Coherence in Legal Interpretation

  • Marisa Iglesias Vila
Part of the Law and Philosophy Library book series (LAPS, volume 49)


In the previous chapter we presented Dworkin’s theory as an attempt to render the existence of objective restrictions to judgements about the legal practice compatible with the denial of semantic realism. Dworkin’s strategy is based on two ideas: on the one hand, that every legal judgement is theory-dependent and internal to a particular social practice; on the other, that the correctness of legal judgements can and ought to be sought within the theory or interpretive scheme that gives rise to them. But hitherto we have only outlined the negative aspect of Dworkin’s position, and the conclusion intended by the thesis of internal objectivity, namely, that there are genuine restrictions to interpretation. We shall now turn to investigate how Dworkin justifies this conclusion. It was pointed out that the justification of this conclusion requires addressing some issues regarding the problem of knowledge and the truth of legal propositions. We contended that neither empiricist holism nor Davidson’s coherentism warrant objectivity, and therefore none of them overcome scepticism.


Conceptual Scheme Propositional Content Language Game Legal Practice Coherence Theory 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    On this point see Putnam’s (1995c, 74–75; 1996, 9–12) and Dworkin’s (1991b, 360–364, 366–369) criticism to Rorty as a disillusioned metaphysical realist. For Rorty’s sceptical position, see (1991b, 8995).Google Scholar
  2. 2.
    See, for instance, Putnam (1992, 177–179).Google Scholar
  3. 3.
    See Putnam (1990, 32) and Dworkin (1996, 88; 1983, 302).Google Scholar
  4. 4.
    See Putnam (1995a, 6–7) and Dworkin (1985, 137–145; 1991a, 88–90).Google Scholar
  5. 5.
    Note that the first formulation of the principle of charity is due to Wilson (1959, 531).Google Scholar
  6. 7.
    Putnam (1987, 19). In a similar sense, Dworkin (1985, 172) observes that ‘we should account to ourselves for our own convictions as best as we can, standing ready to abandon those that do not survive reflective inspection. We should make such arguments to others, who do not share our opinions, as we can make with good faith and break off arguing when no further argument is appropriate. I do not mean that this is all we can do because we are creatures with limited access to true reality or with necessarily parochial viewpoints. I mean that we can give no sense to the idea that there is anything else we could do in deciding whether our judgements are “really” true’.Google Scholar
  7. 8.
    Putnam (1981, 55–56; 1990, 41–42, 114–115). Note that Putnam assumes the equivalence of truth and assertability by relating truth to rational acceptability in ideal epistemic conditions. The notion of truth as idealised assertability faces some difficulties to reject metaphysical realism and scepticism. The reason for this lies in that an idealisation of what is rationally acceptable dissociates truth from the agents’ epistemic abilities and yet truth remains something internal to agents. This position breaches the thesis of internal realism for it renders truth theory-dependent but inaccessible. This, in turn, means giving in to sceptical criticism. However, Putnam view appears ambiguous on this point. We will take advantage of this ambiguity to present his internal realism as a position that postulates a theory-dependent and accessible notion of truth insofar as it is meaningful for the agents to think and act as if they were in ideal epistemic conditions. Such a position permits an intelligible notion of truth internal to a particular language game in which as if is the criterion to assess the correctness of judgements. This criterion is meaningful as long as as if cannot be replaced with something external to the language game that indicates whether or not we really are in ideal epistemic conditions, or know that we are in such conditions, or they obtain at all. Hence, we will pursue Putnam’s remark (1981, 55) that `we speak as if there were such things as epistemologically ideal conditions, and we call a statement “true” if it would be justified under such conditions’. In addition, we will assume his thesis that the specification of ideal or sufficient epistemic conditions is a question that can only be answered in the very process of inquiry. Putnam (1981, 55–56).Google Scholar
  8. 9.
    Dummet (1978, xxvii).Google Scholar
  9. 10.
    It is important to note that coherence among beliefs can be adopted as a theory of truth only if truth and assertability are regarded as equivalent. Otherwise, coherence can only be thought of as an epistemic theory that often occurs in conjunction with what we will call ‘optimistic semantic realism’. This position assumes that a coherent structure of beliefs yields the correspondence between judgements and the world. We have previously noted that Davidson’s first approach to coherence theory as well as Quine’s empiricist holism fit into this line of thought. In the legal context, Michael Moore is the most representative thinker. On this issue, see Moore (1985, 309–313). See also Peczenik (1989, 184–186) and his idea that if a theory is perfectly coherent, then it corresponds to truth-makers, that is, to the world.Google Scholar
  10. 11.
    Raz (1992, 275–276).Google Scholar
  11. 12.
    Raz (1992, 275, 282–283).Google Scholar
  12. 13.
    Raz (1992, 283).Google Scholar
  13. 14.
    Raz (1992, 284).Google Scholar
  14. 15.
