II. Domestic analogy and the rule of law in the relations of states

The use of domestic analogies in structuring thinking about the nature of international relations, as well as the potential for legal rules to regulate these relations, has a long, even if not uncontroversial pedigree. ^{Footnote 30} Historically, the analogy was explicit in Thomas Hobbes’ view of the problem of international (dis)order, wherein he compared the misery of the state of nature to the continual interstate hostilities of the international realm,^{Footnote 31} just as it featured in the works of those following Hobbes in the social contract tradition, such as John Locke,^{Footnote 32} Jean-Jacques Rousseau,^{Footnote 33} and Immanuel Kant.^{Footnote 34} Whilst by the twentieth century there would be an increasing aversion to the use of the analogy amongst international lawyers and international relations theorists alike, particularly those keen to point to the unique, or sui generis character of the international system,^{Footnote 35} analogical reasoning has continued to play a prominent role in structuring international thought. Not only has the analogy provided much of the background framing for successive proposals for institutional reform, from the Hague conferences at the end of the nineteenth century to the creation of the League of Nations and UN in the wake of successive world wars,^{Footnote 36} but it has also informed much of the scepticism that has greeted such institutional reforms, insofar as they are seen as failing to deliver the kind of constitutional architecture underpinning the realisation of the rule of law in the domestic context.^{Footnote 37} In this regard, one might well agree with Hidemi Suganami that the domestic analogy is necessarily assumed by ‘any contemporary writer on international affairs who attributes the instability of the international system primarily to its decentralized structure’.^{Footnote 38}

The analogy is thus as ubiquitous as it is controversial, but we should perhaps not be too surprised in this respect. It is not, or not just, that we international lawyers or theorists are drawn inevitably to analogical reasoning through habit or experience – that we are all, in Joseph Weiler’s words, ‘habitual sinners in this respect’.^{Footnote 39} Rather, this reasoning is so recurrent precisely because what modern international law actually is as a self-constructed project of (largely European) legal imagination presumes the domestic analogy. For as long as international law has been thought of as a broadly consensual, institutional practice, its core doctrines (sovereignty, statehood, self-determination, sources, etc) assume something like the rule of law as the basis of mediating the ‘individual’ sovereignty of states.^{Footnote 40} As Martti Koskenniemi argues:

As such, the rule of law has entered international legal imagination as a necessary presumption of the very idea of international law as an institutional practice, the legitimacy of which is related to the self-authorship (non-arbitrariness) and autonomy (objectivity) of international legal rules. However, insofar as this understanding has always seemed somewhat problematic when applied at the international level, where the decentralised institutional structure of international law seems to undermine its autonomy and objectivity in practice,^{Footnote 42} the rule of law emerges also as an equally important reformist ideal, stressing the important work to be done to ‘complete’ such a vision of ordered international relations.

I will come back to this apparent tension in the concept in due course. My more immediate point, however, is simply that we should not be too quick to dismiss the importance of analogical reasoning in gaining an understanding of the nature of modern international law. Where the resort to analogy might become more problematic, however, is in the move from descriptive comparison to normative prescription. In other words, one should be cautious where the differences revealed through analogical reasoning are treated simply as faults in the object of description (international law), rather than themselves revealing important features that individuate the international legal form or, indeed, demonstrate necessary truths about law more generally. As such, the warnings of those such as Bull remain prescient in highlighting the ever-present danger that the analogy merely masks certain taken-for-granted, though by no means uncontroversial, normative presumptions about the state and/or the international system. This is a point well made by Thomas Poole, who acknowledges the harmless, comparative aspects of analogical reasoning, whilst cautioning against the kinds of argumentative ‘short cuts’ that occur when analogies buttress normative assertions. As he notes, ‘[t]here may be plenty of very good reasons, even if such-and-such is like so-and-so, not to treat the two in the same way. It depends.’^{Footnote 43}

Poole’s caution thus serves to put us on alert to the presence of the ‘negative’ in addition to the ‘positive’ moment in the analogical inference. In fact, we might take this reasoning further to argue that analogical (as opposed to inductive) inferences are defined by the interplay of both positive (apparent similarities) and negative (perceived differences) assertions.^{Footnote 44} As Chiara Bottici also acknowledges, the negative assertions tend to weaken or qualify the positive, such that the analogy is rarely if ever applied in an ‘all’ or ‘nothing’ fashion, that is, to argue for a global leviathan or superstate.^{Footnote 45} There thus arises a kind of normative disjuncture in the chain of reasoning that, in something like the social contract, starts from the idea of moral autonomy and leads towards justifying the centralised authority of the state. Patrick Capps has labelled this normative disjuncture the discontinuity thesis, though the thesis is really just a rather loose assemblage of different reasons why a universal state cannot or should not materialise, ranging from the pragmatic (that such an ordering is simply not feasible in the conditions that pertain in the international system) to the more normative and prescriptive (that such an ordering would be illegitimate bearing in mind the distinctive nature of these conditions).^{Footnote 46} Whilst the discontinuity thesis might at times be deployed as a way of demonstrating the futility of the very idea of international law, such arguments are rather extreme and nowadays quite rare.^{Footnote 47} More likely, the thesis just highlights the need to take seriously the unique characteristics of international relations and how these characteristics might impact our understanding of legal ordering at the international level.^{Footnote 48}

