Reflective Legal Positivism

1. Introduction

A large part of contemporary general jurisprudence seems to be a living scientific rem nant similar to that of the history of science, which refers to theories of the “Newtonian age”. By analogy to the classical approach in physics, traditionally accepted approaches to the concept of law – like legal positivism or naturalism – are supposed to serve as the final answers to the (allegedly) essential question of jurisprudence: What are the mechanics of law and what are its basic concepts, including the general concept of law? The main line of criticism with regard to the classical doctrines is similar to the criticism of Newtonian mechanics during the 19th century. The increasing number of experimental results (so-called anomalies) did not fit the predictions of the classical physical theory. This meant that the ability of Newtonian mechanics to explain and predict natural phenomena had been exhausted, the theoretical framework could not account for new anomalies and better, more inclusive alternatives needed to be found. Such an alternative was later presented, among others, by Einstein in the form of his general theory of relativity.

Why do I discuss problems of scientific methodology in a jurisprudential paper? This is because I believe that most methodological questions in jurisprudence are analogous to questions posed by scientific thinkers. The important difference, as we shall see, is that the evidence that underpins legal theories is very often not of an empirical kind. Thus, the difference between science and jurisprudence lies not in methods but in the nature of evidence. Common platitudes rather than observations serve as the backbone for legal theorizing. The primary methodological commitment to holism (or: reflective, holistic testing of theories) is shared by modern legal positivists, whose approaches I discuss later on. The analogy between the position of classical mechanics and general jurisprudence has its limits. General legal theories are more philosophical than scientific theories.

Although legal theories rely to some extent on empirical data (e.g., socially accepted platitudes about the concept of law or social externalities), they are not usually capable of conclusive verification or falsification since there is no operative, widely established criterion of demarcation in jurisprudence. General legal theories which end in the presentation of a normative judgement are used as basic premises in deductive legal reasoning rather than as a means of prediction. In legal theory, operative empirical data gained in experimental circumstances is less significant in evidential terms than general folk platitudes about the nature of law and its functions.³ Such “intuitive” platitudes often ground standard legal conceptions and methods and are believed to be a solid departure point for legal theorising; they also generally delimit the scope of acceptable theories about law and its interpretation.⁴ I am convinced the claim that law is essen tially agonistic is one of the important folk platitudes which is unjustly neglected by tra ditional legal doctrines. The other but related putative feature of traditional general theories of law is that they appear to commit to representationalism.⁵ As a result, it seems that the disagreement between such theories is the disagreement over the “truth of the matter” (which disagreeing peers might get things right or wrong). But what about the fact that such stubbornly sought legal truth, which is in part conventional and conceptual, influences both the creation of institutional (legal) reality and its under standing? In this context, the traditional theories very weakly commit to the view that law is agonistic: its agonistic character stems not from the nature of law itself but rather from lawmakers’ mistakes or the epistemic failures of law’s interpreters. This solution is not, however, the only philosophical way of accommodating the “agonistic platitude”.

Folk intuitions embedded in platitudes constitute the common understanding of law, and in my view, “genetically” delimit the scope of admissible theories of law. This is accepted by positivists as a meta-philosophical pedigree test in jurisprudence.⁶ Positivism aspires to include and develop as many platitudes as possible, although certain plat itudes collide with others; for example, the conventional platitude related to the institu tion of law being public, reason-giving and stable collides with the “agonistic platitude”. Since platitudes serve as a basis for further multidirectional development of theories, it is no surprise that theorists who assume different conceptual priorities of platitudes may end up accepting substantially different accounts of law. The possible diversity of platitudes partially explains the reason why theoretical disagreements must eventually arise in legal practice.

