It would be truly daunting to try to respond to the many good points made in the above articles – daunting but also perhaps unnecessary. I am struck by the continued critical attention that this work from the ‘over-theoretical 1980s’ continues to receive as a contribution to ‘theory’. To be sure, many abstract discourses are woven into its fabric. But it is still in my mind above all an intervention in the practical world of international law that seeks to understand how to separate good and bad arguments in that field and how that competence might be used to deal with a profoundly unjust world. The articles above rightly stress the many ambivalences in From Apology to Utopia (FATU). I welcome the effort to think beyond the work and to draw new generations of international lawyers and other internationalists to engage critically with today's many injustices. Even as these are not identical with the injustices of the 1980s, the difficulties of dealing with them have not changed very much. On the contrary, it seems to me that the type of analysis employed in that work is still far superior to any rival I am aware of to the extent that its objective would be, as it was in the 1980s, to examine international law's role in the reproduction of the political and economic conditions of the world and to find out what room there might be for progressive transformation.

So I want to use this occasion to spell out as clearly as I can the power of the type of analysis employed in that work that may be called ‘structuralist’. This is a form of analysis that separates phenomena of social life that are immediately visible from others that are usually ‘hidden’ but in some way contribute to producing the former so that once the operation of that ‘hidden’ background is revealed, we feel we ‘understand’ the more familiar phenomena better and are better able to deal with problems they are associated with. Some of the best twentieth century social thought employed the metaphors of ‘surface’ and ‘deep structure’ to describe this kind of analysis and while this may still be appropriate, care must be taken not to draw conclusions that would be too fixed and linear to be credible. The purpose is not to lay out law-like causal invariables to be used for predicting or controlling social events but to understand the limits and possibilities offered by engagement with international law: How does the way the world is organized condition how international legal institutions work? Is it possible to use international law to change those institutions or their distributive outcomes?

Much of the difficulty about dealing with injustice efficiently through international law – or any other expert vocabulary – results from the kind of ‘realism’ that accepts things that are immediately visible because they are somehow unproblematically ‘there’. This feeling of immediate ‘there-ness’ is an effect of specific literary, journalistic, cinematic and other narrative devises, including expert discourses and commercial marketing that underlie everyday debates about politics, the economy and social life. One type of ‘structural’ analysis that arose in the twentieth century aimed to make explicit the rules of production of that ‘there-ness’, the sense in which we end up feeling that something is so ‘true’ that we allow it to determine the way we live. According to this type of analysis, of which FATU is a specimen, learning to know how such ‘truths’ are produced would release us of their power so as to take action in order to deal with problems that otherwise seemed intractable (because they were based on ‘truths’) and allow us to lead in some sense better lives.

There is also an unproblematic international legal ‘there-ness’ that has to do with ‘states’ that ‘exist’ and ‘consent’ and exercise ‘jurisdiction’ over ‘territories’ and ‘peoples’ that have ‘rights’ and so on. Powerful institutions take action and divide resources to give effect to what those truths are supposed to say. The ambition of FATU was to examine how they were produced so as to think critically about whether the conventional ways in which international lawyers used them were really compelling. The metaphors of ‘surface’ and ‘deep-structure’ were expressed there in terms of competent international legal ‘speech’ and its underlying ‘grammar’. The point was to make that ‘grammar’ explicit so as to question the self-evident normative power of authoritative international legal speech. How was ‘consent’ constructed? What did ‘sovereignty’ entail, and whose sovereignty would be decisive? Could ‘autonomy’ and ‘interdependence’ both be true? Hence the very extensive use of materials from international courts and doctrinal debates. The analysis showed that the ‘grammar’ actually allowed the production of many different kinds of reality so that how legal institutions decided, what interests or values were preferred was not a result of the legal ‘grammar’ (although the grammar dictated how the outcome was to be presented) but of something that could be called the ‘structural bias’ of the institution. By shifting attention from the argumentative grammar to the structural bias it was hoped that the limits and possibilities of using international law to transform unjust international relationships would become clearer. The routines of the international legal world would no longer have the feel of natural necessity but would come to seem problematic and contestable. And lawyers could use the law's resources to help change those routines and thereby the world.

