by Professor Jennifer Richards, Newcastle University
With thanks to Kate Court, James Cummings, Fiona Galston, Kathryn Garside, Ian Johnson, Nick Holliman, Tiago Sousa Garcia, Mark Turner, Paul Watson (Newcastle University)
What will the future of the digital book look like? In 2014, Johanna Drucker, Professor of Bibliographical Studies at the University of California, Los Angeles, and a pioneer of digital humanities, proposed it will ‘be fluid [...] a richly networked portal, organized along lines of enquiry in which primary source material, secondary interpretations, witnesses and evidence, are all available, incorporated, made accessible for use’. What was remarkable about Drucker’s vision for the book of the future is that it was inspired by the organisation of the books – manuscript and print – of the past. At Newcastle University, we thought we could take this idea one step further, to think about how the digital books of the future might also recover – and be shaped by – both the lives and liveliness of the material books of the past. That idea was the foundation for Animating Text Newcastle University (ATNU), a collaboration between text-based humanities scholars, computing scientists and software engineers. In this contribution to ViewFinder, I will tell you about what we made – and are making – as well as what we learned. The story I have to tell is about learning to work with and listen to each other in new ways, about collaboration and co-creation, about how humanities can and should engage with AI, and about imagining multi-disciplinary futures.
But let me begin at the beginning. The year was 2015, and the REF2014 results had been announced. English literature and creative writing at Newcastle University had done well. Colleagues were doing all sorts of exciting work, with strengths in four areas: creative writing, children’s literature, postcolonial literature, and scholarly editing. Scholarly editing, the production of high-quality texts for further study, already had a venerable tradition at Newcastle: we had been home to two of the great Shakespeare editors of the 20th century: Peter Ure and E.A.J. Honigmann. In 2013, two new editions were published: Tom Cain’s and Ruth Connolly’s The Complete Poetry of Robert Herrick, in 2 volumes, and Michael Rossington’s contribution to The Poems of Shelley, Volume 4: 1820-1821. I was about to join this distinguished group. I had been awarded AHRC funding for the collaborative Thomas Nashe Project, editing the works of this contemporary of Shakespeare for Oxford University Press.
All of the editions mentioned are hefty volumes, full of detailed information – manuscript and print histories, and textual annotation – making the texts accessible for modern scholars. They are usually published in hardback by University presses who can provide and pay for professional copy-editing at scale, and converted into digital format; they are usually sold in multi-volume sets to university libraries in small numbers – these are prestigious, field-changing rather than profit-making projects. Nashe, for example, will appear in 6 volumes. In his case, however, I am very much aware of the oddity of this: after all, Nashe wrote mainly short, cheap, provocative prose pamphlets; his writing is energetic, fleet-of-foot, and vocal, and he experiments with the lively, even 3D possibilities of the still new media technology of his day: print. I imagine him shuddering at the very thought of a hefty scholarly edition.
Alongside this new work, I was already developing a parallel interest in ‘digital text’, and the possibilities of the digital medium for early modern books, which usually exist in multiple versions, and/or, like Nashe’s works, are playful with the medium in which they are made. I would probably have taken this interest no further if not for a chance conversation with Professor Paul Watson, then Director of the University’s Digital Institute, now the Director of the National Innovation Centre for Data, on a train journey back to Newcastle from London following a UKRI interdisciplinary workshop. Unsurprisingly, we talked about our work. Paul was amazed by the sheer amount of dull, mechanical labour the collation of texts for an edition involves, and we talked about whether the fast-developing field of AI might offer a solution. Surely there must be a better way, he asked, to identify variants in copies of printed books to understand the relationship between them? The answer to this question was and currently still is ‘no’. Even the Traherne Digital Collator, developed by physicists in Oxford with EPSRC funding, which uses the MATLAB mathematical programming language and environment, proved too time consuming to set up for our poor-quality printed texts, and so I still prefer the traditional, so-called ‘Wimbledon method’, collating by eye, word by word. At the very least that slow method gives me a chance to think about what I am seeing.
What did develop, though, was a multi-disciplinary collaboration, ATNU, which set out to explore what the digital can offer traditional, text-based scholarship, while also creating a cross-university culture open to the possibilities of the digital humanities. Thanks to University funding we were able to appoint two new colleagues, James Cummings, previously the Senior Academic Research Technology Specialist in the Research Support Team at the University of Oxford, and a world expert in text encoding, and Tiago Sousa Garcia, our Research Associate. Tiago had studied computer engineering in Portugal and worked as a software developer, before switching to write a PhD on early modern literature and translation at the Universities of Porto and Kent. He became our essential cross-disciplinary translator. Once we added Mark Turner, the head of the University’s Research Software Engineering (RSE) team, RSEs Kate Court and Fiona Galston, as well as Nick Holliman, Professor of Visualization, and Ian Johnson, Head of Special Collections and Archives, the ATNU team was complete and ready to go.