The interesting question is: Why should any “priority thesis” of a conceptual kind be assumed? Why should the conventionalist platitude be dogmatically preferred to the agonistic platitude? However, these questions might pose a false dilemma in that the selection of and commitment to certain platitudes can be pragmatic rather than dog matic. Pragmatist advice is to believe not in what is true but in what is useful (warranted acceptability). Thus, one good prima facie reason to take the conventionalist platitude for granted is tradition: we did it this way and it worked! Still, this does not exempt us from carefulness; any suspicion with regard to the usefulness of our conceptual premise should trigger reflection. I believe that the nature of law is being determined by conventional folk beliefs at the very highest level of generality (the level of artefactual kind). Every step down the metaphysical jurisprudential ladder may lead to changing perspectives and related ontological commitments. Thus, we should not seek any global solutions to answer the question what law is and how it exists (general jurisprudential truth) but, rather, we should allow for a more piecemeal methodology. Here, as in other cases of philo sophical inquiry, there is a strong connection between the superficial agreement, the complicated nature of things, and the ubiquitous controversy about the ultimate under standing of the nature of law. The right way to begin is not to assume agreement but to understand the reasons why views on law are becoming inherently controversial much like views on other socially relevant phenomena (e.g., morality). I will next examine how contemporary legal positivism could constructively deal with the jurisprudential anomaly: by recognizing the pervasiveness of theoretical disagreements.

2. Theoretical disagreements

The “argument from theoretical disagreement” (ATD) was introduced by Ronald Dworkin, who used it as a weapon against conventionalist theories of law (e.g., legal positivism) as he tried to demonstrate that the special kind of disagreement regard ing the determination of the “grounds of law” played a central role in legal practice.⁷ Dworkin thought that judges, parties to legal disputes, and legal academia did not merely empirically disagree about whether certain generally recognized grounds of law existed (were in force). He observed that robust and consequential alternative theories of the grounds of law were frequently assumed in legal disputes.

Dworkin held that Hartian legal positivists believed that all disagreements between lawyers (especially judges) were “empirical”. Such disputes arose when lawyers agreed on the grounds of law but disagreed over whether those grounds were satisfied in a particular case.⁸ Positivists viewed the grounds of law as established by a social rule (convention). Such a rule might have been a very complicated social phenomenon, but according to the standard view, judges (officials) generally shared the attitude of accept ance towards a certain set of sources of law (“rule of recognition”). Dworkin described legal positivism as a theory “that law depends only on matters of plain historical fact, that the only sensible disagreement about law is empirical disagreement about what legal institutions have actually decided in the past”.⁹

In response, Hartian positivists¹⁰ have generally argued that the most important legal disagreements are empirical disagreements. They saw disputes about the criteria for validity to be about repairing existing law rather than about what law was. Hartians fiercely defended the conventionalist thesis.¹¹ Others argued that theoretical disagree ments (TDs) did not play any significant role in legal systems since they appeared only in the minority of cases, for example, in the highest courts (on “the pinnacle of the [legal] pyramid”¹²). Some time ago, Scott Jonathan Shapiro noticed that these responses simply overlooked the problem; as a positivist himself, Shapiro suggested that TDs should be taken seriously. He vindicated Dworkin’s claim about ATD being a “serious threat to legal positivism”.¹³ The folk theory of law and legal practice does not determine the content of the concept (theories) of law but rather underdetermines such theories (like evidence underde termines physical theories).¹⁴ If this is so, TDs are inevitable and there are many possible ways to discriminate between relevant and irrelevant truisms, to set them in order and to develop them into a theory of the nature of law. We may think of the positivist theory of law as one intelligible theory of an important concept; it may be the most popular one, but it is certainly not the only possible one. What is the merit of Dworkin’s ATD then? There is certainly more than the argu ment that there are fundamental conceptual disagreements in jurisprudence. In this form, the argument would be trivial. Dworkin’s most notable example of a TD was the Riggs v. Palmer case (“Elmer’s case”). In Law’s Empire, Dworkin wrote that “the dispute about Elmer was not about whether judges should follow the law or adjust it in the interests of justice”¹⁵ but about “what the law was, about what the real statute the legislators enacted really said”.¹⁶ Dworkin thought that this case was an instance of a general phenomenon. However, critics forcefully argued that judges in this case did not disagree over “what the real statute of wills really said” because all of them accepted that the statute of wills entitled Elmer, who murdered his grandfather, to inherit after him.¹⁷ The genuine problem was whether “equitable principles could temper the applic ation of the statute of wills even though the plain language of the statute permitted wills to be altered only by prescribed means”.¹⁸

