Virginia Woolf in Circulation: The Hogarth Press Order Books, Modernist Bookselling and Digital Praxis

Virginia Woolf’s To the Lighthouse famously opens with six-year old James Ramsay, “sitting on the floor cutting out pictures from the illustrated catalogue of the Army & Navy Stores.” [1] With the awkward dexterity of a young child, James carefully guides his scissors around “a picture of a refrigerator,” endowing the image “with heavenly bliss” under the loving but watchful gaze of his mother. The scene has long captured critical attention, but less so the consumerist allusion to a popular shop on the book’s first page. The Army & Navy Stores was a well-known London department store, over a quarter-century old by the time James shreds its marquee catalogue.[2] Founded in 1871 as an officer’s cooperative, the shop supplied domestic commodities at reduced rates, including books priced at a twenty-five percent discount.[3] By February 1872, it added a grocery wing and moved into stationery, drapery, and tailoring, with a chemist’s shop and a gun department.[4] Catalogues propelled its products far and wide, not just to the fictional Hebrides but also to outlets in Paris, Leipzig, Plymouth, and Mumbai, until its global brand receded after 1945.[5]

Today, it is no longer possible to shop at the Army & Navy Stores at 103 Victoria Street, London.[6] For modernist scholars and book historians, however, its unexpected role—not just as an obscure historical allusion in Woolf’s fiction, but also as a contemporary bookseller of Woolf’s own publications—is becoming newly visible, appearing as one shop name, nested amongst hundreds of others, in the holographic sales ledgers of Woolf’s private press, The Hogarth Press, founded in 1917 (fig. 1). 

These ledgers—the so-called Hogarth Press Order Books—comprise eighteen manuscript folios housed in Special Collections and Archives at the University of Reading. They capture all book orders that flowed daily into the Hogarth Press, primarily from booksellers, but also from private individuals, societies, and institutes, for single, multiple, or bulk copies of the over 500 Hogarth Press titles published from 1920–1946, including all of Woolf’s after 1920. Within the Order Books lies a vast unstudied storehouse of cultural, economic, social and book historical data, both quantitative and qualitative. Tabular entries capture the date an order was received, the purchaser’s name, the number of copies ordered, the date of shipment, receipt of payment, running tallies of sales, and the occasional bookkeeper’s shorthand, for instance “c.c.” for “colonial cloth” when a book was bound for imperial distribution and “LL” or “IL,” a more runic shorthand that may have been tied to payment schedules.[7] Each Order Book page contains about thirty-six lines of data and 400 pages, so the potential cumulative data set for all Hogarth Press publications over a quarter century exceeds 200,000 lines.

These records collectively document the sales and distribution of thousands of Hogarth Press books that would come to line the shelves of countless shops, railway stalls, steamship lines, libraries, educational institutes, schools, and private homes across the United Kingdom and around the globe, from London to Manchester, Edinburgh to Tokyo, Sydney to Cape Town, Birmingham to Mumbai.[8] As a faculty-staff-student collective, we have spent the past several years transcribing and visualizing their contents under the auspices of a much larger digital humanities project, The Modernist Archives Publishing Project (MAPP), a critical digital archive designed to open up the papers of The Hogarth Press and other twentieth-century publishers.[9] The nature of our work on these Order Books—collaborative, distributive, educative, intergenerational, and interdisciplinary—reflects the longstanding ethos of MAPP where the design of the archive itself engages cooperation and reciprocity as “a site of exchange between multiple archivists, cultural heritage institutions, academics, students and users, all of whom inform the character of, and access to, the resource.”[10] That we have needed to come together to interpret the rich complexity of the Order Books—one extensive cache of materials within MAPP’s database—builds on an important feminist aspiration within MAPP to make “visible and legible the labour involved . . . in the creation of digital archives,” including metadata protocols which aim to surface the employees who worked at the Press, alongside Woolf, many of whom were women.[11] Transcribing their (anonymous) hands in these records, recognizing at once the intimacy and distance between our work and theirs, helped to confirm our commitment to fully credited work coming out of our own collaborations, even if it made us aware just how much more historical, archival, and digitally-informed research needs to be pursued into theirs.[12]

In what follows, therefore, we examine the disciplinary stakes and practical challenges of transcribing and computationally analyzing the Order Books alongside literary-cultural and book historical interpretations of their contents. We explain how and why our workflow model led us to focus exclusively on Virginia Woolf’s sales, arguing that the Order Books not only contain rich empirical data about Woolf’s local, regional, and global circulation, but also allow us to contextualize and enumerate her role as a publisher in new ways, raising questions we might not otherwise be able to ask, let alone answer, without them: who bought Woolf’s books, in what quantities, where, and under what conditions of sale and distribution?  Woolf had founded her own press to be free of editorial oversight and intervention, in terms that implicated the patriarchal underpinnings of the book publishing industry itself, asserting that in owning the Hogarth Press, she was the “the only woman in England free to write what I like.”[13] She had once thought of running a bookshop out of the basement of Hogarth House, so shops and publishing were twinned in her entrepreneurial imagination from the start. And even though she never hung out a shingle for a shop as she and Leonard did for their publishing house, the Order Books testify to the shop-like atmosphere that often pertained in a space chock full of book stocks for shipping.[14] So we argue that resurfacing these Order Books helps to reconfigure and to deepen our understanding of Woolf’s role as a modernist publisher, including their influence, implicit and explicit, on her feminist theories about women, writing, and work. They give crucial empirical evidence for the global circuitry of her cultural influence and public intellectual impact in ways hitherto inaccessible to scholarly interpretation and critique. In taking up the challenge, in Lise Jaillant’s words, that “[m]ore work remains to be done in the [Hogarth] press archive, which contains a wealth of quantitative information,” we also document our methodological choices, detailing how our complex but necessarily collaborative work on these rare and generally overlooked types of archival objects throws new light on the complexities and ambivalences of Woolf’s own creative work environment, including, as we have mentioned, the clerical work of her Press employees, whose collective labors over many years ensured—and, crucially, documented—her sales and distribution within the global book trade networks in interwar Britain.[15]

