Skip to navigation – Site map

HomeNuméros28Legal MethodologyLegal disagreements

Legal Methodology

Legal disagreements

A pluralist reply to Dworkin’s challenge
Lorena Ramírez Ludeña
p. 11–32


In this paper I analyse the problem of legal disagreements, initially raised by Ronald Dworkin against Hartian positivism. According to Dworkin, disagreements are pervasive, since law is an argumentative practice in which participants invoke normative arguments. Positivists, who claim that law depends upon agreement among officials, have difficulties to make sense of the fact that lawyers frequently disagree. I first present the main arguments in the debate. I then go on to distinguish different levels at which lawyers disagree. Taking these levels into consideration, I articulate a pluralist reply that shows that the fundamental positivist tenets remain untouched by Dworkin’s challenge.

Top of page

Full text

1 Introduction

  • 1 Hart 1994: 82 and ff.

1There are widespread philosophical beliefs about law that are seemingly beyond question. It is commonly accepted that, in order for a legal system to exist, certain social facts have to obtain. However, many characterizations of these changeable facts have been attempted. According to the Hartian model, law is dependent upon a convergence in certain individuals’ conduct and attitudes. In particular, officials share the same criteria to identify the law of their legal system, and they are committed to them.1 At the same time, it is also difficult to dispute that there are disagreements among lawyers, for example about the interpretation of the law or the relevance of morality to deciding cases.

2The problem of legal disagreements as sketched by Ronald Dworkin seeks to identify the difficulties that positivists such as Hart face when they attempt to maintain the conventional nature of law (in the sense mentioned) while recognizing the fact that there are disagreements – at least apparently – about what the law establishes. The problem for positivists is even worse, since they defend that the beliefs and attitudes of participants must be taken into account and, in many cases, they see themselves as having meaningful disagreements regarding the law. However, the very fact of the officials’ disagreement shows, according to Hart, that there is no unique legal answer for the case. The idea of parties disagreeing on which answer the law really requires would make no sense.

  • 2 See Marmor 2009 and Vilajosana 2010.

3It is not easy to determine in what way positivism is committed to a conventionalist conception of the law.2 In fact, it can be said there are as many “positivisms” as conventionalist theories. I will focus on those conceptions that take law as depending on the existence of a convergence of certain kinds of conduct and attitudes. My main aim is to examine the merits of Dworkin’s critique of legal positivism. I first describe the main arguments that have been offered in the debate over legal disagreements. I then distinguish different levels at which officials disagree; I conclude the paper with the presentation of a pluralist reply to Dworkin’s challenge.

2 The Hart-Dworkin Debate

4The Hart-Dworkin debate has persisted for quite some time, changing course many times and involving a wide range of scholars who disagree even on the very object of the dispute. However, it is unquestionable that the problem of disagreements in law is one of the main aspects of the debate.

  • 3 Although Shapiro (2007) and Leiter (2007) identify it as a new challenge, Leiter argues that it is (...)
  • 4 I do not want to deny that the kinds of disagreements emphasized by Dworkin have changed. In this s (...)

5Some authors held that the problem does not have the importance that Dworkin and others claim it has; others argued that it is indeed an important challenge for positivism.3 I claim that it is a very important critique, deserving of a reply from positivists; however, I argue that Dworkin’s challenge has remained constant: what has changed are the explanations he has offered as to why positivism is incapable of reconstructing hard cases on which officials disagree.4

  • 5 Dworkin 1977: 17 and ff.
  • 6 Dworkin 1977: 81 and ff.

6In Taking Rights Seriously, Dworkin points out that, if we have a look at what actually happens in a courtroom, it is not difficult to notice the fundamental role of principles in resolving cases. Positivists, who assume that legal norms are identified by criteria related to their pedigree, cannot accommodate principles because their legality depends on substantive considerations.5 Dworkin claims that individuals have rights that do not depend for their legality on their prior recognition by legal systems, and that it is the judges’ duty to guarantee them. They have to determine who has the right to win the case by using arguments based on principles. In contrast to rules, principles are non-conclusive standards that are quite often in conflict. Their application depends on the relative weight of the principles involved and, according to Dworkin, there is always a right answer provided by the correct balance of the values at stake. So, lawyers often disagree about what the law establishes even if one side in the disagreement is wrong.6

  • 7 Dworkin 1986: 45 and ff.
  • 8 Dworkin 1986: 93.

7Some years later, Dworkin further developed many of his previous ideas in Law’s Empire, emphasizing the interpretative nature of legal practices. Legal argumentation is creative and constructive. Much as it happens in the case of art, law requires the exercise of creative interpretation, in which participants attempt to interpret something created by individuals as an entity distinct from them. For that reason, determining what the practice requires is different from determining the participants’ beliefs. Legal reasoning is also an exercise of constructive interpretation. It is a matter of imposing a purpose on an object or practice in order to make it the best possible example of the form or genre to which it is taken to belong. Dworkin maintains that that purpose determines what the practice requires. However, it is not the case that “anything goes”, because the way in which a practice or object has previously developed restricts the number of possible interpretations.7 Following Dworkin, the purpose of law is the justification of state coercion by law, taking into account individuals’ rights and responsibilities that derive from past political decisions.8

8The disagreements that make the argumentative nature of law clear are not restricted to general disputes about whether morality is a condition for legal validity, nor are they restricted to the role of principles in legal reasoning. They are also apparent, for example, when participants advocate for different ways of interpreting legal statements. Why is legal positivism unable to offer an accurate reconstruction of these cases? According to what Dworkin emphasizes in his later work, the problem is that, for positivists, the truth of legal propositions (statements concerning what the law of a particular legal system establishes) depends exclusively on certain historical facts that constitute the grounds of law. Positivism may be able to reconstruct empirical disagreements about whether certain historical facts obtain, but is unable to account for disagreements about which elements are legally relevant. Therefore, they cannot offer an accurate reconstruction of theoretical disagreements, and they are forced to understand them as disagreements about what the law should be.

  • 9 Dworkin 1986: 31-37.

9Dworkin claims that legal theorists often rule out the possibility of theoretical disagreements because they subscribe to semantic theories of law, which see meaning of words as dependent upon shared criteria. In law, this means assuming that officials use the same criteria in order to consider something as law. According to Dworkin’s reconstruction of positivism, the very meaning of the word “law” makes law dependent on shared criteria and, therefore, agreement on the grounds of law is fundamental. This is the well-known semantic sting argument, introduced by Dworkin to explain why positivists require that the criteria that we employ to determine what counts as law be established by agreement.9 People agree, if they are competent in the use of the term, about the grounds of law. Controversies would not make sense because people would be using the same word (“law”) with different meanings and, if meaning determines the object of controversy, they would be arguing about different things.

