I. Introduction

In Judging Positivism (^{Martin 2014}, henceforth JP), Margaret Martin advances an interesting and multilayered challenge against legal positivism. In her view, legal positivism is a theory of the nature of law in which “discerning the content of law is a fact-finding mission and not an evaluative one” (JP, 7), that is closely connected with a “descriptive”, “conceptual”, or “nonnormative” methodological project that disconnects law from politico-moral considerations (JP, vii). Martin’s argument is advanced through an examination of the work of Joseph Raz, focusing on his Practical Reason and Norms (¹⁹⁷⁵, henceforth PRN); The Authority of Law (¹⁹⁷⁹, hereafter AL); The Morality of Freedom (¹⁹⁸⁶, henceforth MF); Ethics in the Public Domain (¹⁹⁹⁴, henceforth EPD); and some aspects of Between Authority and Interpretation (²⁰⁰⁹, henceforth BAL).

The “instability” of the positivist project that sees legal theory as an “independent, nonnormative project”, Martin claims, is “best viewed from the inside of Raz’s complex account” (JP, viii). She argues that the Razian theory has evolved and that these changes have revealed theoretical failures and deep inconsistencies. Razian positivism fails because it cannot properly explain a substantial part of legal practice (i.e., common law) and the role of moral reasoning in adjudication. Moreover, the inconsistencies and contradictions of Raz’s theory, which led him to support some stances of its main rivals (i.e., Dworkinian non-positivism and legal realism), show the implausibility of the positivist separation between law and morality. As a result, the failures of the Razian theory are instructive because they open new paths of inquiry that consider “contestable value-laden assumptions” and are informed by the “the work of lawyers and judges” (JP, viii).

This comment critically examines the conception of legal positivism that informs Martin’s challenge against the substance and method of this intellectual tradition. My central claim is that her characterization of the substantive theory of legal positivism sets aside a more fundamental and explanatory prior dimension, namely, the positivists’ account of the existence of legal systems and legal norms. I also argue that her characterization of the positivist descriptive methodology as a nonnormative project is too demanding and overlooks the different ways contemporary positivists allow for relationships between law and morality, and the pivotal distinction between internal and external inquiries. These clarifications provide resources to begin to address some of Martin’s objections against the Razian project.

Although I do discuss briefly some aspects of her interpretation of the views of contemporary legal positivists, I should note from the outset that this examination is not an exegetical inquiry about the best understanding of Hart’s and Raz’s texts and legacies. Instead, my main goal is to rescue the central commitment of the positivist project (i.e., a theory of law’s existence) which has been neglected or misconstrued by several trends in the contemporary theoretical discussion from which Martin’s argument borrows. Furthermore, part of the controversy arises because Raz did not scrupulously follow some of these prescriptions which resulted in substantial theoretical confusion.

II. Legal Positivism as a Substantive Theory of Law

The expression “theory of law” is used in different senses. Sometimes it is used to refer to a theory of the existence of a legal system, namely an account of the conditions under which a community has a legal system rather than a non-legal institutionalized system, or even a haphazard collection of norms. This inquiry is different from an attempt to provide a theory of the existence of individual legal norms, that is, an account of the conditions under which we can say that a legal norm, or a law, is said to exist. Questions about the existence of systems and laws are different from the family of questions about the system’s content, namely, which standards are part of a certain legal system and not others. These questions are in turn different from questions about the content of individual norms, namely, what an individual norm requires from its subjects, also called “legal facts” (^{Sciaraffa 2016}) or “first-order legal claims” (^{Toh 2019}).

For example, the question “Was there a Roman legal system?” queries the system’s existence, while “Did Roman law have a rule compensating property owners for damage caused by third parties?” questions the system’s content. In turn, the question “Was Lex Aquilia a rule of Roman law?” is a question about a law’s existence, while “Did Lex Aquilia protect only the property of Roman Citizens?”, “Which kind of property did the law protect?”, and “Did the law provide compensation for fault only, or did it also allow for strict liability?” are questions about the content of that law. Positivist theories of law can be theories of existence or theories of content. Non-positivism, which in this understanding includes natural classical law theories with a basic theory of basic goods, is the familiar name of many approaches that deny the positivists’ claims.