The structuralism of FATU was designed to destabilize the sense that the immediate aspects of international law were true and fixed and action-determining. They were produced, instead, by techniques and arguments that were quite contested and even contradictory. Revealing this did not make the world any less unjust. But it seemed a useful and perhaps a necessary preliminary for using the resources of international law for supporting efforts to change them. I still regard this a useful ambition. Therefore, instead of using the space allotted to me for responding to the above articles, I will argue for the continued power of the type of analysis employed in FATU.

There is no limit to the times I have sat down with doctoral students desperate over the difficulty of producing a set of rules – new rules, better rules – on their chosen research topic, or coming up with a novel policy proposal – a new proposal, better proposal – than all the proposals contained in prior research. The despair is understandable. The research started out with a sense that the present world, understood as a world of rules or policies, was in some way flawed and that this flaw was so important that it merited spending countless hours, ultimately years of study to figure out what was wrong and how to make things better. And yet, after all that research, the rule or the policy the student finally proposed as the ‘result’ of that anxious study seemed terribly weak and vulnerable to all kinds of objections from alternative rules or policies. What point was there in all those years of study if it did not produce a reasonably impressive statement de lege ferenda? At the end the student was well aware that other researchers would be able – and perhaps likely – to draw from the same archive quite different conclusions, rules and policy-recommendations. Was it all just a waste of time (apart from the fact that having that doctoral degree might still help in furthering the candidate's career)?

One thing that unites the above articles is their shared sense that the approach to international legal research suggested by FATU, as well as David Kennedy's International Legal Structures came to the archive of legal materials from a very different angle. They did not propose new and better rules or policies but were stunningly inadequate from the perspective of the kind of ideals that seemed to be ruling the field. They did not try to ‘help’ decision-makers by proposing better interpretations, rules or policies to manage this or that problem. They did not imagine research as a simulacrum of a trial: the handing out of verdicts about this or that type of action. Nor did they imagine themselves as part of some great project of resolving problems of global governance. No doubt, some regarded them as wholly irresponsible specimens of academic intellectualism, a kind of wordplay, unconcerned about the reality of the world's problems and the pressing need to devote the diminishing funds for academic research to their resolution.

And yet, some of the more satisfying responses to these works have come from practicing lawyers, mid-level UN officials, experts in environmental or human rights law, people concerned on a daily basis with the ‘resolution’ of global problems. Precisely owing to their long practical involvement, such people acutely feel the difficulty of finding good solutions to the problems they are requested to deal with. For such people, it may be a relief to notice that whatever else these works might represent, at least they do not participate in the hubris of believing that ‘real world problems’ could be resolved by a few years of sitting in libraries reading stuff other people sitting in libraries have written. As David Kennedy has put it, global governance is a ‘mystery’. I agree, but would like to extend this from expert rule to the global academy. In all the fields of social science I am aware of, even the very fundaments (especially the fundaments) are the object of intense intellectual controversy. Hegemonic positions are challenged by counter-hegemonic ones, orthodoxy meets heterodoxy. For every theory, there is a counter-theory. With what confidence can the academic world believe that the latest rule or policy it has just exgurgitated is finally able to resolve some problem that the leading lights of the field have been preoccupied with since years? The best practitioners often resemble the doctoral students I mentioned. They, too, feel that the rules of their trade are quite weak and often unhelpful, that the countless policy-proposals landing on their desk are really not that impressive when examined more closely. More or less explicitly, they, too, believe that we know actually very little about the world, and that it is therefore frustrating, not to say revolting, to constantly have to deal with research speaking in voice of high self-confidence about rules or policies which, when put to practice, are likely to show themselves no better or worse than past rules and policies when they were first being excitedly applied.

