II. Positivism and Formalism

The first difficulty, recall, has to do with the association of formalism with a combination of a rigid doctrine of the separation of powers and an understanding of the judicial role that has judges confined to bringing out the content that the law in fact has, which is the way Hart understood the formalist position. Hart, however, indignantly rejected any association of legal positivism with the view that legal reasoning consists “in deduction from premises in which the judges' practical choices or decisions play no part”. It was, he thought, plainly absurd to conceive of the judge as a “formalist”, “automatic”, a “slot machine”. Indeed, his positivist predecessors, Hart said, had insisted that, in cases where it was controversial what the law required, judges are in a “penumbra” of unsettled law and they have to legislate an answer; and John Austin, unlike Jeremy Bentham, had welcomed judicial legislation.Footnote ¹¹ Moreover, Hart suggested that the formalist understanding of legal reasoning and its accompanying rigid doctrine of the separation of powers were part of the natural law and common law traditions.Footnote ¹²

However, in an influential essay on conceptions of the rule of law in public law, Paul Craig argued that such conceptions are either formal or substantive and, in his view, Joseph Raz's legal positivism provides the exemplar of a formal theory.Footnote ¹³ Craig finds that are four marks of a formal theory. I shall show that all are important to a theory of public law form, albeit in a way that legal positivists for the most part would deny.

According to Craig, the marks of a formal theory it that it addresses

In contrast,

Raz's account of the rule of law is formal, in Craig's view, because it avoids making the mistake of collapsing an account of the rule of law into an account of the rule of good law and thus preserves its “independence” from political theory. Raz thus delivers the “important” “message” that, if you wish to defend a right, you should do so directly through an argument in political theory, and not through a claim that the right is somehow part and parcel of the rule of law.Footnote ¹⁶

In Craig's account, then, form is, first, a property of a theory of the rule of law that is independent of political theory and takes no stand on the issue whether the content of the law is good or bad. I shall call this “form as independence”. Second, a formal theory focuses on marks of law that have not to do with law's substance but with law's validity – whether the person or body who made the law was entitled to do so and whether he or it did so in the right way. I shall call this “form as pedigree”, as it has to do with the manner in which the law came about. The third has to do with the clarity of the law that is promulgated and it does have to do with content – that is, the law must have a content that is clear enough to guide those subject to it. I shall call this “form as fact”. Finally, there are characteristics of law that have to do with the legal quality of the law that is produced by those with authority to make law – for example, that it is directed at future conduct. I shall call these characteristics “form as legality”.

I agree with Craig that these four marks are among the marks of a formal conception. He is also correct to observe that the phrase the “rule of law” is “dependent upon what one understands by the term ‘law’”, so that “what ultimately divides the formal and substantive conceptions of the rule of law is disagreement about the way in which we identify legal norms”.Footnote ¹⁷ But he goes wrong, in my view, because the important division in not between formal and substantive conceptions of the rule of law, but between different formal conceptions, which differ most fundamentally at the level of normative argument about the appropriate design of legal order.Footnote ¹⁸

Despite Hart's claim that it is plainly absurd to conceive of the judge as a “formalist”, he relied throughout his published work on the idea that law properly so called governs when and only when it has a content that can be determined by factual reasoning. And this reliance carries over into Raz's “Sources Thesis”, which Craig refers to in rebuttal of Dworkin's claim that legal positivism cannot account for the way in which judicial interpretation of the law must refer to moral principles that are embedded in the law:

In other words, legal positivists acknowledge that judges often have to rely on moral principles in interpreting the law, but they insist that, in such cases, the law on the question is unsettled – that is, judges have to legislate or exercise a discretion. As Raz makes clear, the Sources Thesis requires that the content of the law be determinable “without using moral arguments”; it has, that is, to be determined by resort to facts.Footnote ²⁰ Hence the distinction between settled and unsettled law presupposes precisely the formalist view of law that Hart was so anxious to disown.