With this clarification in mind, I will argue that Martin’s argument does not adequately present the object of her criticism, namely, legal positivism. In her view, the substantive theory of law used in contemporary positivism is primarily represented by the “sources thesis”, which she characterizes as follows:

There are two central theses that are often associated with legal positivist thought: the separability thesis and the sources thesis. Those who champion the former maintain that there is no necessary connection between law and morality. While this claim has been a topic of considerable discussion, I will focus only on the second thesis. According to the sources thesis, which Raz famously defends, one need not rely on moral arguments in order to establish the content of a given law. The law of a given system consists of an aggregate of factually ascertainable norms. In other words, the law consists of an aggregate of positive laws. The label ‘legal positivism’ seems appropriate insofar as it captures this idea (JP, vii).

Martin questionably identifies the separability thesis with the claim that there are “no necessary connections between law and morality”. As discussed in the next section, Raz described such a view as “manifestly false” (AL, 315). For now, let us focus on the sources thesis, which, per Martin’s formulation, is an account of the content of legal norms that entails an account of the content of legal systems. Given this presentation, Martin’s argument can be regarded as an indirect defense of the non-positivism of legal content, a theory of law in which discerning the law’s content is an evaluative mission based on contested politico-moral values, not a “fact-finding” one.

Many contemporary positivists believe the sources thesis best represents the intellectual tradition (e.g., ^{Gardner 2001}; ^{Green 2003}). However, Martin’s description of the thesis sets aside an additional element of the positivists’ project, namely, their theories of law’s existence. Consider Raz’s formulations:

I shall rename the strong social thesis ‘the social thesis.’ (…) A law has a source if its content and existence can be determined without using moral argument (but allowing arguments about the people’s moral views and intentions, which are necessary for interpretation) (AL 47, italics added).

A law is source-based if its existence and content can be identified by reference to social facts alone, without resort to any evaluative argument (EPD, 210-11, italics added).

Two significant differences should be noted. On the one hand, as Raz describes it, the thesis is not merely a negative claim about which facts do not matter to determine the existence and content of law (i.e., moral argument). Instead, it is a positive statement of the “facts by virtue of which it [a law] is valid and which identifies its content” (AL, 47-8), in other words, its sources, such as (but not limited to) Acts of Parliament, precedents, or customs. Moreover, as we will further explain in the next section, Raz’s formulation of the thesis allows for participants’ moral arguments in order to interpret laws but rejects the need to engage in politico-moral arguments about the underlying reasons that led to a particular enactment or settlement. He writes:

Legal sources…. are capable of being identified in ways which do not rely on the considerations they are meant to decide upon. An income-tax statute is meant to decide what is the fair contribution of public funds to be borne out of income. To establish the content of the statute, all one need do is to establish that the enactment took place, and what it says. To do this one needs little more than knowledge of English (including technical legal English), and of the events which took place in Parliament on a few occasions. One need not come to any view on the fair contribution to public funds (EPD, 221).

In this sense, observers can determine the existence and content of law by reference to social facts only.

On the other hand, and more importantly, Martin overlooks the second element of Raz’s characterization (i.e., a theory of legal existence) that is more fundamental and explanatorily prior to the theories of legal content. It is a central insight of contemporary positivism that any account of the existence and content of individual laws presupposes an understanding of the existence of legal systems. As Hans Kelsen put it, “it is impossible to grasp the nature of law if we limit our attention to a single isolated rule” (^{Kelsen 1945}, 3). The “context principle”, as Kevin Toh names this methodological prescription, is also part of the philosophical project of H. L. A. Hart: Instead of analyzing individual norms in isolation, laws need to be studied in the relevant context in which they are meaningful, namely, as part of the normative system to which they belong (See ^{Hart 1983}, 26; ^{Toh 2015}, 337–38; ^{Gardner 2004}, 179–81; ^{Waluchow 2011}, 363–66). Following Kelsen and Hart, Raz claims in his opera prima that “every law necessarily belongs to a legal system” and that “it is a major thesis of the present essays that a theory of legal system is a prerequisite of any adequate definition of ‘a law’” (^{Raz 1980}, 1, 2).

