Skip to navigation – Site map

HomeNuméros44The closure of systems of penal s...

The closure of systems of penal sanction norms as systems of constitutive rules

Juan Pablo Mañalich R.
This article is a translation of:
La clausura de los sistemas de normas de sanción penal como sistemas de reglas constitutivas [es]

Abstract

This article analyses the status of the principle of penal legality as a residual closure rule. It does so in the same sense in which Alchourrón and Bulygin use this notion but rejects its characterization as a rule that would qualify every action that is not “prohibited according to penal law” as “permitted according to penal law”. The crucial premise supporting this rejection can be found in the proposition that penal sanction norms have the status of constitutive rules, the function of which, in Hohfeld’s terminology, is to institute liabilities-to-punishment that are correlative to punitive powers. This position finds decisive support in Hart’s conception of rules that establish legal sanctions as secondary rules of adjudication, which back up some corresponding set of primary rules of obligation. On this basis, to serve as a residual closure rule for a system of penal sanction norms, the principle of legality must also be understood as a constitutive rule

Top of page

Full text

1 Introduction: the closure of the systems of penal sanction norms

  • 1 Arriagada 2021a: 60 (italics suppressed).

1In an important recent work, Arriagada follows Alchourrón and Bulygin (“A&B”) in making sense of the “principle of [penal] legality” as a residual closure rule. In this regard, she claims that the systems of norms susceptible of being closed through such a rule ought to be thought of as systems of regulative rules, inasmuch as the closure rule in question is construed as establishing “that all that is not penally prohibited is penally permitted”.1 It is rather unlikely, however, that such an interpretative reformulation is plausible.

  • 2 See Alchourrón & Bulygin 1971: 20-21, 116-117.

2 Like the concept of a normative gap, the concept of a closed (in the sense of “necessarily complete”) system of norms is relational. As A&B were able to show, the existence or inexistence of a normative gap is relative to a given universe of (generic) cases, as well as to a universe of normative solutions.2 This latter point raises the question of which is the universe of normative solutions to which the closure in question would be relative. Arriagada’s suggested interpretative reformulation makes use of the expressions “penally prohibited” and “penally permitted” to identify an alleged dyad of “deontic qualifications” that, as such —i.e., in the language level of the formulation of the rules that correlate any of those two deontic qualifications with some action type— are contradictory, that is, interdefinable by means of negation.

3 That the closure rule in question operates as a second order rule in relation to the rules that make up the system thus closed means that it imposes a “residual” deontic qualification upon every action type that is not correlated with some deontic qualification under the respective system of first order rules. At the language level of so-called “normative propositions”, the effect of the principle of legality, taken as a closure rule, would consist in making of φ an action type that is penally permitted in the strong sense, when under the system of first order rules φ is penally permitted in the weak sense.

  • 3 Alchourrón & Bulygin 1971: 142.
  • 4 See Alchourrón & Bulygin 1971: 142-143.
  • 5 Alchourrón & Bulygin 1971: 143.

4 Yet, there are good reasons to resist the postulation of deontic qualifications specifically associated with the phrases “penally prohibited” and “penally permitted”. To understand why, one may consider how A&B explain what the nature of the closure rule under analysis would be: “[t]he rule ‘nullum crimen sine lege’ may also be understood as a rule of closure permitting all the actions not prohibited by penal law”.3 A few lines later, however, A&B note that the given system of “penal law” would thus become closed in relation to a universe of normative solutions integrated by “penal prohibitions”, namely, prohibitions linked to a sanction of a determinate kind.4 They then assert: “The fact that an action is penally permitted means that it does not incur penal sanctions, but this does not exclude the possibility that it may be prohibited by some other norm”.5

  • 6 For an early rebuttal of the thesis postulating the existence of such thing as a “penal unlawfulnes (...)
  • 7 Hart 2012: 97-98; hereto Mañalich (2012: 580-582). On this issue, see infra, sec. 3.
  • 8 Raz 1980: 24, 155-156.

5 One may now reflect upon what ought to be inferred from this last assertion, namely, that the fact that an action type is “penally permitted” would have no relation whatsoever with it being permitted, but would rather only imply that its instantiation “does not incur penal sanctions”.6 Thus, the expression “penally prohibited” does not designate some “complex” deontic qualification, but only the conjunction of a deontic characterization resulting from the applicability of a given norm, on the one hand, and the punitive reinforcement of that same norm, on the other. As Hart pointed out, sanctions capable of being imposed in virtue of the transgression of some “primary rule of obligation” —i.e., of a regulative rule that may have the character of a prohibitive norm— are specified through a subclass of those secondary rules that Hart himself termed “adjudication rules”.7 Hence, the phrase “penally prohibited” works as an ellipsis that indicates that the form of behavior in question is placed under a (primary) prohibitive norm, which in turn is backed by a (secondary) sanction norm, so that between the two norms a “punitive relation” à la Raz holds.8

  • 9 Véase Alonso 2006: 171-172.
  • 10 Alonso 2006: 172.
  • 11 Alchourrón & Bulygin 1971: 141-142.

6 This raises a further question concerning the nature of the rules that may give form to a normative system capable of being closed through the principle of penal legality. This question has been carefully answered by Alonso when analyzing the status of the legality principle, for which he makes use A&B’s theoretical framework.9 According to Alonso, from this principle it would follow that “if a conduct is not defined as a criminal offense, the judge is under the obligation of acquitting the defendant”, which would license the claim that “contemporary systems of penal law are complete with regard to the conduct of imposing a penal sanction”.10 Alonso thus draws upon a possibility not further considered by A&B. For although A&B explicitly notice that the nullum crimen sine lege maxim may be taken as expressive of a “directive give to judges”,11 they do not explore the possibility of characterizing it as a residual closure rule when understood in that last sense.

  • 12 Alchourrón & Bulygin 1971: 148, 151-153.
  • 13 This characterization is widely extended in criminal law scholarship; for a sufficiently representa (...)

7 Yet, from the fact that the corresponding closure rule must be framed as referring to secondary norms, i.e., norms that belong to what A&B call the “judge system” (as opposed to the “subject system”),12 it does not follow that the proper understanding of those secondary norms would amount to taking them as regulative rules, i.e. as rules that would ground a duty for the competent judge or court to impose the sanction respectively provided by them.13

  • 14 Binding 1872: 7-14.
  • 15 This is fully compatible with the possibility of a judge being obligated by some regulative rule to (...)
  • 16 Binding 1922: 19-25.

8 In the development of a theory of norms for the systems of penal law, the history of the problem just posed is extremely intricate. In the first edition of the first of the four volumes of his Die Normen und ihre Übertretung, appearing in 1872, Binding defined what we currently call a “penal sanction norm” —in his own words: a “penal law”— as an “imperative” addressed to the state as such, which would thus become the subject of a (conditional) “duty to punish”.14 This relied on a persuasive refutation of both the claim that such a norm would impose duties upon those that, by fulfilling its application condition, become exposed to being punished, and the claim that such a norm would impose duties upon the judicial organs enabled to actualize its legal consequence.15 Yet already in the second edition of the same first volume from 1890, Binding renounced that very conceptualization. He instead replaced it with the thesis that a “penal law” would not have the character of a norm capable of grounding duties, but rather the character of an “affirmative legal proposition”, the function of which would be to fix the relation emerging between the state and the criminal as a result of the transgression, imputable to the latter, of a punitively reinforced norm.16

  • 17 Hereto, see Mañalich 2012: 581-582.

9Although the following was certainly not a theoretical possibility available to Binding at the time, the most plausible reconstruction of his thesis is to read it as a characterization of penal sanction norms as constitutive rules.17 The goal of the present contribution is to offer a detailed defense of this last claim, so that we can grasp its implications for the understanding of the principle of penal legality as a residual closure rule.

2 Power-conferring rules as constitutive rules

2.1 The Hohfeldian conception of legal powers

  • 18 Kurki 2017a: passim.

10Kurki offers a suitable point of departure for the inquiry thus sketched. He notes that, under Hohfeld’s definition, the term “power” has a considerably wider extension than the one attributed to the term “competence”, when the latter is used to designate the kind of legal position correlative to a liability.18 It may be worthwhile here to cite Hohfeld in extenso:

  • 19 Hohfeld 1919: 50-51

A change in a given legal relation may result (1) from some superadded fact or group of facts not under the volitional control of a human being (or human beings); or (2) from some superadded fact or group of facts which are under the volitional control of one or more human beings. As regards the second class of cases, the person (or persons) whose volitional control is paramount may be said to have the (legal) power to effect the particular change of legal relations that is involved in the problem.19

  • 20 Hohfeld 1919: 50.
  • 21 Kurki 2017a: 31. It should be noted that in the context of the sentence “a homeless person could co (...)
  • 22 Kurki 2017a: 44.

11To analyze the implications of this “approximate explanation”,20 Kurki makes use of the following example: “a homeless person could commit a crime in order to spend the night in jail, and this would constitute an exercise of a legal power according to Hohfeld’s account”.21 For “[a] duty is in most cases accompanied by a power-to-subject-oneself-to-legal-sanctions-through-acting-in-breach-of-the-duty, and a corresponding liability, as with the vagrant case above”.22

  • 23 This reformulation suggest that to characterize the vagrant’s situation as one in which he brings a (...)