But why accept the positivist way of making the distinction between settled and unsettled law – that the law is unsettled when judges have to rely on their moral acumen? The answer to this question is, according to Raz, supposed to come out of a theory of authority and a feature of authority that Hart described when he said that the authoritative nature of legal reasons resides in their “peremptory” quality: that such reasons are intended to get one to act as they stipulate without consideration of the merits of the content of the reasons.Footnote ²¹ Hart sought to capture by this term the feature of authority that Hobbes introduced into political and legal theory when he said that the distinction between command and advice is that with command one obeys because the commander says so, not on the basis of one's evaluation of the content of the command, whereas advice offers one a reason to evaluate.Footnote ²²

Hart suggested a second feature of authoritative reasons – that they are “content independent”. By this term, he meant that one who has authority to command can issue commands with very different content, but in each case the content is “intended to function as a reason independently of the nature or character of the actions to be done”.Footnote ²³ The term “content independence” is however rather misleading. In order for a reason to function as Hart suggested, it has to have a very particular kind of content – one determinable as a social fact. In other words, as soon as one has to rely on moral acumen to work out the content of a reason, the reason ceases to function as an authoritative reason. And it this account of authority that Craig presupposes in his adoption of Raz's framework for understanding a formal conception of the rule of law.

However, this account of authority elides a general feature of authority (that one acts because the authority said so) with the positivist claim that a law offers an authoritative reason only when it has a particular kind of content – one that can be determined by factual tests mandated by the law.Footnote ²⁴ In making this elision, Hart and Raz fail to see that this general feature of authority is consistent with the family of interpretive theories whose most prominent member is Dworkin. That moral acumen is required to work out the content of the law permits the stance that one should obey the law with that content because it is the law and not because one would endorse that content. Even when, as Dworkin advises, judges interpret the law in a way that seeks to show the law in its best moral light, the result might well be morally far from ideal. But we can still think that the law has to be obeyed because it is the law of our land – because our legislature spoke. And that leaves intact the distinction between command and advice. That thought, of course, must be backed by some further reason for supposing that the law of our land has to be obeyed and here only some normative argument will suffice. It follows that something like moral acumen is required to supply the reason why we think that we are under an obligation to obey the law, even when its content is morally far from ideal.

Hobbes himself set out a list of 19 laws of nature, “the true and onely Moral Philosophy”,Footnote ²⁵ and argued that judges are under a duty to interpret the positive law in their light. Moreover, when he distinguished between legal authority and advice, he said “Law in general, is not Counsell, but Command; nor a Command of any man to any man; but only of him, whose command is addressed to one formerly obliged to obey him”.Footnote ²⁶ Hence, not only did Hobbes think that this feature of authority is consistent with moral acumen being required to work out the content of the law; he also thought it was consistent with the view that a legal authority is such only if there is a basis for holding that legal subjects are under a prior obligation to obey the authority.

On my interpretation of Hobbes's legal theory, the two issues are connected in that one of the reasons that one has a prior obligation to obey legal authority is that legal authority is constituted in such a way that its exercise is conditioned by Hobbes's laws of nature.Footnote ²⁷ Here it is important to note that one of Hobbes's major innovations in legal theory was to conceive the laws of nature not as God-given moral principles external to legal order used to judge the content of human positive law,Footnote ²⁸ but as what we more readily think of today as principles of legality or of government according to law. Put differently, principles of legality constitute the form government takes when it is according to law. Together, such principles provide the basis for the construction of a theory of the public law form.

Craig is correct that contemporary legal positivism provides such a theory, but he does not take into account, first, that it is in contention with other theories of the public law form, second, that the way the Sources Thesis shapes its understanding of the marks of form is irreducibly dependent on a normative argument about how best to understand law. That dependence is consistent with the feature of authority that one obeys not because of the merits of the content of the law, but because it is the law. But it is not consistent with Craig's contrast between form and substance since it does not permit the distinction between a theory of the rule of law and political theory that Craig envisages, since a theory of law is always at base a political theory. And, because neither correspondence with brute facts about the law nor features of authority can ground the positivist conception of law, the obvious candidate is a normative argument, which in the history of British political and legal thought is supplied by Bentham's utilitarianism.

For Bentham, the way to calculate overall utility is to have the masses elect representatives to Parliament who will translate the preferences of the majority into legislation. It is thus illegitimate for judges to superimpose their moral views on law that has a content that in fact reflects the outcomes of legislative calculation. As I have argued elsewhere, the positivist distinction between law and morality, or Craig's distinction between a theory of the rule of law and political theory, is not best understood as making a claim that there is no necessary connection between law and morality.Footnote ²⁹ Rather, it is a relic of Bentham's utilitarianism, which for reasons of political morality argued that the content of the law should be identified by factual tests. Hence, a substantive political morality – democratic utilitarianism – lies behind the positivist tradition's commitment to the Sources Thesis.

