1. Introduction
The launch of the Journal of Computational Literary Studies’ first issue is a big deal for everyone involved. We’re happy this has finally happened; we’re proud of our authors, editorial board, reviewers, editorial team, and infrastructure partners; we’re excited to find out how the community will receive this first issue and the journal; and we’re optimistic that JCLS will find its place within the broader and currently quite dynamic Digital Humanities journal landscape.
What makes us optimistic are the contributions we were able to attract for this first issue: They demonstrate a rich and productive breadth and variety of research questions, approaches and objects of study currently in play in Computational Literary Studies. There are papers investigating, for instance, the reception of literary works, concerns of gender and ‘queerness’, issues of knowledge transfer in drama, eventfulness in narrative, similarity of short texts, alignment of translations or limerick generation. There are papers using quantitative approaches, for instance topic modeling, keyness or word embeddings, but there are just as well papers relying on rule-based approaches or qualitative annotation schemes. Some papers are applications of computational methods for the investigation of well-established literary concerns, such as the difference between Realism and Modernism, but there are also many papers investigating new methods or providing a benchmark dataset for Computational Literary Studies. There are contributions on translations from ancient Greek, on French and Spanish Enlightenment-period periodicals, German 19th-century drama, English or German modernist novels or contemporary British and French literature. In short, the diversity of research questions, approaches and objects of study is impressive.
With this first step taken, we feel like it is a good moment for a short recap and some reflection.
2. Some background
Even on the day of the publication of our first issue, JCSL already has a history. Just for the record, we would like to offer a brief summary of it here.
In October 2020, Fotis Jannidis approached us with the idea to launch an annual CLS conference that would regularly publish its proceedings, with support from the DFG Priority Programme Computational Literary Studies (SPP 2207). We thought of it as a way to support the establishment of best practices in CLS research, consolidate and expand the CLS community, and make CLS research visible as a distinct field within the broader Digital Humanities. We all thought this was an exciting idea and very quickly agreed to try and do this together. Our first online meeting took place on November 13, 2020. It quickly emerged that what we really wanted to do was to found a journal with an annual conference attached to it. We agreed that the journal should be “Open Access” and committed to the idea and ideals of “Open Science”, that it should use state-of-the-art editorial processes and publication technology, and that it should be open to a wide range of research in CLS, with a decidedly international scope.
It took us several months to develop this idea further, for instance, designing the dual track mode, with a journal+conference track and a journal-only track and their respective timelines from submission to publication. We also weighted the pros and cons of various technical solutions before settling on the combination of an online, collaborative editor using LaTeX (first ShareLaTeX, then Overleaf) with the journal management and publication system Janeway (open source software developed by the Open Library of the Humanities). We found an excellent infrastructure partner, the University and State Library Darmstadt, where our Janeway instance is hosted, where some of the format conversion work is coordinated, and whose team (Anne-Christine Günther, Matthias Kerekes) provided and still provides invaluable support. We learned more about Janeway from the wonderful folks at the Open Library of Humanities (most memorably, Andy Byers). Last but not least, we established a distributed team of editorial assistants at our own departments as well as an international, knowledgeable and diverse editorial board. The latter, once constituted, supported us in working out further aspects of the journal, such as the peer reviewing process.
A series of monthly meetings later, it was summer 2021 and we had at the ready: An awesome logo, a first version of our LaTeX template (thanks to Dominik Gerstorfer), a mission statement and the text of our first call for papers. In addition, we had a commitment from Evelyn Gius to host the first annual conference at Technical University Darmstadt, a website and a Twitter account (now supplemented by a Mastodon account). In short, we were ready to go! So on July 8, 2021, the first Call for Papers went live (see Figure 1). The response to the call was very good and, after a first round of reviews, 12 papers were presented at the First Annual Conference of Computational Literary Studies (#CCLS2022) in June 2022. A hefty dose of revisions to papers, reviewing of the results, improvements of the workflows and refinement of the LaTeX template ensued, finally leading us to November 24, 2022, when the first three articles in the inaugural issue of JCLS were published. We managed to publish all 12 articles of the first issue before the end of the year.
Figure 1: JCLS announces its first call for paper on Twitter. See https://twitter.com/jcls_io/status/1413121051832225799 and https://web.archive.org/web/20210708130226/https://twitter.com/jcls_io/status/1413121051832225799.
