“Fantasy Upon Fantasy”: Some Reflections on Dworkin’s Philosophy of International Law

This article offers a critique of Ronald Dworkin’s article “A New Philosophy for International Law”, (Philos Public Aff 41: 1–30, 2013). It begins by showing that Dworkin’s moralised theory of law is built on two highly questionable background assumptions. On the one hand, a descriptively implausible characterisation of a positivist-voluntarist view of international law as the reigning “orthodoxy”. On the other hand, the methodologically questionable assumption that a theory of international law must discharge the dual function of explaining the validity of international law in a manner that underwrites its presumptive legitimacy. In its core part, the article then offers a sustained criticism of Dworkin’s moralised account of the validity and legitimacy of international law. Various problems are identified with the “principle of salience” that Dworkin offers in place of consent as a ground for international law. A key concern is the difficulties that stem from Dworkin’s willingness to proceed on the “fantasy” assumption that his theory needs to get off the ground, i.e. that there is an international court with compulsory jurisdiction and reliable mechanisms for enforcing its judgements. Finally, the article concludes with some thoughts on how Dworkin’s “fantasy-based” approach led him to over-estimate the degree to which international law can be a vehicle for the global spread of liberal democratic values. More minimalist ambitions for international legal order, along the lines suggested by John Rawls in The Law of Peoples, seem more realistic.

1 3 "A New Philosophy for International Law" was published posthumously 8 years ago in Philosophy and Public Affairs. 1 It offers the most sustained exposition of Ronald Dworkin's thought on international law, drawing on the theory of law as integrity presented in Law's Empire (1986), but also on later work on human rights and political legitimacy in Justice for Hedgehogs (2011). His article develops three phases of argument. In the first phase, Dworkin sketches the theoretical understanding of the validity and legitimacy of international law that he believes is dominant among practitioners and scholars of international law. On this view, which I call the positivist-voluntarist account (PVA), the validity and legitimacy of international law both stem from the consent of sovereign states within a broadly Westphalian international legal order. Dworkin then explains why he believes the PVA is radically flawed and needs to be jettisoned. In the second phase, he offers his own theory of international law's validity and legitimacy, one that explicitly grounds international law in moral principles that justify subjecting states to international legal norms to which they have not themselves consented. A remarkable feature of this moralised view is that it proceeds on the "fantasy upon fantasy" (14), as Dworkin himself describes it, that an international court exists with compulsory jurisdiction, one whose rulings can be effectively enforced throughout the globe. Finally, against the background of his novel theory, Dworkin proposes liberalising the legal regime governing military intervention and instituting a majoritarian procedure for the enactment of international law. I shall discuss each of these phases of argument in turn, although giving greater attention to the first two phases.

I
Dworkin's article begins by sketching the doctrinal theory of international law he supposes is "generally accepted" by practitioners and scholars in the field (5). It is important to recognise that the kind of theory he is concerned with is supposed to perform a double duty. It must (a) explain the conditions of the validity of international legal norms, such as treaty or customary norms, and it must do so in a way that (b) secures their presumptive legitimacy, i.e. their general tendency to impose defeasible moral obligations of obedience on states in virtue of their legal validity. The theory that Dworkin identifies as the orthodox view is one I shall call the positivist-voluntarist account (PVA) which holds that "a sovereign state is subject to international law but… only so far as it has consented to be bound by that law" (5). The PVA is a voluntarist theory because it locates the ultimate source of the law's legitimacy in the freely given consent of its subjects. In so doing, it purports to justify how states, for all the sovereignty accorded them within a Westphalian conception of international society, may nonetheless be morally bound by international law. And it is a positivist theory insofar as the determinants of legal validity are "contingent historical facts" that can be ascertained without any need for moral evaluation. That a state has consented to a given norm is a contingent historical fact, one that can be identified and described as such without making any moral judgement, even if a state will typically present itself as having a moral justification for giving its consent to the norm in question.
Dworkin identifies three familiar kinds of problems with the PVA. One set of problems concerns its indeterminacy. It fails to give adequate guidance on crucial matters, including how to prioritise the potentially conflicting claims of different legal norms, or how to determine which states are sufficiently "civilised" (Dworkin's stipulation) to participate in norm creation. Moreover, in the case of customary international law, there is indeterminacy as to the level of state acceptance that suffices to generate a norm (6-7). Another set of problems concerns the PVA's seeming lack of fit with established international legal doctrines. Prominent here are jus cogens norms, such as those prohibiting slavery, genocide, and torture, which purport to bind states not only in the absence of their consent but even despite their having persistently objected to those norms during their formation. The attempt to subsume such non-voluntarist doctrinal developments within the PVA threatens the "axiomatic place of consent in the [PVA] scheme" (7).