Whilst I do not intend to delve deeply into the question of exactly what are the unique characteristics of the international legal order, it is worth briefly highlighting two recurrent structural claims that are taken to impact international theory generally and which, I believe, have particular relevance to the debate about the international rule of law: the first being that states and individuals are quite distinct kinds of moral agents, and the second – in part following on from the first – is that the environment in which states find themselves is different in certain fundamental ways from the kind of pre-contractual, Hobbesian state of nature, hypothesised between individual human beings.^{Footnote 49} These points of differentiation are at least as old as the social contract tradition itself. Indeed, despite using the domestic analogy to highlight the starkness of the state of nature, Hobbes himself acknowledged that sovereigns simply do not pose the same kind of danger to each other as individual humans do within the state of nature.^{Footnote 50} Meanwhile, Kant used similar arguments to reject the idea of the universal state in favour of a League of Peoples, which he understood as a kind of congress of nations under law.^{Footnote 51} In more recent times, international lawyers have deployed similar reasoning in order to explain and justify the specific character of the international legal order,^{Footnote 52} its particular function,^{Footnote 53} or the ‘horizontal’ allocation of authority therein, wherein states are understood to be both legal subjects and systemic ‘agents’ or ‘officials’ at the same time.^{Footnote 54}

Without necessarily endorsing any of these arguments specifically, I think the observations on which they are based are indeed plausible understandings of certain key differences revealed through such analogical reasoning. My more immediate concern, however, and one upon which I do wish to reflect more fully, is how these points of differentiation should impact on theorising the rule of law at the international level. As I show in the next section, in fact, very few scholars have attempted to argue that the international rule of law is simply the domestic rule of law ‘writ large’.^{Footnote 55} Instead, the unique characteristics of the international legal order tend to be understood as necessitating some kind of adjustment, loosening or stripping back of the concept at the international level.^{Footnote 56} Nevertheless, in highlighting some of the methodological flaws in this form of disjunctive reasoning, I suggest that the question of the rule of law’s application cannot be decided by simply choosing between or otherwise modifying domestic models of the rule of law. Rather, I argue that the rule of law can only take on specific meaning in the international context on the back of an evaluative account of what international law is, or what it is for. Nevertheless, as my argument progresses, I will also suggest that this question is itself complicated by a tension at the very core of international law as a practice, which means that the rule of law is necessarily caught between two plausible, yet seemingly irreconcilable idealisations.

III. Discontinuity and disjuncture in theorising the international rule of law

Two interim conclusions can be drawn from the preceding analysis. The first is that the rule of law is inescapably important to the coherence of modern international law, both as an idea that structures international legal practices and normative expectations and as an ideal in the self-understanding of international law as a kind of collective project. The second is that the concept of the rule of law as it has developed in the domestic context cannot easily be translated to the international legal order due to the structural differences perceived and, arguably, justified in the unique circumstances of international relations. It is for this reason, in fact, that those who have pondered the question explicitly have tended to suggest that the application of the rule of law to the international level necessitates some qualification or contextual adaptation. Certainly, this has tended to involve prioritisation of ‘thinner’ over ‘thicker’ conceptions of the rule of law,^{Footnote 57} of procedural safeguards over substantive protections,^{Footnote 58} but there has also been a tendency to look beyond precise institutional requirements in favour of broadly directive principles,^{Footnote 59} or a more general understanding of the rule of law by reference to its purpose or function.^{Footnote 60}

A common tactic in this regard has been to take a loose tripartite formula – perhaps adapted from Dicey or Hayek, for example – which coalesces around the principled generality, equality and certainty of legal rules, and to test international law accordingly. For instance, Stéphane Beaulac argues that the rule of law can be meaningfully ‘externalised’ beyond the state by applying it simply as three broad principles, which he defines as ‘(1) the existence of principled normative rules, (2) [that are] adequately created and equally applicable to all legal subjects and (3) enforced by accessible courts of general jurisdiction’.^{Footnote 61} Chesterman, similarly, though perhaps more cautiously, posits a looser tripartite formula which requires (1) ‘non-arbitrariness in the exercise of power’, (2) ‘supremacy of the law’, and (3) ‘equality before the law’ – though even these principles have to be understood in a somewhat differentiated, strained form in comparison to their usual application within the state.^{Footnote 62}

Whilst I do not intend to give a full survey of the many formulations of the international rule of law that have been countenanced in recent literature, it seems to be the case that all of these kinds of looser, adapted, or more minimal ‘criterial’ approaches tend to still result in an inevitable frustration due to the decentralised institutional structure of international law. Authors tend to highlight significant institutional failings, chief among them being the consensual, fragmented and somewhat uncoordinated international judicial function.^{Footnote 63} Others have highlighted the way in which international law is created and/or identified as itself particularly problematic from a rule of law perspective.^{Footnote 64} More broadly, a central preoccupation would appear to be international law’s failure to constrain the political choices of states and other actors.^{Footnote 65} If the institutional structure of international law is not seen as inherently problematic, the suggestion would appear to be that the international legal order is relatively underdeveloped and in need of further strengthening through institutional reinforcement.^{Footnote 66} In fact, following this logic one might easily agree with Chesterman either that ‘there is presently no such thing as the international rule of law’ or ‘that international law has yet to achieve a certain normative or institutional threshold to justify use of the term’.^{Footnote 67}

