THE CONCEPT OF A LEGAL SYSTEM

An Introduction to the Theory of Legal System SECOND EDITION

JOSEPH RAZ

CLARENDON PRESS · OXFORD

First published 1970

Second edition 1980

Reprinted 1990, 1997

VIII THE IDENTITY OF LEGAL SYSTEMS

요약

다음은 Joseph Raz의 『The Concept of a Legal System』 제8장(제2판 기준)의 핵심 내용을 정리한 것입니다. 이 장은 법체계의 정체성(identity) 문제를 다루며, 순간적 법체계(momentary legal systems)와 비순간적 법체계(non-momentary legal systems)를 구분하고, 각각의 동일성 기준과 구성요건에 대해 분석합니다.

1. 비순간적 법체계의 동일성 (VIII.1)

문제의 핵심: 두 순간적 법체계가 동일한 법체계에 속하는지를 결정하는 연속성(continuity)의 기준.
Austin의 견해: 동일한 최종 입법자(sovereign)가 법을 제정한 경우 동일한 법체계로 간주 (불충분함).
Kelsen의 견해: 후속 법이 선행 법에 의해 헌법적으로 승인되었는가 여부로 판단. 그러나 이는 새로운 “원초적 법(original law)”이 기존 체계를 자동적으로 단절시키는 것은 아님.
Raz의 비판:
- 헌법적 연속성은 유의미하나, 결정적인 요소는 아님.
- 법체계는 종교, 국가, 부족 등 사회적 실체의 일부이며, 그 실체의 동일성에 따라 법체계의 동일성도 결정됨.
- 따라서 법체계의 동일성은 법철학적 기준만으로는 판단할 수 없고, 사회학적·정치적 요소도 반영되어야 함.

2. 순간적 법체계의 동일성과 구성요건 (VIII.2)

동일성 기준 제시:
1. 주어진 규범 명제 집합의 모든 명제가 동일한 순간적 법체계를 부분적으로 기술할 것.
2. 동일한 순간적 법체계를 기술하는 모든 규범 명제는 이 집합에 의해 함의되어야 할 것.
결정적 개념: ‘소속(membership)’
- 어떤 규범 명제가 주어진 집합과 같은 법체계를 기술하는지를 판단하는 기준이 필요함.
기존 접근 비판:
- Austin: 법의 창설 방식(기원)에만 의존.
- Holland: 법의 집행 방식에만 의존.
- 둘 다 불충분함. Salmond의 입장을 참고해 법체계의 제도적 성격과 ‘승인(recognition)’의 개념을 중심으로 새 기준 제시.
Raz의 제안: ‘기본 법집행 기관(primary law-applying organ)’ 개념을 도입.
- 이 기관이 어떤 규범을 승인(recognize)하는가에 따라 그 규범이 해당 순간적 법체계에 속하는지를 판단.
- 승인이란: “해당 규범이 제기되었을 때 그 기관이 그것에 근거하여 작동할 것”이라는 반사실적 조건을 충족.
기본 기관의 조건:
- 제재 집행 여부를 결정할 권한을 가지고,
- 사회 내에서 비교적 안정적이고 지속적으로 기능함.
결과:
- 하나 이상의 기본 법집행 기관이 존재하면, 법적 제도와 비법적 제도의 경계가 생김.
- 기본 기관의 판단이 규범 체계를 구성하며, 이로써 순간적 법체계가 정의됨.

3. 승인 규칙(rule of recognition) 비판적 고찰 (VIII.3)

Hart의 견해 요약:
- 승인 규칙은 법체계 내부에 존재하며, 그 법체계의 규범들을 식별할 수 있게 해주는 규칙.
- 이는 법집행 기관(법관, 관료 등)의 관행적 적용에 의해 존재함.
Raz의 비판:
- Hart의 승인 규칙은 ‘법으로서 존재하는지’의 판단을 또 다른 법(규칙)에 의존시키며, 결국 순환 논리에 빠짐.
- 승인 규칙은 궁극적으로 법이 아닌 법에 관한 일반이론적 기준(jurisprudential criterion)에 의해 설명되어야 함.
핵심 주장:
- 모든 법이 승인 규칙에 의해 존재하는 것은 아님.
- 일부 법은 법집행기관의 실천을 통해 승인되며, 이는 규범적 규칙에 의한 것이 아닌 사실적 기준(factual practice)에 근거함.
- 법집행기관이 특정 규범을 기반으로 판단을 내릴 것이라는 충분한 정황이 존재하면, 그 규범은 법으로 간주될 수 있음.

