A review of London Stage Database, a digital project that remediates the London Stage Information Book, directed by Mattie Burkert
London Stage Database
Mattie Burkert, University of Oregon
Kalle Westerling, The British Library
On a given night during the long 18th century (1660-1800), hundreds of spectators gathered in each of London’s playhouses. There, they took part in elaborate theatrical events: full-length plays, operas, and concerts interspersed with prologues and epilogues, afterpieces, pantomimes, songs, and dances. The London Stage Database (LSDB) contains records of nearly 52,000 such events, made up of over 117,000 individual dramatic and musical performances. Users can search, filter, and sort events based on a range of parameters, including keyword, date, theater, play title, author, and performer name. Search results and the full dataset are available to export in multiple formats.
The LSDB is at once a scholarly resource and a media archaeological experiment that addresses the crisis of project sustainability in digital humanities. It remediates and revitalizes an artifact of the early humanities computing era, the London Stage Information Bank (1970-1978), which was in turn based on a series of reference books. In 2013, project director Mattie Burkert began an archival and forensic investigation of the surviving files. The recovered data formed the basis for a relational database and website, developed from 2018 to 2019 with funding from the NEH Office of Digital Humanities. (A white paper details the process and individual team members’ contributions.) In 2020, the website moved with Burkert to the University of Oregon. There, the search system was overhauled to improve speed and accuracy; in keeping with the project’s media archeological ethos, however, the legacy search remains online.
In reconstructing incomplete and badly bit-rotten .txt files, the LSDB team minimized traditional data cleaning in favor of a more forensic approach. The underlying dataset retains the original flat-file format and elements of the pre-TEI markup schema. New fields and missing data were entered following the logic from the original project’s documentation. A copy of each record at every major stage of transformation is preserved within the updated flat-file database. Those version power an image gallery on each event page that captures key moments in the data lifecycle, foregrounding the data’s situatedness and making visible the layers of mediation that interfaces too often obscure. Likewise, the “toggle query” option on the search results page resists the tendency toward opacity in algorithmic culture. Taken together, UI elements like these invite users to engage with the database as an object of cultural inquiry and critique rather than a transparent window onto empirically verifiable phenomena.
Since it launched, the LSDB has been reviewed in ABO and Renaissance & Reformation and federated into the 18thConnect and NINES nodes of the Advanced Research Consortium (ARC). Scholarly citations reflect its uptake by humanities researchers, professors, academic librarians, and archivists. Analytics and linkbacks, however, provide a glimpse of a wider audience. To date, the site has received over 60,000 page views from more than 4,900 individuals, including K-12 teachers, genealogists, writers, data scientists, and mainstream journalists. As the team continues to develop additional content and features, the project’s reach offers an opportunity to advance a humanities-informed approach to data and algorithmic culture beyond the walls of academe.
The London Stage Database, conceived and directed by Mattie Burkert, sits in an interesting intersection between archival project and media archaeology. It also occupies an important position as one of few well-funded and large-scale digital humanities projects in the field of theatre and performance history today.
In a multi-layered history that goes back to the 1930s, the database stands on the shoulders of a theatre historical resource that details English theatre performances in the long 18th century: The London Stage, 1660-1800: A Calendar of Plays, Entertainments & Afterpieces, Together with Casts, Box-Receipts and Contemporary Comment. Compiled from the Playbills, Newspapers and Theatrical Diaries of the Period. This 8,000-page volume was first created by William B. Van Lennep, long-time curator of the Harvard Theatre Collection, in collaboration with a group of seven other theatre historians. When it was finally published as 11 volumes in the 1960s, it was widely recognized as an important work for scholars working on the period. However, it was a mastodon of information, difficult to process systematically.
Enter: The London Stage Information Bank (LSIB), a 1970s project that attempted to create a computerized index of the volumes by Lawrence University professor Ben R. Schneider. The LSIB was a well-funded project that aimed to create an easier way to navigate the copious pages of the original volumes using flat text files and a search tool written in the programming language PL/I. Yet, despite its intentions to providing a new model for research in the humanities and social sciences, its resulting files had fallen into disuse and technological obsolescence by the time the index had been created. In an essay in Digital Humanities Quarterly, Burkert has quoted Schneider’s disappointed words that the database fell unused by the theatre research community because “The kind of thing you can do now by computer is not the kind of thing that [researchers in theatre history] ever did, or felt the need to do.”1 The detailed work that had gone into creating the LSIB was meant to be stored in perpetuity on tapes preserved by Lawrence University Archives. However, in the process of the transfer to the Harvard Theatre Collection in 1983, the tapes were lost.
Enter: The London Stage Database, Burkert’s project recovering the material traces of the LSIB through a critical analysis with a strong foothold in contemporary digital humanities discussions. Burkert speculates that the positivistic orientation of the LSIB stood as an example of a dying tradition in the “cultural turn” and the poststructuralist theoretical upswing in the 1980s. Recovering the data turned out to be something of a nightmare: a gruesome story of salvaging data from antiquated disk media and deteriorating files from bit rot, and decoding hex values with custom-made Python script. All of this is detailed in the London Stage Database’s “About” page.
Burkert points out that while the 1970s database would have had difficulty representing any ambiguity that is normally present in theatre history — and a large part of theatre historian’s work — today, we have a different set of tools and technologies that can make us work better with messy data. Part of those tools and technologies aligns with Johanna Drucker’s ideas — making the interface less slick, less opaque to the user: in each entry, the database allows for the user to see the original, cleaned, and parsed “views” on the data that they read. The contrasting SQL query that generated the search result can be unfolded on the page as well, making clear the object’s transport (being served) by the PHP and the SQL servers.
In contrast to the state of the 1970s database, Burkert has called for us to “ensure that our outputs are designed to be accessible for years and decades to come.”2 Interoperability is ensured by flat, open-access file types like XML, JSON, and CSV files provided through the website’s clear and accessible button “Full Data Sets” in the main menu. That said, I had a difficult time reading the data into Pandas DataFrames or creating SQLite databases from the files provided. JSON files loaded without a problem into a Pandas DataFrame. The “Full Data Sets” page would benefit from some documentation of how the formats are structured and which document encodings are applied in the files.
Overall, the London Stage Database project is not aimed at a general audience, but researchers and teachers with special interest in the long 18th century’s performance and entertainment cultures. As such, it will prove a useful resource both for classroom and large-scale research, especially with the ability to download the datasets and run larger algorithms on local machines.
Yet, the website is clear and accessible for most imagined users. Researchers today will be familiar with interfaces to search databases, from Google Scholar to JSTOR and Project MUSE. The London Stage Database follows the familiar search prompt on the first page that already suggests some example queries for the user to try. For any users who are still in doubt, the clearly labelled User Guide in the main menu provides less of the “quick guide” that is promised and more of an extensive introduction to key terms, date conventions, and authorship in 1700s English theatre.
While the “front page” of the project provides a quick access to the data, which general audiences might find appealing, one could easily imagine an augmented version of the website with some more content developed for a general audience that could enrich the experience for those with less of a particular query in mind: subject matter expert’s comments on theatres, a map of their locations, or perhaps newspaper articles that address the performers, venues, or composers.