  • 10 Hart 1994: 81.
  • 11 Hart 1994: 15.
  • 12 Hart 1994: 246.

10If we take into account Hart’s work as a whole, Dworkin’s critique generates some perplexity: Hart did not try to provide a definition of the word “law”, but rather an analysis of the concept of law,10 and he explicitly rejects semantic theories that relate words to necessary and sufficient conditions.11 Moreover, although he claimed that every legal system has a rule of recognition that specifies the criteria for identifying the law, he clarified that it is not part of the meaning of the word “law” that the rule of recognition is present in every legal system.12

  • 13 That is, it concerns the question of how to determine the semantic content of words and expressions (...)

11Dworkin’s critique may be responded to by using two groups of arguments: 1) he associates Hartian positivism with criterial semantics, but there are no elements in Hart’s work that lead us to that conclusion and there are alternative semantic models which are more plausible than the one attributed by Dworkin; 2) the connection that Dworkin establishes between the semantic position (or, strict speaking, the metasemantic position13) regarding the word “law” and the criteria for considering something as law is questionable.

  • 14 Regarding the relevance of a family of descriptions, see Searle 1958 and Strawson 1959. Supporting (...)

12If we focus on the first question, Dworkin’s critique seems easy to refute: positivistits do not reach their conclusions by analysing the meaning of the word “law”; they attempt to understand the nature of law, and often proceed by analysing the concept of law. Even if we accept that positivism is concerned with the analysis of the word, Dworkin’s conception may be objected to on the grounds that he attributes a naïve criterial model to positivists. This model assumes that we associate some descriptions with certain words, that they are transparent to us, and that they determine the objects to which the words apply. This is too demanding with respect to the knowledge that speakers have, since their knowledge is often very poor and is frequently wrong. Moreover, the model cannot make sense of disagreements, given that individuals would be associating the same word with different descriptions, which would refer to different objects. In contrast, there exist more plausible models, for instance, the model that claims that we possess a family of (indeterminate and vague) descriptions and direct reference theories, which reject the necessity of descriptions to refer to objects.14

  • 15 Coleman-Simchen 2003: 8.

13Even if we were to accept that the positivist model is concerned with the analysis of the word “law”, and that positivists hold something like a criterial model, Dworkin’s critique still would not be well grounded. He assumes the existence of a relation between the conception that positivism holds regarding the word “law”, and the criteria used by officials in order to consider something as law. That relation is, however, problematic: determining the meaning of the word is a different undertaking from determining the grounds of law of a particular legal system. There may be a disagreement about the criteria that determine the extension of “law” without there being any disagreement with respect to the grounds of law, and vice versa. For example, linguistic practice in the United States may agree on the meaning of the word “law”, understanding (for example) that it is a group of principles and norms, which express an idea of justice and order, that regulate human relations, and whose observance may be imposed coercively. Nonetheless, individuals may disagree about the grounds of law of the American legal system. It is also possible that the opposite occurs: widespread agreement about the grounds, but significant disagreement about which is the meaning of the word.15

  • 16 Shapiro 2011: 13 and ff.
  • 17 Raz 2001: 6 and ff.
  • 18 Raz 2001: 14 and ff.

14Let us consider that positivism is trying to capture the main features of legal practice by analysing the concept of law. Dworkin could be read as claiming that positivism understands that individuals share the concept that they are trying to elucidate, and that to share the concept requires sharing the criteria for applying it. In the case of law, this means the necessity of convergence about the grounds of law of particular legal systems. This requires assuming three theses: 1) conceptual analysis is based on the identification of shared criteria for the application of the concept; 2) people cannot disagree about the criteria that they share; 3) the criteria for applying the concept of law are the criteria that officials use to identify the law. Although it may be controversial in its details, conceptual analysis is not based on the elucidation of shared and transparent criteria about the concept of law. If it were so, its job would not be very different from linguistic analysis. Positivists try to clarify the main features of law, taking into account obvious truths about law and elaborating theories that accommodate them in a kind of reflective equilibrium.16 In addition, a good explanation of the concept does not require a commitment to an analysis in terms of necessary and sufficient conditions for its application, but instead often emphasizes features that are not unique to it, defeasible conditions, and it offers a reconstruction of its relation to other concepts.17 Even if we accept the relevance of shared criteria, individuals may disagree given the opacity and the anti-individualist character of those criteria.18 In fact, it is very usual to hold that an individual has acquired a concept even if she has very deficient knowledge about its criteria of application. So, it can be said that Dworkin mischaracterizes the idea of criterial semantics and criterial explanations. Finally, it is at least strange to say that the shared criteria for using the concept are the criteria that determine the grounds of law. Individuals from two different legal systems may employ different criteria to identify the law, but may share the same concept. And although we may sometimes assert that an individual has acquired the concept of law, it is possible that she does not know which criteria determine what counts as law in her legal system.

  • 19 I do not want to assume a specific form of conceptual analysis, nor to attribute it to positivism. (...)

15The semantic sting argument, then, is not a problem for positivists. Convergence in the identification of the law is relevant not because of the kind of word or concept involved (criterial or not), but because it is the result of conceptual analysis. Using conceptual analysis, positivism concludes that convergence in the identification of the law is relevant, but this does not follow from the claim that words or concepts require the existence of shared criteria.19 Let us now turn to analysing the critique that positivism, which requires convergence, is unable to provide an accurate characterization of disagreements in the criteria of legal validity, the sources of law, and its interpretation.

3 Three Replies

16If we take into account legal practice, individuals seem to disagree about what the law claims. They do not typically disagree about empirical questions, nor do they seem to disagree about how to decide cases when the law does not provide a solution or when they consider that the answer that it provides is unfair. Positivism, which concedes a central role to convergence, seems unable to properly reconstruct these cases. Many attempts have been made to respond to these criticisms while preserving the basic positivistic tenets. I will briefly analyse three of them.

3.1 Disagreements Are Marginal

17In order to determine whether the problem posed by Dworkin is powerful enough to undermine positivism, it is important to note that nearly all of our daily actions are regulated by legal norms. Even if this kind of “intuitive statistics” may be seen as problematic, it is hard to argue that the number of disputes that end up in court is very small if we compare it with our relationship to norms.

  • 20 Leiter 2007: 1228 and ff.; Vilajosana 2010: 173 and ff.

