Otago Law Review
For all the deep injustices perpetuated by law, there remained a real sense in which the techniques and procedures of law remained independent from the gross manipulation of the executive and in which justice was seen to be done. No account of these years would be accurate if it were not accepted that justice was done and seen to be done in some cases. In this way, principles and values central to the rule of law and a just legal system were not entirely lost. ... The maintenance of such values during the years of apartheid facilitated the transition to a constitutional democracy and provided an important foundation for the legal system in that democracy.
Written Submission to the Legal Hearing of the Truth And Reconciliation Commission by A Chaskalson (President of the Constitutional Court), I Mahomed (Chief Justice of the Supreme Court of Appeal), P Langa (Deputy President of the Constitutional Court) HJO van Heerden (Deputy Chief Justice of the Supreme Court of Appeal), and M Corbett (former Chief Justice).
As we seek order, we can meaningfully remind ourselves that order itself will do us no good unless it is good for something. As we seek to make our order good, we can remind ourselves that justice is impossible without order, and that we must not lose order itself in the attempt to make it good.
Lon L. Fuller, ÒPositivism and Fidelity to Law – A Reply to Professor HartÓ, (1958) 71 Harvard Law Review 630, at 657.
One of the fastest growing areas of academic study in the Humanities, Social Sciences and Law is the set of inquiries now catalogued under the heading “Transitional Justice”. A rough list includes: How does a society produce a stable and just political order out of a state of instability and injustice? Is it better for a society composed of groups which have done terrible things to each other in the past to confront as fully as possible past atrocities or to suppress the memory of the atrocities and get on with the job of living together? If the past is confronted, should the confrontation be a formal one in the court room or are other means of confrontation more effective?
In addition, there is a host of quite practical problems that attend any answer to one of these questions. For example, if one argues for criminal trials as the means to confront past atrocity, how does one cope with the fact that often those with the expertise to organize and prosecute such trials will not only come from the group who held power in the old regime, but will have been directly involved in enforcing its oppressive law. And following on from that problem, there is the rule of law issue about whether from the perspective of the present one can legitimately prosecute acts which were legal or arguably legal under the law of the old regime.
However, as can be seen from this list, these topics are as old as academic inquiry itself. So what is novel is not the inquiries, but the idea that together they make up parts of one compendious inquiry into something called transitional justice. The causes of transitional justice becoming a topic in its own right are clear. Within a relatively short period, there occurred the break up of the Soviet Union and Yugoslavia and the changes in political regime of South Africa and countries in Central America. But, as is usually the case when a new field of inquiry is opened up, its title is not considered by its practitioners to be a label of convenience. Rather, the field is constituted by assumptions that shape the various discrete inquiries undertaken under that title, which is what underpins any claim to novelty.
Here I want to contest that claim by casting doubt on the constitutive assumptions of transitional justice. My argument is very narrowly focused, as it deals with one transitional situation – South Africa – and for the most part with one problem in that transition – the role of independent judges in upholding the rule of law. But I want to suggest that the main point of the argument is capable of general application – that the problems addressed under the label “transitional justice” are best addressed when one sees that that they are not exotic and thus not deserving of their own academic discipline. Rather, they are more dramatic manifestations of problems faced by all stable societies.
The idea that there is a discrete subject of academic inquiry called “transitional justice” is underpinned by at least three assumptions. First, there is an assumption that the nature of the problem one encounters in the period between the collapse of an authoritarian regime and the establishment of a democratic society is a problem about achieving justice. It is not, that is, a problem about achieving a transfer of power from one elite to another in the most orderly fashion possible.
Second, there is an assumption that there is something special about the justice that one achieves in managing a successful transition from authoritarianism to democracy. The most common claim in this respect comes out of the experience of South Africa’s Truth and Reconciliation Commission (TRC), and is found in its Final Report, that the mechanisms of the TRC achieved “restorative justice” — the kind of justice most appropriate to transitions.
A third assumption is that transitions are radically discontinuous with both the past and the future, discontinuous with the past because the transition is a radical break and discontinuous with the future because, at the end of the transition, one has in place the institutions that mark its end and thus issue in the new era. This assumption clearly supports the second, because if transitions are exceptional because of radical discontinuity, it might seem to follow that if any conception of justice can inform the transitional process, that conception will be different from our ordinary conception.
I am quite sceptical about the claims made on behalf of restorative justice, and I will later sketch a much more modest and I think more plausible account of what that conception seeks to achieve. My scepticism in this regard extends to a more general question about whether there is anything special about transitions. The very idea that there is something special to be studied seems to imply that the contrast class for transitional societies is one of societies that are not in transition — static societies. So where I want to differ from the transitional justice model is not in regard to its first assumption — that the problem of transitions should be viewed as one about justice and not merely about orderliness — but in regard to the second and third assumptions. My view is that the study of transitions can shed light on justice, but the light it sheds will help in better understanding our common or garden variety of justice, rather than in revealing something entirely, or even largely new. However, my point of departure is not this more general debate, but rather the role of judges during transitions and, more particularly, the topic of the relationship between judicial independence and the rule of law.
Transitions raise a curious, perhaps even paradoxical, problem for judges. This problem seems not to arise for the politicians who oversee the transition, because their history of opposition to the old regime is what qualifies them for the task of leading the new. The new politicians’ authority might seem then charismatic, an authority in virtue of their past — sometimes heroic — deeds. But the authority of judges is commonly thought to be something impersonal — it flows from their independence from politics.
Three factors complicate the judges’ task of achieving the independence on which their authority rests. First, it is likely that many of the judges will be judges who served the old regime and so are thought to be deeply compromised from the start. Second, many of the newly appointed judges are likely to be appointed for the same reasons that qualify the new political leaders — their record of opposition as “political” or “human rights” lawyers. But then they seem no less compromised in terms of impersonal authority than the judges from the old order. Third, judges in a transition are often faced with deciding deeply political matters, which force the issue of their authority and their independence to the surface. For example, they might have to decide on the legality of the transitional regime’s attempts to use the law to deal with the wrongs of the past, which are arguably not legal wrongs according to the laws of the past. These wrongs might even seem to have amounted to legal duties, as in the case of the East German border guards who shot and killed people attempting to flee to the West. Moreover, the judges often will be faced with interpreting a new constitutional document that is one of the markers of contemporary transitions. They will do this in a society unaccustomed to a culture of constitutional judicial review, so that they might have to consider striking down as invalid a statute of the very popular new regime.
Since transitional judges are doubly politicised — both by appointment and task — it seems that they cannot claim the mantle of impersonal authority, but somehow have to create the conditions under which such a claim can be made. Put differently, since they have a role in putting in place the conditions for their independence, they are essentially engaged in a bootstrapping exercise — pulling themselves up by their bootstraps into the place they had to occupy before they could get started.
This problem is only one manifestation, though a very important one, of the allegedly paradoxical role of law in transitions. As Ruti Teitel describes it, “[i]n its ordinary social function, law provides order and stability, but in extraordinary periods of political upheaval, law maintains order even as it enables transformation. Accordingly, in transition, the ordinary intuitions and predicates about law simply do not apply. In dynamic periods of political flux, legal responses generate a sui generis paradigm of transformative law.”
However, just as I have suggested that no special conception of justice is required in transitions, so I want to argue that law plays no special role — or at least that, by and large, the “ordinary intuitions and predicates about law” suffice. We just have to get right the content of those intuitions and predicates. And, once we have done that, we will see that paradox largely, perhaps wholly, disappears.
In the next section, I set out the basis of my argument — an account of some issues about judicial independence which arose during South Africa’s transition. I will then try to draw out the implications of that account for debates in legal theory about the rule of law. Finally, I will show how the version of the rule of law which I will argue emerges also provides the appropriate conception of justice for transitional societies. But since that version is also appropriate for ordinary stable societies, there is no special kind of transitional justice. Indeed, the account of the kind of justice that is the best candidate for the account of transitional justice is by and large no more than an account of the justice of the law — the morality of the rule of law.