Are these plausible conclusions to reach about the possibility of an international rule of law? Perhaps, but it seems difficult to argue as such without some broader conceptual account of international law’s central function and importance to international legal participants. In other words, drawing on the cautionary remarks of the preceding section, it is not enough to just abstract by analogy the idea of an international rule of law and assert that the apparent ‘faults’ of the international legal order can be fixed by it more closely resembling its domestic counterpart. To simply water down the rule of law as it is understood in a domestic context, to derive and abstract certain transferable principles, appears methodologically problematic.^{Footnote 68} Such reasoning would seem to suggest (though without adequately defending) the conclusion that the international legal system is merely a more loosely organised or less developed version of its domestic counterpart. This view might well turn out to be true – though, I should note, it sits uneasily with the discontinuity thesis, as outlined above – but it surely cannot just be presumed to be the case.

To look at this problem another way, if – as I have suggested – the decentralised character of international law is deemed important to its legitimacy as such, and if that character necessarily implies a certain degree of ‘auto-interpretation’ by states as an important component of this legitimacy, then arguing that international law should be institutionally strengthened vis-à-vis a centralisation of authority appears logically and normatively problematic.^{Footnote 69} This is a point made most emphatically by Ian Hurd, who dismisses these kinds of derivative rule of law formula as ‘conceptually inconsistent and empirically unrealistic’.^{Footnote 70} Specifically, he argues that the kind of prescriptive institutional desiderata outlined on such accounts cannot be seen simply as neutral ‘procedural requirements’ capable of being abstracted and applied outside of the state context, for these desiderata presume ‘a substantive commitment to dividing political power in a certain way’.^{Footnote 71} I myself have made similar claims, noting how the ‘lack of fit’ between this rule of law idealism and the structural conditions of international relations tend to fundamentally undermine the rule of law on these terms.^{Footnote 72} In other words, the kinds of abstracted formula outlined above necessarily presume the state, and would inevitably lead towards justifying something like a universal state (or a rough approximation thereto). It may well be possible, indeed even warranted to argue counterfactually that international law should be reformed towards a more effective institutional architecture, that is, one more able to regulate the conduct of international politics. However, what such an architecture might look like will inevitably depend upon a prior (and sufficiently convincing) account of the nature and normative orientation of international relations as a particular kind of political organisation.

For this reason, Hurd himself argues for a version of the international rule of law more sensitive to what he sees as the deeply instrumental role of law in international relations, whereby rule of law compliance is measured ‘in the degree to which states feel the need to account for their policies in terms of international law’.^{Footnote 73} On this account, there is an ‘inescapable connection between law and power’, where the use of legal argument as a means to justify state conduct suggests both a constraining and empowering aspect to the international legal form.^{Footnote 74} Whether or not one accepts or agrees with Hurd’s own formulation, the underlying point of his analysis is to show how the structural differences between international and domestic political contexts should change rather than simply undermine our expectations of what the rule of law might plausibly mean at the international level.^{Footnote 75}

At the same time, one might question whether Hurd’s analysis too easily drains the concept of the international rule of law of its critical, counterfactual content. Although his thesis is particularly useful in drawing out the peculiarities of international law’s institutional form, it is nonetheless difficult to see how his own view of the international rule of law is any more preferable to the accounts he criticises. The problem would seem to be that the kind of realist-instrumentalist view of international law upon which his argument is premised is simply posited and presumed, rather than cogently defended in ideal terms. The question of who is correct – Hurd, or the kinds of authors he criticises – cannot be resolved simply by assessing which account more adequately describes the conditions of international politics. This conclusion follows simply from acknowledging that the object being described, the rule of law as applied to international relations, should not just aim to mirror the extant practices of contemporary international law. Even if those practices themselves reveal important truths about how the rule of law might manifest at the international level, the theorist’s role is to attempt to construct an ideal account of the role of legal practices in the conduct of international relations.^{Footnote 76} Indeed, and as I claimed at the beginning of this article, given that those engaged in such practices – academic and professional lawyers as much as state representatives – tend themselves to deploy the concept in this counterfactual sense, it surely behoves the legal theorist to attempt to construct a meaningful yet sufficiently robust account of legality to best reflect this critical ambition. In other words, such idealisation and abstraction is, I believe, necessary in order to adequately distinguish a concept of the rule of law from a concept of law more broadly.

