by Matthew H. Kramer. Cambridge: Cambridge University Press, 2007. 262pp. Hardcover. $75.00/£45.00. ISBN: 9780521854160. Paper. $27.99/£15.99. ISBN: 9780521670104. eBook format. $23.00. ISBN: 9780511287398.

Reviewed by Stephen L. Elkin, Department of Government & Politics, University of Maryland. Email: selkin [at] gvpt.umd.edu.


Is the rule of law a good thing? Matthew Kramer does not think so, at least not if the question is posed in this way. It might be good but that depends on moral-political considerations, in particular the purposes to which the rule of law is put. The rule of law itself is simply the result of a set of conditions, and its existence is morally and politically neutral: law is simply what a legal system properly defined produces, whether that is good or bad.

Readers familiar with the debate between H.L.A. Hart and Lon Fuller will recognize the provenance of this argument. It parallels the case made by Hart, but Kramer, strikingly, builds much of the book on arguments made by Fuller. He does so, by dividing Fuller into two theorists as it were: a positivist who argues, with Hart, that law is what comes out of a legal system defined by certain characteristics, and a normative theorist who talks about the “inner morality of the law.” In spite of Fuller spending a good deal of his intellectual life arguing that this division cannot be made – the empirical and normative are necessarily joined with regard to the rule of law – Kramer thinks little if anything is lost and much is gained by doing it. The result is too often unconvincing, with the result that his effort to outline, at least partly, a philosophy of law is substantially weakened.

But before getting to these matters, it is important to note that the book is presented as an introduction to the topic of the relation between objectivity and the rule of law, and is part of a series of Cambridge University Press introductory volumes on philosophy and law. One wonders, however, for whom this particular volume is meant as an introduction. Certainly American undergraduates, even very bright ones, are unlikely to get very far into the book once they realize they are going to be dragged through a bevy of distinctions (about 100 pages worth) on types of objectivity which, if they are going to follow them, will require that they remember with some precision the meaning of terms such as “determinate concreteness” and “transindividual discernibility.” Of course, it might be that such a detailed – and I am sorry to say rather clumsily written – discussion might be necessary for a competent understanding of the rule of law, but, as I will say in a moment, this is apparently not the case. So, the book will not well serve such students. Alas, it is also not likely to be a very useful introduction for scholars who want to get a sophisticated overview of a field about which they know little. Not the least of the reasons is that Kramer regularly says that he cannot go into this particular argument because space forbids it. But too often this is exactly what a scholarly reader [*284] will want. The book is then not for him or her. What then about graduate students? A case could be made in behalf of the book in this respect, but, for myself, I would be disinclined to assign a book to a graduate student trying to work up a field in law and philosophy that uses Fuller’s argument in exactly the way he argued against and that fails to give any real sense of why he argued as he did. Better to assign Hart and Fuller themselves.

Again, before turning to the division of Fuller, there is a puzzle in the way that Kramer wends along his positivist way. On page 109 he says that Fuller’s eight elements of what makes for a legal system are necessary if it is to exist, while on page 143 he says they are necessary and sufficient. Quite apart from the obvious ways in which the difference between the two matters, it is worth considering what the difference means for matters of practice. Of considerable importance in this regard, it is worth noting that if the elements of a rule of law system are necessary but not sufficient, those who wish to realize it will likely run into the following problem: they can make progress in serving some of the other non-necessary conditions but at the cost of reaching whatever threshold is needed for elements that define the rule of law. Concretely, a common problem for those concerned with promoting the rule of law is that the kind of democratic politics that holds a regime together – a variety of favors and political deals – will often make it harder to put in place elements of the rule of law such as implementing the law on the books. Now Kramer in his discussion in this book of what a legal system is, is not concerned with the problem of creating one, but if he is otherwise so concerned, then it matters very much which it is: the elements are necessary, or necessary and sufficient. Perhaps the matter could be put this way: if the rule of law requires something more than the standard elements such as lack of contradictions, generality, and so on – these are necessary but not enough – and the rest of what is needed includes features of democratic political life, then not only will positivist accounts of the rule of law need to look different than they typically do, Kramer’s included. These accounts will also soon become embroiled in the kinds of questions of good practice that they are designed to avoid as they go about defining just what a legal system is.

An additional problem – not in the first instance a matter of practice – arises if the elements of a legal system are necessary but not sufficient: why single out the particular elements Kramer does and not others? Although it is certainly possible that Kramer could show that there are no other necessary conditions for a legal system to exist, it would be nice to have an argument to this effect since it might be that when closely examined, the conditions that make for sufficiency – that are not only necessary – have (again) a certain normative character. Might we not end up having to mix positive and normative elements in our definition of a rule of law system as the distinction between necessary and sufficient proves difficult to maintain? My worry here leads me to raise the more general and important question of whether dividing Fuller’s soul into positivist legal scientist and normative legal philosopher, and with him, much of legal theory is both compelling and valuable. I will proceed by example. [*285]

