I would like to thank Professors Richard Singer and George Thomas for their extraordinarily helpful comments.

1. In a comprehensive survey, one commentator concludes that “the dominant view appears to be that in the Anglo-American culture, the use of strict liability crimes is arbitrary and unreasonable.” See Levenson, Laurie “Good Faith Defenses: Reshaping Strict Liability Crimes” (1993) 78 Cornell L. Rev. 401 note 7 at 403.Google Scholar

2. Notable exceptions include Brady, James “Strict Liability Offenses: A Justification” (1972) 8 Crim. L. Bulletin 217;Google Scholar Paulus, Ingeborg “Strict Liability: Its Place in Public Welfare Offences” (1977–78) 20 Crim. L. Quart. 445;Google Scholar and Nemerson, Steven “Criminal Liability Without Fault: A Philosophical Perspective” (1975) 75 Colum. L. Rev. 1517.CrossRefGoogle Scholar

3. Some legal philosophers suggest that strict liability in tort is inevitable, inasmuch as someone must bear the costs of those accidents in which no one is at fault. Requiring plaintiffs to bear their own accident costs when they are not at fault might be called “strict victim liability.” See Coleman, Jules Risks and Wrongs (New York: Cambridge University Press, 1992) ch. 11.Google Scholar

4. The Model Penal Code imposes liability without culpability for a category of offenses described as “violations.” These are distinguished from “crimes” in that they cannot result in a sentence of probation or imprisonment. See Model Penal Code sec. 2.05 (1985).

5. “In such cases…the courts can afford to disregard the individual in protecting the social interest.” Sayre, Francis “Public Welfare Offenses” (1933) 33 Colum. L. Rev. 55 at 69–70.CrossRefGoogle Scholar

6. “The precise meaning of “strict liability offenses” is unclear.” Levenson, supra note 1 at 417. See also Ashworth, Andrew Principles of Criminal Law (Oxford: Clarendon Press, 1991) at 135–36:Google Scholar “There is no clear convention about when criminal liability should be classified as ‘strict.’”

7. Sometimes reservations about the justifiability of strict liability in the criminal law are expressed almost as a matter of definition. “It is clear that whatever sort of liability “strict liability” may be, it is not criminal liability.” Hall, Jerome General Principles of Criminal Law, 2d ed. (Indianapolis, IN: Bobbs-Merrill, 1960) at 326.Google Scholar

8. A few theorists have suggested that some crimes “require greater strictness in their enforcement” than others, but have not explained how degrees in the strictness of liability are possible. See Perkins, Rollin & Boyce, Ronald Criminal Law, 3d ed. (Mineola, NY: Foundation Press, 1982) at 899.Google Scholar

9. My conclusion that liability that is strict to the maximum conceivable extent has never actually existed should be contrasted with the conclusion that such liability is literally a conceptual impossibility. For an apparent defense of the latter position in the context of tort law, see Perry, Stephen “The Impossibility of General Strict Liability” (1988) 1 Can. J. of L. & Jurisprudence 147.CrossRefGoogle Scholar

10. See Robinson, Paul & Grail, Jane “Element Analysis in Defining Criminal Liability: The Model Penal Code and Beyond” (1983) 35 Stanford L. Rev. 681.CrossRefGoogle Scholar

11. Model Penal Code sec. 221.1(1).

12. “An offence may carry strict liability in one respect but not in all.” Williams, Glanville Textbook of Criminal Law, 2d ed. (London: Stevens & Sons, 1983) at 927.Google Scholar Williams defines a strict liability offence as an offence “in which some element does not require proof of fault.” Ibid. If Williams’ definition of “element” is as broad as that of the Model Penal Code sec. 1.13(9), encompassing such matters as jurisdiction or venue, most or all offenses will qualify as strict liability offenses. Perhaps Williams’ use of “element” corresponds to the Model Penal Code definition of “material element” in sec. 1.13(10).

13. Sometimes a commentator indicates his awareness of various kinds of strict liability by proposing a new term to describe one kind or another. The most commonly used such term is “absolute liability.” According to one theorist, “strict liability characterizes offenses that are usually, though incorrectly, said to entail liability without culpability; absolute liability is the term used to characterize offenses for which there actually is liability without culpability. Gross, Hyman A Theory of Criminal Justice (New York: Oxford University Press, 1987) at 343.Google ScholarPubMed The definitions used to draw this distinction, however, are clearly inadequate. Whether or not a given offense satisfies this definition of strict liability depends on what unnamed persons usually but incorrectly “say” about it. On my account, strict liability is a real property of offenses, not a property that persons mistakenly ascribe to offenses.

