Elusive normativity

Stefano Bertea, The Normative Claim of Law (Oxford: Hart Publishing, 2009), 307 p.

• 1 Introduction

  This article deals with the nature of normativity, a topic which is rather popular nowadays, both in ethical theory¹x Christine M. Korsgaard, The Sources of Normativity (Cambridge: Cambridge University Press, 1996); Ralph Wedgwood, The Nature of Normativity (Oxford: Clarendon Press, 2007). and in the philosophy of law.²xSylvie Delacroix, Legal norms and normativity: an essay in genealogy (Oxford: Hart Publishing, 2006); Stefano Bertea and George Pavlakos (eds.), The Normativity of Law (Oxford: Hart Publishing, 2011). It starts with a review of Stefano Bertea’s book The Normative Claim of Law (2009).³xStefano Bertea, The Normative Claim of Law (Oxford: Hart Publishing, 2009). One of the main assumptions of this book is that the normativity that the law, according to Bertea, has, cannot be reduced to matters of fact, such as the psychology of the law users. The second part of this paper is mainly devoted to a scrutiny of this assumption. Whether and in what manner such a ‘reduction’ is possible are the questions which guide this second part.

• 2 Internalism, externalism and the dual nature of the law

  The law has a seemingly dual nature. On the one hand it contains guidelines for action; it tells its addressees, amongst others,⁴x This ‘amongst others’ is important, because the law creates a whole ‘world’ of institutional facts. Neil MacCormick, Institutions of Law. An essay in legal theory (Oxford: Oxford University Press, 2007); Jaap C. Hage, ‘A model of juridical acts: part 1: The world of law,’ Artificial Intelligence and Law 19 (2011): 23–48. Only part of this world concerns the issue what to do. Many authors consider this part to be the most important one; the rest would only be auxiliary. A prominent example of this view is Hart, who called duty-imposing rules primary rules, and the rest secondary rules. Herbert L.A. Hart, The Concept of Law, 2^nd ed. (Oxford: Clarendon Press, 1994, 1^st ed. 1961), chapter V. what to do and what not to do. On the other hand, it has the ring of objectivity, because most, if not all, of the law seems to exist in a way that can be established by means of our senses, namely by studying the sources of law. This dual nature, if it really exists, is attractive because it is not at all obvious that something can be objective and nevertheless guide behaviour. Under the influence of, amongst others, Hume, who suggested that an ‘ought’ cannot be derived from an ‘is’⁵x David Hume, A Treatise of Human Nature, Selby-Bigge edition (Oxford: University Press, 1978. Original text 1740), Book III, section 1. and Kant, who argued that real normativity is autonomous,⁶x Immanuel Kant, Grundlegung zur Metaphysik der Sitten (Hamburg: Felix Meiner Verlag, 1906. First edition 1785), 432–433. The references use the page numbers of the Akademie Ausgabe. the impression was created that guidance of behaviour cannot be combined with objectivity. Because of its dual nature, the law seems to be a counterexample to this view.

  However, the view that normativity cannot be objective is persistent. If the law is really objective, does this not imply that it is not really normative? Ethical theory has developed terminology which facilitates this kind of discussion. It makes the distinction between internalist and externalist theories about the nature of morality. Internalism with regard to (moral) ought-judgements is one of the following views:

  1. A person who sincerely believes that he ought to do X will normally be motivated to do X.

  2. A person who sincerely believes that he ought to do X will normally also believe that he has a reason to do X.

  3. A person who ought to do X has a reason to do X.⁷x The different variants of internalism were inspired by Michael Smith, The Moral Problem (Oxford: Blackwell, 1994), 61–62, but do not fully coincide with the distinctions made by Smith, however.

  It is also possible to take an internalist point of view with regard to reasons for acting, and then it would amount to something like the following:

  1. A person who sincerely believes that he has a reason to do X will normally pro tanto be motivated to do X.⁸x The ‘pro tanto’ clause is inserted, because having a reason to do A does not imply that one does not have reasons against doing A. The motivation issue will normally be determined by the balance of all contributory reasons. The relation between a single reason and motivation is merely ‘pro tanto’ (as far as this single reason is concerned), because the reason itself will normally have a motivating effect, but this effect may be canceled out by contrary motivating effects. Compare it to Buridan’s ass, which did not move, being motivated by a stack of hay and a pile of water at different sides of it.

  Externalism is the view that such a necessary connection between what one should do and the motivation or reasons for doing so does not exist. It would allow, to borrow an example from Smith⁹x Smith, The Moral Problem, 60. that a persons says: ‘I admit that I should give to famine relief, but I do not see that I have a reason to do so.’

  Given this terminology, one might allow that the law is objective, but combine this view with an externalist stance towards the law (a ‘detached ought’).¹⁰xJoseph Raz, The Authority of Law (Oxford: Clarendon Press, 1979), 153. It can objectively be established what one should do legally, but this does not imply that one has also reason to do it (externalism). Or one might reason the other way round: One has reason to do what the law prescribes (internalism), but precisely therefore, it cannot be established objectively what the law prescribes.

• 3 Bertea’s argument

  It is this discussion in which Stefano Bertea’s book The Normative Claim of Law takes a position. In this book Bertea aims to do two things. First, he defends the view that the law makes a normative claim on us. And, second, he aims to found this normative claim. In doing these two things, Bertea hopes to increase our understanding of the law.

  3.1 That the law makes a normative claim

  What is this normative claim of the law? That is the claim, allegedly made by the law, that it guides conduct by providing its addressees with reasons for acting as directed.¹¹x Bertea, The Normative Claim of Law, 62. So the law would claim that it should be approached from an internalist perspective.¹²x Readers familiar with the work of Hart will notice that taking this internalist perspective is at least similar to taking the internal point of view in Hart’s sense. In the first chapter of his book Bertea argues that the law can indeed make such a claim. It might be doubted whether the law can make any claims at all, since the law is not human and claiming seems to be an illocutionary act, while performing illocutionary acts seems to be a typical human activity.¹³xJohn L. Austin, How to do things with words, 2^nd ed. O. Urmson and Marina Sbisà (Oxford: Oxford University Press, 1975), 98–132; John R. Searle, Expression and meaning, studies in the theory of speech acts (Cambridge: University Press, 1979), 54–71. Bertea addresses this problem by representing the law (people, norms, interactions, practices, processes and states of affairs) as a unitary system of interconnected illocutionary acts and attributing the normative claim to this system as a whole.

  That the law makes a normative claim on its addressees seems to me to be above discussion. At least part of the law consists of rules that indicate what its addressees should do. It is pointless to have a practice which contains such rules if it is not the aim of this practice to guide conduct. This is in my opinion so obvious that the complex argument presented by Bertea makes the impression of being needlesly complicated. The conclusion of the argument is correct, though.

  In the second and third chapter Bertea intends to spell out the essential traits of the normative claim. He does so by arguing that neither inclusive legal positivism (chapter 2), nor exclusive legal positivism (chapter 3) does full justice to the normative claim of law.

  Inclusive legal positivism is the view that legal norms are identified by means of a convention. This convention may, but does not have to, refer to moral considerations. According to inclusive positivism, legal institutions can guide action by providing legal officials with practical reasons of a distinctive legal kind. These reasons are not reducible to moral reasons; they hold independently of their content, and they preclude further deliberation as to how we ought to act.¹⁴xBertea, The Normative Claim of Law, 75.

  Although Bertea sees much of value in the inclusive-positivist account of the law, he nevertheless finds this account defective on three scores: 1. The necessary status of the normative claim is insufficiently established. 2. The notion of normativity is wrongly characterised. 3. The scope of normativity is constructed too narrow.¹⁵xBertea, The Normative Claim of Law, 78.

  Bertea substantiates the claim that inclusive positivism suffers from the first of these ‘defects’ by reference to the view of Kramer, who is taken as a representative of inclusive positivism. According to Kramer, legal prescriptions are imperatives rather than prescriptions. Whereas, according to Bertea, prescriptions create reasons to act and therefore inhabit the realm of ‘ought’, imperatives are rather instances of a ‘must’, deriving their force from the penalties attached to non-compliance.¹⁶xBertea, The Normative Claim of Law, 76.