We can now appreciate why legal positivism is committed to a particular understanding of form as pedigree. If judges and other legal officials are to avoid imposing their moral views on the content of statutes, they should stick to tests that have to do not only with “the manner in which the law was promulgated (was it by a properly authorised person, in a properly authorised manner, etc.)”,Footnote ³⁰ but also which sustain a claim about what that person or set of persons as a matter of fact intended. Similarly, when it comes to form as legality, whatever principles of legality are identified, their function is to help in ensuring that a factual content is produced, identified, and implemented.

The way that the Sources Thesis shapes contemporary legal positivism's understanding of form creates two major problems. First, the view of law as unsettled when it cannot be determined by factual tests has the result that large areas of what public officials do, including judicial supervision of other public officials, must be described as a discretionary or law-making activity because the officials have to rely on non-factual considerations. These are from the positivist perspective areas of official arbitrariness, ungoverned by law. Such a result is intolerable to judges and other public officials, and so, if they are drawn to a positivistic conception of law and the rule of law, they will deny that they have discretion and try to deploy tests for determining the content of the law that correspond to the Sources Thesis. Contemporary legal positivism's life in practice is thus in severe tension with its theory of that practice.

Second, the Sources Thesis permits laws with a content that is in gross violation of form as legality. A law that clearly requires retroactive punishment is perfectly valid from the positivist perspective, but much it violates form as legality. Now Raz claims that this feature of legal positivism is a virtue of the theory because it shows that law does not, as Fuller claimed, have to conform in any significant way to legality.Footnote ³¹ But this claim is in tension with the guidance function Raz attributes to both law and to the rule of law to which we saw Craig refer – that the law should be “sufficiently clear to guide an individual's conduct so as to enable a person to plan his or her life”.Footnote ³² As suggested above, on a positivist understanding of the principles of legality, they exist in order to make law a more effective instrument of guidance. But, in order for law to guide in this way, it has to conform to both the Social Thesis and the principle against retroactive punishment. A judge faced with a law that seemed to require retroactive punishment and who regarded it as her duty to uphold the rule of law must seek to show that the content of the law conforms to legality despite its apparent non-conformity. Put differently, in a case in which the content of the law seems to conflict with legality, a legal problem arises that a judge must try to resolve in an exercise of interpretation driven by a sense of fidelity to principles of legality. The issue here is not that what I called above “form as fact” is unimportant, but that the facts of the law are facts disciplined by the public law form.

If such facts seem incongruous with that form, the problem that arises is a moral as well as a legal problem. The interpretative exercise to determine such facts involves moral acumen because it requires delicate choices to be made between reasons for taking seriously what appears to be a clear legislative statement and the reasons that underpin maintaining respect for the rule of law. Indeed, Raz recognises that the rule of law is a necessary condition of individual autonomy or dignity as these values require the ability to plan ahead and law that conforms to the rule of law is an essential basis for such planning.Footnote ³³ But, since he thinks that it is obvious that the law can on its face violate dignity and autonomy, and, moreover, can do so whilst conforming meticulously to the rule of law, he also thinks that compliance with the rule of law is not sufficient for law to have any moral quality to it.Footnote ³⁴ However, as I shall now show, his argument on this score is riddled with ambiguities that also plague Hart's legal positivism.

III. Fuller's Formalism

Craig does not refer in his essay to Fuller, whose analysis of the rule of law is one of the major contributions in the twentieth century to our understanding of the idea of government according to law. Fuller set out eight “desiderata” or principles of the rule of law: generality, promulgation, non-retroactivity, clarity, non-contradiction, possibility of compliance, constancy through time, and the one which Fuller took to be the most complex, congruence between official action and declared rule.Footnote ³⁵ A system that fails completely to meet one of these requirements, or fails substantially to meet several, would not, in his view, be a legal system. It would not qualify as government under law – as government subject to the rule of law. Fuller's claim is that compliance with the eight principles expressed in the desiderata imbues law with an “inner” or “internal” morality. He thus argues that compliance with the principles makes a positive moral and substantive difference to all legal systems. Even a tyrant who wanted to govern through the medium of law would have to comply with the principles, and this would preclude rule by arbitrary decree and secret terror, which, Fuller says, is the most effective medium for tyranny.Footnote ³⁶

Now Fuller worked up these principles from a fable of a sovereign, Rex, who wished to govern his subjects in accordance with law, but whose ineptness in various respects leads to the identification of one of the principles. By and large, these principles are accepted as items that should figure on any list of rule-of-law principles – for example, an equivalent of each appears on Raz's own list. But Raz rejects Fuller's claim that compliance with the principles gives law a moral quality and also finds it necessary to add to the list.