“Absolute liability” is sometimes used in a different sense. According to one commentator, “liability should only be described as absolute where there is no defence available at all to someone who is proved to have caused the prohibited event.” Almost certainly no real offenses satisfy this definition. The commentator immediately adds: “What this shows, above all, is the inadequacy of common terminology to give simple expression to the numerous permutations of conditions for liability.” Ashworth, supra note 6 at 137.

Yet another confusing term used to contrast with strict liability is “partial absolute liability.” See Mueller, Gerhard “On Common Law Mens Rea” (1958) 42 Minn. L. Rev. 1043 at 1060.Google Scholar

14. According to some commentators, English common law examples of strict liability include public nuisance, criminal libel, contempt of court and blasphemy. See Smith, J.C. & Hogan, Brian Criminal Law, 5th ed. (London: Butterworths, 1983) at 87–88.Google Scholar

15. Two qualifications about these examples are required. First, these examples only manifest strict liability with respect to one or more of their elements. Liability for the offense of statutory rape, for example, is strict only with respect to the age of the victim. Second, these examples only manifest one of the several kinds of strict liability I distinguish below.

16. There is some doubt about the constitutionality of all forms of strict liability. See the careful analysis in Singer, Richard “The Resurgence of Mens Rea: III-The Rise and Fall of Strict Criminal Liability” (1989) 30 Boston Col. L. Rev. 337 at 397–403.Google Scholar

17. Some commentators allege that “in practice, there is no such thing as strict criminal liability.” See Singer, ibid. at 391. This extraordinary result is a consequence of adopting an inadequate definition of strict liability. According to Singer, “the premise of strict liability is that the defendant is held guilty no matter how careful and morally innocent he or she, or one for whose acts he or she is responsible, has been.” Ibid. at 356. Singer is correct that no existing offense satisfies his definition, but is incorrect about “the premise” of strict liability.

A more common observation is that “we have no offenses enforced on the basis of absolute liability.” See Perkins&Boyce, supra note 8 at 898.

18. See Gardner, Martin “The Mens Rea Enigma: Observations on the Role of Motive in the Criminal Law Past and Present” (1993) Utah L. Rev. 635 at 652–53.Google Scholar

19. This account of the central feature of strict liability might be construed as a description of the concept (as opposed to the conception) of strict liability. For a development of this distinction, see Dworkin, Ronald Law’s Empire (Cambridge: Harvard University Press, 1986) at 70–72.Google Scholar

20. It is noteworthy that no term exists to describe that which contrasts with strict liability in the criminal law. In non-legal contexts, “strict” is typically juxtaposed with “lenient” or “loose.” However, no commentator has ever referred to liability as lenient or loose. On my account, liability for an offense is not strict when a defendant cannot be convicted of that offense unless his degree of fault is substantially equal to that of the typical perpetrator of that offense. But I do not attempt to introduce a new term to describe such offenses.

21. One possible alternative would be to identify the baseline as the degree of fault of the paradigm perpetrator of any offense. Yet there is no escape from my general point that (1) whether an offense imposes strict liability must be assessed relative to some norm or baseline, and (2) the totally faultless person cannot be used to represent this baseline.

22. My claim that each offense (such as burglary) involves a paradigm offender with a quantum of fault should not be construed as a statistical generalization (about the average burglar). A robin may be a paradigm case of a bird, even though most birds are not robins. Instead, a paradigm case functions as a heuristic device to represent concepts to oneself and to others. My use of the “paradigm offender” should be understood in a similar fashion. Appropriate persons—such as teachers of the criminal law—would employ the paradigm offender of a given offense to depict the clearest examples in which that offense had been perpetrated. See the discussion of the strengths and weaknesses of the uses of paradigm examples in Hanfling, Oswald “What is Wrong with the Paradigm Case Argument?” (1991) 91 Proceedings of the Aristotelian Soc. 21.CrossRefGoogle Scholar

23. Some theorists recognize that no actual offense satisfies their own definition of strict liability, but do not remedy this deficiency by providing a more satisfactory definition. Consider, for example, Abrams, Norman “Criminal Liability of Corporate Officers for Strict Liability Offenses-A Comment on Dotterweich and Park” (1981) 28 U.C.L.A. L. Rev. 463 note 1 at 463:Google Scholar “In this paper, I use the words “strict liability” to describe offenses for which there is liability without culpability. I then proceed to indicate different ways in which the courts manage to introduce the idea of culpability even for such offenses.” This description of the concept of strict liability is incoherent, as can be seen by asking to what the words “such offenses” refer. According to this account, a strict liability offense has two incompatible characteristics: it is an offense (1) “for which there is liability without culpability;” and (2) “in which the courts manage to introduce the idea of culpability.” See also Singer, supra note 16 at 391.