18Many cases that are discussed before a court of law concern problems of proof or other considerations that cannot be characterized as disputes about the grounds of law. Parties sometimes use the legal system as an instrument to preserve their interests, for example to delay a payment. Only in a limited number of cases, especially in procedures that reach the higher courts, do disagreements arise that appear to involve the question of what the law establishes.20

19Moreover, not all disagreements concerning what the law requires are problematic for positivism. The relevant disagreements (which would throw into question the idea of convergence defended by Hart) are only those that take place between officials. That is, although other individuals, for example attorneys, use shared criteria to identify the law and may disagree about them, they are not part of the relevant practice, and, as a result, their arguments do not show that the practice does not exist. In addition, the fact that a group of individuals who are officials argue about the law does not in itself undermine the conventional nature of law, which requires a generalized practice of recognition, not unanimity.

  • 21 Leiter 2007: 1239.

20Therefore, one line of response to Dworkin’s critique is to claim that, given that positivism to a large extent clarifies the phenomena, we should not abandon it because of what turns out to be a minor problem. When we compare the explanatory importance of different theories, it is preferable to choose the theory that offers a simpler explanation, that explains more aspects of the phenomena, and that leaves well-established beliefs and theories untouched. The fact that a theory does not explain all the facts does not commit us to abandoning it.21

  • 22 Ratti 2008: 308 ff.

21It could be argued that many disagreements that are seemingly about what the law establishes are in fact disagreements about how to interpret legal texts, and that these disagreements do not threaten positivism, which only requires an agreement on the legal sources. Consequently, even if it were claimed that disagreements about the interpretation of the law are frequent (because there are multiple canons that allow multiple interpretations, and frequently the rank between them is not pre-established) it could be replied that such disagreements do not threaten positivism. In this way, the number of disagreements that need to be explained is considerably reduced.22

22I claim that some level of agreement about the content of the sources is necessary in order to make sense of the function performed by the sources and the criteria for identifying them. I also claim that disagreements about interpretation are not as widespread as is sometimes assumed.

23On the one hand, the existence conditions of a legal system cannot be exhausted by the practice of identifying texts without a critical-reflexive attitude toward certain ways of attributing meaning to them. Let us imagine that an individual considers the U.S. Constitution to be part of the American legal system but also believes that its meaning should be determined by using a computer program that assigns meaning by chance. Let us imagine another individual who believes that the content depends on what her son says. Finally, a third individual understands that what is expressed by the Constitution depends on ordinary language. Would we say that there is an agreement among these individuals regarding the fact that the Constitution is part of the law? If these individuals hold an interpretation radically different from the others in relation to what the Constitution establishes, invoking it becomes superfluous. In other words, if interpretive activity is not constitutive of legal activity, convergence with respect to sources may bring about the same results as its absence (a complete disagreement about which norms are valid) and so may be entirely irrelevant. It seems, then, that the convergence characteristic of the positivistic model would be deprived of sense were there no agreement about how to interpret the sources.

  • 23 Guastini 2011.
  • 24 The most accurate description of the situation is that ordinary language plays a fundamental role i (...)

24On the other hand, those descriptions that emphasize the availability of several interpretative options exaggerate the controversial character of legal interpretation. It has been claimed that the existence of a plurality of interpretative instruments in the legal systems we are familiar with means that the interpreter has discretion to choose between several possible norms.23 I think it very difficult to question that, in contemporary legal systems, there are many interpretive instruments, that they depend on the practice of interpreters, and that they can change if the practice changes. However, if we consider every individual judge from the synchronic perspective, there are interpretations that are correct, and others that are not. To argue that there is always a framework of possible interpretations, and that judges always have discretion, is a distortion of legal practice, which exaggerates the controversial character of some cases.24

25For these reasons, it is necessary to take into account disagreements about interpretation, but this should be done without exaggerating the number of disagreements that the different instruments produce.

  • 25 Shapiro 2011: 290.
  • 26 It may be argued that the number and importance of disagreements is irrelevant for Dworkin because, (...)

26Positivism, then, would not be fatally weakened by Dworkin’s critique because it sheds considerable light on legal phenomena. However, in attempting to gauge the significance of disagreements one must take into account not only their number, but also their relevance to the legal system as a whole and to the matter at hand. Moreover, Dworkin’s critique threatens one of the main tenets of positivists, since they emphasize the relevance of convergence as a central feature of legal systems.25 If the Dworkinian critique sheds light on a relevant feature and threatens one of the central arguments of the positivistic model, the small number of disagreements does not render the problem marginal. Consequently, to retain positivism as a good theory of law it is necessary to look for plausible alternative explanations for these cases. Moreover, positivism should be tested against the possibility of pervasive disagreement in law, not merely against factual limited disagreement. However, the small number and significance of disagreements is not entirely inconsequential, since it facilitates the discovery of alternative explanations for these limited number of cases, as we will see in the next section.26

3.2 They Are Not Genuine Theoretical Disagreements

  • 27 Leiter 2007 and Vilajosana 2010: 173–175.

27Two main strategies have been employed to defend that disagreements are not genuine disagreements about the law. On the one hand, it has been claimed that those individuals that disagree are mistaken. On the other hand, that they are disingenuous in that they are aware that there is no right answer, but are trying to conceal what are essentially normative arguments.27

28An error theory with a general scope has been defended in moral theory. To reconstruct certain cases that take place in law as cases of error seems to be somewhat easier than holding that every moral judgement is somewhat mistaken. It is not implausible to say that sometimes individuals are wrong because they believe there is a legal answer where there is none. In other cases, individuals seem to be aware that the solution they have chosen is not the one established by the law. Officials decide according to what they think the law should be, because they do not like the law, or because there is no law.

  • 28 Shapiro 2007: 42.

29Still in need of explanation, however, is the issue of why disagreements take place as if they were genuine. The obvious answer seems to be that those who are in error are not aware of it, and that those that are disingenuous do not want to acknowledge the real situation. If this is so, why have these facts not been discovered yet? A possible answer may be that individuals do not have much knowledge about law, they may feel intimidated by those they consider to be experts, or they may just defer to them. However, lawyers who are not judges are also part of the debate and are aware of it, which makes it more difficult to accept that disagreements about law are always cases of error or disingenuity.28 Even if many cases may be said to be cases of error and disingenuity, other cases require an alternative explanation. To understand that error and disingenuity explain all the cases offers an image of the practice that many participants would reject. Therefore, it cannot be a good explanation, at least as a matter of internal analysis, because it does not take into account the participants’ perspective.

3.3 Positivism Can Account for Theoretical Disagreements

30Many cases can be said to be located in the penumbra of the rule, where the judge has discretion to decide. That does not imply that the decision is arbitrary, only that the law does not provide a unique solution. For example, at the level of the sources of law, it may happen that, in a specific legal system, individuals believe that the results of the activities of parliament are law, but they may doubt whether those decisions can bind future parliaments.

  • 29 Vilajosana 2010: 173 and ff.

