Mohamed Moussa: The ‘Absent Word’ Canon and Asymmetrical Sovereignty

Introduction: The Supreme Court’s Textual Fidelity 

The UK Supreme Court (UKSC) recently issued its unanimous judgment which found the draft Scottish Independence Referendum Bill to be outside the legislative competence of the Scottish Parliament. While the facts of this case are distinctively different from previous Scottish cases, a common theme remains in the Court’s insistence on ‘ordinary meaning of words’ as its ‘general approach to the interpretation of the Scotland Act’. According to such an approach, the Court prioritises ‘the language carefully chosen by the Parliamentary drafter and enacted by Parliament’ as ‘[t]he best way of ensuring a coherent, stable and workable outcome’. The purpose of this blog post is not to explore the facts of this particular case. Rather, it focuses on the UKSC’s repeated stress of textualism. A similarity is found in the Court’s textual interpretation of Section 28(7), which was understood to affirm the doctrine of parliamentary sovereignty and served as the crux of its previous cases on devolution. The aim of the post is to show the questionable nature of the Court’s textual fidelity and highlight that it runs counter to foundational canons of interpretation. For space constraints, my argument focuses mainly on Section 21 from the United Nations Convention on the Rights of the Child (Incorporation) (Scotland) Bill Reference (UNCRC Incorporation Bill case) after briefly discussing section 17 of the Scottish Legal Continuity Bill

The ‘Absent Word’ Canon 

In the Legal Continuity Bill, the UKSC found ultra vires, inter alia, Section 17 of the Bill which required Westminster ministers to secure the consent of Scottish ministers when exercising delegated powers to modify retained EU law which touches on a devolved domain.  In the UNCRC-Incorporation Bill case, the Court examined whether the Scottish Parliament could incorporate: (1) The United Nations Convention on the Rights of the Child (Incorporation) (Scotland) Bill (“UNCRC Bill”) and (2) the European Charter of Local Self-Government (Incorporation) (Scotland) Bill into domestic law. The concern was whether such bills would place obligations on the UK government ministers and affect the UK parliament’s power to make laws for Scotland in reserved areas which would be contrary to the devolution settlement. Section 21 of the Bill is largely a replica of Section 4 of the HRA. Section 21 empowers Scottish courts to issue ‘incompatibility declarators’ in respect of UNCRC-incompatible legislation. This applies to UK legislation within devolved competences that receives Royal Assent on or after the day on which section 21 enters into force. The Court found section 21 of the Scottish Bill, unlike section 4 HRA, to be inconsistent with parliamentary sovereignty. 

The Court’s conclusion on section 21 of the UNCRC Bill drew on its reasoning of section 17 in the Continuity Bill which was framed as a textually grounded reading of the ‘ordinary meaning’ of the Scotland Act. (Continuity Bill case paras 12 & 52; UNCRC Bill para 54). To scrutinise the Court’s textual interpretation, two canons of interpretation are particularly relevant: one is a logic-based canon and the other semantical. My argument is that the Court’s reasoning escapes inconsistency with the logical maxim of interpretation but only at the cost of running counter to a foundational semantic canon.

a) The maxim ‘Cui licet quod majus non debet quod minus est non licere’ (4 Co. 23.): He who has authority to do the more important act shall not be debarred from doing that of less importance (briefly: the greater contains the less). If the Scottish parliament has the greater power to repeal or amend Acts of Parliament within devolved competence, a fortiori, it should have the lesser power of declaring them incompatible (albeit acting via courts). To escape this logical objection, the UKSC invoked a non-delegation argument whereby the Scottish parliament was found to lack the competence to delegate such power rather than exercising it by itself.  

In the Incorporation Bill case, the UKSC made this point in two steps, rather than one. An outright invocation of the non-delegation argument would have been contra textual because it would explicitly add delegation to the list of reserved/protected matters . Instead, the Court anchored its interpretation against delegation through a medium of parliamentary sovereignty. Namely it read section 28(7) of the Scotland Act to contain an ‘unqualified’ legislative power that such delegation would unduly ‘modify’ and thus violate the limits of devolved competence.  However, this very notion of ‘unqualified legislative power’, whilst obviating logical inconsistencies, runs counter to the following semantical canon.

b) The ‘absent word’ or ‘implied word’ canon bans courts from adding words to statutes when the language is clear. The canon has a time old pedigree in the common law, and has long been applied by the UK (Beal’s Cardinal Rules of Interpretation, 3rd edn, p 381) as well as US courts. Yet it seems to have eluded the UKSC in its UNCRC case and the Continuity Bill case. Section 28(7) of Scotland Act speaks plainly of Parliament’s power ‘to make laws for Scotland.’ The Court added the word ‘unqualified’ as an extratextual adjective on the section to articulate a threshold of ‘unqualified power’ to’ make laws’ (para 31 UNCRC case) or ‘unqualified legislative power’ (para 40&43). This ‘unqualified’ power is wider and more taxing than parliamentary sovereignty itself as Elliott and Kilford have shown. Namely, it ‘imposes limits on devolved competence that transcend those that follow straightforwardly from sovereignty itself.’ 

