1. Soft law and the concept of international law: conundrums and challenges
It is a well-known fact that international organizations, formal ones like the United Nations as much as informal ones like the G8, more and more frequently adopt rules which their drafters do not consider to be ‘legally binding’, although they otherwise have all the textual characteristics of binding international treaties or binding resolutions of international organizations.Footnote 1 For instance, the OECD Export Credits Arrangement stipulates in a law-like fashion all the detailed requirements, substantive and procedural, that states need to observe when extending credits or credit guarantees to their enterprises in order to facilitate exports to developing states. Politically, the Arrangement has been a great success. Most states observe it most of the time.Footnote 2 Nevertheless, Section 2 of the Arrangement stipulates that it is only a non-binding ‘Gentlemen's Agreement’.Footnote 3 On a doctrinal level, the ramifications of non-binding international rules, generically referred to as ‘soft law’, are well understood. Soft law may evidence the formation of customary law, guide the interpretation of treaties, authorize action by international organizations, and give rise to duties of good faith such as a duty to consider.Footnote 4 On the other hand, the breach of soft law does not entail the same legal consequences as violations of binding international law, commonly referred to as ‘hard law’. In particular, it does not trigger state responsibility, nor does it give rise to a cause of action before the International Court of Justice.
On a theoretical level, however, soft law has remained a conundrum, even though it is by no means a brand-new phenomenon – in fact, the earliest examples of soft law are almost as old as the earliest modern multilateral treaties.Footnote 5 It implies a dual challenge for the concept of law. First, why should soft law be excluded from the definition of international law if it looks like international law and basically functions like international law? The concept of international law is not a natural given. Its definition is a prudential question and depends on the specific purposes for which one wants to define international law.Footnote 6 Given that, for a vast number of violations of international law, state responsibility and judicial proceedings are merely theoretical options, and given that some binding international treaties contain provisions that are too vague to effectively guide state behaviour or to give rise to state responsibility, one could question the wisdom of current standard definitions of international law. What is a definition of law good for which stresses differences that are arguably of little practical impact, or relegates soft law to the realm of ‘political’ or ‘moral’ rules, although it often resembles much more a refined legal regime than political or moral rules that are normally very abstract and general? Second, since soft law has been under much criticism for its alleged illegitimacy or inferior legitimacy, does – or should – the concept of law draw a meaningful line between more or less legitimate norms? In the tradition of liberalism, the designation of a rule as legal normally implies some degree of legitimacy.
Both of these challenges might require some additional clarification. As regards the first challenge, the present standard definition of binding international law in international legal practice basically corresponds to the sources triad of Article 38(1) of the ICJ Statute (which is understood as also comprising ‘binding’ resolutions of international organizations and unilateral acts). Given that soft law has become a ubiquitous governance instrument that plays in many cases the role of a functional equivalent to binding international law, this definition might appear awkward and arbitrary. The functional equivalence of soft law might also justify putting it on a par with binding international law in order to better reflect in the conceptual framework the reality of contemporary international governance. There seems to be little point in making a conceptual distinction between two kinds of norm that has little practical significance, especially because enforcement of ‘binding’ international law is not always readily available.
An obvious counterargument to this proposal would be to point out the fact that states wanted these two classes of norm to be different. This argument could be rejected by questioning its axiomatic presumption, namely that the definition of international law is only to reproduce the intentions of states. Nevertheless, putting hard law and soft law on a par because of their functional equivalence would call into question the distinction between legal instruments, soft and hard, on the one hand and entirely non-legal instruments like information, statistics, etc. on the other. The latter type of instrument may also assume functions and produce effects in international relations that are similar to those of international law. For example, the OECD Programme for International Student Assessment (PISA) consists of a survey of 15-year-old students that is published every three years in the form of country rankings. The survey is based on a particular concept of education. Although the published reports fall short of making any recommendation, the survey effectively drives states towards adopting that particular concept of education unless they want to be put to shame every three years.Footnote 7 Now, if the concept of law is to be meaningful at all, the PISA reports cannot be considered law. Law establishes, confirms, or destroys normative expectations, not only cognitive expectations.Footnote 8 The PISA reports first of all address cognitive expectations. True, they might trigger public debates that give rise to normative expectations. However, unlike normative expectations, cognitive expectations might be trumped by countervailing cognitive expectations, namely by proving the PISA reports wrong. They do not need to be formally abolished. Thus, functional equivalence cannot be the sole reason that should guide the definition of the concept of law.
For these reasons, some deem it better to replace the categorical, binary distinction between law and non-law with a sliding scale, or a multi-step model that is able to grasp the gradual differences between the various types of instrument issued by international organizations.Footnote 9 But would it be worth sacrificing the particular rationality that might be inherent in a binary distinction between law and non-law purely in order to have a more adequate reflection of practice?
Regarding the second challenge: many approaches in legal theory recognize an innate connection between law and legitimacy. Accordingly, law claims to be an expression of legitimate authority.Footnote 10 This view is even shared by exclusive positivists like Joseph Raz. Although he considers the morality of a legal norm to have no impact on its validity, each legal norm claims to express legitimate authority.Footnote 11 Others again contend that the legal system as a whole raises this claim.Footnote 12 However, the legitimacy of binding international law faces challenges – even more so the legitimacy of soft law. Traditionally, the basis of international law's legitimacy is state consent. Now, state consent itself is problematic as a basis of legitimacy if one takes individuals, and not states, as the ultimate entities from which any legitimate decision needs to derive.Footnote 13 But, even if one accepts state consent as a device for legitimizing law, one has to recognize its shortcomings in present-day international relations. State consent is only very indirect in case of non-plenary bodies like the UN Security Council, in case of long chains of delegation, or in case of majority votes. Majority votes are particularly common for adopting soft law. Further, unlike international treaties, soft law does not need to be ratified. Nevertheless, it is often effective and sometimes even has an impact upon third states that did not actively participate in its adoption. This might be because other institutions like the WTO or the international financial institutions integrate soft law into their legal framework or because a particular soft-law instrument comes to frame the discourse within a given issue area.Footnote 14 In light of these considerations, one could think about defining the concept of international law in a way that ensures greater legitimacy of international law. One might only consider as law those instruments that meet certain other conditions in addition to, or instead of, state consent. This would endow the concept of law with the function of identifying legitimate exercises of authority and distinguishing them from those that are less legitimate.
In reaction to the emergence of soft law, a great variety of positions has formed in international legal scholarship. Each of them proposes a specific concept of international law. The purpose of this article is, first, to systematize these positions in accordance with the particular tradition of international legal thought that they represent. Second, the article analyses how each of these positions responds to the two challenges set out above. It finds that recent approaches that stress the publicness of public international law are very promising. However, I argue that these approaches will ultimately only be successful in tackling those two challenges if the concept of law is dissociated from, and complimented by, a concept of international public authority. This idea draws on Michel Foucault's insight that, in the course of modernity, state activity expanded more and more beyond law-making. By now, it comprises a great variety of instruments that enable governments to influence the behaviour of the population by recognizing and influencing their desires and their way of thinking. Among these instruments are incentive schemes, statistical information, or educational programmes.Footnote 15 Since the end of the Second World War, and even more so since the emergence of globalization, the activities of international institutions have moved in this direction, too, producing more and more soft law and entirely non-legal instruments. In light of this, I argue that the concept of law should be discharged from the function of defining and delimiting the realm of international public authority. Only under this condition will the concept of law regain a meaningful function that is in line with the two challenges mentioned.
The following analysis distinguishes legal traditions by a mix of historical and analytical factors. Regarding the former, I consider the critical legal studies movement and its reception in international law as a turning point in the intellectual history of international legal scholarship (section 4). Some of the positions hardly take the implications of critical legal studies into account (sections 2 and 3), while others do (section 5). Analytically, following the terminology coined by H. L. A. Hart,Footnote 16 I distinguish positions that allow approaching international law from an internal perspective (legal positivism) from positions that are not interested in the internal point of view and concentrate instead on its external, real-life effects (sociological positivism or legal realism). Internal perspectives are interested in the normative prerequisites and effects of law, namely in questions of legal obligation. External perspectives are interested in the social prerequisites and effects of law, namely in empirical questions. Further, the positions take different views concerning the existence of an international order and its basis of legitimacy. Some consider states to be the be-all and end-all. Others give a stronger role to individuals, sometimes going as far as to endorse the existence of an international community of citizens. Finally, some positions consider law as a binary category that follows an on–off scheme, while others accept relative normativity. The concept of law thus defined by one particular position also determines the view that its adherents take on soft law. It should therefore not surprise the esteemed reader that this article is, to a large extent, about diverging concepts of law.
