AustLII Home | Databases | WorldLII | Search | Feedback

Macquarie Law Journal

Macquare Law Journal (MqLJ)
You are here:  AustLII >> Databases >> Macquarie Law Journal >> 2004 >> [2004] MqLawJl 9

Database Search | Name Search | Recent Articles | Noteup | LawCite | Help

Stewart, Iain --- "Men of Class: Aristotle, Montesquieu and Dicey On 'Separation of Powers' and 'the Rule of Law'" [2004] MqLawJl 9; (2004) 4 Maquarie Law Journal 187




Article XXX in the first part of the Constitution of Massachusetts 1780 reads like a textbook case:

In the government of this commonwealth, the legislative department shall never exercise the executive and judicial powers, or either of them: the executive shall never exercise the legislative and judicial powers, or either of them: the judicial shall never exercise the legislative and executive powers, or either of them: to the end it may be a government of laws and not of men.

At first sight, John Adams has provided in nuce the scheme that in 1787 would be basic to the Constitution of the USA: a separation of powers, guaranteeing the rule of law. The Article is also, at second sight, nonsense. How will the three ‘departments’ be kept in line if not by each other, which is to suppose that their respective functions overlap - that a decision made by one department can in some sense be changed by another? And how will laws govern men unless men use and, potentially, abuse them? What underlying sense might be traced?


When William Twining toured the Great Juristic Bazaar (stretching beyond the horizon in every direction), he came upon a Pavilion of the Apprentices. The first lesson for apprentices in jurisprudence was entitled ‘Cocktail Party’:

Each student was provided with a list of 100 jurists and, associated with each name, a single word or phrase – for example, Kelsen – basic norm; Savigny – volksgeist; Hart – union of primary and secondary rules. As the title suggested, students were required to circulate and engage in interchanges which took the form of student A dropping the name of a famous jurist and student B responding with a key word, or vice versa. There was a strict system of scoring ... . Copies of the list were on sale in the pavilion and were specially recommended for reluctant teachers of jurisprudence.[1]

The Bazaar is devoted to A General Theory of Law for the Modern Age. However (as Twining would surely recognise), there are similar establishments (with a tendency to set their own horizons) in which one can observe the Cocktail Party game being played in more specific spheres, such as that of public law.

In the study of public law, indeed, the game is also played at an intermediate level – and sometimes, remarkably, by craftsmen. At this level, the player who hears the phrase ‘separation of powers’, for instance, is supposed to run to a wallful of pigeonholes labelled by author and whip from the hole labelled ‘Montesquieu’ a brief passage by the great man. The proper passage is from his famous book The Spirit of Laws:

6. Of the Constitution of England. In every government there are three sorts of power: the legislative; the executive in respect to things dependent on the law of nations; and the executive in regard to matters that depend on the civil law.

By virtue of the first, the prince or magistrate enacts temporary or perpetual laws, and amends or abrogates those that have been already enacted. By the second, he makes peace or war, sends or receives embassies, establishes the public security, and provides against invasions. By the third, he punishes criminals, or determines the disputes that arise between individuals. The latter we shall call the judiciary power, and the other simply the executive power of the state.[2]

To produce this passage is sufficient to gain the point. Half a point is given for being able to summarise it plausibly. An extra half point is available for observing that this Frenchman knew woefully little about England.

In response to the phrase ‘the rule of law’, the answer that gets the banana is ‘Dicey’ plus reference to a chapter of Introduction to the Study of the Law of the Constitution.[3] Since Dicey’s discussion there is rambling, a mixture of summary and quotation is deemed to suffice. There is an extra half point for adding that his fears about ‘administrative law’ have proved unrealistic.

The remarkable thing is that so many craftsmen manage to score with these passages alone or summaries of them, even after being admitted as masters. Textbook after textbook of public law repeats them without locating them within the author’s œuvre, still less his life and times.[4] With the Montesquieu passage, it is even evident that our master craftsmen have not troubled to read the chapter or even the section from which it is taken – for Montesquieu quickly goes on first to modify this threefold scheme and then to reject it in favour of another. And this superficiality about Montesquieu and Dicey is a serious business, because the masters have been taken at their word both by the framers of modern constitutions and constitutional reforms and by supreme courts in constitutional interpretation.


So, had our editors’ invitation to discuss ‘the way in which the principles of dividing and balancing power can be used to advance rule of law values’ been set as an essay topic, it would surely have been condemned as a stinker. The student expects to have to show that they understand the terms of the question, but not to be confronted with a question whose terms are inadequately explained in the reading material. Perhaps juristically we should give up on both ideas.

Jennings was one who declined to join in the permanent ovation for either of them. He classified the idea of separation of powers as a policy rather than a constitutional principle:

The existence of an elected legislature necessarily implies a separation of powers, not because it is possible to distinguish functions of government into three classes, but simply because an assembly is not a suitable body to control detailed administration or to decide whether the laws have been broken or not.[5]

As to the idea of the rule of law, after examining what Dicey and others had meant by this phrase or apparently equivalent expressions such as Rechtsstaat Jennings was inclined to give up:

The truth is that the rule of law is apt to be rather an unruly horse. If it is only a synonym for law and order, it is characteristic of all civilised States; and such order may be based on principles which no democrat would welcome and may be used, as recent examples have shown, to justify the conquest of one State by another. If it is not, it is apt to express the political views of the theorist and not to be an analysis of the practice of government. If analysis is attempted, it is found that the idea includes notions which are essentially imprecise. If it is merely a phrase for distinguishing democratic or constitutional government from dictatorship, it is wise to say so.[6]

The problem with the idea of the rule of law is that it seems to be a juristic chocolate factory, a category with no definite content apart from law itself and hence open to almost any content. Thus Raz complained in 1977 that the already classic New Delhi Declaration of the International Commission of Jurists 1959[7] mentions or refers to ‘just about every political ideal which has found support in any part of the globe during the post-war years’.[8] He spiced this criticism by proposing for theoretical purposes an entirely formal concept, which in application might identify occasions when the content of a claimed ‘rule of law’ would be morally deplorable.[9] Perhaps, as has been lamented more recently, the idea of the rule of law is just ‘the will-o’-the-wisp of constitutional history’.[10] Or possibly things are worse and the idea is a platonic ‘noble lie’ in need of rescue through an immanent critique which would remould it upon its claimed values.[11] Or are those values nonetheless so elusive that what purported to be an immanent critique would involve heavy importation? That would seem to be the main risk if one were tactically to treat ‘the rule of law’ as equivalent to ‘Rechtsstaat’,[12] if already by 1934 it could be said that the word Rechtsstaat ‘is as worn out as a coin whose relief has become almost unrecognizable through daily use’.[13] Certainly if, as Neumann goes on to recount, the claim that any preferred form of state was a Rechtsstaat was an act into which even Nazis were trying to get. However, Rechtsstaat as usually understood is a substantially different concept and I shall return to it.

The value of a ‘rule of law’ depends, moreover, on the value of law; and that depends on how law is conceived. When famously pronouncing that the rule of law is ‘an unqualified human good’,[14] Thompson drifted toward a very broad conception of law – in effect, all binding social rules, whose value is nearly identical with that of society itself. But the expression ‘the rule of law’ ordinarily assumes a much narrower conception of law, the binding norms issued by the state, and the value of the state is disputed. Even if law in general has a positive value, equality before it – which for Dicey was central – is valueless unless the particular laws are good. There is substantive equality before the law if, for instance, every request that could legally be granted is refused.[15] And there is procedural equality (not to mention despatch) if all petitioners are commanded to remain silent so that they may be heard in turn.[16] The issue is not the value of the content that is given to the idea of the rule of law, but why those elements should be given that form.

Maybe, though, the ideas of separation of powers and the rule of law fall apart only when they are stretched too far. Perhaps they continue to make sense within the smaller confines of individual cases for decision or of juristic commentary upon the judgments. If so, one could expect the best sense in the lands with most experience of them. But, as to the rule of law, his homeland still has not got over Dicey[17] and the jury is still out on how far England has ever known (or wanted to know) separation of powers.[18] In the land of the free, one is told in 1991 that among constitutional scholars there is ‘near unanimity’ that ‘the Supreme Court’s treatment of the constitutional separation of powers is an incoherent muddle’; it is not that the Court ‘has gotten it wrong’ but that the Court ‘has not really “gotten” it at all’ – the Court ‘has adopted no theory, embraced no doctrine, endorsed no philosophy, that would provide even a starting-point for debate’ on separation of powers.[19] To this reputation, the Court continues to live down.[20] A more recent survey agrees: ‘on the whole, both the scholarship and jurisprudence on separated powers is marked by its inconsistency and lack of synthesis’.[21] In the USA, Kairys maintains, the idea of the rule of law remains in ‘an incredible fuzziness’.[22] Yet both there and globally, he goes on (and I shall discuss later), the main problem with the concept is not over-inclusion but exclusion.

Framer, reformer, judge and scholar, it seems, are all in a pickle that is far from pretty. All seem basically not to know what they are talking about. To investigate national traditions here would take impermissible space. Instead, to concentrate on Montesquieu, Dicey and their most originary predecessor will both respond to the juristic literature and be an economical path into key issues.[23]


Is there a Square One for either of our ideas? Not a first occurrence, which it is usually fruitless to seek, but an earliest point back to which current debate appears to hark – the first square in the game being played?

The terminology of the three ‘powers’, at least, is as old as Isaiah: ‘the Lord our judge, the Lord our lawgiver, the Lord our King – he himself will save us’.[24] This being the opposite of separation, it can serve us as little more than a reason to abolish God. Yet a Square One can be found for both of our ideas, as for so much else, in the works of the Philosopher.

Aristotle divides the sciences into the theoretical (the study of how things are), the practical (the art or craft of getting things done) and the productive (the art or craft of making things). He classifies politics as a practical science. His political science is thus a search (although through empirical evidence) for an image of the ideal state (ie a city-state) that will serve as a regulative ideal in political action. The arguments relevant here appear in the empirical phase of his inquiry.

In terms of Aristotle’s theory of causality, a state has four types of cause. The material cause of a state is its citizens (together with its natural resources), the formal cause is its constitution (politeia), the efficient cause is its ruler or ruling group and the final cause is the good at which the state aims. The constitution is not a document but the organising principle of the state seen by analogy as an organism; it is implicit in the citizens’ ongoing way of life.

The constitution, as formal cause, determines what is to be the efficient cause, the form of government.[25] In his Politics,[26] Aristotle identifies three ‘true’ forms of government. They are distinguished according to whether government is by one, by a few or by many. These forms of government are, respectively, monarchical, aristocratic and constitutional (that is, where the constitution and citizens’ behaviour directly coincide). All of them are government for the common good. The respective ‘deviations’ from them are tyranny, oligarchy and democracy. These are government solely in the interest of the monarch, of the rich or of the poor.[27] Of these, tyranny – being the deviation from the most ‘divine’ form of government – is the worst. Oligarchy is little better, while democracy is the most ‘moderate’ of the three.[28]

The key issue about monarchy, for Aristotle, is whether it should be absolute or limited by law. If the absolute ruler is actually a man of supreme excellence it will be unjust to subject him to law, which would be to treat him as equal with his inferiors.[29] On the other hand, monarchy limited by law is not really a form of government but rather a ‘lifelong generalship’.[30] Yet, since absolute monarchy tends to turn into tyranny, the option of a monarchy limited by laws should be seriously considered. The question is ‘whether it is more advantageous to be ruled by the best men or by the best laws’.[31]

Aristotle summarises the answer of those who fear that absolute monarchy will descend into tyranny. For them, ‘it is preferable for the law to rule (nomon archein) rather than any one of the citizens’. Law (nomos) should rule (or govern) (archein), because law is rational:

He ... that recommends that the law shall govern seems to recommend that God and reason (nous) alone shall govern, but he that would have man govern adds a wild animal also; for appetite is like a wild animal, and also passion warps the rule even of the best of men. Therefore the law is wisdom (nous) without desire.[32]

The divine, the rational and the just coincide: ‘when men seek for what is just they seek for what is impartial; for the law is that which is impartial’. Aristotle immediately adds, apparently supposing that customary norms are the most impartial of all: ‘Again, customary laws are more sovereign and deal with more sovereign matters than written laws, so that if a human ruler is less liable to error than written laws, yet he is not less liable to error than the laws of custom.’[33]

Aristotle assumes, however, that under an aristocratic or constitutional form of government written laws will have most weight: for governance has to proceed in an organised form. Written laws, nevertheless, neither supplant a constitution nor even set out its principles. Rather: ‘a constitution is the regulation of the offices of the state in regard to the mode of their distribution and to the question what is the sovereign power in the state and what is the object of each community’, while laws ‘are distinct from the principles of the constitution, and regulate how the magistrates are to govern and to guard against those who transgress them’.[34] In this case, he implies, if there is a coincidence of the divine, the rational and the just it is located in the constitution.

