In early twentieth-century South Africa, the Gordonia School Board refused admission to two children on the grounds that they were not of “European parentage.”<a href=”https://juris.jotwell.com/can-law-speak-against-itself/#easy-footnote-bottom-1-3315″ title=”This decision was ultimately upheld in Moller v. Keimoes School District 1911 AD 635.” target=”_blank”>1 In determining the validity and meaning of the relevant empowering legislation, Chief Justice Lord De Villiers invoked “public history”: the first “civilized” legislators in South Africa regarded its indigenous peoples as “an inferior race” whom they were “entitled to rule over.” The legal order De Villiers was reasoning within presupposed that European legislators had the legal right to rule over all inhabitants of South Africa,<a href=”https://juris.jotwell.com/can-law-speak-against-itself/#easy-footnote-bottom-2-3315″ title=”Many South African judges would invoke the common-law principle that all persons are equal before the law to deny or temper this norm. Whether that principle of equality could plausibly be reconciled with the country’s institutional history was, however, deeply contested. As Judge Beyers observed in Minister of Posts and Telegraphs v. Rasool 1934 AD 167: “The proposition that in the eyes of the law everyone is equal cannot be unconditionally accepted. It is undoubtedly subject to qualification;” and, he argued, at least with respect to the Transvaal province, it had never been true (P. 177). Given the history, logic, and presuppositions of that legal order, Beyers had a point. The question of how the judicial role could be fulfilled within a legal system premised on profoundly unjust norms lay at the core of the 1984 debate over whether South African judges should resign. See Raymond Wacks, Judges and Injustice 101 S. African L. J. 266 (1984); John Dugard, Should Judges Resign–A Reply to Professor Wacks 101 S. African L.J. 286 (1984); Raymond Wacks, Judging Judges: A Brief Rejoinder to Professor Dugard 101 S. African L.J. 295 (1984).” target=”_blank”>2 and the society of which he was a part continued to accept the soundness of that proposition. In his personal capacity, he might have disagreed. But, qua judge, he could either accept that norm for purposes of legal reasoning or abandon legal reasoning altogether. And if South African society came to disagree with that foundational norm, it would face the same choice: accept the norm while reasoning within the legal order or reject the norm along with the legal order itself.<a href=”https://juris.jotwell.com/can-law-speak-against-itself/#easy-footnote-bottom-3-3315″ title=”If they were to succeed in changing the legal order, that change would be, as Stephen Sachs puts it, an “unauthorized change to the law.” See Stephen E. Sachs, Originalism as a Theory of Legal Change 38 Harv. J.L. &amp; Pub. Pol’y 817, 843 (2015).” target=”_blank”>3

That, at least, is the standard view: a legal order’s highest norm might be changed by social and political processes, whether through outright revolution or via a series of “pious fictions.”<a href=”https://juris.jotwell.com/can-law-speak-against-itself/#easy-footnote-bottom-4-3315″ title=”H.L.A. Hart, The Concept of Law, 184 (2012).” target=”_blank”>4 But there can be no legal authorization for changing the highest norm of a legal order. To revise a legal order’s highest norm is to abandon it in favor of a new one. As Stephen Sachs has written, adhering to our current law means rejecting unauthorized changes to it.<a href=”https://juris.jotwell.com/can-law-speak-against-itself/#easy-footnote-bottom-5-3315″ title=”Stephen E. Sachs, Originalism as a Theory of Legal Change 38 Harv. J.L. &amp; Pub. Pol’y 817, 844 (2015).” target=”_blank”>5 Manish Oza invites us to reconsider this standard view in his recent Article, Can We Legally Revise the Highest Legal Rule?.

Oza targets a conceptual puzzle: if legally valid change must be authorized by a higher legal norm, every legal order must contain a highest legal norm that cannot itself validly be changed (P. 273).6 We could try to dissolve this puzzle in several ways, but Oza addresses each attempt convincingly. First, we would avoid the puzzle if law emanates from an extra-legal sovereign lawmaker. If so, all laws can change at their say-so. Oza sets this view aside (P. 271), with good reason. Hart’s critique of the command theory of law is convincing on this front: lawmaking authority must be understood as a consequence of law, not a precondition for it. Second, we might argue that the highest legal norm provides for its own amendment. But this either implies that, once amended, the norm authorizing that amendment no longer exists, or that there was an even higher norm governing the validity of such amendments—thereby reproducing the puzzle at a different level.

Third, we might deny that there is a highest legal norm. Two routes are available. One is to posit an infinite series of ever-higher legal norms. The other is to argue that a legal order has a non-hierarchical “circular” structure in which no norms are immune from revision. Oza argues that we should reject the former route: on this approach, our account of law loses all connection to the knowledge and attitudes of those whose law it purports to capture (P. 274). And on the latter route, the puzzle persists: even if all law is revisable, there remains a distinction between legal and non-legal revision calling for explanation (P. 287).

A final strategy doesn’t dissolve the puzzle but seeks to minimize it: why care about legal change at all? Either the legal order changes or it doesn’t. Nothing turns on whether that change is legal. Oza responds by pointing to the political value of legal change in a society characterized by deep disagreement (P. 290). He might also have cited the influence of originalist theory and practice as evidence of the political salience of the distinction between legally authorized and politically efficacious change.

