With the first phase of our Project at an end and our second phase now well underway, it seems an appropriate moment to look back at our work thus far on EMLO and to return to the dataset that lies at its core: the ‘Index of Literary Correspondence’ in the Bodleian Library, a card catalogue which occupies an imposing set of wooden filing drawers at the ‘Selden End’ of the Duke Humfrey’s Reading Room.

This remarkable free-standing resource, describing a significant percentage of Bodley’s rich holdings of sixteenth-, seventeenth-, and eighteenth-century correspondence, had been accessible until recently only to those working on-site; as per our brief, it has been scanned and digitised, and now its 48,691 unique records can be searched and browsed online within EMLO, radically improving the discoverability and manipulability of the cards and the letters to which they refer. However, until now, we have known nothing about the creation of the analogue card index beyond a farrago of hearsay, local anecdote, and library lore; those who compiled the records, and the intellectual principles on which they, and their supervisors, operated, were proving elusive. This frustrated both our attempts to understand and appreciate the overall shape of the catalogue in terms of what is and is not included, as well as our efforts to credit those epistolary Stakhanovites on whose painstaking work our digital effort has depended.

We are pleased, therefore, to be able to  share a near-complete narrative of the genesis and evolution of the card index, based on completely fresh archival research.[1. For leads and information on the individual cataloguers, I would like to thank in particular Bruce Barker-Benfield, Mary Clapinson, Wilma Minty, and David Vaisey; had it not been for a tip from Bruce, we might never have known of the existence of their working notebook. Thanks are due also to Colin Harris and his team of helpful staff in the Bodleian’s Special Collections, in particular to Vicky Saywell, and to Project colleagues past and present — Philip Beeley, Howard Hotson, Kim McLean-Fiander, Leigh Penman, and Richard Sharpe — for their input, expertise, and comments on drafts. Our Digital Project Manager James Brown supervised my research and helped shape the piece, especially the introductory and concluding sections on library history. Every archivist I have approached regarding our cataloguers has provided efficient and informed assistance: thanks especially to Sian Astill at the Oxford University Archives; Suzanne Foster, archivist of Winchester College; Katherine Phillips at the Imperial War Museum, London; and Anna Sander, archivist of Balliol College, Oxford. Most of all, I would like to thank Professor Julian Hunt for responding to an email which arrived out of the blue and for providing a wealth of information about his grandfather — without him it might not have been possible to identify for certain the man who has made the greatest contribution of all — and Christopher Legge for sharing his childhood and family memories.] The pursuit of such questions is not only of local, antiquarian, or Project interest; the recent archival (or technical) turn across the humanities and social sciences, now manifesting in conferences and research networks, has established repositories not only as sites for research but also as ‘fascinating objects of study’ in their own right.[2. Randolph C. Head, ‘Preface: historical research on archives and knowledge cultures: an interdisciplinary wave’, Archival Science, vol. 10, issue 3 (Sept. 2010), pp. 191–4.] It is recognised now that the ‘material conditions, infrastructures, and mediations’ of archives have a constitutive role in knowledge production,[3. See Ben Kafka, ‘From the Desk of Roland Barthes’, published online: 10 February 2011.] with cataloguing practices in particular shedding light on the intellectual ambitions of organisation within particular institutional contexts, and influencing the discoverability of documentary objects, and — in turn — the kinds of histories which ultimately can be made.

Indeed, the card catalogue itself — direct descendent of ‘Harrison’s Indexes’ (as described by intelligencer Samuel Hartlib)[4. Noel Malcolm, ‘Thomas Harrison and his “Ark of Studies”: An Episode in the History of the Organization of Knowledge’, The Seventeenth Century, 19, no. 2, pp. 196–232.] and linchpin of repositorial efforts to marshal ‘stuff about stuff’ since the early twentieth century — has recently been given monographic treatment in Marcus Krajewski’s Paper Machines, an excellent sociology of card index technology which traces its journey from scholarly experiments in early modern Europe to the libraries and archives of the twentieth century via the French Revolution and corporate America.[2. Markus Krajewski, Paper Machines: About Cards and Catalogs, 1548–1929 (Cambridge, MA, 2011).] Krajewski figures these durable workhorses as ‘universal discrete machines’, [3. Krajewski, p. 3.] but they were of course populated by human hands within specific archival ecosystems. By cross-referencing curatorial documents with the Bodleian’s reports and records, we have reconstructed these processes for the Library’s epistolary card index, and are able at last to tell this catalogue’s story while contributing a small footnote to the history of information science.

The creation of the ‘Index of Literary Correspondence’ took place across three distinct phases, falling between 1927 and 1963, and mapping on closely in each case to changes in senior management within the Library.

These can be described in turn:

A blog post is not the place to elaborate a full range of conclusions emerging from this brief case study — although a second installement will follow in which I intend to consider exactly what was catalogued when, as well as which manuscripts were not selected for inclusion — but what is perhaps most interesting about the creation of the Bodleian’s Index of Literary Correspondence is its ad hoc, iterative, and almost casual nature. It was compiled, piecemeal, by enthusiastic and talented amateurs and its progress over thirty-five years was characterised by false starts, abrupt changes of direction, moments of disruption and caesura, and luck. Whilst tied intimately to changes within the Library prefecture, supervision of the project appears generally to have been light; there was no attempt to standardise ontologies between each of the three phases, and it seems that even those Keepers who took the liveliest interest in the enterprise (Craster and Richard Hunt) may have guided the indexers orally rather than providing extensive formal written instructions and schedules for data-capture.

Inevitably this has implications for the functioning of our particular paper machine, and those who have explored the index either in situ or in the context of EMLO will have encountered idiosyncracies that would not exist in more ruthlessly professionalised resources. It is also highly selective; the ‘euphoria of totality’ engendered by card catalogues is misleading in this case, as a significant number of the Bodleian’s epistolary guardbooks were not included, reflecting the intellectual predilections of both indexers and directors (as in the initial focus on Hearne). In all these senses, the index perhaps bears (in Krajewski’s phrase) closer relation to a subjective and personally generated ‘scholar’s machine’ (or learned excerpt collection) than a conventional institutional catalogue (the latter a sort of universal search engine ‘designed to register everything randomly’), problematising this hard and fast distinction between the two kinds of resource.[35. Krajewski, p. 50.]

These eccentricities notwithstanding, the Index of Literary Correspondence is a sensationally interesting and useful resource, and — in line with the general electrification of libraries — lives on in its new incarnation within the digital realm of EMLO, forming the core around which our virtual machine will grow. Not only are its records now freely available to users worldwide and more siftable and discoverable than ever before — the mobile carriers of the card catalogue find their logical extension in the online relational database — but we have used EMLO’s powerful editorial environment (much more flexible than anything enjoyed by our predecessors) to link them to an ever-wider range of cognate records and resources, and to deal sympathetically with some of the peculiarities of the catalogue likely to be most frustrating to the modern user, in particular the wildly divergent expression of proper nouns (which we have standardised and de-duplicated using our merging toolset).

Alongside these new digital impulses, we have tried to preserve the integrity of the paper machine, not least by reproducing in the front-end scanned images of the original index cards alongside the cleansed metadata. In a very real sense, the Project’s digital work on the index since 2010 has been a triangulation between ourselves, the senders and recipients of the letters being catalogued, and — it turns out — just three indexers, to whom now we are delighted to be able to give names and stories.