1.1. Changing environment, changing discovery
Libraries across Europe are responding to the challenges posed by rapidly changing technology; emergence of new media; digitisation of old media; demands for interoperability across sectors and domains and increased user expectations.
Particularly exciting for cataloguers is the prospect of the new opportunities for resource discovery in a Web environment where (in principle) anything can be related to anything else for any reason. In order to take advantage of these possibilities, libraries have been reassessing the models and standards underlying cataloguing. This process began with the work to develop the FRBR model (IFLA, 1998) which was subsequently extended to name authorities (IFLA, 2009) and subject authorities (IFLA, 2010). Also in 2010, RDA; Resource Description and Access (RDA, 2010), was published as an implementation of FRBR and FRAD. This article reviews the response to RDA in Europe, based mainly on recent surveys by EURIG1, the European RDA Interest Group.
1.2. Changing the Rules: from AACR2 to RDA
RDA was published online and in print in 2010 to address perceived shortcomings of the Anglo-American Cataloguing Rules, 2nd Edition (AACR2, 2002). As Delsey (1998–1999) reported, the logical model of AACR2 is based on classes of materials and does not clearly distinguish between content and carrier. This ambiguity means that AACR2 is ill equipped to respond to the migration of content from physical carriers like volumes, sheets, disks or reels to digital media. Some of these problems were evident even before the development of the Web and mass digitization and were among the factors motivating the Toronto conference on the Principles and future development of AACR2 (Weihs, 1998). The Toronto conference recommended development of a new edition of the Anglo-American Cataloguing Rules, but, in 2005, the Joint Steering Committee for Revision of AACR announced its decision that an entirely new product, RDA, would be developed instead (JSC, 2005). RDA was published in 2010 as an online toolkit and in print and subsequently as an e-book. The RDA element set has also been published as open, linked data, for re-use by libraries and the wider community (RDA, 2014).
RDA’s development has been controversial and changes to AACR2 provoked considerable debate, particularly in the United States of America. Prior to formal adoption of RDA by Library of Congress, National Library of Medicine and National Agriculture Library, in 2013, the US community undertook a substantial programme of testing. During 2013 RDA had also began to be implemented outside North America; by the British Library, Cambridge and Oxford Universities in the UK and by libraries across Australia.
AACR2 has been widely adopted beyond the English speaking community. It has been translated in at least forty languages, including major European languages. Despite its success, AACR2 makes linguistic and cultural assumptions which are not universally applicable and which have to be interpreted or adapted to satisfy community needs. RDA aspires to be an international code and several European national libraries and rule making bodies participated in its development, by commenting on drafts and reviewing proposals. The Deutsche Nationalbliothek was among the agencies that provided feedback to the JSC2 on the early drafts and has taken the lead in promoting RDA to the German speaking community. In 2012 it announced its commitment to implement RDA and in 2012 became the first JSC member from a non-English speaking constituency.
European libraries have been understandably cautious about committing themselves to RDA. At the 2010 joint seminar in Copenhagen (EURIG, 2010), organised by the Joint Steering Committee for Development of RDA (JSC) and the European RDA Interest Group (EURIG), speakers from Europe’s major national libraries were asked about their plans regarding RDA. The majority said they were waiting to see how Library of Congress’s implementation went. The meeting agreed that EURIG provided a useful network and should be formally established.
EURIG (European RDA Interest Group) is a member organization which acts as a focus for discussion of RDA in Europe. Originating as an informal group, formed by the British Library, Deutsche Nationalbiblothek and Biblioteca Nacional de España, EURIG was formally established in 2011 and has subsequently become an active part of the RDA community. EURIG has 32 members representing 20 countries. Members are mainly national libraries, but also include library networks, bibliographic agencies, standards bodies and companies.
1.3.1. Membership Surveys
EURIG has followed up on the informal poll taken in Copenhagen by surveying its members in 2012 (EURIG, 2012) and 2013 (EURIG, 2013) to evaluate changing needs and intentions regarding RDA implementation. It is important to keep in mind that EURIG is not representative of Europe as a whole and only members were surveyed. The surveys therefore cover most countries in Western Europe, but exclude large parts of Eastern and Southern Europe. EURIG members are predominantly national libraries or other major bibliographic agencies and the responses should be interpreted as reflecting the situation of the agency, rather than the national position. A more nuanced picture may emerge from national surveys.
In 2013, the same questions were also put to libraries in the UK, providing a snapshot in a country where implementation is already underway (Danskin, 2013). Only 47 valid responses were received and, it should be noted that the respondents were predominantly research and academic libraries (75%). Only 8.5% of responses came from Public Libraries. It is not clear if this reflects lack of awareness of the survey or a perception that it is not relevant, as few public libraries in the UK retain responsibility for their own cataloguing.