Archives, Book Sales Records, Clerical and Digital Labor

Despite a recent resurgence of the “archival turn” within modernist studies, publishers’ sales records have not featured highly within the field, nor in cultural and book historical studies more broadly.[16] As Laura J. Miller argues in her landmark study of late twentieth-century bookselling, Reluctant Capitalists: Bookselling and the Culture of Consumption, “[s]ociological or social historical work that examines book publishing in the twentieth century has been relatively sparse, with an even greater lack of empirical research on bookselling.”[17] If this absence of empirical research on bookselling has been a problem for sociologists of the book trades, then it might explain why, even years after Lawrence Rainey urged us to rethink the “institutions of modernism” via its publication practices and readership networks,  empirical data on modernist publishing itself has been so lacking.[18] Still, we need to ask: what impacts for modernist literary and cultural historical study arise from not reading bookselling records? Why do they matter?  What stories do they tell that cannot be found elsewhere? Having spent the past five years as a faculty-staff-student collective, transcribing, aggregating, and analyzing 25,000 lines of Woolf’s sales data contained within her own Order Books—a hefty number but still only about ten percent of the total Hogarth Press sales corpus—we argue that the paucity of broadscale empirical research into modernist bookselling may seem self-evident: these records are difficult to access if they exist at all; their contents are messy and hard to read; there are costs—academic, material, and affective—to making them legible. As Joshua Kotin wryly remarks of another modernist publisher’s business archive, that of Ulysses’s publisher, Sylvia Beach: “[a] team of forensic accountants: this is what it would take to reconstruct the finances of Shakespeare and Company.”[19] Yet, we argue that not to attempt to read the Order Books despite their own forensic difficulty represents a missed opportunity, not just to illuminate their contents, but also to better understand their conceptual importance as information rich archival objects that can offer new ways of historicizing modernism and of reexamining Woolf’s role as a modernist publisher, including how her books circulated within a global marketplace, their distribution driven predominantly and perhaps surprisingly by large scale wholesalers, chain stores, lending libraries and book societies (fig. 3).[20]

While critics have not shied away from interrogating the high/low, coterie/commercial, local/global, authorial/collective tensions within modernist literature itself, Woolf’s Order Books train a meta-critical lens on these tensions because they tap so directly into what Miller calls the “structural ambivalence” of the place of the book object within cultural consciousness where books are so often, but complexly, figured as both “sacred objects” and commodities:

Throughout the twentieth century and into the twenty-first century, there has been a consistent belief in the distinctiveness of the book, an uneasiness with viewing books as ‘products’ to be bought and sold like any other commodity. . . . [A]ny enterprise [i.e. bookselling] that transgresses the boundary between the incommensurable sacred and the marketable profane must cope with structural ambivalence (Miller, Reluctant Capitalists, 19).

Clearly, the Order Books are a different genre of “book” than Miller is referencing here, but beyond the institutionalizing vernacular of the book trade industry inscribed in their very title, there is the obvious fact, worth highlighting, that these manuscripts are books that chronicle, with structured informational data, the sales of other books. Put bluntly, the Order Books showcase books as cultural commodities, their own sacral functionality moot in part because they are not even interpolated as book objects visible to (or worthy of) critical attention. Yet, despite their desacralizing emphasis on sales metrics and numerical itemization, the Order Books document constant structural transgression, quantitative records of purchases sitting intimately, column by column, with qualitative histories—stories of lost bookshops and distributive networks, large and small—that lend more than numeric value to the spaces and places of the bookselling world in interwar Britain, a world paradoxically much harder to envision if we cannot access their informationally rich contents.

Bookstores, themselves hybridized sites of commercial, cultural, and human connection, were then (as now) spaces where buyers, sellers, authors, and booklovers met in shared pursuit of the next great read, as well as the next great deal.[21] Indeed, in her own bookshop perambulations, Woolf was keenly aware of the ways that space, place, money, and desire commingled in bookshops:

I went on the top of a bus . . . to Nutt’s shop to get a Leopardi; then to Mudies where I bought Mill on Liberty; then to Charing Cross Road where I bought the Happy Hypocrite, by Max Beerbohm; & Exiles of the Snow, by Lancelot Hogben. In this way I laid out 7/. But I was amused to find that the lust after books revives with the least encouragement. . . . The bookseller shared my lust which increased it; . . . after all, nothing gives back more for one’s money than a beautiful book.[22]

This diary entry, from March 1918, recounts a delightful escapade, flaneur-style, where Woolf namechecks several well-known bookshops, conspiring with a bookseller, both in hot pursuit of “a beautiful book.” It also notably marks the one-year anniversary of the founding of the Hogarth Press. The press creatively liberated Woolf’s work, but it also reoriented her relationship with booksellers, making her not just a rapturous book buyer, but a publisher selling directly to the same sellers. For book historian Robert Darnton, “the bookseller might be considered the most important middleman of the entire [publishing] system, for he operated in the crucial area where supply met demand.”[23] So, how we read these Order Books changes how we understand Woolf as a publisher. Our project reconfigures her sales records not as ephemeral or dismissible because “desacralized” archival objects but as repositories of vast networks of untold stories about art, life, writing, publishing, professionalism, money, and desire.