  This argument as an argument against inclusive legal positivism in general suffers from at least two weaknesses. First, it is dubious to base criticism of a legal-philosophical current such as inclusive positivism on alleged defects of only one adherent of this current. And second, it is not at all clear why imperatives backed up by sanctions could not generate reasons for acting. It is obvious that they are not the Kantian-style reasons which Bertea will later defend, but to demand that they would be is begging the question. A person who is ordered to do something and threatened with a sanction in case of non-compliance still has some choice whether to comply with the order or not. Arguably the threat of the sanction does not leave this choice a free choice, but even an unfree choice is a choice and not a must. If a robber puts a gun against my head, I have a very strong reason to do what he asks from me, and it is little comfort to know that it might not be a ‘real’ reason.

  The second defect of inclusive positivism according to Bertea is that the normativity of law is characterised as being based on acceptance. A legal norm would bind, in the end because it is accepted as binding. This would make bindingness a socio-psychological category instead of the moral category it is according to Bertea. Whether this is a mistake will be one of the main topics of the following sections of this paper. In my opinion it is a strength, rather than a defect of inclusive positivism.

  The third defect would be that according to inclusive positivists, the law binds officials, and not all legal subjects. Bertea ascribes this view to inclusive positivists because of Coleman’s position that the (ultimate) rule of recognition is based on a shared cooperative activity of legal officials. One may wonder, however, whether it follows from the fact that officials are bound by the rule of recognition, that other legal subjects are not bound by the law. If the law is normative, provides reasons for action, this normativity should be looked for in the internal point of view taken up by (most of) the participants in the legal system, and not in the practice of legal officials.

  Therefore, in order to substantiate his third objection against inclusive legal positivism, Bertea should have argued that according to this view only officials are bound by the law, and in particular by the (ultimate) rule of recognition. Some support for this view might be found in Hart’s Concept of Law, but Bertea provides us with no such argument.¹⁷xHerbert L.A. Hart, The Concept of Law, 2^nd ed. (Oxford: Clarendon Press, 1994, 1^st ed. 1961), 116–7. And even if he had done so, he would still have had another obstacle to overcome. According to Hart, the ultimate rule of recognition is a social rule, and therefore not part of the law.¹⁸xHart, The Concept of Law, 108–9. So, if the officials are only bound by the ultimate rule of recognition, they are still not bound by the law.

  Probably it is best not to ascribe to inclusive legal positivism in general any view concerning the issue who are bound by the law, and then the third objection cuts no ice. That does not subtract, however, from the fact that Bertea rightly objects against the view, whether it is implied by internal legal positivism or not, that the law would only bind officials.

  The exclusive positivist account of the law as exemplified by the work of Raz fares little better in Bertea’s eyes. Bertea mentions (amongst others) the following characteristics as essential for Raz’s position.¹⁹xBertea, The Normative Claim of Law, 107.

  1. Legal reasons are practical reasons and are therefore in this respect (similar to) moral reasons. This aspect of Raz’s view is applauded by Bertea.

  2. Legal reasons are authoritative reasons which exclude other reasons. Exclusion means in this connection that the other reasons are not counted anymore in determining what should be done. Bertea disagrees with this position. Legal reasons may be (very) strong reasons, but they do not exclude everything else.

  3. Because legal reasons are authoritative, they derive their binding force from the authority who issued them and for that reason they are content independent. Here Bertea disagrees, basically for the same reason as under 2. Yes, the law is based on authority, and is in this sense formal, but this authority does not replace the underlying substantive reasons. Authorities are authoritative because they tend to be substantively right. Substantive rightness may be pushed to the background by legal authority, but is never completely replaced by it.

  3.2 Whether the normative claim is founded

  In the second part of his book, Bertea changes focus. The questions whether the law lays a normative claim on us and what the nature of this claim is, is replaced by the question how the normative claim can be justified. Why does the law bind? In this connection Bertea addresses three questions, namely whether the normativity of law can be reduced to purely factual characteristics, how normativity in general can be founded, and what this implies for the normativity of the law.

  The first of these three questions is the topic of chapter 4. In this connection Bertea argues that the normative cannot be reduced to the factual, basically because norms, unlike facts, are not only efficacious or inefficacious but also valid or invalid.²⁰xBertea, The Normative Claim of Law, 170. This last characteristic is interpreted by Bertea as the meaning content of a normative standard, as an interpretative standard that gives determinate sense to events.²¹xBertea, The Normative Claim of Law, 153. Moreover, even if validity could be reduced to the factual, this would be a reduction of the wrong kind, because ‘factual validity’ might show how the law (and morality) can compel action, but not how it can do so unconditionally, that is independently of anyone’s willingness to accept the guidance of the law.²²xBertea, The Normative Claim of Law, 170. In Kantian terminology, it cannot show how factual validity could become a categorical, instead of merely a hypothetical imperative.

  Bertea’s first argument against reductionism has a strong ring of circularity. To begin with, not only norms are valid or invalid, but also arguments, goals (in soccer), attempts (in high jumping), and all juridical acts, such as contracts, legislation, administrative dispositions etc. Their validity consists in their being in agreement with the standards that hold for them. So validity in general has nothing to do with being the meaning content of a normative standard.

  Apparently the validity of norms is a special kind of validity. But what might this special kind of validity be? The suggestion is strong that validity is precisely the special nature of normativity that is assumed to be non-factual. However, if that suggestion is taken seriously, Bertea’s argument is plainly circular: normativity cannot be reduced to facts, because facts are not normative.

  Bertea’s second argument will be discussed later, because I will argue that the only kind of normativity which is understandable is what Bertea would call factual normativity. This factual normativity has nothing to do with hypothetical imperatives, however.

  My formulation of Bertea’s criticism of reductionism, ‘Factual validity can at best become a hypothetical imperative, never a categorical one’, preludes on the turn which Bertea’s argument takes in the fifth chapter of his book in which he addresses the foundation of law’s normativity. Following Kant he looks for this foundation in the rational nature of man. Because Bertea hesitates to swallow the Kantian metaphysics according to which the will is free because man partakes in the noumenal world, he offers a ‘modified Kantian account’ of the foundation of normativity.

  The first step in his argument is that the normativity of law is a species of the more general normativity of practical reason, because the law is a praxis.²³xBertea, The Normative Claim of Law, 172. Later Bertea elaborates a bit on this point, when he writes that at least three features of the law account for its membership in the conceptual family of practical reason. These reasons are that:

  1. the basic vocabulary and fundamental ideas of the law, such as ‘obligation’, ‘permission’ and ‘right’, are forged by practical reason itself;

  2. several legal practices are concerned with finding rational ways to cope with practical issues and set standards of required conduct;

  3. rationality plays a crucial role in the law.²⁴xBertea, The Normative Claim of Law, 175.

  As we will see in section 4, this short passage – too short given its importance – is crucial for much of Bertea’s argument.

  The normativity of practical reason is, according to the author, based on the self-conception of human beings as reflective, autonomous and rational beings.²⁵xBertea, The Normative Claim of Law, 202. This self-conception is in Bertea’s words a practical necessity, a necessity based on the status of humans as agents, as subjects capable of action.²⁶xBertea, The Normative Claim of Law, 207. This status is presupposed by the very idea of practical reason, namely that normativity is based on reasons for action.²⁷xBertea, The Normative Claim of Law, 205. To state it simply, human agency is acting for a reason, and this is only possible for a being that reasons (reflective and rational) and is capable to act on the basis of reasons (autonomous). Therefore we have a duty to comply with the demands of practical reason, unless we are ready to give up our distinctiveness as human agents.²⁸xBertea, The Normative Claim of Law, 210.

  Later we will see that this line of argument, which strongly reminds of classical natural law theory, because it derives guidelines for acting from (presuppositions concerning) human nature, is an instance of epistemic foundationalism, and suffers from the drawbacks attached to it.

