English Analytical Positivism PDF

70 4 English Analytical Positivism 4.1 Introduction This chapter identifies that part of legal theory which is known as analytical positivism, and considers the work of two of its leading English exponents, namely Austin and Hart. More particularly, it will show how the work of the latter r...

70 4 English Analytical Positivism 4.1 Introduction This chapter identifies that part of legal theory which is known as analytical positivism, and considers the work of two of its leading English exponents, namely Austin and Hart. More particularly, it will show how the work of the latter remedies certain deficiencies in the work of the former. The tradition of English analytical positivism may be traced back to the work of Jeremy Bentham (1748–1832). Although some of Bentham’s work was published in his lifetime, the manuscript of his major work on legal theory was overlooked for more than a century, with its first publication occurring as late as 1945, under the title The Limits of Jurisprudence Defined, before appearing in definitive form in 1970 as Of Laws in General. There was, however, no such delay in the publication of the work of his follower, John Austin (1790–1852). In 1826 Austin became the first Professor of Jurisprudence in the newly founded University of London, which subsequently became University College, London. (Before this time, access to university education was restricted to members of the Church of England, and University College was founded specifically to widen access to university education for people of other – or no – religious persuasions. This objective, together with the fact that it was – and indeed still is – based in Gower Street, not only explains why members of the college are commonly called ‘the godless of Gower Street’, but also made it a peculiarly appropriate environment for a positivist professor of law.) Austin’s work, The Province of Jurisprudence Determined, which consisted of the text of six lectures which he had delivered in the college, was first published in 1832, but his version of what is known as the command theory of law continued to dominate the English teaching of jurisprudence well into the 20th century. 4.2 Austin’s Command Theory of Law Introduction As the title of his lectures (The Province of Jurisprudence Determined) indi- cates, Austin is concerned to identify the limits of the subject of jurisprudence.  English Analytical Positivism 71 This necessarily involves eliminating a great deal of material which may in one way or another be said to be law, without being law for the present purposes. Austin begins with an unequivocal statement of his positivist position: ‘The matter of jurisprudence is positive law: law, simply and strictly so called: or law set by political superiors to political inferiors.’ (The Prov- ince of Jurisprudence Determined, 1832, p. 9.) Further refinements are then introduced. First, the political superior must be a sovereign: ‘If a determinate human superior, not in the habit of obedience to a like superior, receive habitual obedience from the bulk of a given society, that determinate superior is sovereign in that society, and the society (including the superior) is a society political and independent.’ (Op. cit., p. 194.) The next stage is to explain the concept of command, which may be either ‘direct or circuitous’. (Op. cit., p. 134): ‘The ideas or notions comprehended by the term command are the following. (1) A wish or desire conceived by a rational being, that another rational being shall do or forebear. (2) An evil to proceed from the former, and to be incurred by the latter, in case the latter not comply with the wish. (3) An expression or intimation of the wish by words or other signs.’ (Op. cit., p. 17.) It will be apparent that, for these purposes, the idea of a sanction (i.e. the evil which is imposed in the event of non-compliance with the command) is inextricably involved in the idea of the command itself. Finally, the status of law is reserved for a command which ‘obliges gener- ally to acts or forbearances of a class’. (Emphasis added.) In summary, therefore, Austin’s version of law requires a command which is both express and general, which is issued from a sovereign to a subject, and non-compliance with which results in the sovereign imposing a sanction on the subject. We must now consider each of these elements in more detail. The nature of the command The requirement that the command must be express may seem obvious enough, and is reflected in Lord Diplock’s dictum that ‘Parliament, under 72 Legal Theory our constitution, is sovereign only in respect of what it expresses by the words used in the legislation it has passed’. (Black-Clawson International Ltd v Papierwerke Waldhof-Aschaffenberg AG 1 All ER 810.) Nevertheless, in practice it is by no means certain that Parliament has always expressed that which the courts hold to be the law, since interpretation will always be necessary, at which point factors other than the plain words of the statute (for example, the court’s perception of the statutory purpose, or of judicial policy) may come into play. The case of Kammins Ballrooms Co Ltd v Zenith Investments Ltd 2 All ER 871 provides a classic modern example of the way in which the purposive approach to interpretation may radically affect the meaning