As a collaborative team working to establish the Order Books as codices worthy of autonomous critical attention within modernist studies, while not simply strip mining them for decontextualized information, we have been struck too by how the “structural ambivalence” Miller describes as a salient feature of bookselling history itself reflects multiple (not irresolvable but ever-present) metatextual axes in our collective work to transcribe, aggregate, and interpret their contents. For instance, in analyzing the sales records of a now-canonical modernist writer who co-founded her own Press based in part on the desire to publish writings rejected by commercial presses, we show how the Order Books not only document the global circuitry of Woolf’s cultural influence, confirming her public intellectual impact with new empirical data, but also reveal how large commercial distributive networks paradoxically dominated the circulation of her work, however coterie and noncommercial the early motivations to found a private at-home press might have been. Moreover, in learning to decipher the holographic traces of the dozens of (mainly female) workers who recorded the sales numbers which substantiate our empirical claims—in order, that is, to transform their would-be lost labor into computer readable scripts—the intimacy of our access to their collective, anonymous, and years-long clerical practices refracted, sometimes ambivalently, our own. After all, the scope as well as the parameters of this work which privileges Woolf’s records, precisely because of her canonicity, over the hundreds of authors she published at the Hogarth Press could only have been done collectively, even within that selective parametrization. At the same time, it was the need to collectivize such work that also critically attuned us to the complex, hybridized business/domestic environment in which Woolf wrote, oversaw Press employees, and cultivated her own deep engagement with how booksellers’ orders reflected her growing cultural significance.

Our proximity to the lost labor of Woolf’s press employees, and the traces of their presence within the Order Books, were partly what informed our decision, as we discuss in more detail below, to transcribe the Order Books in spreadsheets through a multi-step process that involved replicating individual shorthand, misspellings, and orthographic irregularities in a diplomatic or character-by-character transcription, and only thereafter cleaning and regularizing the sheets to make a machine-readable copy thereof. There is a paradox or ambivalence here too, however: as much as we were drawn to the lives of the Press employees whose clerical labor (literally) underwrote our research and whose highly idiosyncratic and individualized orthography a diplomatic transcription could preserve, we then had to clean these spreadsheets iteratively to operationalize the historical data they contained. We also had to rationalize (and prioritize) our own clerical and digital labors, expecting that in eventually sharing digital images of the Order Books themselves alongside publication of this article, we would catalyze more and different forms of scholarly engagements with these multivalent archival records, perhaps opening the door to fully annotated and encoded digital editions of them, a different form of digital manuscript transformation that could capture more of the ambiguity and messiness of the day-to-day labor of Press employees.

We hope, however, that our foundational work on these Order Books partakes in and provides the resources for future scholarly collectives to pursue what the digital humanist Kathleen Fitzpatrick calls the “generous thinking” needed to radically rethink both argumentative discourse within the academy and new structures of work to sustain new knowledge paradigms: “generous thinking might . . . lead us to place a greater emphasis on—and to attribute greater value to—collaboration in academic life, and to understand how to properly credit all our collaborators.”[24] As faculty-staff-student coauthors on this article, we can name our labors, but only by sharing our results can the many named (booksellers, individual buyers, and distributors) and unnamed (clerical workers) entities in these records be made visible to historical posterity and available for further research.[25]

The Order Books and Re-envisioning Woolf as Publisher

Given the size and material complexity of the Order Books, we had to confront, a little like James Ramsay, cutting up the Army & Navy catalogue, how to crosscut them, digital scissors in hand, in ways that honored their materiality as time-bound, culturally situated codices that also contained rich aggregable data. Could we mine them for information while being ever mindful, in media theorist Lisa Gitelman’s popular formulation, that there is no such thing as “raw data”?[26]  If, as Andrew Piper has argued, “the paper archive affirm[s] the material differences of writing, [whereas] the digital archive flattens everything into identical objects,” could we operationalize (some) of the data they contained while not losing sight of the books themselves?[27]  Keeping the archival Order Books close to mind as “phenomenal” or whole texts in the words of medievalist Elaine Treharne had epistemological importance to us.[28]  As argued above, the Order Books are, after all, rich humanistic aide-mémoire for the often invisible, but deeply networked cultures of modernist book production, including the lost bookshop worlds Woolf navigated as reader, writer, and publisher and the largely invisible press workers who helped to build her publishing business. Our examinations of the Order Books at some scale also provided the opportunity to both leverage and interrogate contemporary practices in the digital humanities whose material infrastructures—collaboration, interdisciplinarity, data collection and curation—can sometimes fail, not dissimilarly, to make visible the complex, distributed, and collective labor that generates new disciplinary knowledge.

Yet the desire to recognize the totemic materiality of the Order Books even as we labored to make digital transcriptions of them helps ultimately, we argue, to deepen critical appreciation for Woolf as a publisher in her own right. For, despite more nuanced recent portraits of Woolf as an independent publisher, the image of the fragile, etherealized female novelist removed from the messy business of book production and hardscrabble economy has historical traction, deeply enmeshed as it is in gendered mythologies about technology, business, money, and power: trenchant cultural stereotypes within the print culture of Woolf’s era, no less than in the digital culture of ours. Being able to envision these Order Books in toto as the collaborative outputs of an entrepreneurial business that Woolf co-founded not only allows us to glimpse Woolf poring over them to track the ups and downs of her own sales, but also offers material proof of the busy work environment that surrounded Woolf as she wrote. As John K. Young reminds us, “Woolf worked not as an author isolated from textual production but as one immersed in what she once called ‘life on tap down here whenever it flags upstairs.’”[29] Woolf’s pointed deixis, “here,” refers to the domestic situatedness of the Hogarth Press, founded in 1917 as a small hand press on the dining table of the Woolfs’ home, Hogarth House, in Richmond, Surrey, and by the mid-1920s, occupying the basement—the literal foundation of her home—at 52 Tavistock Square, London.  It was a place she often visited to restore her writerly spirits, and where eventually she based her study, perhaps to access more directly the hive-like activity of its below-stairs work environment.