24. See, for example, Packer, Herbert The Limits of the Criminal Sanction (Stanford: Stanford University Press, 1968) at 13. See also Ashworth, supra note 6 at 83.Google Scholar

25. Thus my analysis cannot be expected to resolve all controversies about whether to categorize given offenses as examples of strict liability. I can only hope to explain why such controversies arise. Fortunately, criminal justice practitioners can live with the resultant uncertainty about how to classify a given offense. The concept of strict liability is important to the criminal theorist, but has little relevance to persons who earn their livelihood in the administration of criminal justice. No issue of practical significance depends on whether a given example should or should not be categorized as an instance of strict liability.

26. See supra notes 10 and 12, and accompanying text.

27. Some commentators question whether strict liability is always a property of offenses. One theorist distinguishes “strict liability offenses” from “strict liability rules or doctrines’. He defines a “strictliability doctrine” as “a rule of criminal responsibility that authorizes the conviction of a morally innocent actor.” Dressier, Joshua Understanding Criminal Law (New York: Matthew Bender, 1987) at 117.Google Scholar

As so defined, I do not believe that strict–liability doctrines actually exist. No rule of criminal responsibility authorizes the conviction of a defendant who is “morally innocent,” that is, totally without fault. Some rules, however, allow the conviction of a defendant who is less at fault than persons in the relevant comparison class. The felony-murder doctrine, for example, does not result in the conviction of a defendant who is completely blameless, but rather results in the conviction of a person who is substantially less at fault than paradigm murderers. On my analysis, the application of such doctrines to some offenses (e.g., murder) but not to others allows the former offenses to be categorized as (one kind of) a strict liability offense. Those offenses to which these rules and doctrines apply might be said to be strict liability offenses because they are subject to these rules and doctrines.

28. For a defense of a desert-based theory of punishment, see von Hirsch, Andrew Censure and Sanction (New York: Oxford University Press, 1993).Google Scholar

29. The U.S. Sentencing Commission supposes that each offense suggests a “heartland” (or what I call a “paradigm”) that merits a sentence of a given degree of severity to be located on a grid. Aggravating or mitigating factors that remove a given offense from this “heartland” can justify an upward or downward departure from the sentencing guidelines. However, the fact that the sentence of some offenders is mitigated or aggravated does not entail that the offense is an instance of strict liability. An offense is not an instance of strict liability unless it allows a defendant to be convicted even though he is substantially less at fault than the perpetrator of the “heartland” offense.

30. The most common oversimplification is to treat fault and mens rea as interchangeable and synonymous. According to this oversimplification, all inquiries about the fault of the defendant are exhausted by determining whether she has or lacks mens rea. A broader understanding of fault expands the inquiry to include all questions of whether and to what extent the defendant deserves to be blamed and punished. See Gardner, supra note 18.

For a discussion of many of the complexities of the role of fault in the criminal law, see the symposium in (1994) 5 J. of Contemporary Legal Issues 1.

31. See some of the examples discussed in Paul Robinson, “Imputed Criminal Liability” (1984) 93 Yale L. J. 609. The felony-murder and misdemeanor–manslaughter rules are perhaps the most obvious devices that allow liability to be imposed on a defendant who is substantially less at fault than the paradigmatic perpetrator of murder and manslaughter. I do not examine these devices here because of their application to single offenses. See supra note 27.

32. Of course, the distinction between substance and procedure is notoriously elusive. To the extent that this distinction is difficult to draw, the contrast between strict substantive liability and strict procedural liability will be difficult to draw as well.

33. In re Winship, 397 U.S. 358 (1970).

34. It seems obvious that some of the facts necessary to constitute a given crime pertain to fault, and others do not. This distinction is notoriously difficult to draw, and I make no systematic effort to do so here. Yet no one should doubt the existence of the distinction itself. Such a distinction is presupposed by the Model Penal Code’s contrast between “elements of an offense” and “material elements of an offense.“ See Model Penal Code sees. 1.13(9) and 1.13.(10).

