The digital Lyrical Ballads, co-edited by Bruce Graver and Ron Tetreault, was first published by Romantic Circles in 2000. It was hosted separately, not on the main server space for Romantic Circles maintained by the Maryland Institute of Technology (MITH), and, for unknown reasons, the digital Lyrical Ballads (2000) disappeared even before the transition of Romantic Circles from the University of Maryland to the University of Colorado in 2018. One can see the original Romantic Circles publication via the Wayback Machine, now maintained by the Library of Congress. Actual access, however, takes persistence and patience.

It uncannily apt that the Lyrical Ballads, produced at a critical transition in the history of print — between coterie and mass printing, between dynastic and corporate publishing houses — should also have first appeared as a digital edition very early in the history of web publishing, only to suffer from massive medial transformations: an unprecedented evolutionary acceleration that could render a digital architecture "new" and then unusable within five or ten years (Schweikert; Weaver).

During the summer of 2024, the Director of Romantic Circles, Thora Brylowe, emailed Bruce Graver and me about republishing Lyrical Ballads. While Bruce had the files that he had originally created when transcribing copies from various libraries, the original server had disappeared without prior notice, and so he had saved none of the code for his digital edition. You can see all that remained of the original digital publication by Romantic Circles (Figure 1). Could we revive the digital edition?

Reader, it was possible.

Here we present the second Romantic Circles edition of Lyrical Ballads, revived and expanded: the second release presents all 25 copies of Lyrical Ballads that Bruce Graver transcribed, illustrating various states of the first four editions (1798-1805), including an unauthorized edition printed in Philadelphia. This kind of resurrection and augmentation were made possible for one reason: Bruce had recorded information about the copies he saw using the recommendations of the Text Encoding Initiative Consortium.. He could of course have added editorial information in other formats, but TEI made that information systemmatic and thus easily transformable via programming. Organized in 1986, the TEI's mandate was to provide guidance for creators of digital resources so that their scholarship could be expressed in code that would not obsolesce. Its promise has proven true.

Most certainly, simple typed text, whether inside code or not, would survive multiple phases of web development, although ripping the text from an antiquated CD-ROM, the original format proposed by Cambridge University Press, might have been costly. And a plain HTML version would certainly have survived technological innovation: the Penn Frankenstein provides an excellent example. But text and web formats do not easily afford recording elements of interest for scholarly editors and researchers: the editors would have had to devise a system or write prose comments about various features, and then explain their methods to a programmer who would have to implement an idiosyncratic method for displaying them that might not be legible to programmers of the future. While the base code Bruce used, SGML, is no longer recommended for use by the TEI guidelines for creating digital editions, the TEI tags preserving his meticulous editorial work remained fundamentally stable. TEI tags in SGML made possible recording whether the insertion of a correction appears below the line or in the margin, whether it was written in Dorothy or William Wordsworth's hand. And it is a short skip and a jump from SGML to the XML now used for TEI tagging.*

The field of Digital Editing has learned much about preservation primarily through the school of hard knocks. The Endings Project at the University of Victoria discovered through research that even extremely well-funded digital resources typically did not last longer than five years. An original infusion of funding allowed for hiring a programmer who could design a beautiful site, with all the bells and whistles; but funding runs out, and even one PHP upgrade on the hosting server can break everything.

When the University of Colorado first started hosting Romantic Circles, the Drupal version from Maryland was almost unusable: as project manager Jessica Tebo puts it, "it was hanging by a thread."† It took four years to move everything into a new Drupal version simply because Drupal modules created by the open source community are not upgraded for every new version of Drupal, and often they are not upgraded at all. The commercial server host currently used by Romantic Circles is planning a PHP upgrade for October 2026 that will break the current Drupal instance hosting Romantic Circles, even before all the resources from the previous version have been ingested into this new version. It is for this reason that we are recreating Romantic Circles resources according to Endings-Project principles. Truly, the history of Romantic Circles makes it a poster child for why these best practices are necessary. The first Romantic Circles digital resource to fall victim to the relentless drive for technological innovation, Lyrical Ballads, is also the first Romantic Circles edition to be published according to best practices for preservation.

This digital edition also adheres to Stanford University Library's concept called LOCKSS—Lots of Copies Keeps Stuff Safe. Especially as a researcher interested in eighteenth-century women writers, knowing that all the remaining copies from small print runs can be destroyed by a Blitz and that there were writers whose work we will never know, I see proliferating copies as key to survival. So we offer here with a Creative Commons Zero license, available to any and all, the TEI encoded documents (now upgraded to XML), as well as the transforms, javascript, and css used to present TEI in usable and sustainable HTML. The technical goal of this edition is to ensure that the enrichments that the editor spent years meticulously recording, now visible here, can persist in the future research world, safe and sound.

Back to top