Bringing reason to a messy doctrine: The distinction between mistake of law and mistake of fact

ABSTRACT

This article is concerned with the distinction between mistake of law and mistake of fact in criminal law defences. The article criticises the traditional view that distinguishes between mistake of law and mistake of fact by the criterion of perceptibility. The authors argue that the traditional view is over-inclusive and, if applied consistently, would assimilate all cases of mistake into the category of mistake of fact. The modern ‘prohibition-centred’ approach is also criticised, since it suffers from ambiguity as to what counts as ignorance of, or a mistake about the criminal prohibition or its meaning. Following a critical discussion of the rationales for the differential treatment of mistake of law, the authors propose and justify an alternative test that is based on the rationale of promoting compliance with the law. The alternative test is a two-stage test. First, the criminal prohibition must be identified as the conduct rule that derives from all signs that: (a) are made primarily for the purpose of changing the legal situation; and (b) are addressed by the state or its officials to an indefinite group of persons. Second, the two types of mistake must be distinguished as follows: (a) a mistake of law is a failure to infer that the conduct rule, which was identified in the first stage, applies to a given set of facts, where that inference can be made solely on the basis of the meaning of the conduct rule and the facts determining its temporal and territorial effect; and (b) a mistake of fact is a mistake in regard to the existence (or a lack of awareness of the existence) of the set of facts.

Notes

1 In most common law jurisdictions, there is no general defence to criminal responsibility in the case of a mistake of law. In those jurisdictions, the applicable rule remains, in most cases, the traditional common law rule that ‘ignorance of the law is no excuse’. Where mistake of law has been recognised as a defence to criminal responsibility, it has generally been casuistic and limited, eg as the result of a finding that the mental element of a particular offence entails knowledge of the law: Cheek v United States 498 US 192 (1991) (US Supreme Court) (concerning lack of knowledge in regard to a tax offence). It may be recognized in a manner along the lines proposed in § 2.04 of the Model Penal Code (MPC), which lists instances in which it would be proper to recognise mistake of law as a defence, such as a mistake of law deriving from official advice: American Law Institute, Model Penal Code and Commentaries: Official Draft and Revised Comments, (American Law Institute 1985).

While the situation in Continental law is different, even there mistake of law is treated more severely than mistake of fact. For example, § 17 of the German Strafgesetzbuch states:

If during the commission of an act the perpetrator lacks the appreciation that he is doing something wrong, he acts without guilt if he was unable to avoid the mistake. If the perpetrator could have avoided the mistake, the punishment may be mitigated pursuant to § 49(1).

See Stephen Thaman (tr), The German Penal Code as Amended as of December 19, 2001 (William S Hein 2002) 8. See also Heribert Schumann, ‘Criminal Law’ in J Zekoll and M Reimann (eds), Introduction to German Law (2nd edn, Kluwer Law International 2005) 387, 399–400.

In France, the new Penal Code that was enacted in 1992 and came into force in 1994 incorporated a general defence of mistake of law that is deemed one of the new code's most far-reaching deviations from its predecessor. Under Article 122–3:

A person is not criminally liable who proves that he believed, because of a mistake of law which he was not in a position to avoid, that he could legitimately carry out the act.

See Catherine Elliott, French Criminal Law (Willan 2001) 124.

For a broad, if somewhat outdated, comparative survey of the situation in various countries see Paul K Ryu and Helen Silving, ‘Error Juris: A Comparative Study’ (1957) 24 University of Chicago Law Review 421, 435–39, and their update, Paul K Ryu and Helen Silving, ‘Comment on Error Juris’ (1976) 24 American Journal of Comparative Law 689.

2 For the most part, although there is a theoretical difference between not knowing and erring, in the context of criminal law defences the consequences of mistake and lack of knowledge are the same. See for example, Rollin M Perkins and Ronald N Boyce, Criminal Law (3rd edn, Foundation Press 1982) 1028. Accordingly, the term ‘mistake’ is used in this article to refer both to mistake in the narrow sense and to lack of knowledge.