  In the sixth and last substantive chapter of his book, Bertea aims to show what the implications are of his views that the law is normative to the extent that it complies with the standards constitutive of human agency and that this normativity has the modified Kantian grounds which he advocated. First Bertea defends, amongst others through a discussion of the opposite views of Fish,²⁹xStanley Fish, Doing what comes naturally (Oxford: Oxford University Press, 1989). that the law derives its normativity from human agency. In short: the law can only be normative to the extent that it supports human agency in the sense of reflectivity, rationality and autonomy.³⁰xBertea, The Normative Claim of Law, 232–3. This support is not an all or nothing matter, but rather a matter of degrees. Moreover, it may involve weighing and balancing and therefore requires argumentation.³¹xBertea, The Normative Claim of Law, 244.

  With regard to the support provided by the law to human agency, Bertea mentions rules which right out support agency either by being deducible from it, or by prescribing what is justified by human agency or prohibiting what is not justified by it.³²xBertea, The Normative Claim of Law, 234. There are also rules which contradict human agency, and if this contradiction is sufficiently serious, such rules lack normativity, although they may still be legal rules.³³xBertea, The Normative Claim of Law, 214, 255. And finally there may be rules that are indifferent to human agency. Such rules may have indirect normative force because they belong to a system which as a whole has normative force because it supports human agency.³⁴xBertea, The Normative Claim of Law, 248. Bertea discusses some examples of rules that allegedly support or contradict human agency, such as rules against racial discrimination, rules that create strict liability, and rules that allow torture.

  3.3 Evaluation of the book

  The nature of the law is one of the central topics of legal philosophy, and deserves the attention it has received, mostly in the English speaking part of the world since Hart’s ground breaking work The Concept of Law. Bertea’s book fits in this tradition. Moreover, it takes a position which he himself calls non-positivist, but which might in my opinion just as well be called a version of classical natural law theory. Although it is possible to criticise Bertea’s views, the author offers an intelligent version of such a theory, in which he does not avoid discussion with both recent versions of legal positivism, his Kantian source of inspiration, and other recent authors. For this reason, The Normative Claim of Law deserves to be read by anyone interested in the modern discussion about the nature of law.

  Bertea is very explicit about his claims and the reasons he offers to support them. For instance, he states at the beginning of his book what his main line of argument will be, follows this line in the six main chapters, and summarises his findings in the conclusion of the book. The same approach is taken in each individual chapter, where the argument of that chapter is both previsaged, then given, and finally summarised. This has the advantage that the reader can easily keep track of where he is in Bertea’s line of reasoning. The unavoidable disadvantage is that the text is rather repetitious.

  The main question, however, is whether Bertea convinces. I have already argued that his view that the law claims to be normative is correct, even obviously so. Whether this is precisely the kind of normativity that Bertea has in mind is a different matter, though. More disputable are Bertea’s views that the law, to the extent that it satisfies the demand that are constitutive of human agency, is normative, that this normativity is the normativity of practical reason in general, and that it is based on human agency. It is to these issues that I will turn in the rest of this paper.

• 4 Two views of the law

  To answer the question whether the law is normative we must be clear what we understand by ‘law’ and by ‘normativity’. There are many different views possible about the nature of law, but I think they can sensibly be grouped under two headings, namely positivist and normative views of the law. Under positivist theories of the law I understand those views according to which law is by definition – and not merely as a matter of contingent fact – positive law, without additional demands. The rules of law are on this view by definition those that exist in social reality. In modern legal systems this means that the legal rules stem from a limited number of sources by means of which the law is laid down. A number of centuries ago, most positive law was customary by nature.³⁵xSometimes, custom is categorised as one of the sources of law, as is done in art. 38 of the Statute for the International Court of Justice. This is problematic since rules must already be counted as law in order to be categorised as customary law. They are custom, because they are law, and not the other way round. Customary law is by definition not source-based. Arguable, customary law is not positive law in the strict sense, because it was not ‘posited’, laid down. With positive law in this more narrow sense, customary law has in common that it exists as a matter of social fact. And since this criterium – existence as social fact – is also used to identify positive law, I classify customary law under positive law in a broad sense.

  On the assumption that the law is by definition positive law in this broad sense, it is not on beforehand clear why this law would be normative. Then Austin’s memorable phrase ‘The existence of law is one thing; its merit or demerit is another’³⁶xJohn Austin, The Province of Jurisprudence Determined (London: Weidenfeld and Nicholson, 1954), 184. Reprinted in 1998 by Hackett Publishing Company. seems to take on. If one makes a sharp distinction between the factual and the normative, as Bertea does, it seems that the law as positive law can only receive its normativity from the ‘outside’. In Bertea’s argument, the contingent nature of law’s normativity is exhibited in the fact that the law is only normative to the extent that it complies with the standards constitutive of human agency. We can therefore conclude that a positivist view of the law seems to lead to an externalist view of law’s normativity. A person who sincerely believes that a particular line of action is prescribed by the law, does not need to be motivated accordingly, nor must he believe that he has a reason to comply, nor does the presence of a legal duty imply a reason to comply with the duty.

  In the next section I will argue that on a different, and in my view better, conception of the relation between the factual and the normative, the normativity of positive law does not need to be contingent in this way. By way of a brief detour, I want to argue that the normativity of law is rather uncomplicated if one has a normative view of the law itself.
  Normative views of the law consider the law to be the answer to some ‘normative’ question, some version of the question what to do. And it is true, necessarily true it would seem, that the answer to the question what to do tells one what to do. That the law is first and foremost an answer to the question what to do is nowadays far from obvious. The view that the law is positive law has become so pervasive that it has become customary to treat evaluative and normative questions in relation to the law as questions what would be desirable law, or which law we should have. Evaluative and normative standards are not standards to determine what the law is, but standards to measure the independently existing law against. And yet this almost standard approach cannot explain why so much legal reasoning is evaluative and normative by nature. On a normative view of the law, however, it is easy to explain why legal reasoning so often relies on the study and interpretation of social reality in general and special texts (legislation, case law, treaties) in particular. Legal certainty is highly desirable and positive law is a major source of legal certainty. Therefore it is highly desirable to have positive law and for that reason (and not by definition or convention) positive law is law. The positive law is on this view not necessarily part of the law, but there are good reasons to assume that even on a normative view of the law, positive law forms the bulk of the law. Authors who mentioned these good reasons include Thomas Aquinas, Gustav Radbruch, and Lon Fuller.³⁷xThomas Aquinas, Summa Theologica I,2 Q. 96, art. 4; Gustav Radbruch, Rechtsphilosophie, 8^th edition, ed. Erik Wolf and Hans-Peter Schneider (Stuttgart: Koehler Verlag, 1973), 165; Lon L. Fuller, The morality of law, revised edition (New Haven: Yale University Press, 1963), ch. II.

  We have different normative practices, such as the prudential one in which individuals ask what is wise for them to do, and the moral practice, in which the question is what one should do, taking not only one’s own interests into account, but also those of others.³⁸xI want to leave the question open whether morality also serves values which are not merely instrumental towards interests. The law neither coincides with prudence nor with morality. It concerns guidelines for behaviour which tend to be enforced by collective means, nowadays mostly by state organs. Taking this relation to collective enforcement into consideration, one might characterise the law as those rules which should be enforced by collective means. This characterisation of the law makes most sense if there exists a social practice of collectively enforcing a set of rules. These rules would be rules of positive law. The existence of positive law would then be a precondition if the normative standard for the identification of law tout court is to make sense.

  Since the desirability of enforcement is closely connected to the desirability of the behaviour which is enforced, the step from the premise that a particular rule should be enforced with collective means to the conclusion that this rule should be complied with is a small one. And therefore the normativity of the law on this normative view of the law is quite obvious.

  One might ask whether this view implies that all positive law can rightfully claim to be obeyed. The answer is negative, because rules that do not deserve to be collectively enforced, not even because of their positivity, are not legal rules. They may be positive ‘law’, but in the end they are not law. Here we have arrived at the natural law position which is nowadays not even defended anymore by self-acclaimed natural law theorists such as John Finnis.³⁹xJohn Finnis, Natural Law and Natural Rights (Oxford: Clarendon Press, 1980), 276–281. Nevertheless this position seems well defendable if one does not start from the assumption that the law is a social phenomenon.

