1.Introduction

Decisions of legal officials, promises, contracts, wills and taxes are considered as some examples of legal power exercises.¹ Yet despite the relatively widespread agreement among contemporary legal theorists and legal practitioners on what can be described as legal power exercises - numerous questions arise when one asks about the very concept of legal power. As stated by Alexy, this is a clear clue about the existence of a philosophical problem. A concept can be used in everyday life; however, when one begins to analyze it, it can become quite doubtful (^{Alexy 2008, p. 43}).

Two prevailing theoretical approaches have dealt with the concept of legal power. On the one hand, it has been argued in well-known theories that legal power is reducible to deontic modalities. Those theories which affirm that legal power-conferring rules are reducible to norms of conduct that are either a kind of duty, or a kind of permission belong to this tendency. Austin’s,² Kelsen’s,³ Ross’s,⁴ and Spaak’s,⁵ as well as Bentham’s,⁶ von Wright’s,⁷ Kanger and Kanger’s,⁸ and Lindahl’s⁹ theories can be considered as examples of this reductionist tendency. On the other hand, those theories that elucidate legal power as a normative social, pragmatic or institutional phenomenon may be identified as examples of an approach that has shown that legal power is not reducible to deontic modalities. In this vein, Hohfeld’s,¹⁰ Hart’s,¹¹ Raz’s¹² theories, as well as the position of Ruiter¹³ and MacCormick¹⁴ can be identified as examples of this trend.

Eugenio Bulygin has defended two different approaches in this field. In his early works developed with Carlos Eduardo Alchourrón, he defended a reductionist theory, according to which legal power- conferring rules are a special kind of permissive rules that perform a constitutive function at the same time (^{Alchourrón and Bulygin 1971, p. 151}). Notwithstanding, by rejecting the normativist position about the nature of law,¹⁵ these authors have defended a non-reductivist position in order to define legal power-conferring rules. According to this second approach, such rules must be considered constitutive rules -or better said, definitions (^{Alchourrón and Bulygin 1991, p. 463})- in an exclusive manner, since they do not prescribe any behavior as obligatory, prohibited, or permitted (1991, pp. 462-463). This account has also been defended by Eugenio Bulygin in his article from 1991 titled “On Norms of Competence”.

The goal of this paper is to explore two objections raised against Eugenio Bulygin’s second approach to the definition of the nature of legal power-conferring rules. According to the first objection, inasmuch as Bulygin’s second account maintains that these rules define a variety of heterogeneous phenomena, such an account is vague about what is defined by legal power-conferring rules qua constitutive rules. This problem of vagueness is rooted in the lack of a suitable definition of legal power. In this paper, I shall be arguing for the reduction of the complexity of the definientia through positing a comprehensive concept of legal power which allows one to explain it as a species of competence. According to the second objection, this non-reductive approach cannot explain the normativity of this kind of rules. In other words, the non-reductive approach cannot explain the authoritative feature of legal power. In this paper, I shall be arguing for a dual function of legal power-conferring rules as constitutive rules of legal practice. According to this position, determining the deontic and the definitional function of this kind of rules is a matter of perspective. Unlike the mere activities, the practices are associated with internal normative evaluation standards which are coded by means of rules. These rules govern how to run a practice correctly (^{Rescorla 2009, p. 101}). In this way, from the theoretical point of view, legal power-conferring rules play a definitional role, due to the fact that they are necessary conditions for the identification of the practice. In contrast, from the practical point of view, legal power- conferring rules carry out a deontic function, due to the fact that they are necessary conditions for achieving the aims associated with the practice.

Our inquiry will be divided into three parts. The first part will deal with Bulygin’s theory of legal power-conferring rules. The second part will be devoted to the analyses of the problem of vagueness and the exposition of a method for the reduction of the complexity of definientia. The third part addresses the problem of the normativity of the legal power-conferring rules qua definitions.