However, it should be noted that it has been suggested that mistake of law and ignorance of law should attract different consequences. See for example, Edwin R Keedy, ‘Ignorance and Mistake in the Criminal Law’ (1908) 22 Harvard Law Review 75, 88–96 and IHE Patient, ‘Mistake of law—A Mistake?’ (1987) 51 Journal of Criminal Law 326, 326–28. Kumaralingam Amirthalingam also suggests attaching different consequences to ignorance and mistake of law, but frames the distinction between the concepts differently: see Kumaralingam Amirthalingam, ‘Distinguishing Between Ignorance and Mistake in the Criminal Law in Defence of the De Blom Principle’ (1995) 8 South African Journal of Criminal Justice 12, 15–16. And cf Douglas Husak and Andrew von Hirsch, ‘Culpability and Mistake of Law’ in Stephen Shute, John Gardner and Jeremy Horder (eds), Action and Value in Criminal Law (Clarendon 1993) 157, 157 fn 1, referring to a distinction between ignorance and mistake similar to that proposed by Amirthalingam.

3 We do not think it necessary in this article to consider the question of what precisely is ‘law’. We do note that it is possible to think of at least two different meanings of ‘law’ as the possible object of a mistake. One is law in the descriptive sense. According to this meaning, the answer to the question of what is the law will depend on the manner in which the authorised institutions of a given legal system (eg the courts) decide a given legal question. The second meaning connotes law in the normative sense. This is the sense in which we employ the term ‘law’ when we refer to it as something that is meant to direct the authorised institutions in deciding a given legal question.

The question, thus, is whether the object of the mistake of law in criminal law is law in the normative sense or law in the descriptive sense, or perhaps some combination of the two. In our view, the more reasonable possibility is that the subject of the mistake of law is the law in its descriptive sense. A mistake in regard to law in the normative sense seems less relevant, to the question of guilt (and, of course, the possibility of a mistake in this sense, as opposed to the descriptive sense, is contingent on the position of the chosen jurisprudence, particularly if we accept the single legal answer thesis).

4 A fact in the broad sense is the basis upon which a proposition is deemed true or false. Note, however, that this is not limited to the correspondence theory of truth. Even if we adopt the coherence theory of truth, determining the truth or falsity of a particular statement still requires a criterion for truth, which in this case is coherency. Cf Duncan Sheehan, ‘What is a Mistake?’ (2000) 20 Legal Studies 538, 540, pointing out that the concept of mistake has no meaning without a corresponding concept of truth. There is, therefore, no linguistic reason not to call something that satisfies such a truth criterion a fact.

5 For the distinction between types and tokens see Linda Wetzel, ‘Types and Tokens’ (The Stanford Encyclopedia of Philosophy, Spring edn, 2014) <http://plato.stanford.edu/archives/spr2014/entries/types-tokens> accessed 30 October 2015.

6 Larry Alexander, ‘Inculpatory and Exculpatory Mistakes and the Fact/Law Distinction: An Essay in Memory of Myke Balyes’ (1993) 12 Law and Philosophy 33.

7 See Kenneth W Simons, ‘Mistake of Fact or Mistake of Criminal Law? Explaining and Defending the Distinction’ (2009) 3 Criminal Law and Philosophy 213, 223–34 for a discussion of various borderline cases.

8 (n 1).

9 See Ronald Dworkin, Law's Empire (Harvard University Press 1986) 52–53.

10 See for example Ronald J Allen and Michael S Pardo, ‘The Myth of the Law–Fact Distinction’ (2003) 97 Northwestern University Law Review 1769.

11 Simons, ‘Mistake of Fact or Mistake of Criminal Law’ (n 7) 219.

12 Various legal systems and commentators make a distinction between two categories of mistake of law based on the object of the mistake. The distinction can be characterised generally as a distinction between mistakes concerning the law governing the offence (ie the law that creates the offence or the defences thereto) and mistakes concerning non-governing law (eg mistake as to property law or as to marriage law), although different other terms have been used to denominate the same distinction. According to this doctrine, mistake of a non-governing law is treated like a mistake of fact. See Kenneth W Simons, ‘Mistake and Impossibility, Law and Fact, and Culpability: A Speculative Essay’ (1990) 81 Journal of Criminal Law and Criminology 447, 456–59. See also Kenneth W Simons, ‘Ignorance and Mistake of Criminal Law, Noncriminal Law, and Fact’ (2012) 9 Ohio State Journal of Criminal Law 487; George P Fletcher, Rethinking Criminal Law (Little, Brown 1978) 739–41; Rollin M. Perkins ‘Ignorance or Mistake of Law Revisited'1980 Utah Law Review 473, 475–76

13 This is merely a definition. Where the doctrine of mistake of non-governing law applies, there is nothing to prevent defining mistake of fact to include a mistake of non-governing law (as the rule is identical), while defining mistake of law as comprising only mistakes in regard to the governing law. In such a case, the definition that we suggest will be incomplete and would have to be supplemented by a delineation of the boundary separating a mistake of governing law and a mistake of non-governing law.