    Raz (1992, 275, 283). Raz discusses the consequences of a possible reconciliation between these two positions. In this sense, Raz (1992, 283) points out that `the claim would be that if justified beliefs form a coherent whole then so does the reality they represent. The subjective and the objective, the beliefs and the reality they are about must, when the beliefs are true, mirror each other’. If rendering epistemic coherence compatible with constitutive coherence leads to this conclusion, this notion of constitutive coherence is not the one advocated by a coherence theory of truth. No advocate of a coherence theory of truth would admit that coherence permits access to the reference of judgements. In his definition of constitutive coherence, Raz seems to attribute to it a notion of truth as correspondence according to which everything that is external to judgements is coherent - Raz is not proposing a notion of truth that replaces correspondence. This ascription is most evident when he observes that the success of this theory relies heavily on whether law is in fact coherent or sufficiently coherent, Raz (1992, 284, 290–297).Google Scholar
  15. 16.
    Raz (1992, 284).Google Scholar
  16. 17.
    Raz seems to hold this position hroughout his article (1992, 282–287, 290–297). It may be claimed that Raz’s reconstruction of constitutive coherence is obscure because he simultaneously describes and criticises it: in addition to describing constitutive coherence, he reformulates its problematic elements such as, for instance, the assumption that the relation of fit within a body of beliefs can be taken as a criterion of truth for their propositional contents. Raz seeks to extend coherence to law as an uninterpreted phenomenon, and not to the theories or reconstructions of law as interpretations of this phenomenon. He thus establishes the distinction scheme-world that coherentists have long tried to deny. This enables Raz to distinguish propositions of law, i.e. those that report on what is prescribed by law, from propositions about law, those which reflect the contents of our legal beliefs or legal theories.Google Scholar
  17. 18.
    Raz claims that constitutive coherence is one way of accounting for the nature of law. This, according to this author, is equivalent to accounting for the nature of things independently of what agents think of them. For this reason, this position aims at showing that the legal phenomenon is, by nature, coherent. See Raz (1992, 283).Google Scholar
  18. 19.
    Raz (1992, 290–297).Google Scholar
  19. 20.
    His distinction between epistemic coherence as a subjective justification of beliefs on the one hand, and constitutive coherence as something that is predicated independently of conceptual schemes on the other, can only be grasped in these terms. Note that this version of a theory of truth as correspondence seems to be defended in the legal context by Peczenik. Peczenik (1989, 186) assumes the presence of an essential connection between correspondence and coherence: if a theory corresponds to the world, then it is highly coherent. He asserts that `this thesis would be false had the world been chaotic. But if one assumes that the world is relatively ordered and stable, then it is plausible’.Google Scholar
  20. 21.
    Raz (1992, 289–290).Google Scholar
  21. 24.
    In line with Alston, the following aspects characterise epistemic justification: a) it applies to beliefs; b) it is an evaluative concept, that is, it implies that there is something satisfactory or correct about the fact that an agent holds certain beliefs; c) its evaluative dimension is from an epistemic point of view only. Epistemic justification is a tool to assess beliefs that merely takes into account their adequacy to their `aim at mazimizing truth and minimizing falsity in a large body of beliefs’; d) it is a matter of degree. For instance, if we relate justification to an agent’s evidence for the truth of what she asserts, the extent to which her beliefs are justified depends on the strength and amount of evidence she possesses. See Alston (1989, 83–84).Google Scholar
  22. 25.
    Henceforth, whenever we consider coherence within an epistemic structure we will essentially focus on the relationships among beliefs. True, other mental states such as desires, intentions, and preferences are also part of human psychology. But it is of utmost importance to distinguish beliefs from these other states of mind when undertaking an analysis of theoretical reasoning. Theoretical reasoning involves a particular epistemic attitude and hence relates to our views of how things are. Here, beliefs and their interrelationships play a crucial role. Practical reasoning, by contrast, refers to actions and is therefore connected to what is the thing to do. For this reason, practical reasoning is directly linked to intentions, desires and preferences. Notwithstanding, the relationship between beliefs and intentions and desires is not to be overlooked or neglected. This connection will become clear when we inquire into the role of coherence in a practice such as the legal practice, in which participants conceive the existence of law as a reason for action, that is, as a criterion to evaluate their own behaviour as well as that of their peers. The epistemic standpoint, however, has some conceptual priority over the normative one insofar that, along Harman’s line of argument, the formation of intentions presumes holding some beliefs. On this point and on the distinction between these two kinds of reasoning, see Harman (1986, 68, 113, 116).Google Scholar
  23. 28.
    On this point, see Putnam’s parallelism (1987, 79, 85–86) between ethical and scientific statements. See also Dworkin (1996, 117–118).Google Scholar
  24. 29.
    We will pay little attention to this distinction because to say of a system of beliefs that it is coherent requires that there be coherence relationships among the different beliefs. On this point, see Bender (1989, 2–3).Google Scholar
  25. 30.
    See Bender (1989, 2–5). About coherence as a property of a system of rules, see MacCormick (1984, 235236), Peczenik (1989, 159–161), Gianformaggio (1990, 403–405), and Villa (1990, 432–433).Google Scholar
  26. 31.
