"The Limits of the Imaginable":Women Writers' Networks during the Long Nineteenth Century
The rapid rise of the digital humanities over the past decade has transformed literary study, helping us to discern broader patterns in print culture and media history. Franco Moretti's and Matthew Jockers's respective introductions to distant reading and macroanalysis have fundamentally altered the way in which many scholars now approach literary research.1 In light of their impact on the digital humanities field, I wondered how these methodologies might help us address longstanding critical questions regarding women's social and literary networks during the long nineteenth century. To what extent were women's relationships with fellow women writers—their networks of connectivity—important to their success in a male-dominated publishing marketplace? If, as some researchers have suggested, women formed "alternative networks" to the masculine clubs, universities, and editorial establishments that informed patriarchal print culture, how might we begin to understand these relationships on a broad scale (Easley 112–13)? As Joanne Shattock observes, "Women's literary networks were less obvious and less public" than men's and are therefore more difficult for current scholars to trace and assemble ("Professional Networking" 134). To date, studies of individual writers have revealed the ways in which women's clubs, salons, and other social relationships informed their engagement with popular print culture.2 To understand the ways women functioned within a complex network of private and professional relationships, however, it is necessary to go beyond a single case study approach, which often has the effect of rounding up the usual suspects—canonical women writers—and interpreting their experiences as representative of the field.
Literary scholarship has historically tended to focus on a small canon of writers, leaving their "rivals," as Moretti terms them, to become part of the "great unread" (66–67). Following the work of John Burrows, I hope to bring to light those writers who have "escaped our attention because of [the] sheer multitude" of existent women within the publishing industry in the nineteenth century and to discover why some writers succeeded more than others in creating connections with fellow female authors and in achieving canonical status over the long term (Jockers 26). While outlining his introduction to network visualization, J. Stephen Murphy surmises that "any [End Page 39] medium that groups writers together has the potential to turn writers into conduits through which other writers can be discovered" (iv). Indeed, my macroanalytic approach to network analysis aims to increase our appreciation of the highly influential, hyperconnected writers who often operated behind the scenes of print culture.
Engaging with various digital humanities methodologies such as data mining, distant reading, and network analysis, my essay investigates what new insights can be gained from viewing women's relationships with each other on a comprehensive scale rather than simply viewing the individual network or the network of "important" or "canonical" writers associated with a particular literary period or movement. My macro-network graph reveals how certain women writers functioned as highly visible and centrally located "nodes" within these publishing networks and how these heretofore overlooked writers have been surprisingly influential in the history of women's authorship. Studying women's social and professional networks on a broad scale leads to a deeper understanding of the various ways in which they connected with each other and with print culture during their careers and how this might be correlated with their canonical status, past and present. Of course, any visualization of women's literary and social networks is necessarily a partial one; I conclude with a brief reflection on the gaps and silences encoded in my digital archive source and issues of canonicity in the continually evolving field of digital humanities.
In their quantitative study of reviewer-contributor connectivity in modernist periodicals, J. Stephen Murphy and Mark Gaipa define the term "network" as "a structure of relationships among entities" (52; my italics), and Friedrich Kittler likewise delineates a network as "a structure, the technic whereby cultural exchange takes place" (qtd. in Brake 116; my italics). However, invoking the notion of structure in reference to relationships suggests a state of static kinesis and rigidity. Nathan K. Hensley cautions his reader that there is an inherent danger that our data analysis techniques do not create anything new but merely function as tools for "re-circulating of existing content" and reinforcing what we already know about pre-established relationships between writers within the publishing industry (377). Therefore, I propose that it is perhaps more useful to conceive of networks as "informal, open, multiple, competing, and dynamic" systems, as Simon Potter suggests (622), or as ever-expanding organisms within which nuclei form through temporal and contextual bonds. Taking a distant approach that incorporates nearly seven hundred writers, this study builds on this notion of the network as a fluid model from which new trails of scholarship can be mapped rather than a stagnant structure of evidentiary support for pre-existing arguments. Such an approach, Matthew Jockers argues, draws "attention to general trends and missed patterns" that must be explored "in detail and [accounted] for with [End Page 40] new theories" (29). After all, "the study of literature should be approached not simply as an examination of seminal works but as an examination of an aggregated ecosystem or economy of text" (32). To this I would add that the study of networks should be approached not simply as an examination of seminal authors but as an examination of an aggregated ecosystem or economy of writers.