2. Legal positivism: soft law looked at from the internal point of view
Legal positivist views all share an internal point of view.Footnote 17 Legal positivist positions also agree that rules may guide behaviour and discard the doubts raised against this view by critical legal studies. But they take different views on the subjects and the mechanisms from which international law derives its legitimacy. This leads to different concepts of law and different responses to the two challenges.
2.1. Sovereign or democratic statehood: traditional legal positivism
As traditional legal positivism, I consider certain approaches in international legal theory that understand international law strictly as the product of state consent. There is an older and more recent strand of this approach. Each of them has a different motive for emphasizing state consent. The older strand goes back to the late nineteenth and early twentieth centuries.Footnote 18 It takes its origins in the writings of predominantly German authors such as Bergbohm, von Holtzendorff, Jellinek, or Triepel. These authors faced a particular challenge: they had to provide a theory of international law on the basis of a Hegelian idea of the state, the prevailing ideology in their country at the time.Footnote 19 This idea of the state might seem incompatible with the idea of international law. It sees the state as an indispensable requirement for individual freedom and an ethical (sittliches) life.Footnote 20 Consequently, the state is the only possible subject that might create legitimate law, and no legitimate normative order may exist beyond the state. Not surprisingly, Hegel denies the existence of international law proper. All he accepts is an ‘external law of the state’, which does not have binding force vis-à-vis other states.Footnote 21
Not so the jurists of the older strand of traditional legal positivism. They argue that it is possible to base an international legal order solely on sovereign statehood, namely on state consent.Footnote 22 This bridging of the seemingly unbridgeable poles of absolute state sovereignty and binding international law requires a great number of theoretical wrenches. Koskenniemi opines that only a specifically German understanding of consent, characterized by Kantian deontology, could have done so.Footnote 23 The authors take different routes to reach their goal. For one, Triepel idealizes (or transcendentalizes, if you want) the corresponding consent of two or multiple states, from which he believes that a new, independent common will (Gemeinwille) should arise.Footnote 24 By contrast, Jellinek considers the binding force of consent as a psychological phenomenon – with the important qualification that this phenomenon would only arise inasmuch as the consented rule is in line with the objective purpose of the state.Footnote 25 These constructions enable the authors to reject both natural-law theories of international law and the views of those who deny the existence of international law.Footnote 26 They argue that international law exists and is valid if consented. This implies that international law should not be gapless, but strictly limited to rules to which consent was given. This position found expression in the famous Lotus judgment of the Permanent Court of International Justice.Footnote 27
More recent authors who seem to follow this strand of thinking emancipate themselves from the Hegelian idea of the state. Not a concern for sovereignty, but one for democracy is the guiding motive for their insistence on state consent as the one and only foundation of international law. Since international institutions do not meet domestic standards of democracy, binding legal norms always have to be linked to democratic decision-making processes on the domestic level.Footnote 28 Otherwise, international law would become the instrument of unchecked executive power.Footnote 29 Others go one step further and hold that the creation of an international law without state consent would require a social substratum in the form of an international community to which this law could be attributed (like the nation in case of domestic law), which, though theoretically possible, they do not believe to exist at present.Footnote 30 The ability of state consent to give rise to binding legal obligations, which created so many worries for earlier generations of jurists, seems to be less of an issue in the recent literature following this tradition. Jan Klabbers takes it for granted that legal agreements have binding force for lack of a better alternative. Moral and political agreements could not be binding in the same way. Klabbers considers it irreconcilable with the nature of morality to negotiate and adopt moral principles like one negotiates and adopts international treaties. Also, he argues that politics has no normative system except for law. For this reason, there could not be a political agreement without there being a legal one, too.Footnote 31
Authors following the point of view of traditional legal positivism make a categorical, binary distinction between law and non-law. There is no relative legal normativity for them. Either an agreement is part of binding international law or it is no law at all, but merely of political or moral significance.Footnote 32 Others put it a bit more mildly, classifying soft law as a fait juridique, a legally relevant fact.Footnote 33 What are the reasons for this strictly binary concept of international law? It does not seem to result from a concern about the lack of centralized enforcement for soft law. For this is also the reality for a large number of binding international rules. Some authors give reasons that are not fully convincing and therefore might not reveal the true reasons. In The Concept of Treaty, Klabbers makes a purely negative argument in favour of a binary concept of law. He thinks the advocates of relative normativity would not reach their goal by accepting soft law as a form of law. In that case, a new binary distinction between soft law and non-soft law would emerge.Footnote 34 However, this reasoning does not take into account that it should make a difference whether one recognizes only one single binary structured type of law or two or multiple ones.
For these considerations, I suppose that there must be a deeper reason for the aversion of traditional legal positivism against relative legal normativity. This reason might have to do with the challenge of combining sovereignty (or domestic democracy) with the idea of an international legal order. The idea of a sovereign not subject to the law, as it was coined during absolutism,Footnote 35 is only compatible with the idea of an (international) legal order if one considers the sovereign obliged to exercise its power through law,Footnote 36 while imagining law as being autonomous from politics, morality, and religion. This strategy found its most obvious expression in the metaphorical idea of the ‘sources of law’, which is widely shared among proponents of this position:Footnote 37 according to sources theory, just as the water pours out of a mountain at one point in a spring (in German, Quelle, i.e., source), legal rules have only one point of origin. The afflux inside the mountain is invisible, confined to the will of the sovereign. By contrast, from the moment of its creation at the source, the rule becomes a legal one, autonomous from politics, morality, and religion. Its interpretation and application are considered a matter of legal craft, of correct understanding, and not a political, moral, or religious issue. Thus, under the theory of the sources, there is only a punctual link between law and other normative systems, and this link is under the full control of the sovereign. Not the procedure, not the affluxes, only the consent of the sovereign is what legitimizes a legal rule.Footnote 38
The idea of the sources of law still has a prominent place in the works of Montesquieu and Austin, even though they consider the people as the true sovereign and thereby reinterpret the theory of the sources in a democratic fashion.Footnote 39 I believe that this is also the viewpoint of contemporary representatives of traditional legal positivism. As a consequence, they consider soft law to be a threat to the sources doctrine – thus, as a threat to the autonomy of the law and an undesirable danger to domestic democracy. In corroboration of this view, they tender into evidence the fact that soft law is sometimes susceptible to majority decisions, which might favour powerful states,Footnote 40 or that it does not require ratification by domestic actors. Such considerations informed some philippics against soft law that culminate in the claim to dispense with it entirely.Footnote 41
By contrast, for this position, the ideal international agreement is the treaty. Some maintain this view even though they recognize the difficulties in distinguishing between binding and non-binding norms: one needs to derive the intention to be bound from its external manifestations, which are never straightforward.Footnote 42 This position also runs into trouble with respect to customary law, general principles of law, and peremptory norms. Such norms derive only indirectly from the consent of sovereign states.Footnote 43 Moreover, the answers this position gives to the two challenges enumerated at the outset are less than satisfactory. The call for the elimination of soft law reminds one of Don Quixote's fight against windmills. It also glosses over the fact that soft law might have advantages, like increased flexibility or the possibility of testing some rules before concluding an international treaty – however one might balance them against the risks. And, by delegitimizing soft law entirely, it overlooks that binding international law often faces similar challenges.