    Note that from an intemalist point of view, probability is to be interpreted in terms of confidence or of mutual support among beliefs. On the properties mentioned in the main text, see Goldman’s account of BonJour’s theory, Goldman (1989, 125). Similarly, Harman observes (1986, 32) that coherence `includes not only consistency but also a network of relations among one’s beliefs, especially relations of implication and explanation’. On this point see also the correlation Kress establishes between the higher or lower degree of coherence of a theory or system of beliefs on the one hand, and, on the other, the extent to which this theory or system displays consistency, comprehensiveness, completeness, monism, unity - internal relationships of justification, implication or mutual support, articulateness, and justifiability. Kress addresses each one of these properties separately and analyses the extent to which they can be considered conceptually linked to the idea of coherence. See Kress (1996, 533–534, 539543).Google Scholar
  27. 32.
    Peczenik (1989, 158–159; 1992, 183). See also Alexy and Peczenik (1989, 130–131), Aamio (1987, 199), and Villa (1990, 433).Google Scholar
  28. 33.
    Rawls’ reflective equilibrium fits into this line of thought. We shall later explore this claim further. Coherence as a property of a system of beliefs can be taken to be the global equilibrium we are after when we readjust our initial and provisional justificatory basis in the process of incorporating new beliefs into our cognitive scheme. See Rawls (1971, 14–21, 43–53, 578–582).Google Scholar
  29. 34.
    Raz (1992, 277–278). See also Sosa (1993, 395).Google Scholar
  30. 35.
    Raz (1992, 277–278).Google Scholar
  31. 36.
    Foundationalism, in Harman’s words (1986, 29), `holds that some of one’s beliefs depend on others for their current justification; these other beliefs may depend on still others, until one gets to foundational beliefs that do not depend on any further beliefs for their justification’. Broadly speaking, foundationalism is usually viewed as an epistemic position that seeks to justify our common beliefs in terms of something firmer and more stable than the beliefs themselves, be it beliefs that act as axioms, sensations, sensory stimuli, or brute facts. Thus, every form of foundationalism seeks to identify some sort of support for our epistemic structure that can stand by itself. See, for instance, Davidson (1993b, 312–313), Alston (1989, 19–32) and Kress (1996, 535).Google Scholar
  32. 37.
    See MacCormick (1984, 235; 1978, 106–107). See also Villa (1990, 433).Google Scholar
  33. 38.
    Here, we take firmness only as an example to illustrate the notion of vertical coherence. Surely other criteria are conceivable as a means of defining different hierarchical levels. For example, one could establish hierarchies according to the degree of abstraction, or the fundamentality of the contents of beliefs. However, throughout this chapter, firmness of beliefs and of a structure of beliefs will be portrayed as a main element in a coherentist perspective. We pointed out in the previous chapter that the concept of belief as a disposition to assert that p does not necessarily require that beliefs be, exclusively, those mental states of absolute certainty about the truth of what we assert. This allows us to consider that some beliefs may act as a justification to other mental states that are beliefs too. In this sense, Harman (1986, 22) points out that beliefs are a matter of degree because one can feel more strongly about certain beliefs than about certain others. On the concept of belief as a disposition to assert p, see Alston (1989, 290).Google Scholar
  34. 39.
    According to Mosterín, to be certain about something is not something than can be said of an object or idea, but only of an agent in accordance with his or her degree of confidence in the truth of a proposition or idea. Mosterín points to the importance of distinguishing between ‘certainty’ and ‘what is certain’. Certainty is a property of the agent that results from certain subjective feelings and perceptions, whereas ‘what is certain’ refers to some objective truth, independent of the agent’s opinion. See Mosterín (1987, 132–133). Regardless of whether Mosterín’s distinction can be articulated similarly from an antirealist point of view, I would like to stress that certainty is: a) a property of the agent that b) refers to the truth of his assertions and c) is precisely the maximum degree of confidence about the truth of what is being asserted. Note that from the standpoint of truth-assertability, claiming that coherence ensures confidence in the truth of what is being stated does not mean introducing a notion of truth as correspondence of Davidson’s kind. Confidence about the truth of a proposition can be understood in terms of confidence that we are legitimated to assert that proposition.Google Scholar
  35. 40.
    We have devoted no attention to coherence as intelligibility, for instance. The reason for this is that I understand intelligibility not to be an alternative way of understanding coherence, but one of its consequences or advantages. Intelligibility indicates merely that something is within the reach of our understanding, that is, we can access it. Recall that this is the minimal condition required by the principle of charity to undertake any epistemic task. However, intelligibility should not be mistaken for justification - it is merely a pre-condition for justification. This can be easily grasped if we consider that to justify is to assess something as correct from a specific point of view. We can only do so if we understand what we are talking about. On coherence as intelligibility, see Raz (1992, 276–277) and Harman (1986, 65).Google Scholar
  36. 42.
    Dworkin (1991b, 377; 1996, 119–120; 1983, 293–294; 1985, 169–170).Google Scholar
  37. 43.
    Peczenik (1990, 300).Google Scholar
  38. 44.
    Aamio (1997, 103).Google Scholar
  39. 45.
    See, for instance, Kirkham’s criticism of Blanshard, Kirkham (1992, 104–112). See also Moser’s remarks (1993, 63–75) on Dummet’s theory, and Marmor’s comments (1992, 78–84) to Dworkin’s interpretive point of view.Google Scholar
  40. 46.