Kramer notes, with many others, that legal norms must be “addressed to a general class of persons” (p.110). But is it really the case, as he argues, that this is a morally or political neutrally matter? Such a norm, he says, must and will be at work in a legal system with malevolent purposes as well as in good ones. Simplifying a bit, Kramer’s argument here is that, since legal systems are designed to coordinate behavior – any sort of behavior, including slavery and the like – a system that lacks this kind of generality simply cannot function and as such will not be a legal system at all. Really? Many readers will be able to think of legal systems – that is, systems that indeed coordinate behavior above the kind of threshold Kramer regularly invokes (without being very specific) – and that conspicuously do not regularly address legal norms to such a class of persons. The American system of slavery is an obvious example: behavior was nicely coordinated, one might say, but by definition there was one law for the slaves and one for the free. How about caste systems? The answer is plausibly much the same. It is easy enough to see what the problem is: by talking about ‘coordination” of behavior and the like, Kramer means to say that generality is needed not because it is a good thing – a legal system can be used for evil purposes – but for the morally neutral purpose of , we might say, efficient behavior. But once it is clear that we can have lots of coordination and limited generality – at least with regard to what most people would consider the most important areas of life, e.g. the ability to buy and sell property (historically denied to classes of people) as opposed to say traffic laws where slaves might be addressed in the same fashion as the free – then the natural conclusion is that we prefer generality for normative reasons. That is, we prefer it for just the reasons for which proponents of the rule of law have generally argued: it is unjust to use the law to pick out certain people or class of peoples and treat them badly. More importantly, it tells us that some kinds of failures of generality are a good deal more important than others for moral-political reasons, and thus why we should set the threshold for defining whether we have the rule of law in one place rather than another.

Much the same point can be made with regard to whether the legal rules can be readily known by those to whom they are addressed. Kramer comments (p.113) that if they are not generally known “the ostensible legal system would be thoroughly inefficacious in channeling people’s behavior.” Well, yes, if they are little known across all or most domains of behavior. But how about if the rules concerning what a political crime is are mostly secret while the rules about crossing the street, getting a residence permit and who is eligible for what state benefits are well known. And suppose that the authorities secretly define what is covered by the idea of a political crime very broadly, are people’s behavior channeled? Yes indeed, except that some non-trivial number of these people are being channeled right to the gulag. But that is, of course, not the real point, except for those who get sent there. It is rather that talk about channeling behavior does not settle the question of how much and indeed what kind of publicity of legal norms is needed if there is to be a legal system. Once again, the old time religion of the rule of law suggests that how much publicity about what legal norms is best [*286] settled by looking to normative-political criteria.

In the end, I think the question is partly – to stretch a term – ideological. The real choice is how we prefer to think about the world. As a matter of fact, I think, legal positivism cannot be sustained for reasons suggested above. But I would rather, for the moment, put the point this way. What is to be gained by talking about a malevolent rule of law system. The answer presumably is clarity: the rule of law is one thing, a good rule of law another. But if this pushes us into saying that either there was or was not a rule of law system at work in the United States in 1860 (I use this example to avoid the outrage engendered on both sides of the legal positivism debate when Nazi Germany is introduced), then I am dubious about the gain in understanding the rule of law or the United States. It would be a lot more helpful to say that in the 1860, the US had a very flawed rule of law: it worked reasonably well for white people and barely at all for black people, and that, starting with the Civil War amendments, the US moved closer to a full realization of the rule of law – and that was altogether a good thing. In short, it is clearer to say that the rule of law just is both a normative and empirical concept – and it is thus both easy and natural to talk about more or less of the rule of law and why we want more of it. I do not mean to say here that positivists cannot usefully talk about good practice with regard to the rule of law. I just prefer fewer intellectual handstands than they habitually make in efforts to do so. I do not myself find that I have any trouble keeping clear which are normative and which empirical factors when I say they are joined together in the rule of law. It does not seem any harder than remembering that there can be a malevolent rule of law system. Similarly, I do not ordinarily get baffled looks when I say things like, the generality of law helps coordinate behavior and is also necessary for justice. I would even go so far as to say that I might get looks of comprehension if I said that any plausible account of what it means to coordinate behavior leads us to questions of efficiency and Pareto optimality – and that these in turn are species of utilitarian judgment. I might even get a smile when I conclude that even if coordination could be separated off from justice or fairness, it would thus still be a normative judgment.

In this context, it is worth emphasizing that Fuller was a lot clearer than Kramer (although not as clear as he might have been) about the kind of normative purposes the rule of law serves. He was concerned not with some abstract conception of coordinating behavior – assuming that were possible to define, which I doubt it is. Rather, Fuller was a liberal and ultimately valued law as a (perhaps the) way of making it possible for individuals to pursue their purposes in effective ways. Law is the servant of liberty – and the internal morality of law is inextricably linked to the external morality of liberalism. Now, it is of course possible to have a measure of the rule of law in broadly non-liberal regimes. But, alas, for the proponents of such regimes, the rule of law inevitably provides a measure of individual liberty within them. Fuller, I believe, would have agreed.

Perhaps a good way to formulate the difference in view between Kramer and me (not to mention between Kramer and [*287] Fuller as he wrote as opposed to the dismembered Fuller Kramer presents) is whether the moral-political value of the rule of law is external to it or internal to it, it’s inner morality. I will only add in this regard that I suspect that we would all be better off if we stopped writing about the virtues of positivist and other views of the rule of law, and examine what difference it makes for our understanding of, say, the Polish legal system in 1960 if we approach it a la Kramer or Fuller. I would say much the same thing about how much is to be gained in such understanding if we traverse Kramer’s discussion of objectivity – the connection to his positivism being that both stem from an aspiration to be scientific: opinion and normative matters are to be sternly put aside. Kramer tells us after the almost 100 pages on objectivity, and about 40 pages from the finish of a 232 page book, that “nothing of practical importance will be settled by reference to” his discussion of “mind-independence.” This is the first aspect of objectivity he discusses in the book.

© Copyright 2008 by the author, Stephen L. Elkin.