To explore this broader economy, I focus on a much larger sample of women writers than has typically been studied in existing scholarship on women's networks, aiming to provide a broader understanding of the nineteenth- century woman writer through an examination of contextual biographical and connectivity trends within wider networks of association. Although my connectivity study of nearly seven hundred British women writers captures only a small fraction of the large number of women who lived and wrote during the long nineteenth century, it goes a long way toward expanding our understanding beyond the narrow set of authors usually investigated in women's history (i.e., Jane Austen, George Eliot, and Virginia Woolf—all of whom are included within my data though they are not the focus of this essay). By focusing on a broad sample of women working in the long nineteenth century, I also aim to dislodge assumptions about women's authorship associated with defined literary phases (i.e., Romanticism, Victorianism, Modernism, etc.), which artificially categorize and contextualize women writers according to predetermined assumptions about history, gender, culture, and literary "periods" within the publishing marketplace. Rita Felski calls scholars to fight against lazy usage of literary periodization as a shorthand for the specific relationships informing textual production (573–74). "History is not a box," she warns, and we should not view authors or texts "only as cultural symptoms of their own moment" (574–75).
In order to understand women's networks from a macroanalytic perspective, I mined bio-data from Cambridge University Press's Orlando Project: Women's Writings in the British Isles from the Beginnings to the Present to produce a spreadsheet that tabulated the personal and professional connections between 684 British women writers of the long nineteenth century. I selected the particular writers included in this study from the total of 1,325 entries on the Orlando site based on two criteria: first, they must have been born or have died during the nineteenth century; and second, each must have had at least one signed publication.3 In order to determine each woman writer's connections, I pored through the biographical tabs for each author, noting the names of other women writers that appeared in these narratives; I then cross-referenced these names with the already formulated connections list found under the "Friends, Associates" and "Family" sections, paying special attention to any discrepancies I found between the two—a topic I will revisit in a later section.
The range of connections between writers incorporated into these biographies, as well as into my analysis, included social affiliations such as [End Page 41] friendships, familial relationships, and social/political/literary memberships; professional connections such as co-authorship and editor- contributor relations; and the bonds between reviewers and their authorial subjects. I included any connections between women writers noted in Orlando, no matter how small (e.g., the exchange of just one letter) or how large (e.g., a lifelong friendship), thus attempting to safeguard against the chances of overlooking small but significant connections between writers in my sample. As I worked, I realized I would need to make qualitative decisions about what I felt constituted a "connection" between authors. Though previous studies have often elected to analyze only the professional relationships between authors (i.e., their editing or writing collaborations), I chose to include every possible known connection linking one writer to another.4 Like Bruno Latour, I operated according to the belief that every activity can "be related to and explained by the same social aggregates behind all of them" (8). In other words, all activities, or the means through which connections are formed and measured, are meaningfully "linked in a way that does produce a society" (Latour 8). Social connections between writers, in particular, must be taken into account since they afforded women alternative routes for entering the literary field.
After transferring my aggregated data from Excel to the open-access software program Gephi, I produced a visualization of the complex range of connections between the writers in my study (see fig. 1). In this graph, each node represents a single woman writer and each edge indicates a connection between two nodes (writers). Rather than using relationally weighted edges, which would require me to assign each relationship an arbitrary value of relative significance, I instead selected a visual representation of each connection as either one-sided or reciprocal. For instance, if one author reviewed another but the two writers had no other apparent connection, I marked it as a one-way association originating with the reviewer and with the arrow pointing toward the reviewee. For reciprocal connections, such as familial or social relationships, I used a double-pointed arrow as the visual link between the two authors. In displaying these seemingly simplistic categorizations, I highlight the ways in which women's network formations shifted over the years. In the early decades of the century, my visualization showcases how reciprocal relationships were most prevalent, displaying the alternative ways women entered "into the male-dominated" literary marketplace through connections with other women writers and how this changed over the course of the century as women writers became more accepted and embedded within the publishing industry (Van Remoortel 131). I was also interested in developing metrics for measuring the idea of influence, noting which writers had reciprocal relationships that might indicate influence over one another's writing or career choices and which women were engaged in the one-way relationships often associated with reviewing and [End Page 42]
retrospective writing, activities that contributed to the important work of documenting women's literary history.
