On legal reasoning

Language, discourse and reasoning are important dimensions often overlooked by these debates, with exceptions such as the abovementioned Hartley or Rasmussen. But it was in the mid-1990s that the charge of activism was linked to the court’s defective reasoning. The criticisms coming from political science, from international law, and even from constitutional law were expanding into theories of legal reasoning, a process that Conway very aptly describes in his section on ‘Debating the Court’.Footnote ⁸

Presented this way, there is a risk of distorting the legal-reasoning dimension of the Court’s jurisprudence: how does the Court actually justify its decisions? How does it operate and deliberate? What type of law does it deal with and what are the main features of the legal field in which it operates and which it shapes? Does it all boil down to a question of being conservative or liberal or should we also have a view as what makes a judgment and its reasoning acceptable? Perhaps legal argumentation is not just a function of assuming the institutional role assigned to one institution, but also of engaging in practical reason.

Practical reasoning involves the practice of justifying one’s normative stances and critiques by giving reasons and engaging in rational discourse with interlocutors or audiences. When the normative domain is related to the law, this practice can become institutional and institutionalised, and different actors with different functions engage in legal reasoning: legislators, policy makers, administrators, lawyers, legal scholars, amongst others. In the context of the EU, the duty to give reasons concerns all institutions issuing legal acts and member state authorities when acting within the scope of EU law. Article 296 TFEU provides: ‘Legal acts shall state the reasons on which they are based and shall refer to any proposals, initiatives, recommendations, requests or opinions required by the Treaties’. Article 41 of the Charter of Fundamental Rights of the Union relates the duty of the administration to give reasons for its decisions to the right to good administration. The good-reasons approach is the antidote to free-rein discretion. But what are good reasons?

When judges engage in this practice in courts, deciding on disputes strictly speaking we obtain judicial reasoning. Taking the part for the whole, this judicial reasoning is most commonly referred to as simply legal reasoning, or sometimes legal argumentation. For our purposes when the Court of Justice engages in the practice of justifying – giving reasons or arguments for – its decisions, we talk about the legal reasoning of the Court. This applies to the Court including its judges and Advocates General. To what extent and under what conditions the reasons given and the ways these reasons are linked are considered valid or legitimate or rational or acceptable forms of argumentation is the subject of the different theories of legal reasoning. By and large, these theories can be descriptive and/or prescriptive, analytical and/or critical.

The four recent books reviewed in this contribution, Gunnar Beck’s and Gerard Conway’s both from 2012, Elina Paunio’s and Suvi Sankari’s from 2013, focus on the legal reasoning dimension and engage in these debates from the perspective of jurisprudence and EU law. The four books are well worth studying closely and are based on thorough research and knowledge of the case law and of EU law, and also of the relevant literature. Three books are based on the authors’ doctoral theses.Footnote ⁹ The four books reviewed engage with the most relevant literature on legal reasoning displaying a very good grasp of the main issues involved in the theoretical discussions. They also show very thorough knowledge of the most relevant cases. They are a great contribution to the study of the ECJ and of EU law generally.

Perhaps excepting Conway, the authors do not so much pass a final verdict on the legal reasoning of the ECJ falling into one or another textualist or teleological model as they enhance a realistic understanding of the Court, its reasoning, its multilingual context (Paunio) and context of discovery or steadying factors (Beck), its heuristics, its limits and discretion, the methods available to it, the arguments to which it resorts, the outcomes of its interpretation (Sankari). However, at some points in the work, Beck also engages in normative assessment of the Court’s reasoning and Conway even suggest alternative approaches that seek a greater respect of the separation of powers by the Court.

Some basic features of the Court of Justice

This review article will provide a summary of each of the books with a critical appraisal of each,Footnote ¹⁰ and will conclude by rounding up the discussion of the common themes and the major issues concerning the legal reasoning of the ECJ. But before exploring these works, it is worthwhile remembering some basic features from the institutional context of explanation of the ECJ as special type of judicial decision-making institution.

The Court of Justice consists now of 28 judges, one from each member state, and nine Advocates General who assists the Court (from October 2015 there will be 11 AGs) giving individual reasoned opinions marking the end of the submissions phase of the proceedings and the opening of the deliberation phase.

The Court acts always as a college sitting either in plenary session consisting of all the judges for which the quorum is 17, in the Grand Chamber of 15 judges with a quorum of 11 or in chambers of five or three judges. No dissenting opinions are expressed or published. Dissent is absorbed in the judgement, and the resulting style is impersonal and highly structured,Footnote ¹¹ notwithstanding the important role of the judge-rapporteur. All judgments must be reasoned and the reasoning is best perceived when both the opinion and the judgment are contrasted, which is when one can detect the relevant ‘silences’ and divergences, something that Sankari’s book has sought to analyse in relation to the citizenship cases.

The ECJ is a multilingual court that deliberates and drafts its judgments in French with the help of different sorts of multilinguistically minded jurists. The ECJ is multilingual as regards input and output (21 languages), but francophone as regards throughput. Yet English is by far the largest semiotic context for the academic discussion of the Court’s jurisprudence throughout Europe and the whole world. Of the four authors reviewed only Paunio is systematically aware of multilingualism in the Court. EU law is multilingual law and no linguistic version overrides the rest. It is not possible to know which version conveys the ‘intended’ meaning, if such a concept makes any sense at all (pace Conway).

Dialogue is central to decision-making of the ECJ, especially in references for preliminary rulings, where the ECJ does not decide the case but only interprets EU law. The Court does not pick its own cases, but is seized by national courts or by direct parties, and reacts to their claims or queries. It does obviously have some leeway in the way it frames the preliminary questions put to it but even if it were to be considered activist, it would still be reactive, always reacting to the cases brought before it. But in some of its judgments, it has also been proactive, in that its interpretations on the effectiveness and autonomous meaning of EU concepts has generated further references as they are applied to new situations by ‘creative’ jurists. Whether this automatically turns the Court into a political actor is another issue,Footnote ¹² which largely depends on one’s concept of the political. Beyond this dialogue between courts and legal professions, the ECJ also establishes a certain dialogue with itself over time in different cases, or with the General Court on appeal and the AG in the same case, and with other institutions, the parties to the cases and other repeat players.Footnote ¹³ Precedents are particularly important, because they can lead to a particular type of cases or disputes being considered as ‘clear’.

The distinction between clear cases and hard cases is therefore related to the existence of precedents, but there are also inbuilt procedural norms that point to the importance of the distinction before the Court. To begin with, from the perspective of the referring courts, all references for preliminary rulings on interpretation originate, by definition, in hard cases. This does not mean they will also be hard for the Court. Thus it is not necessary for the AG to give an opinion if the case raises no new point of law, and in fact about one third of the judgments are delivered without one. Also the fact that a Chamber of the General Court can delegate the decision to a single judge where the questions of law or fact raised present little difficulty points to the a priori importance of the clear/hard case distinction.

Finally, the Court ensures the observance and the ‘effectiveness’ of the ‘autonomous’ legal ‘system’ of the EU and this also means that it is not expected to nor supposed to approach EU law from any particular national legal mindset, not even a combination of national outlooks, but rather with an ethos that could be defined as communautaire. Ignoring or overlooking these and other relevant features might lead to subjective approaches to the Court’s reasoning or critiques of its ‘style’ modelled upon national judicial cultures.

Beck: uncertainty, steadying factors, and a cumulative approach

In this impressive book of 451 pages on the Legal Reasoning of the ECJ, Gunnar Beck distinguishes two levels of uncertainty in the law: primary uncertainty exists at the level of legal rules that require interpretation, but the rules of interpretation are themselves uncertain and this gives rise to the secondary level of uncertainty. He considers the problem of legal uncertainty as ultimately incapable of judicial or even doctrinal resolution. This is explained by several reasons. First, because of linguistic vagueness, itself a result of the open-ended nature of many, essentially contested concepts; ambiguity of meaning and context-dependence; and also imprecision. As Chapter 2 explains, rational disagreement can be genuine concerning such concepts. Next, because of value pluralism and norm conflict, there being no hierarchy of norms governing the relation amongst the multiple treaty or judge-made objectives, values and principles, as in the conflict between progressive economic and political integration versus subsidiarity. Standards like proportionality are not decisive when deciding either way, because the values are incommensurable, often incompatible, uncertain, and they are not ranked hierarchically, as Chapter 3 shows. Also, because of gaps in the law, where crucial choices regarding the future course and depth of European integration were left to the judicature, and the instability of precedent, which are pervasive features of EU law. Finally because there is no methodological certainty or overarching formula governing the application of the linguistic, systemic and purposive interpretative topoi.