31On the other hand, if disagreements about the identification of the law were widespread, we would probably acknowledge that it is a pathological legal system, and perhaps even doubt that it is in fact a legal system at all.29

32However, despite the fact that not everything is always debated and that disagreements may usually be understood as marginal, sometimes disputes are about pivotal cases and they seem to represent opposing conceptions emphasizing different features of the same phenomenon. We have seen that Dworkin conceives law as an interpretive practice, in which a purpose is attributed to the practice, and that what the practice requires depends on that purpose. When officials disagree what is implicitly being discussed is what best justifies state coercion. For that reason, disagreements are intelligible and inherent in legal practice. In Dworkin’s conception, then, disagreements are not problematic, but in fact show that his theory reconstructs the practice correctly. However, they are a threat to positivism.

33Two main strategies which attempt to accommodate theoretical disagreements have been developed. They call either for a refinement of the conventionalist model, or for its abandonment.

  • 30 Bayón 2002: 76 and ff.
  • 31 For example, it is not very plausible to argue that the debate regarding what is a cruel punishment (...)

34With respect to the first strategy, deep conventionalism has been defended. Positivism seems to face a dilemma regarding the reconstruction of social rules and the (well-known) problem of following them. Both horns of the dilemma imply that it cannot offer an accurate account of the problem of disagreements. If we claim that a conventional rule is exhausted by explicit agreement on its correct application, theoretical disagreements are not intelligible, because the lack of agreement implies that there is no answer provided by the rule; if we claim that agreement extends only to which texts are important, we would be assuming a very poor conception about the relevant convention. In contrast, Professor Bayón emphasizes the relevance of agreement in paradigmatic cases, which shows the existence of public criteria that are not limited to those applications.30 According to Bayón, acknowledging that there are paradigmatic cases implies mastering a technique. This, however, requires no more than tacit knowledge of the criteria for the correct application of the rule, which need not be transparent to every individual. Generalized agreement neither guarantees that the correct answer has been identified, nor does lack of agreement necessarily imply that there is no correct answer. However, leaving aside the difficulties of relying on the existence of a conventional answer despite disagreement, these positions are committed to the claim that disagreements are about what the convention is, which does not seem to be the case in many situations.31

  • 32 Shapiro 2011: 357 and ff.

35Following the second strategy, Shapiro holds a position in many respects similar to Dworkin’s.32 Shapiro rejects supporting interpretive conventions. Like Dworkin, he believes that to make sense of disagreements, it is fundamental to take into account the purpose of the practice. Both claim that the best interpretive methodology in a legal system depends on which one better fits its objectives. However, he claims that his conception is positivistic. He does not think that the attribution of a purpose requires an exercise of moral and political philosophy, but the search for legal facts. The interpreter has to determine which political purposes the designers of the legal system tried to achieve. In order to discover these purposes the interpreter has to analyse the institutional structure and determine the objectives and values that better explain the form of the system. The correct interpretive methodology will be the one that best harmonizes with these objectives. The relevant purposes are those that explain the practice, not those that justify it, and they may be morally deficient.

  • 33 Shapiro 2011: 336 and ff.

36Shapiro claims that his motivation for insisting on the relevance of social facts is not to preserve positivism at all costs. He claims that it by paying attention to certain social facts regarding the designers of the system that we can make sense of having authorities. We use the law to try to achieve complex purposes; given the difficulties we face in determining partial objectives that contribute to the satisfaction of further purposes, the motivational deficiencies of some individuals and the incapacity that some of them have in order to develop their roles, it would be very hard to satisfy complex purposes without the law. There are therefore deficiencies related to the trust in individuals that would make it very difficult to achieve the objectives, and the law tries to compensate for them. In this sense, law enables the achievement of complex objectives determining a distribution of roles by virtue of the trust relative to the capacities and character of the different participants. For this reason, the proper interpretive methodology in a concrete legal system depends on the attitudes of trust and distrust of those who designed it. For example, literal interpretation fits better with distrust in some individuals’ ability to fulfil their role in the shared activity, than does an interpretation that concedes more freedom. Not taking into consideration the distribution of trust of the system would threaten the point of having authorities to achieve complex objectives and would likely prevent achieving them.33

  • 34 Shapiro 2011: 367.

37Disagreements are then intelligible because individuals may be discussing: a) what are the general and partial purposes of the system; b) what are the roles of the different individuals to achieve the objectives; c) what is the distribution of trust in the system; d) what levels of trust are most consistent with the different interpretive methodologies; and e) what interpretive methodologies are coherent with the purposes and distributions of the system. In these cases, disagreements are genuine theoretical disputes that depend not only on the mere determination of facts, and obey the same principles usually adopted in the elaboration and evaluation of scientific theories.34 In this way, Shapiro’s reconstruction not only makes disagreements intelligible, but also explains why they are as prevalent as they are. It also has the virtue of making room for theoretical disagreements but not as conflicting conceptions about the grounds of law.

  • 35 Shapiro 2011: 383.

38In some cases, since there are many designers, it is possible that there is no single ideology underlying the system, or that it is of such little importance that it does not determine interpretive debates. It may also be possible that the designers’ position relative to the distribution of trust is not too stable. Nevertheless, Shapiro claims, we will eliminate some possibilities if we take into account the different steps just presented, and, even more relevant to the issue of disagreements, their intelligibility does not depend on the existence of an answer. He explains why they take place and are prevalent; the existence of a correct answer is a different and contingent question.35 Some of the features related to Shapiro’s planning theory of law that have just been analyzed will be useful in formulating the pluralist reply to be introduced at the end of the paper.

4 Levels of Disagreements

39Taking into account the arguments that have appeared in the debate, it is possible to distinguish different levels at which disagreements take place. The list does not pretend to be exhaustive, but it will be made clear that, in the debate about disagreements in law, a range of arguments at different levels have been offered. As a consequence, the problem seems to be more difficult to overcome than it actually is. I will argue that it is not possible to offer a single answer to Dworkins’s critique, but different arguments that take into account the level of disagreement under consideration. Although none of the answers to Dworkin is conclusive, a combination of them may be so.

  • 36 Anyway, I do not mean to claim that positivists have presented the best theory of conceptual analys (...)

40On the methodological level, discussions have been focused on what type of concept the concept of law is. On this level, Dworkin has defended his interpretive model and has attributed the criterial model to positivism. As we have seen, positivists have opposed to that characterization by claiming, among other things, that their project does not focus on the analysis of the word “law”, or that to understand the nature of law does not require identifying shared and transparent criteria to apply the concept. In general, it may be stated that Hartian positivism has confidence in the capacity of conceptual analysis to apprehend the content of the concept of law and, in this way, to clarify the legal phenomenon.36

41Second, on the level of the central elements of law, disputes are about what law in general is. Along these lines, once we adopt an interpretive methodology, the advantages and disadvantages of defending conventionalism or law as integrity may be discussed. Or, in the framework of non-interpretive conceptions, it is possible to claim the relevance of certain social facts but debate about what these facts are. The position adopted on this level is not determined by the one adopted on the methodological level. It may be possible to defend similar positions on this level even if different methodologies are endorsed.