2.2. International community of states: modern legal positivism
For some decades, a modernized form of legal positivism has provided the standard concept of international law in practice and scholarship. Like traditional legal positivism, it emphasizes consent and state sovereignty as the foundations of international law. But it modifies these assumptions in a decisive point. It holds that binding law requires a hierarchically superior norm, a kind of Grundnorm from which to derive the binding force of law. Alfred Verdross made this point long ago.Footnote 44 In contrast to Jellinek, this position accepts a superior normative order above the state, namely an international community of states. Opinions diverge as to the actual structure of this order, covering the entire spectrum of the universalistic tradition in political theory.Footnote 45 Some see the basis of the international order in human nature, similar to Grotius's idea of the appetitus socialis,Footnote 46 others in universal values,Footnote 47 or in reason like Kant.Footnote 48 Others again, like Kelsen and Verdross, consider pacta sunt servanda as the Grundnorm whose validity requires no further justification.Footnote 49
This starting point does not deprive sovereign statehood of its significance for international relations, particularly for the creation of international law. However, its relative importance decreases. Sovereignty is not an absolute given any more, but an integral part of the idea of a community of states.Footnote 50 This accounts for a different understanding of law. The importance of state consent decreases. For example, this position has no difficulty closing lacunae in international law by way of analogy.Footnote 51
Still, this position by and large follows a binary understanding of international law, which excludes soft law from the ambit of binding law. An international agreement is only legal if its authors agreed on its legally binding character, which may be inferred from its form, its content, and the context of its conclusion.Footnote 52 For example, it is a relatively safe indicator of the legal nature of an agreement if it contains a clause that grants the parties the right to initiate proceedings before the international Court of Justice in case of a violation of the agreement by the other party.Footnote 53 Any agreement that does not have binding legal character cannot be considered law unless one changes the Grundnorm. Therefore, while soft law might have political or moral significance,Footnote 54 the term is ultimately considered a misnomer.Footnote 55
What is the reason for the insistence on a binary concept of law? It seems to me that there must be more to this view than the circularity of the arguments of those who appeal to ‘logic’Footnote 56 or any axiomatic and therefore question-begging concept of law.Footnote 57 Also, purely negative lines of argument are usually inconclusive. For example, Blutman discards the idea that soft law that functions like hard law should be recognized as law. He claims, first, that states do not feel obliged to follow soft law (which begs the question); second, that it would be difficult to distinguish between soft law and other ‘non-binding norms’ (which could be solved by establishing clear formal or procedural standards for soft law); and third, that the anti-formal tendency of soft law is irreconcilable with the idea of law (which overlooks the high degree of formality of much of today's soft law).Footnote 58 He further argues that the indirect legal repercussions of soft law, such as its role in the ascertainment of opinio juris, were not enough to justify the attribution of legal character to soft law. All of these repercussions could be explained while maintaining a strict separation between law and non-legal (i.e., soft) agreements.Footnote 59 This, however, misses what I have called the first challenge to the concept of law: although one might not like soft law, it often seems to be functionally equivalent to hard law.
In fact, my impression is that advocates of this position are truly concerned about the effectiveness of international law. They fear that a relative concept of law would relativize the significance of international law and reduce its impact on state behaviour.Footnote 60 What might explain this anxiety is the context of the time at which the concept of law advocated by this position was formulated. At its origin, this concept of law stood in opposition to the sceptical view of international law that was put forward so powerfully by Carl Schmitt.Footnote 61 Even later, after the Second World War, when soft law became a recurrent phenomenon for the first time, the advocates of this position saw their views questioned by the realist school of thought. In those coldest days of the Cold War, many deemed the realist revival of the theories of Thucycides,Footnote 62 Hobbes,Footnote 63 and Schmitt the preferable explanation for the state of international relations.Footnote 64 Against this background, those favouring a modern legal positivist view had every reason to distinguish international as clearly as possible from anything resembling ‘politics’, like soft law. Considering soft law as a valid form of law would have been water on the mills of the realists.
Thus, in line with this assumption, representatives of modern legal positivism do not reject soft law as a dangerous development like traditional legal positivism.Footnote 65 First, they recognize that soft law represents at least a certain form of state consent, though an inchoate one compared to binding law. Second, state consent itself becomes less important on the premises of this theory, as law is ultimately based on a rule (or a fact) that precedes consent. As the significance of state consent decreases, that of the various stages leading to a binding rule or, in terms of the source metaphor, that of the affluxes increases. In this process, soft law might play a role as a ‘material’ source that contains a certain rule, though not in the form required for hard lawFootnote 66 or as a source for the recognition of binding legal rules.Footnote 67
Still, modern legal positivism does not consider soft law as international law proper. Just as physicists view dark matter only through its interplay with visible matter, modern legal positivists grasp soft law purely through its interplay with hard law. Thus, other than being an indicator for the emergence or development of customary law, soft law might guide the interpretation of binding law or lead to cases of estoppel. These aspects of soft law have received extensive coverage in the literature, both by the advocates of modern legal positivism and by more traditional positivists who are sceptical about the virtue of soft law but ultimately too realistic to think that wishing it away would do the job.Footnote 68
Nevertheless, such an indirect approach to soft law misses out some important functions of soft law. First and foremost, it tends to overlook the fact that many non-binding instruments are drafted like binding international law, provide a basis for legal reasoning and disputes, and have a significant impact on the issue area concerned. In short, they do not really differ from binding international law, particularly from the binding secondary law of international organizations,Footnote 69 except for the fact that they are officially ‘non-binding’.Footnote 70 Rather, there is much soft law that looks more like ‘law’ as we know it from the domestic level than some full-grown international treaties that are kept in very vague terms.Footnote 71 Second, how should one explain the frequently recognized binding effect of soft law within an international organization?Footnote 72 Is the internal order of an international organization not part of international law? Third, the indirect approach cannot explain the dynamic character of many soft-law instruments. It only sees them as the avant-garde of legalization in a process that may lead from soft to hard law.Footnote 73 However, this is only one function of soft law. The formation of customary international law is a long and hazardous process. If this was all that soft law was about, it could not be explained why international organizations like soft law so much and also update their instruments frequently. Understanding this requires recognizing soft law as a functional equivalent to binding international law.
In sum, modern legal positivism cannot give a satisfactory response to the various roles assumed by soft-law instruments in contemporary international relations. While covering some of the functions and significance of soft law, it still misses out a lot. Also, modern legal positivism cannot make a meaningful contribution to the debate about the legitimacy of soft law. Essentially, it only relies on state consent. If this was convincing, there would not be a debate about the legitimacy of soft law.
2.3. Towards an international community of citizens: positivist reform proposals
In light of these reasons, several authors propose reformulations of the positivist concept of law and understand certain types of soft law as law. Most of these reformulations are motivated by the transformations of the international order since the end of the Second World War,Footnote 74 particularly by two developments: first, decolonization and the end of the Cold War created an international community of states that is less European and more heterogeneous; second, the individual entered stage as a subject of international law. Individuals now enjoy rights and bear obligations. Taken together, these two developments considerably downsize the idea of state sovereignty.Footnote 75 Not only states, but all individuals together form an international communityFootnote 76 – an idea anticipated by Immanuel Kant long ago.Footnote 77
According to this view, soft law allows for a broader and at the same time more nuanced articulation of the will of the international community as expressed in, for example, majority decisions or decisions adopted without any intention to create ‘binding’ international law. The advocates of this position therefore consider it unsatisfactory to confine soft law to the level of purely political or moral obligations.Footnote 78
In addition, those maintaining this position point out the practical insignificance of the distinction between hard and soft law. Certain ‘binding’ international treaties might, if violated, hardly trigger consequences other than diplomatic protests or reputational losses. Their provisions are too vague or too controversial for another state to claim damages or launch reprisals. All that might follow are reputational sanctions – which can be triggered by soft law as well. Certain soft-law instruments are enforced by mechanisms like peer-review procedures. An example of this would be the Millennium Declaration of the UN General Assembly.Footnote 79 Every five years, states review the progress achieved towards the attainment of the Millennium Development Goals. Also, both hard and soft law usually result from formalized law-making procedures, which distinguishes them from moral obligations based on reason, nature, discourse, etc. Relegating soft law to the field of political or moral obligations therefore does not seem to be a promising strategy.
Like modern positivism, in this view, upgrading soft law to the rank of law does not compromise sovereignty, as the latter is merely considered a conceptual tool for the delimitation of powers between states and international organizations. But, in contrast to modern legal positivism, positivist reform proposals do not fear the risk that considering soft law as law would cheapen hard international law and compromise its effectiveness. Rather, they hold that the chances that soft law provides for the international community outweigh the risks.
Nevertheless, these considerations do not explain what kind of law soft law is exactly. This requires a reconsideration of the concept of international law. At this point, the various proposals choose different routes and legal theories. Some of them maintain the traditional binary distinction between law and non-law, while others define soft law as one or several separate categories of law (dualist and pluralist approaches).