This distinction is established most easily by examples. A requirement of mens rea is perhaps the clearest example of a matter that is relevant to fault, and a jurisdictional requirement is perhaps the clearest example of a matter that is irrelevant to fault. Several of the issues that pertain to the fault of a defendant are explored in the subsequent treatment of various kinds of strict substantive liability in Sections IV–IX infra.

35. Courts and commentators rarely question the requirement that guilt be proved beyond a reasonable doubt. But see Reiman, Jeffrey & van den Haag, Ernest “On the Common Saying that it is Better that Ten Guilty Persons Escape than that One Innocent Suffer: Pro and Con ” in Paul, Ellen, Miller, Fred & Paul, Jeffrey eds, Crime, Culpability, and Remedy (Oxford: Basil Blackwell, 1990) at 226.Google Scholar

36. Some commentators have proposed strict procedural liability as a “compromise” (or as a “halfway house”) between the positions of those who favor and those who disfavor strict substantive liability. See Levenson, supra note 1 at 452. Another commentator remarks that strict procedural liability “falls midway between imposing liability without fault and dealing with the statute just as if words of fault were written therein.” See LaFave, Wayne & Scott, Austin 2 Substantive Criminal Law (St. Paul: West Publishing Co., 1986) at 345.Google Scholar But another commentator observes that the "complexities introduced by changes in the burden of proof give rise to “dispute about whether [such offences] are properly termed “strict liability” offences.” Ashworth, supra note 6 at 136–37.

37. My categorization of strict procedural liability as a kind of strict liability is not unusual. According to one commentator, “strict liability offenses should be viewed as procedural mechanisms—rebuttable presumptions that relieve the prosecution from proving mens rea in its case-in-chief.” Levenson, supra note 1 at 449–50.

38. The mistaken supposition that such offenses do exist may result from equating fault with mens rea. Some offenses dispense with mens rea altogether, or create what might be called a conclusive (or irrebuttable) presumption of mens rea. See Section IV infra. But no offense creates a conclusive presumption of each of the myriad kinds of fault that is relevant to criminal liability. No offense, for example, creates a conclusive presumption that the defendant’s conduct is unjustified, and that the defendant lacks an excuse, and that the defendant’s conduct includes a voluntary act, and the like.

39. Patterson v. New York, 432 U.S. 197 (1977). This interpretation allows states to convict defendants who plead insanity without requiring that prosecutors prove sanity beyond a reasonable doubt. See Leland v. Oregon, 343 U.S. 790 (1952); and Rivera v. Delaware, 429 U.S. 877 (1976).

40. New York Laws, sec. 125.20(2) (1965).

41. Patterson, supra note 39.

42. See Note, “Winship on Rough Waters: The Erosion of the Reasonable Doubt Standard” (1993) 106 Harv. L. Rev. 1093.

43. 42 Pa. Const. Stat., sec. 9712(a) (1982).

44. McMillan v. Pennsylvania, 477 U.S. 79 (1986).

45. “Strict liability offenses”…are viewed most often as crimes for which liability is imposed…without a mens rea requirement.” Levenson, supra note 1 at 417. See also Dressier, supra note 27 at 117; Hart, H.L.A. Punishment and Responsibility (New York: Oxford University Press, 1968) at 90;Google Scholar and Hall, supra note 7 at 325.

46. Model Penal Code sec. 2.02.

47. Ibid. sec. 2.02(d), and Commentaries at 240.

48. One leading commentator defines “strict liability to mean liability imposed for an act or omission in violation of the law, without considering at trial whether the defendant may exculpate himself by proving a mistake or accident bearing on the wrongfulness of his violation.” Fletcher, George Rethinking Criminal Law (Boston: Little, Brown & Co., 1978) at 716.Google Scholar

49. Typically, the debate about whether liability for negligence is a form of strict liability takes place indirectly, under the guise of whether negligence is a kind of mens rea. If strict liability is liability without mens rea, and mens rea excludes negligence, then liability for negligence is a kind of strict liability. See Hall, supra note 7 at 133–41.