  • 37 Marmor 2005: Ch. 8.

42Third, on the level of abstract interpretation, it is possible to hold a general position (e.g. emphasizing the relevance of the legislators’ intentions), or a particular position in relation to some groups of cases (e.g. regarding the question of how to attribute meaning to moral terms in law). Disagreements at this level will often involve disputes about the different standards of interpretation, their content and their abstract hierarchy. There may be agreement at this level and disagreement at previous ones, but theorists quite often derive their position on this level from what they claim about the central elements of law. For example, a scholar who focuses on the authoritative nature of law will probably take into consideration the legislators’ intention.37

  • 38 In fact, Dworkin reconstructs Hart’s answer in the Postcript, where he defended the conventional na (...)

43Is the existence of disputes on these levels a problem for positivism? Dworkin seems to assume that it is. He conceives law as an argumentative practice in which individuals disagree. Not only officials disagree about the law of a particular legal system, but also theorists have an interpretive attitude towards it. This is so because, according to Dworkin, either we understand that theorists endorse semantic theories, which implies not being able to reconstruct disagreements, or they are conceived of as proposing competing normative theories in the framework of an interpretive conception about law. In this way, he holds that the debate he maintains with Hart and other theorists shows the argumentative nature of law and that the different theories of law compete on the normative level.38 The main questions to consider are whether the very existence of disputes between theorists at the previous levels shows that law is argumentative, and if the fact that Dworkin may be able to make sense of other positions within the framework of his interpretive conception constitutes an argument in favour of his position.

  • 39 I do not want to deny that participants disagree about the identification of the law because they d (...)

44Let us think about the practice of obtaining knowledge. In that practice there are individuals who, following the scientific method, develop scientific theories about the world. Let us imagine that a group of shamans say that we obtain knowledge by reading coffee grounds. In doing so, shamans believe that they are invoking gods, and that gods manifest themselves by giving different forms to the sediment, in this way allowing knowledge about different aspects of the world to be acquired. Shamans think that the practice of obtaining knowledge consists precisely in invoking gods, and they believe that scientists have developed a different way in order to do that. They could then discuss with scientists about the best way to invoke gods. In contrast, under scientific standards, the discussion with shamans makes no sense. If we evaluate both conceptions, the fact that one of them is able to explain why controversies between scientists and shamans make sense, and the other is not, is inconclusive. This example attempts to show why making sense of disagreements between theorists cannot count as an argument in favour of one theory over another. It does not count in Dworkin’s favour the fact that, in his reconstruction, positivism may be able to adopt an interpretative attitude and so he is able to make sense of disagreements between theorists. In fact, it is not surprising that legal theorists disagree. This is an invariable feature of philosophical reflection. What would threaten Hartian positivism are disagreements between participants, that is, officials, regarding the identification of the law.39

45On the three levels just presented, disagreements are about law in general. The following four levels that I will mention relate to disagreements that take place within specific legal systems.

46On the level of the identification of the law we find disputes which deal with the shared criteria of validity that are part of the rule of recognition of a specific legal system.

47There may also be disagreement about what the concrete legal sources of a specific legal system are and about their organization into a hierarchy. Here, it is fitting to distinguish between the sources-as types level and the sources-as tokens level.

48Regarding the sources-as types, it may be questioned whether customs, or precedents, are sources of law or not. It may be thought that these types of disagreements are nothing more than disagreements about the rule of recognition, but they are disagreements of a different kind. There may be agreement about the sources-as types, but disagreement about the rule of recognition. For example, it is possible for a group of judges to understand that custom is a source of law because that is what the Constitution establishes. Another group may understand that the law of the legal system is constituted by the content of laws and customs. If the Constitution were modified and no longer mentioned customs, the first group would not consider them a source of law, but for the second this fact would not change the status of customs as a mechanism to generate legal norms. This shows that judges identify the same sources but the rule of recognition they use is different.

49Regarding sources-as tokens, we should take into consideration that, even if we agree on the rule of recognition and the sources-as types, officials may disagree about whether a concrete custom is part of commercial law, for example. Let us suppose that nobody doubts that custom in general is a source-as type, but it is controversial whether the factual conditions necessary for the generation of a specific custom have been instantiated. In any case, this seems to be a purely factual disagreement, which does not affect the positivistic conception.

50The controversy may also be focused on the meaning of the sources of a specific legal system. On the one hand, there may be disagreement about which canons of interpretation are valid and how they are organized into a hierarchy. As we saw, a certain degree of agreement at this level is decisive. In addition, disagreement may be about the meaning of the sources-as tokens; that is, even in the presence of agreement about everything else, there may be a disagreement about the meaning of a concrete legal statement. However, these disagreements do not seem to be theoretical in nature. In this sense, if there is agreement regarding the criteria of interpretation, but disagreement about the content of a concrete statement, we should conclude that either there is no real agreement on the previous level (the criteria are in fact controversial), or the disagreement is factual (over whether the agreed criteria are fulfilled).

  • 40 Anyway, this dispute would be practical, about the best way to decide, and not theoretical, about w (...)

51Another level includes the solution of a specific case. It is important to make a distinction at this level because there may be agreement in meaning but disagreement about how to solve a case (for example, because it is acknowledged that there are multiple criteria of interpretation that may be considered, but there may be discussions about which way to solve a case is best40), and there may be agreement about how to solve a case but disagreement about specific criteria of interpretation (in fact, different criteria very often coincide in the solution).

5 A Pluralist Answer

52In the discussion about disagreements, a range of examples belonging to different levels have been employed. However, I think that the focus should be set on whether there are frequent disagreements among participants on the level of the legal solution of cases that arise as a result of a disagreement on the level of the criteria of validity or interpretation. This is so because those disagreements may be understood as questioning the convergence that constitutes, according to positivism, one of the main elements of law.

53If we consider what has been previously explained, it is easy to formulate a partial reply to Dworkin. First, the number of disagreements is small if we consider the incidence that law has in our lives, which facilitates finding alternative explanations to those cases in which disagreements seem to take place.

  • 41 This is only a simplification; in fact, since legal systems should be understood as a web of numero (...)
  • 42 This is, again, a simplification. It is possible to imagine a legal system with an awkward rule of (...)