Thus, one may argue that the significant additional element of Dworkin’s own con ception of TDs is that all theories that disagree have to somehow resolve the moral question with respect to the grounds of law. Any theory has to give a positive or neg ative answer to the question of whether certain moral principles that could be applied to a case at hand exhaust the grounds of law. Dworkin’s construal of the problem deems every TD to be a moral disagreement; that is, a claim which has to be denied by positivists who try to accommodate the idea of TDs. As a result, they are prone to say that such a question need not be asked in a genuine TD. I suggest that the difference between positivist accommodations of ATD and Dworkin’s own idea is a difference in the view about what types of theories are admissible as genuine parties to theoretical dis agreement. Dworkin would demand fully-fledged theories that are able to solve complic ated moral issues, whereas positivists look for the easiest operational conception avail able. The irony of traditional positivism is that it does not see that what is operational in certain widely recognized institutional circumstances might not work in other ones. Therefore, traditional positivists excessively generalize their default conception (and, in turn, they make normative recommendations in novel contexts rather than descriptive statements). The new “reflective positivists” that I discuss seem to be more careful: their conceptual starting point is not the only conceptual solution which may prove useful. In what follows (section 4), I introduce three recent and refined positivist approaches that take ATD seriously: 1) Scott Shapiro’s idea of meta-interpretive disagreements, 2) Alani Golanski’s institutional account of TDs, and 3) Kaarlo Tuori’s critical positi vism. All these theories argue that TDs can be structurally incorporated into the popular mechanics of legal positivism. The question is whether the cost of such an upgrade is too high and whether refined “positivisms” still belong to the positivistic family. In the end, I argue that such positivist answers to ATD may in fact be viewed as relying on “reflective”, “holistic”, and “meta-philosophical” assumptions. This is the evolutionary change that makes contemporary legal positivism more methodologically conscious and less dogmatic. Such a holistic turn is no surprise given that Dworkin’s methodology is also holistic. I conclude, however, that the holistically augmented legal positivism – being a conscious close neighbour of legal realism – is a more reflective theory of law than the Dworkinian one.

3. Positivism and disagreement

Can legal positivists accept any viable account of TDs? There are two issues to discuss here. First, even simple traditional legal positivism as a theory of obvious law not only provides guidance in easy cases but also practically advises judges to make the best con sidered judgement in hard cases. In the latter case, legal reasons given by obvious law do not suffice to answer a legal question. Thus, it follows that a judge has to engage in practical deliberation.

Legal positivism of this kind is a narrow theory about the nature of law. It cannot allow for any disagreement between rival theories of law since such a concession would be self-undermining. The subject-level differences are the differences between theo ries of the same kind that address similar questions and problems. Such theories can not accommodate the fact of a TD (as observed from a meta-theoretical perspective). Instead, they have to consume or undermine other theories that deal with the same level of problems. Any such theory is a position in disagreement. It can be accepted only at

the expense of denying the truth of its adversaries. In other words, from the subject-level perspective, the fact of disagreement is not a reason for theoretical conciliation.

Second, the transition from a subject-level theory of law to a legal discourse level and TDs is a metatheoretical step. It resembles stepping up a theoretical ladder. One builds such theories by leaving open theoretical possibilities, that is, by allowing competing theoretical answers to certain questions. Positivists who want to take TDs seriously must embrace that meta-perspective. As a result, their theories operate on two differ ent levels: 1) the subject level of understanding and explaining the coherent practice of officials and providing a descriptive answer to the main question of “what is law” and 2) the metatheoretical level of discussing subject-level theories that disagree and identifying the scope and nature of both present and possible disagreements of this kind. The question is whether such positivist theories do not betray their basic subject-level assumptions.¹⁹ There are three interesting recent positivistic theories that move up the meta- -philosophical ladder: 1) Shapiro’s planning theory; 2) Golanski’s institutional account, and 3) Tuori’s critical legal positivism. Shapiro perceives a legal system as a complicated social plan realized through a joint commitment of lawmakers and agents. Law is a solution to society’s moral problems. It is described and morally evaluated. Each plan aims at excluding deliberation by agents who otherwise would seek the best possible action on their own.²⁰ On Shapiro’s account, TDs are disagreements between “interpretive methodologies”. Plans determine many types of social interactions, but they are never completely determined and leave place for discretion (the more sophisticated the plan, the less place for discretion).²¹ Plans do not necessarily determine legal outcomes but are the only grounds of law. Thus factors not included in a given plan are extra-legal. The planning theory of law is supposed to decide which interpretive methodology best fits the actual circumstances of legality. Planning depends on a distribution of trust. The more trust placed by planners in the reflective abilities of subjects, the more dis cretion provided to them, thereby allowing them to apply more demanding interpretive methodologies. Conversely, in cases in which trust is limited, plans are detailed. This theory states that the right way to resolve TDs is to analyse the institutional structure of a plan to see how trust is divided within it. If lawgivers do not trust officials, then a meta-interpreter must exclude demanding methodologies like Dworkin’s theory.²² It follows that moral TDs – as disagreements between morally robust interpretive theories (which back up serious moral choices) – are possible only in cases in which a lot of trust is placed in officials so that they can morally deliberate. However, purely institutional TDs of a non-moral kind are also possible in cases in which there is a lack of trust. Golanski presents related reasoning but framed it in Searlean institutional terms. He noted that one can understand TDs in either a narrow or wide sense. Narrow