Performing this quantitative analysis has more than numeric value, even if it reminds us to take Woolf’s own numeracy and figuration of the numeric more seriously (her second novel, Night and Day, featured a female mathematician, Katharine Hilbery, as protagonist).[30] It also complements how Woolf speaks qualitatively of the Press, her powerful, agentive words often ignored in both popular and academic renderings of Woolf as publisher.[31] Woolf is, after all, keenly attentive to the realities of the Press as a business enterprise which generates material wealth for her household, including her employees, a fact that often gets lost in more popular renderings of her presswork as a therapeutic hobby. “For the first time we have made over £400 profit”, she writes in spring 1929, after the unexpected success of Orlando, “And 7 people now depend on us; & I think with pride that 7 people depend, largely upon my hand writing on a sheet of paper.  That is of course a great solace & pride to me.  Its [sic] not just scribbling; it keeps 7 people fed & housed” (Woolf, Diary, vol. 3, 221). That her hand does not appear in the Order Books themselves (Leonard’s does occasionally) does not imply she was hands-off; her diaries repeatedly track her sales, the boom and bust of the market, and how bookseller orders presage commercial success or failure. For instance, when in 1921 she worries about the reception of her first Hogarth Press short story collection, Monday or Tuesday, she is heartened by a telephone order from the wholesaler Simpkin & Marshall: “There was an Affable Hawk [Desmond MacCarthy] on me in the New Statesman which at anyrate [sic] made me feel important (and it’s that that one wants) and Simpkin & Marshall rang up for a second fifty copies. So they must be selling” (Woolf, Diary, vol. 3, 220). To get that number—the specificity of its having been a repeat bulk order and from Simpkin & Marshall, her leading distributor according to our analysis—she is not only asking questions of her staff; she is likely poring over the records for herself, watching the numbers as they grow. The Order Books allow us to read those increments as she would have seen them.

The rich and granular detail of these Order Books therefore adds depth and texture to critical accounts of Hogarth Press sales that, we argue, have been overly reliant on synoptic and retrospective narratives of profit and loss ledgers.[32] Given Woolf’s own proximity to them, it is perhaps not surprising that account books appear as crucial sites of feminist revisionist history in A Room of One’s Own (1929), a book more deeply indebted to and rhetorically imbricated within Woolf’s self-identification as a publisher than critics have acknowledged. As our analysis shows, it also sold exceedingly well—a fact that surprised Woolf less than did the success of its predecessor Orlando (1928).[33] Orlando, a parody biography, confounded genre categories, as reflected in booksellers’ confusion about where to shelve it in their stores (Battershill, Biography and Autobiography, 93–5). Conversely, it seems clear—and the post-publication numbers bear this out—that Woolf knew Room would generate a large audience, speculating that readers “will be feeding off Women & Fiction [Room] next year for which I predict some sale. It has considerable conviction” (Woolf, Diary, vol. 3, 221). Rarely remarked about Room’s form, in fact, is that a solicitous and cooperative editorial voice helps shape the narratorial voice (Staveley, “Virginia Woolf and the Hogarth Press,” 259). Invited to speak to an audience of women undergraduates about “women and fiction” in her role as acclaimed novelist, Woolf switches hats, her narrator sounding more and more like an acquisitions editor, one of Woolf’s many duties as Press owner. Channeling that role, Woolf’s narrator encourages the young undergraduate women in the audience to unearth the “mass of information” lying about in scattered archives where they might find, most suggestively, the “account books” needed to write new feminist history:

What one wants, I thought—and why does not some brilliant student at Newnham or Girton supply it?—is a mass of information; at what age did she marry; how many children did she have as a rule; what was her house like; had she a room to herself; did she do the cooking; would she be likely to have a servant? All these facts lie somewhere, presumably, in parish registers and account books; the life of the average Elizabethan woman must be scattered about somewhere, could one collect it and make a book of it.[34]

If our work on Woolf’s own account books helps solidify a portrait of her as a publisher in her own right, one audience member, E. E. Phare, who heard the original talk on which Room was based, explicitly recalled Woolf acting in that role, encouraging her audience to write books and send them to the Hogarth Press for publication consideration, a touching reminder that wherever Woolf went, so went the Hogarth Press.[35] 

In 1929, Woolf reached back in time to insist on the hard work needed to excavate scattered archives to find the empirical evidence to rewrite women’s history, within the framework of a forward-looking, pedagogically driven feminist treatise, in Melba Cuddy-Keane’s words, “to encourage and empower women in the [1928] audience by inscribing them as the culminating point in a progressive narrative.”[36] Almost a hundred years later, as a faculty-staff-student team, we have come to see our work on Woolf’s own Order Books as inheriting this collaborative and intergenerational, but non-teleological colloquy: one requiring new configurations for academic work, communal forms of “thinking in common,” and digital technologies that enable us to aggregate scattered fragments to reveal new patterns in old archival data, all while abiding by a feminist digital humanities ethic to “show [our] work”, so that we don’t replicate the kinds of historical erasures Room itself critiques, especially in the new digital economy where, as Catherine D’Ignazio and Lauren Klein argue “the work of data entry is profoundly undervalued in proportion to the knowledge it helps create” (Woolf, Room, 65).[37]

How Should One Read (and Clean) the Order Books?^[38]

Pre-dating our collaborative work (2017–2021) was a faculty-student pilot project at the University of Reading in 2014 that transcribed initial sales records for three Woolf novels (The Years, Between the Acts, and To the Lighthouse) to better understand her colonial distribution markets.[39]  Extending this work, in summer 2017, we began with a single Order Book, MS2750/A27, for which we had received a complete set of digitized page images.[40] A27 became our beta book, allowing us to establish protocols for textual transcription for this kind of mixed cultural (names of people and of book distributors) and numeric (dates, prices, quantities) data.  Materially, its contents were bibliographically evocative, binding sales records for Virginia Woolf and Vita Sackville-West, not just onetime lovers but also the Press’s two highest sellers.