Consequently, legal positivism necessarily begins with a theory of the existence of legal systems. For example, according to Hart’s influential theory, a community is regulated by a legal system if (i) a normative complex regulates the subjects’ behavior; (ii) that complex has social rules that regulate the creation and application of rules and empower officials for those tasks; (iii) norm-applying officials practice a social rule identifying the system’s norms (i.e., rule of recognition); and, (iv) subjects widely comply with the norms that norm-appliers identify and their settlements about them (^{Hart 2012}, henceforth, CL, ch. 5). Save for some reservations regarding (iii), Raz largely accepted Hart’s conception of legal systems. Still, he suggested that Hart’s view could not differentiate legal systems from other institutionalized orders (i.e., complexes of norms with norm-creating and norm-applying officials), such as those of associations, clubs, and universities. Hence, he added additional conditions: (v) legal systems are those that display three claims or self-understandings exhibited in the language of the participants: a claim of comprehensiveness (i.e., the capacity to regulate any behavior); a claim of supremacy (i.e., the power to regulate the establishment and application of other institutionalized systems by its subject-community); and a claim of openness (i.e., the ability to give force to norms that do not belong to the legal system) (PRN, 149-154; AL, 116-20; cf. CL, 249). This explanation warrants a commitment to the following thesis as the key statement of contemporary legal positivism:

(LP1) The existence of a legal system is a matter of social facts (i.e., i-v), not of moral merit.

The first part of the statement is often called the “social fact thesis;” the second is the “separability thesis”, as positivists understand it.

In keeping with the “social fact thesis”, positivists also hold that legal norms are brought to existence by means of other norm-generative social facts (i.e., certain speech acts, deeds, publications, inscriptions, recognition, etc.) that presuppose the presence of those facts specified in LP1. For example, if Roman norm-appliers recognize consul Aquilio as the communal legislator, and he claims in the proper context “I hereby declare this rule that grants compensation to such and such victims”, such statement creates a new norm that did not exist before the utterance. In this sense, most contemporary positivists concur in the following statement:

(LP2) The existence of individual norms is a matter of social fact (i.e., creation or recognition by the community’s officials), not moral merit.

Significantly, positivists disagree on several aspects of LP2, among others, on how to interpret those social facts that recognize (non-social-based) moral standards such as human dignity or due process. Inclusive legal positivists advocate what Raz calls the “weak social fact thesis”, according to which moral norms recognized by social facts can be incorporated or included into the legal system (e.g., ^{Waluchow 1994}). Since the recognition of norm-applying officials can transform moral standards into legal ones, the existence of a norm is ultimately a matter of social fact. By contrast, Raz and exclusive legal positivists advocate the “strong social fact thesis” or “sources thesis”, in which the existence of legal norms is solely a matter of social facts, namely, its source. Moral considerations, in the strong interpretation, are “extra-legal” considerations relevant to the interpretation of legal material and the resolution of conflicts.

The positivists’ division runs deeper when they turn to the theories of the content of legal systems and norms. Doubts arise because inquiries about existence use external statements that describe and explain certain social facts (regularities of behavior and pro-attitudes) that give rise to practice-based norms, while questions about content are internal inquiries that use the norms fixed by social facts in particular situations (CL, 88-91, 102-106). For example, a claim such as “these players follow a rule forbidding them to touch the ball with their hands” is an external statement typically, but not necessarily, issued by observers to explain and describe the game’s social rules. In contrast, an internal statement such as “Penalty!” (issued by a referee of the game after one player touches the ball with her hand) uses the norm of practice in a specific situation. Internal statements are typically, but not necessarily, issued by participants, namely, agents that accept the practice’s standards or that comply with the settlements of those who accept them. According to the Hartian description, internal statements presuppose, without stating, the state of affairs described by external statements. In other words, external inquiries describe the “normal, though unstated, background or proper context” —to use one of Hart’s expressions (CL, 85)— in which internal statements can be issued.

