Helen Shenton, Deputy Director, College Management (Preservation), The British Library

Introduction

Preservation is often seen as the most difficult area in the management of digital material. The need for expertise in this area was identified in Margaret Hedstrom’s study for RLG as a major factor in the successful management of digital material.¹ She makes the comment that skilling and competencies are obviously dictated by what you want to do.

What does the British Library want to do? The BL wants to manage the long term access of the digital material it is purchasing, receiving under voluntary legal deposit and digitising itself. It wants to manage its electronic records. It wants to embrace the advantages of the ‘hybrid’ approach to surrogacy of microfilming/digitising. Essentially, it wants to manage its digital assets and ensure their continued survival.

There is a growing number of collaborative digitisation initiatives within the BL and the Library already has an estimated four terabytes of digitised information. All the technological and publishing developments will not be rehearsed here, but as snapshots, legal deposit of electronic material is estimated to be two years off and in the next 10 years, Mark Bide’s research for the BL’s Policy Unit predicts that ‘ by 2010 ... between 65% and 95% of journals will be published only digitally.'²

It is a truism by now to say that the preservation of digital material is essentially a distributed process, that traditional demarcations do not apply, and an interdisciplinary approach is necessary. But what does that ‘distributed process’ mean? Does it mean totally new, different types of jobs or does it mean that a computer expert needs to be aware of the basic philosophy of preservation; a preservation expert needs to understand the basic issues of digital material; a cataloguer needs to understand all the life cycle of an item? Can this different way of working happen in one leap or are there intermediate steps?

The Council of Europe’s Council for Cultural Co-operation Culture Committee’s report on Training Qualifications and Professional Profiles for New Information Technologies ³ advises ‘against producing ‘hybrid’ jobs, but to place the emphasis on teamwork of different professionals, with an understanding of each others’ cultures and attitudes’. It goes on to discuss the change in professions and skills as a mix of redefinition of the relative weights in the ‘old’ ones and the creation, from scratch, of brand new ones, stressing that the divide between more traditional situations and brand new ones also depends upon factors other than technology alone.

The importance of people and managerial matters are underlined in an interview with Kevin Guthrie in RLG Diginews on ‘Developing a Digital Preservation Strategy for JSTOR.’⁴ in which he says ‘ I think the important issues regarding archiving are organizational’ ... ‘I don’t regard the archiving issue as being about technology really. There is no black box fix.’

British Library context

In the immediate short term within the BL, a new Chief Executive arrived on 1st July 2000 with e-strategy, specifically highlighting digital preservation, as the highest priority and an ever greater emphasis on external collaboration. On 18^th September 2000 the Digital Library System design and implementation started. The contract was awarded to IBM. The Digital Library System will be the infrastructure to guarantee the long term access for selected electronic material acquired, deposited or created by the BL. From a staffing point of view, how did we get to this point, and what do we need to do next?

For digital library preservation and collection management issues, the BL had been reliant on one or two forward-thinking, polymathic IT professionals. The expertise was not strongly located in the business areas of Collection Management. The need for a digital library infrastructure had been identified by the Digital Library Programme Board. A Digital Storage Project was established in 1998, which was to form the core of what became the digital library system specifications for the Invitation to Tender. The mantra of the project was ‘the core is the store’.

This project was originally sponsored by the BL’s representative on the Working Party on Legal Deposit, so the project was already owned by the ‘business’ part of the BL.

The objectives of the Digital Storage Project were :

1. To elicit and capture requirements for an integrated hardware, software and operational environment for the storage of all the Library’s digital collections, accommodating the range of formats and volumes of acquisition anticipated through deposit, purchase and digitisation.
2. To assist the definition of standards for metadata associated with objects in the collections, specifically where this relates to the storage, processing, management and retrieval of digital objects.
3. To contribute the set of requirements for a Digital Store to the Digital Library System procurement process.

Phase I - Digital Storage Project

The method of defining the digital store, was to have five working groups or ‘user’ groups represented by an ‘ambassador’, to think and work through the issues, in areas such as cataloguing, preservation and the reading rooms. A ‘model office’ methodology was used.

The following are extracts, specifically from the ‘people’ point of view, from the end of project report analysing what went well and what did not,.