In a network visualization, a node derives its "authority, roughly based on how many other nodes … [it] is linked to" (Murphy and Gaipa 53). Nodes with numerous convergent edges, then, appear larger than less- connected ones and are colour coded to reflect their higher "authority." In the full-colour representations of fig. 1, hyperconnected writers Virginia Woolf and Harriet Martineau are signified by red-coloured nodes, while writers with fewer associations are symbolized by a slightly cooler shade of orange. This pattern repeats through lessening degrees of connectivity and nodal sizes, ending with purple-coloured points representing writers generally detached from the main groupings of other female authors (those with few or no connections). Thus, Gephi identifies and accordingly maps network "hotspots"—dense clusters of connectivity focused around particular authorial nodes. We can see that Woolf's and Martineau's relationships span the graph both spatially and temporally, indicating the authors' high level of connection with writers from a multitude of micro-networks and their tendency to make reference, through retrospective writings, to authors of earlier time periods. Because Woolf and Martineau were highly connected in print culture—reviewing, critiquing, editing, and, in Woolf's case, running a printing press—it is easy to see why they commanded such network authority. Though they belonged to different generations, both Woolf and [End Page 43] Martineau had a retrospective trans-temporal tendency in their writing to reflect on past women writers through reviews, biographies, and other forms of historical writing, multiplying the number of their associations and thus surpassing many of their peers in connectivity.
Besides revealing the connections between Woolf and Martineau, my network visualization revealed other, more surprising, "hotspots." In the following section, I focus on three non-canonical authors who emerged as important nodes in my analysis—Joanna Baillie (1762–1851), Geraldine Jewsbury (1812–80), and Margaret "Storm" Jameson (1891–1986). Their lives and work are less visible than those of Woolf and Martineau, whose careers and networks have been studied at length and whose canonical status is a commonplace in feminist scholarship, yet Baillie, Jewsbury, and Jameson were among the most highly connected writers of their respective generations, as we can see both from their central positions within the macro-network as well as from the weight and colour of their representative nodes (see fig. 1). It could even be supposed, based on this data visualization, that they functioned as the nuclei of distinct, clustered micro-networks around which other nodes seem to have positioned themselves. It is interesting to note that in all three cases, high connectivity did not, it seems, translate into canonicity. Their texts are rarely taught in college classrooms or included in anthologies of women's literature.5 It is my hope that my own digital humanities project can help to recover and reclaim the work of these writers, which often took place behind the scenes in the form of reviewing, editing, participating in writers' groups, and engaging in various forms of nineteenth-century social networking. While this activity may not have ensured their lasting presence in the literary canon, it nonetheless was influential in the history and production of women's writing during the long nineteenth century.
Until this point, I have made an argument for a macroanalytic approach to the study of women writers' history and biography. In the next section, I address the lives and network structures of three individual writers; however, I would like to emphasize the difference in my approach to these case studies. Rather than utilizing a traditional, singular case-study entry point into network analysis and taking an already well-known or "significant" writer and producing a network of known (or heretofore undiscovered) associates, or mapping the interconnectivity of a group of writers who were part of the same literary circle or wrote for a particular periodical, I discovered the objects of my case studies and their subsequent importance organically via my production of a macro-network of women writers. Using macroanalysis as the entry point to the study of women's connectivity allowed me to then effectively reverse-engineer an individual case study from a larger set of aggregated data. As a result, I was not only able to "discover" these women writers for myself (as I had not previously encountered their names in my own scholarship) but was also able to discern how surprisingly connected [End Page 44] and central they were to the formation of women writers' networks in the long nineteenth century. In the following sections, I take a biographical approach to three nodes in my visualization to explore how the authors they represent came to be so connected, and thus so seemingly influential, in print culture.
Joanna Baillie (1762–1851)
During the latter half of the eighteenth century, when Joanna Baillie came of age as a writer, the literary marketplace became increasingly female dominated. Upper-class women found themselves with an abundance of leisure time, and the print industry responded to fill their void with a flood of novels catering to what Ian Watt terms an "easy vicarious indulgence … in sentiment and romance" (45, 290). Baillie's literary aspirations took root in this romantic atmosphere while at a Glasgow boarding school, despite attempts at suppression by her deeply religious father. After his death, Baillie travelled with her mother and sister to London at the age of twenty-one to keep house for her brother. This fortuitous move brought Baillie into contact with her aunt, Anne Home Hunter (a published poet of small renown), who introduced her to literary society. Baillie quickly became a regular at her aunt's weekly literary salon, which included attendees such as Anna Laetitia Barbauld, Frances Burney, and Elizabeth Montagu. Encouraged by these examples of successful female authorship, Baillie published her first volume of poetry, Poems: Wherein It Is Attempted to Describe Certain Views of Nature and of Rustic Manners, in 1790.