Because statements about content are internal and have an indirect relationship with social facts, it is unclear whether contemporary legal positivism entails a theory of the system’s content or needs to develop one. Statements about the content of the legal system use, without mentioning it directly, the rule of recognition—the social rule practiced by legal officials establishing the criteria of validity that identify the norms that belong to the legal system. While Hart suggested that the content of the legal system is determined by the shared perspective of norm-applying officials (CL, ch. 6), there are doubts about the cogency of such a view (e.g., ^{Toh 2005}, 90–91; ²⁰⁰⁸, 483– 87; ^{Duarte d’Almeida 2016}, 194–95; ^{Sciaraffa 2016}; ^{Toh 2019}). Any positivist view of legal content needs to provide an argument about the external description of the participant’s statements, and these critics believe that such burden has not been adequately met. As a result, one relevant interpretation of legal positivism has abandoned the project of creating theories of the content of the legal system and legal norms. Some of these positivists (i.e., advocates of LP1 and LP2) have suggested that the project is compatible with non-positivist accounts of legal content (for an explicit argument, see ^{Sciaraffa 2012}). Conversely, some self-styled non-positivists have limited their explanations to legal content (e.g., ^{Brink 1985}), or moral criticism of existing law (e.g., ^{George 1996}, vii–viii; ²⁰⁰¹, 228), while endorsing the positivist account of law’s existence.

While some contemporary positivists have rejected the task of providing a theory of legal content, such is precisely the task of the second half of Raz’s “sources thesis” that Martin highlights in her characterization. This view is the following:

LP3: The content of a norm is a matter of its sources, not of moral merit.

Like LP2, this thesis allows for strong and weak readings. According to the former, advocated by Raz and fellow exclusive legal positivists, the norm’s content is solely a matter of its source; according to the latter, the norm’s content is ultimately a matter of its source. In the weak reading, merit matters for the determination of the norm’s content when a source establishes that (^{Waluchow 1994}). Martin recognizes the possibility of a weak reading of LP3, but he dismisses it as a mere “conceptual possibility” (JP, 3-4, n.3), disregarding arguments of the advocates of the inclusive reading who consider it as a better description of the practices in contemporary democracies. In any case, Raz and other defenders of the strong reading LP3 are aware that this view does not follow directly from LP1 and LP2, so they have offered arguments to defend it against both non-positivists and those positivists who reject it. The most important argument for this view is Raz’s famous argument of authority (AL, ch. 2; EPD, ch. 9; BAL, ch. 5), but there are others, such as the argument of the “practical difference” (^{Shapiro 2001}). Neither of these arguments is universally accepted among legal positivists. In other words, the strong reading of LP3, that Martin introduces as the central to legal positivism, is a contested claim within the positivist camp.

Further, it is critical to note that LP3 is not an internal thesis of how participants of a legal practice (subjects and officials who accept the system’s constitutive rules) identify or should identify the content of laws. Instead, in keeping with the social fact thesis, Raz presents it as an observer’s description of participants’ behavior and pro-attitudes. The success of the strong reading of LP3, then, presupposes the possibility of an external description of internal statements. To meet this challenge, Raz introduced the notion of “detached statements”, a third kind of statement made by speakers who adopt, without sharing, the internal perspective that accepts the norms and uses them as standards of behavior (PRN, 172-3; AL, 140-3, 153-7, 305). Raz’s favorite example is the case of an orthodox Jew who asks for the advice of his Catholic friend, who is an expert in Rabbinical law. The friend needs to adopt, without endorsing, the perspective of Jewish law to provide his advice. Like the Catholic friend who is not subject to Jewish law but can describe how the participants of that practice have determined its content; legal theorists do not take the place of participants in identifying the norm’s content. They do, however, describe how these participants have specified it in the past. Since the detached determination of legal content needs only knowledge that certain enactments and deeds took place in the past and knowledge of what they mean, the legal content remains a matter of social facts.

While the possibility of detached statements was accepted by Hart and other prominent theorists (^{Hart 1983}, 14–15), recent literature has raised substantial doubts about its success (e.g., ^{Toh 2007}; ^{Duarte d’Almeida 2011}; ^{Bulygin 1982}). For these critics, there is no distinctive detached point of view, but merely non-participants’ internal statements that use the norms so this notion does not adequately explain the possibility of external describability of internal statements required to maintain LP3. This skepticism explains why some advocates of LP1 and LP2 have suggested imagining legal positivism beyond the Razian pale (^{Toh 2015}, 691).

We can now see that Martin’s description restricting legal positivism to the strong reading of the theory of legal content stated in LP3 is problematic. First and most clearly, Martin’s characterization has set aside a significant part of Raz’s statements about the sources thesis (i.e., the strong reading of LP2), and in doing so, it excluded a whole dimension of positivist theory (i.e., theories of the existence of legal systems and individual norms).

