Mandatory rules as peremptory reasons and principles as non- peremptory reasons; the 'closed' or 'open' configuration of the conditions of application

In chapter I we have adoptedAlchourron/Bulygin's characterization of norms as correlations between generic cases (sets of properties) and normative solutions, on the one hand, and Hart's characterization of authoritative legal reasons as peremptory and content-independent reasons, on the other.

1.1. Prieto's critique

Let us begin with Prieto's critique which we will quote in extenso and then ana­lyse step by step. According to a certain interpretation, Prieto writes,

"what distinguishes rules from principles is said to be that a priori, before their application, we can say precisely in what cases a norm should be observed, since the order foresees, or should foresee, all possible exceptions to its application, whereas principles do not and cannot have such a clause [...] In consequence, the difference is said to be that while we can know with certainty when the solution foreseen in a norm will be imposed, there is a margin of doubt concerning the solution supported by a principle."

Among those who hold this thesis, Prieto names Dworkin and ourselves: ’’Al­though they differ from Dworkin in important ways, I think that among us M.

”[A]ctually, in legal reasoning rules often appear as hermeneutic criteria and not as specific rules exhaustively treating the case at hand; they also must often be combined with other legal stan­dards (principles or norms), thus changing their extent and their very scope of application; and finally, their meaning is always open-textured or potentially vague, so they do not completely and absolutely enumerate a catalogue of possible exceptions for their application. Thus, the indi­cated position seems to be overly confident in the logical finiteness, if not of the entire system, at least of the system of rules. [Because] it seems rather improbable that someone could be capable of completely enumerating all possible exceptions to the application of a normative solution: on the one hand, because through legislation or jurisprudence new exceptions can always arise — e. g., through the appearance of new exempting or extenuating circumstances — and, on the other, because changes in the interpretation of the law also can, in the guise of old exceptions, open the door to new circumstances — e.

In order to reply to this critique, we must try to separate the different questions implied in it:

1. A first distinction that is important here, and which Prieto seems to pass over, is the one concerning the difference between generic and individual cases. Generic cases are nothing but sets of properties. Therefore, a system of rules that contains an exclusive general rule permitting everything that does not fall under the prohibitions particular rules correlate with the descriptions of the generic cases they themselves contain, would have no normative gaps, i. e., would be a complete — or, as Prieto prefers to say, a logically finite — system with respect to the generic cases. Quite another matter is that in the description of generic cases there may be problems of semantic indeterminacy such that with respect to an individual case (or a case less generic than that described in the rule) doubts may arise about what generic case it belongs to.

The idea that the conditions of application in rules are configured in a closed way (in contrast to what is the case with principles) can be elucidated, somewhat paradoxically, by looking at the work of authors who, like Al­chourron and Bulygin, do not have much sympathy for the rules-principles dis­tinction. In an article first published in 1988 ('Conditionality and the represen­tation of legal norms', in Alchourron/Bulygin 1991), Alchourron asks how legal norms can be adequately (that is, in a way compatible with the common intui­tive interpretation) formalized. The problem is that norms make the stipulation of obligations or the conferring of rights depend on the satisfaction of certain conditions. However, norms usually do not explicitly mention all the conditions required for the corresponding obligation or right to arise; generally, they tend to omit those negative conditions whose presence prevents a right or obligation from emerging. This leads to problems of the following kind. Suppose a nor­mative system includes the two norms:

a) Judges shall punish those who have committed murder.

b) Judges shall not punish persons under age.

Using bivalent propositional logic and standard deontic logic, they would be expressed through the following logical formula:

c) p Oq.

d) r—> O-q.

But from this it follows that in the case when the conditions of applica­tion of the first norm (having committed murder: p) and the second norm (be­ing under age: r) both hold, then judges face a conflict of obligations. Because (with the law of strengthening the antecedent) from c) one can derive p.r —> Oq, and from d) one can derive p.r —> O-q. That means that the legal system forces a judge to conclude that he has the obligation to punish as well as the ob­ligation not to punish those who commit murder while being under age.

In order to solve this problem, Alchourron, on the one hand, distingui­shes between the language of the norms of the system and the meta-language that describes the obligations resulting from the system in question, and, on the other, he introduces a relation that under certain conditions gives priority and preference to one norm over another. In his view, then, there is in fact a conflict of obligations — on the level of norms — in the case just described, but the conflict is solved — at the level of the meta-language — because norm d) has priority over norm c). Thus, the answer to the question "What should a judge do in the case of p.r?" is simply that he should not punish.

Now, Alchourron shows that in legal systems which have the type of ordering relation just presented (a transitive and asymmetric relation), under no condition (except an impossible one) can there be conflicts of obligation, since what is obligatory is determined not only by the conceptual content of the norms but also by their rank order. What he does not seem to see, however, is that in a legal system there may be norms stipulating incompatible normative solutions that do not stand in such a relation, and where this is not a simple case of antinomy either. And that is precisely what, in our view, happens with prin­ciples: they stipulate obligations, configuring their conditions of application in what we have called an open way — 'p is obligatory unless that obligation is overruled by a principle which in the case at hand has higher weight' — and without establishing an order, because the system does not predetermine a rank order (the 'relative weight') for the case of conflicting principles. Therefore, in order to solve a case that involves principles, an intermediate operation is re­quired, that is, a new rule must be stipulated (on the basis of those principles). This operation, which consists in transforming principles into rules, is usually called specification. We will come back to it later, in the context of what we have called the expansive force of principles which, as Peczenik has pointed out to us, we did not sufficiently explain.