More promising, it might seem, is the recent argument of Jeremy Waldron, who is critical of the view that the rule of law concept can just be loosened or adapted in the sense suggested. Specifically, Waldron develops an explicit version of the discontinuity thesis, as outlined above, in order to suggest that the usual form of the domestic analogy necessarily breaks down. As he claims,

At the international level, then, Waldron’s central point – that ‘the state remains a creature of law’^{Footnote 78} – means that the international rule of law should not be understood as something distinct from its domestic counterpart, as some externally imposed constraint. Rather, it has to be understood as reflecting the normative constraints that should limit governmentals in their external as much as internal actions. For this reason, then, Waldron argues that the state should not benefit from the rule of law in the way that an individual does in its relationship with the state. In fact, the apparent freedom that a state possesses, the exact extent of its agency and autonomy, can only be determined by its participation within the international legal system.^{Footnote 79}

Nevertheless, whilst Waldron’s account is, I believe, more nuanced and self-aware of the dangers of reasoning by analogy than many others, his own attempt to reconfigure and argue for a ‘new analogy’ in which the state is seen more as a kind of external, systemic agent of the international legal order, is neither particularly novel, nor informative. This new analogy would seem to be simply a restatement of older ideas already absorbed into international legal theory: for instance, Georges Scelle’s well-rehearsed theory of dédoublement fonctionnel, or ‘role-splitting’.^{Footnote 80} The problem with reverting to this kind of ‘states-as-agents’ argument, as insightful as it might be, is that it itself serves to demonstrate what is commonly perceived to be one of the principle obstacles to the realisation of the rule of law at the international level, that is, the ‘auto-interpretation’ of legal obligations by states and the concordant lack of adequate external institutional constraint. This might matter less if Waldron was more sensitive to the ways in which the structural limitations of the international legal order function to condition and shape the choices of state-agents, but by focussing his critique more towards domestic foreign policy he tends to overlook such inter-national structural constraints.^{Footnote 81} In this way Waldron not only misses the way in which the discontinuities in the ‘old analogy’ play out to create unique mechanisms of accountability – a point, in that sense, better made by Hurd – but, in doing so, he arguably also fails to capture an account of the rule of law that accords with the self-understanding of legal participants themselves.

Although these brief reflections on both views cannot do justice to the importance of their contribution to international rule of law debates, I counterpoise both here more to illustrate how each acts as a critical counterpoint to the other and how, in their confrontation, one gets a clearer sense of the seemingly conflicting urges that materialise in the attempt to translate the rule of law to the international context. Hurd’s account suggests important ways in which the rule of law is already presumed within international legal practices and acts as an important measure of accountability, giving legitimacy to international legal doctrines and institutions. At the same time, Waldron is correct to point to some of the difficulties in theorising international relations in this way and, indeed, of the need to rethink the domestic analogy underlying this theorisation of international legal practice. As such, his own intervention acts as an important counterfactual plea to bring such practices in line with already well-established rule of law values. In the two sections which follow I will aim to flesh out further how we might better understand and theorise these conflicting urges in terms of two distinct idealisations of the international rule of law. However, I will suggest that, rather than choose between them, both idealisations should be harnessed as important, yet competing sides of an overarching account of the rule of law at the international level.

IV. The rule of law as the ‘basis of association’ in international relations?

With this background in mind, we begin to get a clearer picture of what is required, methodologically speaking, to forge a more plausible understanding of the international rule of law. Rather than adapting a list of abstract institutional requirements by analogy to the state context, we must determine what is valuable about international law in the context of a different kind of political relationship; we must relate how the rule of law is understood as meaningful amongst international legal participants themselves, but also abstract from the likely many resulting ideas and idealisations in order to construct a rationally defensible account of the rule of law in this specific context. As Terry Nardin has argued on this point,

As such, I do not think that the theorist can escape the necessary evaluative judgement and normative commitment required to conceptualise the international rule of law on these terms. Arguably, as I have claimed elsewhere,^{Footnote 83} this conclusion follows for any conceptual account of legal phenomena, insofar as such judgement and engagement is necessary to cut through competing and incommensurable viewpoints.^{Footnote 84} However, in the current context I wish simply to stress the particular evaluative burden that arises due to the critical, counterfactual role of the rule of law in engaging with extant legal practices.

On this basis, then, Nardin’s own response to this challenge is to argue that the international rule of law should not be thought of as some set of determinate institutional characteristics by which to evaluate the performance of international law in achieving certain ends. Rather, it can be derived from the form of international law as an ideal representation of a particular kind of political relationship. In other words, he argues that the rule of law functions as the ‘basis of association’ in international relations.^{Footnote 85} It is the moral expression of the kind of ‘deep-structural mutuality’^{Footnote 86} that is encapsulated within many of the core doctrines and institutional structures of international legal practice. However, this is not to suggest some kind of apologia to the status quo, or the raison d’état, and instead expresses the importance of international legal order as a peaceable end in itself. Contra Hurd – or, at least, contra his own description of his account – this view of the rule of law is understood in distinctly non-instrumental terms: a commitment to international ‘legality’ as expressed through the autonomy of the legal form itself.

