Multilingualism and its Consequences in European Union Law

Abstract

When discussing the nature of EU law it is common usage to highlight the principles of direct applicability and of primacy as the two fundamental features that distinguish EU law both from classic international treaty law and, to a certain extent, from domestic law. Increasingly reference is also being made to the consequences of the principle of sincere cooperation, which, contrary to direct applicability and primacy, is being expressly mentioned in the EU Treaties, i. e. in Art. 4.3 TEU. Strangely enough most of the EU law literature does not dwell on a fourth principle, which to my view is of fundamental importance to the nature of EU law, namely the principle of multilingualism. This is the reason why I thought it necessary to dedicate a whole section to multilingualism in my own handbook of EU law, in the chapter on Fundamental Features of EU Law. Interestingly most colleagues, albeit acknowledging the originality of my approach, told me that you could not put multilingualism at the same level as direct effect or supremacy. Albrecht Weber, this book’s addressee, is not one of the latter colleagues; on the contrary, throughout his major work on European Comparative Constitutional Law, Weber clearly shows his consciousness of the importance of language differences, including the fact that “the signification of a concept may break away from its etymological sources and bring forward a different content; even concepts such as Constitution, Verfassung, Grundgesetz, Grondwet (the Netherlands) show differences in their significations.”

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Multilingualism as a Fundamental Feature of EU Law

When discussing the nature of EU law it is common usage to highlight the principles of direct applicability and of primacy as the two fundamental features that distinguish EU law both from classic international treaty law and, to a certain extent, from domestic law. Increasingly reference is also being made to the consequences of the principle of sincere cooperation, which, contrary to direct applicability and primacy, is being expressly mentioned in the EU Treaties, i. e. in Art. 4.3 TEU. Strangely enough most of the EU law literature does not dwell on a fourth principle, which to my view is of fundamental importance to the nature of EU law, namely the principle of multilingualism. This is the reason why I thought it necessary to dedicate a whole section to multilingualism in my own handbook of EU law, in the chapter on Fundamental Features of EU Law.¹ Interestingly most colleagues, albeit acknowledging the originality of my approach, told me that you could not put multilingualism at the same level as direct effect or supremacy. Albrecht Weber, this book’s addressee, is not one of the latter colleagues; on the contrary, throughout his major work on European Comparative Constitutional Law,² Weber clearly shows his consciousness of the importance of language differences, including the fact that “the signification of a concept may break away from its etymological sources and bring forward a different content; even concepts such as Constitution, Verfassung, Grundgesetz, Grondwet (the Netherlands) show differences in their significations.”³

One issue is more important than the theoretical question of where to put the principle of multilingualism: the multilingual nature of the EU is only too often ignored by scholars who refer to only one language version of EU primary and secondary law, and – maybe worse – of CJEU case law. This is more and more common to the authors who use only the English language version for references, albeit writing in another language. Such an attitude sometimes leads to important misunderstandings of the concepts used in EU law. Furthermore, due both to its multilingual nature and to the multiplicity of concepts to be dealt with in view of the growing number of EU Member States, scholars who are excellent specialists of domestic law very often encounter problems with EU law, due to the fact that the same words do not have the same meaning when used in EU law and in domestic law. This chapter will recall and develop the legal notion of multilingualism of EU law and its consequence, before discussing the autonomy of EU law concepts with regard to concepts that are being used with the same wording in the domestic law of EU Member States.

In way of introduction, it is worthwhile to give a few illustrations of misunderstandings generated by multilingualism. The first of those misunderstandings relates to a somewhat strange line of literature, which developed at the end of the 1990s in the framework of the discussion of a European civil code: some scholars started to explain that there could be no harmonization of property law because the EC had no competence in that field (which is true), in quoting Art. 295 EC,⁴ according to which “[t]he Treaties shall in no way prejudice the rules in Member States governing the system of property ownership”; such a reasoning was implying that on the other side the EC had a competence for the harmonization of contract law or torts law (which is wrong), as there was no provision comparable to Art. 295 EC with respect to contracts or liability law.

Such a discussion seemed totally out of place to French or Italian native‐speaking EU law specialists, but not too many others. A comparison between the four language versions of 1957 reveals the source of the misunderstanding. Art. 345 TFEU (ex‐Art. 295 EC, wording unchanged since 1957) says “[l]es traités ne préjugent en rien le régime de la propriété dans les États membres” and “[i] trattati lasciano del tutto impregiudicato il regime di proprietà esistente negli Stati membri”, making it obvious that the underlying issue was that of public enterprises and nationalizations, which existed in different forms in all the founding Member States, a hot issue of home political debate in all the Parliaments that would eventually have to give their authorization for ratification. The German version “[die] Verträge lassen die Eigentumsordnung in den verschiedenen Mitgliedstaaten unberührt” undoubtedly reflected the same idea. However the Dutch version said “[d]e Verdragen laten de regeling van het eigendomsrecht in de lidstaten onverlet”, a wording that might be read in literal terms as the Treaties shall not prejudice the regulation of property law in the Member States. I submit that any EU lawyer accustomed to the methods of interpretation of EU law would come to the conclusion that the English version – which was produced in 1972 – did its best to reflect the idea of the French, German and Italian versions, which as a matter of fact was in no way contradicted by the Dutch version, and hence deduct that Art. 345 TFEU has nothing to do with the issues of conferral. On the contrary, as happens for contract law or torts law, if a legal issue arises that falls into the scope of EU law, there is no impediment for harmonization in order to solve that specific issue. This being said, the misunderstandings I am referring to with this example have not spilled over from academia to practice, to my knowledge.

A second example shows how multilingualism can raise political problems. The Constitutional Treaty of 2004 contained a provision that has been sacrificed in the Lisbon Treaty in order to save the reforms on which there had been agreement during the IGC of 2003–2004, and which had been proposed by the European Convention 2002–2003: Art. I‐6 of the Treaty establishing a Constitution for Europe stated: “The Constitution and law adopted by the institutions of the Union in exercising competences conferred on it shall have primacy over the law of the Member States”, “Die Verfassung und das von den Organen der Union in Ausübung der der Union übertragenen Zuständigkeiten gesetzte Recht haben Vorrang vor dem Recht der Mitgliedstaaten”, “La Constitution et le droit adopté par les institutions de l’Union, dans l’exercice des compétences qui sont attribuées à celle‐ci, priment le droit des États membres”, “La Costituzione e il diritto adottato dalle istituzioni dell’Unione nell’esercizio delle competenze a questa attribuite prevalgono sul diritto degli Stati membri”. It is worthwhile noting that the French text of the European Convention’s draft Treaty (Art. 10) was saying “La Constitution et le droit […] ont la primauté sur le droit des Etats membres”, which was identical to the English version – knowing that the European Convention was discussed on drafts that were presented in English and French. Whereas the General Secretariat of the Council, which prepared the text that was submitted to the representatives of Member States during the 2003–2004 IGC, preferred to submit a French version that it deemed more elegant – a position that I do not share – the English version remained unchanged; that is probably not by chance. Indeed, discussion in Member States in the framework of the preparation and adoption of the Constitutional Treaty⁵ showed that – at least for non‐English native speakers – the word supremacy, which had been hitherto used by English language literature for the concept of primauté/Vorrang raised a lot of political opposition, while the word primacy raised less opposition, maybe because it was not used in US or UK law, contrary to what was happening with the wording of Article Six of the United States Constitution: “This Constitution, and the Laws of the United States which shall be made in Pursuance thereof; and all Treaties made, or which shall be made, under the Authority of the United States, shall be the supreme Law of the Land; and the Judges in every State shall be bound thereby, any Thing in the Constitution or Laws of any State to the Contrary notwithstanding”.⁶ As a matter of fact, the Spanish Constitutional Court constructed an entire reasoning upon the difference between supremacía and primacía in its Ruling of 13 December 2004 on the Constitutional Treaty,⁷ in order to reconcile as much as possible the supremacy of the Constitution in the domestic order with the primacy of EU law.

A third example shows how scholarship can be abused by non‐existing conceptual differences in case law, due to varying translations. For some reason, the case law of the CJEU – at least before the adoption of the Charter of Fundamental Rights – contained variations in the translation of the words bonne administration: whereas the French version of the relevant rulings (which are always drafted in the French language) invariably said bonne administration, in other languages different terms are being used. For instance, in the German version of the Burban ruling⁸ (a case where the language of the proceedings was also French, hence the only with authentic value), the translators preferred ordnungsgemäße Verwaltung instead of gute Verwaltung, proper administration instead good administration, sana amministrazione instead of buona amministrazione. The English translators sometimes used sound administration or even good governance for the same bonne administration, the Italian translators buon andamento dell’amministrazione, etc.⁹ Much ado for nothing? The point is that there is a temptation for scholars who do not work on the French version to elaborate absurd typologies on the basis of non‐existing differences. In order to offend nobody, I prefer not to give any references, but I can testify that I had to deal with doctoral students’ works in the European University Institute, which showed the reality of such a risk.

Last but not least one such difference in wording is certainly at the root of some of the divergences between on the one side the CJEU and most of the non‐German speaking scholars and on the other side the Bundesverfassungsgericht and a significant number of German Staatsrechtslehrer. The issue is the wording of Art. 51.1, first sentence EUCFR on the scope of application of the Charter. The English version says “[t]he provisions of this Charter are addressed to the institutions, bodies, offices and agencies of the Union with due regard for the principle of subsidiarity and to the Member States only when they are implementing Union law”, the French “[l]es dispositions de la présente Charte s’adressent aux institutions, organes et organismes de l’Union dans le respect du principe de subsidiarité, ainsi qu’aux États membres uniquement lorsqu’ils mettent en œuvre le droit de l’Union”; but the German says “Diese Charta gilt für die Organe, Einrichtungen und sonstigen Stellen der Union unter Wahrung des Subsidiaritätsprinzips und für die Mitgliedstaaten ausschließlich bei der Durchführung des Rechts der Union”. It is well known to specialists that the German wording was personally chosen by Roman Herzog, Chairman of the Convention which drafted the Charter, that it was one of the rare cases where the texts discussed in the Praesidium were not drafted only in the French and English versions, and furthermore that the English word implementing as well as the French mettent en oeuvre were chosen on purpose because they admit a broader interpretation than executing and exécutent. As a matter of fact the “Explanations” in the Charter¹⁰ state that “as regards the Member States, it follows unambiguously from the case law of the Court of Justice that the requirement to respect fundamental rights defined in the context of the Union is only binding on the Member States when they act in the scope of Union law [wenn sie im Anwendungsbereich des Unionsrechts handeln] […] The Court of Justice confirmed this case‐law in the following terms: ‘In addition, it should be remembered that the requirements flowing from the protection of fundamental rights in the Community legal order are also binding on Member States when they implement Community rules [müssen bei der Durchführung der gemeinschaftsrechtlichen Regelungen aber auch] […]’”. Those Explanations perfectly reflect the differences in positions in the Praesidium and can be used in order to support either interpretation: the restrictive one preferred in Germany and by the Bundesverfassungsgericht ¹¹ or the broad one preferred by the CJEU in Åkerberg‐Fransson ¹² but also by most of the commentators in other languages.

2 The Multilingual Nature of EU Law

The multilingual nature of EU law cannot be compared with the experience of any other international organization and even less to that of most multi‐lingual States, as a result not only of the number of languages involved, but also the fact that no language has a legal status superior to others.

Art. 3 TEU on the objectives of the Union provides that “[the Union] shall respect its rich cultural and linguistic diversity” and Art. 4 TEU that “[t]he Union shall respect the equality of Member States before the Treaties as well as their national identities”. There is quite some discussion about the meaning of “national identities”, but there are no divergences on the fact that language is a fundamental part of national identity. As a consequence, it not admissible to give preference to one of the official languages, whether English, French or any other language.

2.1 The Origins of EU Linguistic Regime

The ECSC Treaty of 1951 was drawn up in French only. It did not contain any provision on the use of languages of the Community institutions. However, the subsequent recognition of Dutch, French, Italian, and German as official languages and working languages of the ECSC was an easy decision for the Foreign Ministers of the Member States in the light of the practical necessities encountered in managing the ECSC that foreshadowed the language regime of the two Communities that would be born a few years later.

Within the framework of the Treaties of Rome of 1957 the language issue was addressed and solved right at the start. Both Treaties were signed in the official languages of the Contracting States: Dutch, French, Italian and German. Art. 248 EEC (now Art. 55 TEU) as well as Art. 225 Euratom, listed the official languages of the Treaties (nowadays 24, since the accession of Croatia) and further provided that “the texts in each of these languages being equally authentic, shall be deposited in the archives of the Government of the Italian Republic, which will transmit a certified copy to each of the governments of the other signatory States”.¹³ Art. 217 EEC (now Art. 342 TFEU) provided furthermore that “by means of regulations the languages of the institutions of the Community [now Union] shall, without prejudice to the provisions laid down in the rules of the Court of Justice, the Council, acting unanimously.”

The first Regulation ever adopted by the Council of the EEC is Regulation No. 1/58 determining the languages to be used by the European Economic Community,¹⁴ which laid the foundations for full multilingualism. Art. 1 of Regulation 1/58 provided that “[t]he official languages and the working languages of the institutions of the Community shall be Dutch, French, German and Italian”. Those were the four official languages of the Member States of the Communities until 1972. Dutch was the official language not only of the Netherlands but also one of the official languages of Belgium; French was the official language not only of France but also one of the official languages of Belgium and Luxembourg, and an official language in the Valle d’Aosta in Italy; German was not only the official language of the Federal Republic of Germany but also one of the official languages of Belgium and Luxembourg, and an official language in the Trentino‐Alto Adige autonomous provinces in Italy.

On the occasion of the successive enlargements of the Communities and EU, the principle of multilingualism has always been maintained, and EEC Regulation 1/58 has been updated from time to time to take account of new accessions, and therefore the number of treaty languages has grown from four to seven in 1973, eight in 1981, ten in 1986, twelve in 1995, twenty one in 2004, twenty three in 2007 and twenty four in 2013. With future enlargements the same technique will be used as was used so far, i. e. adding a new paragraph specifying “by virtue of the accession treaties, are equally authentic versions of this Treaty language […]”, and correspondingly amending Regulation 1/58.

In 2005, the Council amended Regulation 1/58.¹⁵ A special characteristic of this latest change was that it was not needed due to a new enlargement, but it only intended to “regularize” the position of the Irish language, which had until then a somewhat special status, as Irish has been mentioned as one of the authentic treaty languages since 1973 but was not established as a fully fledged official/working language in Regulation 1/58. Ireland requested that Irish be added in Regulation 1/58 as a consequence of the acknowledgment of Maltese as a treaty and official/working language, with the revision of Regulation 1/58 that occurred as a result of the 2004 enlargement. In Malta, as in Ireland, there are two official languages: Irish or Maltese, and English. Albeit having requested and obtained the relevant amendment to Regulation 1/58, the Irish government has so far favoured maintaining the exemption – established since 1973 – of an Irish language translation for all the acts of the Union.

Art. 41.4 EUCFR provides for the right – also embedded in Art. 20.2 lit. d TFEU – to address the institutions offices, bodies and agencies of the Union in the language of one’s choice and to obtain a reply in the same language. Provided it is one of the official languages of the Union. Art. 2 of Regulation 1/58 already provided in its first version that “[d]ocuments which a Member State or a person subject to the jurisdiction of a Member State sends to institutions of the Community may be drafted in any one of the official languages selected by the sender” and that “[t]he reply shall be drafted in the same language.” The Regulation provided moreover in Art. 3 that “[d]ocuments which an institution of a the Community sends to a Member State or to a person subject to the jurisdiction of a Member State shall be drafted in the language of that State.” The Treaty of Amsterdam in 1997 introduced into primary EU law the rule of Art. 20.2 lit. d TFEU according to which EU citizens enjoy “the right to petition the European Parliament, to apply to the European Ombudsman, and to address the institutions and advisory bodies of the Union in any of the Treaty languages and to obtain a reply in the same language.”

To summarize the language regime of EU institutions, bodies, offices and agencies¹⁶ in a nutshell, the principle is that all 24 languages are official and working languages, according to the provisions of Regulation 1/58. That Regulation applies in the absence of more specific provisions established by the regulation establishing an office or agency – as is the case for OHIM, the EU’s agency for trademarks, whose working languages are only five.¹⁷ It has to be stressed that the Commission’s Rules of Procedure¹⁸ do not establish what its working languages are, but Art. 25 empowers the Commission to establish rules of application of the Rules of Procedure; as a consequence the working languages of the Commission are identical to the official languages, albeit the main de facto working language is English (for about 65 % of non‐binding documents which the Commission produces), followed by French (less than 20 %), German, Spanish and Italian (less than 15 % on the whole). The CJEU has a language regime of its own, established in Protocol No 3 on the Statute of the CJEU: the authentic language of Court rulings is the language of procedure of the case, i. e. normally the treaty language chosen by the referring party or the language of the addressed Member State in case of a procedure for failure to fulfil its obligations. The provisions of Protocol No. 3 on the language regime are amongst those that can only be amended by the treaty amending procedure of Art. 48 TEU, while most of the procedural provisions of the Protocol can be amended through the ordinary legislative procedure (Art. 281 TFEU).

The issue of the cost of multilingualism is often being used instrumentally in order to request a reduction of the working languages or even of the official languages of the EU. As for costs, no precise figures are available; anyway the costs in time devoted to translation should be added to the financial costs. Complaining about these costs is rather sterile, especially in the light of more important values such as the rule of law and citizenship; after all, also democracy has important financial costs. At any rate, the principles of direct applicability and of uniform application of EU law have a necessary consequence, i. e. the availability of binding EU law texts in all the official languages of the Member States. If EU institutions were not to carry the costs of the work that is needed to establish all language versions of EU law, the relevant costs would have to be borne by citizens and legal persons of the countries whose language would not be an official language of the Union. Those budgets would have to bear not only the direct cost of translation, but also the costs arising from the absence of authentic value of the translated versions of EU acts. Indeed the absence of an authentic value of home‐made translations would presumably lead to a multiplication of ill‐founded court actions by individuals and businesses, and to a multiplication of infringement actions against Member States. It is better therefore to consider multilingualism as a fact the costs of which one can only work on marginally. It would be in any way wrong to think that the language problems in the EU may be reduced in the future with greater integration of Member States: in the United States of America the growth of multilingualism is becoming a novel issue in the discussion on democratic institutions and amongst others for participation rights.¹⁹

2.2 The Principle of Equal Value of all Linguistic Versions of EU Legal Texts

The principle of equal legal value of all language versions of EU acts is the central feature of the language regime of the EU. Multilingualism, as guaranteed by the Treaties and the Charter and by Regulation 1/58, is deeply rooted in the very nature of EU law.

As indicated by the CJEU in its Skoma‐Lux ruling, the Court “has held that the principle of legal certainty requires that Community legislation must allow those concerned to acquaint themselves with the precise extent of the obligations it imposes upon them, which may be guaranteed only by the proper publication of that legislation in the official language of those to whom it applies […].²⁰ In addition, it would be contrary to the principle of equal treatment to apply obligations imposed by Community legislation in the same way in the old Member States, where individuals have the opportunity to acquaint themselves with those obligations in the Official Journal of the European Union in the languages of those States, and in the new Member States, where it was impossible to learn of those obligations because of late publication. Observing fundamental principles of that kind is not contrary to the principle of effectiveness of Community law since the latter principle cannot apply to rules which are not yet enforceable against individuals.”²¹ The quoted judgement implicitly refers also to its jurisprudence in Kik, where the Court, contrary to the Court of First Instance held that “the Treaty contains several references to the use of languages in the European Union. None the less, those references cannot be regarded as evidencing a general principle of Community law that confers a right on every citizen to have a version of anything that might affect his interests drawn up in his language in all circumstances.”²²

In case of divergence between the various language versions of the EU Treaties or Acts, the Court has clearly established that none of the versions prevails over the other; and this is true even if a version could be identified as an “original” text in which the document was drafted before being translated. Since the entry into force of the Treaty of Accession of Denmark, Ireland and the UK Danish and English versions of the previously existing law, including treaties, are also in force, which have the same value as those in French, Italian, Dutch and German versions, in spite of the fact that it is easy to show that the Rome Treaties were negotiated primarily in French, Italian and German and afterwards translated into Dutch in view of the signature that took place on March 27th 1957.

The ruling of the ECJ in Stauder is very interesting and very illustrative of language issues in the interpretation of EU law. The issue that was raised before the Court of Justice stemmed from a difference between versions of a Commission decision that restricted the supply of butter at reduced prices to beneficiaries of certain forms of social assistance. In the German and Dutch versions of Art. 4 of that decision, it was stated that “the Member States must take all necessary measures to ensure that beneficiaries can only purchase the surplus butter on presentation of a coupon indicting their names (auf ihren Namen ausgestellt)”, whilst in the other versions it was only stated that a “coupon referring to the person concerned” had to be shown. The Court stated in this regard that “when a single decision is addressed to all the member states the necessity for uniform application and accordingly for uniform interpretation makes it impossible to consider one version of the text in isolation but requires that it be interpreted on the basis of both the real intention of its author and the aim he seeks to achieve, in the light in particular of the versions in all four languages. In a case like the present one, the most liberal interpretation must prevail, provided that it is sufficient to achieve the objectives pursued by the decision in question. It cannot, moreover, be accepted that the authors of the decision intended to impose stricter obligations in some member states than in others.”²³

There is thus an abundant case law on the interpretation of treaties and secondary legislation in case of divergence between the various language versions. The consequences of the principle may be summarized as follows. Since all language versions have the same authentic value, including those which are a translation subsequent to the adoption of the rule – as happens usually as a result of enlargement to new Member States – interpretation requires a comparison of all language versions. All versions have the same weight, irrespective of the statistical significance of the population that uses the language. In case of discrepancies, the norm must be interpreted in the light of the purpose and logic of the general rules of which it forms part. The ambiguity of a language version can be solved by resorting to other versions if they are clear and if the comparison can reveal errors in the formulation of one of the languages. These principles apply to the interpretation of legal acts, which are, in principle, published in all official languages, unlike decisions addressed to single subjects: the latter are translated into other than the original authentic languages only when published for the information of possible interested parties. As regards the judgments of the Court of Justice only the language version in the language of the case will prevail in case of divergence, even though, in practice, the use of the French version can be of great help to the interpreter, since French is the working language of the Court, and especially the language in which the judge rapporteur draws up the text of the judgment.

A special procedure ensures that the drafting of any legislative proposal by the Commission has been set up within its legal service in order to guarantee the possibility of finding an equivalent for each legal concept that is understandable in all official languages.²⁴ The Council Legal Service has established a procedure that leads to the same result for the final version of the texts approved by it. Differently, it should be noted that the amendments of the Treaties have always been negotiated in one or two languages (usually in French, and to a lesser extent also in English). After their political approval – usually during a meeting of the European Council – the other language versions are prepared before the formal signature of the Treaty, work which explains why usually one to two months are needed between approval of a new Treaty in principle and its signature. The Constitutional Treaty of 2004 and the Treaty of Lisbon of 2007 have been occasions of wide ranging checks of the correspondence between the different versions of the Treaties; for this purpose some changes of primary legislation were only made in some versions.

Even in the absence of obvious differences between the language versions of the Treaties or EU acts, as well as of the judgments of the Court of Justice, the multilingual character of EU law has an important impact on the interpretation of EU law. It is worth mentioning the role of the principle of equal legal value of linguistic versions in the application of the acte clair theory. In the 1982 CILFIT Case, regarding an issue raised by an Italian importer of wool who had to pay fees for health inspections of wool imported from non‐EU countries, the Court has established the conditions under which the national courts may refrain from submitting to the Court of Justice a question for interpretation of Community law. As the Court explained “the correct application of community law may be so obvious as to leave no scope for any reasonable doubt as to the manner in which the question raised is to be resolved. Before it comes to the conclusion that such is the case, the national court or tribunal must be convinced that the matter is equally obvious to the courts of the other member states and to the court of justice. Only if those conditions are satisfied, may the national court or tribunal refrain from submitting the question to the court of justice and take upon itself the responsibility for resolving it. However, the existence of such a possibility must be assessed on the basis of the characteristic features of community law and the particular difficulties to which its interpretation gives rise. To begin with, it must be borne in mind that community legislation is drafted in several languages and that the different language versions are all equally authentic. An interpretation of a provision of community law thus involves a comparison of the different language versions.”²⁵

As already indicated, the national courts refer usually only to the version in their own language. However, there are exceptions. For example, the Finanzgericht Hamburg, in connection with a request for a preliminary ruling in the proceedings in Stolle, compared different language versions of a note attached to a Regulation²⁶ and pointed out that “the German version of that note differs from the French and English version of it as regards the definition of the concept of ‘drawn’. In particular, although the French and English versions state that, in order for a poultry carcass to be classified under subheading 0207 12 90, the carcass must be ‘completely drawn’, the German version states that all of the giblets must be removed (‘sämtliche Innereien entfernt sind’). Consequently, the referring court wonders whether the German version of those notes correctly reproduces the intention of the European legislature.” The CJEU answered that “although the expression ‘sämtliche Innereien entfernt sind’ in German does not constitute a literal translation of the words ‘complètement vidés’ in French and ‘completely drawn’ in English, the specification in the German version that all the organs of a carcass coming under subheading 0207 12 90 must be removed is not really different in meaning from that of the French and English versions specifying that the carcass must be completely drawn”²⁷ Ergo, multilingualism is a difficult issue to solve for a national judge, and a reference for preliminary ruling is usually the only way to deal with linguistic differences, in order to understand whether or not they represent an issue.

2.3 Multilingualism as an Issue for EU Law and Policies

Next to the language regime of the EU institutions, bodies and agencies, the issue of multilingualism is also of relevance in relation to the law and policies of European integration. One important issue is the need to mediate possible conflicts between, on the one hand, Member States’ law concerning the use of languages and, on the other hand, the fundamental freedoms of movement of persons, goods, services and capital, and the prohibition of discrimination on grounds of nationality.

Three cases are of particular interest as illustrations of this issue. In its ruling of 1989 in Groener, the Court stated that “the EEC Treaty does not prohibit the adoption of a policy for the protection and promotion of a language of a Member State which is both the national language and the first official language. However, the implementation of such a policy must not encroach upon a fundamental freedom such as that of the free movement of workers. Therefore, the requirements deriving from measures intended to implement such a policy must not in any circumstances be disproportionate in relation to the aim pursued and the manner in which they are applied must not bring about discrimination against nationals of other Member States” and concluded that “it is not unreasonable to require [the teachers] to have some knowledge of the first national language”, even if it is not being used in daily life by most of the population, as is the case for the Irish.²⁸

Conversely, the issue of labelling, presentation and advertising of foodstuffs generated important case law already in the 1980s, on the basis of Art. 14 of Directive 79/112/EEC.²⁹ In Piageme, the ECJ held that the Community rules “preclude a national law from requiring the exclusive use of a specific language for the labelling of foodstuffs, without allowing for the possibility of using another language easily understood by purchasers or of ensuring that the purchaser is informed by other measures”.³⁰ This abundant case law pushed the Commission to issue an Interpretative Commission communication concerning the use of languages in the marketing of foodstuffs in the light of the judgment in the Peeters case. ³¹

There are also significant rulings in relation to the use of language and the freedom of movement of persons. Two cases relating to the free movement of persons are of particular relevance to the German language. In Angonese ³² an Italian citizen who was a German native speaker and resided in the province of Bolzano had been a candidate in a competition for a job at the Cassa di Risparmio. He did not possess the required certificate of bilingualism that was required in the Province. The Court held that “where an employer makes a person’s admission to a recruitment competition subject to a requirement to provide evidence of his linguistic knowledge exclusively by means of one particular diploma, such as the Certificate, issued only in one particular province of a Member State, that requirement constitutes discrimination on grounds of nationality contrary to Article 48 of the EC Treaty”. In Bickel and Franz,³³ a German and an Austrian citizen had problems with the police authorities and justice in the province of Bolzano. The Court held that “the exercise of the right to move and reside freely in another Member State is enhanced if the citizens of the Union are able to use a given language to communicate with the administrative and judicial authorities of a State on the same footing as its nationals. Consequently, persons such as Mr Bickel and Mr Franz, in exercising that right in another Member State, are in principle entitled, pursuant to Article 6 of the Treaty, to treatment no less favourable than that accorded to nationals of the host State so far as concerns the use of languages which are spoken there.” Albeit a criminal trial was at stake in that case, it is undisputed that the same reasoning applies to administrative proceedings involving European citizens in the German‐speaking Trentino‐Alto Adige, as well as European citizens of the French language in the Valle d’Aosta or tongue Slovenian in Friuli‐Venezia Giulia – to take just the case of bi‐ or multilingual Italian regions. The same reasoning applies obviously in any other bi‐ or multilingual region of an EU Member state where the relevant languages are EU official languages. It is clear that there is no obligation to allow the use of these languages beyond the boundaries of the regions in which there is a specific bi‐ or plurilingual status in national law, nor to allow the use of other EU languages that do not enjoy a particular status according to the relevant member State Law.

3 The Autonomy of EU Law Concepts in Relation to Concepts of Domestic Law

3.1 Interpretation of EU Law and the Diversity of National Legal Systems

The diversity of national legal systems of the Member States affects in a major way the interpretation of EU law, as well as its multilingual character. In the aforementioned CILFIT ruling, the Court of Justice pointed out in para 19 that “it must also be borne in mind, even where the different language versions are entirely in accord with one another, that community law uses terminology which is peculiar to it. Furthermore, it must be emphasized that legal concepts do not necessarily have the same meaning in community law and in the law of the various member states.” The CILFIT ruling notwithstanding, a substantial number of scholars – and also quite some number of judges – only too often make the mistake of assessing the interpretation of EU law given by the CJEU according to the parameters of their own national law, forgetting that these parameters change from one EU country to the other.

Even among countries that share the same language, there are important differences between legal systems. For example, the concepts of German law are often quite different from those of Austrian law; civil law was codified in different periods and with different tools: the relevant codifications came into force in 1794 for Prussia, in 1811 for Austria, and in 1901 for united Germany. Administrative procedure was already codified in 1925 in Austria, but only in 1975 in Germany; Austria has had a system of judicial review of the constitutionality of laws since 1920, Germany since 1949, and so on. As a result the intuitive approach to EU law is different for a lawyer who has been educated in Austria and for a lawyer who has been educated Germany, also because the Austrian tradition of hermeneutics have been marked by formalism much more than the German tradition. Similar considerations also apply to the countries that have French in common as the official national language; while also having in common the Civil Code of 1804, the civil law of Belgium and that of Luxembourg often differ from French civil law. In addition, there are also many differences between Belgium and the Netherlands who share the Dutch language; these two countries also shared the same Civil Code as France and Luxembourg until 1987, at which point the Netherlands replaced the Code Napoléon of 1804 with an entirely new one. Last but not least, it may be mentioned for example, that in France the law of contracts and torts of public authorities is generally subject to a specific discipline developed independently by administrative courts, while in Belgium civil courts apply the Civil Code to such matters.

The concepts used in EU law, while they often find their origin in the legal systems of the Member States have to be interpreted autonomously, since their original meaning is different from one Member State to another. A typical example is that of the concept of “subjective right”. Such a concept is in use in most of the EU Member States in order to identify a right belonging to a person (natural or legal). The scope of the concept is however quite different from one legal system to another and the term is hardly used in the UK and Ireland as the notion of “subjective right” has hardly any relevance in common law. In Italy and Germany, the concept of subjective right is crucial in order to determine which courts – ordinary courts or administrative courts – are competent to deal with a case; or what type of judicial review can be undertaken, and in order to know what rules and principles apply to a case. In France the concept of subjective right is not useful at all in order to establish the competence of ordinary or administrative courts, or the available remedies; and that concept has a very limited impact on the content of the applicable rules of law, differently from – again – German and Italian law. This being said the concept of “subjective rights” is certainly not identical in German and Italian law.

Differences in the understanding of concepts are reflected, for instance, in the famous Van Gend en Loos ruling,³⁴ as far as the determination of the rights guaranteed by the EEC Treaty is concerned. When reading the opinion of AG Karl Roemer ³⁵ – who had been a practicing attorney in Germany until 1953 and was AG of the ECJ from 1954 until 1973 – it appears that his reasoning derives from the notion of German subjective right, and that such a reasoning quite logically leads him to deny the existence of a right based upon Art. 12 of the EEC Treaty, which contained a stand still clause on customs duties in the Member States. On the contrary, in reading the ruling, it appears that the majority of the Court was in favour of a much broader conception of the notion of right, which almost coincides with that of a protected interest. There are also interesting differences between the language versions of the Van Gend en Loos ruling. For instance, the Italian translation uses the expression diritto soggettivo whereas the French version says droits individuels. The reason was that judge Trabucchi wanted to make it clear to Italian lawyers that the rights that were referred to in the ruling were not mere interessi leggittimi – that enjoyed a lower degree of protection in the Italian legal system as far as liability was concerned. Furthermore, the Dutch and German versions of the ruling easily show that the Court took over the vocabulary of the reference of the Amsterdam duties court, whereas for the French or Italian doctrine, the vocabulary of the ruling departs from that of the reference: it is therefore not astonishing that French and Italian speaking scholarship – and later on also the English speaking one – have seen far more judicial activism in the Van Gend en Loos ruling than the Dutch or German scholarship.³⁶

This being said Community law included from the outset – i. e. in the Treaties themselves – a series of concepts that were autonomous with respect to the legal system of each Member State. A typical example is that of the competition rules contained in the Treaties. These rules relating to private companies, now contained in Art. 101 and 102 TFEU, had no equivalent in 1957 in national law of five out of six founding states. Only Western Germany had adopted antitrust legislation at the request of the Allies. That legislation was clearly inspired by US American legislation that had already a long tradition that began with the Sherman Antitrust Act of 1890. Apart from Germany, there were at that time hardly any other European countries with similar legislation – even outside of the Community Member States; competition law consisted almost only in protecting traders against “unfair competition” with the prohibition of sale at a loss, and so on. The ECSC Treaty already contained competition rules that were also inspired by American antitrust law.³⁷ It is therefore logical that the ECSC High Authority and the Commission, as well as the Court of Justice, have developed a legal framework in an autonomous way with respect to the law of Member States, which turned out as a series of new concepts for almost everyone in Western Europe.

In the EEC Treaty itself, there is a quite striking example, namely that of service of general economic interest. The concept is contained in Art. 106.2 first sentence TFEU (unchanged since 1957) according to which: “Undertakings entrusted with the operation of services of general economic interest or having the character of a revenue‐producing monopoly shall be subject to the rules contained in the Treaties, in particular to the rules on competition, in so far as the application of such rules does not obstruct the performance, in law or in fact, of the particular tasks assigned to them”. French scholarship has for a long time criticized this wording, i. e. the use of the terms service d’intérêt (économique) général instead of service public (économique), unlike under Art. 93 TFEU in relation to transport policy (also unchanged since 1957), according to which “Aids shall be compatible with the Treaties if they meet the needs of coordination of transport or if they represent reimbursement for the discharge of certain obligations inherent in the concept of a public service”. The problem is that in France service public is a central concept of administrative law. In addition, the concept of service public économique et commercial has been clearly defined in the case law since 1921.³⁸ However, whereas this concept of public service was utmost clear to a French lawyer and commonly used in everyday language (albeit with a different and somewhat ideological connotation as opposed to its legal meaning), it was and remains unknown in the law of the majority of other Member States, even those such as Italy, which use the same expression of servizio pubblico. In the Dutch version of the Treaty it says “openbare dienst” with quotation marks within Art. 93 TFEU: that was the way the Dutch lawyers chose to indicate that a literal translation of public service does not match with Dutch law, as openbare dienst means really civil service, in French fonction publique as does by the way the German “öffentlicher Dienst”: the German version of the Treaties should probably also use inverted commas for Begriff des öffentlichen Dienstes or put it into the plural öffentliche Dienste in order to make it correspond to service public at least in daily language.

3.2 EU Law and the Diversity of Legal Families

With the growth in the number of Member States and in the extension of the material scope of EU law, differences between national legal systems that are relevant to the interpretation of EU law have increased, while in the same time there has been an increasing impact of EU law on the Member States’ legal systems, which led to their approximation in many fields.

The impact of enlargement on the issue of differences between national legal systems tends to be overestimated, both as regards the enlargement of 1973 with the accession of two common law countries, and that of 2004/2007 with the accession of ten countries that have had for forty years a “communist” regime. Enlargements have led to problems, yet I submit that those problems are not due to differences between legal systems of the Member States.

There is as widespread but mistaken idea according to which there would be a summa divisio between legal systems (even at global level) opposing the so‐called common law systems and civil law systems, and that this opposition would be a problem for the coherence of EU law. As is well known, the expression common law refers to the legal systems of Great Britain and Ireland, of the United States of America and Commonwealth countries – including Australia, Canada and New Zealand – as well as other former British colonies. Those legal systems have in common their origin, i. e. in the law of English Crown Courts before independence of the United States of America in 1776. The concept of civil law is usually being used with the Code Civil of 1804 in mind, in order to refer to the legal systems of continental European countries and their former colonies, countries in which the law has been to a large extent codified, especially since the beginning of the nineteenth century. For a long time already has the age‐old Anglo‐French opposition assumed the form of an ideological opposition between lawyers, as demonstrated by the famous statements of Dicey according to which droit administratif would be the contrary of the rule of law. Nowadays there is, a mythical view of the common law as the law best adapted to a free market economy, as demonstrated in the field of “law and economics”; there is also a no less mythical view of civil law as the law of State interventionism. Both views ignore the fact that the first “modern” codifications had been adopted significantly prior to the French Civil Code of 1804, as soon as 1687 in Denmark, 1734 in Sweden and 1794 in Prussia.

In a nutshell we can say that if it is true that there are some important differences between the legal systems of the common law family and legal systems of the civil law family, there is often more in common between a common law country and a civil law country – e. g. England and France – than between two countries of the same family – e. g. France and Germany – when it comes to certain concepts or ways of reasoning. A basic mistake that is often made is to think that the system of sources of law is dramatically different between common law countries, where law would be essentially jurisprudential, and civil law countries where it would be essentially based upon statutory law. This error is leading part of the scholarship to write that the importance of judge‐made law in the development of Community law would be due to the influence of common law thinking, disregarding the evidence, i. e. that the landmark decisions in van Gend en Loos (1963) and Costa v. ENEL (1964) were adopted about ten years before accession of Ireland and the UK to the European Communities, and furthermore in a time in which the prospects for UK accession had been frozen by de Gaulle’s refusal in 1962 to continue negotiation on the matter.

With regard to the legal systems of the countries of Central and Eastern Europe, it has to be remembered that previous legal traditions are probably far more important than the legal system of the communist period. The traditional influence of Austrian law in Czechoslovakia, Poland, Hungary and part of Yugoslavia (Slovenia, Croatia and Bosnia‐Herzegovina today) as well as that of German law in the Baltic countries (Estonia, Latvia and Lithuania) and in a large part of Poland, and even that of French law in Bulgaria and Romania, has been reinvigorated after the end of the communist regime in 1989 and has probably more importance in terms of legal concepts and methods than the fact that those countries have experienced a communist regime.

The administrative law systems of the Member States have been of particular relevance to the development of EU law since the beginning, since the drafting of the ECSC Treaty, the adoption of the first pieces of ECSC secondary rules and decisions, and since the first rulings of the Luxembourg court. The problems to be addressed by the Court and other Community institutions were very similar to those which Member State courts had to deal with when judging cases derived from the relations between public administration and private parties. Such a starting point is reflected in the wording of Art. 215 EEC, now Art. 340 TFEU: “In the case of non‐contractual liability, the Union shall, in accordance with the general principles common to the laws of the Member States, make good any damage caused by its institutions or by its servants in the performance of their duties.” This wording recalls that – beyond differences between systems, such as the fact that administrative courts adjudicated on State liability in some countries and civil courts in other – what matters is the fact that courts are faced with the same problems, and often come to the same solutions.

The first AG to the Community Court was Maurice Lagrange – soon joined by Karl Roemer. Lagrange was a member of the French Conseil d’Etat and had already drafted the ECSC Treaty Articles dedicated to Court proceedings. Lagrange had authored conclusions in more than 60 cases from 1953 to 1964, including in the leading case Costa v. ENEL in 1964. There is no doubt that he has brought to the Luxembourg Court the traditional pragmatism and traditional functionalist and teleological reasoning of the commissaires du gouvernement of the French Conseil d’Etat, thus compensating for the influence of legal tradition of civil courts and civil law scholarship of the Romano‐Germanic legal system. In this context it may be useful to recall for instance the judgment of 7 February 1947 in d’Aillières ³⁹ (Lagrange was a member of the Conseil d’Etat from 1923) where the French Supreme administrative court did not hesitate to contradict the wording of a statutory provision, by stating that it could have neither the purpose nor the effect to exclude the possibility of a remedy for annulment of an administrative decision, while the relevant statute, an ordonnance adopted by the provisional government after the liberation of France from German occupation had expressly provided that the decisions of committees having to adjudicate on acts of collaboration with occupying forces “shall not be subject to any appeal.” Typically many of the solutions developed in the second half of the twentieth century by the UK Courts – where there is no institutional separation between ordinary courts and administrative courts – are close to those made by the Conseil d’Etat. This influence is clearly due to a large extent to the regular meetings between the justices of the House of Lords and those of the French Conseil d’Etat, which have taken place since the 1970s. French administrative law had developed since the 19th century, not on the basis of the Civil Code but mainly by way of pragmatic judge‐made law, a situation that was familiar to the British courts.

By way of conclusion I submit, with all due respect, that differently from Latin until the end of the 17th century the time has not yet come where the English language might be considered as the lingua franca of EU law. If any, the German language is probably more appropriate for legal science, due to its properties as an agglutinative language, with the proviso that whereas it is usually not too difficult to translate from another language (like English, French, Italian or Spanish for instance) into German – the reverse is not true: German is a difficult language for foreigners and for legal translators. The study of EU multilingualism⁴⁰ should therefore be more developed and would certainly also benefit the consolidation of common European legal thinking.