  This brief discussion of normative views of the law was meant to illustrate that it does not need to be a difficult issue to determine whether the law is normative. On a normative view of the law, the law is normative by definition. As we have seen, normative views of the law are not very popular nowadays, and we will ignore this approach for the rest of this paper. The central question for most of the remainder will be how a positivist view of the law relates to the normativity of law. It is in this connection that we will go into some detail concerning the nature of normativity.

• 5 The nature of normativity

  We have seen that Bertea ascribed a view of normativity to inclusive legal positivists which he thinks is wrong, because according to this view normativity is based on acceptance. Such a socio-psychological category would not suffice; the normativity of law should be the normativity of morality. Whether this is correct depends not only on the nature of the law, but also on what the normativity of morality amounts to.

  According to Bertea, the normative claim of law is the claim that the law guides conduct by providing its addressees with reasons for acting as directed.⁴⁰xBertea, The Normative Claim of Law, 62. So far so good, but this explication only shifts the need for elucidation from the normativy of law to the reason-giving nature of law. What does it mean that the law guides conducts by providing reasons for acting? As we will see, the notion of a motivating reason plays a crucial role in this connection. Therefore we need to pay some attention to the phenomenon of reasons.⁴¹x My account in the following is based on the discussion of reasons in Jaap C. Hage, Reasoning with Rules (Dordrecht: Kluwer, 1997) chapter II (Of reasons). Other treatments of reasons, especially in relation to the law, are Raz, Practical Reason and Norms (London: Hutchinson, 1975) and Cristina Redondo, Reasons for Action and the Law (Dordrecht: Kluwer, 1999).

  5.1 Causal reasons

  The notion of a reason plays not only a role in the normative disciplines, but also in discourses which are not normative at all. Reasons are also used to explain things, such as the fact that the train was late or the fact that the sun was eclipsed by the moon, and to predict events, such as a drop of the price of government bonds in the stock exchange, or that a rope will break. All these examples are examples of reasons in the sense of causes, causal reasons.

  Another example of a causal reason is that the reason why I take the bicycle to my work is that I want some exercise. Here my behaviour (an event) is explained by means of my mental state (my want of some exercise plus the belief that taking the bicycle will lead to the desired exercise) which is assumed to have caused it. On the one hand, this example is similar to the other ones: a reason is adduced to explain an event, on the assumption that the reason is (part of) the cause of the event. That the reason is a mental state does not change this, because it was (implicitly) assumed that mental states can cause behaviour. On the other hand, this example is different, because its plausibility depends on it that I consider the fact that it gives me exercise is a guiding,⁴²xRaz, Practical Reason and Norms, 16. or a normative⁴³xSmith, The Moral Problem, 94. reason to take the bike to my work. This has to do with the fact that mental states that are used to explain behaviour are usually motivating reasons.

  5.2 Guiding reasons

  Causal reasons for human behaviour will often be mental states. It is the belief that it is raining that causes my motivation and intention to take an umbrella with me, and it is the fear that I might get hurt that causes me to take safety measures when I have to walk on icy streets. These causal reasons are in a sense special, because they hang together with a different kind of reasons, with guiding reasons.

  Guiding reasons are different from causal reasons in at least three ways. First, whereas causal reasons can be reasons for anything that can be caused, including events which have no mental aspect whatsoever, guiding reasons always relate to behaviour, be it that this category can be taken broadly so as to include mental behaviour such as believing or deciding something.

  Second, if a causal reason is a reason for acting, it will by far most of the times be a mental state, such as a belief, or an experience, such as hunger, anger etc. Guiding reasons are always facts, and by far most of the times they are not facts about mental states.⁴⁴x An example of a guiding reason which is a mental state is the fact that somebody has grief is a guiding reason for consoling him. It is not the belief that it is raining that is a reason for me to take an umbrella; it is the fact that it is raining which is such a reason. Therefore, a person may have a reason to do something, even if he is not aware of it, and somebody may be motivated (by a false belief) to do something, even if he has no reason to do it.

  And third, causal reasons operate in chains of cause and effect, governed by causal laws, while guiding reasons operate in chains of justification, governed by normative laws. So, apparently, guiding reasons and causal reasons, even if they are reasons for acting, are ‘worlds apart’.

  And yet there is a close connection between the two. The connection is the following: If a person P considers some fact to be a (guiding) reason to do X, then the belief of P that this fact obtains explains why P did X (if he did it), or predicts that P will do X. Suppose that Jane considers beautiful weather to be a reason to make a trip to the mountains. Then Jane’s belief that it is beautiful weather explains why Jane made a trip to the mountains, or predicts that she will make such a trip. It is the mental state (the belief) that explains or predicts behaviour; it is the fact expressed by the propositional content of this mental state (the fact that the weather is beautiful) which is the guiding reason.

  This relation between guiding reasons and motivating reasons is precisely what makes internalism with regard to normative judgements so attractive. Beliefs about the presence of guiding reasons tend to motivate believers to do what they believe they have reason to do, and that makes that these beliefs explain behaviour, or help predicting it. Jane’s belief that it is beautiful weather can only explain her trip if there is a general connection between such beliefs and the motivation for such behaviour. Beliefs of the type [It is beautiful weather] should normally cause mental states of the type [motivation to visit the mountains]. This has nothing to do with reasons for acting in particular, but is an immediate consequence of how we see causal relations. Such relations hold between individual facts or events, but must be based on lawlike relations between types of events or facts.⁴⁵xCarl G. Hempel, ‘The function of general laws in history,’ The Journal of Philosophy 39 (1942), 35–48. Brian Fay, ‘General Laws and Explaining Human Behavior’ in Readings in the Philosophy of Social Science, ed. Michael Martin and Lee, C. McIntyre, (Cambridge: MIT Press, 1994), 91–110.

  5.3 Motivating reasons

  I will call the contents of those mental states which stand in such a general causal relation to the motivation to perform a particular kind of act motivating reasons. A motivating reason for doing A is a fact, or rather a possible fact, the belief of which normally leads to the motivation to do A. In Jane’s case the state of affairs ‘It is beautiful weather’ is a motivating reason for visiting the mountains.

  Motivating reasons are like guiding reasons in that they are both facts. They are like causal reasons in that their role is defined in terms of cause and effect. In particular from an internalist point of view, such a view as mentioned in the introduction, there must be a close connection between motivating reasons and guiding reasons. It seems weird if a person sincerely believes that he has a guiding reason to donate to famine relief, and does not even feel the beginning of a motivation do do so. Still the relation between explanatory reasons and guiding reasons is not straightforward. It is for instance not the case that if the belief that F normally motivates a person to do A, facts like F are for this person guiding reasons to do A.

  To begin with, a fact can only be a guiding reason if the person for whom it is a reason knows that it functions as a guiding reason. Suppose that seeing a commercial for beer tends to motivate me to drink a beer, but that I do not know that this connection exists. Then the fact that I see such a commercial is not a guiding reason for me to have a beer.
  Moreover, a fact can only be a guiding reason for a person if this person somehow approves of it. (I am deliberately vague here.) For instance, if I know that seeing commercials about beer tends to motivate me to drink beer, but wished that this were not the case, I do not consider the fact that I see such a commercial as a guiding reason for drinking a beer. If being thirsty motivates me in the same way, and if I have no objections against it, then arguably I consider being thirsty as a reason for drinking a beer.

  However, even if I know that I am being motivated by a particular kind of fact or event and if I approve of being thus motivated, this might still be ‘wrong’. One might argue that, if I were wise, I would disapprove of being motivated to drink beer if I am thirsty, and that this shows that my disposition to be motivated, even if I approve of it, does not show that being thirsty is a reason for me to drink beer. I only think it is a reason, but I am wrong. Real reasons can stand the scrutiny of critical investigation, and the unhealthy effects of drinking beer, one might argue, make that my disposition cannot stand such scrutiny and this shows that I do not have a real reason to drink beer when I am thirsty.