Sometimes, however, the conclusion that liability for negligence is a kind of strict liability is explicitly affirmed by definition: “I shall use the terms “strict liability offence” and “negligence offence” interchangeably.” Brudner, Alan “Imprisonment and Strict Liability” (1990) 40 Univ. of Tor. L.J. 738 note 5 at 738.CrossRefGoogle Scholar

50. But see also Section IX infra.

51. See Section V infra.

52. See Section VI infra.

53. The number of such offenses may grow, however, as courts and legislators respond to public anxiety about crime. According to one commentator, “no limit appears to exist on the crimes for which a strict liability solution strike some as justified.” See Coffee, John “Does “Unlawful” Mean “Criminal”?: Reflections on the Disappearing Tort/Crime Distinction in American Law” (1991) 71 Boston Univ. L. Rev. 193 note 106 at 219.Google Scholar

54. See Morissette v. U.S., 342 U.S. 246 (1952).

55. Model Penal Code sec. 213.2

56. Ibid., sec. 213.6.

57. See the rationale invoked in Regina v. Prince, [1874–80] All E.R. 881.

58. Courts and commentators have struggled, however, to clearly identify the class of offenses which possess this rare characteristic. One familiar attempt invokes the problematic distinction between offenses mala in se and offenses mala prohibita. See City of Seattle v. Gordon, 342 P.2d 604 (Wash., 1959).

59. See the account of how the meaning of mens rea has evolved from “moral guilt” to a “much narrower and more confining approach” in Brett, Peter “Strict Responsibility: Possible Solutions” (1974) 37 Modern L. Rev. 417 at 419.CrossRefGoogle Scholar See also Gardner, supra note 18.

60. Some commentators may insist on reserving the generic term “strict liability” to refer to what I take to be merely one kind of strict liability. If so, my project can be reconceptualized as an exploration of the various ways that liability can be imposed on a defendant who is substantially less at fault than the typical perpetrator of that offense. This project would be facilitated by devising a name to refer to these kinds of liability. I use “strict liability” for this purpose. Ultimately, however, little is gained by debating whether an altogether new term should be coined to serve this useful purpose.

61. Article 3 of the Model Penal Code provides for a number of justificatory principles. See the various “justification defenses” listed by Robinson, Paul 2 Criminal Law Defenses (St. Paul: West Pub. Co., 1984) ch. 4.Google Scholar

62. The conduct of a justified defendant is better construed as permissible than as commendable. See Husak, Douglas “Justifications and the Criminal Liability of Accessories” (1989) 80 J. of Crim. L. and Criminology 491.CrossRefGoogle Scholar

63. There are better reasons for concluding that a defendant who is justified has committed a criminal offense for which he is not liable than for concluding that he has not committed a criminal offense at all. See Husak, Douglas Philosophy of Criminal Law (Totowa, NJ: Rowman & Littlefield, 1987) at 190–92.Google Scholar

64. The principle that a defendant who has done nothing wrong must be substantially less at fault than the typical perpetrator of that offense seems easy to apply in most cases. But some theorists argue that there is an exception to this principle. Consider a case in which the objective facts give rise to a justification, but the defendant is unaware of these facts. Such a defendant may have done nothing wrong, and yet be just as much at fault as the typical perpetrator of his offense. Arguably, a defendant cannot be justified in committing a criminal offense unless he is subjectively aware of those facts that give rise to his justification. See Robinson, Paul “A Theory of Justification: Societal Harm as a Prerequisite for Criminal Liability” (1975) 23 U.C.L.A. L. Rev. 266.Google Scholar For a contrary view, see Fletcher, George “The Right Deed for the Wrong Reason” (1975) 23 U.C.L.A. L. Rev. 293.Google Scholar

65. See Model Penal Code sec. 3.02(l)(a) (called “Justification Generally: Choice of Evils”).

66. The most well-known case is Regina v. Dudley & Stephens (1884), 14 Q.B.D. 273.

67. Thus the Model Penal Code rejects this common law limitation, holding that “it would be particularly unfortunate to exclude homicidal conduct from the scope of the defense [of necessity].” Model Penal Code, Commentaries, sec. 3.02 at 14 (1985).

68. The alternative to recognizing necessity as a justification for homicide is to recognize it as an excuse. See von Hirsch, Andrew “Lifeboat Law” (1985) 3 Crim. Justice Ethics 88 (1985).CrossRefGoogle Scholar But if necessity does not justify such homicides, then the acts of persons who kill under necessity is just as wrongful as the acts of persons who kill for more typical reasons.