2. Cataloguing in Europe
2.1. Plans to implement RDA
EURIG asked respondents whether they had taken a decision to implement RDA and, if so, when they would implement. Those who had not yet taken a decision were asked whether they had a timescale for making the decision.
The responses illustrate increasing commitment to RDA. In the 2013 survey 13 organisations (57% respondents) indicated they had definite plans to implement RDA. This is a significant increase on 2012, when only 4 organisations (20% respondents) actually had plans to implement the standard. It is also notable that, in 2013, 4 respondents were already implementing RDA; another 4 were planning to do so in 2015; and 1 other in 2016 and the remaining respondents had not yet set a date. The feedback in the 2013 survey was markedly more definite than in 2012, when the majority of respondents had yet to decide on when to implement RDA although they did expect to make this decision in the following few years.
The same question was put to UK libraries by CILIP Cataloguing & Indexing Group. The vast majority of respondents (39/47 or 85%) have decided to implement RDA. Of these, 16 (34%) have already implemented and 14 more expected to implement within 18 months. None has decided not to implement, but 7 have yet to decide. The relatively high number of implementations in the UK can probably be attributed to the absence of a language barrier and the high dependency of all libraries on copy cataloguing from Library of Congress, OCLC, BDS (Bibliographic Data Services) and the British Library, which had all given advance notice of their intention to implement RDA.
2.2. Reasons for implementing RDA
Respondents who had indicated that they intend to implement RDA were asked to give their reasons.
All the respondents in 2013 cited interoperability with others as the key reason for implementing RDA. Other most cited reasons for adopting the standard were the requirement to revise or replace the current cataloguing code and also the desire to implement FRBR. ‘Cost effectiveness’ and ‘the range of materials covered by RDA’ were seen as less significant considerations by the respondents.
Interoperability was also the main reason for adopting RDA in the UK; 70% of respondents gave this as a reason, as opposed to only 17% who selected FRBR. In the longer term (more than 24 months) there was a greater expectation of improved resource discovery and more effective cataloguing of non-print resources. There was little expectation that RDA would improve productivity or generate efficiency gains.
2.3. Reasons for not implementing RDA
Respondents who had indicated that they had no plans to implement were also asked to indicate their reasons for not implementing RDA. It is notable that no respondents had taken a decision not to implement RDA.
In the 2013 survey a small percentage of respondents (7 organisations/29% respondents) indicated reasons for not adopting RDA and all these respondents said they were waiting for evidence of how RDA works for other organisations. Other reasons, such as: cost of subscription; lack of suitable translation; or lack of cost effectiveness, were cited by too few respondents to be of any significance.
As is Europe, UK libraries which have not implemented RDA, are mainly waiting to see how others get on, but there are also concerns about the availability of training and the need to change systems. In RDA familiarisation sessions held in the UK, many smaller institutions have said that the subscription cost and the uncertainty surrounding the benefits make it difficult to sell the case for transition from AACR2 to their parent institutions.
2.4. Expected benefits from using RDA
The 2013 survey asked respondents about the benefits they expected from using RDA in the short term, viewed as 18–24 months, and in the longer term, seen as over 2 years. This question had a high response rate of 83%.
In the short term, most respondents expected to achieve benefits in the areas of interoperability, improved resource discovery and increased responsiveness to change. These respondents did not anticipate gains in enhanced productivity or cost efficiencies over this time scale.
Longer term, a high percentage of respondents expected benefits to be evident in terms of cost efficiencies and enhanced productivity, and less so in interoperability and responsiveness to change. This question was not asked in the 2012 survey.
2.5. Impact of RDA on productivity
Respondents who had implemented RDA (in 2013) were asked to comment on the impact of RDA on productivity. Only 4 organisations (16%) supplied information and consequently this information can be considered to be of limited significance. However the respondents did make some interesting, rather mixed, comments. One said it was ‘too early to assess’, another that they ‘need to adjust their workflows’ and finally that there were ‘no significant changes for most categories of material, but there was an increase in the quantity of authority work required for conference proceedings and other grey literature’. This question was not asked in the earlier survey.
The British Library has now been using RDA for over a year and has not seen any significant impact on productivity as a consequence of the migration to RDA. The Library has reported an increase in the time being spent on authority control. This has been attributed to the relaxation of AACR2’s arbitrary limit of three added entries. RDA imposes no such restriction and therefore more controlled access points have been created for compilations and conference proceedings. The Library has addressed this by applying the core requirements to this kind of material.
Other UK agencies that have been using RDA for a similar or shorter period mostly said it was too early to tell, others reported either neutral or slightly negative impact on productivity, but commented that RDA is just one of many changes being experienced in a period of financial retrenchment and technological change, so that the “RDA effect” is hard to isolate.
