The Anglo-British imaginary and the rebuilding of the UK’s territorial constitution after Brexit: unitary state or union state?

ABSTRACT Brexit has foregrounded radical divergences between the accounts of the UK’s constitutional order advanced by the UK Government and the devolved governments, with the distinctions coming into sharp relief in debates over legislation to sustain the UK’s internal market. This article examines the limits to the roots of the UK Government’s insistence that the UK is a unitary state, and not a union-state, in the textbook tradition of constitutional scholarship. Writers from A. V. Dicey to S. A. de Smith asserted that the UK was a unitary state largely as an adjunct to their accounts of parliamentary sovereignty. We also examine how the received orthodoxy of unitary accounts of the UK’s constitution came under increasing pressure after the advent of devolution, but that the UK’s membership of the European Union, and the operation of the principle of subsidiarity within European law, forestalled a considerable amount of constitutional contestation. The need to replace European law as a foundation of the UK internal market, and the UK Government’s attempts to exert control over this transition, has produced a sustained debate about what the union means after Brexit.


INTRODUCTION
In 2014, the year of the Scottish independence referendum, former UK Supreme Court President Lord Neuberger delivered a lecture on the topic of the 'The British and Europe'. There he made the following statement: 'since 1066 the UK has never been successfully invaded by a foreign power … 950 years without a single foreign occupation is a record which I think no other European country can claim' (para. 12). He also said: since Wales was effectively united with England in the 13th century, there have been no changes to the boundaries of the countries of Great Britain: there has been a union with Scotland in 1707, but that was Having identified its Anglo-British imaginary, we question the concepts, historical narratives and sociological tropes through which territoriality appeared within the textbook tradition. We turn to interrogate: the limited treatment of the 'union' concept; the deployment of the historic unions though which the UK was constructed as underpinnings for parliamentary sovereignty; the unitary state ideaevanescent but never wholly absent; and constitutional lawyers' treatment of UK nations and national identities. The territorial constitution we reconstruct, imagined throughor behindlaw's textbook tradition is both ever present and hidden in plain sight. Although subject to fundamental challenge through the mid-20th century decades, the period opened and closed with an ascendant orthodoxy of a parliamentary sovereignty within a unitary state.
Our second section turns to the 'Millennium Settlement' (Keating, 2021, forthcoming). Under the momentum of European integration, European human rights norms and devolution, the UK swung towards a more differentiated constitutional form. Drawing on a wider range of sources than for the earlier period, we ask: How did the new constitutional elements relate to older ones? Parliamentary sovereignty and the unitary state were never wholly removed from English constitutional discourse. However, they coexisted and sometimes competed with other ways of imagining the UK's constitutional order. For example, the union-state concept, rooted in comparative political analysis, was taken up by some legal analysts (Anderson et al., 2012) and also featured in parliamentary debates. Ambiguity and complexity seemed to be making space for pluralistic conceptions of the UK's constitutional order.
The final section considers the territorial implications of UK withdrawal from the EU. Concentrating on the UK Internal Market Act 2020, it interrogates the image of territory that underpins the legislation and questions its implications for the future of devolution and the territorial constitution. The White Paper -The UK Internal Marketthat preceded the legislation provided a highly stylized history of the Union, an exercise in motivated reasoning designed to lead to a preordained, centralizing conclusion. It included an explicit definition of the UK as 'a unitary state' (BEIS, 2020, para. 16). The Scottish and Welsh Government roundly rejected the UK Government's unitarist vision (Miles, 2020;Scottish Government, 2020). Each had an established view that the UK is a union-state, dependent on the ongoing consent of the constituent parts (Keating, 2018;McHarg, 2018;Wincott et al., 2021). Normally, explicit legislative consent is required from the devolved legislatures for any UK legislation that encroaches on their competences. Despite not gaining consent from any of the three devolved legislatures, the UK Government passed its internal market legislation into law.
Notwithstanding differences in tone, emphasis and substance within and across our three periods, the Anglo-British imaginary never faded awayit has, indeed, generally been dominant in UK constitutional discourse. Equally, English law, especially English constitutional law, is central to the imaginary. On the one hand, debates among mainstream constitutional lawyers are largely framed within the Anglo-British imaginary. On the other, images of constitutional law, and arguments made by lawyers and judges, occupy a central place within it, exemplified by A. V. Dicey's An Introduction to the Study of the Law of the Constitution (hereinafter The Law of the Constitution (1915 -first published 1885). While it is easy to treat Dicey's later influence as inevitable, we reconstruct the historical contexts in which his text emerged and exercised influence.
Our focus on territoriality clarifies the constitutional assumptions shared by the Anglo-British imaginary, assumptions otherwise occluded in the textbook tradition and wider constitutional debates. Typically revealed only in fragments, these shared territorial assumptions allow the Anglo-British imaginary to encompass both the Diceyean orthodoxy and its English critics Murray & Wincott, 2020;Wincott et al., 2021). Equally, the Anglo-British imaginary has long coexisted with distinct peripheral visionssometimes including alternative or overlapping imaginariesof the multinational UK state. Its relationship to the territorial legal jurisdictions in Scotland and Northern Ireland is complex, neglected and misunderstood. Understood in this way the grip of the textbook tradition and Anglo-British imaginary on constitutional thought was never truly broken. Ambiguity consequently swirls around the foundational matter on which our analysis is focused: the territorial character of the state.