2.3.1. Binary approaches
In the view of those choosing binary approaches, there is only one kind of international law. Nevertheless, they count some types of soft law as international legal instruments. Some writers provide a voluntaristic foundation for this concept of law. Asamoah argues that some resolutions of the UN General Assembly express a form of state consent that is identical to the consent required for binding international law.Footnote 80 Abi-Saab reaches the same result, but emphasizes the decision-making procedure leading to some resolutions instead of the final vote. He deliberately leaves open the question of the source of their binding nature.Footnote 81
Godefridus van Hoof proposes a non-voluntaristic theory. He follows H. L. A. Hart's idea of law as the rules identified by a rule of recognition that rests on acceptance by the international community.Footnote 82 According to Hart, international law is what the rule of recognition accepted by state practice identifies as an international legal norm. Van Hoof argues that current state practice recognizes a rule as a rule of international law if it is based on the consent of the states participating in its creation. Consent may manifest itself in different ways, not just through treaty law and customary law.Footnote 83 Given the mentioned geopolitical transformations, it may also be expressed in resolutions of the UN General Assembly. Van Hoof proposes five ‘points of recognition’ that guide the search for such state consent in a particular instrument,Footnote 84 namely the level of abstraction of the instrument (which he considers to be particularly high in the declarations of the General Assembly); the travaux préparatoires; the text of the instrument, particularly its designation; the existence of follow-up mechanisms; and state practice subsequent to the adoption of the instrument.Footnote 85
Thus, each of the binary views concludes that certain types of soft law like resolutions of the UN General Assembly – clearly the phenotype of soft law getting the most attention – may be part of binding international law. The beauty of this view is that it is compatible with the sources doctrine. There seems to be widespread agreement that Article 38 of the ICJ Statute does not stipulate an enumerative list of sources, but only reflects the state of international legal doctrine at the time of its creation.Footnote 86 The ICJ itself confirmed this view by recognizing unilateral acts as sources of international law not mentioned in Article 38.Footnote 87 In addition, most of the mentioned authors propose generic definitions, which have the advantage that not every new instrument beyond the traditional sources of international law requires a reinvention of the wheel.Footnote 88 Nevertheless, these views still leave the nature of a rather large number of soft-law instruments in the dark, even though they might be functionally equivalent to international treaties. This is a consequence of the binary structure of the proposed concept of international law. Treating all international legal instruments equally means restricting the ambit of instruments that might be classified as legal. Some of the legal consequences of this reclassification, particularly the duty to pay damages in case of a violation, are simply inappropriate for a number of soft-law instruments that are otherwise functionally equivalent to international treaties. For example, avoiding damages might be the very reason why the Export Credits Arrangement declares itself as a ‘gentlemen's agreement’, though being otherwise very similar to an international treaty.Footnote 89 Binary approaches are therefore, in many respects, not in a position to meaningfully facilitate the legal conceptualization of soft law.
Regarding issues of legitimacy, the binary positivist view does address some of them, as it attributes the status of binding law to some declarations considered non-binding by traditional doctrine because of their high degree of legitimacy. Other questions of legitimacy, however, such as those raised by soft-law instruments that affect states and individuals that did not have a say in the decision-making, tend to be neglected.
2.3.2. Dualistic approaches
The adherents of dualistic positions consider soft law a separate category of international law. Unlike hard law, it does not entail specific legal consequences of violations, such as the duty to pay damages, or the right to take recourse to reprisals or proceedings before the International Court of Justice. Otherwise, soft law is deemed to share all the characteristics of legal rules, such as their binding nature, and the full range of general international law is applicable to it.Footnote 90 Most authors base their view on a voluntaristic line of reasoning, emphasizing that states deliberately choose to create instruments of different quality.Footnote 91
This concept of law takes into account the fact that soft law might be functionally equivalent to hard law. Unlike the binary view, it allows classifying a large number of soft-law instruments as law. Nevertheless, putting all soft-law instruments into the same basket yields a relatively crude result and does not account for differences in the social effects of various soft-law instruments. Thus, technical soft-law instruments that are enforced by review mechanisms are put on a par with highly political, vague, and aspirational norms.
As concerns legitimacy issues, there is no difference between this position and the binary view. It does not link the qualification of an instrument with its legitimacy, except that soft law might be adopted by the majority. The dualist approach appears as nothing more than a semantic shift from modern legal positivism. While the latter distinguishes law and non-law, dualist approaches distinguish hard and soft law. All objections against the legitimacy of soft law remain applicable.
2.3.3. Pluralistic approaches
Another group of authors modify the dualist view. Although they stick to the broad distinction between binding and non-binding international law, they split non-binding norms into further subcategories. This amounts to a pluralistic definition of legal norms.
René-Jean Dupuy proposes declaratory and programmatory law of international organizations as two categories of non-binding law. Declaratory law refers to resolutions of international organizations that confirm customary law, and programmatory law covers aspirational instruments that are aimed at the further development of international law. For the identification of programmatory law, he proposes a set of criteria, such as a voting pattern that shows widespread support for the instrument, the degree of precision, and the presence of follow-up mechanisms.Footnote 92 Compared to dualist approaches, this proposal further specifies the legal conceptualization of soft law. However, Dupuy focuses on the interrelationship between soft and hard international law. Soft law that does not aim at creating binding law is not covered by this proposal.
Hillgenberg takes up the challenge to understand as law even those soft-law instruments that are not meant as contributions to the formation of hard law, but operate independently of such effects. Rather, he holds that every soft legal instrument establishes a self-contained regime that produces specific legal and factual effects, and to which a specific set of rules of general international law are applicable.Footnote 93 From a legitimacy perspective, this approach seems promising. It would allow setting specific criteria for each regime that ensure its legitimacy. However, Hillgenberg does not tap this potential and entirely relies on state consent for the legitimacy of soft law.
Hillgenberg's ‘compartmentalization’ of soft law also suffers from a very low degree of conceptual abstraction. It considers each instrument individually and does not produce meaningful larger groups of instruments. This requires reinventing the wheel for each instrument anew. There is no template that would indicate which legal principles are applicable to a specific soft-law instrument. Hillgenberg thus forgoes the rationalizing potential of classifying an instrument as legal. In addition, the transfer of the doctrinal concept of self-contained regimes to the level of theory does not work. A new source of law cannot be adopted like an international treaty setting up a self-contained regime. Rather, it requires rethinking the rule of recognition.Footnote 94 Hillgenberg does not explore these questions, but assumes them away in a questionable manner. For example, why should non-binding international agreements have the quality of law, but not the resolutions of the UN General Assembly?Footnote 95 By taking the concept of self-contained regimes as a compass, Hillgenberg overlooks the implicit theoretical challenge.
2.4. Intermediate conclusions
The positivist reform proposals offset the shortcomings of modern legal positivism only to a limited extent. They do not second-guess state consent as the basis of the legitimacy of international law. Also, most of them do not do justice to the whole spectrum of soft-law instruments and propose relatively crude distinctions that do not seem to match the complex social reality of soft law.
3. Sociological positivism: soft law looked at from the external point of view
The difficulty of positivist approaches to grasp reality led to an important strand in legal thinking that ultimately gives up the legal positivist paradigm and the internal perspective with its focus on validity. I refer to them here as sociological positivist approaches. Some prefer to speak of legal realism. Like sociologists, proponents of sociological positivism understand law in the first line as a social factFootnote 96 and usually define it by its effects. Beyond this, however, the various views do not have much in common and strongly disagree about the concept of international law. Again, such disagreement results from diverging ideas about the properties of the international order.
