A functional approach: Power-conferring rules as reasons for action

3.3.1. The distinction we have just drawn between definitions and norms, on the one hand, and between regulative norms, purely constitutive norms and power-conferring rules, on the other, will be clarified — or confirmed — if we move from the structural to the functional level and look at the role any of these entities plays in practical-legal reasoning.

Obviously, the primary function of norms is that of motivating or guiding the behaviour of people. Undoubtedly, that function is fulfilled by mandatory norms: the (action) norm punishing murder claims, with respect to people in general, to deter that type of behaviour and, with respect to judges, to indicate what they should do in a case of murder. Here, behaviour is, so to speak, guided directly, i. e., by specifying what is prohibited, or what is obli­gatory under certain circumstances (given certain states of affairs) and some­times by ordering sanctions for contrary behaviour which then work as auxi­liary reasons for reluctant addressees (that is, for Holmes' bad man who does not accept norms as guides of behaviour).

In the case of power-conferring rules, the motivation for a certain be­haviour works in an indirect (or, as Raz says, indeterminate) way:^[30] they do not say directly how we should behave under certain circumstances, but how we can obtain some normative result X; the norm conferring the power to get married, for example, shows the steps to be taken in order to bring about a normative result that can be seen as facilitating a partnership through certain guarantees of stability, economic security, etc. Like technical rules, power-con­ferring rules are doubly conditional (Conte 1985a, 357; 1985b, 184; Azzoni 1988, 123): they say how we should behave if certain conditions obtain and if we want to bring about a certain result. Power-conferring rules do, of course, presuppose regulative norms (they would have no functional sense if through them it would not become possible to introduce, derogate, apply, etc., regula­tive norms), whereas regulative norms make sense by themselves; but that ob­viously does not mean that one can say that power-conferring rules do not have the function of guiding or orienting behaviour.

3.3.2. Now, while regulative norms and power-conferring rules are — albeit different types of — guides for action, the same cannot be said of definitions (and, as we will see shortly, with respect to purely constitutive rules, whether or not they are guides for action depends on the nature of the state of affairs operating as the antecedent of the normative change). Again we must turn to the paper by Alchourron and Bulygin already quoted so often, and again we must say that where they err is in assuming that the only function of power­conferring rules is to identify norm-formulations, and that they do not also have a practical function. We can say that Alchourron/Bulygin's error originates in the fact that they see the law exclusively from the point of view of someone re­constructing it theoretically — and who is, therefore, interested in determining what sentences belong to a particular legal system — and not from that of its addressees, such as an ordinary citizen who can be a 'contracting party', a 'testa­tor', etc., a member of parliament who may successfully promote a bill, or a judge who can change the normative situation of certain individuals. They all regard the law as a mechanism that enables them to bring about normative changes; and in our view, one of the criteria for evaluating the (epistemic) quality of a legal theory is — as Hart has pointed out and we already mentioned earlier — whether it is able to account for such points of view. Alchourron/ Bulygin's reductionist attempt looks rather poor when seen from the prism we are looking through now, that is, when norms are regarded as reasons for action. Because while the rule conferring the power to get married does offer a reason for action, i. e., for taking the course of action resulting in 'being married', it is difficult to think that the definition of 'insidiousness' or 'major prison term' are reasons for acting in some way.

To sum up the essence of our view of mandatory norms, power-con­ferring rules, purely constitutive rules and definitions from the point of view of reasons for action, we can say the following. The first — mandatory norms — operate in practical reasoning as categorical imperatives, because for someone who accepts such a norm (for instance, the norm ordering the judge to impose such-and-such a punishment in the case of murder) that is sufficient reason for acting according to it (if the conditions for the application of the norm obtain). Power-conferring rules (and also purely constitutive rules), on the contrary, give rise only to hypothetical imperatives: they are reasons for action if their subject wishes to reach a certain end (a certain normative result). And defini­tions are no reasons for action at all, but criteria enabling us to understand (or identify) norms (reasons for action).

3.3.3. These distinctions can be shown in a more concrete and — it seems to us — more convincing way if we consider the following two patterns of practical reasoning:

E_t) a) If state of affairs X obtains, then it is obligatory for Z to do Y.

b) In case C, X obtains.

c) Hence, in case C, Z should do Y.

E₂) a) If state of affairs X obtains, and if and only if Z does Y, then institutional result R is produced.

b) Z wishes to reach result R.

c) Hence, since state of affairs X obtains, Z should do Y.