Raz rejects the claim about moral quality because, on his account, the principles of the rule of law serve to make law into a better instrument for transmitting to legal subjects the content injected into the law by those with authority to make law. Compliance with the rule of law thus makes the law a more efficient instrument of rule. But, since law that so complies merely transmits the content as determined by the rulers, the content can be morally atrocious as well as morally meritorious.Footnote ³⁷ In Kelsen's dramatic formulation, law can have any content.

Raz's additions to the list have to do with the interpretation and implementation of the law: that judges should be independent and impartial, that officials including judges should observe the principles of natural justice, and so on.Footnote ³⁸ Further, Jeremy Waldron has elaborated an even longer list of such “procedural principles” in order to contrast them with what he calls the “formal” principles on Fuller's list.Footnote ³⁹ However, as Waldron notes,Footnote ⁴⁰ Fuller thought that such principles are implicit in the principle of congruence. Indeed, their presence there is a large part of the reason Fuller took the principle of congruence to be the most complex of the eight.Footnote ⁴¹ And, as Waldron also notes, it is clear from Fuller's other work that he had such principles and more in mind, notably in his essay “The Forms and Limits of Adjudication”.Footnote ⁴²

Waldron argues, however, that Fuller made a couple of mistakes that obscured his contribution to our appreciation of the rule of law. First, Fuller himself called the principles on his list “procedural” in order to provide a contrast with “substantive”, thus failing to see that there are two different contrasts in play – the procedural/substantive contrast as well as the formal/substantive contrast. This “patent” error is not merely terminological, since Fuller neglected the procedural dimension in his debate with Hart about what law is and the relationship of law to morality. Fuller's focus on “legislative form” rather than “judicial procedure” allowed, Waldron says, Hart to “set the agenda” in that debate because if we are to establish a connection between law and morality, it is more likely that we will find such a connection in the institutional procedures of the law than in substantive theories about how to interpret the content of legal rules.Footnote ⁴³

It is not altogether clear why Waldron thinks that the formal features of law are so lacking in resources to establish a connection between law and morality. As he points out, Fuller argued that “to embark on the enterprise of subjecting human conduct to the governance of rules involves of necessity a commitment to the view that man is, or can become, a responsible agent, capable of understanding and following rules, and answerable for his defaults”. Further, “every departure from the principles of law's inner morality is an affront to man's dignity as a responsible agent” – for example, to judge someone by “unpublished and retrospective rules, or to order him to do an act that is impossible, is to convey … your indifference to his powers of self-determination”.Footnote ⁴⁴ Waldron suggests that the connection Fuller establishes here between “his formal principles and dignity can be said to apply even more to the connection between procedure and dignity” because the “essential idea of procedure” is that there are two parties who confront each other in open court before an impartial judge who is under a duty to attend and respond appropriately to their arguments and the evidence.Footnote ⁴⁵

These remarks occur in a little-read chapter of Fuller's book, “The Substantive Aims of Law”, and, as Waldron acknowledges, there is a full treatment of the procedure–moral connection in the “The Forms and Limits of Adjudication”. There, Fuller emphasises the importance of impartial adjudicators in a rule-of-law order, in which the issues submitted to the adjudicators “[tend] to be converted into a claim of right or an accusation of guilt. This conversion is effected by the institutional framework within which both the litigant and the adjudicator function”.Footnote ⁴⁶ The process of reasoned argument involved requires that the claim be presented as more than a “naked demand”. It has to be presented as a “claim of right” – that is, as “supported by a principle”. And that has the consequence that “issues tried before an adjudicator tend to become claims of rights or accusations of fault”.Footnote ⁴⁷ Thus, Fuller regards courts and other adjudicative institutions as “essential to the rule of law”. The “object of the rule of law is to substitute for violence peaceful ways of settling disputes. … There must be some agency capable of determining the rights of the parties in concrete situations of controversy”.Footnote ⁴⁸

Notice that, in order for this adjudicative conversion process to take place, the enacted law has to be convertible and that requires a prior conversion process. The latter process involves the reduction of a political programme to the explicit terms of a statute and thus a conversion of policy into public standards, which produces a kind of legal surplus value. By this I mean that the legitimacy of official action in compliance with the statute does not simply stem from its compliance with a political policy that the demos or polis has determined to be appropriate. It is also the case that this conversion process adds value because it brings into being a particular type of public standard – one that permits the operation of the principles identified by Fuller as the desiderata of the inner morality of law, and which enables claims of right based on legal principle to be adjudicated. And, if, as is often (even usually) the case in the administrative state, the statute delegates in large part to public officials the task of developing the policy of the statute, the officials will be responsible for producing in appropriate form the public standards that will govern their administrative regime. Hence, there is also often an administrative conversion process that mediates between statute and judicial review.