69. Buckoke v. General London Council [1971], 2 All E.R. 254 at 258.

70. This example and its implications for views about the nature of strict liability are described in Fletcher, George “Book Review: Corrective Justice for Moderns” 1993 106 Harv. L. Rev. 1658 at 1663.CrossRefGoogle Scholar

71. Model Penal Code sec. 2.1 l(2)(a) allows consent as a defense for conduct that causes bodily injury only if such injury is not serious.

72. At least one commentator, however, apparently defines strict liability (in tort) in terms of the absence of a justification: “We can then define strict liability in torts as liability that is defeasible by neither excuses nor justifications.” Coleman, supra note 3 at 220 (emphasis in original).

73. Article 4 of the Model Penal Code provides for a number of defenses predicated on the lack of responsibility. See the various “excuse defenses” listed by Robinson, supra note 61 ch. 5.

74. This characterization of excuses is defended in Husak, Douglas “The Serial View of Criminal Law Defenses” (1992) 3 Crim. L. Forum 369.CrossRefGoogle Scholar

75. “The excuses to which the law turns a deaf ear when there is strict liability are all claims of mistake or accident.” Gross, supra note 13 at 343.

76. “Presumably no court would convict a psychotic person…of any offense.” Hall, supra note 7 at 342. But some theorists propose that the insanity defense should excuse persons for some offenses but not for others. See Wexler, David “After Hinckley: New Directions for the Insanity Defense” in Feinberg, Joel & Gross, Hyman eds, Philosophy of Law 3d ed. (Belmont, CA: Wadsworth Pub. Co., 1986) at 578.Google ScholarPubMed

77. See Model Penal Code sec. 2.09.

78. Prior to the writing of the Model Penal Code, most state statutes governing duress did not allow the defense for “most serious crimes.” See Model Penal Code sec. 2.09, Commentaries at 368.

79. Many (but not all) jurisdictions allowed duress to reduce the culpability of an illegal homicide from murder to manslaughter. See Dressier, supra note 27 at 264. But the use of this “imperfect” defense is flawed. The person who deliberately kills under duress exhibits not only substantially less fault than the paradigm murderer, but also substantially less fault than the paradigm perpetrator of manslaughter.

80. But not always. Occasionally this excuse is recognized despite the fact that it is not explicitly authorized by statute. See Lambert v. California, 355 U.S. 225 (1957).

81. Model Penal Code sec. 212.3.

82. The culpability provision in this statute applies to all material elements pursuant to ibid. sec. 2.02(4).

83. The second assumption is defended in Husak, Douglas & von Hirsch, Andrew “Culpability and Mistake of Law” in Shute, Stephen, Gardner, John, & Horder, Jeremy eds, Action and Value in Criminal Law (Oxford: Clarendon Press, 1993) at 157.Google Scholar

84. My claim—that disregarding a defense of ignorance of law for a given offense renders liability for that offense more strict—is novel. Still, it is not entirely unprecedented. See Fletcher, supra note 48 at 730–31: ‘The maxim that ignorance of the law is no excuse is so well entrenched in many legal systems that one is not likely to think of this form of mistake as a factor bearing on culpability… [Yet there is no denying that in such a case] the court imposes liability regardless of the actor’s culpability in violating the statute.” See also Johnson, Phillip in Kadish, Sanford ed., “Strict Liability: The Prevalent View” 4 Encyclopedia of Crime and Justice (New York: Free Press, 1983) at 1521:Google Scholar “There is an element of what might be called strict liability throughout the criminal law, because of the doctrine that ignorance of the law is no excuse.”

85. If strict liability is defined as liability that does not require mens rea, it is easy to see why liability despite ignorance of law would not be regarded as a kind of strict liability. Ignorance of law is not generally thought to preclude mens rea. See LaFave & Scott, supra note 36 at 340.

86. An exception to this generalization is that liability for this offense is not defeasible by the excuse of ignorance of law.

87. See Hart, supra note 45.

88. At least one commentator, however, apparently defines strict liability (in tort) in terms of the absence of excuses: “We can then define strict liability in torts as liability that is defeasible by neither excuses nor justifications.” Coleman, supra note 3 at 220 (emphasis in original). See also the more general definition of strict liability as liability that is not defeasible by excuses: “Liability is strict in the sense that it does not depend upon mental states or character defects and does not allow excuses.” Murphy, Jeffrie & Coleman, Jules Philosophy of Law (Totowa, NJ: Rowman & Allanheld, 1984) at 132.Google Scholar

89. See 1 LaFave & Scott, supra note 36 at 353.

90. I will not attempt to defend this judgment of relative fault, although such judgments are necessary if my account (or, indeed, any account) of strict liability is to be applied. See Section II supra.

