Feminist Modernist Collaboration, Then and Now: Digitizing Hope Mirrlees’s Paris

When we teach Paris: A Poem, we find ourselves repeatedly facing the same pedagogical question: what do we want our students to see when we read this poem? Hope Mirrlees’s text is at once a personal, lyric exploration of post-war Paris and a work of printed visual art. It is a modernist long poem written by a single author as well as an example of the kind of feminist modernist collaboration possible in small, independent presses. When faced with a poem like this that showcases its intentionally challenging design, how do we frame the relationship between reading and seeing? Between poetic text and visual construction? While these questions apply to many difficult modern texts, they are complicated here by the poem’s minor status coupled with its proximity to the works of canonical literary figures. Should we play up the fact that Paris appeared two years before T. S. Eliot’s formally groundbreaking The Waste Land, a move that almost inevitably results in extended comparisons between Mirrlees’s long-overlooked poem and its unquestionably famous counterpart? Or, from an appealingly feminist pedagogical perspective, should we highlight the material history of the text, thereby tethering the contemporary significance of the poem to its status as an early publication of Leonard and Virginia Woolf’s Hogarth Press and all but ensuring that an almost starstruck attention will be focused primarily on the corrections that appear in Virginia Woolf’s hand? Alternatively, should we let the poem stand on its own, in all its complexity, divorced from the high modernist contemporaries so often referenced in discussions of Mirrlees? A few years ago, seeking a way forward that would make any of these choices possible while avoiding the overvaluation of either Eliot or Woolf in the poem’s literary history, and hoping to make Mirrlees as easy to assign in our classes as either of them, we embarked upon the creation of a digital edition of Paris.[1]

As a formally experimental depiction of a woman’s movement across the postwar City of Lights over the course of a single day, Paris has always challenged the media available to convey it. Set during the Paris Peace Conference of 1919, the poem is not only visually complex, it also asks us to see and hear as we read—to call up particular details of a world that exists beyond the page. It is full of period-specific imagery (advertisements, street signs, posters in a Metro station), sounds (bits of conversation, several bars of music), and architectural references, and several sections incorporate the techniques of concrete poetry. Yet despite the foundational introduction (“modernism’s lost masterpiece”) and accompanying scholarly annotations provided by Julia Briggs in Bonnie Kime Scott’s Gender in Modernism (2007), it remains difficult to bring the poem to life in the classroom without the use of other media.[2] This is what we wanted to assemble for our digital edition: images, sounds, maps, and so on. Since the publication of Briggs’s annotations, Sandeep Parmar’s edition of Mirrlees’s Collected Poems (2011), and Faber’s paperback centenary edition of Paris (2020), scholarly attention to Mirrlees has grown steadily. Yet much of her work remains out of print and Mirrlees’s presence in the digital realm remains limited.[3] Unlike, say, The Waste Land, which was the object of one of the earliest e-lit hyperlink projects and which was repackaged as an expansive digital edition via Touch Press in 2011, there has not yet appeared “an app for that” that would make Mirrlees’s poem, and the world it inhabits, more approachable to contemporary readers.

With an explicit understanding that adaptation is a form of interpretation and collaboration, we (ambitiously! naively!) set out to change that. Working together with an interdisciplinary team of scholars, staff, and students, all of whom were invested in feminist digital methods, we embarked upon an edition of Paris that would preserve the original layout of the poem, while also offering access to Briggs’s notes and a range of contextual materials that ground the poem in the world of interwar Paris. Yet, perhaps unsurprisingly, this process of remediation has turned out to be vastly more complicated than we had initially imagined. Just as Paris challenged printing conventions in its time—incorporating graphic elements, vertical poetic lines, musical notation, and lots of variant spacing—the poem challenges our common systems for representing texts digitally today. Although modern web-development tools make it fairly straightforward to add historical imagery or maps, recreating the poem’s complex original layout proved to be quite difficult. Scholarly editions in general lift poems out of their original material contexts, but this effect can be intensified in digital editions, which remove texts from the realm of print altogether, instead embedding poetic lines and stanzas in standardized html boxes, often with very little formatting.[4] However, because Paris’s typographic layout is so crucial to its poetic meaning, we could not ignore its printed form. Thus, rather than detaching Paris from its original material context, the process of digitizing the text has reinforced to us that this form—and the poem’s contention with the limits of form—are crucial to our understanding of the text. We thus consider the process of remediating Paris to be a form of digitally-assisted close reading, one that reminds us that the formal complexity of modernist poetry serves as a challenge to conventional understandings of the relationship among poems, books, and images, even in a digital age.