54Second, disagreements about the criteria of validity are not frequent, and may be explained in a plausible way as marginal cases in which judges have discretion, which is compatible with the convergence necessary for positivism. Moreover, if disagreements about the criteria were common within a community, we would have doubts about whether it did in fact have a legal system.41 We would think that it is a pathological practice, unable to satisfy most of the functions we commonly associate with law. On the other hand, disagreements about the sources may be reconstructed, either as empirical disagreements, or as disagreements about the criteria of validity.42

  • 43 For this kind of cases, see the metalinguistic response offered by Plunkett & Sundell 2014.

55Third, regarding interpretive disagreements, one cannot avoid the problem by saying that there is convergence in the sources. Some degree of convergence regarding the correct way to interpret them is necessary for the law to accomplish its function as a guide for conduct and to make sense of the rule of recognition. These disagreements are also marginal and, if they were extensive, we would doubt whether it is a legal system. In any case, to acknowledge the existence of interpretive instruments does not entail that everything is controversial: there is broad agreement about how to solve many cases.43

  • 44 Leiter 2007: 1232.

56In this area, it may be observed that individuals sometimes invoke different interpretive criteria in order to support different solutions for the same case. Are these genuine theoretical disagreements that question the positivists’ thesis? It is important here to determine if what makes the most sense of disagreements is a reconstruction à la Dworkin, in which participants try to offer the best justification of the legal practice. If we take into account the arguments used by lawyers, and the way in which they discuss with each other, Dworkin’s position does not seem to be right. Lawyers often use different arguments to defend their position, trying to maximize their likelihood of winning the case. They accumulate arguments of different kinds, which makes it difficult to understand that they are assuming a coherent iusphilosophical position. And it does not seem that the relevant agents, judges, are essentially different in this regard. Very often they adhere to an interpretive canon that they leave aside in other decisions, without emphasizing any distinctive feature of the case that would justify the change.44 An analysis of jurisprudential repertoire shows that judges, far from engaging in theoretical disputes, present a façade of justification to ground their decisions. I think it is appropriate to argue that in problematic cases we have conventions which are eminently legal and which enable the defence of different positions. Moreover, although in these cases individuals present their opinions by claiming that law establishes that solution, error theory and disingenuity seem to explain what really occurs. Additionally, in many cases the best interpretation of what individuals say and do leads us to conclude that, even if they assume that there is no law, they believe it is part of their function to adopt a decision for the case, and they try to identify the answer that fits best with the system, the one that they think to be more defensible on moral grounds, the one that they consider to have the best consequences, etcetera.

  • 45 Kripke 1980 and Putnam 1975. I do not need to assume that those are the only theories available, bu (...)

57A final group of cases may be problematic. Sometimes the discussion has to do with the meaning of a word, in the sense that different participants in the dispute present conflicting conceptions regarding the main features of the object to which the word refers. If, according to the positivistic model, the truth value of legal propositions depends on the existence of a convergence among participants with respect to the interpretation of legal statements, these disagreements would not make sense. The occurrence of the debate would show the absence of a pre-existing answer. However, I believe that positivism may reconstruct these cases if we take into account the arguments advanced by direct reference theories.45

  • 46 This is so even if, at the end, we discover that there is not a unique essential feature. The model (...)

58In a way similar to what happens in our daily linguistic practices, some words in legal texts show certain features emphasized by the defenders of direct reference. Sometimes we use words to refer to a kind that we assume has a deep nature, and we are capable to refer even if we have deficient information about it. We are able to refer successfully since we are part of a chain of communication, which is ultimately related to exemplars of the kind, even if our beliefs regarding the objects that pertain to the kind are very poor and fallible. In order to have a meaningful discussion, individuals do not have to agree on information that identifies the object, but to take part in the same chain of communication, which goes back to instances of the same kind. Disagreements are intelligible because individuals try to discover the nature of the kind. According to the defenders of these theories, the identification of the relevant features may depend on ordinary speakers or on experts. In any case, since the determination of these features requires theorization, disagreements may be explained as competitive arguments that try to identify the fundamental features of the kind.46

  • 47 Even if the argument has only been sketched in the text, I hope to have established that direct ref (...)

59I find it difficult to question that law tries to guide our conduct and it is to a large extent expressed in ordinary language. If ordinary language is sometimes better reconstructed in the way the defenders of direct reference claim, it is reasonable to believe that sometimes the language of the law may be reconstructed in the same way. In addition, since the impact of direct reference theories depends (in general, but also in the legal field) on how these words are used, taking them into account does not threaten the convergence relevant for positivism. That is, direct reference theories will be a good reconstruction of the phenomena only in those cases in which legal interpreters share certain assumptions while using legal terms: they assume that the information they have about the object may be deficient and that the features of the object to which they refer may transcend them. In short, the conventionality of law need not imply a conventional conception of our linguistic practices that requires the existence of shared and transparent criteria.47

60In these cases, individuals are discussing about the meaning of a legal statement. Accordingly, contrary to what has been previously defended in this paper, there may be disagreements at the level of the meaning of the sources-as tokens even if there is an agreement at the previous level. We may accept that ordinary language is the interpretive standard for the case, but disagree, for example, on whether fungi are or not plants, what toxicity is, or about the nature of cruelty. These disagreements are not (merely) empirical disagreements, but theoretical.

61On the other hand, other kinds of disagreements are also intelligible if we consider direct reference theories. First, a question may arise as to whether we may reconstruct the use of a certain word according to direct reference theories or under the conventionalist scheme: to determine that individuals are using a word according to direct reference theories requires not only empirical verification, but also an analysis of our assumptions regarding the use of words, which is made evident in our reactions to counterfactual situations, and the conventional character of a word does not have to be transparent to those who use it.

  • 48 Therefore, even if it is controversial whether fruits and vegetables are natural kinds, I think tha (...)

62Second, there may be two different chains of communication, one that goes back to the common use of the word and another to the experts’ use, resulting in a dispute about which chain of communication the term of the law belongs to. The determination of the proper chain of communication requires theorization, so there may be meaningful disagreements about those matters. This would be the case, for example, if a law introduces a tax on fruit and it is debated (since it is evident that there are differences between the ordinary and the expert’s use) whether tomatoes are included in the regulation or not.48

  • 49 This case was discussed by German judges and ended in a decision by BGH (25. 10. 2006). See Montiel (...)

63Third, when the word is included in a law, it may be debated if a new chain of communication has been generated, and whether or not the meaning of the word in the law is different from the extra-legal meaning. Let us imagine that a law forbids the trafficking of hallucinogenic plants and that there is a dispute about whether hallucinogenic fungi are or not included. The fact that there are doubts regarding whether or not fungi are regulated makes it clear that it is controversial whether legal language assumes the use in ordinary language (which, in this case, defers to the experts’ use and excludes fungi) or if it may be understood that a new chain of communication, strictly legal, has taken place (and so fungi are, according to the law, plants).49

  • 50 Dworkinians may point out that, according to my reconstruction, positivism is affected by the seman (...)