The first epigraph to my paper comes from a written submission by five of South Africa’s most eminent judges to the “Legal Hearing” held by the TRC in October of 1997. While the TRC is best known for its inquiry into and report on the “gross human rights violations” of the apartheid era, its statutory mandate was to establish “as complete a picture as possible of the causes, nature and extent of the gross violations of human rights”. In fulfillment of that mandate, the TRC went beyond the hearings in which perpetrators of gross human rights abuses sought amnesty from criminal and civil liability for their deeds and the hearings in which victims and their relatives testified. For the TRC held several professional and institutional hearings which were inquiries into how professions and institutions, which on the face of it seemed no different from their counterparts in Europe or North America, were deeply implicated in apartheid.
Archbishop Desmond Tutu, Chair of the TRC, said in his opening address to the Legal Hearing that it was the “most important of the professional hearings”, almost as important as the “victim/survivor hearings”. He told of what it was like to grow up as a black child in a world of daily humiliation, not only of oneself but, inevitably more painful, of one’s parents. That humiliation, he pointed out, was enshrined in the law of the land, laws whose violation demanded the sanctions of the criminal law, branding as criminals people attempting to exercise basic human rights. However, Tutu did not think that all that the Hearing could reveal was the way in which law had been used to oppress; he suggested that it was just as much the task of the Hearing to inquire into the way in which law was and could have been used to resist oppression. Indeed, his reason for according the Legal Hearing special importance was his view that the law is usually thought of as the instrument of justice yet, during the apartheid era, it was used most often as the instrument of injustice.
Tutu made this point when he dealt with the fact that not one of South Africa’s judges accepted the TRC’s invitation to attend the Hearing, though several made written submissions. The judges claimed that to submit themselves to an official inquiry into judicial conduct would compromise their independence. Many were also concerned about the possibility that judicial participation in such debates would rupture the fragile collegial relationships which existed between the judges of the old order, who — like all state employees — had kept their jobs under the terms of the transitional negotiations, and the new order judges — often appointments from the small group of lawyers who had devoted their careers to the struggle for human rights.
Tutu excoriated the judges. They had, he said, been faced with moral choices under apartheid and generally they had made the wrong ones. In the transition they had been faced with another choice — whether to appear before the TRC — and again they had made the wrong choice. This showed, he said, that they “had not yet changed a mindset that properly belongs to the old dispensation ...”.
Tutu’s first claim that judges had made the wrong moral choice under apartheid is not the claim that judges should have refrained from taking office because of the moral evil of the law they had to enforce, nor that they should have resigned in protest at any particular time. If that were his claim, then their failure would have been the same as that of any white South African who chose by and large to participate in and benefit from apartheid rather than to resist it. Nor is his claim that they should have lied about what the law required in order to mitigate the injustice of apartheid.
Rather, Tutu’s claim is that the judges failed as judges. They failed to make a moral choice available within the law. And that choice would have made a difference. It would have made the law something more than a mere instrument of the policy of apartheid by bringing the law more into line with an ideal of justice. And we should note that South African judges swore an oath of office which required them to “administer justice to all persons alike without fear, favour or prejudice, and, as the circumstances of any particular case may require, in accordance with the law and customs of the Republic of South Africa.” So one can put Tutu’s claim in the following way: the judges were in dereliction of the duty they undertook when they assumed office.
Now during apartheid judges had the formal guarantees of independence — life tenure, salary, administrative autonomy — that judges in the United States of America, Canada, the United Kingdom, New Zealand or Australia had. It is in seeing why it was the case that apartheid-era judges for the most part lacked independence even though they had its formal trappings that we see that judicial independence is also a kind of dependence; it depends on something positive — the judicial pursuit of the justice of the law. One has to ask not only what judges have to be shielded from in order to be independent, but what we want them to be independent for.
Tutu’s second claim — that the judges had also failed to make the right moral choice in refusing to attend the Hearing — might seem to follow naturally from his first. If the virtue of judicial independence is that it protects judges so that they can carry out their duty to do justice within the law, and they are in dereliction of their duty, they should not in principle be able to claim their independence as a bar to appearance before an official body. As a (dissenting) Canadian judge once put it when judges in Canada raised a defence of total immunity to a summons to answer questions before a commission of inquiry about their role in an injustice, “[w]hen there is a real risk that judicial immunity may be perceived by the public as being advanced for the protection of the judiciary rather than for the protection of the justice system, the public interest ... requires that the question be asked and answered.” And in regard to the judges’ thought that collegiality would be compromised if some of their number appeared at the Hearing, one has to take into account the possibility that the kind of collegiality bought at the expense of an open and honest debate about the substance of judicial independence might be a very shallow one, unlikely to sustain a judiciary which carries the burden of large expectations.
However, there is another factor that has to be taken into account before Tutu’s second claim can be said to follow from the first. And here it is important to consider a speech by a man who has become the most important jurist of the new South African legal order — Arthur Chaskalson, the first President of South Africa’s Constitutional Court.
From the early 1960s, Chaskalson had played a role in representing accused in political trials which made him one of the two leading “political” or human rights advocates at the South African Bar. Chaskalson went on to became the first director of the Legal Resources Centre, which from 1979 moved political lawyering from a defensive and reactive stance to active challenges to apartheid legislation.
In a speech in 1989 — “Law in a Changing Society” — to mark the tenth anniversary of the Centre, Chaskalson contemplated, though not with great confidence, the possibility of the dismantling of apartheid and the creation of a democratic society. Although he could not predict the end of apartheid, he thought it important to set out a vision of the positive role law could play in establishing a democratic order.
Chaskalson had to contend with the fact that law had been used as the instrument of apartheid. Unlike the new South Africa where the Final Constitution explicitly requires that statutes be consistent with constitutional values, the Parliament of apartheid South Africa was supreme — unlimited by any constraints on its legislative power other than constraints to do with the “manner and form” of its legislation.
Chaskalson said that fact this created an “almost schizophrenic approach by courts” to interpretation of apartheid law. The common law heritage of the judges required them to interpret statute law, in so far as this was possible, in the light of principles developed by judges in their decisions which “deny all forms of discrimination” and which seek “to protect fundamental rights and freedoms”: “They were at one and the same time being asked to articulate and give effect to equitable common law principles, and to uphold and enforce discriminatory laws: at one and the same time to be an instrument of justice and at another to be an instrument of oppression”.
Here Chaskalson is referring to the fact that the South African legal order is part of the family of common law legal orders. Thus, embedded in the South African common law are judicially crafted presumptions, which have their roots both in the English common law tradition and in Roman-Dutch sources. Examples are: the presumptions that statutes should be interpreted so as to give maximum effect to individual liberty and to a principle of equality of all individuals before the law; the presumption that anyone whose rights or interests will be affected by an official’s decision taken on the basis of a discretion afforded by statute has the right to be heard before the decision is made; the presumption that individuals who have been deprived of their liberty have the right to seek legal advice and to contest their detention before a court in order to establish its legality.
However, Chaskalson’s claim that judges were under a duty to resort to common law presumptions in a legal order where Parliament is supreme and under the control of a government determined to use the law to implement its racist ideology is not unproblematic. Even when statutes are not clear, so that one can make a case that there is an “ambiguity” in the statutory language which common law presumptions or principles might clarify, judges have to be predisposed towards regarding those principles as having very strong presumptive force before they will have any effect. Judges have to adopt a view of the rule of law which requires judges whenever possible to interpret statute law so as to make it consistent with fundamental principles of the common law, because it is those principles that make law morally legitimate.
In short, Chaskalson’s common law understanding of the rule of law is controversial because many judges are hostile to the idea that their moral sensibilities should have any impact on the process of interpretation. As they understand it, their duty as judges is to interpret the law as it was in fact intended by the legislators to be interpreted. Such judges suppose themselves to be legally required to decide a controversial case of statutory interpretation in accordance with what they take to be facts about the legislative intentions that explain the statute, whatever their personal moral views about the statute. And this “plain fact” understanding of judicial duty is ultimately based in a political theory about the relationship between the judiciary and the legislature. That theory says that judges act appropriately when they defer to a particular understanding of what makes up facts of the matter about legislative intention. Thus, they should adopt interpretative tests which seek to find such facts before resorting to other sources of legal authority, such as the common law.