But, for Dworkin, much more consequential is a third set of problems which are normative in character. The fundamental objection here is that agreement to treat something as morally binding law is neither a necessary nor a sufficient basis for its legitimacy (10). Even if states consent to treaty or customary norms, the question still arises as to what reasons they had to do so. It is these underlying reasons that will be crucial in determining whether or not there is a genuine obligation of obedience, including an obligation conditional on state consent. Just as philosophers seek to explain why promises create moral obligations by invoking deeper moral principles, such as those of reliance, so too philosophers of international law must invoke underlying principles to justify the moral force of treaty and customary norms. This conclusion sets the stage for the development of Dworkin's own "moralised" account of international law, since its aim is precisely to ground international law in deeper moral principles. Now, the rejection of the PVA, especially its voluntarist component, is in reality fairly commonplace among philosophers who have engaged with international law in recent years. 2 In truth, Dworkin's critique adds nothing significantly new to a familiar set of objections. Indeed, if anything, Dworkin overplays his hand. For when it comes to explaining the moral force of treaty obligations, voluntarism has a lot going for it, since consent to a treaty norm is a necessary condition of its bindingness qua treaty norm (as opposed to a treaty norm that has transformed into a general customary norm). Indeed, it is doubtful that the bindingness of treaties can be explained in the same terms as the bindingness of customary international law. The former needs an explanation akin to that for promissory obligations; the latter, by contrast, more readily fits the shape of an explanation of legitimate authority, whereby an institution's "right to rule" imposes moral obligations on others to obey its directives. At most, Dworkin is entitled to the conclusion that the PVA, as elaborated by him, is an incomplete account of international law, and this in two ways. First, it does not obviously explain the bindingness of non-treaty-based international law and most prominently norms of jus cogens. Second, even with respect to treaty obligations, it does not offer an underlying ground (e.g. reliance) for treating state consent as their source.
Two other broad deficiencies in Dworkin's critique are worth elaborating at greater length. The first consists of descriptive deficiencies in his account of international legal doctrine and the views prevalent among international law scholars and practitioners; the other relates to difficulties with the methodological setting in which he situates the encounter between the PVA and his own theory.
Descriptive Deficiencies Dworkin's contention that the PVA is the dominant account of international law among scholars and practitioners is, at best, highly misleading. Admittedly, when pressed about their theoretical commitments, many international lawyers will occasionally emit distinctly positivist and voluntarist noises. And, as just observed, voluntarism surely has something to teach us about the moral force of treaty norms, even if it is not the whole truth, and even if Dworkin is right that we are still owed a philosophical explanation of why consent binds. But even accepting this point, the reality of international law since the period of decolonisation era belies any consent-based rhetorical façade that international lawyers might be disposed to erect when subjected to philosophical interrogation. This becomes apparent once we look to other sources of law, such as customary international law. Compelling evidence to this effect comes from the most articulate defender of the PVA in recent decades, the great French international lawyer Prosper Weil.
Weil's classic article "Towards Relative Normativity in International Law?" remains one of the most highly cited writings in the discipline, and deservedly so, given its clarity, depth, and argumentative rigour. 3 Weil's overall thesis is twopronged: first, that conformity with positivism and voluntarism is essential to international law's ability to fulfil its primary objectives of securing co-existence and co-operation among states, and, second, that both of these features of international law have been dangerously compromised in the post-decolonisation period by a trend towards "relative normativity". In consequence, Weil's entire analysis is suffused with a sense of dark foreboding about pathological developments which, albeit motivated by admirable moral aspirations, seriously threaten to render the international legal order dysfunctional. 4 Relative normativity, according to Weil, is a complex amalgam of two closely related processes. First, a dilution in the scope of international legal norms, i.e. in the range of states that are bound by these norms. As a result of dilution, norms are increasingly taken to bind states they did not consent, or to which, in the case of jus cogens norms, they persistently objected during the process of their formation through the accumulation of state practice and opinio juris. The second process is the generation of normative hierarchies in which some norms enjoy a higher status than others -for example, jus cogens norms versus ordinary customary norms, or international crimes versus delicts -with correspondingly augmented forms of normative significance. The two trends are united by the role that they accord to value judgements in determining the existence, scope, and rank of international legal norms. The value-laden process involved in identifying these norms contravenes the requirement of positivism, which excludes valuejudgements as a basis of legal validity. Moreover, relative normativity also contravenes voluntarism, because it asserts that certain categories of norms, e.g. jus cogens norms, can bind states regardless of their consent or even despite their persistent opposition.
Note that Weil's article appeared in 1983, 30 years prior to Dworkin's incursion into the field of international law, and its defence of the PVA is cast as a lament for a world already slipping away. Moreover, the developments that Weil condemned only gained in momentum and strength during the intervening three decades prior to the publication of Dworkin's article. This is exemplified perhaps most powerfully by the value-dependent conception of custom adumbrated by the International Court of Justice in the Nicaragua case and by the enshrinement of jus cogens as part of doctrinal orthodoxy. Even if it is going too far to assert that "relative normativity" is now the reigning theoretical selfunderstanding among international lawyers, it has secured a sufficiently strong foothold to undermine any pretensions of the PVA to orthodoxy. Hence, Dworkin is largely wrestling with a straw man when he casts the PVA as the "generally accepted" theory of international law. Given the essentially dialectical nature of philosophy's development, there is the worry that Dworkin is not pitting his own view against the most widely accepted or most plausible alternative theories on the current scene.
