Theory and Concepts: Responding to Alexy and Bulygin

Joseph Raz

We all know that the law changes, sometimes by deliberate action, over time.

Explanations of how the law changes are challenging and

controversial. But the fact that it changes is not. The ability of the law to change is generally welcome. Disputes about whether and how it should change are taken as natural, perhaps even beneficial. Not so with theories about the law. Of course, they too change. New theories emerge all the time, and those we have are always controversial. Contemporary academic culture cultivates change almost for its own sake. Yet the existence of disputes about which theory, if any, is correct is often mentioned disapprovingly, The persistence of disagreement is often thought to cast doubt about the enterprise. Science relies on a methodology which commands agreement and allows for progress. Philosophical theories, about the law, or anything else, resolve nothing for they do not command agreement. This makes progress impossible. One cannot build on secure foundations established by previous theorists. One is always starting from the beginning. Does that not show that the enterprise is flawed and futile? A lot can and need be said in response to such doubts. A lot has been written on the subject. My article ‘Can there be a theory of law?’ attempts, among other things, to explain why we can expect the theory of 1

law to keep changing, and why that does not cast any doubt on the cogency of any theory. Put in different terms: I tried to show that there is no final theory, that the task of explaining the nature of law is never-ending, and yet that all steps along the explanatory road are objective, subject to evaluation as true or false, as successful or unsuccessful. The account does so by connecting theory and concept, more particularly by pointing to the connection between the theory of law and the concept of law. Roughly speaking my thought was as follows: the part of the philosophy of law that I was concerned with is the explanation of what law is, or, to put it in other words, the explanation of the nature of law. I suggested that with some important differences an explanation of the nature of law is also an explanation of the concept of law. That establishes the desired link between theory and concept. It also opens the way to explain some of the inevitable, and far from undermining, factors which make for the perpetual persistence of disagreements about the correctness of accounts of the nature of law. Two factors are relevant here: First, that the concept of law changes over time, naturally leading to the need for different explanations. Second, that people’s interests, and their sense of puzzlement change over time, mainly because of changes in other aspects of our culture, including changes in other concepts, leading to new questions about the relations between the concept of law and those other concepts. Hence, again, a need for new explanations of the law. The changes referred to can be continuous. They are, at times, subterranean, escaping easy detection. Hence often it is not easy to attribute the disagreement to these factors, even when it is due to them. And of course it is not my claim that

2

all disagreement is due to such factors, only that their presence explains why the task of that part of legal philosophy is never ending. By and large, neither Alexy nor Bulygin is concerned to challenge directly my conclusions on these points. Rather they dispute other aspects of the article, aspects relied upon in the argument above. For in writing the article I had another goal as well. I argued that, contrary to the view of some writers, like Dworkin, an account of what law is is not an account of the meaning of ‘law’, and that contrary to some other writers, like Hart, such an account is not primarily an account of the concept of law. Rather, it is an account of (some of) the essential properties of law, those whose explanation helps to dissolve the puzzle(s) motivating the theory. In order to clarify this point I had to explore the relations between concepts and essential properties. I tried to reconcile several generally accepted points: First, the identity conditions for properties differ from those for concepts. Second, we explain concepts (in part) by reference to properties that they designate. Following the precept that concepts are to be explained by setting out the conditions for their possession or mastery I exploited the difference between minimal and complete command of concepts to explain the relations between concepts and properties, according to which while complete mastery of concepts involves knowledge of all their essential properties, the difference between concepts is accounted for by the difference between the minimal conditions for their possession. Among other advantages of this account is its ability to explain how people who share a concept can disagree about the correct account of it. I will allow myself to deviate a little and expand on the last point as it was not fully explained in the article: Concepts are explained by setting out 3

the standards which govern their correct use, which enable us to distinguish between a correct and an incorrect application of a concept. There are, I claimed two sets of such standards. Observing one of these sets, the one which corresponds to the minimal condition for the possession of the concept, shows that one has the concept, though one may not have a good understanding of it. Only observing the second set establishes full mastery of the concept. And there are, of course, many intermediate stages. Notice that mastery of the concept implies implicit knowledge of the standards governing its use and application. But it does not require explicit knowledge of them, nor knowledge that one knows them. The result is that people who have the concept may disagree even on what are the minimal conditions for its correct use, for while they know them they may not know what they know. Most of these points are not addressed by Bulygin and Alexy. But they do dispute the basic understanding of the relationship between concepts and properties. As I see it each of them raises two main points of disagreement. I will not say much about Alexy’s comment, primarily because I do not understand it. But I do understand, and I think that he was right to point out that even if many concepts are historical creatures some may be necessary for the very possibility of thought. I should have mentioned that possibility. He does not claim that status for the concept of the law. His essay focuses on two points: First he says that we “will prefer the concept that best captures its [that is the law’s – JR] essential features over all available alternatives.” Second, he presents some arguments about the 4