TERRITORIALITY AND THE CONSTITUTIONAL LAW TEXTBOOK TRADITION
Delineating a tradition A new academic legal discipline focused on the common law in England emerged in the late 19th century. Its founding figures provided synoptic textbooks for various fields of law, since common law itself generates no comprehensive statements. These texts founded the distinctive 'textbook tradition' in legal education and legal practice. The academic lawyers walked a 'narrow ledge' (Sugarman, 1986) between the academy and the legal professions. Although at mid-19th century many saw common law as an 'empire of chaos and darkness' (p. 107), the academics deferred to judges' legal authority. Even so, textbooks sought to detect and elaborate a systematic underlying logic to the law. Although little known beyond legal studies, the textbook tradition retains some power over academic lawyers, privileging the 19th-century world-view of its proponents (Stychin, 2019). Even active critiques of the tradition's orthodoxies are, after all, often channelled into its modes of discussion. Notwithstanding the diversity it contained, the tradition was embedded across a corpus constitutional law texts: later authors use earlier texts as a foil for new arguments and as points of reference internal to the tradition.
Dicey helped found the textbook tradition, both in general and for constitutional law in particular. The constitution became a bastion for the tradition. The absence of a codified constitution, a reliance upon governing practices and a dearth of constitutional litigation created more space for the tradition in constitutional law, relative to other areas . For Dicey (1915), so-called 'writers of authority' helped fill the space. Their legacy strengthened the emerging academic tradition in constitutional law.
Dicey's Law of the Constitution (1915Constitution ( [1885), swiftly followed by Sir William Anson's The Law and Custom of the Constitution (1886), initiated the textbook tradition in constitutional law. Frederic Maitland's History of the English Constitution (1908) was based on lectures given at the University of Cambridge in 1887 and 1888. Thereafter, the tradition was shaped by four further texts, which appeared in multiple editions. The first two began to appear in the 1930s: Wade and Phillips's Constitutional Law (1931) and Ivor Jennings' Law and the Constitution (1933). Some scholars place Jennings' work outside the textbook tradition (e.g., Loughlin, 1992). But Jennings' analysis is an explicit counterpoint to Dicey. As we analyse it, the main tradition up to about 1980 is rounded out by Owen Hood Phillips's The Constitutional Law of Great Britain and the Commonwealth (1952), and Stanley de Smith's Constitutional and Administrative Law (from 1971). Dicey's influence is owed in part to the position of The Law of the Constitution as the initial source of English constitutional law's textbook tradition.
Generally, these lawyers eschewed explicit engagement with constitutional concepts; instead, they organized their works around institutions and their history. Dicey provided an acknowledged exception to this rule. He structured The Law of the Constitution (1915) around concepts parliamentary sovereignty, the rule of law and conventions. In turn, Jennings (1933) was organized as a critique of Dicey. In calling Dicey 'an artist' while describing himself as 'a surveyor', Anson (1886, p. v) captured this distinction. Dicey's conceptual work has proven more resilient than the stylized descriptive surveys.
Parliamentary sovereignty is a central theme through these debates. It is at the heart of a constitutional orthodoxy closely associated with Dicey (1915). Equally, the constitutional law mainstream encompasses sharp disputes over sovereignty. Powerful, anti-sovereignty counterarguments seem to have been in the ascendency during the 1930s-50s and between the mid-1990s and 2016. Territorial themes about union or the unitary state are ubiquitous, but often they serve as second-order manifestations of the Westminster Parliament's sovereignty. Despite their omnipresence, these concepts play a revealingly limited part in legal-constitutional narratives. In one sense fundamental, territorial themes are given inconsistent and limited treatment, marked by ambiguities and silences, gaps and occlusions. They rarely come into sharp focus.
Between the 1880s and 1980s, all our key writers engaged with the constitution's territoriality, if to varying extents. They did so in three main ways. First, the texts invariably engage with the idea of union, mostly through the Acts of Union that created Great Britain and then the UK. Many, but not all, discussed these Acts in staged histories of the UK, typically in the context of a wider discussion of the British Empire or Commonwealth. Second, some texts made use of other ostensibly territorial concepts, particularly federalism and the unitary state. A few used both territorial concepts and staged histories. Third, most also made observations about the UK's 'nations' and characteristics purportedly associated with them. We will consider these three perspectives in turn.

Concepts of union
All our constitutional law texts use the term 'union', mostly in the context of the Acts of Union between England and Scotland (1707) and Great Britain and Ireland (1801). Very few, however, even start to interrogate what union might mean conceptually. Starting with Anson (1892), most address the Acts of Union as way-points in the history of the UK's creation. In turn, they set the UK in a wider Imperial (and latterly, Commonwealth) context. They describe a process of territorial expansion, with England and English law at its core. Although sometimes prefigured by a brief initial historical sketch (Anson, 1886, pp. 11-30), the bulk of this discussion was generally positioned relatively late in the texts. The essentials of constitutional law could be established before consideration of the non-English parts of the UK state, it seemsand largely without reference to them.
Anson provided the template for this approach. His short initial outline was a highly stylized history of England, encompassing the Tudor and Stuart monarchies, without mentioning Wales or Scotland (only looking beyond England in its final paragraph). His main engagement with territoriality came much later in the text. It described the expansion of Crown powers from England to Wales followed by the Act of Union which 'settled the relationship of Scotland with England and then the Act of Union of Great Britain and Ireland' (Anson, 1892, pp. 208, 215). After discussion of the Home Office, Local Government Board and the 'adjacent islands', he proceeded to address the Colonial Office and territorial aspects of the Empire: the Crown's various colonies and protectorates. For our purpose, though, the absence of any major conceptual apparatus in Anson's approach is noteworthy. His legal history presented as a straightforward narrative description, albeit in highly stylized form. The term 'union' carries little conceptual weight or development, with one notable exception. Anson called England and Wales a 'complete union', but only after 1830, when 'the separate jurisdiction of the Welsh courts of great session was taken away, their judges abolished, and Wales brought wholly under the jurisdiction of the Westminster courts' (p. 208). If only implicitly, he seems to have imagined a set of distinct, asymmetrical internal unions.
