1. Introduction
There seem to be two clear truths about positivism in international law: it is widely regarded, in both its voluntaristFootnote 1 and formalistFootnote 2 manifestations, as being deeply flawed; and it continues to hold sway. Positivism fails, in the eyes of many critics, to provide a means for linking validity, or legality, with legitimacy, ethics, or justice. Yet, both these versions of positivism have their defenders, among whose number are found a great many scholars who are in fact concerned with finding links between legality and legitimacy, with virtue and law's inner morality, and with democratic principles. A common thread running through this scholarship is a concern with maintaining the distinctive nature of law – with avoiding its conflation with morality, politics, or other normative or social institutions. Both versions of positivism (voluntarism and formalism) find defenders in scholars who see the maintenance of the binary distinction between law and not-law as being normatively grounded.
This scholarship will be explored through the lens of international scholarship on soft law. Soft law poses serious challenges to the binary distinction between law and not-law and, in the eyes of many, to international law itself. This is particularly true of a certain strand of scholarship on soft law that would grant a role in law creation to non-state actors. Some scholars, notably Benedict Kingsbury and Gunther Teubner, seek to articulate conceptions of law that attribute jurisgenerative capacity to non-state actors while maintaining a formalist approach to the validity of international legal rules and therefore maintaining the binary distinction between law and not-law.
My central concern in this paper is to explore the challenges that soft law poses to public international law by focusing on the question of the boundary between law and not-law. I examine various ways in which this boundary is treated in the literature: as something real and important that is nevertheless porous; as real and important but in need of relocation and reconceptualization; or as something that could be done away with altogether. Following this introduction (section 1), the paper addresses the most common approaches to describing and defining soft law, and presents a preliminary definition, adopting a formalist conception of international law's validity (section 2). I then turn to the potential threats posed by soft law to international law's validity (section 3) before examining approaches that seek to link legality with legitimacy without proposing a dramatically revised rule of recognition, focusing on work by Jutta Brunnée, Stephen Toope, and Jan Klabbers that draws on Fuller's internal morality of law (section 4). I then turn to approaches that pose a greater challenge to the rule of recognition and that would open up significantly more space for the participation of non-state actors in processes of international-law formation and implementation: Benedict Kingsbury's concept of international law as inter-public law (section 5) and Gunther Teubner's global law without the state (section 6). I then present some brief comments on the utility of the term ‘soft law’ to international law and legal scholarship (section 7), before concluding (section 8).
2. Soft law: the nature of the category
The many and varied phenomena described as soft law in international legal literature can be roughly divided into three categories: binding legal norms that are vague and open-ended and therefore (arguably) neither justiciable nor enforceable; non-binding norms, such as political or moral obligations, adopted by states; and norms promulgated by non-state actors.Footnote 3 Authors do not necessarily restrict their definitions of soft law to one or another of these categories. The boundaries of the category may be drawn so as to include all three types of norm;Footnote 4 only norms, whether legally binding or not, promulgated by states;Footnote 5 non-binding norms promulgated by states;Footnote 6 non-legally binding norms, regardless of authorship;Footnote 7 or vague and general norms contained in international legal instruments,Footnote 8 to mention the most prominent examples. The various definitions are generated by a series of criteria, sometimes used alone and sometimes in combination with others. The criteria are ‘normativity’ (or justiciability), enforceability, precision, and formal legal status.
2.1. Normativity and enforceability
The term ‘normativity’Footnote 9 is used by certain authors to refer to the right- or obligation-creating character of legal norms. For some authors, normativity refers both to the creation of a legal right or obligation and to the availability of a sanction in the case of violation,Footnote 10 while others appear to treat enforceability as a separate criterion.Footnote 11 Both groups of authors take essentially functional approaches to law: law is law because it accomplishes certain things. There are two problems with a functional approach. First, legal rules perform various functions. Nicholas Onuf identifies three: directive, assertive, and commissive.Footnote 12 Only the first is captured by a definition of legal rules as obligations backed by sanctions. Other essential functions of law, such as constituting authorities, granting powers, or conferring competencies, are not adequately captured by this definition. The assertive function is found in definitions: the portion of the ocean within 12 nautical miles of the baseline is the territorial sea. This function is also filled by judgements: the assault by the troops of state A on those of state B was an act of self-defence. The commissive function is evidenced in secondary rules that govern the creation of legal obligations: state A accepts to be legally bound by a convention.
