1. See Tammelo, Ilmar, Modern Logic in the Service of Law (Vienna: Springer-Verlag, 1978) at 117–18Google Scholar. For Tammelo, the “Paradox of Parliamentary Sovereignty, whereby on the one hand it is postulated that no law is beyond Parliamentary competence, while on the other hand it is claimed that one kind of law is beyond such competence” arises from “[t]wo fundamental principles of constitutional law … : (1) Parliament can make or unmake any law whatsoever. (2) Parliament cannot bind its successor” (ibid). See also JC Hicks, “The Liar Paradox in Legal Reasoning” (1971) 29:2 Cambridge LJ 275.

2. One is reminded of the paradox of the barber who can shave himself if and only if he cannot shave himself, notably referenced by Bertrand Russell, “The Philosophy of Logical Atomism” (1919) 29:3 Monist 345 at 354-55. Stated simply, the paradox supposes a society where all men keep themselves clean-shaven either by (1) shaving themselves or (2) being shaved by the barber. If the society’s one, clean-shaven barber is defined as one who “shaves all those, and those only, who do not shave themselves” (ibid at 355), then by extension, the barber would shave himself if and only if he did not shave himself.

One popular solution to this paradox is to point out that the barber is a woman. Translated into the language of constitutional amendment, this could mean that the ultimate constituent authority is not a legal authority and cannot be changed in legal terms (which roughly corresponds, as will shortly appear, to Alf Ross’s solution). See in general Laurence Goldstein, “Four Alleged Paradoxes in Legal Reasoning” (1979) 38:2 Cambridge LJ 373.

3. Or, perhaps more aptly, dissolution. On this solution, see JL Mackie, “Evil and Omnipotence” (1955) 64 Mind 200 at 212; Hart, HLA, The Concept of Law (Oxford: Clarendon Press, 1961) at 145Google Scholar-49 [Hart, Concept]; Hart, HLA, “Self-Referring Laws” in Lejman, Fritjof, ed, Festskrift tillägnad professor, juris doktor Karl Olivecrona (Stockholm: PA Norstedt & Söner, 1964) 307Google Scholar at 315 [Hart, “Self-Referring”].

4. On the proposition that the choice is somehow unavoidable between the two notions of sovereignty, see George Winterton, “The British Grundnorm: Parliamentary Supremacy Re-examined” (1976) 92:4 Law Q Rev 591. This is not to say that in-between positions are excluded; they are simply qualified versions of one of the basic positions (for instance, continuing sovereignty cum manner-and-form exception, which could also, conceivably, be expressed as self-embracing sovereignty cum limits-on-content exception). See also George Winterton, “Is the House of Lords Immortal?” (1979) 95:3 Law Q Rev 386.

5. Hart, Concept, supra note 3 at 145.

6. Ibid.

7. See generally Marshall, Geoffrey, Parliamentary Sovereignty and the Commonwealth (Oxford: Clarendon Press, 1957) at 60–75Google Scholar. For a recent discussion of the assumption, see Peter C Oliver, “Sovereignty in the Twenty-First Century” (2003) 14:2 King’s College LJ 137 at 149-52.

8. Dicey, AV, Introduction to the Study of the Law of the Constitution, 8th ed (London: MacMillan, 1915) at 65, n 3 [emphasis added]Google Scholar. Dicey believed, however, that Parliament could conceivably “abdicate”, so long as the abdication was complete, so that the “logic” of continuing sovereignty could then be applied to the next “sovereign”: ibid.

9. Alf Ross, “On Self-Reference and a Puzzle in Constitutional Law” (1969) 78 Mind 1 [Ross, “Self-Reference”]. The general issue of self-reference taken as a logical absurdity, which was also dwelt upon by Ross, will only be addressed indirectly. See also Alf Ross, On Law and Justice, translated by Margaret Dutton (London: Stevens and Sons, 1958) at 78-84 [Ross, Justice]. More generally, see Barwise, Jon & Etchemendy, John, The Liar: An Essay on Truth and Circularity (Oxford: Oxford University Press, 1987).Google Scholar

10. In a context of parliamentary sovereignty, this corresponds to self-embracing powers.

11. Ross, “Self-Reference”, supra note 9 at 5.

12. Ross, Justice, supra note 9 at 82.

13. See Hart, “Self-Referring”, supra note 3 at 314; J Raz, “Professor A Ross and Some Legal Puzzles” (1972) 81 Mind 415 at 420.

