Digital Libraries

Mr Ceynowa, the Bavarian State Library is the first German library to cooperate with the search engine Google. Did you approach Google or did Google approach you?

To be honest, I can't remember because the BSB already had many contacts with Google in other areas. What I can say is that we were very active in purusing the talks because we were greatly interested in a cooperation like this.

What benefits will such cooperation bring you?

On a very fundamental level, as a library it is our job to bring people and information together. As an international research library we don't have a mandate geared towards a specific user group. Rather, with our unique collection we have a commitment to research and teaching, to studying and science. In this respect, the Internet would seem the most natural medium for a traditional research library like ours. It is one of our prime objectives to make our holdings – provided it is permissible under copyright law – available digitally worldwide as soon as possible.

The digitalisation of such a large collection cannot be financed by the public sector alone, unless we are prepared to let this project drag on for over 20, 30 or 40 years. With a public-private-partnership, on the other hand, we can get it done in a few years – and Google is taking on the entire costs of the digitalisation.

In addition, as the central regional and archive library of the Free State of Bavaria we have the obligation to ensure the long-term preservation of our collections. Digitalisation also allows us to solve the urgent problem of preserving information for collections at risk of paper degradation.

The books that Google is scanning are all publications that are no longer subject to copyright protection – in other words, the authors have already been dead for 70 years or more. What criteria are used to select the books?

The aim is to digitalise everything in our collection that is not bound by copyright. Exceptions are manuscripts, incunabula, very valuable materials or items that are already heavily damaged, for which we will continue to use the customary digitalisation methods via the German Research Foundation (DFG) and EU funding programmes. Hence the materials are not selected according to content, but based on conservation and restoration criteria.

Do you have any influence on the type of digitalisation and indexing that Google carries out?

Yes. The metadata of the titles that are being digitalised are already available at our library. It is very much in our interests that these qualitative data are used by Google.

In addition, we are establishing the conservation criteria for digitalisation suitability together with Google and I must say we are constantly pleasantly surprised by the level of awareness Google shows in this regard.

Your library is known for its work in the area of digitalisation. Does it not upset your staff to know that Google will just be laying the books on the OCR scanner?

No. And anyway, what does "just" mean? The procedure that Google is using is one we use ourselves in many projects. It is not as if there are two completely contrasting worlds here: the procedure used by our Munich Digitalisation Centre on the one hand and the evil Google world with its McDonald's-style digitalisation – as was so maliciously claimed – on the other.

Which procedure will be selected depends much more on the technology, the material and, of course, on a quite fundamental level, on the aim of each individual project. If you are looking after a very limited section of a collection then you can use a procedure in which every page is typed twice and then checked by hand. But then you also have costs of 50 cents per 1000 characters. As soon as you are dealing with genuine mass digitalisation, a procedure like this is simply no longer financially viable.

The German Research Foundation (DFG) wants to spend 10 to 20 million euros a year from 2008 on digitalising the collections of academic libraries. Will that be enough to take away the incentive to enter cooperation agreements with other business enterprises?

No, of course that won't be enough to make public-private-partnerships no longer necessary. Nevertheless, we are very happy about it and we're taking advantage of the programme, which will run for five years initially. But the DFG sponsors digitalisation in very specific areas. The programme is largely limited to German prints up to the year 1800, archival finding aids and selected special collection areas of the DFG. In addition, it is not just aimed at the large universal libraries, but at all German academic libraries so there will be many mouths to feed, as it were.

On the other hand our cooperation with Google gives the DFG more latitude because whatever we do within this cooperation doesn't have to funded by the DFG.

Will your library receive a copy of the digital versions which you can make accessible from your library portal or from higher-level national portals as well?

Yes, we will receive a copy and by contract we have the right to offer the digital materials not just on our website, but also – via the meta data – in associated databases, in the German Digital Library and in the planned European Digital Library.

Are you allowed to use this copy however you wish for as long as you like?

Yes, the copy is the property of the BSB. We can make it available through the services we offer on a permanent and unlimited basis – and we can and may do this free of charge as well.

Agnès Saal, Vice President of the Bibliothèque Nationale de France, said she was disappointed about your decision to work with Google. She thinks Google's conditions are unacceptable and criticises the contracts for not being transparent. How do you respond to this criticism?

With three arguments. Firstly, the constant comments that the contracts are not transparent are starting to get on our nerves. It's not a public sector project that we are dealing with here and non-disclosure agreements are simply standard procedure with public-private partnerships. Those who don't like it just haven't understood the essence of these partnerships.

Secondly, people are alleging that we have a very limited scope in what we can do with our copy of the digital versions. But that doesn't reflect the reality of our contractual agreement at all.

And, finally, the Bibliothèque Nationale de France fears being swamped with culturally dominant Anglo-Saxon literature, since Google supposedly favours English-language publications. However, that wasn't even true of the first cooperation agreements that Google closed with libraries in the USA and the UK. Analyses have shown that just 49 % of the titles being digitalised there are English-language titles.

We have the feeling that overall there is a relatively short-sighted anti-American and anti-commercial sentiment lurking behind the protests and that some are trying to exploit this for political means. The arguments based on facts are fairly threadbare.