The Weak Discretion Model

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


Weak discretion was introduced in Chapter 1 as one of the possible uses of ‘judicial discretion’. It was broadly defined as the need for reasonableness and good judgement in identifying the course of action prescribed by the law when there exists a right legal answer. The different approaches to weak discretion arise in connection with what the nature of the link between right answer and the epistemic abilities of interpreters is taken to be: if the right legal answer is assumed independent of interpreters’ considerations, difficulties in identifying it lead to a need for weak discretion; by contrast, if the right legal answer is considered to be dependent on the epistemic abilities of the interpreting agents, weak discretion merely embodies the implementation of a particular theory or conception of law. Before embarking on a full analysis of what I will call ‘the weak discretion model’, I would like to stress that these are the two approaches underlying the debate on the existence and the extent of weak discretion in legal adjudication.


Social Practice Legal Theory Paradigmatic Case State Coercion Legal Practice 
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. 2.
    See Raz’s, Mackié s, MacCormick’s, Greenawalt’s and Woozley’s papers in Cohen (1991).Google Scholar
  2. 4.
    Dworkin (1986, 90; 1986, 412). Although Dworkin always associates adjudication to the judicial application of norms, he occasionally argues that the link between theory and practice goes beyond the relationship between theory and judicial adjudication. According to Dworkin, the arguments of citizens and jurists (in other words, the statements of any participant) are also to be considered part of legal practice, Dworkin ( 1986, 413 ).Google Scholar
  3. 8.
    A statement such as `Law is what the legislator dictated intentionally’ is what Dworkin has in mind at this point. This statement differs from ‘judges think that law is what has been intentionally dictated by the legislator’. Dworkin claims that legal theory is concemed with the first kind of statements. On this point, see his response to Hart and to Hart’s view that legal theory describes what the legal community in general calls ‘law’, Dworkin (1994, 2–8). See also Hart (1994, 239–244).Google Scholar
  4. 10.
    On this point see Marmor’s (1992, 47) and Coleman’s (1995, 53) considerations on Dworkin’s legal theory.Google Scholar
  5. 12.
    Notwithstanding, his methodological proposal seeks to have implications on all aspects of an approach to law, and not just on the structure of judicial reasoning. See Dworkin ( 1986, 12, 413 ).Google Scholar
  6. 13.
    This standpoint is clearly reflected in the controversy between Dworkin and originalism, which has mainly turned on what is the correct reading of the American Constitution. While originalism argues that a constitution should be understood as a collection of standards with a specific contents primarily determined by the intentions of its founders, Dworkin insists that a constitution is in fact a collection of abstract principles that reflect an abstract legislative intention to be interpreted under the light of a theory about the rights and obligations of the citizens, i.e. a theory that offers the best possible justification for state coercion. See Dworkin (1993, Ch. 5; 1986, 179–181, 355–379 ).Google Scholar
  7. 14.
    Murphy (1990, 90) has objected to the possibility of adopting such a political conception in continental law.Google Scholar
  8. 15.
    See Dworkin (1986, 6–15, 413). ‘Practice’ is to be understood as regularity in behaviour or human interaction, assessable in terms of its adequacy to the purpose or value justifying its existence.Google Scholar
  9. 18.
    See Dworkin (1986, 6–15, 31–33, 37; 1985, 131). Although we shall not concern ourselves with a detailed analysis of Dworkin’s main argument against Plain Fact Theory, let us just indicate that it turns against the assumption that the facts conventionally associated with the existence of law are ‘brute facts’, i.e. facts whose existence and meaning is independent of any conceptual construct and which therefore carry their own identifying labels, Dworkin (1986, 7, 66; 1985, 137–138, 167–168 ).Google Scholar
  10. 19.
    Dworkin (1986, 31–33). Dworkin considers that positivism has traditionally defended this general conception in spite of important divergences about what the conventional criterion is. Austin, for instance, takes the existence of commands of a sovereign to whom obedience is customarily paid as the conventional criterion. For others, Hart among them, the conventional criterion is the social acceptance of a rule of recognition defining the validity criteria for legal norms. On this point, see Dworkin (1986, 3334). As was pointed out in the previous chapter, it should be borne in mind that Hart’s semantic theory does not accord with a semantic theory based on the notion of necessary and sufficient conditions. Notwithstanding, it can be considered a conventionalist semantic theory. See Hart ( 1994, 124–136; 1983, 89–98 ).Google Scholar
  11. 21.