disagreements are about which methodology of interpretation should be applied to legal materials in particular circumstances (originalism, evolutionism, intentionalism, con sequentialism etc.), as is the case with Shapiro’s TDs.²³ Nonetheless, Golanski claims that Dworkin did not perceive TDs in such a narrow way. Dworkin’s TDs are disagreements over the soundest interpretation of legal practice.²⁴ However, taking TDs seriously does not require the acceptance of Dworkin’s interpretive methodology.

Golanski argues that contemporary institutional philosophy has provided us with analytic tools sufficient to analyse the wider TDs. He refers to two features of Searle’s “institutional logic”: 1) the description of institutional structure comprised of constitutive and regulative rules and 2) an account of how to recognize power relations and different forms of commitments as reasons for actions within that structure.²⁵ Golanski writes: “[a]n understanding of the logic of institutional power and authority shows that ‘theoret ical’ disputes in law are, in the first instance, best understood as controversies over the standards for determining whether the existing legal materials are sufficiently directed at the present circumstances, and whether they provide a solution to the new matter with sufficient exactness”.²⁶ Although such TDs concern the relation of (Dworkinian) fit between existing legal materials and new contexts, they are not necessarily moral disagreements since they are not necessarily disagreements about the justification of that relation. The “institutional” account allows us to discern a basic sphere of TDs as con troversies over 1) the standards of exactness of legal decisions and 2) their intentionality. There are different parties and officials who could provide different answers to questions such as “What are the social facts which, having determined such-and-such forms of law, determine such-and-such legal outcomes?”²⁷ Thus, institutional TDs are disagreements over the criteria for the assessment of legal propositions (i.e., over the constructive fea tures of the institutional reality). They may also be understood as disagreements over “ontological presuppositions” of legal decisions. Disagreeing lawyers may tacitly assume differing views about what makes legal propositions true. They must, however, be aware of the consequences such presuppositions yield while being ready to revise these presup positions whenever new jurisprudential contexts require so. Golanski also argues that many instances of Dworkinian TDs are “institutional” rather than “moral” in character: Even in Riggs (…) the outcome-determinative disagreement did not center on the moral issue. The entire appellate panel agreed that morality would frustrate Elmer’s scheme. The judges disputed whether the case should be decided in accord with the morally required outcome. They struggled to delimit their theoretical disagreement to the standard by which to weigh the presumed collective intention of the legislators against the precision of their stat utory language regulating the making of testamentary documents. (…) Even for the limited range of cases arising from the unsettled construction of constitutional or statutory clauses, and in which different interpretive methodologies are available, the issue of whether prior legal assertions and stipulations are sufficiently directed to the question typically takes pri ority over, and often preempts, any theoretical debate concerning interpretive approaches.²⁸

Both Shapiro and Golanski accept the possibility of Dworkin’s moral TDs when offi cials disagree about whether some moral principle is included in the grounds of law. However, they insist that substantial moral questions with regard to the grounds of law need not be asked in institutional contexts in which TDs typically arise. Institutional forms of TDs are more basic than the moral ones.²⁹ The third theory is Tuori’s “critical legal positivism”. Tuori argues that law is a multi layered phenomenon in which social practices are combined with normative thought. Law is never exhausted by concrete legal materials (individual regulations and court decisions) but also includes “sub-surface” layers: legal culture and deep structures of the law. With this in mind, Tuori describes problems of traditional legal positivism.