A second problem with functionalism is that many other kinds of norm carry out the same functions as legal rules.Footnote 13 Functional approaches are therefore both over- and underinclusive.Footnote 14 This point can be illustrated with reference to Anthony D'Amato's argument that legal rules are enforceable rules backed by sanction. ‘The essence of any soft-law rule,’ he argues, ‘is that it is unenforceable.’Footnote 15 He goes on:
A soft-law system will allow an infraction to be cost-effective: that is, a violator of a norm of soft law may suffer a reputational loss, but reputational damage may be well worth the benefits that are derived from non-compliance with the norm. By contrast, a hard-law system must, without exception, endeavour to make every violation cost-ineffective.Footnote 16
Soft law, then, is ‘a naked norm, whereas hard law is a norm clothed in a penalty’.Footnote 17 On these terms, provisions such as the definition of a treaty found in the Vienna Convention on the Law of Treaties could be described as soft law. At the same time, the myriad consequences that might befall an actor that disregards rules, commands, or threats of a non-legal nature are ignored: the only cost that counts is that imposed by a legal penalty. If cost-effectiveness of violations is the criterion for distinguishing hard from soft obligations, the lines around international law would probably have to be redrawn altogether; indeed, law would become indistinguishable from the exercise of power.Footnote 18
The criteria of obligation creation and enforceability seem to refer to a third approach to positivism – one more closely related to social sciences than to legal science, which places emphasis on the effectiveness of law – on its measurable impact on behaviour and outcomes. It might be assumed that among the most effective legal rules are those that clearly communicate an obligation, but, even if this is a sound assumption, which is open to question, the validity of a legal rule is and must be a separate issue from its effectiveness.Footnote 19 Clearly, law's effectiveness and the pathways through which law has an impact on the world are issues of great concern to jurists as well as to scholars in cognate disciplines. But criteria based on effectiveness are less helpful – though certainly far from irrelevant – for identifying the bases of law's validity.Footnote 20
2.2. Precision
Many authors categorize norms as ‘soft’ due to their lack of precision.Footnote 21 This can overlap with the criterion regarding the creation of an obligation – it is argued by many that vague provisions may set out objectives but cannot create obligations – but is nevertheless distinct. For example, a provision calling on parties to endeavour to reduce greenhouse gas emissions or to improve literacy rates among girls and women creates an obligation, though one whose fulfilment is difficult to measure. Such obligations, like the obligation in Quebec civil law ‘to abide by the rules of conduct which lie upon [one], according to the circumstances, usage or law, so as not to cause injury to another’,Footnote 22 are certainly vague, and any attempt to specify the location of the threshold between legal and illegal behaviour, as defined by this provision, would fail. Fortunately, judges are not required to locate this threshold; they are rather required to determine whether it has been passed or not.
The precision of an obligation depends not only on the legal text itself, but also on the thickness of shared understandings that support the rule. The obligation regarding negligence drawn from the Quebec Civil Code, cited above, is expressed in vague and open-ended language, and its application in particular cases is often extremely difficult but, in Quebec society, there is a dense network of shared understandings regarding applicable rules of conduct on which citizens, lawyers, and judges can draw in evaluating behaviour. In international law, in which shared understandings are much thinner and more fragile, vague and open-ended legal provisions may be more problematic. Particularly troubling is the use of vague language as a deliberate strategy to create the illusion of agreement and resolution.Footnote 23 Prosper Weil refers to deliberately vague rules as ‘“precarious” norms’ and states that the proliferation of such norms ‘does not help strengthen the international normative system’.Footnote 24 But he also notes – correctly, in my view – that ‘[a] rule of treaty or customary law may be vague, “soft;” but . . . it does not thereby cease to be a legal norm’.Footnote 25
2.3. Formal status
A third criterion for identifying soft law, which is adopted here, focuses on the manner in which a rule comes into being. If the rule meets the criteria contained in the rule of recognition of positive rules of law, it is a legal rule; if not, then it is not. Of course, there are different accounts of what the rule of recognition is. A ‘useful starting point’Footnote 26 is the list of sources in Article 38(1) of the Statute of the International Court of Justice, but it is probably no more than a starting point.Footnote 27 Increasingly controversial, however, is the proposition that the rule of recognition is based on state consent – a position that Jean d'Aspremont argues has been central to international-law scholarship since the nineteenth century and has come to be challenged only in recent decades.Footnote 28 Even among scholars committed to formalism in international law, a rule of recognition phrased in terms of a state's consent to be bound encounters resistance. For example, d'Aspremont and Klabbers, while recognizing the centrality of intent to be bound to the boundary between law and not-law, remind us that it is the rule of recognition and not the simple expression of consent that confers legally binding status on a rule.Footnote 29 This question will occupy a significant portion of the discussion below. For present purposes, the point to be made is that, following a formal approach, ‘soft’ law is not-law, though that need not be the end of the story. There may be reasons to identify a subset of non-legal rules, norms, or standards and place them under the rubric ‘soft law’. Proponents of this approach seek to identify criteria to distinguish soft international law from the mass of norms, rules, and standards that may exist at any given time: the term ‘soft law’ is taken to refer to a body of norms that are relevant to international law in some way, such as because of their close resemblance to law, the extent to which they are taken up in legal discourse, the extent of consensus around them, their influence on the behaviour of international actors and outcomes in international society, and other such factors.Footnote 30
As a step in the direction of a definition of soft law, then, soft law is not law, but is somehow of relevance to law. The reasons for adopting this approach have already been identified in the above discussion regarding the range of criteria used in the literature to define and describe soft law, and can be summarized as follows. Definitions of law based on function – approaches that focus, for example, on the command-like structure of a rule and the availability of sanction in case of violation – are over- and underinclusive and do not help us to understand how rules of international law differ from other normative statements, commands, or threats. For these reasons, approaches that focus on the normativity or justiciability of rules to distinguish between soft and hard law must be set aside. As for approaches that focus on the precision of a rule, they must be rejected, for similar reasons. Lack of precision may indeed cause a rule to be non-justiciable, particularly when the vague and general language in which the rule is expressed is not undergirded by shared understandings in the legal community. But the meaning to be ascribed to a legal rule often develops gradually and indeed can change significantly as jurists and laypersons seek to interpret and apply it.Footnote 31 Therefore, while lack of precision may in many cases affect the quality of a rule, its effectiveness, its influence, its compliance pull, etc., one should not too quickly reach a conclusion that deprives the rule of its rule-ness.