12 In the vocabulary favored here, Kurki’s observations can be reformulated as follows: by —imputably— fulfilling the application condition of a penal sanction norm, the perpetrator would exercise a legal power, since the offense’s perpetration puts him in a position of liability-to-punishment.23 This would follow from the conjunction of the following two propositions. (1) By perpetrating the offense and thus fulfilling the antecedent condition of the respective sanction norm, the agent brings about an alteration of the positions that shape a legal relation, namely, (a) the immunity-to-punishment susceptible of being conferred by the relevant sanction norm, in which every person who does not perpetrate an offense of the corresponding kind stands, and (b) the correlative disability affecting the state organ (or organs) that would otherwise hold the power to impose such punishment. (2) Under the imputation rules that shape a model of penal law subordinated to the (subjective) responsibility principle, the offense’s perpetration, which results in the transformation just described, is defined as being under the agent’s volitional control. Otherwise the lack of (subjective) responsibility for the instance of felonious behavior would preclude the transformation of her position of immunity into a position of liability-to-punishment.

  • 24 Kurki 2017a: 45.

13 This reformulation of Kurki’s point has the advantage of making transparent the irrelevance of the fact that by perpetrating the offense the vagrant is seeking to be incarcerated. This matters because one of the most salient features of the “approximate explanation” of the concept of a legal power that Hohfeld gives us, and which Kurki endorses tout court, lies precisely in its lack of any reference to intentions, desires, or other attitudes that may accompany the exercise of a thus understood power. Therefore, to sustain the claim that in that very case the vagrant would be exercising a legal power, the observation that he would be “availing himself of a possibility to manipulate the normative reality in a way beneficial to him”24 is rather uncalled-for. For if the concept of a legal power is defined in terms of the “approximate explanation” offered by Hohfeld, then every perpetrator of an offense would, when perpetrating it, exercise a legal power, since its perpetration would correspond to the performance of a voluntary act that would result in the suppression of an immunity position regarding the punishment specified by the sanction norm that criminalizes the offense in question.

2.2 Attitudinalistic reconstructions of the concept of a legal power

14A well-known strategy for avoiding the conclusion that someone who perpetrates a criminal offense would eo ipso be exercising a legal power consists in rejecting the Hohfeldian definition of “power” and replacing it with a more demanding definition. The most notorious versions of such a strategy directly or indirectly incorporate an attitudinal component in the characterization of the class of acts through the performance of which someone may exercise a legal power.

  • 25 Halpin 1996: 141.

15 A direct reference to an attitudinal factor emerges in Halpin’s proposal. His starting point lies in the observation, that he takes to be “banal”, that for producing a transformation of a given legal situation it would always suffice to bring about that an instituted condition for such a transformation be fulfilled, whereas the condition would consist in the occurrence of an event of a certain kind.25 Yet this raises the following question: why is it necessary, in some contexts, to identify the fulfillment of the relevant condition with the exercise of a legal power?

  • 26 Halpin 1996: 145-147.
  • 27 See Raz 1990: 65.
  • 28 Halpin 1996: 145. A somewhat proximate, but by no means identical analysis of the concept of a lega (...)
  • 29 Halpin 1996: 145.

16 After analyzing some possibilities explored by other theorists, Halpin claims that an adequate answer to this last question would involve detecting the significance that the notion of decision would have for the analysis of the concept of a legal power.26 One should note here that adopting a decision is tantamount to forming an intention,27 which enables the acknowledgment of the attitudinal component of the concept of a legal power, according to Halpin’s approach. Thus, it would be a defining feature of a legal power that the relevant condition, upon the fulfillment of which the transformation of a legal situation depends, be strictly identified with some person’s decision to bring about that transformation.28 In a case such as the vagrant’s, this would suffice to conclude that his perpetration of the offense leading to his incarceration would not be coupled with the exercise of a legal power. For although the vagrant may contribute to him being incarcerated by perpetrating the offense, it is not the decision to produce that very transformation of his legal situation that would count as a condition for him being punished.29

  • 30 MacCormick 2008: 98. See also MacCormick 1972: 62-63, 73-78.
  • 31 Thoroughly on this, MacCormick 1972: 75-76; MacCormick 2008: 100-101. In opposition hereto, see Raz (...)
  • 32 See MacCormick 2008: 97-98.
  • 33 MacCormick 2008: 97.

17 In contrast, an indirect recourse to the attitudes of the potential holder of a legal power appears in the definitions proposed both by MacCormick and Raz. According to MacCormick, “[p]ower is conferred by a rule when the rule contains a condition that is satisfied only by an act performed with the (actual or imputed) intention of invoking the rule”.30 The first feature of this definition that merits highlighting is that it finds support in the notion of power-conferring rules,31 which justifies the labeling of MacCormick’s appeal to the attitudes of the agent who appears to be exercising a legal power as (only) indirect. But a further aspect also stands out and concerns the fact that, for an act to be interpretable as the exercise of a power, it would be necessary that it be performed with the intention of invoking the power-conferring rule.32 Yet MacCormick sees himself forced to introduce a disjunctive characterization of the required intention as either “actual” or (merely) “imputed”, in order to avoid the conclusion that, in cases in which someone behaves in a way that would justify the ascription of the intention of invoking the power-conferring rule to her, though without her having formed such an intention, one should rather negate the exercise of the power in question.33 The concession is significant because a (merely) “imputed” intention is, according to MacCormick, an intention that need not join the performance of the act identified as the exercise of a power.

  • 34 MacCormick 2008: 99.

18 With regard to what is of immediate concern here, it is clear that MacCormick can hold that in the case already considered, the vagrant who perpetrates the offense would not be exercising a legal power leading to his incarceration. Indeed, the condition for the transformation of his legal situation would not be identified with the performance of an action (actually or at least “constructively”) accompanied by an intention to invoke the rule that connects the offense’s perpetration with the punitive consequence.34

19 In turn, Raz’s proposal also indirectly appeals, although in a different sense, to the attitudes of the potential holder of a legal power. In his own words:

  • 35 Raz 1972: 81. In reference to the (broader) concept of normative powers, see Raz 1990: 98-104.

The solution of the difficulty lies not in the intention with which power-exercising acts are performed but in the reasons for which they are recognized as effecting a legal change. An action is the exercise of a legal power only if one of the law’s reasons for acknowledging that it effects a legal change is that it is of a type such that it is reasonable to expect that actions of that type will, if they are recognized to have certain legal consequences, standardly be performed only if the person concerned wants to secure these legal consequences.35

  • 36 Critically hereto Halpin 1996: 143-144; Kurki 2017a: 33.
  • 37 See Raz 1972: 81.

20Raz thinks that this way one would avoid referring to the intentions of the power’s holder, by focusing on the reasons in virtue of which the given legal system would attribute transformative efficacy to its exercise.36 Yet, by demanding that the action type instantiated through the exercise of some power be such that its instances are identified with actions “standardly” performed by someone who wants to bring about the legal change in question, Raz makes the recognition of a power conceptually dependent on an attitude that would characteristically accompany its exercise. And this leads him to conclude that, in the case of the vagrant who perpetrates an offense so that he is incarcerated for it, the vagrant would not be exercising a legal power. The action type in the instantiation of which the thus perpetrated offense consists, would not be an action type in reference to which one may reasonably expect that the actions that instantiate it be standardly performed with the purpose of bringing about the agent’s own incarceration.37

2.3 Legal powers as positions instituted by constitutive rules

21The problem faced by the previously considered versions of an attitudinalistic strategy consists in that they all directly or indirectly confer relevance to a factor the locus of which are the intentional states of the agent who may exercise a legal power. This threatens the categorial homogeneity of the class of legal positions capable of being organized through a conceptual framework like the one presented by Hohfeld. For none of the further seven classes of “fundamental” legal positions needs to be defined in terms that make any reference to some attitudinal factor.

  • 38 Kurki 2017a: 43-44.

22 The difficulty is conditioned, at least to some extent, by the tenor of the “approximate explanation” of the concept of a legal power offered by Hohfeld. The goal aimed at by those who endorse an attitudinalistic strategy consists in qualifying or amending the requirement that the circumstances capable of bringing about the transformation of the respective legal situation be under the volitional control of the potential holder of a legal power. A definition of the concept of power thus oriented, however, faces the risk of turning it into a theoretically trivial concept. This becomes patent with Kurki’s suggestion that someone who commits suicide would also be exercising a legal power, since suicide “is a volitional act that effects numerous changes in legal relations”.38

23 To consider the point more thoroughly, we may ask about the consequences that follow from taking Hohfeld’s “approximate explanation” verbatim, as Kurki does, with regard to a case such as this:

While holding a knife, A approaches V and is ready to attack her. In order to prevent the imminent and potentially lethal aggression, V hits A’s head with a heavy object, knocking her out.

  • 39 Hereto, see Mañalich 2014: 504-516. Note that the generation of the respective tolerance duty for A (...)
  • 40 On this, see Mañalich 2014: 495-503.

24The case is paradigmatic of constellations of self-defense. To begin with, it should be uncontested that A’s imminent attack against V brings about a change in their respective legal positions: while V becomes the holder of a Hohfeldian “privilege” (or “liberty”) to defensively harm A (insofar as this is rationally necessary as a way of neutralizing the attack), grounded upon the permissive norm of self-defense, A is placed under a duty to tolerate —i.e., of not preventing— V’s defensive reaction, grounded by whatever prohibitive norm A would break if she did not refrain from preventing V from defending herself (which would in turn depend on the harm that would result for V if she were prevented from displaying her defensive maneuver).39 Secondly, it should also be clear that V’s privilege and A’s duty are not correlative, since only a “no-right” can be correlative to a Hohfeldian privilege, whereas the correlative position to a duty must be a right. This is consistent with the fact that V’s privilege and A’s duty are positions grounded by different norms.40

  • 41 As Kramer 1998: 58-59 has argued, the Hohfeldian axiom of the correlativity of rights and duties ne (...)