A broadly Anson-style treatment of English territorial expansion appeared late on in a majority of the subsequent texts (de Smith, 1971;Hood Phillips, 1952;. The tradition changed decisively only during the 1970s, with de Smith (1971) providing one pivot. UK territorial structure began to be separated from consideration of the Commonwealth (which was itself subsequently often dropped altogether). Constitutional lawyers thereafter became more focused on the UK state, and discussion of its territorial structure appeared much earlier in their texts (e.g., Bradley et al., 1977, pp. 32-39). De Smith offered characteristically brusque, and again highly stylized, observations on this change. The sun, he said, was 'setting on the Empire that was to last a thousand years', with 'dark strangers … in our midst to stay' (De Smith, 1971, p. 30). Here, de Smith indulged in an ersatz sociology, adopting the language of the then contemporary accounts of 'race relations' (Waters, 1997), but without directly citing them. This tendency was not necessarily unique within the textbook tradition: Coffey (2020, p. 204) has observed an earlier aversion to immigration into the UK in Wade and Phillips (1931, pp. 178-179).
Dicey and Jennings eschewed Anson's historical-narrative approach. They engaged with the Acts of Union in a different way. Both focused on Westminster's powers, and the apparent restrictions that the terms of union placed on Parliament. The Acts of Union were pivotal to Dicey's argument for parliamentary sovereignty, since each provided putatively permanent protections for major institutions in Scotland or Ireland. Both provided for a distinct legal system. In Scotland, the position of the Presbyterian majority church and the education system associated with it were also protected. In Ireland, the establishment of minority Anglicanism in unity with the Church of England was entrenched. Protecting key institutions 'forever after' and 'in all time coming', potentially made the legal texts underpinning the Anglo-Scottish union fundamental constitutional texts for the UK. On the contrary, Dicey argued, Scotland's Parliament and parliamentary tradition had been subsumed within and incorporated under Westminster. His 'strongest proof' of Westminster's sovereignty was that it had reformed some of these seemingly permanent protections (Dicey, 1915, pp. 21-22).
For Dicey, this single breach shattered the whole structure of apparently permanent constitutional protections for Scotland. If changes to the terms of union had been forced through Parliament over objections from Scotland or most Scottish MPs, then Dicey's position might have been compelling. His worked example concerned ending a condition on professorial positions at Scottish universities, which had restricted them to people who subscribed to the Confession of Faith (1915, p. 22). This change, however, reflected ongoing processes of social changewhich flowed through Scotland's four 'ancient' universities before making a mark upon Oxford and Cambridge. Dicey's exemplary breach to the terms of the British-Irish union, the disestablishment of the Church of Ireland under the Irish Church Act 1869, is at least as peculiar. This reform reflected Ireland's majority religion and increased territorial differentiation within the UK; it better corresponds to a union-state model.
In contrast to Anson, Jennings focused only on the domestic state and touched early on its territorial formation. While he built his reputation on a critique of Dicey, the Acts of Union were peripheral to his analysis. Without endorsing Dicey's view of sovereignty, Jennings described the UK as formed by 'extending some English institutions to Scotland and in other respects leaving Scottish institutions intact' and then 'applying British institutions to Ireland (Jennings, 1937, pp. 7-8).
By contrast, de Smith (1971) marked a textbook reinvestment of Dicey's position. His reason for dismissing the idea of the Act of Union as 'fundamental law for the United Kingdom' was as revealing as it was blunt: 'the general consensus of opinion among constitutional lawyers (at least south of the border)' rejected claims that it had a special character (p. 32).
Aside from Anson's brief allusion to 'complete union', lawyers devote very little attention to union as a concept, despite the attention given to the Acts of Union. Hood Phillips is the only exception, explicitly contrasting different types of union. Starting in 1952, but dropped by 1967, he compared unitary and composite state forms, and provided a concise discussion of four types of composite state: Federalism, Personal Union, Confederation and Real Union. According to Hood Phillips, Real Unions (which share some governmental institutions) tend to emerge from Personal ones (where two distinct countries are ruled by one monarch). Like de Smith's allusion to dark strangers, this discussion has the quality of ersatz social science. It may owe something to John Marriott, an early 20th-century, Oxford-based political scientist. Marriott's English Political Institutions (1913), contrasted unitary and composite states, but delineated the latter in only confederal, personal union and federal forms.
The idea of 'union' was variously invoked within stylized historical narratives, contested readings of the Anglo-Scottish union, or, sporadically, as one of several, poorly theorized state forms. It was ever present yet ambiguous and thoroughly neglected in the textbook tradition.
Unitary, unitarian and non-federal states The same tendency within the tradition manifests in various conceptions of the UK state. In the Law of the Constitution, Dicey called '[t]he principle … which gives its form to our system of government' 'unitarianism' ('a foreign but convenient expression') (Dicey, 1915, p. 74). For all his work's enduring influence, this term has not lasted. Dicey later also used a different term, one more familiar today: the unitary state (pp. xcii, xciv, xcv, xcvii). However, his main text retained unitarianism: the two appear synonymous. As an English constitutional law concept, the unitary state's origins are obscure and it was not picked up after Dicey for several decades. Until Hood Phillips in 1952, it was not widely used in constitutional law textbooksand only in the 1973 edition did Hood Phillips explicitly categorize the UK as unitary. de Smith (1971) was the first leading textbook after Dicey to define the UK as a unitary state.
By describing the state as a single unit, the term 'unitarianism' or 'unitarism' suggests a territorial meaning. On investigation, though, the concept often has little substance. It is defined partly by negation: unitary since it is not federal (Dicey, 1915, pp. xcv, 80;de Smith, 1971, p. 33;or for Hood Phillips, 1952, not 'composite'). These writers were concerned with the disposition of powers, not the government of territory. Under federalism powers are divided, dispersed and balanced; unitary arrangements concentrate them.