What went well

• Open communication initiated by the project was ‘refreshing’ and well-received, particularly as it applied across directorates and with external organisations (e.g. CEDARS).
• User areas appreciated the level of consultation and involvement and the willingness to listen to particular concerns. The use of the ‘Ambassador User’ role was a notable success.
• The storyboard, process modelling and Use Case workshops were helpful to users in clarifying their issues and their own requirements.
• Users commented that project management hit a happy medium between firm control and openness to input. Risk analysis ... was beneficial.

What did not work well

• ‘Jargon’ was seen as a problem — both OAIS and modelling jargon — from the user side. This was possibly exacerbated by introducing two new sources of jargon at the same time.
• Members of groups were learning as they went along — although the project accurately assessed the extent of unknown territory being entered into.

This way of working on the specifications generally drew in more and more people across the Library into the arena of digital material, from cataloguers to reading room managers.

From a specifically preservation perspective, a Digital Preservation Working Group had been established in 1999, to address all the preservation aspects of digital activity, that is not just the preservation of digital material. This comprised people from Preservation, the National Preservation Office, Reprographics, Information Systems, Cataloguing and Acquisitions.

For the preservation component of the Digital Storage Project, the group produced a specification using CEDARS draft metadata. A very useful exercise was to compare CEDARs, NLA, Nedlib and BIC metadata. The CEDARs draft metadata was used for the specification.

Phase II Digital Library System

The Library awarded a contract in September 2000 to IBM, following an EU Procurement process. Design work is underway. The DLS is based on the OAIS Reference Model proposed as an ISO standard by NASA. The DLS will be based on a CORBA-based component architecture and will interact with the Library's systems providing the British Library Public Catalogue and a new Reading Room Catalogue. Initial access will be in the Library's Reading Rooms.

The Digital Library system board comprises the IBM project manager and senior supplier, the BL project manager and programme manager, the senior ‘user’ (who represents the interests of all the Library’s personnel) and the project chair (or senior responsible officer, ahead of new Cabinet Office guidelines on governance of projects). The former are IT specialists, the latter two, from Collection Management.

Preservation elements

The first phase of the project is the detailed functional specification with an intense period of workshops between IBM and BL. Because preservation cannot, nor should not, be readily isolated from other issues, an ‘e-team’ was formed with people seconded from cataloguing, acquisitions and the preservation (NPO). They had by the end of September 2000 produced candidate metadata elements for DLS integrating the requirements of Preservation, Accessions and Cataloguing.

There is an interesting parallel with the National Library of Australia’s experience, in that it established 5 people in an ‘Electronic Unit’ to develop ‘selection guidelines and a set of business principles ... to put boundaries around the task so that we wouldn’t be swamped by the enormity of what lay before us ... this enabled working at two levels, the conceptual and the practical. Each informed the other.⁵ This is what is happening at the BL. In an ideal world, the logical order is to have collection development policies, retention policies and a full suite of digital preservation policies and strategies, all in place before any practical steps are taken. There was a draft digitisation policy, draft criteria for digitisation projects and draft digital preservation policy but the practical and the strategic are, in reality, informing each other, given the urgency of developing a store in particular in anticipation of the legal deposit of electronic materials.

The dependencies of collection development, selection and retention policies for electronic material have been mapped onto the workshops and development work and is driving work in Collection Development and Collection Management.

Collaboration

Strategically, the BL is putting great emphasis on co-operation and partnership, and again it is a axiomatic by now to say the subject of digital preservation particularly lends itself to collaborative working. The BL was a major case history for the AHDS workbook, described elsewhere in the conference, which is a digestible and realistic manual for managing digital preservation.

A major collaboration in the DLS is with the Koniklijke Bibliotheek (Dutch Royal library) which has a contract with the same supplier as the British Library, for the same Digital Library System, having gone through a parallel EU procurement process.⁶ The BL and KB are working closely on their respective developments. The project organisation is again similar, with interesting divergences, for example, the project ownership at the KB is with the Head of IT and Facilities Management, at the BL, it is with Collection Management.