To argue that the rule of law is a ‘non-instrumental’ ideal is to suggest a particular way of conducting international politics by known, general, and non-arbitrary rules. To be clear, this commitment is not to reduce the international rule of law to some criterial formula, for the ends associated with the concept – generality, equality, non-arbitrariness, and so on – are simply what follows from the relative autonomy of a well-functioning legal system. Likewise, it is not to suggest, as does Hurd (at least to some degree), that the international rule of law is satisfied simply when the international legal form is used by states and other actors. Rather, the rule of law on these terms represents a more substantive commitment to ‘legality’ as a counterfactual abstraction of the international legal relationship: an ethos of playing by the rules which, though manifesting in terms of a kind of empty, procedural formality, is not itself devoid of moral content and will function to place substantive and not merely procedural limitations on the interpretive freedom of legal participants. This point is perhaps best expressed by David Dyzenhaus, who develops a kind of neo-Hobbesian reading of the rule of law in international relations:

Dyzenhaus’s point here has clear echoes of Waldron’s argument, considered above, yet he is arguably far more attentive to how the structural and relational aspects of international legal practice help to maintain in place a procedural guarantee of accountability.^{Footnote 88}

What I have in mind, therefore, would be closer to a kind of Fullerian ‘inner morality’ of law,^{Footnote 89} a view which was more recently given expression in an international context in Jutta Brunnée and Stephen Toope’s ‘interactional’ account of international law, whereby they stress a ‘practice of legality’ which leaves scope for legal participants ‘to pursue their purposes and organize their interactions through law’.^{Footnote 90} More broadly, this argument has clear affinity with EP Thompson’s famous defence of (a somewhat minimal conception of) the rule of law as an ‘unqualified human good’.^{Footnote 91} Despite seeing the potential for legal rules to perpetuate injustices, Thompson also saw how the legal form placed both substantive and procedural limitations on the exercise of political power, whilst creating an equalising framework from which alternative interests could find a platform.^{Footnote 92} At the international level, a similar ethic is expressed by Koskenniemi in terms of, alternatively, a ‘culture of formalism’^{Footnote 93} and as a ‘constitutional mindset’,^{Footnote 94} whereby international law provides the ‘shared surface’ for legal participants to channel their normative disagreements under an ‘assumed universality’.^{Footnote 95}

Nevertheless, the most explicit and well-thought out version of this ideal is given by Nardin, who argues that the international rule of law should be conceived as:

Nardin’s account is particularly useful in the current context as he grounds it historically and conceptually in the emergence of the modern understanding of international law as a positive, systemic practice – as I have also described this transition, above.^{Footnote 97} At the same time, his aim is not simply to describe the impact of liberal political doctrines in structuring international legal thought, but to actively theorise how such an understanding might be plausibly realised in international relations in ideal terms. To do so, he draws extensively on the approach of the English political philosopher, Michael Oakeshott, whilst relating his ideas to the theory of ‘international society’ developed by the English School in international relations, which offers a distinctive defence of the pluralist character of international legal and political institutions.^{Footnote 98} Nardin’s view of the international rule of law thus reflects what he calls the ‘practical conception’ of international society. According to this view, states are understood as related to one another ‘in terms of common practices, customs, and rules’, which in turn allow ‘for making judgments of just and unjust international conduct, for advancing claims ... and for seeking vindication and redress when rules are violated, rights infringed, and duties ignored’.^{Footnote 99} Accordingly, Nardin follows Oakeshott in understanding the rule of law as forming the very basis of this association, in the sense that it represents an ideal expression of a shared moral commitment to a system of common rules, the authority of which stems from this commitment alone and not from any sense of higher purpose or instrumental benefit.

Although Oakeshott himself did not see international relations as capable of being theorised by reference to his account of civil association,^{Footnote 100} and although Nardin is not alone in nonetheless attempting to transpose Oakeshott’s ideas to the international context,^{Footnote 101} his is arguably the clearest and most extensive elaboration of the potential of Oakeshott’s philosophy in better understanding the nature and point of international legal practices. In fact, Nardin argues that Oakeshott’s understanding of political community is more, rather than less relevant in the international context, precisely because of the broadly pluralist character of international society.^{Footnote 102} As such, it is worth briefly recounting the core tenets of Oakeshott’s understanding of civil society and the rule of law underpinning it in order to more fully understand the commitment to autonomous legal order as a ‘non-instrumental’ good.

Oakeshott’s philosophy is not easy to distil in brief, particularly bearing in the mind the somewhat idiosyncratic terminology he employed to make sense of the state and political community more broadly. The core tenets of his theory are found in his most complete work of political philosophy, On Human Conduct,^{Footnote 103} though also elaborated more directly in an extended essay on the rule of law published some years later.^{Footnote 104} Within these works Oakeshott defends an understanding of civil association as distinguishable by its non-voluntary and non-instrumental character: the former quality premised on the acceptance of the authority of the rules contained within the association; the latter explaining the particular kind of ‘freedom’ to which this kind of ordering gives rise. To explain this latter quality more clearly, Oakeshott distinguishes two types of non-voluntary association: civil association and what he somewhat disparagingly refers to as ‘enterprise association’: a type of political relationship in which the constituent members are joined together in view of the fulfilment of some substantive goal(s) or purpose(s). Under ‘enterprise association’, one might say that the members are free in the sense that their participation remains voluntary, and dependent in particular on the enterprise continuing to serve the ends for which it was constituted. He claims that this is the most familiar form of human association and is therefore commonly, but erroneously, assumed to form the basis of civil association.^{Footnote 105} In contrast, Oakeshott’s view of civil association, or civitas, can be understood as an order in which the constituent members (cives) are free to pursue their own ends;^{Footnote 106} it is resistant to the realisation of any particular conception of the common good; and in its prioritisation of political equality it denies that any one member of the political community is free to impose their own conceptions of the good on others.