25 Yet Kurki would also have to claim that, by imminently attacking V, A would be exercising a Hohfeldian power. On the one hand, A’s attack generates both V’s privilege to defensively harm A, as well as A’s duty to tolerate V’s defensive reaction.41 On the other hand, A’s aggression against V appears to be under the former’s volitional control. According to Hohfeld’s “approximate explanation”, the conjunction of these two propositions should suffice for regarding A to be bringing about, through the exercise of a legal power, a change in both hers and V’s legal positions. But this is a problematic conclusion in that the norms that ground each of those two legal positions are clearly regulative rules, whereas power-conferring rules are constitutive rules.

  • 42 MacCormick 1972: 76-77.
  • 43 See specially Raz 1972: 87-92.
  • 44 MacCormick 2008: 100-108. On the theoretical need for the articulation of pragmatically sensitive c (...)

26 With regard to the permissive norm of self-defense, which grounds the privilege that V holds against A qua aggressor, Kurki could perhaps argue that that very norm would at the same time have the nature of a power-conferring rule. That one discrete legal norm could simultaneously impose duties and grant powers was a claim made early on by MacCormick, according to whom one and the same norm may be taken as a rule of one or the other kind, depending on the agent from whose point of view it is contemplated.42 The objections raised against this early position of his,43 however, led MacCormick to recognize the need of preventing the distinction between rules of the one and the other kind from collapsing. For this distinction would be indispensable for the individuation of the elements of any institutionalized legal system.44

  • 45 Arriagada 2018a: 17-22.

27 Thus, taking Hohfeld’s “approximate explanation” literally seems to be incompatible with holding onto the distinction between regulative rules, which ground duties and rights as well as privileges, and no-rights, and constitutive rules, some of which institute powers and liabilities as well as disabilities and immunities. If that explanation was to be taken literally, then every time a first order Hohfeldian position —such as a duty or a privilege— is grounded by a regulative rule structured as a conditional norm, and to the extent that the fulfillment of that norm’s application condition is placed under some agent’s volitional control, we could not but conclude that that very rule would at the same time be a power-conferring one. Yet, this conclusion needs to be avoided if one notices, as Arriagada does, that Hohfeld’s conceptual framework presupposes the distinction between regulative and constitutive rules precisely because the relation of logical independence between the classes of first and second order positions depends on it.45

28 It is therefore likely no coincidence that when analyzing a hypothetical sufficiently similar to the self-defense case previously considered, Hohfeld describes the transformation of the legal positions involved without mentioning the possible exercise of a legal power by the attacker:

  • 46 Hohfeld 1919: 94.

If, e.g., R threatens bodily harm to X, R’s right that X shall not strike him becomes thereby extinguished, and a no-right in R substituted; or, correlatively, in such contingency, X’s duty to R ceases, and X acquires a privilege of self-defense against R.46

  • 47 Kurki 2017a: 44.

29Notwithstanding that Hohfeld seems to be regarding self-defense as a category of private law, it is crucial that he does not come close to suggesting that the changes in R’s and X’s legal positions could result from the exercise of a power held by R qua aggressor. Hence, it is not enough to concede, as Kurki does, that “Hohfeld might […] have been envisaging a more limited conception [of what a legal power is] than his account encompasses”.47 Since that approach presupposes the very distinction that becomes threatened through a literalist reading of Hohfeld’s “approximate explanation”, it seems more cautious to regard that explanation as overinclusive, and as such in need of a restrictive correction aimed at preserving the internal coherence of the Hohfeldian conceptual framework.

2.4 Power-conferring rules as specific kind of constitutive rules

  • 48 Hart 2012: 27-42, 79-81. Further hereto Hart 1982: 210-219.

30The most promissing path for resisting the collapse of the distinction between regulative rules, capable of grounding duties or privileges, and constitutive rules, a subset of which is made up by rules that institute powers or disabilities, was already anticipated. It involves endorsing a non-reductionist strategy for the individuation of rules, such as the strategy outlined by Hart when articulating his insight into the “union of primary and secondary rules” as “the key to the science of jurisprudence”.48

  • 49 See Hart 2012: 80-81, 94.

31 Hart clearly suggests that the distinction between primary rules of obligation, on the one hand, and secondary rules of recognition, change and adjudication, on the other, ought to be taken as a qualification of the more basic distinction between duty-imposing and power-conferring rules, respectively.49 Yet, this does not imply that each rule that belongs to one of the three subclasses of secondary rules must have, when considered in isolation, the status of a power-conferring rule. Rather, as Hart makes explicit by depicting those three subclasses as “remedies” for the three defects that every system of obligation rules not yet institutionalized would exhibit —namely: the uncertainty concerning which rules belong to it, its static nature, and the diffuse social pressure exerted to achieve conformity with them—, the rules of recognition, change, and adjudication cannot be simplistically reduced to just one category of rules:

  • 50 Hart 2012: 94.

The remedy for each of these three main defects in this simplest form of social structure consists in supplementing the primary rules of obligation with secondary rules which are rules of a different kind. […] Though the remedies consist in the introduction of rules which are certainly different from each other, as well as from the primary rules of obligation which they supplement, they have important features in common and are connected in various ways. […] They specify the ways in which the primary rules may be conclusively ascertained, introduced, eliminated, varied, and the fact of their violation conclusively determined.50

  • 51 It should be noted that, in contrast to an extended trend in the exegesis of his The Concept of Law(...)

32From this passage we can extract the following three claims: (1) secondary rules are “of a different kind” than primary rules; (2) the three subclasses of secondary rules are “certainly different form each other”; and (3) secondary rules have in common, at least, that they specify action possibilities related to the identification and the production or suppression of primary rules, as well as to the ascertainment of their eventual transgression.51

  • 52 Hart 2012: 28.

33 This is consistent with the fact that, in his preliminary characterization of rules conferring private or public powers, Hart highlights the functional diversity of the rules thus labeled, which may therefore be organized into different kinds. Hence, in the case of rules that confer powers to “private individuals” for entering into a contract or otherwise generating or altering some legal relation, we would at least have to differentiate rules about capacity or “minimal personal qualification” for their exercise, rules that define the “manner and form” in which the given power is to be exercised, and rules that delimitate the variety, and eventually the maximal or minimal temporal extension, of “the structure of rights and duties” that may be thus generated.52

  • 53 Hart 2012: 29.
  • 54 See Hart 2012: 30-31.

34 When it comes to rules that institute judicial powers, a similar differentiation becomes pertinent: rules specifying the “subject-matter” and the scope of the given judicial power; rules that indicate how judges are to be appointed, the qualifications they must meet, as well as the corresponding ranks associated to their appointment; and rules that fix canons for the exercise of such powers and determine the procedure according to which a trial should take place, with the validity of the respective judicial decision depending upon their satisfaction.53 The diversity of the classes of rules involved would be even greater in the case of rules that confer legislative powers.54

35 It is now possible to state, with somewhat more precision, the following two thesis: the characterization of some set of rules as power-conferring rules needs to be understood as a global characterization, in the sense of aiming at the function jointly performed by the thus characterized rules, without this implying that each of the rules belonging to the corresponding set, taken as a discrete entity, ought to be characterized as power-conferring; and (2) the property individually exhibited by each of the rules that belong to a set of power-conferring rules consists in it being a constitutive rule.

  • 55 Hart 2012: 29.
  • 56 Hart 2012: 31.
  • 57 Hart 2012: 76 (italics in the original). On this, see Marmor 2011: 45-48.

36 Hart provides us with several tokens of textual support for this last proposition. For instance, when noting that it would be perfectly conceivable that the rules that confer judicial powers be supplemented by rules that obligate judges to behave in certain ways, he observes that the latter rules, capable of imposing legal duties upon judges, are “additional” to the former, which reflects the fact that the rules that confer legal powers “define the conditions and limits under which the court’s decision shall be valid”.55 On the other hand, and with regard to rules that institute legislative powers, Hart suggests that the satisfaction of the conditions fixed by them may be assimilated to a “move” in a game such as chess, in the sense that from it follow “consequences definable in terms of the rules”.56 More significantly, in the context of his celebrated critique of Austin’s model of “sovereign and subject”, Hart argues that the rules that fix the conditions for an election taking place, or for the enactment of a law, are “constitutive of the sovereign”.57

3 Penal sanction norms as constitutive rules

3.1 Penal sanction norms as rules of adjudication

  • 58 It is worth noting that, in the Postscript to his 1961 monography, Hart identifies the provision of (...)

37We can now focus our attention on the specific subset of secondary rules that Hart calls “rules of adjudication”. These can be characterized as constitutive rules that specify the conditions, the form, and the consequences of the institutionalized application, and enforcement of the rules that make up the corresponding legal system.58 Although the issue tends to be sidestepped in the interpretation of Hart’s writing, he explicitly asserts the membership of penal sanction norms to that very subsystem of secondary rules. Indeed, Hart says:

  • 59 Hart 2012: 97-98 (italics added).

Most systems have, after some delay, seen the advantages of further centralization of social pressure; and have partially prohibited the use of physical punishments or violent self help by private individuals. Instead they have supplemented the primary rules of obligation by further secondary rules, specifying or at least limiting the penalties for violation, and have conferred upon judges, where they have ascertained the fact of violation, the exclusive power to direct the application of penalties by other officials. These secondary rules provide the centralized official ‘sanctions’ of the system.59

  • 60 Supra, sec. 1.