Different philosophies

All preservation is about managing risk; minimising the risk of damage to the collections in whatever form and format, and maximising their useful life. The balance has to be struck between managing the risk, and taking a risk. Generally, it seems unfashionable to be cautious, but the history of preservation and conservation is littered with extremely well-intentioned treatments or policies which entail future, further work to rectify, for example the use of soluble nylon to consolidate flaking pigment in manuscript illumination — it is now known to cross-link and turn yellow and it cannot be removed due to the cross-linking. The timeframe of preservation thinking is 200 or 500 or 10,000 years, as opposed to the ever accelerating half-life of technological development. This made for some very interesting debates during the Digital Library System preservation workshops between the different professions.

The preservation approach to digital material is that it is different, but not that different. For example the AHDS workbook nicely summaries the preservation strategies into primary and secondary, which can be seen as parallel to preventive and interventive conservation. Preservation involvement at the point of creation is a preventive activity. In fact this new realm is the opportunity to do what is very logical, and often talked about, but rarely done in many libraries, namely the whole life cycle management of digital material with preservation decisions taken at the beginning or pre-acquisition, which could act as a catalyst for best practice with analogue collections.

Language

This is an underestimated issue. Lorcan Dempsey comments that ‘a feature of change is that we have no settled vocabulary. Some terms may have partial or sectoral associations, which are not commonly shared.⁷ The phrase ‘Digital Library’ seems to have as many meanings as questions it poses. What is meant by ‘archiving’ differs depending on the sectoral background.

Terms are being invented, reinvented, and honed as concepts develop and it can be a barrier for people’s understanding. Some terms seem to be settling down, for example, ‘born digital’ is commonly used in the UK, (though it was new to a colleague from Australia in the summer) and it has the healthy hallmark of being comprehensible to those outside the subject. However, the combination of jargon, acronyms, new concepts groping for expression and words shifting meaning make for an alienating fog which can enshroud the development of skills and competencies.

Preservation components of digital activities

The Digital Library System at the BL is only one, albeit significant, aspect of the digital preservation activity — other areas range from the conservation of the physical item such as CD-ROM and the microenvironment of the CD-case, to digital technologies in the service of curatorial and conservation decisions and treatments.

What constitutes digital preservation at the BL?

1. Preservation of digital, electronic material
   See above
2. Conservation and preservation of the physical carrier of the digital object.
   Issues here include the physical characteristics and longevity of, for example, CD-ROMs, their handling and storage requirements. The concepts are essentially the same as for traditional material, so, for example, the interaction between a physical item (CD-ROM) and its microenvironment (CD-ROM case) and whether there is offgassing over time which would affect the item, is basically an extension of current knowledge and approaches.
3. Conservation and preservation of the original ‘physical’ material which is being digitised ie an electronic copy is being made.
   The conservation of, for example, Dunhuang material to stabilise it before it is digitised, is essentially little different to conservation before an exhibition, requiring already known skills. It is the reason for doing the conservation which is different, so this is basically a resource and programming issue. This category also entails preventive conservation during a digitisation project, for example, in designing cradles to prevent damage during the process. Again, the reason is different, but there is substantial similarity to, for instance, exhibition work. Again this is a resource issue.
4. Digital technology in the service of curatorial and conservation decisions and treatments.
   There is a project at the BL in conjunction with the University of Kentucky to enhance and recover information that is otherwise invisible or illegible on a fire-damaged Cotton manuscript.⁸ This uses Ultra Violet and Infra Red lighting methods, and 2-dimensional and 3-dimensional digital image processing. Whilst the technique is innovative, there is no substantive difference to any other analytical or examination technique, such as ramen laser spectroscopy for pigment analysis for manuscript illumination or a magnifying lens or an astute pair of eyes. As this is being done by a conservator who has developed in this direction, by being involved in the seminal Electronic Beowulf project, again it is essentially a resource and development issue.
5. Hybrid’ surrogacy
   Whether to digitise at the same time as preservation microfilming is a major, well-rehearsed issue, which will potentially have an enormous impact on the reprographics functions, from a technical through to an organisational point of view. In addition, the BL is currently investigating the feasibility of developing the "Register of Preservation Microforms" to incorporate records of digital material besides records of microform material.
6. Policy, strategy, procedures
   A full suite of digital preservation policies, strategies, procedures and guidelines underscore all the activities outlined. At the BL, they are being addressed within the Preservation function, which comprises part of Collection Management. It is interesting to compare how their formulation is being addressed in different libraries and archives, in different sectors.