원문 참조

하트(Hart)가 승인 규칙(rule of recognition) 이론을 채택하게 된 근본적인 이유는 다음 문장에서 표현된다. “어떤 제안된 규칙이 법적으로 유효한지 여부가 문제로 제기되면, 우리는 그 질문에 답하기 위해 다른 어떤 규칙에 의해 제공된 유효성 기준을 사용해야 한다.” 문제가 되는 것은 바로 이 전제이다. 내게는, 어떤 특정한 법이 하나의 법체계에서 법으로서 존재하는지를 묻는 질문에 답하기 위해서는, 결국 어떤 법을 가리키는 것이 아니라(refer not to a law) 법에 관한 일반이론적 기준(jurisprudential criterion)을 가리켜야 하는 것으로 보인다. 궁극적으로 이 질문에 답하는 자가 가리켜야 할 것은(refer to), 어떤 개별 법(a law)을 기술하는(describe) 것이 아니라 법에 대한 일반적 참(general truth about law)을 기술하는 일반적 진술(a general statement)이다. 일부 법체계에서는 특정 조건을 충족하는 모든 법을 적용하도록 어떤 기관에 의무를 부과하는 법이 있을 수 있고, 실제로 그런 법이 해당 법체계 내의 모든 법을 포함하는 경우도 있을 수 있다. 하지만 그러한 법이 존재한다고 해도(그리고 그것은 항상 그런 것은 아니다), 그러한 법들로 구성된 법체계가 해당 법체계에 속하는 이유는 그러한 승인 규칙(rule of recognition) 때문이 아니라, 그 법들이 모두 1차적 법집행기관(primary organs)에 의해 승인되었기 때문이다. 모든 법체계에 그러한 승인 규칙이 하나씩 존재하는 것이 아니라는 점은 다음 두 가지를 통해 확인할 수 있다:

하트가 어떻게 해서 모든 법체계에 단 하나의 승인 규칙만이 존재한다고 보는 견해를 정당화하는지 불분명하다. 왜 여러 종류의 공무원 각각에게 적용되는 여러 승인 규칙이 있다고 보지 않는가? 왜 다양한 승인 규칙들이 다양한 유형의 법을 승인하도록 규정하고 있다고 보지 않는가?

앞 절에서 언급한 바와 같이, 기본 기관의 행위는 소속(membership)의 기준에서 핵심적인 요소이다. 그러나 하트의 이론이 전제하는 것처럼, 이 기관들이 항상 어떤 의무의 이행으로서 행동한다고 볼 이유는 없다. 어떤 법을 승인할 의무가 있을 수는 있지만, 동시에 다른 법들에 대해서는 승인할 법적 자유가 있을 수 있다. 예컨대 하트가 주장하듯, 어떤 법이 제재나 법적 구제수단 없이도 두루퍼진 비판적 반응(diffused critical reactions)만으로 D-법(의무 부과 규범, duty-imposing laws)이 될 수 있다 하더라도, 기본 기관들이 어떤 법을 승인하지 않거나 다른 법을 새롭게 승인하기 시작한다고 해서 반드시 비판적 반응에 직면해야 하는 것은 아니다.

The fundamental reason which moved Hart to adopt his doctrine of the rule of recognition is expressed in the following sentence: ‘If the question is raised whether some suggested rule is legally valid, we must, in order to answer the question, use a criterion of validity provided by some other rule.’ [1] It is this assumption that is questionable. It seems to me that to answer the question whether a certain suggested law exists as a law in a certain legal system one must ultimately refer not to a law but to a jurisprudential criterion.[2] Ultimately one must refer to a general statement that does not describe a law but a general truth about law. In some legal systems there may be laws which oblige certain organs to apply all the laws fulfilling a certain condition, and it may be that these laws are in fact all the laws of the system. But even when such laws exist, which is not always the case, the laws of the system belong to the system not because of this rule of recognition, but because they are all recognized by the primary organs.That it is not the case that in every legal system there is one such rule of recognition can be seen by considering the following two points:

1. It is not clear on what Hart bases his view that there is only one rule of recognition in every legal system. Why not say that there are various rules of recognition, each addressed to a different kind of officials? Why not say that various rules of recognition prescribe the recognition of various types of laws?
1. Though, as has been suggested in the previous section, the behaviour of primary organs is the key to the criterion of membership, there is no reason to think that they always act in fulfilment of duties, as Hart’s theory entails. They may be under a duty to recognize some laws and yet be legally free to recognize or not recognize certain others.[3] Suppose, as Hart does, that legal laws can be D-laws merely in virtue of diffused critical reactions, without being backed by sanctions or other legal remedies. Even so, it need not be the case that primary organs will meet with critical reactions if they stop recognizing certain laws or begin recognizing certain other laws.

1 CL, p. 103. My italics. 2 It should be remembered that we are concerned with the ordinary man’s point of view, not with that of a judge faced with the question ‘which law ought to be recognized?’. 3 Cf. Ch. II, sect. 4 above.