3. Soft law: threats and challenges
The potential dangers posed by soft law depend on one's approach to defining it. Vague provisions in legally binding instruments may be regarded as a waste of valuable time and effort, or as creating the illusion of agreement among parties and resolution of a problem.Footnote 32 When such vague provisions come before third-party dispute-settlement bodies, the wide discretion that they confer on those bodies may be cause for concern.Footnote 33
The main concern regarding definitions of soft law as non-binding agreements concluded by states – political declarations, unilateral statements by political authorities, non-binding resolutions, recommendations, and decisions adopted by inter-governmental bodies – appears to be a muddying of the waters. Potentially applicable norms proliferate, some of which may be mutually incompatible.Footnote 34 The clarity provided by a binary approach to the definition of law is lost; actors can no longer be sure which rules apply, or with what force, or with what consequences in the case of violation.Footnote 35 Key functions of legal systems – the provision of order, predictability, and stability – are compromised.
A further concern lies with democratic process: soft-law instruments may seem preferable because they are easier to adopt, particularly if the rigours of debate and approval through formal law-making processes can be dispensed with.Footnote 36 It is frequently observed that states negotiate non-binding declarations, and even press releases, with almost as much care as they do binding instruments.Footnote 37 Often, however, the appeal of soft law may lie in the relative ease with which it can be created because democratic processes and other formal procedures required for the creation of legally binding rules can be circumvented.Footnote 38
The version of soft law that presents the most serious challenge to international law's rule of recognition is that which encompasses norms promulgated by non-state actors.Footnote 39 However, many of the authors who include such norms in their definitions of soft law do not wish to efface the boundary between law and not-law or even to relocate it: they place soft law outside the boundary of international law. For many such authors, norms produced by non-state actors are of interest because of their lex ferenda character – their influence on the development of international law. They may ask what conditions seem to favour this transformation,Footnote 40 or whether the creation of a strong consensus around a norm before it crosses the boundary has an impact on its effectiveness or on perceptions of its legitimacy once it is transformed into law.Footnote 41 Alternatively, they may be interested in the influence that discourse, debate, and consensus formation in civil society or in more specialized fora have over processes of law creation, interpretation, and application.Footnote 42 For these authors, soft-law norms are legal facts that, as d'Aspremont notes:
can still produce legal effects . . . [such as] partak[ing] in the internationalisation of the subject matter, provid[ing] guidelines for the interpretation of other legal acts, or pav[ing] the way for further subsequent practice that may one day be taken into account for the emergence of a norm of customary international law.Footnote 43
Soft law is often seen as a means to address certain acknowledged weaknesses of the international legal system: the limited effect of many legal norms on state behaviour and the relative paucity of sanctions for violations, the democratic deficit,Footnote 44 the slowness and reluctance with which international legal institutions respond to grave problems in international society,Footnote 45 and the woeful inadequacy of many of those responses.Footnote 46 Many critics of soft law acknowledge these weaknesses but are nevertheless concerned to preserve a binary definition of law, not because of a deep-seated commitment to it, but because they conclude that it is in fact a more effective means of pursuing principles such as democracy, rule of law, and collective self-determination than any readily available alternatives. For example, Weil refers to the ‘two essential functions’ of international law: ‘to reduce anarchy through the elaboration of norms of conduct enabling orderly relations to be established among sovereign and equal states . . . [and] to serve the common aims of members of the international community’.Footnote 47 He then argues that the fulfilment of this function requires international law to retain a number of essential features: voluntarism, neutrality, and positivism.Footnote 48 These, importantly, are not presented as ends in themselves, but rather as features essential to international law's ability to carry out its twin functions of coexistence and common aims. He freely acknowledges that ‘neither the basis nor the ultimate justification of international law is to be found in the normative system as such’ but argues that ‘it is still necessary for that system to be perceived as a self-contained, self-sufficient world’.Footnote 49
Klabbers's conclusion that soft law is ‘undesirable’ rests in large part on the latitude that soft law gives to actors who already enjoy extensive power and influence in international society – power and influence that formal processes of law creation serve, in some measure, to constrain. Soft law, Klabbers argues, is not autonomous of morality and politics and risks being ‘a fig leaf for power’.Footnote 50 In a similar vein, Anna Di Robilant summarizes a line of critique of soft law as follows: ‘“Pluralistic participation” in soft governance processes is limited to visible and powerful social actors, reinforcing and asserting existing power structures and cleavages rather than encouraging openness. Soft rhetoric, they [critics of soft law] contend, masks hard practices’.Footnote 51 She is speaking of the European context, but this critique becomes all the more powerful when we move to the international level and consider undemocratic states: any attempt to represent at the international level the voices of the citizens of authoritarian states would have to rely heavily on what participants imagine those citizens would say if they could speak up.