of an apparently clear, express legislative text. The facts arose under the scheme of security of tenure for business premises contained in Part II of the Land- lord and Tenant Act 1954. Briefly, the statutory scheme requires that a tenant who wishes to retain the tenancy of premises on the expiry of the existing tenancy should ask the landlord to grant him a new one. If the land- lord refuses to comply, the tenant can then apply to the court, where the matter will be resolved. In the context of this procedure, s. 29(3) of the Act provides that ‘no application... shall be entertained unless it is made not less than two nor more than four months after... the making of the tenant’s request for a new tenancy’. In Kammins the tenant’s application to the court was made outside the statutory period, but the House of Lords held that this did not necessarily invalidate it. Lord Diplock acknowledged that: ‘Semantics and the rules of syntax alone could never justify the conclu- sion that the words “No application... shall be entertained unless” meant that some applications should be entertained notwithstanding that neither of the conditions which follow the word “unless” was fulfilled.’ (Original emphasis.) Nevertheless, Lord Diplock pointed out that the purpose of the Act was to persuade landlords and tenants to proceed by agreement wherever possible, and that the time limit in question had been enacted for the protection of landlords. It followed that landlords should be allowed to waive compliance with the time limit if they so wished. (The fact that, in all the circumstances of the case, this particular landlord was held not to have waived compliance is irrelevant to the principle of interpretation.) It will be readily apparent that the law, as formulated by the House of Lords, did not rely on the express words contained in the statutory command. The decision of the House of Lords in Jones v Secretary of State for Social Services 1 All ER 145 provides an example of the court’s perception of judicial policy (in this case, in relation to the operation of the doctrine of binding precedent) being crucial to the outcome. The case was decided  English Analytical Positivism 73 under a scheme of industrial injuries legislation which provided that claims for long-term benefit were determined in two stages and by two different tribunals. In the first place the question was whether an injury had been caused by an accident arising out of employment. If this question were answered in the affirmative, the second stage was to decide the extent of the disablement, and therefore the amount of benefit payable. The legislation stated that the answer to the first question was to be ‘final’. In Re Dowling AC 725, the House of Lords had held that the tribunal determining the second question could not re-open consideration of the first question. In Jones, on legislation which was substantially the same as that which was before the court in Dowling, the majority of a seven-member House of Lords decided to follow the earlier case. The relevance of the case in the present context is that even though four Law Lords held that, as a matter of statutory interpretation, Dowling had been wrongly decided, only three were willing to exercise their discretion to depart from that decision. Several reasons were given for the willingness to follow a case which had been wrongly decided, thereby consolidating the error, but for the present purposes the point is simply that the command (i.e. the relevant statutory provisions) required to be interpreted, and that different judges not only produced different interpretations, but also had different opinions about the effect of those interpretations on the final outcome of the case. Once it is accepted that cases such as these show that in reality the courts may treat a statute, albeit one enacted by a legislatively supreme Parliament, as no more than a starting point when identifying and applying the law, the proposition that the express command of a sovereign is an essential element in law must be open to serious doubt. Austin’s requirement that a command must be general before it qualifies as a law may be thought to neglect the facts of legal life, because in practice all sorts of individualized commands clearly have legal effect. For example, orders of a court restraining a defendant from continuing to commit a tres- pass, or prohibiting the republication of a libel, are both clearly commands, backed by the sanction of imprisonment for contempt of court in the event of breach, and yet they are equally clearly individualized in the sense of being addressed to specific defendants. However, Austin would regard these as being merely applications of law to particular cases, rather than being law strictly so-called, and on this basis they provide no challenge to his theory. The nature of the sovereign The requirement that the sovereign must be human clearly excludes divine law from Austin’s area of inquiry, but it would be unfair even for natural lawyers to lay this exclusion specifically at Austin’s door by way of criti- cism, since it is a matter which is common to positivist theories generally. 