The DFG National Research Collection System (Sondersammelgebietssystemplan -SSG) for Print and Electronic Publications

Germany (the sum of the sixteen states of the Federal Republic) is historically and politically a decentralized country. The influence of the states is very significant, particularly in cultural and educational affairs (Kulturhoheit der Länder). There are only few organisations that are jointly financed by states and the federal government. One of the most important of these is the Deutsche Forschungsgemeinschaft (DFG), the German Research Foundation. It is mainly driven by the research community, and a highly efficient evaluation system, managed by a well-organized office in Bonn. Founded as the ‘Notgemeinschaft der deutschen Wissenschaft’ after the Second World War, the DFG began a library acquisitions programme, mainly for foreign research material, in 1948 (Schmidt & Oertel, 1966). The aim was and is to provide a copy of every publication of research interest in at least one German library. There was and is no real national library in Germany. Leading libraries such as the Prussian State Library [1] were divided after the Second World War, or were severely damaged like the Bavarian State Library. As a result, the organisational scheme was simple and yet sophisticated:


INTRODUCTION
Germany (the sum of the sixteen states of the Federal Republic) is historically and politically a decentralized country. The influence of the states is very significant, particularly in cultural and educational affairs (Kulturhoheit der Länder). There are only few organisations that are jointly financed by states and the federal government. One of the most important of these is the Deutsche Forschungsgemeinschaft (DFG), the German Research Foundation. It is mainly driven by the research community, and a highly efficient evaluation system, managed by a well-organized office in Bonn. Founded as the 'Notgemeinschaft der deutschen Wissenschaft' after the Second World War, the DFG began a library acquisitions programme, mainly for foreign research material, in 1948 (Schmidt & Oertel, 1966). The aim wasand is -to provide a copy of every publication of research interest in at least one German library. There was -and isno real national library in Germany. Leading libraries such as the Prussian State Library [1] were divided after the Second World War, or were severely damaged like the Bavarian State Library. As a result, the organisational scheme was simple and yet sophisticated: • to seek the most important holdings, • to give the library with the strongest collection special funding to acquire foreign material (nowadays about 66% of costs); and • to ask the library to take responsibility for buying foreign material as well as the complete range of German publications, and the staff required for acquisitions processing and interlibrary loan services.

A SUCCESS STORY
More than 20 leading German libraries and a lot of smaller special libraries receive support through this funding scheme for special collections [2] Over the last fifty years some major fields have been taken over by national central libraries (Medicine, Engineering and Economics) [3] The German unification brought new participants in the Eastern Part of Germany. The divisions between subjects and the funding regulations for acquisitions are quite complex in order to avoid financing duplicates, but -with a well-organised interlibrary loan facility in the background -the system operates rather well. [4] The combination of a relatively small central investment with local financial and personal contributions is shown in the statistics of library costs developed by the Working Group of SSG libraries: in 2003 the DFG accounted for 63% of the cost of the acquisitions programme. [5] The ratio is turning up, if the staff costs are included: the libraries expenditures are 68% of the full amount: These figures show how successfully DFG funding acts as an incentive for additional decentralized investment. But do the libraries have an interest in sustaining the system? The most positive effect of DFG funding is the opportunity for a library to buy a more or less complete range of materials in one of its strongest collecting fields by allocating funds at the level of an 'ordinary' research library. The 'Sondersammelgebiet' can be an asset to the library and -if it is a university library -to the university as well. But the reduced market power of libraries (mainly as a consequence of increases in journal prices) has caused stringent reductions in acquisitions in many libraries. Sometimes additional cuts have, for example, made it necessary for libraries to cancel their acquisitions programme for monographs. But the expenses for the Sondersammelgebiet have to be accepted. As a consequence, an increasing percentage of the budget has to be allocated to the DFG special collections. That's good for library patrons at a national level: special research literature is undoubtedly acquired. However, the local library may get into a position of conflict with the university over reduced acquisitions in fields of stronger local interest. For example, the University Library of Tübingen had to step down from the SSG South Asian studies. But the South Asia Institute of the University of Heidelberg was willing to take over the responsibility, by this means acquiring additional funding for its main collecting field. This shows that the SSG system is quite sustainable: • the shared budgeting scheme makes local costs accountable, • the shared selection scheme with a trusted group of well-selected subject specialists makes the German National Research Collection system internationally unique. Comparative studies in the Netherlands have demonstrated this from a national point of view. [6] DRAWBACKS But there are drawbacks in the system too. The increasing number of publications and their growing costs are becoming a problem for the DFG budget as well as for the local library. As a consequence, there are discussions about strategies for reducing costs. An evaluation study carried out by Freiburg University Library has shown that there is no significant overlap between the funded collections.
[7] But how 'special' are these collections? How much unique material do they provide? The Münster evaluation study on system usage asked this question. It came to the conclusion that in selected subjects 19.7% to 41.5% of monographs and 18.2% to 54% of journals were unique (Nutzungsanalyse, 2003). But perhaps more potentially alarming -at least at first glance -were the usage figures. These showed that • the number of interlibrary loans satisfied by libraries with special collections is not very high (23.8%), • a lot of 'ordinary' libraries provide material on special subjects in addition to the SSG library, • users are not generally aware of the SSG library in their special field.