Reviews of her work appeared in periodicals such as the Monthly Review and the Eclectic Magazine, and the affirmative reception of her poems as "true and lively pictures of nature" gave Baillie the confidence to start working in a new medium: drama (Brown, "Joanna Baillie"). She began incrementally, writing introductions and epilogues for friends' productions before eventually moving into producing complete plays of her own. From 1798 to 1812, Baillie wrote and published three texts that, when later collected, would comprise her most successful work to date—a series of verse dramas entitled Plays on the Passions. After the book of the first play appeared under the name "Anonymous," London was abuzz with speculation regarding the writer's identity, with most assuming male authorship until Baillie came forward in 1800. Though the playscripts themselves garnered mixed critical reception, their successful staging—produced on Drury Lane with John Philip Kemble and Sarah Siddons in leading roles—led Baillie to "fame almost without parallel," as Harriet Martineau later wrote, becoming a dramatist "second only to Shakespeare" in both talent and fame (358).
Baillie's biography reveals an author for whom the formulation of relationships with other women was instrumental in building her success. Unlike her hyperconnected successors who emerged later in the century, Baillie engaged in relationships with fellow women writers that were less [End Page 45] likely to be the one-way encounters enabled by professional opportunities in the literary marketplace (e.g., through book reviews, editorials) but were more reciprocal in nature: built through mutual correspondence, social encounters at her aunt's literary salon, and collaborations with fellow members of the theatrical world (see fig. 2).6 Baillie's hyperconnectivity is likely more attributable to her decision to work primarily as a dramatist rather than as a poet or novelist. Writing for the stage necessitated in-person connections with actors, directors, and other collaborators, some of whom also worked in the publishing industry. Georgiana Cavendish, Duchess of Devonshire, wrote an epilogue to one of Baillie's early productions, De Montfort. Baillie collaborated with Felicia Hemans on numerous occasions to secure actors for their respective shows, and she frequently attended and reviewed Jane Porter's plays of the period. Later, Baillie's work as the editor of a little-known collection of Scottish poetry, A Collection of Poems, Chiefly Manuscript, and From Living Authors (1823), also afforded her an opportunity to develop both singular and reciprocal connections to the authors included in the volume (McLean).
Simon Potter notes that the networks of late eighteenth- and early nineteenth-century women writers tended to be "only loosely structured," based on personal connections and often characterized by a "tendency toward homogeneity" owing to the fact that they were less public than
[End Page 46] male literary communities (634). During the early portion of the nineteenth century, even as professional opportunities in the public sphere expanded, writing was a means through which women could work safely in the private sphere of the home, "with only the … product of the author being necessarily in the public domain" (Thompson 69). This characterization held true for Baillie's early network connections, formed in a private literary salon with authors of similar backgrounds and experiences as herself. However, as her career progressed she made more public and diverse connections through her work as an editor and dramatist, and her network became far less homogenous than might otherwise be assumed. Yet, as important as her networked connectivity was, it was not yet considered the sort of work that led to canonical status. When the plays and volume editions of her works went out of print and out of fashion, her legacy was also largely forgotten, remembered only as a footnote in accounts of her broader, mixed-gender literary network, especially her relationship to the canonical novelist Sir Walter Scott. A networked analysis of Baillie's career as a dramatist and editor restores her status as a highly connected writer who was engaged with fellow women writers in ways far more substantial than attendance at a literary salon and who played an influential role in facilitating and interpreting their contributions to literary history.
Geraldine Jewsbury (1812–80)
Geraldine Jewsbury was born in Derbyshire twelve years after her sister, Maria Jane Jewsbury (1800–33), whom she eventually followed into the publishing business—producing a large number of novels and journalistic pieces, and later working as a publisher's reader. She also served as an editor and as a prolific critic and reviewer for the Athenaeum, for which it is estimated she produced well over two thousand reviews.7 Jewsbury faced significant trials early in life; the death of her mother and sister left her, at the young age of twenty, as the sole caretaker for her severely ill father. Experiencing a profound crisis of faith, Jewsbury discovered comfort in the literary texts that Maria had left her, especially the works of Thomas Carlyle. Boldly writing him a letter regarding her similar views of the themes of religious doubt in his work, she established a lifelong connection with Carlyle and, more importantly, with his wife, Jane Welsh Carlyle.