Admittedly, Dworkin himself stresses that the PVA has great difficulty accommodating key doctrines of international law, such as jus cogens, treating these doctrines as aberrations within the PVA-based self-understanding of contemporary international law. But this not only underestimates a prevalence of such doctrines, it also fails to notice that they are not standardly presented by their adherents as compliant with voluntarism. For example, customary norms apply to a state even if it has not consented to them, because the state came into existence only after the norm was crystallised or else because the state failed persistently to object to the norm during the process of the latter's formation. Again, jus cogens norms apply to states irrespective of whether they have consented to them, or even despite their persistent objection to them, as in the case of apartheid South Africa's rejection of the jus cogens norm prohibiting systematic racial discrimination.
In both cases, jus cogens and customary norms emerge as a result of sufficient consensus (as exemplified by state practice and opinio juris), and this normative consensus can bind states that are not themselves a party to it. 5 It may be that a state can avoid subjection to a customary norm by means of persistently objecting to it during the process of its formation, thereby exercising a consent-based opt-out. But the standing of the 'persistent objector rule' is itself disputed, and in any case, no amount of persistent objection will avail a state if the norm in question has achieved the status of jus cogens. Had the significance of consensus in contemporary interpretations of custom and jus cogens been brought more clearly into view, important similarities between contemporary international law and Dworkin's own principle of salience (discussed in Section II, below) would have come to light, calling into question the extent to which Dworkin's theory is a radically new proposal in departing from voluntarism and positivism.
In short, in order to engage the leading theoretical ideas embodied in international law and scholarship Dworkin needed to go beyond the PVA and the associated Westphalian conception of "pure unrestricted sovereignty" (17). The failure to do so no doubt confers an unfair dialectical advantage on Dworkin's own theory within the confines of his article and exaggerates the extent of its originality in departing from positivism and voluntarism. Moreover, Dworkin is prevented from considering how positivism and voluntarism can be departed from in ways that stop short of embracing the radically moralised conception of international law that he defends (see II, below).
Methodological Deficiencies I turn now to the methodological deficiencies in Dworkin's approach. The starting point here is Dworkin's methodological requirement that a theory of international law has to discharge a double duty: it must (a) set out the grounds of the validity of international legal norms, and do this in a way that (b) confers legitimacy upon valid international legal norms, in the sense that states are presumptively morally bound to obey them in virtue of their status as valid law. This double duty requirement explains the PVA's yoking together of two logically distinct positions, positivism and voluntarism, as a package deal. After all, it is logically coherent to be a legal positivist whilst rejecting voluntarism as an account of the legitimacy of law. 6 Equally, one can be a voluntarist about law's legitimacy whilst nonetheless subscribing to the view that the identification of legal norms is necessarily a value-laden process. Perhaps some international legal scholars and practitioners implicitly share this double duty methodological demand. Indeed, this is one possible explanation of why the consensus required in customary norm-formation is often conflated with consent: the idea that operates as a ground of validity silently metamorphoses into one thought capable of grounding legitimacy. But, even if some international lawyers and scholars accept this requirement, many others do not. All this exacerbates the "straw man" quality of Dworkin's characterisation of the PVA.
For Dworkin, the double duty constraint is an implication of the contrast between a sociological theory of law and a doctrinal theory of law. The first kind of theory offers a scheme for classifying a set of social norms as legal norms -as opposed to norms of etiquette, morality, religion, etc. -whilst scrupulously abstaining from any moral judgements either in identifying or ascribing legitimacy to them. But, according to Dworkin, sociological theories are of scant interest to scholars and practitioners, who seek above all a doctrinal theory of law, one that enables them to identify what the law of a given jurisdiction requires. And the central thesis of Law's Empire is that any such theory must be conceived as an exercise in constructive interpretation, one that both fits a given object of interpretation and presents it in its most attractive light as an instance of the genre to which it belongs. At the most abstract level, doctrinal theories of law offer different conceptions of a general concept of law that forms the plateau on which they fight out their interpretative battles. That general concept is identified by Dworkin in the following terms: Law insists that force not be used or withheld, no matter how useful that would be to ends in view, no matter how beneficial or noble these ends, except as licensed or required by individual rights and responsibilities flowing from past political decisions about when collective force is justified. 7 This concept of law, as specified by Dworkin, gives prominence to two ideas: that law is a matter of standards back up by coercion and that it is presumptively legitimate. Law as a form of social ordering involves the presumptively justified upholding of rights and responsibilities embodied in past political decisions by means of collective force. Different theories will provide rival interpretations of this concept which are to be assessed in terms of requirements of fit and justification with the legal practice of a given jurisdiction. And, of course, according to Dworkin, the best constructive interpretation of law so understood is his theory of law as integrity.