One of the threats posed by soft law is to the boundary between law and politics. Many soft-law norms are in fact political or ethical values presented in the language of law. Klabbers, drawing on the writings of Hannah Arendt, highlights problems with this approach, notably with its tendency to conflate politics and law to the detriment of both. Arendt, Klabbers notes, insists on the public nature of politics, which is seen not as the mere aggregation of interests, but rather as debate and deliberation in a public sphere in which people gather to make collective judgements.Footnote 52 As Klabbers puts it, Arendt seeks to ‘[develop] a style of thinking about politics where individuals jointly, in all their plurality, take care of the world, and assume responsibility for it together’.Footnote 53
Arendt's concept of natalityFootnote 54 is key, first as it relates to plurality: the collective judgements made in the public sphere are not the result of a gradual homogenization of interests or values through processes of discourse, but rather of an agonistic confrontation among people, each of whom is utterly unique.Footnote 55 Each human action unleashes chains of events on the world that are utterly unpredictable and quickly escape the control of their authors.Footnote 56 The only institutions available to place some bounds on the uncontrollability and unpredictability of human interaction are promises and forgiveness.Footnote 57 Arendt's focus on promises manifests itself in an interest in contract, and more generally in positive law.Footnote 58 Her approach to law focuses closely on the contract as a means of creating ‘“islands of predictability,” or “guideposts of reliability”’.Footnote 59 Klabbers states:
If it is the case that force, domination, and rule are not authentically political, then any attempt to bring legislation and politics (in the Arendtian sense) together would have to stress the consensual nature of law, all law, including legislation.Footnote 60
Of course, international political processes that lead to law formation bear little resemblance to Arendt's deliberation in the public sphere. Domestic legislative processes, even in robust democracies, fall short as well. But, in domestic democratic systems, there are at least attempts to ensure the existence and operation of ‘sluices’ between informal processes of will and opinion formation on the one hand, and formal legislative processes on the other.Footnote 61 It is probably fair to say that international society does possess something like a public sphere (or rather spheres), but it is much harder to argue that these spheres have strong links with formal processes of law-making. This is precisely one of the objectives of certain proponents of soft law – to render processes of law formation more genuinely public. The question is whether, and how, this could be done.
4. Diminishing the danger
At the heart of much literature on soft law is a desire to move beyond a description of international law's validity that depends directly (as in voluntarism) or indirectly (as in formalism) on state consent. In the words of Benedict Kingsbury:
The idea that international law should speak to the whole of society is evident in the continuous efforts to nudge the field beyond states-will theories of sources, beyond bilaterality and opposability, toward community norms, beyond a focus on managing disputes and adversarial proceedings, toward a deeper structure of normative enunciation and claims arising from neighbourhood and impact rather than contract and technical legal interests. It appears in the idea of jus cogens –peremptory norms applicable to all, which no group of states can contract out of – and in other modern natural law ideas. It appears in the frequent resort to ‘general international law’ rather than simply the specific agreement made by the parties in a dispute.Footnote 62
The weaknesses of voluntarism have been well understood for decades, and many authors express surprise at the durability of this approach to international law's validity.Footnote 63 But the reason is not far to seek, as Kingsbury notes:
Much of the effort of international lawyers in the century since Oppenheim wrote has gone into broadening the functioning legal conception of international society from the narrowly statist one of Oppenheim's ‘Family of Nations’. But it is difficult to argue that a robust theory of international law has as yet accompanied these newer accounts of more and more inclusive and complex international society with disaggregated states, an infinite diversity of non-state actors, private or hybrid rule-making, and an ever expanding range of topics covered by competing systems or fragments of norms. The extensive cognitive and material reconstruction required to actualize emancipatory projects such as that of Philip Allott is indicative of the scale of the challenge. However unappealing Oppenheim's approach has seemed, its coherence and manageability are normative attractions that make its continuing political influence intelligible.Footnote 64
More particularly, Kingsbury notes that a renewed interest in voluntarist approaches is spurred by a desire to connect international law-making processes to democratic principles.Footnote 65 Nevertheless, Kingsbury, along with a number of other authors, has sought to rethink international law's rules of recognition. The challenge, as Klabbers and Koskenniemi make eminently clear,Footnote 66 is to do so in a manner that does not threaten to undermine law by conflating it with cognate social systems such as politics, ethics, or economics. Lon L. Fuller's conception of the internal morality of law provides an excellent starting point for discussions of legality and legitimacy in the heterogeneous setting of international society. Fuller proposed a set of criteria that allow us to grasp the particularity of law while at the same time linking legality and legitimacy.Footnote 67
Fuller's approach is apt for international law, with its decentralized, horizontal structure and its dependence on its subjects’ adherence to its rules. The notion that law's legitimacy depends on its capacity to achieve acceptance among the members of the society to which it is addressed is central to Fuller's notion of a legal system that supports the self-determination of its addressees. In the first place, Fuller rejects the idea that law's effectiveness can be based on the notion of public order, on the use or threat of force, or on a formal hierarchy of authority.Footnote 68 Fuller argues that legal systems cannot be regarded as structures of authority to which their addressees are subject,Footnote 69 but rather consist of ‘the enterprise of subjecting human conduct to the governance of rules’Footnote 70 – an enterprise in which legislators, the administrators of law (administrative authorities and judges), and addressees participate.Footnote 71 Gerald Postema refers to this as Fuller's vertical interaction thesis:Footnote 72
First, law regulates or guides actions of citizens by addressing reasons or norms to them. Rather than altering the social or natural environment of action, or manipulating (nonrational) psychological determinants of action, law seeks to influence behavior by influencing deliberation. It addresses norms to agents and expects them to guide their actions by those norms. Moreover, it expects those norms to figure in deliberation not as contextual features setting the environment or parameters of choice, but as reasons for deliberate choice. Thus, rules are intended to be ‘internal’ in two respects: (a) they figure in the deliberation of agents, and (b) they figure as reasons for, and not merely parameters of, deliberation and choice.
Second, law seeks to influence deliberation in a wholesale fashion, not through detailed step-by-step instructions, but through general norms that agents must interpret and apply to their specific practical situations.Footnote 73
The effectiveness of rules is thus not based on their ability to inform their addressees precisely what forms of behaviour are required of them, or the ability of a judge to apply the rules without having to engage in interpretive processes.Footnote 74 Once the rule has been articulated, its addressees and those charged with its application must begin the process of determining what it means in individual cases. In international law, this task tends to fall to the addressees themselves, as third-party adjudication is not often resorted to.