2. Eugenio Bulygin’s Theory of Legal Power-Conferring Rules

In his theory, von Wright coined the name normative action in order to refer to a peculiar mode of action by virtue of which an authority “orders, permits, or prohibits certain subjects to do certain things on certain occasions” (1963, p. 75). This peculiar mode of action is ruled by a special kind of norms which have been termed “norms of higher order” (p. 190). According to von Wright, subjects of norms of higher order are themselves authorities of norms of lower order (p. 192). The norms belonging to a higher order must be distinguished from those whose contents are other than normative actions, i.e. norms of the first order (p. 191). Whilst among norms of the first order commands and prohibitions hold the most prominent position, in norms of higher order permissions are of peculiar interest and importance (p. 192). In this vein, according to von Wright’s account, legal power-conferring rules are norms of higher order which permit a certain authority to issue norms of a certain content (p. 192). According to Alchourrón and Bulygin, the main interest of jurists lies in discovering the solutions that law provides for certain cases (1971, p. 68). This activity amounts to “the determination of the normative consequences of a set of legal sentences for a certain problem or topic” (p. 68). This determination involves “the construction of an axiomatic system, adopting these [legal] sentences as axioms” (p. 68). In order to achieve that construction, jurists must define the criteria that will be used to identify those sentences (p. 68).¹⁶ These identification criteria define the notion of a valid legal sentence (p. 73) by setting out which requirements a legal statement must meet to be valid (p. 72). For the reason that these identification criteria are conceptual rules -that is, they merely regulate the use of the concept of validity- (p. 73) they must be distinguished from those norms which establish that an action is obligatory, forbidden or permitted -that is, from norms of conduct (^{Alchourrón and Bulygin 1991, pp. 462-463}). Alchourrón and Bulygin abandon a certain legal positivist tradition, which identifies the validity criteria of a legal sys- tem with its norms of competence¹⁷ -rectius: legal power-conferring rules (1971, p. 73, n. 2)-¹⁸ by maintaining that the identification criteria must not be confused with such norms, since, in contrast to those criteria, they must be considered as a special class of norms of conduct which “[permit] the creation of new norms” (1971, p. 73).

For Alchourrón and Bulygin, the centralization of the function of jurisdiction in the hands of specialized officials is a plausible reason for the existence of two normative systems which run parallel but that are interdependent (1971, p. 148). Along with a primary system -or better said, a subject system- which is made up by norms which regulate the behavior of the subjects of law, there is a secondary system -or better said, a judge system- which is comprised of norms which regulate the behavior of judges and other jurisdictional organs (p. 148). These latter norms may be either norms of competence or norms which establish obligations and prohibitions for the judges (p. 151). Based on von Wright’s theory, Alchourrón and Bulygin affirm that norms of competence are “norms of conduct for the judges, if we regard them as permissive norms establishing the permission to perform certain acts in certain circumstances” (p. 151). Notwithstanding, they argue that these permissive norms are a special kind since, at the same time, they are constitutive of judicial authority. In fact, “nobody becomes a judge except on account of a norm of competence and to the degree and extent that this norm specifies” (p. 151). These norms “are part of the meaning of the term ‘judge’ ” (p. 151).

Various criticisms have been raised against the position that posits the reduction of legal power-conferring rules to permissive rules.¹⁹ On the one hand, Makinson has sharply pointed out that the difference between these kinds of rules lies in the necessary constitutive feature of the former. Indeed, according to him,

[i]f a person does something that is not permitted, then the action is done, although it may render the person liable to punishment. On the other hand, if a person tries, say, to celebrate a marriage or issue a passport without having the power to do so, then we say that he has not in fact celebrated a marriage or issued a passport (for emphasis: has not issued a valid passport) but has only gone through the motions or given the appearance of doing so. (^{Makinson 1986, p. 411})

On the other hand, based also on the constitutive nature of the legal power-conferring rules, Makinson has recognized a structural difference between legal power and permission that has been less noticed. Although it makes perfectly good sense to say, that a person has permission not to do something, it is very strange to say that someone has the legal power not to perform it (1986, p. 412).

But even more importantly, Bulygin himself has accepted that “[t]he idea that competence norms are essentially permissive is difficult to reconcile with legal practice” (^{Bulygin 1991, p. 205}).²⁰ In fact, against von Wright’s position, Bulygin has maintained that “[i]f competence norms are permissions, then a prohibition to make use of the power conferred by such norms would generate a contradiction, but this is not how such situations are in fact treated by jurists” (1991, p. 206). Indeed, legal practice shows that a main argument, which can be raised against the reduction of legal power-conferring rules to permissive rules, reveals that while there is a normative conflict between a permissive rule and a prohibitive rule referring to the same action, a normative conflict between a legal power-conferring rule and a prohibitive rule to exercise such legal power cannot take place, given that the constitutive nature of legal power-conferring rules entails that every rule that prohibits one from exercising the legal power presupposes a rule that ascribes it.²¹

As is well known, one of the fundamental tenets of Alchourrón and Bulygin’s theory is the rejection of the normativist position about the nature of law.²² Based on this tenet, Alchourrón and Bulygin abandoned the reductive thesis that they had defended until their article from 1983 titled “Definitions and Norms” (1991 (1983)). According to their second approach, legal power-conferring rules are definitions, or better said, conceptual rules -that is, rules that “merely regulate the use of a concept (or a term)” (^{Alchourrón and Bulygin 1971, p. 73}). This account has also been defended by Eugenio Bulygin in his article from 1991 titled “On Norms of Competence”.