[T]raditional positivism has to abandon either the strict separation between the “Is” of empir ical social facts and the legal “Ought” (Hart) or presuppose at the top of the hierarchically -structured posited legal order a non-positive, hypothetical norm (Kelsen).³⁰ A plausible positivistic alternative “should be capable of providing a solution which does not include the assumption of universal and immutable normative principles”.³¹ Since the positivity of law “entails that the substantive limits of modern law, as well as the yardsticks for its legitimacy, have to be found within the positive law”,³² any admissible theory of law must include “the possibility of an imminent normative criti cism of positive law”.³³ Tuori distinguishes between the descriptive social and norma tive layers of law by “emphasising the constant interaction between the law as a sym bolic normative phenomenon and the legal practices producing and reproducing this phenomenon”.³⁴ Tuori does not refer explicitly to ATD, but his position is a direct answer to Dworkin’s challenge. Critical legal positivism argues for an institutionally imposed framework for TDs; here, opposing positions result from various forms of criticism of applied concepts and interpretive methods. Tuori distinguishes between fundamental and imminent legal criticism. The former “is suspicious of the justifiability of all law”³⁵ and “tends to renounce every form of legal regulation of society”.³⁶ This form of criticism autonom ously “draws its grounds upon somewhere else than upon the positive law itself”.³⁷ The latter form of criticism, “despite its critical nature, also contributes to the repro duction of its object: it sustains the law both as a normative order and as specific legal practices”.³⁸ However, as Tuori argues, this criticism cannot be of a “fundamental” kind: there are limits of admissibility which are determined by legal culture and law’s deep structure.³⁹ A “positivistic” label for this account refers to the limiting premise that law is primarily an institutionalized practice.

All three theories rely on basic shared intuitions about law’s relation to moral ity, law’s positivity, and its institutional character. They also point to legal culture as important background. A detailed analysis of these theories leads to the con clusion that even if questions about substantial moral grounds of law are excluded by default, they do not deny the possibility of further reflection about the questions. In certain circumstances, institutional TDs could transform into moral TDs. I argue that such transformation is backed up by a commitment to pragmatic holism on the methodological level. At the end of this section let me elaborate on one more feature of the refined ver sions of positivism indicated above, namely their flexibility or spontaneity. Positivistic theories of law are by design “minimalistic”. This is manifested in the claim that the idea of the rule of recognition reflects some kind of “common meaning” of the word law.⁴⁰ Thus, the intuition that most legal cases are “easy cases” is reflected in the positivist theory of the “obvious law”,⁴¹ seen as a default theory accepted by continental lawyers. Tuori explains how such a default shared understanding constitutes an important part of the identity or consciousness of most lawyers: The positivity of modern law corresponds – so I venture to maintain – to how a typical (contin ental European) lawyer conceives of the law. [This] formalistic narrative corresponds to the average self-understanding of (continental European) lawyers, although they may only rarely feel a need to make it explicit.⁴² He adds that this minimalist positivist theory of law that is reflected in typical lawyer’s consciousness is “spontaneous”: The account of the law attributed here to the typical lawyer represents a kind of spontaneous positivism. Positivistic legal theory can be understood as the reflexive level of this spontan eous positivism: positivistic theory has given an explicit and systematic expression to the self-understanding of the typical lawyer.⁴³ The fact that the default theory is “spontaneous” implies that it is weakly justified, even though it is the best general conceptual tool at hand; however, it not only fails to exclude other alternatives – it seems to require them! Such default conceptual basis is practice -oriented, so whenever the default conceptual setting produces recalcitrant conclusions, it would be wise to refine it. Spontaneous solutions are good for the time being – they are not eternal truths. Spontaneity is at root pragmatic: if you are unsure about which theory would work in given contexts, trust your gut and refer to good traditions. But at the same time, beware of becoming a dogmatist!