Second, LP3 is not representative of the entire intellectual tradition. On the contrary, the “sources thesis” or “strong social fact thesis” that comprises the exclusive readings of LP2 and LP3 is one iteration of the general commitment of legal positivists, the “social fact thesis” as articulated in LP1 and LP2. This reductive characterization of legal positivism would not be an issue if Martin’s argument focused on Raz’s work individually, and not as a representative of an intellectual tradition. In fact, Raz himself would have preferred that his claims be judged on their own merits and not as illustrations of a label or intellectual tradition (AL, 335; BAL, 373). However, Martin extracts general lessons about legal positivism from her objections. She writes: “The instability of [positivists’] non-normative ‘conceptual’ or ‘descriptive’ project…is best viewed from the inside of Raz’s complex account” (JP, vii). I doubt her success in this more general enterprise, given that she has not identified the most representative view of the contemporary discussion of legal positivism.

Finally, Martin has not only excluded important parts of the positivist project, including its theories of existence, internal/external distinctions, and arguments for the external describability of internal statements but, more importantly, she has altered the positivists’ order of explanation: any theory of the content of individual norms presupposes an account of the norms’ existence, which in turn presupposes an understanding of the legal system. To inquire about theories of content without a theory of existence is a decontextualization of the positivists’ view. Because of this reductive presentation of legal positivism, Martin ends up with an idiosyncratic discussion of Raz’s views that neglects or misrepresents relevant arguments of his project, including the theories of law’s existence he developed in The Concept of the Legal System and PRN.

The most important examples of this decontextualization occur in her discussion of norm-applying officials. Martin suggests that the presence of norm-applying officials is the “necessary feature” that defines a legal system (JP, 6, 17, 23). She then criticizes Raz for his inability to explain common law based on what she takes to be the distinction between non-legal “systems of pure discretion” and legal systems with “judges” (JP, 19). Since common law adjudication allows judges to alter the system’s rules at times, common law has “more in common with instructions given to judges in systems of absolute discretion than those in legal systems” (JP, 25). Martin further claims that because of the failure of this theory, Raz develops novel accounts of adjudication in AL and EPD, in which he “abandon[ed] his claim that judges have a duty to apply the law” (JP, 4, 26, 30).

However, all these claims are dubious as matters of interpretation. In my view, PRN does not develop “a clear vision of the way in which law creates order” (JP, 3) that Martin suggests. Instead, it provides a general positivist theory of the existence of practice-based norms and normative systems (including games, associations, clubs, etc.) in terms of social facts (PRN, chs. 1-4), of which law is only an illustration (PRN, ch. 5). In this account, neglected by Martin, the presence of norm-applying officials is not the “necessary feature” that distinguishes between law and non-law, but a central feature applicable to legal and non-legal “institutionalized systems”. A complex of interrelated norms (i.e., a system) becomes “institutionalized” when it has agents or institutions in charge of the creation of norms (i.e., norm-creators, such as codifiers or domestic legislators) and the application and enforcement of norms (i.e., norm-appliers, such as referees or state judges). To further explain norm-appliers, Raz contrasts them with systems of pure discretion. Unlike purely discretionary agents, like Khadi judges of Weberian examples who decide according to their own assessment of the situation (^{Weber 1978}, 976–78), norm-appliers make decisions according to pre-existing rules. For example, a soccer referee needs to apply soccer rules, and state judges need to apply state norms. The application of norms is constitutive of the office of norm-appliers; if they cease to do so, and make decisions based on their best judgement, they cease to be judges or referees.

Hence, Martin’s suggestions about the abandonment of the judicial duty to apply the law form a problematic interpretation of the Razian theory. As I explain in the next section, these judges can have significant discretion and still be bound by pre-existing rules. The distinction between law and non-law is not connected to the difference between discretionary agents and judges, as Martin suggests, but to the trio of claims mentioned above: comprehensiveness, supremacy, and openness (PRN, 149-154; AL, 116-20). Martin’s argument never discusses these elements. The third of these claims, as I suggest below, is critical in understanding the role of moral principles in legal practice.

In sum, Martin’s argument does not adequately portray the intellectual tradition of legal positivism or the Razian project. Because her reading eschews an essential part of the positivist project and her critique turns its explanatory structure upside down, it is not surprising that she finds the project internally contradictory.