4. 부록: ‘종이 법(paper laws)’ 문제에 대한 해결 제안

Kelsen: 최소한의 효력(minimum efficacy)이 있어야 법으로 간주됨.
Raz: 중요한 것은 법집행기관이 인식하고 작동하는가 여부.
- 대중이나 경찰이 무시하더라도, 법집행기관이 해당 법에 따라 판단하고 제재할 준비가 되어 있다면 여전히 법임.
- 따라서 효력보다는 승인이 법의 존재 조건.

핵심 정리

구분	요점
동일성(비순간적 법체계)	사회 실체의 정체성과 연속성에 의존, 법 자체의 형식적 창설 방식으로는 불충분
동일성(순간적 법체계)	모든 규범이 동일 시스템을 기술하고, 다른 모든 해당 규범을 포함함
소속 기준	‘기본 법집행 기관’이 해당 규범을 인식하는가 여부에 따름
법의 존재 조건	법집행기관의 실질적 인식이 핵심, 규범의 실제 영향력이나 창설 기원은 부차적
Hart의 승인 규칙에 대한 비판	모든 법체계에 단일 승인 규칙이 있다는 전제를 부정하며, 규범의 존재는 실천과 경험에 의한 인식으로 파악해야 함

이러한 Raz의 접근은 법체계를 단지 규칙의 총합이 아닌 제도적으로 구현된, 법집행 기관과 연계된 실천 체계로 이해하려는 시도입니다. 특히 “소속” 개념과 법의 존재 조건을 명확히 구분하고 분석하려는 점에서 Hart의 이론과 뚜렷이 차별화됩니다.

원문

VIII THE IDENTITY OF LEGAL SYSTEMS

VIII.1: THE IDENTITY OF NON-MOMENTARY LEGAL SYSTEMS

Discussion of the structure of legal systems in the previous two chapters has assumed that the content of a given legal system is known, and the proper way to represent it as a system of interrelated laws has been explored. This chapter is concerned with the criterion for determining the content of legal systems, i.e. with the criterion of identity of legal systems. The key to the problem of structure is the criterion determining which of the complete descriptions of a legal system is its proper description. The problem of identity is the problem of finding a criterion determining whether a given set of normative statements is a complete description of a legal system.

Corresponding to the distinction between a legal system and a momentary legal system there are two criteria of identity–one specifying the way in which the identity of legal systems can be ascertained, the other providing a way in which the identity of momentary legal systems can be ascertained.

Apart from a few remarks in this section, this chapter is concerned only with the problem of identity of momentary systems. The crux of the problem of the identity of (non-momentary) legal systems is the question of continuity, namely what events disrupt the continuous existence of a legal system, bring about its disappearance, and, perhaps, result in the creation of a new legal system in its stead. In other words, the crucial problem is the problem of deciding whether two given momentary systems belong to the same legal system.

In previous chapters the solutions of this problem implied by the theories of both Austin and Kelsen were considered and rejected. They admit that the fact that two momentary systems contain many laws which are identical in content is no proof that they belong to the same legal system. Austin implies that the identity of the ultimate legislator of the laws of the two

-187-

momentary systems is both necessary and sufficient for them to belong to one legal system.

Kelsen avoids some of the weaknesses of this position; for him the constitutionality of the creation of the laws of the one momentary system relative to the other is the deciding factor. According to his theory, two momentary systems A and B belong to the same legal system if, and only if, the creation of all the laws of B which are not identical with laws of A was authorized by the laws of A. The creation of a law is authorized by another if the law-creating acts creating the first amount to the exercise of legislative powers conferred by the second, or by another law the creation of which is authorized by the second.

If the creation of a law is not authorized by another law it can be called an ‘original law’. Kelsen’s criterion for the identity of legal systems presupposes that no momentary system contains an original law which does not belong also to the first momentary system of the same legal system. But this presupposition, as has been argued in Chapter V, section 4 above, is not justified. The continuity of a legal system is not necessarily disrupted by the creation of new original laws. Nor is the fact that the creation of a law is authorized by a law belonging to a certain legal system a sufficient proof that the authorized law belongs to that system. A country may be granted independence by a law of another country authorizing all its laws; nevertheless, its laws form a separate legal system.

The ‘constitutional continuity’ of the laws is only one, and not the most important, factor in determining whether two momentary systems belong to the same legal system. Another factor is the content of the non-authorized law. The creation of a new original law disrupts the continuity of a legal system only if it is a constitutional law of great importance.

But neither the ‘constitutional continuity’ of the new laws nor their content are necessary or sufficient conditions for establishing the continuity or lack of continuity of legal systems. Legal systems are always legal systems of complex forms of social life, such as religions, states, regimes, tribes, etc, Legal systems serve as one, but only one, of the defining features of these.