38This passage is decisive for sustaining the characterization of penal sanction norms as constitutive rules, and more precisely, as secondary rules of adjudication.60

  • 61 See Alchourrón & Bulygin 1971: 34-43, who explicitly note that their definition only covers conditi (...)
  • 62 Vilajosana 2010: 24.
  • 63 Searle 2010: 96-97. Note that, as suggested by Ludwig 2020: 183 and 203, the constitutive nature of (...)
  • 64 See Searle 1995: 50, who offers the following example of a possible penal sanction norm qua constit (...)

39 To see why, it may be appropriate to provide a more precise definition of the concept of a constitutive rule, such us the one offered by Vilajosana. If, in accordance with the terminology introduced by A&B, it is assumed that a (conditional) regulative rule consists in the correlation of a generic case with a normative solution,61 then by contrast we can say that a constitutive rule consists in the correlation of “a generic case with another generic case”.62 This fits the canonical formulation that, according to Searle, would be distinctive of a constitutive rule, corresponding to a sentence of the kind “X counts as Y in (the context) C”.63 Making use of the conceptualization proposed by Vilajosana, we may translate that formulation into the following: given the context C, the generic case X counts as the generic case Y. Now we can see how this approach puts us in the position to categorize penal sanction norms as constitutive rules.64

  • 65 The lack of specification of some context here is due to the fact that, as a general strategy for t (...)

40To state the point more sharply, let us take the norm stated in art. 391 No. 2 of the Chilean Penal Code (“ChPC”), which criminalizes so-called “simple homicide” and may be restated as follows: “in any given context, someone who kills another human being counts as punishable with imprisonment from 10 years and one day to 15 years”.65 According to this formulation, the sanction norm expressed in art. 391 No. 2 of the ChPC correlates the generic case consisting in killing another human being with the generic case consisting in being punishable with imprisonment from 10 years and one day to 15 years. But what property defines the latter generic case? The answer is: the (dispositional) property of being liable to being punished with imprisonment from 10 years and one day to 15 years.

  • 66 On this, see Mañalich 2019: 420-422.
  • 67 The proposition that penal sanction norms are supplemented —among others— both by imputation rules (...)

41It is crucial to note that a sanction norm such as the one stated in art. 391 No. 2 of the ChPC is not individually sufficient for determining the jointly sufficient conditions upon which someone may be (legally) in position to suffer the imposition of the corresponding sanction, nor for fixing the concrete punishment that may be imposed upon a person declared guilty of (a particular instance of) homicide. On the first of these two issues, for the satisfaction of the description “the killing of another human being” to lead to someone being exposed to the imposition of a penal sanction, a particular instance of that description’s satisfaction must be attributable to the person in question according to the criteria fixed by the respective imputation rules.66 On the second issue, it is clear that a legal consequence specified as “imprisonment from 10 years and one day to 15 years” is nothing more than a punitive frame, on the basis of which the competent judicial organ will have to individuate the sanction to be imposed. For this, the judge or court will normally have to apply a set of criteria established by grading and sentencing rules. Thus, both imputation rules and grading and sentencing rules functionally supplement the relevant sanction norm qua constitutive rule.67

3.2 Penal sanction norms as rules that institute liabilities (and correlative powers)

  • 68 On the categorization of “punishability” as a kind of liability correlative to a punitive power, se (...)
  • 69 Hart 2012: 97.
  • 70 Así Binding 1922: 17, 28.

42On the conception presented, the applicability of a penal sanction norm consists in that someone whose (imputable) behavior has fulfilled that norm’s antecedent condition comes to occupy a determinate institutional position, which in turn consists in being exposed to the imposition of a penal sanction of a certain kind. In Hohfeld’s terminology, that position corresponds to a liability, the correlative position of which must be a power.68 This understanding optimally fits Hart’s characterization of sanction norms as secondary rules of adjudication, i.e., as rules “specifying or at least limiting the penalties for violation” of the rules of obligation that they reinforce.69 This means, more precisely, that sanction norms institutionalize the centralized display of pressure for conformity to duty-imposing rules. Individually considered, each penal sanction norm is nothing but a constitutive rule that, by correlating the realization of its antecedent condition with the imposition of a penal sanction, institutes —in Binding’s words— a “punitive equivalence”.70

43 Further clarification of this thesis crucially requires settling two points, namely: (1) the relation existing between penal sanction norms and other secondary rules of adjudication that also have an impact on the grounding of punitive powers; and (2) the relation existing between penal sanction norms and a specific subset of rules of obligation, the function of which consists in regulating the exercise of the powers instituted by rules of adjudication in general.

  • 71 Hart 2012: 97-98.
  • 72 Supra, sec. 2.4.
  • 73 See Binding 1922: 15.
  • 74 See Altenhain 2002: 309-310, though holding onto Binding’s terminology, according to which the rele (...)

44On the first issue, one should recall Hart’s observation that the specification of the sanctions that may be imposed on those who break some punitively reinforced rule of obligation will normally be joined by the attribution of “the exclusive power to direct the application of penalties by other officials” to judges.71 We can grasp the sense of the thus suggested differentiation if we understand it as an implication of the proposition that although secondary rules of adjudication can globally be taken as power-conferring rules, it does not follow that each of the rules thus classified, when considered individually, may be characterized as a power-conferring rule.72 Hence, a judicial power the exercise of which may lead to the imposition of punishment upon the basis of some penal sanction norm cannot be identified, strictly speaking, with the punitive power instituted by that sanction norm.73 Rather, such a judicial power consists in a “decision power” through the exercise of which the respective punitive power may be exercised, insofar as the thus obtained decision results in the defendant’s conviction.74

  • 75 See Ferrajoli 2001: 539.
  • 76 This makes apparent the convenience of not treating “power” and “competence” as synonymous, as Arri (...)

45Such a (contingent) attachment of the exercise of the power to impose a penal sanction to the exercise of a decision power conferred to a judicial organ is usually regarded as the organic dimension of the principle of “penal judicial reserve”.75 That such a judicial power may be exercised without a punitive power also being exercised, however, becomes clear with the possibility that the decision adopted by the competent judge or court be one of acquittal. And let it be noted that the institutional position occupied by the judge or judges in question will be fixed by further rules of adjudication à la Hart; different from the ones that confer punitive powers, and the function of which is to institute some group of officials as competent for producing decisions through which penal sanction norms, among others, may eventually be applied.76

  • 77 For this notion of the incorrect application of a legal norm (in the sense of the application of a (...)

46It is nevertheless worth mentioning that for a person to acquire the institutional status of a convicted person, it is not necessary that she have actually occupied the liability position in which, under the relevant sanction norm, someone who (imputably) realizes its antecedent condition puts herself. The former and latter statuses are logically independent precisely because it is possible that a penal sanction may be (validly) imposed on someone who has not perpetrated the corresponding offense. In this last case, the relevant sanction norm will have been incorrectly applied,77i.e., incorrectly identified as a premise for the decision to convict and to sentence the defendant.

  • 78 Within a constitutional Rechtstaat, this will normally involve the suppression of the fundamental r (...)

47The fact that being exposed to being punished is equivalent to being exposed to suffering the transformation of one or more deontic positions incidental to the infliction of the sanction imposed through the corresponding sentence, shows that the position occupied by an agent whose behavior realizes the antecedent condition of a penal sanction norm corresponds to a liability in Hohfeld’s terms. Indeed, it is a defining feature of a judicial decision convicting a person of some offense conviction that it results in the imposition of a duty on the convicted to either do something or to tolerate something done to him. Under a regime of public penal law, that duty will be correlative to a right (in the sense of “claim”) of the state, the satisfaction of which will correspond to the infliction of the sanction imposed, for instance, the duty to transfer a certain amount of money, by means of which a criminal fine may be paid, or the duty to tolerate incarceration during some period of time, by means of which a prison sentence may be served.78

48Thus understood, a penal sanction norm specifies the content of the decision through which a competent judicial organ may impose some instance of punishment on a person, which will then be actualized by further officials in charge of its infliction. It is the power to impose the corresponding sanction what, stricto sensu, ought to be identified with the punitive power exercised against the person (declared to be) guilty of having perpetrated a criminal offense.

  • 79 Véase supra, sec. 2.3.
  • 80 Binding 1922: 15-16. Following his path, Altenhain 2002: 307-308.
  • 81 Hart 1982: 219 (italics in the original).

49This leads us directly to the second of the two problems previously stated: in what sense may one say that a judicial organ applies a penal sanction norm when it passes the sentence that imposes a certain instance of punishment on the defendant? The constitutive nature of such norm qua secondary rule of adjudication excludes the possibility of asserting that it would simultaneously function as a regulative rule for the judge or judges called to apply that sanction norm.79 As was already anticipated, Binding clearly noticed this point.80 His approach to the problem may be reformulated, without distortion, as follows: every “penal law” presents itself as a rule that provides a parameter for the corresponding judicial decision, yet without imposing any duty. This description comes close enough to the one offered by Hart in reference to power-conferring rules: “They are more like instructions how to bring about certain results than mandatory impositions of duty”.81

50Whether the competent judge (or each member of the competent collective judicial body) is subject to a duty to adjust her decision to the parameter provided by the relevant sanction norm qua (constitutive) rule of adjudication —i.e., qua “decision rule”—, will depend on whether, as a contingent matter of fact, a rule imposing such a duty belongs to the corresponding legal system. What Hart more generally predicates of rules of adjudication applies to penal sanction norms:

  • 82 Hart 2012: 97.