74 Legal Theory However, other aspects of Austin’s concept of a sovereign are open to specific criticism. The requirement that the sovereign shall ‘not [be] in the habit of obedience to a like superior’ may be reasonably accepted as a characteristic of sovereignty, while the proposition that the sovereign must ‘receive habitual obedience from the bulk of a given society’ may seem to be no more than a realistic acceptance that no legal system can expect to experience total compliance with its laws. However, this latter element does conceal a difficulty, since it is a commonplace observation that many laws are accorded habitual obedience even though the same cannot be said of the sovereign which enacted them, for the simple reason that that sovereign no longer exists. For example, the Offences Against the Person Act 1861 is unquestionably still law, despite the fact that no members of the Parliament which enacted it are still alive. Admittedly there are various ways in which it is possible to seek to avoid this objection. For example, you may wish to argue that it is the institution Parliament, rather than a specific and transitory assemblage of parliamentarians, which is sovereign. While this contention obviously makes sense in general terms, and would no doubt satisfy most people, it must be said that according sovereignty to an institution fails to take account of Austin’s requirement that the sovereign shall be human. Alternatively, you may wish to argue that, at any given time, any current Parliament’s failure to repeal earlier statutes may be taken to be tacit endorsement of them, and that therefore all unrepealed statutes must be taken to have the authority of every successive Parliament. This attempted defence of Austin, however, can fare no better than the previous one, because it fails to take account of his requirement that there must be an ‘expression or intimation of the [sovereign’s] wish by words or other signs’, which clearly precludes the concept of an implied command. Furthermore, Austin runs into difficulty during those periods when one Parlia- ment has been dissolved and another has not yet been summoned. Who, then, is the sovereign? Austin’s own analysis seeks to make this question irrelevant, by saying that, even when Parliament is sitting, the electorate is the true source of the House of Commons’ contribution to the sovereignty of Parliament as a whole: ‘Adopting the language of most of the writers who have treated of the British Constitution, I commonly suppose that the present parliament, or the parliament for the time being, is possessed of the sovereignty: or I commonly suppose that the king and the lords, with the members of the commons’ house, form a tripartite body which is sovereign or supreme. But, speaking accurately, the members of the commons’ house are merely trustees for the body by which they are elected and appointed: and, conse- quently, the sovereignty always resides in the king, and the peers, with the electoral body of the commons.’ (Op. cit., pp. 230–1.)  English Analytical Positivism 75 Once again, this seems more than a little strained, especially when it is remembered that part of the notion of sovereignty is habitual obedience by a political inferior to a political superior. Although admittedly Austin is not saying that the electorate is the sovereign, his analysis does seem to suggest that an essential element of sovereignty is that the electorate should be habit- ually obedient to itself; which is somewhat odd, to say the least. The nature of the sanction The need for a sanction fits the idea of criminal law quite well; or at least it did so at the time when Austin was writing, since the imposition of a sentence of imprisonment was the normal consequence of a conviction. (It would be churlish to criticize Austin for not taking account of alternative means of dealing with offenders, such as probation orders, which were not available in his day.) However, the general requirement of a sanction constitutes a serious weakness in Austin’s theory, and did so even in his own day, since it appears to disqualify the whole field of civil law from his definition of law. The point may be illustrated most clearly by reference to areas of law which confer powers on people, such as the law relating to the making of wills and contracts. Suppose I make a will but fail to observe the statutory requirements as to the witnessing of my signature. The consequence is that my will is a nullity, and therefore the beneficiaries I have named will not inherit. While Austin is content to describe nullity as a sanction, it is difficult to see how this can be said to fit the facts. Even if, in the abstract, it makes sense to speak of nullity as being a sanction, in this particular example I shall be dead before the consequences of the nullity occur, and therefore there is a degree of artifici- ality in saying that I am subject to a sanction. Leaving aside the example of an improperly executed will, and taking a situation in which I am still alive at the relevant time, it may still be difficult to see nullity as a sanction. Suppose, for example, that I have a credit account with a bookmaker and that I place a bet on a horse. By virtue of the Gaming Act 1845, the contract is void; therefore if I lose before I pay the stake, and refuse to pay it thereafter, I cannot be sued. On any ordinary use of language, can I really be said to be subject to a sanction in this situation? 