Significant and unconstitutional changes of law may be

-188-

important enough to establish an alteration in the identity of the social entity of which the legal system is a part, but other factors have to be taken into account as well. The identity of legal systems depends on the identity of the social forms to which they belong. The criterion of identity of legal systems is therefore determined not only by jurisprudential or legal considerations but by other considerations as well, considerations belonging to other social sciences.

Not wishing to trespass on other fields, I shall confine myself henceforth to the problem of the identity of momentary systems. It should not, however, be assumed that momentary systems can be analysed independently of the legal systems to which they belong. It will become clear from the arguments of this and of the next chapter that both the identity and the existence of momentary systems can be determined only by reference to other momentary systems of the same legal system. But the solution of these problems does not presuppose an ability to determine the precise boundaries of legal systems.

VIII.2: IDENTITY AND MEMBERSHIP IN MOMENTARY LEGAL SYSTEMS

The criterion of identity of momentary systems can be formulated as follows: A set of normative statements is a complete description of a momentary legal system if, and only if, (1) every one of the statements in it describes (part of) the same momentary system as all the others, and (2) every normative statement which describes (part of) the same momentary system is entailed by that set.

This formulation presupposes a criterion of membership in momentary systems, i.e. a criterion for ascertaining whether a given normative statement describes (part of) the same momentary system as the one which is described by a given set of normative statements. It is with the problem of membership in momentary legal systems that the rest of this section is concerned.

The definition of ‘a complete description’ of a momentary system [1] makes it clear that, if a normative statement is entailed by a set of normative statements, then it describes the same system as is described by the set. The difficulty in finding a

________ 1 Cf. Ch. III, sect. 1 above.

-189-

criterion of membership is in discovering a condition for a given normative statement partly describing the same system as a given set of normative statements, even though it is not entailed by it. There are various ways in which one may try to tackle this problem. In Chapters II and V we have criticized two attempted solutions based on the principle of origin, i.e. making the facts creating laws the sole factor determining the legal system to which they belong. Here we shall attempt to sketch a more hopeful line of approach, which can be said to be based on the principle of authoritative recognition.

The defects of Austin’s criterion of membership were soon discovered by other legal theorists. Some attempted, without much success, to perfect his criterion, and remained faithful to the principle of origin. Others abandoned this principle altogether. One of the latter was Holland, who defined laws as ‘a general rule of external human action enforced by a sovereign political authority’.[^1] Holland, as opposed to Austin, thinks that it is not the way in which laws are created but the way in which they are enforced that determines their nature as ‘positive’ laws, as distinguished from non-legal custom, etc.; and, by implication, it is this that determines also to which legal system a law belongs.

His definition, however, is deficient in many respects. Not least among them is that it accepts Austin’s theory of sovereignty in only slightly modified form, and that it assumes that every law can meaningfully be said to be enforced. It is clear from the discussion in the previous two chapters that: (a) Not every law which is a ‘rule of behaviour’, i.e. a norm, can be enforced. Power conferring laws cannot be enforced. (b) No law which is not a norm can be enforced.

Both these shortcomings are avoided by Salmond’s definition of a law. Salmond explains that, whereas there are various ways in which laws are created, ‘all law, however made, is recognized and administered by the courts, and no rules are recognized by the courts which are not rules of law. It is, therefore, to the courts and not to the legislature that we must go in order to ascertain the true nature of the law.’ Accordingly he defines law as follows: ‘The law consists of the rules recognized and acted on by courts of justice.’[^2]

2 Salmond on Jurisprudence, p. 41.

-190-

Unlike Austin’s definition of law, Salmond’s definition does not include an embryonic theory of legal system. Salmond’s aim is merely to point to the difference between legal systems and other norms and normative systems. He therefore assumes the normativity of the law, disregards its coercive nature, and bases his definition on the fact that legal systems are institutionalized. It is this that distinguishes them from other normative systems. The institutionalized nature of the law is manifested, according to Salmond, in the existence and operations of the courts, i.e. of certain law-applying organs. In this he differs from Bentham, Austin, and Kelsen, who all concentrate on and emphasize the importance of the law-creating organs, and regard their particular mode of creating laws as one of the distinguishing features of legal systems.

Salmond’s main contention is sound. Not every law is created by law-creating organs, and though the importance of legislation as a law-creating method is characteristic of modern legal systems, it is not characteristic of every legal system, nor is any other law-creating method. On the other hand, it can be said that every legal system institutes law-applying organs which recognize every law of the system. Having said this, we must admit that there are many borderline cases. Some such cases will be mentioned in the sequel. But borderline cases are unavoidable in problems like this, and their existence does not in itself detract from

________ 1 Jurisprudence, p. 40.

the value of general statements on the subject, though these should not be interpreted too dogmatically.