Dicey and de Smith thus used 'unitarianism' or the unitary state to describe and underscore parliamentary sovereignty. Dicey's unitarianism meant 'the habitual exercise of supreme legislative authority by one central power … the British Parliament' (Dicey, 1915, p. 74), although he located sovereignty in three institutions: the Crown and the two Houses of Parliament (pp. 27, 62, 65), arguably more a trinity than a unity. Were the UK federal, for de Smith, Parliament 'would not be omnicompetent' (de Smith, 1971, p. 33), and the constitution, therefore, must be unitary.
From a political science perspective, Marriott made these arguments more sharply than any lawyer. He identified the UK as 'a unitary state of the ordinary kind' (Marriott, 1913, p. 306). What is more, 'England, and indeed, the British Empire must be assigned formally to the [unitary] category (p. 17), reflecting 'not … practical working but … legal form'. 'Legislative sovereignty' is key here. It is 'vested for the whole British Empire in the "Imperial" Parliament, i.e., in King, Lords and Commons, sitting at Westminster' (p. 16). In other words, like Dicey, Marriott placed a primarily legislative unit formally at the heart of politics and the constitution. This 'unit' was capable of embracing hugely diverse territories (with formal variation from this model for the Dominions only coming, under such accounts, under the Statute of Westminster 1931).
Later elected as an MP, Marriott returned to these issues in that role. He distinguished constitutional theory and 'political practice': the 'regular text books', he said, 'perfectly accurately describe … a unitary Constitution', but in 'political practice' there 'is a very great deal more of what is loosely called federalism in the working of the United Kingdom at present than is commonly supposed' (Marriott, 1919). He pointed to the UK's three judicial systems and the character of its legislation: the 'Legislature may be unitary; the resulting legislation is not.' Foregrounding the concept's territoriality elementuniform government, administration and lawweakens UK claims to unitary status.
Aside from Anson's early allusion to the complete union of England and Wales from 1830, English constitutional lawyers paid remarkably little attention to the UK's multiple legal jurisdictions. Bradley et al. (1977, p. 38) came to emphasize the constitutional significance of legal diversitythe three distinct jurisdictionsonly in the 1970s. They were not, though, persuaded of the relevance of federalism: '[f]or many purposes' they argued the UK was 'unitary … since there is no structure of federalism' (p. 38). During the 1960s, constitutional law texts began to appear from outwith England. They expressed similar uncertainty about UK state classification. From a Scots Law perspective, Mitchell described the constitution as 'neither federal nor strictly unitary' (Mitchell, 1964, p. 2). While emphasizing Northern Ireland's distinctiveness, Calvert rejected a federal frame as 'likely to lead to misunderstanding' (Calvert, 1968, p. 60). Available concepts did not fit UK-wide territorial-constitutional practice well.

England and Englishness
Ostensibly engaged in legal analysis, many constitutional lawyers grounded their work in an ersatz sociology of the nations and national identities within the UK, particularly of England and Englishness. For Dicey, England meant 'the country known, and famous, as England before the legal creation either of Great Britain or of the United Kingdom'. He conflated 'the greatness of England with the prosperity of the United Kingdom and the greatness and good government of the Empire'. Dicey also refuted Scottish complaints that 'England' and 'Great Britain' were often being conflated: these 'prejudices which … kept Scotsmen and Englishmen apart, have in reality vanished' (Dicey, 1915, p. cvi).
Wade and Phillips also grounded their work in sweeping assumptions about the English character: 'Englishmen are correctly credited with a capacity for self-government', but 'abstract questions … are repugnant to them' and 'they are impatient of analysis of their methods' (Wade and Phillips, 1931, p. 4). With similar Anglo-centrism, de Smith said 'England (or Britain, or the United Kingdom) has been a sovereign state for many centuries [so] the question of adopting a constitution on independence has not arisen' (de Smith, 1971, p. 29). His imagined Anglo-British state also had a 'high degree of ethnic homogeneity'it is one 'intractable political (and religious) problem was Ireland': 'partitioning' made it 'possible to sweep this embarrassing problem under the carpetuntil 1969 (pp. 29-30).
The tendency to conflate England with Britain and the UK, while downplaying forms of territorial differentiation, also helps to explain the ascendancy of the unitary state view. By privileging England, it facilitates an understandingan imaginaryundisturbed by peripheral idiosyncrasies.
Overall, then, we see how territorial concepts are evanescent in the textbook tradition. They appear as stylized vignettes within historical 'surveys' or as ersatz and deracinated social science. More conceptual (or, for Anson, 'artistic') treatments press the Acts of Union and unitary-type state concepts into a service role for their primary preoccupation: parliamentary sovereignty. Their use has thus been inconsistent and superficial. English legal perspectives, charged with Anglo-centric visions of the UK nations and the putative national characters of their peoples, dominated this scholarship. Although riven with ambiguities and abeyances, a remarkably resilient Anglo-British imaginary connects these elements of UK constitutional history.

MAKING THE MILLENNIUM SETTLEMENT, 1997-2016: FAREWELL TO THE UNITARY STATE?
At the turn of the century, the Anglo-Britain conjured up by the textbook tradition looked increasingly anachronistic. Nevertheless, the two decades that followed would bear witness to its enduring influence.
By 1997, the constitutional landscape had changed significantly (Keating, 2021, forthcoming). The growth of administrative law from the late 1960s saw courts asserting supervisory jurisdiction over government action after a prolonged period of constitutional 'irrelevance' (Stevens,  Wade (1996). The pace of constitutional change accelerated dramatically under New Labour after 1997. Following referendums, bespoke regimes of legislative devolution were established in Scotland, Wales and Northern Ireland. The Human Rights Act 1998 ushered in a new system of human rights law centred on the European Convention on Human Rights. By 2005, senior judges were openly expressing doubts about the unqualified reach of parliamentary sovereignty. In 2009, the UK Supreme Court replaced the Judicial Appellate Committee of the House of Lords and Judicial Privy Council. The various constitutional functions accumulated by the courts since the 1970s were consolidated in one UK-wide institution.