Like the other secondary rules these are on a different level from the primary rules: though they may be reinforced by further rules imposing duties on judges to adjudicate, they do not impose duties but confer judicial powers.82

  • 83 See Binding 1905: 551-555. For a description of the offense’s structure according to German law, se (...)

51These “further rules imposing duties on judges to adjudicate” are precisely the regulative rules capable of obligating those who are competent for the application of penal sanction norms to adjust the content of their authoritative decisions to those norms. And when a regulative rule of this kind is, in turn, itself backed by a penal sanction norm, a judge who is guilty of its transgression may become punishable as the perpetrator of an specific offense of “bending the law”.83

3.3 Penal sanction norms as rules that institute powers “to commit crimes”?

52We have thus completed the argument in favor of the characterization of penal sanction norms as constitutive rules that institute liabilities-to-punishments and correlative punitive powers. It now seems appropriate to go back to the question that motivated the inquiry into the nature of power-conferring rules in the first place: does it make sense to claim that someone who perpetrates an offense would thus exercise a legal power to bring about her own liability-to-punishment and the correlative punitive power?

  • 84 Supra, sec. 2.3.

53 Qua conditionally structured constitutive rules, penal sanction norms correlate some antecedent condition (that corresponds to the offense’s abstract description) with an abstractly determined legal consequence. The fact that the realization of the norm’s antecedent condition is necessary, yet not sufficient, for the generation of the respective liability-to-punishment and its correlative punitive power does not imply, however, that the agent whose behavior satisfies the antecedent condition would be eo ipso exercising a legal power. Otherwise, and as was already argued, we would have to commit ourselves to the conclusion that every conditionally structured legal norm, the antecedent condition of which may be satisfied as the result of some agent’s “voluntary” behavior, would count as a power-conferring rule. This would lead to the collapse of the very distinction between regulative and constitutive rules, which lies beneath the differentiation of legal positions consisting in duties, rights, “privileges”, and no-rights, on the one hand, and legal positions consisting in powers, liabilities, immunities, and disabilities, on the other.84

  • 85 Very clearly in this direction, see Orunesu & Rodríguez 2019: 19-21, who following Kelsen define “i (...)
  • 86 Its characterization as a genuine legal power should appear less extravagant if we notice that, tra (...)

54 This last remark is by no means incompatible with the acknowledgment that there is a sense in which one may indeed speak of a “power to act criminally”, taken as a status attributable to a person. To grasp that sense, it is essential to insist that the realization of the sanction norm’s antecedent condition is not sufficient for the generation of a liability-to-punishment. For this to be the case, it is equally necessary that the realization of the norm’s antecedent condition be imputable to the person that may thus become liable to punishment. It follows from this that “imputability”, understood as a “capacity for felonious agency”, is a legal power instituted by the imputation rules that supplement the respective set of sanction norms.85 In this regard, the incidence of sanction norms reduces to specifying the forms of behavior in reference to which that capacity may be exercised.86

  • 87 As was very clearly stated by Binding 1922: 4: “Thus far from breaking the penal law under which he (...)

55 As should be clear, the fact that by imputably realizing the antecedent condition of a penal sanction norm, the perpetrator exercises a legal power (identified with her capacity for delinquent agency) does not affect the circumstance that, by realizing that antecedent condition, she also infringes a duty, namely, the duty imposed by the regulative rule punitively reinforced by the corresponding sanction norm. That the exercise of a legal power conferred by the relevant set of imputation rules may be coextensive with the infringement of a legal duty, grounded by the regulative rule for the transgression of which a person may be punished under the respective sanction norm, should not give rise to any perplexity, once we acknowledge the relation of logical independency that holds between the rules of the former and latter kind.87

4 The principle of penal legality as closure rule

4.1 The requirement of categorial homogeneity

56Having clarified the status of penal sanction norms as constitutive rules that institute liabilities-to-punishment and correlative punitive powers, we can now go back to the initially stated problem, which consists in fixing the exact sense in which the principle of penal legality may be regarded as a residual closure rule (“RCR”).

  • 88 Arriagada 2021a: 60 (italics suppressed).
  • 89 See Arriagada 2021a: 74.

57 When drawing our attention to the rather exiguous theoretical interest put into the question concerning the possible closure of systems of “competence norms”, Arriagada claims that the prevailing answer would be that, in general, those systems have the property of being closed. This rests on the fact that among the elements of such systems we would normally find an RCR, the formulation of which could read as follows: “all those actions by normative authorities that have not been explicitly permitted by a superior normative authority are prohibited”.88 Arriagada denounces the inadequacy of such a rule as a possible RCR for any given system of competence norms precisely because a rule formulated as such could not possibly close a system of competence norms. An RCR rule formulated this way would have the character of a regulative rule, since it would impose the deontic qualification of “prohibited” on every “normative action” not qualified as permitted by some other rule belonging to the same system. Whereas a rule capable of closing a system of competence norms would itself need to the exhibit the character of a competence norm.89

  • 90 Supra, sec. 1.
  • 91 Supra, sec. 3.1.
  • 92 An implication of this is that for a system of rules to be closed by an RCR, it is necessary that a (...)

58 This last proposition can be understood as the implication of a more general thesis, according to which the rules of the system closed by an RCR and the RCR itself must be categorially homogeneous: a rule capable of closing any given system of rules needs to be of the same kind as the rules that make up the system closed by that rule. The theoretical grounding for the thus stated thesis can be obtained from generalizing A&B’s finding concerning the relational character of the concept of a closed normative system,90 one dimension of which points to the respective universe of solutions. If we now recall that, according to the definition already considered, constitutive rules correlate generic cases with other generic cases,91 it should be clear that for a rule to function as an RCR for a system of constitutive rules, it must specify a generic case that thus becomes “residually correlated” with every generic case itself not correlated with some other generic case by any of the constitutive rules that belong to the system under consideration.92

  • 93 Of which the rules that Arriagada labels as “competence norms” are a subset, at least if one acknow (...)

59 If we now ask what follows from the requirement of categorial homogeneity for systems of constitutive rules that institute powers and liabilities,93 as penal sanction norms do, the answer is this: an RCR capable of closing such a system of rules, in relation to some universe of generic cases, would need to consist either in a rule that institutes a power and a correlative liability, or in a rule that institutes a disability and a correlative immunity.

4.2 Necessary completeness of “Hohfeldian” normative systems?

  • 94 Biasetti 2015: passim.

60A greater awareness of the problem of the closure of what Biasetti calls “Hohfeldian normative systems” has recently become recognizable.94 That problem is generated by the distinctive recursivity allegedly exhibited by normative positions of power, liability, disability, or immunity.

  • 95 See, among many, Duarte d’Almeida 2016: 558-559; Arriagada 2018a: 12.
  • 96 Biasetti 2015: 954.

61In his analysis of the problem, Biasetti notes that the relation that holds between each of the dyads of positions that Hohfeld characterizes as “opposites” —duty and “privilege”, right and no-right, power and disability, liability and immunity— is, in logical terms, a relation of contradictory opposition,95 which would be equivalent to calling it a relation of exclusive disjunction.96 Hence, if we think about the relation existing between two persons, P1 and P2, with respect to a possible action φ by means of which P1 could bring about a transformation of some legal position of P2, we can assert that either P1 has a power to φ in relation to P2 or P1 has a disability to φ in relation to P2: tertium non datur.

  • 97 Biasetti 2015: 955.

62From this premise, Biasetti reaches the conclusion that Hohfeldian normative systems would stand out not only for the possibility of being incoherent —i.e., of containing antinomies—, but also for being necessarily “complete”. Since the opposition between any two normative positions is to be understood in the sense of a contradictory opposition, it would follow, going back to our example, that the given legal system would necessarily contain either a rule instituting a power for P1 to φ in relation to P2 or a rule instituting a disability for P1 to φ in relation to P2. And once it is assumed that every Hohfeldian normative system would also be complete at the level of its power-conferring or disability-imposing rules, it ought to be concluded that such a system will be —in Biasetti’s terms— “necessarily unclosed”, which means, that the set of normative positions instituted by that system would be necessarily infinite.97

  • 98 Alchourrón & Bulygin 1971: 116-119.

63To prevent a possible terminological confusion, it should be noted that the sense in which Biasetti makes use of the adjective “unclosed”, in reference to some normative system, is not equivalent to the sense in which, following A&B, one should use that very expression. According to A&B, to say that a normative system is “closed” is equivalent to saying that it is “necessarily complete”,98 whereas for Biasetti the “unclosed” character of a normative system consists in the fact that the positions instituted by the rules that integrate it are infinite, which would be compatible with —because implied by—that system being necessarily complete.

64Now, Biasetti’s explanation of the purported necessary infinity of the set of positions instituted by every Hohfeldian normative system lies in the acknowledgment of the recursivity of the generation of the corresponding second order positions. For every pair of first order normative positions —for instance, a duty and the correlative right—, the following would be the case: either at least one individual has a power to alter that duty, and thus a power to alter the correlative right, so that the subject of the duty and the holder of the right stand in corresponding positions of liability correlative to the powers in question; or at least one individual has a disability to alter that duty, and thus a disability to alter the correlative right, so that the subject of the duty and the holder of the right stand in corresponding positions of immunity correlative to the disabilities in question. Hence, if at the first level we found two positions —a duty and a right—, in the second level we will find four positions —either two powers and two liabilities, or two disabilities and two immunities—, with the geometric progression extending infinitely in virtue of the completeness predicated of the corresponding system of norms.