The following two arguments could be a possible interpretation of the two pat­terns:

Let us now look at these cases somewhat more closely.

i) While in Ej) and E_r), it is premise a), that is, the norm (the regulative mandatory norm) that functions as an operative reason, in E₂) and E₂·), the operative reason is premise b), that is, not the premise stating the norm (the power-conferring rule), but the one attributing certain ends to certain subjects. In E₂) and E₂·), the normative premise (expressing a power-conferring rule) is simply an auxiliary reason.

ii) Ej), E_r), E₂) and E₂·) are complete arguments, i. e., acceptance of the respective premises necessarily leads to the acceptance of the conclusion, with­out any need for further premises; in each case, there is an operative reason and an auxiliary reason which, together, constitute a complete reason.

Now, to answer the first question necessarily implies embarking on a new argument with new auxiliary and operative reasons (e. g., art. 405 of the Spanish Criminal Code punishes murder with a major prison; judges ought to obey the valid legal norms; being married to B will give A economic security; A and B want A to have economic security); the second question, in contrast, can be answered simply with a clarification (e. g., 'of age' means being at least 18 years old; 'murder' is the killing of a person under certain aggravating circum­stances, e. g. with malice aforethought etc.); in the second case, we do not ob­tain reasons for accepting the premises, but — if you wish — for understanding them; what we want to decide with our answer is not whether we should accept the premise in question; rather, we are interested in something prior to that, i. e., in the sense or meaning of the premise. Of course, the question about 'of age' could be posed in another way; we could ask why those who are at least 18 years old are of age, to which one would have to answer that they are of age because some provision thus stipulates it, and that this provision is binding for legal operators. But in that case, what we have done is precisely to modify the nature of the question: we then do not try to understand a sentence, but to justify why we understand it as we do.

Hi) Ej) and E₂) are different schemes of practical reasoning. The funda­mental difference between them is that the first, but not the second one, is a deontic pattern. That means that in Ej), premise a) is a norm — not a norm­proposition — making it obligatory, prohibited or permitted to perform a cer­tain action, and the same can be said of conclusion c). This is not the case in E₂): E₂) a) is not a deontic sentence stipulating that some conduct is obligatory, permitted or prohibited. In a statute, a sentence like 'Persons of age may get married' actually is ambiguous. On the one hand, it expresses a power-con­ferring rule in which the word 'may' does not have a deontic character but only indicates the capacity of changing normative states of affairs; on the other, the sentence also expresses that the exercise of that power is facultative, that is, that the use of it is regulated by a (regulative) deontic norm modalizing the action in that way.^[33] One could say that premise a) is an abbreviated reference to an entire set of norms (those concerning the institution of marriage) many of which are of a regulative kind (e. g., the one stipulating the mutual obligation of spouses to support each other). But to this we can answer that those regula­tive norms are not used but only — implicitly — mentioned in the argument, because one does not need to accept them for the argument to be true. In other words, pattern E_}) is valid only if a regulative norm is used (in the sense that it is accepted); in E₂), regulative norms are also involved, but the argument is valid irrespective of our attitude towards those norms.

The conclusion of E₂) — sentence E₂) c) — is not of a deontic kind either: the word 'should' in this case has only a technical meaning,^[34] as when one says that if you want the water to boil then you should heat it to 100 de­grees. Thus, for example, A and B can come to the conclusion that they should take the stipulated course of action because they want to get married in order to provide a legal protection for A who has no savings, no stable employment, etc. However, if A's situation suddenly improves, or if a befriended lawyer suggests another way of providing the desired type of economic security, then it may well be possible that they no longer should take the course of action resulting in marriage; that is, whether or not they should do so does not depend on what regulative norms (or power-conferring rule) stipulate, but on the ends the agents set for themselves.

iv) The agents' ends, however, can be affected by mandatory norms (rules or principles). That is, premise b) of pattern E₂) — that Z wishes to attain result R — is something the legal order can make depend exclusively on the individuals will, or can stipulate as obligatory (as in the case of judges for whom it is obligatory to produce a result, consisting in a legally founded sen­tence, in the cases they hear). We would then have the case that a mandatory norm becomes the reason justifying the operative reason of that pattern of rea­soning. But that, of course, does not impede that E_t) and E₂) must be seen as two different patterns of practical reasoning which, obviously, can be inter­twined in a great number of ways and present different degrees of complexity. Just as the law — considered in a static way — cannot be fully understood if we try to reduce all its sentences to one and the same form, legal practical rea­soning cannot be reduced to one single pattern.