III. The Methodological Approach of Legal Positivism

Since Martin believes that substantive theories about the nature of law are intrinsically related to certain methodological views about how legal theory should proceed, the problems noted in the previous section also affect her characterization of the “methodological approach” of legal positivism. For Martin, legal positivism of legal content, as represented by the strong reading of LP3, goes hand in hand with a “descriptive”, “conceptual”, or “non-normative approach” that excludes politico-moral considerations from legal theory. This methodological project, created by Hart and now “dominant” in the contemporary discussion (JP, 5), relies on two sets of distinctions:

The tenability of the Hartian project rests on the viability of the sharp distinction between non-normative conceptual analysis and normative philosophy. The distinction between a non-normative theory of law and a normative theory of adjudication is eventually erected to protect the first more foundational one. If this project is going to work, value assumptions about the nature of law must be kept out. In particular, if a theory of adjudication is fundamentally moral in nature, then it must remain quarantined in a separate sphere. I will call into question the stability of both sets of distinctions, thereby raising serious doubt about this particular methodological approach (JP, 6).

The first distinction is between “descriptive” or “non-normative” conceptual analysis, advanced by legal theories, and “normative philosophy”, offered by political and moral philosophers. Per this view, Hart aims to separate the conceptual project of legal theory from the studies of political and moral philosophy, in contrast to a venerable tradition in legal philosophy that preceded him (JP, 5-6).

In my view, the now-familiar contrast between “descriptive” or “non-normative” and “normative” or “evaluative” is controversial and obscure (cf. ^{Dickson 2004}; ^{Rodriguez-Blanco 2006}; ^{Priel 2007}). I believe that the contrast can be best understood in terms of “politico-moral argument:” whereas normative or evaluative theories are based on substantive politico-moral arguments, descriptive theories avoid them. Moreover, since “normative philosophy” includes conceptual elements concerning the analysis and elucidation of basic concepts (i.e., state, democracy, etc.) in addition to substantive politico-moral claims, one can suggest a better distinction between a “conceptual theory of law” and “substantive politico-moral philosophy”. Therefore, the first thesis of the “methodological approach of legal positivism” (methodological positivism, for short) that Martin’s argument opposes could be reconstructed as follows:

(MLP1) A (conceptual) theory of law describes social facts and does not engage in substantive politico-moral argument.

Martin further argues that to defend this project, positivists introduce a second distinction between “theories of law” and “theories of adjudication”, namely theories about the judicial role, application of norms to specific cases, and resolution of disputes. This second distinction is the following:

(MLP2) A (conceptual) theory of law does not need to consider politico-morally-based adjudication.

Implicit in this distinction is the claim that positivists exclude adjudicative concerns because they regard them as politico-moral issues, so they end up excluding considerations that matter to legal operators and subjects. Since her case against legal positivism “calls into question the stability of both sets of distinctions”, she indirectly advocates for a type of methodological non-positivism, namely, conceptual theories of what law is based on politico-moral argument and closely connected with adjudication, which would, in turn, assign a more prominent role to legal operators and subjects.

I have doubts about these two distinctions as characteristic of the methodology of legal positivism. Regarding MLP1, Martin’s formulation suggests that politico-moral considerations have no role at all in positivist “theories of law” (i.e., the positivist account of the whole legal domain). For Martin, positivists hold MLP1 as a methodological correlative of what she believes is the positivists’ substantive “separation thesis”, namely, that “there are no necessary connections between law and morality” (JP, viii). However, such a formulation is too strong and neglects the various senses of “law” discussed above. As stated in LP1 and LP2, positivists claim that the existence of legal systems and norms does not depend on politico-moral considerations (^{Green 2008}). For example, positivists claim that an observer can ascertain that the Roman Empire had a legal system because it met conditions (i-v) listed above (i.e., it was a complex of rules with standards to identify, create, and apply norms as well as legal officials who claim comprehensiveness and supremacy over other institutional orders). Consequently, the presence of iniquitous laws regulating slavery (i.e., unjust laws enacted or recognized by the relevant officials) did not affect the legal nature of the system or the norms.