Predictably, Dworkin's opponents protest at being incarcerated within this methodological set-up. Legal positivists, in particular, have insisted on the distinctness of the questions of law's grounds of validity and law's legitimacy, and have denied that law is inherently either coercive or a presumptively legitimate institution. 8 Faced with Dworkin's choice between a merely sociological or a fully interpretive theory of law, they elaborate methodological options captured by neither of these 7 Dworkin 1986, 93. 8 Another contestable assumption in Dworkin's set-up is the invocation of a concept of law as inherently coercive. Strangely enough, although this assumption would be accepted by classical legal positivists like Thomas Hobbes, Jeremy Bentham, and John Austin, it is rejected by mainstream contemporary legal philosophy, both positivist and anti-positivist, e.g. H.L.A. Hart, John Finnis, and Joseph Raz. alternatives, options in which legal validity can be perspicuously addressed independently of legitimacy. The cultural anthropologist Clifford Geertz once spoke of a method of interpreting a people's way of life "which is neither imprisoned within their mental horizons, an ethnography of witchcraft as written by a witch, nor systematically deaf to the distinctive tonalities of their existence, an ethnography of witchcraft as written by a geometer". 9 Many contemporary philosophers of law, especially those of a positivist bent such as H.L.A. Hart and Joseph Raz, regard Dworkin's interpretative theory as sailing alarmingly close to the Scylla of ethnography-by-witch. They seek, instead, a hermeneutically sensitive theory of law that apprehends the institution of law from a suitably internal point of view without being morally committed to it, including any claims of presumptive moral legitimacy.
There is no space to go deeper into these methodological issues here. It is enough to observe that the double duty thesis is an eminently contestable one, rather than an "undeniable" fact as Dworkin describes it (11), and that its organising role in his article lends further weight to the nagging worry that he unfairly rigs the rules of the game to the disadvantage of rival theories. The two main skewing effects that double duty constraint has on Dworkin's discussion are the following: (a) it leads to the presentation of positivism and voluntarism as a package deal, which obstructs an examination of their independent merits as accounts, respectively, of the nature and legitimacy of law, and (b) in relation to voluntarism, in particular, it forces that doctrine into the mould of an ambitious, all-purpose justification of the legitimacy of international law, rather than a more local justification that is predominantly confined to treaty law.

II
Leaving aside his disappointing encounter with the PVA, let us turn now to Dworkin's own moralised conception of international law. In line with the methodological strictures we have just discussed, Dworkin insists that any theory of law offers a legal answer to the following question of political morality: We ask: which political rights and obligations of people and officials are properly enforceable on demand through institutions like courts that have power to direct coercive force? (12) Even if we accept the controversial linkage of law in general with coercive measures, attempting to answer this question in relation to international law poses a glaring problem. There are no international courts with compulsory jurisdiction over all states and generally effective mechanisms to enforce their judgements, nor is there any realistic prospect of such courts being established in any reasonably foreseeable future. Yet precisely this institutional set-up is what makes the crucial distinction, for Dworkin, between a normative theory of what the law is and a normative theory of what it ought to be. One familiar response at this point would have been for Dworkin to join the legion of legal theorists, such as Thomas Hobbes, John Austin, and H.L.A. Hart, who question the credentials of international law to be designated law in a true or full sense. In order to avoid this sceptical conclusion, however, Dworkin takes a counterfactual turn, one that in his own words piles "fantasy upon fantasy" (14). To answer the interpretative question in relation to international law, Dworkin says, we must imagine the existence of an international court enjoying compulsory jurisdiction over all states. We must further imagine that cases can be fairly easily brought before such a court and that effective sanctions exist to enforce its rulings (14). 10 The imagination having furnished the interpretative stage in this way, we can then address the following "tractable question of political morality": What tests or arguments should that hypothetical court adopt to determine the rights and obligations of states (and other international actors and organisations) that it would be appropriate for it to enforce coercively? (14) To answer this question, Dworkin begins by invoking a general obligation incumbent upon all states to undertake steps that enhance their own political legitimacy. Since international law is part of the coercive system that states impose upon their own citizens, this duty includes an obligation to enhance the legitimacy of the international legal order itself: If a state can help to facilitate an international order in a way that would improve the legitimacy of its own coercive government, then it has a political obligation to do what it can in that direction. (17) In order to meet this general obligation, and enhance their own legitimacy, states must work to ameliorate the failures and risks of the "pure unrestricted sovereignty that the Westphalian system gives them". They can do so by taking steps to promote an international legal order that mitigates the "flaws and dangers of the Westphalian system" by means of: (a) upholding human rights at home and abroad, thus helping prevent each state's own possible future lapse into tyranny; (b) guaranteeing the assistance of the international community against invasion and other external threats to a state's independence; (c) encouraging international cooperation, especially by using the threat of coercion to solve collective action problems in order to prevent forms of economic, medical, or environmental disaster; and (d) enabling people to have some "genuine, even if minimal and indirect, role in their own government", including forms of political participation beyond state boundaries, so as to enable those most affected by prospective measures to have a say (18). 11 10 In construing Dworkin's resort to fantasy, in this way, as originating from the deeper commitments of his wider theory of law, I disagree with David Dyzenhaus' suggestion that his hypothesis of an international court betrays the influence of a Hartian approach to law focussed on municipal legal systems, see Dyzenhaus 2020. 11 It is not obvious how these four ways of improving a state's legitimacy through engagement with the international legal order square with the account of legitimacy presented in Dworkin 2011. The problem is that the latter account is formulated exclusively in terms of compliance with human rights, whereas human rights figure as only one of four possible ways of improving legitimacy on this new account. The answer may be that compliance with human rights is necessary, and perhaps sufficient, for legitimacy, but that the other three factors, insofar as they cannot be simply comprehended under human rights, help strengthen the obligation to obey the law of a legitimate state.