14 Glanville Williams, Criminal Law: The General Part (2nd edn, Stevens 1961) 287.

15 Simons, ‘Mistake and Impossibility, Law and Fact, and Culpability: A Speculative Essay’ (n 12) 469.

16 Indeed in a later article Simons seems to acknowledge the problems with his former formulation of the distinction and revises it by eliminating the reference to perception or empirical judgments: Simons, ‘Mistake of Fact or Mistake of Criminal Law’ (n 7) 222. His later definition of mistake of fact is residual:

[W]e should, when drawing the distinction, begin with an account of ignorance or mistake of law, and then treat all other claims that ignorance or mistake is relevant to criminal liability as involving ignorance or mistake of (nonlegal) fact.

17 Alexander (n 6) 37–38.

18 cf Peter Tillers, ‘The Value of Evidence in Law’ (1988) 39 Northern Ireland Legal Quarterly 167.

19 John R Searle, Speech Acts: An Essay in the Philosophy of Language (Cambridge University Press 1969) 51.

20 Indeed legal facts themselves are institutional facts. See Neil MacCormick, ‘Law as Institutional Fact’ in Neil MacCormick and Ota Weinberger (eds), An Institutional Theory of Law: New Approaches to Legal Positivism (D Reidel Publishing Co 1986) 49, 51.

21 See for example the more recent definition in Simons, ‘Mistake of Fact or Mistake of Criminal Law’ (n 7) 220:

The fundamental distinction is between:(1) M Law: a mistake about what the state prohibits (including a mistake about how state officials, including judges, authoritatively interpret the prohibition); and

(2) M Fact: a mistake about the instantiation of that prohibitory norm in a particular case, where the mistake does not flow from the first type of mistake.

Similarly, Gerald Leonard's approach can be classified as a prohibition-centred approach: see Gerald Leonard, ‘Rape, Murder, and Formalism: What Happens If We Define Mistake of Law?’ (2001) 72 University of Colorado Law Review 507, 527–29. The prohibition-centred approach is essentially the same as the approach adopted by the MPC (n 1) in § 2.02(9) which states:

Neither knowledge nor recklessness or negligence as to whether conduct constitutes an offense or as to the existence, meaning or application of the law determining the elements of an offense is an element of such offense, unless the definition of the offense or the Code so provides.

The main difference is that § 2.02(9) of the MPC seems to deal only with mistakes of governing law: cf Simons, ‘Ignorance and Mistake of Criminal Law, Noncriminal Law, and Fact’ (n 12) 497.

22 The hypothetical was originally proposed in Alexander (n 6) 50. We changed the facts to make it a case of exculpatory mistake.

23 This is the way Westen defines the criminal prohibition: Westen (n 5) 535, fn 34.

24 This is the way Simons defines the criminal prohibition: Simons, ‘Mistake of Fact or Mistake of Criminal Law’ (n 7) 232–33, fn 48.

25 See the text after n 20.

26 The roots of this approach can be traced back as far as Aristotle's Nicomachean Ethics 1113b30–1114a3.

27 Husak and von Hirsch (n 2) 161–65. Other scholars suggest that a defendant who is mistaken about the law may be blameworthy because she breached a duty to know the law: Ronald A Cass, ‘Ignorance of the Law: A Maxim Reexamined’ (1976) 17 William and Mary Law Review 671, 692–93; Andrew Ashworth, Principles of Criminal Law (5th edn, OUP 2006) 233. This rationale is particularly problematic, inter alia for the reasons brought by Husak in his convincing criticism of this line of justification: Douglas N Husak, ‘Ignorance of Law and Duties of Citizenship’ (1994) 14 Legal Studies 105.

28 Jerome Hall, General Principles of Criminal Law (2nd edn, Bobbs-Merrill 1960) 382–87. A similar rationale is proposed by Boštjan M Zupančič, ‘Criminal Responsibility Under Mistake of Law: The Real Reasons’ (1985) 13 American Journal of Criminal Law 37, 57–58, who takes the view that the reason for the rule that ignorance of the law does not excuse is the interest in preserving the objectivity of the law.