Thus understood, positivists do not deny that politico-moral considerations have a role in legal theory as MLP1 suggests; they suggest that it does not determine law’s existence. This thesis is compatible with the influence of politico-moral considerations in many other aspects of a theory of law, including concerns about content and adjudication. Most legal positivists are happy to recognize that, setting aside the theories of law’s existence: “a theory of law includes an undeniable normative dimension”, as Martin suggests (JP, 57). In fact, Raz explicitly recognizes that several aspects of a theory of law, including adjudication and interpretation of legal norms, involve substantive politico-moral activities, as we will discuss below (AL, EPD, 192-3, 219-221; BAL, 111-5).

Furthermore, the connection between substantive and methodological positivism is dubious. As stated in LP1 and LP2, positivists deny only one relationship between law and morality, namely, that the existence of legal systems and laws depends on moral merit. However, positivists explicitly recognize other connections (as noted by ^{Gardner 2001}; ^{Green 2003}; 2008), some of which are necessary. Kelsen claims that legal norms must deal with moral matters, unlike other normative systems (¹⁹³², 34). In turn, Hart suggests that all legal systems need a “minimal content of natural law”, including rules governing violence, property, and kinship since it aims for the survival of agents like us (CL, 193-200). He also claimed legal systems are morally risky because they create possibilities for oppression and alienation not possible in non-legal situations (CL, 204). Raz also suggested two essential connections between law and morality. First, law’s existence is morally valuable (PRN, 165-170), and second, law makes moral claims (i.e., moral claims are part of officials’ self-representations of legal practices), such as the claim to authority (AL, ch. 2; EPD, ch. 10). In studying these connections, positivists make substantive politico-moral arguments about law. As a result, virtually all positivists, including Raz, recognize that politico-moral considerations have a crucial role in many aspects of a theory of law other than law’s existence; none of them is committed to MLP1 as Martin formulates it.

Now, given that the target of Martin’s objection is the strong reading of LP3 (i.e., positivism as a source-based theory of content), one can suggest a more restrictive interpretation of the methodological claims she aims to reject:

(MLP1*) A theory of legal content is a matter of describing the sources of law, not of politico-moral considerations.

However, I do not think the objection is successful even in this more limited formulation. Consider Raz’s formulation of the thesis again:

I shall rename the strong social thesis ‘the social thesis…’ A law has a source if its content and existence can be determined without using moral argument (but allowing arguments about the people’s moral views and intentions, which are necessary for interpretation) (AL, 74, italics added).

As we noted above, the highlighted passage makes clear that participants’ moral arguments matter for Raz. To understand this passage, we need to recall that the sources thesis provides a detached description of the content of a legal system, not an internal account of how participants determine the content of a norm in a given situation. In this view, observers can identify the content of a legal system (e.g., Roman law) by relying on certain “facts” (i.e., the legislation, precedents, and customs), without worrying about the merit of these norms or politico-moral reasons that justified their enactment.

This argument is connected to Raz’s conception of the role of norm-applying institutions, another element of his account of law’s existence that Martin does not discuss. Per this view, when norm-appliers resolve specific cases, they both solve individual disputes and authoritatively establish the norms’ content. Whereas other officials and community members might have “opinions”about what the communal norms should be before the settlement,as Raz puts it, norm-appliers authoritatively establish the shared public standards that guide the community and determine the content of these rules (PRN, 135; AL, 108-9). Hence, norm-appliers are the community’s authoritative identifiers of rules and their content. In this sense, the sources thesis allows for external descriptions of the content of a legal norm once norm-applier’s decisions identify them.

For example, observers can use social facts to determine the existence of a given law (i.e., Lex Aquilia is a law because the relevant norm-applier used it) and its content (e.g., in their precedents, norm-appliers limited the application of Lex Aquilia to Roman Citizens only). Before the settlement, the law’s content was not yet authoritatively established; once norm-appliers utter the relevant settlements, the content became identifiable to external observers. Very differently, the process of reaching these decisions, or “deliberative stage” (EPD, 206), concerns interpretation and adjudication, which, for Raz, are substantive moral activities (BAL, 223-370). Raz even allows participants’ moral views and intentions to determine the system’s content, but these are internal issues that are ex post facto described by observers. As a result, even in the most robust version of the social fact thesis, the politico-moral considerations have some role in determining the norm’s content. Consequently, MLP1* is not accepted by Raz’s version of positivism.