One of the main reasons Paris was difficult to move online is that in digital systems, text and images are usually kept separate. When we read historical texts online, we are often offered two ways of viewing them: as PDF scans of the original pages or as plain, unformatted text usually generated through Optical Character Recognition (OCR).[5] The former represents the text as a series of images, while the latter represents it as machine-readable text. Yet Paris resists this treatment because it exists as both text and image, and it would be utterly changed by any attempt to represent it without formatting. At the crux of this issue is whitespace. Paris is shaped throughout by whitespace, but when text is processed by computers, the dimensions of whitespace are not usually taken into account; it is used simply to mark the end of clusters of characters and to mark the boundings of words.[6] Because of this, we could not use OCR to create a transcription of Paris. Instead, we had to transcribe it by hand, which initially did not seem like too tall an order, given the length of the poem: just under 500 lines printed across 23 pages. However, we still had to find a way to account for the variant whitespace, which appears both within and between lines and varies on a smaller scale than can be represented through standardized spacing in modern digital typefaces. We considered trying to approximate the spacing, but this seemed wrong, particularly because our team initially included Jaleen Grove, a historian of illustration and graphic design, who encouraged us to view each page as a unique image in its own right. The difficulty of representing the poem as both text and image was compounded by the fact that we were encoding it in the markup language XML and following the Text Encoding Initiative (TEI) guidelines, which required us to choose between categorizing our transcription of Paris as a text or as a facsimile—the former would have allowed us to represent it as a poem, but the latter would have allowed greater possibilities for encoding formatting information. We needed both.

These decisions were important because in Paris, as in so much other modernist literature, whitespace has poetic meaning. And it is also an important site of collaborative feminist labor. It was by adding spacing with physical pieces of type that Woolf, who typeset the poem herself, worked to achieve Mirrlees’s poetic vision. And 100 years later, it was accounting for this spacing that took up much of our labor in reproducing the poem. Despite our intention not to overvalue Woolf’s contribution to the text, this work in turn led us back to the close consideration of the poem’s initial visual and material construction, and we began to wonder what the difficult process of digitizing Paris now could tell us about the collaboration underlying its initial production. As the only two extant page proofs demonstrate, Mirrlees was extremely particular about how her poem was laid out. But she and Woolf were communicating through written letters, and Woolf was ultimately the one who controlled where each character was printed on the page. At the time, Woolf was still new to typesetting; Paris was only Hogarth’s fifth publication. As a novice printer and fellow writer, her efforts to translate Mirrlees’s words onto the page involved a significant amount of improvisation and ingenuity, and in its original form, the book has a distinctly amateur quality. It is this quality that is hardest to recreate via digital systems, which favor standardization and consistency and can make it difficult to represent the feeling of play and invention that is apparent in the Hogarth edition. But this amateur quality is a key part of the impact of the poem, which visually stages a contention with the possibilities of print, while within the poetic lines Mirrlees interrogates the limits of language. Without distracting attention from Mirrlees’s significance as a poet, it is clear that Woolf’s work as typesetter was creative labor that, in some sense, was constitutive of the work not just as a book but as a poem. This focus on Paris as a collaboration put more pressure on our decision about how to record the work in XML. If we adhered to TEI standards for recording verse, it would be difficult to capture all the complex formatting decisions about which Mirrlees had been so insistently particular. However, if we created a facsimile transcription, that would shift our primary focus to the decisions made by Woolf about the layout of the printed text. Either way, we would lose important aspects of the piece.

Ultimately, what helped us to solve this problem was a return to the methods that Woolf herself had used when publishing Paris. Woolf printed the book on a small press in her dining room, and all she had to use when translating Mirrlees's manuscripts and letters onto the printed page were pieces of moveable type that, when inked, made impressions on rectangular surfaces. When we looked at the pages with letterpress printing techniques in mind, we realized we needed to move our focus from the words of the poem to the whitespace between them. In letterpress printing, whitespace is created through the insertion of individual pieces of spacing material. These blocks of metal function just like regular type; however, since they are not type-high (that is, not as tall as the other pieces of type), they do not make a mark on the page. Wherever Woolf wanted to create spaces in Mirrlees’s poem, she had to insert spacing material, and in order to copy her layout while preserving the poem as text, we would have to insert individual spacing “material” in our edition as well. So, we measured the dimensions of each vertical and horizontal space on each page of the text, and then we transferred those measurements to idealized pages with standardized dimensions.[7]

After gathering all this information, we encoded each space using categorized TEI “space” elements, which we included within and between poetic lines. We later wrote CSS corresponding to each space element, so a standard web browser would be able to interpret them. In this way we were able to represent the poem as text while preserving its original visual layout.

Once we had created a diplomatic, machine-readable transcription of the poem and its layout, we then had to make decisions about how it should display on a screen. Should the poem be presented as a continuous scroll or flipping pages? Should we preserve the original page margins or just the internal spacing? What about other printed marks on the pages, like the imprint statement or page numbers? What about Woolf's hand-written corrections? Each of these questions carried weight as we thought about Paris as a book, a poem, and now a digital artifact. But in the end, we had to make decisions in order to mount the poem online. We created two digital layouts for the poem: one that presents it as unbroken scrolling text and another that presents the text within the dimensions of the original pages and includes page numbers, an indication of the original margins, etc. Over these displays, we layered options to view both Briggs’s notes and a set of contextual images selected by our team. Offering these different views, of course, shifts the reading experience, creating opportunities for interaction that further destabilize the relationship between text and page. 