2. The assertion that rules configure their conditions of application in 'open' form, thus, does not deny, but rather is perfectly compatible with admitting that the description of those conditions may present a zone of penumbra where the subsumption of particular individual cases may be doubtful. And, of course, we also do not think that the fact that "through legislation or jurisprudence there can always arise new exceptions" to the applicability of a rule poses any prob­lem for our conception. That the normative authorities of a system may change that system by restricting the conditions of application of some normative solution is obvious; the idea that rules present their conditions of application in 'closed' form is, of course, limited to the rules existing at a given moment and does not imply any thesis about changes in the normative system.

3. Although it is somewhat more of a problem, it also does not seem to bring any special difficulty for our conception if it is admitted that a case that is not normatively regulated by the system is solved by analogy. What the admissi­bility of the argument by analogy shows is precisely that when an individual case cannot be subsumed under the conditions of application of an existing rule (and precisely because such conditions of application are of a 'closed' charac­ter), as a basis for his decision, the judge must construct a general rule corre­lating other conditions of application with the same normative solution which an existing rule correlates with conditions of application the judge sees as sub­stantially similar. And such a relation of similarity can not be asserted (nor de­nied) on any basis other than the principle that explains and justifies the exist­ing rule. That means that the argument by analogy always implies the use of principles, although it does not mean that recourse to analogy and to principles are the same thing. Actually, analogy is an argument — or, more precisely, a structure of reasoning —, whereas principles are a material that must neces­sarily be used in that type of reasoning.^[10]

4. Prieto does not object against our thesis that principles present their condi­tions of application in 'open' form. But he asserts that this has a consequence which invalidates our characterization of rules, when he writes: "[...] if one as­serts that the exceptions to principles, and therefore also their conditions of ap­plication, cannot be enumerated, and if one asserts, in contrast, that principles can ground exceptions from a rule, then the logical consequence is that we also cannot know the cases of exception from a rule". Here, Prieto is right; but that does not have the fatal consequences for our proposal that he claims, although it does, of course, force us to specify — as we have already done in ch. I — the scope of our thesis that rules present their conditions of application in 'closed' form and constitute — when those conditions of application apply — peremp­tory reasons.

In our view, it is true that the applicability of any rule is conditioned on its application not being in conflict with a principle that has greater weight with respect to the relevant properties of the case. Once this is recognized, we could, of course, conclude — as realists do — that rules are merely cute little toys and that legal reasoning is always radically open, and then dedicate ourselves to verbal games about indeterminacy, like those the authors of Critical Legal Stu­dies are so fond of. The problem with this kind of orientation, as has often been explained, is that it clashes with the evidence that in the vast majority of indi­vidual cases taken to the courts, the subsumption under the generic case treated in some general rule — and the issuing of a judicial decision thus grounded — does not give rise to any kind of controversy in the legal community.

In order to be able to account for both circumstances — the subordina­tion of the applicability of rules to principles and the fact that the great majority of cases is solved through an uncontroversial application of rules — we pro­posed to understand the way in which the law guides the reasoning of its organs of application as a two-level structure: on a first level, we said, the law ’’im­poses on the judicial organs the duty to find a balance of reasons, where only legal standards count as reasons and where other reasons may be taken into ac­count only to the extent that the legal standards themselves allow this”; on a second level, ’’that balance of reasons in most, though not in all, cases requires the use of a legal rule, that is, a peremptory reason, as the basis for decision”. Thus, that rules present their conditions of application in 'closed' form (and that rules themselves operate as peremptory reasons), in our view, applies only at the second level.^[11] Of course, this second level is reached when the principle that 'One ought to do what legal rules prescribe' is not overruled, on the first level, by a principle that in the case at hand has higher weight. But when that principle is not overruled, then in the reasoning of the judicial organ the legal rule under which the individual case is subsumed functions as a peremptory reason for issuing a decision whose content is in accordance with the rule.

1.2. Peczenik's critique

Aleksander Peczenik rejects our idea that principles are merely first-order rea­sons for deciding in a certain way, whereas rules are 'peremptory' or 'protected' reasons. Thus, he writes that "not only principles, but also some rules create merely prima facie duties and, therefore, require deliberation". And he goes on:

"I think that 'legal standards' allow us to take into account all morally relevant considerations [...] Established legal norms constitute prima facie reasons that must be weighed and balanced against other reasons. Those prima facie reasons are first-order reasons for performing a certain action, A, and, at the same time, second-order reasons. In this latter capacity, they indicate that the reasons for not doing A may prevail only if they are particularly strong, that is, clearly stronger than they would need to be in a free moral debate."