4.3 Hart’s Concept of Law Introduction Despite its evident deficiencies, Austin’s command theory of law dominated the field of English legal theory until well into the 20th century. The work which transformed English legal theory was H.L.A. Hart’s The Concept of 76 Legal Theory Law, first published in 1961, with a second edition appearing posthumously in 1994. (It is worth noticing that the second edition takes the rather unusual form of reproducing the first edition in practically identical terms, with the new part being confined to a postscript in which Hart responds to his critics.) Hart describes his book as ‘an essay in analytical jurisprudence’ which is intended to clarify ‘the general framework of legal thought’ rather than presenting a critique of law or legal policy. (The Concept of Law, 2nd edn, 1994, p. v.) Additionally, he says that the book is also ‘an essay in descrip- tive sociology’, by which he means that a great deal of light can be shed on social situations or relationships by examining the way in which we speak about them, and that although this will often depend on the social context, that dependence may well be left unstated. Whatever labels are attached to it, one thing is clear about The Concept of Law: it makes no claim to offer a definition of law. Indeed, Hart rejects the attempt to provide a useful definition of law at all. He begins by identifying ‘three recurrent issues: How does law differ from and how is it related to orders backed by threats? How does legal obligation differ from, and how is it related to, moral obligation? What are rules and to what extent is law an affair of rules?’ (Op cit., p. 13.) Having commented, inter alia, that ‘definition... is primarily a matter of drawing lines or distinguishing between one kind of thing and another’ (ibid.) Hart concludes that ‘the history of attempts to provide concise defini- tions has shown’ that ‘nothing concise enough to be recognized as a defini- tion’ (op cit., p. 16) can prove satisfactory. Nevertheless, he says that bringing the three issues together ‘has not been misguided; for... it is possible to isolate and characterize a central set of elements which form a common part of the answer to all three’. (Ibid.) Hart proceeds to devote three chapters to a detailed, critical analysis of Austin’s theory, but his conclusions (see The Concept of Law, 2nd edn, 1994, p. 79), may be summarized as follows. First, even the example of criminal law, which most clearly fits the idea of commands backed by sanc- tions, fails to fit Austin’s theory fully, because it generally applies to those who enact it, as well as to others. Secondly, there are other types of law, such as those conferring powers on public bodies and those giving private persons the power to regulate their own legal relationships which cannot accurately be seen as commands backed by sanctions. Thirdly, some laws arise other than by way of express command. Finally, there is the problem of the continuity of law’s validity, despite the limited life of the sovereign which enacts it. Hart concludes that, after a study of Austin’s theory, there is ‘plainly need for a fresh start’ and proceeds to locate ‘the root cause of failure’ in the fact  English Analytical Positivism 77 that Austin’s theory ‘cannot... yield, the idea of a rule, without which we cannot hope to elucidate even the most elementary forms of law’. (Emphasis added. Op. cit., p. 80.) Law as a system of rules There are several strands to Hart’s argument that law consists of a system of rules, each of which needs to be understood individually before their joint effect may be properly appreciated. These strands may be summarized as the distinction between personal habits and social rules; the distinction between being obliged and being under an obligation; the distinction between the external and the internal aspects of rules; and the distinction between primary and secondary legal rules. We shall consider each of these distinctions in turn. The distinction between personal habits and social rules Although it is obvious that certain personal habits are highly individual (such as always wearing a lucky charm for superstitious reasons), it is equally obvious that many are very widespread. Clearly, therefore, we must distinguish between widespread habits and social rules. For example, as a matter of personal habit, I may invariably drink coffee after my dinner. Furthermore, the practice of after-dinner coffee drinking may be very widespread. Nevertheless those who prefer to drink tea, or not to drink anything, will not find themselves subjected to criticism by other members of society. In other words, although drinking coffee after dinner may be an example of convergent behaviour, when correctly analysed it is nothing more than a large number of instances of individual behaviour, and therefore cannot be called a rule. By way of contrast, the convergent behaviour of standing up when the national anthem is played is more than simply a matter of personal taste, not only because most people conform, but also, and crucially, because those who do not do so are likely to incur the disapproval of other members of society. In other words, people feel themselves to be under an obligation to join in this example of convergent behaviour, and consequently it may be called a rule. In passing, it may be worth saying that although the existence of an obligation