circumstances in which the concept of law is part of the law. I know what it is for a concept to be a legal concept. Legal concepts are those used primarily to express the content of legal norms and their applications, and only secondarily and derivatively in other contexts. But I do not know what it is for the concept of law, or for any other concept, to be part of the law. The law, as we both understand it, is a normative system, and the parts of a normative system are rules, standards, norms, and the like, but not concepts. Alexy’s discussion of this topic occurs in the context of his reply to my claim that legal philosophy is a distinct activity, distinct from legal scholarship, and of course from the activities of legal authorities, and that its conclusions are not part of the law. But that point seems to have nothing to do with any concept being or not being part of the law. That is why I cannot comment on Alexy’s second main point. The first of Alexy’s points assumes that the law has essential features which the concept of law may fail to capture. But if there is a concept that does not ‘capture’ the essential features of law then it is not the concept of law, or at least that is part of the view I advanced. There is no question of choice of concepts. My paper allows for a different, but in some ways analogous possibility. It has to do with the conditions for minimum mastery of concepts. They can be satisfied even when people’s criteria for determining where the concept applies are contingent and only partially valid. That is, while the criteria are reliable guides for the application of the concept in some circumstances, they may lead to error in others. An obvious concomitant is that people may be able to improve their criteria, discard some in favour of more reliable ones, thus improving in that way as well 5

their understanding of the concept. This assumes that the properties designated by concepts remain the same. Only people’s grasp of them improves as it does when they come closer to complete mastery of it. Alexy says nothing to challenge this point, or to give sense to his alternative. Turning to Bulygin’s comment, I think that contrary to his own view, there is no fundamental disagreement between us regarding his first main point. Bulygin writes: Has a thing properties which are essential per se, independently of the concept used to identify this thing? This seems to be Raz´ opinion. I have serious doubts about this point. Which are the necessary properties of the thing I am looking at in this moment? If we identify it as a lamp, then the property of illuminating is essential; but the stuff it is made of is accidental. But if we identify it as a bronze, then the material it is made of is essential and its illuminating capacity is accidental. Minor divergences notwithstanding, I agree with Bulygin on the main point, and have not written anything inconsistent with it. Roughly speaking the case is as follows: Properties are primarily properties of things, of objects, and secondarily also properties of properties (and relations), of functions, operations, and more. Being a law is a property of an object. I will think of it, as Bulygin does, as a property of a normative system. Some normative systems are legal systems. Others are not. To say of a property, e.g. to be backed by sanctions, that it is an essential property of law is to say that necessarily, anything which has the property of being law also has the property of being backed by sanctions (that necessarily, if a normative system is a legal system then it is a normative system which is backed by sanctions). In other words, things can have essential properties in as much

6

as they are things of a kind. What is essential about the essential properties is that without them the thing will not be of that kind. This formulation brings us close to the core idea: essential properties are essential to the preservation of the identity of the thing concerned. Things are what they are, have an identity, because of the properties that they have. Not every property determines identity. Suppose you identify a location in space and time: the area within a 5m radius, and within the last 5 minutes from here. Suppose you now ask: how many things occupy this location in space-time? The question admits no answer. We need to know what kind of things? How many items of furniture? How many molecules? How many living organisms? The one thing which is clear is that we cannot say: one because we specified one space-time continuum. Being in a certain space-time location does not establish a thing. I believe, however, that Bulygin learns the wrong lesson from this story. He says: No property of a thing is essential per se, but only in relation to a concept. If this is so, then there are no essential properties of a table or of the law, independently of the concepts of a table and of the law that we are using. In other words, it is the concept of a thing that determines which are its essential properties and so, the nature of this thing and not vice versa. He is right of course to say that we identify the kind whose essential properties are in question by using concepts. But what could it mean that ‘there are no essential properties of a table … independently of the concept of a table’? Surely Bulygin does not mean that, for example, if we did not have the concept of a molecule or of inflation, molecules or inflation would have different essential properties than those they do have. Our acquisition of the concepts enables us to think about molecules and, 7