  • 99 Kurki 2017b: 144.
  • 100 Arriagada 2021a: 74-75 (italics suppressed).

65Assuming the thesis of the ad infinitum recursivity of the positions grounded by the rules that make up any Hohfeldian normative system, which would lead to describing such a system as a “recursive tree”, Kurki observes the following: “Except perhaps in very odd normative systems, in any Hohfeldian recursive tree there is always a limited number of powers and an infinite number of disabilities”.99 This observation is, however, nothing but a generalization of a normative proposition by means of which we could describe the content of the RCR that, according to Arriagada, would tend to be postulated by theorists and scholars of public law. Such an RCR could be stated as follows: “every individual whose normative status is not explicitly made liable by the constituent authority to the power of a constituted authority, possesses an immunity vis-à-vis the constituted authority”.100

  • 101 See Alchourrón & Bulygin 1971: 119-127. Hereto Arriagada 2020: 10-13.
  • 102 Hereto, see Arriagada 2020: 22-27; see also Guastini 2016: 86-87. The fact that the existence of le (...)

66According to Arriagada, who adapts A&B’s analysis of the so-called “principle of prohibition”,101 whether such an RCR belongs to the corresponding normative system is a contingent matter of fact. For from the fact that a person occupies, for instance, a position of immunity “in the weak sense” it does not follow that she occupies a position of immunity “in the strong sense”. For the former to be the case, it suffices that no rule conferring a power correlative to the liability in the inexistence of which that (weak) immunity consists belongs to the normative system under consideration. For the latter to be the case, it is necessary that a rule instituting the corresponding (strong) immunity in fact belongs to the system in question.102

  • 103 Arriagada rightly observes that Hohfeld’s presentation of the eight classes of normative positions (...)
  • 104 See Alchourrón & Bulygin 1971: 121. On the significance of that distinction for the clarification o (...)

67This is enough to deactivate the premise leading Biasetti to assert the necessary completeness of every Hohfeldian normative system. From the fact that, à la Hohfeld, the relation between two opposite legal positions is one of contradictory opposition, it follows that if P1 does not have a power to φ in relation to P2, then P1 has a disability to φ in relation to P2. Yet this does not imply that a rule instituting P1’s disability and P2’s correlative immunity belongs to the respective normative system.103 Biasetti overlooks the need for distinguishing between norms (or rules) on the one hand, and normative propositions, on the other.104 With this distinction, we can now reach the conclusion that, eventually, P2 has an immunity towards φ in relation to P1, without this implying that such a relation be grounded on a rule instituting a position of immunity. This will be precisely the case when the immunity and its correlative disability have the character of “weak” (and not of “strong”) positions, which, by definition, will imply that the normative system under consideration be incomplete in relation to the corresponding universe of cases. This suffices for concluding that that same system will not count as a closed one.

4.3 The principle of penal legality as constitutive meta-rule

68We can now put an end to the present inquiry by putting together the results already obtained. The clarification of the status of penal sanction norms as constitutive rules acquires utmost significance once we acknowledge that for a rule to serve as an RCR for a system of penal sanction norms, that rule needs to be categorially homogenous with the rules that make up the system in question. Hence, if penal sanction norms have the character of constitutive rules that institute liabilities-to-punishment and correlative punitive powers, then as an RCR for a system of penal sanction norms only a constitutive rule will come into consideration. More precisely: a constitutive rule that “residually” institutes either a liability-to-punishment and thus a correlative punitive power, or an immunity-to-punishment and thus a correlative punitive disability, in reference to every form of behavior that does not correspond to the antecedent condition of at least one of the sanction norms that belong to the system thus closed.

69Likewise, it has also been shown that at the level of second order “Hohfeldian” positions it is critical to distinguish between “strong” and “weak” positions, i.e., positions grounded by rules that belong to the normative system under consideration, on the one hand, and positions resulting from that same system not containing rules that ground the corresponding contradictorily opposite positions, on the other. This leads to rejecting the claim of the necessary completeness of “Hohfeldian” normative systems. For that claim results from not making the distinction between norms and normative propositions.

  • 105 Note that the negation of liability-to-punishment expressed in the antecedent clause of the thus st (...)

70 This enables us to state the following: in contrast to the claim made by A&B, the principle of penal legality cannot be understood as an RCR for a system of regulative rules, but rather as an RCR for a system of constitutive rules. As such, it determines that every immunity-to-punishment in the weak sense be redefined as an immunity-to-punishment in the strong sense, as well as that every punitive disability in the weak sense be redefined as a punitive disability in the strong sense. Its canonic formulation could thus be: “anyone who does not incur in behavior that under a legislatively formulated sanction norm counts as the antecedent of the imposition of punishment is exempt from punishment”. Or more briefly: “anyone who is not punishable under a legislatively formulated sanction norm is unpunishable”.105

  • 106 For a very sharp suggestion of the (theoretical) intelligibility of an RCR of the inverse content, (...)

71 A corollary of the analysis thus offered consists in the reminder that the membership of an RCR to the relevant legal system is a contingent matter of fact, which is also true of the principle of penal legality.106 That jurists tend to regard it as a genuine axiom of the operation of contemporary systems of penal law, reflects the cultural success of its postulation, which we owe to the reformers of the Enlightenment.

Acknowledgment.— I would like to thank María Beatriz Arriagada, as well as two anonymous referees, for several observations that hopefully contributed to a clearer presentation of the arguments developed in the present article.

Top of page

Bibliography

Alchourrón, C. & Bulygin, E. (1971). Normative Systems. Springer.

Alonso, J. P. (2006). Interpretación de las normas y derecho penal. Didot.

Altenhain, K. (2002). Das Anschlußdelikt. Mohr Siebeck.

Arriagada, M.B. (2018a). Fundamentality, interdefinability, and circularity. Revus. Journal for constitutional theory and philosophy of law, 35, 7-44.

Arriagada, M.B. (2018b). Fundamentalidad, interdefinibilidad y circularidad. Revus. Journal for constitutional theory and philosophy of law, 35, 45-84. (Spanish version of Arriagada 2018a.)

Arriagada, M.B. (2020). Inmunidades fuertes y débiles: el imperio contraataca. Revista de Derecho (Valdivia), XXXIII(1), 9-29.

Arriagada, M.B. (2021a). The closure of the systems of legal norms of competence. Revus. Journal for constitutional theory and philosophy of law, 43, 59-84.

Arriagada, M.B. (2021b). La clausura de los sistemas de normas jurídicas de competencia. Revus. Journal for constitutional theory and philosophy of law, 43, 85-112. (Spanish version of Arriagada 2021a.)

Beling, E. (1906). Die Lehre vom Verbrechen. J.C.B. Mohr (Paul Siebeck).

Beling, E. (1908). Die Vergeltungsidee und ihre Bedeutung für das Strafrecht. Wilhelm Engelmann.

Biasetti, P. (2015). Hohfeldian Normative Systems. Philosophia, 43, 951-959.

Binding, K. (1872). Die Normen und ihre Übertretung, vol. I (1st ed.). Wilhelm Engelmann.

Binding, K. (1905). Lehrbuch des Gemeinen Deutschen Strafrechts. Besonderer Teil, vol. II. Wilhelm Engelmann.

Binding, K. (1915). Strafrechtliche und strafprozesualle Abhandlungen, vol. II. Duncker & Humblot.

Binding, K. (1922). Die Normen und ihre Übertretung, tomo I (4th ed.). Felix Meiner.

Duarte d’Almeida, L. (2016). Fundamental Legal Concepts: The Hohfeldian Framework. Philosophy Compass, 11(10), 554-569.

Dubber, M. & Hörnle. T. (2014). Criminal Law. A Comparative Approach. Oxford University Press.

Ferrajoli, L. (2001). Derecho y razón. Teoría del garantismo penal (5th ed.). Trotta.

Guastini, R. (2016). La sintaxis del derecho. Marcial Pons.

Halpin, A. (1996). The Concept of a Legal Power. Oxford Journal of Legal Studies, 16, 129-152.

Hart, H.L.A. (1982). Essays on Bentham. Clarendon Press.

Hart, H.L.A. (2012). The Concept of Law (3rd ed.). Oxford University Press.

Hohfeld, W.N. (1919). Fundamental Legal Conceptions. Yale University Press.

Kramer, M. (1998). Rights Without Trimmings. In: Kramer, M., Simmonds, N.E. & Steiner, H. A Debate over Rights (pp. 7-111). Oxford University Press.

Kramer, M. (2019). Power-Conferring Laws and the Rule of Recognition. Jerusalem Review of Legal Studies, 19, 87-99.

Kurki, V. (2017a). Legal Competence and Legal Power. In: McBride, M. (ed.). New Essays on the Nature of Rights (pp. 31-47). Hart Publishing.

Kurki, V. (2017b). Hohfeldian Infinities: Why Not to Worry. Res Publica, 23, 137-146.

Lindahl, L. & Reidhav, D. (2017). Legal Power: The Basic Definition. Ratio Iuris, 30, 158-185.

Ludwig, K. (2020). The Social Construction of Legal Norms. In: Garcia-Godinez, M., Mellin, R. & Touomela, R. (eds.). Social Ontology, Normativity and Law. De Gruyter.

MacCormick, N. (1972). Voluntary Obligations and Normative Powers. Proceedings of the Aristotelian Society. Supplementary Volumes, 46, 59-78.

MacCormick, N. (2008). H.L.A. Hart (2nd ed.). Stanford University Press.