3. Where we stand today
First of all, we believe in the mission of JCLS to be both a venue where a broad variety of research in CLS can be published and read, and a space for the community of researchers in CLS to meet and share ideas. We are glad that, thanks to our infrastructure partner and our distributed team of editorial assistants, we are able to publish JCLS as a Diamond Open Access journal, meaning that all articles are free to publish, read, download, mine, translate and transform without cost to neither authors nor readers. The fact that we can build on Janeway and have a pro-bono infrastructure partner also means we can do all of this as an entirely scholar-led journal, independently from a traditional publishing house. We are convinced this is an important, even if small, step towards taking back the scholarly publishing process from large for-profit, data-hungry corporations.
It is equally important to us that we are able to publish this journal in a state-of-the-art manner, with articles being made available both in a reader-friendly PDF format as well as in HTML (for in-browser reading) and XML-JATS (for long-term archiving and text mining purposes). Another aspect of this is the inclusion of authority data for authors and reviewers (via ORCID), but also for institutional affiliations (via ROR) and author contributions (via CRediT). We’re proud as well that many of the authors contributing to this first issue have followed our call to make their code and data openly available, for instance using a Github repository. Finally, we think that the combination of a journal with an annual conference has already proven its worth, because it brought together a part of the CLS community to meet and discuss their work and in this way provided an occasion for additional collaborative, discursive and open peer review. We’re looking forward to continue and expand these practices.
4. Our plans for the future
However, the past two years of planning the journal and putting together the first issue have also taught us innumerable things about the ways in which our current procedures and technical workflow could be improved. We are keenly aware, in addition, that much is happening in the world of scholarly publishing with respect e.g. to single-source publishing workflows as well as to reproducibility and executable papers; developments that we embrace but have learned to approach with some humility.
One of our most ambitious aims is in fact something that would hardly be noticed by readers, but all the more so by authors and editors, because it concerns the way in which texts proceed through the reviewing, revision, copyediting and typesetting pipeline. Only a few of the many steps currently happen online: In the reviewing phase, for instance, reviewers have the option of (anonymously) making comments and suggestions in the Overleaf document containing the article, where authors can see them and revise their text accordingly, right in the online editor. Similarly, the copyediting phase also happens online, in the Overleaf document. However, the entire pipeline still involves multiple steps of uploading and downloading files into and from Janeway, starting with the submission, the upload of a revised version, or the upload of the final copyedited version.
As a consequence, a stronger integration of the online editor (not necessarily Overleaf, but possibly also Fidus Writer or Quarto, among several options) with the submission management system (Janeway) is high on our agenda of improvements to our pipeline. This should help avoid some time-consuming and error-prone file management steps and, coupled with a Github integration, will also enable us to make batch-updates e.g. to the template or the BibTex files. It would also be a welcome occasion to take Overleaf, a for-profit entity, out of our workflow. It is, however, not a small problem to solve, because it would require Janeway to modify its article processing workflow quite substantially.
Another plan is no less ambitious than the first, in part because it involves substantial modifications in our production and output data formats and the technical infrastructure that supports them. This plan is our desire to better support the integration of executable code in the papers published in JCLS. Without necessarily making them fully executable papers, we would very much like to support the integration of code segments that accept a dataset (e.g. in the form of a table), apply some calculations to it, and then create a visualization of the results, all within the paper itself. Readers would not only be able to closely understand how the data is analyzed, but would also be able to modify the code and see how this affects the resulting visualization. In this configuration, the current in-browser view would become much more central as an interactive reading environment than it currently is, while the PDF version would become a secondary, static, offline reading format.
Some more modest adjustments to the current publication processes are in the works as well. For example, we plan to support authors better in preparing properly-structured bibliographic data so that the BibTex files can be published as supplementary files for copying, scraping or citation analysis. Similarly, the next round of submitting authors will have the possibility to opt for a specific, additional reviewing step focused on the data and code accompanying a paper (see our information on the Code and Data Review). The aim here is to make sure papers present truly reproducible research, something that also includes concerns for copyright issues when publishing literary datasets and the long-term archiving of code and execution environments. Finally, we are currently discussing whether we could not offer, in addition to the double-blind pre-publication review model, a track where submitted papers are published as a preprint directly after a formal editorial review, but before going into a post-publication, single-blind peer review. This would greatly speed up availability of the latest research while awaiting the results of the formal peer-review process.
5. Conclusion
In conclusion, we are both happy, grateful and proud about what JCLS has become within two years, and excited about what it can become in the next, say, twenty years. We hope that, with feedback and advice from our editorial team, the editorial board members and the CLS community at large, we will be able to shape this journal in a way that it responds to the needs of the community as best it can.
6. Author Contributions
Christof Schöch: Conceptualization, Writing – original draft, Writing – review & editing
Peer Trilcke: Conceptualization, Writing – original draft, Writing – review & editing
Evelyn Gius: Conceptualization, Writing – original draft, Writing – review & editing