To the extent that FATU, International Legal Structures and other such works have been received positively by the men and women working ‘in practice’, I believe this has been because they do not carry the pretence that what is needed are more rules or policies but a better understanding of what goes on under the façade of rule-application or policy-implementation, especially in places conventionally thought of as ‘legal institutions’. To make this point in the way I have made it to many students with whom I have had this conversation, instead of asking yourself ‘what would be a better rule or policy’, try to answer the question ‘what does it take to believe that this rule or this policy is the better one’. In FATU, I called this ‘regressive analysis’. Instead of trying to find out what we should do now, I took a step backwards in order to examine what people who make these suggestions (of better rules or policies) think about the world, and formalize that ‘thinking about the world’ in a series of propositions that would then account for the ‘structure’ of standard international law (both theory and practice) in the field of research. In the language of structuralism I used in From Apology, the idea would be to descend from the level of legal speech (parole), the routine production of better rules and policies to the language, the conditions within which such rules or policies may come to seem ‘better’, that is to say more just, more coherent, more valid, more efficient and so on.Footnote ¹ The result would not be new rules or policies but an account of rule and policy production that would make explicit the conditions under which specific proposals seem professionally plausible and are accordingly either approved or rejected.

I think those who have liked this sort of analysis have liked it owing to the way it takes seriously the difficulty of legal work and tries to give an explanation to the experience that even the best rules and policies often turn sour, that legal practice and writing are much more improvisation and groping in the dark than the ‘application’ of straight-forward directives. It is not that lawyers fail in what they are doing but that the legal ‘field’ is so constructed that its abstract rules and principles always prove somehow insufficient and instead point the lawyer to the ‘structural bias’ of the institutional context, its inherited thumb-rules, preferences and ways of operation. The experience is that much of the talk of ‘justice’, ‘validity’, ‘coherence’, ‘efficiency’ and other such official criteria is just that – ‘talk’ – and that any worthwhile scrutiny of the operation of official rules and policies will show that they can be bent this way and that way. Although they are necessary parts of the justification game the institution is expected to play, they are really unhelpful as an account of how lawyers in the end choose. It is those choices that critical analysis hopes to make express, especially as they crystallize in stable patterns: which interests or actors do they support? In this sense, I think it is right to write, as Jean d'Aspremont does, that structural analysis helps lawyers reflect on what they already feel as the complexity of their practice and the fragility of the outcomes it helps to produce.Footnote ²

My hope is that understanding legal work in its complexity – the ‘mystery’ of it – might also be helpful for the anguished doctoral student who would no longer have to think that he or she will have to pretend to know how to carry out those practices better than those who have been out there for years. There is no need to think that lawyers ‘out there’ are either incompetent or evil because their work does not fit the official criteria offered by the field to assess how it is doing. Akbar Rasulov is right that FATU was written to ‘eradicat[e] the gap between the discourse of academic theory and the discourse of practical knowledge’Footnote ³ and to do this by giving an academic form to practitioner experience – especially the experience of indeterminacy, namely that the field's official criteria (rules, policies) both justify and enable the critique of established practices. This is not to say those criteria – or the international law they represent – are useless. They keep the conversation going, account for reproducing the institutional practices of the field, create mainstream and deviant approaches, and point to constantly new features of the world that could be thought of as ‘problems’ for which the field could then search ‘solutions’. Akbar speaks of legal concepts as ‘crystalline-style structure of ordered antagonisms’.Footnote ⁴ That is very well said. They offer the interpretative prism that legal work uses to produce its images of the ‘true’ and ‘right’. But the legal criteria and concepts are not action-determining so that it would suffice to pay attention to them, only, instead of to ‘choices’ influential men and women make by using them, to explain the way the world is.