In accordance with this second approach, legal power-conferring rules cannot be reducible to norms of conduct, due to the fact that they only “define patterns of behavior, which do not exist outside these rules” (^{Bulygin 1991, p. 211}). In other words, legal power- conferring rules must be considered conceptual rules or definitions in an exclusive manner (p. 211), since they do not prescribe any behavior as obligatory, prohibited, or permitted (p. 211). This second position of Bulygin is based on Searle’s account, according to which while regulative rules regulate an activity whose existence is logically independent of the rules, constitutive rules constitute an activity the existence of which is logically dependent on the rules (^{Searle 2009 (1969), p. 34}). In this way, Searle affirms that whereas regulative rules characteristically take the form of, or can be paraphrased as imperatives, constitutive rules can be construed as analytic by means of the form “X counts as Y in context C” (p. 34). As stated by Bulygin, a non-reductive definition of legal power-conferring rules provides a suitable conceptual tool in order to grasp the rather obscure concept of “rechtliches Können” (1991, p. 215).

3. The Problem of Vagueness

Despite its virtues, the non-reductive approach defended by Bulygin is not without problems. Two important criticisms may be raised against it. First, as stated by ^{Ferrer Beltrán (2000, pp. 97-99)}, Bulygin’s account from 1991 (p. 215) is not clear about what is defined by legal power-conferring rules qua constitutive rules. According to Ferrer Beltrán’s objection, Bulygin posited four possible candidates of definientia. Legal power-conferring rules could define not only the form of entering into a contract, or a will in order to make it valid, but also, in a more general way, some kind of legally relevant behaviors. Legal power-conferring rules could define not only the concept of legislator and make legislating possible, but also, in a more general way, what a contract, a promise or a will is. This vagueness shows one of the biggest obstacles that must face anyone who addresses the problem of the definition of legal power-conferring rules: The problem of the complexity of the definientia. This problem is rooted in the lack of a suitable definition of legal power.

I shall be arguing for the reduction of the complexity of the definientia through positing a comprehensive concept of legal power that allows one to explain it as a species of competence.²³ As Ross has pointed out, the normative conditions for the exercise of legal power

usually fall into three groups: (1) those which prescribe what person (or persons) is qualified to perform the act which creates the norm (personal competence); (2) those which prescribe the procedure to be followed (procedural competence); and (3) conditions which prescribe the possible scope of the created norm with regard to its subject, situation, and theme (substantial competence). (^{Ross 2009, p. 130})

Although Ross has stated that these three groups of conditions are necessary for the exercise of legal power, procedural and substantial competence presuppose that the competence has been ascribed at least to a person. In this sense, personal competence has an analytic priority over the other two necessary conditions.²⁴

Based on this analytic priority, competence can be considered as a lexical attributive concept. As stated by Werning based on Millikan, a basic distinction is that between attributive and substance concepts.

Attributive concepts represent features of objects that are volatile in the sense that one and the same object can fall under different attributive concepts at different times: An object may, e.g., change its color, size, or speed, but still continues to exist. [Blue] thus is a paradigmatic attributive concept. Substance concepts, in contrast, are governed by the identity conditions of objects: A mug ceases to exist when it no longer falls under the substance concept [mug], say, because it has been shattered. Substance concepts serve to re-identify things over time in spite of their contingent changes of attributes and so allow us to gather, store and update information in a systematic and enduring way. (^{Werning 2008, p. 70})

The question thus becomes, what is attributed to an agent by means of the concept of competence?²⁵ The ability²⁶ to follow rules is attributed to an agent by means of the concept of competence, that is, the practical possibility²⁷ of behaving as a result of understanding and applying rules (^{Williams 2010, p. 192}) -or in other words, the practical possibility to engage with reasons.²⁸ Since this rule-obeying behavior implies knowledge²⁹ and intention (^{Baker and Hacker 1984, p. 299}), it can be distinguished from that behavior that as a matter of fact conforms to some rule or instantiates a law in a causally determined way (p. 192).³⁰