However, a complete coincidence – in the best constitution – will not occur through either leadership or legality. It will occur when all the citizens are virtuous.[35] Since, Aristotle acknowledges, this must be an aspiration rather than a realistic aim, it is necessary to consider the second-best constitution. This will be either constitutional government or a mixed constitution. The latter will combine ‘features of democracy, oligarchy, and aristocracy, so that no group of citizens is in a position to abuse its rights’.[36] This notion of the mixed and balanced constitution was to reach down through Montesquieu and beyond.

The idea of a rule of law can thus be found in Aristotle, but not – as is sometimes supposed – almost the very expression. Jennings, who makes this attribution, is misled by Jowett’s translation: ‘The rule of the law is preferable to that of any individual’.[37] Jowett translates nomon archein as ‘the rule of the law’, whereas we have seen that Rackham more literally has ‘the law to rule’. Dicey studied classics under Jowett, a renowned scholar whose translation of the Politics appeared in the same year as Dicey’s Introduction. It would be intriguing to learn whether either man influenced the other here. What is found in Aristotle, however, is not Dicey’s notion of the subjection of all equally to the law but only his derived point (even if it was of greater concern to him), a subjection of rulers to law – the idea of a ‘government of laws and not of men’. This is not yet Dicey’s doctrine of the subjection of rulers and ruled alike to ordinary positive law. Aristotle envisages, rather, the subjection of all to what he nebulously conceives as a constitution and of the ruler to specific (one would say today, specifically constitutional) positive-legal norms. I will refer to Aristotle’s doctrine as that of a ‘government of laws’.

A distinction between legislative, executive and judicial functions is also found in Aristotle. However, it occurs quite discretely:

All forms of constitution ... have three factors in reference to which the good lawgiver has to consider what is expedient for each constitution; and if these factors are well-ordered the constitution must of necessity be well-ordered, and the superiority of one constitution over another necessarily consists in the superiority of each of these factors. Of these three factors one is, what is to be the body that deliberates about the common interests, second the one connected with the magistracies, that is, what there are to be and what matters they are to control, and what is to be the method of their election, and a third is, what is to be the judiciary.[38]

The deliberative factor is ‘the sovereign power in the constitution’.[39] It ‘is sovereign about war and peace and the formation and dissolution of alliances, and about laws, and about sentences of death and exile and confiscation of property, and about the audits of magistrates’.[40] The magistrates (archons) are the officers of state, its managers. In a large state the functions to be performed may be divided among a large number of officials, but in a small state several functions may have to be given to each official.[41] The judiciary determines disputes, except that the first of Aristotle’s eight types of law court is a ‘court of audit’[42] – which today would be counted, in the shape of an Auditor-General, as an executive organ. Aristotle goes on to examine the variations on these themes to be found in the different forms of government.

This is a doughnut of an exposition, formed around a central obscurity as to the logical status of the ‘three factors (tria moria)’. The language is concrete: a morion is a component part, a piece.[43] The ideas denoted, however, are abstract. The three factors are empirical generalisations from Aristotle’s constitutional research.[44] Having stated the three as such, he proceeds to elaborate on the varieties of their realisation. Simultaneously, he shifts from the empirical to the practical, so that those varieties appear as political options. The concreteness of the language persistently tempts one to think of three discrete organs or sets of organs – yet Aristotle clearly states that the three factors may be allocated to a single organ or distributed among a plurality of organs or in various ways combined.

This is a division of labour in the political community. It is not specifically designed for checking and balancing, although some of that will inevitably follow from the mere fact of division. The deliberative body is necessarily unitary: the men must be together in order to deliberate. But it cannot be in permanent session and therefore must delegate or at most devolve day-to-day executive and judicial functions to the magistrates and the judges. The deliberative body does not divest itself of these functions, nor do the magistrates and judges have any authority over the deliberative body. The deliberative body will check the activities of the magistrates and judges – not because it will compete with them but because it will be their boss. Magistrates will be subject to the judges as to their private actions, but as to their actions as magistrates it is not suggested that there will be any recourse except to the deliberative body.

Aristotle is concerned, however, that in this distribution of tasks all classes shall get a share and in that sense there will be a balancing of class interests.[45] Albeit that the poor may participate only in deliberation and adjudication, and then only because otherwise they would surely turn into enemies.[46] There is, here, a further principle of distribution of power.

As to allocation of powers, then, Jennings and Aristotle are in agreement that it is a matter of political policy and not of legal principle.

Today, the options as to the overall form of government have been transformed. For Aristotle, there are three options: rule by one, by a few or by many. Each of these possibilities has both a true and a deviant form. He ignores the logically evident fourth possibility, rule by all.[47] His ‘democracy’ is only a perverse form of rule by many, that in which the many are the poor. There is logic in this. Royalty, an aristocracy and rich commoners all have an economic interest which they wish to maintain. The principal economic interest of the poor, however, is to abolish poverty. In terms of class identity, therefore, government by a monarch, by aristocrats or by wealthy commoners can be stable, while government by the poor cannot be, since in abolishing poverty the poor would abolish themselves as a class. Yet the question of stability here is only logical, only about class identity. Underneath it remains the question of whether to maintain the relation of exploitation.


The famous passage in Montesquieu, and his further discussion, can be read as an update on Aristotle. But is Montesquieu’s take on the issues worthwhile?

On the evidence of that passage, Charles-Louis de Secondat, Baron de la Brède et de Montesquieu (1689-1755), is open to accusation upon four counts – two of ignorance and two of naïvety. He may be accused: of particular ignorance, in that he seriously misunderstands the British constitution;[48] of general ignorance, in that he was just an unworldly scholar; of theoretical naïvety, in that he thinks of three powers that are ‘separate’, which could never work; and of empirical naïvety, in that the coverage of his three powers omits the sphere of domestic administration – the very sphere, some would think, that a separation would be most designed to control. I will argue that he is innocent on all of these counts and guilty of something different.

A Particular Ignorance

As to particular ignorance, of English politics and law,[49] I have shown elsewhere that Montesquieu knew English politics quite well and something about English law. He spent a year and a half in England and was welcomed in the highest circles. His interpretation of the British constitution was acclaimed by British contemporaries.[50]

B General Ignorance

He first achieved fame as a novelist, with the Lettres persanes (Persian Letters) which gained him admission to the Académie Française. But he did not, like Montaigne, retire to his château. On the contrary, he took a European tour on which his status as author and académicien gained him access to practically any ruler or minister he wished to meet. He then spent as much time as possible in the salons of Paris, where he would have found a great deal of politicking, both intellectual and raw. Indeed, he may in a covert way have achieved his early ambition to become a diplomat. Not only his surviving manuscripts and letters[51] – notably those to and from the Young Pretender – but also mysterious matching gaps in his own Nachlass and in those of influential English friends, Lords Chesterfield and Hervey, suggest that he had very active contacts not only with French political figures but also with leaders on both the ruling and the Jacobite sides of British politics. This was hot stuff – in this period Britain and France were occasionally at war - and some of his letters contain suggestions of caution. It seems quite possible that Montesquieu, loyal Frenchman and (at least nominally) Catholic, was an active liaiser. If so, however, his degree of success is unknown.

There is a PhD topic in these lines of inquiry. Not often does a scholar get to deal with a tale involving missing bodies, missing manuscripts, two Lords Chancellor, international Freemasonry, Bonnie Prince Charlie and Pickle the Spy. One of the missing bodies is Montesquieu’s own: his grave was desecrated during the Revolution – after all, he was a baron.

C Theoretical Naïvety

If one reads a few more pages on from the famous passage,[52] one finds that Montesquieu’s preferred scheme of checks and balances is not the three famous powers but the established English scheme of king, lords and commons.[53] The transition from his initial statement has two stages. He first cuts out the judicial power. He conceives it not as a professional or even permanent body but as a sort of occasional assembly, which he appears to model upon a romanticised image of the witanagemot as the true manifestation of the English judicial ‘spirit’. Not being permanent, it cannot check or balance the legislative power or the executive power. In any case, it has no creative rôle: Montesquieu swallows whole the declaratory theory of adjudication - that judges are merely, he says, ‘the mouthpieces of the law’. The judicial power so conceived is ‘invisible’ and ‘in a way, a nullity’.

Next, Montesquieu distributes both the legislative power and the executive power among all of the three key organs of English government: the King, the House of Lords and the House of Commons. He adopts the traditional English notion of the ‘mixed and balanced constitution’. More specifically, his preferred position is essentially that of the opposition and Jacobite leader, Bolingbroke. Montesquieu’s concern is not that the three powers must each be in separate pairs of hands but that no two of them should be placed in a single pair of hands. He achieves this, and thinks that England achieves it, by placing both of two powers in three pairs of hands. It is a solution to the problem of the tendency of power to corrupt. If the possession of power always tends to encourage bastardry, one must get the bastards to keep each other honest.

Montesquieu applies, that is, the principle of distribution of power which we have seen in Aristotle. Whatever types of power are exercised by a state, each type must be distributed or its exercise will be liable to corruption. By ‘corruption’ he understands a lack not just of honesty but of integrity and, indeed, of strength. That power be distributed is a matter of principle, but how it should be distributed is a matter of local circumstance. Montesquieu does not discuss federations – depending on definition, there might have been none to discuss. However, his principle is applied when legislative, executive and judicial power are all distributed among central and regional authorities, whatever the relationships among those powers might be in the hands of each authority. A generation later, Blackstone took a similar view as to the principle; he was expressing the common understanding of the age.[54]

Montesquieu is therefore not guilty on the count of theoretical naïvety.

By rider to that verdict, however, one should note the mode of his sophistication. Montesquieu shares with Aristotle a commitment to both division of labour and distribution of power. But, while the terms of Aristotle’s commitment are entirely political, Montesquieu’s terms are legalised. Aristotle’s processes of deliberation, execution and adjudication are replaced by types of legal decision: legislation, execution and adjudication. And it is authority to make these types of decision that Montesquieu considers how to distribute.

D Empirical Naïvety

There remains the charge of empirical naïvety, that Montesquieu’s initial scheme has no place for domestic administration – the very thing that today is said to constitute the bulk of ‘the executive’.[55] To address this charge requires close reading.

De l’esprit des lois[56] (fully, On the Spirit of Laws) first appeared in 1748. Anglophones are accustomed to reading it in (or as quoted from) Thomas Nugent’s translation, published in 1751.[57] This translation is fairly accurate: Montesquieu, at least, was pleased with it. It is of the first edition of De l’esprit des lois, but the final text is not very different as to the elements in question here and a new English translation,[58] which is of the final text, is to my mind not a great improvement. Both translations neglect Montesquieu’s usage of terminology taken from Roman law. If one attempts to remedy that defect, a novel possibility emerges.

I shall give the famous passage in French (from the final text)[59] and then in my own translation:

6. De la constitution de l’Angleterre

Il y a dans chaque État trois sortes de pouvoirs: la puissance législative, la puissance exécutrice des choses qui dépendent du droit des gens, et la puissance exécutrice de celles qui dépendent du droit civil.

Par la première, le prince ou le magistrat fait des lois pour un temps ou pour toujours, et corrige ou abroge celles qui sont faites. Par la seconde, il fait la paix ou la guerre, envoie ou reçoit des ambassades, établit la sûreté, prévient les invasions. Par la troisième, il punit les crimes, ou juge les différends des particuliers. On appellera cette dernière la puissance de juger, et l’autre simplement la puissance exécutrice de l’État.

6. On the constitution of England

In every State there are three modes of governance: legislative power, executive power over matters subject to the law of nations and executive power over those subject to the law of the particular state.