A few counterarguments are worth exploring here. First, it could be argued that without adding any ‘absent words’ the ordinary meaning of ‘making law’ entails a notion of sovereignty which is intolerant of the need to express repeal or declarations of incompatibility.  Likewise, it could be argued that the key part of the sentence under interpretation is not ‘making law’ but the part of s28(7) that ‘does not affect the power of the Parliament’ because, arguably, the well-established meaning of parliamentary sovereignty is that is legally unqualified. These objections rest on the presupposition that the need for express repeal is a qualification of sovereignty. However, as McHarg and McCorkindale, among others, have shown, this sits uncomfortably with the established case law where sovereignty was found to be compatible with the requirement of express repeal. One of two things could be true. Either the need for express repeal in a statute modifies and qualifies sovereignty, or it does not. If the former is true, then the Court’s earlier jurisprudence accepts a modification of sovereignty. If it is a case of the latter, then the Scottish Parliament is not guilty of ‘constitutional solecism’ and the relevant sections of the UNCRC /Continuity Bill  should have been upheld. 

Defenders of the Court’s approach would have a second arrow in their quiver which is that express repeal is ordinarily inconsistent with sovereignty and can only be allowed when self-imposed by the parliament itself. However, what is overlooked here is that express repeal is allowed by the statutory scheme of the Scotland Act itself. In fact, the possibility of express repeal of devolved legislations distinguishes devolution from the mainstream federal model where the central legislator lacks such power. In a devolutionary context, the possibility of express repeal is an affirmation rather than negation of sovereignty. How would the UK parliament act in areas falling within the devolved competence and where devolved parliament have already legislated. The UK parliament would simply introduce a new Act repealing and overriding the devolved acts. It is generally accepted that this method does not limit sovereignty, but if that is the case, then why would the requirement of express repeal of the relevant sections in the two bills limit sovereignty.

A third possible defence of the Court’s interpretation can be premised on the argument that the specific language of the Scotland Act itself imposes an exceptionally ‘wider concept than parliamentary sovereignty’. Yet, if the Court is to live up to its asserted textualism, it should pinpoint some textual authority to justify this bidimensional concept of sovereignty when applied in devolution compared to other constitutional matters. When shall lawyers expect ‘ordinary’ parliamentary sovereignty to bite (where the need for express repeal or incompatibility declarations are permissible) and when is ‘unqualified legislative power’ applicable (where such measures are impermissible) and based on what textual criteria? Namely, what are the exact words whose ‘ordinary meaning’ trigger the higher threshold of ‘unqualified legislative powers’. The Court leaves us with no concrete answer, contrary to its promise that using ‘ordinary meaning’ ensures laws are ‘constant and predictable’. 

Realism or Textualism?

Surely, one cannot discuss the canons of interpretation without reference to the American jurisprudential school of realism and its famed attack on textualism. A central claim of realism is that  judges reach decisions through reliance on non-legal factors (e.g., personal/moral views), then strategically recourse to textual analysis to make their judgements seem legally inevitable. Notwithstanding the accuracy or the universality of such a claim (on which much ink has been consumed), the UKSC’s approach to devolution may expose it to the crossfire of realist and critical objections. 

Rather than the—unwavering yet often confusing—insistence on the ‘ordinary meaning’ of the Scotland Act, perhaps a more consistent framework to explain the Court’s approach is institutional bias. Namely, that the Court (consciously or not) reads sovereignty in a manner that is co-expansive of its own institutional power. In a horizontal dispute (i.e., when the UK legislature and judiciary are pitted against one another) the Court has a vested institutional interest to contract the definition of sovereignty to allow more room for its own power as much as constitutionally possible. By contrast, in a vertical dispute between the UK and the devolved governments, expanding parliamentary sovereignty usually travels in tandem with expanding the Court’s own oversight. For instance, to allow Scottish courts to issue an incompatibility declaration would be to concede the possibility of pluralistic or higher standards of judicial protection of rights in Scotland that may behove the UKSC to play catch-up in this domain.  

These are admittedly mere speculations which the Court may attempt to distance itself from. However, to do so will require the Court to have a more consistent analysis of sovereignty. Without such reasoning, the current bidimensional or asymmetrical notion of sovereignty unjustifiably distinguishes vertical and horizontal disputes. This sits uncomfortably with settled caselaw and may open the Court’s textual approach to accusations of being ‘a false aura of objectivity’.

The author would like to thank Alison Young and Michael Gordon for written comments. 

Mohamed Moussa is a College Lecturer in Law, Murray Edwards College, University of Cambridge. His research focuses on public law and comparative federalism.

(Suggested citation: M. Moussa, ‘The ‘Absent Word’ Canon and Asymmetrical Sovereignty’, U.K. Const. L. Blog (20th December, 2022) (available at