SOME THEORETICAL REFLECTIONS ON STATUTORY INTERPRETATION WITH AN EMPHASIS ON PURPOSIVE INTERPRETATION IN LAW AND ITS LIMITS

Transcription

1 [Draft - Ph.d. seminar, Bergen, 7-9 June 2010] SOME THEORETICAL REFLECTIONS ON STATUTORY INTERPRETATION WITH AN EMPHASIS ON PURPOSIVE INTERPRETATION IN LAW AND ITS LIMITS Róbert R. Spanó Contents: I. Structure of Paper II. Theoretical Premises of Statutory Interpretation III. Hermeneutics and Context in the Interpretation of Legal Texts IV. The Concept of Purposive Interpretation a. Preliminary remarks b. Aharon Barak s Theory c. Reflections and Three Examples from Case Law V. Limits of Purposive Interpretation a. Textual basis of purposive interpretation and legal texts as political compromises b. Levels of abstraction in constructing the purpose of a legal text c. Legal interpretation or non-interpretive doctrines Professor of Jurisprudence and Statutory Interpretation, Vice-Dean, Faculty of Law, University of Iceland, Cand. Jur (Univ. of Iceland), Mag. Jur (Univ. of Oxford). Acting Parliamentary Ombudsman of Iceland from 1 January 2009 (Althingets Ombudsmand).

2 2 I. Structure of Paper Statutory interpretation involves the interpretation of a legal text, domestic or international, with the aim of deciding on its substantive meaning in general, or with respect to a defined set of facts. In this paper I will in chapters II and III discuss some theoretical premises of statutory interpretation and its relationship with hermeneutics, the classical philosophical school of textual interpretation. I will attempt to make and defend two claims: The first is as follows: Each time a person attempts to interpret a legal text, the method used reflects a normative choice that must be justified in light of the nature of legal texts as authoritative, emanating from an institution that society accepts has the power to promulgate law. The second claim is this: The interpretation of a legal text does not involve the search for a preexisting truth or fact, but consists of the normative construction of a rule based on a legal text. On this theoretical basis I will in chapters IV and V discuss a distinct type of interpretive method which is known as purposive interpretation, or formålsfortolkning in the West-Nordic languages. Firstly, I will explain as possible what I mean by purposive interpretation. I will draw heavily on the general theory of purposive interpretation argued forcefully by the eminent Israely justice, Aaron Barak, in his seminal book on the subject. 1 Furthermore, I will introduce three examples drawn from case-law from trans- and supranational jurisdictions in Europe - examples which I will attempt to use to clarify my arguments. I will subsequently argue that a purposive method of statutory interpretation has inherent limits. These limits emanate from a concept of statutory law that derives its normative force from the general acceptance in its jurisdiction that the rule in question is an explicit manifestation of a decision by a legislature. The legislature in this sense being a lawmaking organ distinct from a judiciary, irrespective of whether the legislature is national or international, democratic or based on state-consent as in the case of most international treaties. II. Theoretical Premises of Statutory Interpretation It is common for a Ph.d. student s project summary to include a section on methodology which states that the student intends to base his analysis on the traditional legal method, or dogmatic method based on the theory of legal sources, rettskildelæren, or the descriptive approach or method. But what does this mean? For sure, we lawyers share a common linguistic understanding of these methodological concepts. But if 1 Aaron Barak: Purposive Interpretation in Law. Princeton University Press, Princeton (2005).