4. Jurisprudential holism

Traditional theories of law cannot both be subject to TDs and at the same time fully accommodate TDs. Such a view is acceptable only on the condition that there is a strict distinction between a theoretical subject level and meta level. Although I have already

appealed to that distinction, I do not think that it is a priori or metaphysically solid. Rather, I take it to be a pragmatically justified way of speaking, a provisional tool used to compare theories. In this context, Dworkin denies this distinction. He claims that there is no divide between theory and meta-theory of law. This is a result of his acceptance of holism, the same methodological attitude that was embraced by his teach ers at Harvard, namely John Rawls, Willard Van Orman Quine, and Morton Gabriel White. Thus, Dworkin thinks that theories are “webs of beliefs” or reflectively organized wholes. In one of his latter works, he wrote:

If we are better to understand the non-instrumental integrated values of ethics, we must try to understand them holistically and interpretively, each in light of the others, organized not in hierarchy but in fashion of a geodesic dome. We must try to decide what friendship or integrity or style is, and how important these values are, by seeing which conception of each and what assignment of importance to them best fits our sense of the other dimensions of living well, of making a success of the challenge of a living life. Ethics is a complex structure of different goals, achievements, and virtues, and the part of these plays in that complex structure can only be understood by elaborating its role in an overall design fixed by the others. Until we can see how our ethical values hang together in that way, so each can be tested against our provisional account of the others, we do not understand any of them. Two of the most overworked of philosophical images are nevertheless apposite here. In value as in science we rebuild our boat one plank at a time, at sea. Or, if you prefer, light dawns slowly over the whole.⁴⁴ The domain of ethics which, according to Dworkin, is tied to the domain of law since every legal decision as an imposition of coercion has to be justified. In reflective justi fication normative moral judgments are set as default premises. But it is not necessary to commit to the Dworkinian view even on the assumption that normative reasoning is essential for determining and/or justifying legal outcomes. Positivists may well accept this general “normativity thesis” and claim that there are possible theories of the grounds of law that include normative premises that are not moral. The debate here may resem ble the one between neo-pragmatists, some of whom settle for merely epistemic and superficial normativity (Quineans), whereas others argue for a robust moral component in our everyday thinking (Whiteans, Putnamians). Subject-level positivistic theories that limit themselves to the presentation of a set of obvious descriptive statements about socially determined grounds of law may enter the domain of reflective thinking as normative premises applied in “normal circumstances” of legality. Positivistic statements about grounds of law in the reflective process of a judge facing a TD may be treated either as 1) sufficient and necessary or 2) insufficient or unnecessary. In the first case, the positivistic criteria (the theory of grounds of law) may compete with other such criteria provided by alternative, competing theories. In the second case, such criteria may be either consumed by another theory (as insufficient) or deemed irrelevant (as unnecessary). Dworkin, who himself defends a reflective attitude in law, accepts the positivistic criteria in the second sense. These criteria may be but need not necessarily be imple mented into the holistic, normatively structured web of beliefs since there may be other theories (natural law) which establish the grounds of law that are descriptively better in most circumstances. Although Dworkin obviously needs some criteria of this kind

– a preconception of law to be applied at the pre-interpretive stage – he is ready to aban don it in the process of ethically-oriented reflective thinking. According to Dworkin, legal decisions cannot ever be made without indulging in moral considerations. Legal positivism as a theory that tries to exclude moral considerations and establish descriptive grounds of law as a starting point for fruitful legal reflection is not sufficient to provide fruitful legal outcomes. Positivism can be a starting point but can never be accepted as a conclusive theory of the grounds of law. Dworkin’s reflective judge in resolving a hard case can at best recognize legal positivism as a partial or undeveloped theory. The main point here is that Dworkin himself does not take the TD with legal positiv ists seriously. If a theory of law has to present the grounds of law that would justify legal decisions and make propositions about legal duties and obligations true, then the only theories with which one can seriously disagree are also normative, reflective and interpretive in character. Traditional descriptive legal positivism is not a partner in such debate!

