"It is almost as if we were asking that a building be created with structural integrity, yet be capable of accommodating a varied multiplicity of spatial requirements in the course of its existence."

The role of a machine based authority file in an automated bibliographic system / S. Michael Malinconico, 1975.

In addition to the fact that the two library warehouses in this study hold rare and valuable goods, in terms of process they are doubly atypical, as warehouses, in that any item going out will at some point come back. Thus there is no ‘turnover’ of stock. Helen Andrews commented that, in watching the containers accelerated away into the dark depths of the void, one can easily imagine that some are going in, never to return. The BSF and ASB were planned recently, at a time when it was reasonable to think that the production of printed matter would, in the foreseeable future, decline. The British Library’s Pat Oddy, writing in 1996, quoted “A recent poll carried out for the BBC’s ‘Bookworm’ programme [which] discovered that two out of three people believe that books will be obsolete by the year 2010” (p.vii). This was long before the eBook, indeed, right at the beginning of the UK public’s introduction to the World Wide Web. Perhaps it was the related excitement that skewed the results. Either way, the expected decline is still expected. As speculated in the introduction to this study, it could well be that the BSF and ASB represent the end of a great period of library expansion in the West, preserving information carriers produced by a technology, and in a format, that changed the world but whose time is up; if not now, at least within the lifespan of those living today.1

The next big question, which is already being asked, is which has the greater likelihood of being readable in 500 years’ time – the last paper book to be ingested into a library warehouse or its eBook or scanned version, now readily available online but with no mechanism for its preservation? And might it even take an archaeologist to discover the printed edition, as Helen Andrews playfully feared? After all, it has been demonstrated that the library warehouse is dependant for its operation on computer hardware and software, which is itself surely potentially as vulnerable in the longer term as online access to books.

Michael Williams’s story about some of the Bodleian’s librarians being upset at the dispersal of their collections within the BSF is equally instructive. The catalogue, or LMS/WMS, is now the unifying factor, not the shelf arrangement. Browsing is online, via metadata2, and storage is freed from the need for collocation. This means that finding is more than ever reliant on metadata and, in the case of the ASB, robots – subject to the algorithm that tells them where to place each tray of books. Fascinatingly, and unlike at the BSF, where once a shelf location for a particular tray is chosen it is expected to be kept in perpetuity, at the ASB the robots should, in time, replace the most accessed trays in the foremost bays. As the photographs in Appendix B show, these are still empty after three years of operation. Perhaps those location-specific British Library shelfmarks have a hold on the books after all.

This study has shown that the successive book floods of the modern era – defined as 1500 to the present day – have significantly changed the way libraries are designed, and the way they work. It has also found continuities, and real similarities between 21st-century and early-modern methods. Barring disasters, the internet will become more and more embedded in daily life. Progress towards a semantic web of distributed data linked by embedded metadata that allows computers to manipulate and reformulate it, not only offers the promise of efficient and robust catalogues; it also suggests that cataloguing may finally be truly automated. However, linked data is a way of making information processable. The work of preparing the metadata is likely to remain a human activity for the foreseeable future. But, as the pioneers of card-based excerpting and cataloguing realised, once the initial work is done the possibilities of how that information can be combined and reused expand. Although their readers might not realise it, many webpages today have their content collected automatically thanks to these linked data techniques in a manner very similar to the excerpting and composition techniques described in this study.

Let us hope for an environmentally and technologically sustainable future in which the library buildings that accommodate and perpetuate books are as fluid and dynamic as their organising catalogue systems are stable. After all, it is such buildings and information processing systems that have successfully brought us to this point.



1. This study has not considered the advance of printing technology beyond moveable type up to the digital systems of today, which make it possible for the production of an individual book (ie, print on demand) to be commercially viable. Such technology is of particular value to libraries. The assumption, however, is that, despite the screen-based options for reading available today, printing is easier and cheaper than ever. back

2. SOLO has a ‘BROWSE’ feature that attempts to reproduce the act of looking along a shelf for related works. back