We are left, therefore, with the puzzle: can a legal order authorize changes to its own foundations? Oza draws a parallel between this question and one in logic: whether it is possible, while using a given logic, to rationally conclude that one ought not use that logic. Hartry Field argued that it is,<a href=”https://juris.jotwell.com/can-law-speak-against-itself/#easy-footnote-bottom-7-3315″ title=”Hartry Field, Logic, Normativity, and Rational Revisability (2008).” target=”_blank”>7 and Oza offers a parallel argument for law. The solution lies in distinguishing between two steps in the reasoning. The first is concluding that the highest norm ought not to be followed. This is possible: one can validly derive, in a logic, the conclusion that one ought not reason according to that logic. Similarly, Oza suggests, one can conclude by legally valid reasoning that the highest norm of a legal order ought not to be followed. He writes, “There’s nothing impossible about using Canadian legal reasoning to conclude that Canadian law is invalid. If a legal order includes rules about the conditions of its own validity, then it may also include the conclusion that those conditions are not met” (Pp. 281-282).8

The second step is to replace the highest legal norm with a new norm while still reasoning legally (or rationally, in Field’s case). Clearly, the replaced norm can’t be dictated by the previous highest norm. But, Oza argues, the previous norm can “ground its own replacement” (P. 282). The conclusion that law is invalid is itself a legal reason, which in turn supports the adoption of a revised highest norm. That revision isn’t an extra-legal change: we are “responding to a specific legal reason that is generated by the legal rules” (P. 282). In other words: a change to the highest norm of a legal order can be legally grounded, even if not legally authorized.

I am unsure of Oza’s solution. I worry that the parallel with rationality and logic is being asked to do more than it can. Field is intervening in a complicated debate about the relationship between logic and rationality. And even if that relationship were straightforward, I worry that the idea of “rational thought” does not sufficiently illuminate the idea of “legal thought” in Oza’s parallel. Do legally valid norms relate to legal reasoning in the same way that logic relates to rational thought? I don’t know. I am left with this question: What makes reasoning appropriately legal? It is in foregrounding this question that Oza’s article is especially welcome, and I hope it will prompt further discussion.

Oza argues that legal reasoning is not exhausted by drawing valid inferences from existing legal norms. A formally coherent system of norms can generate the conclusion that its own norms ought not to be followed. If such conclusions are to have any legal significance, that significance must come from some further standard of legal reasoning that is not itself captured by formal validity. Articulating that standard is an important theoretical task.

There are two further implications of his argument that merit attention. First, Oza’s argument exposes a problem for theories of law that identify what judges ought to do according to the law with valid legal content. If the law’s content can include self-undermining directives, then what is legally required might not be exhausted by what valid law directs.

Finally, Oza’s argument invites us to rethink a prevalent but overly crude picture of the relationship between judicial lawmaking and “gaps” in the law. In that picture, judges are authorized to make law only where the law explicitly directs them to, or where it is silent. Oza’s argument suggests this framing is incomplete. Judges might be legally directed to change the law not because it is silent, but precisely because it speaks against its own foundations.

Download PDF
  1. This decision was ultimately upheld in Moller v. Keimoes School District 1911 AD 635.
  2. Many South African judges would invoke the common-law principle that all persons are equal before the law to deny or temper this norm. Whether that principle of equality could plausibly be reconciled with the country’s institutional history was, however, deeply contested. As Judge Beyers observed in Minister of Posts and Telegraphs v. Rasool 1934 AD 167: “The proposition that in the eyes of the law everyone is equal cannot be unconditionally accepted. It is undoubtedly subject to qualification;” and, he argued, at least with respect to the Transvaal province, it had never been true (P. 177). Given the history, logic, and presuppositions of that legal order, Beyers had a point. The question of how the judicial role could be fulfilled within a legal system premised on profoundly unjust norms lay at the core of the 1984 debate over whether South African judges should resign. See Raymond Wacks, Judges and Injustice 101 S. African L. J. 266 (1984); John Dugard, Should Judges Resign–A Reply to Professor Wacks 101 S. African L.J. 286 (1984); Raymond Wacks, Judging Judges: A Brief Rejoinder to Professor Dugard 101 S. African L.J. 295 (1984).
  3. If they were to succeed in changing the legal order, that change would be, as Stephen Sachs puts it, an “unauthorized change to the law.” See Stephen E. Sachs, Originalism as a Theory of Legal Change 38 Harv. J.L. & Pub. Pol’y 817, 843 (2015).
  4. H.L.A. Hart, The Concept of Law, 184 (2012).
  5. Stephen E. Sachs, Originalism as a Theory of Legal Change 38 Harv. J.L. & Pub. Pol’y 817, 844 (2015).
  6. As Oza notes, this is not only a puzzle for positivists (P. 275).
  7. Hartry Field, Logic, Normativity, and Rational Revisability (2008).
  8. I worry that we may have slid into a different sense of ‘validity’ here. But it is precisely this question about the different senses in which a conclusion might count as legally valid that Oza’s article so admirably brings to the fore.