indicates the existence of a rule, Hart concedes that language is sometimes used in such a way that the converse does not neces- sarily apply. In other words, although we may speak of the rules of matters such as grammar and etiquette, and we may even express those rules in terms of what you should or should not do (for example, always use a capital letter at the beginning of a sentence or don’t put your elbows on the dining table), 78 Legal Theory it would not accord with the normal use of language to say that there is a duty to comply with such prescriptions and prescriptions. As Hart puts it: ‘Rules are conceived and spoken of as imposing obligations when the general demand for conformity is insistent and the social pressure brought to bear upon those who deviate or threaten to deviate is great.’ (Op. cit., p. 86.) It is clear, therefore, that Hart is using rules in a relatively strict sense. However, further insight may be gained by considering more closely his view of the nature of obligation, as clarified by the distinction between being obliged and being under an obligation. The distinction between being obliged and being under an obligation Although the phrases being obliged and being under an obligation to do something will commonly appear to be interchangeable, Hart argues (op. cit., pp. 82 et seq.) that there is, in principle, an important distinction between them. He illustrates the distinction through the example of a gunman, A, who demands money from a victim, B. According to the ordi- nary use of language, we would say that B is obliged to hand over the money (because he fears the consequences if he does not do so), but we would not say that B is under an obligation (or owes a duty) to comply with A’s demand. Hart develops the point by saying that we would not say that B is even obliged to comply with the demand if the harm which A threatens to inflict on B is merely trivial; or if the harm is serious but there are circum- stances making it unlikely that A will actually inflict it. Moreover, just as it is possible to be obliged without being under an obligation, so it is possible to be under an obligation without being obliged. For example, if the law requires that all able-bodied males within a specified age group shall undertake compulsory military service, we would say that X is under an obliga- tion to undertake such service (provided, of course, that he is male, able-bodied and within the specified age group) even though, in fact, he has gone abroad, and will, therefore, never actually be obliged to perform his obligation. Of course, in practice, the two situations will commonly overlap, so that, for example, someone who has been fined following conviction for an offence will be both obliged, and under an obligation, to pay the money. Nevertheless, this aspect of Hart’s argument provides a useful insight into the distinction between commands which may be backed by nothing more than naked power and rules which create obligations. The distinction between the external and internal aspects of rules The third of Hart’s distinctions which we must consider is that between the external and the internal aspects of rules.  English Analytical Positivism 79 As the expression suggests, the external aspect of a rule is apparent from the outside, so that, for example, systematic observation can be used to predict when an habitual after-dinner coffee-drinker will drink coffee. The internal aspect, on the other hand, is apparent only to those who are subject to the rule. Hart (op. cit., p. 90) takes the example of a person who observes the way in which motorists respond to red traffic lights. Since it is an exter- nally observable fact that most motorists stop at red traffic lights, it is reasonable to predict that any given motorist will do so. However, viewed internally (from the point of view of the motorist) the fact of a red traffic light is not simply evidence on which to base a prediction, but actually imposes an obligation (or a duty) to stop. This internal aspect of the rule may appear to be simply a feeling on the part of the person subject to it, but Hart rejects this analysis. More particularly, he argues: ‘What is necessary is that there should be a critical reflective attitude to certain patterns of behaviour as a common standard, and that this should display itself in criticism (including self-criticism), demands for confor- mity, and in acknowledgments that such criticism and demands are justi- fied, all of which find their characteristic expression in the normative terminology of “ought”, “must”, and “should”, “right” and “wrong”.’ (Op. cit., p. 57.) Since there is nothing in the after-dinner coffee-drinking example which can be said to correspond to this internal element, it follows that that example of convergent behaviour remains at the level of a habit, rather than being elevated to the status of a rule. The distinction between primary and secondary legal rules We are now equipped to proceed to the core of Hart’s theory, namely the distinction between primary and secondary legal rules, and the need for both to exist before there can be said to be a legal system. Primary rules are those which impose obligations. Such rules may be either positive or negative, and would embrace both the rule which requires you to pay income tax and the rule which requires you to refrain from murder. However, it will be readily apparent that the idea of primary rules presents only a partial view of the law, since there will frequently be uncertainty as to the content of the primary rules, coupled with a need to change them from time to time, and a need to resolve disputes arising from the application of the rules in particular cases. It is to deal with these matters that Hart formulates the concept of secondary rules, which he classifies under three headings. The first category of secondary rules deals with the problem of uncer- tainty. Here, Hart says, there must be a rule of recognition, so that actors 80 Legal Theory within the legal system can recognize those rules which are law and distin- guish them from those which are not. The rule of recognition operates ‘by reference to some general characteristic possessed by the primary rules. This may be the fact of their having been enacted by a specific body, or their long customary practice, or their relation to judicial decisions.’ (Op. cit., p. 95.) In the English system, therefore, the rule of recognition has at least two limbs: those things are law which are enacted by, or under the authority of, Parliament, as are the decisions of the courts which acquire authoritative status under the doctrine of binding precedent. It may be possible to formulate it in such a way that more limbs appear. For example, the rule of recognition may extend to laws emanating from the European Community, and it may still be possible to regard custom as a source of English law. However, the first of these propositions stems directly from the European Communities Act 1972, while the second stems from a number of judicial decisions to this effect, and therefore the application of Occam’s razor should leave the original formula- tion unadorned by any additions. (Occam’s razor is the name given to the principle attributed to William of Occam, a 14th century Franciscan philoso- pher, which requires superfluous elements in any argument to be discarded. He is said to have dissected arguments as if he were taking a razor to them.) At this stage, therefore, it becomes plain why the obligation to stop at a red traffic light is law (because it is contained in the road traffic legislation), whereas the obligation to stand up when the national anthem is played is not (because there is neither parliamentary nor judicial authority that it is). In passing, it may be worth commenting on the oddness of Hart’s classi- fying the rule of recognition as being secondary, when common sense would seem to indicate that its being the very foundation of the legal system should make it primary, but we will introduce needless complication if we try to redefine his terms. The second category of secondary rules, namely rules of change, has two dimensions. The first dimension deals with the need to change the primary rules, which, in the context of the English legal system, means overruling judicial decisions in order to terminate their status as binding precedents, and repealing or amending statutes. Obviously, therefore, these rules are intimately connected with the rule of recognition, because their application enables us to decide which of two apparently conflicting primary rules will prevail over the other. The second dimension of the rules of change is that they enable people to change the way in which the primary rules of obliga- tion apply to their own personal situations, through such means as making contracts and wills ‘and many other voluntarily created structures of rights and duties which typify life under law’. (Op. cit., p. 96.)  English Analytical Positivism 81 The third category of secondary rules, which deals with the need to resolve disputes, consists of rules of adjudication, which not only confer upon public officials the power to determine legal disputes, but also govern the way in which that power is exercised and give special legal status to orders which are made in consequence of adjudication. (Op. cit., pp. 96–7.) It is, of course, the function of the rule of recognition to enable actors within the legal system to identify the other rules which make up that system as a whole. Therefore, while it may be stating the obvious to say that a rule must exist before the rule of recognition can be applied to it, it is essential to realize that those rules which satisfy the rule of recognition go beyond mere de facto existence and exist in the additional sense of having legal validity. Hart makes this point by way of comparison with the rule which requires Christian men to bare their heads when entering a church. The question of the validity of such a rule does not arise independently of its effectiveness. Therefore, if people were to stop complying with it, the rule would cease to exist, beyond which point it would make no sense to ask whether the rule was still valid. In the case of a law, on the other hand, mass disobedience of a rule (for example, of a speed limit) has no effect on the rule’s validity. This quality of validity-irrespective-of-observance stems from the rule of recog- nition, which provides the criterion for granting or withholding the status of law from any given rule. More particularly, the rule of recognition itself ‘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’. (Op. cit., pp. 109–10.) Two final