Notwithstanding, given that knowledge and intention are mental states which are only accessible by the subject who is experiencing them, the actual use of a rule is the only warrant for the fact that it is understood (^{Williams 2007, p. 74}). In this way, the ability to follow a rule has to be understood not as a hidden mental state or process³¹ but as a know-how that has to be displayed in actions (^{Medina 2002, p. 179}).³² Thus, according to Kripke:

Our entire lives depend [ . . . ] on the ‘game’ of attributing to others the mastery of certain concepts or rules, thereby showing that we expect them to behave as we do. This expectation is not infallibly fulfilled. It places a substantive restriction on the behavior of each individual, and is not compatible with just any behavior he may choose. [ . . . ] We can restate this in terms of a device that has been common in philosophy, inversion of a conditional. (^{Kripke 1982, p. 93})

As stated by Riesenfeld based on Wittgenstein’s view of rules, “[a]n inverted conditional retains the causal structure of the original conditional but reverses the order of ‘cause’ and ‘effect’. [ . . . ] ‘We do not say that 12 + 7 = 19 and the like because we all grasp the concept of addition; we say we all grasp the concept of addition because we all say 12 + 7 = 19 and the like’ ” (^{Riesenfeld 2010, p. 46}).

In this way, the answer about the normative nature of the starting point of our practices³³ must be found in the mutual ascription of competence by means of which subjects achieve their own constitu- tive status as normative entities, that is, as agents capable of making commitments and being accountable (^{Villa Rosas 2015, p. 55}). Indeed, from the practical point of view, every single practice requires that each participant should acknowledge with respect to others what he acknowledges of himself, that is, the ability to follow rules (p. 55). And hence, from the theoretical point of view, the definition of the participants -as a necessary condition for every single practice- amounts to the ascription of the ability of following the rules that constitute and regulate such a practice to certain agents. In this vein, every single rule presupposes the ascription of the ability to follow it to an agent -or in other words, every single rule presupposes a rule that constitutes a competent agent.³⁴

Given that a rule is a kind of norm (^{Alexy 1994, pp. 71-104}), that a norm is the meaning of a normative statement, and that a normative statement expresses that something is required, permitted, or prohibited (1994, pp. 39-53), the ascription of the ability to follow a normative content to an agent is a speaker’s existential presupposition associated with the expression of deontic modalities as conventional specific triggers.³⁵ Although premised on a prescriptive conception of norms, by analyzing the relation between “ought” and “can”, Hare has noticed that their relation is

analogous to that Mr. Strawson has claimed to exist between the statement that the King of France is wise, and the statement that there is a King of France. If there is no King of France, then the question whether the King of France is wise does not arise. And so, by saying that the King of France is wise, we give our hearers to understand that we think, at least, that the question arises to which this is one possible answer, and that, accordingly, there is a King of France. And similarly, if we say that somebody ought to do a certain thing, and “ought” has its full (i.e. universally prescriptive) force, then we give our hearers to understand that we think that the question arises to which this is a possible answer, which it would not, unless the person in question were able to do the acts referred to. (^{Hare 1963, pp. 53-54})

At this point, we need to distinguish between the analysis of the context of ascriptions of responsibility -which is closely related to the problem of free will- and the analysis of meaning of normative statements. Whilst the first analysis deals with the relation between the agent and the rule, the second analysis addresses the propositional content of normative statements as units of meaning.³⁶ Notwithstand- ing, these analyses are closely interwoven.

On the one hand, the analysis of the ascription of responsibility takes for granted that there is a normative content that could be followed by every agent who is under the conditions required by such content -Tatbestand- and accordingly, it may be based on the requirement that only such agents who were able to follow this content can be blamed. This argument is closely linked to the Principle of Alternate Possibilities (PAP), according to which “[a]n agent is morally blameworthy for what she has done only if she could have done otherwise” -as is well known, Frankfurt famously argued that this maxim is false (^{Frankfurt 1998 (1969), p. 1}).³⁷ However, as stated by Stern, this argument, when accepted, shows merely that “blame implies can” (^{Stern 2004, p. 46}).