  5.4 Foundationalism

  This line of argument makes sense. Apparently it may be the case that a person tends to be motivated by particular state of mind, approves of that, and therefore believes that he has a reason to do something, but is nevertheless wrong. It may be this insight which makes authors like Bertea reject the idea that guiding reasons and normativity can be defined in psychological or sociological terms. It would also be this reason which makes them reject a ‘reductionist’ analysis of normativity and the acceptance-based view of it held by inclusive legal positivists. Are they right in doing so, or do they merely think that they have a reason for these rejections?

  The beginning of an answer to this question can be found in the observation that the issue at stake is whether guiding reasons and motivating reasons belong to completely different categories, the one normative and the other psychological, and that the argument why this is so has to do with the justification of guiding reasons. We will have a closer look at this justificatory issue to see whether it can lead to the conclusion that guiding reasons and motivating reasons are fundamentally different.

  Suppose that the government proposes to decrease taxes and justifies this measure by the reason that it increases employment. This reason can be questioned in two ways. First it may be asked whether a decrease in taxes really increases employment. This question concerns the truth of the reason.⁴⁶x Actually it is not the reason that is true or false, but the sentence which expresses the state of affairs that would count as a reason. The reason itself obtains or not. And second, the question may be asked whether this fact, if it is a fact, counts as a real reason for decreasing taxes. This is a question after the relevance of the alleged reason. Are the effects on employment relevant for the issue whether taxes should be decreased (or raised)? Because this paper deals with normativity, I will only focus on the second question.⁴⁷x However, in my opinion the justification of ‘factual’ isues is basically the same as that of normative issues, so the following discussion would also apply to the question about the causal connection between decreasing taxes and increasing employment. Jaap C. Hage, ‘The Method of a Truly Normative Legal Science’, in Methodologies of Legal Research, ed. Mark van Hoecke (Oxford: Hart Publishing, 2011), 19–44.

  A fact can only be a guiding reason for a position (e.g. the taxes should be decreased) if it is relevant. If this relevancy is challenged, it can be justified by giving reasons. These reasons should normally not only concern the relevancy of this particular fact for this particular position but the relevancy of this type of facts for this type of position. Obviously these reasons should be relevant too, and it is possible – at least in theory – to support this relevancy by the adduction of still other reasons. The relevancy of these reasons may be challenged too, and … so on.

  As this brief example illustrates, justification is burdened by a problem, which was vividly baptised the Münchhausen-trilemma by Hans Albert, after the famous baron who simply lifted himself out of the morass by pulling on his pigtail.⁴⁸xHans Albert, Traktat über kritische Vernunft (Tübingen: Mohr, 1980), section 2. Because the relevancy of the alleged reasons in a justificatory argument would need to be justified themselves, there seem to be only three possibilities:

  1. the relevancy of some facts are dogmatically accepted as true or justified; this variant is called foundationalism;

  2. the need to justify the relevancy of the reasons leads to an infinite regress, because the relevancy of the reasons given to justify the premises also depends on reasons which need to be justified, and so on …;

  3. the reasons given in a justificatory argument are indirectly justified by the conclusion of the justificatory argument; in other words the justification would be circular.

  Clearly, Bertea did not intend to accept anything dogmatically (foundationalists always have reasons why the foundation really can function as such), but it seems that he nevertheless took recourse to the first strategy when he tried to found the normativity of law in human agency. Apparently it is obvious that normativity requires agency and that agency requires reflection, autonomy and rationality. Let me be clear, this is not to argue that reflection, autonomy and rationality are not important for normativity. That is an issue I do not want to enter into. My purpose here is to point out that Bertea, presumably without realizing it, picked some starting points as ones which are not in need of further justification anymore, and attempted to build his account of normativity on this foundation.

  5.5 From points of view to foundations

  In this connection Bertea used a strategy which one encounters more often, namely to make the step from the nature of a particular point of view to the foundation on which the standards of this point of view are based. One example of this strategy can be found in a book by Geoffrey Warnock, The Object of Morality.⁴⁹xGeoffrey J. Warnock, The Object of Morality (London: Methuen, 1971), 26. Warnock starts his argument by pointing out that the human predicament is not very attractive, partly because of the limited sympaties which humans have for each other’s interests. And then he writes:

  ‘Now, the general suggestion that (guardedly) I wish to put up for consideration is this: that the “general object” of morality, appreciation of which may enable us to understand the basis of moral evaluation, is to contribute to betterment … of the human predicament, primarily and essentially by seeking to countervail “limited sympathies”, and their potentially most damaging effects.’⁵⁰xWarnock, The Object of Morality, 26. Italics in the original.

  What Warnock does in this passage is to start from a view about the nature of morality (and since morality is in his opinion purposive, this is also the object of morality) and end with a starting point from which moral positions can be argued. This starting point is somehow above dispute, because another starting point would not be moral anymore.

  We seem to encounter a similar move in Kant, who starts his argument in the Grundlegung zur Metaphysik der Sitten ⁵¹xKant, Grundlegung zur Metaphysik der Sitten, 393. by dogmatically assuming that the only thing which can be held good without any reservations is a good will. This assumption seems to gain some respectability if one sees it as the characteristic that sets of moral goodness from other kinds of goodness, such as beauty, prudence or legality, to mention some possible alternatives. Morality would then more or less by definition be based on a good will. And from this ‘definition’ it would ‘follow’ – by means of an in my eyes dubious line of argument – that the foundation of morality is the categorical imperative.

  Bertea’s strategy is also similar, although he does not seek a foundation for morality but for practical reason. His argument (171–175) runs as follows: Law has to do with deciding between courses of action. Law governs action by providing reasons for it. And reasons cannot be separated from reasoning, and since this reasoning concerns action, it is practical reasoning. Therefore the law essentially belongs to the sphere of practical reason. This is the first part of the argument, that places the law, given its nature, in the sphere of practical reason. The next part seeks to find a foundation of practical reason in human agency; we already discussed that.

  This approach, which moves from the nature of a particular point of view (morality, the law) to a foundation for reasoning within that point of view, has its attractions, but is open to a line of criticism that was well formulated by Philippa Foot:⁵²xPhilippa Foot, ‘Morality as a System of Hypothetical Imperatives’ in Virtues and Vices (Oxford: Basil Blackwell, 1978), 157–73. if the norms of a particular point of view are derived from the nature of that point of view, one can still raise the question whether we should act in accordance with that point of view. Should we really act morally? Or – in Foot’s formulation – is not morality a set of hypothetical imperatives too? Does not morality only guide those who want to act morally?

  It will not do to say that morality is different from (other) points view because it deals with the question what one should do all things considered. It will not, because it is either one of two possibilities:

  • either morality is a point of view like others and possibly allows to derive its normative foundation from its definition, but then its normativity is merely ‘hypothetical’,

  • or it is not such a point of view, and then it does not allow to derive its normative foundation from the nature of the moral point of view.

  The derivation of a normative foundation from the definition of a point of view is, at least in the sphere of the normative, the only plausible option to make foundationalism work. In fact it might work if one does not demand that the derived norms are ‘categorical’ in the Kantian sense. If, however, one follows Kant and – in Kant’s footsteps – Bertea, in demanding categorical norms for the normativity of morality respectively the law, the foundationalist track does not fly.

  An infinite regress is obviously not a serious possibility to justify normative judgements. The third alternative, if interpreted as a coherentist view of justification, theoretically works,⁵³xKeith Lehrer, ‘A Companion to Epistemology’ in Coherentism, lemma, ed. Jonathan Dancy and Ernest Sosa (Oxford: Blackwell 1992), 67–70; Jaap C. Hage, Studies in Legal Logic (Dordrecht: Springer, 2005), chapter 2. but is useless from a practical point of view because the coherent set that needs to be constructed is infinitely large. The Münchhausen-trilemma does not exhaust the spectrum of possibilities, however. There is a fourth alternative, based on ideas from the so-called Erlanger Schule.⁵⁴xOswald Schwemmer and Paul Lorenzen, Konstruktive Logik, Ethik und Wissenschaftstheorie (Mannheim: Bibliographisches Institut, 1973). As we shall see, this alternatives also makes room for the possibility that guiding reasons are defined in terms of motivating reasons.