Although the duty to mitigate the imperfections of the sovereign-state system is the most general structural principle underlying international law, Dworkin admits that it suffers from considerable indeterminacy. There are numerous rival ways of improving the existing international legal order, and there is reasonable disagreement among states as to which of these ways is best. Here, Dworkin invokes another basic structural principle that of salience: If a significant number of states, encompassing a significant population, has developed an agreed code of practice, either by treaty or by other form of coordination, then other states have at least a prima facie duty to subscribe to that practice as well, with the important proviso that this duty holds only if a more general practice to that effect, expanded in that way, would improve the legitimacy of the subscribing state and the international order as a whole. (19) The result is that a state will be "prima facie" 12 bound to comply with an emerging consensus, of the kind described, if doing so will enhance its legitimacy, and it will enhance its legitimacy if compliance would help ameliorate some of the four problems of the state system identified above. Applying this principle, Dworkin believes that the basic elements of the UN system, including the UN Charter, the Geneva Conventions, the conventions on genocide, and the Rome Statute of the International Criminal Court, have acquired salience of this kind, such that all states now have an obligation to obey them as law, irrespective of whether they have individually consented to them. The obligation arises simply because these multilateral treaties are pathways to a more legitimate international legal order (20). Moreover, he claims that the principle of salience offers a better account of the traditional sources of international law, as set out in Article 38 of the Statute of the International Court of Justice, than the PVA. Custom, treaties, general principles, and jus cogens norms are sources of legitimate law not because states consent to them but because "it contributes to international order to continue to treat those provisions [of Art 38] as sources of international law". Finally, the principle of salience is said to yield an acceptable interpretative strategy for international law, unlike the PVA; i.e. the documents and practices referred to by the principle should be interpreted "so as to advance the imputed purpose of mitigating the flaws and dangers of the Westphalian system" (22).
Although the principle of salience no doubt offers some gain in determinacy, many difficult matters remain unresolved. First, the principle does not tell us how much of an improvement needs to be achievable in order to generate an obligation of compliance, especially when set against any costs involved in securing those improvements. The latter include costs to the state's own self-determination and the opportunity costs of making improvements on other fronts. Second, the principle, as stated, seems to countenance a state being presumptively bound to comply with an emerging consensus even though the latter secures only a modest improvement in legitimacy compared to other arrangements that would achieve significantly greater improvements at no extra cost or perhaps at a lesser cost. It is implausible that an obligation of obedience arises in such circumstances, especially if morally suspect motives explain the choice of the radically sub-optimal arrangement. Third, as formulated by Dworkin, the principle of salience seems insensitive to the fact that states may bear differential moral responsibilities with respect to securing improvements to the international order. These differences will be rooted in such factors as their disparate levels of wealth and development or morally-salient differences in their histories, e.g. whether they have perpetrated environmental depredations, gross human rights violations, or illegal invasions. Fourth, the principle does not explicitly confront the problem of trade-offs between a state's legitimacy and the legitimacy of the international order. For example, some scholars have suggested that when a liberal democracy joins a multilateral human rights treaty, this may help advance overall compliance with human rights across the globe, but only at the price of human rights "backsliding" at home. 13 How are the rights of fellow citizens to be weighed against the rights of foreigners in a scenario of this kind? Should we take literally the terms of Dworkin's proviso -"improve the legitimacy of the subscribing state and the international order as a whole" (19, my italics) -which appears to disallow domestic sacrifices of human rights for the greater global good? Or should we be prepared to countenance a loss of domestic legitimacy in order to promote greater legitimacy among states transitioning towards liberal democracy?