Through the first, the prince or magistrate makes laws of temporary or permanent duration and revises or repeals those already made. Through the second, he makes peace or war, sends or receives diplomatic missions, provides security and prevents invasion. Through the third, he punishes crime or judges disputes between individuals. One may call the last the power to judge and the other simply the executive power of the State.

Over ‘constitution’ we need not tarry: Montesquieu has simply transplanted the English usage into French (he was the first to do so) and what he goes on to say does not depend on its meaning. Nugent’s rendering of État as ‘government’ seems to be an attempt to incorporate into the sentence the sense in which pouvoir means ‘government’. I have found a different solution in today’s favoured expression ‘governance’, although that too is not without ambiguity. The phrasing also reproduces the Aristotelean doughnut and that seems to have made it possible to read this scheme as stating the existence of three separate organs.[60] Then the edge of the hole is rendered crumbly with Montesquieu’s overlap of meaning between ‘pouvoir’ and ‘puissance’. Both of them directly translate the contemporary English use of the word ‘power’, by Locke and especially by Bolingbroke; at the same time, ‘puissance’ carries a nuance of ‘force’ as in physics. Montesquieu both follows Aristotle’s empirical views and attempts a more modern scientific method.

Montesquieu’s scheme appears, however, to have a yawning gap: domestic administration. While certainly there is more domestic administration in the modern state than in Montesquieu’s day, even in his day there was a lot of it and it has not yet made an appearance in his scheme. It is evidently not in his judicial power. Nor is it in his executive power, if it be accepted that the category ‘security’ does not stretch so far. Therefore either it is missing from the scheme, which would be extraordinarily naïve, or it is in the legislative power.

One way to handle this is to note that Montesquieu appears to contradict himself, when shortly afterwards he says that when political liberty is at stake:

All would be lost if the same man, or the same body of leaders – whether from among the nobles or from among the people – were to exercise these three powers (pouvoirs): that of making laws, that of executing public resolutions (celui d’exécuter les résolutions publics) and that of judging crimes or disputes between individuals.

The ‘public resolutions’ would presumably be legislation: hence the opening reference to executive power over matters subject to the law of the particular state did after all include an internal ‘executive’ power. (And presumably the external executive power does not appear in this list because it does not ordinarily affect political liberty[61] – so that Montesquieu actually states four powers.) This looks like a glaring contradiction, twice committed. Yet, evidently, to Montesquieu it did not glare. I shall suggest that his discussion is clumsy rather than contradictory.[62]

Let us take Montesquieu’s Roman-law background very seriously. By his day, almost all texts of Roman law had been lost except the Byzantine codifications commissioned in the fifth century by the emperor Theodosius II and, superseding that work, in the sixth century by the emperor Justinian I. In Justinian’s codification, law (ius) is first divided into written and unwritten. Then written law is divided into six types: statute (lex), plebeian legislation (plebiscitum), senate resolution (senatusconsultum), constitution (constitutio), praetorian edicts or honorarian law (ius honorarium) and juristic answers (responsa prudentium).[63] By the sixth century, however, the only type of law still being made was the ‘constitution’. This category, as Justinian defines it, is both autocratic and inclusive. The Roman people has by statute conferred upon the emperor all of its ‘authority and power’, so that whatever the emperor decides has the force of statute (quod principi placuit, legis habet vigorem).[64] An imperial decision can take the form of a letter (epistula), of a judgment after hearing a case or of ordaining something by edict (edictum). Although all these are called constitutions, some ‘are clearly personal and cannot be used to support a general rule because the emperor never intended them in that way’ while others are general and bind everybody.[65]

Neither Justinian nor Montesquieu assumes that a legal order is a deductive hierarchy, in which relatively general norms are applied through the creation of individual or relatively particular norms. Justinian explicitly includes individual norms. Montesquieu’s overall definition of law (loi) as ‘human reason’ privileges the general over the individual:

Law, in general, is human reason, in that it governs all the peoples of the Earth; and the political and civil laws of each nation ought to be only the particular instances where that human reason is applied.[66]

Yet this is to say only that positive law is rational, not that what is rationally established will always be a deductive hierarchy. It is, rather, Locke who in the late seventeenth century insists that the legislative power is a power solely to create general norms.[67]

Montesquieu’s lois and puissance législative are etymologically related to lex (plural leges). But since this power, like the others, is exercised by the ‘le prince ou le magistrat’ the type of law referred to must be the constitutio. And Justinian defines the category constitutio so widely as to include measures of domestic administration. Montesquieu’s ambiguous ‘pour un temps ou pour toujours’ may be read both as I (and Nugent) have translated it or as reflecting Justinian’s distinction between ‘personal’ and ‘general’ constitutions; one would read ‘temps’ as ‘moment’. One may also note that Montesquieu’s puissance législative, although explained as law-creation, is already characterised as a mode of governance (pouvoir), so that law-application might be assumed to be included. These interpretations do not entail that Montesquieu has included domestic administration within the legislative power, but they make it possible to suppose that he was not so dumb as to have left it out of his scheme. Nor is this to suppose that clear distinctions among the powers were known to Roman law, as they were not. It is to suppose only that Montesquieu might not have separated domestic administration out from the legislative power. If domestic administration is part of the legislative power, however, it does follow that, if a particular constitution can be understood in these terms, ‘delegated legislation’ is not a problem of separation of powers: there is delegation within the legislative power, not a delegation from one power to another.[68]

Generations of jurists, therefore, seem not only to have attributed to Montesquieu a scheme that he actually rejected but also to have misunderstood even that scheme. His preferred scheme should be understood as spreading around among King, House of Lords and House of Commons both a legislative power that includes domestic administration and an executive power concerned mainly with foreign relations and defence.

On the count of empirical naïvety, therefore, Montesquieu is not guilty.

E L’aristo

Suspicions of ignorance and naïvety cannot, therefore, affect a further charge which has been too rarely made, that Montesquieu’s account of the ‘powers’ is a bid for constitutional security of his own class, les aristos.

The French nobility – and Montesquieu was proud to declare that his noblesse was centuries old – were losing ground to the rising bourgeoisie. The monarchy, though still absolute, was increasingly appointing bourgeois to office because they were more docile than the nobles. Montesquieu’s adoption of the English constitutional model offers a solution to this problem.[69] Indeed, it is a double solution. It secures the aristocracy, through membership of an upper house whose existence denies ultimate political power to the bourgeoisie, represented in a lower house. At the same time, this scheme secures the existing political system against the common people: it is strengthened against their challenge and shuts them out of its membership.

True, Montesquieu speaks enthusiastically of ‘political liberty’ in this part of The Spirit of Laws and in his ‘Notes on England’. But he means a liberty in which the aristocracy and the bourgeoisie will be free from oppression by a despotic monarch, free from conflict with each other and free from overthrow by what he refers to in the ‘Notes’ as ‘the rabble’.

This perspective permits an answer to the still pending question: if Montesquieu preferred the scheme of mixed and balanced government (gouvernement modéré as he termed it, following Locke), why did he introduce the three-power scheme at all? He does not give a source for that scheme and he was the first to produce just this formulation, but the terminology is ancient. We have seen it in Aristotle and, like him, Montesquieu does not elevate it into a legal principle.

But Bolingbroke had,[70] and Montesquieu’s handling of it may be read as addressing a danger in doing so. If there are three powers and they are in three separate pairs of hands, any one of the powers could be monopolised by a single class. At that time in England, indeed, the judiciary was threatening to become a bourgeois monopoly. And what if the Diggers, Levellers, Ranters and their loud grubby ilk had in the previous century succeeded in dominating the House of Commons? One way to read Montesquieu’s initial scheme, then, is to read it as a straw man, a future that he is determined to prevent. The bourgeois ascendancy in the judiciary then appears as a reason to follow Bolingbroke in pushing the English judicial spirit out of the sphere of government and into the oaken forests of yore. The balance of king, lords and commons is then maintained.

F After Montesquieu

The framers of the US Constitution had no use for the British distribution. They had defeated mad George, they would decide not to have an aristocracy and they would register a distrust of the lowly not within the structure of government but in the system of indirect election. Montesquieu’s initial scheme, however, provided them with what seemed an up to date framework within which the structure of government should be designed. They, and their French successors, gave that scheme two sorts of interpretation, which are today distinguished as ‘formal’ and ‘substantive’. Very briefly: in a formal interpretation there are three quite separate powers which ought not to have anything to do with each other (as we have seen with Adams), while in a substantive interpretation the three powers are functions that are not distributed discretely and hence may check and balance each other.[71] To Montesquieu, complete separation would not work at all and, so far as it was attempted, would so insulate each branch of government from the others as to foster corruption in them all.

Because the Americans also chose a federal system, the federal distribution of power was a different and to an extent a substitute guarantor against corruption. That distribution, however, removed any hope of complete separation within federal government as soon as the federal supreme court asserted itself to be the guardian of the federal balance and, to that end, able to strike down legislation as unconstitutional.[72] Since then, the federal balance itself has been upset by the remorseless transfer of fiscal and consequently economic power from the states to the centre. Sensitivity to this has lain behind the Supreme Court’s recent and vehemently split decision that even the war power does not contain a review-free zone. The majority rejected the government’s plea for such a zone, on the ground that it ‘serves only to condense power into a single branch of government’.[73] Simultaneously, however, the principle of distribution as Montesquieu envisaged its application remains valid for a distribution among a head of state and the chambers of a bicameral legislature, autonomously of the nature of the head of state’s office and of the reasons for bicamerality.


Montesquieu’s balance concerns legislative and executive power. If there is a ‘rule of law’ in that scheme, it is restricted to constitutionality. Far from securing equality before the law, it expressly preserves inequality. Nor, certainly, was Aristotle any friend of the ‘mob’. The idea of ‘the rule of law’ as classically expressed by Dicey, in contrast, is centrally committed to equality before the law. It is necessary to examine whether the ideas of separation of powers and of the rule of law are mutually supportive or, perhaps, even incompatible.

A The Rule

The expression ‘rule of law’, Dicey says, ‘has three meanings, or may be regarded from three different points of view’. The first meaning, or appearance from a first point of view, is ‘the absolute supremacy or predominance of regular law as opposed to the influence of arbitrary power’; it excludes arbitrariness, prerogative and even ‘wide discretion’. The second meaning or appearance is ‘equality before the law, or the equal subjection of all classes to the ordinary law of the land administered by the ordinary law courts’; this excludes privileging officials by allowing them to appear only in special administrative courts or tribunals, such as those of the French droit administratif. Thirdly, the expression ‘rule of law’

may be used as a formula for expressing the fact that with us the law of the constitution, the rules which in foreign countries naturally form part of a constitutional code, are not the source but the consequence of the rights of individuals, as defined and enforced by the courts; that, in short, the principles of private law have with us been by the action of the courts and Parliament so extended as to determine the position of the Crown and of its servants; thus the constitution is the result of the ordinary law of the land.[74]

This scheme is so evidently an expression of English legal piety, and has contributed so mightily to the maintenance of the same, that it is appropriate to begin with the man and his armchair.

B Albert Venn

The central fact about Albert Venn Dicey (1835-1922), according to one of his biographers, is that he was ‘a man whose basic values were fixed in youth, when he imbibed the conventional beliefs of the mid-Victorian generation unquestioningly’.[75] Salient among these beliefs were English nationalism and an enthusiasm for laissez faire.[76] In addition, Dicey’s youth was spent in the ambit of the Clapham Sect, whose mixed spirit – conservative, Anglican evangelical and yet rationalist – can be found throughout his life and works.[77] He may have laboured among dreaming spires, but he lived out his own dreams energetically, with political campaigning both in print and by correspondence. That spirit appears in his passionate commitments to both the existing social order and, within those limits, formal equality. The Clapham Sect had been a leader in the campaign to abolish slavery: William Wilberforce had been its most prominent member. The same attitude was taken to equality of the sexes: witness Dicey’s enthusiasm for the equalisation of married women’s property and, at first, for female suffrage.[78] Likewise Dicey was all for uplifting the virtuous worker – he devoted much time to the Working Men’s College in London. As to the less than virtuous worker, he was keen to see the forces of order use all necessary measures against workers who, failing to appreciate strict equality before the law, favoured an immunity in tort for inducing breach of a contract of employment in the course of industrial action. Dicey the evangelist could not accept shades of grey.[79] His lifelong involvement with the Irish Question exhibits a wholehearted devotion at first to the nationalist cause and then, disturbed by its violence, to that of the loyalists.[80] At the end of his life, he wrote that he had noted ‘the course of English history’ from what ‘I am inclined to call the Whig point of view’.[81] But whiggery was always a spectrum and, as the spectrum was in his day, he had been at both ends of it.