These topoi are themselves applied consequently but cumulatively in the three-stage model rather than according to any clear order of preference or by default. Hence the cumulative approach based on the three-stage model of MacCormick and Summers,Footnote ¹⁵ which is the subject of Chapter 9. Courts commonly consider linguistic arguments first; if the doubt persists these arguments give way to or are complemented by systemic arguments, and these again by purposive arguments.Footnote ¹⁶ The Court employs broadly the same accepted canon of interpretative topoi to resolve primary legal uncertainty as appellate and constitutional courts in national legal systems.Footnote ¹⁷ The specificity of the ECJ is that the linguistic, systemic and purposive arguments are normally combined rather than ranked in order of preference and applied only if the doubt persists.Footnote ¹⁸ These topoi are thoroughly examined in part Two of the book and in Chapter 7, which tends to confirm previous works on the methods of interpretation of the ECJ, including my own.Footnote ¹⁹ The sections on the absence of a formal doctrine of ratio decidendi in the Court of JusticeFootnote ²⁰ are especially interesting.

This does not however mean that judges openly decide in accordance with political or personal preference or full discretion for they must justify their decisions according to the cumulative weight of those topoi and available precedents. The ethos guiding the Court is a restrained integrationist agenda, largely following from the telos of the Treaties, and their meta-telos of the ever closer union. Rather than following a scientific method of legal reasoning which would deliver predictability of judicial decisions, the guiding method is heuristic where the Court follows patterns that underlie the official justificatory discourse and takes into account steadying factors of judicial decision-making in order to persuade the audiences about the correctness of its constructions: the period style, communautaire interpretation, judicial deference factors, motivational factors. Judicial behaviour is thus a subcategory of social behaviour with its own professional, doctrinal or institutional constraints and traditions.Footnote ²¹ The resulting reasonable reckonability of decision-making would lie somewhere between full discretion and predictability.

Beck accepts the distinction between clear cases and hard cases and argues that syllogistic reasoning is sufficient for clear cases, but in practice most cases are harder and this requires a secondary legal justification based on the above topoi of interpretation. The relevance of what Karl Llewellyn called steadying factors allows Beck to systematise the factors that contribute to limiting and reining in the pervasive uncertainty of the law, and thus to check the Court’s discretion. ‘All the various steadying factors – whether “legal” or “extra-legal” – are shorthand for the underlying determinants of the Court’s reasoning and decision-making.’Footnote ²² These include personal motivational factors, legal factors like the mentioned legal topoi or accepted terms of legal argument, the principle of non-contradiction, the importance of precedent following, the ordinary use of language creating a basic interpretative plausibility, but also professional and institutional constraints, or public confidence in the substance of judicial decisions, political constraints imposing judicial self-restraint, the key issue of deference to member states or other actors, ideas of political correctness and dominant values and institutional ethos. These steadying factors are the subject of Chapters 1 and 11. Systematising these factors is no easy task, but this attempt is perhaps Beck’s major contribution to the study of the Court’s legal reasoning.

Given all the factors enhancing judicial discretion like the prevailing value pluralism, vagueness and indeterminacy and precedent uncertainty where judicial decisions designed to elucidate general principles often merely replace one set of uncertainties with another,Footnote ²³ and given Beck’s carefully chosen heuristic approach to method ‘which does not hold out the promissory notes of legal certainty’,Footnote ²⁴ clearly discarding the normative level that prescribes how judges ought to decide cases,Footnote ²⁵ it is quite surprising that the epilogue should take such a strong attack on the ECJ in its Pringle judgment on the ESM Treaty.Footnote ²⁶ The ECJ is criticised for adopting a straightforward contra legem interpretationFootnote ²⁷ ignoring the ‘natural meaning’ of Article 125 TFEU, against a wording of the law that is ‘express’ and disregarding objectives such as price stability and the principles ‘no bond buys’ and ‘no bail out’ that are specific.Footnote ²⁸

As mentioned above, disagreements about legal interpretation can easily be projected on the forum of politics. Instead of simply considering that its interpretation has favoured systemic and dynamic considerations over the textual ones, which are deemed clear, the Court is now accused of overstepping its role, even of suspending the rule of law.Footnote ²⁹ For Beck, in situations of emergency like the financial crisis, ‘the Rechtsstaat is effectively suspended, as the normative steadying factors of legal reasoning cease to function as effective constraints’ for ‘it is difficult to see how the Court’s finding can be squared with the ordinary meaning, or indeed any plausible meaning, of the words “assumption of liability for any commitment” and “financial assistance”’Footnote ³⁰ as contained in Article 125.Footnote ³¹ One is left wondering about the normative punch of the steadying factors which now are expected to constrain considerably more than in the main text of the book. One also wonders what is left of the ‘vagueness and value pluralism which in part embody political compromises between the member states, and a pervasive “teleological” theme written into the texture of the Treaties and legislation themselves’.Footnote ³²

A more coherent take on Pringle could be attempted on the lines suggested by Kaarlo Tuori, who considers that the Court is relying on a double telos, one being the purpose of Article 125 to circumscribe fiscal liability to a particular member state and prevent bail outs of that state, and the other, the more general objective of the Economic and Monetary Union seeking stability of the eurozone.Footnote ³³ After all, Beck himself admits that the use of purposive arguments would not only be justified as having a literal basis in the Treaties and even in EU legislation, ‘but also on the grounds that they form part of the general scheme of EU law [...] the EU Treaties are value-based and purpose-oriented.’Footnote ³⁴

A final idea worth noting from this book, from the point of view of the Court’s style of reasoning, is to be found in the section on multilingualism, where Beck has more to say about style than about multilingual interpretation: ‘[t]he reliance on whole paragraphs and recurring turns of phrases tends to give the Court’s judgement a certain formulaic and even mechanistic appearance: building blocks from previous judgments including familiar statements and principles are used to emphasise and suggest continuity but are in fact applied to different factual scenarios and thus in effect extend, limit or even modify the rule they contain’. This reliance on prefabricated materials from previous judgments ‘gives rise to rule instability which exacerbates precedent uncertainty.’Footnote ³⁵ It would have been really interesting to instantiate such sweeping statements with some of the examples the author mentions later:Footnote ³⁶Grzelczyk, von Colson and Kalman, Marleasing and Pupino.

Conway: the constitutional case for limits or interpretative restraint

Conway’s Limits of Legal Reasoning keeps predictability as a desirable feature of judicial decision-making and displays a rare faith in legal certainty, ordinary meaning and original intention. By contrast to Beck, Conway believes predictability could be achieved if the Court followed the law and stuck to ‘relatively clear, shared criteria of interpretation’.Footnote ³⁷ Consistently with his critical focus, Conway advocates a normative theory that ‘contrasts interpretation with law-making and related legitimate interpretation to a close textual analysis of the most relevant specific laws (in priority over more general provisions) and an effort to retrieve original intention. This understanding is linked to the idea of separation of powers’Footnote ³⁸ further developed in Chapter 4: ‘It may be argued that a modern approach to the separation of powers envisages a dialogical relationship between the branches, thus rendering judicial interpretive licence less problematic. However, the context of the EU of the “unusually permissive environment” in which the ECJ operates (in the sense that it requires coordination from all of the Member States to reverse ECJ interpretation of the Treaties) renders problematic the notion of inter-institutional dialogue.’Footnote ³⁹

Conway makes a selection of cases where ‘doctrines and new rules have emerged of general application that far transcend their immediate facts and so that can make these cases analogous to legal provisions of a legislative character, in violation of an understanding of the relative distinctness of the legislative and judicial functions.’Footnote ⁴⁰ The cases he selects for study all tend to confirm ‘the prospective, developmental character of teleological interpretation employed by the Court, in contrast to a retrospective, originalist analysis’Footnote ⁴¹ and ‘this choice of cases shows that judicial interpretation can have far-reaching and innovative effects [...] where the ECJ has established quite new rules as opposed to giving a broad interpretation of existing rules.’Footnote ⁴² He does not seem to mean this as a compliment. Conway is very critical towards the Court. We shall only mention some of his chosen lines of cases.