Finally, let us now consider MLP2, namely, the distinction between a “theory of law” and a “theory of adjudication”. From the outset, note that if “adjudication” is an integral part of legal practice, a “theory of adjudication” is part of a “theory of law”. In the light of the distinctions developed in the previous sections, we can better formulate the claim as follows:

(MLP2*) A theory of the existence and content of law need not consider adjudication, which is based on politico-moral considerations.

If this is correct, it is strange that Martin has chosen to characterize this distinction as a distinctive positivist’s methodological commitment given non-positivists scholars widely accept it (e.g., ^{Moore 1985}, 274; ^{Brink 1985}, 364–65; ^{Soper 1992}, 2414; ^{Murphy 2003}, 264–67). Furthermore, some non-positivists even claim that the presence of politico-moral elements does not abolish the divide between conceptual legal theory and politico-moral philosophy while some non-positivist scholars also attempt to “describe” the law, not to justify it; and the fact that the “description” of the nature of law includes politico-moral considerations does render it a non-conceptual enterprise or transform it into an exercise of substantive politico-moral discourse (e.g., ^{Murphy 2006}, 8–9; ²⁰¹³, 8–9).

In any case, most positivists will qualifiedly agree with the claim, that the theory of law’s existence is different from theorizing how judges should decide. For example, it is one thing to determine whether Romans had a legal system according to facts (i)-(v); it is a different one to explain how Roman norm-appliers should decide in specific cases. However, pace Martin, the distinction between “theories of existence” or “theories of adjudication” is not “erected to protect” the division between “descriptive” and “normative jurisprudence” or to shield legal theory from politico-moral considerations. Very differently, the distinction is closely related to the divide between internal and external inquiries mentioned above. Theories of adjudication are concerned with internal statements, namely, statements of participants who accept and use the system’s norms, not with the external statements of social facts, including the detached description of internal statements presupposed by Raz’s characterization of LP3. Positivists do not suggest that adjudication is irrelevant; they only mean that it is at a different theoretical level.

To be clear, inquiries about existence and content need to identify the relevant norm-applying officials and the criteria of validity they share. However, such theories do not need to study doctrinal questions about how their interpretation proceeds, let alone resolve doctrinal questions about the content of norms (e.g., that certain standards apply to Roman citizens only, or certain types of property are protected and not others). Far from a distinction based on “taste” (JP , viii), as Martin suggests, positivists have offered a theoretical argument to questions of content and questions of adjudication that the objection fails to consider.

In the end, Martin’s arguments are primarily directed towards an extreme methodological view that separates law and morality in all dimensions. However, this is not one that modern positivists advocate. Moreover, her formulation fails to consider a critical distinction between internal and external affairs: while theories of adjudication are concerned with internal statements, theories of existence and content are advanced through external statements, including external descriptions of internal statements about adjudication.

IV. Conclusions

In this paper, I have argued that Martin’s challenge did not adequately characterize legal positivism’s method and substance. In her view, positivism is a theoretical “castle” that aims at explaining legal existence, content, adjudicative procedures, and authority (JP, vii-viii). The failures of the Razian account show the impossibility of an explanation of all these aspects without politico-moral considerations. As a result, the separation between law and morality fails, and the castle “collapses from within” (JP, 7, 9, 29, 40, 43, 55, 124). In response, I suggest that the intellectual tradition of positivism is not a comprehensive theory of legal phenomena but revolves around a modest account of law’s existence that needs to be complemented with further theoretical considerations. In this view, “no legal philosopher can be only a legal positivist”, as ^{Leslie Green (2003)} reminds us. While authors who endorse the positivist’s theories of law’s existence have their views on those other matters, there is little reason to believe that these additional components are also part of the intellectual tradition or entailed by it. There is no requirement that positivist authors explain every aspect of legal practice in terms of social facts, nor is it incoherent that defenders of LP1 and LP2 explain other features of law by relying on politico-moral values.

If these ruminations are correct, we end up with a more sophisticated conception of the theoretical task: We are better off by studying the “castles” —to preserve the metaphor— built by individual authors, not intellectual traditions, while recognizing that these castles can be composed of positivist and non-positivist materials. In our studies, we should be more concerned with the truth of their specific components, not with their coherence to loosely related projects or their allegiance to certain agendas. This is, I think, how the projects of legal positivism and non-positivism should be described and judged.