3.1. The international community of states: functionalism and compliance theories
Functionalist approaches follow an idea of the international legal order similar to that of modern legal positivism. Both consider the international community of states as the central actor on the international plane. In matters of legal theory, however, functionalist approaches stand in the tradition of American legal realism. According to the latter, legal rules are much more indeterminate than legal positivism usually admits. Law alone does not provide courts with a basis of decision. Rather, judges follow their own ideas of fairness. Only after a decision has been taken do they try to justify it with legal reasoning. Legal scholarship should therefore analyse which reasons were really determinative for a decision.Footnote 97 Against this background, functionalist approaches think of public international law and international organizations as tools of states which serve a particular purpose that legitimizes them.Footnote 98 They focus on the politics in and of international law and institutions instead of questions of validity or interpretation.Footnote 99 Why do states and other international actors conclude soft or hard agreements?Footnote 100 When or why do states comply with binding or non-binding international norms?Footnote 101 It should not come as a surprise that some term this approach ‘managerial’.Footnote 102
There is no common functionalist concept of international law. One frequently encounters a relative concept of legal normativity, which is rarely theorized.Footnote 103 The terms ‘soft law’ and ‘hard law’ are often used in a heuristic sense and should not be taken as signals of a dualist understanding. Rather, some writers propose a fluid continuum of legal normativity, while others favour broader categories.
Transnational legal process is located on the more fluid side of the spectrum. Although this approach historically evolved from the New Haven school, it does not share the normative conviction of the latter that law should be used for the realization of social goals.Footnote 104 Transnational legal process assumes that law is constantly in a process of evolution in which the normativity of specific rules increases and decreases. Legal scholarship is not about characterizing a certain rule in a certain point of time, but about understanding this process of change and its effects on social reality.Footnote 105
Most functionalist approaches, however, see law as a more stable social phenomenon and try to conceptualize different categories of international law.Footnote 106 Kenneth W. Abbott and his co-authors propose three different stages of legalization. This quantification of legal normativity provides the conceptual framework for their research on the conditions under which states choose specific forms of co-operation.Footnote 107 Each of the three stages is meant as an ideal type and defined by the degree of obligation, precision, and delegation involved in a legal regime. ‘Hard legalization’ features a high degree of obligation and delegation, such as the rules and regulations of domestic public law. All regimes that do not meet this threshold are considered ‘soft legalization’ as long as at least one of the parameters is strong. Otherwise, there is no ‘legalization’, which constitutes the third ideal type.Footnote 108 In the end, this seems to be little more than a reproduction of the traditional distinction between hard and soft law.
Theoretically, functionalist approaches should be able to produce a more nuanced categorization or typology of soft and hard law than legal positivism because they focus on social reality and the effects of rules, not on their formal qualities. As far as I can see, however, there are no proposals that lie in between the amorphous concept of law favoured by transnational legal process and the broad categories of the legalization school. Maybe this lack of meaningful typologies is due to the focus on social effects, which is fraught with a host of difficulties. There is little merit in defining as law only those rules that actually influence state behaviour. It is never possible to say whether states follow a rule because it is a rule, because the rule happens to be in their interest, or because the rule prescribes what they have always done anyway. Most functionalist authors therefore consider as a legal norm every promise that generates an expectation of compliance and thereby increases the chance of actual compliance.Footnote 109 Yet, this approach is hard to apply in practice. The generation of expectations is an intersubjective matter, depending both on the issuer and the recipient of a promise and on their interpretations of the textual basis of the norm.Footnote 110 The double contingency of law seems to defeat the objectivist aspirations of functionalism.Footnote 111
Authors who identify law as a certain practice or mode of argumentation avoid these difficulties.Footnote 112 However, this leads to an entirely amorphous concept of law, as the example of transnational legal process shows. Such a concept of law is of little practical use.Footnote 113 The key function of the concept of law – the distinction between legal and illegal acts – requires an assessment of a legal rule at a certain point in time. A diffuse concept is therefore of no use to lawyers except if they are ready to give up their core competence.
Finally, functionalist theories do not necessarily deal with issues of legitimacy, but tend to focus on efficiency. Their theoretical foundations do not compel them to call into question how, and by whom, the policy decisions implied in soft law should be made.
3.2. The international community of values: the New Haven school
The New Haven school appears to be a normatively richer alternative within the sociological positivist strand of thinking. It focuses on the choice of policy goals as part of a world social process, whose participants strive for the maximization of values like power, welfare, respect, etc. In this process, law designates those ‘authoritative and controlling decisions’ which serve the realization of the goals thus selected.Footnote 114 The New Haven school does not care much about questions of validity, but about the policy implications of law. This has the advantage that its representatives do not consider soft law as pathological, but rather as a regular component of democratic governance in liberal societies.Footnote 115 However, the lack of concern for validity exposes this view to the same criticism as functionalist theories: it does not allow for a clear distinction between law and politics. Further, from a legitimacy perspective, the New Haven School does not take into account that the values and goals underlying a certain policy might not be universally shared. Rather, the foundation and application of the ‘base values’ have the status of self-evident truths.
3.3. Sovereignty: neo-realism
Within the spectrum of sociological positivist approaches, there is finally one view that denies that international law has the capacity to autonomously influence state behaviour. Jack Goldsmith and Eric Posner base their viewFootnote 116 on historic antecedents like Hobbes and Hegel, who held that there was no public order above the domestic level.Footnote 117 Consequently, they assume that states do not have a preference for complying with international law.Footnote 118 They see the only functions of treaty law, first, in the facilitation of co-ordination among states provided that they have a pre-existing preference for such co-ordination and, second, in the facilitation of co-operation in the situation of a prisoner's dilemma, where international law might signal which moves count as co-operative, while leaving incentives to defect unaffected.Footnote 119 Goldsmith and Posner also deny a moral duty to obey international law. Even if states were legitimate addressees of moral duties, it is submitted that the shortcomings of state consent such as majority decisions or consensus would, in many cases, prevent the emergence of a moral duty to follow a specific international legal rule.Footnote 120
Since it is insignificant for this position whether a rule is binding or not, there is no categorical difference for them between soft law and hard law.Footnote 121 The only advantage of hard treaty law consists of the insights into domestic power struggles that the ratification procedure allows other states parties to take, which might give important clues about the other party's sincerity. Also, ratification enables domestic courts to apply the treaty.Footnote 122 This results in a two-tiered concept of law. Further distinctions are not of interest for this position, since the social impact of legal rules is believed to be limited anyway. This is also the reason why the legitimacy of international law raises no particular concerns in this view.
3.4. Intermediate conclusions
In principle, sociological positivist approaches should be able to conceptualize all sorts of soft legal instruments in a nuanced way, distinguishing them by their social effects. However, no complex categorizations have been proposed yet. Rather, the available proposals either stick the legal positivist distinction between hard and soft law or they give up any attempt to classify instruments at a certain point in time, emphasizing instead the process character of the law. This might result from the inability of sociological approaches to distinguish law from morals or politics. Also, most sociological positivist approaches do not deal with questions of legitimacy. The New Haven school, the only approach that situates its concepts within an elaborate normative framework, does not tackle current challenges to the legitimacy of soft law and is insensitive to the contingency of the base values. It did not take long for this to provoke more critical perspectives.
4. Critical legal studies: discarding positivism
Like sociological positivism, critical legal studies have their roots in American legal realism. However, its adherents criticize the objectivist bias of sociological positivism, which, in their opinion, overlooks the fact that legal rules are the result of power contestations and that the quest for efficiency does not equal the quest for the goals to be pursued with a certain policy.Footnote 123 They also criticize the formalist bias and rule fetishism of legal positivism, which they believe overlooks the contingencies involved in the application and interpretation of law.Footnote 124 In sum, critical legal studies fundamentally questions whether law is a workable instrument for achieving justice in society.
David Kennedy and Martti Koskenniemi are among the main proponents of a critical approach to international law.Footnote 125 In their view, sociological positivism serves only powerful states, which it helps to effectively ‘manage’ international society according to their interests.Footnote 126 The legal positivist paradigm overlooks the fundamental uncertainty of international law and the ‘structural bias’ resulting from it: assuming that legal concepts may have different meanings depending on the context in which one uses them, only the act of application decides their meaning and, consequently, the political preferences of the persons applying them.Footnote 127 International institutions are believed to be structured in a way that makes sure that the interests of powerful states prevail in the application of the law.Footnote 128 In the eyes of the ‘Crits’, international law is therefore rather useless as a means of achieving justice, at least to the extent expected by legal positivists. Law can only make a contribution to justice if one understands it as a specific kind of political discourse. According to Koskenniemi, legal argumentation as a discourse is characterized by a ‘culture of formalism’. It stands for universal values like responsibility, equality, and fairness. This culture of formalism can be used for emancipatory purposes. It allows phrasing a claim in universalistic terms – in terms that cannot be discarded easily by hegemonic strategies, because they require a response in just the same terms.Footnote 129 Of course, this makes the value of legal arguments hinge on the professional ethos of lawyers and their inclination to stick to the rules of legal argumentation.