After corresponding with the Carlyles, Jewsbury was inspired to begin her own publishing career. She began as a novelist, producing several books throughout the late 1840s and early 1850s, including Zoe: the History of Two Lives (1845), The Half Sisters (1848), and Marian Withers (1851), before transitioning to a career as a reviewer for the Athenaeum, among other journals, and a reader for Bentley and Hurst & Blackett. Her work as reader and reviewer tended to foster one-way connections, distinguishing her from Baillie, who, as we have seen, was more likely to form mutual connections with those in her literary and theatrical orbit. Though Jewsbury experienced reciprocal [End Page 47] relationships, such as those with Jane Welsh Carlyle, Harriet Martineau, Mary Russell Mitford, and Elizabeth Barrett Browning, we can also see a greater influx of one-way (lighter-coloured line) relationships in her personalized network (see fig. 3).
As a reader, Jewsbury evaluated and made recommendations on the work of writers such as Ouida, Mary Elizabeth Braddon, Rhoda Broughton, and Ellen Wood, who appear as smaller nodes in her network visualization. During her prolific reviewing career, Jewsbury also formulated one-sided connections with Charlotte Brontë, Ellen Mary Clerke, and Charlotte Yonge, among others—as was typical of the reviewing trade at mid-century. Jewsbury still engaged in networking as part of her personal relationships, through letters and visits (most notably with the Carlyles), as was largely the case for Joanna Baillie, but she was more likely to engage in one-directional, professional relationships with her fellow women writers.
One possible reason for this connective shift during the mid-century was the "number of new and enabling public spaces for women writers [that] emerged from the 1840s," which provided an increasing number of opportunities for networking among women writers (Shattock, "Researching Periodical Networks" 61–62). New opportunities in the publishing field—writing, editing, and reviewing—and the rise of the "celebrity author" in the latter half of the nineteenth century changed not only the literary marketplace and the publication practices of the period but also the ways in which authorship, especially for women, was defined. Publishers began "abandoning the commitment to anonymity" and instead used "by the author of …" or authors' initials, which allowed readers to identify their favourite writers
[End Page 48] (Jordan). Later in the period, editors regularly advertised authorial identities when marketing their periodicals. Alexis Easley explains how this policy "made it increasingly difficult for women to engage in low-profile literary careers" (5). Thus, we are able to trace the roots of the shift from personal to blended personal-professional relationships for Jewsbury, as other authors—like Isabella Banks and Eliza Lynn Linton—were able to identify her reviews toward the latter half of her career.
This fundamental shift from the largely reciprocal composition of Baillie's more personal network to the heterogeneity of Jewsbury's professional relations follows Simon Potter's postulation that "from the mid-nineteenth century onward … technological and commercial changes modified earlier patterns of interconnection and privileged particular links at the expense of others" (634). Even though Jewsbury held a reputation "among her contemporaries as a major influence on Victorian literature, her contributions as author and critic have faded into obscurity" (Brown, "Geraldine Jewsbury"). Feminist scholars such as Elaine Showalter have attempted to rescue Jewsbury from obscurity by directing attention to her novels. Yet focusing on Jewsbury's fictional works does not do justice to the highly connected nature of her literary practice. Only through careful study of her behind-the-scenes, often anonymous, networked labours in the publishing industry can we begin to assess her broader contribution to the history of women's writing.