To begin with, the direct-effect cases constitutionalising the Community. Van Gend was not obviously supported by the wording of the Treaty, which appeared only to attribute such effect to Regulations but ‘the Court used the term “direct effect” in contrast with “direct applicability” as used by the Treaty in relation to Regulations.’Footnote ⁴³ In Lütticke ‘it was enough that the text did not explicitly rule out direct effect for the ECJ to extend the direct effect doctrine.’Footnote ⁴⁴Walrave and Koch further extended the direct effect doctrine, horizontally rendering it applicable to cases between private individuals and in Grad the ECJ extended direct effect to a decision ‘where only Regulations appeared to be attributed with such effect under the Treaty.’Footnote ⁴⁵ But ‘what really gave the direct effect doctrine its significance was its combination with a new principle of the supremacy of Community law over the law of the Member States, first set out in Costa v. ENEL. This decision […] resulted in a doctrine of fundamental significance, without explicit Treaty support.’Footnote ⁴⁶

Conway argues that a departure as significant as the supremacy doctrine compared to general international law was not intended by the member states, given that there was no express reference to a supremacy principle and ‘the Treaty of Rome had not explicitly addressed the issue of the hierarchical relationship between Community law and that of the Member States.’Footnote ⁴⁷Internationale Handelsgessellschaft confirmed that Community law takes precedence over all national law, including constitutional law, relying on effectiveness and uniformity, also the justification for Foto Frost, where ECJ jurisdiction to give preliminary rulings on validity was only expressly exclusive under the ECSC Treaty.Footnote ⁴⁸

Conway adds to supremacy, a competence creep, ‘[c]onsistently with a tendency to adopt a pro-integration interpretation in questions of competence, the ECJ has made limited use of the principle of subsidiarity, which was intended, it seems, to counter an assumption that integration of competences was necessarily desirable as an end in itself’Footnote ⁴⁹ at the expense of other important values like member state regulatory autonomy, the principle of conferral and democratic consent.

On the free movement cases, Conway reminds us that Dassonville and Cassis de Dijon were framed in consequential terms; ‘the ECJ did not seek to recover the original intention of the Member States, for example, as to the meaning of what was an ambiguous provision’ i.e. ‘measures having equivalent effect’.Footnote ⁵⁰ Also Keck was extra-textualist and explicitly consequentialist in noting the effects of the wide interpretation in Dassonville.Footnote ⁵¹ As regards Bachmann, or Kraus, or Gebhard, ‘by establishing a category of justified restrictions outside the Treaty-based categories, the ECJ was establishing an alternative set of criteria to that established in secondary legislation for the application of Treaty-based exceptions. The Gebhard-Kraus-Thieffry line of case law clearly illustrates how, through smaller incremental steps in a series of cases, an overall doctrine of great significance can emerge and how this process of development is rendered less than transparent through a declaratory, superficial style of reasoning.’Footnote ⁵²

Conway’s argument ‘attempts to elaborate on the idea that the interpretative perspective of the judiciary should in general be aligned with that of the law-maker and that of other reasonable participants, i.e. ordinary citizens, in the legal system, and that this can be achieved in the elaboration of a hierarchical scheme of interpretation that reflects “ordinary constraints on interpretation”.’Footnote ⁵³

Conway is interested in the normative level of justification. He has a clear view as to when judicial interpretations are justified and what the limits or constraints are. He is well aware that the Court has opted for a different type of reasoning, indeed has institutionalised an alternative purposive interpretation but ‘the institutionalisation of a certain method of reasoning by the ECJ does not dispose of the normative question of how interpretation and legal reasoning should be conducted.’Footnote ⁵⁴

In this sense Conway is an outlier amongst the four books under review: he is not interested in descriptive accounts of the Court’s legal reasoning. He is well aware of these approaches, for in Chapter 2 he reviews the main literature on the Court’s reasoning, and is not happy overall: ‘the comment that the language of love could be used to describe how EU lawyers look upon the ECJ remains largely true.’Footnote ⁵⁵ True, laudatory excesses are never helpful, but flattery and love are rather different things, I would say. Conway is more interested in ‘the relatively few attempts at directly engaging with the issue of a normative theory of interpretation for the Court of Justice in all of the extensive literature on its role.’Footnote ⁵⁶ Perhaps he does not agree with the method of rational reconstruction, which allows for internal critical appraisal of the Court’s practice on its own standards and from an internal point of view. Conway seems to think rational reconstruction, and ‘this is most obvious in the work of Bengoetxea’, amounts to simply accepting ‘the fundamental structure of the Court's reasoning’ as it is.Footnote ⁵⁷

Conway advocates a conserving interpretation based on input legitimacy from the lawmaker whose intention is grasped by direct subjective originalist interpretation. To be legitimate and consonant with the rule of law, interpretation must be faithful to and thus constrained by a pre-given core of meaning contained in the law.Footnote ⁵⁸ Conway reminds us that, in the context of the doctrine of interpretation of domestic law in conformity with directives developed in the cases of von Colson and Kalman and Marleasing, ‘the ECJ itself has occasionally stated interpretation cannot be contra legem.’Footnote ⁵⁹ But this is not enough, for this would not rule out evolutive interpretation. Conway discards evolutive interpretation for political reasons related to ‘predictability as a key feature of formal legality’ and to the ‘democratic authority of the law-maker and the comparative lack of democratic or representative legitimacy of the judiciary’. Instead, Conway advocates conserving interpretation,Footnote ⁶⁰ which he does not see as political: ‘originalist interpretation evidences no particular political preference; it simply respects the authority of the law-maker in any legal system.’Footnote ⁶¹

Conway is aware that some concepts are essentially contestedFootnote ⁶² but argues that it is a mistake to believe that democracy or the rule of law are ‘so contested that they have no fixed content.’Footnote ⁶³ This leads to downplaying the distinction between hard and clear cases: ‘In the present work, the distinction between hard and easy cases becomes less significant because all cases can be subjected to the interpretative framework proposed, and difficult or hard cases to which a solution is not found when this model of reasoning is applied can be the subject of a finding of non liquet, save where innovation may be necessary to avoid extreme injustice.’Footnote ⁶⁴ Well, it seems, after all, that there are some limits to limits, and innovation is not ruled out. But what Court would pronounce non liquet and how would it proceed in such extreme cases? This is unsettling, for outside the limits there are no limits, full discretion à la Hart would seem to creep back in, and democracy and the rule of law might be in trouble again before the exception.

But Conway does have a normative model of legal reasoning for the ECJ, which he calls the rule-bound theory of interpretation, ‘emphasizing five main criteria of interpretation: (i) the centrality and authority of the constitutional text and the normative priority of its ordinary meaning; (ii) the application of the lex specialis principle for structuring systemic or integrated interpretation; (iii) the resolution of indeterminacy resulting from abstraction through originalist interpretation, primarily through reliance on Member State legal traditions or relevant preparatory materials indicating the intentions of the legal authors or ratifiers; (iv) a preference for dialectical reasoning and the explication of interpretative assumptions; and (v) the relevance of the argument from injustice only in exceptional cases. This can be understood as a principled (not principles-driven) approach to legal reasoning, in emphasising a method and process as having priority over a more pragmatic orientation to intuition and consequences.’Footnote ⁶⁵

Conway does not deal with multilingualism. The closest he comes is to say that ‘problems of translation in the EU’ can complicate ordinary meaning, but are ‘to be resolved on the basis of evidence, not as conceptually recasting the enterprise of legal reasoning as a pretext for judicial law-making.’Footnote ⁶⁶ One wonders how evidence can solve problems of translation, but presumably there will be one right answer, the true meaning, the evidence of which would need to be properly reflected in all the languages.