  5.6 Is there a gap between motivating and guiding reasons?

  Before continuing the argument about the justification of reasons and normative judgements it may be useful to take a step back and consider how the justification issue relates to the presumed gap between motivating reasons and guiding reasons. This gap between motivating and guiding reasons can be seen as one of the many manifestations of the often assumed gap between is and ought. Motivating reasons are defined in terms of motivation, which is a psychological notion and easily assumed to belong to the category of the ‘is’.⁵⁵x As the reader may notice from the cautious formulation ‘easily assumed’, I am quite hesitant to write about the distinction between ‘is’ and ‘ought’, because these two categories are seldom if ever defined well enough to make the distinction clear. Guiding reasons are not defined in psychological terms (they are seldom defined at all), but are taken to belong to the realm of the ‘ought’ because they determine what ought to be done. The relation between motivating and guiding reasons has therefore direct impact on the alleged gap between ‘is’ and ‘ought’.

  We have seen that somebody can be critical about the facts that tend to motivate him. Does one, for instance, really want to be motivated by the fact that one is thirsty, to drink a beer? If one comes to think of it and considers the effects of alcohol on one’s health, one might be motivated to get rid of this motivating relation between being thirsty and drinking a beer. However, one may also be critical about the relation between beliefs about one’s health and the consumption of stimulants. And then one may also be critical about the facts that one finds relevant in this connection, … and so on. Here we encounter again the regress that we found in connection with foundationalism, but now applied to motivational connections instead of guiding reasons. Moreover, it is clear that any relation between beliefs and motivations can be questioned. The temptation is then to conclude that no such connection can be a solid ground for the motivation to act and that the ‘real’ foundation must therefore be found somewhere else. This ‘somewhere else’ cannot be any motivational connection, because it may be questioned whether this connection should be accepted. Apparently we cannot base any decision about what to do on tendency to be motivated and ‘therefore’ there must be an ‘ought’ or ‘should’ outside the sphere of motivation which makes that we should do something.

  If this explanation of the distinction between motivating reasons and guiding reasons is correct, the adoption of guiding reasons and ‘oughts’ as separate categories that do not belong to the sphere of the ‘is’ is the result of the search for a foundation for decisions about what to do. Since any foundation in motivating factors may be questioned, we look for a foundation in something fundamentally different, and the sphere of the ‘ought’ seems to provide us with it. The search for a foundation can in this way be ended if one assumes with Kantians that it is possible to find a basic ought in reason and humanity, with utilitarians that it can be found in happiness, or with Bertea that it can be found in human agency. The alleged gap between ‘is’ and ‘ought’ and, in its continuation, between motivating and guiding reasons, is then an outflow of epistemic foundationalism and the assumption that motivational connections can never be a sufficient foundation for the relevancy of other facts for decisions what to do.

  5.7 Criticism

  The above section contains a very condensed argument to the effect that it is not possible to define guiding reasons in terms of motivation. If such a definition were possible and if one assumes that guiding reasons belong to the realm of the ‘ought’, the alleged gap between ‘is’ and ‘ought to do’⁵⁶x I write here about ‘ought to do’ instead of a only ‘ought’ because many oughts have nothing to do with action, and the present discussion only deals with action. A general account of the relation between ‘is’ and ‘ought’ should preferably be general, and not be confined to cases of ‘ought to do’. The beginnings of such a general account can be found in Jaap C. Hage, ‘Legal Transactions and the Legal Ought,’ to appear in an edited volume in 2011. would turn out not to be a gap at all, because what ought to be done could then in last instance be defined in terms of the motivation to act.

  It seems to me that the argument of the previous section makes a category mistake. It starts from the assumption that any motivational link between a belief and an action may be questioned. This questioning is a psychological phenomenon. It consists of wondering whether to accept or to reject the motivational link at issue. At least in theory, the process of accepting or rejecting motivational links and wondering whether such acceptance and rejection is based on a motivational link which itself may be accepted or rejected, and so on, can continue infinitely. However, inifinite or not, we are still talking about psychological decisions and dispositions, and not about an ought which is taken to belong to a completely ‘different world’. The observation that there is no firm foundation in motivation, because any motivational link may be questioned at most shows that the foundationalist project has problems. It does not provide any reason why a psychological account of reasons must be replaced by a normative account which presumedly cannot be reduced to a psychological account.

  5.8 Constructivism

  If the search for a foundation for decisions what to do is the cause of postulating a separate category of guiding reasons next to motivating reasons, an argument why such a source is unnecessary would also be an (additional) argument why it is not necessary to postulate unreducible guiding reasons. Such an argument can be provided on the basis of ideas from the so-called Erlanger Schule. The traditional model of justification holds that some thesis (including a decision what should be done) is justified if it can be derived (usually: deduced) from justified premises. This view of justification presupposes a basis of justified premises, the justification of which does not depend on derivation from other justified premises. The derivation of the thesis or decision is nothing else than making explicit what was already implicit in the original premises.

  The alternative view of justification that was proposed by Schwemmer and Lorenzen,⁵⁷xSchwemmer and Lorenzen, Konstruktive Logik. Ethik und Wissenschaftstheorie. members of the so-called school of Erlangen,⁵⁸x Cf. http://de.wikipedia.org/wiki/Erlanger_Konstruktivismus (last consulted on February 16, 2011). is that the basis of justification is assumed as long as it is not brought up for discussion. For instance, it is possible to justify the decision to give money to a beggar by the (presumed) fact that the beggar will use this money to buy alcohol and the goal to promote alcohol consumption of beggars. As long as these premises are not questioned, the justification succeeds and the decision to give money to the beggar is considered to be justified. It remains possible, however, to question one or both of the premises, and when this happens, these premises must be justified. Such a justification makes in turn use of premises that are temporarily assumed, but that can always be questioned and brought up for discussion. On this view there is no basis of justified premises. Which premises are needed for the justification depends on course of the justificatory argument. This argument is not a timeless derivation, but rather a process that takes place in time. It is constructed, and with the construction of the argument, the justification of the final thesis – final in the sense of the end of the argumentation chain, but not as the last step in time – is also constructed.

  Regarding reasons for acting this view brings along that it is not necessary to justify the relevance of a motivating reason, unless this reason is questioned. Questioned by whom? First and foremost by the person who tends to be motivated by the facts which are considered to be guiding reasons. If this person has reasons, motivating reasons, to question whether a reason for acting should really be considered thus, she must think about the issue, and this thinking consists in considering which reasons plead for and against the behaviour guiding force of the questioned reason. Why is the fact that the beggar will use the given money to buy alcohol a reason for giving him money? Would not this contribute to the alcohol addicton of the beggar? The reasons why some facts should be reasons can be questioned themselves too, and then it is possible to give reasons, for reasons, for reasons … etc. The difference with the infinite regress is, however, that this process of giving reasons can stop if there are no doubts anymore. In this connection it is the actual doubts that count; not whether there would be reasons to doubt. Unquestioned reasons can stand by themselves, without foundaton in still other reasons.

  According to the above account, a guiding reason is essentially a fact the awareness of which tends to motivate and of which the tendency to motivate survived any actual critical scrutiny. This account of normativity has two important advantages. First, it clarifies the relation between guiding reasons and the motivation of behaviour. The account is, to state it in modern terminology, internalist. And second, it avoids the infinite regress and the dogmatization of assumptions that threaten foundationalist accounts of normativity.

  There is also a disadvantage, because reasons for acting are not necessarily rational anymore in the sense that they are only acted upon if they would survive any critical scrutiny that would be possible. They only need to survive any critical scrutiny that has actually taken place.⁵⁹x See in this connection also the distinction between dialogical and dialectical theories of justification in Hage, Studies in Legal Logic, 243.

  5.9 Social normativity

  It is only natural that a persons tends to be motivated by the reasons he has to do something, at least if he knows that these reasons actually obtain. For instance, if I am thirsty and being thirsty is a reason for me to have a glass of water, this will normally motivate me to have a glass of water. In other words, if reasons for a person are concerned, an internalist view of these reasons is appropriate.