The political justification for the plain fact approach is at one level about the importance of parliamentary supremacy and the doctrine of the separation of powers. At a deeper level, it is about the way in which parliamentary supremacy is the best institutional expression of democracy; and in South Africa that view held sway, though with the qualification that democracy should be reserved to the elite capable of responsibly exercising the franchise, that is, the white population. The theory is not, then, politically neutral or wholly independent of judges’ moral sensibilities since judges who adopt it will think that it is morally and politically superior to its competitors. Rather, the kind of political theory it is precludes judges’ moral sensibilities from playing a role in interpretation.
Chaskalson, then, in depicting the dilemma of South African judges caught between the “facts” about the intention of apartheid statutes and the pull of common law principles neglects that class of judges who did not find themselves in that dilemma. And they did not find themselves in that dilemma because their understanding of their duty confined them to what they took to be facts of the matter about legislative intention. And this point requires some clarification of my initial interpretation of Tutu’s assertion that judges had made the wrong moral choice under apartheid. More precisely described, the moral choice that faced judges was not so much one within the law as about the law, about what conception of the law and of the rule of law one should adopt.
Chaskalson, of course, was well aware that most South African judges had, in Tutu’s words, made the wrong moral choice. He said that in “reviewing the history of the courts of this country some writers have criticized the way in which South African judges have discharged their duties over the years”. But he went on:
That they could have done better than they did is I think now clear. But that is true of all of us, and little is to be gained by lamenting the past. What is important is the future, and it is here I believe that we will come to appreciate that we owe much to our judges, and a great deal to some of them. For despite all the paradoxes they have somehow held to the infrastructure and have kept alive the principles of freedom and justice which permeate the common law. True, at times no more than lip service has been paid to these principles, and there have been landmark cases where opportunities to give substance to and uphold fundamental rights have been allowed to pass without even an expression of discomfort, let alone a vindication of the right. Yet the notion that freedom and fairness are inherent qualities of law lives on, and if not reflected in all of the decisions, is nonetheless acknowledged and reinforced in numerous judgments of the courts. That is an important legacy and one which deserves neither to be diminished nor squandered.
Chaskalson’s suggestion that in the event that South Africa could implement a true democracy, the past ought not to be lamented must be taken seriously. He rightly enjoys the personal authority of a lawyer who dedicated his professional life to the oppressed, an authority now greatly enhanced by his Presidency of the Constitutional Court. Thus, his statement was frequently quoted at the Legal Hearing because it starkly raised and answered the question whether it is healthier to leave a traumatic past largely behind, dwelling only on its positive moments, in order to go forward into a healthy future. Clearly, those who established the TRC answered “no” to this question and so the little-use-in-lamenting-the-past stance might seem in direct conflict with the rationale for establishing the TRC.
However, this conflict has to be described, at least initially, in somewhat tentative terms because of a special problem which a TRC-type hearing encounters when it inquires into a legal order, especially into the role of judges. One might think that a TRC-type inquiry is justified as long as it confines its inquiry to investigations of perpetrators and victims, but that an investigation of judges risks politicising the judiciary in a way that compromises their role for the future. Even in those transitions where it was deemed fit to adopt a policy of lustration or of purging officials compromised by the past, judges have been seen to present a special case for exemption from lustration.
So we now have in place the factor that complicates evaluation of Tutu’s second claim. It is not exactly the issue of principle — that judicial independence requires that judges never submit to an official inquiry into their role. Nor is it exactly the issue of collegiality — that fragile relationships between judges would be ruptured by some judges testifying against others at a public hearing. Rather, it is that if some basis of respect for the rule of law was maintained by some lawyers and some judges in the period prior to the transition, that basis will provide a precious resource in the transitional period which can only be compromised if one dwells on who did what either to sustain that basis or to undermine it. From the perspective of the future, what matters is that the basis is there. Further, Chaskalson suggests that in the circumstances of South Africa no participant in the apartheid legal order will emerge untarnished from an inquiry into the past. And the actual course of the Legal Hearing might seem to support his suggestion.
For the absence of the judges dominated discussion of the Hearing and anger at that absence extended to the new order judges as well as to those few old order judges who had done their best to maintain respect for the principles of the common law. Further, the Hearing served to underline not only how few judges and lawyers had played a role in opposing apartheid, but also how it was the case that those who did oppose, especially the judges, could not help but be complicit in sustaining apartheid. Even the judges who gave the most innovative common law judgments of the apartheid era had for the most part to give the stamp of legality to apartheid law — if the government were prepared to be completely explicit about the “facts” in its drafting of legislation, judges had to defer. And even when they were successful in mitigating statutes by resort to the common law, they lent legitimacy to the apartheid government’s claim to be part of the family of legal orders that is committed to the rule of law. In sum, even the best efforts by some lawyers to blunt apartheid both involved them in the business of implementing injustice through law and lent legitimacy to the regime.
It was also perhaps inevitable at the Hearing that the representatives of the professional associations of advocates and attorneys, as well as those who made submissions in their own right, would succumb to the temptation to manage memory — to try to present their conduct in its best light, thus exposing just how few of their number had been serious about upholding the rule of law, and even exposing those few to criticism about particular incidents where they could have done better.
It may then seem that Chaskalson’s little-use-in-lamenting-the-past claim is right. The fact that there were so few lawyers who were prepared to take the role of upholders of the rule of law would hardly give the public confidence in a legal profession and in a judiciary still staffed by a vast majority of lawyers who had either been enthusiastic about apartheid or at least content to be its beneficiaries.
In addition, even if the plain fact approach were so wrong-headed that lawyers who believed in it were subscribing to a myth, one has to take into account that certain myths might be thought worth preserving because they enable the virtuous activity of the few. Thus, if the government had harassed in too crass a fashion the few lawyers and judges who upheld the rule of law, it would have violated the formal conditions of judicial independence and indeed the independence of the legal profession generally.[ 18]
And there was clearly more restraining the government than its tactical desire not to act in ways that would test its internal as well as its international sources of support because it could no longer say that it was a member of the family of “rule of law abiding” nations. As has often been remarked, commitment to the rule of law was part of the self-understanding of many of the Afrikaner men who presided over apartheid. In sum, respect for the rule of law, on a very particular understanding of it, was embedded in the psyche of lawyers as well as in many of the ruling politicians, and that respect made it possible for the alternative rule of law understanding to have a rather precarious toehold.
Of course, no judge or lawyer who had adopted the plain fact approach sought to justify themselves in terms of the utility of myth. Rather, they claimed that their professional obligations required them to be independent, required them to be plain fact lawyers who take heed of the law alone and not of morals. It seemed to them to follow that if one wanted lawyers to show fidelity to the law of the new order, one should not deprecate the tradition of independence which they had maintained under the old, since it was that same tradition of plain fact lawyering which would make fidelity to the law possible under the new. And this argument was usually put in tandem with another — that because of the sovereignty of Parliament under the old order, judges and other lawyers had no option but to do as the majority of them in fact did, just as the fact that the legislative endorsement by South Africa’s first democratic Parliament of the Final Constitution gave lawyers and judges no option but to be faithful to the legal values of the new order.
This line of thought leads to an even more potent objection to my argument about the plain fact approach and its associated focus on the formal attributes of judicial independence. It is that my critique of the plain fact approach is misconceived. There is no myth since the plain fact approach captures the essence of judicial independence and of the rule of law, and thus even though that approach might have had bad consequences during apartheid, it is what is required both for transitions and for the post-transitional society. This objection breaks down into three related parts.