29 Laurence D Houlgate, ‘Ignorantia Juris: A Plea For Justice’ (1967) 78 Ethics 32, 38–41; George P Fletcher (n 12) 733–35; Ryu and Silving ‘Error Juris’ (n 1) 433.

30 John Austin, Lectures on Jurisprudence or the Philosophy of Positive Law (Robert Campbell (ed), John Murray 1880) 238–39. It would appear that this rationale did not originate with Austin: see for example John Selden's comment on the rule that ignorance of the law does not excuse: ‘not that all men know the law, but because ‘tis an excuse every man will plead, and no man can tell how to confute him’: Table Talk, Law, quoted in Williams (n 14) 290; and John Selden, Table Talk (Sharpe 1821) 77; See also Paul H Robinson, Criminal Law Defenses, vol 2 (West Publishing 1984) 376, which supports this argument, adding that this is a defence that has no limit or abnormality that can be foreseen or refuted.

31 For criticism of this rationale, see Oliver Wendell Holmes, The Common Law (Harvard University Press 2009) 45.

32 ibid 46:

[T]o admit the excuse at all would be to encourage ignorance where the law-maker has determined to make men know and obey, and justice to the individual is rightly outweighed by the larger interests on the other side of the scales.

It should be emphasized that we understand Holmes as saying that the purpose of the rule that ‘ignorance of the law is no excuse’ is to encourage objective compliance with the law—reducing the incidence of external conduct that the legislature deems harmful and undesirable in terms of the public welfare—and not theoretical study of the law. See for example Cass (n 27) 689, who also seems to have understood Holmes in this manner. Cass writes simply: ‘The second major justification ( … ) advocated most vigorously by Holmes, is that adherence to the maxim deters crime’ (emphasis added).

Quite a few authors have understood Holmes incorrectly as establishing that the rationale underlying the rule is the encouragement of knowledge of the law. On the basis of this misunderstanding, Hall criticises Holmes, saying that the statement implies that knowing the criminal law is of value, when, in fact, it is irrelevant: Hall (n 28) 380–81. According to Hall, it is difficult to accept that the main purpose of the criminal law is to encourage legal education. Hall goes on to say that if we adopt Holmes's approach, we must find that a defendant who makes an effort to verify his legal obligations but violates them nonetheless, would be entitled to at least some mitigation of punishment since, after all, he did learn the law. But, of course, this is not the case. On the contrary, such circumstances are rather an aggravating factor.

As we have pointed out, Hall's criticism appears to be based upon a mistaken understanding of Holmes. Holmes should not be understood as establishing that the purpose of criminal law is to encourage the theoretical study of the law as an objective in and of itself. Had that been his intent, Hall's criticism would have been appropriate.

33 Dan M Kahan, ‘Ignorance of Law Is an Excuse—But Only for the Virtuous’ (1997) 96 Michigan Law Review 127, 133–35 argues:

In fact, if the goal were truly to maximize private knowledge of law, a negligence standard would be unambiguously superior to a strict liability standard. This is so because the value of learning the law is always higher when the law excuses reasonable mistakes of law than when it doesn't.

This criticism does not undermine our argument. As we said, it is enough that the rationale of encouraging compliance justifies a negligence standard for mistakes of law where mistakes of fact are held to a lower standard. Moreover, it seems that Kahan, too, misinterpreted the rationale of encouraging compliance. The purpose of the doctrine of mistake of law is not to encourage legal study. Kahan is right that strict liability would in some cases provide less of an incentive to learn the law than the negligence standard. Nevertheless, the strict liability standard might still encourage people who are in doubt to walk on the safe side and avoid potentially illegal activity, even if they choose not to ascertain the law because of the costs of learning the law.

34 See Ben-Zion Lahav, ‘I Yedia'at Ha-din’ (1988) 6 Mehkarei Mishpat 165, 180 (in Hebrew).

35 cf Rene Descartes, Meditations on First Philosophy- With Selection from the Objections and Replies (John Cottingham (tr), Cambridge University Press 2013) 39–41; Bertrand Russell, The Problems of Philosophy (2nd edn, OUP 2001) 19.