The Millennium Settlement struck at two elements that had animated lively debates among English constitutional lawyers: 'parliamentary sovereignty as an authoritative legal doctrine and … the institutional centralism of the parliamentary state' (Walker, 2014, p. 535). The UK, it seemed, was being rendered 'more multipolar in its sources of authority and less institutionally concentrated' (p. 536). Bogdanor (2009, 53-231) argued that a 'New British Constitution' was emerging. The Human Rights Act represented the 'cornerstone' of this new order, while devolution had 'turned Britain from a unitary to a quasi-federal state ' (pp. 53, 116). They heralded a 'transition from a system based on Parliamentary sovereignty to one based on the sovereignty of a constitution, albeit a constitution that is inchoate, indistinct and still in large part uncodified' (p. xiii).
Devolved institutions' powers deepened over time. In Wales, the National Assembly was gradually transformed from a hybrid institution with limited law-making powers into a full legislature with significantly wider competences. The Scottish Parliament gained further competences under the Scotland Acts (2012 and, particularly, 2016; after the 2014 independence referendum and Smith Commission recommendations). The Belfast/Good Friday Agreement played an innovative constitutional role in Northern Ireland. Marking a new beginning in 1998, it enhanced stability and supported the peace process. Equally, devolved government operated only intermittently. Todd (2017, p. 301) argues it failed to emphasize constitutional principles or embed them in the politics of Northern Ireland. After the 2006 St Andrews Agreement, the Executive returned to operation. Further powers were then transferred, including over justice and policing in 2010. New political identities did emerge around these institutionsand seem likely to grow in importance with generational change. In the meantime, the framework has relied heavily on informal British and Irish oversight. It is at risk if the attention of either state slips away.
Legislative guarantees wove the devolved institutions further into the UK's constitutional fabric. The Scotland Act 2016 declared Scotland's devolved institutions 'a permanent part of the United Kingdom's constitutional arrangements'. It placed the Sewel Conventionby which the UK Parliament ordinarily refrains from legislating with respect to devolved competences except with explicit consent from the relevant devolved legislatureon a statutory footing. The Wales Act 2017 has parallel provisions. It established a reserved powers devolution model, bringing Wales into line with Scotland and Northern Ireland.
Alternative conceptualizations of the UK state, old and new, became more commonplace: not unitary but 'quasi-federal' (Bogdanor, 2009), a 'union-state' or a 'state of unions' (Douglas- Scott, 2016;Mitchell, 2006). After the Scottish National Party (SNP) established a minority government, opposition parties triggered the creation of the Calman Commission. It directly rejected the unitary state conception of the UK (Pittock, 2012, p. 17, 18). While observing no consensus over 'what type of state the Union embodies', the House of Lords Select Committee on the Constitution likewise pronounced the traditional unitary account dead (albeit while acknowledging that a more nuanced view was available): 'Where it may once have been correct to say that it was a unitary system (with power centred in Westminster), that is no longer accurate' (Constitution Committee, 2016, p. 15). For the devolved governments devolution was permanent; devolved institutional consent was required to make legitimate changes to the devolution settlements; and they claimed a general constitutional presumption in favour of devolving (as opposed to reserving) power (Scottish Government, 2016;Welsh Government, 2017).
Nevertheless, a unitary state image continued to mark UK Government thinking before the 2016 referendum (Douglas- Scott, 2016). Prime Minister David Cameron repeatedly emphasized the UK's singularity, including his historical allusion to 'our ancient democracy' following the 2014 Scottish independence referendum. The UK Government used conventions and concordats to manage devolution, at least where EU law was not at issue: no attempt was made to build formal legal relations around a concept of subsidiarity (Scott, 2001). Despite sweeping constitutional change, core tenets of the Anglo-British imaginary, inherited from older constitutional discourses, remained alive and well.
Before the 2016 referendum, these ambiguities were occluded by the terms of EU membership. Each arrangement imposed a bar on devolved legislation or making policy contravening EU law (Scotland Act 1998, s.29(2)(d), s.54 and s.57(2); Government of Wales Act 2006, s.108(6)(c) and s.80(8); Northern Ireland Act 1998, s.6(2)(d) and s.24(1)). They were obliged to uphold European Single Market rules (promulgated under the Treaty on the Functioning of the European Union, Articles 34-35). No domestic rules were needed to prevent barriers to trade emerging. The House of Lords' European Union Committee concluded that 'the European Union has been, in effect, part of the glue holding the United Kingdom together since 1997. … In practice, the UK internal market has been upheld by the rules of the EU internal market' (HL EU Committee, 2017, para. 277). The EU law concept of subsidiarity gives priority to relevant domestic legislatures' implementation of certain rules, particularly where EU Directives provide broad frameworks for law-making (Hunt, 2010). EU glue thus provided a flexible bond which allowed devolved institutions to create distinct legal arrangements within the UK.
Scotland introduced minimum pricing for alcohol under the Alcohol (Minimum Pricing) (Scotland) Act 2012. It was based on flexibility permitted in EU law with regard to free movement of goods. Single Market protections did not 'preclude prohibitions or restrictions on imports, exports or goods in transit justified on grounds of … the protection of health and life of humans' (Treaty on the Functioning of the European Union, Articles 36). Although ostensibly adopted for health purposes, a coalition of drinks manufacturers and distributors challenged the measure as a disguised trade restriction (including from other parts of the UK). The UK Supreme Court recognized that the Scottish Parliament and government had put a greater weight upon 'combatting alcohol related mortality' than on 'the benefits of free EU trade and competition' (Scotch Whisky Association, 2017, para. 63). The court concluded it 'was a judgment which it was for them to make, and their right to make it militates strongly against intrusive review by a domestic court' (para. 63). Ironically, this clear statement of the flexibility for devolution under EU law came after the 2016 Brexit referendum, as the rupture of the Millennium Settlement's EU-based legal structure loomed.