First, my critique at most applies to the plain fact approach when it is employed in a context such as apartheid South Africa, where it led to the judges facilitating the implementation of evil policies of an illegitimate legislature. But if the legislature is legitimate both because it is democratically elected and because it generally enacts statutes that are not morally obnoxious, then there can be no objection to the plain fact approach. Second, where a legitimate legislature is the vehicle for entrenching a constitution protective of human rights, the plain fact approach requires judges to have regard to the fact of constitutional incorporation of these values. Third, my critique of the plain fact approach cannot account for one fact of the matter which all interpretative approaches must accept — that legislative intention must exert some weight. As Sidney Kentridge, the lawyer who ranks with Chaskalson in eminence for his human rights practice during apartheid, put it when a Justice of South Africa’s first Constitutional Court, a legal text “does not mean whatever we might wish it to mean” for else “the result [would be] not interpretation but divination”. Recall in this regard my recognition of the fact that even the most innovative judges of the apartheid era who used the common law to resist the injustice of statutes had to give way at a certain point to facts about legislative intent.
As I will now show, an attempt to answer this objection is best made at the level of legal theory, and it is at that level that we can then start to come to terms with the other complexities in this section.
The theoretical choice here can be depicted very starkly as one between a positivistic and a natural law understanding of the rule of law. The positivistic understanding is that fidelity to the rule of law amounts simply to regular compliance with clear, general and validly enacted rules. When the rules have a pernicious content, then adherence to the rule of law amounts to enforcing those rules. It follows that the value of regular compliance — the value of the rule of law — is entirely parasitic on the contingent content of the positive rules of the legal order. When the content is morally sound, there is some added value to compliance, since those subject to the law will know in advance what the framework of lawful activity is; and that provides the stability and security which is the prerequisite for a successful individual life. It follows that whatever remedies one seeks in a transitional situation for the abuses of the past are likely to be political rather than legal, except in so far as the officials of the old order acted in violation of the positive law. But basically what is required is effectively enforced new positive law, not a new understanding of the rule of law. Discontinuity of unjust positive law, continuity of the rule of law is then the positivist recipe for a successful transition.
Recall my claim that independence cannot be conceived in wholly formal terms, as independence from. One also has to ask what independence is for. Positivists do answer this second question, though their answer is that independence is important so that, as positivists from Thomas Hobbes through to Joseph Raz have suggested, judges will take notice only of positivistically conceived law. But that answer does not move much or even at all beyond formal terms, since it assumes that what is important about independence is the way in which the amoral process of law determination is best ensured.
Of course, legal positivism is not logically committed to the claim that upholding the rule of law is all that judges do. Many contemporary positivists say that when law lacks determinate content, judges have to make new law by exercising discretion. For these theorists, an exercise of discretion is a deeply political act of legislation, though within more confined limits than those under which a legislature acts.
I do not want to be drawn too deeply in this paper into the debate between positivists and others about discretion. But recall that the old order judges who adopted the plain fact approach sketched earlier did not find that they had discretion. Rather, in so far as the law presented interpretative difficulties, they resolved the difficulties by searching for the best available proxies for plain facts about legislative intent, and thus sought to avoid making political decisions about what the law ought to be.
So plain fact judges and positivist theorists would disagree about the extent to which their common understanding of the rule of law governed adjudication. But what they have in common would require both to suppose that in so far as judges during transitions have to resolve problems that arise out of the transition, the judges will, by and large, have to make explicitly political decisions — decisions beyond the governance of the rule of law.
Now, from the theoretical perspective of contemporary positivism, it is an advantage to conceive of the judicial role during transitions as overtly political. For one thing, the insight that judges are exercising a highly political discretion precludes them from hiding behind the “childish fiction” that judges never make law, and thus makes it easier to hold them accountable for what they do. For another, the insight avoids complicating or mystifying the parsimonious but allegedly accurate positivist understanding of the rule of law.
But from the practical perspective of plain fact judges, this politicisation of the judicial role during transitions is threatening for the very same reasons that plain fact judges reject the idea that in normal times they have discretion. They wish to conceive of their role as apolitical, albeit for political reasons. In other words, the political basis of the understanding of their role and of fidelity to the rule of law requires them to stay away from politics in that role. They are thus disabled by their conception of role from playing a role as judges during transitions in all those cases which cannot be brought under the governance of their conception of the rule of law. But notice that legal positivism is no less disabled from assisting judges qua judges, though particular theorists might be sanguine, even welcoming of this fact.
What positivist theorists should perhaps be less sanguine about is the affinity between their account of the rule of law and that of the plain fact judges. They will naturally attempt to disown such judges on the basis that authentic positivistic judges would recognise that they had discretion. But it is the case that the combination of political justification with a plain fact understanding of the law seems to lead to the plain fact interpretative approach. And, to put it bluntly, the alternative justifications available for legal positivism, which are purely conceptual in nature, and which have the consequence that much adjudication is discretionary, are of little interest to judges. For judges of any stripe are after an understanding of law that is both politically sound and that gives them a solid guide as to how to resolve even the hardest of cases.
The natural law understanding of the rule of law is no more helpful than legal positivism’s. It argues for a moral content to the rule of law that transcends positively enacted rules, and so claims a vantage point for judging illegal mere compliance with positive rules when the rules have violated that moral content. This position has its standard exposition in Gustav Radbruch’s “formula” that laws that reach a certain pitch of injustice fail for that reason to have the character of law. But this natural law understanding hardly differs from the positivistic one. Whatever arguments it can muster for its conclusion as to the content of intolerable injustice, it does not dispute the claim that fidelity to the rule of law amounts to regular compliance with clear, general and validly enacted rules. It simply subtracts the category of intolerably unjust rules, and that subtraction makes no difference to the positivist recipe for transitions.
Both of these understandings of the rule of law can be termed substantive, in that the value of the rule of law is, as I suggested, parasitic on positive law having a morally sound content. But whether the content is morally sound or not, what judges do in upholding the rule of law is the same — they determine and then apply (positive) content.
Ronald Dworkin’s theory of law as the product of appropriately conducted interpretation is also substantive in this sense. While it complicates the question of what counts as law, the limit as to what should count is set no differently from Gustav Radbruch’s — it is set by an argument about intolerable injustice. This position comes about not because Dworkin neglects the role of procedural values, including fairness, but because he regards procedural values as ultimately trumped by substantive values. In addition, he seems to equate fairness or process with the democratic reasons that require judicial deference to legislative intent. Thus, at the limit, justice trumps fairness, which means that law is stripped of its character not because of any procedural flaws but because it is intolerably unjust. And Dworkin, unlike Radbruch, does not bite the bullet on the issue of whether law that oversteps the limit should be disqualified as law. He has vacillated between the position that moral and legal duty come apart in this situation, so that judges should lie about their legal duty in order to comply with the moral law, and the position that there is no law.
More promising, in my view, is Lon L. Fuller’s account of the rule of law, as the rule of an inner or internal morality of law made up of eight principles — generality, publicity, non-retroactivity, clarity or intelligibility, non-contradiction, possibility of compliance, constancy through time, and congruence between declared rule and official action.
Fuller is often classed as someone who argued for a procedural account of the rule of law — that the value of the rule of law lies in the procedures necessary to produce valid law rather than in some predetermined substantive content against which positive law cannot offend without losing its character as law. This classification provides a useful contrast with substantive accounts. But it invites the positivist response that these principles have no value independent of the substantive content of the law and it neglects the degree to which process and substance are interwoven in Fuller’s position so that the very distinction between the two is blurred. Process is important because it brings substance to the surface — the principle of publicity. But process also plays a role in the determination of substance — the principle of intelligibility.
Consider a couple of examples that involved some of the common law presumptions and principles mentioned in the last section. South Africa’s highest court — the Appellate Division — overruled a judge who held that the statutory discretion given to officials to set aside areas for residence by one racial group had to be exercised reasonably or in accordance with a principle of equality, so that the areas set aside for whites could not be vastly superior to those set aside for other races. The Appellate Division held that legislators who initiated the vast social experiment initiated by the Group Areas Act must have contemplated that it could be implemented in ways that created vast inequalities.