For Fuller, the uniqueness of law is due not to formal criteria, but to its internal morality. In the words of Jutta Brunnée and Stephen J. Toope, ‘[w]hat distinguishes law from other types of social ordering is not form, but adherence to specific criteria of legality: generality, promulgation, non-retroactivity, clarity, non-contradiction, not asking the impossible, constancy, and congruence between rules and official action’.Footnote 75 This procedural approach, note Brunnée and Toope, allows a link to be established between legality and legitimacy, without assuming a thick set of shared values or an international community.Footnote 76
One potential contribution of this interactional approach to international law is to shed light on the role of non-state actors in the project of law. Because a legal rule can only function as such if it generates a sense of legal obligation, rules concluded under pressure of time, power, or economic clout may qualify on a positivist account as rules of law, but may fail to generate fidelity. Rules rooted in understandings shared by many international actors, on the other hand, may be capable of doing so.Footnote 77 Indeed, Brunnée and Toope acknowledge that certain non-binding rules may have a greater capacity to generate fidelity than certain binding rules.Footnote 78 A rule may become capable over time of generating a sense of legal obligation, or may lose that capacity: for example, states formally bound by the rule may not respect it, or may ignore it in formulating legal arguments about their own or others’ behaviour. The extent to which a norm generates a sense of legal obligation can be influenced by non-state actors as they critique (or, less frequently, praise) state behaviour or make representations in legal and political arenas.Footnote 79
Brunnée's and Toope's insistence on the importance of links between legitimacy and legality, the role of non-state actors in the project of international law, and their welcoming attitude towards the influence of non-binding norms on legal arguments do not lead them to deny the existence or importance of a boundary between law and not-law.Footnote 80 Similarly, Klabbers, while embracing Fuller's internal-morality arguments, concludes that formal criteria are still needed to distinguish law from other forms of normativity.Footnote 81 He acknowledges that ‘state consent cannot explain the binding force of the legal system’ but adds that ‘it can – or rather could – explain most individual rules of international law in most individual settings’.Footnote 82 To attenuate the weaknesses of this approach, Klabbers argues for ‘“presumptive law”: normative utterances should be presumed to give rise to law, unless and until the opposite can somehow be proved’.Footnote 83
These approaches certainly take their distance from (certain versions of) positivism, and may create space for various kinds of soft law or, more precisely, for appeals to soft-law norms for some purposes, but, at the same time, they acknowledge the ongoing need for a formal rule of recognition based on state consent. The two approaches considered below – international law as inter-public law and global law without the state – rely on a distinction between law and not-law but propose to redraw that boundary.
5. Public international law as inter-public law
Benedict Kingsbury proposes a modified rule of recognition for international law – one that incorporates ‘publicness’ as a criterion for formally binding legal rules. He describes ‘publicness’ as ‘the claim made for law that it has been wrought by the whole society, by the public, and the connected claim that law addresses matters of concern to the society as such’.Footnote 84 He further argues that ‘“[p]ublicness” is a necessary element in the concept of law under modern democratic conditions’,Footnote 85 presenting a modified version of Hart's rule of recognition that ‘include[s] a stipulation that only rules and institutions meeting these publicness requirements immanent in public law (and evidenced through comparative materials) can be regarded as law’.Footnote 86 Kingsbury fleshes out his notion of publicness with reference to a series of general principles, presented as an indicative list: the principle of legality, according to which authorities ‘are constrained to act in accordance with the rules of the system’; the principle of rationality, or the requirement that reasons be given for decisions; the principle of proportionality between means and ends; rule of law, understood in a procedural sense; and human rights.Footnote 87 He argues that:
in choosing to claim to be law, or in pursuing law-like practices dependent on law-like reasoning and attractions, or in being evaluated as a law-like normative order by other actors determining what weight to give to the norms and decisions of a particular global governance entity, a particular global governance entity or regime embraces or is assessed by reference to the attributes, constraints and normative commitments that are immanent in public law.Footnote 88
This approach does more than alter the basis on which international law is recognized as law; it also makes it possible for non-state entities to possess jurisgenerative capacity.