C Dicey’s Method?

His Introduction to the Study of the Law of the Constitution appeared in 1885 and is still in print. The text finalised by himself is that of the seventh edition 1908. An edition of 1959, with an immense Introduction by E C S Wade, went into paperback and is on every constitutional lawyer’s bookshelves.[82]

The book was timely. The forms of action had been abolished only in 1875; English law was only beginning to shed its procedural carapaces. The phrase ‘law of the constitution’ reflects the circumstance that constitutional law was not yet an established category.[83] The Vinerian chair of law at Oxford, to which Dicey was appointed in 1882, had been occupied with little distinction and sometimes in absence since the end of Blackstone’s brief tenure of it – the first chair of common law in England – in 1780. Not that English law wholly lacked science, but it was not securely established as an academic discipline or, indeed, as a discipline even outside the universities.[84] Its condition when Dicey entered it is reflected in the subtitle of his first book, which had appeared in 1879: The Law of Domicil as a Branch of the Law of England, Stated in the Form of Rules. Likewise, his inaugural lecture was temeritously titled ‘Can English Law be Taught at the Universities?’. His answer, harking back to and sadly little advance upon his first predecessor, was of course that it could. He stressed the need for study of law to be a science – complementary, he was careful to say, to study in chambers for those who wished to proceed to practice.[85] In publication, he did his university proud, producing successful books in both public and private law. The Introduction, based on his lectures in constitutional law, came out in 1885. The Law of Domicil was expanded into a treatise on conflict of laws which is still going in its thirteenth edition.[86] And he turned to a sort of political science in his 1905 Lectures on the Relation Between Law and Public Opinion in England during the Nineteenth Century,[87] the product of what seems to have been a highly successful lecture tour in the USA.

The philosophical background against which the Introduction was written was dominated by the analytical method formulated by John Stuart Mill. That method had suffused John Austin’s The Province of Jurisprudence Determined,[88] which after a slow start on its publication in 1832 had become a fount of orthodoxy in English legal theory.[89]

If Mill and Austin were Dicey’s background, it is tempting to assume[90] that in the Introduction he adopted the analytical method. Yet, in stark contrast with Austin, there is little of that method’s primary emphasis on a breaking down of ideas into component elements. Dicey derives from Mill values[91] but not method. Actually, there is little sign of any decided method at all. Dicey’s approach is instead merely, as Loughlin puts it, ‘mechanistic’: ‘Law is viewed as a datum to be analysed and classified and a descriptive account provided of how its various divisions fit together to provide an ordered whole.’[92] There is analysis, but there are equally synthesis and functional matching.[93] On the other hand, so far as there is history in the Introduction it barely rises above a narrative of laws. As to methodological quality the Introduction constitutes a serious regression from Austin and even a regression from Montesquieu.[94]

Which, no doubt, was the secret of its success. Few law students of Dicey’s place and time wanted rigour, still less Theory. Even Austin’s ideas were taught not from his works as such but from vulgarised summaries and as abridged in student editions, from which his ‘discussion of the bearing of morality on law and the relationship between positive law, divine law and utility were expunged’.[95] Not until the 1970s, indeed, would there be a sustained push for theory among English legal academics. Nor have many English law students, at any modern time, been interested in the romanist mainstream of the western legal tradition.[96] And, although Dicey’s level of theorisation could have benefited considerably from consideration of Continental constitutional theory and he read widely in French, he mentions the French mainly to show how wrongheaded they are and the Germans – whom Austin had studied extensively[97] – barely get their heads in at all. Dicey’s Austin is the students’ Austin. His book was admirably tailored to the stature of its audience and has moulded generations to the shape of its own attainment.

D Parliamentary Sovereignty

The Introduction centres on two doctrines, parliamentary sovereignty and the rule of law.[98] Equally important within Dicey’s œuvre, however, is his insistence (agreeing, for once, with Bentham and Austin) that judges make law – indeed, that there is judicial legislation.

Dicey introduces his sovereignty doctrine by defining a ‘law’ as ‘any rule which will be enforced by the courts’.[99] These rules are, first of all, legislative. Legislation is the exercise of ‘parliamentary sovereignty’. ‘Parliament’, in the legal sense of that word, is composed of the Queen, the House of Lords and the House of Commons. This tripartite Parliament is ‘sovereign’, in a double-sided sense. On the positive side, it has ‘under the English constitution, the right to make or unmake any law whatever’. Furthermore: ‘Any Act of Parliament, or any part of an Act of Parliament, which makes a new law, or repeals or modifies an existing law, will be obeyed by the courts.’ On the negative side, ‘no person or body is recognised by the law of England as having a right to override or set aside the legislation of Parliament’. ‘There is no person or body of persons who can, under the English constitution, make rules which override or derogate from an Act of Parliament, or which (to express the same thing in other words) will be enforced by the courts in contravention of an Act of Parliament.’ The doctrine of parliamentary sovereignty, Dicey insists, is as to both of its sides ‘fully recognised by the law of England’.[100]

His argument that there is such recognition relies largely upon doctrine and statute. As to the positive side he relies upon Coke, Blackstone and seventeenth-century statutes.[101] As to the negative side he is more historical. He reflects upon the transfer of legislative power from the King in Council to Parliament. He contends that there is no popular sovereignty but, rather, the ‘sole legal right of electors under the English constitution is to elect members of Parliament’. While a consolidation of precedent can amount to ‘judicial legislation’, the fact that judges cannot override statute while statute can override precedent shows that such precedents are only ‘subordinate legislation’.[102] A resolution of a House to commit for contempt is not law, because judges will not enforce it; yet it is an exercise of parliamentary sovereignty, because it is purported to be binding and judges will not interfere with it.[103] Parliamentary legislative power is not limited by morality or by public international law, no element of the royal prerogative is immune from legislative override and no parliament can bind its successors.[104] Parliamentary sovereignty is ‘an undoubted legal fact’; it is ‘the very keystone of the law of the constitution’.[105]

Dicey then tackles the apparent difficulty that his conception of parliamentary supremacy is drawn from Austin, yet Austin considers sovereignty to lie in the monarch, the houses of parliament and the people – the last, according to Austin, because a relationship of trust is presupposed by the ideas of delegation and representation.[106] Here, Dicey contends, Austin has confused sovereignty in a ‘political’ sense, in which ultimate power may indeed lie with the people, and in a legal sense, in which sovereignty is confined to the monarch and the houses of parliament. His evidence for this is that no judge has ever recognised or ever would recognise the will of the people except as expressed in legislation.[107] Political sovereignty, Dicey admits, has an external limitation in that the sovereign’s laws might be disobeyed or resisted and an internal limitation in that the sovereign inevitably legislates in accordance with its own ‘character’, which is moulded by contemporary social circumstances and morality.[108] These are limitations on political sovereignty, while legal sovereignty remains unlimited.

Yet Dicey doubly misunderstands Austin. First: by defining sovereignty in terms of ‘habit of obedience’, and law in terms of rule and sanction, Austin was trying to express the idea of sovereignty and law in terms that today would be called sociological.[109] Austin does not confuse political and legal sovereignty conceptually. Second: in Austin’s view, members of the House of Commons enjoy an ‘absolute’ delegation from the electorate – probably because the ‘trust’ reposed in them ‘is tacit rather than express’ and is ‘simply enforced by moral sanctions’ (so that this part of ‘constitutional law’ is actually only positive morality).[110] Thus Austin implies that a confusion between political and legal sovereignty occurs within the law as his subject matter. What Dicey achieves by deleting any element of popular sovereignty is to deny argumentational space to the rising labour movement’s insistence upon mandated delegation. The key difference between Austin and Dicey here is that the membership of that adaptable category ‘the people’ had changed.

E Judicial Ultimacy

In Law and Public Opinion, Dicey fully acknowledges that there is ‘judge-made law’, which he equates with ‘judicial legislation’ and finds to be little different in practice from parliamentary legislation.[111] By declining the veil that was the declaratory theory of adjudication, he revives the threat of judicial arbitrariness that Bentham identified in ‘dog-law’.[112] Dicey’s argument that the two types of legislation differ little in practice contains, however, a partial answer to such an objection. He contends that judges pay more attention to logic and principle than may appear while parliamentary legislators pay more attention to needs of the moment than they may admit. The list of objections to judicial legislation which he considers includes an objection on the ground of separation of powers but not the objection, which today would be paramount, that in Britain a parliament is elected while judges are not.

The Introduction, however, already contains a more subtle resolution of the dilemma over arbitrariness. Dicey’s expression the ‘rule of law’, as has been seen, is ‘a formula for expressing the fact that with us the law of the constitution, the rules which in foreign countries naturally form part of a constitutional code, are not the source but the consequence of the rights of individuals, as defined and enforced by the courts’. The constitution, therefore, has been produced through consolidation of precedent (a point, Dicey seems eventually to have seen, with which the idea of judicial legislation as subordinate is inconsistent). A central feature of this constitution is the doctrine of parliamentary sovereignty. It thus seems clearly to be assumed in Dicey’s argument that this doctrine is a judicial product – from which it would follow, as Allan and some senior British judges argue, that it may be judicially altered.[113] Dicey’s assumption is incorrect both historically and for his own day,[114] but what is important here is that he makes it.

The passage continues, as has been seen: ‘in short, the principles of private law have with us been by the action of the courts and Parliament so extended as to determine the position of the Crown and of its servants; thus the constitution is the result of the ordinary law of the land’. Dicey provides no warrant for assuming, further, that the consolidated precedents that have produced the constitution have been those of private law and not, by implication, those of public law. He might be right quantitatively, but the qualitative matters here. In his day, indeed, the summa divisio into private and public law which pervades the mainstream of the western legal tradition was not emphasised in England. His logically gratuitous preference for private law effects a constitutional privileging of private-law values over those of public law. Central to that would be a privileging of property over civil liberties. This is indeed Dicey’s preference:

Interference with public rights is at bottom a less striking exhibition of absolute power than is the interference with the far more important rights of individuals; a ruler who might think nothing of overthrowing the constitution of his country, would in all probability hesitate a long time before he touched the property or interfered with the contracts of private persons.[115]

Such is ‘the position of Parliament in regard to those private rights which are in civilised states justly held specially secure or sacred’.[116] Dicey is not unconcerned with public rights: he examines the legal defences against wrongful arrest and imprisonment, making much of the writ of habeas corpus and its supersession by the Habeas Corpus Acts – although he makes light of the Habeas Corpus Suspension Acts, saying nothing about the circumstances in which they had been readily passed. He is likewise sanguine about freedom of speech and of assembly.[117] Concerning assembly, he goes so far as to assert that a public meeting cannot be prohibited solely on the ground that it may result in a breach of the peace and that the ‘right of public meeting’ has become ‘a part of the law of the constitution’.[118] But that has never been English law with regard to picketing and in later public order legislation a likelihood that there may be ‘serious public disorder’ has been exactly the ground on which a meeting could be at least temporarily prohibited.[119]

Dicey simply would not have considered restrictions on picketing to be a problem for right-thinking people. His notion of equality before the law was entirely formal: he strongly objected to legislation that would introduce legal inequality in order to even up a substantive inequality. Thus he was distressed by the Trade Disputes Act 1906, which gave trade unions the aforementioned immunity in tort. This was to confer upon the unions, he thundered, ‘a privilege ... opposed to every principle of justice’.[120] The legislation went through and unions took advantage of it. During a miners’ strike in 1912, Dicey wrote to a friend: ‘Privilege, whether it be given to a King or a trade union, is simply a decent name for despotism. I am in my own mind firmly convinced that nothing will go right until we get back to the rule of equal law.’ This use of ‘force’ by the miners should be met with force, even if the result was bloodshed.[121] In terms of his distinction between the internal and external limits to sovereignty, it becomes clear, the external limit is to the state’s capacity to suppress. If workers push that limit, through exercising a privilege that is already inconsistent with the proper exercise of legal sovereignty, the worse for them. More generally, what is excluded here is legal expression of the socialist programme, involving legal inequality for the amelioration of economic inequality.[122]

Dicey also had a backup framework. The paramount interest, he argued, was that of the ‘nation’. The trade unions and those who legislated to privilege them were subverting the nation. So was the House of Commons when it asserted an ultimate right to approve budgetary legislation: the Lords should assert their traditional rôle as defenders of the national interest. With the rise of the Labour Party, which had entered Parliament in 1900,[123] Dicey saw the horrible prospect of ‘socialist revolution’ – he equated all collectivism with socialism.[124] And the prospect of actually slicing up the British lion by giving Home Rule to the Irish drove our mature Albert to apoplexy.[125]

In Dicey’s thinking, then, there is a coincidence of ultimates: property, nation and judicial legislation. The common link was something that would have been very familiar to him: that the English judiciary, from the High Court up to the Judicial Committee of the House of Lords, was selected exclusively from the Bar and the Bar consisted exclusively of scions of wealth.[126] They were reliable men. Though occasionally paternalistic, they would understand that the nation’s interests rested upon private property and ‘freedom’ of contract. While Dicey acknowledges that judges tend to reflect public opinion of the day before yesterday, in his view that is not so different from Parliament, which acts less upon the public opinion of today than upon that of yesterday, and in any case ‘beliefs are not necessarily erroneous because they are out of date’. Up to this point he is only stating commonplaces, but he immediately adds that ‘there are such things as ancient truths as well as ancient prejudices’.[127] He does not reveal what he thinks these are, but we have seen which rights he counts as ‘sacred’.