No doubt, as a matter of plain fact, the Court was right. It was also likely that upholding the lower Court’s judgment would have provoked a legislative response that explicitly permitted officials to act unreasonably. But that legislative response would at least have had the virtue of forcing the government to declare explicitly in its legislation that it was not willing to abide by a central value of the rule of law.
Similarly, if judges consistently interpreted security legislation to as to try to render it consistent with the presumption of liberty, or with the principle that one is entitled to contest the basis of one’s detention, and that to do so one is entitled to access both to a lawyer and a court, the legislature would have likely responded — and on occasion did — by making it clear that it did not respect liberty and did not wish detentions to live up to requirements of legality. But that clarity also involved a public declaration that the law of the land was not designed to uphold values deeply associated with the rule of law.
At the same time as the principle of publicity would require the legislature to declare its lack of commitment to values of the rule of law, so the principle of intelligibility is imperilled. To see this one has to see that intelligibility has not to do with mere understanding, but with communication by legal authority to legal subject in a way that makes sense to the subject of her subjection to the law; that is, that subjection is necessary to establish the relationship of reciprocity that serves the interests of the subject. A law that tells one that one can be deprived of liberty without justification, or that one is not equal before the law because of one’s race, is not intelligible on this understanding of intelligibility, because one’s subjection to authority is in fact against one’s interests. Law has become — to use Fuller’s expression — a mere “one-way projection of authority originating with government and imposing itself on the citizen”. Inevitably, other principles are thrown into doubt — for example, the principle of congruence is not served if detentions fail to live up to requirements of legality, because there is nothing with which administrative decision-making can be congruent. The discretion to detain is, as administrative lawyers put it, “unfettered”.
The distinction between Fuller’s understanding of the rule of law and the natural law accounts is, then, that law does not lose its character as such principally because of its violation of some set of standards of justice external to the law, but because those with power choose to govern outside of the rule of law. Judges should take their stand not on their views about the political content of legislation, but on whether that legislation is capable of compliance with the inner morality of law. From the perspective of the rule of law, the list of substantive interests that are protected by the common law is less important than the list of legal requirements someone with public power has to satisfy before it is established that the violation of an interest is legally justified.
It is of course the case that the greater the weight accorded to a substantive interest, the more difficult it will be to discharge the onus of justification. And this connection is one between the procedural components of the rule of law and the substantive moral values that are incorporated into the (positive) law. If the positive law makes it increasingly explicit that values such as equality and liberty are disdained rather than respected by those with legislative power, it will be harder for judges to make that connection. But at the point where the relationship disintegrates, the application of the positive law becomes unjustifiable to its subjects, and so loses its claim to have the authority of law. And to the extent that the authority of law is conditional upon its potential to be justified to the subject, the idea of law presupposes the distinctly liberal idea of that subject as a free and equal citizen. To that extent, the substantive values of the legal order must necessarily include the values of liberty and equality, even if these are included in a fashion offensive to most, perhaps even all, liberal accounts of what I have called substantive morality.
For example, well before the Appellate Division decided the Group Areas Act matter, black South Africans had an abundant basis in the statutory regime, much of which long predated apartheid, for knowing that they were not regarded as free and equal citizens in the country of their birth. But what made that decision especially important is that it shattered both the principle of equality before the law and the ideology of grand apartheid that South Africa was to be divided up in such a way that served the interests of all racial groups. “Separate but equal” might be rejected as a slogan by most, even all, liberal accounts of equality, but it is still an interpretation of the value of equality that can plausibly be represented as a conception of the principle that all those subject to the law are equal before the law. In contrast, “separate and unequal” is not an interpretation of equality, but a violation of the idea of equality on any account, and thus also of the idea of equality before the law.
It is important to emphasize that the conclusion to the argument that particular laws violate the inner morality of law is not that judges are therefore permitted to invalidate these laws. Whether judges have such a review authority depends on the institutional facts of the legal order. Indeed, it will look odd to claim that a law remains on the books even though it violates the rule of law only if one accepts the positivist view that what matters most for legal theory is whether laws are valid or not. Thus Dworkin and other critics of legal positivism (Gustav Radbruch, for example) have erred in accepting the positivist challenge that rests on examples of particular laws that are technically valid but morally offensive.
For Fuller, the issue is not validity or invalidity, but the strain that any law that is suspect from the perspective of the inner morality of law puts on the legal order as a whole. As the examples multiply, so the legal order will start to lose its character as such, as it will no longer be an order which can be plausibly claimed to aspire to legality. So the point for judges who share that aspiration is to orientate themselves towards legality in their decisions, in so far as this is possible.
It is, in my view, the Fullerian account of the rule of law that makes sense of the claim that the old order South African judges were in dereliction of their duty. These judges were independent in order that they could uphold the justice of the law — the inner morality of the rule of law — which they failed to do, mainly because they assisted the apartheid governments to avoid declaring their hand fully on the extent to which they were prepared to rule outside of the rule of law. Moreover, the interpretative approach I call the plain fact approach was the one which these judges adopted; and so, at least in the circumstances of apartheid South Africa, that approach should be rejected.
To go back to the objection outlined at the end of the last section, I want also to argue that that approach should be rejected even when there is an authentic democratic justification for it, or when a transitional society is well on the way to putting such a justification in place. It should be rejected, first, because it views law as a one-way projection of authority and thus requires that inevitable, interpretative problems are resolved not in accordance with the rule of law, but in terms of proxies for legislators’ intent. It thus precludes the intervention in the interpretative process of those rule of law considerations that make law into a relation of reciprocity between ruler and ruled. And it precludes such intervention just at the point where the subject actually encounters the law. If these rule of law considerations were antithetical to democracy, such preclusion would be in order. But they are far from being antithetical. While their existence and application do not in themselves make a legal order into a democratic one, they do make the legal subject into an individual closer to the model of the democratic citizen than the plain fact approach.
Second, it is of course the case that when a legal order incorporates a written constitution that explicitly protects fundamental values, and gives to judges guardianship of those values, the plain fact approach will require judges to advert to those values. But there are a series of problems raised by such incorporation for the plain fact approach, each of which makes the approach ill suited to addressing such a constitution.
Most important here is that given the political justification for the plain fact approach, it is a grave political mistake to put such a constitution in place. For it is the nonsense-upon-stilts writ large that Bentham and other positivists have often attributed to the common law. Once such a constitution is in place, the plain fact approach is confronted by questions to which it has no answer. Are the constitutional values to be interpreted as one-way projections of authority by dint, for example, of a doctrine of original (or plain fact) intent? Alternatively, does the constitution create a space for the unfettered discretion of judges, so that that space should be as restricted as possible? And if the space is not so restricted, does the fact of incorporation require the plain fact approach to notice that fact and then as it were liquidate itself?
Third, the Fullerian understanding of the rule of law recognises that legislative intent has weight and that there is a distinction between interpretation and divination. But it is also the case, as Fuller said, that a “statute is seen, not as a message cast into a void, but as a message whose meaning is dependent upon the interpretation its addressee, as a reasonable and sensible man, would put on it.” If the legislative intent is such that no such interpretation is possible then the legislature — or rather the government that controls it — has chosen to govern free of the constraints of the rule of law, and thus through a medium different from law.
In the light of the above, there is something to the claim of plain fact lawyers that during apartheid they helped to maintain the idea of the rule of law and a tradition of independence. But they helped to maintain these only because an attack on the formal attributes of independence would have destroyed even the façade of the rule of law, and, as long as there is the façade there will be some room for interpretation in accordance with the inner morality of law. That fact does not of itself justify the choice of the truly independent lawyers and judges to continue in their role, since to the extent they are successful they also help to maintain the façade.