Kingsbury's dissatisfaction with a voluntarist approach arises from the difficulty of explaining why states should have jurisgenerative capacity, and why other types of actor should not.Footnote 89 He notes that ‘the concept of the state as a juridical unit . . . does not adequately reflect the quality of states as public law entities, a quality that distinguishes them from mere “rational actors”’. Second, ‘the jus inter gentes model of international law does not account adequately for the burgeoning activities of regulatory entities that are neither states nor simple delegates of states’.Footnote 90 As to the first point regarding the reasons why states should be seen to have jurisgenerative capacity, Kingsbury does not seek to question this capacity, but rather to rethink its conceptual grounding:
When states – as public law entities and committed to publicness in law – come together with each other in an international legal rule-making and decision-making normative process, the results are not identical in form or meaning to what would result from a comparable process among unitary rational non-public actors.Footnote 91
Actors possessing jurisgenerative capacity on this understanding include states, but not as such – rather, because they are ‘entities that are themselves public – operating under their own public law, and oriented toward publicness as a requirement of law’.Footnote 92 But so are certain other actors.Footnote 93
Kingsbury's dissatisfaction with voluntarism does not lead him to the conclusion that one can or must treat ‘every normative assertion in transnational governance as international law, on condition only that it is made with a claim to authority and establishes a sense of obligation’.Footnote 94 He notes that ‘a convincing rule of recognition for a legal system that is not simply the inter-state system has not been formulated’Footnote 95 and lays the groundwork for a new approach to developing such a rule, locating the point of origin of ‘the normative content of law . . . in the public nature of law itself’.Footnote 96
Kingsbury's approach goes a great distance towards answering many of the concerns of critics of soft law. It takes seriously the distinction between law and not-law, purporting to identify a uniform rule of recognition that allows one to identify valid legal rules. The distinction is formal, based neither on the function of rules nor on their content. The criteria proposed to distinguish law from not-law are robust: an actor seeking to demonstrate the existence of a legal rule will be required to produce evidence of the rule's validity and will not be able to rely on wishful thinking. Finally – and this is a central contribution of this approach – the normative criteria proposed for distinguishing between law and not-law are much more robust than the references to democracy that have come to serve as a justification for voluntarism.Footnote 97
Kingsbury does not seek to ground his rule of recognition in principles of representative democracy.Footnote 98 The qualification of public entities as such does not depend, in Kingsbury's conception, on the extent to which they actually represent the will or interests of specific publics.Footnote 99 This approach has been criticized as giving rise to a ‘fundamental legitimacy crisis’ that Kingsbury seeks to resolve by ‘locat[ing] its legitimacy [of law] outside democratic control’.Footnote 100 This is true, on the whole, but, as a practical matter, the legitimization of international law through democratic control is very difficult to conceive of.Footnote 101 If Klabbers and Koskenniemi are right about the threats posed by soft law to law and politics, then attempts to ground international law on genuinely democratic foundations might pose more of a threat to than an opportunity for international law and respect for democratic principles. Kingsbury's assessment of the medium-term prospects for cosmopolitan democracy at the global level leads him to search for other, firmer ground for law's legitimacy.
Kingsbury's approach has been criticized for relying more heavily on natural law, and less on positivism, than he claims. Alexander Somek makes a couple of different points in this respect. First, he notes that global administrative law (GAL) has brought within its compass a range of processes that go beyond the exercise by rule-making and rule-applying bodies of authority delegated by a legal authority, which, he argues, are of doubtful relevance to administrative law ‘if they do not give rise to the adoption of legally binding administrative acts’.Footnote 102 Somek calls on scholars of GAL to ‘explain which of the phenomena it studies are to be described as law’.Footnote 103 For his part, Kingsbury has done so, in two different ways. First, he explains why the kinds of procedure, rule, standard, etc. to which he and colleagues refer should be understood as forming part of a body of GAL, referring to David Dyzenhaus's distinction among constitutive, substantive, and procedural administrative law.Footnote 104 Second, he acknowledges that he is not limiting himself to a discussion of law, if law is to be understood as based on state consent; he states that the term ‘global’ was preferred to ‘international’:
to avoid implying that this is part of the recognized lex lata or indeed lex ferenda, and instead to include informal institutional arrangements . . . and other normative practices and sources that are not encompassed within standard conceptions of ‘international law’.Footnote 105
He seeks a new rule of recognition, and the categories of norms to which he refers do or could meet the criteria of this new rule.Footnote 106
Another aspect of Somek's criticism is that the principles in light of which administrative acts are to be judged seem to be derived using natural-law, not positive-law, methodologies. ‘The underlying idea,’ writes Somek:
appears to be that no people with ‘mature reason’ (Kant) would adopt laws incompatible with freedom (hence, rationality and proportionality along with the rule of law) or equality (hence, the principle of legality). The regulative principles that flesh out the meaning of what are necessary components of the volonté générale are presented as principles underlying any law, and hence, any public law.Footnote 107
These principles may be compatible with natural-law thinking; in other words, they may be susceptible to justification in natural-law terms. But Kingsbury's methodology is empirical, perhaps influenced by certain comparative approaches: he looks to ‘diverse substantive regimes of global regulatory practice’ for insights into principles that may be emerging; his own contribution is not to derive them from first principles, as in natural law, but to proceed in an inductive fashion,Footnote 108 observing the practice of diverse public authorities, and to categorize and evaluate those practices in light of his version of a Hartian rule of recognition that incorporates publicness as a criterion.Footnote 109 Furthermore, there is no indication that Kingsbury is presenting these principles as the final word on GAL: he makes it clear that this is an emerging field.Footnote 110 The principles ought to be understood as works in progress and not as a constitution, to my mind.
Yet another approach to the drawing of a boundary around law seeks to implicate individuals and social groups directly in international law-making processes, rejecting the centrality of the state in those processes but at the same time maintaining – or, more accurately, working towards – a clear distinction between law and other social systems such as autopoiesis. This approach, based on autopoietic theory as developed by the sociologist Niklas Luhmann and the jurist Gunther Teubner, identifies an emerging body of ‘global law without the state’ and argues that this could be the future for law under conditions of globalization.
6. Global law without the state
Some scholars of autopoietic theory, notably Gunther Teubner and Karl-Heinz Ladeur, seek through global law without the state to carve out an even greater role for non-state actors in the generation of legal norms. Autopoietic theory takes a formal approach to the definition of law: the legal system is distinguished from other social systems through its use of the code legal/illegal lawful/unlawfulFootnote 111 to distinguish itself from and communicate with its environment. But the formalism of autopoiesis is obviously different from that of positive international law, in which state consent is generally regarded as being a formal criterion, and a fundamental one at that. For theorists of global law without the state, state consent is not a requirement for the validity of legal rules. This emerging global law is law, argues Teubner, in the sense that it constitutes a social system distinct from morality, politics, economics, and other social systems,Footnote 112 but not in the sense of emanating from national legal processes or the sovereign will.Footnote 113 Rather, it ‘emerges from various globalization processes in multiple sectors of civil society independently of the laws of the nation-states’.Footnote 114 Teubner rejects functional or structural criteria for defining law generally, and global law more in particular, in favour of the nature of the binary code used by the legal system to distinguish itself from its environment and to identify that which is relevant for the legal system.