In this perspective, soft law could not be considered but a threat. Not only does it not require domestic ratification and is therefore exempt from democratic control,Footnote 130 more than that, the very idea of relative normativity constitutes a frontal attack on the culture of formalism. The power of legal argument rests on the very distinction between law and politics. Thus, soft law is considered illegitimate and not worthy of further doctrinal consideration.
5. Post-postmodern reconstructions
As compelling and uncompromising as this critique of soft law might sound, it misses important aspects of soft-law instruments. In particular, it does not always seem to occur to its proponents that soft law might sometimes serve emancipatory purposes. When the choice is not between having a hard or a soft agreement, but between having a soft agreement or no agreement at all, a soft agreement might indeed be valuable and afford some protection to important public interests such as environmental safety or financial stability, or the interests of the Global South, like attempting to establish a New International Economic Order in the 1970s. It would therefore be short-sighted to condemn soft law as the tool of imperialist interests and as the coffin nail of the ethical, emancipatory underpinnings of international law that emerged in the late nineteenth century.Footnote 131 In addition, demonizing soft law seems all the more meaningless, as it is hard to imagine how soft law should be abolished. In international relations, like anywhere else, one cannot unscramble eggs. Therefore, it appears a better strategy to try to tame soft law as much as possible. A number of recent approaches take the points made by the Crits into consideration, but strive towards better legal disciplines for soft law. Still, they disagree about the right concept of international law.
5.1. Neo-formalism
The position that I designate as neo-formalism has been advanced by Jörg Kammerhofer and Jean d'Aspremont. This position has some affinity with traditional legal positivism, but does away with some of its flaws, like the ideal of a coherent system of rules,Footnote 132 or the claims of political neutrality and the strict separation between facts and norms.Footnote 133 Instead, d'Aspremont accepts that law is ultimately grounded in social facts.Footnote 134
In the end, however, this position does not seem to be fundamentally different from modern legal positivism. Even though d'Aspremont recognizes that there has been a ‘pluralization of global law-making’ involving actors other than states, such as non-governmental organizations,Footnote 135 he only recognizes states as ultimate law-makers.Footnote 136 The authors also insist on a strictly binary concept of law identified by formal criteria.Footnote 137 Otherwise, law would cease to function as an autonomous order of commands.Footnote 138 I do not see the reason why one could not imagine a non-binary legal order, namely one that recognizes different types of legal rule, some of which trigger liability in case of their violation, while others may only entail reputational sanctions.
Further, this position does not call into question the legitimizing function of state consent. D'Aspremont argues that legitimacy concerns mainly arise from the fact that some lawyers attribute a law-creating function to non-state actors.Footnote 139 But depriving instruments created by non-state actors or any other type of soft law of the status of international law would not eliminate their actual social effects. This strategy shows probably not enough sensitivity for power – something that a theory branded as postmodern might want to avoid.
5.2. The universality of language: post-positivist approaches
Friedrich Kratochwil and Nicolas Onuf's proposals aim at combining aspects of legal and sociological positivism.Footnote 140 Following the Crits, they understand international law as a specific kind of communication. In contrast to the Crits, however, they do not see legal argumentation just as another mode of interest-oriented realpolitik. Rather, they believe that the specific features of legal communication distinguish law from politics or morality.
Kratochwil tries to avoid the fallacies of legal positivism by focusing on rule application instead of rule making and those of sociological positivism by focusing on regularities of language instead of behaviour. Accordingly, he considers international public law a particular ‘style of reasoning’. Legal reasoning is more precise than political reasoning, defining not only the ends of a policy, but also its means. It is more precise than moral reasoning, independent of the conscience of the applier, and limited by procedural constraints and the necessity to take a decision. This definition leads to some overlap between law, politics, and morality at the margins, but, on the whole, it remains possible to distinguish these adjacent discursive styles.Footnote 141 In light of this approach, soft law may be law in the proper sense of the term, as long as it gives rise to legal reasoning. Kratochwil gives the example of soft law that contains a commitment to continue negotiations. The legal character of a norm does not hinge on its form.Footnote 142
Yet, the problem with this approach is the focus on rule application. If the legal quality of a rule depends on the qualities of the reasoning surrounding its application, and if this reasoning is specific to the institutional context, it is, in the end, the institution applying a rule and not discourse that decides the legal quality of the rule.Footnote 143 Thus, Kratochwil's idea of legal reasoning closely follows the reasoning before courts or tribunals in which two opposing parties litigate in front of a neutral third party with decision-making power.Footnote 144 In contrast, he does not consider the arguments exchanged in bilateral bargaining situations to be legal reasoning, even if the parties argue about the applicability of rules.Footnote 145 These rules count as politics, not as law. Kratochwil thus excludes large parts of soft law from the definition of law, just like large parts of domestic administrative law such as internal guidelines. This exposes Kratochwil's theory to similar criticism as that directed against modern legal positivism. Ultimately, he uses a formal (i.e., institutional) criterion for the distinction between law and politics or morality.Footnote 146 This theory has nothing to offer for the questions raised by the insight that international treaties and soft law can be functionally equivalent, since soft law is usually not applied by courts or tribunals, or at least not directly, only in order to facilitate the interpretation of other norms. It does not take the ‘administrative’ character, the technical nature of many soft-law instruments, into consideration.
The legitimacy of soft law or hard law is no issue for Kratochwil, either. Only a rule's degree of precision, only external patterns of reasoning, not its legitimacy, matter for its character as law. Like sociological positivism, this purely external perspective cannot capture the ongoing debates about the legitimacy of soft law.
Nicholas Onuf tries to overcome the purely external perspective. In order for an internal perspective to be credible, he needs to identify a Grundnorm-like reason for law's validity. Onuf finds it in language, following speech-act theory by Austin and Searle as well as their transformations into theories of society by Jürgen Habermas.Footnote 147 Accordingly, each speech act is characterized by an illocutionary element that designates a claim raised by the speaker. An example would be the postulate that what the speaker utters is a rule to be obeyed by others. The addressee of such a claim has to accept it if the speech act is ‘performatively sufficient’, namely if it fulfils the required conditions of validity. Those conditions follow from the use of language or have been explicitly agreed upon in advance.Footnote 148 In short, Onuf sets out a theory of recognition. In contrast to H. L. A. Hart, recognition has a purely linguistic significance and does not require any form of social acceptance beyond language.Footnote 149 A legal rule can be a legal rule even if there is no social pressure inducing compliance, or rather if the only available sanction is the threat that the acts and speech acts of a person rejecting a legitimate rule will not be understood. Institutions are not necessary for the establishment of rules, but merely stabilize the conditions of validity and ensure that they do not change too rapidly.Footnote 150
This theory allows understanding virtually all kinds of soft law as law in the proper sense, as the legal nature of a norm is merely a function of its linguistic properties, not of its formal character. The concept of ‘conditions of validity’ ensures some level of abstraction and generality to the conceptual framework. The only drawback is that Onuf's theory does not seem to work on closer inspection. It equates moral and legal rules. While rational discourse might produce agreement on abstract moral principles, experience tells us that this does not have to be the case for legal rules. People actually disagree and disobey. While systematic disobedience would leave moral rules intact, the same cannot be said about legal rules.Footnote 151
5.3. Fragmentation and differentiation of global society: systems theory
The fundamental difference between systems theory and critical legal scholarship is that the former sees law and politics as separate, autonomous communicative systems. Unlike legal positivism, however, it does not hold that law has a direct impact on politics. Understanding this separation between law and politics requires a grasp of Niklas Luhmann's view of society. Luhmann takes an extreme view on the grand themes of modernity, namely the differentiation of society and the alienation of the individual. He ultimately dissolves the individual as a sociological entity and discards any philosophy focusing on the subject. Instead, he holds society to be made up of several separate communicative systems. Each system is characterized by a specific binary code that determines which communicative acts belong to the system.Footnote 152 In the case of law, the relevant code is legal/illegal.Footnote 153 Only concrete, specific acts of application form part of the legal system, not abstract legal rules. The validity of legal norms is only a symbol that guides legal communication.Footnote 154 Politics is a different system with a different code. Communication between two systems is erratic and coincidental because they follow different discursive logics. Even ‘structural couplings’, such as constitutional discourse, which has significance for both law and politics, do not ensure that one and the same communicative act is of identical significance for both systems.Footnote 155
Gunther Teubner and Andreas Fischer-Lescano apply Luhmann's systems theory to present international relations in times of global governance. Globalization, they say, leads to a functional differentiation of society, which prevails over the earlier territorial differentiation.Footnote 156 Functionally differentiated systems, like world trade or world finance, tend towards developing legal regimes that are adapted to their specific rationalityFootnote 157 and connected with them through structural couplings.Footnote 158 Soft-law or private-law regimes can be at the core of such regimes.Footnote 159
By virtue of its functional, communicative approach to law, this theory considers all soft law as law, regardless of whether it has any significance for the formation of customary law or the interpretation of treaty law. However, this theory is unable to properly deal with legitimacy concerns. Such concerns presuppose a world-view that puts the subject centre stage. Systems theory may only observe the effects of such debates on the system of law, but it cannot engage in them.Footnote 160 The primary normative concern of Teubner and Fischer-Lescano seems to be the preservation of the functionality of the legal system. In order to ensure some basic level of coherence for it, they suggest the establishment of punctual connections between the various auto-constitutional regimes. Hierarchical or institutional solutions that re-establish the unity of the law are unavailable to this theory. All that can be done is to try to cut losses.Footnote 161
A different approach based on systems theory by Calliess and Renner introduces an institutional threshold into the concept of law, as opposed to morality: law presupposes third-party dispute settlement.Footnote 162 It is not only difficult to see how this view could be compatible with systems theory, since the latter focuses on communication and defies institutional requirements for the definition of a system. It is also exposed to the critique that has been advanced against Kratochwil's theory, namely the limits of the focus on courts.