36 Assuming that we construe consent in accordance with its normal meaning as a mental fact.

37 Indeed some authors have argued that negligence liability can be justified at least in certain cases on retributive grounds: see for example Kenneth W Simons, ‘Culpability and Retributive Theory: The Problem of Criminal Negligence’ (1994) 5 Journal of Contemporary Legal Issues 365; Peter Arenella, ‘Character, Choice, and Moral Agency: The Relevance of Character to our Moral Culpability Judgments’(1990) 7(2) Social Philosophy and Policy 59, 74–75. However, unless one subscribes to a strict retributivist position (according to which moral desert is not only a necessary condition for punishment but also a sufficient reason for it), then the retributive considerations should be accompanied by a consequentialist justification for punishment in negligence cases.

38 Miriam Gur-Arye, ‘Reliance on a Lawyer's Mistaken Advice—Should It Be an Excuse from Criminal Liability?’ (2002) 29 American Journal of Criminal Law 455, 463–64.

39 As regards the social conventions about consent to sexual relations, see Douglas N Husak and George C Thomas III, ‘Date Rape, Social Convention, and Reasonable Mistakes’ (1992) 11 Law and Philosophy 95, 102–103. As Husak and Thomas point out, social conventions, such as those concerning how consent to sexual relations is expressed, constitute a fact that must be empirically determined, and are subject to change over time. A person who errs in regard to such social conventions is, indeed, mistaken in regard to an empirical fact, but this is a behavioural generalisation, and therefore his ‘scope of freedom’ is quite wide.

40 For the purpose of this hypothetical case we assumed that the meaning given to consent in the framework of rape is that of some mental fact. Accordingly, its external expressions, or the lack thereof, are merely an empirical indication of the existence of that fact. However, there is nothing to prevent giving the term ‘consent’ a different meaning in the context of the crime of rape. In this regard, see Lucinda Vandervort, ‘Mistake of Law and Sexual Assault: Consent and Mens Rea’ (1987) 2 Canadian Journal of Women and the Law 233, 267, proposing that consent to sexual relations be construed as ‘an act of communication between persons in the social world, and not the thoughts and mental state of the person who gives consent’.

41 It would not be out of place to make a brief comparison of the shared and different elements of the proposed approach and Welzel's final-conduct doctrine (finale Handlungslehre), ‘the goal-directed theory of human action’ that George Fletcher refers to as the ‘teleological theory’: Fletcher (n 12) 433–39. Welzel proposed reorganising the relationship between blameworthiness and mens rea. According to his approach, mens rea becomes part of the actus reus (that is, its subjective element), and is not part of the question of guilt. This conceptual approach allowed Welzel to separate the question of knowledge of the illegality, which relates to blameworthiness, from the question of knowledge of the factual elements of the offence, thus permitting a differential approach to the two (On the relationship between Weltzel's theory of human action and his theory of blameworthiness see Gunther Arzt, ‘Ignorance or Mistake of Law’ (1976) 24 American Journal of Comparative Law 646, 654–55.)

A certain similarity may be seen between Welzel's theory and the approach proposed in this article in regard to portraying human action as essentially purposeful. But there are also substantive differences. First, the reasoning is different. As we understand it, Welzel's approach was premised upon a conceptual examination of the concept of action. This way of reasoning was criticised, justifiably, in Helen Silving, Constituent Elements of Crime (Thomas 1967) 361–62, and see also J R du Plessis, ‘Hans Welzel's Final-Conduct Doctrine—An Importation from Germany We Could Well Do Without’ (1984) 101 South African Law Journal 301. We also think that justifying the differential treatment of mistake of law by conceptual arguments is mistaken. We are free to design our concepts in accordance with the rationale that we intend to base them upon, and we are not bound by them. While our approach is also based upon the teleological character of human action, we do not argue that this character is part of the concept of human action. When we argue that human action is teleological, we are just making a descriptive claim. Based on this descriptive claim we explain the substantive difference that justifies the differential treatment of mistake of law.

Second, Welzel's conceptual approach leads to an undesirable result. According to his approach, because mens rea is part of the criminal act, knowledge must relate only to the elements that constitute the definition of the offence. The result is that a mistake of fact in regard to something relevant to the defence to criminal liability (eg self-defence) must be evaluated according to an objective standard, similar to a mistake of law, since it, too, falls within the scope of guilt. We do not think this distinction is necessary. The teleological character of human action provides motivation for verifying the facts that form the basis for a defence to the same extent that it provides motivation for verifying facts that form the elements of the offence.