Unlike many large publishers’ archives which are organized chronologically, according to their day-to-day business operations, the Hogarth Press is, as Nicola Wilson argues, an author-based archive organized around the books produced, which can lead to some quixotic arrangements where books are indexed both by genre—poetry, essays, novels, psychoanalysis, lectures—and by alphabetical author.[41] Thus, some author records are dispersed across different Order Books. While many of Woolf’s records are in A27, for instance, nested with Sackville-West’s, her overall entries are scattered in different Order Books. This disaggregation has the salutary, democratizing effect of integrating her within the Hogarth Press lists: Woolf is but one amongst the hundreds of writers the Press published and whom she, as director and editor, cultivated. But for us, not working in situ in the Reading archives, we found it harder to get a bird’s-eye overview of Woolf’s collected works, especially since, when we first began, we were unsure whether we would be able to get digitizations for all her dispersed works, an important consideration for us in making empirically accurate assessments of her overall distribution. Over three years, however, we eventually digitized, transcribed, and cleaned sales data from Woolf’s entire corpus from To the Lighthouse (1927) onwards; we also collected information from some earlier, chronologically incomplete sales records, including her landmark essay, “Mr Bennett and Mrs Brown” (1924), The Common Reader First Series (1925), and Mrs Dalloway (1925). While we have retained these transcriptions in our records, for our data visualizations, we have only included the sixteen texts for which we had largely complete sales runs.[42] 

To the Lighthouse marks the beginning of a steady, largely unfaltering rise in both Woolf’s sales and her fame, a mere decade after the Press’s founding. Sales records do exist for the early years of the Press (1917–1922) when it was more coterie, and these Order Books remain open for future study.[43]  During those first five years, the Hogarth Press operated a subscription service with A subscribers paying one pound and receiving all publications and B subscribers receiving a publication list from which to select book choices (Willis, Leonard and Virginia Woolf, 48). While this early business model favored individual buyers, whose names predominate, these lists do nonetheless contain the names of a few bookstores and wholesalers (e.g. J. Bain; Simpkin Marshall) who appear with increasing frequency as the decade progresses and the Woolfs expanded their distribution channels. But booksellers and wholesalers were there from the start, contrary to the “coterie consumption” argument that the Woolfs only published for their friends.[44] This mixture of the private and the public constitutes the Order Books’ vernacular.[45] Woolf herself captured the hybridized buying landscape when she wrote to Vita Sackville-West about sales of Sackville-West’s travelogue, Passenger to Teheran; Woolf’s dry wit mingling (once again) with numerical evidence she is reading incoming orders,“ . . . it is doing very well: 6 sold to Smiths [W. H. Smith], one at the door to an old woman with long teeth.”[46] Even when bookshops and institutions come to dominate the records, many individual buyers continued direct purchase, as did some bookshops whose orders fell, similarly, into the category of “1 copy”.[47] 

While individual numbers were intriguing and we knew they would tell their own story once we could visualize the patterns they revealed, numeric script was also more legible; we had to work much harder to decipher handwritten words. As argued earlier, access to these kinds of archival records is rare, so we had few if any methodological models to follow. As Corinna Norrick-Rühl, in Book Clubs and Book Commerce, notes, “questions of bookselling and distribution have not been highlighted [in book history], although they figure as . . . important actors in the standard book historical models.”[48] And whereas “publishers’ archives are better preserved and more easily available than sources relating to distributors and distribution,” the Order Books, as part of the larger Hogarth Press publishers’ archive, were themselves sequestered from the production papers and uncatalogued until the past decade (8).

 Compounding these overlapping obstacles of access, method, and archive is, as argued earlier, the reconstitution of academic work needed to examine them. Still, we wanted to avoid the pitfalls of what Spencer D. C. Keralis calls (out as) the “labor economy of digital humanities” where students do the grunt work and faculty, or “PI,” publish the results for sole credit.[49] Much has been written about the opportunities and challenges of collaborative work in the field, ranging from the need for collaborators’ manifestos that encourage transparency and equitable work practices to biases against co-authorship based on the “tacit notion of scholarly credit as a zero-sum game.”[50] Most of these debates center on graduate and postgraduate work. Yet undergraduates not only make superb digital humanists, but their training, under conditions of nonexploitative institutional oversight, can also form a pedagogically rich throughline during their university years. The time it has taken to responsibly process these Order Books, for instance, has tracked student progress from first year to graduation (and beyond). As paid interns at the Center for Spatial and Textual Analysis (CESTA) at Stanford University, they worked at a central on-campus location, with shared cohort skills training, socialization, and presentation opportunities. Even so, from the start of this project, and precisely because the students were responsible for the lion’s share of transcription, data entry, and data visualization, we committed to producing this coauthored article as one key part of what Keralis calls “ethically managed student labor”: visible culmination of years of transcription, weekly seminar-style discussions on modernism, Woolf studies, and digital humanities, group writing by all team members, and online synchronous and asynchronous exchange (“Disrupting Labor,” 274).[51]

Laboring on both the tasks and the interpretative challenges of transcription work, we confronted a truism about the digital humanities (and data science in general): eighty percent effort for twenty percent visualization. The manuscript data was messy, so how to standardize, or clean, it was our first concern.  Some who argue for the practice, but against its name, contend that the word itself is self-limiting:

the collective acceptance of a connotative term, “cleaning,” suggests two assumptions: first, that researchers in many domains consider the consequences of whatever is done during this little-discussed (80 percent) part of the process as sufficiently limited or bounded so as not to threaten the value of any findings; and second, relatedly, that there is little to be gained from a more precise description of those elements of the research process that currently fall under the rubric of cleaning.[52]

While critics are right to highlight for interrogation the actual work of cleaning, changing or erasing its name betrays a problematic desire to avoid the feminization attending both the phrase and the practice. For us, in fact, the communal, iterative labor required to clean these records conjured the “Time Passes” section of To the Lighthouse. When the Ramsay family home is about to tip into oblivion from the ravages of time and neglect, it is saved by the ministrations of Mrs. McNab and Mrs. Bast and the neighboring handyman, George, but not before Mrs. McNab, on first viewing the task, turns away from it, ruminating: “It was too much for one woman, too much, too much” (Woolf, To the Lighthouse, 141). The too-muchness of our work had to be confronted by collective endeavor, so we embraced the data cleaning terminology. The iterability of the task, the fact it would have to be returned to, and would never be finished, embodied an affective feminist reclamation. 