Mañalich, J. P. (2012). Reglas primarias de obligación. Las “reglas del derecho penal” en El Concepto de Derecho de H.L.A. Hart. Zeitschrift für Internationale Strafrechtsdogmatik, 11, 571-585.

Mañalich, J. P. (2014). Normas permisivas y deberes de tolerancia. Revista Chilena de Derecho, 41, 473-522.

Mañalich, J. P. (2017). Responsabilidad, autoridad y democracia. Una exploración crítica de la filosofía del derecho penal de Antony Duff. Discusiones, XVII, 167-219.

Mañalich, J. P. (2019). The Grammar of Imputation. Jahrbuch für Recht und Ethik, 27, 411-428.

Mañalich, J. P. (2019). Justicia, procedimiento y acción de revisión. Ius et Praxis, 26(1), 28-56.

Markovich, R. (2019). Rights and Punishment: The Hohfeldian Theory’s Applicability and Morals in Understanding Criminal Law. IFCoLog Journal of Logics and their Applications. Available at: https://orbilu.uni.lu/bitstream/10993/41661/1/Rights%20and%20Punishment_IFCoLog.pdf.

Marmor, A. (2011). Philosophy of Law. Princeton University Press.

Molina, F. (2001). Antijuridicidad penal y sistema del delito. Bosch.

Navarro, P. (2014). Hechos y normas aplicables. Comentarios en torno a una propuesta de Ricardo Caracciolo. Isonomía, 40, 147-159.

Navarro, P. & Moreso, J.J. (1997). Applicability and Effectiveness of Legal Norms. Law and Philosophy, 26, 201-219.

Nino, C.S. (1980). Los límites de la responsabilidad penal. Astrea.

Nino, C.S. (2008). Fundamentos de Derecho Penal. Gedisa.

Orunesu, C. & Rodríguez, J.L. (2018). Una revisión de la teoría de los conceptos jurídicos básicos. Revus. Journal for constitutional theory and philosophy of law, 36, 81-110. (Spanish version of Orunesu & Rodríguez 2019.)

Orunesu, C. & Rodríguez, J.L. (2019). A revision of the theory of fundamental legal concepts. Revus. Journal for constitutional theory and philosophy of law, 39, 7-35.

Rawls, J. (1999). A Theory of Justice (2nd ed.). The Belknap Press.

Raz, J. (1972). Voluntary Obligations and Normative Powers. Proceedings of the Aristotelian Society, Supplementary Volumes, 46, 79-102.

Raz, J. (1980). The Concept of a Legal System (2nd ed.). Clarendon Press.

Raz, J. (1990). Practical Reason and Norms (2nd ed.). Princeton University Press.

Searle, J. (1995). The Construction of Social Reality. The Free Press.

Searle, J. (2010). Making the Social World. Oxford University Press.

Vilajosana, J. (2010). El derecho en acción. Marcial Pons.

Von Wright, G. H. (1963). Norm and Action. Routledge & Kegan Paul.

Top of page

Notes

1 Arriagada 2021a: 60 (italics suppressed).

2 See Alchourrón & Bulygin 1971: 20-21, 116-117.

3 Alchourrón & Bulygin 1971: 142.

4 See Alchourrón & Bulygin 1971: 142-143.

5 Alchourrón & Bulygin 1971: 143.

6 For an early rebuttal of the thesis postulating the existence of such thing as a “penal unlawfulness”, see Beling 1906: 129-131.

7 Hart 2012: 97-98; hereto Mañalich (2012: 580-582). On this issue, see infra, sec. 3.

8 Raz 1980: 24, 155-156.

9 Véase Alonso 2006: 171-172.

10 Alonso 2006: 172.

11 Alchourrón & Bulygin 1971: 141-142.

12 Alchourrón & Bulygin 1971: 148, 151-153.

13 This characterization is widely extended in criminal law scholarship; for a sufficiently representative of this stance, see Molina 2001: 614-632, who asserts that “[t]he citizen’s sanction norm is, at the same time, the judge’s conduct norm”.

14 Binding 1872: 7-14.

15 This is fully compatible with the possibility of a judge being obligated by some regulative rule to apply a penal sanction norm, taking the latter as the major premise for deciding a case which that judge is called to adjudicate, whereas the transgression of such regulative rule may, contingently, be constitutive of an offense of “law bending”; hereto infra sec. 3.2.

16 Binding 1922: 19-25.

17 Hereto, see Mañalich 2012: 581-582.

18 Kurki 2017a: passim.

19 Hohfeld 1919: 50-51

20 Hohfeld 1919: 50.

21 Kurki 2017a: 31. It should be noted that in the context of the sentence “a homeless person could commit a crime in order to spend the night in jail”, “could” is being used in the sense of “has the practical possibility of”, and not in the sense of “has a legal power to”.

22 Kurki 2017a: 44.

23 This reformulation suggest that to characterize the vagrant’s situation as one in which he brings about that he “spend the night in jail” as a consequence of perpetrating an offenses involves a massive oversimplification. For someone who has committed an offense to be put in jail immediately after having committed it, it will normally be necessary that rather complex conditions be satisfied, whereas these (substantive and procedural) conditions will be fixed by rules different from the sanction norm that criminalizes the corresponding form of behavior.

24 Kurki 2017a: 45.

25 Halpin 1996: 141.

26 Halpin 1996: 145-147.

27 See Raz 1990: 65.

28 Halpin 1996: 145. A somewhat proximate, but by no means identical analysis of the concept of a legal concept, leading to a definition that includes the requirement of a “manifestation of intention”, is offered by

Lindahl & Reidhav 2017: 169-181.

29 Halpin 1996: 145.

30 MacCormick 2008: 98. See also MacCormick 1972: 62-63, 73-78.

31 Thoroughly on this, MacCormick 1972: 75-76; MacCormick 2008: 100-101. In opposition hereto, see Raz 1972: 83-85. On this problem, infra, sec. 2.3.

32 See MacCormick 2008: 97-98.

33 MacCormick 2008: 97.

34 MacCormick 2008: 99.

35 Raz 1972: 81. In reference to the (broader) concept of normative powers, see Raz 1990: 98-104.

36 Critically hereto Halpin 1996: 143-144; Kurki 2017a: 33.

37 See Raz 1972: 81.

38 Kurki 2017a: 43-44.

39 Hereto, see Mañalich 2014: 504-516. Note that the generation of the respective tolerance duty for A follows from the fact that, V’s defensive action becoming permitted, an eventual defensive action by A against V’s defensive action could not be permitted qua self-defense, since the antecedent condition for such a permission consists in the unlawful —i.e., unpermitted— aggression attributable to some other agent.

40 On this, see Mañalich 2014: 495-503.

41 As Kramer 1998: 58-59 has argued, the Hohfeldian axiom of the correlativity of rights and duties need not be set aside when it comes to “public duties”, among which we find the duties the infringement of which constitutes a criminal offense. Such duties would be correlative to equally public rights, i.e., rights held by “a collectivity (the state, the nation, the community)”; along the same lines, see Markovich 2019: 10-12. For a defense of the thesis that the right correlative to a penally reinforced duty has the state as its holder, see Mañalich 2014: 500-503. Thus, and in contrast to the suggestion made by Duarte d’Almeida 2016: 560, as well as by Arriagada 2018a: 25-26, the existence of such duties does not affect the scope of Hohfeld’s correlativity axiom.

42 MacCormick 1972: 76-77.

43 See specially Raz 1972: 87-92.

44 MacCormick 2008: 100-108. On the theoretical need for the articulation of pragmatically sensitive criteria for the individuation of rules (qua discrete entities) that be sensitive, see Raz 1980: 140-147.

45 Arriagada 2018a: 17-22.

46 Hohfeld 1919: 94.

47 Kurki 2017a: 44.

48 Hart 2012: 27-42, 79-81. Further hereto Hart 1982: 210-219.

49 See Hart 2012: 80-81, 94.

50 Hart 2012: 94.

51 It should be noted that, in contrast to an extended trend in the exegesis of his The Concept of Law, it does not follow from what Hart says that the rules of recognition and the rules of change may not equally refer to secondary rules. Scholars usually overlook the fact that Hart is thus emphasizing how, in the transition towards the emergence of a properly institutionalized legal system, a set of obligation rules may become supplemented by a further set of rules that at least refer to the rules of the former set.

52 Hart 2012: 28.

53 Hart 2012: 29.

54 See Hart 2012: 30-31.

55 Hart 2012: 29.

56 Hart 2012: 31.

57 Hart 2012: 76 (italics in the original). On this, see Marmor 2011: 45-48.

58 It is worth noting that, in the Postscript to his 1961 monography, Hart identifies the provision of “secondary rules for the identification, change, and enforcement of its standards” as one of the two distinctive features of every institutionalized legal system; see Hart 2012: 249. Among other considerations, one reason for taking a legal system’s rules of adjudication as rules concerning law’s application and imposition is that, besides the rules that institute judicial powers in a narrow sense, one should also include here the rules —that remain rather overlooked in Hart’s analysis— that institute powers of administrative application of legal rules; see Kramer 2019: 93.

59 Hart 2012: 97-98 (italics added).

60 Supra, sec. 1.

61 See Alchourrón & Bulygin 1971: 34-43, who explicitly note that their definition only covers conditionally structured norms, and not categorical norms, the structure of which would correspond to the structure of normative solutions as such, which in turn consists in the deontic qualification of a certain action type. (This is consistent with the fact, that in terms of A&B, a normative solution may be characterized as the deontic modalization of some element of the corresponding universe of actions.)