Finally, and implicit in the previous observation, there is a serious empirical issue regarding the extent to which the multilateral treaties listed by Dworkin are indeed "route[s] to a satisfactory international order" (20). We cannot just "read off", as it were, whether they are such pathways simply from the content of their constituent norms. We have to ask the empirical question, whether in fact these instruments serve to enhance the legitimacy of the international legal order in their practical effects. Here, there are serious reasons for doubt that Dworkin does not even begin to acknowledge. I have already referred to controversies about the effects of human rights treaties on actual human rights compliance. But consider an even more troubling case -scepticism about the legitimacy of the International Criminal Court. Is the prosecution of political leaders under the Rome Statute something that contributes to the reduction of human rights violations and inter-state violence? What about the fact that the International Criminal Court exhibits a profound regional bias in its prosecutions, such that leaders of powerful North Atlantic democracies are de facto immune from jurisdiction, whilst its docket is full of vanquished sub-Saharan African leaders? 14 Now, to this last concern, Dworkin might respond by invoking his hypothetical baseline. The question, on this view, is not whether the relevant treaties enhance legitimacy in the world as it is, but whether they would do so in the idealised 13 E.g. Goodman and Jinks (2013) and Linos and Guzman (2014). 14 For an elaboration of such scepticism about the International Criminal Court, see Osiel (2014). circumstances of an international court with compulsory jurisdiction backed up with effective enforcement mechanisms. That question seems, all too easily, to answer itself. But what bearing does any such hypothetical situation have on a state's obligation, here and now, to comply with institutions that are not currently effective in anything remotely like the manner they would be in Dworkin's fantasy world? How can there be even a defeasible duty to comply with international legal norms that are effective in a fantasy world but ineffective, or perhaps even dangerously counterproductive, in the world as it is? Is it enough to say the obligation of obedience genuinely exists, but that it is vulnerable to being defeated by unfavourable real-life circumstances? And even if we accept this get-out clause, how will we ever arrive at obligations tailored to actual conditions if we start from Dworkin's utopian scenario? These concerns raise the wider question of the strong, and somewhat troubling, role of idealisation in Dworkin's theory of international law.
Of course, Dworkin's theory of law as integrity famously introduces a strong element of idealisation in its specification of the interpretative process for identifying law. This is personified by the imaginary Judge Hercules. Dworkin's judicial demigod differs from any real-life judge in having complete knowledge of the political community's legal materials, institutions, and history; an extraordinarily high level of competence in moral and political philosophy; and the limitless leisure and other resources needed to arrive at the unique set of principles that best fits and justifies past political decisions. Many object that the Herculean standpoint is so remote from that inhabited by ordinary judges that it is disqualified from serving as a meaningful account of the grounds of law in real-world adjudication. The discrepancy between the ultimate standard of legality and any decision procedure that might be practically adopted by real-life judges, they suggest, is simply too great. Whatever the merits of this line of criticism, when it comes to international law, Dworkin's idealisation takes a much deeper and even more problematic turn. It does not simply pose an idealised standpoint for determining the grounds of law; it also idealises the object of interpretation, by proceeding on the basis of "ideal conditions" -a court with compulsory jurisdiction and effective sanctions -that are wildly divergent from the current and foreseeably attainable reality of geopolitics. Given the double duty performed by the theory, there is a serious question not only about whether we can meaningfully speak of existing law by reference to this idealised test but also of a genuine obligation to obey this law, given the extreme degree of idealisation involved.
One further complication worth noting here relates to how Dworkin's idealised conception of international law fits with the principle of salience. The latter principle imposes a presumptive obligation to comply with any "agreed code of practice, either by treaty or by other form of coordination" that has been adopted by a significant number of states whose members together comprise a significant portion of the earth's population. The problem here is a mismatch between the consensus and the idealisation involved in Dworkin's theory. When states "agree" to international normative arrangements of any kind, they generally knowingly do so in the absence of an international court with compulsory jurisdiction to uphold those arrangements through effective sanctions. Indeed, it could be that the states that agree to arrangements in the absence of these features would refuse to agree to them if these features were present. To take just one example, it is almost certain that neither the USA nor the UK would have subscribed to the Universal Declaration of Human Rights, or to the subsequent international covenants that gave it legal effect, had they been judicially enforceable by an international court with compulsory jurisdiction. There is a troubling dissonance involved in interpreting a state's agreement to certain arrangements on the basis of the idealised assumption that those arrangements will be judicially enforceable if that agreement was formed without reference to that assumption, or perhaps even in explicit opposition to it. Why should states accept the legitimacy of any such fantasy-based interpretations of the arrangements to which they agreed? Indeed, in what sense has a norm interpreted on this counterfactual basis become "salient"?
The problem of mismatch between salience and idealisation becomes even more acute if we ask about the function the principle of salience is supposed to perform. One function, as mentioned before, is to augment law's determinacy. It helps pick out, from an array of hypothetical arrangements each of which would enhance the legitimacy of international law, the one that states are obliged to comply with. But it is arguable that it serves another valuable function, which is that of improving the effectiveness (and hence, indirectly, also the legitimacy) of international law. The arrangement that has acquired salience might be inferior to other possible arrangements that could have been adopted. Yet, Dworkin would presumably conclude that an obligation still obtains to comply with the salient yet sub-optimal arrangement. One important reason for this conclusion is that the arrangement converged upon is far more likely to be effective in achieving an improvement precisely because it has already attracted the support of a significant number of states. However, it seems very doubtful that the principle of salience can help foster the effectiveness of international law in this way if it licences interpreting norms that states have agreed to on the basis of assumptions that those states do not share or would strongly repudiate.