Cleaning was also an activity that allowed us to recognize the labor of the Hogarth Press workers, to chronicle where their idiomatic expressions survived as proof of labor in the written record. As such, we chose a diplomatic transcription: exact typographic replication of all letters and numbers whatever their orthographic presentation. Our spreadsheets captured the book title, the date of order, cumulative total copies sold, markings (such as check marks), purchaser name, notes (such as the mysterious LL/IL), number of copies ordered, date order received, date order filled, and price (in pounds, shillings, and pence).  We added columns to capture geographic data where it was referenced, often in shorthand (e.g. B’ham for Birmingham).  Wanting a machine-readable version that was as faithful to the original as possible, we retained inconsistencies in names (Foyles vs. Foyle’s) and Press worker miscalculations.[53] If there was a purchaser’s name we were unsure about, we highlighted the corresponding cells, deputizing a team member to unearth the correct name or spelling through sources including Google, Google Books, and publishers’ almanacs and directories. Experimenting with workflow methods, we tried to balance the impulse to historicize with the need to create machine-readable spreadsheets. When two students split up the work vertically—one taking all the purchaser names, the other all the sales numbers—speed was gained at the cost of narrative context. In pursuit of ever more efficient methods of data capture, we discovered that material features of the manuscript page, its tabular arrangement of data, could be manipulated to increase transcription speed; yet not only did interest wane as a result of this division of labor, but the tabular design is itself a stumbling block to improving machine readable tools for similarly mixed quantitative and qualitative data.[54]

We ultimately generated rigorous standardized transcriptions by experimenting with several different cleaning techniques: from eyeballing and grouping similar names to writing dozens of Python scripts to automate what is colloquially called find and replace; to using OpenRefine’s clustering feature, which groups together similar names based on nearest neighbor algorithms.[55] Figure 2 gives a good sense for the challenges we faced: there were, for instance, a dozen different iterations of “International University Booksellers” alone. And not all variants were misspellings.  W. H. Smith, for instance, was often colloquially rendered as “Smith” or “Smiths,” not to be confused with J. Smith, another shop (or person) entirely. In the end, however, our labors were justified: we went from roughly 3,500 alleged discrete purchasers to roughly 1,600. We maintained for the historical record through our diplomatic transcriptions the idiosyncratic naming practices of the original Hogarth Press workers, while ensuring as closely as we could the empirical accuracy of the sales data.

How Should One Visualize the Order Books?

With reasonably clean data, we answered questions hypothesized before transcription began: Who bought Woolf’s novels and essays? In what comparative quantities? How much was gross income per book? What did the buying landscape look like and what could it tell us more generally about the brick-and-mortar experience of British bookselling in the interwar years?  Visualizing Woolf’s bookshop distribution for a comprehensive dataset of Woolf’s sales across numerous Order Books between 1924 to 1946 reveals for the first time where her books travelled before arriving in readers’ hands. It also affords an unprecedented overview of contemporary industry players: a richly hybridized book buying landscape composed of wholesalers, bookshops, book societies, lending libraries, and individual buyers all driving different patterns of access, reception, and status regionally, nationally, and internationally. 

The following visualizations offer several bird’s-eye views of Woolf’s book distribution based on total sales of 146,604 individual units. In what follows, we focus primarily on the “large” buyers, which we define as purchasers who bought over 500 copies of Woolf’s books over about twenty years. Figure 3 documents the relative distribution of all purchasers buying 500 or more copies of Woolf’s works.[56]

Figures four and five compare total copies purchased by size of purchaser, making it clear just how dominant were large purchasers in Woolf’s overall distribution; they suggest that not only did Woolf’s highbrow reputation not limit her market reach, but that her sales were also, crucially, dependent on large, mainstream distributors moving her books into wider markets. These large purchasers, though relatively small in number, moved the vast proportion of her books, while the small purchasers (many but not all of whom were individuals) bought many fewer books but, given their heterogeneity and only as a share of “total number of purchasers,” vastly outnumber their mainstream competitors (figs. 4, 5).

The top five buyers—W. H. Smith, Simpkin Marshall, Boots Book-lover’s Library, the Book Society and the Time’s Book Club—constitute slightly over fifty percent of Woolf’s total market and represent mainstream British book distribution modes in the first half of the twentieth century: wholesalers, bookstalls/shops, lending or circulating libraries, and book societies (fig. 3).[57] These modes, which combined buying with borrowing, lending, or subscription services, dominated the book trade before the rise of paperbacks in the mid-1930s accelerated what Andrew Nash describes as Britain’s “transform[ation]—at last—into a book buying culture” and its relative diminishment as a borrowing or lending book culture.[58] Woolf’s leading distributors, W. H. Smith (founded 1812) and Simpkin Marshall (founded 1814), had different business models but were behemoths of the trade. W. H. Smith gained market clout by developing book stalls along the rapidly expanding Victorian railway networks, while Simpkin Marshall was Britain’s leading wholesaler, distributing books around the nation and abroad.[59] Simpkin’s name also signified cultural capital to Woolf, legitimizing the Hogarth Press as a prestige publisher: “Simpkins said today that many great publishers would be proud to have our list. In 10 years we shall be rather celebrated” (Woolf, Diary, vol. 3, 220).