CLASHING CONSTITUTIONAL VISIONS
Brexit and the territorial constitution The 2016 Brexit referendum flagged up divisions in the attitudes to EU membership across the UK's constituent parts. They have reverberated through the debates that followed. Popular support for EU membership in Scotland and Northern Ireland generated vociferous complaints that both are being pulled out of the EU against their will. The nature of the land border between Ireland and Northern Ireland and the reliance on common EU rules to support North-South The Anglo-British imaginary and the rebuilding of the UK's territorial constitution after Brexit: unitary state or union state? 705 TERRITORY, POLITICS, GOVERNANCE cooperation under the Belfast/Good Friday Agreement ultimately necessitated special arrangements contained within the Ireland/Northern Ireland Protocol (Withdrawal Agreement, 2020). The subsequent struggles over its implementation, however, illustrate the complexity of differentiating market arrangements for part of the UK, even where a sea border facilitates the tracking of goods movements (Weatherill, 2020). There is no easy way for part of the UK to remain within the Single Market for goods and also to enjoy 'unfettered' access to other parts of the UK's internal market. Even so, unless the UK Government opted for an EEA membership-based Brexit, EU Single Market access of that kind was the Scottish Government's explicit aim (Scottish Government, 2016, p. 26).
This put the Scottish Government on a collision course with the UK Government. Prime Minister Theresa May's administration, assured that it could ultimately sideline devolved institutions by the Supreme Court's Miller judgment (R (Miller) v Secretary of State for Exiting the European Union, 2017), maintained that the apportionment of powers within the UK after Brexit would primarily be a matter for Whitehall and Westminster (Keating, 2021). Its European Union (Withdrawal) Bill was thus based on repatriating EU law powers to Westminster, some of which might be later released to the devolved institutions (Constitution Committee, 2017, paras 73-74).
The 2017 general election outcome upset these plans. The Conservative-DUP confidenceand-supply arrangement made negotiating special arrangements for Northern Ireland much more complex. May's precarious position in Parliament, moreover, depended on her 13 MPs from Scotland. If these MPs acquiesced to the return of powers to London, without specific protections for devolution, their seats could become vulnerable. The withdrawal legislation was thus reworked to reflect the presumption 'that powers returning from the EU should sit at a devolved level' and the watchword became the 'framework' nature of the outline arrangements for the UK's internal market (Lidington, 2018). David Lidington's concessions suggested a negotiated approach to the UK's internal market. They indicated substantial scope for differences to emerge (and indeed, continue) across the UK with regard to environmental and product standards. Reflected in amendments to the legislation (European Union (Withdrawal) Act 2018, s.12), these arrangements conditioned the subsequent Intergovernmental Agreement (Intergovernmental Agreement on the European Union (Withdrawal) Bill and the Establishment of Common Frameworks, 2018).
Although Lidington faced criticism from sections of his party for giving ground to the devolved governments, the 2018 Act, as amended, retains Westminster's ability to make overarching arrangements across the whole UK, and to block off potentially conflicting devolved developments (regarding Scotland, see Scotland Act 1998, s.30A(1)). Westminster has to seek, but crucially need not receive, consent from the devolved institutions to impose restrictions on their legislative remit (s.30A(4)). In any field that could relate directly to the UK internal market, or its 'level playing field' underpinnings, the UK Government holds a trump card, provided that it was not used to curtail existing devolved legislative competences (s.30A (2)).
This blueprint for constructing the UK's internal market nonetheless remained contested: the Scottish Government did not accept the Intergovernmental Agreement and Northern Ireland's devolved institutions remained suspended until January 2020. Only the Welsh Government committed to the Agreement's formal intergovernmental process. The process then stalled amid the challenges of finalizing the UK-EU Withdrawal Agreement. Beyond a shortage of institutional bandwidth to manage both tasks simultaneously, the nature of the Withdrawal Agreement would have consequences for the shape of the UK internal market's arrangements.
Under May, the Withdrawal Agreement's 'backstop' had implications for the whole of the UK's market in goods (Draft Withdrawal Agreement, 2018: Protocol on Ireland/Northern Ireland, Articles 6-12). EU law might well have continued to provide some underpinnings for the UK's internal market. Through his efforts to sever the UK market in goods (at least in as it applies to Great Britain) from EU law, however, Prime Minister Boris Johnson brought the necessity of constructing, and indeed imposing, 'common frameworks' back to the forefront of the UK Government's Brexit planning.
In July 2020, the UK Government's Internal Market White Paper heralded the end of the Brexit transition/implementation period as bringing with it the 'single biggest transfer of powers to the devolved administrations in history' (BEIS, 2020, para. 12), but one which is predicated on the UK being 'a unitary state' which would also see an expanded role for Whitehall in overseeing these new arrangements (para. 16). The UK Government fixated upon the risk that the repatriation of powers from Brussels to devolved institutions in Scotland and Wales would produce 'unnecessary regulatory barriers could emerge between the different parts of the UK', requiring Westminster legislation to ensure 'principles of mutual recognition and non-discrimination' (para. 28).
The concept of a 'unitary state' was left unexplained; a presumption which gained little support from the strangulated paragraphs on the history of the Union (BEIS, 2020, paras 58-64). The Act of Union was presented as dissolving all barriers to trade between the UK's constituent parts, even though exceptions were maintained or subsequently developed. Excise officers continued to enforce a revenue border between England and Scotland until 1855 due to differentiated duties on whisky (Denton & Fahy, 1993, p. 3). In the 20th century numerous barriers to trade developed between Northern Ireland and other parts of the UK (Murray & Rice, 2020, pp. 18-19). Restrictions on movements of meat and livestock to Northern Ireland from other parts of the UK, imposed to protect against the spread of foot-and-mouth disease, only ended after UK membership of the EEC (O'Neill, 2000, p. 217). Centralization did not preclude barriers to trade emerging.