Dworkin's moralised theory, I suggest, is problematic not only because it seeks to comply with the methodological double duty but also because it adopts a fantasy-based perspective in formulating claims about the grounds and legitimacy of law. It is highly implausible that claims in the here and now about the existence and content of international law, and especially about obligations to obey it, can be helpfully derived from fantasy premises about an international court with compulsory jurisdiction and effective enforcement mechanisms. The irony is that, had Dworkin adopted a slimmed down version of his interpretative approach, one that drops the double duty demand and the fantasy-based test, he might have arrived at an approach to customary international law and jus cogens that fits current legal practice and presents it in a light that helps foster, albeit not guarantee, its legitimacy. 15 Unfortunately, by setting his moralised approach against the largely straw man PVA, he fails to engage with this more realistic and modest approach.

III
The first phase of Dworkin's argument offers a misleading picture of contemporary international legal practice and scholarship as being in thrall to the PVA. The second phase advances a rival, moralised theory of international law whose basis in "fantasy" assumptions prevents it from yielding the key to the validity and legitimacy of international law. The third phase consists of two radical proposals for the reform of key aspects of the international legal order. First, Dworkin calls for a more permissive regime of military intervention in response to crimes against humanity. Specifically, Dworkin envisages the UN General Assembly taking steps to enable military intervention even in the absence of Security Council authorisation. Such authorisation would be subject to a majority of Security Council members supporting the intervention and an ICJ advisory opinion confirming that the intervention responds to a crime against humanity, and not merely the violation of human rights. (22)(23)(24)(25)(26). Second, Dworkin argues that the moralised conception of international law also allows for the creation of a world legislative body by means of a Global Legislative Convention. He sketches a four-majorities system of international legislation whereby proposed laws receive votes representing states with a majority of the UN General Assembly's members' total populations, a majority of votes in the General Assembly, a majority of votes in the Security Council, and a majority of votes among the Council's permanent members. This quadruple-majority legislative process is to be overseen by the International Court of Justice to ensure compliance with the principle of subsidiarity, thereby avoiding the enactment of global legislation that intrudes into matters properly left to state discretion. (28).
Reading these radical proposals in the last few pages of Dworkin's article, it is hard to avoid reflecting on the fact that the 8 years since its publication has been a very long time in international politics. I do not want to delve into the detailed pros and cons of the two reforms advocated by Dworkin but rather to try to draw a larger moral about his embrace of a broadly liberal ideal of international legal order. This is a legal order in which international law uses coercion to promote values such as human rights and democracy and in which, by implication, only states that are compliant with these values enjoy legitimacy and are full members of the international legal community. 16 We can begin by observing that Dworkin betrays not even the slightest inkling of the significance of recent political developments that would soon stall and even reverse the budding emergence of a liberal international legal order. Among these developments are the aftershocks of the global financial crisis of 2008, the devastation wrought by the policies of economic austerity adopted in response to the eurozone's structural flaws, the rise of right-wing authoritarianism in many parts of the globe, and the outbreak of a "populist backlash" that reasserts the importance of state sovereignty and opposes forms of economic globalisation that have made life for citizens in Western democracies increasingly precarious. These developments have led commentators today to speak not only of worldwide "democratic backsliding" but also, more specifically, of the emergence of "authoritarian international law", as nations like China, Brazil, and Russia assume an increasingly prominent role in shaping international norms and institutions in ways that confront and undermine liberal democratic values. 17 Of course, it would be churlish to condemn Dworkin for lacking powers of geopolitical prophesy that many other prominent intellectuals also lacked. After all, his article came at the tail end of a period of heady liberal aspirations for the international legal order. Prominent international lawyers like Thomas Franck were predicting the eventual global triumph of democracy on the back of the inexorable forces of economic globalisation, 18 whilst others like Anne-Marie Slaughter and William Burke-White presented the European Union as the shining future of global governance: "the Treaty of Westphalia… has given way to the Treaty of Rome". 19 Now, were Dworkin alive today, he might well have argued that what we are currently witnessing are only temporary setbacks, that we need not only to continue "watching the space" of an emerging liberal democratic global order but also to make a renewed effort to "nourish its roots" (30). Perhaps; however, I think a more pessimistic conclusion is worth entertaining, which is that Dworkin's theory of international law, precisely in fundamentally orienting itself by reference to a fantasy world that effaces dramatic contrasts between conditions in liberal democracies and those that obtain globally, oversteps the parameters within which a normative theory of international law can be usefully elaborated. We have already seen this in relation to the fantasy-based test for the validity and legitimacy of international law that he employs, but it is also present more generally in the normative ideal of a liberal democratic world order, equipped with powers of military intervention and majoritarian legislation, whose rough outlines are sketched towards the end of this article.
What is lacking here, I believe, is an adequate grasp of the pre-conditions -economic, institutional, and especially cultural -within which liberal democracy can flourish, combined with a sober realisation that there is no way to fast-track progress towards democracy from the outside, least of all by means of coercive intervention. Thinkers such as Mill, Mazzini, and even Rawls and Nussbaum in our own day have stressed the need for an underlying element of solidarity and common sympathies, a sense of belonging, trust, and even emotional attachments worthy of the name "love", in order to underwrite commitment to, and stabilise long-term compliance with, liberal democracy. If this latter picture is correct, there is a broader point arising from the fact that these underlying conditions are probably best secured in modern conditions, if history is any guide, by the sovereign state. For all the supposed opposition between state sovereignty and human rights, the sovereign state, and all that which sustains its sense of identity as one state among others, is the 17 Ginsburg (2020). 18 See Franck (1997); for a more sombre view questioning the idea that economic globalisation spurs the spread of moral cosmopolitanism, see Ignatieff (2018). 19 Slaughter and Burke-White (2006), Ginsburg (2020), 60. primary mechanism for the fulfilment of human rights. One of the best ways to create a human rights disaster, after all, is to engineer a failed state, whether through military or economic means.