For sceptics who believe that this kind of a view leads to an ‘anything goes’ attitude, I continue to insist on precisely the contrary. The possibilities in the real world of legal work are always limited. Only some arguments go, others do not. Only some rules or rule-applications, some policies are accepted, other are not (in the relevant institutions). The very point of a regressive (structural) analysis such as FATU is to lay out the conditions within which important rules and policies emerge from the mass of legal materials available and within which they come to be interpreted and applied in certain ways instead of other prima facie plausible ways. Let's return for a moment to the proposed research question: ‘What does one need to believe in order to think that rule X instead Y should be applied, and that it should be applied in the way Z instead of W’? Two things about that formulation are noteworthy. First, it rejects the assumption that determining rules or policies emerge from the mass of legal materials somehow automatically or out of their own intrinsic force. They are chosen. Moreover, they are chosen by leading participants in influential and easily identifiable institutions. There is no mystery about the process. Secondly it points to the need to understand how such choices reflect the background conditions within which some choices seem ‘good’ and others ‘bad’, some policies appear plausible while others seem implausible. Owing to its indeterminacy, the legal ‘grammar’ itself does not generate those choices – though it helps to justify them. Instead, the choices reflect background conditions about the epistemic, economic, ideological, psychological and other such ‘truths’ and self-evidences that institutions have come to accept in more or less unthinking terms. It is those background conditions that are revealed by the answer to the question ‘what should one believe in order to think that a rule or policy “X” should be chosen and interpreted in the way “Y”’. Far from assuming that ‘anything goes’, all such analysis is underlain by the firm conviction that such choices are not random, that jurists and professional men and women act in predictable ways and that this fact is one key to their very professionalism. The task of legal research would be to understand legal professionalism not just by examining what institutions say but what makes them choose from equally plausible alternatives the ones they do, and draw from them the conclusions they draw.

The answer to the question: ‘What one should believe in order to think of this rule (policy/interpretation) good (valid/legitimate)’ provides an understanding of the field that is the object of the study. That understanding is in itself disinterested to the extent that it merely describes what people do, how they choose, and on what bases those choices are made. In a sense, it simply makes express what competent lawyers in a field always already know – even if they have not carried out formal research and do not have the vocabulary to express what they know. That vocabulary is offered by the kind of structuralism displayed in FATU. But the research offers more than description. It becomes ‘critical’ of an institutional practice in demonstrating that it is not exclusively captured by the official rules and policies the institution claims to follow. This is so owing to the indeterminacy of those rules and policies, the fact that they have to be interpreted and there are always many alternatives for such interpretation – many ways to choose. Probably the most famous aspect of FATU is precisely its demonstration of the endlessness of interpretation and its basis not in the ambiguity of words (as standard jurisprudence often suggests), but in the openness of the world itself. Because we experience the world in contrasting ways, we project different meanings on the words we use to describe it. The question ‘what one should believe in order to think this interpretation better than that’ aims to bring to the surface that underlying world of beliefs that controls our institutional practices, and accounts for the way decisions are made and resources are distributed. The analysis leads into a contrast between the official justification ‘we apply rule/policy X’ and the results of the research that says ‘but rule/policy X can be applied in many different ways, and you have chosen to apply it in the way Y because you believe that Z’. This ‘Z’ may now be associated with the underlying beliefs that accounts for the structural bias of the institution. In other words, research of this type is (immanently) critical because it demonstrates that the official justifications that legal institutions give of their activity (‘oh we just apply rules / policies / interpretations’) are insufficient for understanding those institutions. And their insufficiency reflects the indeterminacy of their underlying belief systems. An example of this kind of analysis is also provided by Justin Desautels-Stein's above discussion of how ‘rational choice’ falls within ‘classical legal thought’ (what I have called ‘liberal theory of politics’) while simultaneously undermining it (and the liberalism on which it stands).Footnote ⁵ Uncovering the hidden beliefs – the structural bias –provides a critical understanding of the institution, uncovering aspects of its work that are rarely expressed – for example, ‘racism’ and Desautels-Stein's discussion of an apparently ‘rational’ reliance on the ordering force of the ‘Great Powers’.Footnote ⁶ To the extent that those beliefs are contestable, the contested institution may then be critiqued as having been instrumentalized for the values or interests that such beliefs about the world sustain.