From 2017 to 2021 James, Tiago and Mark’s team ran pilots with colleagues in the humanities to represent, visualise and/or animate the dynamic ways texts work. Because the pilots were exploratory, they were as interested in what didn’t work as they were in what did. Together, they built a dynamic map to track the movement of the translations of Mary Wollstonecraft’s writings across the shifting borders of 18th-century war-torn Europe; re-interpreted a (verbose) description of an imagined new voting system in a 17th-century political pamphlet, James Harrington's The Use and Manner of the Ballot, as an algorithm in order to bring to life the accompanying illustration; explored ways of enabling the comparison of multiple textual witnesses, using ‘Must We Go?’, a multi-part poem by the Pakistani-British poet Moniza Alvi, as a case study; started to map the transatlantic networks of one of Romanticism's most important nodes in the exchange of literary and scientific correspondence, David Bailie Warden (1772-1845). In addition, ATNU supported professional development, with James and Tiago offering training in TEI. Tiago convened the Coding for Humanists study group, and now co-convenes Code Community at Newcastle University, while James designed the British Academy-funded JUMPSTART, workshops offering humanities scholars an introduction to coding and machine learning, and I championed the rolling out of the National Innovation Centre for Data’s “Data Innovation Bootcamp” to students in arts, humanities and social sciences, giving them the same opportunity as their peers in the Faculty of Science, Agriculture and Engineering, to learn basic data analysis, and use their disciplinary skills to solve business problems and opportunities shared with them by industry partners as diverse as Opera North and Siemens.
What now in 2023? ATNU continues under the direction of James, now a Reader in the School of English Literature, Language, and Linguistics, leading the Evolving Hands project, overseeing HTR to TEI workflows, and a member of the LEAF-VRE project, creating LEAF-Writer, which gives easy tags-off methods to add editorial and named entity Linked Open Data to TEI files. Tiago has moved to the RSE team, where he is a front-end web developer. The RSE team has expanded. Uniquely in the increasingly STEM-focussed world of HE, they have a dedicated digital humanities interest group. This is a very welcome, even crucial development, I’d argue, as digital experimentation becomes the norm in the arts and humanities; and it should give Newcastle a head start in addressing the challenges of AI. I have two ATNU examples to share to support this bold claim. The first example, Tiago’s and RSE Kathryn Garside’s Creativity Engine, provides an alternative to a new potential challenge, ChatGPT, reminding us what is needed if we are to create a safe environment for AI-supported learning in the digital age. The Creativity Engine, which launches this month, uses a similar generative AI as ChatGPT, but is fine-tuned on a more carefully curated archive than the worldwide web: the words kindly donated to us by the ethnically-diverse children’s authors whose works are held by Seven Stories, The National Centre for Children’s Books. This writing tool was developed with the young people and teachers it aims to support at four Further Eduation colleges in England, thanks to the help of Rebecca Fisher and Elizabeth Draper of The English Association. The second example is more blue skies, but the challenge it aims to address is real enough: the current lop-sided, ‘sci-fi’ style debate about AI, which forgets that we will need collaboration between artificial and human intelligence, and between machine and embodied learning to solve future problems, and that we cannot achieve this while politicians fail to value SHAPE subjects, and when our Universities remain so siloed. This is the challenge the Leverhulme Trust-funded Bee-ing Human project takes up.
The Bee-ing Human project is not an AI project, although its unifying output – an open access book – is born-digital. It involves a multi-disciplinary team that includes literary scholars (Jennifer Richards, Olivia Smith), musicologists and musicians (Bennett Hogg, Magnus Williamson), animal scientists (Balumurali G.S., Vivek Nityananda) and digital humanists (James Cummings, Tiago Sousa Garcia) exploring how bee behaviour has been understood in the past, applying methods developed to understand human cognition to explore the possible emotion-like states of social bees, and using sound to explore the knowledge of beekeepers. But because it involves the original ATNU team, who use a book from the past, Charles Butler’s The Feminine Monarchie or the History of Bees (1609, 1623, 1634), to build a digital book for the future, the thinking I outlined above is never far away.
Butler, a keen apiarist, who was trained as a humanist, and who published books on rhetoric and music, was one of the first to confirm the sex of the dominant honeybee in a hive. He also realised as a result of his research that he needed to understand the social (and emotional) behaviour of honeybees differently. He left his study for the field, and, armed with a wind instrument, listened attentively to the hive, notating what he heard – the tooting and quacking of the queen and the virgin queens – and then he recreated what he heard in print, turning an otherwise traditional manual on bee-keeping into a buzzing hive we enter. We are creating a born-digital edition of Butler’s extraordinary, animatable, interactive book to understand the evolution of his thinking across three editions and his notation of bee sounds, and we are working with scientists to update the conclusions he drew; we are also using his book as a frame for our project, creating a 21st-century digital book in which scientists and arts and humanities researchers and beekeepers and the general public can enter to advance their understanding of our threatened pollinators. And, as we do this, we are also thinking about what the digital books of the future can be: fluid, richly networked portals certainly, as Drucker argued, but also living, evolving artefacts and immersive learning communities – exactly as Butler wanted his printed book to be – as well as the product of collaboration between artificial and human, and just possibly, perhaps, animal intelligences.
You can find out more about ATNU’s work on ourwebsite.
About the Author
Jennifer Richards is the Joseph Cowen Chair of English Literature at Newcastle University, and the Co-Director of Newcastle University Humanities Research Institute. Her study, Voices and Books in the English Renaissance: A New History of Reading (Oxford University Press, 2019), won the 2020 European Society for the Study of English (ESSE) biennial award for Literatures in the English Language. She was the co-lead of the digital humanities Animating Text Newcastle University project, and is the lead of Bee-ing Human: An Interactive Book for the 21st Century (funded by The Leverhulme Trust).