This, then, is the implicit content of his formal assertion, in the closing paragraph of the Introduction, that ‘the supremacy of the law of the land means in the last resort the right of the judges to control the executive government’.[128] It is a judicial right and one that judges have the right to create and maintain. This is the nub of his objection to droit administratif, which he thinks effects a ‘right of the government to control the judges’.[129] He is not concerned that the ordinary law should develop its capacity for judicial review of administrative action: he hopes, rather, that the sphere of state action will not expand so much as to require that. Though, when that expansion did take place, he was content that the ordinary courts embarked on such a development of ‘administrative law’.[130]

His concern with arbitrariness, however, was most selective

a study of European politics now and again reminds English readers that wherever there is discretion there is room for arbitrariness, and that in a republic no less than under a monarchy discretionary authority on the part of the government must mean insecurity for legal freedom on the part of its subjects.[131]

However, no exercise of authority can be so micro-gubernatorial as to involve no discretion at all. Or, rather, the exercise of legal authority is not a strictly or even primarily deductive process but a process of acting according to authorisation, in which all manner of extra-legal factors may and often must be taken into account.[132] This is true of judicial authority as well as of legislative or executive authority. In terms of the concept of discretion, Dicey wants to privilege judicial discretion over executive discretion. He provides no warrant for this: it just happens to give the sound individualists of the bench oversight of a collectivist (and socialist!) executive.[133]

Yet the most discretionary of the three ‘powers’ is the legislative. Dicey was right to insist that, in Britain at that time, there were few if any constitutional limitations on legislative power.[134] As Jennings was to note: ‘There is still, it may be argued, a rule of law, but the law is that the law may at any moment be changed’ and with any content.[135] Jennings rejects both of Dicey’s responses to this problem. Firstly, that legislation requires combined action of the three elements of parliament means that the process must be formal, but it does not mean (as Dicey thinks) that it must also be deliberate – legislation is sometimes passed in great haste. Nor, although the enactment of general rules would tend to be considered formally and deliberately, is the business of legislation confined to general rules.[136] Secondly, while it is generally true that outside periods of revolution parliament has not directly exercised executive power, it sometimes does so and its ordinary forbearance is a matter of convenience rather than of law.[137] In any case, there is no reason to suppose that administrative (executive) discretion is more open to abuse than legislative discretion.[138]

If Parliament is supreme, however, there is nothing ultimately to prevent it from interfering with even the most sacred rights. Dicey’s view appears to be that the issue is not legal but political. Meddling legislation could be kept out of the statute book provided that legislators were sound men and out of the lines of precedent provided that judges were sound men.[139] To the problem of the extreme, however, he had no answer. He quotes with mere approval his cousin Leslie Stephen’s now famous passage on blue-eyed babies:

If a legislature decided that all blue-eyed babies should be murdered, the preservation of blue-eyed babies would be illegal; but legislators must go mad before they could pass such a law, and subjects be idiotic before they could submit to it.[140]

Nor does Stephen himself have a rational solution. He merely asserts that social order is to be maintained, and such abuses of it are to be prevented, through proper social ‘instincts’.[141] Twentieth-century history has taught us that such a perspective tends, to put it mildly, to be self-defeating.[142] But past intentions must not be read simply in terms of their fate. Stephen’s reliance on instinct, like Bentham’s on the feelings of pain and pleasure, was an attempt to bring science – as, or on the model of, physical science – into the study of ethics. Stephen would have seen that simply as modernisation. The problem here is not how this perspective was later used but its emptiness and Dicey’s contentment with that.[143]

This concern may equally be expressed by contrast with the idea of a Rechtsstaat, as usually understood.[144] As a creature of the mainstream of the western legal tradition, it assumes a primacy of legislation over adjudication and hence a primarily legislative ‘legality (légalité, Legalität)’. The idea of ‘legality’ in this sense appears, as we have seen, in Aristotle. It appears throughout the current political spectrum, from Hayek[145] to Habermas[146] and beyond.[147] It is an antithesis, indeed it is intended as an antidote, to the Nazi predilection for illegality as justified by ‘exceptional’ circumstances.[148] However, the Nazi claim, in its developed form, was that to act outside current legality was to act in the name of a superior legality. It was to act outside the current statutory legality, in the capacity of a judge expressing the good sense of the Volk. These were the terms in which Hitler and his juristic champion Schmitt justified the ‘clean up’ of 1934. Schmitt subordinated legislation, with its feeble ‘neutral legality’, to the ‘supreme judgeship’ of the Leader as expressing the nation’s constantly good sense which unfailingly distinguishes its friends and its enemies.[149] When this argument is included in the picture, tyranny appears no longer as an issue of legality versus illegality but as an issue of competing legalities.[150] As Aristotle already saw, the rule of an absolute ruler is itself a kind of law.[151] It would certainly be presented as a kind of law.

Dicey surely has no tyrannical intention: he worries not about competing legalities but about competing content for the current legality. Nor should one risk any historical determinism by supposing that the supersession of competitive by monopoly capitalism[152] inevitably expands the sphere of executive or judicial discretion under permissively vague legislative provisions.[153] Indeed, current law is characterised at least as much by panoptical detail[154] while vaguery is found as much in the emancipatory areas of the law, such as anti-discrimination law or administrative law, as in commercial areas. Any bill of rights is necessarily couched in very general terms. The recent search for implicit ‘principles’ in Australian constitutional law was aimed in part at countering the absence of a bill of rights. International conventions must be expressed very generally, if their provisions are to be enactable into widely differing legal orders. Nonetheless, the spread of the notion of proportionality, especially where there is no bill of rights to provide its criteria, amounts to a considerable arrogation of judicial power.[155] The good intentions of one bench can become hellish in the hands of its successor. And nowhere does that matter more than in stating constitutional norms. In this perspective, Dicey positions the judiciary dangerously. Unbound by any constitutional niceties except such as they may concoct, or in a country with a constitutional document bound only by its necessarily very general terms as to fundamentals, the judiciary can weave into the law whatever is not plainly contrary to the constitution or to statute and they will also be the arbiters of such conflicts.

F Dicey and Separation of Powers

Strikingly, Dicey pays only passing attention to separation of powers. His interest is much more in the ‘distribution of powers’ that is characteristic of a federation. It is merely in the course of discussing that distribution that he contrasts the British and American understandings of separation of powers.[156] Moreover, his emphasis is on ‘balance’ as much as on ‘division’ and on institutions as much as on sources of authority:

We talk indeed of the English constitution as resting on a balance of powers, and as maintaining a division between the executive, the legislative, and the judicial bodies. These expressions have a real meaning.[157]

This is in line with Austin, who had already rubbished the idea that a separation of powers existed in Britain. For Austin, to suppose that legislative and executive powers ‘belong to distinct parties’ is ‘too palpably false to endure a moment’s examination’. One might be supposing that ‘legislative powers’ are ‘powers of establishing laws, and of issuing other commands’, while ‘executive powers’ are ‘powers of administering, or of carrying into operation, laws or other commands already established or issued’. Yet this is ‘far from approaching to precision’, he says with considerable understatement, for laws are most often applied through other laws and through judgments, and legislatures commonly exercise judicial powers. In Britain, indeed, king, lords and commons each exercise all three types of power.[158] He prefers to distinguish ‘the larger political powers’ into ‘supreme and subordinate’:

The former are the political powers, infinite in number and kind, which, partly brought into exercise, and partly dormant, belong to a sovereign or state: that is to say, to the monarch properly so called, if the government be a government of one: and, if the government be a government of a number, to the sovereign body considered collectively, or to its various members considered as component parts of it. The latter are those portions of the supreme powers which are delegated to political subordinates: such political subordinates being subordinate or subject merely, or also immediate partakers in those very supreme powers of portions or shares wherein they are possessed as ministers and trustees.[159]

This echoes Aristotle and supports what I said earlier about delegated legislation. Yet, it leaves no space for an independent judiciary. Austin later confirms that neither ‘statute law’ nor ‘judiciary law’ is a monopoly of any particular organ.[160]

Montesquieu gets a mention on this matter only during Dicey’s attack on droit administratif. Droit administratif rests, Dicey says, upon two ‘leading ideas’:

The first of these ideas is that the government, and every servant of the government, possesses, as representative of the nation, a whole body of special rights, privileges, or prerogatives as against private citizens, and that the extent of these rights, privileges, or prerogatives is to be determined on principles different from the considerations which fix the legal rights and duties of one citizen towards another.

The second is that of separation of powers (séparation des pouvoirs), which is a ‘necessity ... of preventing the government, the legislature, and the courts from encroaching upon one another’s province’. Dicey then traces the idea of separation of powers to Montesquieu and accuses him of having misunderstood the British position.[161]

One would expect Dicey to have read the whole ‘separation of powers’ section of The Spirit of Laws and, since he read widely in French, to have read it in the original. Had he acknowledged what Montesquieu goes on to say, he would have found himself and the baron to be in broad agreement about the British position. Except, however, on the point so crucial for Dicey – the judicial power. For Dicey, the judicial power is both permanent and professional, and accordingly it can be not only part of the scheme of checks and balances but also ultimate. Because of that difference, Dicey has no use for the scheme preferred by Montesquieu. That leaves Dicey with Montesquieu’s initial scheme, as indeed adopted in French law, about which all that he needs to say is that it underpins the dratted droit administratif.[162] What Dicey understands as the rule of law, then, does not rest upon either of Montesquieu’s schemes. Instead, it begins from the traditional British scheme of mixed and balanced government, then modifies that scheme in three ways none of which is to be found in Montesquieu: Dicey acknowledges that democracy has arrived (for men); he acknowledges that the monarchy is no longer a force within the scheme of checks and balances; and he installs the judiciary as a new third force within that scheme, with potential to restrict the consequences of democracy.

One might now find an explanation of why so much attention is devoted to droit administratif in what were, after all, introductory lectures on principles of British constitutional law. The exercise is certainly not comparative: Dicey pays little attention to the English counterpart in which he appears to rest so much hope, judicial review of administrative action; its wobbly keystone, the prerogative writ of certiorari, is not even mentioned.[163] On the other hand, in the first and second editions of the Introduction the critique of droit administratif is integral with the argument for the rule of law.[164] Dicey seems to have inserted this marginally relevant critique in order to gain momentum for his idea of the rule of law. He was able to erect that idea in criticism of the privileges that he asserted were claimed by government in droit administratif on an excuse that the state represents the general interest, when by his day the English battle against the monarch as representative of the general interest had long been won while that against using the state as instrument of the general interest as perceived from the bottom of a mineshaft seemed about to be lost.