5.4. Between sovereignty and international community of citizens: recent public-law approaches
Recent scholarship suggests refurbished legal positivist approaches. They aim at a concept of law that is essentially rule-oriented, like legal positivism, but integrates the insights of sociological positivism into the role of soft law on the international plane, takes the postmodern critique seriously, and cares about legitimacy. Because of this latter aspect, I designate them as public-law approaches. One can distinguish two subgroups within this strand of research. The first subgroup suggests a holistic approach and imagines the international order as an international community of shared values. This provides a basis for universal standards by which one can measure the legitimacy of soft law. The second subgroup emphasizes the plurality and diversity among different collectives and their values, and aims at fleshing out institutional frameworks that respect this plurality.
5.4.1. The new jus gentium: Lon Fuller and global administrative law
The first subgroup within the public-law strand stands in the context of the debate about the emergence of a global administrative law. Jutta Brunnée, Stephen Toope, David Dyzenhaus, and meanwhile also Jan Klabbers stand for a rather non-positivistic approach within this subgroup.Footnote 163 They base their view on Lon Fuller's eight criteria of legality, which any rule needs to meet for it to be considered law. Otherwise, the rule is not a legal rule, namely it is not legally binding but only serves political ends (‘managerial direction’).Footnote 164 Under this theory, universal legal principles decide about the validity of rules as legal rules, while the significance of state consent is lower.
Besides the Fuller criteria of legality, Brunnée and Toope argue that international law, in order to be able to influence behaviour, must be based on shared understandings of the matters and problems that are the subject of the regulation, and that there needs to be a community of practice that maintains and further develops those shared understandings.Footnote 165 Dyzenhaus requires an international legal norm to be adopted by a global administrative organization. Additionally, validity requirements may follow from the law of that organization.Footnote 166 While these proposals have the advantage of understanding soft law as law, they apply the same standard of legitimacy to rules of all types, no matter how different they may be. This only works if that standard is very low and characterized by a high degree of generality. And, indeed, Fuller's criteria do not set a high threshold. It was not without reason that H. L. A. Hart considered Fuller's theory as a contribution to effective law-making, but not to the legitimacy of law.Footnote 167 Many of the neuralgic points of global governance that taint its legitimacy, such as overlapping or unclear competencies, lack of procedural fairness, judicial review, or adequate representation or participation, are not addressed by Fuller's theory. By contrast, if the standard of legitimacy is too high, and the requirement of shared understandings appears to be a high standard for disenchanted, functionally differentiated societies, the entire concept collapses. Only very few rules might pass that test, too few for the needs of a complex world.
Klabbers and Kingsbury avoid this problem by proposing more nuanced sets of criteria. Klabbers starts with the presumption that all normative utterances may be presumed to be law if they have received the consent of those involved in the decision-making procedure.Footnote 168 One may rebut this presumption by showing that the rule does not meet Fuller's criteria, or by additional criteria, such as its content (the rule is entirely discretionary and therefore political, not legal), its context (the authors wanted the rule to be non-binding), its origin (the authors lack the necessary authority to make law), procedural issues (e.g., the authors are not representative of the group or community for which they want to legislate), or its topic (e.g., the rule is about trivial matters).Footnote 169
Benedict Kingsbury relies on ‘principles of public law’ (or publicness) instead of Fuller's eight criteria.Footnote 170 His approach constitutes a delicate balance between positivist and non-positivist elements. On the one hand, he chooses a decidedly positivist point of departure, namely H. L. A. Hart's theory of rules of recognition.Footnote 171 He considers the publicness principles as accepted components of the rule of recognition.Footnote 172 Consequently, he chooses an inductive method and extracts five general principles of publicness from the practice of international institutions, namely legality, rationality, proportionality, rule of law, and human rights,Footnote 173 as well as certain more specific procedural principles for global administrative bodies, like review, reason-giving, participation, and transparency.Footnote 174 On the other hand, the reason for the recognition of these principles seems to be non-positivist. They safeguard individual autonomy and ensure that law corresponds to the volonté générale of society, that it is a rule by society for society.Footnote 175 This idea of autonomy seems to be the reason why the publicness principles are supposed to apply to all actors on all levels, private and public, like a universal rule of recognition, recalling the Grotian idea of jus gentium.Footnote 176 And this idea also explains why Kingsbury does not understand every principle as a mandatory component of the rule of recognition in every instance. Depending on how a specific soft-law instrument affects autonomy, it could also be voluntary in the sense that it reinforces the authority of an instrument the more it is respected.Footnote 177
Both proposals not only secure a high level of legitimacy for soft law. They also avoid applying the same standard to all sorts of soft and hard law. The question is not a binary one, namely whether or not a certain soft-law instrument is in conformity with Fullerian criteria or GAL principles, but to what extent it is so, thereby allowing for meaningful differentiation. If a rule does not meet the required level of conformity, it is not legally binding and only of political significance. Nevertheless, the flexibility in the application of these principles and criteria leads to some drawbacks. As Klabbers admits, his set of criteria might raise more questions than it solves.Footnote 178 By amalgamating the concepts of legitimacy and legality, these proposals would produce a vast number of rules of unclear legal status that fail to pass the legitimacy/validity threshold. Also, the content of the principles constituting this threshold might give rise to serious disagreement. By contrast, law in a legal positivist understanding, which Kingsbury explicitly wants to adhere to, is usually associated with formal rationality.Footnote 179 It seems problematic to include such a high degree of uncertainty into the rule of recognition. Certainly, the establishment of a rule of customary international law is often very messy. But this shows exactly the advantage of treaty law or the soft law produced by international organizations over customary law. The approaches suggested by Klabbers and Kingsbury forgo this advantage. Perhaps those proposals should better be understood as theories about the legitimacy of soft law rather than as legal theories in the narrow sense.
In addition, the basis of validity of Klabbers's additional criteria and Kingsbury's publicness principles is not straightforward. Induction from the practice of international and national courts prevails over deductive or dialectical reasoning. It is hard to explain why national and international courts should have the legitimacy to decide such contingent issues. This method recalls the Grotian understanding of natural law that is reflected in the practice of civilized states,Footnote 180 or the idea of a law of reason. It might be difficult to reconcile with a contemporary, pluralistic understanding of autonomy.Footnote 181 More representative institutions and procedures might be better positioned to decide such issues. Therefore, it would be better to consider those principles initially as political proposals, not as ready-to-apply legal rules. Kingsbury's argument that, in choosing to adopt a legal rule, its authors entered into an obligation to respect those principles of publicnessFootnote 182 seems circular to me, because these principles are meant to define what counts as law in the first place.