In a democratic political order, we should then distinguish between democratic legitimacy and legal legitimacy. The former has to do with the fact that what we can think of as compulsory public policy (the public standards that govern a society) has been determined by the representatives of the people in an open deliberative political process – a legitimacy which might be taken to be transmitted to the policy developed by public officials who have been delegated that task. Legal legitimacy, in contrast, has to do not with the political warrant for the policy, but with its legal warrant – the fact that that policy has been put into legal form.

If legality has is own legitimacy, it follows that the law of an undemocratic legal order will be to some extent legitimate. We might think that legal legitimacy does not suffice for legalised public policy to be fully legitimate, because full legitimacy requires that the policy be made democratically, but still recognise that legalised public policy has some legitimacy just because it has been legalised. But, conversely, we should also think that democratic legitimacy does not suffice for full legitimacy, since democratically determined policy has to be put into legal form before it is fully legitimate.

Now legal positivists are officially committed to denying that legality has its own legitimacy. But they agree, although at times ambiguously, with most of the elements of the Fullerian case for the legitimacy of legality. First, and unambiguously, there is the idea that the institutions of government in a legal order are legally constituted artificial persons. As Hart was to note in 1958, there is no such thing in a legal order as an “uncommanded commander” because “nothing which legislators do makes law unless they comply with fundamental rules specifying the essential law-making procedures”.Footnote ⁴⁹

Second, legal positivists agree that the law made by the institutions of government has to live up to principles of legality and it is here that ambiguities arise. At times, positivists seem willing to accept that compliance with the principles does give law a moral quality, but they assert that law does not have to comply to any great extent with such principles so might lack the moral quality that compliance creates. At other times, they suggest that the compliance has to be quite full but they deny that compliance leads to law having any moral quality to it because, as Hart insisted, it is “compatible with very great iniquity”.Footnote ⁵⁰ Put differently, since law can have any content, and since full compliance serves to transmit that content more effectively, compliance with the rule of law can make things morally worse.

However, Hart considerably qualified that insistence in T he Concept of Law, chapter 9, “Laws and Morals”. In that chapter, Hart says that “a minimum of justice is realised whenever human behaviour is controlled by general rules publicly announced and judicially applied”. If law is to function as a system of social control, its rules “must be intelligible and within the capacity of most to obey, and in general they must not be retrospective”.Footnote ⁵¹ His Hobbesian argument is that both law and morals have to have a minimum content if they are to “forward the minimum purpose of survival which men have in associating with each other” and that, “in the absence of this content”, men would have “no reason for obeying voluntarily any rules”.Footnote ⁵² The point about reason is important because we have to understand law as operating with reasons, not causes.Footnote ⁵³ These reasons must be addressed at least to the natural facts of human vulnerability, approximate equality, limited resources, and limited understanding and will.Footnote ⁵⁴ This “natural necessity”, he thus says, has to qualify the “positivist thesis that ‘law may have any content’”.Footnote ⁵⁵ Moreover, when judges interpret the law, they should avoid “arbitrary or mechanical” choices and seek to display what Hart calls “the judicial virtues”: “impartiality and neutrality in surveying the alternatives; consideration for the interest of all who will be affected; and a concern to deploy some acceptable principle as a reasoned basis for decision.”Footnote ⁵⁶

Hart, therefore, accepts that law has to be convertible in the ways Fuller's theory suggests – it requires the conversion of public policy into legal form and that process in turn makes possible the conversion of law into claims of right. In particular, the form that law takes in the first process or processes makes it possible for particular judgments about how the law applies to individuals to be presented to the individuals as offering them reasons that respond to their interests. Moreover, the officials who are charged with the ultimate responsibility for showing that law is responsive in this fashion – the judiciary – must seek to show how a particular law is reasonable in that it offers reasons responsive to the interests of legal subjects. In doing so, the judges search for some general principle drawing on the resources offered by the law beyond the law that directly is in question. Put differently, the judges must seek to show that the answer given to the individual is one that displays the full law in its best light.

To put things this way is provocative both because it seeks to make Dworkinian interpretivism – a substantive theory of adjudication – fall out, as it were, from law's form, and because it attributes endorsement of such a theory to Hart. The provocation is minor when it comes to Fuller, as he had set out the main elements of such an interpretive theory well before Dworkin developed his interpretivism as a full-blown challenge to Hart's legal positivism.Footnote ⁵⁷ The provocation is more dramatic when it comes to Hart and legal positivism in general.