But from the perspective of a transition, that cost is probably worthwhile. The resource maintained by the few is probably close to being a necessary condition for the rule of law in the future. One cannot neglect here that, in the nature of things, transitional and even post-transitional regimes are often tempted to deal with the vast problems they face by riding roughshod over the rule of law in ways that cruelly mimic the past from which they are trying to escape. Indeed, it is precisely, as it were, the charismatic authority of the lawyers who in the past used independence for the ends which justified their having it that might in the transitional present stand in the way of the regime’s temptation to revert to the lawlessness of the past. But that authority is not free-floating — it stems from the lawyers’ proven devotion to the rule of law in the bad old days. And it is hardly far-fetched to suggest, against the distinction earlier, that the kind of politician best suited to lead the transitional process does not have authority of a purely charismatic kind. While Nelson Mandela is a man of undoubted charisma, the respect he rightly enjoys comes in large part because he devoted his personal qualities to the establishment of the rule of law in South Africa, thus rejecting the idea that the political issues could be reduced to a fight between communism and capitalism or to one between different racial groups.
My argument, as should now be clear, is that fidelity to the rule of law is fidelity to a set of values that are quintessentially political. As Thomas Hobbes so clearly saw, these values involve a commitment to a very particular understanding of how to construct a civil society. And that commitment is no less political in societies that are stable and abide by the rule of law. At most the commitment is a much more implicit part of the social compact. As Fuller described it, in a nation committed to the rule of law “both law and good law” would be regarded “as collaborative human achievements in need of constant renewal” and lawyers would be “still at least as interested in asking ‘What is good law?’ as they are in asking “What is law?’”. And without that commitment one might well have peace, but it will be a “peace without justice”.
Hence there is no paradoxical role for law during transitions, nor is there anything paradoxical about the role of judges who are required to create the conditions for their independence. Judicial authority is not something impersonal — it is deeply dependent on a commitment to a particular vision of political order and it is the ongoing demonstration of that commitment on which their authority depends.
If there is any paradox at all, it is the one described by the editors of an excellent collection of essays on transitional justice, which they call the “paradox of the probable and the unnecessary”: “The impact of truth and justice policies on a new democracy depends on starting conditions or the initial balance of power; in other words, the more likely the implementation of such policies is because of a favourable balance of forces, the less necessary they are to ensure a process of democratisation”. But, as they also point out, whatever the effect of such policies on democratisation, they are “crucially important moral and political demands which are, even if imperceptibly, part of a changing climate that places respect for human rights at the forefront within and between national communities”.
Of course, even if I am right that we do not need a special account of the rule of law for transitions, the justice of the rule of law is not equivalent to justice itself and so we are still left with the possibility that there is a special kind of transitional justice. And so in the last section, I want to suggest that the best candidate for this kind — restorative justice — turns out on closer inspection to be little more than an account of the justice of the rule of law.
It is well known that perpetrators of gross human rights violations who testified before the TRC received both criminal and civil amnesty if the amnesty panels of the TRC were satisfied that the perpetrator had fully disclosed his role and that the violation was committed in pursuit of a political purpose. That is, the elites who negotiated South Africa’s transition decided not to rely on criminal trials as the main mechanism for dealing with the past.
One can view this decision as born of necessity — the old regime retained enough muscle to sabotage the transition and the new regime’s limited resources for mounting successful prosecutions. In these circumstances, justice was unlikely to be achieved, so it was traded for the truth that emerged from the hearings at which victims of violations, or their relatives, testified as well as the hearings at which perpetrators sought amnesty.
Alternatively, there is the claim made by the TRC itself and by its many supporters that truth was not traded for justice. Rather, the way the TRC went about finding out the truth achieved a kind of justice different from — even superior to — criminal or retributive justice. The kind of justice that the TRC achieved is restorative justice, which holds perpetrators accountable for their actions, but restores both them and their victims to the community by recreating the relationship of equality ruptured by the criminal act.
Moreover, it is claimed that at least in a transitional context, restorative justice has many advantages over retributive justice. It promotes a process of truth-finding in which a fuller picture of the truth emerges than would in a series of trials, since amnesty seekers have an interest in making full disclosure, which in turn implicates others who therefore will come forward to seek amnesty. Further, in that process the testimony of victims has a role that goes well beyond serving as an instrument to achieve conviction. It is in taking this role — in telling their stories — that victims might find not only that they can come to terms with the abuses, but also that they are “restored” to a relationship of equality with the perpetrators, so that they develop a sense of agency appropriate for participation in a democratic society. Even more generally, the supporters of the old regime are forced through the confessions of the perpetrators to acknowledge its nature, while those who suffered under that regime, though without suffering gross human rights violations, can — through the experience of the victims — also come to terms with the past and find their agency in a way appropriate for a democratic future.
The actual experience of the TRC should provide a highly cautionary note. First, as Richard Wilson shows, many victims did not get to testify and those who did often found themselves in a micro-managed process in which their testimony was reduced to the empirical data the TRC required. Indeed, he argues that the objection to criminal justice that it instrumentalises victims’ testimony in order to achieve the end of conviction can be turned against the TRC, which instrumentalised testimony in order to assist the project of nation-building. Second, anecdotal evidence suggests that perpetrators often stuck to a script, probably coordinated by the few lawyers who appeared time and again with this group, which disclosed as little as possible and attempted to confine implicating others to implicating security force actors who had died. Third, in relation to the issue of catharsis, perpetrators stuck for the most part to rote apologies if they apologized at all and victims often found the experience of testifying deeply traumatic. There was also deep resentment at the attempt, spearheaded by Tutu, to impose a quasi-religious protocol of expressions of repentance and forgiveness on the Hearings. Finally, the South African government was committed to providing meaningful compensation to victims in accordance with the TRC’s recommendations, but it has largely reneged on this commitment.
Many proponents of restorative justice will react to these points by suggesting that the causes of what went wrong with the TRC lie in the ways in which its design departed from the model of restorative justice, most notably in that there were separate hearings for victims and perpetrators rather than the cathartic confrontations of a single hearing, and in that the burden of reparation was borne by the state rather than by the perpetrators. Such proponents of “strong” restorative justice do not, it should be noted, accept the second assumption of the transitional justice debate that there is something altogether special about transitional justice, since their model of justice is drawn from the experience of aboriginal groups in stable societies with restorative justice. Indeed, they wish to argue that the appropriateness of their model for transitions underwrites the need to reform the criminal justice systems in stable societies along restorative lines.
The restorative justice model might offer much as a project for reforming stable societies. However, it is out of place in transitions where a structural expectation, let alone the requirement of expressions of repentance and forgiveness, is inappropriate, and where the issue is not how best to restore someone to a place in a fairly homogenous community, but how to build a political community in the face of deep division and distrust.
For these reasons, the more convincing accounts of transitional justice as special in virtue of its restorative aspects emphasize the ways in which the choices in the design of the TRC sacrificed criminal justice for the sake of truth and reconciliation. They acknowledge that the sacrifice is a matter of deep moral regret — something fundamental has been lost. But it is not only that the sacrifice was required by realpolitik that justifies it, but that to the extent that the TRC can be understood in terms of a restorative justice model, it did not fail entirely to live up the commitments of criminal justice. That is, the accounts emphasize the commitments shared between restorative justice and retributive justice in transitional contexts, which are, according to the best such account I know, “(1) to affirm and restore the dignity of those whose human rights have been violated; (2) to hold perpetrators accountable by emphasizing the harm they have done to individual human beings; and (3) to create social conditions in which human rights will be respected.”
But this more pragmatic or “weak” account of transitional justice as restorative justice is convincing because of its proximity to an account of transitional justice that not only refrains from any reliance on the restorative model, but even denies, in my view, that there is anything special about transitional justice. As Jonathan Allen has argued, the kind of justice achieved by the TRC is better understood in terms of what he calls “justice as recognition” and “justice as ethos.” In regard to the first, the work of the Committee on Human Rights Violations supports the restoration of the rule of law by drawing attention to the “evil consequences resulting from apartheid and the officially sanctioned transgressions of the rule of law”. The hearings thus “demonstrate the consequences of public commitment to justice and the rule of law and thus show the importance of such a commitment. In this sense the TRC is supportive of legal recognition in a context where law’s equal recognition of all responsible agents has been grossly distorted.”
