Vol. 28 No. 5 · 9 March 2006

What security is there against arbitrary government?

John Gardner

2983 words
Rhetoric and the Rule of Law: A Theory of Legal Reasoning 
by Neil MacCormick.
Oxford, 287 pp., £40, July 2005, 0 19 826878 5
Show More
Show More

The government of Securitania deports some supposed enemies of the people and puts others under house arrest; public scrutiny of these measures in the ordinary courts is denied. Disruptive people against whom no crime can be proved are subjected to orders obtained from magistrates on hearsay evidence that make it a crime in future for them to do ordinary things (such as going shopping) that would not be a crime if done by anyone else. A man thought to be acting suspiciously is shot dead by plainclothes police officers; the immediate reaction of the police chief is to obstruct the routine criminal investigation of the killing. Legislation is underway to criminalise the ‘glorification’ of those using force to resist oppressive policies at home and repressive regimes abroad; objections to the breadth and vagueness of the measure are met with the response that the authorities can be trusted to ensure that those engaging in legitimate debate will not be prosecuted. Do any of these developments make the people of Securitania more secure? Probably not. But in one way they clearly make them a lot less secure: the people of Securitania are progressively being deprived of the rudimentary security of living under the rule of law.

The rule of law is an ideal of good government. By calling it an ideal, I am not suggesting it is utopian. I mean only that it is a cluster of principles from which isolated departures may be inevitable even in a decent regime, but substantial and committed compliance with which is one key determinant of a regime’s decency. The most basic principle is this: when we live under the rule of law, the agents of the state are legally and publicly accountable in the courts like everyone else, and the courts are independent enough of the (other) organs of state to make this accountability a significant check on state power. The government of Securitania does not ignore this principle. On the contrary, it uses its de facto control of the legislative assembly to secure the passage of laws that license it to do most of the things it does. Rarely does it actually break the law. But that is because it only rarely needs to break the law. The laws that it steers through the legislative assembly provide it with generous cover for most of the things that it is inclined to do. The result is that it becomes increasingly difficult for people in Securitania to rely on the law in advance of their actions to predict how they will fare in the event that their actions come to the attention of the authorities. The law gives the authorities extensive leeway to act against people or not as they choose. Those acted against have few and difficult legal avenues for complaint about their official treatment, and there is always a risk that the government will find some technically legal way to frustrate whatever complaints are launched through the courts. Therefore, people generally have to rely on the patronage or forbearance of petty officials to provide them with whatever security they enjoy against governmental abuses of power.

This goes to show that the rule of law is not only an ideal for government; it is also an ideal for law. The government of Securitania cannot pride itself on respecting the rule of law merely by ensuring that it and its officials always act within the law. If Securitanians are to live under the rule of law, the law itself must also live up to certain standards. It must be such as to give Securitanians reliable guidance in advance about where the line falls between what is permitted and what is proscribed, and to enable them to work out what will be the likely official reaction to their failure to stay on the right side of the line. The law therefore needs to be open, clear, stable, general, consistent and not retroactive, and must be complied with impartially (without fear or favour) by the officials charged with administering it. Without law that lives up to these standards, acting within the law is a hollow achievement for the government and a constant struggle for the population.

Failing to live up to the ideal of the rule of law represents one important way in which the law can be bad; but living up to the ideal is not enough by itself to make the law good. There are other important values for it to pursue apart from the reliability of its own guidance. Sometimes these values conflict with the rule of law and warrant departures from it. But it is not just a simple matter of weighing the rule of law against other ideals. That is because of what Bernard Williams called the ‘Basic Legitimation Demand’: government and law lose their legitimacy if they resort to using terror themselves in their efforts to protect us from terror – if they become, as Williams puts it, ‘part of the problem’. The government is clearly becoming part of the problem in Securitania. But we should not lean so far in the opposite direction that we regard the law’s ability to provide accessible guidance in advance as the be-all and end-all of legal policy. The tendency among lawyers to elevate ‘legal certainty’ to this position is one thing that has given the rule of law a bad name in progressive circles, and hence given succour to governments that look on themselves as progressive but have little commitment to the rule of law. Such governments regard the law, as filtered through the ideology of lawyers, as an obstacle to the pursuit of progressive social policies. Neil MacCormick’s Rhetoric and the Rule of Law: A Theory of Legal Reasoning aims to moderate both poles in this debate: to temper the lawyer’s enthusiasm for legal certainty, while insisting on the very great importance of the rule of law as an ideal – for government and law – that should animate and constrain progressive governments no less than others.