Margaret Ethel "Storm" Jameson (1891–1986)
As a woman who memorably described herself as the "invisible aunt of English letters," Margaret Ethel "Storm" Jameson has unsurprisingly received scant scholarly attention to date (Maslen 401). Born in Yorkshire in 1891, Jameson was first published in 1913, at the age of twenty-two. Married at a young age, she was acutely aware of the need to support her family as her husband finished his university studies and began teaching. The drive to work and to write propelled Jameson further into the publishing world than she at first dreamed possible. At the outset of her writing career, she lacked confidence in her abilities, noting that the "singular badness" of her first attempt "proves that I was not a born novelist" (Jameson 3). Nonetheless, Jameson continued writing in order to make ends meet, transitioning into writing copy for an advertising agency, publishing reviews in periodicals such as the Egoist, and performing research work—all of which gave her the background and experience necessary to land the position of editor of the Commonwealth magazine and London representative of publisher Alfred A. Knopf. Jameson's budding political activism, furious work ethic, and strong literary output of novels addressing the injustices of World War I led her to become the first female president of the English PEN (Poets, Essayists, and Novelists) Centre. Through PEN, she forged an ever-widening network of connections to both male and female authors, the latter including Rebecca West, Vera Brittain, Winifred Holtby, Naomi Mitchison, and Virginia Woolf (see fig. 4). [End Page 49]
Given her high degree of professional responsibility, it is unsurprising that Jameson's connections with fellow writers were numerous and diverse. As her network visualization demonstrates, Jameson was linked to a considerable number of other authors, with an even mix of one-sided and reciprocal connections. Jameson's work as a reviewer, like Jewsbury's, led to an early proliferation of one-sided connections with fellow authors. However, by the beginning of the twentieth century, such reviews were more likely to be signed. Thus, even though the reviewer and the reviewed author might never meet, their names would be linked within the publishing world. For example, author Dorothy Richardson was recruited by a publisher to write a hybrid review-response piece to Jameson's essay "Bored Wives," a feminist critique of suburbia as an intellectual desert. Because Richardson's fame was nearly equal to Jameson's at the time, their names were linked, even though they did not interact socially or professionally. Later, as PEN's president and self-appointed organizer, Jameson worked closely with other authors through mutual correspondence and in-person relationships, formulating the basis for her reciprocal connections.
If, as Murphy and Gaipa suggest, "modernism's emergence" directly resulted from the "circulation and connectivity" of period authors and their literary associations (31), Jameson seems to exemplify its success as a literary movement. So why has her self-described legacy been rendered all but invisible in literary history? One possible answer may be found in the nature of her connections with her contemporaries. As president of the English PEN Centre, Jameson was a fully engaged leader—organizing and running meetings, proposing and enacting policy change, and recruiting and retaining [End Page 50] organizational membership. Jameson was also wholly invested in supporting Allied efforts during both world wars—speaking to politicians and the public, writing and publishing pieces to fight back against Nazism and fascist ideology, and even raising funds for writers fleeing the terror and persecution of Europe during World War II. Jameson was abundantly engaged with and rooted in the specific issues of her time, which was reflected in her writing. Consequently, Birkett and Briganti note, when her first-generation Modernist peers insisted that truly great literature must be taken "out of history and [cut] off from its political moorings," "Jameson was easily nudged out of a literary canon" (10). Cast as an ephemeral novelist of a particular cultural moment, Jameson faded from literary history in favour of more well-known peers such as Virginia Woolf, whose novels still addressed the most pressing issue of the time—the war—but more fully embodied Modernism's ideals of timelessness and experimentation.
Though both Jameson and Woolf also engaged in various other forms of networked connectivity, such as reviewing, publishing, and writing about fellow women authors, Jameson's work was deemed too topical for lasting greatness, and her other networked activity within print culture, though arguably just as significant as Woolf's, resultantly disappeared from visibility. Though Virago Press reprinted Jameson's best-known novel, the autobiographical Journey to the North (1960), and second-wave feminism briefly rekindled scholarly curiosity in Jameson's life and work during the 1980s, interest in Jameson's career and texts ultimately failed to gain traction, and she, like Baillie and Jewsbury, was cast into the shadow of her (largely male) contemporaries. Instead of being recognized as central figures in nineteenth-century women's literary print culture, all three women have been reimagined as minor figures in the narratives and networks of canonical writers such as Sir Walter Scott, Thomas Carlyle, and T.S. Eliot.
Reading Broader Patterns in the Long Nineteenth Century
"Visualizing large data sets can reveal structural transformations as they took shape" in a variety of literary eras (Murphy and Gaipa 33). In graphing the network of seven hundred interconnected women writers, I was able to examine a temporal span long enough to see repeating patterns. As fig. 1 illustrates, three distinct clusters of networks emerge from the overall figure, centred around similarly sized nodes, representing three individual writers distributed over three nearly equidistant periods of forty to fifty years. Each micro-network represents a cycle within the macro-level whole, as Gephi is additionally able to measure the modularity of a set of network data, allowing it to measure the "degree to which a network is divided into smaller communities" (52). In graphing such a large span of time, this feature was particularly useful not only in foregrounding nodal hotspots of individual [End Page 51] authorial connectivity but also in distinguishing temporal connectivity pockets clustered around key women writers who flourished from 1750–1800, 1850–1900, and 1900–50. A new network arises as one generation of writers gives way to another and new women writers become central nodes within a cycle of densely connected, intertwined women's writing communities.