INTERLIBRARY LENDING & DOCUMENT DELIVERY (ILL/DD)
Does the system of Research Collections fail? A brief look at the interlibrary lending and document delivery system in Germany is necessary in order to understand the figures better.
Interlibrary lending is normally arranged hierarchically (Ordnung, 2003): • local holdings have to be exhausted -including open access departmental libraries without lending services, • regional holdings (which are easily searchable via the centres of regional networks with electronic union catalogues and ILL-ordering systems), • national holdings (KvK).
As a consequence, highly specialised demand is delivered by SSG libraries if it cannot be satisfied by any other library -a quite sophisticated solution as long as one copy of the printed book is in the holdings of the SSG library: the SSG system is a 'last resort'-an arrangement with the communities of local and regional libraries at the forefront, and the SSG libraries relatively hidden at the end of the service chain.
The SUBITO document delivery system is quite different. A small number of about 30 libraries offer direct online ordering from their holdings. The system began with document delivery of copies of journal articles; now books may also be ordered. However, the most significant type of order is for electronic copies, where it does not matter if you have an item as the source copy in the background. The SUBITO service normally provides these copies direct to the end user (in between there is an additional library service that plays a minor role). The service fee for direct delivery is higher than the costs for interlibrary loan services, but the SUBITO service is much faster. Most SSG libraries participate in both interlibrary lending and the SUBITO service, and in its Green Paper on the future development of the SSG system, the DFG will make participation compulsory.

FUTURE DEVELOPMENTS
The DFG Green paper deals with ideas and alternative approaches for the acquisitions system for monographs, journals, electronic material and the development of services (DFG-System, 2004). The group of SSG libraries (a subgroup of the German Library Association (DBV), Section 4 -Research Libraries) has discussed this and, with the assistance of funding from the DFG, is preparing research proposals to optimise the system.

Monographs
The Green Paper proposes that the SSG library should act as 'collection manager' for the subject for which it is responsible. This seems to be an unachievable task: • The SSG library has no responsibility for the collecting policies of other libraries, and it, therefore, has no ability to manage the sustainability of the system.
• It is not possible to reduce acquisitions to unique material only: the uniqueness cannot be foreseen. If the SSG library, for example, waits for one or two years to assess an item's uniqueness, the publication may by then be out of print.
• Searching for unique material is as consumptive of staff time as preparing a complete collection.
• Increasing costs for minor amounts of funding will reduce libraries' willingness to take on responsibilities in the National Research Library System.
To sum up: for printed material, completeness of collection is necessary in the library that is funded. But it may be meaningful to develop strategies for collection mapping. The SSG libraries group is therefore preparing more detailed research in this field.
Collection mapping projects have been developed in different countries. SCONE, the Scottish Collections Network Extension, uses descriptions to identify strong collections. The RSLP Collection Description scheme was developed in England but is not widely used. One of the most impressive collection description activities is the 'Handbuch der historischen Buchbestände in Deutschland' in 27 volumes, developed by Bernhard Fabian (Fabian, 1992), and extended by the 'Handbuch deutscher historischer Buchbestände in Europa' (Fabian, 1997).
But none of these projects offers real assistance for collection management activities at national level. The most promising scheme is OCLC/LACEY, which is used in different library environments, for example, in Illinois.
[8] The holdings information of 86 libraries with 24,236,228 titles were mapped against DDC and the LC classification. The results demonstrate the level of overlap and uniqueness. They can show a library how strong the collection is in a special subject, and help to identify collections that need support. Another evaluation study of OCLC LACEY is prepared by CURL. [9] The OCLC tool cannot immediately be used in Germany. The lack of DCC or LCC classification in German holdings information is the main drawback. But a German collection mapping project for monographs is planned, using the OPACs of SSG holdings or, as a first step, the acquisition lists of the last decade. The titles will be compared with the holdings of other ILL libraries to investigate uniqueness and completeness.
This data may be used for the discussion of alternative collection (and perhaps collection management) policies for subjects with a high number of academic staff and student users, and highly specialised subjects with a small population (the so-called 'Orchideenfächer' -'orchid' subjects). It may be that the percentage of DFG funding can be reduced if it is clear that many libraries 'normally' spend a higher amount than the 35% rate the SSG library has to add to the DFG grant (with additional full costs for German publications): vice versa, the funding for smaller subjects had to be higher, compared with the average amount spent in the research library community.

