Behind the scenes Collections Long read Series

Palladium: Research Potential of the Carcanet Email Archive

Research in the Carcanet Email Archive

Paul Carlyle writes:

Regular readers of this blog will know that since the beginning of the year we have been working hard on a project called Palladium (Providing Access to Large Literary Archives in a Digital Medium). For this project we are looking at the email archive of the Carcanet Press, the celebrated Manchester-based literary publisher, and thinking about how we accession emails, manage the archive’s long-term preservation, provide access and support research. (If you are new to the project or in need of a reminder, please see our earlier blog posts: ‘Palladium: a new project begins at the Rylands’ and ‘Palladium: appraisal and sensitivity review of the Carcanet email archive’.)

Carcanet’s email archive is part of a much larger collection. Carcanet first deposited its records at the Library in 1978 and has continued to do so ever since. The collection as a whole has much to tell us not only about Carcanet and its poets but also literary culture in the twentieth and early twenty-first centuries. It is a rich resource for anyone looking to understand, among many other things, the emergence of various literary debates and Carcanet’s participation in them (through its activities as publisher as well is in the pages of its long-standing independent journal PN Review); the evolution of publishing over several decades; the influence of arts organisations and cultural programmes; and trends in academic research.

Our changing means of communication are also well documented in the archive. The shift from letters to emails is the clearest and most abundant example. Letters are still sent and received, of course, and correspondence is also conducted digitally by means other than email, but email remains for many the dominant and most persistent form of professional communication. Michael Schmidt, founder and Publisher/Managing Director of Carcanet, reflecting on his own letter-writing habits, writes, ‘An epistolary habit is hard to shake, even when e-mail comes along. Indeed, e-mail can make the habit an addiction’ (Fifty Fifty: Carcanet’s Jubilee in Letters (2019)). Between 2001 and 2019 Carcanet sent or received several hundred thousand emails, and we noted in a previous blog how the number of emails sent globally, a frightening number now, is likely to increase in the next few years. For archivists this presents many serious challenges but also opportunities.

Researchers in the digital humanities have been applying computational methods and related developments, such as machine-learning and data science, to traditional humanities subjects. Digital humanities departments and courses have become well-established, with researchers and students from different disciplines collaborating on projects which explore archive, library and museum collections in new ways. At the University of Manchester, researchers from across the five Schools of the Faculty of Humanities have used digital technologies on various large-scale projects, including the Mary Hamilton papers and the Manchester Together Archive.

We are looking at how methods used in the digital humanities, such as data visualization, could be applied to the Carcanet email archive. The Palladium project began earlier this year but it builds on the award-winning email preservation work of colleagues at the Library from a decade ago, who experimented with data visualization tools on information extracted from Carcanet’s emails. (See here for a summary of those experiments and screenshots of various charts and network diagrams.) They found that data visualization tools could be used to present information from large datasets in different ways and to different audiences; that the ability to anonymise information potentially means the emails could be made accessible without releasing messages in their entirety (which otherwise might be restricted, for example, on data protection or copyright grounds); and that new areas of research might be revealed. We have spoken to academics who have specific research questions in mind, and which we intend to investigate: the frequency of contact between Carcanet and its poets; how often different poets (published by Carcanet or not) are mentioned; how often individual poems or collections are mentioned; Carcanet’s engagement with poetry in different countries; how marked the influence of particular poets or editors was over an extended period of time. There will no doubt be more.

The main tool we are using to manage the email archive is ePADD, email preservation software developed and maintained by Stanford. ePADD allows us to extract data from the email archive and present it in visual form. It has some basic visualization tools that can be used to create graphs showing the number of emails sent or received by individuals over set period. Here is an example, anonymised, showing the number of emails sent by different correspondents (they are colour-coded) between 2002 and 2019:

A wide range of information can be extracted from ePADD, such as correspondents, message headers (dates, subjects), lexicons (words or phrases, sentiments) and ‘entities’. ePADD identifies ‘entities’ using a form of natural language processing:

This image has an empty alt attribute; its file name is Capture-2.png

These entities include people, places and events, which can be exported to a CSV file and used with many different visualization tools. While these tools can be used on the archive as a whole, it is also possible to analyse a specific series of exchanges between, say, one poet and Carcanet or correspondence from a single year, week or day. We are planning to experiment with all of these features.

Here, data from ePADD has been used with DataWrapper to plot the geographical distribution of 478 poets published by Carcanet in the past fifty years:

It is a simple, slightly reductive example (it does not include, for instance, the many poets published in PN Review) and place of birth or residence is not necessarily an indication of the literary tradition within which a poet writes. But it shows what might be possible. The same could be done for the many editors, translators, novelists and non-fiction authors with whom Carcanet has worked in that time.

Please keep an eye on the blog for further project updates.

0 comments on “Palladium: Research Potential of the Carcanet Email Archive

Leave a Reply