The research as proposed so far has produced a description of a legal practice and proposed an ‘understanding’ of it by uncovering the normally unarticulated belief-system that dominates that field of practice. It has shown how from a mass of rules, policies and interpretations some are chosen owing to the kinds of beliefs that control professional activity. And it has produced a potential critique of that activity by showing that depending on what one believes true of the world in general, ‘it could always have gone the other way’. But it has not proposed a ‘better’ rule or policy. Is that not frustrating? Doctoral students often begin their work by hoping to correct some injustice or ‘wrong’: the victims do not get justice, the environment is spoiled, asylum seekers are treated unfairly, trade perpetuates inequality, human rights are breached, post-conflict governance supports neo-colonialism. Critical structuralism such as that offered by FATU does not provide a ‘theory of justice’ that would finally validate such intuitions. Instead, it enables their critical examination in the course of the research and suggests ways in which these intuitions might be turned into practical ‘choices’ in legal work. What this might mean is easiest to see if it is contracted with a common experience in problem-solving research. The researcher is invited to take a problem that some institution regards as important, scrutinize past ways of dealing with it, and then propose a new rule or policy to resolve it. This is a receipt for frustration. Once research has been completed, and the researcher has received an intimate understanding of the problem, its resolution will inevitably seem very difficult indeed. Sometimes, perhaps often, acquaintance with the complexities may have led to the dilution of the original intuition about the ‘wrong’. The rule or policy is proposed (because this is required) without much faith in it. Why should it be any better than a contrasting rule or policy that is being proposed perhaps at the very same time in some other research institution? The research works only as initiation to a practical sensibility – ‘oh it is really hard’ – which is that of a competent lawyer, and ‘competent lawyers’ have modest ambitions. The sense of ‘wrong’ with which research once began is diluted into blasé professionalism.

Critical research tries to avoid this common predicament. Instead of guiding the researcher to produce solutions to problems given from the outside, it examines the way some things emerge as ‘problems’ worthy of the time of researchers and the resources of legal and academic institutions. Instead of merely teaching students to produce new rules and policies within the field's well-established discourses and biases (though it does this, too), it is interested in the field itself as part of the problem. If the legal vocabulary really is so open-ended and contradictory, why do legal institutions nevertheless regularly end up supporting the same actors or interests? How come the law itself is always implicated in one way or another in perpetuating the problems it seeks to deal with? What really is the relationship between, say, humanitarian law and the justification of killing, environmental law and the persistence of massive pollution, trade and investment law and the increase of massive inequality in the world? What is the relationship between the rule of law, development and postcolonial relations of dependence and clientelism? By uncovering the ‘deep structure’ of assumptions about the world that control the activities of influential legal institutions – or indeed makes some institutions ‘influential’ and keeps others marginal – critical research contributes to the understanding of the way the world operates, including by producing the ‘wrongs’ that motivated the research from the scratch. Sahib Singh in his article pointed to the importance of the question of the legal ‘subject’.Footnote ⁷ Structural research of the kind displayed in FATU tries to keep alive the political intuitions of the researcher by demonstrating that there really is no safe ground of ‘mere professionalism’ where attitudes of blasé neutrality would be appropriate. On the other hand, by making express the rules that provide for legal competence, such research seeks to empower the critical researcher to operate in actually existing institutions in potentially influential ways, aware of the structural constraints but also of the malleability, gaps and loopholes of their official rhetoric. This is a world of compromises, and as legal institutions are no doubt amenable to change, so are the lawyers – critical or not – working there. FATU ended with a discussion of the dangers of routine and the possible frustrations of trying to change the contexts in which routines are formed.

I cannot go into the strategies of power and survival that must accompany the critical spirit when it turns away from the safety of the academy to work in the legal field. Instead I want to end by stressing the need to support the intuition about the ‘wrong’ or the injustice that triggers best research and professional practice. John Haskell warns about what he regards as the over-emphasis in FATU of linguistic analysis.Footnote ⁸ It is too cool, too distant. While I think that analysis of the law's operating structures can only be predominantly linguistic, I agree that maintaining a sense of the ‘wrong’ that motivates best academic and practical work requires constant engagement with the politics, economy, sociology and history of the international world. The formation of the critical sensibility, the road that leads from an unanalyzed malaise about the state of the international world – aroused perhaps by daily reporting of crises – to engagement needs to rely on some larger frame of thinking about the way in which the world has organized itself and is being governed, perhaps a historical sociology, as Haskell suggests. The discourse may be random, but the world surely is not. And to find the frame that controls the routines, such as the daily production of international law at international institutions, the intuition of the ‘wrong’ is immensely more helpful than any technical skill of the jurist.