Either the carapace of the 'unitary state' provided by the textbook tradition was thought so flimsy that an extended explanation would just serve to expose holes, or the White Paper's account is so controversial in Scotland, Wales and Northern Ireland as to warrant its exclusion. Or both. All of which speaks to the instability of the theoretical underpinnings of the UK's constitutional arrangements. The omissions from the White Paper make for a lopsided account of how the UK's internal market functioned after Brexit.
Although it deliberately emphasized the binding effect of EU law on the devolved institutions until the end of the Withdrawal Agreement's transition/implementation period (BEIS, 2020, para. 64), it left the outline nature of many EU directives and the public interest exceptions to the operation of core Single Market rules wholly unexplained. This transparent effort to underplay the considerable autonomy EU law gave to devolved institutions, enabling measures such as Scotland's minimum alcohol pricing legislation, unsurprisingly provoked the ire of the Scottish Government, which dismissed the proposals a 'power grab' by Westminster and Whitehall (Pooran, 2020). The radically divergent accounts of the White Paper set up a dramatic legislative clash between the Scottish Government and the UK Government over the Internal Market Bill.

The UK Internal Market Act 2020
Just as the introduction of the Internal Market Bill seemed set to ignite this clash, attention was wrenched to an even more controversial element of the legislation. The other particularly pressing internal market challenge facing the UK Government was how to accommodate Northern Ireland within the new UK arrangements when it essentially remained part of the EU Single Market for goods under the Withdrawal Agreement. This aspect of the White Paper lacked much detail, beyond a commitment to 'legislating for full unfettered access for Northern Ireland goods to the UK market by the end of this year' (BEIS, 2020, para. 29). When the Internal Market Bill was presented to Parliament in September 2020, it became evident that this vagueness had concealed the UK Government's willingness to breach its Protocol commitments until The Anglo-British imaginary and the rebuilding of the UK's territorial constitution after Brexit: unitary state or union state? 707 this key point in EU-UK negotiations over the future relationship and the Protocol's implementation. Justifying this breach in Parliament, Johnson reached for a unitary account of the UK's constitutional order, laced with jingoism: [t]he EU is threatening to carve tariff borders across our own country, to divide our land, to change the basic facts about the economic geography of the United Kingdom and, egregiously, to ride roughshod over its own commitment under article 4 of the protocol. (Johnson, 2020) These histrionics glossed over the exceptions built into the operation of Article 4 and the ongoing efforts of the Withdrawal Agreement's Joint Committee to navigate these issues. They also neglected how Johnson's government had agreed terms which supposedly permitted such developments.
Notwithstanding the Part 5 of the Bill's proposed assault upon the Protocol's terms on exit summary declarations, state aid and onward movement of goods through Northern Ireland into the Single Market, its provisions left goods produced in Northern Ireland subject to the standards required by the European Single Market; the UK Government has sought no power to constrain the application of those standards (Murray & Rice, 2020). Northern Ireland's institutions would also, like those of Scotland and Wales, be subject to the mutual recognition and non-discrimination principles of the UK internal market and thus obliged to allow goods from Great Britain into the market in Northern Ireland. As the product standards diverge after Brexit, this will produce increasing strains on the coupling between the two legal orders.
If certain product standards become lower in Great Britain, these principles mean that businesses in Northern Ireland will face being undercut by goods produced in other parts of the UK or imports entering the UK internal market under trade deals. If other product standards are lowered in the EU Single Market, then the UK Government's commitment to unfettered access for goods from Northern Ireland would be open to exploitation; traders could use this commitment to move goods from the Single Market through Northern Ireland and into other parts of the UK. The consequences of regulatory divergence remain, and are not negated by UK Government pledges that its arrangements mean that 'new barriers to trade in goods and services within the UK are avoided' (BEIS, 2020, para. 146).
Soon after the Joint Committee reached an agreement over the application of the Protocol the UK Government withdrew the law-breaker clauses from Part 5 of the legislation. The provisions that remained reflected the agreed UK-EU position on state aid and on exit summary declarations for goods moving from Northern Ireland into the remainder of the UK. Michael Gove crowed that the UK's hard-ball approach had won these concessions, and that '[h]aving put beyond doubt the primacy of the sovereignty of this place [the UK Parliament] as we leave the EU, we rest safe in the knowledge that such provisions are no longer required' (Gove, 2020). There is, however, little to suggest that these technical issues would not have been resolved through the Joint Committee process without applying such pressure, with all of its attendant damage to trust in the UK's international commitments.
In December 2020, Gove, however, had another good reason to be happy. The UK Government's manufactured crisis over Part 5 had sucked much of the oxygen out of debates over other aspects of the Internal Market Bill. This is not to say that there was not vigorous opposition to the approach adopted by the UK Government within Parliament. Major sections of the Bill were turning the loose arrangements and common frameworks approach under the Withdrawal Act 2018 into a process of much harder edged centralized control, but the media attention upon the Bill, remained focused upon the struggles over its impact on the Ireland/Northern Ireland Protocol.
Under the new legislation, the devolved institutions would be able to enact their own product safety and environmental standards, as had been promised by David Lidington in 2018, but section 1(2) ensured that mutual recognition and non-discrimination would operate to ensure that no devolved institutions can use such standards to restrict the sale of goods produced in other parts of the UK. Section 2(1)(a) also provides that the devolved institutions will not be able to impose their own higher standards to block imports under post-Brexit trade deals; '[e]nsuring the UK remains a coherent and integrated economy will be key to fostering all the opportunities in trade' (BEIS, 2020, para. 45). The protections for the internal market, moreover, extend in section 8 into indirect discrimination, which occurs where a devolved measure: displays no direct discrimination but applies to, or in relation to, the incoming goods in a way that puts them at a disadvantage, has an adverse market effect and cannot reasonably be considered a necessary means of achieving a legitimate aim. (Constitution Committee, 2020, para. 92) The combination of these provisions means that many of the powers being extended to the devolved institutions under the 2018 Act are neutralized, in that they can only raise standards applicable to businesses in their own jurisdictions, which would amount to creating competitive disadvantages for them with regard to businesses in other parts of the UK or imports. Minimum alcohol pricing exemplifies the dispute over this shift from the protections of the EU Single Market to the UK Internal Market. The enactment of this measure in Scotland restricted trade in affected products from EU countries and from other parts of the UK. It was nonetheless justified on public health grounds. Public health grounds remain a legitimate aim by which an indirectly discriminatory measure can be justified under section 8(6), but many of the broadest legitimate aims under EU lawsuch as 'public policy'are not replicated.