5.4.2. Taking pluralism seriously: deliberative approaches
The second subgroup of public-law approaches emphasizes the establishment of deliberative mechanisms in order to ensure the legitimacy of soft law.Footnote 183 They believe it advantageous to conceptually separate validity and legitimacy in order to preserve the degree of legal certainty expected from legal positivist approaches. Legitimacy is also considered too controversial, in an international community that is rather heterogeneous and pluralistic, to serve as a criterion for the validity of legal rules.
Samantha Besson advocates such an approach based on Joseph Raz's theory about the authority of law.Footnote 184 According to Raz, a claim to legitimate authority is intrinsic to each legal norm, even though questions of legitimacy do not necessarily affect the validity of law.Footnote 185 This claim to legitimate authority should not be understood as the pretension of a moral duty to obey a particular rule, but as a functional aspect of law that strengthens compliance. Under the ‘normal justification thesis’, a legal norm is legitimate if there are reasons to presume that obeying the law would generally be more rational than relying on one's own, idiosyncratic reasoning.Footnote 186 Besson gives a deliberative edge to the normal justification thesis. A rule of international public law, she argues, presumptively provides rational guidance if it results from a deliberative process involving all groups whose essential interests the rule affects. This is what she calls ‘global demoi-cracy’.Footnote 187
Whether one shares this view or not,Footnote 188 it is an important step towards an idea of legitimacy that is commensurate with pluralistic societies. However, this positive aspect is set off by a binary, source-like conception of international law. Besson does not consider soft law as a qualitatively different form of law, but only as a stage in the development of a ‘real’, binding international legal rule.Footnote 189 Her statement that legal normativity can be relative should therefore only be understood in an evolutionary sense: from the initial proposal to the final, binding legal rule, law passes through several stages at which it might have some indirect legal effects.Footnote 190 This resembles the modern legal positivist stance. By this move, Besson forgoes some of the potential of her approach to tackle the legitimacy issues surrounding soft law. If soft law is not considered law properly speaking, it does not claim legitimate authority and therefore does not need to respect the standards of global deliberative ‘demoi-cracy’.
Klaus Günther suggests another discursive approach. At first sight, his concept of law resembles Luhmann's idea of law as a binary-coded communicative system: even in a fragmented international order, he opines, law is characterized by a ‘universal code of legality’, a meta-language that is characteristic for all legal rules. But, from there, Günther takes a markedly different route that leads him away from Luhmann and in the direction of Habermas. Unlike Luhmann's binary code, he sees the function of the universal code of legality not confined to drawing a distinction between legal and illegal acts. Rather, the universal code of legality is to ensure that legal norms are just. For this purpose, Günther enriches the universal code with a set of normative principles – certain ideas that the use of law as a form of communication evokes, such as fundamental rights, due process, and the possibility of sanctions.Footnote 191 To be ‘legal’, norms must respect these principles. Thus far, Günther's theory strongly resembles Klabbers's and Kingsbury's approaches. But there is a fine, though remarkable, difference. Günther recognizes the fundamentally uncertain and essentially contested character of these principles to a greater extent than do Klabbers and Kingsbury. He therefore does not consider them from the outset as criteria of validity. Initially, their function is only to inform and stimulate debate about the democratic legitimacy of public authority. The specific content of these principles in specific situations still has to be established in deliberative processes. In those discourses, the various actors involved in a particular regime would entrap themselves by and by through the positions they take with respect to these principles. This would eventually concretize the universal code of legality with respect to that regime. This approach seems well tailored for a fragmented, pluralistic international order. Nevertheless, Günther argues that the universal code of legality might induce a process of constitutionalization, namely of concretization of the various principles and their transformation into positive validity criteria.Footnote 192 This, he argues, presupposes the development of a system of individual rights.Footnote 193
Günther's approach indeed takes into account the variety of soft law exhibited by international relations in times of global governance. He also links his concept of law with legitimacy issues, without compromising the conceptual distinction between validity and legitimacy. Nevertheless, it is unclear how non-legal instruments, such as the PISA reports, could enter the picture. The intrinsic link Günther establishes between the concept of law and the application of principles for the legitimacy of power compromises the ability of this theory to go one step further and look at exercises of authority beyond the form of law.
6. Towards a concept of public authority
The public-law approaches seem to offer an adequate reaction to the two challenges outlined at the beginning of this article. In particular, Klaus Günther's proposal devises a sophisticated grasp of soft law and does not only consider it a proto-form of binding law. It is inherently linked with questions of democratic legitimacy in a way that neither compromises legal certainty nor relies on principles of publicness that are reminiscent of natural law. Nevertheless, it is not apt to deal with non-legal instruments, such as pure information, and the legitimacy issues they raise. The root cause of this problem lies in the central role that this approach and other public-law approaches attribute to the concept of law. For them, the only recognized form of authority is legal authority. Any other form of authority is deemed to be legally insignificant. In light of this, it should not be a surprise that, until recently, legal scholars have hardly paid much attention to non-legal instruments. While mechanisms of compliance control such as reporting procedures have received some coverage, mostly with a practitioner's focus, other non-legal instruments, such as indicators, only recently came under scholarly scrutiny.Footnote 194 What seems to be lacking is a uniform approach that allows consideration of both soft-law and non-legal instruments, namely every instrument that does not pertain to the traditional sources of international law.
One should therefore follow in international law the route devised by Foucault for political theory and ‘cut off the head of the king’ in international legal thinking.Footnote 195 Rather than putting the concept of law into the centre, one should think about authority in a broader way that is commensurate with the activities of contemporary international institutions. What is needed is a legal conceptualization of international public authority. The concept of international public authority should be understood as being wide enough to encompass all instruments of international institutions, legal and non-legal, formal and informal, private and public, that affect individual and collective liberty and therefore need to be legitimated by a public-law regime. By contrast, instruments that do not compromise collective liberty, but only the individual liberty of those who consent to them, such as purely private contracts, would not fall under the concept of international public authority.Footnote 196
Because this concept of international public authority will necessarily be very broad and general, it should be complemented by a non-enumerative set of standard instruments, of instrumental ideal types. One could imagine the traditional sources of binding international law as examples of how standard instruments work. They need to be complemented by new standard instruments for typical soft-law and non-legal instruments. These new standard instruments would have the function of linking specific forms of public authority with specific instrumental, procedural, and substantive standards that ensure an adequate level of legitimacy for the type of public authority concerned.Footnote 197 They should be sufficiently general and abstract in order to bring order into the fluid molasses emanating from international institutions, and sufficiently concrete in order to ensure the legitimacy of each specific instrument that falls under them. In the long run, this doctrinal conceptualization might allow striking a balance between the need to adapt to changing circumstances and the values underlying the formal rationality of positive law. The development of proposals for standard instruments is, first of all, a task for legal scholarship. Those proposals might feed discursive processes in international institutions and lead to the adoption of standard instruments in the practice or positive law of international institutions, and eventually perhaps to the formation of customary law. This emphasis on the distinction between non-valid proposals and valid legal rules that enable an internal perspective firmly grounds this approach in the positivist tradition.
Within this framework, the concept of law would play a modified role. It would be discharged from the function of identifying authoritative from non-authoritative acts, but it would continue to serve the identification of authoritative acts that immediately give rise to normative expectations, as opposed to authoritative acts giving rise to cognitive expectations.Footnote 198 Legal standard instruments would function in ways similar to the traditional sources, with the only difference being that they would also comprise soft law.
The concept of international public authority would not necessarily be linked to any specific theory of legitimacy, although the approaches by Samantha Besson and Klaus Günther demonstrate the potential of discourse theory for pluralistic settings. While the standard instruments that international institutions might adopt one day are likely to be a blend of different theories, practices, and interests, it would be useful for the scholarly proposals that initiate their formation to be based on a theory that allows for a plurality of world-views. Discourse theory enables the inclusion of opposing views. It starts from the assumption that most people disagree about almost anything at first sight. This makes it especially apt for dealing with the essentially contested nature of justice in a pluralistic international order.