However, the drama has little to do with disagreement about law. In this respect, the only real disagreement is in regard to the positivist claim that official discretion is required to resolve unsettled law. But, in retrospect, debate about this issue might appear a tremendous waste of energy. The debate makes sense if with Bentham one argues that judicial interpretation should be marginalised to the extent possible in legal order because, from the perspective of democratic utilitarianism, such interpretation is an arbitrary intervention in the law-making process. But from Austin on legal positivists have conceded to judges a legitimate role in settling questions of unsettled law by resort to standard common law methods of interpretation. And, as Hart's take on the judicial virtues shows, discretion seems to vanish from the positivist vocabulary when it comes to describing what judges do. As I shall now argue, the real disagreement has to do with legal positivism's conception of morality.

IV. The Moral Quality of Formal Public Law

Fundamentally at stake is not positivism's conception of law, but its conception of morality as a set of standards external to law.Footnote ⁵⁸ These standards, “used in the moral criticism of society itself”,Footnote ⁵⁹ provide the tools for evaluating the content of the law. They are the standards of what Hart called “critical morality” in contrast to the “social” or “positive morality” that one finds in social conventions and the law.Footnote ⁶⁰ Since the latter is entirely contingent on what people happen to think, the former might well condemn it as immoral.Footnote ⁶¹

Note that positivists can agree that the standards are external to law, whilst disagreeing about what the standards are – they could be what utility demands (Bentham), what the demos decides (Waldron), or what a liberal political theory prescribes (Raz). And they can also agree that law has to be conceived as a vehicle for content determined outside of the law and that principles of legality are those principles that help to transmit such content. But, as we have seen, contemporary legal positivism cannot stick to this rather flat conception of law in which law's normative force comes from the correspondence of its content with external standards of morality. And it cannot do so because Hart set out to understand the way in which legal order generates a kind of internal normativity or authority. That journey took him very close to developing the main elements of the account of law of his critics, though he tended to obscure this because of the drag of the Separation Thesis on this development.

Consider in this regard that Hart alludes to the basis for establishing the connection between the processes of conversion in his discussion of the legal constitution of ultimate legislative authority. His argument on this score in both “Positivism and the Separation of Law and Morals” and in The Concept of Law proceeds from private to public law. The first step is that law includes legal rules that provide “facilities more or less elaborate for individuals to create structures of rights and duties for the conduct of life within the coercive framework of the law”.Footnote ⁶² The second step is that such “power-conferring”Footnote ⁶³ rules are not confined to rules that confer power on private individuals, since they are to be found also in public law in the rules that confer power on public officials to legislate, adjudicate, and administer.Footnote ⁶⁴ In the public law context, if officials are to legislate, adjudicate, and administer, they must abide by these rules, just as a private individual has to act in accordance with the law of contracts to succeed in contracting. It follows that even the highest legislative authority has to be understood as legally constituted.

Hart thought that this kind of rule can explain not only the way that legislation is a rule-bound or legally regulated activity in the same way that promising is, but also how legislation can be self-binding in the same way as promising. He did note some differences. The rules governing legislation

Compare this passage to Fuller who, after setting out his eight principles of the inner morality of law, says that the morality is better understood in terms of a morality of aspiration than a morality of duty and hence that its “primary appeal must be to a sense of trusteeship and to the pride of a craftsman”.Footnote ⁶⁶ Both Hart and Fuller, in appealing to these private law ideas, evoke the attempts by early modern scholars to come up with a juridical notion of public power that appealed to ideas from Roman private law. The common thread that ran through these early modern ideas was that the sovereign, the ultimate legal authority, did not have dominium or absolute rights over his power, but rather exercised that power on behalf of another – “the people” who are subject to that power. One of the problems they encountered was that, in private law, if a usufructuary goes beyond his right to use the fruits of someone else's property or a guardian does not act in good faith to serve the interests of his ward, a judge could be asked to enforce the rights of the owner or the interests of the ward. When translated into public law, these ideas can no longer be understood as rules of enforceable positive law or lex, so they are understood in terms of unenforceable standards of public right or ius, which boils down to the sovereign's sense of what will best secure the safety of the people. Salus rei publicae suprema lex – the safety of the state is the supreme law – becomes in later theorists the organising principle of public law.Footnote ⁶⁷ The sovereign is recognised as the ultimate interpreter of what the safety of the people demands, perhaps with a nod (as in Locke) in the direction of “the people” asserting their judgment in an act of revolt against a sovereign who steps too far out of line with popular sensibilities.