Building the critical mass

So the preservation of digital material has many elements which are the same as, or simply extensions of current preservation work, and some new elements. The competencies needed for these different roles in a national library were analysed earlier this year (see appendix I for extracts).

Concurrent with this work, the Preservation Department was also addressing questions about the conservation and preservation needs of the BL’s collections in the future, and what traditional skills will need replenishing and what new skills will need developing.

Very simply, on the one hand more books than ever in the history of the world are being published and the numbers of volumes of legal deposit are rising, whilst the laws of entropy dictate that existing organic collections are deteriorating no matter how much environmental conditions are improved and particularly given increased use (half-a-million reader visits to the St Pancras reading rooms per year)..

On the other hand, four terabytes of electronic material already exist, and developments from JSTOR to Highwire Press to electronic patents provision to electronic Ordnance Survey maps to e-journals are changing the nature of the collections to be preserved. The quid pro quo is being examined of, for example, the impact of electronic scientific journals on the need to spend money on binding science periodicals. In tax terms this is a straightforward instance of hypothecation.

There is an issue of succession-planning for traditional bookbinding and conservation skills against a backdrop of a crisis in the provision nationally of such training (should the BL be a repository and training ground for those skills?) What is the optimum balance between preventive and interventive conservation of the collections? There is a lot of discussion about having to reallocate resources in a digital age and the haemorrhaging of funds from preservation into digitisation. In 1999 when the digital preservation working group was established, the Preservation function of the BL comprised some 130 people. Even including the National Preservation Office (which is part funded by the BL and has the national remit for digital preservation) less than the equivalent of one person was working full-time on digital preservation. By the end of 2000, with the appointment of a digital preservation co-ordinator, there will be the full time equivalent of about four people working in the all the preservation aspects of the digital arena, from digitisation to strategy. This costs. In real terms, given the multi-strands of digital preservation activities analysed above, there will be less interventive conservation due to this shift in resource.

Conclusions for staffing for the preservation of digital material at the BL

• Mixed economy
  The analysis of competencies (appendix I) concluded that the BL should create a mixed economy of home grown and bought in competencies. For areas where digital preservation forms only part of a person’s job, such as digitisation projects, it was recommended to develop the individual in post. Therefore the NPO has been commissioned to deliver two-tier training for all in Preservation, comprising firstly a general awareness day and secondly a more advanced course, along the lines of the HATII week course.

  The Council of Europe Culture Committee comments that

  ‘...the problem at stake in cultural institutions in teaming different people with different culture, skills and mental attitudes to co-operate for the purpose of accomplishing a complex task is not to create hybrid skills and professions, but to develop organisations supporting co-operation between people with different culture, skills and mental attitudes. In this perspective the main problem in training is to enable different people to understand each other not only in the field of ICT but in the working process as a whole ... that is a very difficult and sophisticated accomplishment.'⁹

• Home-grown versus bought-in
  The availability of skills in digital preservation is at an interesting point. External data archiving specialists with recent experience of recruiting in this area, indicate that this field is seen by computer professionals as an extremely interesting one, which is likely to be ‘solved’ in 2-3 years, and therefore attractive to be involved with now. With a sufficiently large critical mass of in-house people with knowledge of digital preservation issues, this external body was successfully recruiting general computing professionals and training them in-house in a few months. Those who already had knowledge of these issues, tended to have gained such experience from seismic survey work, oil climate modelling, weather prediction and other areas of long-term data modelling.

  The issue for the BL is whether there is that sufficiently large critical mass of in-house people to train and develop others within the organisation. Margaret Hedstrom comments that ‘interestingly, those [organisations] with the highest levels of staff expertise are most likely to acquire expertise from outside sources.’¹⁰

  As for immediate availability of these competencies, there are a growing number of relevant research projects and a small number of dedicated digital preservation posts are being created (see appendix II for some recent posts). In the medium-term, an MPhil course in ‘Digital Management and Preservation’ has just started in Glasgow (HATII). Other national libraries are tackling these issues and have documented their experiences well.