  Apart from reasons for a person, facts which are considered by a particular person as reasons (for acting), there can also be reasons for a social group perspective or from a point of view. For instance, serious mental suffering is in the eyes of some individuals a reason to allow euthanasia, while in the eyes of the community at large this may be different. Whereas reasons for persons may be defined in terms of a disposition to be motivated, this is more difficult for reasons that count within a social group. Assuming an individualist psychology, motivation is primarily a characteristic of individuals and at most in a derived sense of groups. Nevertheless it is possible to define what counts as a guiding reason in a social group. The following would be an attempt:

  A fact is a guiding reason in a social group if it is a guiding reason for sufficiently many members of this group and if sufficiently many members accept⁶⁰x Acceptance here includes both the belief that something is the case and not having (decisive) objections against this being the case. Notice that a fact can only be a guiding reason if it is accepted in this sense as a guiding reason. that this is the case and that sufficiently many members accept the same.

  This is not the place to go into details concerning group behaviour.⁶¹x The reader who is interested in this topic can find a lot of literature on it. Raimo Tuomela, The Philosophy of Sociality (Oxford: Oxford University Press, 2007) is a good starting point. Here I only want to point out that the internalism, which is so natural in connection with reasons for a person, is not so natural anymore for social reasons in the sense of reasons that are recognised within a social group. Not every member of the group needs to consider the social reasons that are recognised by the group as reasons for him or her personally. Lovesickness may not be recognised in the Netherlands as a reason for self-mutilation, but a particular lovesick person may think differently about that, even if he knows that his views are not generally recognised.

  What holds for social reasons also holds for reasons which belong to a particular point of view. Suppose that Kant were right and that it is unconditionally morally wrong to lie. Then a lover should, morally speaking, not tell his girl friend that she is beautiful if he does not really believe this. For him, however, the fact that telling her so would make her happier is a sufficient reason to lie a little. After all, nobody gets hurt. The moral reason does not motivate him.⁶²x I assume that the moral reason is not merely overridden by the altruistic reason, but completely set aside in the sense that it does not motivate at all. This lover takes the external point of view towards at least part of morality. Clearly if most people would not be motivated by moral reasons, the moral point of view would not make sense and morality would not be a social phenomenon.

  A similar story can a fortiori be told about legal reasons. Given the law of the Netherlands, a judge should legally order the deportation of asylum seekers with economic motives, but this does exclude the possibility that in a particular harrowing case this legal reason does not motivate him at all even though he knows what the law is. As Raz⁶³x Raz, The Authority of Law, 153. wrote: legal judgements can be ‘detached’. Again it holds that if most people are not motivated by legal reasons, the law does not make sense. Or, to state the same thing differently, this society would not recognise law at all, or does not recognise the rules of a particular system that pretends to be the law.⁶⁴x This last demand boils more or less down to the Kelsenian demand that a legal system as a whole must be by and large effective in order to exist and the Hartian demand that at least the officials of a legal system treat the ultimate rule of recognition as a social rule, meaning that they adopt the internal perspective towards it. Hans Kelsen, Reine Rechtslehre, 2^nd ed. (Vienna: Franz Deuticke, 1960), 219; Hart, The Concept of Law, 110, 117.

• 6 Some conclusions

  The nature of normativity is a topic that cannot be dealt with adequately in one paper. The above account can therefore not offer more than a glimpse of an approach. Nevertheless, I will briefly summarise some of the results that were (too) briefly argued.

  1. The temptation to separate guiding reasons strictly from motivating reasons may be caused by a combination of, on the one hand, an attempt to find a solid foundation for guiding reasons and, on the other hand, the (Kantian) rejection of the possibility that such a foundation can be found in personal characteristics such as a disposition to be motivated.

  2. By adopting a constructivist (instead of a foundationalist) theory of justificaton, it is possible to define guiding reasons in terms of motivating reasons, thereby establishing a firm connection between the sphere of normativity and individual psychology.

  3. The connection between guiding reasons and personal psychology makes it evident that if a person believes he has a guiding reason to do something, he will normally be motivated to do so. In other words, this connection strongly supports internalism with regard to guiding reasons.

  4. This support for internalism only holds for personal reasons, not for social reasons or reasons that belong to a particular point of view such as the legal point of view.

  Concerning the second main question which Bertea answers in his book, namely whether the law is, under certain conditions, normative in the sense that it provides reasons for acting, the answer must be balanced. First it is necessary to distinguish between a positivist and a normative view of the law.

  On the normative view of the law, the law is normative by definition, because the question what the law is, is a specific variant on the question what should be done.

  On the positivist view, according to which the law is by definition a social phenomenon, the normativity of law on this view is not obvious, but still there are reasons to assume that the law is normative. Since the law has as one of its main function to prescribe behaviour, it provides its addressees with legal reasons. So, if the law precribes to wear a red hat on Monday mornings between 4 and 5 AM, the subjects of the legal system in question have a legal reason to wear such a hat then. Whether they will be motivated to act accordingly is quite a different issue, especially if this rule is not enforced. Moreover, whether this legal reason is also a guiding reason in the sense that persons would reject it if they were not motivated to act on it is also highly dubious. It seems therefore that on the positivist view of the law, the claim that the law is normative in the sense of reason-giving is either obviously true (it gives legal reasons), or not generally true, because the law does not always motivate, and people do not always reject it if they are not motivated by legal reasons.

  If we look at Bertea’s argument concerning the normativity of law, it is not clear whether he adopts the positivist or the normative view of the law (or still some other view). Given his elaborate argumentation, one would take it that he starts from a positivist view, because otherwise his conclusions might have been more easy to arrive at. However, in his arguments he presupposes that proper law, which takes the demands of human agency seriously, must be normative. Such a presupposition can only be made if one starts from the normative view of the law. That Bertea starts from the normative view seems to be supported by Bertea’s brief argument why the law belongs to the ‘conceptual family of practical reason’. One might therefore argue that Bertea either made it to easy for himself, or too hard. He made it too easy if he started from the positivist view, because then he should have paid more attention to his conclusions that the law also belongs to the conceptual family of practical reason and that proper law provides reasons for action. He made it too hard if he started from the normative view, because then he would not have needed so much argumentation based on human agency and its prerequisites.

Noten

• * This review article has benefited substantially from comments by Stefano Bertea on a draft version. It is not to be expected, however, that Bertea agrees with all, or even most of my comments on his work. Neither is he responsible for any remaining misunderstandings on my side.

• 1 Christine M. Korsgaard, The Sources of Normativity (Cambridge: Cambridge University Press, 1996); Ralph Wedgwood, The Nature of Normativity (Oxford: Clarendon Press, 2007).

• 2 Sylvie Delacroix, Legal norms and normativity: an essay in genealogy (Oxford: Hart Publishing, 2006); Stefano Bertea and George Pavlakos (eds.), The Normativity of Law (Oxford: Hart Publishing, 2011).

• 3 Stefano Bertea, The Normative Claim of Law (Oxford: Hart Publishing, 2009).

• 4 This ‘amongst others’ is important, because the law creates a whole ‘world’ of institutional facts. Neil MacCormick, Institutions of Law. An essay in legal theory (Oxford: Oxford University Press, 2007); Jaap C. Hage, ‘A model of juridical acts: part 1: The world of law,’ Artificial Intelligence and Law 19 (2011): 23–48. Only part of this world concerns the issue what to do. Many authors consider this part to be the most important one; the rest would only be auxiliary. A prominent example of this view is Hart, who called duty-imposing rules primary rules, and the rest secondary rules. Herbert L.A. Hart, The Concept of Law, 2^nd ed. (Oxford: Clarendon Press, 1994, 1^st ed. 1961), chapter V.

• 5 David Hume, A Treatise of Human Nature, Selby-Bigge edition (Oxford: University Press, 1978. Original text 1740), Book III, section 1.

• 6 Immanuel Kant, Grundlegung zur Metaphysik der Sitten (Hamburg: Felix Meiner Verlag, 1906. First edition 1785), 432–433. The references use the page numbers of the Akademie Ausgabe.

• 7 The different variants of internalism were inspired by Michael Smith, The Moral Problem (Oxford: Blackwell, 1994), 61–62, but do not fully coincide with the distinctions made by Smith, however.