64In addition, it will be frequently debated which individuals are responsible for identifying the main features of the kind. The considerations pointed out by Shapiro regarding the distribution of trust are important in determining who the relevant individuals are. For example, regarding the term “causality”, what distribution of trust is expressed in the system? In criminal law, conduct is usually regulated in a precise way and the judge merely applies the law. However, the legislator (normally) does not establish criteria for the correct application of “causality”, even if it has the opportunity to do so. Judges, on the other hand, (normally) use in their decisions the reconstructions developed by criminal law theory. They take into account several theories, assuming that as time goes by they make more sense of problematic cases, and that they are progressive attempts to discover the nature of causality. When the legislator has modified other parts of the regulation, it has not introduced new considerations regarding causality, it has not assumed a specific conception of causality. It may be concluded that the system shows an attitude of trust towards criminal law theory regarding the meaning of some words. The determination of the appropriate experts will require analysis, but this does not ultimately prevent the investigation from looking for social facts.50

65These kinds of disagreements are possible in a positivistic framework because the way in which the words are used by officials is decisive. They are not conflicting views about the grounds of law, but disputes about the nature of specific objects to which legal statements contingently refer.

Acknowledgment. I thank the participants of the IV Workshop Oxford-Girona on Legal Philosophy (2013) for their comments, especially to Diego Papayannis, Pablo Rapetti and Nicos Stavropoulos. I also want to thank Josep Maria Vilajosana, Sebastián Agüero, Alberto Carrió, José Juan Moreso, Íñigo Ortiz de Urbina, Eze Páez, Verónica Rodríguez-Blanco, Guillermo Lariguet, Samuele Chilovi, Nicola Muffato, Álvaro Núñez and two anonymous reviewers appointed by Revus for their suggestions to a previous version of this paper. This is a revised version of the paper I published in Spanish in Doxa (Ramírez Ludeña 2012).

Top of page


Juan Carlos BAYÓN, 2002: Derecho, convencionalismo y controversia. La relevancia del derecho. Eds. Pablo Navarro & Cristina Redondo. Barcelona: Gedisa. 57–92.

Jules COLEMAN & Ori SIMCHEN, 2003: Law. Legal theory 9 (2003). 1–41.

Ronald DWORKIN, 1977: Taking rights seriously. Cambridge: Harvard University Press.
——, 1986: Law’s Empire. Oxford: Hart Publishing.

Riccardo GUASTINI, 2011: Rule-scepticism restated. Oxford Studies in Philosophy of Law, vol. 1. Eds. Leslie Green & Brian Leiter. Oxford: Oxford University Press. 138161.

Herbert HART, 1994: The Concept of Law. 2nd ed. New York: Oxford University Press.

Saul KRIPKE, 1980: Naming and necessity. Cambridge: Harvard University Press.

Brian LEITER, 2007: Explaining Theoretical Disagreements. The Chicago Law School Review 76 (2007) 3. 12151250.

Andrei MARMOR, 2005: Interpretation and legal theory. 2nd ed. Oregon: Hart Publishing.
——, 2009: Social conventions. From language to law. Oxford: Princeton monographs in philosophy.
Juan Pablo MONTIEL & Lorena RAMÍREZ LUDEÑA, 2010: Von Biologie Studenten als Kellner bis zu Richtern als Biologen. Zis Online. URL:
Jose Juan MORESO, 1997:
La indeterminación del derecho y la interpretación de la Constitución. Madrid: Centro de Estudios Políticos y Constitucionales.
——, 2010: Tomates, hongos y significado jurídico. Los desacuerdos en el derecho. Jose Juan Moreso, Luis Prieto Sanchís & Jordi Ferrer Beltrán. Madrid: Fundación coloquio jurídico europeo. 1547.

Ian PHILIPS, 2014: Cetacean Semantics. Analysis 74 (2014) 3. 379–382.
David PLUNKETT & Timothy SUNDELL, 2014: Antipositivist Arguments from Legal thought and Talk. The Metalinguistic Response.
Pragmatism, Law, and Language. Eds. Graham Hubb & Douglas Lind. London: Routledge. 56–75.
Hilary PUTNAM, 1975:
Mind, Language and Reality: Philosophical Papers. Vol. 2. Cambridge: Cambridge University Press.

Lorena RAMÍREZ LUDEÑA, 2012: Los desacuerdos en el derecho. Doxa (2012) 35. 125155.
——, 2015: Diferencias y deferencia. Barcelona-Madrid: Marcial Pons.
Giovanni Battista RATTI, 2008: Los desacuerdos jurídicos en la jurisprudence anglosajona. Analisi e Diritto (2008). 301331.
Joseph RAZ
, 2001: Two views of the nature of the theory of law: a partial comparison. Hart’s postscript. Ed. Jules Coleman. New York: Oxford University Press. 137.

John SEARLE, 1958: Proper names. Mind 67 (1958). 166173.
2007: The Hart-Dworkin debate: a short guide for the perplexed. Ronald Dworkin. Ed. Arthur Ripstein. New York: Cambridge University Press. 2255.
——, 2011: Legality. Cambridge: Harvard University Press.
Peter STRAWSON, 1959:
Individuals. An Essay in Descriptive Metaphysics. London: Routledge.

Josep Maria VILAJOSANA, 2010: El derecho en acción. Barcelona: Marcial Pons.

Top of page


1 Hart 1994: 82 and ff.

2 See Marmor 2009 and Vilajosana 2010.

3 Although Shapiro (2007) and Leiter (2007) identify it as a new challenge, Leiter argues that it is not a serious one.

4 I do not want to deny that the kinds of disagreements emphasized by Dworkin have changed. In this sense, I do agree with Ratti (2008), who distinguishes between disagreements about the sources, to which Dworkin makes reference during the first period, and interpretive disagreements, relevant during the second one. What I want to defend here is that Dworkin’s challenge (regarding the existence of hard cases and disagreements) has remained, in an important sense, constant.

5 Dworkin 1977: 17 and ff.

6 Dworkin 1977: 81 and ff.

7 Dworkin 1986: 45 and ff.

8 Dworkin 1986: 93.

9 Dworkin 1986: 31-37.

10 Hart 1994: 81.

11 Hart 1994: 15.

12 Hart 1994: 246.

13 That is, it concerns the question of how to determine the semantic content of words and expressions, which is different from debates about specific conceptions of the meaning of the word “law”.

14 Regarding the relevance of a family of descriptions, see Searle 1958 and Strawson 1959. Supporting direct reference theories, Kripke 1980 and Putnam 1975.