The meta-interpretive versions of legal positivism, like those presented by Shapiro, Golanski or Tuori, do not merely state what the grounds of law are but also what kind of reflection could be invited to determine certain legal outcomes. These theories are comprehensive and holistic. Shapiro’s plans may allow different interpretive method ologies and developments, but they are not sacrosanct and may be abandoned if they are thoroughly wrong and not prone to reinterpretation. It is questionable whether Dworkin’s theory is ready to make any concession with respect to the method of legal reflection. TDs may be about “the soundest interpreta tion of legal practice”, but Dworkin seems to think that there is only one such interpret ation. His theory is based on a very peculiar view of what the soundest interpretation is. Apart from calling for quite pragmatic methods of reflective thinking in law and a focus on substantial disagreements, Dworkin tends to be the greatest “steadfaster” amongst contemporary legal scholars. Thus, his theory faces a paradox: ATD is intro duced to denote subject-level versions of legal positivism but Dworkin’s own theory cannot accommodate certain types of TDs. ATD cannot be invoked when it is argued that only a narrow scope of theoretical positions is acceptable (like Dworkin’s own morally-oriented interpretive theory of law). Ultimately, a TD is a disagreement between rival theories of law. Theories cannot be treated as serious rivals or peers if they are deemed inferior immediately after the statement that TDs are an important phenomenon in legal practice. The initial power of ATD comes from its focus on the philosophy-laden conflict before the court. This was Dworkin’s main point when he wrote his famous phrase: Any practical legal argument, no matter how detailed and limited, assumes the kind of abstract foundation jurisprudence offers, and when rival foundations compete, a legal argument assumes one and rejects others. So any judges’ opinion is itself a piece of legal philosophy, even when the philosophy is hidden and the visible argument is dominated by citation and lists of facts. Jurisprudence is the general part of adjudication, silent prologue to any decision at law.⁴⁵ ATD is a significant self-standing argument which invites a deeper, reflective and hol istic attitude towards legal thinking. I think that positivists who discuss it seriously,

such as Shapiro, Golanski and Tuori, understand it in this way. Their efforts are con sistent with an overall holistic programme, although they do not explicitly refer to it (as Dworkin does).

5. Reflective legal positivism

Shapiro’s book Legality is in part an elaboration over the division of trust within the US legal system after the Civil War. Shapiro states:

The Fourteenth Amendment was based on a different, and conflicting, set of relative judg ments. Insofar as traditional attitudes toward the trustworthiness of the state legislatures and their ability to protect basic liberties could no longer be maintained after the Civil War, the Fourteenth Amendment not only imposed substantive limitations on state legislation through the Equal Protection Clause, but also empowered Congress to protect equality and other constitutional rights through federal legislation. The Fourteenth Amendment thus assumes attitudes of trust that are contradicted by the attitudes of trust that underlay the first twelve amendments. How should such conflicts be resolved? My suggestion is that we approach the synthesis of conflicting trust judgments in roughly the same manner that philosophers of science treat the revision of inconsistent theories. For example, when a scientist faces highly credible evidence that contradicts an accepted theory, she will be forced to adjust her theory in light of the recalcitrant data. The generally accepted method in the philosophical literature for synthesizing these conflicting elements is called “minimal” revision: the scientist ought to give up as little of the theory as possible in order to achieve consistency. Thus, the scientist should attempt to retain the central premises of the theory – the “hard core” in Lakatos’s helpful terminology – but jettison more peripheral elements in order to reestablish consistency. It may not, of course, be possible to keep the hard core intact, in which case certain central elements will have to be revised away. But in most situations, such revolutionary change will not be necessary, and adjustment of more marginal elements will ensure a consistent synthesis.⁴⁶ The second part of this quote, which refers to “the synthesis of conflicting trust judgments”, is the most important one. Although Shapiro points to Imre Lakatos’s “hard core” and “periphery” terminology, it is accurate to frame his argument in more pragmatic and holistic terms. The programme of “holistic pragmatism” was consciously applied by Rawls in his method of “reflective equilibrium” and defended by White, who extended Quine’s version of scientific holism to other domains of culture (including the normative domains of ethics, politics and law). According to White, the holistic method boils down to the following set of theses. That doctrine was succinctly formulated by Quine in his famous 1951 paper, “Two Dogmas of Empiricism”, when he wrote: “Each man is given a scientific heritage plus a continuing barrage of sensory stimulation; and the considerations which guide him in warping his sci entific heritage to fit his continuing sensory promptings are, where rational, pragmatic”. In several ways, this statement is especially significant. First of all, it is about the behaviour of human beings and their heritage, and is for that reason about a cultural phenomenon. Second, a scientific heritage is regarded as a conjunction of many beliefs rather than as one noncon junctive belief, thereby indicating that the view is holistic. Third, the reference to a barrage of sensory stimulation or a flux of experience indicates the empiricism of the view. Fourth, the reference to the pragmatic warping of a scientific heritage to fit sensory promptings shows that the view is in the tradition of pragmatism. According to holistic pragmatism, scientists’