In what Richard Friedman has referred to as the ‘ambiguity thesis’,^{Footnote 107} Oakeshott was clear that the modern (European) state was in actuality an amalgam of both civil and enterprise association, and in that sense its actual character was capable of idealisation according to either logic. However, in Oakeshott’s view, the rule of law manifests only in the former mode of association, where civil order is maintained through a system of non-instrumental laws.^{Footnote 108} In his own words:

Much more could be said about Oakeshott’s understanding of civil association, as well as its relation to the rule of law specifically, but for current purposes the key point is that his account provides a moral justification for the idea of civil association under an autonomous system of law.^{Footnote 110} Thus, where the rule of law adheres there arises a kind of ‘self-sufficient (although not self-explanatory) system’, in the sense that law is understood (in somewhat Kelsenian terms) as identifying or defining its own jurisdiction.^{Footnote 111} This point requires some clarification, however, as it is clear that Oakeshott did not believe that such law could ‘run of itself’, operating as some kind of deus ex machina. His account necessarily presupposes a system of rules in which ‘umpires’ are deployed to mitigate unending conflicts of interpretation, and there is submission to some arbitrator to mediate between individual agents in their pursuit of more instrumental purposes through the legal form.^{Footnote 112} For Oakeshott, then, adjudication is a necessary and not merely contingent condition of civil association.^{Footnote 113}

For this reason, one might be sceptical that this rule of law ideal might be feasibly realised in the decentralised institutional conditions of international law. However, this is a point considered at some length by Nardin, who argues that there is nothing in Oakeshott’s account that requires centralised institutions as such, but only agreed-upon procedures for authoritative interpretation and dispute resolution in particular circumstances. There is clearly evidence of this kind of systemic agency in international law, of a ‘variety of means for applying international law’ as well as ‘rules delimiting their jurisdiction and defining the scope and significance of their authoritative determinations’, and there is therefore nothing problematic, conceptually speaking, in further institutional reform to realise more effective international procedures in this regard.^{Footnote 114}

The main point of applying his theory to international law is that Oakeshott did not see such institutional roles as necessary for the achievement of some collective purpose; they are there to help realise the conditions of civil association itself. As such, the key innovation in this understanding of the rule of law, compared to arguably more recognisable, but intrinsically more instrumental accounts (eg Raz), is that Oakeshott does not see the rule of law as something that acts on government to limit its arbitrary power in order to realise freedom. Rather, the rule of law acts as an idealised embodiment of a certain type of freedom: it is a prior, assumed commitment which, where it pertains, nonetheless has the effect of guiding and limiting the range of permissible actions of the sovereign.

As such, what we arrive at is precisely a philosophical justification for a view of the rule of law that inheres in the form of legal order itself and which, in the international context, presumes a kind of inter-relational limitation to the freedom of sovereign states, as well as the institutions they create. This theory gives moral grounding to a popular disciplinary narrative surrounding the necessary autonomy of international law, which is seen as important to make international law an objective condition for the conduct of international politics.^{Footnote 115} In other words, Nardin’s extrapolation of Oakeshott’s ideas gives a particular expression to what Christian Reus-Smit has described as a pervasive ‘discourse of institutional autonomy’ in international law, a discourse which has played out in the disciplinary effort to defend the relevance and reality of legal rules in the conduct of international politics.^{Footnote 116}

Nevertheless, recognising that the rule of law is expressed through the ideal of international law’s autonomy only serves to highlight a fundamental ambiguity in what we mean when we speak of the autonomy of law on these terms. Indeed, for as long as this discourse has played out, there has always been a rival articulation of the ambition to subject international politics to objective legal rules.^{Footnote 117} As Oscar Schachter once argued, to say that international law possesses a degree of autonomy cannot just be to present it as a kind of ‘specialized or symbolic language to describe behaviour’. Rather, he claimed, the law must function as ‘a means of independent control that effectively limits the acts of the entities subject to it’.^{Footnote 118}

On these terms, then, the objectivity of law manifests in its ability to constrain the political choices of states in order to realise agreed-upon objectives – ends which are seen as equally intrinsic to the idea of international legal order, whether peace, order, friendly relations, and so on. Accordingly, the rule of law is understood by reference to the institutional conditions necessary for law to manifest this kind of objective normative constraint. Rather than being reflected within international law’s institutional form, then, it is more likely now that the rule of law is seen to be hindered by the apparent limitations of a decentralised legal order, thus necessitating certain institutional reforms.