3 3 students are attempting to proclaim that they intend to pursue a form of legal science, in the sense of describing objectively what the law is in the relevant area, without making any normative assessments or choices along the way, I would argue that they are misleading themselves. For within that realm of legal science, where lawyers attempt to address what the law is in a particular area, inevitably normative choices of method will have to be made. Furthermore, if there is such a thing as the traditional legal method or legal dogmatic approach, it is important to bear in mind that this method is relative to the legal source one is using. For example, one can argue that the flexibility one has in infusing external value choices to an analysis of existing law, when dealing with statutory written law, is much more limited than where the legal source being analysed is unwritten general principles of law or customs. So the legal method one uses, or is justified in using, varies with the legal source one is dealing with. I now turn to the topic in question, the interpretation of legal texts, i.e. constitutional provisions and statutes in domestic law of a primary nature and/or provisions of international treaties having the force of law. Before dealing with the theoretical issues of the interpretation of legal texts stricto sensu it is important to elaborate somewhat on the concept of legal interpretation that I am using. The concept of interpretation has various meanings in law which can be, for our purposes, classified as narrow or wide. The former, the one used in this paper, comprises the concept of legal interpretation as a rational actitivity that gives meaning to a legal text. The latter, in accordance for example with professor Ronald Dworkin s expansive notion of interpretation, defines law itself as an interpretive process, irrespective of the source being interpreted. 2 Modern, developed, legal systems are in general regulated ever more by legal norms in the form of written texts, promulgated by an organ having legislative powers, as opposed to the law developing through unwritten rules or customs. It is therefore one of the fundamental tasks of lawyers to apply rules constructed from legal texts to factual situations. A theoretical understanding of the process of statutory interpretation is thus increasingly important so lawyers can assess whether the methods being used are legitimate or must be cast aside as being in non-comformity with the essence of this enterprise. This last distinction, between legitimate and illegitimate methods of statutory interpretation, of course begs the question: How does one distinguish between the 2 Ronald Dworkin: Law as Interpretation, 60 Tex. L. Rev. 529 (1982): Legal practice is an exercise in interpretation not just when lawyers interpret documents or statutes but also generally. Propositions of law are not simply descriptive of legal history, in a straightforward way, nor are they simply evaluative in some way divorced from legal history. They are interpretive of legal history, which combines elements of both description and evaluation, but is different from both.

4 4 two? For sure, if there is such a thing as the traditional legal method or dogmatic approach, which involves the interpretation of statutes, one has to assume that the tradition being invoked has developed on the basis of a consensus that the method has distinguishing characteristics different from other possible methods for deciding on the meaning of legal texts. What then are these characteristics? Some texts are legal because they fulfill the following condition: They are a manifestation of a decisional process by an organ that society has by consensus granted the power to issue norms that are to have authoritative force in social relations. In this sense a text in a work of literature, in a newspaper article or even in a wider sense the signs and symbols on an historical artifact, are not legal texts. When one is to construct the meaning of a legal text the process is therefore by its very nature and purpose different than the process in other areas of textual interpretation. The method one uses to interpret legal texts must thus in general comprise objective factors that have as their main purpose the enforcements of those norms that the legislator intended to create for the regulation of society. 3 In this respect the most important criterion for statutory interpretation is the accurate preservation of the normative link between the meaning (rule) given to the text and the language used by the legislator when the text was promulgated. The limits of the text therefore set the limits of interpretation in law, and the limits of language set the limits of the text. An activity is interpretive if it confers meaning on a text that is consistent with one of its (explicit or implied) meanings in the (public or private) language of the text. Giving a text meaning beyond its semantic meaning is not an act of interpretation and it must therefore rely on non-interpretive doctrines as I will further discuss in chapters IV and V. Interpretation thus ends at the point at which language ends. Furthermore, interpretation can give a text an expansive or a narrow meaning. However, for an interpretation of a legal text to be legal at all, it must at least confer upon the text a meaning that it is capable of bearing in the language in which it is expressed. 4 III. Hermeneutics and Context in the Interpretation of Legal Texts A legal text has no true meaning. The most to which we can aspire is the proper or legitimate meaning of the text. Within the limits of the text, the normative message that arises is ascertained through interpretation of the text, which cannot therefore be proven true or false. 5 As described above a text must fulfill the condition, for it to be legal, that it be a manifestation of a decisional process by 3 Róbert R. Spanó: Túlkun lagaákvæða [The Interpretation of Statutes]. Codex, Reykjavík (2007), Barak: supra, note 1, at Barak: supra, note 1, at 9-10.