• 8 The ‘pro tanto’ clause is inserted, because having a reason to do A does not imply that one does not have reasons against doing A. The motivation issue will normally be determined by the balance of all contributory reasons. The relation between a single reason and motivation is merely ‘pro tanto’ (as far as this single reason is concerned), because the reason itself will normally have a motivating effect, but this effect may be canceled out by contrary motivating effects. Compare it to Buridan’s ass, which did not move, being motivated by a stack of hay and a pile of water at different sides of it.

• 9 Smith, The Moral Problem, 60.

• 10 Joseph Raz, The Authority of Law (Oxford: Clarendon Press, 1979), 153.

• 11 Bertea, The Normative Claim of Law, 62.

• 12 Readers familiar with the work of Hart will notice that taking this internalist perspective is at least similar to taking the internal point of view in Hart’s sense.

• 13 John L. Austin, How to do things with words, 2^nd ed. O. Urmson and Marina Sbisà (Oxford: Oxford University Press, 1975), 98–132; John R. Searle, Expression and meaning, studies in the theory of speech acts (Cambridge: University Press, 1979), 54–71.

• 14 Bertea, The Normative Claim of Law, 75.

• 15 Bertea, The Normative Claim of Law, 78.

• 16 Bertea, The Normative Claim of Law, 76.

• 17 Herbert L.A. Hart, The Concept of Law, 2^nd ed. (Oxford: Clarendon Press, 1994, 1^st ed. 1961), 116–7.

• 18 Hart, The Concept of Law, 108–9.

• 19 Bertea, The Normative Claim of Law, 107.

• 20 Bertea, The Normative Claim of Law, 170.

• 21 Bertea, The Normative Claim of Law, 153.

• 22 Bertea, The Normative Claim of Law, 170.

• 23 Bertea, The Normative Claim of Law, 172.

• 24 Bertea, The Normative Claim of Law, 175.

• 25 Bertea, The Normative Claim of Law, 202.

• 26 Bertea, The Normative Claim of Law, 207.

• 27 Bertea, The Normative Claim of Law, 205.

• 28 Bertea, The Normative Claim of Law, 210.

• 29 Stanley Fish, Doing what comes naturally (Oxford: Oxford University Press, 1989).

• 30 Bertea, The Normative Claim of Law, 232–3.

• 31 Bertea, The Normative Claim of Law, 244.

• 32 Bertea, The Normative Claim of Law, 234.

• 33 Bertea, The Normative Claim of Law, 214, 255.

• 34 Bertea, The Normative Claim of Law, 248.

• 35 Sometimes, custom is categorised as one of the sources of law, as is done in art. 38 of the Statute for the International Court of Justice. This is problematic since rules must already be counted as law in order to be categorised as customary law. They are custom, because they are law, and not the other way round. Customary law is by definition not source-based.

• 36 John Austin, The Province of Jurisprudence Determined (London: Weidenfeld and Nicholson, 1954), 184. Reprinted in 1998 by Hackett Publishing Company.

• 37 Thomas Aquinas, Summa Theologica I,2 Q. 96, art. 4; Gustav Radbruch, Rechtsphilosophie, 8^th edition, ed. Erik Wolf and Hans-Peter Schneider (Stuttgart: Koehler Verlag, 1973), 165; Lon L. Fuller, The morality of law, revised edition (New Haven: Yale University Press, 1963), ch. II.

• 38 I want to leave the question open whether morality also serves values which are not merely instrumental towards interests.

• 39 John Finnis, Natural Law and Natural Rights (Oxford: Clarendon Press, 1980), 276–281.

• 40 Bertea, The Normative Claim of Law, 62.

• 41 My account in the following is based on the discussion of reasons in Jaap C. Hage, Reasoning with Rules (Dordrecht: Kluwer, 1997) chapter II (Of reasons). Other treatments of reasons, especially in relation to the law, are Raz, Practical Reason and Norms (London: Hutchinson, 1975) and Cristina Redondo, Reasons for Action and the Law (Dordrecht: Kluwer, 1999).

• 42 Raz, Practical Reason and Norms, 16.

• 43 Smith, The Moral Problem, 94.

• 44 An example of a guiding reason which is a mental state is the fact that somebody has grief is a guiding reason for consoling him.

• 45 Carl G. Hempel, ‘The function of general laws in history,’ The Journal of Philosophy 39 (1942), 35–48. Brian Fay, ‘General Laws and Explaining Human Behavior’ in Readings in the Philosophy of Social Science, ed. Michael Martin and Lee, C. McIntyre, (Cambridge: MIT Press, 1994), 91–110.

• 46 Actually it is not the reason that is true or false, but the sentence which expresses the state of affairs that would count as a reason. The reason itself obtains or not.

• 47 However, in my opinion the justification of ‘factual’ isues is basically the same as that of normative issues, so the following discussion would also apply to the question about the causal connection between decreasing taxes and increasing employment. Jaap C. Hage, ‘The Method of a Truly Normative Legal Science’, in Methodologies of Legal Research, ed. Mark van Hoecke (Oxford: Hart Publishing, 2011), 19–44.

• 48 Hans Albert, Traktat über kritische Vernunft (Tübingen: Mohr, 1980), section 2.

• 49 Geoffrey J. Warnock, The Object of Morality (London: Methuen, 1971), 26.

• 50 Warnock, The Object of Morality, 26. Italics in the original.

• 51 Kant, Grundlegung zur Metaphysik der Sitten, 393.

• 52 Philippa Foot, ‘Morality as a System of Hypothetical Imperatives’ in Virtues and Vices (Oxford: Basil Blackwell, 1978), 157–73.

• 53 Keith Lehrer, ‘A Companion to Epistemology’ in Coherentism, lemma, ed. Jonathan Dancy and Ernest Sosa (Oxford: Blackwell 1992), 67–70; Jaap C. Hage, Studies in Legal Logic (Dordrecht: Springer, 2005), chapter 2.

• 54 Oswald Schwemmer and Paul Lorenzen, Konstruktive Logik, Ethik und Wissenschaftstheorie (Mannheim: Bibliographisches Institut, 1973).

• 55 As the reader may notice from the cautious formulation ‘easily assumed’, I am quite hesitant to write about the distinction between ‘is’ and ‘ought’, because these two categories are seldom if ever defined well enough to make the distinction clear.

• 56 I write here about ‘ought to do’ instead of a only ‘ought’ because many oughts have nothing to do with action, and the present discussion only deals with action. A general account of the relation between ‘is’ and ‘ought’ should preferably be general, and not be confined to cases of ‘ought to do’. The beginnings of such a general account can be found in Jaap C. Hage, ‘Legal Transactions and the Legal Ought,’ to appear in an edited volume in 2011.

• 57 Schwemmer and Lorenzen, Konstruktive Logik. Ethik und Wissenschaftstheorie.

• 58 Cf. http://de.wikipedia.org/wiki/Erlanger_Konstruktivismus (last consulted on February 16, 2011).

• 59 See in this connection also the distinction between dialogical and dialectical theories of justification in Hage, Studies in Legal Logic, 243.

• 60 Acceptance here includes both the belief that something is the case and not having (decisive) objections against this being the case. Notice that a fact can only be a guiding reason if it is accepted in this sense as a guiding reason.

• 61 The reader who is interested in this topic can find a lot of literature on it. Raimo Tuomela, The Philosophy of Sociality (Oxford: Oxford University Press, 2007) is a good starting point.

• 62 I assume that the moral reason is not merely overridden by the altruistic reason, but completely set aside in the sense that it does not motivate at all.

• 63 Raz, The Authority of Law, 153.

• 64 This last demand boils more or less down to the Kelsenian demand that a legal system as a whole must be by and large effective in order to exist and the Hartian demand that at least the officials of a legal system treat the ultimate rule of recognition as a social rule, meaning that they adopt the internal perspective towards it. Hans Kelsen, Reine Rechtslehre, 2^nd ed. (Vienna: Franz Deuticke, 1960), 219; Hart, The Concept of Law, 110, 117.