It was a grasp of these sorts of considerations that led John Rawls, a more cautious and reliable guide on these matters than Dworkin, I believe, to stress that a realistic global utopia could neither consist in global norms that simply reproduced the demands of liberal democracy nor that restrict full membership to liberal democratic states. In virtue of the ineradicable persistence of ideological pluralism at the global level, Rawls thought that we needed to fashion -precisely for good liberal democratic reasons -an international legal order that was not itself liberal democratic. Such an order could not treat liberal democracy as a condition either of internal legitimacy, in the sense of the moral bindingness of a state's law on its own population, nor external legitimacy, in the sense of immunity from justified intervention. In the same vein, Rawls recoiled from the idea of a world government, echoing Kant's fear that it would deteriorate into either despotism or anarchy, and expressed prescient misgivings about the European Union's federalist project, with its tendency to corrode the sources of national solidarity and operate in the service of powerful corporate interests. 20 Of special relevance here is the strong faith exhibited by Dworkin, throughout the article, in the ability of the International Court of Justice or some other such international judicial institution to maintain the ground rules of a liberal international order by means of judicial review. Among the many problems with this faith in juristocracy, it is worth mentioning that one thing that has been made vivid in recent years is the symbiosis between technocracy -rule by experts, including judges -and populism. Both seek to remove political decision-making on important matters from the domain of democratic contestation: in the case of technocracy, in the name of the superior knowledge of experts, and in the case of exclusionary forms of populism, in the name of the "real people" whose interests and values are channelled by charismatic leaders. And, as Jan-Werner Mueller has pointed out, 21 the populist backlash is often a reaction to technocratic overreach of this sort, with the overall result being the marginalisation of democratic politics which as a matter of empirical fact, appears to be a vital background condition for securing human rights. 22 I realise that these are very large, impressionistic claims that stand in need of a fuller defence. Still, my suggestion is that Dworkin's enthusiasm for a liberal democratic world order, and his preoccupation with combatting a largely mythical Westphalian ideal, makes him overlook a more minimalist, pluralist, and feasibly attainable conception of international legal order. Such an alternative could be defended by someone who believes that liberal democracy is, all 20 See Rawls 1999 andLittle 2011. For a similar critique of misplaced faith in mechanisms of "global governance", including the EU, as instruments of progressive values, rather than corporate interests, see Kuttner (2018), 137. 21 Mueller (2018). 22 On the empirical link between democracy and human rights compliance, see Sikkink (2017). things considered, the best regime type in contemporary circumstances. As I have already suggested, Rawls offers one vision along these lines. Like Dworkin, however, he apparently retains the demand that states that are members in good standing of the international legal order (or, in his jargon, "the Society of Peoples") must be internally legitimate, even if they are not liberal democratic. Rawls is able to encompass non-liberal democratic regimes within his Society of Peoples by setting a comparatively low bar for internal legitimacy, one in which his ultraminimalist schedule of human rights plays an important role. As a matter of logic, however, Dworkin's own position may not be all that distant from Rawls'. Dworkin takes himself to differ significantly from Rawls because he believes that compliance with a more demanding schedule of human rights is a condition of internal legitimacy. Human rights, for Dworkin, are that sub-set of rights that are honoured by any "good faith" or "intelligible" effort to respect the dignity of citizens, understood as the equal value of each person's life and their special responsibility for determining what a worthwhile life is for them. 23 Elsewhere, I have cast doubt on whether this "intelligibility" criterion yields anything like the robust list of human rights Dworkin supposes it does. 24 If so, his position starts veering in the direction of Rawlsian minimalism.
But another, more promising way forward, I believe, is to abandon a fundamental assumption that appears to be common to both Rawls and Dworkin. This is the postulate that full membership in the international community is conditional upon internal legitimacy (where compliance with the full schedule of human rights, whatever that may be, is taken as a condition of internal legitimacy). Such a view would separate internal and external legitimacy, recognising that there are pragmatic and principled reasons to allow certain categories of non-liberal and non-democratic states to be full members of the international legal order. It would allow for human rights to be enshrined in international law without necessarily treating them as benchmarks of internal legitimacy or triggers for external intervention in the case of their extensive violation. Either way, my suggestion is that the space we need to watch is not one of a liberal democratic world order that acquires elements of a federal world state but rather some version of a more minimalist and pluralist alternative that is workable in light of geopolitical realities. 25 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. 23 Dworkin (2011), 327-344. 24 Tasioulas (2013). 25 For one suggestive approach to one form of minimalism of this sort, wee Wuerth (2017).