The White Paper had made explicit assurances over the maintenance of devolved competences; 'the devolved administrations would retain the right to legislate in devolved policy areas that they currently enjoy' (BEIS, 2020, para. 32). There is a pale reflection of this pledge in section 4 and 9 of the Act, which protects existing devolved measures from the operation of the mutual recognition and non-discrimination principles, but does not extend this protection to future enactments of a similar nature. The UK Government did make a concession under section 3 to confirm that the legislation would not affect minimum alcohol pricing, and to provide for consultation with the devolved governments should the UK Government seek to use delegated legislation to alter the Act's market access principles, but with attention so focused on Part 5, it was able to resist other amendments to these provisions. Taken as a whole, the Internal Market Act imposes greater restrictions upon the competences of the devolved institutions than the provisions of the EU Single Market which it replaced, in spite of pledges to use common frameworks to address these issues. Lord Hope, responsible for many of the leading judgments relating to the first two decades of devolution, regarded the legislation's terms as deliberately confrontational; 'this Parliament can do what it likes, but a different approach is essential if the union is to hold together' (Hope, 2020).
The controversy over the Act ensures that its safeguards for devolved competences are certain to be litigated, with the devolved legislatures seeking to persuade the courts to interpret them generously. The established jurisprudence of devolution, taken alongside ministerial pledges over Brexit providing a power surge to the devolved institutions, could encourage such an approach. The Courts have long emphasized their respect for the broad legislative competence of the devolved institutions (Reference re The UK Withdrawal from the European Union (Legal Continuity) (Scotland) Bill, 2018, para. 25). For devolved legislation to be called into question because it relates to a reserved matter, a provision must have 'more than a loose or consequential connection' to that matter (Martin v Lord Advocate, 2010, para. 49). This position rests upon the underlying statutory purpose of devolution being to allow for autonomous legislative action The Anglo-British imaginary and the rebuilding of the UK's territorial constitution after Brexit: unitary state or union state? 709 TERRITORY, POLITICS, GOVERNANCE (de Mars et al., 2018, pp. 125-129). This commitment could influence the interpretation of the UK's internal market arrangements, with the 2020 Act's terms allowing as much scope for judicial action as EU law before it. The Internal Market Act was enacted, as has become the pattern with Brexit legislation, notwithstanding rejections of legislative consent from the devolved legislatures. The stark dichotomy between power surge/power grab readings of the legislation is the product of very different ways of thinking about the UK's constitutional order. It results from long unresolved tensions between understandings of that order, and its convoluted provision means that for all that the 2018 Withdrawal Act might allow the devolved institutions to exercise new competences from 31 December 2020, these are effectively circumscribed and overseen from London. The devolved governments are justifiably wary that the UK's internal market legislation has been enacted to prevent them from pursuing the sort of distinct policies once facilitated by EU law.

CONCLUSIONS
Dicey's emphatic rejection of Home Rule for Ireland provided a dry run for many of the criticisms which, were more than a century later, ultimately levelled at devolution. For Dicey such a scheme amounted to 'semi-independence', which 'makes it easy … to attribute every mishap to the absence of absolute freedom' (Dicey, 1973, p. 184). He believed that any move towards devolution would precipitate the UK's collapse. No devolution scheme could give the other parts of the UK sufficient 'legislative independence' to satisfy their aspirations, but which would also 'leave to England as much supremacy as may be necessary for the prosperity of the United Kingdom, or for the continued existence of the British Empire' (p. 281). England was the linchpin of his Union, and the Union and Empire were necessary because they increased England's heft on the world stage.
After Dicey, these preoccupations lived on in the Anglo-British constitutional imaginary. Its legal components have proved flexible and capacious, even as other aspects of constitutional discourse ebbed and flowed. The imaginary survived while key elements of Dicey's framework were rejected in the 1930s and 1940s. After being robustly reasserted in the early 1970s, from the 1990s Diceyean orthodoxy coexisted uneasily with the UK's EU membership and devolution. Though ambiguous and tentative, there were some signs then that the Anglo-British imaginary might be transcended. Brexit has brought it, and Dicey's orthodoxy, back with a vengeance.
Accounts of the UK as unitary state subject to parliamentary sovereignty must be open to reassessment. It is no good claiming, as Dicey did, that devolution is too difficult to operationalize. It is now a fact of the UK constitutional order. The decades-long effective operation of this order under the Millennium Settlement was, however, heavily reliant upon EU structures. It was embedded in EU law and its concept of subsidiarity. Once the textbook writers helped to create the constitutional tradition they purported to describe. By emphatically reasserting Diceyan sovereignty and the state's unitary nature the UK Government may now be helping to make a devolved territorial constitution unsustainable. If that territorial constitution is to be sustained without EU support, the Millennium Settlement now requires significant adaptationincluding a decisive move beyond the Anglo-British imaginary. It is hard to see how the UK can remain intact, unless the UK Government changes its historic disposition. Loose talk of framework arrangements must give way to an explicit and legalized commitment to the operation of subsidiarity post-Brexit. Alternatively, of course, the UK could break-up. Even that transformation would inevitably require further consideration of shared governance arrangements for and across the Atlantic archipelago.