14. Ross, “Self-Reference”, supra note 9 at 20.

15. Goldstein, supra note 2 at 376. See also Norbert Hoerster, “On Alf Ross’s Alleged Puzzle in Constitutional Law” (1972) 81 Mind 422.

16. Thus put, the inference violates the canon of Ross’s logical system that premises determine the conclusion instantaneously.

17. I would suggest that Goldstein’s strategy only differs formally from the time-based strategy. What he is saying in substance is that the first premise applies to the constitution found at time₁, and the norm contained in the conclusion applies to the constitution found at time₂. Goldstein otherwise asserts that the time-based strategy is invalid.

18. Hart, “Self-Referring”, supra note 3 at 315.

19. Ross, “Self-Reference”, supra note 9 at 19.

20. Ibid at 6-7.

21. These are sometimes called “Boolean operators”.

22. The traditional rule of first-order logic is that predicates can take as an argument only a term, not a sentence (a “sentence” being the representation of a “proposition” that contains a subject and a predicate). A “statement about a sentence” takes the proposition as subject and adds a further predicate (introducing thereby a “second-order” of predication). See Davis, Ernest, Representations of Commonsense Knowledge (San Mateo: Morgan Kaufmann, 1990) at 83–86.Google Scholar

23. The same applies to the problem of self-reference, taken in isolation, which cannot be translated directly into first-order logic. Consider the classical liar sentence: “This sentence is false”. Basic first-order logic has no equivalent of the demonstrative “this” used self-referentially because the latter implies a “second order” of predication. See ibid.

24. Ibid. This is apparently a common mistake.

25. Note that Ross otherwise rejects that approach to deontic logic which treats directive propositions descriptively in order to ascribe to them truth-value (in terms of legal validity), an approach which directly puts a logic of norms outside the scope of first-order logic: see generally Wright, Georg H von, Norm and Action: A Logical Inquiry (London: Routledge & Kegan Paul, 1963) [von Wright, Norm]Google Scholar. To Ross, “deontic” logic is a system parallel to “indicative” logic, the latter “dealing with the formal conditions for [meaningful] indicative discourse, and the other dealing with the formal conditions for [meaningful] directive discourse”; logic is thus “concerned with the conditions under which the posing of one [directive or indicative proposition] is compatible with the posing of another one [respectively directive or indicative]” (Alf Ross, Directives and Norms (London: Routledge & Kegan Paul, 1968) at 180, 29-33, 177-82). For his constitutional law puzzle, at least in the 1969 version, Ross treats his inference as a “directive”, or “deontic”, inference and concludes, presumably, that taken in isolation it is contradictory and therefore fails to meet the conditions for meaningful direction.

26. Those additional connectives have to meet a number of formal criteria in order to qualify as “extensional connectives”, that is, connectives that can function in legitimate first-order logic. For instance, they have to commute with the traditional connectives: Davis, supra note 22 at 52-59.

27. Davis, supra note 22. Extensional time operators have been discussed in artificial intelligence research since the late fifties. See McCarthy, John, “Programs with Common Sense” in Mechanisation of Thought Processes: Proceedings of a Symposium Held at the National Physical Laboratory on 24th, 25th, 26th and 27th November 1958 (London: Her Majesty’s Stationery Office, 1959) vol. 1, 75Google Scholar. A complete set of time operators should at least include a set of “tense operators” covering all the possibilities envisaged by a traditional grammar.