The book forms the second volume of MacCormick’s grand project on ‘Law, State and Practical Reason’, of which the first volume, Questioning Sovereignty, was published in 1999. The project as a whole – intended to fill four volumes – addresses the most persistent general problems in the theory of law as they intersect with problems concerning state power. This second volume reprises a range of topics that MacCormick originally tackled in his 1978 classic, Legal Reasoning and Legal Theory. In the new book, he responds to his critics and to other relevant writings that have appeared in the meantime, and draws attention to the lessons that he has learned from his friends. MacCormick is constantly welcoming of new ideas and gives credit freely to others for seeing things that he missed. He sees insight and inspiration in everyone else’s work, sometimes to a fault. For example, the book begins with the claim that it is ‘a contribution to the “new rhetoric” pioneered by my late friend and respected senior colleague Chaim Perelman, along with Lucie Olbrechts-Tyteca’. This explains the reference to ‘rhetoric’ in the book’s title, but in the end nothing turns on this conceptualisation of the book’s project, and one wonders whether, apart from the opportunity for a kindness, the book would have lost anything by making no mention at all of rhetoric, new or otherwise.

The upshot of the book’s wide range of reference and graciousness towards others is that it does not entirely work as a self-contained study of legal reasoning. In many cases one needs to trace the arguments backwards to see why the protagonists, including MacCormick, are saying what they are saying. It is partly to lend some unity to the many fragments of debate thus collected that MacCormick foregrounds the ideal of the rule of law. He deftly represents the main debates about the nature of legal reasoning as debates about the way legal reasoning is affected by the demands that the rule of law places on it. The rule of law requires that the law not be retroactive; but surely judges often arrive at new rulings which make a retroactive change in the law applicable to the parties appearing before them? The rule of law requires that the law be consistent; but surely legal reasoning often points in at least two inconsistent directions, so that either side in a dispute might reasonably expect to win? The rule of law requires that the law be stable and clear; but isn’t it a ‘commonplace’, as MacCormick says, that ‘the law on any point is perennially arguable’? As MacCormick brilliantly illustrates with case-law from several countries, such tensions are inevitable in any developed legal system, and cannot but be confronted by its lawyers and judges every day. Whatever the rule of law ideally expects the law to be like, it cannot expect absolute consistency, stability, clarity, generality and so on. Such absolutes would be irreconcilable with various characteristic and (up to a point) desirable features of legal reasoning itself.

What makes these features of legal reasoning desirable? Often, as MacCormick points out, they are made desirable by other aspects of the ideal of the rule of law. In many cases, in other words, the desiderata of the rule of law conflict among themselves, and insisting on one of them cannot but be at the expense of another. MacCormick offers a good example. Because the stakes are especially high, the rule of law demands that criminal laws be especially stable and predictable in their application. Yet it also insists on the right of the defence in a criminal trial to challenge all aspects of the prosecution case, including the interpretation of the law. ‘There is no security against arbitrary government,’ MacCormick observes, ‘unless such challenges are freely permitted.’ Such challenges tend, however, to destabilise the criminal law and make its application less predictable. So the rule of law is not just the rule of stability and predictability. So-called legal certainty ‘is not the only value of, or present in, the rule of law, though it is one benefit which people rightly look to legislators and judges to confer on them so far as possible’. ‘So far as possible’ here means something like ‘so far as the other desiderata of the rule of law allow’. MacCormick’s point, reprised several times, is that the tensions endemic in legal reasoning – between generality and specificity, between stability and flexibility, between clarity and arguability, and so forth – often turn out to reflect conflicts not between the rule of law and some other ideal of legal policy, but between different demands of the rule of law itself. So the demands cannot possibly be interpreted as absolute: the rule of law needs to be understood as capable of compromising with itself.

MacCormick gives particular attention, and particularly lively attention, to the question of how much generality in legal reasoning the rule of law demands. The Securitanian trick of inventing special criminal offences for named individuals strays a long way from the acceptable. On the other hand, judges do need to be able to make rulings applicable to specific named individuals, and sometimes in the process they need to adapt the law to suit the unanticipated facts of a novel case. MacCormick sets the scene for this topic with Solomon’s famous judgment in the ‘Whose baby?’ case. The case is problematic as an illustration of the special dilemmas of legal reasoning because it arguably involves no law at all. Arguably, Solomon in the story is not a judge but an arbitrator, free to decide on the raw merits of the case. Judges, as MacCormick notes, are never quite that free. They are always bound by at least some relevant law and (to add to their woes) the rulings they issue in particular cases often yield new law that will bind future judges. Modern case-law reveals the plight of judges trying to cope with Solomonic questions complicated by the fact that they are acting as both receivers and givers of law.