62 Vilajosana 2010: 24.

63 Searle 2010: 96-97. Note that, as suggested by Ludwig 2020: 183 and 203, the constitutive nature of a rule is always relative to the description specifying that which is constituted by it. From this it follows that, in terms of the formulation favored by Vilajosana, the given constitutive rule is constitutive of the generic case Y, but not of the generic case X. As should be clear, this is compatible with the possibility of the generic case Y corresponding to a fragment of (institutional) reality constituted by one or more rules different from the rule that constitutes Y.

64 See Searle 1995: 50, who offers the following example of a possible penal sanction norm qua constitutive rule: “Killing, under certain circumstances, counts as murder, and murder counts as a crime punishable by death or imprisonment” (italics added).

65 The lack of specification of some context here is due to the fact that, as a general strategy for the individuation of legal norms, it is preferable to assume that the criteria that determine whether a norm is applicable as the result of the solution of an antinomy or a normative redundancy, or in virtue of a factor impacting its temporal or spatial validity, function as criteria of external applicability. On the distinction between the conditions of internal and external applicability of legal norms, see Navarro & Moreso 1997: 202-210.

66 On this, see Mañalich 2019: 420-422.

67 The proposition that penal sanction norms are supplemented —among others— both by imputation rules and by grading and sanction rules, rests upon a more general approach to the individuation of the rules that make up a system of penal law. This problem cannot be further considered here.

68 On the categorization of “punishability” as a kind of liability correlative to a punitive power, see Nino 1980: 232-234; Nino 2008: 123-126; also Mañalich 2017: 172-173.

69 Hart 2012: 97.

70 Así Binding 1922: 17, 28.

71 Hart 2012: 97-98.

72 Supra, sec. 2.4.

73 See Binding 1922: 15.

74 See Altenhain 2002: 309-310, though holding onto Binding’s terminology, according to which the relevant sanction norm would ground a (public) “right” of the state to modify the regal relation that exists between the state and the offender. For an early denunciation of the inadequacy of this terminology, see Beling 1908: 40-41, whose argument clearly anticipates the conceptualization of what Binding named a “punitive right” rather as a power.

75 See Ferrajoli 2001: 539.

76 This makes apparent the convenience of not treating “power” and “competence” as synonymous, as Arriagada does; see Arriagada 2021a: passim. On this, see further infra, footnote 93.

77 For this notion of the incorrect application of a legal norm (in the sense of the application of a norm to a case to which that norm is not applicable), see Navarro 2014: 154-157. This possibility follows from the fact that, according to Rawls (1999: 73-78), a “criminal trial” corresponds to a context of imperfect procedural justice; hereto, see Mañalich 2020: 31-36. The acknowledgment of the possibility of an incorrect, yet valid application of a penal sanction norm explains that Binding spoke of the “constitutive” (as opposed to “declarative”) character of erroneous convictions; see Binding 1915: 284-290.

78 Within a constitutional Rechtstaat, this will normally involve the suppression of the fundamental right (in the sense of “claim”) that, had she not been convicted, the person in question would have against the state imposing upon her the form of treatment by which the corresponding punishment is actualized. The suppression of that right will be correlated with the transformation of the respective “negative” duty of the state into a “privilege” for bringing that very person under such form of treatment.

79 Véase supra, sec. 2.3.

80 Binding 1922: 15-16. Following his path, Altenhain 2002: 307-308.

81 Hart 1982: 219 (italics in the original).

82 Hart 2012: 97.

83 See Binding 1905: 551-555. For a description of the offense’s structure according to German law, see Dubber & Hörnle 2014: 633, 638-640. On the relevance of the not infrequent lack of punitive reinforcement of the rules that regulate the behavior of judicial organs with regard to the application of legal norms, as an example of the possibility of there being legal norms the transgression of which is not attached to the imposition of sanctions, see Raz 1980: 152-154.

84 Supra, sec. 2.3.

85 Very clearly in this direction, see Orunesu & Rodríguez 2019: 19-21, who following Kelsen define “imputability as “the power (not authorization) to perform an illegal act”.

86 Its characterization as a genuine legal power should appear less extravagant if we notice that, traditionally, the debate about the capacity for felonious agency of so-called “legal persons” (i.e., of corporations and the like) has been framed in terms of whether the maxim societas delinquere non potest ought to be replaced by the contradictorily opposed maxim.

87 As was very clearly stated by Binding 1922: 4: “Thus far from breaking the penal law under which he is judged, to be punishable he must have acted, rather, according to the first part of such penal law”.

88 Arriagada 2021a: 60 (italics suppressed).

89 See Arriagada 2021a: 74.

90 Supra, sec. 1.

91 Supra, sec. 3.1.

92 An implication of this is that for a system of rules to be closed by an RCR, it is necessary that among the criteria for identifying the rules that belong to that system we encounter the requirement that the system be exclusively integrated by rules of a same kind, i.e., by categorially homogeneous rules.

93 Of which the rules that Arriagada labels as “competence norms” are a subset, at least if one acknowledges the convenience of reserving the concept of competence for identifying a subclass of powers; hereto, see Kurki 2017a: 43-46, who —as was already shown— problematically endorses Hohfeld’s “approximate explanation” of the notion of a legal power. In contrast to Arriagada 2020: 21-22, according to whom “a competence [would] always [be] positive, for it is the possibility of validly altering a normative status”, one should note that, as a conceptual matter, the notion of a “negative” power is perfectly intelligible, as is shown by the example of a power of veto, the structure of which may be analyzed by means of the framework provided by Von Wright’s theory of action; see Von Wright 1963: 56-69. In these terms, a change operated on a legal relation can result either from the performance of an action that produces (or brings about) the change in question or from the forbearance of an action that would prevent that very change. Someone who holds a veto power is in position to prevent the transformation of a legal relation and may thus contribute to that transformation taking place by not exercising the power in question. It would be odd, however, to claim that someone who has not exercised a veto power that she could exercise has thus exercised a “positive” power.

94 Biasetti 2015: passim.

95 See, among many, Duarte d’Almeida 2016: 558-559; Arriagada 2018a: 12.

96 Biasetti 2015: 954.

97 Biasetti 2015: 955.

98 Alchourrón & Bulygin 1971: 116-119.

99 Kurki 2017b: 144.

100 Arriagada 2021a: 74-75 (italics suppressed).

101 See Alchourrón & Bulygin 1971: 119-127. Hereto Arriagada 2020: 10-13.

102 Hereto, see Arriagada 2020: 22-27; see also Guastini 2016: 86-87. The fact that the existence of legal rules instituting positions of immunities is perfectly conceivable —and by no means infrequent— suffices for dissipating the possible objection that, pragmatically considered, it would be senseless to postulate the existence of positions of disability “in the strong sense”. According to the Hohfeldian scheme assumed here, a rule that institutes an immunity correlatively institutes a disability. We find a sufficiently clear example of this in rules of tax invariability: the position correlative to the immunity in which the person favored by the invariability rule is placed is, indeed, a disability that will typically affect one or more organs or official bodies that would otherwise have a normative power (conferred by one or more constitutive rules different from the one instituting that specific disability) to alter the corresponding tax regime. On this, see Hart 1982: 190-192, in reference to immunities resultant from constitutional limitations of legislative powers; further Hart 2012: 66-71. In situations of this kind, one will have to acknowledge the existence of an antinomy between constitutive rules that institute (opposite) positions of power and disability —and, correlatively, of liability and immunity—, such that, if the characterization of that very situation as one in which the organs or official bodies in question occupy a position of disability in the strong sense were correct, this would rest upon a solution of the antinomy in favor of the rule instituting the disability (and the correlative immunity); on the possibility of such inconsistency of systems of constitutive rules, see Arriagada 2020: 22-23. This should suffice for acknowledging that the set-up of a disability in the strong sense through the promulgation of a corresponding constitutive rule ought not to be confused with either the derogation of a rule that institutes an “opposite” power, which would rather generate a “weak” disability, or with the promulgation of a prohibition of exercising such a normative power.

103 Arriagada rightly observes that Hohfeld’s presentation of the eight classes of normative positions he analyzed seems to be neutral with respect to the distinction between “weak” and “strong” positions, since his framework is displayed at the language level of normative propositions; hereto, see Arriagada 2021a: 65.

104 See Alchourrón & Bulygin 1971: 121. On the significance of that distinction for the clarification of the structure of “fundamental legal positions”, see Orunesu & Rodríguez 2019: 22-25.

105 Note that the negation of liability-to-punishment expressed in the antecedent clause of the thus stated rule (“… is not punishable”) has the status of an external negation, whereas the negation expressed in the consequent (“… is unpunishable”) has the status of internal negation.

106 For a very sharp suggestion of the (theoretical) intelligibility of an RCR of the inverse content, see Beling 1906: 47-48.

Top of page

References

Electronic reference

Juan Pablo Mañalich R., “The closure of systems of penal sanction norms as systems of constitutive rules”Revus [Online], 44 | 2021, Online since 08 September 2021, connection on 20 October 2021. URL: http://journals.openedition.org/revus/7439; DOI: https://doi.org/10.4000/revus.7439

Top of page

About the author

Juan Pablo Mañalich R.

Full Professor of Penal Law, Law Faculty, Universidad de Chile (Chile).

Address: Av. Santa María 076 – 7520405 Providencia – Santiago – Chile

E-mail: jpmanalich (at) derecho.uchile.cl

Top of page

Copyright

All rights reserved

Top of page
  • OpenEdition Journals
Search OpenEdition Search

You will be redirected to OpenEdition Search