In this way, the modern discipline came to place its hope for the realisation of an international rule of law in the expectation that the international legal order will develop institutional structures more able to effectively secure the restraint of arbitrary political power. Indeed, Nardin himself conceded as much, arguing that since the creation of the UN the rule of law has come to be thought of – wrongly, in his view – less as a pre-existing basis of association (from which one might engage in such institutionalised cooperation) and more as a goal to be achieved through enhancing the capacity of such institutions to protect community interests.^{Footnote 119} Nevertheless, as I show in the next and final section, this alternative idealisation of an international rule of law is not necessarily misguided or wrong-headed, as Nardin believes, but simply reflects an alternative, competing, but equally important logic of the rule of law as it plays out in the international context. To fully understand the international rule of law one must thus appreciate the antagonistic, yet mutually-constitutive and constructive relationship between both idealisations.

V. The rule of law as institutional ideal: The emergence of global governance

The account of the international rule of law I have just provided reflects the collective experience of the modern discipline in trying to work out a feasible and legitimate system of positive laws as a means of structuring international political relations. I believe that this idealisation of international law as an autonomous, systemic legal order is plausible and persuasive, reflecting familiar and important aspects of international legal practice. I also believe, therefore, that Oakeshott was wrong to deny the possibility of explaining the international legal order on these terms. Nevertheless, the differences that he perceived between the state and international relations do arguably highlight a way in which international law might appear inherently more instrumental, not just in how rules are created and used by states and other participants, but also in the discipline’s desire to subject international law to the governance of objective legal rules more broadly. Let me now explain this point further.

As previously noted, the effort to make sense of international law as an autonomous legal order by reference to an ideal of legality derived from domestic legal experience has to be thought of as a very deliberate disciplinary transition, precisely because such autonomy is seen to be critical in subjecting states to more effective legal regulation. This is precisely why, to begin with, I have claimed that international law has always been as much a project as a practice, as much a promise of a future order to come as a reflection of institutional reality. This condition creates something of a tension as far as the international rule of law is concerned, however, for the pluralist justification for law’s systemic autonomy competes with a more rational, functional justification that sees this autonomy as important to the achievement of law’s regulatory purpose.^{Footnote 120} On one view, then, the differences revealed in our analogical inference – the specific characteristics of international law as a decentralised legal order – are seen as important aspects of the legitimacy of international law as a regulatory system between diverse political communities; on another view, however, they look much more like a kind of ‘constitutional deficiency’ that undermines the determinacy of international law and thus prevents the full realisation of the rule of law.^{Footnote 121}

One can see this latter view laid out most explicitly in the early writings of Hersch Lauterpacht – certainly one of the twentieth century’s most celebrated international lawyers, and one of the most vocal champions for the advancement of the international rule of law. As he states:

Lauterpacht’s argument is not to deny the legal character of international law outright, but it is to make the question of international legality somewhat relative to the institutional maturity of the international legal order. Thus, whilst international law’s apparent imperfections might not call for the installation of a universal state, for Lauterpacht, they did necessitate that international lawyers work towards securing a more effective global architecture, that is, one more able to effectively realise collective ambitions against recalcitrant states and other powerful actors.^{Footnote 123}

This more instrumental understanding of an international rule of law has been a recurrent rhetoric in the modern discipline, but particularly since international law’s ‘move to institutions’ at the start of the twentieth century.^{Footnote 124} Each new post-war project of international organisation, from the League of Nations to the creation of the UN, has been interpreted, at least in part, as an attempt to counteract or fix the apparent constitutional weaknesses of a decentralised legal order. This is evident in particular in Lauterpacht’s reading of the League Covenant, which he saw as an effort to centralise the international system through a kind of proto-constitution.^{Footnote 125} He was sensitive to the fact that the actual legal form of the Covenant as an interstate treaty reduced the cogency of the constitutional analogy,^{Footnote 126} but the key point was precisely to look beyond the legal form, to actively interpret the institutional structures which it created in a more functional way, and thus necessarily changing the nature of international law in practice.^{Footnote 127} Lauterpacht was by no means alone in advocating this kind of internationalist and functionalist reading of the League and its impact on the constitutional structure of the international legal order. Arnold McNair, the author of the leading text on the law of treaties at the time, argued that the constitutive instruments of international organisations could be thought of as a kind of decentralised form of ‘legislation’, allowing the international community to create institutional structures more able to effectively protect community interests.^{Footnote 128} Similarly, James Brierly, a contemporary of Lauterpacht, was quite explicit on the necessity of such a functional reading in order to fulfil collective ambitions for an international rule of law:

This sort of functional reading of the project of international organisation continued well into the latter years of the twentieth century, fuelled by the creation of the UN and the proliferation of institutions, regimes, courts and dispute settlement mechanisms that followed in its wake. Whilst few would go on to see the UN Charter as an actual constitution for the international system,^{Footnote 130} it seems to be more broadly accepted that the UN system and associated institutions are essentially ‘gap-filling’ for the lack of actual constitutional organs at the international level.^{Footnote 131} At times this kind of functional reading has been explicitly advocated in distinctly anti-formal terms.^{Footnote 132} More often, however, it has manifested simply in the presumption that the advancement of the functional autonomy of international institutions will help constrain the arbitrary choices of states in order to advance international law’s regulatory ambitions.^{Footnote 133}

