Copyright confusion dogs European digitisation push
The cultural life of Europe will suffer unless more effort is made to clarify what libraries can do with so-called orphan works, says a study.
The British Library looked into ways to speed up the digitisation of books, journals and other printed materials held by Europe's libraries.
It considered 10 works from every decade between 1870 and 2010.
About 43% of the sample were orphan works suggesting a large part of Europe's media may never go online.
"The law does not allow us to use any copyrighted material if we cannot find the rights holders," said Ben White, head of intellectual property at the British Library. Such works where rights holders cannot be found are known as orphan works.
That meant, he said, that across Europe enormous amounts of material would have to be left on the shelf as nations push on with mass digitisation projects.
The study considered a wide variety of printed works including illustrated children's book from the 1920s, travel guides, political pamphlets from the 1960s and 1970s and early "fan fiction" from the 1980s.
Many of the orphan works uncovered in the study were published more than 50 years ago but a surprising number came from more recent times.
For instance, the 1980s produced more than half of all the in-copyright orphan works considered by the study.
The potential legal problems involved with publishing orphan works that are subsequently found to have a rights owner, has led libraries to concentrate their digitisation efforts on older and historical works they are sure are out of copyright.
The study was carried out in support of the European Arrow project which wants to automate the process of finding who owns the rights to a work.
Automation is going to be essential if Europe is to meet its targets for putting books and other printed materials online, suggests the study.
"The costs around the time it manually takes to look for rights holders could be much much higher than actual cost of the digitisation projects," he said.
Manual clearing of rights takes on average about four hours for each text, found the study.
By contrast, an automatic system like Arrow which links together rights databases can reduce the search time to five minutes.