The importance of law-applying organs in the law is manifested in various ways which should be carefully separated. One of the most important aspects of their role, the aspect which more than any other can be connected with a ‘definition of law’, is the relevance of law-applying organs to the criterion of membership in momentary legal systems. The criterion formulated below follows the philosophical tradition established by Salmond, but it deviates from the particulars of his position to a considerable extent. A statement will be said to describe a primary organ if it identifies it and ascribes to it primary legal powers.

A given normative statement describes (part of) the same momentary system as is described by a given set of normative

-191-

statements, if it describes a primary law-applying organ that recognizes1 the laws described by this set of normative statements; or if a primary law-applying organ described by the set of normative statements recognizes the law which the given statement describes. Accordingly, it can be said that a momentary legal system contains all, and only all, the laws recognized by a primary law-applying organ which it institutes.It may be found necessary to conceive of a legal system not as a set of laws recognized by all the primary law-applying organs instituted under it–as is implied in this criterion–but as a set of substantially, but only partly overlapping, sets of laws, each recognized by one or more of the organs instituted under it. The proposed criterion of membership can be relaxed to take account of that need. The various ways in which this can be done will not be explored here.The meaning and implication of this criterion of membership should be carefully scrutinized. It assumes that every momentary legal system institutes at least one law-applying organ of a type the nature of which will be explained below. The criterion thereby presupposes and manifests a certain view of the institutional nature of law. Furthermore, the criterion assumes that the identity and actions of primary law-applying organs are essential in establishing the membership of a legal system. Given only two normative statements, neither of which describes a primary lawapplying organ, it is impossible to determine whether or not they describe part of the same momentary legal system.The criterion is based on two key concepts, ‘primary lawapplying organs’ and ‘recognizing a law’. The following remarks are only the beginning of an explanation of these concepts. A primary law-applying organ (‘ primary organ’ for short) is an organ which is authorized to decide whether the use of force in certain circumstances is forbidden or permitted by law. This concept is indeed the key to the criterion. It was argued in Chapter VII, section 4 that:

Every legal system prohibits the use of force in certain conditions (at least when force is used to obstruct the execution

[1] ‘Recognizes’ here means ‘is certain to recognize, if the question be raised before it’. Or, in so far as the system described no longer exists, it means that the organ did recognize the laws in question when it existed.

–192–

of sanctions) and permits (or prescribes) the use of force in certain other circumstances, i.e. in the course of executing certain sanctions.

In every legal system all the laws have internal relations to the laws forbidding the use of force or to the laws permitting or prescribing the use of force in the execution of sanctions.

A primary organ may be authorized to decide only whether a certain specific law forbidding the use of force was violated; but in deciding this, it is liable to have to decide whether the use of force was justified as being the execution of some sanction or other, or whether it was justified on other grounds (mistake, etc.). In recognizing certain acts as the execution of sanctions, the organ, explicitly or implicitly, also recognizes other laws forbidding the use of force, namely those laws for the violation of which the sanctions were prescribed. Similarly, by admitting a defence of mistake (e.g. accepting that the agent acted under a misapprehension as to his rights or the rights of other people) other laws may be recognized.

Thus a primary organ, an organ which is authorized to decide whether certain acts using force were a violation of a certain law, is liable to recognize in the exercise of its powers other laws permitting the use of force in the execution of sanctions, and other laws prohibiting the use of force which belong to the same momentary system. A primary organ may also recognize the existence of other law-applying or lawcreating organs, and the validity of their actions and of the laws regulating their actions. Thus a primary organ recognizes, directly or indirectly, explicitly or implicitly, all the laws of the momentary system.

For the purposes of this criterion of membership an organ can be said to be authorized to decide whether a law forbidding the use of force was violated, if in certain circumstances its decision is a condition for the execution of a sanction against the violator, either by that organ itself or by others.

One can imagine a society in which there exist certain rules prescribing behaviour and backed by coercive sanctions, but in which there exist no primary law-applying organs. The decision that a law was violated and the application of sanctions may be entrusted to the injured party or his family or to any

-193-

one in the society, provided that the violator has not already been punished by another person, etc. Primary organs make their appearance the moment the power to decide upon the application of such sanctions is concentrated in the hands of relatively few people, who are appointed to or otherwise entrusted with this task primarily because of their supposed abilities, or because they deserve it, and not because of their relation to the injured party or to the violator of the law, provided that they have such powers for a relatively long period and may use them in an indefinite number of cases. Primary organs may be the organs executing the sanctions, or their decision that the law was violated may be in certain circumstances a necessary condition for the application of the sanctions.