G After Dicey

Dicey’s views on droit administratif have, however, become a central issue in the reception of his idea of the rule of law, with the mushrooming of administrative tribunals. The procedure of a major tribunal is often barely distinguishable from that of a court. A major specialist tribunal may possess greater expertise than any court even as to the relevant law. The ‘rule of law’ issue is then what justification remains for judicial review of those tribunals’ decisions and the arguments have become thin.[165] Beneath that, however, is a ‘separation of powers’ issue that has long waited in the wings. As long as, with Aristotle, the focus is on processes as politically understood, commonsense understandings of those processes’ identities can be serviceable. But, once as with Montesquieu the focus is shifted to types of decision as legally understood, common sense no longer serves.

This shift is intensified when Montesquieu’s initial scheme, or some application of it, is expressed within a constitution. Then begins an exercise in disintegrative reflection: each of the three constitutionally established organs tends to find its identity less through literal interpretation of the constitution, or as time goes on by recourse to framers’ intention, than by just looking in a mirror. Since the mirror will show each of them not what they ought to do but what they think they are already doing, they conclude that they are doing it very well and that any apparent ugliness is due to injury by one or both of the others. The judiciary, being the only authorised reporters of what is seen in the mirror, attracts especial attack from the other two organs. It loses, or cannot be accorded, the ascendancy required by Dicey.


A Multiplication

Our two ideas have multiplied. The idea of separation of powers has given place to

(1)a supposition that government caries out three main functions: legislation, execution and adjudication – or four, if one distinguishes between internal and external execution;

(2)a policy that effects a division of labour among organs or officers of state – usually, although not necessarily, following (1);

(3)a principle of distribution of power (to keep bastards honest) – within government and/or, in a federal system, between governments.

The idea of the rule of law has given place to

(4)the idea of a government of laws (where officials are subject to law, even if their legal position remains one of legal privilege as with a constitutional monarch);

(5)an idea of universal positive legality (where everybody, including officials, is equally subject to law).

Yet the identities of all five remain uncertain and, as to how any of them may be realised in combination, we are (as the Germans say) as clever as before. The itch gets worse when one finds it necessary to add, as I shall discuss, a use of the expression ‘the rule of law’ as an intensifier of (4) or (5). Especially, we do not yet have an explanation of how the five have been condensed into two.

B Reduction

All of the five are, from the start, external reflections upon positive law and not elements of it: (1) is a product of Enlightenment, quasi-materialist general theory of law, while (2) to (5) are products of political science or political philosophy. Yet today we meet two ideas, ‘separation of powers’ and ‘the rule of law’, as components of any respectable positive legal order. More exactly, they appear as legal ‘institutions’.

The notion of a legal ‘institution’ is first of all, in Roman law, that of an interdependent set of norms within a legal order, having a common type of content.[166] These sets were later promoted to the status of ideal forms within a positive legal order.[167] Currently, this idealism has been superseded by an account of legal institutions as ‘institutional facts’ in Searle’s sense.[168] But legal norms are not that innocent. Their legal character is that they are tacitly closed to alternatives and appear as binding through that default.[169] With this tacit invisibility of alternatives, their deletion from discursivity, the norms characterised as legal, together with their contents, are elevated beyond discursivity; they come to appear as suprahuman.

So also with their agglomeration in institutions. This illusive appearance is real in the sense that it is a held belief. In its exclusivity, however, it monopolises the understanding of those norms. On one side, it permits only an internal understanding, by way of exegesis. On the other, it suppresses the possibility of an external understanding, by way of social description. In this perspective, firstly material events are kept out of the picture, secondly the semantic relations among institutions or sectors of institutions appear as real (ideal) interactions and thirdly the material events reappear as instances of those interactions. Thus one is led to debate relationships among the three ‘powers’ as if they were real (ideal) interactions and to neglect the real (material) relations of power. Since this is also to deprive those powers of prior raisons d’être, they appear as merely given and as a set less of identities than of barriers.[170] Likewise one is led to debate ‘the rule of law’ as a relation between a benevolently suprahuman set of norms and ‘the individual’ in which the latter appears only as a natural legal person, an instance of the legal institution ‘legal person’, formally equal by definition and standardisation. One is distracted from such issues as whether the norms composing a rule of law (or guaranteeing a Rechtsstaat) are actually implemented by officials.[171] The original externality of the five ideas to which those of separation of powers and the rule of law can be traced is obliterated. They appear no longer as policies or political principles but as having always embodied legal principles. The five, so reduced and elevated, are transformed into the familiar two.

This internality is an impediment to general theory. Generalisation in idealist theory of law either achieves externality to positive legal orders through their transcendence by a natural legal order or is confined to a noting of intuitively common factors which is constantly sucked down into the categories and habits of the particular order of positive law that is the primary focus of attention. This may explain the paradox that Montesquieu swiftly abandons his initial scheme, in one of the oldest and therefore presumably most worked upon parts of The Spirit of Laws. The initial scheme looks like a project for a set of ideal types,[172] but the set does not develop into a theory before it is overwhelmed by the very different British model.[173]

At the same time, the elevation of legal institutions to a suprahuman status impedes a major option for a basis of judicial independence at least where there is a formal constitution: that, since a supreme court is the final interpreter of the meaning of the constitution, and since the constitution stands above the three powers, that interpretive rôle is not an exercise of the judicial power as such but a distinct activity to which the judiciary is best suited.[174] In this rôle the judiciary is not the least dangerous branch of government but not a branch of government at all. And to perform that rôle it has to be independent – not only of the legislative and executive powers but of all other interests bar that of the people as expressed in the constitution and capable of re-expression through the people’s ultimate capacity to amend or replace it. This picture is distorted if either the institutions established by the constitution acquire an ideality that usurps the reality of the document or the document itself is idealised.

C Dual Objectivity

The re-presentation, however, is not solely idealist. There is already ambiguity in Aristotle, between ‘parts’ in a semantic and a material sense. Montesquieu begins to restate legal institutions in Enlightenment materialist terms: we have seen the ambiguity of his expressions ‘pouvoir’ and ‘puissance’. Likewise, Dicey’s shift from Aristotle’s ‘the law to rule’ to ‘rule of law’ cuts out the material power relations that Aristotle only summarises in saying ‘the law to rule’, in favour of making a ‘rule of law’ appear as an actual situation to which There Is No Alternative.[175] In this sort of reasoning, an ideal objectivity is veneered with a material objectivity. The stake is not to choose between idealism and materialism, but to use both of them, with their common though differently formed metaphysic of objectivity, to deny historicity and its infinitely various alternatives.

D The Outer

McDonald is accordingly correct in warning against a ‘dangerous temptation’ that, in utilising the idea of the rule of law, one will ‘invoke its rhetorically powerful language as an argument stopper, thus avoiding the substantive arguments that the ideal inevitably involves’.[176] I think that this is nearly right, except that the argument-stopping lies first of all not in the expression ‘rule of’ but in the expression ‘law’.[177] In that case, however, what is there for ‘rule of’ to add?

What is added, it seems, is the appearance of a material situation to which there is no alternative. As Kairys observes, there is ‘a flip, a turning of things upside down’: democracy is seen not as requiring the rule of law but as a requirement of the rule of law, hence confined to what current law requires. ‘This is quite a significant flip, and it is gaining momentum without any real debate or discussion.’ One thinks: ‘When my side wins, the rule of law has prevailed’; this ‘move’ has the same form as an appeal to ‘God’. Appeal to ‘the rule of law’ then becomes a menace in both domestic and foreign (or imperialist) affairs.[178] The very suggestion that there could be an alternative is already defined as irrational. And whoever proposes an alternative is not a friend but an enemy of society.[179]

We do need a ‘new constitutional anatomy’, created through interdisciplinary examination not only of the legal texts that describe constitutional arrangements but also of the various texts and practices that actually constitute those arrangements.[180] But there is no royal road to that anatomy through other disciplines. The question is, rather, both to draw upon them and to invent - in order to reinvent legal science.

[*] Senior Lecturer, Division of Law, Macquarie University, Sydney 2109, Australia; <> I am grateful to the Macquarie Law Journal’s three anonymous reviewers of a hasty first version.

[1] William Twining, ‘The Great Juristic Bazaar’ (1978) 14 Journal of the Society of Public Teachers of Law (ns) 185, 190.

[2] Montesquieu, The Spirit of the Laws (Thomas Nugent trans, 1949 ed), Book 11, ch 6. I prefer to translate ‘des lois’ as ‘of Laws’, since this is a work of general theory.

[3] A V Dicey, Introduction to the Study of the Law of the Constitution (first published 1885, 10th ed, 1959) ch 4.

[4] Eg, Tony Blackshield and George Williams, Australian Constitutional Law and Theory (3rd ed, 2002) 105-6 (Dicey), 603 (Montesquieu).

[5] Ivor Jennings, The Law and the Constitution (first published 1933, 5th ed, 1965) 61.

[6] Jennings, above n 5, 60.

[7] Quoted, Blackshield and Williams, above n 4, 109-10.

[8] Joseph Raz, ‘The Rule of Law and its Virtue’ (1977) 93 Law Quarterly Review 195, 195. Cf Geoffrey de Q Walker, The Rule of Law (1988) ch 1; David Kairys, ‘Searching for the Rule of Law’ (2003) 36 Suffolk University Law Review 307.

[9] Raz, above n 8; cf Paul Craig, ‘Formal and Substantive Conceptions of the Rule of Law: an Analytical Framework’ [1997] Public Law 467.

[10] Alan C Hutchinson and Patrick Monaghan, ‘Democracy and the Rule of Law’ in Alan C Hutchinson and Patrick Monaghan (eds), The Rule of Law: Ideal or Ideology? (1987) 97, 99.

[11] Ian Harden and Norman Lewis, The Noble Lie: the British Constitution and the Rule of Law (1986).

[12] See especially, Franz L Neumann, The Rule of Law (1933 thesis, published 1986). William E Scheuermann, The Rule of Law under Siege: Selected Essays of Franz L Neumann and Otto Kirchheimer (1996).

[13] Franz Neumann, ‘Rechtsstaat, the Division of Powers and Socialism’ in Otto Kirchheimer and Franz Neumann, Social Democracy and the Rule of Law, ed Keith Tribe (1987) 66-74, 67.

[14] E P Thompson, Whigs and Hunters: the Origins of the Black Act (1975) 266; cf Bob Fine, Democracy and the Rule of Law (1984) 169-89.

[15] Franz Kafka, ‘The Problem of Our Laws’ (1920) in his Description of a Struggle and Other Stories (Malcolm Pasley trans, 1979 ed) 128-30.

[16] As is recorded to have been the practice of the legendarily just Louis IX: Jean de Joinville, ‘Life of Saint Louis’ in Joinville & Villehardouin: Chronicles of the Crusades (M R B Shaw trans, 1963 ed) 161-353, 177.

[17] Lord Bingham, ‘Dicey Revisited’ [2002] Public Law 39; an address by the senior law lord. See also the Introduction centenary essays in [1985] Public Law.

[18] British discussion of separation of powers currently revolves around the Constitutional Reform Bill 2004, which would abolish the triple-hatted position of Lord Chancellor and replace the Judicial Committee of the House of Lords with a Supreme Court separate from the legislature. Discussion of the rule of law focuses on the Human Rights Act 1998, which gives qualified domestic force to the European Convention on Human Rights, and on the jurisdiction of the European Court of Human Rights under the Convention. Cf Jeffrey Jowell, ‘The Rule of Law Today’ in Jeffrey Jowell and Dawn Oliver (eds), The Changing Constitution (4th ed, 2000) 3-22; Jeffrey Jowell, ‘Beyond the Rule of Law: Towards Constitutional Judicial Review’ [2000] Public Law 671.

[19] Rebecca L Brown, ‘Separated Powers and Ordered Liberty’ (1991) 139 University of Pennsylvania Law Review 1513, 1517-18.

[20] See the vehement language in the three decisions handed down on 28 June 2004 (542 US ___ (2004)), on habeas corpus petitions by prisoners captured in Afghanistan: Rasul v Bush, Hamdi v Rumsfeld and Rumsfeld v Padilla. Cf the strongly worded disagreement between McHugh and Kirby JJ in the Australian immigration detention case Al-Kateb v Godwin [2004] HCA 37.