28. Davis, supra note 22 at 187.

29. This is the same as saying that “constitution” is a time-varying term in the inference, which the basic system of representation cannot handle. One commentator encapsulates Ross’s problem as follows: “His intransigence suggests the assumption that all logical matters are concerned with the timeless and changeless interplay of concepts, which then of course cannot accommodate amendment. This resistance can be met by embedding a logic of process within the scope of logic, or by removing the stigma from the epithet ‘illogical’ when applied to action under norms, in case one continues to insist upon the narrower scope of logic” (Christopher Berry Gray, “Amendment: Legal Continuity and Ongoing Revolution” in Elspeth Attwooll, ed, Shaping Revolution (Aberdeen: Aberdeen University Press, 1991) 47 at 51).

30. I have so far assumed with Ross that his art. 88 does refer to itself but the assumption is not provided by the text.

31. One commentator has claimed that the equivalent of continuing (unchangeable) powers could be formally written into a constitutional document as follows: “All rules concerning the making and change of the rules of the constitution are unchangeable,” and that this can be taken as “solving the paradox of the rule of constitutional change” (Nikolas HM Roos, “The Identity of Legal Systems in the Light of Some Paradoxes of Constitutional Law” in Elspeth Attwooll, ed, Shaping Revolutions (Aberdeen: Aberdeen University Press, 1991) 56 at 66-68). The paper takes it as “self-evident that any system in its proper sense must have a core of rules that do not change and that define the limits of change compatible with the system” (ibid at 66). I would suggest that this is a change-reluctant practical judgment that simply happens to match the “staticity” of first-order logic. The author is mistaken in thinking that a “negatively recursive” rule can therefore be formally self-referential in a first-order logical construction.

32. Note that in other respects self-embracing sovereignty is more “static” than continuing sovereignty. For an original argument showing that the reason most commonly advanced to defend continuing sovereignty (wisdom of future generations) usually involves a self-refutation, see John Finnis, “Scepticism, Self-Refutation, and the Good of Truth” in Hacker, PMS & Raz, J, eds, Law, Morality, and Society: Essays in Honour of HLA Hart (Oxford: Clarendon Press, 1977) 247 at 254–56.Google Scholar

33. For a recent effort at handling time in the modelling of legal rules, see Palmirani, Monica, Governatori, Guido & Contissa, Giuseppe, “Modelling Temporal Legal Rules” in Proceedings of the 13th International Conference on Artificial Intelligence and Law (New York: ACM, 2011) 131.Google Scholar

34. Major problems have yet to be overcome for this assumption to prove satisfactory: see generally the special issue (1991) 4:3 Ratio Juris; von Wright, Norm, supra note 25; Weinburger, Ota, Law, Institution and Legal Politics: Fundamental Problems of Legal Theory and Social Philosophy (Dordrecht: Kluwer Academic, 1991) at 77–89Google Scholar. Some prominent problems are as follows. First, the applicability of the principle of contradiction to normative systems remains a controversial matter: see Kelsen, Hans, General Theory of Norms, translated by Hartney, Michael (Oxford: Clarendon Press, 1991) at 211–14Google Scholar. Second, a sufficiently differentiated set of extensional sentence operators remains to be developed: see Georg H von Wright, “Is There a Logic of Norms?” (1991) 4:3 Ratio Juris 265 [von Wright, “Logic”]; Tammelo, supra note 1. Third, the use of even the most basic truth-functional operators with normative propositions gives rise to serious problems: see Carlos E Alchourrón & Eugenio Bulygin, “Limits of Logic and Legal Reasoning” in Antonio A Martino, ed, Expert Systems in Law (Amsterdam: North-Holland, 1992) 9 at 21-22. Fourth, attempts at providing a formal representation of the conceptual structure of reasoning from norms have so far succeeded only with the most primitive patterns: von Wright, “Logic”, above, at 266. A significant portion of the efforts dedicated to the formal representation of legal phenomena is now deployed in the area of “defeasible” logic, or “non-monotonic” logic, on which see Christian Strasser & G Aldo Antonelli, “Non-monotonic Logic” in Stanford Encyclopedia of Philosophy (2 December 2014), online: Stanford Encyclopedia of Philosophy http://plato.stanford.edu/.

35. See generally Suber, Peter, The Paradox of Self-Amendment: A Study of Logic, Law, Omnipotence, and Change (New York: Peter Lang, 1990) at 25–32.Google Scholar

36. With respect to logic, I refer to this context as the “pragmatics” of legal change (“pragmatics” is a term in contemporary linguistics indicating the branch of the discipline that deals with language in use).