The study of elementary forms of primary organ may shed much light on normative systems which are on the border between a legal and a pre-legal status. Understanding the various types of such transitory normative systems is more important than deciding for each of them whether it is a legal system or not, a question which in many cases is completely barren. The exploration of various forms of primary organ is, however, clearly beyond the scope of this study.

The moment one primary organ exists, even if it is empowered to decide upon the violation of only one norm, it is possible to distinguish between the norms recognized by it and other norms which are not recognized, either because they have no bearing on the use of force or for some other reason. The fact that the organ’s decisions are based, however remotely or indirectly, on a certain group of norms justifies regarding them as constituting a normative system. It is this sense of ‘normative system’ which is determined by the criterion of membership formulated above.

It is arguable that not every such normative system is a legal system; that in a legal system the decision on the violation of more than one norm should be entrusted to primary organs. As has been pointed out above, the essential role of institutions in the law is not confined to their role in the criterion of membership. We need not, however, concern ourselves here with this problem, for, whatever its solution, the correctness of the proposed criterion of membership is not affected by it.

-194-

A primary organ’s decision can he declared, and is sometimes accompanied by an order or a permission to apply a sanction, or do some other acts. The decision may not even be declared. Upon reaching it the organ may simply proceed to act accordingly, i.e. to apply the sanction or abstain from applying it. In developed legal systems primary organs arrive at their decisions usually after a process of litigation regulated by law, and the decision itself is often accompanied by an explanation of the reasons which led to it. In such circumstances it is relatively easy to find which are the laws recognized and acted on by the organs. It is more difficult to do this when the decision is not explained or justified by the organ.

It is, however, possible to tell a person’s reasons for an act or a decision, even if they are not revealed by that person himself. Where legal organs are concerned, the task is made easier by the fact that laws exist only as part of legal systems that are widely accepted by a certain population. Thus the primary organ is usually expected by the population or the legal profession to recognize certain laws, and if its decisions do not conflict with them it may be assumed to have acted on their basis.

It goes without saying that laws are only part of the reasons for the decisions of primary organs. Other reasons may be certain findings of fact, considerations of justice, etc.

Primary organs may not only act on previously existing laws, they may sometimes create new laws and apply them. As they recognize the laws which they create, the distinction between laws which exist prior to their recognition by the primary organs and laws which are created and applied by the primary organs is not essential to the understanding of the criterion of membership. It is, however, of the utmost importance to understand that the view that all the laws are recognized by primary organs (or the view that they are all recognized by courts), or that their membership in a legal system is determined by this fact, does not entail the view that primary organs create all the laws (or that the courts do this).

Whether their creation can be attributed to the primary organs (or to the courts) depends upon various factors. Thus, if a law was created by the exercise of powers conferred by another PLlaw recognized by the primary organs or the

courts, then it is recognized by them as made by whoever has these powers. Furthermore, if the reasons given for recognizing a law are reasons justifying its content, then we shall tend to think that it was created by the courts or the primary organs. If, on the other hand, the reasons for recognizing a law concern the authority of some person or body of persons who laid it down, then it is likely that the court or the primary organ is applying a previously existing law. This view will be strengthened if the law is held to be valid from the time it was laid down by that person or group, and even more so if it was practically certain before the organ’s decision to act on this law. When these conditions obtain, the law recognized can be an original law, i.e. a law created by a person or group who were not authorized by law to create it.

Saying of a primary organ that it recognizes a law means that, had the question been raised before it, in the proper exercise of its powers it would have acted on that law. It is a counterfactual statement which raises many philosophical questions that need not concern us here. The evidence for such statements is the past behaviour of the primary organ, the attitudes and opinions of the population and of the legal profession, etc. The evidence can be and is largely indirect. Of special importance is the fact that recognizing a PL-law logically entails recognizing all the laws created by the exercise of powers conferred by it. This fact is responsible for the importance often attached to PL-laws in discussions of the problem of membership. Of equal importance is the fact that one primary organ’s recognition of the norms creating another primary organ (or a court) gives considerable support to the supposition that the first organ will recognize all the laws recognized by the second, though it does not entail it.

This short explanation of the criterion of membership leaves many unanswered questions, though I believe that all of them can be satisfactorily answered. We may conclude this section by enumerating some of them.