It is difficult to grasp – and, for this author, even more difficult to describe – the nature of law's binary code. This is largely because it rests on a tautology: law is law because the legal system says it is law. Let us look at this from the point of view of participants in a legal argument. We can readily accept that lawyers representing plaintiffs and defendants, presenting arguments to a judge, are engaging in legal argumentation. It is probably not difficult to accept that two businesspeople involved in negotiations and making reference to legal rules as they understand them are also engaging in legal argumentation. But, one might object, if they are referring to rules that do not in fact exist (i.e., that have not been formally adopted in conformity with the legal system's rule of recognition), then they are not really engaged in legal argumentation. Perhaps, although autopoietic scholars would probably conclude that, in fact, they are – that this negotiation does indeed constitute an operation of the legal system.
To take this line of argument one step further, what if actors were not arguing about the application of legal rules they believe to exist, but rather about the rules whose operation allow legal rules to be created? In other words, actors may make claims about the nature and content of the rule of recognition. Can they actually create a rule of recognition in this manner? Yes, according to autopoietic theory, though the mere assertion by a small group of actors that a rule of recognition exists is not sufficient.
Global law's validity rests, argues Teubner, on a ‘paradox of self-reference’:Footnote 115 the rule is valid because its authors declare it to be valid. This is a tautology, of course, but Teubner argues that it does little harm because it is concealed through a combination of ‘time, hierarchy and externalisation’.Footnote 116 Time is relatively straightforward: the circumstances under which a rule emerged are, with time, forgotten; the rule takes on a self-evident character. As to hierarchy, Teubner describes the phenomenon of the ‘self-regulatory contract which goes far beyond one particular commercial transaction and establishes a whole private legal order with a claim to global validity’.Footnote 117 The presence within this ‘legal order’ of ‘an internal hierarchy of contract rules’, including primary and secondary rules, further obscures the paradox.Footnote 118 Third, the process of externalization involves referring processes of validation, interpretation, and application to other institutions that appear to be external to the contract but that in fact are created by it. The most prominent example is commercial arbitration, ‘which has to judge the validity of the contracts, although its own validity is based on the very contract the validity of which it is supposed to be judging!’.Footnote 119
Teubner is acutely aware of problems relating to the ‘democratic deficit’ of global law. Discussing this problem in light of lex mercatoria, he notes that this body of non-state-based law:
is extremely vulnerable to interest and power pressures from economic processes. Since there is no institutional insulation of its quasi-legislation and its quasi-jurisdiction, the relative autonomy and independence which historically national legal orders have been able to achieve will probably remain something unknown. For the foreseeable future lex mercatoria will be a corrupt law – in the technical sense of the Latin word corrumpere. At the same time, lack of institutional autonomy makes this law vulnerable to political pressures for its political ‘legitimation’.Footnote 120
Could the corruption of lex mercatoria be cured by granting this body of law greater autonomy from powerful economic actors? This emphasis on the autonomy of law is instructive, not least because it recalls arguments made by Klabbers and Koskenniemi about the independence of law from politics and vice versa.
Teubner does not believe that global law without the state is inevitably corrupt. Rather than seeking to protect state-based law from ‘private’ governance, he asks, what would happen if these private governmental activities were seen as jurisgenerative? This might prompt us, he suggests, to ‘ask more urgently than before the question: What is this “private legal regime's” democratic legitimation?’. It would be ‘naïve’, he argues, to think of the democratic legitimation of such regimes in the narrow terms of parliamentary democracy: ‘Rather, we are provoked to look for new forms of democratic legitimation of private government that would bring economic, technical and professional action under public scrutiny and control.’ He describes this as the:
liberating move that the paradox of global law without the state has actually provoked: an expansion of constitutionalism into private law production which would take into account that ‘private’ governments are ‘public’ governments. And the potentially fruitful analogy to traditional political democracy might lie in the rudimentary consensual elements in contract, organization and other extra-legal norm producing mechanisms. Is a democratization of these rudimentary consensual elements feasible?Footnote 121
Teubner is as alive as are Kingsbury, Klabbers, and Koskenniemi to the weaknesses of the current sources of international law's validity. But he is also fully aware of the power that these private forms of governance often wield. His proposed approach is reminiscent of Arendt's observation that the only thing to counter power is more power – that trying to hem power in with rules and constraints is not nearly as effective as creating more sites of power.Footnote 122 Arendt's rather idiosyncratic description of power – that it ‘springs up between men when they act together and vanishes the moment they disperse’ – is evocative in this context. But Teubner's proposition involves a leap of faith: if we invite lex mercatoria into the house, will it throw a huge corporate bash and wreck the place?