In regard to the second, justice as ethos involves the demonstration of how under apartheid people’s sense of justice was corrupted by their equation of justice with the political ideology of the group of which they were members. This equation in turn led to an impoverished view of public discourse — politics is either a dirty business or a noble calling in which the end justifies the means. But neither of these views permits any operation for a sense of injustice, and that sense is required if the transitional government is to preserve legitimacy and a commitment to the constraints of the rule of law, constraints which make it possible for citizens to call government to account for injustice.
As I understand them, Allen’s ideas of justice as recognition and justice as ethos are not meant to be discoveries of new kinds of justice. Rather, they are labels for processes that serve the transformation of an unjust society into a just one. But the justice they serve is little more than the justice of the rule of law, the kind of justice which has to be in place before order becomes something worth having, and which also makes it possible for a society to decide other kinds of political issues in a civil fashion.
It follows that most of the debate about transitional justice should be deflated to a debate about how to achieve the rule of law. The inflated debate purports to be about what justice is, retributive, restorative, or whatever. The deflated debate is about our ordinary understanding of the rule of law, and about how the rule of law is established even under very difficult conditions. The deflated debate is not, however, one with the wind knocked out of it. It is an important debate about the content of one of the achievements of Western civilization, and one which seems more universally valued than others.
In this article, I have tried to sketch an account of what is achieved when the rule of law is established, though I am of course aware that any such account is controversial. I hope to have gone some way to showing that the issues raised in transitions might shed a light that helps in taking that controversy forward. But, in my view, the light which transitions shed is probably much the same as that shed by confronting our intuitions about the ordinary legal situation from the perspective of any extraordinary one. For instance, we learn something about legality when we confront the situation of the proper rule of law response to a declaration of emergency, to a coup d’état, or to a terrorism statute. And if that is right, we can conclude that in all such confrontations we will find that our understanding of what is ordinary changes, just as the extraordinary seems less unfamiliar.
[*] Law and Philosophy, University of Toronto. This paper draws on two unpublished papers, one presented at Mahmood Mamdani’s seminar at Columbia University in 1999 and the other at the Conference on the Act of Settlement (Vancouver) in 2001. I thank all those present for discussion, especially Ian Shapiro, the commentator for my session at Columbia. An earlier version of this paper was presented at the meeting of the American Political Science Association in San Fransisco in 2001, and I thank all those present for discussion, especially Farid Abdel-Nour, the panel discussant, and Cheryl Misak. It has been revised further in the light of comments at a session of the Conference on Transformative Justice, held at the National Centre for the Humanities at Chapel Hill in October, 2001, and at a Political Theory Workshop at Yale University. I thank all present for discussion, especially Bruce Ackerman, Allen Buchanan, Paul Kahn, Elisabeth Kiss, David Luban, David Lyons, John McCormick, Gerald Postema, Mark Tushnet and Jeremy Waldron.
 (1998) 115 South African Law Journal 21 at 29.
 For a wider-ranging, general argument, see my review article, “Transitional Justice” in the forthcoming first volume of the International Journal of Constitutional Law.
 Hence the very deliberate choice in South Africa of the death penalty as the first case for the Constitutional Court – S v Makwanyane (1995) 2 SA 391 (CC). This case afforded the South African public the drama of the Court invalidating a statutory provision which while itself very popular was part of the legacy of the apartheid regime and a feature of the legal order of which Nelson Mandela clearly disapproved. A case that combined these issues was the constitutional challenge to the legality of the amnesty provisions of the TRC, which the Constitutional Court rejected in Azanian Peoples Organisation (AZAPO) and Others v. President of RSA and Others 1996 (8) BCLR 1015 (CC).
 Ruti G. Teitel, Transitional Justice (New York: Oxford University Press, 2000) 6.
 See my Judging the Judges, Judging Ourselves: Truth, Reconciliation and the Apartheid Legal Order (Hart Publishing: Oxford, 1997) for the full account of the Hearing from which much of this section is drawn.
 See Truth and Reconciliation Commission of South Africa Report (Cape Town: Juta & Co., 1998), volume 4, chapter 4.
 Transcript, 3-7.
 “You see, in dealing with human rights violations, you are really concerned with justice, law, order, the disposition of power and authority and how these are regulated within conventional parameters so that they are not abused”; Transcript, 2.
 Transcript, 8-9.
 Supreme Court Act (1959), section 10(2)(a), my emphasis.
 Judge Bertha Wilson in her partial dissent in MacKeigan v. Hickman  2 SCR 796, 808-9. The case arose out the attempt by the Nova Scotia Court of Appeal to pin the blame for the wrongful conviction of Donald Marshall — a seventeen year old native Canadian — on Marshall. The Court of Appeal’s comments were widely perceived to be a continuation of Canadian racist attitudes. Moreover, one of the judges had been Attorney General at the time of Marshall’s trial. The judges refused to attend a subsequent commission of inquiry and their right to refuse was upheld by the majority of the Supreme Court. For discussion, see Judging the Judges, 138-50.
 Arthur Chaskalson, “Law in a Changing Society” (1989) 5 South African Journal on Human Rights 293.
 Ibid., 294.
 The common law heritage of South Africa includes principles derived from Roman-Dutch law, the legacy of the first white settlers. The place of Roman-Dutch law has been a subject of academic controversy, in particular whether the old authorities on Roman-Dutch law are just part of the common law or in some way a superior source of law. Such controversy is often tinged by nationalism, by Afrikaner v. British national sentiments. However, since the Roman-Dutch authorities endorsed roughly the same set of principles as the common law ones to which Chaskalson referred, this controversy does not affect the rule of law issue canvassed in the text. Indeed, in the heyday of apartheid, Afrikaner nationalism favoured the plain fact approach described below, one which marginalised both the Roman-Dutch authorities and the common law.
 It is interesting that John Austin in his major work on philosophy of law found that no other than Thomas Hobbes had been too optimistic about the possibility of educating the masses and thus put forward a version of legal authority suitable for rule by elites of the ignorant crowd: Austin, Lectures on Jurisprudence, 2 vols. (London: John Murray, 5[th] edn. 1885, rev. and ed. R. Campbell), vol. I, 281-3.
 Chaskalson, “Law in a Changing Society”, 295.
 Neil Kritz, “The Dilemmas of Transitional Justice”, in Kritz (ed.) , Transitional Justice: How Emerging Democracies Reckon with Former Regimes (Washington, DC: United States Institute of Peace, 1995, 3 vols.) vol. I, xix at xxv-xxvi.
 “Crass” has to be understood here as a term relative to the South African context, where political opponents of apartheid were subject to detention, torture in detention, and assassination. The government often issued veiled threats against such lawyers and judges and the lawyers were often subject to harassment by the security forces. Late in apartheid the State Security Council, which at that stage ran the country, put together a list of the most prominent lawyers accompanied by the kind of jargon that suggested that they were to be subject to extreme forms of harassment or worse and one prominent human rights lawyer was assassinated in Namibia. But generally it seems safe to say that judges had nothing to fear, except that they might not be promoted, and lawyers enjoyed an immunity from severe forms of harassment. One exception to the point about promotion was the elevation of Judge Milne, the active Judge President of the most liberal Bench in the country (the Natal Bench) to the highest court – the Appellate Division. This promotion both weakened the Natal Bench and had no impact on the Appellate Division since the Chief Justice carefully kept him (and other untrustworthy judges) off panels that decided public law issues to do with resistance to apartheid.
 These are culled from Alfred Cockrell’s perceptive review of Judging the Judges, “Who’s Afraid of the ‘Plain Fact View’”? (2000) 3 Legal Ethics 76. (For some related points, in response to my Hard Cases in Wicked Legal Systems: South African Law in the Perspective of Legal Philosophy (Oxford: Clarendon Press, 1991) see also Anton Fagan, “Delivering Positivism from Evil” in David Dyzenhaus, ed., Recrafting the Rule of Law: The Limits of Legal Order (Oxford: Hart Publishing, 1999) 81 .)