Lastly, negligence offences do not pose a problem for our approach, because the teleological nature of human action is not deemed a condition for its definition as an action, but merely points to a fact that is significant in our context.

42 By ‘moral motivation’ we mean the honest desire to obey the law, not out of fear of punishment but, rather, owing to a moral position that requires it. This desire may, of course, serve as a motivation to verify the legal situation in order to comply with the law.

43 By this we mean ordinary people as opposed to professionals (eg jurists) who have a practical motivation to learn the law for their work.

44 One might be able to challenge this, arguing that, at times, the defendant has no motivation to check contingent facts for fear of discovering facts that would prohibit his intended conduct. The answer to this is that as long as the defendant does not subjectively suspect in regard to a specific act (or category of acts) that checking might reveal contingent facts that would constitute an offence, he also has no reason to refrain from checking as dictated by his normal motivations in similar situations. In other words, he knows of nothing that would distinguish this situation from other, similar situations he has experienced, and will experience. He might adopt a general policy of refraining from factually checking in any case but, clearly, the practical significance of this is too severe. If, on the hand, he does harbour a subjective suspicion, or if (what is effectively the same) he is aware of information that distinguishes the specific case and categories it as one that poses a higher likelihood for the commission of a crime, then the matter could be resolved simply by recourse to such existing doctrines as wilful blindness.

45 See Meir Dan-Cohen, ‘Decision Rules and Conduct Rules: On Acoustic Separation in Criminal Law’ (1984) 97 Harvard Law Review 625.

46 It should be noted that aggravating circumstances are also part of the conduct rule since they are addressed to the public and give guidance as to degree of wrongdoing.

47 It is, nevertheless, important to note that even among the intra-legal motivations one can distinguish between the fear of punishment and other intra-legal motivations such as using the institutions offered by the legal system for one's own purposes (for instance by imposing legal duties on oneself or others by such institutions as contracts, property, etc). We are of the opinion that this distinction among the various types of intra-legal motivation is the basis for fixing the boundary between a mistake in regard to the governing law and a mistake in regard to non-governing law. In the context of this article, as explained in the introduction, we limited the scope of our examination to mistake of law, while including both mistake regarding the governing law and mistake regarding non-governing law under that rubric.

48 It is has been argued that a mistake as to the existence of a traffic sign is a simple example of a mistake of fact: see Lahav (n 34) 169–70. According to that view, a driver who does not see a traffic sign is no different from one who does not take note of a school or children standing alongside the road, which is a relevant fact in regard to the existence of an elevated duty of care. Thus, just as it cannot be said that the children in the latter example constitute the law, because law is ‘established’ by the ‘physical act’ of their walking alongside the road, the same is true for the traffic sign; and so a driver who fails to note either fact is not making a mistake of law.

This analogy is not persuasive. This should be clear from what has been explained above. There is a significant difference between the presence of children on the roadside and a traffic sign. The children exist independently, and they cannot be said to have a purpose of changing the law. Their normative consequences are merely incidental. The same would hold true if a driver were mistaken as to the presence of a school, which imposes a duty to drive slowly. The school has an independent purpose; it is not there solely in order to cause drivers to slow down. The same cannot be said of traffic signs. They are not independent, and their sole purpose is to change the legal situation. The intuitive tendency to categorise a traffic-sign mistake as a simple instance of mistake of fact derives from traffic signs being concrete physical objects. But, as demonstrated above, the attempt to distinguish mistake of law from mistake of fact on that basis can yield paradoxical results.

49 This however is not always true even in the case of signs that are addressed to a definite person. Consider for example restraining orders that can be in force for a considerable amount of time.

50 Leonard (n 21) 528 defines mistake of law as a failure to recognise that ‘the defendant's experience would instantiate the meaning of the statutory term’. However, in effect, he ignores the important distinction between empirical inferences and inferences by virtue of meaning. This can be seen in the far-reaching implications he draws from his definition. Take for example his argument that a mistake of fact can never be unreasonable. According to Leonard, an unreasonable mistake must be a mistake of law, inasmuch as the fact-finder could have inferred from the facts known to the defendant that a certain statutory term had been satisfied: ibid 564–75. The revision of fundamental distinctions of the positive law that is required by this conclusion is so profound that it constitutes, on its own, a powerful argument against Leonard's approach.