generally, about the things they are concepts of. But acquisition of concepts does not change in any way the things of which they are concepts. As change is change in properties, it follows that acquisition of a concept does not change the properties of the things to which it applies. True, some things are social in a way which means that they cannot exist if their concept does not exist. Marriage may be an example.1 But that is not true generally, and not true of the law. Things of most kinds have the nature they have independently of whether we have concepts enabling us to refer to them. We can test this quite simply: does physical reality change when we acquire the concept of a molecule? Does economic reality change, in the relevant respect, once we acquire the concept of inflation? Clearly not. But then inflation and molecules were there and had the same essential properties whatever concepts we had. This was the reason for proceeding the way I did in the article. To express the view which I put forward there I had to rely on the fact that properties, which, with a few exceptions, things have whether or not we are, or even are capable of being aware of them, differ from concepts, which are cultural products. It seems to me that Bulygin is mistaken in attributing to me a belief in essentialism. There are various things philosophers mean by essentialism. The common one is associated with the thought that some common nouns are rigid designators, that is that the meaning of these common nouns is determined by the way reality is divided (in ways independent of speakers).

1

Though one should be careful here. Usually cultural products, like marriage, can exist so long as a closely related concept exists.

8

It has been popular to think that natural kind terms are of this kind: we identify a certain substance, say by pointing at a sample of it, and baptise it ‘gold’. Thereafter ‘gold’ just means a substance of this kind, and the boundaries of the kind are fixed by the world, not by any way in which we think of it. The nature of a natural kind, that fact that gold is a natural kind, and that it has the properties it has is not up to us. This is just how things are. No one believes that all words are natural kind words, and very few people – perhaps only Michael Moore – believe that ‘the law’ is a rigid designator. My discussion of the way the concept of law changes shows that I hold beliefs inconsistent with such a view. The second major point Bulygin makes does, however, mark a fundamental disagreement between our approaches. 2 Bulygin seems to think that there are many different concepts of law, among which we can choose: I would say that in a certain sense the different concepts of law are products of the theory of law. They try to elucidate the structure of the social institution called “law”, but they do it in different ways. This does not mean that all concepts (or all definitions of the concept of law) stand on the same foot. A concept of law can be better than another, either because it is more exact or because it does not lead to distortions.

2

There are several other minor points Bulygin raises. He is surprised that I think that historical properties can be essential properties of things of certain kinds. It raises in his mind doubt about the kind of necessity I have in mind when saying that essential properties are properties possession of which is necessary for the identity of their possessor (as law, or chess, or whatever). There is nothing special in my use of identity, and there should be no surprise that historical properties can be essential properties. This is true of a variety of social phenomena like nations, and various religions, which are identified historically. Bulygin also ‘has the impression’ that ‘law’ is ambiguous and that concepts and meaning are closer to each other than I allow. But as he has no arguments in support of these claims, and no argument against my attempted explanation of the univocal sense of ‘law’ I cannot comment.

9

This time we encounter a fundamental difference between Bulygin’s view and mine. Bulygin is true to the logical-positivist tradition. He regards the concept of law in the way they understood scientific concepts: It is an old analytic tradition to call the process that leads from a concept to a better, i.e. more exact, concept rational reconstruction or explication (Carnap). The different theories of law strive to formulate a more exact and more suitable concept of law according to some theoretical criteria, as simplicity of presentation, fruitfulness and even elegance. Whatever the merit of this way of understanding scientific concepts, and there is a good deal to recommend it, it is altogether inappropriate to concepts like that of the law, or any other concept which is used widely by people to understand their own and other people’s situation, especially their social situation. Molecules and tigers do not use concepts. We use concepts to enable us to express our thoughts about them, and our knowledge of them. We have reason to think that the various natural sciences are our best source of knowledge. In our practices we often give scientists a privileged position in forging our concepts. We take the concept to be whatever it is in the hands of the scientific community working in the area concerned. Things are different with social concepts such as gifts, ownership, marriage, law, rights, duties, which are used by all to understand themselves and others, and their position in the world. These concepts are not merely tools of understanding, they are part of what shapes the social world we are trying to understand. They, the concepts themselves, are what we are trying to understand, and not tools of explanation. If we could change them at will we would be changing the social reality we were trying to understand. Such a change would not have constituted an explanation of the reality which changed. it would have displayed failure of understanding. Sometimes it is more important to 10

change reality than to understand it. But my paper was not about what is most important, and was not about ways of changing reality. It was only about explaining it.

11