Extraction and databases -- the ECJ rules

Earlier today the Court of Justice of the European Communities (ECJ) pronounced judgment in Case C‑304/07, Directmedia Publishing GmbH v Albert-Ludwigs-Universität Freiburg, a reference from the Bundesgerichtshof, Germany, for a preliminary ruling on database rights and the meaning of Article 7 of Directive 96/9 of the European Parliament and of the Council of 11 March 1996 on the legal protection of databases.

Right: Freiburg -- spires to inspire aspiring litigants

This decision has been delivered virtually four months to the day after Advocate General Sharpston published her Opinion (noted here by the IPKat), which is a couple of months earlier than usual. Well done, ECJ, for speeding things up.

The facts are as follows. Professor Dr Ulrich Knoop, of the German Department of the Albrecht-Ludwigs-Universität Freiburg, directed the ‘vocabulary of the classics’ project, which led to the publication of the Freiburg Anthology, a collection of verse from 1720 to 1933, as part of this project, Knoop drew up a list of verse titles that was published on the internet under the heading ‘The 1,100 most important poems in German literature between 1730 and 1900’. This list, which served as a basis for the Freiburg Anthology, set out, in order of the frequency with which the poem is mentioned, the author, title, opening line and year of publication for each poem.

What was the professor's methodology? From some 3,000 published anthologies of poetry, 14 were selected. That selection was supplemented by the bibliographic compilation of 50 German-language anthologies by Ms Anneliese Dühmert under the title ‘Who wrote that poem?’, which yielded some 20,000 poems. Poems which were listed in at least three anthologies or were mentioned on at least three occasions in Ms Dühmert’s bibliographic compilation qualified for inclusion in the list. As a precondition for that statistical analysis, the titles and opening lines of the poems were standardised and a list of all verse titles compiled. Finally, the poems were referenced to the works in which they were published and the year of their composition was established. This task, completed under Knoop's direction, took around two and a half years and cost the University EUR 34,900.

Directmedia sold a CD-ROM called ‘1,000 poems everyone should have’, which appeared in 2002. Of the poems on the CD-ROM, 876 were from 1720 to 1900 and, of these, 856 appeared in the list of verse titles generated by the ‘vocabulary of the classics’ project. In selecting the poems for inclusion on its CD-ROM, Directmedia used the list of verse titles from the ‘vocabulary of the classics’ project as a guide. It examined attentively the selection made by Knoop, omitting some of the poems listed and adding others of its own choice. Directmedia took the actual verse texts from its own digital resources. Professor Knoop and the University took the view that, in reproducing and distributing its CD-ROM, Directmedia had infringed the copyright of Professor Knoop as compiler of an anthology and the related right of the University as maker of a database. They sought an order requiring Directmedia to desist from reproducing and/or distributing the CD-ROM ‘1,000 poems everyone should have’, as well as damages and delivery up for destruction.

The Landgericht ruled in favour of Knoop and the University. Directmedia’s appeal to the Oberlandesgericht failed so it appealed on a point of law to the Bundesgerichtshof. In its initial judgment the Bundesgerichtshof dismissed the appeal against the judgment for copyright infringement granted in favour of Knoop. It then proceeded to consider the appeal against the database infringement judgment granted in favour of the University. After hearing the arguments of both sides the Bundesgerichtshof referred the following question to the Court:

"Can the transfer of data from a database protected in accordance with Article 7(1) of Directive 96/9/EC and their incorporation in a different database constitute an extraction within the meaning of Article 7(2)(a) of that directive even in the case where that transfer follows individual assessments resulting from consultation of the database, or does extraction within the meaning of that provision presuppose the (physical) copying of data?"

Elearnor Sharpston advised the ECJ to rule as follows:

"Extraction’ ... does not presuppose the (physical) copying of data;

– In order to constitute an ‘extraction’ within the meaning of Article 7(2)(a) ..., it is immaterial whether the transfer of data from a database protected in accordance with Article 7(1) of the directive and their incorporation in a different database takes place following individual assessments of the data after consulting the database".

The Court's ruling goes as follows:

"The transfer of material from a protected database to another database following an on‑screen consultation of the first database and an individual assessment of the material contained in that first database is capable of constituting an ‘extraction’, within the meaning of Article 7 of Directive 96/9 ... to the extent that – which it is for the referring court to ascertain – that operation amounts to the transfer of a substantial part, evaluated qualitatively or quantitatively, of the contents of the
protected database, or to transfers of insubstantial parts which, by their repeated or systematic nature, would have resulted in the reconstruction of a substantial part of those contents".

In other words, the answer is "yes", says the IPKat, but it's a long yes that had to be worked for rather than a short, sweet one. Now the case goes back to Germany for a final determination of the original dispute. Merpel says, I noticed this little bit in the court's ruling

"47 ... it is of little importance that the act of transfer in question is for the purpose of creating another database, whether in competition with the original database or not, and whether the same or a different size from the original, nor is it relevant that the act is part of an activity, whether commercial or not, other than the creation of a database .... Moreover ..., the transfer of all or a substantial part of the contents of a protected database to another medium, which would be necessary for the purposes of a simple on‑screen display of those contents, is of itself an act of extraction that the holder of the sui generis right may make subject to his authorisation.

48 In its reference for a preliminary ruling, the referring court draws attention to recital 38 in the preamble to Directive 96/9 [which] refers to the case of the contents of a database being ‘copied and rearranged electronically’, it could, in the referring court’s view, militate in favour of an interpretation of the concept of extraction which is limited to acts based on a process of copying by technical means.

49 However, as the Advocate General pointed out ..., the recital in question seeks to illustrate the particular risk for database makers of the increasing use of digital recording technology. It cannot be interpreted as reducing the scope of the acts subject to the protection of the sui generis right merely to acts of copying by technical means,...".

This is significant in terms of interpretational technique: if examples in the preambles to Directives were allowed to be employed as a means of limitative interpretation of the substantive provisions, it would then be necessary to provide lists of examples that were exhaustive, thus rather undermining the point of having examples -- which reflect the content of a category rather than define it -- in the first place.