On the other hand, the analysis of the ascription of responsibility requires the analysis of the propositional content of a normative statement as a unit of meaning. Indeed, ascribing responsibility to an agent entails making explicit the general presupposition of the ascription of the ability to follow a normative content in relation to this particular agent -which involves the subsumption of the rule referred to by the normative statement. The analysis of the relation between the normative content and such a presupposition leads to the analysis of the relation between “ought” and “can”. As stated by Gardner, “[t]he thesis that ‘ought’ implies ‘can’ [ . . . ] is often traced back to Kant. The attribution is correct but misleading. Those who invoke [‘ought’ implies ‘can’] today typically argue that, because A lacks the ability to Φ, it cannot be the case that A ought to Φ. What one ought to do cannot be established without first establishing what one can do” (^{Gardner 2013, p. 63}). Indeed, since this interpretation of the maxim “ ‘ought’ implies ‘can’ ” entails positing that “can” is a necessary condition for “ought”,³⁸ it implies a form of a naturalistic fallacy, in the sense, that factual circumstances referring to what the subject “can” determine what the subject “ought”.³⁹ Nevertheless, as affirmed by Gardner, “Kant invoked [‘ought’ implies ‘can’] to argue in the opposite direction. His thought was that, because A ought to Φ, it cannot be the case that A lacks the ability to Φ.” In fact,

in so far as Kant thinks that nothing can prevent the moral law commanding us, he does so because his conception of our agency is such that he holds us to be essentially capable of acting as right requires, not because our capacities as human agents naturalistically conceived put limits on what the moral law can comprise, so that no such gap can arise based on his conception of us qua exemplary agents, and not merely on his conception of “ought implies can”. (^{Stern 2004, p. 57})

In this vein, a Kantian interpretation of our ability to follow rules -i.e. competence- must be based “also nicht nach der empirischen Kenntniß, die wir vom Menschen haben, wie sie sind, sondern nach der rationalen, wie sie der Idee der Menschheit gemäß sein sollen” (<AAVI: 405> (^{Kant 1797, vol. VI, p. 404})). In other words, as stated by Kripke, “what is important here is that the notion of ‘competence’ is itself not a dispositional notion. It is normative, not descriptive” (1982, pp. 30-31, n. 22).

For the sake of clarity, let us imagine two strangers who want to start a game of chess. The one who is playing with the white pieces must ascribe the ability to follow the rules of chess to the one who is playing with the black pieces and vice versa. Moreover, the rules of chess must not be adapted to the actual abilities of the players of this concrete game but, on the contrary, every single rule of chess presupposes the ability that every single player must have in order to follow it. It is precisely the ascription of the ability to follow the rules of chess as a whole that allows a third person, who is looking on these agents taking part in the game, to define them as chess players.

Accordingly, given that the concept of grammar presupposes the concept of competence, this latter concept can also account for the depth of legal practice. In this way, since the multi-layered character of competence extends over multiple levels of abstraction, it may account for the continuity between semantics, pragmatics, practical reason, and law.⁴⁰

In this vein, it is possible at this point to distinguish between legal competence in a broad sense and legal competence in a narrow sense -or legal power. By means of the concept of legal competence in a broad sense is attributed the ability to follow some kind of rules that we called legal rules that is, the practical possibility of behaving as a result of understanding and applying legal rules- to an agent.⁴¹ Legal power is a special case of legal competence; more specifically as held by Ross, legal power is “the legally established ability to create legal norms (or legal effects) through and in accordance with enunciations to this effect” (^{Ross 2009, p. 130}).⁴²

In fact, changing a legal position may⁴³ involve, firstly, a normative status defined by a legal rule ascribing to a person the ability to follow certain rules (rectius: personal competence), and secondly, the fact that this person performed a special kind of act according to the rules that regulate her ascribed ability to follow such rules (rectius: procedural competence) (^{Spaak 1994, p. 10}). In this sense, it is possible to distinguish between legal power-conferring rules and procedural rules of legal power as necessary normative conditions for the exercise of legal power.⁴⁴ Whilst the former rules ascribe to a person the ability and disposition to follow certain rules in order to create legal norms -or legal effects- the latter rules regulate the use of this ability.⁴⁵ This common intuition allows the distinction between the creation and the exercise of legal power. Indeed, as Lindahl suggests, having legal power means having the practical possibility to change legal positions, no matter if this possibility is never exercised (^{Lindahl 1977, pp. 206-210}). Given that if there is not a competent agent, then procedural rules of legal power cannot be followed, and, moreover, that if there is not any rule that ascribes the ability to follow such procedural rules to someone, then there is not a competent agent, therefore every single procedural rule of legal power presupposes a legal power-conferring rule, which constitutes a competent agent in order to follow it.