2.6. Database structure
In both surveys, respondents were asked about their database structure because systems architecture has a bearing on the benefit that can be derived from FRBR implementations, such as RDA. The respondents were asked to select between options corresponding to the three implementation scenarios defined by the JSC (2009):
- Scenario 1: Fully Relational/Object Oriented;
- Scenario 2: “Standard ILS” with related bibliographic, authority and holdings records
- Scenario 3: Flat File.
Perhaps not surprisingly, given the relatively short interval, the responses were consistent in both surveys – a large majority, over 70% of respondents described their current database structure as ‘linked bibliographic and authority records‘; corresponding to RDA implementation scenario 2. Around 20% considered their current database structure as fully relational/object-oriented.
In the UK a higher percentage of respondents (85%) described their systems as corresponding to “scenario 2”. Only 2 respondents had fully relational systems and 10% of respondents are using flat file systems, corresponding to implementation scenario 3.
Post implementation of RDA the number of organisations expecting their system to be enhanced to a fully relational/object-oriented one almost doubled to 7 with a corresponding drop in those with linked bibliographic and authority records database structure.
FRBRisation refers to the explicit representation in bibliographic metadata of the entities defined by FRBR, in particular the Work, Expression Manifestation and Item entities, which are fundamental to the FRBR model. There was no significant difference in responses between 2012 and 2013. Just over half of EURIG respondents (13 organisations) have yet to decide whether to FRBRise. A small minority (3 organisations) did not plan to do so. The remaining respondents had decided to FRBRise their data.
As an alternative to restructuring legacy data, algorithms can be used to present a FRBRish view of bibliographic data in the user interface. When asked about the FRBRisation of their user interface, the majority are not decided, but around a third responded positively.
Many UK respondents commented that they are entirely dependent on system suppliers to make configuration changes to their systems, which obviously colours their view of potential benefits of FRBR. It was also evident from the comments that some UK librarians remain sceptical (or poorly informed) about the FRBR model.
2.8. Translations of RDA
The 2012 survey questioned respondents about translations of RDA. Almost half of the responding organisations indicated they had plans to translate RDA. Translations of RDA were planned for French, Finnish, German, Portuguese, and Norwegian. The remaining organisations were split almost equally between those who had no plans for translating RDA and those undecided.
At the time of writing, the RDA Toolkit includes translations of RDA into German and French. Spanish, Italian and Swedish translations are in progress and are expected to be incorporated into the toolkit. Translations into Chinese and Japanese are also in preparation, although there are no plans at present to include these in the RDA Toolkit. ALA publishing is also in talks regarding translations into Arabic, Catalan, Finnish, Korean and Norwegian. The translation of the text of RDA is a major undertaking in itself; on top of which is the translation of associated documentation, such as change proposals, and commitment to the maintenance of the translation in line with RDA updates and releases.
It is not surprising that some countries have considered cheaper alternatives. The Netherlands has decided to implement RDA without a full translation of the text. Instead, a workflow or interpretation layer in Dutch is being developed which it is expected will be sufficient to meet the needs of most cataloguers, most of the time. The Dutch workflow will contain links to the English text of RDA, enabling cataloguers to check the instructions when more guidance is required. The RDA glossary definitions are also being translated into Dutch, but the intention is to retain the English terms so that the cataloguers are familiar with these in the instructions. This model could also suit other countries where English is widely spoken.
2.9. RDA training materials
Respondents in 2012 were asked about the availability of training materials for FRBR and RDA. Then, 7 organisations (35% respondents) had prepared or were in the process of preparing training materials with just under half of these making their material available on the internet. The remaining respondents were not preparing such material. This question was not repeated in 2013. Work is going on independently in different countries to support national or institutional implementations and some documentation has been published on the Web. Some institutions have signed up for distance learning courses run by American universities. Feedback received during workshops and seminars indicates that there is a need for a range of different training options, particularly to meet the needs of smaller institutions and solo cataloguers.
EURIG membership is strongly oriented towards Western and Northern Europe, with very limited penetration into Southern and Eastern Europe. It is not clear whether this reflects differing priorities or a lack of interest in RDA. Among EURIG members, it is clear that RDA is seen as a potential solution to the problems of updating aging cataloguing codes and as a means of implementing FRBR. However, changing cataloguing code is a significant undertaking, which most members have approached cautiously, adopting a wait and see attitude. The implementation of RDA by the Library of Congress and the British Library in 2013 has encouraged others to follow, largely on grounds of interoperability of derived catalogue records. Different solutions to implementation have been adopted in different countries. The German speaking countries are working together on a community wide implementation, whereas in the United Kingdom implementation has been left to the discretion of individual institutions.
The translation of RDA into European languages is removing one of the barriers to implementation, but there remain significant technical barriers to realising the long term benefits for discovery and description. Most European libraries remain dependent on MARC for data exchange and use systems that are not optimised for the linked data technologies that RDA is embracing. Many European national libraries are working on linked data and the next challenge is to bring this work together with RDA.