Nor does Conway distinguish different levels of interpretation; there is little nuance: ‘what is necessary is to avoid a mechanical literalism, but at the same time to respect the priority of ordinary meaning.’Footnote ⁶⁷ One wonders where the line may lie between literal and ordinary meaning. As a practical guide to finding the ordinary meaning this is what Conway has to commend: ‘applying, first, ordinary linguistic interpretation of the most specific legal provisions. Where that process does not remove ambiguity, originalist interpretation can be resorted to: ambiguity is resolved by relying on relevant, public legal tradition as to the meaning of legal concepts or on evidence of the will and understanding of the law-maker. This process should be marked by dialectical reasoning to the extent that it still entails choices, though not in a way that reopens very general, systemic considerations about the legal system, except where fundamental rights or extreme injustice are at issue.’Footnote ⁶⁸ How concepts brought from ‘legal traditions’ or notions of ‘fundamental rights’ and avoiding ‘extreme injustice’ will bring more certainty to decision-making is an intriguing query. This normative guide is then tested against a number of cases in Chapter 5. In the case of citizenship cases the ECJ is said to rely on ‘meta-teleological interpretation to enhance integration,’Footnote ⁶⁹ something Sankari’s study will partly deny. As we can see below, both Sankari and especially Paunio are interesting antitheses to Conway.

Sankari: Context of reasoning and systemic criteria of interpretation

Beck’s ‘steadying factors’ and Conway’s ‘limits’ become Sankari‘s ‘constraints’.

Legal reasoning of the Court of Justice becomes too general a category when analysed as a whole. The importance of focusing on a line of cases like citizenship of the EU might illustrate the general theory of the Court’s general interpretative strategies but it can show how the type of legal reasoning and the interpretative methods might change. Suvi Sankari shows this by exploring a particular line of cases in a concrete subject area like citizenship of the Union and in a type of action: preliminary rulings on interpretation as opposed to direct actions. The type of action and the type of case has an impact on the type of reasoning and the interpretative arguments. The finding is the prominence of systematic arguments in order to ensure coherence of an area of law where citizenship is framed in Treaty articles, in relevant directives and where free movement also sets the benchmarks.

Suvi Sankari’s Legal Reasoning in Context opts for a legal-reasoning rational reconstruction approach to the analysis of judicial decision-making of the Court, based on its own standards or self-understanding,Footnote ⁷¹ precisely the method Conway seems to abhor. This is a proper and necessary approach to understand and critically appraise the Court’s decision-making in individual cases, but also for the analysis of judicial decision-making strategy in a longer run, diachronically, an area where alternative approaches lean to more political-science or constitutional analyses of judicial activism, all too often neglecting the idiosyncrasy of judicial justification and institutional and legal cultural factors. Sankari’s approach is neither normative nor predictive.Footnote ⁷² The main tools for analysis are Bengoetxea’s first and second order criteria,Footnote ⁷³ while ‘also paying attention to silence and distinguishing between individual cases and lines of cases.’Footnote ⁷⁴

The book starts from the recognition of the importance of justification of judicial decisions in the European legal and constitutional culture. Sticking to the facts and circumstances of the cases and the preliminary references, as the courts are institutionally expected to do, implies that the range of options open to the courts are probably narrower than assumed by alternative analyses based on the activism/restraint dichotomy; which also suffer from a ‘pre-determinacy’ problem, a tendency to consider that the proper role for a court is to stick to a pre-determined view or conception of European integration, either favouring the member state interests or those of integration and market access. This pre-determinacy is sometimes projected on theories of meaning, as though the literal meaning of provisions in EU law would be fixed and objective if the ‘proper’ meaning were to be given to the terms and concepts used by such law, be they autonomous concepts or not.Footnote ⁷⁵ The ‘activism’ charge would then consist in departing from the obvious, clear, pre-given meaning (Rasmussen). These simplistic approaches blur the actual analysis of justification of the particular case. The ex-post approach on activist decision-making based on measuring the types of reactions to court decisions after they are rendered also seems simplistic and random and does not really clarify the reasoning alternatives available to the decision-maker.

Suvi Sankari engages with the relevant legal literature suggesting different approaches (e.g. when discussing Somek, Hailbronner, Hartley or Tridimas). The activist critique of the Court, when translated into a legal reasoning approach to the Court’s interpretation of EU law, results in a tendency to emphasise or overemphasise the dynamic categories of interpretative reasoning, more particularly the teleological method. If one paid heed to the activist charges made by major critics as regards the citizenship cases (Articles 17 and 18 EC or 20 and 21 TFEU), one would expect to find, when analysing the Court’s interpretative method and practice, extraordinary recourse to the dynamic methods but if one were to find that the Court sticks to its usual, standard methods of interpretation, the charges of activism would be deprived of their punch and reconstructed into politically informed criticism of the Court’s orientation. The Court would, in other words, be carrying out its usual task, its ordinary business by taking the Treaty provisions seriously and considering the whole of the legal order governing citizenship according to systemic criteria where the Treaty provisions are lex superior, lex generalis and lex posterior as regards secondary law contained in the 1990s directives or in the older free movement regulations.

As the citizen’s right to free movement and residence is the general rule, the provisions of secondary law that regulate the exercise of this right are to be interpreted strictly to the extent that they constitute exceptions to the general rule, and checked against the necessary principle-based justification and against the proportionality requirement. Indeed, the initial directives were facilitating the free movement of persons and the institution of citizenship of the Union was later superimposed on the pre-existing law and case law in order to complete, not to compete with, the free movement of persons. Although the Court may follow a functional approach wherever it needs to secure the effectiveness of the right to free movement in each particular situation, it cannot be said from the outset exactly how it will strike the balance between the right to free movement and non-discrimination on the basis of nationality and the legitimate member state interests in a concrete situation. This enterprise of interpretation is applied not only to the Treaty (and the relevant 2004 directive) but also to the previous case law of the Court. The juristic community is thus engaged in ideal discourse situations and communicative action, and interacts in an on-going dialogue, where the Court has to give persuasive, convincing, acceptable reasons as the cases evolve; and this necessitates transparency, sincerity, and dialogue with the relevant audience (rational acceptability).

The contribution here lies in highlighting the systemic approach to interpretation that features prominently in the types of cases analysed, citizenship, over the more dynamic purposive, teleological interpretation as the standard doctrine of interpretation and reasoning. The application of this major finding to the standard formulation of the Court in the Grzelczyk preliminary ruling – ‘Union citizenship is’ ‘destined’ or ‘intended’ ‘to be the fundamental status of national of the Member States, enabling those who find themselves in the same situation to enjoy the same treatment in law irrespective of their nationality, subject to such exceptions as are expressly provided for’ – gives the axis of the book (section 3.1.1.). The first part of the standard is clearly dynamic not only because of its use of the term ‘destined’ but because it recognises a general right to equal treatment as well as a fundamental status of citizenship, whereas the whole of the formula invites systemic examination of the exceptions contained in the secondary legislation.

Another interesting contribution of the book is suggested around the idea of the elements that go unsaid in a judgment, its silences. This idea goes further than the mere elliptical reasoning where parts of the premises are not stated in the reasoning but can be derived or induced from the general scheme of the reasoning. In this case it relates rather to two possible strategies of focusing the decision: (a) the level of generality when determining the ratio decidendi, wherein the narrower the ratio, the fewer cases it will cover; (b) the arguments and contentions that might have been made in the written and oral debate over the case but which are not addressed by the Advocate General or addressed by the Advocate General but not by the Court, not even obiter, presumably because they are unnecessary to decide the case. A third possibility of silence would be (c) an argument that has featured only in the deliberations of the Court but which does not reflect on the written justification, the Court having preferred to skip it or avoid it for different strategic reasons, probably because agreement concerning those issues was not reached in deliberation or was only reached insufficiently (pace SunsteinFootnote ⁷⁶). These silences can be very telling indeed. The problem is methodological: how to detect them and elaborate on them. Always aware of the differences between the context of discovery and the context of justification, Sankari goes some way at least to showing instances where these silences might have been in operation.

Connected to this notion of silences, Sankari contributes to the mainstream theory of legal reasoning with this idea of streams of interpretative reasoning in lines of cases. The types of reasons used in the series of cases can be illustrative of the silences. A good example could be the cases of Ruiz Zambrano and McCarthy, both decided in 2011. The very scant reasoning of Ruiz Zambrano becomes slightly clearer through McCarthy. As the thesis suggests, it might be that the Court is aware that it is going to decide McCarthy and that it might be easier to reach consensus and a more complete reasoning in this case as regards the purely internal situation than in the much more thorny Ruiz Zambrano.

Together with silences and streams of cases and their reasoning, the book’s other major contribution to the theories on legal reasoning and even of the Court’s legal reasoning, is the analysis of substantive arguments of interpretation, as it is sometimes predicated of the standard formulation of Grzelczyk (see below). Substantive reasoning often related to reasonableness or equivalent topoi, is an important criterion replacing or counterbalancing the relevance of dynamic criteria. Thus, where silences, streams of cases and substantive reasoning are combined we get new avenues for developing an adapted theory of legal reasoning. This is a very interesting theoretical finding, explored passim in section 4.3.

As regards citizenship of the Union, Sankari also makes interesting contributions. The Treaty additions on citizenship of the Union would complete the picture of free movement as regards those persons who did/could not benefit from any of the specific categories of persons – workers, professionals, services recipients and providers, students, pensioners – classically affected by the free movement provisions. Citizenship would then be conceived of primarily attending to the freedom to move and reside within the Union. However, another feature of citizenship, as developed by the interpretation of the Court is to recognise a general right not to be discriminated against on the basis of nationality, and generally regulated in the Treaty. The provisions on citizenship do not make any direct connection with the prohibition of (national) discrimination, but provide a general blanket ban, which is then instantiated in each of the categories of freedom of movement. However the Court made such a connection in Martínez Sala first (obiter) in 1998 and more decisively in Grzelczyk (ratio) later in 2001: what citizenship of the Union allowed was to invoke the general prohibition of discrimination. In doing so, some commentators have insisted that citizenship of the Union creates a different sort of category on top of the classic four freedoms (the issue is discussed passim but more specifically in section 3.2.2).

Citizenship would therefore operate as a ‘metaphor’ for something else, perhaps for a preference in favour of market access as opposed to fundamental freedoms. Suvi Sankari has very convincing arguments in favour of the thesis of the completion of the free movement of persons. On the other hand, in section 3.1.2, Sankari delves on the distinction between having a right and enjoying or exercising a right, which does point in the direction that one can simply ‘be’ a citizen of the Union in a purely internal situation, and potentially be a holder of the citizenship rights without necessarily exercising them, and this would call for the separate category, or even citizenship as a fundamental institution; a ‘constitutional’ move for the Union that would permeate the ‘purely internal situation’ conundrum.

Starting with Grzelczyk, again, but continuing with an important line of cases, it becomes clear that the recognition of free movement and residence to citizens of the Union is not to imply the imposition of any financial burden on the host member states, a caveat already foreseen in the secondary law before even the entry into force of the Treaty provisions on citizenship. The differences between ‘burden’ (in the directive provision) and ‘unreasonable burden’ (in the recitals of the Directive’s preamble) is analysed in section 4.1, with interesting and far-reaching consequences. Burden is an all or nothing condition whereas unreasonable burden requires a dimension of weighing and balancing, an application of the method of proportionality on which the book also has an important contribution to make by suggesting the application of the three-pronged proportionality analysis of necessity, suitability-reasonableness and degree. The implication for the Court of Justice is that the institution of citizenship requires the acceptance of a certain degree of financial solidarity among the citizens of the host member state as regards citizens of the Union who might happen to undergo temporary difficulties.

As regards activism, Sankari adopts a critical and neutral approach toward the Court, whenever shortcomings can be detected in the judgment’s reasoning and style. She addresses the activist critique and its challenge in the form of a provocative question: what could (should) the Court have done differently?Footnote ⁷⁷ The idea behind the activism critique is that the Court should have given preference to the secondary law restrictions over the general provisions of the Treaty, understating the fact that the Treaty articles were lex posterior and more to the point lex superior stating constitutional rights to be interpreted broadly whereas the secondary law provisions were exceptions to be narrowly interpreted. Other critics might say the Court has been too restrictive in not abolishing the exceptions altogether, so that its failure to fulfil its role to ensure that in the interpretation and application of the Treaty the law is observed, would have consisted in excessive restraint. Sankari correctly reformulates the question and insists on the legal reasoning approach: sometimes the Court has gone very far in addressing issues it was not really asked to respond rather than answering the questions. On other occasions it has not addressed relevant arguments of the parties.

What the Court could have done differently is to have better reasoning, more complete and coherent reasoning, which implies a notion of internal reasoning coherence: not only coherence in the interpretation of the law but also coherence in the reasoning of the judgment itself and coherence within the stream of cases dealing with similar issues. The book is so preoccupied with activism charges, that it often places systemic interpretation above dynamic interpretation. In contrast with Beck, the systemic and the dynamic methods are well differentiated in Sankari (section 3.2). More stress on dynamic reasoning would have probably led to a different approach on citizenship of the Union, as a separate, fifth category capable of combating national discrimination and all related prejudices, on the importance of financial solidarity and effet utile. But in favour of Sankari, it must be said that it is easier to find counterexamples and refutations of the dynamic reasoning than it is of the systemic interpretation.

Paunio: Enhanced uncertainty in multilingual law

There is a contrast or tension between multilingualism of EU law, which is an undeniable reality, and legal certainty, which is a fundamental component of the rule of law, obviously also of EU law. Certainty is ‘a fundamental principle according to which the addressees of laws must know the law in order to be able to plan their actions in accordance with it.’Footnote ⁷⁸ The solution of this tension requires a serious rethink of legal certainty from an initial, almost pre-interpretative understanding of legal certainty – understood as objectivity and the natural, pre-given, obvious ordinary meaning of legal norms – to a hermeneutic, interpretative conception of legal certainty based on argumentation and rational acceptability, enhancing aspects like the relevant juristic audience and the conditions for rational communication.

This new conception of legal certainty is further reinforced not only by a philosophy of language that questions any such thing as a natural or objective meaning of words, also applicable to legal language, but especially by the multilingual character of EU legal language, where not only is there no objective meaning of the legal provisions; even more to the point, there is no uncontested language reference amongst the equally valid official language versions in which the law is expressed.

Paunio’s Legal Certainty in Multilingual EU Law is a major contribution to the field of European legal reasoning and the understanding of decision-making, interpretation and justification at the European Court of Justice and to multilingual law and legal thinking. The methodology consists of an examination and critical discussion of the relevant literature in different fields like EU law, legal theory and legal reasoning, multilingualism, philosophy of language, theory of translation together with an analysis of the case law concerning multilingual interpretation of EU law, and some other official sources. All these sources are discussed and processed into a theoretical fabric woven with interpretation and discourse theories.

Another interesting point is the book’s interdisciplinary approach in the relevant fields such as EU law, jurisprudence, theories of judicial decision-making, philosophical theories of meaning, translation studies, theories of legal interpretation, discourse theory, systems theory, even venturing into ‘connection theory’ in an intelligible and enlightening way. Elina Paunio skilfully mixes these approaches without losing sight of the legal reasoning of the Court, which is characterised, perhaps not surprisingly, as more teleological than multilingual: ‘the Court must “arbitrate” linguistic conflicts arising from multilingual legislation by using other interpretative criteria.’Footnote ⁷⁹ What is multilingual is the object law but the meaning of this multilingual EU law emerging in judicial decision-making is the result of a complex interplay between texts, context, purpose and discourse.

As with Beck, there is enhanced indeterminacy in the multilingual context but Paunio really draws important conceptual consequences. Already in Chapter 1, she revisits two received views: the comparison of language versions as a tool that might clarify meaning and the ordinary meaning thesis, where Paunio’s book stands in clear contrast with Conway’s.

Moving on from the type of certainty one might intuitively relate to linguistic-semantic methods of interpretation, i.e. predictability and reaffirming the relative indeterminacy thesis, the book puts forward a conception of legal certainty where purpose, telos, and dynamic methods of interpretation are the major sources of meaning construction in a procedural setting that makes dialogue possible (Chapter 2). Purposive, teleological interpretation is the standard here.

As certainty is also a function of rational acceptability, the juristic community is engaged in (ideal) discourse situations and communicative action, and interacts in an on-going dialogue, where the Court has to give persuasive, convincing, coherent and acceptable reasons; and this necessitates not only transparency, sincerity, and dialogue with the relevant audience (rational acceptability) but also procedural rules that favour this dialogue. Yet, in this process, the relevance of other rhetorical devices is not neglected, and the reasoning approach is enriched with other elements calling to ethos and pathos: the connection between Habermas and Perelman, through essential theories like those of Gadamer or MacCormick is very well argued here. Much of Habermas’ theory of communicative action is predicated on a monolingual context where multiplicity of language is not really at stake. It therefore seems challenging to use, as Paunio does in Chapter 3, the theory of rational discourse and ideal consensus when suggesting a new conception of legal certainty based on acceptability, as it would be obtained in a rational discourse situation. In the EU, this would be a multilingual (and translated/interpreted) ideal speech situation.

Interestingly, translation and shared foreign languages are helping bridge the gap and develop a special type of, less spontaneous but still self-conscious plurilingual demos, an issue so dear to Habermas. Anyone who has been at a hearing of the ECJ, or at Parliamentary committees will know that mediated communication between jurists is possible. Translation is one essential tool for inclusion. There is no need to replace one demos for another larger one; after all why should there be only one demos and one single identity, and why cannot there be multiple demoi corresponding, for instance and not exclusively, to multiple language communities and their constitutional speech situations? Special requirements and inclusion strategies will be needed to ensure that all citizens of the Union, and not only the professional jurists, lawmakers and lawyer-linguists can actually engage in this multilingual communication or speech situations. This can be inspired by the inclusive and participatory view of citizenship – one of the other foundations of the EU polity (and this would add a new edge to Sankari’s discussion of citizenship as status) – and communication advocated in Paunio’s book when analysing the audience where acceptability is explored; the more inclusive the audience, the greater the degree of ‘certainty’ obtained.

Indeed, multilingualism is one of the foundations both of legal thinking and of this polity in the making, which is the EU. It is mostly when coming from language environments like Finnish or, in my case, Basque, which are not those of dominant or majority languages, that one realises how fundamental language officiality becomes for the EU as a polity. And it is especially when one is in the theoretical or practical business of comparing language versions of relevant judgments and laws, or when one tries to analyse theories of EU law that have been developed in dominant languages into one’s own minority or lesser-used language, that one becomes aware of the distinctive nature of multilingual interpretationFootnote ⁸⁰ as compared to that carried out in monolingual contexts, as many Anglophone and Francophone authors are prone to do. Paunio focuses directly on the ECJ rather than on the difficulties, even impossibilities for national courts to respect the ECJ’s CILFIT criteria.Footnote ⁸¹

Thus, multilingualism in EU law both requires and boosts a new, more sophisticated understanding of legal certainty. A proper understanding of the ECJ’s interpretative strategy would therefore require moving somehow ‘beyond the words’ in the CILFIT decision, or at least not taking them at face value; after all the Court itself is not very serious about its own CILFIT directives regarding the systematic comparison of language versions in cases where a possible divergence has not been raised as an argument by any of the parties in the main proceedings, the referring court, the intervening parties or the amicus curiae. I would doubt whether the Court would ever follow the Köbler line of member state extra-contractual responsibility for clear/serious judicial breaches of EU law, i.e. for not following the CILFIT criteria as regards the comparison of language versions, but I would not exclude the possibility of the Köbler (or Traghetti) jurisprudence being applied to a stubborn national court of last instance abusively declaring acte clair in situations that should have been submitted for preliminary ruling. Arguably, the main point of CILFIT was precisely to indicate that the interpretation of EU law is not as obvious as some national courts had pretended. The interpretation of EU law requires going beyond words,Footnote ⁸² or at least beyond the plain, seemingly objective meaning of a word in one particular language version, and looking for a more autonomous, cross-language or harmonised meaning, taking into account context, system, objectives, effectiveness and consequences.

The interpreter needs to go beyond the words and look into other dimensions of normative communication to find and determine the meaning, namely to intentions, purpose, telos. Beyond the words is the telos. The standard introductory quotation in the book is from Lewis Carroll’s Through the Looking Glass and its classical dialogue between Humpty Dumpty, who makes words mean what he wants, and Alice: a dialogue ending with the wry reply: ‘the question is who is to be master.’ As Alice could have retorted, why should we recognise the authority of that master, why should we recognise the ECJ as the master? We will need to be rationally convinced (find it rationally acceptable), even rhetorically persuaded. Looking beyond the words of legislative text may help courts configure their decisions to people’s predictions, people’s expectations on purpose and consequences. Thus certainty is also linked to telos and dynamic methods of interpretation as sources of meaning construction.

There are serious theoretical objections to identity of meaning between language versions, and serious practical obstacles to equivalence of meaning, yet the multilingual system seems to work, from a pragmatic point of view. For the system to work perhaps the juristic community operating it has to learn to go beyond words, and an internal cultural explanation, where actors share normative expectations, is in order.Footnote ⁸³ Again, the important distinction between clear cases and hard cases, essentially a pragmatic distinction, is worth remembering.

Going beyond words in a relatively democratic, open system, and in situations inspired by ideal discourse is different form sticking to words in other contexts (say, in dictatorial situations, or in hard-negotiated contracts and agreements, even in punitive contexts). In other words, going beyond words requires the actors to share cognitive and normative expectations and to engage in the ideal discourse situation. Where the actors cannot be relied upon, or where, as in some international treaties, an important cultural distance exists between the actors and cognitive or more principled normative expectations are not shared, perhaps sticking to literal, dictionary, meanings or interpretations might be safer; but it can, arguably, be held that in the EU context we are happily beyond that level of mutual and structural mistrust; we are building a complex community, beyond the very important but limited common market vision, where discussions on the finalities is a shared practice. This could be a response to the democratic objections put forward by Conway’s limits.

Certainty is seen either as objectivity, or as acceptability, but it can also be seen as factual, sociological, certainty, as a result. If there is objective meaning, there is certainty and controllability, and the rule of law seems secured. But because of vagueness, generality and ambiguity of words, of some words in some pragmatic contexts, indeterminacy pervades and a new approach is suggested to recover certainty. A certain uneasiness remains with the result of the switch to this higher level of certainty beyond the literal, or to a deeper level, depending on how one sees it, as regards the rule of law values of predictability and knowledge of the law. Paunio’s twofold distinction of certainty has left aside the third factual dimension, and has opted for a rational acceptability – legitimacy – approach rather than a factual-acceptance legitimation. This choice is much in line with her research question and her methodology.

Translation is another major theme in the book. Translation is the official language of the EU, as translations are law, and the law is translations. The constitutional equality of the official languages carries a legal principle of existential equivalence of language versions, which is still a fiction if meant as identity. Translation is thus better seen as approximation,Footnote ⁸⁴ but approximation to what if there is no clear reference-norm to grasp? The answer is: approximation between language versions. Take the following examples: manifest breach and primacy. The first relates to one of the criteria to determine whether there is extra-contractual liability, i.e. existence of a damage or harm, existence of a breach of EU law and relation of causality and imputation. The breach has been described by the Court in its working language, French, as ‘manquement suffisamment caractérisé’, and has been translated into English as ‘sufficiently serious breach’. Seriousness seems not to be a feature of the breach but of the consequences it might provoke, but caractérisé lacks a clear equivalent in English, especially when qualified by ‘sufficiently’. The use of ‘manifest breach’ has been favoured more recently and this itself has contributed to a clearer understanding of the very terms used in French, suffisamment caractérisé, as manifest – clear or blatant – and this shows how translation is an approximation as interactive feedback between languages and as attempts to modulate and find the right terms within one language over time and across a line of cases. The Court has sought a reflective equilibrium of approximations in a combination of its different players – members and collaborators, lawyer-linguists, readers.

The second example is primauté, which corresponds with primacy of EU law, as developed in Costa v ENEL and Simmenthal. It basically means setting aside or dis-applying domestic law that is incompatible with EU law. Interestingly Declaration 17 to the Lisbon Treaty uses the term ‘primacy’, probably a neologism, with much caution and care, but much of the EU legal doctrine or dogma still clings to the term ‘supremacy’, which does not exactly have the same connotations – sense and reference – as primauté. The result is rather confusing and has important consequences. The debate on constitutional pluralism, a late favourite of much of the EU literature textbooks and articles in English, largely delves into the understanding of primauté as supremacy, which brings along a claim to sovereignty and thus a clash of claims to ultimate authority, national versus European. If the debate had focused on primacy, defeasibility and dis-application of an antinomy, less grand theoretical implications would have been drawn. The approximation of primacy and supremacy in this case has been primarily a work of the academics and commentators and of Eurosceptic politicians and shows a lack of multilingual awareness typical of dominant language communities.

Interesting implications follow from multilingualism for law-making and judicial decision-making. The view that translation is no exact science but an art of approximation and of finding a harmonised meaning has enormous implications for the elaboration of multilingual law. Legislation is a collective enterprise also in monolingual contexts, but even more so in multilingual ones, since the different political and technical inputs at the level of drafting and negotiating, amending and redrafting legal proposals is mediated by translations. There is a constant interaction between language versions at all levels of law-making. The end result is not any specific language, but a combination of these. This is sometimes expressed by the term co-legislation and implies sharing linguistic inputs at all stages of the law-making process rather than following different parallel tracks for each language and then contrasting the end result.

But of course, multilingualism has enormous implications for the judicial application of EU law as well. To begin with, it is not possible to identify an authentic language version of the law, because there simply is none. Next, each party will invoke the linguistic version closest to their professional use, and will do so without contrasting or comparing other versions. The Advocate General might do the same, and then the Court will work from the French but occasionally compare other versions. Going beyond the literal meaning of the words means implicitly finding a sort of cross-linguistic meaning determined by other factors, notably dynamic ones: purpose, effectiveness, results. The doctrine of the autonomous meaning of the terms used by EU law can be understood in this light.

The most radical consequences of the multilingualism of EU law affect the theory of legal norms and ultimately for the theory of law. When postulating the ‘no ordinary meaning’ thesis or even more clearly when one holds the ‘radical indeterminacy’ thesis, there is no pre-determined meaning of the provisions which are the object of interpretation; there is no pre-interpretative sense of the norm: only after the provision gets interpreted and a specific meaning is opted for, can we say that a norm has been constructed from the provision, or from the legal principle. The legal text, the provision, contains norm candidates, and the interpretation determines the norm for the instant case. This theory is much clearer in multilingual law, where the different language versions can potentially contain different norms, but a common, shared, harmonised or superimposed meaning is extracted from them all to create the EU norm.

If norms are only determined at the judicial application stage, one cannot hold that there is any possible interpretation that goes beyond or against the law, praeter or contra legem, because the lex in the original provision to be interpreted is itself not an objective given but rather the result of an interpretation itself. We simply cannot know what the norms are beforehand, we need to interpret them, perhaps under a more intuitive understanding, but still interpret them in the sense of attributing a meaning to their terms. But if this is so, how can we rely on certainty of the law, or on knowledge of the law? We need a higher or deeper conception of certainty, a more sophisticated theory and concept of law as a hermeneutic or interpretative enterprise. This is the concept advocated by Dworkin, and later on by Alexy, or MacCormick.Footnote ⁸⁵

The hermeneutic concept of law moves away form the classical positivist approaches to the norms, legal certainty and objective meaning. As postulated above, this concept comes with a certain political philosophy, a democratic, dialectical democracy where the rationality and validity lie in the process of rational discourse rather than on predetermined normative positions. The EU and its basic constitutional principles set a proper framework for this theory of law to unfold and deliver its discursive potential through dialogue, conversation seeking overlapping consensus. Law is seen as communication, it offers a basis for communication, and it is itself founded on communication, as the preliminary reference procedure before the ECJ shows. The procedural rules facilitate this communication and it would be interesting to analyse them from this functional point of view. The key issue is of course whether this communication goes in all directions, or whether only one authority is sending the normative messages and not receiving inputs about its interpretative decisions. The question is not only who gets access to the Court, but also: to what extent do the immediate and the wider audiences participate in the ideal speech situation? Legal certainty and ideal speech situations are regulative ideals, but inclusive procedures and communicative action at the Court can contribute to an approximation to the ideal.

It must be said that these theoretical niceties are not routine at the Court. As we said at the outset and as Paunio recognises, normally interpretation and decision-making proceed without too many major disagreements, even where cases are disputed and by definition: if they reach all the way to the ECJ, it must be because they contain difficulties. But perhaps this legal-political-cultural understanding that the Court will always go beyond words, and will be expected to do so by parties and the juristic community, is sufficient to explain why certainty is seldom seen as seriously compromised. The rule of law is taken to a higher level, removed from textuality into contextuality and argumentation. This context sensitivity is an important contribution of the book, and it relates again, to rational acceptability as a key criterion of substantive certainty.

Justification consists in giving good reasons for the decision. There is a connection between rational acceptability and justification since the ‘goodness’ of the reasons can be defined precisely as rationally acceptable reasons in a given audience. But justification can have connotations close to objectivity, and this is probably why justification been deliberately avoided by Paunio when constructing this theory of substantive, discursive certainty concerning interpretation and meaning. Teleological reasoning from the purposes of a legal provision that links individual interpretations to a broader systemic context exposes this theory to debate within the legal community rather than to the Court justifying its decisions (which is also relevant, of course). Meaning is harmonised or stabilised by communicative practices of interpretation between the juristic communities involved in the judicial resolution of disputes in this multilingual community.

The innovative focus on substantive certainty linked to rational acceptability and ideal discourse could have also been matched by a more forceful and innovative approach to authority at the Court, in relation to the possible discussion of separate opinions or at least more transparent motivation. We know how important it is for the Court as an actor to engage in discourse both within the individual case – with the parties, interveners, referring courts, amici curiae – and across cases – with national and international courts, the juristic community of law professionals, the EU legislators, academics. But the Court is a collective agent and the issue is how to guarantee internal discourse, dialogue and debate within the Court: between the General Court and the Court of Justice, between the Advocates General and the Court, between chambers within the Court, but especially between the judges in the instant case and the chamber.

Paunio has been understandably cautious in this regard. Separate opinions are only part of the question, but more information about the debates within the Court could be a modestly welcome move requiring no constitutional amendment, but only a new practice of transparency. The main argument against separate opinions and for not disclosing information concerning internal debate or differences of views inside the Court are normally related to authority: there would be an erosion of authority on the part of the Court if it showed its weaknesses and internal differences. Quite to the contrary, Paunio could have said, had she taken a more innovative approach to authority as she has done towards certainty.Footnote ⁸⁶ We could move to a higher or deeper conception of authority linked to transparency and sincerity, and to recognition that an issue is moot and that different views are theoretically and practically possible but that a majority within the Court has opted for one particular interpretation. After all Paunio herself sees legal concepts as the results of communicative processes, and it would help to know about the internal communicative processes, discursive practices and speech situations inside the Court. This discourse ethics need not erode legitimacy, as one might detect from a comparative analysis of the European Court of Human Rights.

Closing comments - the legal reasoning of the ECJ and European legal culture

Having analysed the four books on the legal reasoning of the Court, it might be worthwhile to briefly comment on some of the contentious issues. As we have seen, approaches to the Court’s reasoning are roughly descriptive or normative. But the descriptive focus can be superficial or analytical and reconstructive. In a superficial analysis, the Court is described as using one argument or another in order to interpret a provision or a concept. This analysis requires little or no theoretical background and does not aspire to have any predictive impact. Case notes typically follow this line of discourse as a prelude to the analysis of the ratio decidendi and the analysis of the substantive arguments from a dogmatic point of view. The normative focus, as advocated by Conway, is not interested in this descriptive dimension and moves directly to postulate how the Court has to decide its cases and to impose from the outside the standards upon which it is to be evaluated: sticking to the ordinary meaning, paying heed to the subjective intention of the law-maker and (thus) respecting the rule of law. But, arguably, that is already a partie prise. It is founded on the threefold premise that ordinary meaning is a relatively straight-forward matter, that, when it is not, the subjective intention is at least as clear and decisive, and that the Court does not respect the rule of law when it resorts to teleological or dynamic methods. Situations of doubt and disagreements on meaning are minimised or downplayed and judicial methodology becomes like following a manual, simply following the rules. Between the superficial descriptive and the normative approaches we find the rational reconstruction, hermeneutic or interpretative approach. It remains descriptive, but adds an internally critical dimension that allows the observer to evaluate the Court’s decision-making on its own standards, as analytically reconstructed from a series of sources amongst which the Court’s own doctrine of interpretation as expressed in its own case-law features prominently. Beck, Sankari and Paunio adopt this approach, and Beck is explicit in rejecting the normative level.

Secondly, the distinction between the context of discovery and the context of justification remains relevant in any analysis of the Court’s reasoning. The frontier between the two contexts is not always clear when it comes to silences or to the methods of interpretation. As regards silences, almost by definition, they will not feature explicitly in the judgment. Silences are to be identified elsewhere, in a different level of analysis: discovery. How the Court decides to frame a legal question, the classification or qualification of a case as falling within free movement or environmental protection, the level of generality of the ratio, the arguments that are put forward by some parties but not addressed to the extent that we can get to know them, the unstated consequences of the decision: these are all interesting elements that blur the distinction between discovery and justification, or rather that require a more sophisticated research into and theory of discovery.Footnote ⁸⁷ Sankari’s monograph is a welcome move in this direction.

As regards interpretation, criticisms of the Court’s activism basically rely on a discovery-related interpretation, i.e. that it is pursuing a given agenda and that before it moves on to the analysis of the issues of interpretation raised by the case and by the parties and interveners, it has already made up its mind on how it is to be decided. Thus, Conway criticises the Court for following an agenda that is openly communautaire.

On the other hand, analysing the context of justification, it is possible to reply that the Treaties themselves, taken as a system, and the specific articles, as well as most EU legislation is explicitly and openly pursuing certain objectives, and the Court is resorting to these in its interpretation and doing so transparently. As Sankari shows in her analysis of the citizenship cases, if the integrationist agenda were straightforward it would be easy to predict the Court’s decisions. Beck has stressed how futile it is to expect predictability from the ECJ, or indeed any court. And Paunio underlines one very credible explanation for this, i.e. the idea that there is no pre-given ‘natural’ meaning. Even Conway, as we have seen, does not plainly equate literal and ordinary meaning. If it is all a matter of interpretation, it seems difficult to picture the Court, a collegiate body working in a multilingual context, as having pre-decided the case and simply window-dressed the decision by a careful choice of methods of interpretation and an elaborate reasoning. But even if we assumed the decision was already made, we still have the reasoning to see if it can be justified, first according to its own standards of interpretation (rational reconstruction again), and secondly, by analysing the general acceptability of the judgments. This second point is not blunt. It may be that in the early years the Court profited, if we can use this optimistic term, from a ‘benign neglect’. Perhaps this was generally the case for all the institutions. But at any rate, this is clearly not the case now.Footnote ⁸⁸ The Court is well aware of reactions to its case-law and by analysing the lines of cases and decisions on a given subject like anti-discrimination or consumer protection to mention but two,Footnote ⁸⁹ we can see how the Court refines its interpretation by distinguishing or limiting the ratio.

This is why the neglect was perhaps not benign after all. Courts need to be subjected to criticism. Conway has a point when he shows discomfort concerning all the praise the Court receives. Maybe he exaggerates in calling this the ‘language of love’. Yet, it is not clear that the Court is at its best, from the point of view of legal reasoning, whenever it feels it is being closely watched, when member states and all the parties to the case intervene, when the media are closely following a case. Beck (and he is not alone in this) is very critical of Pringle, for instance, a case that raised much media attention. Perhaps there is consensus that the context of discovery in this case accounts for the less than optimal reasoning in this case. The quality of the Court’s reasoning depends on other factors, as I have tried to analyse elsewhere.Footnote ⁹⁰ But critical reactions to the Court are an essential input for quality of judicial output, just as in the academic context. The Court has long known it cannot ‘run wild’.Footnote ⁹¹

Debate and open rational discourse are crucial not only for the quality of legal reasoning. A new hermeneutic concept of law develops from this critical dialectical dimension, a concept of law based on following rules, but also on contributing to their recognition and development by citizens who adopt a critical internal point of view. In this sense, it is worth remembering how studies of legal reasoning took off in the late 1970s and early 1980s. The debate in legal theory was still dominated by the contrast between natural law, legal realism, legal positivism and issues of interpretation were secondary. Where the rules did not clearly resolve a penumbra situation, the judge was seen as having full discretion in either Hart’s or Kelsen’s positivism. This meant full-blown practical discretion. Yet, legal realism had shown there were important steadying factors, or limits to discretion. Natural law also provided some normative guidance, but from an extra-legal dimension. Dworkin’s response was to take rights seriously and enhance the forum of principles, understood as cohering and integrating the law. Jurists seemed to share many of these principled approaches to the law in the modern constitutional culture Dworkin was reconstructing.

In Europe the emphasis was on reason in law, and the analysis of theories and arguments of legal interpretation, decision-making and topoi. The Brussels school of rhetoric (Perelman, Olbrechts, Tyteca), the impact of hermeneutics (Gadamer) and of discourse theory in Germany (Habermas), the study of legal cultures and legal argumentation (Tarello, Guastini and the Genoa school, but also Conte, Jori, Pattaro and many others) and analytical legal philosophy applied to judicial decision-making (Wróblewski) all paved the way for the first ground-breaking analyses of legal reasoning as a special case of practical reasoning in philosophy with a group of authors that would eventually converge in the so-called Bielefelder Kreis: MacCormick, Summers, Wróblewski, Aarnio, Peczenik, Alexy, and Taruffo.Footnote ⁹²

Perhaps these theories came at a turning point in constitutionalism where the legislature was losing the leading role it had in the codification age and even in the highly regulatory post-war welfare state where the will of the legislator (voluntas) and the technical and bureaucratic know-how of the administration became prominent and the judiciary was confined to a lesser role applying the legal syllogism as la bouche de la loi, on the basis of their traditional authority (autoritas) and their knowledge of the law, at least on the Continent. In fact, this was especially the case on the Continent, where EC law was emerging as a special case of administrative and international law. The EC added a further, quasi-federal, dimension to the mediation of judicial authority by institutional and political processes.Footnote ⁹³ Then came the importance of fundamental rights, the European Court of Human Rights, and of constitutional courts. The European Court of Justice also experienced these developments,Footnote ⁹⁴ especially after the accession of the common law jurisdictions.

Legal reasoning has gained ground due to law’s and the courts’ central role in contemporary liberal democracy as both enabling and constraining political power and mediating politics with other social systems and with society at large. In this sense the requirement to justify judgments and other decisions in law by proving good reasons – acceptable reasons according to the legal culture – is a factor constraining decision-makers’ discretion, and limiting the arbitrariness of decision-makers’ decisions, and ultimately making them accountable to citizens and society at large. Courts could no longer rely exclusively on their authority, and even though courts remained resilient to democratic models of governance, they at least had to display ratio, their ultimate source of legitimacy when applying the law.Footnote ⁹⁵

This analytical interest in reasoning was accompanied by an interest in conflict resolution where courts have become a central element in a complex system of dispute resolution, a juridical field, to use Bourdieu’s term.Footnote ⁹⁶ Legal reasoning, providing good reasons, shows the decision-makers’ respect for the addresses of their decisions thus inducing their compliance with the decision while at the same time enabling their criticism of the decision.Footnote ⁹⁷ This argumentative dimension has contributed to an enhanced recourse to, and trust in, the courts; in a context where other players – parties, repeat players, institutions, the media – are also interacting.

The ECJ has been no stranger to these developments. Indeed, one can say that it has brought along a culture of legal reasoning and argumentation, a given dynamic style, that has influenced national courts’ approach to law, at least when interpreting and applying EU law. And thanks to books like the ones under review in this contribution, we are gaining an ever more sophisticated understanding of its legal reasoning and a modest epistemological ground from which to evaluate it critically.