[21] Bruce G Peabody and John D Nugent, ‘Toward a Unifying Theory of the Separation of Powers’ (2003) 53 American University Law Review 1, 3-4. The authors understand ‘separation of powers’ to include bicameral and federal distribution.

[22] Kairys, above n 8, 308; recent US works on the idea of the rule of law are listed at 311.

[23] On separation of powers in Australia, see, eg, Fiona Wheeler, ‘Due Process, Judicial Power and Chapter III: an Evolving Guarantee’ in Australian Association of Constitutional Law, The Australian Constitution in Troubled Times (Proceedings, Annual Public Law Weekend 2003) 33-58.

[24] Isaiah 33.22 (New English Bible).

[25] Aristotle uses the word politeia to mean the state, the ground rules of the state and his third ‘true’ form of government. He also asserts that politeia means the same as politeuma, ‘government’: Politics 1279a. Commentators have wrestled with this shaky terminology.

[26] Aristotle, Politics (H Rackham trans, 1932). This is the bilingual Loeb edition. Rackham’s is a relatively literal translation, appropriate for the present purpose. A concise introduction to the Politics is Fred Miller, ‘Aristotle’s Political Theory’ (2002) in Stanford Encyclopedia of Philosophy <> at 6 September 2004.

[27] Politics, above n 26, 1279a-b.

[28] Ibid 1289a-b.

[29] Ibid 1284a.

[30] Ibid 1285a.

[31] Ibid 1286a.

[32] Ibid 1287a.

[33] Ibid 1287b. Rackham notes on ethos (customary laws or laws of custom): ‘i.e., the rules of duty and of manners that are customary but not embodied in legislation’.

[34] Ibid 1289a.

[35] Ibid Books 7-8.

[36] Miller, above n 26. See further Fred D Miller Jr, Nature, Justice, and Rights in Aristotle’s Politics (1995) ch 7.

[37] Jowett’s translation of the Politics was for long the standard translation for the non-classicist reader and has appeared in various editions.

[38] Politics, above n 26, 1297b-1298a.

[39] Ibid 1299a.

[40] Ibid 1298a.

[41] Ibid 1299a-b.

[42] Ibid 1300b.

[43] Liddell and Scott, Greek-English Lexicon <> at 6 September 2004.

[44] This use of concrete language is consistent with Aristotle’s immanent idealism, but there is not enough abstract purchase here to read the structure of this exposition off from the metaphysics elsewhere.

[45] Miller, Nature, above n 36, 259-60.

[46] Politics, above n 26, 1281a-b.

[47] Richard Robinson (trans), Aristotle’s Politics Books III and IV (1962) 24-5.

[48] Though Montesquieu treats the British constitution as English, as in substance it was.

[49] Eg, Jennings, above n 5, 20-33. Holdsworth, discussing the history of the idea of a mixed and balanced constitution, accuses Montesquieu of ‘a very superficial study’ of the British constitution while citing from the beginning and the end, but not the more instructive middle, of Montesquieu’s ‘powers’ section: William Holdsworth, A History of English Law, x.718-19 (1938). Lord Simon, 597 HL Debs 719 (17 February 1999), argued that eighteenth-century Britain had not separation of powers but balance of powers, which I shall contend is what Montesquieu actually said.

[50] Iain Stewart, ‘Montesquieu in England: his “Notes on England”, with Commentary and Translation’ (2002) Oxford University Comparative Law Forum 6 < articles/montesquieu.shtml> at 6 September 2004. Detailed references to The Spirit of Laws are given there. New here, however, are broader reflections on Montesquieu’s political life and a suggestion that domestic administration is part of his legislative power.

[51] Most of the letters are in the Œuvres complètes edited by André Masson (1950-55) 3 vols. The accuracy of Masson’s transcriptions has, however, been doubted.

[52] As francophones as well as anglophones seem rarely to have done. Among anglophones, even Vile barely notices that the initial scheme is actually replaced: M J C Vile, Constitutionalism and the Separation of Powers (1967) 83. As to francophones, see Michel Miaille, L’État du droit (1980) 212-19.

[53] Cf Vile, above n 52; W B Gwyn, The Meaning of the Separation of Powers (1965).

[54] For Blackstone, in England the terms ‘sovereignty’ and ‘legislature’ are mutually ‘convertible’: both of them indicate ‘the supreme power’, which is distributed among king, lords and commons while the executive power is subordinate and is concentrated in the hands of the king: William Blackstone, Commentaries on the Laws of England (1765-9) i.50-2 <> at 6 September 2004.

[55] Eg, Jennings, above n 5, 29-33.

[56] Or, in Montesquieu’s now antiquated spelling, ‘loix’.

[57] Above n 2.

[58] The Spirit of the Laws (1989).

[59] In Montesquieu, Œuvres complètes R Caillois (ed), (1949-51), Bibliothèque de la Pléiade.

[60] Montesquieu was surely aware of Aristotle’s scheme: although he does not refer to it (though he adopts its idiom), he refers to several other passages in the Politics. On this plane we may find a reason why Montesquieu does not explain why there should be three elements - neither more nor less. It might be because of an underlying dialogue with Aristotle, who was fond of threesomes. Alternatively, Montesquieu may have had in mind that three is the minimum number of entities that could carry out checking and balancing at all and that many more than three would be a self-defeating crowd. That view, however, would be prescriptive: the descriptive question of whether states do contain neither more nor less than three ‘powers’ needed to be answered independently.

[61] Ordinarily: Montesquieu’s ‘Notes’ show that he was aware of the distrust of a standing army.

[62] These four powers had also been distinguished, and also not felicitously, by Locke – although he placed the internal and external executive powers in the same hands: John Locke, Two Treatises of Government (first published circa 1690, Peter Laslett (ed), 1988) §124-58.

[63] Institutes 1.1.3-8; cf Digest 1.1.6-9, 11.

[64] Cf Justinian in Code, dated 529. No statute conferring the people’s authority and power upon an emperor (lex regia) has ever been found.

[65] Indeed, all laws should be expressed in general terms, not directed to a particular case: Digest 1.3.3-6, 8, 10. This requirement is addressed sometimes to law generally, sometimes only to written law; it seems to have been considered unnecessary to apply it specifically to unwritten law.

[66] De l’esprit des lois Book 1, ch 3 (my translation). Montesquieu means by ‘political laws’ those concerning relations between the rulers and the ruled, and by ‘civil laws’ those concerning relations among the citizens themselves: ibid. He appears to be distancing himself from Roman legal terminology in order to formulate general theory, although the elements of his theory remain principally romanist.

[67] Locke, above n 62, §142.

[68] Blackstone’s perspective was similar: above n 54. This view appears to correspond to practice, which may explain why the High Court of Australia found it so awkward to come to terms with the evident: Victorian Stevedoring & General Contracting Co Pty Ltd and Meakes v Dignan [1931] HCA 34; (1931) 46 CLR 73. Cf Denise Meyerson, ‘Rethinking the Constitutionality of Delegated Legislation’ (2003) 11 Australian Journal of Administrative Law 1.

[69] Miaille, above n 52, 219.

[70] Discussed in Stewart, above n 50.

[71] Cf Jennings, above n 5; Brown, above n 19, 1522-31. Eisenmann similarly distinguished between ‘juristic’ and ‘political’ interpretations: discussed in Stewart, above n 50.

[72] Marshall CJ in Marbury v Madison 5 US [1803] USSC 16; (1 Cranch) 137 (1803), 176-80.

[73] O’Connor J for the majority in Hamdi v Rumsfeld, above n 20.

[74] Dicey, above n 3, 203.

[75] Richard A Cosgrove, The Rule of Law: Albert Venn Dicey, Victorian Jurist (1980) xiv-xv. Cosgrove’s is mainly an intellectual biography; and see review by David Sugarman, (1983) 46 Modern Law Review 102. Trowbridge A Ford, Albert Venn Dicey: the Man and his Times (1985) is mainly a personal biography. Ford thinks that Cosgrove has been fooled: that Dicey posed as ‘naïf, diffident and clumsy’ and Cosgrove ‘took the whole bait’ ([iii]). My interest is in the Introduction’s manifestations of legal ideology - an issue that has emerged much more recently. No such criticism is noted in E C S Wade’s ‘Introduction’ to the tenth edition or in the simultaneous defence of Dicey by F H Lawson, ‘Dicey Revisited’ (1959) 7 Political Studies 109 and 207.

[76] He would advocate that the ‘natural bias’ toward state intervention, under the attraction of its beneficial effects, which being direct are visible, and insouciant of its ‘evil’ effects, which being indirect are hidden, could be counteracted only by ‘a presumption or prejudice in favour of individual liberty - that is, of laissez faire’: A V Dicey, Lectures on the Relation Between Law and Public Opinion in England During the Nineteenth Century (first published 1905, 1962 ed) 257-8.

[77] The group’s focal figure was John Venn, rector of Clapham in south London, after whom Dicey was named.

[78] He would conclude that the ‘weaker’ sex should not, after all, have the vote because on the Irish Question they would not be sufficiently staunch.

[79] Cosgrove, above n 75, 8. ‘In all conflict, Dicey wanted total victory’: Ford, above n 75, 284.

[80] Ford, above n 75, throughout.

[81] ‘Autobiographical Fragment’ in Robert S Rait (ed), Memorials of Albert Venn Dicey (1925) 1, 4-5.

[82] Dicey, above n 3.

[83] F W Maitland, The Constitutional History of England (1919) 526-39.

[84] On law at Oxford in Dicey’s day see F H Lawson, The Oxford Law School 1850-1965 (1968).

[85] Rait, above n 81, 86-7.

[86] Digest of the Law of England with Reference to the Conflict of Laws (1896); Dicey and Morris on Conflict of Laws (13th ed, 2000).

[87] Dicey, above n 76, 257-8.

[88] John Austin, The Province of Jurisprudence Determined (1832) reprinted as the first volume of Austin, Lectures on Jurisprudence (first published1861; reprinted 1970).

[89] On Austin, see W L Morison, John Austin (1982). More factual study of law was confined to legal history, which took two forms. Maine’s Ancient Law, grounded in the evolutionist metaphysic, was already a standard textbook: Henry Sumner Maine, Ancient Law (1861); cf Roger Cotterrell, The Politics of Jurisprudence (1990) 41-51. The more empirical variety of historical jurisprudence remained antiquarian – the masterwork by Pollock and Maitland would appear in 1895: Frederick Pollock and Frederic William Maitland, The History of English Law Before the Time of Edward I (first published 1895, 1968 ed). Sociology had only recently come into existence and social anthropology, then being solely about ‘primitives’, was evidently unsuitable for the study of English law.

[90] As Wade does: E C S Wade, ‘Introduction’ in Dicey, above n 3, xvii-cxcviii, clxxxviii, cxci.

[91] Haig Patapan, ‘The Author of Liberty: Dicey, Mill and the Shaping of English Constitutionalism’ (1997) 8 Public Law Review 256.

[92] Martin Loughlin, Public Law and Political Theory (1992) 17; Loughlin calls this approach ‘analytical’.

[93] McEldowney struggles to find in what sense Dicey’s method is ‘analytical’. See John F McEldowney, ‘Dicey in Historical Perspective – a Review Essay’ (reviewing Cosgrove’s biography) in Patrick McAuslan and John F McEldowney (eds), Law, Legitimacy and the Constitution: Essays Marking the Centenary of Dicey’s Law of the Constitution (1985) 39, 54-8. Even Dicey’s defender Lawson admits, on one key point, ‘a fumbling approach, characteristic of a pioneer, though a pioneer of genius’: above n 75, 121.

[94] It is therefore hard to agree even with Lord Bingham’s qualified praise, ‘Dicey was a genius, but a complex genius, a man subject to contradictions and blind spots’, above n 16, 51.

[95] David Sugarman, ‘Legal Theory, the Common Law Mind and the Making of the Textbook Tradition’ in William Twining (ed), Legal Theory and Common Law (1986) 26, 43.

[96] Though they are now forced in that direction by British membership of the European Union and a statutory partial incorporation of the European Convention on Human Rights.