The lending libraries were significant players in the interwar book distribution world, often as market “embeds” within larger business structures, and we see their influence in our data. Boots Book-lover’s Library (founded 1898) was what Leah Price calls the “bibliotherapy” brainchild of Florence Boot, wife of Jesse Boot, founder of the Boots pharmacy chain, which cross marketed books and comestibles and offered affordable subscription services to attract, as Nicola Wilson has argued, lower middle-class and middle-class customers.[60] Even W. H. Smith had a lending library arm whose origins lay in its competition with Mudie’s, the “leviathan” Victorian lending library; when Mudie’s refused Smith’s overtures to act as its lending library agent, Smith’s started its own.[61] Yet it is easy to forget how long Mudie’s survived into the twentieth century, despite such competition. Our data visualizations reveal the diminished, but still significant, presence of Mudie’s , for instance, as Woolf’s thirteenth highest buyer with 2,212 purchases, far behind Smith’s 17,378. But the relevance of Mudie’s is not negligible, even into the late 1920s (it closed in 1937). Notably, when we break down proportional purchases by buyer, to ask—Which book of Woolf’s did each establishment purchase most frequently, in proportion to their total inventory?— the largest purchase by Mudie’s is A Room of One’s Own. Despite its reputation for circulating fiction to its subscribers, Mudie’s, as Robert Altick notes, invested heavily in nonfiction, a category aligned with Room.[62] It also has the distinction of having one of the highest proportionate shares of Room sales compared with the same distributive view in other top buyers (fig. 6). So, while Room was a leading seller across the board, its particularly high proportion on the Mudie’s list indicates that the Victorian lending library was still an active book distributor in the first half of the last century, with Woolf herself, as evidenced in her book- buying spree quoted earlier, frequently stopping in to get a bargain in their secondhand book department, a feature of their business model since the nineteenth century (Griest, Mudie’s Circulating Library, 170).

Moreover, what Figure 6 reveals is that Room represents a large slice of almost every shop’s Woolf inventory, particularly, perhaps most evocatively, that of the Army & Navy Stores (whose proportion is about equal to Mudie’s). Such dominance makes it striking when, for instance, in the case of Zwemmer’s, Room constitutes only a very small slice of its inventory, with preference given to Woolf’s slender essay “The Art of Walter Sickert”. This distinction, however, not only aligns with Zwemmer’s target market as a bookshop-cum-art-gallery, noted, as Matthew Chambers writes, for its trade in “high-end art books”, design, and architecture; it also highlights how our data, while focusing on Woolf’s sales, goes beyond Woolf, helping to uncover more obscure features of contemporary bookshop marketing, inventory purchasing decisions, and target audiences which all structured the interwar book trade markets (Chambers, London and The Modernist Bookstore, 62).

Unique among Woolf’s top distributors is the Book Society—neither bookshop, nor wholesaler, nor lending library, but rather a subscription book-of-the-month club—with its outsize presence in her sales due solely to one book: Flush (1933), Woolf’s biography of Elizabeth Barrett Browning’s dog. Then as now, dog books were a popular genre form, and, as Nicola Wilson has argued, when the Book Society selected Flush as their October 1933 book of the month, Woolf knew she had hit pay dirt, writing punningly: “We should net £2000 from that six months dogged & dreary grind.”[63] While actual sales were £3,590, Woolf’s estimate, which speaks explicitly of net sales, not gross, might well have been accurate, showing once more just how attuned she was to the business side of the industry (fig. 9).[64] In breaking sales down by year, filtering for copies ordered by the top ten purchasers between 1927 and 1946, we can see just how dominant and how sought after the Book Society endorsement could be to a publisher: Flush overshadows all other sales in 1933 (fig. 7).

Given that our data captures not just year of sale, but month and day, we were curious if we could filter in terms of pre-publication sales as an indication of market interest and anticipated reception. Our results confirmed, indeed, that Flush is again a standout, with sixty-four percent of its orders coming in before publication date, proof of just how much the Book Society could drive sales. As we’ve seen, Woolf recognized the importance of incoming orders as a sign of uptake, so this data point helped us understand why. The average for all her books’ pre-orders is about thirty percent, a fact that brings into view other fascinating apparent anomalies in the data. Orlando’s seven percent pre-orders appears then to be another exception to a general rule, confirming just how confused booksellers were by its genre, and helping to explain why Woolf (once again closely tracking sales) wrote two weeks before its publication specifically about the lack of bulk orders, “the news of Orlando is black. We may sell a third that we sold of The Lighthouse before publication—Not a shop will buy save in 6es & 12s. They say this is inevitable. No one wants biography” (Woolf, Diary, vol. 3, 198). Equally interesting is the relatively low (eleven percent) pre-order rate for Room, a book that will, as Woolf expected, strike a deep cultural chord. It appears that Woolf’s intuition and the market’s readiness might have been opposed. Yet Woolf’s initial positive instincts proved correct. Room generates such high demand that the Hogarth Press will have to print five impressions in the first year alone.[65] Its cheaper price point—5 shillings instead of the standard book price of 7 shillings 6 pennies —can only partly explain this surge (Jaillant, Cheap Modernism, 126).[66] Notably, it contrasts with the pre-orders for Three Guineas (1938). Woolf had conceived of Three Guineas as a sequel to Room, and although it was a more difficult book to market, its pre-orders were more than double Room’s. Whether this was due to the Woolfs’ having hired a feminist marketing agent whose first job was to promote Three Guineas or because Woolf had a much stronger reputation by the late 1930s, or some combination of the two, cannot be answered by numbers alone.[67] Either way, in the pre-order stakes, Three Guineas had more initial market visibility (though not as much as Woolf’s novels); priced higher than Room, it sold fewer copies, but still generated almost equal income in aggregate. Odder still is Between the Acts which appears to have almost no pre-orders (perhaps owing to difficulties marketing a posthumous novel during the war) while the essay On Being Ill, long considered a specialty “luxe” edition limited to 250 copies, appears to have attracted over 900 orders, far more than a limited edition might have foretold, and which the Press could not in fact fulfill.[68]

Shifting our attention from bookshops to money, our datasets can also address pointed questions about Woolf’s earnings: how did Woolf’s book sales compare to one another and what income did they generate, individually and cumulatively (figs. 8, 9)?