91. I do not discuss whether D2 might be guilty of some other offense for which his liability would not be vicarious.

92. See U.S. v. State, 535 F.2d 512 (9th Cir., 1976).

93. For a contrary view, see Singer, supra note 16 at 355: “In a case of vicarious liability, the issue is whether the defendant, although totally innocent of any wrongdoing (indeed of any actus at all), is to be held responsible for the acts of someone else who was in fact knowledgeable and hence morally culpable.”

94. A similar rationale explains the temptation, usually resisted, to hold parents vicariously liable for the criminal acts of their children.

95. Fault may not be involved in the principles used by some theorists to identify the defendant on whom tort liability should be imposed. Some theorists propose that liability should be imposed on the cheapest or best “cost avoider,” even though he seemingly bears no degree of fault for an injury. See Calabresi, Guido The Cost of Accidents (New Haven: Yale University Press, 1972).Google Scholar

96. Many moral philosophers have argued that the failure to intervene to prevent a harm is one way to merit blame for that harm. See, for example, Harris, John, “The Marxist Conception of Violence” (1974) 3 Phil. & Publ. Affairs 192.Google Scholar

97. 320 U.S. 277 (1943).

98. 421 U.S. 658 (1975).

99. LaFave and Scott describe the decision in Dottenveich as “outrageous” because they claim that it “upheld the conviction of the president of a company that shipped misbranded or adulterated products in interstate commerce notwithstanding the fact that he had nothing to do with the ship-ment.” Supra note 36 at 355. This decision would indeed by outrageous if the defendant had “nothing to do” with the harm. However, the defendant was not a total stranger with no relationship to the employees who shipped the adulterated products, but a corporate officer with supervisory authority over them.

100. 320 U.S. at 281 (1943).

101. 421 U.S. at 670 (1975).

102. Ibid. at 674.

103. Ibid. at 673.

104. The “concept of a responsible share…imports some measure of blameworthiness.” Ibid. at 673.

105. At least not in Anglo-American law. Applications of such a theory, however, are not unprecedented. Vicarious liability enabled Nazis to effectively occupy territories in conquered states. Any offense against the Nazis was punished severely, with little regard to whether the person punished had actually committed the offense. The use of such vicarious liability, although effective in ensuring compliance with rules, is surely unjust.

106. “Vicarious liability…amounts to strict liability.” See Bayles, Michael Principles of Law (Dordrecht: D. Reidel Pub. Co., 1987) at 236. LaFave and Scott treat “strict and vicarious liability” as “two types of liability without fault.” See supra note 36 at 352.Google Scholar

107. See Patient, Ingrid “Some Remarks About the Element of Voluntariness in Offences of Absolute Liability” (1968) Crim. L. Rev. 23.Google Scholar

108. I will follow most commentators and discuss the nonvoluntariness defense in the context of offenses that do not include mens rea. Of course, I do not follow most commentators in believing that the nonvoluntariness defense has significance only in the context of strict liability offenses. On my account, strict liability offenses should not be defined as offenses that dispense with mens rea; the latter offenses are merely one of several varieties of strict liability.

109. See Husak, supra note 63 ch. 4.

110. For an alternative rationale for allowing excuses that defeat agency to preclude liability in tort, see Coleman, supra note 3 at 261–66.

111. See the examples of bodily movements that are not voluntary acts in Model Penal Code, sec. 2.01(2)(a), (b), and (c).

112. Model Penal Code, sec. 2.01(2)(d) includes a “catch-all” category of “a bodily movement that otherwise is not a product of the effort or determination of the actor, either conscious or habitual.”

113. Model Penal Code, sec. 2.01(1)

114. 571 P.2d 65 (1977).

115. See the cases cited in the discussion of “the grand schemer” in Robinson, supra note 61 at 39.

116. Some theorists claim there to be some sort of inconsistency in basing liability on fault for offenses said to be cases of strict liability. See Alexander, Larry “Reconsidering the Relationship Among Voluntary Acts, Strict Liability, and Negligence in Criminal Law” in Paul, , Miller, , & Paul, , eds, supra note 35 at 84.Google Scholar But see Husak, Douglas & McLaughlin, Brian “Time Frames, Voluntary Acts, and Strict Liability” (1993) 12 L. and Phil. 95.CrossRefGoogle Scholar

117. For a more skeptical view, see Kelman, Mark A Guide to Critical Legal Studies (Cambridge: Harvard University Press, 1987) at 92–93.Google Scholar

118. See Alexander, supra note 116; and Kelman, ibid.

119. The distinction between act-types and act-tokens is developed by Peirce, C.S. Collected Works (New York, 1935) at 537.Google Scholar

120. The thesis that the criminal law should be used only to punish persons for harmful conduct is typically attributed to John Stuart Mill. This thesis is amended, embellished, and defended most ably by Feinberg, Joel Harm to Others (New York: Oxford University Press, 1985).Google Scholar Feinberg argues that a person harms another in the relevant sense when his conduct is a violation of a right.