These theses of Peczenik — which in the paper to which we are replying here only are stated — are developed more fully in his important book On Law and Reason (Peczenik 1989) to which he himself refers. In that book, in a section titled 'Weighing Rules', one can read the following:

"However, not only principles but also some rules create a merely prima facie duty. This is true about both moral and legal rules. For example, one ought not to kill people. The moral rule for­bids prima facie all killing but to state that a given individual, all things considered, ought not to be killed, one must also pay attention to other rules, stipulating exceptions; for instance, in a de­fensive war, one may kill the aggressors. [...] All socially established legal norms, expressed in statutes, precedents etc., have a merely prima facie character. The step from prima-facie legal rules to the all-things-considered legal (and moral) obligations, claims etc. involves evaluative in­terpretation, that is, weighing and balancing." (Peczenik 1989, 80 f.)

In our view, several objections can be raised against this conception of Pe­czenik:

1. The first is that his example of a rule creating merely a prima facie duty (the prohibition to kill), according to our definition, is not a rule, but a principle. This is so because that norm does not determine its conditions of application other than in what we have called open form (which we regard as characteristic for principles): "It is prohibited to kill, unless there is another principle pertinent to the case at hand that has more weight". Of course, one can object against drawing a conceptual distinction between types of legal norms, like the one we have drawn between principles and rules, on the basis of whether the conditions of application are determined in an 'open' or 'closed' way, by saying, e. g., that this characteristic is not very relevant, that it does not reflect ordinary usage of the terms 'rule' and 'principle', etc. But if one accepts our distinction, then it is clear that (in contrast to the rules on homicide contained in the Cri­minal Code) 'Thou shalt not kill' is not a rule, but a principle. Besides, even ac­cording to Peczenik's own characterization of the distinction between rules and principles one would have to conclude that the prohibition to kill is to be under­stood as a principle. This is what seems to be implied by his remarks that "Un­like a principle, the rule [...] does not express a single value but a compromise of many values (and corresponding principles)." (Peczenik 1989, 81) and that "the main source of the justificatory force of principles is their one-to-one link to the corresponding values. Every principle corresponds to some value [...]" (Peczenik 1992).

2. The second objection refers to Peczenik's thesis that legal standards allow one to take into account all morally relevant considerations, that established legal norms constitute prima facie reasons which in their dimension of second- order reasons only indicate that the reasons for not performing the action re­quired by them can prevail only if they are especially strong. In our view, this thesis is ambiguous, because it can be understood i) as a thesis that intends to account for the claims of the law, or ii) as a thesis that intends to account for the attitude a practically rational subject should adopt with respect to such claims. Neither of the two cases seems acceptable to us. If it is meant to refer to the claims of the law, it is clearly false. To give an example that can hardly be controversial: To restrict someone's basic rights on the grounds of his race contradicts the morally relevant consideration that all human beings should be granted such basic rights for the simple fact that they are human beings. But, obviously, there have been legal orders that did not allow one to take this con­sideration into account. And, on the other hand, if Peczenik's thesis is under­stood as referring to the attitude a practically rational subject should take with respect to the claims of the law, then the thesis concedes too much. To say that a practically rational subject should always consider existing legal norms as prima facie reasons for performing the actions prescribed by them would mean that a practically rational subject should accept that there is a prima facie gene­ral moral obligation to obey the law. This is something Aleksander Peczenik has defended repeatedly (cf. 1989, 238 ff.; 1990, 96 ff.), but in our view, his arguments are not convincing. As he himself says, the "central point" of his theory is the following: "There exists a general prima facie moral obligation to obey the law because general disobedience would create chaos" (1989, 246). Peczenik justifies this thesis not by an appeal to the causal effect acts of dis­obedience would have on the maintenance of the system, but by an appeal to the requirement of universalization. Peczenik admits that there are acts of dis­obedience that do not at all increase the probability of other acts of disobedi­ence, and he thinks that Raz's critique of attempts to ground the thesis of the existence of a prima facie general moral obligation to obey the law on the cau­sal effect (of a 'bad example') is correct. Peczenik himself grounds his thesis on the following "universal premise" that "is a consequence of [the] universalisa- ble character of morality": "I have a prima facie moral obligation to act in such a way that my action could be repeated by everybody without creating morally wrong consequences" (Peczenik 1989, 246). But if the principle according to which I act is that 'It is morally legitimate to disobey the law when my act of disobedience does not in any way increase the chance of other acts of dis­obedience and, therefore, does not causally affect the conservation of the sys­tem', then it is analytically true that everyone could act on that principle (that is, that the principle could be universalized) without morally incorrect conse­quences (i. e., an increase in the probability of other acts of disobedience, and, in consequence, the collapse of the system) (cf. on this point Bayon 1991a, 708 f.). In fact, then, Peczenik's argument does not serve as a justification of his thesis, just as it would not serve as a justification of the judgment that 'There is a prima facie general obligation not to drive by kilometer X of road Z at H hours' to say that if everyone would drive by kilometer X of road Z at H hours, traffic would collapse. But this already surpasses the limits of our present topic concerning principles in the law.