[97] Albeit that in Austin’s own day German jurisprudence was in a state of ‘transition from natural law to the historical school’, the latter being the idealist outlook of Hegel, Hugo and Savigny: Andreas B Schwarz, ‘John Austin and the German Jurisprudence of his Time’ (1934) 2 Politica 178, 183.

[98] Dicey, above n 3, 473. Constitutional conventions, which Dicey emphasises can be more important than laws, initially receive equal priority but eventually are disqualified from the category ‘law’: ibid 35, 439-40.

[99] Ibid 40. A view common in his day: cf Pollock and Maitland, above n 89, i.xcv.

[100] Dicey, above n 3, 39-40.

[101] Statutes such as the Act of Settlement, the Acts of Union, the Septennial Act and the Indemnity Acts; ibid 40-50.

[102] This is misleading: a consolidation of precedent is not subordinate legislation; that is, it is not application of primary legislation. If it is legislation, it is another type of legislation and hence might be considered equal with or even superior to statute.

[103] Dicey, above n 3, 50-61.

[104] Ibid 61-8.

[105] Ibid 69-70.

[106] Austin, above n 88, i.201-6. Dicey cites the fourth edition of the Province. Austin’s view is, of course, also the American tradition.

[107] Dicey, above n 3, 70-6. At that time, English law did not provide for referenda and Dicey advocated the introduction of this potentially populist device.

[108] Ibid 76-85.

[109] An aspect of his approach that summaries of him still ordinarily neglect: see further Cotterrell, above n 86, ch 3.

[110] Austin, above n 88, i.202-4.

[111] Dicey, above n 76, 361-98, 483-94. He traces in detail the parallel judicial and parliamentary movements toward equality of married women’s property, claiming that parliament caught up with the courts rather than the other way around: ibid 371-95.

[112] Jeremy Bentham, ‘Truth versus Ashhurst’ in John Bowring (ed), Works (1792) v.231-7, 235-6; and in the texts online with the Bentham Project, <> at 6 September 2004.

[113] T R S Allan, Law, Liberty, and Justice (1993); see, for a general critique of this, Richard Ekins, ‘Judicial Supremacy and the Rule of Law’ (2003) 119 Law Quarterly Review 127.

[114] Jennings, above n 5, 39, 313-17; Jeffrey Goldsworthy, The Sovereignty of Parliament (1999).

[115] Dicey, above n 3, 48-9.

[116] Ibid 48.

[117] Ibid chs 5-7.

[118] Ibid 283.

[119] Public Order Act 1936 (UK) s 3; this ground was extended in the Thatcher government’s Public Order Act 1986 (UK) ss 11-16.

[120] Letter to the Times, 29 October 1906; quoted, Cosgrove, above n 75, 209.

[121] Cosgrove, above n 75, 209-10.

[122] Cf Karl Marx, ‘Critique of the Gotha Programme’ (1875) in David Fernbach (ed), The First International and After (1974) 341, 346-7; Peter Schöttler, ‘Friedrich Engels and Karl Kautsky as Critics of “Legal Socialism”’ (Iain Stewart trans) (1986) 14 International Journal of the Sociology of Law 1.

[123] It entered government (in a coalition) in 1916, became the official opposition in 1922 and formed a government in 1924.

[124] Dicey, above n 76, 17, 259.

[125] Ford, above n 75, throughout. Dicey was so worried that he sided with the Conservatives in advocating as late as 1913 that the monarch could refuse the royal assent to the Home Rule Bill, a power not exercised since 1707: V Bogdanor, The Monarchy and the Constitution (1995) 132. His likely state of excitation over current British devolution is canvassed by Stephen Livingstone, ‘Dicey and the Celtic Nations: a Nightmare Come to Life?’ in W John Morgan and Stephen Livingstone (eds), Law and Opinion in Twentieth-Century Britain and Ireland (2003) 194-215.

[126] Cf Ralf Dahrendorf, ‘Law Faculties and the German Upper Class’ (1964) in Vilhelm Aubert (ed), Sociology of Law (1969) 294-309; Walter Ott and Franziska Buob, ‘Did Legal Positivism Render German Jurists Defenceless During the Third Reich?’ (1993) 2 Social and Legal Studies 91.

[127] Dicey, above n 76, 369.

[128] Dicey, above n 3, 472.

[129] Ibid.

[130] ‘The Development of Administrative Law in England’ (1915) in Dicey, above n 3, 493-9.

[131] Dicey, above n 3, 188.

[132] Hans Kelsen, Pure Theory of Law (2nd ed, 1960; Max Knight trans, 1967) 101-7; Iain Stewart, ‘Kelsen Tomorrow’ (1998) 51/1 Current Legal Problems 181, 187-8.

[133] Dicey also downplays the extent to which judicial power in England was still fragmented among a range of types of court: cf H W Arthurs, ‘Special Courts, Special Law: Legal Pluralism in Nineteenth Century England’ in G R Rubin and David Sugarman (eds), Law, Economy and Society, 1750-1914 (1984) 380-411 and notes.

[134] Perhaps the Treaties of Union.

[135] Jennings, above n 5, 57.

[136] Dicey, above n 3, 407; Jennings, above n 5, 57-8.

[137] Dicey, above n 3, 408-9; Jennings, above n 5, 59-60.

[138] Jennings, above n 5, 307.

[139] That they would be men was the realistic prospect. Women did not get the parliamentary vote in Britain until 1918 and not equally with men until 1928.

[140] Dicey, above n 3, 81; quoting Leslie Stephen, The Science of Ethics (1882) 143.

[141] Stephen, above n 140, 137-47.

[142] Cf Erich Fromm, The Sane Society (1965; reissued 1998) and other works.

[143] Stephen’s observation was debated inconclusively when the High Court of Australia explored whether there is any implicit legal limitation on the races power, Constitution s 51(xxvi): Kartinyeri v Commonwealth (the Hindmarsh Island Bridge Act Case) [1998] HCA 22; (1998) 195 CLR 337 (transcript, 5-6 February 1998, in <> at 6 September 2004).

[144] Cf Michel Troper, ‘The Limits of the Rule of Law’ in Cheryl Saunders and Katherine Le Roy (eds), The Rule of Law (2003) 81-97.

[145] F A Hayek, The Constitution of Liberty (1960), pt 2. On Hayek’s conception of the rule of law, see Norman P Barry, Hayek’s Social and Economic Philosophy (1979), 90-5, 108-11.

[146] Jürgen Habermas, Between Facts and Norms(1992), William Rehg trans (1996); Wesley Shih, ‘Reconstruction Blues: a Critique of Habermasian Adjudicatory Theory’ (2003) 36 Suffolk University Law Review 331.

[147] Cf David Dyzenhaus, Legality and Legitimacy (1997). See also, Dyzenhaus, ‘The Justice of the Common Law’ in Saunders and Le Roy, above n 144, 21-51.

[148] Still more generally, an appeal to legality as against arbitrariness is not necessarily an appeal to legislation. In defence of aristocratic against bourgeois economic rights, Savigny contrasted the stability of customary law with the arbitrariness of legislation: Friedrich Carl von Savigny, Vom Beruf unsrer Zeit für Gesetzgebung und Rechtswissenchaft (1814) in Hans Hattenhauer (ed), Thibaut und Savigny: ihre programmatischen Schriften (1973), 95-192, 105.

[149] ‘Der Führer schützt das Recht: zur Reichstagsrede Adolf Hitlers vom 13. Juli 1934’ (1934) 39 Deutsche Juristen-Zeitung 945.

[150] Cf Adam Podgorecki and Vittoria Olgiati (eds), Totalitarian and Post-totalitarian Law (1996). Cf Saddam Hussein, during his first day on trial: ‘Are you a judge? You are a judge. And judges, they value the law. And they rule by law, right? Right? Right is a relative issue’, Sydney Morning Herald, 3-4 July 2004.

[151] Politics 1284a.

[152] On which see, for example, Bob Fine, above n 14; Bob Fine et al (eds), Capitalism and the Rule of Law (1979).

[153] Cf Franz Neumann, ‘The Change in the Function of Law in Modern Society’ (1937), Klaus Knorr and Edward A Shils trans, in Neumann, The Democratic and the Authoritarian State (1957), 22-68. See further Roger Cotterrell, ‘The Rule of Law in Transition: Revisiting Franz Neumann’s Sociology of Legality’ (1996) 5 Social and Legal Studies 451.

[154] The English judiciary abandoned its attempt to approve prosecutions for conspiracy ‘to corrupt public morals’ (Shaw v DPP [1961] UKHL 1; [1962] AC 220, 268, 290-1) or ‘to outrage public decency’ (Knuller v DPP [1973] AC 435, 456).

[155] Cf Hartmut Maurer, Staatsrecht (1999) 233-6.

[156] Dicey, above n 3, 156. Likewise, he contrasts the British and French understandings of separation of powers only in the course of attacking the idea of droit administratif: ibid 472.

[157] Ibid 156.

[158] Austin, above n 88, i.206-8.

[159] Ibid i.210.

[160] Ibid ii.320-1.

[161] Dicey, above n 3, 336-8.

[162] Ibid 472.

[163] Lawson, above n 75, 125-6; Wade, above n 90, cxvi-cxvii.

[164] Lawson, above n 75, 111.

[165] Especially in the hopeless attempt to make, or find an alternative to, a distinction between jurisdictional error (reviewable) and intra-jurisdictional error (unreviewable): Anisminic v Foreign Compensation Commission [1968] UKHL 6; [1969] 2 AC 147 (HL); Craig v South Australia [1995] HCA 58; (1995) 184 CLR 163 (HCA).

[166] As in the second-century Institutiones of Gaius and its sixth-century development, the Institutiones of Justinian.

[167] Albert Broderick (ed), The French Institutionalists (1970).

[168] Neil MacCormick and Ota Weinberger, An Institutional Theory of Law (1986).

[169] Iain Stewart, ‘Critical Approaches in Comparative Law’ (2002) Oxford University Comparative Law Forum 4 < articles/stewart.shtml> at 6 September 2004.

[170] Cf Oskar Werner Kägi, Zur Entstehung, Wandlung und Problematik des Gewaltenteilungsprinzipes (1937) 238-44.

[171] Cf Otto Kirchheimer, ‘The Rechtsstaat as Magic Wall’ in Kurt H Wolff and Barrington Moore Jr (eds), The Critical Spirit: Essays in Honor of Herbert Marcuse (1967) 287-312.

[172] Georg Jellinek, Allgemeine Staatslehre (first published 1900; 3rd ed, 1914) 41; Max Weber, ‘“Objectivity” in Social Science and Social Policy’ (1904) in Max Weber, The Methodology of the Social Sciences (Edward A Shils and Henry A Finch trans, 1949 ed) 49-112, 90.

[173] The situation becomes dire when the positive legal ‘order’ is one of common law, beset with similarity induction rather than deducing from elaborated general concepts: cf Vladimir M Sloutsky and Anna V Fisher, ‘Induction and Categorization in Young Children: a Similarity-Based Model’ (2004) 133 Journal of Experimental Psychology: General 166.

[174] Especially in a federal system, where this rôle may derive instead or additionally from the federal division of power: Marbury v Madison, above n 72..

[175] Cf Iain Stewart, ‘Positivist Natural Law in Spencer’s Social Darwinism’ (1997) Archiv für Rechts- und Sozialphilosophie, Beiheft 70(4), 78. The exclusionary expression (or slogan) ‘there is no alternative’ originates with Spencer; in his concept of natural law, that to which there appears to be no alternative is binding by default.

[176] Leighton McDonald, ‘Rule of Law’ in Tony Blackshield et al (eds), The Oxford Companion to the High Court of Australia (2001) 610, 612.

[177] Cf Stewart, above n 169, section 3.2. I propose there that the word ‘law’, when referring to a norm or a normative practice, is a kind of ‘dark performative’ (‘performative’ in J L Austin’s sense) - a type of locution that is the mechanism of closure and is neither logical (because its nature is hidden) nor rhetorical (because there is no attempt to persuade).

[178] Kairys, above n 8, 314-18. On ‘flipping’ see further Stewart, above n 132, 198.

[179] Cf Jenny Hocking, Terror Laws (2004) 2-3.

[180] V F Nourse, ‘Toward a New Constitutional Anatomy’ (2004) 56 Stanford Law Review 835.

AustLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback