The production process of EU directives and their transposition into national law - additional considerations reviewing "Observing Eurolects. Corpus analysis of linguistic variation in EU law"
by Laura Mori (ed.), Amsterdam/Philadelphia, John Benjamins Publishing Company 2019
24. November 2019
2. Some remarks on the transposition of EU directives into national law
3. Period and languages chosen for the Eurolect Observatory Project
4. Some remarks on the corpora of national transposition measures (corpus B)
5. Some remarks on legal matters
6. Some remarks on the corpora of EU directives (corpus A)
6.1. Corrigenda of EU directives
6.2. Three different kinds of EU directives: three different production processes
6.2.1. Directives of the European Parliament and the Council
6.2.2. Council directives
6.2.3. Commission directives
7.1. Directive structure: preamble, enacting terms and annexes
7.2. Drafting guidelines
8. Research results
9. Second phase of the Eurolect Observatory Project
The author would like to thank staff of the European Parliament’s Directorate for Legislative Acts, in particular the Dutch, English, French, German, Italian, Latvian, Maltese, Polish and Spanish teams, for many fruitful discussions. Thanks also to Council and Commission staff who took the time to answer my queries. All shortcomings remain the responsibility of the author.
< 1 >
The book under review presents the first results obtained in the framework of the ambitious Eurolect Observatory Project, set up in 2013 at the Università degli Studi Internazionali di Roma (UNINT) and coordinated by Laura Mori. In this first phase, 19 researchers from 11 European universities analysed corpora containing EU directives and national implementing measures in 11 languages (Dutch, Finnish, French, English, German, Greek, Italian, Latvian, Maltese, Polish and Spanish) with a view to testing and/or confirming “the hypothesis that there are EU legal varieties (“Eurolects”) which were born and developed within the linguistic dia-systems” of a specific Member State, and to highlighting “the differences between each Eurolect and its corresponding national legal variety, in order to provide elements that might be useful to the study of metalinguistic and translational issues”1. It was hoped that this would also provide reference data to stakeholders and interested parties such as “the EU language services, national and regional parliaments, governing bodies of autonomous regions” and “produce results that may help improve the quality of legal drafting, both nationally and supranationally”2. So the bar is set very high indeed. Here, I have to declare an interest as a practitioner: I worked, until July 2018, in the European Parliament’s directorate housing the lawyer-linguists and at the time, may well have been involved in the German version of one or other EU directive which is the subject of the corpus studied.
< 2 >
After a useful preface by Ingemar Strandvik, quality manager in the European Commission’s translation services - which provided operational support to the project - who is therefore very familiar with the challenges of EU multilingualism, the introductory chapter by Laura Mori presents the details of the project and the methodology followed, in particular the common research template covering the lexical, morphological, morphosyntactic, syntactic and textual level which the different teams used for their analyses. This is followed by a chapter on the construction of the corpora used as well as the query tools (Marco Stefano Tomatis). Chapters 3 to 13 were drafted by the different research teams and vary in length (from 14 pages on Netherlandic Dutch to 41 pages on Italian). All chapters, including the conclusion by Laura Mori, are drafted in English but contain examples in the language concerned.
< 3 >
Two corpora are used for the quantitative and qualitative analyses: corpus A which comprises 660 EU directives which were published between 1999 and 2008 and which were still in force in 2014 (when that corpus was downloaded from the EUR-Lex website3) and corpus B which contains the corresponding national transposition measures. As EU directives are “binding, as to the result to be achieved, upon each Member State to which it is addressed, but [leaves] to the national authorities the choice of form and methods” (Article 288 of the Treaty on the Functioning of the European Union), the corpora are well chosen.
2. Some remarks on the transposition of EU directives into national law
< 4 >
Member States’ approach to transposition varies. The UK government recommends4: to “always use copy-out for transposition where it is available, except where doing so would adversely affect UK interests”5 and to depart from copy-out and use language that differs from the wording of the Directive only where necessary in order to clarify its meaning for legal or domestic policy reasons6. The German Federal Ministry of the Interior refers in its recommendations7 to the difficulty of complying with the logic and terminology of EU directives in the German legal system and concludes that copying the directive by using more or less the same words is often not an appropriate transposition method8. In Belgium (not covered in this first phase of the Eurolect Observatory project), the Council of State9 notes that copy-out may be the best method if a directive is very detailed and precise, but does not work in every case, as the transposition measure must fit the domestic legal system. If the wording of the EU directive is different from the wording normally used in national law, the drafter, according to the Belgian Council of State, has two options: either use the wording from national law or stick to the wording of the EU directive and adapt the terminology in all other acts of domestic law where the same concept is used.10 Copy-out minimises the risk of incomplete or incorrect transposition which could lead to infringement proceedings against the transposing Member State by the European Commission. This fact, as well as time pressure and workload, has contributed more recently to an increased use of copy-out even in Member States which traditionally tried to transpose EU directives in such a way as to ensure that the provisions would fit naturally and seamlessly in the national legal framework. For obvious reasons, national transposition measures using copy-out are less interesting for the purpose of the project.
3. Period and languages chosen for the Eurolect Observatory Project
< 5 >
The choice of the period studied (1999-2008) corresponds to EU legislation adopted before the entry into force of the Lisbon treaty on 1 December 2009 which brought about a considerable extension of the codecision procedure. With the Lisbon treaty, codecision became the default legislative procedure, a fact which is reflected in its new denomination: the ordinary legislative procedure. The Lisbon treaty also introduced a clearer demarcation between legislative acts on the one hand and binding non-legislative acts (which, confusingly, may also take the form of regulations, directives or decisions) as well as delegated and implementing acts on the other. However, the choice of period also had to take into account that the deadline for transposition of directives is usually two years or more (not to mention the fact of frequent late transposition by Member States), so the balance between a period providing for a sufficiently large corpus and sufficiently up-to-date data was always going to be a difficult one. Annalisa Sandrelli - author of the chapter on English - rightly points out that it would be necessary to refine the research diachronically in order to ascertain whether there have been any significant changes since the 2004 enlargement (p. 89). A diachronical analysis might well reveal such changes, including those which are not linked directly to the 2004 enlargement, and the subsequent replacement of French by English as lingua franca, which is today almost exclusively used for intra- and interinstitutional negotiations on legislative proposals.
< 6 >
The languages analysed can be split into different groups. Dutch, French, German and Italian are languages of the founding Member States and French, in particular, had a decisive influence on the look of EU legislation. That said, both French and Dutch have covered more than one Member State right from the start. French-speaking drafters, translators and, lawyer-linguists have been recruited from Belgium, France and Luxembourg; Dutch speakers may come from the Netherlands or Belgium, each with their own cultural, linguistic and legal-linguistic baggage. German, on the other hand, has had to grapple with the Austrian variety only after the accession of Austria in 1995, when German Eurolect was already well established. Expanding the analysis of French, German, English and Greek to the national transposition measures of those Member States not covered in the first phase (Belgium, Austria, Cyprus, Ireland) is one of the many plans to develop the Eurolect Observatory project (p. 4).
< 7 >
The UK joined the EU in 1973, Greece in 1981 and Spain in 1986, so that in all three cases, it can be assumed that their EU legal language was well established before the period chosen for the analysis. Not so for Finland, which joined the EU in 1995, so that EU legal language during the first couple of years of the research period may still be somewhat fluid. Finnish, together with Maltese, is also the only non-Indo-European language analysed.
< 8 >
The group of the three remaining languages - Latvian, Maltese and Polish - only became official EU languages on 1 May 2004, so any directive adopted before that date was part of the so-called acquis. The translation of the acquis into the new official languages is the responsibility of the acceding Member State, even if translations are revised - more or less thoroughly, depending on workload and time available - by the lawyer-linguists of the EU institutions and published by the EU’s Publication Office. Łucia Biel, in her chapter on Polish, distinguishes three phases of development for the Polish Eurolect: the early pre-accession phase corresponding to the translation of the acquis; the transition phase starting with accession when units of Polish translators and Polish lawyer-linguists were formally established in the EU institutions; and finally the later, post-accession phase by which time Polish Eurolect had achieved terminological and stylistic stabilisation (p. 296 et seq.). According to Biel, that third phase started from 2008 to 2010 which means that it is not covered by the research period. Biel also highlights the adverse conditions of the finalisation of the acquis when 4,500-page translation corrections were sent to the institutions six months after accession (p. 296). This is symptomatic of the great pressure under which the acquis was translated, and which led to an increasing lack of thorough revision the closer the accession date came. All the 2004 accession languages faced similar problems. Overall, by June 2004, finalisation of the acquis had reached just 72 % on average, with great differences between the new languages concerned11. A Slovene lawyer involved in the revision of the translated acquis told me that in the last two years before accession, EU texts were classified into three categories: (1) texts which had to be fully revised; (2) texts which were considered suitable for a light revision; and (3) texts which were not revised at all. As Biel points out, these circumstances were to have an impact on later stages of the research period and were overcome only in the third phase as from 2008. For the 2004 accession languages, it could therefore be interesting to split the data into two periods - before and after accession - and to compare with later periods starting in 2009.
< 9 >
Among the 2004 accession languages, Maltese is a special case. At the request of the Maltese government, and due to the lack of Maltese translators, a derogation was granted by Council Regulation (EC) No 930/2004 of 1 May 200412 for three years during which the EU institutions were not bound by the obligation to publish all acts in Maltese. In autumn 2006, it was decided that acts which had not been published in Maltese by 30 April 2007 would have to be published in that language by 31 December 200813.
< 10 >
In his chapter on corpus construction14, Tomatis notes that two directives15 were not available in Maltese or Polish (p. 28) – in fact, these two directives are not available in any of the 2004 accession languages. According to the Legal Service of the European Commission16, amending Directive 2001/53/EC was already obsolete before the accession date as it was superseded by Commission Directive 2002/75/EC of 2 September 2002 which amended Council Directive 96/98/EC on marine equipment on the same points as Directive 2001/53/EC. Directive 2002/75/EC was translated into the 2004 languages and as it remained in force in 2014, it was probably included in corpus A in all languages. On the other hand, amending Directive 2001/97/EC was not included in the list of acts to be translated for the 2004 accession – it was implicitly repealed by Directive 2005/60/EC17 which entered into force on 15.12.2005 with a transposition deadline of 15.12.200718. As amending directive 2001/97/EC was no longer in force in 2014, it should not have been incorporated in any language version of corpus A.
< 11 >
So the reason for missing translations has nothing to do with the subject matter – Tomatis seems to think that the number of directives may vary as “they indicate the policies that different EU Member States should put into force” and thus, that Poland and Malta were “probably” not obliged to transpose Directives 2001/53/EC and 2001/97/EC (p. 28). Not so. All EU directives have to be published in all official languages, including the official language of any Member State that has opted out of the subject matter concerned. There is, for example, a Danish version of Directive (EU) 2017/137119 , although Denmark did not take part in the adoption of that directive and “is not bound by it or subject to its application”20.
As the quality of corpus A and corpus B is crucial for the research results, a closer look at their construction, as described by Tomatis, is appropriate.
4. Some remarks on the corpora of national transposition measures (corpus B)
< 12 >
The construction of corpus B presented a major challenge not only from the technical point of view and was certainly far from straightforward. The national transposition measures were identified on the basis of the information available in EUR-Lex and then downloaded from the relevant national government websites. However, the information in EUR-Lex is based purely on the information transmitted by the Member States, and there is a reason why the Publications Office which manages the site makes an explicit disclaimer21. As Tomatis points out, the team constructing the corpus was confronted with many cases where the reference to the national transposition measure was incorrectly reported or missing (p. 34). Vilelmini Sosini, Katia Lidia Kermanidis and Sotirios Livas, in the chapter on Greek, note that only 564 directives - instead of 660 - “were transposed into Greek law” (p. 176): they overrate the quality of the information contained in EUR-Lex. It is far more likely that Greece did not communicate the necessary information in many cases even if it transposed the directives concerned. The European Commission regularly launches infringement proceedings against Member States precisely for failure to notify national transposition measures22. In July 2019 for example, the Commission database on infringement proceedings registers 849 active cases just for non-communication of national transposition measures23 (and 1,964 active cases for other reasons, such as late, incorrect or incomplete transposition).
< 13 >
In the period covered, some 35 EU directives correspond to codifications, i.e. cases where a new directive replaces and repeals a previous directive (and subsequent amending acts) without any substantive modification. Such directives do not contain an article on transposition. Directive 2006/1/EC24 for example replaces Directive 84/647/EEC25 which had been amended once in 1999. Spain and Greece indicated that Directive 2006/1/EC did not require any national transposition measures - obviously because Directive 84/647/EEC and the 1990 amendment thereto (Directive 90/398/EEC26) had already been covered by national transposition measures. Italy, France, Finland, Malta and the UK did not transmit any information whatsoever, whereas three transposition acts are cited for Poland and two for Latvia, with four of these acts having been adopted prior to 1 May 2004, i.e. in the framework of the translation of the acquis. Germany on the other hand indicated a text from 1998 which does not contain any reference to any of the EU directives concerned so that the link remains rather unclear.
It is not clear whether codified directives were taken into account or not for the construction of the corpora, nor whether they were excluded or not from corpus A.
< 14 >
For another example for the difficulties concerning the construction of corpus B: France informed the Commission - at least according to the information available in EUR-Lex - that no specific transposition measures were needed concerning Directive 2008/99/EC on the protection of the environment through criminal law27, presumably because the content of that directive was already covered by French national law. Would that directive then have to be excluded from corpus A as there were no matching transposition measures in corpus B?
< 15 >
The construction of the Maltese corpus B was particularly difficult. According to Article 74 of the Maltese constitution, every Maltese law is published in both English and Maltese, “save as otherwise provided by Parliament”. However, the reality is more complex, as “certain main legislation allows for subsidiary legislation made under it to be published in the English language only”28. This is surely the reason why some (subsidiary) law is available only in English, and not the supposed “recent adoption of an independent [Maltese] alphabetic system”, as speculated by Tomatis (p. 35). In fact, the current standardised Maltese orthography goes back to 192429. It is the creation of a standardised Unicode classification for the special characters used in Maltese which is indeed much more recent, and a Maltese keyboard was launched only in 2005 together with Windows XP Service Pack II. As late as 2000, the special characters were rendered as a graphic design on a case-by-case basis and could therefore not be easily reproduced by other users30.
< 16 >
Even where Maltese transposition measures were available on the Maltese Government’s website, these were available only in PDF-format and conversion from PDF into a usable format proved to be problematic: Out of 449 texts initially downloaded, only 139 - corresponding to the transposition of 77 EU directives - could be used for quantitative analyses (p. 35). As Sergio Portelli and Sandro Caruana point out in their chapter on Maltese, their research results and conclusions should therefore be interpreted with caution (p. 292). Nonetheless, I found their chapter particularly interesting as it shines a light on a non-Indo-European language with a Semitic stratum and Italian and English adstrata.
5. Some remarks on legal matters
< 17 >
Where EU matters are concerned, the text would have benefitted from a review by a specialist in EU law in order to avoid a number of inaccuracies. The Council and the European Council are two separate bodies31 and must not be confused.
It is not true that “any European legislative instrument comprises 24 language versions” (Patin and Megale, p. 126). Irish – one of the 24 official languages – has a special status. There has always been an Irish version of the treaties, but the first secondary legislation (regulations of the European Parliament and the Council) in Irish was not published until 2007, and the derogation for Irish is being phased out very gradually so that by January 2021 all newly adopted EU legislative acts, including Commission regulations, directives and decisions, should be available in Irish32 unless the Council decides to revise that date in the meantime. The translation of the acquis into Irish remains a tricky issue due to the lack of (human) resources. It is also rather misleading to state that “all EU-related information is processed into Latvian” (Dilāns, p. 245) as for pragmatic reasons a number of Commission reports, communications and staff working documents are generally published only in English, French and German and some are published only in English.
European Commissioners do not “serve to the Parliament” (De Sutter and De Bock, p. 49), even if the Commission, as a body, is accountable to the European Parliament. The principle of subsidiarity33 seems not have been clearly understood by Patin and Megale when they write that the ”supranational normative apparatus, according to the principle of subsidiarity, is meant to be obligatory” (p. 145). Directives are binding by definition, whether they concern areas of exclusive competence (where the principle of subsidiarity is irrelevant) or not. And the Maltese acronym KE in Directive 2008/8/KE34 does not stand for Kunsill (Council), as Portelli and Caruana believe (p. 273), but for Komunità Ewropea (European Community).
6. Some remarks on the corpora of EU directives (corpus A)
< 18 >
The downloading of texts for corpus A in a machine-readable format was much more straightforward as it could be done directly from EUR-Lex. However, the simple direct download meant that some choices were made - possibly without further analysis of the potential impact - which should be spelt out. The case of codified directives has already been mentioned, but there is also the issue of corrigenda.
6.1. Corrigenda of EU directives
< 19 >
From the examples given about problems or errors encountered, it seems clear that corrigenda of directives were not taken into account35. Indeed, the publication of corrigenda raises a difficult methodological issue: corrigenda may be published any time, sometimes even a number of years after initial publication, to correct for example a linguistic discrepancy between language versions, a translation error in one or more language versions or an erroneous reference in all language versions of a text. However, the full text of a directive may also be re-published in the form of a corrigendum concerning all language versions within a very short period after initial publication. In those cases, it would have been more appropriate to examine the text of the corrigendum. In the period 1999-2008 this applies to a total of 19 directives, 17 of which were initially published in April 2004, just before enlargement, with corrigenda published mostly in June 200436. What was the reason for the high number of full text corrigenda published in June 2004? From 1 May 2004, all texts published in the Official Journal had to appear in all the official languages, including those of the new Member States (with the exception of Malta, see < 9 > above). For those 17 directives agreed by the legislator shortly before enlargement, this would have meant a lengthy publication delay as the text would have to be translated from scratch, no translation memory based on the initial Commission proposal being available for those new languages. This would have considerably delayed the signature of the acts, their entry into force and thus their transposition deadlines. In order to avoid such a situation, it was decided to proceed with a rushed publication of a version which was not fully revised and finalised in all official languages, and to re-publish the full text as a corrigendum once the texts were duly finalised. It is the text of the corrigendum which is the authentic version and prevails over the text that was the subject of the initial publication. In the case of full-text corrigenda, I would therefore argue that the text of the corrigenda rather than the version of the text published initially should have been used, whereas it was acceptable for practical reasons to exclude corrigenda concerning only specific languages. Given the size of corpus, this may not be statistically relevant, but it should at least be explained.
6.2. Three different kinds of EU directives: three different production processes
< 20 >
By downloading all the directives without distinguishing their author, corpus A mixes three different categories of text which are characterised by very different procedures and by a different set of actors. Based on the overall statistics from EUR-Lex, it can be assumed that some 50 % of the directives downloaded are directives adopted by the Commission alone, 15 % are directives adopted by the Council alone and 35 % are directives adopted by both the European Parliament and the Council37. Although Mori refers to the multi-layered and reiterative ‘drafting-translation-revision’ process which may trigger backtracking, i.e. changes to the source text or master copy, most authors do not seem to have a very clear understanding of the complex production processes, the role of lawyer-linguists and the impact of national experts in particular on acts adopted jointly by the European Parliament and the Council. Once a Commission proposal is transmitted to the European Parliament and the Council under the codecision procedure, Parliament’s lawyer linguists – and to an increasing degree Council’s lawyer-linguists as well – follow the legislative process from beginning to end, advise negotiators on alternative or compromise wording throughout the process and are not limited to verifying only, “at the very end of the translation pipeline” (p. 245), compliance with EU drafting rules, as stated by Gatis Dilāns in the chapter on Latvian. In the chapter on Greek, Sosoni, Kermanidis and Livas report back from interviews with EU staff that “it is often lawyer-linguists and not translators who have the final word in cases of translated legislation” (p. 180). In order to fully understand this remark of interviewees to the Greek research team, it is necessary to understand the respective production processes.
6.2.1. Directives of the European Parliament and the Council
< 21 >
Directives which are adopted jointly by the European Parliament and the Council acting as co-legislators under the codecision procedure (now called the “ordinary legislative procedure”) are subject to a complex revision and finalisation process. The legislative procedure starts with the Commission proposal38, which at the level of the Commission is already subject to inter-service consultation and thus input from different stakeholders within the Commission and which is translated by the Commission’s translation units. The final act is the result of negotiations between the European Parliament and the Council in which the Commission acts as a mediator and a technical expert. The Parliament’s negotiating mandates are prepared by its parliamentary committees, where the initial amendments to the Commission proposal may be drafted in any official language, although since the 2004 enlargement compromises between political groups are usually negotiated in English. Council’s negotiating mandate originates in Council working groups - composed of delegates of the Member States - and is established by Coreper ('Committee of the Permanent Representatives of the Governments of the Member States to the European Union) or Council. Compromises agreed between the Parliament and Council may lead to significant changes in the final text agreed as compared to the initial proposal. The agreed text is then translated by the translation units of the Parliament or the Council and fully revised and finalised by Parliament’s and Council’s lawyer-linguists, including a lawyer-linguist in each institution who is an English native speaker, in a two- or three-steps procedure, where the text is sent from one institution to the other, and the wording - whether in the source text or any other language version - can be changed only if the lawyer-linguists and other stakeholders of both institutions, as well as the Commission, agree. Ambiguities detected in that process due to difficulties in rendering the text in other languages may lead to redrafting of the English source text, if all negotiating parties involved agree. The agreed source text is also discussed at an internal meeting at the Council with the national experts. It is not only the drafters in the institutions who are exposed to a language contact scenario, but also these national experts. They generally comprise officials of the national ministry concerned who also sit in the Council working group where the initial proposal and draft mandate with changes to that proposal were discussed, and who may also be involved later in the preparation of the national transposition measures. All language versions are transmitted to the Member States for comments before that meeting. The role of these national experts – who are exposed to a language contact scenario – cannot be underestimated: they intervene with very concrete suggestions for changes to wording, in particular at lexical level. Requests for changes coming from the Member States are taken on board in the final version of the text unless the Parliament’s or the Council’s lawyer-linguist object to the change suggested on behalf of their respective stakeholders.
< 22 >
The language used in EU directives adopted jointly by the European Parliament and the Council is therefore influenced by direct input from the Member States, by the same national officials who participated in the legislative process from the initial discussions in the Council working groups to the final Council meeting on the compromise text agreed with the Parliament’s negotiators. In such EU legislation which undergoes the complex production and finalisation process with direct involvement of the Member States, the boundary between Eurolect and national language is somewhat blurred at least as far as the lexical level is concerned, apart from obvious Europeism such as the frequent use of “Member States”, the addressees of EU directives. As the Lisbon treaty provided for a near doubling of the percentage of acts to be adopted under that procedure, it could be very interesting to look at future developments in this area. According to Mori, an extension of the time-span analysed is planned in future phases of the project, in particular for the 2004 enlargement languages Latvian, Maltese and Polish as well as for Finnish (p. 4 et seq.)39.
At the end of the production process – and that applies also to Council directives and Commission directives, which are discussed below – the Publication Office’s proofreaders intervene. They ensure, for example, that the use of brackets in numbered subdivisions, quotation marks, and other typesetting elements correspond to the typographical standards as defined in the Interinstitutional Style Guide40.
6.2.2. Council directives
< 23 >
Directives adopted by the Council as sole legislator undergo a somewhat lighter finalisation process: the lawyer-linguists of the European Parliament do not, as a rule, intervene in such finalisation. Member States may send comments to the Council’s lawyer-linguists, but their intervention is not a formal part of the finalisation schedule, as it is for the finalisation of codecision acts. Member States may request that Council organise a meeting with national experts in order to discuss the text before it is signed and published, but this does not happen very often. It is in particular in the politically sensitive area of tax law that Member States have asked for such a meeting.
6.2.3. Commission directives
< 24 >
Commission directives, on the other hand, are fully finalised within the European Commission, without intervention of Parliament’s or Council’s lawyer-linguists. Commission lawyer-linguists will normally revise the (mostly English) draft as part of the inter-service consultation, but they are not necessarily involved in the finalisation of all language versions (this is at least the current situation where only corrigenda to autonomous Commission acts are revised by the lawyer-linguists in all language versions).
< 25 >
However, these autonomous Commission acts pose another conceptual issue as such Commission directives are not legislative acts sensu stricto, but, rather, so-called comitology acts41. In such acts, the European Commission, assisted by committees of experts from the Member States, acts on the basis of powers delegated to it by the legislator - either Council alone or the European Parliament and Council acting jointly as co-legislators -, generally in order to adopt (technical) implementing measures, to adopt or amend (often very technical) annexes to the basic act or to adapt an act to technical progress42. Where corpus B contains also regulatory measures of implementation, as detailed by Lorenzo Blini (p. 334) concerning acts issued by the Spanish government (“Real Decreto”), or by Fabio Proia (p. 148) for German concerning not only acts approved by the Bundestag, but also Verordnungen issued by the Federal government, this is unproblematic. However, it could be argued that Commission directives should have been excluded from corpus A in those cases where acts of a similar level were excluded in corpus B. For the case of France for example, Stéphane Patin and Fabrizio Megale excluded “décrets” adopted by the French government without involvement of the French parliament (p. 125), whereas corpus A still included Commission directives; the same applies to the chapter on Finnish where Mikhail Mikhailov and Aino Piel consider only acts of Parliament in corpus B (p. 96).
< 26 >
That said, it should be noted that Member States are free to choose the instrument by which to transpose directives. Directive 2000/14/EC43 for example, was adopted jointly by the European Parliament and the Council under the codecision procedure and transposed into German law by a Gesetz (adopted by Parliament) and a Verordnung (issued by government) and therefore taken into account in the German corpora A and B. In France, that directive was transposed by an arrêté du ministre de l'aménagement du territoire et de l'environnement, du ministre délégué à l'industrie, aux petites et moyennes entreprises, au commerce, à l'artisanat et à la consommation et de la secrétaire d'Etat au budget, and thus, the latter was not taken into account in the French corpus B, whereas the directive was kept in corpus A and used for concrete examples (p. 128, figure 3 in Patin and Megale).
7.1. Directive structure: preamble, enacting terms and annexes
< 27 >
As Mori states in the introduction, the analysis focused on the normative part of the directives, excluding, as a rule, the preamble and annexes (p. 17). Preambles in EU legislation contain citations and recitals. As indicated in the Interinstitutional Agreement between the European Parliament; the Council and the Commission of 22 December 1998 on common guidelines for the quality of drafting of Community legislation44 “[t]he purpose of the recitals is to set out concise reasons for the chief provisions of the enacting terms, without reproducing or paraphrasing them. They shall not contain normative provisions or political exhortations”45. Any EU legislative act must contain such recitals, but they certainly do not contain prescriptive rules, which is why non-mandatory language is used (e.g. “should” instead of “shall” in English, “sollte” instead of the present tense in German etc.). In national law, depending on the system used by the Member State concerned, such reasons may be found only in travaux préparatoires published separately. In those cases, it would seem entirely appropriate to exclude the preamble also from corpus A in order to avoid comparisons between texts of a different nature, in particular for any quantitative analysis. According to Mori, data from the preambles could be taken into account only for lexical description (i.e. not a quantitative analysis) or for analysing intra-textual differences, i.e. comparisons between the recitals on the one hand and the enacting parts on the other (p. 17), and if data from preambles were used by a research team, this should of course be indicated.
< 28 >
Annexes to EU directives are characterised by differing linguistic and textual features. The presentation of annexes varies considerably and is far less standardised than the presentation of the articles in the main body of the act. The Joint Practical Guide46 contains only a rather vague instruction: “Although there are no specific rules governing the presentation of annexes, they must nonetheless have a uniform structure and be subdivided in such a way that the content is as clear as possible, in spite of its technical nature”47. It is likely that further digitalisation in text production, such as the introduction of XML editors, may eventually lead to some degree of standardised presentation in the future. In the current state of affairs, Mori’s conclusion that annexes should be studied separately is a wise decision.
< 29 >
That said, although at first glance it might seem quite straightforward to exclude annexes from corpus A, the enacting articles of an amending act may well contain text which concerns annexes, but which is not easily identifiable and is highly likely to have been included in corpus A. An example may be found in Commission Directive 2005/51/EC48 which does not contain any annex, but where Article 1 contains a new wording of the first sentence of point 1(a) of Annex XX to Directive 2004/17/EC49.
7.2. Drafting guidelines
< 30 >
Most authors introduce the chapter on their language with some very welcome more general information about legislation, legislative drafting and legal language at national level as well as existing guidelines, if any, and the transposition process. Some information on EU guidelines is also included which inevitably overlaps partially with the preface by Ingemar Strandvik. However, not all the guidelines or recommendations have the same standing or are consulted by all the actors involved in the – often lengthy – legislative process. Internal guidelines and recommendations prepared in the Commission translation units, such as The English Style Guide: A handbook for authors and translators in the European Commission50, available on the Commission’s website, are “intended primarily for English-language authors and translators, both in-house and freelance, working for the European Commission”51. Although they may be used by other services and institutions as well, interinstitutional guidelines such as the Interinstitutional Style Guide published by the Publications Office of the European Union are more widely spread, and as far as legislative acts adopted under the codecision procedure are concerned, lawyer-linguists are more likely to consult the Joint Practical Guide (drafted by the Legal Services of the European Parliament, the Council and the Commission) and the Joint Handbook for the presentation and drafting of acts subject to the ordinary legislative procedure, drafted by the lawyer-linguists of those three institutions (which, for those acts, replaces the Council’s Manual of Precedents). Guidelines drafted at national level may of course be consulted by EU staff and may provide useful inspiration, but not more than that. On the other hand, national drafters dealing with national transposition measures are obviously expected to respect their national guidelines, which prevail, as far as domestic law is concerned, over EU guidelines.
< 31 >
De Sutter and De Bock in the chapter on Dutch refer to two interinstitutional guidelines, to a Commission document and to the Aanwijzingen voor de regelgeving, published on a Dutch government website (p. 51). All of these are useful, but the guidelines and “Writing tips” are – as is clear from their title – not binding. On the other hand, the Aanwijzingen voor de regelgeving contain instructions for lawmakers in the Netherlands, and only for them – Belgian lawmakers preparing transposition measures in Dutch are not bound by them. De Sutter and De Bock then present seven rather specific recommendations (p. 51) which seem to come from internal documents from the Commission’s translation unit, but are given without any context, and which may have led the research team down the wrong path in their lexical analysis as will be shown later in this review.
< 32 >
So yes, there is a hierarchy of guidelines, but that hierarchy varies in accordance with the role of the stakeholder concerned. In his chapter on German, Proia (p. 150) refers to an interesting document describing such a hierarchy of guidelines (Hierarchie der Leitfäden für die Abfassung von Rechtstexten52) – this seems to be an internal Commission document and would not necessarily be known or consulted in the other institutions.
8. Research results
< 33 >
The aim of this review is not an in-depth discussion of every aspect found by the research teams, but rather to look into some particularly interesting, instructive or debatable issues.
A preliminary remark on sentence length which is referred to by some research teams (e.g. for Dutch [p. 59], Finnish [p. 109], Greek [p. 184]) as one of the elements linked to readability: in EU acts, the subdivision of a text into sentences is normally the same in all language versions. There are practical reasons for this rule, in particular to ensure that references to a specific sentence are identical in all language versions. The Joint Handbook for the Presentation and Drafting of Acts subject to the Ordinary Legislative Procedure states: “When this principle causes difficulties in a language, the sentences should be broken down into smaller sentence-like units, separated by semicolons.”53 This may also explain the frequent use of the semicolon in EU directives noted by Sosoni, Kermanidis and Livas for Greek (p. 181).
< 34 >
Contact-induced features, in particular loanwords and calques, are analysed by most research teams54. Contact-induced features appear in all languages, but at different levels (lexis, morphology, syntax) and to varying degrees. Mikhailov and Piehl note that Finnish is characterised by the more limited use of loanwords as compared to some other languages, as adapting foreign words to Finnish phonology and grammar is not straightforward (p. 98). However, as far as Eurolects generally are concerned, Blini rightly sounds a note of caution: he argues that “it is likely that the role of language contact is overestimated as far as Eurolects are concerned […] and underestimated” as far as national legal Spanish is concerned (p. 362). Proia also notes that “in some domains, English non-adapted loanwords have established themselves in many national languages and are now part of their specialised lexis” (p. 156). This is certainly true in particular in the financial and banking sector. For German, Proia finds that Eurolect “shows a stronger attitude to accept and include [loanwords] in the word-formation process […] in comparison to national drafters” (p. 165). A spot check of the word “Recycling” and compounds in Directives 2000/53/EC55 and 2006/66/EC56 and the corresponding German transposition measures confirms his conclusion. However, recent practical experience has shown that nowadays it is often the national experts who ask for the use of English terms in the German version of EU texts (e.g. Hotspot57, Back-to-back-Transaktion58, Front-Office59 …). This seems to be a trend in other languages as well, as German, Hungarian, Latvian and Spanish lawyer-linguists of the European Parliament have told me, and considering the role of national experts in the legislative process as discussed above, this is not really a surprise.
< 35 >
I was particularly interested in Proia’s discussion of “Clearingstelle” (Clearing house) (p. 156) which is used in EU texts mostly in dealing with the financial sector60. In fact, the term is currently used quite widely in Germany for a number of advisory and/or dispute settlement bodies in different areas61. A full text search in German legislation62 shows 15 results, from a protocol to the 1990 Unification Treaty between the Federal Republic of Germany and the German Democratic Republic63 to more recent texts such as the 2015 Kraft-Wärme-Kopplungsgesetz (combined heat and power generation) or the 2017 Erneuerbare-Energien-Gesetz (renewable energy). This does not invalidate Proia’s research results (p. 155) – that the German legislator preferred to use other terms when transposing a specific EU directive in order to fit the transposition measure to the logic of the existing German system. But it shows, in my view, that an extension of the analysis to a wider corpus of national law as well as to more recent texts could deliver interesting results in future phases of the project.
< 36 >
Blini for Spanish finds a lower lexical diversity in EU directives and concludes that Spanish national transposition measures might use synonyms more frequently (p. 337). He points rightly to the Joint Practical Guide which recommends that “the use of synonyms and different expressions to convey the same idea should be avoided”64. That principle also applies in national legislation65 as the use of a single term to express a specific content ensures legal certainty (“same form, same content”). The overarching principle of legislative texts is - or should be - clarity and consistency. The Austrian drafting guidelines66 put it very nicely:
“Allgemeine Regeln für den Sprachstil (z.B. die Unterlassung von Wortwiederholungen) sollten bei der Formulierung von Rechtsvorschriften nicht überbewertet werden. Jedenfalls muss der Eindeutigkeit und Übersichtlichkeit der Norm der Vorrang vor der Ästhetik des Textes eingeräumt werden.”67
In 1901, the eminent Swiss jurist Eugène Huber wrote:
“nous avons désignés toujours par les mêmes termes les notions qui se répètent [...]. L’application des lois existantes a surabondamment démontré dans quelle perplexité une simple divergence de texte, due à l’inadvertance du législateur, peut plonger le juge.”68
If national transposition measures indeed make a lesser use of repetitions and use other mechanisms such as pro-forms and ellipsis instead (p. 233, Mori on Italian), this would be a matter for further research. Paraphrases, synonymy, hyponymy or hyperonymy on the other hand would not avoid the potential interpretative problems pointed out. Mori, in her chapter on Italian, notes a similar phenomenon in national legislative Italian “in compliance with the rhetorical principle of variatio” which should have no place in legislative language. She adds that the Italian legislator seems to prefer anaphoric pronouns instead of repetition even though this may cause ambiguity when the referent is not clearly recognisable (p. 233). This is one of the factors quoted by Mori in her conclusion that Italian transposition measures are potentially less accessible to the Italian citizen than the Italian version of the corresponding EU directives (p. 238).
< 37 >
Legal certainty is – or should be – the aim of any legislative drafter. That said, it is worth noting that EU law is negotiated law (negotiated between Member States and, in the case of codecision, between the European Parliament and the Council). Ambiguity is sometimes the prize paid by the negotiators in order to achieve a political compromise. The insertion of phrasemes such as “where/as appropriate”, “wherever practical”, ”where possible” is often due to such political compromise which, in directives, provides the limit that Member States must respect when filling the gap opened by such vagueness. Therefore, I would expect that these and similar expressions are more frequent in EU directives than in national transposition measures. Something to analyse in future research?
< 38 >
Blini notes that as far as Spain is concerned, “EU legal language is considered exclusively on the terminological level; even amongst jurists” (p. 334). In the codecision procedure, the predominant focus of national experts on terminology points in the same direction in all languages. The research results are thus of particular interest where they go beyond the purely lexical level and look into other linguistic features.
< 39 >
Proia refers to the fluctuating use of genitive vs “von + dative” for certain prepositions and prepositional phrases which “may possibly point to a growing tendency in contemporary German to prefer the dative over the genitive” (p. 159). Proia finds inter alia a slight predominance for the use of im Sinne von instead of “im Sinne + genitive” and concludes that EU drafters seem to have ignored the recommendation contained in the Commission’s Übersetzungshandbuch69 according to which the use of “von + dative” should be avoided in prepositional phrases introducing references to legal sources (p. 161). Again, it is worth noting that the Übersetzungshandbuch is a purely internal document of the Commission’s German translation unit and remains unknown and unused by Parliament’s and Council’s lawyer-linguists. However, the issue of the decreasing use of the genitive has received wide popular interest in Germany and is taken up regularly by newspapers, following Bastian Sick’s 2004 bestseller Der Dativ ist dem Genitiv sein Tod, and empirical evidence is always most welcome. The prepositional phrase “im Sinne” is particularly interesting as it is frequently used in legislative and legal drafting (in the latter often abbreviated as i.S.v. or even iSv and less often as i.S.d. for the use with genitive). Is there a pattern which characterises the use of the genitive or the von-construction? Random checks of EU directives do not show a fully coherent picture, but the genitive is always used in expressions such as “im Sinne dieser Richtlinie/dieses Absatzes”, “im Sinne der Richtlinie/Verordnung/Empfehlung [number]”. There seems to be a slight preference for the construction with von followed directly by nouns without article such as “im Sinne von Artikel 3 Absätze 4 und 5” where the word Absätze is used in the nominative case. But the picture is indeed far from clear, and that seems to be the case in German domestic law as well. I have looked at a randomly chosen recent German law – the KiQuTG of 19 December 201870 which uses both the genitive and the von-construction in close proximity in § 1:
“Kindertagesbetreuung im Sinne dieses Gesetzes umfasst die Förderung von Kindern in Tageseinrichtungen und in der Kindestagespflege im Sinne des § 22 Absatz 1 Satz 1 und 2 des Achten Buches Sozialgesetzbuch bis zum Schuleintritt. Maßnahmen nach § 2 dieses Gesetzes sind Maßnahmen, die frühestens ab dem 1. Januar 2019 begonnen werden und […] Maßnahmen im Sinne von § 22 Absatz 4 des Achten Buches Sozialgesetzgebung sind […]“.
This is of course anecdotal evidence and it could be interesting indeed to see the results of a diachronic research in more recent and larger corpora.
Incidentally, the last example illustrates a German usage which is not discussed by Proia: the use of singular or plural in references. Whereas the German KiQuTG refers to § 22 Absatz 1 Satz 1 und 1, or, a bit further in text, to § 90 Absatz 3 und 4, any EU act would use Sätze 1 und 2 and Absätze 3 und 4.
< 40 >
Most research teams have looked in particular at deontic modality which characterises EU legislation. In her chapter on English, Annalisa Sandrelli discusses in particular the use of “shall” to express an obligation which is still used in the enacting terms of any EU act but has disappeared in most (but not all) English-speaking countries (p. 79 et seqq.). Parliament’s lawyer-linguists71 confirm that this issue has been discussed repeatedly. However, changing current practice in English which is used as lingua franca at EU level, could be a rather delicate operation: it would require that the relevant European Parliament, Council and Commission services agree to such a general and far reaching change; that all guidelines, standard texts, electronic templates etc. be changed accordingly; and that all drafters be made aware accordingly – difficult to implement when the legislative machinery keeps running more or less on an ongoing basis! That said, small changes have been introduced over time. For example, in the introductory sentences of amending acts, the declarative use of the present tense is recommended since 2012 (e.g. Article 5 is replaced by the following instead of shall be replaced).
< 41 >
When discussing Biel’s chapter on Polish with lawyer-linguists of the European Parliament, they were surprised by the high frequency of “musi” (“must”) in corpus A (p. 320), as the present tense is normally used in the enacting terms72. Random spot checks show that musi is often (but not always) used where the English version has must73. Directives provide for obligations on the addressees, i.e. the Member States, but it would be rather clumsy and uneconomical to start each article with the full formula “Member States shall ensure that [other actor + action]”. As EU regulations are directly applicable and provide for obligations on the individuals and entities addressed, it would be interesting to see whether must/musi is used less frequently in the enacting terms of regulations (and extending the research to a corpus consisting of EU regulations is planned for a future phase of the Eurolect Observatory project).
< 42 >
In their chapter on French, Stéphane Patin and Fabrizio Megale conclude that the future tense is overrepresented as compared to the present tense (p. 136) although the latter should be used to express obligations74. That may be so, but I remain to be convinced. It is true that in the examples shown in figure 10 (p. 136), the present tense could have been used in most cases. However, a closer look at these examples reveals that 4 out of the 14 examples shown are from French law and not EU legislation75. Figure 10 thus mixes examples from corpus A and corpus B without this being specified, which at the very least casts some rather serious doubts on the validity of the examples and the conclusions drawn by the research team.
< 43 >
In the same chapter, table 3 on concordance extracts of deontic modality (p. 134) also contains five examples with verbs in the future tense: two examples from annexes, one from a preamble, and the remaining two from enacting terms of an EU directive. This is also not specified; on the contrary, Patin and Megale indicate explicitly in their section 2 (Research aim and methodology) that, in accordance with the common research protocol, neither preambles (i.e. citations and recitals) nor annexes were taken into account (p. 123; cf. also < 27 >). I was therefore somewhat surprised to see examples taken from preambles and annexes as in table 3 (p. 134). That table also contains 10 examples for “il est, par conséquent, nécessaire” - all coming from recitals and not from enacting terms. Three out of the four examples for “il faut que” in the same tableare taken from annexes. So they were taken into account after all?
< 44 >
On the other hand, the examples shown in figure 3 (concordance’s extracts of hapaxes in corpus A) (p. 128) seem indeed to be limited to the enacting terms of EU directives. At least some of these rather technical cases, however, are misleading due to exclusion of annexes. Débroussailleuses, pelleteuses, chargeuses etc. all occur again in the annex, not surprisingly, as the enacting terms contain an explicit reference to the annex76. Articles 12 and 13 of Directive 2000/14/EC contain long lists of equipment subject to noise limits, or to noise marking only, with further definitions describing such equipment in the annex.
< 45 >
Incidentally, the French transposition measures77 follow exactly the same presentation, so what do these examples prove, if anything? If the French transposition measure - an administrative arrêté - was excluded from corpus B, the question is rather whether Directive 2000/14/EC should not have been excluded from corpus A as well, as it could not be checked against the national transposition measure.
< 46 >
In other examples of hapaxes (p. 128, figure 3), the term used (agenti di cambio, parts de capital non fongible, yachts) is explicitly excluded from the scope of the directives concerned and therefore appears only once in each directive. The Italian term “agenti di cambio” is actually used in all language versions of Directive 2004/39/EC78 as it refers to a term of Italian law79 (and therefore does not appear in any national transposition measure other than Italian). It is interesting to note that EUR-Lex indicates 15 (sic!) French transposition measures for that directive - two lois, three ordonnances, three décrets and seven arrêtés. This is a good illustration of the difficulties of constructing a reliable corpus B.
On the purely lexical level, some of the results must be taken with caution or are based on incorrect assumptions.
< 47 >
In their chapter on Netherlandic Dutch, De Sutter and De Bock analyse a set of lexical preferences for general concepts, such as STRANGER, JOB, SCOPE, SOCIETY etc. (p. 54 et seqq.) However, some concepts and the lexical preferences linked to such a concept are problematic. I will look at three examples which illustrate different kinds of problems – the concepts SOCIETY, SCOPE and INHABITANT.
SOCIETY is discussed on the basis of a juxtaposition of maatschappij and samenleving (p. 55). However, contrary to samenleving, maatschappij can also stand for “company” (business structure). The statistical juxtaposition would be reliable only if each instance where maatschappij = “company” has been excluded. Was it? The reader is not told.
< 48 >
De Sutter and De Bock discuss the term “SCOPE” (toepassingsgebied / werkingssfeer) and conclude that Dutch transposition measures prefer werkingssfeer thus going “against the clear writing guidelines” (p. 55). An interesting example indeed, but the conclusion is wrong. The authors are presumably referring to an internal Commission document which recommends that “scope” be translated as “toepassingsgebied”. This purely internal (!) recommendation is of no interest whatsoever for the Dutch legislator when transposing an EU directive into the domestic law of the Netherlands. The specialised legal drafter in the Netherlands80 takes account of the Aanwijzingen voor de regelgeving which refers to “werkingssfeer”, but also to “toepassingsbereik”. That term has not been looked at by the research team.
< 49 >
So why do EU directives then use “toepassingsgebied” and not “werkingssfeer”? The 1998 Interinstitutional Agreement on common guidelines for the quality of drafting of Community legislation81 and the Joint Practical Guide based on those guidelines use “toepassingsgebied” which, in the interest of consistency, is thus the term to be used on EU level. Moreover, that term is also understood and used in Belgium. We do not know whether the translator or the legal-linguistic reviser of the Dutch version of the 1998 Interinstitutional Agreement was Dutch or Belgian82, but the wording had to take into account both varieties: Dutch as used in the Netherlands as well as that used in Belgium where “werkingssfeer” is not commonly used.
< 50 >
Belgian transposition measures were excluded in this first phase of the project. Although De Sutter and De Bock refer to lexical and grammatical differences between Belgium and the Netherlands, in general, (p. 50) this example shows that it is impossible to completely disregard the issue even when limiting the discussion to Netherlandic Dutch.
< 51 >
Concerning the concept of INHABITANT, corpus A “has a preference for the more neutral onderdaan while the implementation laws corpus prefer the” – according to De Sutter and De Bock – “somewhat more old-fashioned ingezetene” (p. 55). However, this is comparing apples and oranges: onderdaan refers to a person’s country of nationality; ingezetene refers to a person’s country of residence. These are two different legal concepts: Council Directive 2003/109/EC83 for example concerns “the status of third-country nationals who are long-term residents” (Richtlĳn 2003/109/EG betreffende de status van langdurig ingezeten onderdanen van derde landen). Or in other terms: I have a German passport, so I am a German onderdaan, but I am living in Brussels, so I am ingezetene van België. Ingezetene may be administrative language (used for example in the “verblijfskaart voor langdurig ingezetene” delivered to long-term non-EU residents in Belgium), but that does not mean that it is old-fashioned.
< 52 >
The term “resident” is also discussed by Mori for Italian: she compares soggiornante or soggionarnte di lungo periodo to straniero/stranieri which is less frequently used in corpus A because, she opines, “… the latter could be perceived more in conflict with the pivotal idea of European citizenship” (p. 216). It is true that straniero is defined under current Italian law as covering non-EU citizens and stateless persons, but the term is ambiguous in non-legal language (e.g. in reports from Istat, the Italian National Institute of Statistics, on tourism in Italy, it stands for non-Italians84). In corpus A, straniero is used for example when referring to the law relating to aliens. But not every straniero is a soggiornante within the meaning of the directives analysed.
< 53 >
The juxtaposition of these two terms reveals a different type of problem: Straniero/stranieri appears in 11 EU directives from 1999 to 2008, soggiornante only in two directives, but in both directives it is a key concept which is defined and then re-used repeatedly throughout the text85. In Council Directive 2003/86/EC86 on the right to family reunification, Article 2(c) defines “soggiornante” - corresponding in the English version to “sponsor” - in the following terms: "soggiornante": il cittadino di un paese terzo legalmente soggiornante in uno Stato membro che chiede o i cui familiari chiedono il ricongiungimento familiare87; it is used a total of 33 times in the act. Council Directive 2003/109/EC88 determines the terms for conferring and withdrawing long-term resident status granted by a Member State in relation to third country nationals legally residing in its territory, and the rights pertaining thereto. “Soggiornanti di lungo periodo” is a concept defined in Article 2(b)89 and used a total of 103 times in the text, including 82 times in the enacting part. Straniero/stranieri on the other hand appears, as we have seen, in a higher number of directives, but within the text of those acts, it occurs usually only once or twice, as for example in Directive 2001/40/EC90 which refers in its Article 3 to the national rules on the entry or residence of aliens (normative nazionali relative all'ingresso o al soggiorno degli stranieri).
< 54 >
As soggiornante has an exceptionally high frequency in only two EU directives, the statistical results are skewed in favour of that term. Normalisation of the raw frequency to distribution per 1 million was done in order to allow for comparisons independently of the differing sizes of corpus A and corpus B, but does not address the problem described here.
< 55 >
The juxtaposition of Italian figlio/figli and minore/minori (p. 216) raises similar issues as these are two different legal concepts: figlio refers to a relationship, minore to age. Council Directive 2003/86/EC on the right to family reunification for example refers to “minor children, including adopted children, as well as the adult unmarried children”91. On the other hand, Council Directive 2005/85/EC92 on minimum standards on procedures in Member States for granting and withdrawing refugee status, refers inter alia to unaccompanied minors93 in a context in which the use of the word “figlio” would not made have sense. A qualitative analysis would have to look much more closely into the wider context of each usage in order to check whether the concepts have been used consistently in corpus A and corpus B. This may not have been possible within the timeframe and with the resources available for the first phase of the project.
Still, there is ample food for thought be found in each chapter.
9. Second phase of the Eurolect Observatory Project
< 56 >
The second phase of the Eurolect Observatory project is already underway and is just as ambitious. The Hungarian language will be included, and the corpora extended to include national legislation unrelated to EU law (corpus C, already available for Polish), national varieties of languages that are official languages in more than one Member State or other countries, such as Dutch (Belgium, Netherlands), English (UK, Ireland), French (Belgium, France, Luxembourg, Switzerland) German (Germany, Austria, Switzerland), Greek (Greece, Cyprus) (corpus D), more recent data for newer Member States where EU legal language has been significantly affected by translation of the acquis, such as Finnish; Greek, Latvian, Maltese and Polish (corpus E), and other sources of EU law such a regulations or primary law (corpus F) (p. 4 et seq.)94. One research area deals with gender in legislative languages and the first results concerning five languages, published in 201995, will be the subject of a separate review in these pages.
< 57 >
May this review contribute to clarify some methodological issues and legal issues linked to the complex legislative procedures of the European Union and thus help avoiding similar pitfalls in the future of the project which deals with a difficult, but fascinating area at the crossroads of law and linguistics.
1 See First phase – Objectives, on the UNINT website of the project https://www.unint.eu/en/
research/research-projects/33-page/490-eurolect-observatory-project.html. (Access 9.11.2019).
3 The EUR-Lex website, managed by the EU Publications Office, provides access to all EU law, to preparatory documents, to Official Journal of the European Union, to EU case-law etc. https://eur-lex.
4 HM Government. Transposition guidance: How to implement European Directives effectively. 2013, last updated February 2018. https://assets.publishing.service.gov.uk/government/uploads/
system/uploads/attachment_data/file/682752/eu-transposition-guidance.pdf (Access 9.11.2019).
5 Ibid., p. 3.
6 Ibid., point 2.24 at p. 11.
7 Bundesministerium des Inneren, Handbuch zur Vorbereitung von Rechts- und Verwaltungsvorschriften, 2012. https://www.verwaltung-innovativ.de/DE/Gesetzgebung/Projekt_eGesetzgebung/Handbuecher_Arbeitshilfen_Leitfaeden/Hb_vorbereitung_rechts_u_verwaltungsvorschriften/hb_vorbereitung_rechtsvorschriften_node.html (Access 9.11.2019).
8 Ibid., point 262: „Besonders bei umfangreichen bzw. innovativen Richtlinienwerken der EU tritt immer wieder die Frage auf, wie die deutsche Rechtsordnung deren Systematik und Terminologie gerecht werden kann. Eine bloße, mehr oder weniger wortgleiche Übernahme von EU-Richtlinien ist dabei häufig kein geeignetes Mittel der Umsetzung.“
9 Conseil d’État/Raad van State, Principes de technique législative/Beginselen van de wetgevingstechniek, 2008. http://www.raadvst-consetat.be/?page=technique_legislative&lang=fr. (Access 9.11.2019).
10 Ibid., point 186.
11 Commission européenne, Direction Générale de la Traduction, La traduction à la Commission: 1958-2010. Études sur la traduction et le multilinguisme 2/2009. 2014, at p. 53. https://op.europa.eu/en/publication-detail/-/publication/b691a6cf-7fe7-476d-87a5-8584250000c4. (Access 9.11.2019).
12 Council Regulation (EC) No 930/2004 of 1 May 2004 on temporary derogation measures relating to the drafting in Maltese of the acts of the institutions of the European Union.
13 Council Regulation (EC) No 1738/2006 of 23 November 2006 amending Regulation (EC) No 930/2004 on temporary derogation measures relating to the drafting in Maltese of the acts of the institutions of the European Union.
14 Tomatis refers to the corpus collection in 8 languages (Dutch, English, German; French; Italian; Maltese, Polish and Spanish) which was done centrally. The Finnish, Greek and Latvian corpora were compiled locally by the language units concerned which had joined the project somewhat later.
15 Commission Directive 2001/53/EC of 10 July 2001 amending Council Directive 96/98/EC on marine equipment; and Directive 2001/97/EC of the European Parliament and of the Council of 4 December 2001 amending Council Directive 91/308/EEC on prevention of the use of the financial system for the purpose of money laundering.
16 Email communication to author on 13 May 2019.
17 Directive 2005/60/EC of the European Parliament and of the Council of 26 October 2005 on the prevention of the use of the financial system for the purpose of money laundering and terrorist financing.
18 Interestingly enough, Directive 2001/97/EC was published in Romanian and Bulgarian for the 2007 accession. Estonia, Hungary and Slovakia had translated that directive as well, but as it was not listed as part of the 2004 acquis, those translations were not published in the corresponding Special editions. Despite the explanations of the Commission’s Legal Service, there is a lack of coherence in the treatment of that directive.
19 Directive (EU) 2017/1371 of the European Parliament and of the Council of 5 July 2017 on the fight against fraud to the Union's financial interests by means of criminal law.
20 See recital 38 of that directive. The details on the areas of the Danish opt-out are reflected in Protocol No 22 annexed to the Treaties and were agreed by the European Council meeting in Edinburgh in 1992 after the referendum on the Maastricht Treaty had been rejected in Denmark.
21 “A reference to national transposition measures does not necessarily mean that these measures are either comprehensive or in conformity. The national provisions are shown as communicated by the member states.” and “The member states bear sole responsibility for all information on this site provided by them on the transposition of EU law into national law.”
22 To take an example: in 2017, the Commission launched such a procedure against 17 Member States concerning the transposition of Directive 2014/66/EU.
23 Database consulted on 30.7.2019. See http://ec.europa.eu/atwork/applying-eu-law/infringements-proceedings/infringement_decisions/?lang_code=en.
24 Directive 2006/1/EC of the European Parliament and of the Council of 18 January 2006 on the use of vehicles hired without drivers for the carriage of goods by road.
25 Council Directive 84/647/EEC of 19 December 1984 on the use of vehicles hired without drivers for the carriage of goods by road.
26 Council Directive 90/398/EEC of 24 July 1990 amending Directive 84/647/EEC on the use of vehicles hired without drivers for the carriage of goods by road.
27 Directive 2008/99/EC of the European Parliament and of the Council of 19 November 2008 on the protection of the environment through criminal law.
28 Bruno VANNI, Ensuring the Quality of Drafting of Legislation in a Multilingual Context: Legislation Process in Malta. Presentation of 20.9.2013 at the European Commission. As examples, he quotes Article 40(1) of the Maltese Product Safety Act and Article 17(3) of the Civil Aviation Act. See https://ec.europa.eu/dgs/legal_service/seminars_en.htm. (Access 9.11.2019).
29 See Ray Fabri, Maltese, in: Langues et littératures modernes, Revue belge de Philologie et d’Histoire, 2010, 88-3, p. 791.
30 Personal communication to author by a Maltese graphic designer and IT expert working in the European Parliament.
31 Since the Lisbon treaty, the European Council has the status of an EU institution. Before that, it designated the so-called summits, i.e. the regular meetings of the heads of state and governments of the EU Member States.
32 For details, see Council Regulation (EC) No 920/2005 of 13 June 2005 amending Regulation No 1 of 15 April 1958 determining the language to be used by the European Economic Community and Regulation No 1 of 15 April 1958 determining the language to be used by the European Atomic Energy Community and introducing temporary derogation measures from those Regulations, Council Regulation (EU) No 1257/2010 of 20 December 2010 extending the temporary derogation measures from Regulation No 1 of 15 April 1958 determining the languages to be used by the European Economic Community and Regulation No 1 of 15 April 1958 determining the languages to be used by the European Atomic Energy Community introduced by Regulation (EC) No 920/200 and Council Regulation (EU, Euratom) 2015/2264 of 3 December 2015 extending and phasing out the temporary derogation measures from Regulation No 1 of 15 April 1958 determining the languages to be used by the European Economic Community and Regulation No 1 of 15 April 1958 determining the languages to be used by the European Atomic Energy Community introduced by Regulation (EC) No 920/2005.
33 First subparagraph of Article 5(3) of the EU treaty: “Under the principle of subsidiarity, in areas which do not fall within its exclusive competence, the Union shall act only if and in so far as the objectives of the proposed action cannot be sufficiently achieved by the Member States, either at central level or at regional and local level, but can rather, by reason of the scale or effects of the proposed action, be better achieved at Union level.”
34 Council Directive 2008/8/EC of 12 February 2008 amending Directive 2006/112/EC as regards the place of supply of services.
35 Tomatis (p. 29) quotes the case of the Dutch version of Directive 2008/101/EC of the European Parliament and of the Council of 19 November 2008 amending Directive 2003/87/EC so as to include aviation activities in the scheme for greenhouse gas emission allowance trading within the Community, initially published as Directive 2009/101/EG on 13.1.2009. A corrigendum rectifying this numbering error was published in the Official Journal (OJ L 26) on 30.1.2009.
36 E.g. Directive 2004/38/EC of the European Parliament and of the Council on the right of citizens of the Union and their family members to move and reside freely within the territory of the Member States. The directive was signed into law on 29 April 2004, published in the Official Journal (OJ L158) of 30 April 2004 and entered into force on the day of its publication. A full-text corrigendum was published in the 11 pre-enlargement languages on 29 June 2004 (OJ L 229). A number of additional corrigenda concerning specific points in a certain language were published in 2005 and even later.
37 From 1999 to 2008, a total of 1,047 of EU directives was adopted (including those which were not in force any more in 2014) corresponding to 550 Commission directives, 152 Council directives and 345 European Parliament and Council directives (data calculated on the basis of statistics available on the EUR-Lex website). It can be assumed that the proportions for the 660 directives of corpus A are very similar.
38 Except for a limited number of cases where the Treaties attribute a right of initiative to another institution (e.g. the Court of Justice or the European Central Bank concerning their statute) or to a group of Member States.
39 See also Second phase on the UNINT website of the project https://www.unint.eu/en/research/
40 European Union, Interinstitutional Style Guide. Luxembourg: Publications Office of the European Union, 2011. Updated continuously at https://publications.europa.eu/code/en/en-000100.htm. (Access 15.11.2019).
41 Since the entry into force of the Lisbon treaty, the title of such acts indicates that it is either a delegated or an implementing act. However, a number of basic acts have not yet been amended to take into account the changes introduced by the Lisbon treaty and in those cases, the Commission may still adopt comitology acts under pre-Lisbon provisions.
42 E.g. Commission Directive 2006/130/EC of 11 December 2006 implementing Directive 2001/82/EC of the European Parliament and of the Council as regards the establishment of criteria for exempting certain veterinary medicinal products for food-producing animals from the requirement of a veterinary prescription; Commission Directive 1999/61/EC of 18 June 1999 amending the Annexes to Council Directives 79/373/EEC and 96/25/EC; Commission Directive 2004/73/EC of 29 April 2004 adapting to technical progress for the twenty-ninth time Council Directive 67/548/EEC on the approximation of the laws, regulations and administrative provisions relating to the classification, packaging and labelling of dangerous substances.
43 Directive 2000/14/EC of the European Parliament and of the Council of 8 May 2000 on the approximation of the laws of the Member States relating to the noise emission in the environment by equipment for use outdoors.
44 OJ C 73, 17.3.1999, p. 1–4. See in particular point 10.
45 See also Article 296 TFEU: Legal acts shall state the reasons on which they are based.
46 Joint Practical Guide of the European Parliament, the Council and the Commission for persons involved in the drafting of European Union legislation. Updated edition 2015. https://eur-lex.europa.eu/content/techleg/KB0213228ENN.pdf (Access 15.11.2019).
47 See point 22.5.
48 Commission Directive 2005/51/EC of 7 September 2005 amending Annex XX to Directive 2004/17/EC and Annex VIII to Directive 2004/18/EC of the European Parliament and the Council on public procurement.
49 Directive 2004/17/EC of the European Parliament and of the Council of 31 March 2004 coordinating the procurement procedures of entities operating in the water, energy, transport and postal services sectors.
50 8th edition: January 2016. Last updated: October 2019. https://ec.europa.eu/info/sites/info/files/
styleguide_english_dgt_en.pdf (Access 18.11.2019).
51 Ibid., p. 4.
52 https://ec.europa.eu/info/sites/info/files/using_style_guides_german_de.pdf (Access 15.11.2019).
53 Point D.4.3, at p. 50. https://www.consilium.europa.eu/media/32619/joint-handbook-en-january-2018_2018_01_25_def.pdf (Access 15.11.2019).
54 Such features are not discussed in detail in the chapter on Dutch.
55 Directive 2000/53/EC of the European Parliament and of the Council of 18 September 2000 on end-of life vehicles.
56 Directive 2006/66/EC of the European Parliament and of the Council of 6 September 2006 on batteries and accumulators and waste batteries and accumulators and repealing Directive 91/157/EEC.
57 E.g. in recital 138 of Directive (EU) 2018/1972 of the European Parliament and of the Council of 11 December 2018 establishing the European Electronic Communications Code.
58 Article 2(92) of Directive 2014/59/EU of the European Parliament and of the Council of 15 May 2014 establishing a framework for the recovery and resolution of credit institutions and investment firms and amending Council Directive 82/891/EEC, and Directives 2001/24/EC, 2002/47/EC, 2004/25/EC, 2005/56/EC, 2007/36/EC, 2011/35/EU, 2012/30/EU and 2013/36/EU, and Regulations (EU) No 1093/2010 and (EU) No 648/2012, of the European Parliament and of the Council.
59 E.g. recital 31 of Decision (EU) 2015/2240 of the European Parliament and of the Council of 25 November 2015 establishing a programme on interoperability solutions and common frameworks for European public administrations, businesses and citizens (ISA2 programme) as a means for modernising the public sector.
60 See definition in Article 2(e) of Directive 98/26/EC of the European Parliament and of the Council on settlement finality in payment and securities settlement systems. “’clearing house’ shall mean an entity responsible for the calculation of the net positions of institutions; a possible central counterparty and/or a possible settlement agent”.
61 E.g. Clearingstelle EEG (Erneuerbare-Energien-Gesetz); Clearingstelle der Deutschen Rentenversicherung Bund; Clearingstelle Mittelstand des Landes NRW bei IHK NRW.
63 Einigungsvertrag vom 31. August 1990, anliegendes Protokoll. Point 8 of the protocol reads: „Die Verwaltungshilfen des Bundes und der Länder beim Aufbau der Landesverwaltungen und bei der Durchführung bestimmter Fachaufgaben werden in einer Clearingstelle abgestimmt, die von Bund und Ländern gebildet wird.“
64 See point 1.4.1 of the Joint Practical Guide.
65 See for example point 74 of the Handbuch der Rechtsförmlichkeit issued by the German Federal Ministry of Justice: „Rechtsnormen sind verständlicher, wenn Wörter oder Wendungen für die gleichen Inhalte immer einheitlich verwendet werden“. http://hdr.bmj.de/page_b.1.html#an_53. (Access 9.11.2019).
66 Bundeskanzleramt, Handbuch der Rechtssetzungstechnik, 1990. https://www.justiz.gv.at/home/
2c94848a60c158380160e4e3747c0c33.de.html (Access 9.11.2019).
67 Ibid., p. 6.
68 Eugène Huber, Code civil suisse: exposé des motifs, Berne 1901. Quoted after Alexander Flückiger, Les racines historiques de la légistique en Suisse, Seminar at the European Commission on 19.10.2007, point 18.104.22.168 at p. 10. https://ec.europa.eu/dgs/legal_service/seminars_en.htm. (Access 9.11.2019).
69 European Commission, Deutsche Sprachabteilung; Übersetzungshandbuch. https://ec.europa.eu/
info/sites/info/files/german_style_guide_de.pdf (Access 15.11.2019).
70 Gesetz zur Weiterentwicklung der Qualität und zur Teilhabe in der Kindertagesbetreuung.
71 Personal communication with English lawyer-linguist on 18 March 2019.
72 See point 2.3.2 of the Joint Practical Guide: „W języku polskim używa się czasu teraźniejszego.”
73 For example in Article 18(3) of Directive 2008/57/EC of the European Parliament and of the Council of 17 June 2008 on the interoperability of the rail system within the Community: ”The notified body shall be responsible for compiling the technical file that has to accompany the ‘EC’ declaration of verification. This technical file must contain all the necessary documents relating to the characteristics of the subsystem […].” Polish version: “Jednostka notyfikowana odpowiada za zebranie dokumentacji technicznej, która musi towarzyszyć deklaracji weryfikacji WE. Dokumentacja techniczna musi zawierać wszelkie niezbędne dokumenty związane z cechami podsystemu […].”
74 See point 2.3.2 of the Joint Practical Guide: « Dans le dispositif des actes à caractère contraignant, les verbes s’emploient en français au présent de l’indicatif. »
75 A full text search in EUR-Lex did not give any results on the four following segments which can be traced to French law: « qui n'auront pas signé de convention à cette date verront à compter de 2006 leur dotation globale de fonctionnement »: Article 199-1 créé par la Loi n°2004-1485 du 30 décembre 2004; « faire figurer la part du contrat que le titulaire attribuera à des architectes, des concepteurs, des petites »: Article L6148-5-1 créé par Loi n°2004-806 du 9 août 2004; « apportera les participations qu'il détient ou viendra à détenir, au capital de la Banque de développement des petites »: Article 31 de la Loi n° 2004-1343 du 9 décembre 2004; « mentionnés au 2° du XLV, les mandats de ces membres viendront à échéance, respectivement, en juin 2012 »: Article 62 de la Loi n° 2009-879 du 21 juillet 2009.
76 E.g. Article 12 of Directive 2000/14/EC: Le niveau de puissance acoustique garanti des matériels énumérés ci-après ne peut dépasser le niveau de puissance acoustique admissible fixé dans le tableau suivant des valeurs limites: [...] - chargeuses (˂ 500 kW) - Définition: annexe I, point 37. Mesure: annexe III, partie B, point 37.
77 Arrêté du 18 mars 2002 relatif aux émissions sonores dans l'environnement des matériels destinés à être utilisés à l'extérieur des bâtiments.
78 Directive 2004/39/EC of the European Parliament and of the Council of 21 April 2004 on markets in financial instruments amending Council Directives 85/611/EEC and 93/6/EEC and Directive 2000/12/EC of the European Parliament and of the Council and repealing Council Directive 93/22/EEC.
79 Article 2(n) of Directive 2004/3/EC: „La présente directive ne s'applique pas [...] aux "agenti di cambio", dont les activités et les fonctions sont régies par l'article 201 du décret législatif italien n° 58 du 24 février 1998.
81 See guideline 13: “Where appropriate, an article shall be included at the beginning of the enacting terms to define the subject matter and scope of the act.” Dutch version: “In voorkomend geval worden in een artikel aan het begin van het regelgevend gedeelte doel en toepassingsgebied van het besluit omschreven.”
82 There are many Belgians both in Dutch and French translation units of the EU institutions as well as among the lawyer-linguists of the European Parliament, the Council and the Commission.
83 Council Directive 2003/109/EC of 25 November 2003 concerning the status of third-country nationals who are long-term residents.
84 E.g. Istat, Anno 2017, Movimento turistico in Italia, 27 November 2018, https://www.istat.it/
it/files//2018/11/report-movimento-turistico-anno-2017.pdf. (Access 9.11.2019).
85 I did not count Directives 2004/114/EC and 2006/123/EC as “soggiornanti” appears in those directives only once as a quotation of the title of directive 2003/109/EC.
86 Council Directive 2003/86/EC of 22 September 2003 on the right to family reunification.
87 "Sponsor" means a third country national residing lawfully in a Member State and applying or whose family members apply for family reunification to be joined with him/her”.
88 Direttiva 2003/109/CE del Consiglio, del 25 novembre 2003, relativa allo status dei cittadini di paesi terzi che siano soggiornanti di lungo periodo.
89 “(b) ‘long-term resident’ means any third-country national who has long-term resident status as provided for under Articles 4 to 7”.
90 Council Directive 2001/40/EC of 28 May 2001 on the mutual recognition of decisions on the expulsion of third country nationals.
91 E.g. in Article 4(3) of that Directive: “figli minori ..., anche adottati, di tali persone, come pure i figli adulti non coniugati”.
92 Council Directive 2005/85/EC of 1 December 2005 on minimum standards on procedures in Member States for granting and withdrawing refugee status.
93 Defined in Article 2(h) : ‘unaccompanied minor’ means a person below the age of 18 who arrives in the territory of the Member States unaccompanied by an adult responsible for him/her whether by law or by custom, and for as long as he/she is not effectively taken into the care of such a person; it includes a minor who is left unaccompanied after he/she has entered the territory of the Member States; Italian version: «minore non accompagnato»: una persona d’età inferiore ai diciotto anni che arrivi nel territorio degli Stati membri senza essere accompagnata da un adulto che ne sia responsabile per la legge o in base agli usi, fino a quando non sia effettivamente affidata a tale adulto, compreso il minore che venga abbandonato dopo essere entrato nel territorio degli Stati membri. Wording used in the Italian transposition measure: «minore non accompagnato»: il cittadino straniero di età inferiore agli anni diciotto che si trova, per qualsiasi causa, nel territorio nazionale, privo di assistenza e di rappresentanza legale”.
94 See also Second phase on the UNINT website of the project https://www.unint.eu/en/research/
95 Stefania Cavagnoli, Laura Mori (eds.), Gender in legislative languages. From EU to national law in English, French, German, Italian and Spanish. Berlin, Frank & Timme 2019.
Aanwijzingen voor de regelgeving (1992). Circulaire van de Minister-President van 18 november 1992. https://wetten.overheid.nl/BWBR0005730/2018-01-01 (Access 9.11.2019).
Bundeskanzleramt der Republik Österreich (1990). Handbuch der Rechtssetzungstechnik. https://www.justiz.gv.at/home/verfassungsdienst/legistik/e-recht-und-legistische-richtlinien/legistische-richtlinien~2c94848a60c158380160e4e3747c0c33.de.html (Access 9.11.2019).
Bundesministerium der Justiz (2008). Handbuch der Rechtsförmlichkeit. Empfehlungen des Bundesministeriums der Justiz für die rechtsförmliche Gestaltung von Gesetzen und Rechtsverordnungen nach § 42 Absatz 4 und § 62 Absatz 2 der Gemeinsamen Geschäftsordnung der Bundesministerien. 3. neu bearbeitete Auflage. http://hdr.bmj.de/
sitemap.html (Access 9.11.2019).
Bundesministerium der Justiz und für Verbraucherschutz (n.d.). Gesetze im Internet. https://www.gesetze-im-internet.de/ (Access 9.11.2019).
Bundesministerium des Inneren (2012). Handbuch zur Vorbereitung von Rechts- und Verwaltungsvorschriften. https://www.verwaltung-innovativ.de/DE/Gesetzgebung/Projekt_eGesetzgebung/Handbuecher_Arbeitshilfen_Leitfaeden/Hb_vorbereitung_rechts_u_verwaltungsvorschriften/hb_vorbereitung_rechtsvorschriften_node.html (Access 9.11.2019).
Cavagnoli, Stefania / Mori, Laura (eds.) (2019). Gender in legislative languages. From EU to national law in English, French, German, Italian and Spanish. Berlin: Frank & Timme.
COMMISSION EUROPÉENNE, Direction Générale de la Traduction, La traduction à la Commission (1958-2010). Études sur la traduction et le multilinguisme 2/2009. 2014, at p. 53. https://op.europa.eu/en/publication-detail/-/publication/b691a6cf-7fe7-476d-87a5-8584250000c4 (Access 9.11.2019).
Conseil d’État/Raad van State (2008). Principes de technique législative/Beginselen van de wetgevingstechniek. http://www.raadvst-consetat.be/?page=technique_legislative&lang=fr (Access 9.11.2019).
Eurolect Observatory Project (n.d.). Website https://www.unint.eu/en/research/research-projects/33-page/490-eurolect-observatory-project.html (Access 9.11.2019).
European Commission, Deutsche Sprachabteilung (n.d.). Übersetzungshandbuch. https://ec.europa.eu/info/sites/info/files/german_style_guide_de.pdf (Access 15.11.2019)
European Commission (2016). German resources: Using style guides — in what order? Hierarchie der Leitfäden für die Abfassung von Rechtstexten. Hinweis für die Übersetzer der Kommission. https://ec.europa.eu/info/files/german-resources-using-style-guides-what-order_en (Access 15.11.2019).
European Commission (2019). English Style Guide: A handbook for authors and translators in the European Commission, 8th edition. https://ec.europa.eu/info/sites/info/files/styleguide_english_dgt_en.pdf (Access 18.11.2019).
European Parliament, Council of the European Union and the Commission of the European Communities (1998). Interinstitutional Agreement of 22 December 1998 on common guidelines for the quality of drafting of Community legislation; https://eur-lex.europa.eu/legal-content/EN/TXT/?qid=1574953645655&uri=CELEX:31999Y0317(01) (Access 15.11.2019).
European Parliament, Council of the European Union, European Commission (2015). Joint Practical Guide of the European Parliament, the Council and the Commission for persons involved in the drafting of European Union legislation. https://eur-lex.
europa.eu/content/techleg/KB0213228ENN.pdf (Access 15.11.2019).
European Parliament, Council of the European Union, European Commission (2018). Joint Handbook for the presentation and drafting of acts subject to the ordinary legislative procedure. https://www.consilium.europa.eu/media/32619/joint-handbook-en-january-2018_2018_01_25_def.pdf (Access 15.11.2019).
European Union (2011). Interinstitutional Style Guide. Luxembourg: Publications Office of the European Union. Updated continuously at https://publications.europa.eu/code/en/en-000100.htm (Access 15.11.2019).
Fabri, Ray (2010). “Maltese”. Revue belge de Philologie et d’Histoire, Langues et littératures modernes, 88 (3): 791-816.
Flückiger, Alexander (2007). Les racines historiques de la légistique en Suisse, Seminar at the European Commission on 19.10.2007. https://ec.europa.eu/dgs/legal_service/seminars_en.htm (Access 9.11.2019).
HM Government (2013). Transposition guidance: How to implement European Directives effectively. Last updated February 2018. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/682752/eu-transposition-guidance.pdf (Access 9.11.2019).
Huber, Eugène (1901). Code civil suisse: exposé des motifs de l'avant-projet du Département fédéral de justice et police. Berne: Imprimerie Büchler & Co.
Sick, Bastian (2004). Der Dativ ist dem Genetiv sein Tod. Köln: Kiepenheuer & Witsch.
Vanni, Bruno (2013). Ensuring the Quality of Drafting of Legislation in a Multilingual Context: Legislation Process in Malta. Presentation at the European Commission 20.9.2013. https://ec.europa.eu/dgs/legal_service/seminars_en.htm (Access 9.11.2019).
Laura Mori (ed.).
Observing Eurolects. Corpus analysis of linguistic variation in EU law.
Amsterdam/Philadelphia, John Benjamins Publishing Company 2018. ISBN: 978-902720170-6.