4. The Lack of Explanation of Normativity

According to the second objection raised against Eugenio Bulygin’s second approach to the definition of the nature of legal power- conferring rules, such an approach cannot explain their normativity.⁴⁶ In other words, the non-reductive approach cannot explain the authoritative feature of legal power. This criticism is related to the problem of the relation between constitutive and regulative rules.⁴⁷ As a matter of fact, although authors like Searle and von Wright have not ruled out the prescriptive function of constitutive rules, others have claimed that these rules do not play any motivational role over the agents, but rather a restricted doxastic role with regard to the observer.⁴⁸ In its stronger form, the objection identifies normativity with prescriptivity and considers that rules cannot be constitutive and regulative at the same time.⁴⁹

As is well known, such a strong version of the objection is defended by ^{Bulygin (1991, pp. 213-214)}. His position rests on three main arguments. First, whilst constitutive rules are analytic inasmuch as they are definitions, regulative rules or norms of conduct are synthetic in order to be able to fulfill their role to guide behavior (1991, p. 213). Second, in contrast to regulative rules, constitutive rules as conceptual rules or definitions always create an impossibility. Finally, whilst we can distinguish the rule prohibiting certain behavior from the provision for penalties to be exacted if the rule is broken, in the case of legal power-conferring rules we cannot logically distinguish between the rule requiring compliance with certain conditions and the nullity (1991, p. 214).

A good example of the problem of the normativity of the legal power-conferring rules qua constitutive rules can be found in the interpretation of one of the most important theories about legal power. I mean especially the groundbreaking work elaborated by ^{Hart (2012(1961))}.

As is well known, Hart argued that law is a set of two kinds of rules (^{2012, p. 116}). While the primary rules’ addressees are required to do or to abstain from certain actions, secondary rules provide their addressees with the power to recognize, create, modify, and eliminate primary rules (2012, pp. 81, 94). According to Hart’s position, the introduction of the secondary rules provides a remedy for the main defects of a simple customary rules system (2012, p. 94). More precisely, the rule of recognition is introduced in the legal system as a remedy against the uncertainty in identifying the primary rules (2012, p. 94). Indeed, according to Hart the rule of recognition exhaustively defines the necessary and sufficient conditions under which a set of rules can be considered legal by a specific community (2012, pp. 116-117). As pointed out by ^{Shapiro (2009)}, despite the clarity of the basic idea of the rule of recognition, it is astonishingly difficult to formulate the doctrine with greater precision.

^{Shapiro (2009)} argues that since the rule of recognition is accepted and practiced -that is, it is a social rule- it does not exist in virtue of any other rule, i.e. it is an ultimate rule. In this vein, if a legal official’s attitude towards accepting the rule of recognition is the only key for identifying legal officials -or in other words, if legal power- conferring rules amount to the mere practice of practical attitude of acceptance of the rule of recognition-⁵⁰ then Hart’s account is unable to explain the normativity of the rule-conferring legal officials’ power. Thus, in contrast to a legal official, although a citizen must have an internal attitude of acceptance towards the primary rules, she does not have to develop this attitude in relation to the rule of recognition. Therefore, even though this citizen must obey the primary rules based on the fact that these rules have been created by legal officials, she cannot find any reason which supports the legal officials’ authority (^{Bernal 2013, p. 173}).

It seems plausible to think that a suitable path for explaining the normativity of legal power-conferring rules qua definitions is based on assuming the distinction between internal and external points of view of rules as components of a normative practice.⁵¹ As is well known, this distinction was used by Hart in order to show that sanction-centered accounts of rules ignore an essential feature of law (^{Shapiro 2006}). Indeed, according to Hart,

Austin among [some theorists] seeing perhaps the general irrelevance of the person’s beliefs, fears, and motives to the question whether he had an obligation to do something, have defined this notion not in terms of these subjective facts, but in terms of the chance or likelihood that the person having the obligation will suffer a punishment or “evil” at the hands of others in the event of disobedience. (^{Hart 2012, p. 83})

As stated by Hart,

[t]he fundamental objection [against this position] is that the predictive interpretation obscures the fact that, where rules exist, deviations from them are not merely grounds for a prediction that hostile reactions will follow or that a court will apply sanctions to those who break them, but are also a reason or justification for such reaction and for applying the sanctions. (^{Hart 2012, p. 84})