Journals
The national provision for journal publications differs significantly in print and electronic environments. The library is the owner of the printed journals to which it subscribes. It may -at least on the basis of German copyright ('Urheberrecht') regulations -copy the materials via interlibrary lending or document delivery in response to requests from users. A flat copyright fee for each copy is paid to the Verwertungsgesellschaft Wort (the German legal collecting society). Through this scheme the SSG library is able to satisfy all demands for copies of SSG journals. On the other hand, anyone who is interested in a defined subject will subscribe to the journals in the research field and use the SSG system as an additional resource for special material.
The situation with electronic journals is rather different. The library is a licensee of the journal, and the publisher retains the ownership. Use for interlibrary lending or document delivery services depends on a licence from the publisher as the rights owner. If a SSG library requests a licence to allow electronic document delivery, the publisher may be suspicious that this will be used as a national licence, thereby significantly reducing the scope for further licences in the country. As a result, the publisher will normally refuse the request, or will demand a high additional licensing fee. Undoubtedly, this cannot be met by the DFG/SSG funding scheme with the objective of securing access for special research purposes. Unlike JISC, the DFG cannot act as a national financing agency for the higher education system.

National licenses
But the seemingly impossible was achieved for the first time in 2005: through the lead taken by four large SSG libraries 18 databases were licensed at national level (Nationallizenzen, 2005). Members of all (mainly publicly funded) universities in Germany have free access to these databases; and in many cases members of research institutes (e.g. Max-Planck-Institute) and even -an internationally unique facility -private individuals with research interests. In 2006 additional databases and digitised backlists of journals of major publishers will be considered for similar national licenses. This seems to be a real breakthrough in the provision of electronic material, but as has been mentioned earlier, there is no chance of securing national licences for the full range of electronic materials, especially for recent issues of costly journals. There are additional schemes under discussion to address this problem, for example, a combination of free access for the backlist and a 'pay per use' approach for recent material.

Collection management
Nevertheless, the idea of collection management for journals should be discussed in more detail. Here, again, the Green Paper proposes a new definition of the tasks of the SSG library: co-ordination and compensation are to be the new duties. The background idea is again to reduce DFG costs. Only titles with an insufficient number of subscriptions will be funded. Tools should be developed to generate holdings statistics automatically from the German Database for Journals (Zeitschriftendatenbank), with additional information provided from the EZB (Elektronische Zeitschriftenbibliothek) for electronic material. These are very significant databases with particularly good service functions. For example, if a user is searching the site of a virtual library and orders a special document, local access information is provided to the user if the library of which he is member has a licence for the journal.

Virtual libraries
The development of virtual libraries is one of the special project areas for research and development activities funded by the DFG.
[10] In terms of additional activities for SSG libraries, the digitisation of journals is a main priority for the DFG programme. DigiZeitschriften -the German JSTOR -with 3,000,000 pages as its goal -provides more than 1,600,000 pages of the 60 top German research journals starting from the first issue.
The virtual libraries provide electronic services with access to online catalogues, online contents tables for journals, guides to websites in special fields and access to additional online databases. Some are joint activities among two or more libraries. The development is often driven forward in close co-operation with the research community. Additional co-operative developments, which are in preparation among virtual libraries, include: • union catalogue of websites, • collecting and archiving of grey literature in electronic format instead of print, • OAI repositories for special research communities

Vascoda
These activities are brought together in the German research portal Vascoda which includes the virtual libraries of SSG libraries, national central libraries and other distributed information service providers.
Overall searches, cluster searches (Engineering Sciences, Life Sciences, Economics, Social Sciences and Humanities with Area Studies), and searches in special subjects are all offered.
The system functions as a metasearch engine for participating services. [11] Search engine access (with Fast technology) is in preparation.

European co-operation
The rapid development of the arrangement of SSG libraries providing special literature for research purposes, and the growing system of virtual libraries in Vascoda, offer many opportunities and pose challenges for international library co-operation.
Fields of co-operative web services developed by European libraries could include, for example: • open access activities, • (subject) cataloguing of websites of research value, • ILL and DD services; and • long-term archiving.
LIBER may offer the best platform for finding the right partners for this future aspect of the development of library services.