121. Justifications are typically interpreted to require a greater good. But see Husak, supra note 62.

122. “Crimes of possession are “inchoate” or incomplete offenses… Their real purpose is to provide law enforcement officials with a legal mechanism by which to stop a person from performing some later, more dangerous act.” Dressier, supra note 27 at 72.

123. Narcotics use may or may not be the consummate harm that the anticipatory offense of narcotics possession is designed to prevent. Perhaps narcotics use is proscribed because of its causal link with property offenses. Unfortunately, it is not always clear what consummate harm(s) an anticipatory offense is designed to prevent. See Husak, Douglas Drugs and Rights (New York: Cambridge University Press, 1992) at 178–95.CrossRefGoogle Scholar

124. See People v. Mijares, 491 P.2d 1115 (Cal. 1971).

125. My “innocent purpose” category is somewhat different from the category of defense Robinson describes as an “offense modification.” According to Robinson, the defendant has an “offense modification defense” when he “has apparently satisfied all elements of the offense charged, [although] he has not in fact caused the harm or evil sought to be prevented by the statute defining the offense.” Robinson, supra note 61 at 77. He claims that this “defense would seem to be theoretically applicable to all offenses.” Ibid. at 77. By contrast, I do not see how the offense of arson, for example, could be committed with an innocent purpose (although it could clearly be justifiable). The “innocent purpose” category has its central application to nonconsummate, inchoate or anticipatory offenses. Perhaps Robinson disagrees because he includes the de minimis infraction defense as an example of an offense modification defense. Ibid. at 79.

126. Some of the problems that arise in attempts to justify the creation and enforcement of such offenses are discussed in Husak, Douglas “The Nature and Justifiability of Nonconsummate Offenses” (1995) 37 Arizona L. Rev. 151.Google Scholar

127. S. 1437, 95th Congress, 1st Sess. Sec. 1327(b) (1977).

128. Model Penal Code, sec. 2.01(4) defines possession so that “the possessor knowingly procured or received the thing possessed or was aware of his control thereof for a sufficient period to have been able to terminate his possession.” The Commentaries explain that this definition is designed to ensure that possessory offenses satisfy the requirement of an act or omission. In addition, this definition goes some distance toward ensuring that defendants will not be held liable for possessory offenses despite an innocent purpose. But the definition does not solve all such problems. Some defendants who illegally possess narcotics may have an innocent purpose even though they have had a “sufficient period” in which to “terminate” their possession.

129. Without a catalogue of harms that the law seeks to prevent, the very distinction between inchoate and consummated offenses cannot be drawn. See Fletcher, supra note 48 at 133.

130. See supra note 45.

131. See Section II supra.

132. Model Penal Code, sec. 2.12(2) provides a defense when “the defendant’s conduct…did not actually cause or threaten the harm or evil sought to be prevented by the law defining the offense.” Of course, problems in identifying the “harm or evil sought to be prevented” make this defense difficult to apply.

133. 133. Even at this juncture, however, I am sure to have overlooked several means by which the criminal law ensures that liability is imposed only on defendants who are at fault. At the very least, a hypothetical example of an offense for which liability could be imposed on a totally faultless defendant presupposes some controversial views about the nature of fault. Suppose, for example, that fault is collective, so that each person is partly responsible for all the wrongdoing in the world. On this assumption, it would be impossible for liability (for any act of wrongdoing) to dispense with fault altogether. Any person on whom liability was imposed would be partly at fault simply in virtue of being a member of the collective.

134. Justifications, for example, are thought to arise when the harm sought to be prevented by the statute is outweighed. If there were no harm to outweigh, it is hard to see how an offense could be justified. Excuses, moreover, are thought to arise when persons are not blameworthy for their conduct. If the conduct is not wrongful, questions of blame do not arise.