 Kentridge AJ, in S v. Zuma  ZACC 1; 1995 (2) SA 642 (CC), at 653.
 Note that perhaps the most innovative of these judges, the late John Didcott, managed to avoid ever imposing a death sentence, even though the South African statute which set out the exceptions to the requirement that the sentence be imposed for certain crimes made such a record seem unachievable. One of South Africa’s most eminent new order judges remarked to me that he, having made the transition from human rights lawyer to judge, now feels that Didcott was in dereliction of his duty because in order to establish his unblemished record, he had to lie about the law. The section that follows is as much an attempt to respond to this remark, made by someone whose stance is informed by profound reflection as well as practical experience, as it is to the theoretical objections I list in the text.
 Even that added value has to be balanced against the value of having rules that are to some extent open textured so that officials can adapt them when necessary to meet changing social conditions; see Joseph Raz, “The Rule of Law and its Virtue”, in Raz, The Authority of Law: Essays on Law and Morality (Oxford: Clarendon Press, 1983), 210. Raz, following HLA Hart, and indeed Hobbes, recognises that rules are inevitably open-textured and, at least with Hart, argues for the desirability of this feature of rules so that it is often for the good that legislators deliberately inject the feature into legislation.
 See Thomas Hobbes, Leviathan (Harmondsworth: Penguin, 1986), chap. 26, 327-28 and Raz, “The Rule of Law and its Virtue”, 216-17, including note 6.
 See HLA Hart, “Positivism and the Separation of Law and Morals”, in Hart, Essays in Jurisprudence and Philosophy (Oxford: Clarendon Press, 1983) 49, quoting from Austin at 66.
 See Robert Alexy, “A Defence of Radbruch’s Formula”, Julian Rivers, “The Interpretation and Invalidity of Unjust Laws”, in Dyzenhaus, ed., Recrafting the Rule of Law: The Limits of Legal Order, respectively and Stanley L. Paulson, “Radbruch on Unjust Law: Competing and Later Views” (1995) 15 Oxford Journal of Legal Studies 489. My discussion here is most indebted to Rivers, 58-65 and also to the account of the border guard trials in A James McAdams, Judging the Past in a Unified Germany (Cambridge: Cambridge University Press, 2001).
 See especially his response to critics in Marshall Cohen, ed., Ronald Dworkin and Contemporary Jurisprudence (London: Duckworth, 1984) 256-60.
 Lon L. Fuller, The Morality of Law (New Haven: Yale University Press, 1969, revised edition), especially the section replying to critics at 187-253.
 Minister of the Interior v. Lockhat 1961 (3) SA 587 (A), discussed in detail in David Dyzenhaus, Hard Cases in Wicked Legal Systems: South African Law in the Perspective of Legal Philosophy, 79-83.
 For discussion, see ibid., chapter 4.
 Fuller, The Morality of Law, 207.
 Here I follow in part the path-breaking treatment of Fuller in TRS Allan, Constitutional Justice: A Liberal Theory of the Rule of Law (Oxford: Oxford University Press, 2001).
 See the discussion of the rule of law in Rome in GEM de Ste Croix, The Class Struggle in the Ancient Greek World (London: Duckworth, 1981), chapter VI.
 In “Positivism and the Separation of Law and Morals”, Hart breaks up challenges to the “separation thesis” into challenges to do with the interpretation of particular laws and challenges that depend on claims about the nature of legal order; this way of carving up the terrain rigs it against Fuller, for whom the two sorts of challenge are intimately connected.
 Allen Buchanan, “From Nuremburg to Kosovo: The Morality of Illegal International Legal Reform” (2001) 111 Ethics 673 premises his argument on the validity/invalidity distinction and so, in my view, misses much of what is valuable in Fuller in his account of the rule of law at 683.
 Lon L Fuller, “The Justification of Legal Decisions”, in Die Juristische Argumentation, ARSP NS 7 (Wiesbaden, 1972), 78.
 This point is fully elaborated in Jeremy Waldron’s sympathetic though critical elaboration of Fuller’s inner morality as an inner morality of fidelity to law — see Waldron, “Why Law – Efficacy, Freedom, or Fidelity?” in (1994) 13 Law and Philosophy (special issue on Lon Fuller, Ken Winston, ed.) 259. At 283, Waldron says that the argument he attributes to Fuller is “precarious ... as well as modest”. It is modest because Fuller does not say that “the forms and institutions of legality are the only ones we should use” and even then the “claims made on their behalf are subject to the vicissitudes of political sociology”.
Waldron does not seem to doubt, however, that political sociology would assist Fuller and, if my own work on South Africa is classifiable as such, I would venture it as providing some support. More troubling for my own argument is the first point. I agree with Waldron that Fuller was open-minded about questions of institutional design, so that he did not think that a legitimate government has to address every social problem through the medium of the rule of law. (Here I think, as Fuller did, that a study of what one might think of as the political sociology of administrative law would be illuminating — see Fuller, “Positivism and Fidelity to Law – A Reply to Professor Hart”, 648.) Fuller must, however, be committed to the view that certain fundamental interests of the legal subject have to be so addressed if they are to be legitimately addressed. Since the point of law is to establish a relationship of reciprocity between ruler and ruled, so that society is governed in the fundamental interests of the individual, law has to address these fundamental interests. It also must be the case for Fuller that law’s character as such depends on the maintenance of its inner morality in so far as collective decisions affect these fundamental interests. Moreover, since on the list of such interests is the citizen’s interest in being treated as free and equal, there is a tighter connection between the idea of fidelity to law and the idea of freedom (and the idea of equality) in Fuller’s work than Waldron suggests.
 Indeed, Hobbes can productively be viewed as addressing the problem of transitional justice. Despite the fact that he starts his argument about the design of political and legal order in the state of nature, his audience is a society in transition and his task is, on the one hand, to show the leaders of the transitional process how best to transform their society into one that is both stable and just and, on the other, to persuade those subject to the new sovereign that there are grounds for deeming the law in a society thus transformed to be legitimate even when they object to its content. I have recently argued that Hobbes’s legal theory is best understood as Fullerian, rather than positivistic — see Dyzenhaus, “Hobbes and the Legitimacy of Law”, (2001) 20 Law and Philosophy 461.
 Fuller, “Positivism and Fidelity to Law”, 648.
 For a detailed account, see Margaret Popkin, Peace Without Justice: Obstacles to Building the Rule of Law in El Salvador (Pennsylvania: Penn State Press, 2000).
 Alexandra Barahona De Brito, Carmen Gonzalez-Enriquez, and Paloma Aguilar, The Politics of Memory: Transitional Justice in Democratizing Societies (Oxford: Oxford University Press, 2001), 314.
 For a fuller account of the issues in this section, see David Dyzenhaus, “Justifying the Truth and Reconciliation Commission” (2000) 8 Journal of Political Philosophy 470.
 In fact the Act required that the violation be shown to be proportional to the purpose, but amnesty decisions did not apply this requirement.
 Richard Wilson, The Politics of Truth and Reconciliation in South Africa: Legitimizing the Post-Apartheid State (Cambridge: Cambridge University Press, 2001).
 Such sentiment surfaces even in the sympathetic account of the TRC by its Deputy Chair: Alex Boraine, A Country Unmasked: Inside South Africa’s Truth and Reconciliation Commission (Oxford: Oxford University Press, 2000).
 For example, Jennifer J Llewellyn and Robert Howse, “Institutions for Restorative Justice: the South African Truth and Reconciliation Commission”, (1999) 49 University of Toronto Law Journal 355.
 Elizabeth Kiss, “Moral Ambition Within and Beyond Political Constraints: Reflections on Restorative Justice” in Robert I Rotberg & Dennis Thompson, eds. Truth v. Justice: The Morality of Truth Commissions (Princeton: Princeton University Press, 2000) 68, at 79.
 Jonathan Allen, “Balancing Justice and Social Unity: Political Theory and the Idea of a Truth and Reconciliation Commission”, (1999) 49 University of Toronto Law Journal 315, at 330-1.
 Ibid., at 336-7.