Should Search Engines Copy less than 11 Words in Search Results?
Recently the European Court of Justive (ECJ) delivered an
interesting judgment that may also have an impact on whether search
engines are systematically infringing copyrights. One of the
questions referred to the ECJ by the Danish court in this case was
whether use of part of a publication - 11 words, to be precise -
could be regarded as a reproduction of that publication, which
would mean that such reproduction would only be allowed with the
prior consent of the copyright owner.
The parties in this case were DDF (a professional association of
Danish publishers) and Infopaq, a company engaged in data
monitoring and -analysis.
Infopaq scans published articles and makes these articles
digitally searchable. The search results are then reproduced by
capturing the five words which come before and after the search
term. For example:
"4 November 2005 - Dagbladet
Arbejderen, page 3: TDC: 73% "a forthcoming sale of
the telecommunications group TDC which is expected to be
acquired"'.
This short fragment of eleven words in total is subsequently
saved and printed by Infopaq.
Does the copying of only eleven words already qualify as a
(copyright-law) reproduction of a publication? The ECJ is of the
opinion that it does. A copyright owner not only has the right to
prohibit reproduction of its work as a whole, but also to prohibit
reproduction thereof in part. To establish whether there is such a
partial reproduction, the ECJ considers that regard should be had
as to whether "the originality of the whole publication
(such as a newspaper article) is conveyed" in that
reproduced part (in this case the 11 words). The copying of
isolated words does not pass this test, but in the opinion of the
ECJ it cannot be ruled out that certain isolated sentences, or even
certain parts of sentences in a text - which would include the
eleven words copied by Infopaq - may be considered a partial
reproduction.
The ECJ gives the impression that this is a broad test, under
which partial reproduction may be quickly established. I think,
however, that in practice this test will turn out less broad than
as presented by the ECJ, for a random selection of 11 words from a
publication is not likely to "convey to the reader the
originality of the whole publication".
What are the consequences of this judgment for the Internet? A
search result of 11 words is quite similar to the results presented
by search engines. This judgment confirms that search engines, when
representing search results in such a way that part of the source
text is copied as well - as they normally do - run the risk of
creating a (partial) reproduction of the original work. Does this
mean that search engines will require the prior consent of the
copyright holder? That is by no means certain.
The Infopaq judgment only deals with the question of whether
copying 11 words constitutes a (partial) reproduction of a
publication as a whole. If this is indeed the case, this does not
necessarily mean that such reproduction leads to copyright
infringement. In many cases, search engines may be able to rely on
one of the limitations of copyright, such as the right to quote or
freedom of the press. If this is the case, even though there may be
a (partial) reproduction under copyright law, this reproduction
will not be infringing.