An additional problem is that, at most, private law provides a fruitful source of analogies for trying to understand the public law relationship between ruler and subject as a reciprocal rather than a unilateral relationship. Any attempt to go further, to try to understand public law in private law terms, risks imposing the inherent libertarianism of private law on public law by restricting public authority to two roles: clarifying though without distorting the corrective justice framework of private law and acting as the enforcement mechanism for the reciprocal relationship. And, to the extent that public law cannot be understood in these private law terms, it seems to be envisaged as an unconstrained realm of distributive justice.

Something like that result might also flow from Dworkin's interpretivism. If the principles that underpin the full law have to be understood in terms of “right as trumps”,Footnote ⁶⁸ then, to the extent that public law cannot so be understood, that is, to the extent that public law is not a matter of judges protecting entrenched constitutional rights, it is a realm of policy decisions with legal effect that escape the discipline of judicially guarded principle.

The demotion of public law to this lesser status in theories that accord a special status to private right in their understanding of legal form is, in my view, a relic of the classical natural law argument that iniusta lex non est lex – an unjust law is not a law. Law that does not correspond to the justice of private law is not law properly so called. However, public law of this sort is ubiquitous, entrenched, and the law that most affects people in their day-to-day lives. Moreover, it has been “colonising” private law regimes for decades. So it is difficult to declare it not law by some standard external to it, even if the standard is argued to be internal to law properly so called – to the regimes of private law.Footnote ⁶⁹

However, iniusta lex non est lex can have a more fruitful afterlife in public law if one adopts a different and more complex account of morality from that suggested by Hart's idea of authentic morality as critical morality – the external unchanging, universal standards for evaluating the contingent, variable, and often immoral content of the positive law.

This conception of morality is different because it identifies another set of moral standards in the internal morality of law. And it is more complex because of the argument that the substantive content of public law is authentically moral. It is the product of contingent agreement, but it is also an agreement that has survived the screening of the internal morality of law in both conversion processes. The content of the law is thus more than Hart's “positive morality”. It has a quality to it that exerts a moral pull on those subject to it which complicates their situation should they judge in good faith that the content of the law fails by what they take to be the external standards of critical morality.

The situation is morally complicated just by virtue of the fact that legal subjects will find themselves in a dilemma about which pull to heed, that of what they take critical morality to be or that exerted by the law. But it is also complicated by the fact that it is not possible completely to disentangle the moral perspectives in play. The moral quality of the content of the law comes about because that content can be presented as offering reasons to legal subjects in a way that makes it plausible for them to understand the law as responsive to their dignity as responsible agents. Even if they think that the content of the law should be very different from the content it has been determined to have, the fact that the content offers them reasons of this sort constitutes a reason within their own perspective to grant the law its moral quality.Footnote ⁷⁰

Moreover, to the extent that those charged with determining the content of the law cannot come up with a determination that offers such reasons, there is a legal as well as a moral problem. Legislators who wish to address the individuals subject to their power as lacking dignity have to adopt one of two strategies. Either they can explicitly state that aim, or they can delegate power to officials that permit the officials to achieve the same end, not because this end is explicitly stated in the empowering statute, but because official implementation of the statute is explicitly stated to be unreviewable by judges. Both strategies use law to place individuals or groups of individuals beyond the reach of the law. But they do so in a way that does not comply with law's form, in the first case by negating generality and its implicit commitment to formal equality before the law, in the second case by ensuring that there is no law with which official action has to be congruent.Footnote ⁷¹ If the legislators adopt one or both of the strategies in not altogether explicit fashion, judges are under a duty to treat the law the legislators make as if it were intended to comply with their legislative ethos. As a result, if the form of law is to some extent respected, to that extent it will be interpretable in a way that respects the dignity of those subject to the law.Footnote ⁷²

This conclusion shows why the principle of publicity exercises a moral discipline on the content of the law and thus, as Fuller argued, why a tyrant who wishes to govern through law will find himself both legally and morally constrained. In addition, it helps to show why it is not merely a contingent fact that where law is present, so there will be interpretive resources available to judges of the sort on which Dworkin staked his challenge to legal positivism. These resources might fall well short of providing the basis for the kind of liberal political theory Dworkin argued was to be found in the full law of the US legal order. But it remains significant that, where law's form is respected, there will be interpretive resources that sustain a claim that law provides a moral kind of reason to those subject to it.