If one hesitates to take on board the full implications of global law without the state – international and domestic law sidelined, if not altogether replaced, by norms emerging through networks of private and public actors – this approach is nevertheless of great value. First and foremost, it presents a different perspective on the function and importance of the boundary between law and not-law, even if it proposes to draw it in a dramatically new way. Second, it presents a theoretical and methodological framework for analysing an extremely important phenomenon: the growth of norms and indeed entire regimes created and implemented by non-state actors. The legitimacy and effectiveness of these norms and regimes are secured – to the extent that they are secured – without resort to the authority of the state, but in ways that nevertheless resemble state-based law. Global law without the state, and autopoietic theory more generally, helps us to understand the logic that can make private authority effective, even when the actors seeking to exercise that authority do not themselves wield immense economic or political clout.
Organizations such as the Forest Stewardship Council (FSC), a non-governmental organization that has created and administers a programme for the certification of forestry industries and related distributors and retailers based on sustainable practices, provide excellent objects for autopoietic analysis. The FSC has no formal authority to certify forestry industries, and therefore must rely on a combination of economic incentive structures and appeals to political and ethical principles to transform its certification into a commodity that firms are anxious to get and keep.Footnote 123 The FSC could potentially qualify as a public actor on Kingsbury's definition: it seeks to provide a public good, namely sustainable harvesting of forest products,Footnote 124 and it has taken care to structure its decision-making process along democratic lines: transparency is striven for, and the three ‘chambers’ representing stakeholder groups are clearly intended to ensure representativeness as well as responsiveness to a constituency wider than the organization's members.Footnote 125 In other words, the FSC has taken pains to structure itself as a public actor, rather than a pressure group seeking to promote the interests of its members. However, global law without the state brings an additional dimension to the analysis, prompting questions about the extent to which the authority which the FSC wields or seeks to wield has been ‘constitutionalized’ – a complex concept that draws attention, in particular, to the extent to which the political and legal facets of its activity are separate from one another.Footnote 126
7. Soft law: a useful concept?
Two separate questions should perhaps be put: first, is soft law a useful concept for international legal scholars? Second, is it useful for international jurists more generally? My answer to the first question would be a qualified ‘no’, and to the second a qualified ‘yes’. For international legal scholars, the term simply means too many different things to different people. Even if one were able to settle on a definition, most such definitions encompass such a wide range of phenomena that the usefulness of the category for legal and scholarly analysis is severely limited. My ‘no’ is qualified because I strongly believe that the phenomena generally referred to by the term ‘soft law’ are extremely important to both scholarship and practice. Certainly, the term ‘soft law’ can be of some service in describing a fairly eclectic body of principles, rules, documents, statements, and various forms of communication that both scholars and practitioners ignore at their peril. But, from an analytical point of view, it makes more sense to use the categories legal act and legal fact, as suggested by d'Aspremont.Footnote 127
Aside from the problem of the eclecticism of the category of phenomena that could count as ‘soft law’, the term is unfortunate in that it evokes a blurring of the boundary between law and not-law, which leads almost inevitably to a blurring of boundaries between law and politics, law and economics, law and science, etc. This is a particular problem for projects such as Kingsbury's and Teubner's. Neither uses the term to describe the law that would issue from a newly constituted rule of recognition, and for good reason: neither suggests that law that has not emerged by way of state consent is any less ‘law-like’ for that. But both acknowledge the need to make distinctions between law and politics. Indeed, Teubner insists on it: law that is insufficiently distinguished from politics is corrupt, or readily subject to corruption.
8. Concluding remarks
A recurring theme in much, though by no means all, of the literature touching on or relevant to soft law is the need for formal criteria for identifying law and explaining its legitimacy. At the same time, formal criteria need not be incompatible with the maintenance of robust links between legality and legitimacy. The boundary around international law is not hermetic. Processes of legal argumentation inevitably move back and forth across it. But the interaction of law with other social systems ought not to lead to the collapse of law into those systems.
Perhaps the greatest challenge to current conceptions of the sources of international law's validity is the assumption by non-state actors of the mantle of legal or law-like authority. Norms and standards promulgated by non-state actors are clearly not-law on most positivist conceptions. Yet, their close resemblance to law – a resemblance that is often desired and deliberately pursued by soft law's authors, state and non-state – compels legal scholars to think carefully about what distinguishes such efforts from positive international law. The answer is not far to seek – state consent for voluntarists and respect for international law's rule of recognition (whatever form it may take) for formalists. But these answers prompt further questions about the adequacy of state consent as a sine qua non for the creation of international law and, most pointedly, about the normative arguments underpinning state consent. As we have seen, the arguments in favour of a positivist framework are often phrased in negative terms – the alternatives provide inadequate guarantees of certainty, order, stability, and clear differentiations between power and law. Yet, these arguments are no less compelling for all that.
Throughout this paper, I have made much of the importance of the boundary surrounding law and the necessity of maintaining a distinction between law and other social systems. This position is strongly influenced by autopoietic theory and is born as much of a concern for the integrity of cognate social systems, notably politics, as it is for the integrity of law. Political processes, when they work reasonably well, permit the telling of stories from various points of view, the consultation of experts, the weighing of different types of objective, reference to languages of efficiency as well as justice, and careful attention to possible consequences. Attempts to package this wide-ranging discourse into the language of legal rights and obligations involve the imposition of significant constraints on the stories that can be told and how they can be related. This is not meant as a criticism of law, which is capable of doing certain things that political processes cannot accomplish, or can accomplish only with difficulty: reaching decisions with some expediency and finality, arriving at solutions without purporting to make determinations about the right and the good, providing actors with means to interact with one another and project their wills into the future, to name a few. The debate about soft law is, on one level, a debate about the relative functions and merits of law and politics and about the interaction of the two. Interaction there must be, but the boundary between the two is of importance, nowhere more so than in international society.