MacCormick points, by way of example, to the contortions of the Court of Appeal in the famous conjoined twins case of five years ago. When doctors must end the life of one conjoined twin to save the life of the other, does the law license the killing? The court struggled to decide the case within the existing law of murder without making any ruling that would be generalisable to other future cases. But how could it do so? Any exception to the general law of murder must be capable of some rational explanation, or else the court has no business making it. But if the exception is capable of rational explanation, the rational explanation is necessarily capable of re-application to any future case that falls under it, involving other people and other incidents. There is nothing that can properly be done to avoid that. There is no legitimate way for judges to avoid getting entangled with the generalisation of their rulings. The rule of law, to put it another way, requires rulings affecting named individuals to constitute applications of general rules affecting indefinite numbers of possible individuals. The Securitanian magistrates defy this principle by ruling that it will be a crime for some named person to go shopping, without explaining how going shopping is an instance of some more general crime that could be committed by other people.

It may be said that this has little to do with the rule of law. It just reflects a general principle of rational intelligibility, applicable to all realms of life. It applies no less to relations between friends or to the appreciation of the arts. ‘What do you like about me?’ ‘What do you see in the work of Jasper Johns?’ These are requests for further explanation of someone’s reactions to specific people and objects. The answer, to be satisfactory, needs to cite facts about those people and objects which could in principle be repeated across an indefinite number of further objects and people. True, it is an interim answer to these questions to say: ‘I don’t know, it’s just the way I feel.’ This is an interim answer because it puts the absence of a further explanation down to ignorance. On the other hand, ‘nothing’ is not even an interim answer. If there is nothing about me that you like – no fact or combination of facts that grounds your liking, and hence that could in principle ground your liking of someone else – then you have no business liking me. Your position is unintelligible as the position of a reasonable human being. The law is no different. Never mind the rule of law: the law should be rationally intelligible.

Thinking along similar lines, MacCormick occasionally finds himself troubled by the thought that there is nothing special about legal reasoning. He writes, for example, that ‘any commitment to impartiality between different individuals and different cases entails requiring that the grounds of judgment in this case be deemed repeatable in future cases,’ and he makes clear that this requirement applies equally outside the law: ‘There is . . . no justification without universalisation.’ But these remarks are apt to mislead the unwary reader. As MacCormick goes on to point out, the rule of law demands much more in the way of impartiality and much more in the way of generality than is required by the ethics of ordinary non-legal life. Under the heading of impartiality, it requires judges who are not only open to reason, but also willing to stand up for the law without fear or favour; and that applies equally when the fear itself is reasonable, such that non-judges (politicians, police officers, members of the public) might naturally be expected to submit to it. Under the heading of generality, the rule of law requires not only rationally intelligible explanations for particular rulings, but also rationally intelligible explanations for particular rulings in terms of rules, considerations that could not only be used by other judges but can also bind them.

So it is never sufficient for the government of Securitania to point out that its proposed courses of action would be reasonable were it not for the law, with the implication that the law should be bending to the government’s will rather than vice versa. If the country is to live under the rule of law – thereby meeting the Basic Legitimation Demand – then the law exists to make a difference to what can reasonably be done, and the judges and lawyers are there to insist on the law’s making that difference even in the face of government hostility and pressure. (Hence the grave fears about the Legislative and Regulatory Reform Bill currently before Parliament. This outrageous bill would allow major changes to Parliamentary legislation to be undertaken directly by the government.)

The whole of this line of thought is present in MacCormick’s book. But it is not present as a line of thought, and it is not conspicuous. It takes shape in fits and starts, almost fortuitously, as the book is pulled this way and that by the uneven and sometimes tangential interventions of other thinkers to whom the ever generous MacCormick feels impelled to respond. The result, although a subtle and thoughtful work by any standard, is a less luminous explanation of the nature of legal reasoning, and a less powerful defence of the rule of law, than we might have hoped for and expected from one of our most publicly engaged legal theorists.

It is not only in Securitania that the rule of law is under threat. Rarely has it been more important for the arguments of the ideal’s ablest defenders to be well marshalled. In this respect Rhetoric and the Rule of Law disappoints: its excellent arguments are not as well marshalled as they could be.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN


Please include name, address, and a telephone number.

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences