The Strong Discretion Model

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


In the previous chapter we outlined the most common uses of ‘judicial discretion’. It became clear that discrepancies in the notions of strong and weak discretion are a consequence of different approaches to the truth conditions of legal propositions. When investigating judicial discretion, legal theorists do not only propose a characterisation of this concept, but also defend a particular thesis on judicial discretion, that is, they hold a specific position about whether or not discretion is necessary in judicial adjudication. Moreover, every thesis on judicial discretion ought to be considered in the wider context of a theory of law and its limits Every such theory involves a certain position on the existence of right legal answers, and on how these answers can be identified. This is an essential point to bear in mind when analysing theoretical disagreement about the existence of judicial discretion. On some occasions, the differences between those who advocate and those who oppose the need for strong discretion are merely a matter of degree. On other occasions, however, they reflect deeper disagreements on how we come to understand normative standards. Incidentally the role that judicial discretion is assumed to play in adjudication may introduce certain changes in the legal theory embraced. Consequently, a scrutiny of the thesis on judicial discretion maintained by a particular legal theory permits us to examine its internal coherence.


Hard Case Literal Meaning Easy Case Semantic Rule Legal Term 
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.
    We will concern ourselves with the version of positivism that offers an answer to the questions what is law, how do we identify it, and what is its structure. In my view, this version turns on Hartian legal theory because current positivism adopts a Hartian approach to law: it seeks to articulate the main aspects of Hart’s perspective in a more sophisticated way. Google Scholar
  2. 2.
    See Raz (1979, 37–51) for a detailed discussion of this thesis. For Joseph Raz, the social sources thesis implies, in the main, that what is considered to be part of law is a matter of social facts, of human activity and, therefore, of the practices and conventions in the community. Thus, the identification of law can be done on the basis of the social sources regardless of the interpreter’s morality. The existence of these sources is the basis for truth of a proposition about what is, legally, the thing to do. In a similar sense, see Hart’s conception of the rule of recognition as a social practice in which certain criteria for legal validity permitting the identification of law are accepted. See Hart (1994, 100–117). See also Pâramo (1984, 379 Google Scholar
  3. 12.
    Note that semantic theories for which the meaning of propositions relies on their truth conditions have traditionally been related to semantic realism, while constructivist or antirealist theories reject a semantic theory based on truth conditions, and favour one based on conditions of assertability. Nevertheless, Baker argues that certain articulations of constructivism accept the semantics of truth conditions but adopt a concept of truth as assertability. See Baker (1977, 50–51) and Dummet (1978, xxi, xxiii). Perhaps this new formulation of antirealism appears in response to scepticism, the theory of radical indeterminacy that rejects the very possibility of talking about truth conditions. Hence, this does not seem to be a merely terminological discussion about whether it is preferable to talk about ‘truth’ or ’assertability’ to refer to what we are entitled to assert. It would be interesting to analyse whether the different formulations of antirealism diverge in some significant way. Unfortunately, such an enquiry is beyond the scope of this work. Google Scholar
  4. 13.
    Shiner (1992b, 87–90).Google Scholar
  5. 14.
    Shiner (1992b, 89–90) points out that, according to sophisticated positivism, one of the rules governing the judicial role takes correspondence to the institutional history to be what actually warrants the assertability of a proposition in easy cases. This idea is rather odd for it seems to imply that only Google Scholar
  6. 21.
    This communication technique is a paradigmatic case in continental legal orders and, for this reason, will be taken as the starting point for the present chapter. This, however, does not by any means imply that there exist no other techniques of communication. For example, legal precedent and custom are extremely relevant factors when accounting for the structure of any normative system.Google Scholar
  7. 22.
    See Hart (1994, 126–127, 135–136).Google Scholar
  8. 23.
    Hart (1994, 126). Hart claims that indeterminacy is an inevitable feature of any attempt to guide human conduct by means of linguistically formulated general rules. It will always be possible to find cases for which no previous definition of the general classifying terms appearing in the rule allows us to decide whether the case falls within its scope of applicability or not. See Hart (1980, 9). Google Scholar
  9. 24.
    See Hart (1983, 33, 35). It is not intended that the truth of a legal proposition depend exclusively on the semantic considerations about the conditions for legal validity coming together with the meaning of each formulation considered independently. Surely, the truth or falsity of these propositions is also conditional on more fundamental semantic conventions about the meaning of ‘law’. Google Scholar
  10. 25.
    See Bix (1993, 33). Hart (1994, 128) considers the relative indeterminacy of purpose as a source of linguistic indeterminacy. It is worth noting the relevance conferred by Hartian positivism to linguistic indeterminacy as a source of normative indeterminacy and, hence, as a source of judicial discretion. On the different types of legal indeterminacy, see Coleman and Leiter (1995, 212–218). On the relationship between semantic and legal indeterminacy in Hart, see Ruiz Manero (1990, 184–189). Google Scholar
  11. 26.
    On Hartian positivism as a semantic theory see, for instance, Coleman and Leiter (1995, 35–36). The above consideration, however, does not seek to reduce Hartian positivism to a semantic theory of law. It wishes to highlight the relationship between the semantic considerations set forth above, and the truth of judgements about what is legally the thing to do. In addition, it does not imply that the thesis on the social sources of law is merely a semantic one. Raz (1979, 37–40) specifies that the thesis on the social sources of law seeks not only to explain the substratum of the meaning of normative statements, but also their normative force. The normative force of law is linked to the idea of internal point of view proposed by Hart. This is the standpoint adopted by the participants in a social practice that is subject to rules, and not just to habits of conduct. Participants view legal rules as providing guidelines to their own conduct and standards for assessing that of their peers. On the notion of normativity and the internal point of view, see Hart (1994, 82–91, 116–117). See also Raz (1990, 154–177). Google Scholar
  12. 27.
    Note that both realism and antirealism maintain a semantic theory based on the conditions for truth of legal propositions. The difference between them is, precisely, what they understand these conditions to be, i.e. what are the specific conditions that make a proposition true. See Patterson (1996, 18).Google Scholar
  13. 28.
    See, for instance, Marmor (1992, 87–90). Google Scholar
  14. 29.
    Marmor (1992, 90)Google Scholar
  15. 30.
    See Dummet (1978, 164–165). Marmor (1992, 88) observes that the relationship between realism and the principle of bivalence is not symmetrical: realism implies bivalence, but the converse is not true.Google Scholar
  16. 31.
    Regarding this point, Hart (1994, 126) asserts that `general terms would be useless to us as a medium of communication unless there were such familiar, generally unchallenged cases’. See also Marmor (1992, 134–135).Google Scholar
  17. 32.
    Marmor (1992, 30) claims that the meaning of expressions, unlike what is actually meant by them, is essentially determined by rules and conventions.Google Scholar
  18. 33.
    See Hart (1994, 126–127; 1983, 106), Schauer (1991, 56) and Bix (1993, 63–70, 73–76).Google Scholar
  19. 34.
    See Bix (1993, 73–74), Marmor (1992, Ch. II) and Schauer (1991, 53–62).Google Scholar
  20. 36.
    See Hart (1994, 127–128). Schauer (1991, 55) points out that the meaning is not reducible to the purpose for which the language is being used on a particular occasion. Language is semantically independent of the communicative intentions of its users. Fuller (1958, 662–669) criticises the position that posits that theinter-subjective use of a word or sentence is central to its meaning. He thinks it pointless to attempt todissociate the meaning of an utterance from its particular context of use and the communicative intentions of the speaker. Google Scholar
  21. 37.
    When uttering a statement, language can be used for different purposes. The illocutionary force of a speech act is the intention or communicative purpose that drove its production. On the distinction between illocutionary, locutionary and pre-locutionary acts, see Austin (1976, 94–120) and Searle (1969, 22–53). Google Scholar
  22. 38.
    See Schauer (1991, 55) and Bix (1993, 75). Hart claims that assigning meaning to a legal expression does not entail describing a psychological fact. See Hart (1983, 33). Google Scholar
  23. 39.
    On the definition and distinction between meaning of and meaning that, see Marmor (1992, 15, 30). Schauer (1991, 53–58) claims that this distinction is necessary to differentiate the meaning of general rules from their justification. Rules are instances of further rules that account for their justification. If the meaning of these instances is limited to the intentions of the agent who enacted them, their meaning is simply the justification as applied to the case at hand. Hence, it would obtain that the meaning of the instance and its justification collapse into one and the same thing, that is, they become coextensive. Grice distinguished between the sentence meaning on the one hand, and the utterer’s meaning, on the other. But it should be noted that Grice, in contrast to what is being presented here, tried to reduce the sentence meaning to the utterer’s meaning. See Grice (1989, 86–137). See also Searle (1969, 42–50). Google Scholar
  24. 40.
    On the semantic autonomy of meaning see, for instance, Davidson (1984b, 164–166). Google Scholar
  25. 41.
    41. Schauer (1991, 55–56).Google Scholar
  26. 42.
    Bix (1993, 75).Google Scholar
  27. 43.
    These sub-communities of use have specific jargons and dialects: the literal meaning of words differs from one context to another. On this point see Bix (1993, 73–76) and Marmor (1992, 26–27). For criticism on a variety of fronts of the idea of sub-communities, see Coleman and Leiter (1995, 260–263). Google Scholar
  28. 44.
    The claim that the literal meaning of a word, unlike its pragmatic meaning, is determined by linguistic rules and conventions is defended by Marmor who, along the lines of Wittgenstein’s considerations on rule-following, distinguishes between understanding the meaning of an expression, and interpreting it. Understanding the meaning of an expression relies on its literal meaning, whereas its interpretation is linked to its pragmatic meaning. See Marmor (1992, 22–30, Ch. VII). Google Scholar
  29. 45.
    Schauer (1991, 58). See also Marmor (1992, 24–28). Searle (1969, 48–49) contends that communication between speaker and listener is made possible by the public dimension of language, i.e. by the fact that both share the same language. His idea is illustrated by the following example: `1. Understanding the sentence “Hello” is knowing its meaning. 2. The meaning of “Hello” is determined by semantic rules, which specify both its conditions of utterance and what the utterance counts as. The rules specify that under certain conditions an utterance of “Hello” count as a greeting of the hearer by the speaker. 3. Uttering “Hello” and meaning it is a matter of (a) intending to get the hearer to recognize that he is being greeted, (b) intending to get him to recognize that he is being greeted by means of getting him to recognize one’s intention to greet him, (c) intending to get him to recognize one’s intention to greet him in virtue of his knowledge of the meaning of the sentence “Hello”. 4. The sentence “Hello” then provides a conventional means of greeting people’. Google Scholar
  30. 46.
    In connection with this issue, see Marmor’s thesis that literal meaning is governed by rules whereas pragmatic meaning is beyond the control of semantic rules. Google Scholar
  31. 47.
    Raz (1996, 266).Google Scholar
  32. 48.
    Raz (1996, 262–266) draws attention to the fact that legislators are often not acquainted with the details of the legislation they vote, or only to a very limited extent. Google Scholar
  33. 49.
    See Raz (1996, 267–268). This standpoint entitles Raz to maintain the thesis that law is the product of acts of institutional authorities without having to assume strong intentionalism, the view that identifying the contents of the texts that have been elaborated by institutional authorities requires elucidating their state of mind and real intentions at the time of writing.Google Scholar
  34. 50.
    Raz claims that the plain or conventional meaning of a text is the only factor restricting the possible interpretations of the legislative intention. Unless one adheres to plain meaning, even a Shakespearean interpretation of the Constitution is imaginable. See Raz (1986, 1103–1119; 1995, 161–162). Google Scholar
  35. 51.
    On reference, see Frege (1993, 24–28). Google Scholar
  36. 52.
    Indexical terms operate as rigid designators for they refer to the same object in every possible world. Kripke (1980, 106–155) has used this approach to account primarily for the meaning of proper names. Putnam (1975, 229–235, 245–247) extended this approach to any term referring to classes of objects whose existence does not result from mere human intervention and intellectual ability. In fact, it could be argued that Putnam’s extension is far more ambitious, for it wishes to extend this view of meaning to most terms, regardless of whether they designate artefacts, or, classes of objects that come into existence only as a consequence of the structure of the world. Google Scholar
  37. 53.
    For this reason, referencialist theories claim that a whole community of speakers may misidentify the meaning of the words it uses. See Putnam (1975, 223–227, 245–247; 1981, 22–25) and Kripke (1980, 121¬139). The most recent attempt to show the indexicality of the term ‘law’ is due to Michael Moore (1985, 277–398). Google Scholar
  38. 54.
    Marmor (1992, 96–97) stresses that the assertion that ‘law’ is indexical is in contradiction with the antirealist position that, in his opinion, characterises positivism. No positivist would agree that the whole of a legal community may be mistaken on the extension of the term ‘law’. The reason for this lies in the fact that positivists, unlike natural theorists, consider that the extension of this term is reducible to social rules, practices and conventions. Google Scholar
  39. 55.
    Searle (1995, 14) writes that functions are not inherent in objects - functions are assigned to or imposed upon objects by conscious users. Hence, functions are always observer-relative. See Searlé s theory on institutional facts, and MacCormick and Weinberger’s further development and extension of this theory to the legal context. The truth or falsity of statements about institutional facts, unlike those referring to brute facts, does not depend on the structure of the world and the causal relationships among its elements. The existence of institutional facts presupposes human institutions and social rules. Thus, statements about institutional facts are true on account of an interpretation of events that is intrinsically bound to human practices and normative rules. See MacCormick and Weinberger (1986, 49–55) and Searle (1969, 50–53; 1995, 27–29). Google Scholar
  40. 56.
    On this point, see Hart (1983, 23, 31–35). For a similar articulation of this idea, see Moreso (1996, 113) and Camap (1971, 308). Google Scholar
  41. 57.
    Moore claims that, whether consciously or unconsciously, we think of words as indexicals. See Moore (1985, 331). Google Scholar
  42. 58.
    See Moore (1985, 294–301; 995, 3–6). Note that the problem we are considering here is not whether the terms we use are conventional or not, but whether objects carry their own identifying labels or whether the differentiation between objects is essentially conventional. On this point see Coleman (1995, 41 footnote 8). Google Scholar
  43. 59.
    Moore (1985, 338). Google Scholar
  44. 60.
    Moore (1985, 287) points out that legal definitions of terms and the conventions about their use can only be viewed as an agreed summary of the real meaning of words. Such legal definitions may misidentify the reference and lead to absurd conclusions about the meaning of terms. See Moore (1985, 329–331). Google Scholar
  45. 61.
    Note that Putnam’s referencialist theory is hard to redirect towards a realist approach to semantics, for the latter presupposes a correspondentist theory of truth. On this point, see Putnam’s argument in favour of what he calls ‘internal realism’, and which he put forward in response to truth as correspondence, Putnam (1981, 1–74). Google Scholar
  46. 62.
    See Stavropoulos (1996, 76) and Bix (1995, 138–141). Google Scholar
  47. 63.
    Coleman and Leiter (1995, 252). Google Scholar
  48. 64.
    Coleman and Leiter (1995, 253) point out that positivism displays antirealist features such as the rejection of the epistemic transcendence of meaning and legal ontology. Moreover, they claim positivism holds an objectivist position because it rejects the idea that the social world may depend only on how a particular individual perceives it. Google Scholar
  49. 65.
    See Hart (1994, 126–127, 131–132) for further consideration of this point. Google Scholar
  50. 66.
    Schauer (1991, 55), for example, claims that the basis for the semantic autonomy of language with respect to particular instances of use can be explained in terms of a socially established reference. Google Scholar
  51. 67.
    Moore elaborates a detailed criticism of positivism that develops along similar lines. He argues that asserting that legal terms are not in correspondence with natural classes of objects but with functional classes is compatible with semantic realism. Hence, no additional theory of language is needed to account for legal terms. See Moore (1992, 206–208). Brink (1989, 183–185) presents a similar argument. Google Scholar
  52. 68.
    Brink (1989, 184) Google Scholar
  53. 69.
    Coleman and Leiter (1995, 249). Google Scholar
  54. 70.
    In his analysis of the social sources thesis, Joseph Raz (1979, 39–40) asserts that `a jurisprudential theory is acceptable only if its tests for identifying the content of the law and determining its existence depend exclusively on facts of human behaviour capable of being described in value-neutral terms […1’. This idea clearly suggests a preference for a correspondentist position about the truth conditions of legal propositions, and for an explanation of mistakes in the identification of law in terms of erroneous descriptions of certain facts. Google Scholar
  55. 71.
    With regard to this point, it may be interesting to consider Marmor’s criticism of Lewis’s thesis that the very existence of a convention requires that its existence be widely known. Following Burge, Marmor points out that people may be mistaken about the conventional nature of the rules they follow. Moreover, it may be the case that people believe to be following a convention when in fact they are not. In effect, whether or not something is a convention is not a matter of convention. A community may follow a convention even if it does not possess the concept of convention. Marmor’s main objection to Lewis is that the he leaves no room for the controversies about the conventionality of any dimension of human conduct. This type of controversies is possible and intelligible because it involves matters of facts which are not necessarily matters of common knowledge. In spite of assuming Burge’s criticism to Lewis, Marmor indicates that this criticism is of limited applicability. It can only be taken into consideration in connection with the question about when is it that we face a convention. In Marmor’s opinion, accepting that there could be a general misunderstanding of the contents of a rule is a far more delicate question. Nevertheless, the only reason put forward not to extend this idea is that it is inconsistent to maintain simultaneously an antirealist position for a certain kind of statements, and to accept the possibility of general mistake about the truth of such statements. Marmor (1996, 354–355). Regardless of how Burge’s criticism affects Marmor’s position about linguistic conventions in the determination of the propositional contents of legal rules, this criticism proves useful to illustrate the possibility to distinguish between shared use, and widespread knowledge of this shared use. Google Scholar
  56. 72.
    Dummet (1978, 162–165) claims that rejecting bivalence necessarily implies antirealist semantics and, hence, a theory of truth as assertability. See also Marmor (1972, 86–93) and Moreso (1998, 49–51). Google Scholar
  57. 73.
    Bix (1993, 9). Google Scholar
  58. 74.
    Hart (1994, 126–130). Google Scholar
  59. 75.
    75. Hart (1994, 129). This explains why, in spite of acknowledging a relevant analogy between the language of rules and that of precedents, Hart points to an important difference between them, namely that the language of rules reflects general criteria of use in far more stable and permanent a way than the case-to¬case language of precedents does. Google Scholar
  60. 79.
    See Schauer (1991, 226). Understanding the meaning of a linguistic formulation seems to be very closely related to following a rule or acting in accordance with it. Given that rules are expressed by means of language, the range of conducts that constitute acts of rule-following will be specified insofar as the meaning of normative formulations is determined. For this reason, ‘the correct use of a term’ and ’following a rule will be considered synonymous henceforth. On this synonymy, see Bix (1993, 43, 47¬49). Google Scholar
  61. 80.
    80. See Kripke (1982). For a sceptical interpretation of rule-following in the legal context see Yablon (1987, 629-636). Google Scholar
  62. 81.
    According to Kripke (1982, 13–22, 53–54), rule-following is not determined by platonic entities nor by any past event such as external behaviours or states of mind. Google Scholar
  63. 82.
    See Wittgenstein (1953, secs. 186–188) and Kripke (1982, 7–13). Google Scholar
  64. 83.
    See Wittgenstein (1953, secs. 185–201). Google Scholar
  65. 84.
    Kripke (1982, 55) and Wittgenstein (1953, Sec. 201). Kripke draws attention to the fact that this is not merely an epistemological problem. The sceptic position does not claim that there is no way of knowing when an agent has complied with a rule or contravened it. On the contrary, any past event taken to be a criterion of meaning is compatible with either conclusion and hence, nothing can establish the meaning of a term or expression. Kripke (1982, 21). Google Scholar
  66. 93.
    Baker and Hacker (1994, 102, 171–172). Google Scholar
  67. 94.
    Opponents of scepticism claim that the search for mediators between the rule and its instances of use is what in fact leads to the thesis of ‘anything goes’. Taking interpretation as such a mediator yields the sceptical conclusion. See Marmor (1992, 149) and Bix (1993, 39, 74). Google Scholar
  68. 95.
    In line with Dummet, Marmor advocates the thesis of the exceptionality of interpretation. Interpretation is required only when a rule cannot possibly be followed, that is, when it is dubitable whether or not a specific action complies with the rule. In this case, we need some interpretation of the rule imposing meaning to it on the basis of somebody’s understanding of what its meaning is, be it the interpreter or the author of the linguistic formulation himself. According to Marmor, this activity is no longer governed by rules and conventions, but guided by paradigms. However, interpretation is an exceptional kind of understanding for it is parasitic on understanding. Thus, every interpretation requires a previous understanding of the instances embraced by the literal meaning of the formulation of the rule. See Marmor (1992, 21–24) and Dummet (1993, 464–472). Google Scholar
  69. 96.
    Hartian positivism acknowledges the importance of the problem identified by Wittgenstein and pointed to by scepticism, but considers that Wittgenstein does not (and need not) offer a sceptical answer to it. See Bix (1993, 38–41). Google Scholar
  70. 97.
    Marmor (1992, 148–152). Google Scholar
  71. 98.
    See Bix (1993, 41), Marmor (1992, 153) and Baker and Hacker (1994, 259–262). Google Scholar
  72. 99.
    99. According to Hacker and Baker, agreement on judgements consists in interpersonal consensus about the truth or falsity of empirical propositions, which, in fact, is just a consensus on how we describe the world. Agreement on definitions, by contrast, is consensus on how we use words, which implies that we use the same words to refer to the same objects. We shall have later to consider the relationship between these two kinds of agreement. On this distinction, see Baker and Hacker (1994, 258-262), Hurley (1989, 30-33), Stavropoulos (1996, 125-128) and Bix (1993, 53-59). Google Scholar
  73. 100.
    Baker and Hacker (1994, 170–174) and Bix (1993, 41). Google Scholar
  74. 101.
    This, however, does not imply that the meaning is ultimately reduced to the particular intention of the speaker, that is, the communicative purpose with which language is used in a particular instance. It seeks to stress that the meaning of an expression is not independent of the epistemic abilities of the individuals applying the semantic rules, nor of the interpreter that evaluates the correctness of somebody else’s application of the rule according to her own understanding of it. Hacker and Baker assert that only two requirements must be satisfied to ascertain that a rule has been followed. Firstly, the agent ought to have rules of an appropriate class, and an awareness of the rule that she claims to be following. Secondly, her conduct ought to exhibit a sufficient degree of complexity to demonstrate the intention and ability that are essential to following a rule. It should be borne in mind that Hacker and Baker think these two conditions to be independent of the existence of either a shared use or a community of speakers. An isolated individual could follow a rule even if there is no community. Similarly, an agent can follow a rule even if there is no general consensus about its use. This is so because, although following a rule requires the existence of a technique or a practice (a criterion of correctness), this practice need not be social. Claiming that there is an internal relationship between the rule and its instance of application amounts to asserting that the agent does not act arbitrarily but in accordance with some criterion of correctness that allows her to distinguish between compliance and violation of the rule she claims to be following. See Baker and Hacker (1994, 170–179). Google Scholar
  75. 102.
    Baker and Hacker (1994, 259). Agreement on the form of life accounts for the similarity of the answers given by different individuals, but does not justify the claim that they are following a rule. On the contrary, answers may coincide but they may not result from following a rule. As pointed out by Hacker and Baker, the only factor that permits the distinction between rule-following and the mere appearance that a rule is being followed is the mastery of a particular technique - the ability to provide an answer to which is the rule that is allegedly being followed, and which acts are required to follow it correctly. See Baker and Hacker (1994, 176–179). Along similar lines, McDowell (1981, 149) states that ‘there is nothing that keeps our practices in line except the reactions and responses we learn in learning them’. Google Scholar
  76. 103.
    Baker and Hacker (1994, 170–171, 233, 243–251). Google Scholar
  77. 104.
    In this sense, Hacker and Baker argue that agreement in behaviour can only be taken as an explanation to what is correct if the idea of an internal relationship between a rule and its instances of use is dismissed. See Baker and Hacker (1994, 172). Note that the claim that a community may be wrong about the contents of its own rules does not inevitably lead to support realist semantics. The community may be mistaken because the rule is the criterion to evaluate conduct. Here rules are understood as the object of debate, and not as metaphysical entities. Hence, a semantic judgement that dissents from the opinion of the majority still makes sense (and may even be right) insofar as it rests on a better characterisation of the standard of conduct than other judgements do. Thus, the debate about the contents of the rule is not descriptive of shared linguistic conventions, but dialectic. Google Scholar
  78. 105.
    Baker and Hacker (1994, 173–179, 243–251). In a similar sense, see Pears (1991, 280) and McGinn (1984, 77–84). According to Stavropoulos, Wittgenstein’s project seeks to maintain a negative thesis. The distinction between the correct and the incorrect application of a rule rests neither in a pattern hidden in a secret book, nor in a particular interpretation, view of the world, or anything we agree on. Wittgenstein’s thesis implies denying that the normativity in the application of concepts may be reduced to something simpler. See Stavropoulos (1996, 154–155). Google Scholar
  79. 106.
    McGinn (1984, 78, 89–91). Google Scholar
  80. 107.
    Schauer (1991, 56 footnote 5, 58–59). See also Bix (1993, 65, 71). Google Scholar
  81. 108.
    Bix (1993, 62, 179). Consensus on the form of life as the criterion for meaning invalidates the thesis of an internal relationship between the rule and its instances of use. This, in turn, leads to the sceptical problem. The situation would be different if resort to this consensus were with a view to explaining why we follow the same rules, rather than to identifying when is it the case that we are following them correctly. Google Scholar
  82. 109.
    Marmor (1992, 19). Google Scholar
  83. 110.
    Hart (1983, 106). See also Hart (1994, 126). Google Scholar
  84. 111.
    Hart (1994, 126–127). Google Scholar
  85. 112.
    In this sense, positivism seems to approach Kripké s idea on the impossibility of private language. Kripke rejects the plausibility of accounting for rule-following in terms of an analysis of either the circumstances or the standpoint of the acting agent without referring to her membership of some community, Kripke (1982, 108–110). Google Scholar
  86. 113.
    On this point see, among others, Bix (1993, 64–65) and Coleman and Leiter (1995, 253). Google Scholar
  87. 114.
    Bayón claims that Marmor’s version of rule-following does not overcome the problem about how to link meaning and consensus because it does not supply sceptics with an explanation on how this internal relationship comes about. See Bayon (1996, 16, 17). It could thus be argued that positivism does not explicitly reject consensus as the mediator between a rule and its instances of use. Google Scholar
  88. 115.
    I already pointed out in previous sections that the idea of a social reference as the determining factor of meaning only introduces a change in the ontology to be identified to grasp meaning. Although it is not assumed that legal terms are indexical (they do not designate natural classes of objects), their reference is based on social facts which are reducible to collective linguistic behaviour. Google Scholar
  89. 116.
    Shiner (1992b, 89–90). Google Scholar
  90. 117.
    On this point, see Hart (1983, 105), Kress (1989, 296–297), Barak (1989, 40–41), Conklin (1981, 231), Schauer (1985, 415) and Parent (1980, 99). Google Scholar
  91. 118.
    It is a common assumption within positivism that the need for discretion is the counterpart of the absence of right answer. See Bix (1993, 26) and Schauer (1991, 222). Google Scholar
  92. 119.
    It should be noted, as indeed Schauer does, that a case may be normatively determined, that is, it may be an easy case, and yet its inclusion in the semantic contents of a rule may be indeterminate. Hence, in order to link normative indeterminacy to semantic indeterminacy a notion of rule that goes beyond a simple canonical formulation is needed. Sometimes a rule may be a combination of a formulation and a first order justification. Then, semantic indeterminacy arises when the contents of both of these elements is undetermined, Schauer (1991, 210–212). See also Hart’s and Raz’s reflections on the role of principles and rules of interpretation as factors reducing the semantic indeterminacy of rules, Raz (1972, 846) and Hart (1994, 126). Note that these observations simply intend to point out that normative indeterminacy does not result from the semantic indeterminacy of a rule as an isolated entity, but from the semantic indeterminacy of law in general with respect to some specific case. Google Scholar
  93. 120.
    It was already pointed out in the previous chapter that these are ‘intermediate cases according to Barak’s terminology. They are thus distinguished both from those for which a legal answer is readily available to the interpreter, and from those for which there is no right legal answer, Barak (1989, 39–40). Google Scholar
  94. 121.
    In connection with this point, see MacCormick (1978, 86–93). Google Scholar
  95. 122.
    It is typical of positivism to reject this notion of ‘hard case’ as a source of relevant difficulty to the knowledge and application of law. Schauer, for instance, argues that if the difficulties in a case depend on a rule leading to unjust results, then the notion of rule itself eventually collapses with that of its justification. For Schauer this does not imply that there are no legal mechanisms to overcome this problem. Most legal orders include what Schauer calls ’rule avoiding norms’. These norms allow the judge to reduce the rigidity of the rules in the event of unjust or absurd results. Nevertheless, both, this type of norms and the rules that justify further rules, are of a legal nature. See Schauer (1988, 515–520; 1991, 53–56, 211). For the positivist thesis on the differentiation between law and morality, see Hart (1983, 49–87; 1994, 185–212) and Raz (1979, 37–52; 1990, 162–170). Google Scholar
  96. 123.
    will not elaborate any further on factually hard cases because my interest here is in the knowledge of law. Google Scholar
  97. 124.
    See Marmor (1992, 127) and Waluchow (1983, 335–336). Google Scholar
  98. 125.
    The reader is surely familiar with the fact that this is one of the most controversial points in legal positivism. It could be argued that the thesis of the social sources does not prevent moral criteria from being part of a rule of recognition. Nevertheless, positivism does not accept that the legal validity of a specific normative act should necessarily depend on its moral justification. Hence, the existence of a right legal answer is independent of its moral evaluation. As pointed out in a previous footnote, this does not exclude the possibility that the legal system may itself contain mechanisms to diminish the morally unacceptable consequences to which the application of legal rules may lead. See Waluchow’s account of inclusive positivism, Waluchow (1994, 80–140). Google Scholar
  99. 126.
    Marmor claims that, unlike rules, paradigms cannot be misunderstood or followed incorrectly. The paradigm may be respected or not, but not respecting it does not presume a lack of understanding. See Marmor (1992, 21–22). Google Scholar
  100. 127.
    See Endicott (1994, 22). Google Scholar
  101. 128.
    Endicott (1994, 22–23). Google Scholar
  102. 129.
    See Chapter 1 for a detailed description of the zone of reasonableness. Google Scholar
  103. Shiner’s assertion that sophisticated positivism assumes a theory of truth as assertability in hard cases is wrong for such a claim would presume that this theory regards the possibility of discussing even these cases in terms of right legal answer, that is, that the truth value of propositions is preserved in hard cases. Let us refer the reader to the link between the notions of hard case, absence of right answer and limited strong discretion pointed to by Bix (1993, 26-27) among others. See also Hart (1994, 131-132). Google Scholar
  104. Dummet (1978, xxi-xxiii, 14–15, 162–165, 175–176, 275–276, 315–316). Google Scholar
  105. 132.
    Dummet (1978, 146). Google Scholar
  106. Recall that it is on these grounds that Marmor claims the relationship between realism and bivalence to be asymmetric. It is possible to hold an antirealist position and accept the principle of bivalence. Marmor writes that it is plausible to envisage an artificial language which respects bivalence without assuming realism. Thus, even though in this language the truth-value of statements will be fixed, their truth or falsity will not be conditional on the existence of an objective reality independent of our knowledge. Marmor (1992, 88; 1995, 186–188). It will be shown in later chapters that the assumption of antirealist semantics without abandoning bivalence can be found in Ronald Dworkin’s theory of law. Google Scholar
  107. Wright (1993, 433–434). Google Scholar
  108. Wright (1993, 434–436, 458–465). Similarly, McDowell distinguishes classical realism from bare realism. While the former maintains the principle of bivalence for statements transcending verification, the latter only claims that the assertion that the truth or falsity of these propositions may transcend our epistemic abilities is plausible. In McDowell’s opinion, these forms of realism are independent of one another, and hence bare realism need not accept bivalence for those statements that transcend verification. See McDowell (1993, 42–66) and Wright (1993, 85, 434). Google Scholar
  109. For this reason, Wright claims that the principle of bivalence as a substratum to realism should be understood in the sense that any statement that is neither vague nor ambiguous is necessarily either true or false, Wright (1993, 458, 464). Similarly, Putnam (1995b, 76) argues that if the principle of bivalence needs to be abandoned simply because the situation encountered has no right answer, e.g. there is no right answer to the question of which of two children is taller if it is the case that both of them are the same height, then the principle of bivalence could never have been accepted as a logical principle in the first place. Google Scholar
  110. 137.
    See, for instance, Redondo (1997, 189–191). Google Scholar
  111. 142.
    Brink (1989, 183). Google Scholar
  112. A number of authors have directed their criticism of Hartian positivism against its inability to account for legal disagreement on the truth-value of legal propositions without having to assume an absence of right answer for these cases. In this sense, see Dworkin (1986, 45–46), Brink (1988, 14–116; 1989, 181¬191) and McDowell (1981, 150–155). Google Scholar
  113. On this association, see Shiner (1992b, 90, footnote 56). 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