15 Coleman-Simchen 2003: 8.

16 Shapiro 2011: 13 and ff.

17 Raz 2001: 6 and ff.

18 Raz 2001: 14 and ff.

19 I do not want to assume a specific form of conceptual analysis, nor to attribute it to positivism. I just want to defend that positivists are not committed to the relevance of convergence because they are committed to the existence of shared and transparent criteria regarding the word “law” or the concept of law.

20 Leiter 2007: 1228 and ff.; Vilajosana 2010: 173 and ff.

21 Leiter 2007: 1239.

22 Ratti 2008: 308 ff.

23 Guastini 2011.

24 The most accurate description of the situation is that ordinary language plays a fundamental role in our understanding of what is expressed by rules, and in many cases the solution given by taking into account ordinary language cannot be dismissed by invoking other instruments, because they all lead to the same solution. See Moreso 1997: 222.

25 Shapiro 2011: 290.

26 It may be argued that the number and importance of disagreements is irrelevant for Dworkin because, even if there were a pervasive agreement, he could claim that participants attribute some purpose to the practice and understand that what the practice requires depends on that purpose also in easy cases. However, it would be a remarkable coincidence. Dworkin would have to show why this reconstruction is better than a simpler one based on shared criteria (which, as Hart points out, can be accepted by officials for all sorts of reasons). Hart's theory is able to explain the legal practice in easy cases without assuming that lawyers are (in some sense) engaged in an exercise of political philosophy. Be that as it may, even if Dworkin has a hard time explaining agreement in law, positivists still need to account for legal disagreements. It may also be argued that, even if there is a pervasive agreement about how to decide many cases, participants do not agree on the specific details of the theory of law they assume. However, in the last part of this paper I will defend the irrelevance of this kind of disagreements.

27 Leiter 2007 and Vilajosana 2010: 173–175.

28 Shapiro 2007: 42.

29 Vilajosana 2010: 173 and ff.

30 Bayón 2002: 76 and ff.

31 For example, it is not very plausible to argue that the debate regarding what is a cruel punishment concerns our deep conventions.

32 Shapiro 2011: 357 and ff.

33 Shapiro 2011: 336 and ff.

34 Shapiro 2011: 367.

35 Shapiro 2011: 383.

36 Anyway, I do not mean to claim that positivists have presented the best theory of conceptual analysis available. Indeed, I think that the reflection on methodological issues in legal theory is still underdeveloped.

37 Marmor 2005: Ch. 8.

38 In fact, Dworkin reconstructs Hart’s answer in the Postcript, where he defended the conventional nature of law and the neutrality of his project, as a substantive conception about legality.

39 I do not want to deny that participants disagree about the identification of the law because they disagree (in some sense) about theoretical issues. What I want to claim here is that disagreements among theorists are irrelevant in the dispute between Hart and Dworkin.

40 Anyway, this dispute would be practical, about the best way to decide, and not theoretical, about what the law establishes (participants agree on the fact that there are, according to the law, several solutions for the case at stake).

41 This is only a simplification; in fact, since legal systems should be understood as a web of numerous interconnections and relations, it is conceivable to have a system where controversies are quite common and important for particular agents, but are not destabilizing the system in general.

42 This is, again, a simplification. It is possible to imagine a legal system with an awkward rule of recognition that establishes that both statutes and morality are law; all officials may agree on that. However, in cases where statutes and morality give opposite prescriptions the controversy would arise. So disagreements about the sources but not about the criteria of validity would take place. However, I think this would not be a theoretical disagreement about the sources, but would have an extra-legal nature and lead us to disagreements about the existence of moral principles. Making the same point but regarding interpretation, see Ratti 2008. According to Ratti (2008: 308 ff.) many disagreements consist of the selection of an interpretation from among a set of different and incompatible, but equally justified, legal solutions. The presence of moral considerations in these interpretive controversies does not require us to concede that Dworkin is right, from Ratti’s point of view, since those disagreements cannot be reconstructed as disagreements about the sources; they are not disagreements about, for example, whether natural law is or is not a legal source. They are disagreements about second-order sources, that is, about extra-legal criteria to choose between antithetic legal solutions that may be used when legal instruments have run out. Taking into account this reconstruction, many disagreements are moral and not legal, and are about what the law should be, not about what it is.

43 For this kind of cases, see the metalinguistic response offered by Plunkett & Sundell 2014.

44 Leiter 2007: 1232.

45 Kripke 1980 and Putnam 1975. I do not need to assume that those are the only theories available, but just that they are useful to provide an answer to Dworkin’s challenge.

46 This is so even if, at the end, we discover that there is not a unique essential feature. The model makes it intelligible why disagreements take place, even though in certain cases there is no unique answer.

47 Even if the argument has only been sketched in the text, I hope to have established that direct reference theories are in principle compatible with positivism since it requires taking into account the conduct and attitudes of legal interpreters. It is compatible with positivism to the extent that, according to my reconstruction, social facts (regarding the rule of recognition and the interpretation of the rules identified by the rule of recognition) are relevant, and there is no necessary connection between law and morality. And it is not assumed that law (itself) is a natural kind. In a similar way, inclusive legal positivists have claimed that there is a contingent connection between law and objective morality. See Ramirez Ludeña 2015.

48 Therefore, even if it is controversial whether fruits and vegetables are natural kinds, I think that the best way to reconstruct disagreements about tomatoes is to assume the existence of two different chains of communication. I am here using an example similar to the case decided in Nix v. Helden (149 US 304, 1893). The case is analyzed in Moreso 2010: 41 and ff.

49 This case was discussed by German judges and ended in a decision by BGH (25. 10. 2006). See Montiel-Ramírez Ludeña 2010; Moreso 2010: 31 and ff; and Philips 2014.

50 Dworkinians may point out that, according to my reconstruction, positivism is affected by the semantic sting argument. However, I am just claiming that, if we take into account how legal interpretation works, sometimes direct reference theories are the best reconstruction of the use of some legal terms.

Top of page


Bibliographical reference

Lorena Ramírez Ludeña, Legal disagreementsRevus, 28 | 2016, 11–32.

Electronic reference

Lorena Ramírez Ludeña, Legal disagreementsRevus [Online], 28 | 2016, Online since 10 June 2018, connection on 18 January 2022. URL:; DOI:

Top of page

About the author

Lorena Ramírez Ludeña

Assistant Lecturer in Legal Philosophy at the Department of Law, Pompeu Fabra University, Barcelona

 Lorena Ramirez Ludeña
Edificio Roger de Llúria (Campus de la Ciutadella)
Ramon Trias Fargas 25-27
08005 Barcelona


Top of page


All rights reserved

Top of page
  • OpenEdition Journals
Search OpenEdition Search

You will be redirected to OpenEdition Search