37. Ross, “Self-Reference”, supra note 9 at 24.

38. He also relied on the idea that an inference from his principle was a delegation, as opposed to a “self-destroying transference of competence” that is postulated from outside the principle: ibid at 22.

39. Eekelaar, JM, “Principles of Revolutionary Legality” in Simpson, AWB, ed, Oxford Essays in Jurisprudence (Second Series) (Oxford: Clarendon Press, 1973) 22.Google Scholar

40. Sampford, Charles, “Coups d’Etats and Law” in Attwooll, Elspeth, ed, Shaping Revolution (Aberdeen: Aberdeen University Press, 1991) 161 at 162–67.Google Scholar

41. For an exhaustive survey of judicial responses to coups d’état in post-colonial common law jurisdictions, see Tayyab Mahmud, “Jurisprudence of Successful Treason: Coup d’Etat & Common Law” (1994) 27:1 Cornell Int’l LJ 49.

42. For an account of the various modes of validity with respect to time, see Bulygin, Eugenio, “Time and Validity” in Martino, Antonio A, ed, Deontic Logic, Computational Linguistics and Legal Information Systems, vol. 2 (Amsterdam: North-Holland, 1982) 65.Google Scholar

43. State v Dosso, [1958] SC 533.

44. Uganda v Matovu, [1966] E Afr LR 514.

45. Mitchell v Director of Public Prosecutions, [1986] LRC (Const) 35.

46. For other cases see Mahmud, supra note 41 at 54-99.

47. Ibralebbe v R [1964] AC 900 at 918.

48. Ibid at 922.

49. Finnis, John, “Revolutions and Continuity of Law” in Simpson, AWB, ed, Oxford Essays in Jurisprudence (Second Series) (Oxford: Clarendon Press, 1973) 44 at 63-65 [Finnis, “Revolutions”]Google Scholar. Finnis says that his formulation makes sense of the practice and offers the further justification of practical reasonableness. French legal philosopher Michel Virally had reached a substantially similar conclusion, though in a mode that never gets beyond the description of practice: Virally, Michel, La pensée juridique (Paris: Librairie générale de droit et de jurisprudence, 1960) at 188–98.Google Scholar

50. Finnis, “Revolutions”, supra note 49 at 63. Compare with Virally who, speaking of legal continuity in cases of revolution, secession, or annexation, writes: “La continuité de l’ordre juridique, dans ces diverses hypothèses, pour remarquable qu’elle soit, n’a rien de mystérieux. Elle va de soi pour toute la partie de cet ordre qui est d’origine coutumière ou jurisprudentielle: sa validité n’a jamais été déduite d’une quelconque norme fondamentale. Elle résulte du fonctionnement de l’ordre juridique, c’est-à-dire de l’ensemble des rapports sociaux, qui ne sont pas immédiatement affectés par un changement d’institutions politiques ou d’appareil étatique, dans la mesure où ils sont compatibles avec les règles dominant les institutions nouvelles. Mais il faut en dire autant du droit écrit. Comme nous l’avons vu, il bénéficie d’une validité permanente à partir du moment où il est créé. Cette validité lui est conférée par l’acte qui l’a posé et qui l’empruntait lui-même à la norme supérieure en vertu de laquelle il a été pris. La validité du droit écrit—c’est là la grande différence avec le droit coutumier—lui est attribuée par l’acte qui le pose, dans l’instant et de façon définitive. Il n’est donc pas nécessaire que la norme supérieure conserve sa validité pour que la norme inférieure soit maintenue dans la sienne” (Virally, supra note 49 at 193 [footnote omitted]).