5 5 an organ that society has by consensus granted the power to issue norms that are to have authoritative force in social relations. Hence, the interpretive system that is legitimate in a democratic regime is not necessarily the legitimate system of interpretation in a totalitarian regime. The system of interpretation furthermore depends on the constitutional requirements in force. 6 The object of legal interpretation is the text. Every text requires interpretation as a text cannot be understood without being interpreted. A text therefore has no preinterpretive meaning. The common determination by lawyers that a text s instructions are plain or obvious, and thus do not require interpretation, is an interpretive determination that succeeds, rather than precedes, the interpretive act. The real question in law is therefore not whether a plain text requires interpretation but rather what rules of interpretation are needed to arrive at the text s plain meaning. 7 As the interpretation of legal texts is only one category of a more general project of textual interpretation it can be helpful to gain insights from hermeneutics 8 when attempting to analyse in more depth the theoretical premises of legal interpretation. Hermeneutics address on a philosophical level how a text is given meaning and analyzes the triangular relationship between the author of the text, the text and its reader. The modern day theories of hermeneutics form a part of an attempt to lay the foundations of a theory of science proper to the social sciences such as law and intended to serve as a reply to and a defence against the rise and claims of the natural sciences and the onslaught of criticism from positivist theories of science. 9 The various hermeneutic theories emphasize the problem created by the gap in time between understanding at the time of creation and understanding at the time of interpretation. An interpreter s horizons are the present, making it difficult for him or her to understand the text whose horizons existed in the past. In his seminal work, Truth and method, the German philosopher Hans Georg Gadamer ( ) further developed this notion be relating his idea of tradition to a reworked notion of prejudice, which he understood as pre-judice or pre-judgment, in other words as that which makes any kind of discrimination possible. A prejudice in this sense is not 6 Barak: supra, note 1, at Barak: supra, note 1, at The Greek term hermeneuein, meaning to interpret, is the root from which the word hermeneutics is derived. Key figurs in the development of early modern hermeneutics are J.C. Dannhauser, the rationalist philosopher Benedict Spinoza, Fredrich Ast, Chladenius and Friedrich August Wolf. They devised rules for the accurate interpretation of biblical, legal and classical texts. These figures may be the founding fathers of modern hermeneutics but it was Friedrich Schleiermacher ( ), the German philosopher and theologian, whose work was to have the biggest impact in the development of Romantic hermeneutics, see Chris Lawn: Gadamer A Guide for the Perplexed. Continuum, London (2006), See Stig Strömholm: Legal Hermeneutics Notes on the Early Modern Development. Scand. St. Law (1978), 213.

6 6 a distorting form of thought that must be shaken off before we see the world aright. For Gadamer prejudices are present in all understanding. Against Enlightenment claims that reason, detached from historical and cultural perspectives gives a test for truth, Gadamer claimed that we are irredeemably embedded in language and culture, and that the escape to unclouded certainty via rational method is a chimera. 10 According to Gadamer s theory, context is therefore crucial in the interpretation of texts creating the hermeneutic circle, a paradox in which the whole cannot be understood without understanding its parts, but the parts cannot be understood without understanding the whole. 11 The hermeneutical approach to statutory interpretation manifested in the hermeneutic circle is thus premised on the argument that the correct reading of a legal text cannot be known unless we know, roughly, the text as a whole; we cannot know the text as a whole unless we know particular passages. We cannot know the meaning of a word unless we know the meanings of the surrounding words and of the text as a whole; knowing the meaning of the whole involves knowing the meaning of individual words. We cannot fully understand the text unless we know the author s (legislators) works as a whole, but this requires knowledge of the texts and other events that constitute his life. We cannot fully understand a text unless we know about the whole (legal) culture from which it emerged, but this presupposes a knowledge of the texts and so on that constitute the culture. 12 Applying normative doctrines of interpretation 13 based on hermeneutics to the enterprise of statutory interpretation therefore demonstrates the importance of context when constructing the meaning of legal texts. The context in this respect is both internal and external to the text in question. 14 The internal context of a legal text involves the textual and grammatical analysis of the text in the elaboration of its semantic frame. The semantic frame thus encompasses all the possible meanings a text can entail. When choosing between possible meanings (or rules) that fall under a legal text s semantic frame, the interpreter must however also be mindful of its external context, which comprises those factors external to the text that can properly be consulted when constructing its correct meaning in light of a particular factsituation (or dispute) Chris Lawn: supra, note 8, at Barak: supra, note 1, at Michael Inwood: Hermeneutics, in Craig, E, and Floridi, L (eds.), Routledge Encyclopedia of Philosophy, Oxford, Blackwell, (1998). 13 Stig Strömholm, supra, note 7, at Róbert R. Spanó: supra, note 3, at See also my discussion of Aharon Barak s theory of interpretation and the importance of context in chapter IV.b in this paper. 15 Róbert R. Spanó, supra, note 3, at