points must be made about the rule of recognition. First, despite its factual nature, Hart specifically acknowledges that ‘the rule of recogni- tion may incorporate as criteria of legal validity conformity with moral prin- ciples or substantive values’ and that therefore his theory is an example of soft positivism. (Op. cit., p. 250. For the distinction between hard and soft positivism, see p. 21.) Secondly, the proposition that the existence of the rule of recognition is a matter of fact creates a difficulty, since it is also, ex hypothesi, a matter of law. Hart responds to this by saying that ‘the ultimate rule of recognition may be regarded from two points of view: one is expressed in the external statement of fact that the rule exists in the actual practice of the system; the other is expressed in the internal statements of validity made by those who use it in identifying law’. (Op. cit., pp. 111–12.) The minimum content of natural law Although it will be apparent that Hart is first and foremost a positivist, he also identifies ‘certain universally recognized principles of conduct which have a 82 Legal Theory basis in elementary truths concerning human beings, their natural environment and aims’ and which, he says, may be said to constitute ‘the core of good sense’ in, and therefore ‘the minimum content’ of, natural law. (Op. cit., pp. 192–4). This notion is based on nothing more sophisticated or obscure than the assump- tion that humankind collectively, and the individuals who constitute the species, wish to survive, and simply consists of the ‘rules of conduct which any social organization must contain if it is to be viable’. In developing this aspect of his theory, Hart relies on five things which he terms ‘generalizations’ or ‘truisms’. These are that while human beings are vulnerable to each other, they are nevertheless approximately equal in their powers, and are commonly motivated by limited altruism, with limitations also being evident in their resources, their understanding and their strength of will. It follows from these that there must be rules restricting violence. The facts of vulnerability and approximate equality and the limited nature of altruism require ‘a system of mutual forebearance and compromise which is the base of both legal and moral obligation.’ (Op. cit., p. 195.) More particularly, the fact of limited resources requires some ‘minimum form of the institution of property (though not necessarily indi- vidual property), and the distinctive kind of rule which requires respect for it.’ (Ibid.) Finally, the facts of limited understanding and strength of will result in everyone being ‘tempted from time to time to prefer their own immediate interests and, in the absence of a special organization for their detection and punishment, many would succumb to the temptation.’ (Op. cit., p. 197.) Hart’s mention of punishment may appear to take us back to Austin’s command theory, in which the role of sanctions is, of course, basic. However, it is both clear and unsurprising that Hart does not intend this reversion to the theory which his own theory is intended to supplant. More particularly, Hart explains that ‘“sanctions” are... required not as the normal motive for obedience, but as a guarantee that those who would voluntarily obey shall not be sacrificed to those who would not. To obey, without this, would be to risk going to the wall. Given this standing danger, what reason demands is voluntary co- operation in a coercive system.’ (Original emphasis. Op. cit., pp. 197–8.)  English Analytical Positivism 83 Despite Hart’s use of the phrase natural law, it can be argued that this part of his theory simply deals with certain aspects of law which are essential to the functioning of society. Therefore, since no moral quality can attach to that which is truly essential, it is difficult to avoid the conclusion that Hart’s use of the phrase natural law is merely an attempt to establish that his subject matter has a moral status which closer consideration reveals to be not only spurious but also fundamentally inappropriate. Oliver Wendell Holmes, writing more than thirty years before Hart, sees the point clearly, even if he expresses it circumspectly: ‘No doubt it is true that, so far as we can see ahead, some arrangements and the rudiments of familiar institutions seem to be necessary elements in any society that may spring from our own and that would seem to us to be civilized – some form of permanent association between the sexes – some residue of property individually owned – some mode of binding oneself to specified future conduct – at the bottom of all some protection for the person. But... the question remains as to the ought of natural law.’ (Natural Law (1918) 32 Harv LR 40, at p. 41.) When the rules run out From time to time, it is inevitable that questions will arise to which the law has no ready answer. This clearly poses a difficulty for the judges who have to decide cases in which such questions arise, because the court must always provide an answer. Judges may protest the difficulty of their function, but they are not allowed to refuse to adjudicate on the issues which are placed before them. (A ruling that the court has no jurisdiction to deal with a partic- ular matter may appear to be an exception, but it is not truly so, since the ruling as to lack of jurisdiction is itself a ruling on a matter of law.) Hart acknowledges that this situation will arise, and takes the example of a legal prohibition on the use of vehicles in a park. Taking some examples, not of all of which Hart himself uses, the prohibition clearly includes motor cars, but what of a toy car which is electrically propelled; and if size is thought to be of the essence, how large can the toy be without falling foul of the prohibition? Similarly, what about bicycles; and children’s push chairs; and wheelchairs; and roller-skates? Hart’s conclusion is that the judge has a discretion, which means that ‘even though it may not be arbitrary or irrational, [the decision] is in effect a choice. [The judge] chooses to add to a line of cases a new case because of resemblances which can reasonably be defended as both legally relevant and sufficiently close’. (Op. cit., p. 127.) In making this choice, the judge will be using the fact that language is open-textured. This quality of language contrasts with systems such as, for 84 Legal Theory example, mathematics and music, whose symbols have precise and univer- sally recognized meanings. More particularly, Hart maintains that the meaning of any given word will be made up of a core and a penumbra. The word penumbra is borrowed from the terminology which is used to describe the different densities of the shadows which occur during eclipses. The penumbra refers to the semi- shadow area which surrounds the deep shadow (or umbra) at the centre of the eclipse. Etymologically, penumbra derives from the Latin meaning almost shadow, but in the present context its meaning is more accessibly conveyed by the vernacular phrase grey area, within which there will be substantial scope for uncertainty. While it would be churlish to criticize Hart for using terminology which, when seen in its original context, could cause confusion by suggesting that the deepest shadow creates the greatest clarity, his use of the idea of the core is open to criticism on another ground. The problem is that Hart seems to assume that a word will have a single core meaning, whereas many cases giving rise to difficulties of interpretation in practice arise from a number of possible meanings, with no intrinsic indica- tion of how they should be ranked. (For Hart's discussion of the core and penumbral meanings, see Positivism and the Separation of, Law and Morals (1958) 71 Harv LR 593, reprinted in Essays in Jurisprudence and Philos- ophy, 1983, in which see p. 49.) To take a straightforward, non-legal example, if I see a poster in a shop window saying Last Week – Everything Half-Price, it would be perverse to think that it refers to an offer which expired at the end of the previous, or most recent, week: it plainly means Final Week – Everything Half-Price. On the other hand, if I ask you to name the last king of England, you will almost certainly interpret my question as referring to the most recent (rather than the final) king, and therefore you will say George VI. Of course, if you were to understand my question as referring to the final king of England, you would perfectly properly reply that the question is currently unanswerable because there may well be many more kings still to come. However, it is very unlikely that you would put this interpretation on my question. The point of this example is simply that neither most recent nor final is the core meaning of last, with the other being relegated to being part of the penumbral range of possible meanings; nor are they both penumbral. In other words, Hart's terminology tends to conceal the fact that, even at the heart of a word, courts must frequently choose between two or more mean- ings, all of which rank equally as possibilities unless and until contextual factors are taken into account. As we shall see in Chapter 7, the conclusion that judges, faced with hard cases, make the law by exercising a choice is one of the main grounds on which Dworkin attacks Hart’s theory, but we will adjourn our examination of this topic until we have considered both another version of positivism  English Analytical Positivism 85 (namely that expounded by the Austrian theorist Hans Kelsen) and the revival of interest in the natural law tradition arising after the Second World War. Summary 1 In the 19th century, Austin’s command theory defines law as a command from a sovereign backed by a sanction. Each of these elements is prob- lematic. 2 In the 20th century, Hart’s concept of law analyses law into primary rules and secondary rules, with the legal system being the product of the union of both categories. The key to the validity of law lies in the rule of recogni- tion, which is one of the secondary rules. 3 Hart emphasizes the open texture of language and relies on the (not wholly convincing) concepts of core and penumbral meanings. 4 When the rules run out, judges must, according to Hart, exercise their discretion. Reading Austin, John, The Province of Jurisprudence Determined, 1832 Bentham, Jeremy, The Limits of Jurisprudence Defined, 1945, republished as Of Laws in General, 1970 Hart, H.L.A., Positivism and the Separation of Law and Morals (1958) 71 Harv LR 593, reprinted in Essays in Jurisprudence and Philosophy, 1983 Hart, H.L.A., The Concept of Law, 2nd edn, 1994 Holmes, Oliver Wendell, Natural Law (1918) 32 Harv LR 40