general exception to the typical ordering of culpable states that seems to apply to paternalism might apply to a few additional cases as well. The ordering of culpable states changes from its familiar sequence not only in the context of paternalism, but also in the context of all cases in which persons do not actually have the interests that the criminal law generally assumes them to possess. But this objection does not undermine my general conjecture. A person who knowingly or purposely kills Green is not more blameworthy than a person who is negligent or reckless with respect to the risk that Green will die. If so, the sequential principle of relative culpability—or at least, the strong version of the sequential principle—must be qualified.