The data confirms some well-known narratives about Woolf’s increasing fame and popularity over time, while enabling the kind of comparative overview that corrects shibboleths about canonicity and reception which the case study approach can inadvertently entrench. It is well known, for instance, that Orlando, for all its shelving problems, became Woolf’s first runaway seller—which the data confirm with sales over 10,000 copies. What is less assimilable without the visualizations, however, is that Orlando’s success marked the beginning of a trend, its sales numbers ultimately lower than almost every other book (and certainly every subsequent novel) Woolf published in the late 1920s and throughout the 1930s. Most strikingly, The Waves, that “mystical eyeless” book that doesn’t make it onto most undergraduate modernist syllabi today, outsold Orlando by 1,336 copies. Testifying to their parity of earning power, despite their wildly distinctive forms, Orlando and The Waves sit side-by-side on the graph of income by title (fig. 9). Flush dominates the graph of orders per title, but so does A Room of One’s Own, coming in as the second most purchased book, at almost 15,150 copies (fig. 8). A Haunted House (1944), a collection of stories published three years after Woolf’s death, is the third highest seller by unit, outselling Between the Acts, Woolf’s posthumous final novel, by almost 2,000 copies, and slightly beating The Years, Woolf’s top seller overall by income (whose popularity landed her the cover of Time magazine in April 1937).[69] Paper rationing during World War II was a serious impediment for publishers, so the Hogarth Press’s decision to allot scarce resources to A Haunted House showed how confident they were in Woolf’s posthumous reception. There is, furthermore, a measure of historical recompense to its success since A Haunted House reprinted a number of groundbreaking short fictions initially published in Monday or Tuesday (1921); that collection had marked Woolf’s foray into modernist experimentalism, where she had hoped critics would see she “was after something interesting” and whose fulfillment Simpkin Marshall’s repeat bulk order, discussed earlier, had seemed to promise (Woolf, Diary, vol. 2, 107). From Monday or Tuesday’s publication in 1921 to A Haunted House in 1944, Woolf made quality literature commercially viable, on sales of £34,446 the equivalent of £2,222,192.29 today.[70] While this is gross income, which doesn’t account for her overall expenditures as a publisher (on book costs or employee salaries), or revenue from publications not included in our data, it is also only a portion of her overall earnings given sales of the hundreds of other authors whose books she, as publisher, brought to market (and which still await transcription), as well as her lucrative career as a literary reviewer.

How Should One Continue to Read the Order Books?

Although we have focused largely on Woolf’s dominant purchasers, our collaborative work traces Woolf’s many distribution networks, highlighting just how extensive is the “mass of information” they proliferate (Woolf, Room, 44). These networked histories—the stories behind the hundreds of stores; the narratives behind the now not quite countless numbers—await further excavation.  But by opening up these Order Books to scrutiny, we have begun the process of making it possible to trace the journeys of some of the thousands of books along distribution routes that reveal the names of once well-known department stores within Britain such as Jarrods (Norwich) and bookshops, large and small, from around the world which act as nominal placeholders for further global modernisms research: Maruzens (Tokyo); The Oxford Bookshop (Mumbai); Nordiska Kompaniet (Sweden); Argus Newspapers (South Africa); the geographic list goes on and on. Some stores lost to time are delightfully whimsical (The Green Frog) while others leave more robust historical footprints, perhaps because they adopted mainstream naming trends. If, for instance, the double “o” bodes cultural authority in the names of Silicon Valley companies today (think Google), in interwar Britain the double-barreled patronymic clearly held gravitas: Bowes & Bowes; Gordon & Gotch; Truslove & Hanson; Douglas & Foulis; Sherratt & Hughes; Goulden & Curry, to name a few. These names tell the histories of fathers and sons, sons and brothers. But not exclusively. There are less obvious, but still compelling, stories needing excavation of women’s bookshop ownership; the Order Books call attention to Collet’s, for instance, the famous leftist bookshop at 66 Charing Cross Road run by manufacturing heiress (and socialist) Eva Collet Reckitt (Rayner, 55–58).[71] Elsewhere, especially in Australia and New Zealand, women leading independent, unconventional lives often led the way in founding bookshops and readership clubs we can now identify as shaping Woolf’s international reception history: Rowlatt and Moore in New Zealand, for instance, co-owned by Eva Moore, founder of the New Zealand Fabian Society; Mary Fisher’s Book Club in Tasmania; or Margareta Webber, who appears frequently in the Order Books, seeming at first to be a private consumer, but who is, in fact, a direct purchaser, placing regular small orders for her bookstore in Melbourne where she employed only women booksellers.[72]

To further interrogate these Order Books, including all author records beyond Woolf’s, requires acknowledging the collective labor such work entails: intensive multidisciplinary and multi-collaborative investment; openness to what technological affordances can bring to the generation and interpretation of humanistic archival data; and the institutional supports needed to confront its logistical challenges. This future work would not only extend the ambitions of our own collaborations, built upon the labor of the Press workers who first chronicled the Press’s business history, but also would reaffirm the importance of sustainability, iterability, and visualizable data within modernist digital humanities.[73] As we publish this article together with  digitizations of some of the Order Books’ pages now on MAPP, in testament to the Order Books as historically rich archival books tout court, we invite others in, so that Woolf’s own “lust for books” might be “revive[d] with the least encouragement” among those eager to follow the destinies of her thousands of British first editions into the global marketplace (Woolf, Diary, vol. 1, 126).