Some further explorations need be made of the structure of legal systems, to prove that the degree of interdependence of laws presupposed by the criterion proposed here really exists. The following are some of the main problems’

There is the question of certain pleas or defences like personal

-196-

immunity, and their effects on the content of legal systems. Total immunity from sanctions can sometimes prevent primary organs from considering certain norms, in which case they must be regarded as not belonging to the legal system that consists of the laws recognized by the primary organs. Thus religious bodies can be immune from state control and be subject to a separate legal system. When the immunity is not complete it does not suffice to isolate the norms governing the behaviour of the immune persons, if there are such norms. But the precise effects of immunity and similar defences need looking into precisely and in much greater detail.Another problem is the question of a ‘conflict’ between the acts of various primary organs. Can one such organ, A, recognize another organ, B, and recognize all the laws applied by B, while B does not recognize A and some of its laws? And how should one account for such a situation?Yet another difficult problem is to distinguish between an organ’s recognition of the laws of its own legal system and its recognition of laws of other legal systems in virtue of the directions of its private international laws.Finally, there is the problem of distinguishing law from fact-Are companies’ regulations recognized by courts as law or as

fact? What is the status of a father’s order to his child, disobedience to which is punishable by law? etc.

VIII.3: ON THE RULE OF RECOGNITION

Many legal theorists based their explanations of law on the activities of law-applying organs. None of them succeeded in offering a satisfactory criterion of membership. This is due, no doubt, at least partly to their failure to formulate clearly the problem, and to separate it from other problems. This feature is so marked that it is often questionable whether they were at all interested in the problems of membership and identity. The criterion formulated in the previous section expresses a difference from the views of most other theorists who share Salmond’s approach in two important respects:

1. It is concerned not with the activities of the courts or of law-applying organs in general but with primary organs only. -197-
1. It is concerned with the actual behaviour of primary organs, not with what they ought to do, yet it is concerned with the activity of these organs as guided by normative considerations. This second point, and the nature of the proposed criterion of membership in general, can be made clearer by reference to Hart’s position on the subject, i.e. to his doctrine of the rule of recognition.

The relevance of the rule of recognition to the problem of membership is made clear in the very first passage in which the concept is introduced. Rules of recognition, Hart explains, ‘specify some feature or features possession of which by a suggested rule is taken as a conclusive affirmative indication that it is a rule of the group’. [1] A rule of recognition is ‘a rule for conclusive identification of the primary rules of obligation’. [^2]

The rule of recognition is a legal rule and belongs to the legal system. It differs from other laws in that its existence is not determined by criteria laid down in other laws but by the fact that it is actually applied:

. . . whereas a subordinate rule of a system may be valid and in that sense ‘exist’ even if it is generally disregarded, the rule of recognition exists only as a complex, but normally concordant, practice of the courts, officials, and private persons in identifying the law by reference to certain criteria. Its existence is a matter of fact. 3

This seems to imply that the rule of recognition is always a customary and not a legislated rule. On the other hand, Hart says:

If a constitution specifying the various sources of law is a living reality in the sense that the courts and officials of the system actually identify the law in accordance with the criteria it provides, then the constitution is accepted and actually exists. It seems a needless reduplication to suggest that there is a further rule to the effect that the constitution (or those who ‘laid it down’) are to be obeyed. 4

The constitution, in such cases, should presumably be regarded as created both by legislation and by custom, a position which is perhaps not impossible, but needs some explaining.

This is a minor difficulty. Of much greater consequence is the difficulty of finding who are the norm-subjects of the rule

________ 1 CL, p. 92. 2 CL, p. 92. 3 CL, p. 107. 4 CL, p. 246.

-198-

of recognition and the doubt as to whether it is a duty-imposing or a power-conferring law. (It should be remembered that, according to Hart, all norms are either duty-imposing or powerconferring norms, and that he assumes that all laws are norms.) Hart says that ‘wherever such a rule of recognition is accepted, both private persons and officials are provided with authoritative criteria for identifying primary rules of obligation.’ 1 This suggests that the rule of recognition is directed to the population at large. Does it mean that all the law-subjects of all the laws which are to be identified by it are its law-subjects?

Hart often contrasts the rule of recognition and other secondary rules with primary rules which are rules of obligation. So presumably rules of recognition do not impose obligations but confer powers. This hypothesis is strengthened by the following half sentence: in certain circumstances ‘ . . . the rule which confers jurisdiction will also be a rule of recognition’. 2 But it is quite clear that this is not Hart’s intention, as he himself confirmed to me. In his book Hart explains only that dutyimposing laws can be customary laws. There is no sense, according to his theory, in which power-conferring laws can be customary laws, 3 unless they are part of a legal system of which they are not the rule of recognition.

Consequently it must be concluded that the rule of recognition is a duty-imposing law. This means, however, that its lawsubjects cannot be the population at large, for there is no duty on ordinary people to identify certain laws and no others (nor, for that matter, do they have a legal power to do so).

The rule of recognition should, therefore, be interpreted as a D-law addressed to the officials, directing them to apply or act on certain laws. Hence only the behaviour of the officials and not the behaviour of the population as a whole determines whether the rule of recognition exists.