Nevertheless, whilst it is difficult today to deny the impact of international institutions in the actual functioning of the international legal system – whether in facilitating the advancement of ‘community interests’ through communal institutional fora,^{Footnote 134} or in enhancing international law’s law-making procedures,^{Footnote 135} or the available means for adjudication and enforcement in practice^{Footnote 136} – it is clear that this impact is seen in increasingly ambivalent terms from the perspective of the international rule of law. Although many international institutions have come to exercise a kind of functional authority and influence that belies their formal grounding in treaty law, such innovations cannot but strain the kind of formalist, non-instrumental logic inherent in the rule of law ideal as defended above.^{Footnote 137} Thus, on the one hand, the increasing activism of the UN Security Council in the post-Cold War period was initially greeted with disciplinary enthusiasm, but this soon gave way to concern over the Council’s apparent institutional overreach, selectivity, and lack of legal accountability.^{Footnote 138} Likewise, although the ICC was initially heralded as helping to end impunity and ensure the more effective protection of global human rights, the institutional design and functioning of the Court in practice increasingly raises concerns over its politicisation and apparent Western bias.^{Footnote 139}

In light of such concerns, it is hardly surprising that the new millennium brought with it an increasing academic focus on the question of the legitimacy of international institutions;^{Footnote 140} policy initiatives aimed at addressing questions of accountability and institutional responsibility;^{Footnote 141} as well as the emergence of new movements self-defined in terms of ‘global constitutionalism’^{Footnote 142} and ‘global administrative law’,^{Footnote 143} the ethos of which is essentially to address the kind of rule of law concerns that arise from the apparent overreach of ‘global governance’ institutions.^{Footnote 144} Certainly, much of the concern surrounding the normative impact of such institutions and regimes stems from a perceived weakening of rule of law protections at the domestic level, as much as the international. However, from the international perspective, two particular concerns have arisen and suggest that the account of the international rule of law set out above remains a still pertinent facet of the overall legitimacy of international law. First, insofar as these different regimes and institutions are seen to develop a range of normative instruments, standards and forms of policy influence which fall outside of the recognised sources of international law, there appears to be a particular concern related to the law’s deformalisation.^{Footnote 145} Second, where such regimes and institutions do apply, interpret or enforce recognised rules of international law, the worry is rather that the structural bias of each may well contribute to the apparent fragmentation of international law, thereby undermining its overall systemic coherence.^{Footnote 146}

Whilst space precludes any deeper inquiry into the global governance phenomenon and the apparently ‘postmodern anxieties’ that it raises,^{Footnote 147} what I do want to suggest in the remaining paragraphs is that the growing embrace of, yet seeming ambivalence towards international law’s institutionalisation demonstrates the interplay between the two opposed idealisations of the international rule of law: on the one side, calling for the development of autonomous institutional structures to overcome the perceived inefficiencies and externalities of a decentralised legal order; on the other side, however, the more autonomous and effective such regimes and institutions become, the more they appear threatening to the rule of law understood by reference to the law’s formal systematicity. The competing formal-pluralist and functional-rationalist logics play off against each other, with the instrumental urge towards institutional effectiveness causing a strain from the perspective of the formal-structural and non-instrumental ‘lex’ of the system, whilst these structural rules at the same time seemingly undermine or limit the attainment of communitarian goals. This tension carries over into the institutional framework of intergovernmental organisations, whereby their functionality may well contribute to enhance or streamline international legal processes (for example, the formation of customary international law, the enforcement of legal frameworks, and so on), but insofar as this kind of functional authority lacks any formal legal grounding other than the constitutive rules of treaty law, such normative influence will always be somewhat contingent, revisable and challengeable from within (by member states) as well as from outside (by interested global publics) the constraining legal framework of the institution.

One might therefore see the structural tenets of a decentralised legal order as giving rise to uncertainty and inefficiency, requiring necessary institutional innovations to secure more effective legal mechanisms, but such institutional innovation will always be tempered by a counter-logic expressed through the formal systematicity implied by the rule of law as the basis of international association. It is this kind of rule of law logic that insists on accountability through formal legal warrant, that ends do not override means, and which creates the formal-legal equality necessary to counteract the pervasive structural inequalities of the global system.

One might equally see global governance as straining the system of international law, or as giving rise to certain legitimacy deficits, precisely because of concern to protect the value of this formal systematicity. Certainly, the concerns surrounding international law’s fragmentation are an acute expression of this urge to protect the systemic autonomy of international law. However, this kind of functional authority is no less important to what contemporary international law is, providing a dynamism to its institutional framework to respond more effectively to arising global challenges, as well as a counterweight to the kind of sovereigntist rhetoric that arguably threatens to abuse legal formalities for the sake of protecting the entrenched interests, biases and policy preferences of powerful actors (whether states or global institutions).^{Footnote 148}