I want thus to suggest that Fuller was right to give formal principles of the rule-of-law primacy in his treatment of the rule of law. It is these principles that give law its moral quality and the procedural principles are principles that are required to put into effect that quality. This is not to say the procedural principles are unimportant. The adjudication of claims of right requires that the legal order have institutions organised around the procedural principles. It is such principles that accord individuals the right to ask an independent official for reasons why the law applies to them in a way that addresses them as persons with dignity as responsible agents, concretely, to ask a judge or other impartial adjudicator “But how can that be law for me?”.

If the adjudicator is able to give an answer to that question, he is able to satisfy what Bernard Williams called “the Basic Legitimation Demand” that every legitimate state has to satisfy if it is to show that it wields authority rather than sheer coercive power over those subject to its rule. In order to meet that demand, he says, the state “has to be able to offer a justification of its power to each subject”.Footnote ⁷³ Williams opposes political theories that make morality prior to politics, either by taking politics to be the instrument of the moral – Bentham and Raz would be examples here – or as structurally constrained by moral principles – here Kant, Rawls, and Dworkin would be examples.Footnote ⁷⁴ However, Williams is willing to countenance that the Basic Legitimation Demand is a moral principle as long as one understands that it is inherent in politics – that is, in the Hobbesian political question of how to secure “order, protection, safety, trust, and the conditions of social cooperation”.Footnote ⁷⁵ This, he says, is the “first” political question because it has to be solved before any other questions can be posed, let alone solved.Footnote ⁷⁶ And, although he barely discusses law, it is clear that he regards law and the rule of law as part of the answer to that question.Footnote ⁷⁷

Whilst I cannot develop the full implications of these ideas for a formal theory of public law here, I think their salience to my argument is clear.Footnote ⁷⁸ A formal theory of public law is independent from moral theory, thus preserving what I called earlier form as independence – the independence of a theory of public law from, say, a theory of human rights. But it cannot be independent from political theory because it is part and parcel of a theory about how to answer the most fundamental political question – how to create and maintain order. Moreover, the part it takes in answering that question explains why the authoritative reasons the law conveys to its subjects have a moral quality to them. That is, the state's decisions found in the content of legal law will have the quality that comes about through the conversion processes sketched above.

Law of that sort will fulfil the guidance function identified by Raz but not quite in the way he sees. My account recognises the importance of the guidance function and that law must be determinate in order for law to be capable of exercising that function. Both of these features are essential to any account of law as serving the interest of legal subjects in being treated as subjects or agents rather than objects. However, the authority of law does not come about because, as in Raz, the content of the legislators’ views injected into law coincides with the demands of morality.Footnote ⁷⁹ Rather, authority comes about because of the normativity that the legislators’ views accrue through conversion – the surplus value I referred to above.

The content of the law is the content that is produced through legislation, which is quite different from the positivist idea that content is injected into the law and then transmitted from law-maker to legal subject. And, whilst the rule of statute law requires that by and large statutes are capable of guiding individuals prior to any judicial determination, what the individuals take from the content of that guidance is always provisional in that it is subject to the other conversion processes. Indeed, even prior to a judicial determination, the content is so subject since, in a rule-of-law order, namely a legal order, individuals should understand the law that governs their lives in accordance with their expectation that law is legal.

Another way of putting this point is to say that legislators, as public officials who have a law-making role, intend that more than a particular policy be implemented. They intend that their policy be implemented through law, hence legally or in accordance with the rule of law. It is for this reason that Hobbes said that it would be a “great contumely” or insult for a judge to suppose that the intention of the legislature is other than to comply with the laws of nature.Footnote ⁸⁰ “He ought therefore, if the Words of the Law doe not fully authorize a reasonable Sentence to supply it with the Law of Nature.”Footnote ⁸¹ But it follows that legal subjects have a legitimate expectation when it comes to understanding the law prior to judicial interpretation that its content is reasonable in this way. If it does not seem understandable in this way, they will be entitled to argue to a judge or other impartial official that the content of the law is different from what might be taken to be its literal or prima facie meaning.

Following the argument of the last section, if the subject cannot get a plausible answer to the question “But how can that be law for me?”, the situation will be legally as well as morally problematic. The law, as Hart was on occasion inclined to recognise, will make a defective authority claim on the individual and, as he was less inclined to recognize, it will also be legally defective. Such a law might be defective to the point of not counting as law, even though it meets the formal pedigreed tests for what counts as law in legal order, such as a bill of attainder. For formal principles can, on occasion, yield conflicting results. But if such conflict arises, it should be resolved in the way most respectful of the dignity of those subject to the law.Footnote ⁸²