Eleven words that axed a thousand trees

Rather late in the day, the ruling of the Court of Justice of the European Communities in Case C‑5/08, Infopaq International A/S v Danske Dagblades Forening, a reference for a preliminary ruling from the Højesteret (Denmark).

In short Infopaq, a media monitoring and analysis business, drew up summaries of selected articles from Danish daily newspapers and other periodicals. Those articles were selected on the basis of subject criteria agreed with their customers. Once the articles had been secured by data capture, they were sent to customers by email. DDF, a professional association of Danish daily newspaper publishers which assisted its members with copyright issues, discovered that Infopaq was scanning newspaper articles for commercial purposes without authorisation from the relevant rightholders. Taking the view that such consent was necessary for processing articles using the process in question, DDF complained to Infopaq about this procedure.

The data capture process was as follows:

1. The relevant publications were registered manually by Infopaq employees in an electronic registration database.

2. Once the spines were cut off the publications so that all the pages consisted of loose sheets, the publications were scanned. The section to be scanned was selected from the registration database before the publication was put into the scanner. Scanning allowed a TIFF (‘Tagged Image File Format’) file to be created for each page of the publication. After the scanning, the TIFF file was transferred to an OCR (‘Optical Character Recognition’) server.

3. The OCR server translated the TIFF file into data that could be processed digitally. During that process, the image of each letter was translated into a character code that told the computer what type of letter it was. That data was saved as a text file which could be understood by any text processing program. The OCR process was completed by deleting the TIFF file.

4. The text file was processed to find pre-defined search words. Each time a match for a search word was found, data was generated giving the publication, section and page number on which the match was found, together with a value expressed as a percentage between 0 and 100 indicating how far into the text it is to be found, in order to make it easier to read the article. Also in order to make it easier to find the search word when reading the article, the five words which come before and after the search word were captured (this being referred to in the judgment as an ‘extract of 11 words’). At the end of the process the text file was deleted.

5. At the end of the data capture process, a cover sheet was printed out in respect of all the pages where the relevant search word was found. The following is an example of the text of a cover sheet’.

Infopaq denied that this procedure required consent from the rightholders and brought an action claiming that DDF should be ordered to acknowledge that Infopaq was entitled in Denmark to apply this procedure without the consent of DDF or of its members. After the court dismissed that action, Infopaq appealed to the Højesteret, which referred 13 questions to the Court for a preliminary ruling.

It was conceded that consent from the rightholders was not required for (i) press monitoring activity and the writing of summaries consisting in manual reading of each publication, (ii) selection of the relevant articles on the basis of predetermined search words, or (iii) production of a manually prepared cover sheet for the summary writers, giving an identified search word in an article and its position in the newspaper. Nor was it disputed that genuinely independent summary writing per se is lawful and does not require consent from the rightholders, or that the data capture process described above involved two acts of reproduction: the creation of a TIFF file when the printed articles are scanned and the conversion of the TIFF file into a text file. It was also common ground that the procedure entailed the reproduction of parts of the scanned printed articles since the extract of 11 words was stored and those 11 words were printed out on paper. The Højesteret however sought guidance on the following issues:

‘(1) Can the storing and subsequent printing out of a text extract from an article in a daily newspaper, consisting of a search word and the five preceding and five subsequent words, be regarded as acts of reproduction which are protected (see Article 2 of [Directive 2001/29, the Directive on copyright and the information society]?

(2) Is the context in which temporary acts of reproduction take place relevant to whether they can be regarded as “transient” (see Article 5(1) of Directive 2001/29)?

(3) Can a temporary act of reproduction be regarded as “transient” where the reproduction is processed, for example, by the creation of a text file on the basis of an image file or by a search for text strings on the basis of a text file?

(4) Can a temporary act of reproduction be regarded as “transient” where part of the reproduction, consisting of one or more text extracts of 11 words, is stored?

(5) Can a temporary act of reproduction be regarded as “transient” where part of the reproduction, consisting of one or more text extracts of 11 words, is printed out?

(6) Is the stage of the technological process at which temporary acts of reproduction take place relevant to whether they constitute “an integral and essential part of a technological process” (see Article 5(1) of Directive 2001/29)?

(7) Can temporary acts of reproduction be an “integral and essential part of a technological process” if they consist of manual scanning of entire newspaper articles whereby the latter are transformed from a printed medium into a digital medium?

(8) Can temporary acts of reproduction constitute an “integral and essential part of a technological process” where they consist of printing out part of the reproduction, comprising one or more text extracts of 11 words?

(9) Does “lawful use” (see Article 5(1) of Directive 2001/29) include any form of use which does not require the rightholder’s consent?

(10) Does “lawful use” (see Article 5(1) of Directive 2001/29) include the scanning by a commercial business of entire newspaper articles, subsequent processing of the reproduction, and the storing and possible printing out of part of the reproduction, consisting of one or more text extracts of 11 words, for use in the business’s summary writing, even where the rightholder has not given consent to those acts?

(11) What criteria should be used to assess whether temporary acts of reproduction have “independent economic significance” (see Article 5(1) of Directive 2001/29) if the other conditions laid down in the provision are satisfied?

(12) Can the user’s efficiency gains from temporary acts of reproduction be taken into account in assessing whether the acts have “independent economic significance” (see Article 5(1) of Directive 2001/29)?

(13) Can the scanning by a commercial business of entire newspaper articles, subsequent processing of the reproduction, and the storing and possible printing out of part of the reproduction, consisting of one or more text extracts of 11 words, without the rightholder’s consent be regarded as constituting “certain special cases which do not conflict with a normal exploitation” of the newspaper articles and “not unreasonably [prejudicing] the legitimate interests of the rightholder” (see Article 5(5) of Directive 2001/29)?’

The Advocate General answered each question separately, but the Court decided to short-circuit the issues by ruling as follows:

"1. An act occurring during a data capture process, which consists of storing an extract of a protected work comprising 11 words and printing out that extract, is such as to come within the concept of reproduction in part within the meaning of Article 2 of Directive 2001/29 ..., if the elements thus reproduced are the expression of the intellectual creation of their author; it is for the national court to make this determination.

2. The act of printing out an extract of 11 words, during a data capture process such as that at issue in the main proceedings, does not fulfil the condition of being transient in nature as required by Article 5(1) of Directive 2001/29 and, therefore, that process cannot be carried out without the consent of the relevant rightholders".

Says the IPKat, this short-circuiting approach is one that the ECJ rather likes, since it redirects the focal point of its rulings from the factual details -- which are in its view the province of the referring court -- to the general principles that national courts are to provide. Apart from the fact that the ECJ is wary of being asked effectively to rule on the facts, the more detailed are its answers to detailed questions, the easier it is to distinguish subsequent cases on the facts. He also muses on how just 11 words can generate so many thousands of words in terms of opinions, judgments etc.

Merpel says, it's an odd situation really: the printing out of the 11-word extracts looks like a non-transient part of a process which is part-transient, part-fixed; economically it is effectively transient but physically it isn't. Tufty adds: the Court made it plain that Article 5 provides derogations from the principle of protection for authors' rights and that, being exceptions to the basic rule that consent is needed, those derogations must be construed narrowly.

Eleven words that sound dirty but aren't, here

Day in a life, summarised in eleven words

Elven words here

Transient cat here