In this vein, in contrast to other behavioral regularities such as habits, social rules are characterized by a practical attitude of rule-acceptance (^{Shapiro 2006}). As affirmed by Shapiro, this attitude manifests itself not only through acting according to the dictates of the rule (^{Shapiro 2006, p. 9}; ^{Hart 2012, p. 55}), but also through the use of the rule in order to justify or to assess one’s own or a third party’s action (^{Shapiro 2006}) by means of an evaluative language.⁵² In other words, the internal point of view corresponds to a practical use of the rules that an insider performs when she decides how she will respond to the law (^{Shapiro 2006}).

Notwithstanding, along with the practical use of rules, they can be also used as a scheme of interpretation in order to understand a practice.⁵³ In this sense, as stated by Shapiro, “an attitude might be external by failing to be a practical attitude at all. Someone whose interest in the law is primarily theoretical, who simply wishes to describe how members of a group regard and respond to a set of rules and, perhaps, to make predictions as well, takes the external point of view” (2006, n.p.). In other words, “the theoretical perspective is that of the observer, who is often but not necessarily an outsider, who studies the social behavior of a group living under law” (n.p.).

As is known, Rescorla has alluded to the particular nature of practices in contrast to mere activities.⁵⁴ Indeed, although both are forms of action, unlike mere activities, practices are associated with internal normative evaluation standards which are coded by means of rules. These rules govern how to run a practice correctly (^{Rescorla 2009, p. 101}). In this way, unlike mere activities regarding which it is only possible to differentiate between actions which belong to the activity from those which do not, every practice allows a threefold division of actions (p. 101). According to this division, it is possible to distinguish, besides the actions which belong to the practices, those actions which are correctly performed.⁵⁵ This threefold division involves the special relationship that one sort of rules has with the practice, due to the fact that this sort of rules allows one not only to identify some actions as precedents belonging to the practice, but also to assess the singular actions as correct or incorrect. From the theoretical point of view, this group of rules is a necessary condition for the identification of the practice. From the practical point of view, meeting this group of rules is a necessary condition for achieving the aims associated with the practice. Due to their features, the rules belonging to this group can be called constitutive.

In this vein, it is possible to grasp the dual function of the legal power-conferring rules as constitutive rules of the legal practice.⁵⁶ According to this position, determining the deontic and the definitional function of the legal power-conferring rules is a matter of perspective.⁵⁷ From the theoretical point of view, the legal power-conferring rules play a definitional role, due to the fact that the definition of the participants is a necessary condition for the identification of the legal practice. In contrast, from the practical point of view, the legal power-conferring rules carry out a regulative role, due to the fact that they are necessary conditions for achieving the aims associated with the legal practice. Firstly, although legal power-conferring rules can be expressed as analytic propositions, their definientia can be expressed as synthetic propositions. In this vein, although, from the theoretical point of view, the legal rules of our inquiry are definitions, their definientia must be regarded as regulative norms from the practical point of view -that is, from a participant’s perspective- in the sense that they give members of a community, officials and law-subjects alike, reasons for acting.⁵⁸ Secondly, and even more importantly, on the one hand, legal-power conferring rules are in themselves legal power exercises of an (superior) authority, whose purposive interpretation reveals their necessary connection to dutyimposing norms that impose the aims of the legal system.⁵⁹ On the other hand, they “make it obligatory to act according to the norms of conduct which have been created according to the procedure laid down in” procedural rules of legal power (^{Ross 2009, p. 118}). As is well known, this aspect refers to the question concerning the problem of individuating norms,⁶⁰ and more specifically, the reduction of legal power-conferring rules to duty-imposing rules (^{Ross 2009, p. 118}).⁶¹

However, the discussion about this issue goes beyond the scope of this paper.

5. Concluding Remarks

As expressed above, the goal of this paper was to explore two objections raised against Eugenio Bulygin’s second approach to the definition of the nature of legal power-conferring rules. According to this second approach, such rules must be considered constitutive rules -or better said, definitions- (^{Alchourrón and Bulygin 1991, p. 463}) in an exclusive manner, since they do not prescribe any behavior as obligatory, prohibited, or permitted (pp. 462-463).