In relation to Paris, this instability is fitting: the poem has always pushed the boundaries of static text. It traces the journey of its speaker as she navigates the metro stations and crowded streets of Paris, and throughout the poem unconventional typography is used to signal embodied and sensory experiences that cannot be conveyed directly in print: words in all capital letters represent the flashiness of advertisements and metro signs, exaggerated justification turns poetic lines into a representation of the layout of the Tuileries as the speaker walks through it, italicized phrases indicate moments of speech and overheard conversations, and embedded bars of musical notation give an even more specific indication of sounds the speaker is hearing. The graphic of ursa major at the end of the poem—created with a mix of six-spoked and eight-spoked asterisks from Woolf’s font of used 9-point Caslon type—gestures toward a vision of the night sky that moves beyond the specific context of post-war Paris and into a more universal, planetary frame. Through the incorporation of these elements, Mirrlees dramatizes a productive contention with the limits of print: she gestures not only verbally but also formally toward things that cannot be directly represented on the page. The grandeur of ursa major at the end of the poem, for example, is cast into relief by the inadequacy of seven asterisks for representing it.

But while Woolf only had a single, incomplete set of type and a small hand-press to translate Mirrlees’s poetic vision onto the page, today we have access to an array of multimedia tools that can convey sights, sounds, and images. For our digital edition, we could have chosen any number of images of ursa major to include as context for the final lines of the poem, from historical drawings to high-definition photographs to contemporary star charts. For that matter, we could have included a video of the constellation in the night sky or of a scholar explaining its significance in classical literature. But how could we incorporate these kinds of contextual materials while preserving some sense of the poem’s initial low-tech construction? As we moved out of the process of recording the poem and into decisions about representing it online, we had to return to our initial questions: what do we want students to see when they read this poem, and how does thinking about Paris as visual art change the way we read and teach it as a poem? What additional contexts could help contemporary readers engage with Mirrlees’s text? How should the infrastructure of the website invite that kind of engagement? And how could we make the collaboration between Woolf and Mirrlees visible when the way it was most apparent to us was in an aspect of the poem usually overlooked—its whitespace?

Emphasizing the importance of whitespace runs contrary not only to established editorial practice in relation to poetry but to long standing conventions in typography, which insist that the labor of printing should be “invisible.”[8] And while book historical scholarship has long emphasized the collaborative production of literary texts assigned to single authors, the identities of printers, except in very particular cases, usually remain unknown, or at least unremarked-upon. Writing about the opacity of the contemporary publishing process, Matthew Kirschenbaum rightly asks, “After all, any author can tell you who published their book. But how many can tell you who printed it? Or where it was printed? . . . What about the person who designed the cover? Who were they? Who did the copyediting?”[9] From our current vantage, ensconced as we are in the global, neoliberal “age of Amazon,” these questions are particularly hard to answer, and, for many scholars of modernism and twentieth century literary culture, their answers still remain difficult to verify. Yet in the case of Paris, these questions are uncharacteristically easy to answer because of Virginia Woolf’s enduring fame. J. Howard Woolmer’s A Checklist of the Hogarth Press, which includes a history of the press by Mary E. Gaither, presents an extensive account of the press’s resources, practices, and publications.[10] This level of attention has extended into the digital realm with the creation of the Modernist Archives Publishing Project, which describes itself as “a critical digital archive of early twentieth-century publishers, beginning with Leonard and Virginia Woolf's Hogarth Press” and currently contains thousands of images of archival material related to Hogarth.[11] Mirrlees, of course, has received no comparable digital attention, and as we thought about how to present her work online, we had to consider what it would mean to once again hitch her writing to one of the few early twentieth-century women writers whose work has moved beyond the “cycles of forgetting” that women writers are so often subject to in print and online.[12]

Our ultimate goal has been to draw out the formal, material, and historical importance of this collaboration between the poet and her printer—Hope Mirrlees, “modernism’s lost hope,” and Virginia Woolf, “icon, celebrity, star”—without reducing the significance of Paris to its contact with the canonized.[13] We are not often faced with texts from celebrity printers and nearly forgotten authors; the balance of history’s gaze tends to tilt the other way. Yet we suggest that the unusual historical circumstances of this poem offer an opportunity to consider the collaborative work underlying modernist texts more broadly. This often invisible (and often feminized) labor—akin to what Sarah Blackwood and Lauren Klein have described as intellectual “carework”—belongs not only to the history of publishing but to the story of poetic production, and to separate these processes is to fundamentally damage our understanding of literary history.[14] Digital remediation is a form of intellectual care work too—an effort to make a historical text more available using new tools, while staying true to some sense of its original form. But while digitizing Paris we found ourselves focusing less on Mirrlees’s poetic vision and more on the layers of labor that have brought it into contact with audiences over the last century–-not only Woolf’s typesetting, but Julia Briggs’s foundational annotations, and the more recent recovery efforts surrounding the piece. By calling attention to the historical and transhistorical feminist collaborations that inform the history of Paris in our edition, we hope not only to make the text more accessible and engaging to students but to illustrate the amount of effort and labor required to keep challenging texts by women available to readers within current systems of textual production and dissemination.