51. The formalization of the rules of “pragmatics” poses the same problem in the logical representation of ordinary, real-life reasoning. For building such rules into a system of representation will usually mean cutting off that system from the real-world situation it is meant to represent. After a period of time, real-world reasoning may evolve into new rules of pragmatics and the system may cease to provide adequate representations. In order to avoid this, one may want to keep the rules of pragmatics separate in the system of logical representation and provide a “window out on the real world” through which those rules can be updated. This is one of the difficulties in evolving a computer expert-system meant to give legal advice. See, e.g., Rissland, Edwina L & Timur Friedman, M, “Detecting Change in Legal Concepts” in Proceedings of the 5th International Conference on Artificial Intelligence and Law (New York: ACM, 1995) 127Google Scholar. This difficulty is also linked to the problem of accounting for soft standards and values in legal decisions, on which see generally Bench-Capon, Trevor, “Relating Values in a Series of Supreme Court Decisions” in Atkinson, Katie M, ed, Legal Knowledge-Based Systems (Amsterdam: IOS Press, 2011) 13Google Scholar; Bench-Capon, Trevor et al, “Argument Schemes for Reasoning with Legal Cases Using Values” in Proceedings of the Fourteenth International Conference on Artificial Intelligence and Law (New York: ACM, 2013) 13.Google Scholar

52. It is significant in this respect that the Interpretation Acts in Canada seem to be largely ignored by lawyers and judges alike.

53. In England, prior to 1850, the common law understanding of repeal included the rule that the repeal of a repealing Act revived the Act originally repealed: Finnis, “Revolutions”, supra note 49 at 61, 63, n 47. Note that this position was peculiar to the relationship between common law and statute law. The position of British imperial law was otherwise consistent with the principle. As expounded in Campbell v Hall (1774), 1 Cowp 204 at 209, 98 ER 1045 (KB), the laws of a conquered or ceded colony were taken to “continue in force, until they [were] altered by the conqueror.”

54. Kelsen, Hans, Pure Theory of Law, 2d ed translated by Knight, Max (Berkeley: University of California Press, 1967) at 226–27.Google Scholar

55. Kelsen, Hans, General Theory of Law and State, translated by Wedberg, Anders (New York: Russell & Russell, 1961) at 115.Google Scholar

56. Hart, Concept, supra note 3 at 120.

57. This contrasts with his staunch insistence as a logician that formal rules of change can be self-reflective and provide for their own amendment. His position on acceptance, as a legal theorist, means that a new ultimate rule of competence is valid whether or not it can be formally brought about by self-reference.

58. Finnis, “Revolutions”, supra note 49 at 59.

59. See generally ibid at 65-70.

60. Hart, Concept, supra note 3 at 116 [emphasis added].

61. One example of this is afforded by British statutes still applied in Canada: the rule that identifies some British statutes as valid Canadian law today is distinct from the ultimate rule of competence that does not recognize British law-making power for Canada.

62. Hart, Concept, supra note 3 at 152.

63. Contrast this with Raz’s second attempt to account for “rules of discretion” (soft standards) in Hartian terms: the normative context is outside “the law”, but is pointed to as relevant in adjudication by formally validated law: Raz, Joseph, “Legal Principles and the Limits of Law: A Postscript” in Cohen, Marshall, ed, Ronald Dworkin and Contemporary Jurisprudence (Totowa: Rowman & Allanheld, 1984) 81Google Scholar. The picture taking shape in the text above reads the other way: it is not the law that points to external considerations as relevant for legal decisions; it is those external considerations reflected upon by human beings in society that point to certain acts and facts as presumptively conclusive for decisions.

64. Finnis, “Revolutions”, supra note 49 at 65-70.

65. Virally, supra note 49.

66. Raz, Joseph, The Authority of Law: Essays on Law and Morality (Oxford: Clarendon Press, 1979) at 100.Google Scholar

67. Ross, Justice, supra note 9 at 83.

68. That conclusion does not involve that self-reference, as a means of constitutional change, is illogical: it means that even if it were free of logical flaws, it could not be sufficient.

69. Finnis, “Revolutions”, supra note 49 at 66.

70. Ibid at 69. The mimeographed work from which Finnis drew this insight is now published. See Voegelin, Eric, “The Nature of Law” in Pascal, Robert Anthony, Lee Babin, James & Corrington, John William, eds, The Collected Works of Eric Voegelin: The Nature of Law and Related Legal Writings, vol. 27 (Baton Rouge: Louisiana State University Press, 1991) 1.Google Scholar

71. Finnis, “Revolutions”, supra note 49 at 69.

72. See François Ost, “Entre ordre et désordre : le jeu du droit ; discussion du paradigme autopoiétique appliqué au droit” (1986) 31 Archives de philosophie du droit 133.