According to the first objection, inasmuch as Bulygin’s second account maintains that these rules can define a variety of heterogeneous phenomena, such an account is vague about what is defined by legal power-conferring rules qua constitutive rules. As we have seen, this problem is rooted in the lack of a suitable definition of legal power. Legal power as “the legally established ability to create legal norms (or legal effects) through and in accordance with enunciations to this effect” (^{Ross 2009, p. 130}) is a special case of competence. In the most comprehensive sense of the term, competence is the ability to follow rules. Since rule-following behavior implies knowledge and intention, it can be distinguished from that behavior that as a matter of fact conforms to some rule or instantiates a law in a causally determined way. Notwithstanding, given that knowledge and intention are mental states which are only accessible by the subject who experiencing them, the actual use of a rule is the only warrant for the fact that it is understood. In consequence, our practices require that competence must be ascribed to the agents. In this vein, the ascription of the ability to follow a normative content to an agent is a speaker’s existential presupposition associated with the expression of deontic modalities as conventional specific triggers. Concordantly, every single rule presupposes the ascription of the ability to follow it to an agent -or, in other words, every single rule presupposes a rule that constitutes a competent agent. From the theoretical point of view, the ascription of the ability of following the rules that constitute and regulate a practice to certain agents amounts to their definition as participants of the practice. From the practical point of view, such a definition amounts to the ascription of the ability of following the rules to certain agents in order to justify or to assess one’s own or a third party’s action by means of an evaluative language.

As a kind of competence, legal power must be attributed to an agent. Legal power is attributed to an agent by means of a special kind of rules. We call these rules legal power-conferring rules. According to this analysis, it is possible to distinguish between legal power-conferring rules -that is, legal rules that ascribe the ability to follow certain rules in order to create legal norms or legal effects to an agent -and procedural rules of legal power- that is, rules that regulate the use of this ability -as necessary normative conditions for the exercise of legal power. Based on the analytic priority of the former over the latter ones, our inquiry must be firstly focused only on them.

According to the second objection, the non-reductive approach defended by Bulygin cannot explain the normativity of legal power- conferring rules. In other words, this approach cannot explain the authoritative feature of legal power. As we have seen above, Bulygin has defended this approach based on a strong version of the thesis of the separation between constitutive and regulative rules. He has rested his position on three main arguments, which can be reduced to the problem of analyticity. According to this problem, while constitutive rules are analytic, regulative rules must be synthetic in order to be able to fulfil their role to guide behavior (^{Bulygin 1991, p. 213}). Nevertheless, although legal power-conferring rules can be expressed as analytic propositions, their definientia can be expressed as synthetic propositions. In this vein although, according to their theoretical function, the legal rules of our inquiry must be assumed as definitions, their definientia must be regarded as regulative norms, according to their practical function.

For the sake of clarity, let us take as an example the point of view of an expert in comparative constitutional law who asks who the “President of the United States of America” is, according to the American Constitution. Now compare his perspective with the point of view of the young Obama, who had to meet each of the consti- tutional requirements for being a President of the United States of America, and the point of view of those subjected to its legal power.

For the young Obama -from the ex ante or before subsumption of the general legal power-conferring rule perspective- firstly, meeting each of the requirements of the definition of this expression -President of the United States of America- was a necessary condition for achieving the aim laid down in the general legal power- conferring rule -which, on the one hand, corresponds to the status created by the definition, and on the other hand, in its turn must be harmonious with the aims of the legal system. And secondly, each of these requirements played not only a regulative (motivational) anankastic role over him, but also a regulative (evaluative) role as standard to assess his actions.⁶² Moreover, for those subjected to his legal power -from the ex post or after subsumption of the general legal power-conferring rule perspective- the particular legal power-conferring rule of Obama’s office served as legal basis for making it obligatory to act according to the norms of conduct which were created according to the procedures laid down in the rules which determined the way he ought to exercise his legal power (^{Ross 2009, p. 118}) -rectius: procedural rules of legal power.

As we have seen, although legal power must be understood as a special case of competence -that is, as the ability to follow certain kinds of rules- legal power is created -or better said, constituted- by means of a special kind of rules -rectius: legal power-conferring rules. From the practical point of view, legal power-conferring rules must be interpreted as conditions for achieving the aims associated with the legal practice. Concordantly, they must be interpreted as legal basis of those norms of conduct, which have been created according to the procedure established by the procedural rules of legal power. In this way, as stated by Ross, from the point of view of those subjected to the legal power, legal power rules must be considered as “norms of conduct in indirect formulation” (1958, p. 50; 2009, p. 118).⁶³