warpings are carried out with concern for the elegance or simplicity of the theory they adopt and with the intention to warp the heritage conservatively – that is, by engaging in what James calls minimum modification of it and what Quine calls minimum mutilation of it.⁴⁷ This method can be applied in the legal domain. Its unconscious applications can be seen in the thought of American legal realists who, at least to a certain extent, relied on the pragmatist, holist, and empiricist ideas of John Dewey or William James. It was James who wrote: “[g]iven previous law and a novel case the judge will twist them into fresh law”,⁴⁸ which is a reference to the same procedure by which a scientist tries to work hypotheses and experimental results into a manageable (i.e., testable) structure of belief. Just as the scientist begins with a conjunction of previous truths, encounters fresh facts, and incorporates fresh truths or revises the old ones, a legal official, espe cially a judge, relies on previous truths and the facts of a case to draw consequences from them by use of certain methods of application and interpretation. However, if the consequences are unacceptable (“recalcitrant experience”), then the whole set of premises is open for revision. It seems to me that the aforementioned “reflective legal positivists” are open to accepting such a holistic attitude in a legal decision-making process.⁴⁹ For them, the positivist concept of law is merely an old truth, an obvious premise or a body of such premises in a complicated reasoning. However, it seems that in novel circumstances or in cases in which rival theories (alternative grounding assumptions) are intelligibly presented, those positivists feel free to deal with these recalcitrant experiences in ways that fit into the boundaries determined by the general institutional structures. Only then do the revisions and readjustments remain intelligible. Recalcitrant legal experi ence occurs when the acceptance of obvious assumptions leads to unacceptable con sequences. It is usually on those occasions that TDs become visible and meaningful. In legal academia, we usually force such recalcitrant experiences in order to discuss different theoretical possibilities with our students. In legal practice, such experiences are rather forced upon parties and judges. Nonetheless, in law, just as in science, there is a tendency to save established theories and recognized solutions. TDs are often solved by applying a pragmatist “minimum mutilation” rule, that is, by embracing the most popular theory with certain necessary adjustments. This is the reason TDs are often presented as disagreements between various subject-level theories rather than as complex theories that operate across levels. Judges try to be predictable. An import ant upshot of the discussion of ATD is, however, that there is a crying need for legal officials, practicing lawyers, and legal academics to realize the “reflectivity” of legal

sources, legal theories, and legal methods. Predictability does not mean being blind: Themis may be blindfolded, but she knows the law. Since the very concept of law is philosophically disputable, jurisprudential work requires achieving holistic, reflective equilibria.⁵⁰ The positivistic theory is a forceful premise in complex legal deliberation. However, a legal thinker is free to give it up if it produces unfortunate consequences. It is a default premise because it has certain practical virtues, such as consilience and simplicity, and is widely accepted. But a different simple descriptive theory can be imagined which could substitute for it as a “ruling theory of law” at a given time and place. In such circumstances, the “institutionalized” account of TDs would not differ much, but it would not be positivistic. In this light, the reflective versions of legal posit ivism do not betray basic philosophical positivistic assumptions but rather contextualize them. An interesting question is which theories laypeople and professionals are ready to accept as peer rivals for such default legal positivism. As indicated earlier, I believe that shared, common beliefs saturate the institutional structures of legal order within which various kinds of TDs may appear.

Golanski wrote: When a legal case begins, or when a legal issue arises, the first step for the parties or the court is to discern whether the existing legal materials – prior decisions, enactments, and so forth – point the way ahead. If the answer is clearly “yes”, law is likely, but certainly not strictly compelled, to accept that outcome and resolve the matter. If not clear, the court will summon some manner of persuasive authority. But either way, the controversy that defines the case will at the outset be characterized by a claim that one outcome or the other is supported by existing institutional facts which are directed at the new situation.⁵¹ The general truth of this statement would not be denied if the sources or grounds of law were described differently by any other popular non-positivist theory of law. Any such theory could play the role of the default premise. In case of recalcitrant experience, judges and parties might indulge in reflection, which may easily lead to TDs. TDs are inevitable and should be recognized as an important legal phenomenon. Recognizing this would positively influence the quality of legal decisions. There is, however, one condition that has to be met if we are to avoid a paralysing conclusion that “anything goes” in jurisprudence. Namely, in order to understand law as a philosophically com plicated discipline and accept judges as reflective, philosophically oriented thinkers, society must itself become well educated in philosophy and be skilled in reflective, critical thinking. It is not my role here to discuss whether that condition is realistic enough to ever be met.