73. Finnis, John, “Natural Law and Legal Reasoning” in George, Robert P, ed, Natural Law Theory: Contemporary Essays (Oxford: Clarendon Press, 1992) 134 at 150.Google Scholar

74. Ibid. On the importance of legal forms, see Finnis, John, Natural Law and Natural Rights (Oxford: Clarendon Press, 1980) at 311-13, 354–62Google Scholar. On the “bridging” of natural law and positive law theories, see ibid at 23-49, 363-66; Neil MacCormick, “Natural Law and the Separation of Law and Morals” in George, Robert P, ed, Natural Law Theory: Contemporary Essays (Oxford: Clarendon Press, 1992) 105.Google Scholar

75. Hart, Concept, supra note 3 at 120 [emphasis added]. He wrote, on the following page: “[W]e have at the end of this development two independent legal systems. This is a factual statement, and not the less factual because it is one concerning the existence of legal systems.”

76. Ibid at 107.

77. See the discussion of NE Simmonds, “Practice and Validity” (1979) 38:2 Cambridge LJ 361 at 369-70. See also Jacques Derrida, who provides an elegant demonstration that declarations of independence are necessarily both descriptive and prescriptive, both fact and law: Derrida, Jacques, Otobiographies : l’enseignement de Nietzsche et la politique du nom propre (Paris: Galilée, 1984) at 27.Google Scholar

78. John W. Salmond is often quoted to justify statements of that type: “Before there can be any talk of legal sources, there must be already in existence some law which establishes them and gives them their authority. The rule that a man may not ride a bicycle on the footpath may have its source in the by-laws of a municipal council; the rule that these by-laws have the force of law has its source in an Act of Parliament. But whence comes the rule that Acts of Parliament have the force of law? This is legally ultimate; its source is historical only, not legal. The historians of the constitution know its origin, but lawyers must accept it as self-existent. It is the law because it is the law, and for no other reason that it is possible for the law itself to take notice of” (Sir John W Salmond, Jurisprudence, 7th ed (London: Sweet & Maxwell, 1924) at para 48 [emphasis added]). This reflects Salmond’s first formulation of this argument: “there must in every system of law be some one or more ultimate principles, the existence of which is not deduced from any other principle; rules which exist, but for the existence of which the law provides no reason” (John W Salmond, The First Principles of Jurisprudence (London: Steven & Haynes, 1893) at 222). He mentions three such rules: one for statutes, one for common law, and one for custom: ibid at 220-23.

79. Hart, Concept, supra note 3 at 152.

80. This is particularly obvious in Joseph Raz’s jurisprudential criterion for a norm’s membership in a legal system. The system is said to contain “only those norms which its primary organs are bound to apply” (Joseph Raz, Practical Reason and Norms (Oxford: Oxford University Press, 1999) at 142). The legal system is thus defined as containing only norms that are “binding” on “primary organs” of the system, where “binding” and “primary organs” are, more or less directly, defined by norms of the system, that is, norms binding on primary organs of the system. Raz provides a semi-formal representation of such a circle of justification: Raz, Joseph, The Concept of a Legal System (Oxford: Clarendon Press, 1970) at 138–40.Google Scholar

81. Hart, Concept, supra note 3 at 152.

82. Kerchove, Michel van de & Ost, François, Legal System between Order and Disorder, translated by Stewart, Iain (Oxford: Oxford University Press, 1994) at 65-72, 118–22Google Scholar. See also MacCormick, Neil, Legal Reasoning and Legal Theory (Oxford: Clarendon Press, 1978) at 54-55, 245–46.Google Scholar

83. Luhmann, Niklas, “The Unity of the Legal System” in Teubner, Gunther, ed, Autopoietic Law: A New Approach to Law and Society (Berlin: Walter de Gruyter, 1987) 12.Google Scholar