51 At this point one might argue that the distinction between empirical inferences and inferences whose validity derives solely from the prohibition's meaning is problematic. Support for such a view may be found in the approach of Willard Quine, who rejected the distinction between synthetic and analytic statements in favour of semantic holism: WV Quine, ‘Main Trends in Recent Philosophy: Two Dogmas of Empiricism’ (1951) 60 Philosophical Review 20, 39–40. If we were to adopt that view, we would not be able to distinguish empirical inferences from inferences whose truth depends upon meaning, and we would have to say that a person who is mistaken in regard to the incidence of the law due to a mistaken meaning of the laws of physics (eg in regard to the laws of optics) is actually mistaken in regard to a question that concerns the meaning of the law.

However, we assume that distinction does exist, although the justification of our position in this matter is beyond the scope of this article. See HP Grice and PF Strawson, ‘In Defense of a Dogma’ (1956) 65 Philosophical Review 141 for a critique of Quine's view. It bears pointing out, however, that the question of which inferences will be analytic (that is, true by virtue of the meaning of the prohibition) and which will be synthetic or empirical may vary according to the context. This variation poses no particular difficulty; it is inherent in the very ambiguity of language. Thus, just as we can accept the ambiguity associated with the extension of a particular concept (eg whether a kick scooter should be deemed a ‘vehicle’ in certain contexts), it is not surprising to find ‘vertical’ ambiguity, namely, ambiguity in regard to the question as to which of the inferences that lead to the conclusion that a particular proposition is true derive directly from the meaning of the proposition in that context, and which are entirely an empirical matter. In both cases, the ambiguity will have to be resolved in accordance with the context.

52 As we noted above (at n 47), in the case of legal institutions (such as property and marriage) there are intra-legal motivations to acquire knowledge of the constitutive rules of the legal institution that are not the fear of punishment: intra-legal motivations such as using the institutions offered by the legal system for one's own purposes (for instance by imposing legal duties on oneself or others by such institutions as contracts, property, etc). This distinction may provide the basis for the distinction between mistake of non-governing law and mistake of governing law.

53 See the Regulations for the Protection of Holy Places to the Jews 5741–1981.

54 It seems that according to Leonard's definition the answer is ‘yes', since based on that person's experience the court would have reached the conclusion that the defendant is at the tomb of Absalom: see Leonard (n 21) 536.

55 Indeed, traffic offences are in many cases strict liability offences. In these cases the legislator does intend to deprive us of the freedom not to be in full possession of our faculties. However, these are cases in which the legislator deems insufficient the extra-legal motivations one has to inquire as to the factual situation.

56 This hypothetical case is based on Alexander (n 6) 49. We changed the facts to make it a case of exculpatory mistake.

57 ibid 51. We changed the facts to make it a case of exculpatory mistake.

58 We assume that in this example the constitutive rules that govern marriage are legal (as opposed to the extra-legal rules in the hypothetical bigamy case). Since the constitutive rules of marriage are addressed to an indefinite group of persons, the definition of marriage includes all those constitutive rules. In other words, if Mr A's mistake were a mistake as to the constitutive legal rules of marriage, his mistake should have been classified as a mistake of law (although it would probably be a mistake of non-governing law). However, Mr A's mistake was not as to those rules, but as to a sign addressed to a definite person.

59 cf Simons, ‘Mistake of Fact or Mistake of Criminal Law’ (n 7) 225.

60 Alexander (n 6) 50.

61 One might argue that temporal and territorial effect can always be reformulated to be part of the conduct rule. Take the example of a statute that proscribes the use of marijuana and has a commencement provision that states that it will come into force on 1 January 2016. We can interpret the conduct rule as stating ‘do not use marijuana on 1 January 2016 and on the days after’ rather than a conduct rule that comes into force on 1 January 2016 and states ‘do not use marijuana’. However, we believe this is not the case. The rationales at the basis of the concepts of temporal effect and territorial effect can help us determine what the correct interpretation of the conduct rule is. In this case there is nothing special about 1 January 2016 when viewed from the rational of proscribing the use of marijuana. The date of commencement is 1 January 2016 in order to provide time for the law-enforcement agencies to prepare for the change in the legal situation, to give the public time to learn of the change etc. These are considerations that are addressed by the concept of temporal effect.