7 7 IV. The Concept of Purposive Interpretation a) Preliminary remarks For practical purposes, I now turn to examining one particular, albeit popular, method of interpreting legal texts, especially in current European case-law, purposive or teleological interpretation. Before examining the conceptual framework of this method, it is important that I stress the context of my argument. I am only dealing with the use of purpose in deciding the meaning of a legal text, that is the substantive meaning to be given to a constitutional or statutory provision or a provision of an international treaty, in light of a factual situation creating a legal dispute. Note, that I do not as a principled matter distinguish here between the methods used when interpreting legal texts adopted in a domestic setting from those adopted on the basis of negotiations between sovereign states. It suffices for my argument that in all of these situations a text is adopted with the intention that it have the force of law. As argued in chapters II and III, an interpretational approach dealing with controversies about the meaning of these texts must thus preserve and enhance their nature and character as legal texts. An interpretion that does not fulfill this condition is not legal in any meaningful sense of that term. b) Aharon Barak s Theory There are several ways to describe the method of establishing the meaning of a legal text on the basis of the interpreter s construction of its purpose. Its general premise is however relatively straight-forward: When deciding on the meaning of a legal text, the primary focus is the construction and implementation of the underlying purpose of the text being interpreted. According to Aharon Barak the purpose of a legal text is a normative concept. It is a legal construction that helps the interpreter understand the legal text. He claims however that the purpose of the text is not a part of the text itself. The judge formulates the purpose, one the one hand, based on information about the intention of the text s author this Barak terms the subjective purpose and the intention of the legal system on the other the objective purpose. 16 Determining the purpose of a legal text always combines objective and subjective purposes, 17 the conclusion being the construction of the ultimate purpose which is used in deciding on the meaning to be given to the legal text being interpreted. Barak furthermore claims that a theory of purposive interpretation presupposes that the interpretation involves analysis of the text, not psychoanalysis of its author. The judge analyzes the text using interpretive criteria 16 Barak: supra, note 1, at Barak: supra, note 1, at 111.

8 8 formulated based on both subjective and objective sources. Then he states that the purpose of a statute is to bring about a certain kind of social change. The interpreter learns about this change from, among other things, the intentions of the members of the legislative body - but, he argues, legislative intent is distinct from purpose. 18 Barak goes on to state that the meaning of a word depends on its context. The context of the language consists of overlapping circles of diverse contexts. Context can be semantic, social, historial, and/or value-dependent, to name a few types. A judge must choose the relevant context of a legal text from this infinite list. He then askes: What is the relevant context of a legal text, and how does it differ from the relevant context of a literary or musical text? His answer is as follows: 19 I say that the relevant context of a legal text is the data that provides information about the text s purpose. We look to the language of the text, other texts, the history of the text, and the general values of the system in order to learn about the purpose of the text. The purpose of a legal text is the only relevant context for its interpretation as a normative creation. [...] The context of a legal text helps explain the meaning of the words from the point of view of their purpose. The different contexts the totality of the text s language, similar texts, history, and the legal system s fundamental values are relevant only insofar as they help the interpreter learn about the purpose of the text. A context that is not connected (directly or indirectly) to the purpose may help interpret a text as a work of literature, but it cannot help interpret the text as a normative construction. Before I analyse this argument, it is necessary first to elaborate somewhat on Barak s exposition on the multiple purposes of a legal text and their different levels of abstraction. Barak argues that purposive interpretation assumes that every legal text has multiple objective and subjective purposes. Sometimes, these purposes exist at identical levels of abstraction, what he calls horizontal purposes. Usually, however, the different purposes exist at varying levels of abstraction a situation he terms vertical purposes. Often, the relationship among the purposes is such that the judge can formulate the ultimate purpose which is used to determine the text s legal meaning without resorting to judicial discretion. Sometimes however, the purposes contradict each other. In those cases, the interpreter according to Barak must resolve the conflict while achieving synthesis and harmony among the various purposes. At times, he or she will have to exercise discretion Barak: supra, note 1, at Barak: supra, note 1, at Barak: supra, note 1, at

9 9 Barak furthermore describes what he means by levels of abstraction when determining the purpose that must guide the interpretive process. He states that two parameters determine the levels of abstraction: First, the level of particularity or generality of the subject regulated or addressed; and second, the extent to which the content at the core of the normative arrangement is value-laden. These parameters, he goes on to say, create a range of levels of abstraction. 21 At a low level of abstraction, the statute s purpose is the individual objective that the statute is designed to achieve. At a higher level, interpreters focus on the objectives that a statute of that particular category or type is designed to achieve. At the highest level are the fundamental values of the system which constitute the general objective purpose shared by all legislation. This last purpose, Barak says, is a kind of normative umbrella that extends to all norms in the system. 22 Barak proposes that interpreters should always prefer purpose at a low level of abstraction to purpose at a higher level. 23 In chapter V, I analyse Barak s arguments on these points in more detail. c) Reflections and Three Examples from Case Law I have now given a summary of the main conceptual elements of Justice Barak s theory of purposive interpretation in law. It is important to bear in mind that his approach to legal interpretation is not some scholarly eccentricity but well established in domestic, and indeed supranational, European jurisprudence in the modern age. Let me now reflect on Barak s arguments by taking three examples from Europe s international and supranational jurisdictions, some of which are relatively well known, others perhaps a novelty to some. Let s first take a trip down memory lane and refresh our memories with the case of Van Gend en Loos, 24 a preliminary ruling issued by the European Court of Justice (ECJ) on February 5, There the ECJ had to answer the question whether article 12 of the EEC Treaty (now Article 30 of the TFEU 25 ) had direct application within the territory of a member state, in other words whether nationals of such a state could, on the basis of the article in question, lay claim to individual rights which the courts had to protect. Article 12, as the current Article 30, prohibited Member States from introducing between themselves any new customs duties on imports or exports or any charges having equivalent effect. The Court began its reasoning towards an affirmative answer, as follows: 21 Barak: supra, note 1, at Barak: supra, note 1, at Barak: supra, note 1, at Van Gend en Loos, Case 26/62, [1963] ECR Treaty on the Functioning of the European Union.

10 10 To ascertain whether the provisions of an international treaty extend so far in their effects it is necessary to consider, the spirit, the general scheme and the wording of those provisions. I think it important to note at the outset that the structure of this reasoning seems, at least at first glance, to be formulated so as to give the spirit of the legal text in question priority within the interpretive process. This can also be inferred from the rest of the ECJ s reasoning in Van Gend en Loos, the next sentence stating as follows: The objective of the EEC Treaty, which is to establish a common market, the functioning of which is of direct concern to interested parties in the community, implies that this treaty is more than an agreement which merely creates mutual obligations between the contracting states. The ECJ decision thus most surely rests firmly and without reservation on a theory of purposive interpretation. The significance of the purposive approach to the substance of the decision is, I think, striking from a methodological point of view in legal interpretation. Remember that the wording of the provision in question was directed exclusively to the member states. This did not trouble the Court which stated that the The fact that under [Article 12] it is the member states who are made the subject of the negative obligation does not imply that their nationals cannot benefit from this obligation. The next example comes from the ECJ s sister Court, the EFTA Court, established under the EEA-Agreement. The date is 10 December 1998 and the case is the one of the Icelander, Mrs. Erla María Sveinbjörnsdóttir. The Court had to answer the question, in an advisory opinion, 26 whether a State would, as a matter of principle, be liable vis-à-vis an individual for not having adapted national legislation when it became party to the EEA Agreement, so that the individual would have a legal right accorded to him by a directive. As is well known, the EFTA-Court answered this question in the affirmative establishing the rule that the Contracting Parties to the EEA Agreement are obliged to provide for compensation for loss and damage caused to an individual by incorrect implementation of a directive incorporated into the EEA Agreement. Before the Court, the Governments of Iceland, Norway and Sweden, and the EC Commission, argued that neither the EEA Agreement nor the EC Treaty contained an express provision on State liability. Under Community law, the Govern- 26 Erla María Sveinbjörnsdóttir, Case no. E-9/97, 10 December 1998.

11 11 ments and the Commission furthermore submitte, that the principle of State liability had been established through the case law of the European Court of Justice. However, since the case law was largely based on special characteristics of the Community legal order that were not found in the EEA-Agreement, this case law could not be transferable to the EEA Agreement by virtue of Article 6 of the Agreement. The Court of course disagreed. The starting-point for its reasoning is important for my subsequent argument. The Court began with these words: The Court notes, firstly, that there is no explicit provision in EEA law establishing a basis for State liability on account of incorrect adaptation of national legislation. In the absence of an express provision in the EEA Agreement, the question arises whether such a State obligation is to be derived from the stated purposes and the legal structure of the EEA Agreement. The general aim of the EEA Agreement, as laid down in Article 1(1) EEA, is to promote a continuous and balanced strengthening of trade and economic relations between the Contracting Parties with equal conditions of competition and the respect of the same rules, with a view to creating a homogeneous European Economic Area. The Court concluded as follows: The Court finds that the homogeneity objective and the objective of establishing the right of individuals and economic operators to equal treatment and equal opportunities are so strongly expressed in the EEA Agreement that the EFTA States must be obliged to provide for compensation for loss and damage caused to an individual by incorrect implementation of a directive. I think one could plausibly ask here the following question: Does the emphasis in the Court s reasoning on the objectives of the EEA-Agreement possibly transgress the limits inherent in the use of the purposive approach when interpreting a legal text. In other words is the EFTA-Court s advisory opinion in Erla María a legal interpretation at all. If not, the opinion would I presume have to be justified by resorting to other theories of law and interpretation, e.g. theories based on the premise that at least some international judicial bodies possibly have some inherent powers of norm-creation that go beyond the traditional judicial role in domestic law? I will comment on this question in chapter V.c. Lastly, let me take an example from the case-law of the European Court of Human Rights. It deals with an interpretation of Article 4 of Protocol 7 to the Convention stating the principle of Ne bis in idem. The provision is as follows:

12 12 No one shall be liable to be tried or punished again in criminal proceedings under the jurisdiction of the same State for an offence for which he has already been finally acquitted or convicted in accordance with the law and penal procedure of that State. One could, I think, very plausibly argue that the text of this provision clearly envisages, and is limited to, the case where a person has either been convicted or aquitted in the traditional sense of these terms according to the law and penal procedure of the State in question, and that person is subsequently liable to be tried or punished again in criminal proceedings for the same offence. This understanding of the provision thus semantically seems to preclude the possibility that a State can violate the provision when a person has, outside the penal procedural sphere in domestic law, for example received a fine or another kind of administrative decision with punitive elements, or vice versa received a favourable decision by an administrative organ where such a measure could have been introduced, and that person is subsequently prosecuted and punished in a Court under the penal procedure of the State in question, as the former procedure was not penal but administrative. Despite this seemingly insurmountable textual hurdle, the European Court of Human Rights has from september 1995 interpreted the provision so as to include a prohibition against the repetition of proceedings, irrespective of whether they are based on administrative law or penal procedural law, on the condition that the administrative measure be of a criminal character in accordance with the so-called Engel Criteria developed by the Court while interpreting Articles 6 and 7 of the Convention. 27 I note that both the Norwegian and Icelandic Supreme Courts have endorsed this understanding in domestic cases. 28 The justification for this approach is, again, the underlying purpose of the ne bis in idem provision and the necessity of construing it in conformity with other convention provisions dealing with offences of a criminal nature. V. The Limits of Purposive Interpretation After having now reflected upon the conceptual elements of a theory of purposive interpretation by using three examples from European case-law, where this approach seems to dominate the interpretational outcome, let me know turn to my second part. 27 For an overview of the ECHR-case law, see the Grand Chamber judgment in Zolotukhin v. Russia, no /03, 10 February 2009, para See Hrd. June 8, 2006, no. 248/2006, and Rt. 2002, bls. 557.

13 13 I will now argue that there are inherent limits to the use of purpose as a tool for the interpretation of legal texts, and attempt to demonstrate that the examples given from the case-law possibly manifest the dangers when these limits are not respected in the practical application of legal texts. My argument has two parts: Firstly, I will argue that Aharon Barak is wrong as to the starting-point for the normative construction of the purposes of a legal text. Secondly, I claim that any legitimate use of purpose as an interpretive criterion must be based on a rather restrictive sense of the levels of abstraction used when the purpose is normatively formulated. Barak is therefore right to propose that an interpreter should always prefer purpose at a low level of abstraction to purpose at a higher level. Now to my first point. a) Textual basis of purposive interpretation and legal texts as political compromises Remember, Barak claims that the purpose of the text is not a part of the text itself. He states that the judge formulates the purpose, one the one hand, based on information about the intention of the text s author and the intention of the legal system on the other. 29 This argument seems therefore to be based on the premise that a text s purpose is a normative construction that is, as it were, not constructed within the confines of the text itself. The purpose is thus a normative concept that is formulated on the basis of sources that are external to the text being interpreted. With this in mind, I think it important to analyse in more detail what an interpreter, for example a judge, is doing when he is normatively constructing the purpose that will subsequently determine the meaning of the legal text forming the subject of his inquiry. I would argue that the construction of the purpose or, possibly and invariably the multiple purposes, behind a legal text is itself an act of legal interpretation. Therefore, the construction of the purpose cannot in my view be guided by methods that would be considered illegitimate when the focus of the interpretation is the text itself. Thus, to the extent that the underlying purpose of a legal text is used, when deciding on the latter s meaning in relation to a defined factual scenario, the purpose must, at least at the outset, be constructed within the semantic confines of the text itself. The purpose of a legal text cannot, therefore, be derived from external sources, as Barak would have it, to the extent that the normative concept of purpose being constructed and used in the interpretation does not have a reasonable basis in the text being interpreted. 29 Barak: supra, note 1, at 110.

14 14 Let me for clarification use the example of the ECHR s case-law on the scope of the Ne bis in idem rule as formulated in Article 4 of Protocol 7 which I discussed in chapter IV.c. Analysing the text of this provision, it seems clear to me that the purpose of the rule is to preclude the repetition of criminal proceedings conducted under the penal procedures of the State in question. If one however constructs the purpose of the provision by looking, as Barak would have it, to sources external to the provision, one could possibly conclude from information on the intention of the authors or from other sources that the purpose of ne bis in idem rules in general is to preclude the repetition of criminal proceedings irrespective of whether they are defined according to penal procedural law or administrative law. The Member States to the ECHR could most surely have drafted the text of the provision more broadly in this way, thus explicitly including under its substantive scope a case of the type I just mentioned, and thus precluding a clash between administrative decision-making and a conviction or acquittal in a Court in accordance with penal processes. But that is not what they did. By constructing the purpose of the provision on the basis of sources outside the semantic frame of the provision, the ECHR has used an interpretational approach that does not preserve and enhance the nature and character of the provision as a legal text emanating from discussions and agreements by the Members States to the Council of Europe. The Court has thus created a rule which only derives its normative legal force from the fact that it is espoused by an institution that is regularly obeyed by the memberstates to the Convention. It is a judicial invention. The rule, one could argue, is not the one drafted and adopted by the Member-States. Its reasoning based on the underlying purpose thus fails to provide an adequate justification for its existence because the formulation of the purpose is not reasonably based on the text of the provision. This brings me to another point that is important for this part of my argument. Legal texts are invariably a manifestation of political compromises by their authors. Of course, external sources can be an aid in the further elaboration of a legal text s purpose that has already been derived from a fair reading of the text. However, it seems impossible to adequately preserve the underlying political compromise of the legal text if its meaning is predominantly guided by the normative construction of purposes that are not, at least as a starting-point, drawn from the text itself and its semantic confines. Furthermore, it is unsound to think that the meaning of a legal text can in difficult cases of statutory interpretation be decided in accordance with the construction of an all-encompassing, rational and coherent purpose. In many cases at least the reason for disputes in hard cases of statutory meaning is, in fact,

15 15 disagreement between parties as to the underlying purpose of the text in question or at least on what level of abstraction the purpose should be formulated. b) Levels of abstraction in constructing the purpose of a legal text This brings me to my second point in my argument on the limits of purposive interpretation of legal texts. When constructing the purpose of a legal text in a search for its meaning, it is a crucial step in the process to formulate the correct levels of abstraction at which the purpose is defined. Barak is right in my view to argue for an approach that invariably prefers purpose at a low level of abstraction to purpose at a higher level. We, however, disagree I think when it comes to describing the process of formulating these levels. As my starting point in the purposive interpretation of legal texts are the range of semantic meanings of the text in question, it seems to me that the level of abstraction given to the text s purpose must be reasonably related to the text itself and not necessarily, or at least at the outset, external sources or information about its meaning. If the wording of the text is formulated narrowly, that should pull the interpreter towards a purpose formulated at a rather low level of abstraction. On the flip-side a very broadly formulated text, for example in some classic human rights provisions in a constitution or an international treaty, possibly justify the formulation of purposes at a higher level of abstraction. Remember, however, that even human-rights provisions can be narrowly formulated. I refer here for example to the textual definition of the ne bis in idem rule in Article 4 of Protocol 7 the European Convention, I discussed above. Bright-line rules are of course not possibly here, but this general approach seems to me to be more in conformity with the inherent legislative character of legal texts in the sense elaborated in chapter II. Furthermore, a lower level of abstraction is more suited to those hard cases where it is clear that the text in question is a manifestation of a nuanced political compromise between its authors. c) Legal interpretation or non-interpretive doctrines This now brings me in conclusion back to the advisory opinion of the EFTA Court in the case of Erla María Sveinbjörnsdóttir. When discussing this case, one must bear in mind at the outset that the Court was faced with the question whether the EEA-Agreement could, as a matter of principle, be interpreted to include a defined rule of tort liability of the Contracting States for incorrect adaption of a directive into national law. Remember, the Court began its reasoning by stating the evident fact that no express provision incorporating such a rule was to be found in the text of the Agreement. The opinion then explicitly goes on to characterise the

16 16 legal question as one whether such a State obligation could to be derived from the stated purposes and the legal structure of the EEA Agreement. As the subsequent reasoning demonstrates the Court arrived at its affirmative answer by constructing the objectives of the EEA-Agreement at their highest levels of abstraction and giving the purposes, thus constructed, explicit normative significance as the legal foundation for the existence of a specific rule of State-liability within the scope of the Agreement. Not only does the Court derive the rule itself as a matter of principle from these general objectives, but on the same basis goes on to identify and define the three conditions that form the substance of the rule, making reading the opinion seem somewhat akin to reading a statutory code. It is important to bear in mind that even though the EEA-Agreement is based on the objectives of homogeneity and of establishing the right of individuals and economic operators to equal treatment and equal opportunities, that does not necessarily mean that these objectives can be a vehicle for their attainment at all possible costs through judicial interpretation. In accordance with a legitimate theory of purposive interpretation, the levels of abstraction accorded to these objectives must, at least as a starting-point, be reasonably related to those rules and measures that the Contracting States decided to include in the document itself. One can therefore plausibly argue that the Court s approach possibly transgressed the inherent limits of purposive interpretation as it did not adequately focus on what the text of the Agreement actually states explicitly and, more importantly, omits, at least implicitly. Remember, the rule on State-Liability in EC Law had been established under the famous Francovich-ruling of the European Court of Justice in It is important to bear in mind that interpreting the objectives and purposes of the EEA-Agreement at their highest levels of abstraction can almost always provide the opportunity for the EFTA Court to include within that document s substantive scope all those rules and measures that the Court s members, at any given time, feel will make the EEA-Agreement what it should be. This is especially so if the clear textual omission by the Contracting States of a defined rule or measure has limited or no weight in the formulation of these purposes and their subsequent normative use in the Court s legal reasoning. The opinion in Erla María must therefore I think be justified by reference to other theories of interpretation or theories relating to the institutional structure or competences of the Court itself. I express no opinion on whether such an argument can be sustained. I stress furthermore that my comments on this issue are a consequence of my exposition above of a theory of purposive interpretation with certain characteristics and inherent limits. However, if a scholar of statutory interpretation, or an interpreter of legal texts, believes he can plausibly

17 17 argue for a different approach, his or her approach could possibly provide a more solid foundation for the EFTA-Court s opinion in Erla María as well as the other examples from the case-law that I have discussed An attempt has been made to provide such an argument in the EC context, see Trevor C. Hartley: Constitutional Problems of the Éuropean Union. Hart Publishing, Oxford, (1999),