  Within the BL, there are new jobs emerging but the majority are developing in post. There are secondments, for example from IT to the curatorial Special collections area, which will have the benefit of skills transference. ‘Home grown’ is a major development issue, because this subject will increasingly have an impact on many people. Training ‘by doing’ sears the knowledge in, but it is uncomfortable and it is time-consuming. Training on a corporate scale takes time. Two areas have bubbled up across the Library recently, namely digital preservation and digitisation training. It will take a year to design and carry out the training.

• Not from a to b, but a to a to ß to b
  It is clear that the BL will not travel from one way of working to another in one leap. Step one, entails one or two individuals highlighting an issue such as digital preservation. Step two, entails learn-as-you-go, project-based work, scooping people in. Step three, entails moving towards a limited number of distinct new jobs, whilst addressing how developments affect more and more people. In a smaller institution, one bold step might be possible – in a large institution, it has to be more evolutionary.
• Be ever more consultative internally and ever more open and outward facing
  The use of wide-scale consultation within the BL via the functional working groups was deemed a successful element of the Digital Storage Project. The BL has opened itself to being studied, for example it is the major case history for the LIC-funded research into ‘Management of Digital Preservation’¹¹ and it became a test site for CEDARS preservation metadata when this was incorporated into DLS specification. The BL is strengthening collaboration with other bodies, for example discussions are underway with LOCKSS, Koniklijke Bibliotheek including specifically on digital preservation, and JISC Digital Preservation Focus, on projects such as, possibly, one on web archiving.

The draft Digital Preservation Policy states that the British Library ‘will build on the work of other comparable organisations engaged in the care of national electronic written and documentary heritage. It will work collaboratively, both nationally and internationally, and with different library and archive sectors, in the development and implementation of its preservation strategy for digital material. It will take a lead where appropriate and play a junior partner where appropriate. The British Library will continue to open itself to being studied whilst addressing this relatively new area’.

Acknowledgements
Kate Streatfield
Michael Alexander
Neil Smith
David Inglis
Dennis Pilling
Janice Swain
Vanessa Marshall
Maggie Jones
Lynne Chivers

Appendix I
Extracts from ‘What are the competencies the British Library needs in order to address the preservation of digital material; where and how can they be acquired? April 2000.

1. Types of competencies
The types of competencies can be divided into three broad categories (as defined in PRO Records Management-Human Resources; standards for the management of Government records)¹²

• Core competencies
  Competencies relating to the preservation of digital material strategic priorities
• Specific functional competencies
  Role specific competencies and abilities, related to professional or technical skills
• Managerial competencies
  Competencies reflecting managerial activities to ensure the continued existence of digital material

Who needs them within BL
Different groups need the competencies throughout the BL

• Within Preservation
• Within Collection Management
• Within BL, eg Collections areas, IS, Reader Services, Reprographics etc

Competencies required within the BL
Core - Competencies relating to the preservation of digital material strategic priorities, supporting the Library’s strategic priority ‘to embrace the Digital Age’ (BL Corporate Plan 1999-2002).

• General awareness of digital preservation
• Preservation of born digital material
• Preservation and conservation implications of digitisation programme
• Hybrid approach to surrogacy
• Technology watch

Specific Functional - Role specific competencies & abilities, related to professional or technical skills

• 'Empowered' user for DLS implementation
• Preservation technician for DLS
• Expert technical/developmental with links to externals
• Conservation of the physical item - eg CD-ROM, longevity of microenvironment etc

Managerial - Competencies reflecting managerial activities to ensure the continued existence of digital material

• Policy and strategy
• Long term planning
• Project management
• Interface with other BL change projects
• Preservation strategy and policy ‘watch’

Appendix II
Examples of recent job adverts

Harvard University Library, ‘Co-ordinator of digital archiving and preservation’

Stanford University Libraries and Academic Information Resources, ‘Head, Media Preservation’

Library of Congress, ‘Life Cycle Management of Digital Data’ [Life Cycle Management and Digital Preservation]

Joint Information Systems Committee, ‘Digital Preservation Focus’

British Library ‘Digital Preservation Co-ordinator’

References to come.