The fundamental reason which moved Hart to adopt his doctrine of the rule of recognition is expressed in the following sentence: ‘If the question is raised whether some suggested

________ 1 CL, p. 97.

² CL, p. 95.

³ The explanation of P-laws in Ch. VI, sect. 5 above admits the possibility of customary Plaws. But the existence of P-laws in general depends on the existence or the possible creation of certain D-laws, and there are no such D-laws to make possible the interpretation of the rule of recognition as a P-law.

rule is legally valid, we must, in order to answer the question, use a criterion of validity provided by some other rule.’ [1] It is this assumption that is questionable. It seems to me that to answer the question whether a certain suggested law exists as a law in a certain legal system one must ultimately refer not to a law but to a jurisprudential criterion. [2] Ultimately one must refer to a general statement that does not describe a law but a general truth about law.In some legal systems there may be laws which oblige certain organs to apply all the laws fulfilling a certain condition, and it may be that these laws are in fact all the laws of the system. But even when such laws exist, which is not always the case, the laws of the system belong to the system not because of this rule of recognition, but because they are all recognized by the primary organs.That it is not the case that in every legal system there is one such rule of recognition can be seen by considering the following two points:

1. It is not clear on what Hart bases his view that there is only one rule of recognition in every legal system. Why not say that there are various rules of recognition, each addressed to a different kind of officials? Why not say that various rules of recognition prescribe the recognition of various types of laws?
1. Though, as has been suggested in the previous section, the behaviour of primary organs is the key to the criterion of membership, there is no reason to think that they always act in fulfilment of duties, as Hart’s theory entails. They may be under a duty to recognize some laws and yet be legally free to recognize or not recognize certain others.[3] Suppose, as Hart does, that legal laws can be D-laws merely in virtue of diffused critical reactions, without being backed by sanctions or other legal remedies. Even so, it need not be the case that primary organs will meet with critical reactions if they stop recognizing certain laws or begin recognizing certain other laws.

-200-

A. Note on Laws and Paper Laws

This chapter is concerned with the way in which the fact that the law is institutionalized helps to solve the problem of membership. The proposed solution accords special importance to certain law-applying organs, without thereby assuming (1) that laws are predictions or descriptions of the acts of these organs; (2) that the laws are directed only to these organs; they are directed to various classes of persons and are simply recognized by these organs; (3) that all the laws are created by these organs; or (4) that the organs are always under a duty to recognize them.

The proposed criterion of membership points the way to the solution of another problem. Kelsen thinks that minimum efficacy is a necessary condition to the validity of every law.1 This is his solution to the problem of ‘paper laws’. The proposed criterion of membership suggests a different approach to the problem: The laws of a system are those recognized by the primary organs. Statutes or regulations, etc., which are disregarded by the law-applying organs are not really part of the legal system. If they are recognized in a modified form they are laws in this modified form. But the fact that they are disregarded by the population is irrelevant to their existence. Moreover, even if they are disregarded by the police, even if no prosecutions are brought for their violation (if they are D-laws), and if interested parties do

not invoke them before the lawapplying organs, they are still valid laws, provided that if raised before primary organs in the proper exercise of their powers they would recognize and act on them.

It must be admitted that ordinary prevalent opinion on the subject is divided, and the view suggested here cannot be justified as explicating current opinion. Its justification is indirect. Firstly, it accords with the views expressed on the problems of membership and identity, which in their turn are explications of common sense and professional opinion on these subjects. Secondly, it expresses another aspect of the decision to attribute a central role to the law-applying organs.

It is true that one of the distinguishing features of norms is that they constitute certain types of reasons for action. But

________ 1 Cf. Ch. III, sect. 3 above.

-201-

norms are always only one of the reasons for action in every situation; their effect and their weight are affected by many other facts that are also reasons for action, and which sometimes enhance and sometimes diminish or even contradict the effect of norms. The fact that the existence of certain laws is not always a decisive reason for acting as they prescribe is generally accepted. An important moral rule contradicting a law may cause the population to disregard the latter, but neither the contradiction nor its effect on the population mean that the law does not exist. Similarly, a low rate of crime detection may cause people to attribute a smaller weight to certain laws, though this does not mean that the laws do not exist.

It is characteristic of norms that the type of reasons for action which they constitute are in general weighty and important reasons for action. Hence the tendency to think that if a law has lost much of its importance as a reason for action, because of its general disregard, it is no longer a law. But where normative systems are concerned it seems reasonable to allow this consideration of the weight of individual norms as reasons for action to be overridden by consideration of the systematic nature of the norms. Because they belong to a system characterized by the organized application of laws by specially appointed organs, even laws largely disregarded and neglected are laws so long as they are recognized by these organs.

-202-