    Dworkin claims that it is a mistake to think that we identify legal institutions by means of a shared definition of what a legal system should necessarily be like. See Dworkin (1986, 45, 46, 91; 1987, 10 ).Google Scholar
  12. 25.
    On the idea of substantial disagreement as a disagreement in judgements see Baker and Hacker (1994, 258–262).Google Scholar
  13. 32.
    It may be worth noting that Dworkin’s position appears to be very close to referencialist semantic theories such as Putnam’s or Kripke’s. It will be shown in later chapters that Dworkin does nevertheless not assume a causal theory of reference. (Note that it is not straightforward to present Putnam’s position towards a causal theory of reference, mostly because of his acceptance of the thesis of internal realism). Dworkin maintains that the reference is never a brute fact, but an interpretive object or state of affairs. As such, it is the product of the best theory about what a social institution is like. Thus, although the truth-value of a proposition depends on the reference rather than on a convention, the reference is limited to a coherent set of convictions about how things are. On the latter point, see Dworkin (1983, 292–303; 1985, 169–174; 1984, 275–278; 1991b, 360–365; 1996, 87–94 ).Google Scholar
  14. 35.
    Aarnio ( 1987, 218) shares the position of Dworkin on this point. He stresses, referring to those who take interpretation to be a mere linguistic matter, that `those who have this view have misunderstood the concept of language. If language, as a semantic category, is seen together with the form of life, and the action element in language is picked out, the analysis of language can specifically be said to be the analysis of matters. The analysis of language is the analysis of the form of life. Through it we can specifically understand our life, the acts of which our form of life is composed’.Google Scholar
  15. 38.
    Dworkin (1986, 50). Note that Dworkin distinguishes these two types of interpretation and establishes a hierarchy between them: constructive interpretation takes conceptual priority over conversational interpretation because the latter can be reconstructed as a form of constructive interpretation. We will defer further consideration to this point till later chapters. However, note that this consideration suggests a tendency to view constructive interpretation as the general way of assigning meaning to any phenomenon or object. See Dworkin (1986, 53) and Marmor ( 1992, 48 ).Google Scholar
  16. 43.
    For a classical discussion of the distinction between concept and conception, see Rawls (1971, 5) and Hart (1994, 246 footnote 26). This question had already been addressed by Gallie (1956, 167–180) though he used a different terminology. It is worth noting that Dworkin uses the same terminology as Rawls andGoogle Scholar
  17. 44.
    The possibility of a private practice of rule-following was discussed in the previous chapter with reference to the conditions that would enable Robinson Crusoe to claim to be following a rule. See Baker and Hacker (1994, 170–179).Google Scholar
  18. 54.
    For this reason, although the idea of paradigmatic cases could virtually draw Dworkin’s position closer to Hartian semantics, a significant difference in approach seems to persist. Hart tackles paradigms from the standpoint of conventional semantics. He seems to suggest that the semantic view on the necessary and sufficient conditions is to be overcome, and that certain necessary (though not sufficient) criteria together with consensus about paradigmatic cases determine the convention establishing the truth conditions of legal propositions. Nevertheless, these elements are not interpretive but belong to the convention determining the meaning. For discussion, see Hart (1983, 89–92; 1994, 244–248, 255–259 ).Google Scholar
  19. 55.
    Dworkin urges to distinguish paradigms of interpretive truth from analytic truths. He does not claim that statements about these paradigms be true by definition, as it is often said to be the case about the statement that ‘single individuals are those who are not married’ in connection with the concept of single. See Dworkin (1986, 72, 92). Dworkin’s position on paradigmatic cases does not imply accepting the distinction between synthetic and analytic truths. On the contrary, his general perspective develops along the lines of those who try to blur this distinction. With regard to this point, see Quine’s criticism to empiricism in Quine ( 1953, 20–46 ).Google Scholar
  20. 58.
    Although a new radical interpretation may appear absurd at a specific moment in time, it may eventually prevail as the best way of understanding the social institution in question. See Dworkin ( 1986, 72, 89 ).Google Scholar
  21. 59.
    The notion of absolute paradigmatic cases can be exemplified with the term ‘game’. It can be argued that the only reasonable way to explain the meaning of ‘game’ is by means of paradigmatic cases that define the concept. Although Wittgenstein’s idea of game seems to develop along this line, he proposes family resemblance as the connection among the members of a family. This connection could also be viewed as the search for a minimal conceptual link between the paradigms, however remote and weak. This would distance him from the strict notion of absolute paradigmatic case. See Wittgenstein (1953, secs. 65–77).Google Scholar
  22. 63.
    Dworkin ( 1986, 65, 91–92 ). The beliefs about these data do not determine the truth conditions of legal propositions. For Dworkin they do not constitute the legal ontology but pre-conditions for a meaningful discussion on the truth conditions. The legal ontology will be conditional on the meaning assigned to these data by means of interpretation.Google Scholar
  23. 64.
    The legal ontology will be conditional on the meaning assigned to these data by means of interpretation Dworkin (1986, 75).Google Scholar
  24. 65.
    In the context of literary interpretation, Dworkin considers that this contextual theory is the theory that determines what the literary work is. Thus, interpreters’ convictions should enable them to recognise that their interpretation turns on a specific literary work and hence, although this judgement may be controversial, their discussion will only be meaningful if they share one and the same contextual theory. For discussion see Dworkin (1986, 66; 1985, 150–151, 167–168 ).Google Scholar
  25. 69.
    This, however, does not imply that a given interpretation should account for all pre-interpretive data, but only that it should have some general criterion facilitating the choice of sufficient relevant data to reconstruct the practice as a whole. See Dworkin ( 1986, 66, 230 ).Google Scholar
  26. 72.
    Dworkin ( 1986, 52–53; 231 ). Recall that the relevant data are those that can be part of a plausible reconstruction of the practice as a whole. Which one of the possible interpretations of the practice best reflects its general justification is, precisely, the object of discussion at the post-interpretive stage.Google Scholar
  27. 73.
    Dworkin appeals to courtesy for an example on this point. Although it may be agreed that courtesy is a way of showing respect and deference, the debate about how best to demonstrate respect may influence what is perceived as a sign of respect in interactions between participants. Thus, removing one’s hat when saluting a woman may initially be taken as a paradigmatic case. At a different moment in time, however, in the light of a more powerful interpretation of the general justification, this case may no longer qualify as a paradigmatic case of courtesy. On this point, see Dworkin ( 1986, 48–49, 66 ).Google Scholar
  28. 75.
    This assumption is needed for the interpreter to view herself as an author in a chain novel rather than the creator of a new literary work. Not allowing for this possibility because of conceptual reasons is equivalent to denying her very role. This is, according to Dworkin, the position defended by internal scepticism, the theory that claims that the best reconstruction of a practice is that which asserts that no such practice exists. This discussion shall be pursued further and the distinction between internal and external scepticism addressed again in later chapters. For discussion, see Dworkin (1986, 78–85, 266–267; 1985, 175–177 ).Google Scholar
  29. 86.
    It is important at this point to highlight the difference between something being the object of discussion, and it being its result. Assuming an interpretive theory ultimately implies accepting that the contents of law is the result and not the object of interpretation. Quite often the term ‘prima facie law’ is used to refer to the object of interpretation. This expression is nevertheless misleading for it may appear to suggest that ‘prima facie law’ is, in fact, the law; thus, the arguments that aim at reconstructing these provisional data can be viewed to fall outside the legal context. They embody proposals about what law should be like in terms of parameters that belong to quite a distinct discursive context. This way of understanding ‘prima facie law’ directly implies dismissing the interpretive theory ab initio. In the context of interpretive theory, prima facie law is the provisional data that, if not interpreted, are meaningless. Some authors consider that Hartian positivism, unlike Dworkin’s theory of law, views law as a social phenomenon consisting of prima facie law solely. For discussion on this point, see Gavison (1987, 29–33) and Guest ( 1992, 30 ).Google Scholar
  30. 87.
    As was already pointed out, the interpreter need not account for every pre-interpretive datum but only for those essential to achieve an omnicomprehensive view of the legal practice. In this sense, Dworkin states that a particular conception of law may exclude data such as old-fashioned legislative decisions on the grounds that they no longer conform to a global reconstruction of the practice. See Dworkin ( 1986, 99, 154 ).Google Scholar
  31. 89.
    Predictability and legal certainty demand that every citizen be able to know his or her legal rights and duties prior to his or her actions. See, for instance, Aamio ( 1987, 3–5 ).Google Scholar
  32. 90.
    Dworkin agrees that pragmatism could be reconstructed as a position that rejects the interpretive concept of law outlined above. In this case, pragmatism would embody a radical form of scepticism about the existence of law and about the very idea of justification of state coercion. Nonetheless, it could also be portrayed as an interpretive conception provided it uses the language of legal rights and duties, at least at the formal level. Pragmatism as an interpretive theory posits that the contents of these rights and duties is fully contained in the convictions of every judge when she makes the actual decision. See Dworkin ( 1986, 94–95, 151–155 ).Google Scholar
  33. 95.
    Dworkin (1986, 167). In Chapter 5 we shall analyse the concept of coherence and differentiate it from derivability and logical consistency. See MacCormick’s (1984, 235–236) for an illuminating discussion of this point.Google Scholar
  34. 101.
    Dworkin ( 1986, 178–184 ). The following may prove a useful example to illustrate the problem of checkerboard or diversified solutions. Let us assume that diverging opinions about whether the consumption of drugs should be allowed occur within a community. Let us further assume that the different opinions occur in distinct geographical areas, e.g. those in the north would permit consumption, whereas those in the south would not. It may be the case that the most equitable decision turns out to be the diversified one, i.e. that which makes consumption legal in the north and illegal in the south. It is thus ensured that the majority of the citizens have similar opportunities to influence public decisions. Moreover, the diversified solution may be preferable, in terms of justice, to a unitary one. Those who are of the opinion that permitting drug consumption is a moral mistake will find it fairer that consumption be allowed in the northern strip only rather than in the entire geographical area. Integrity, by contrast, requires that the same solution be adopted throughout the territory because it does not contemplate the possibility to differentiate between similar cases. In this case, global coherence overrules fairness and justice.Google Scholar
  35. 102.
    Dworkin distinguishes three different models of community: circumstantial community, community of statute, and fraternal community. Circumstantial communities are those whose members perceive their association as a de facto matter, accidental, because they understand there to be no further connection among themselves. In communities of statute, members reach an agreement to establish a convention about how to guide conduct. In fraternal communities, members do not only accept a certain convention to guide conduct, but also share the following beliefs: a) the community is to be governed by shared principles (agreement of conviction); b) each participant is individually responsible for respecting these principles; c) the community speaks with a single voice and hence diversified solutions are unacceptable. Dworkin (1986, 206–216).Google Scholar
  36. 105.
    In this sense, Dworkin distinguishes between two types of integrity relevant to adjudication: inclusive integrity, and pure integrity. The former demands that the judge reach some balance between the three above-mentioned virtues–justice, fairness, and legality. By contrast, pure integrity is achieved by favouring justice over the other two values. Dworkin points out that pure integrity is only possible in an ideal world in which justice would offer the best justification of law. In the real world, however, only inclusive integrity can be achieved. See Dworkin ( 1986, 405–406 ).Google Scholar
  37. 106.
    Alexy, for instance, introduces the requirement of universality in legal reasoning on the basis of the following rule: a speaker is allowed to assert those judgements of value and duty that he or she would assert in any situation that he or she considers to be identical in all relevant aspects to the one under consideration. See Alexy (1989, 190–191, 275–278). The principle of universality may be realised in two demands, one addressed to the legislator and one to the judge. The former is the principle of material equality. It prescribes that the (choice of) scope of applicability of a norm should be justifiable in order to avoid discrimination. The latter is formal equality, which requires judges to give the same legal solution to all cases that they consider equal in all relevant aspects. See Alexy (1993) and Dworkin (1986, 219–224).Google Scholar
  38. 107.
    This terminology is Raz’s, who defines strongly monist coherence as the criterion of adequacy or fit into a set of principles, which are reducible to a single one. By contrast, a weaker form of monist coherence would be adequacy to a small number of principles that are not reducible to a more general one. See Raz (1992, 286) and Kress ( 1996, 542–546 ).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