    We already pointed out that this assumption can be found in Davidson’s epistemic theory (1993b, 307319).Google Scholar
  41. 47.
    Dworkin (1985, 167–170).Google Scholar
  42. 48.
    See, for instance, Harman (1986, 45) and Mattey (1989, 42). Recall that introducing the notion of probability does not necessarily entail adopting a notion of truth as correspondence, since claiming that it is likely or unlikely that our beliefs are true directly reflects our willingness to keep them or to revise them.Google Scholar
  43. 49.
    Harman (1986, 46).Google Scholar
  44. 50.
    On these principles, see Harman (1986, 39–47).Google Scholar
  45. 51.
    See Dworkin (1996, 118).Google Scholar
  46. 52.
    In this respect it is interesting to draw attention to Gärdenfors’ reflections (1988, 20) on the epistemic dynamics internal to a system of beliefs. He observes that ‘from within a belief system there is no difference between knowledge and full belief. From the subject’s point of view there is no way to tell whether she accepts something as knowledge, that is, has full belief in it, or whether her accepted knowledge is also true’. Although this author assumes a theory of truth as correspondence, it is nevertheless interesting to adopt his claim as the main idea of an intemalist point of view about the relationship between the justification of beliefs and truth-assertability. See Gärdenfors (1988, 18–20). On the pre-eminence of confidence above empirical data and logical argument in general knowledge, see, for instance, Hardwig (1991, 693–708).Google Scholar
  47. 53.
    Dworkin (1985, 138; 1986, 4–5; 1991a, 84–85). Hurley holds a similar position in this respect. She focuses her attention on the role played by coherence in the identification of what is, in moral terms, the thing to do. She understands her position to be also applicable to the domain of law. Hurley observes that the correctness or truth of a moral or legal proposition depends on whether or not the theory that displays the greatest coherence among the values to which it applies favours the truth of that proposition. See Hurley (1989, 193).Google Scholar
  48. 54.
    Dworkin (1983, 310–313).Google Scholar
  49. 55.
    Dworkin (1985, 170; 1983, 293–294; 1986, 236).Google Scholar
  50. 56.
    See Dworkin (1983, 294–295, 310–313; 1991b, 376–377).Google Scholar
  51. 57.
    It should be pointed out that, more recently, Putnam has changed his mind on this point. He has thus partially rejected the equivalence between truth and assertability that he had advocated in his early work Reason, Truth and History. Putnam has softened his internal realism and has admitted that occasionally truth may transcend what we are entitled to assert. He thus concludes that a proposition can be true and yet not verifiable or assertable. Notwithstanding, he persists in advocating the equivalence between truth and assertability for statements about social practices such as Ethics or Law. Here, in order to assert the truth or falsity of our judgements we need not appeal to a transcendent notion of truth. In this context, the mere possibility of distinguishing those statements that are acceptable form those that are not suffices to assert their truth or falsity. See, for instance, Putnam (1995a, 16–19; 1995b, 70–75). Thus, for the purposes of our discussion, we will adopt Putnam’s original idea of internal realism for statements about social practices.Google Scholar
  52. 58.
    Putnam (1987, 20–21). From this perspective, and as Villa (1990, 445) points out, `the conceptual dichotomy between “truth as the criterion of the empirical sciences and coherence as the criterion of the interpretive sciences”, breaks down completely because coherence has a very much broader scope, in the sense that it applies even to empirical disciplines’.Google Scholar
  53. 59.
    Putnam (1981, 54–56).Google Scholar
  54. 60.
    Putnam (1981, 54–55, 102, 133–134). This puts Putnam in a position to legitimately claim that ‘we cut up the world into objects when we introduce one or another scheme of description. Since the objects, and the signs are alike internal to the scheme of descriptions, it is possible to say what matches what’. Putnam (1981, 52). See also, Dworkin (1991b, 377).Google Scholar
  55. 61.
    As I have stressed, this kind of holism can be understood in terms of Rawls’ reflective equilibrium. Every interpretation requires that we take as our starting point a quasi-axiomatic basis of convictions. This basis, however, is not strictly foundational. These convictions may be subject to later revisions once we put the whole of our epistemic structure at work to find a balance among our convictions as we incorporate new ones. For this reason, it is important to distinguish this point of view from a position such as Quine’s, for whom statements with stimulus meaning eventually form a permanent foundational basis and hence, sensorial stimuli act as mediators between beliefs and reality. This basis ultimately seeks to set an external limit to the conceptual schemes that indicates the sufficiency of evidence for what we assert. Dworkin’s and Putnam’s holism implies a self-adjusting mechanism, both in the horizontal and the vertical dimensions of it; moreover, no belief holds a privileged position, and hence all of them are potentially subject to be revised. Concurrently, however, this kind of holism requires us to concede that some beliefs are better settled than others. These beliefs act as a test for truth only as long as they are not questioned or revised. See Rawls (1971, 14–21, 43–53, 578–582), Dworkin (1996, 119), Putnam (1987, 85–86), and Weinberger (1984, 231–232). See also Davidson’s criticism of Quine’s and Dummet’s foundationalism, Davidson (1993b, 312–313).Google Scholar
  56. 62.
    Dworkin (1985, 173–174).Google Scholar
  57. 63.
    Aarnio (1997, 109).Google Scholar
  58. 64.
    Dworkin (1985, 172). Dworkin insists (1996, 118) that `I mean that any reason we think we have for abandoning a conviction is itself just another conviction, and that we can do no better for any claim, […] that to see whether, after the best thought we find appropriate, we think it so’.Google Scholar
  59. 65.
    As a language game, a social practice is effectively only possible if we have confidence in something. On this point, see Putnam’s references to Wittgenstein in Putnam (1992, 176–177).Google Scholar
  60. 66.
    We noted earlier that scepticism ultimately supports externalist idealism because it seeks to position itself at a pre-theoretical stage, and claims that beliefs are all that exists. Thus, sceptics advocate an uninterpreted world made of judgements and mental states exclusively. On this point see, for example, Putnam (1990, 32).Google Scholar
  61. 67.
    On this point, see Hawkins’ distinction (1992, 32–35) between the context of negotiation and that of adjudication. MacCormick, for example, argues that the role of a judge is different from that of a mediator because the judge assumes an institutional role, which takes law rather than equity or the particular circumstances of the case as the criterion to settle conflicts. See MacCormick (1978, 242) and Barak (1989, 229).Google Scholar
  62. 68.
    See Dworkin (1991b, 380; 1986, 410–413).Google Scholar
  63. 69.
    This role is that of applying the law and standards of judicial action such as reasonableness, rationality and argumentative honesty. On these standards, see MacCormick (1978, 250–255), Raz (1972, 846–847), Hart (1980, 9) and Soper (1984, 3–9).Google Scholar
  64. 70.
    See Dworkin (1986, 86; 1985, 172; 1996, 118).Google Scholar
  65. 71.
    On Kant’s maxim, see von Wright (1970, 124–131, 137–138).Google Scholar
  66. 72.
    In both cases the judge needs a theory of law that justifies her decisions and that is not arbitrary. As a participant fulfilling an institutional role, every judge needs to provide reasons based on general principles to prefer one legal perspective over any other. On this point, see Dworkin (1986, 265–266, 353–354, 410–413; 1991a, 88, 90).Google Scholar
  67. 73.
    This idea can be more easily grasped in terms of the distinction between the internal and the external justification of a legal argument. Internal justification is the demand that the conclusion of the argument be logically inferred or deduced from the premises invoked. External justification, in contrast, is related to the suitability of the chosen premises to carry out the judicial syllogism. Often, the ‘legal justification’ of a decision has been linked to its internal justification, and its moral justification has been related to the idea of external justification. From an interpretive point of view of law there is more to the legal justification of a judicial argument than just the formal correctness of the logical inference. Moreover, legal justification does not require that the normative premise be morally acceptable. The justification of the interpretive conclusion requires some legal criterion to select the normative premise. Therefore, some kind of external justification is needed. This specifically presumes the possibility of showing that it coheres vertically with a horizontally coherent collection of principles that offer the best justification of the legal practice as a whole. On the classical distinction between internal and external justification, see Wróblewski (1974, 33–34). On the idea that legal justification requires just internal justification when there is no right legal answer and that, otherwise, it demands external justification, see Aamio (1990, 197–200), Atienza (1991, 45–46; 1993, 124–128).Google Scholar
  68. 74.
    Raz (1992, 293–297; 1979, 28–33) is one of the main avocates of this thesis.Google Scholar
  69. 75.
    This idea has been presented under different lights. Hart, for instance, considers that a legal system is made of a set of primary rules (i.e. rules that confer rights and impose duties upon the citizens) and a set of secondary rules (i.e. rules that confer power to introduce or eliminate primary rules, and to determine when primary rules have been broken) which has been granted the authority to guide conduct. However, as such, a system of rules is the product of an essentially judicial practice of acceptance and use of a set of criteria to determine what is law: a rule of recognition. According to Hart, in a complex society the rule of recognition embraces a constitution, legislative decisions, and judicial precedents, but there are other sources too. These criteria are sources of authority in the sense that they reflect who or what can generate binding patterns of conduct. See Hart (1994, 92–99, 100–104, 109–110, 116–117, 202–203).Google Scholar
  70. 76.
    Raz (1992, 284–285, 290–297). This conflict must not be mistaken for the classical debate about the authority of the legislator (as a representative of the citizenry) vis-à-vis the authority of the Constitution (as a collection of principles and values that restrict the freedom of public powers for the sake of the protection of individual rights). A Constitution may help to achieve coherence in the normative production because every legislative act and every judicial decision must conform to that set of principles. However, from a positivist point of view, the Constitution is a further authoritative legal source, which is pre-eminent over other sources. For this reason, the conflict between authority and coherence can also be found at the constitutional level. The problem that persists is, specifically, that either we concede the authority of the constitutional text, or we accept as the constitution only that part of the constitutional text that can be regarded as a coherent collection of standards of conduct.Google Scholar
  71. 77.
    These authors claim that law has normative force in the sense that the participants in the legal practice tend to take law as a reason to evaluate conduct. Hart asserts that this is the criterion that enables us to distinguish between a system of social rules and a set of habits of conduct. Hart draws the following distinction between social rules and habits: a) the existence of a habit simply requires that there be a certain regularity in conduct whereas social rules require, in addition, that deviations from the required conduct meet with criticism or reprobation; b) in contrast with habits, the criticism faced when behaviour deviates from the rule is considered to be justified; c) the existence of a social rule requires that some of those who comply with the rule (namely judges) accept the rule as a guide to conduct for everybody. These differences between habits and social rules emphasise what Hart called ‘the internal aspect of rules’. The internal aspect of rules reflects the critical/reflective attitude of participants. It is necessary that judges display such an attitude for a practice governed by social rules to exist. Hart (1994, 55–57, 88–91, 116–117).and Raz (1990, 170–177).Google Scholar
  72. 78.
    ’To have authority’ is to be able to create binding patterns of conduct. On this point, see Raz (1979, 5–11, 18–21).Google Scholar
  73. 79.
    It could be argued that Hart and Raz consider that for a legal system to exist, certain participants, namely judges, ought to display a particular attitude when facing rules. These two authors seem to agree that this attitude does not entail that judges believe in the moral correctness of the contents of the rule, but simply that they accept rules created by normative authorities as exclusionary reasons for action, i.e. reasons taken as a justification for actions that exclude prior to the deliberation process, other reasons that could have led to a different outcome. Thus, accepting a rule’s binding force does not depend on its contents but on its origin. Hart repeatedly insisted that judges might accept this out of respect for tradition or their desire to identify with others, and not necessarily their moral beliefs. See Raz (1990, 73–76, 141–146; 1979, 153–157) and Hart (1982, 265–268; 1994, 243, 257). However, Raz has changed his mind on this point.Google Scholar
  74. 80.
    Raz seems to acknowledge this problem when he claims that the idea of an authority and its intention when legislating is not relevant to determine the contents of law, but only to determine which statements should be endowed with propositional contents. This position admits the weakness of the idea of law as authority. If we adopt a semantic view of law as positivism typically does, to identify which rights and duties citizens have is to identify the propositional contents of normative texts. And if the propositional contents is no longer determined by the legislative will, the role of normative authorities in the definition of these rights and duties is minimal, or even irrelevant if we recall the sceptical challenge on rule-following. Accordingly, no statement or linguistic inscription can determine its own meaning. On this point, see Raz (1996, 256–268). See also Dworkin (1986, 317–338; 1991a, 87–88; 1985, 43–57, 154–158; 1991b, 374–377).Google Scholar
  75. 81.
    See Hart (1994, 126–129) and Raz (1996, 262–268).Google Scholar
  76. 82.
    See Hart (1994, 124–132).Google Scholar
  77. 83.
    Although it can be argued that this turn is only necessary to answer the question about what the contents of this general use of language is, it takes us back to the initial question: what does law demand we do? As was already pointed out, adopting Davidson’s principle of charity for interpretation does not enable us to dissociate the problem of what is it the members of a linguistic community think from that of what is it they do. In order to understand what they do, the interpreter need assume similarity between her conceptual scheme, her understanding of the world, and that of those whose linguistic conduct she seeks to interpret.Google Scholar
  78. 84.
    Throughout his work, Hart insists that any description of law should take into account the point of view of participants in the legal practice (the normativity of law). Hart argues (1994, 90) that a legal theory that does no account or explain the participants’ normative attitude towards rules cannot properly reflect `a whole dimension of the social life of those whom he is watching’. However, it is worth noting that in contrast with Dworkin, Hart claims that a theory of law may describe the beliefs and normative attitudes of the participants and yet not adopt an internal point of view. To clarify this point, Hart distinguishes between different types of statements about the legal practice on the basis of who produces them and the degree of internalisation they presume. The first such type of statements are those made by an external observer, who need not share the normative attitudes or the beliefs of participants. These are statements about law (descriptive statements), and should be clearly differentiated from internal statements. Internal statements are those statements produced by participants in the practice, that is, by individuals who accept its rules as a guide to assessing conducts and as justifying reasons for action. According to Hart, external statements fall into one of two further categories: those made by an external observer, which describe mere observable regularities in behaviour, and those by means of which the external observer describes the participants’ attitudes and beliefs from an internal point of view. Hart claims that a theory of law does not require the observer to share the participants’ normative attitudes, but its description of law should reflect the internal point of view. In Essays on Bentham, Hart includes a further type of statements: normative statements (which are not statements about participants’ normative attitude). Statements made by lawyers and jurists (as opposed to judges, who are assumed to adopt the internal point of view) belong in this class of statements. Normative language is used in these statements and yet the rules are not taken to be justifying reasons for action. See Hart (1994, 89–91, 115–117; 1982, 153157; 1983, 165–169) and Raz (1990, 171–177; 1979, 154–157). We pointed out in Chapter 2 that one of the main objections to positivism is that it regards legal theory to be something external to the practice, wherefore its statements are not committed to law as a justification for state coercion. The arguments explored after Chapter 2 aim at showing the unfeasibility of externalism as an approach to law. In my opinion, Dworkin would consequently be right in claiming that every description and understanding of the legal practice is interpretive because it involves a descriptive justificatory viewpoint about the object of interpretation. See Dworkin (1985, 146–149; 1994b, 2–3). For evidence for this argument, see Putnam’s views on the interrelationship between fact and value (1981, 127–135). See also Wittgenstein (1992, 9–91).Google Scholar
  79. 85.
    Although these two positions are dependent on one another within a normative practice, we could nevertheless distinguish between epistemic and axiological justifications. Epistemic justification evaluates beliefs to maximise the truthfulness of their propositional contents. Axiological justifications, in contrast, evaluate beliefs in terms of their conformity with certain values or standards of conduct. From an internalist point of view, whether or not a belief is epistemically justified depends on its degree of conformity with a structure of beliefs as a whole, whereas for it to be axiologically justified it needs to accord with an axiological structure or a scheme of values. Every epistemic structure is nevertheless often considered to be interrelated with an axiological structure, particularly as far as phenomena with a normative dimension are concerned, that is, when participants understand their interactions to be binding or due acts. In this context, any belief about which conducts are legally regulated is in fact a belief about what is legally the thing to do. It can thus be claimed that the participants in the legal practice hold a normative point of view about their interactions. A great deal of literature concerned with the normativity of law has emerged and yielded fruitful debate about which kinds of reasons for action some participants need assume if a system of legal rules is to exist. This debate has re-opened the controversy on the connection between law and morality. See Hart (1982, 243–268) and Raz (1990, 162–177; 1979, 153–157). For discussion on these two authors concerning this point, see Nino (1985, 118–137; 1994, 7981, 192–196), Bayon (1991), Garzón Valdés (1990, 119–124) and Redondo (1996). We shall not devote further attention to this issue. However, I would like to point to an interesting question: to what extent does the assumption that participating in the legal practice requires reasons dependent on the contents imply that these reasons are of a moral character? To answer this question we should elucidate the extent to which we can claim the autonomy of justificatory legal discourse. This would break down the idea of unity of practical discourse. See Putnam (1995a, 7) and Fried (1981, 35–58). This is precisely Coleman’s interpretation (1995, 53–54) of the relationship between description and justification in Dworkin’s interpretive theory.Google Scholar
  80. 86.
    Similarly to Raz, Hart considers that for a system of legal rules to exist, a critical/reflective attitude is not required from all of its participants, but just from some of them: judges. See Hart (1994, 114–117) and Raz (1990, 171).Google Scholar
  81. 87.
    In a broader sense, we could even say, following Villa (1990, 446, 447), that at the background of any holistic structure `value judgements cohabit with theoretical statements mixed up with them (…). It is impossible to insulate the evaluative component from the whole theoretical framework’.Google Scholar
  82. 88.
    Dworkin notes that his interpretation of Hercules’ reasoning is no more than what he considers to be his own opinions about which arguments are best. Thus, his claim that law as integrity should combine justice, equity and due process results from his interpretive convictions. Like any other judgement, this one is subject to debate and can be refuted by better arguments. See Dworkin (1986, 239–240).Google Scholar
  83. 89.
    For this reason, Dworkin points out that judgements about the contents of law are interpretive. They are the result of applying a complex theory that includes two dimensions: one tends to be more epistemic; the other, more substantive. It is important to define the character of these judgements on the basis of the idea of tendency because this reflects the thesis of interconnectedness of our judgements. This tendency is the result of interconnectedness and of the location of each judgement within the epistemic structure. On this point, see Dworkin (1985, 147–148).Google Scholar
  84. 90.
    Dworkin (1986, 67–68, 98–101).Google Scholar
  85. 91.
    On this notion of fraternal community, see Dworkin (1986, 206–216). It is not my intention to elucidate to what extent coherence should be adopted as a legal/political virtue. In connection with this point, we will assume Dworkin’s position about what is required from a political community that has made a commitment to certain principles. Hurley’s views on this issue are interesting too. She appeals to coherence as a means to warrant personal autonomy and to determine what is the thing to do in a democratic society. Hurley (1989, Ch. 15).Google Scholar
  86. 92.
    This position would have an extemalist character and would imply scepticism because it would assume that there are just as many right or true positions as there are coherent schemes. However, such a claim is external to the conceptual schemes - and the plausibility of this kind of assertions is what we have been attempting to refute throughout this work.Google Scholar
  87. 93.
    Notwithstanding that this issue falls beyond the scope of this work, I pointed out in Chapter 2 that I find Dworkin’s substantive project appealing. In my view, his argument against the claim that institutional authorities may be mistaken about law is convincing: a decision is not legal just because it has been made by an institution that we regard as authoritative. Accepting authoritative decisions as exclusionary reasons entails precisely what needs to be proven: specifically, that such a reconstruction is epistemically plausible, and that it is better to view the legal practice as a passive practice than as a practice in which participants have their own convictions about which behaviours are legally regulated and assume that some arguments are better than others. I agree with Dworkin’s denial of global scepticism as an acceptable interpretation. We cannot adopt scepticism because we presuppose that we are participants in a collective practice and, therefore, whenever we produce judgements about the law, we have an object to interpret.Google Scholar
  88. 94.
    Dworkin (199la, 88).Google Scholar
  89. 95.
    Dworkin (1985, 144–145; 1991a, 90).Google Scholar
  90. 96.
    See Finnis (1987, 371–374), Putnam (1995a, 6–7), Mackie (1984, 165, 168–169), Marmor (1992, 54–57) and Bix (1993, 96–106).Google Scholar
  91. 97.
    Dworkin denies the possibility of a tie between rival interpretations because in his view this claim makes no sense. Furthermore, he rejects incommensurability because he contends that it is always possible to embrace two interpretations according to the same parameters or criteria. See Dworkin (1991a, 89–90). Note that the purpose of this section is not whether or not it is possible to reject incommensurability between any two interpretations. This issue has been explored earlier when we considered global scepticism. Total incommensurability has advocates in every field of knowledge. Paradigmatic examples of this position are Feyerabend’s critique of the scientific method, and the sceptical interpretation of the problem of rule-following in the context of the philosophy of language. These lines of argument have been adopted by deconstructionist approaches to legal theory such as that of Critical Legal Studies. See Putnam (1992, 124–128, 149–152; 1981, 113–119) for an argument on how global scepticism is self-defeating.Google Scholar
  92. 98.
    This is Finnis’s internal criticism to Dworkin’s thesis on the right answer. See Finnis (1987, 371–374) and Mackie (1984, 165).Google Scholar
  93. 99.
    True, it is always logically possible to establish some comparative parameter between the two, but it would be hard to do so convincingly. In this case, I believe there to be better reasons to defend than to deny incommensurability. It should be conceded, however, that the question about whether we are within one or a number of evaluative contexts is a matter of degree and interpretation. See Dworkin (1996, 132135; 1991a, 89).Google Scholar
  94. 100.
    Marmor (1992, 54–57).Google Scholar
  95. 102.
    See Hurley (1989, 263–270), Bix (1993, 101–102) and Raz (1992, 286).Google Scholar
  96. 103.
    Dworkin (1996, 137–138) replied to incommensurability by appealing to the need for decisions in the legal context. Moreover, he has often insisted that we will always have an opinion to readily choose between two alternatives. See Dworkin (1991a, 90).Google Scholar
  97. 104.
    See, for example, Guastini’s objections to Dworkin’s model in Guastini (1983, 455).Google Scholar
  98. 105.
    Some authors have devised perspectives to eliminate the incommensurability of values but they always assumed or added a structure that includes a meta-value or meta-criterion that allows for an intelligible comparison. The interesting question is whether it is possible to eliminate the problem of incommensurability when there are no such criteria readily available. On these proposals, see Raz (1991, 83–88) and Finnis (1990, 238–241).Google Scholar
  99. 106.
    Putnam, one of the exponents of the theory of truth as coherence, argues that a coherence theory should always assume incommensurability. He occasionally criticised Dworkin for being too naïve in this respect. See Putnam (1981, 72–73; 1995a, 6–7). See also Hurley (1989, 264–270).Google Scholar
  100. 107.
    Dworkin (1991a, 89; 1985, 144).Google Scholar
  101. 108.
    Currie (1993, 421).Google Scholar
  102. 109.
    This problem may be reproduced ad infinitum. See Moreso (1998, 82–83), Raz (1979, 73–74) and Putnam (1975, 217; 1995b, 76–77). For a detailed analysis of this issue, see Sainsbury (1991, 167–182).Google Scholar
  103. 110.
    See Endicott (1995) and Dworkin (1985, 405 footnote 3).Google Scholar
  104. 111.
    Dworkin (1996, 129–139).Google Scholar
  105. 112.
    Coleman (1995, 49). Coleman argues that Dworkin has by now abandoned the optimistic thesis about the existence of right answers. But Coleman’s claim is not based on an explicit turn of Dworkin’s view. Perhaps, Coleman intends to suggest that Dworkin should abandon his optimism because it is unjustified. Nothing in Dworkin’s work suggests that he rejects one of his more controversial theses (and for that matter, one that he defends most fiercely). See, for example, his argument supporting the right answer in one of his most recent papers, Dworkin (1996, 129–139). Like Coleman, Kress (1996, 540541) seeks to dissociate coherence from an optimistic thesis about the existence of right answers.Google Scholar
  106. 113.
    See Dworkin’s distinction between internal and external scepticism. External scepticism is an externalist position along the lines of what I have outlined in Chapters 4 and 5. According to external scepticism, it is pointless to claim the truth or falsity of a judgement. External scepticism concludes that discussions about what argument or interpretation is best simply reflect the expression of incommensurable opinions, or, even just noise. Dworkin (1986, 78–86; 1996, 89–94).Google Scholar
  107. 114.
    On this principle see Harman (1986, 46).Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2001

Authors and Affiliations

  • Marisa Iglesias Vila
    • 1
  1. 1.Department of Legal, Moral and Political PhilosophyPompeu Fabra UniversityBarcelonaSpain

Personalised recommendations