As Moretti observes, "cycles constitute temporary structures with the historical flow" of a network graph (76; original italics), often pointing us to a generalized time period or literary movement, but the very nature of network graphs shows us that the interconnections that occur between authors of such movements often transcend these movements. Romantic, Victorian, or Modernist women writers cannot simply be bundled together and made to fit into predetermined contextual boxes. Though a surface reading of the macro-network graph suggests that repeating patterns of homogenous authorial clusters appear at regular intervals centred around certain women in specified time periods, closer inspection of individual case studies reveals the cross-temporal nature of the connections between these women, suggesting much more complex relationships than have hitherto been understood. The authors simultaneously cultivated connectivity with one another in the present and wrote retroactively about their literary foremothers. Not only were they involved in documenting women writers' uniquely gendered history within the publishing industry but many also engaged in analyzing and critiquing the way that the works of these same women writers had entered, existed, and endured in the marketplace alongside those of their dominant male peers.
In light of this temporal entanglement with both past and present, network analysis, I contend, should be conceived as a tool for investigating multiple points of origin and entry into the study of writers in the nineteenth century rather than as the end-structure proof of pre-existent theories. Networks open lines of inquiry rather than closing off previously held assumptions or hypotheses. My study aims to expand the canon to include, or at least consider, the kinds of important roles women played both on the literary public stage and behind the scenes. Given their capacity for influencing fellow women writers and reclaiming female authors of the past, perhaps networked relationships are a better measure of any given writer's influence in literary history than their published works of imaginative literature. This idea of the importance of influence is complex, referencing matters of literary style and output, celebrity, connectivity, opportunity, critical acclaim, and so on. However, bids for canonicity by women writers have long been made on a variety of literary claims; from saintly genius (Christina Rossetti) to imaginative virtue and brilliance (Charlotte Brontë) to intellectual power and exemplary leadership (George Eliot) (Chapman, "Achieving Fame" 78–83). If nothing else, my study suggests a means for considering concepts of connectivity and influence as other viable criteria for canonicity. [End Page 52]
Archival Gaps and Silences: The Orlando Project
In examining the lives and networks of Baillie, Jewsbury, and Jameson, my goal was to illustrate the way that women's networks evolved structurally and temporally during the long nineteenth century. Of course, the validity of network analysis in defining both presence and absence in women's literary history depends on the quality of the source data used as the foundation for research—in this case, the Orlando Project database. As Lauren Klein contends, we must think of the digital archive "not as a neutral repository of knowledge, but instead as a tool for exposing the limits of our knowledge" (684). As scholars, we must consider the "epistemological assumptions" built into our use and visualization of data and "engage in a critical description" of such tools when putting them into practical use (Drucker 248).
Just as I have mined the digital material of the Orlando database, I must also interrogate the project's gaps and silences. Given the stake I have in the data's validity, "it is vital," as Paul Fyfe argues, to "account for its history" (552). The Orlando Project database is a relatively new resource in digital scholarship, rooted in the compilation of entries on women authors across several eras and aesthetic movements by editors Virginia Blain, Patricia Clements, and Isobel Grundy for their reference volume The Feminist Companion to Literature in English: Women Writers from the Middle Ages to the Present. By the end of the project, they found that the collected content had grown too large to be feasibly contained in a hard-copy text. In 2006, the latter two editors, along with scholar Susan Brown, turned their attention to building a digital project on feminist literary history.8 The first archival iteration, launched in 2010, incorporated entries on just over one thousand British women writers—with British male writers and international women writers occasionally included at the editors' discretion. In 2013, the database underwent another major expansion, bringing the total number of writers listed to 1,325 entries.9 Yet, even with this fairly large cross-sectional sample of women writers, the archive is (and will remain) incomplete. The nearly seven hundred women writers I mined from the Orlando Project database do not begin to approximate the actual number of women writers who published during the long nineteenth century, which numbered in the tens if not hundreds of thousands. Thus, while the Orlando sample is commendably large compared to the number of authors usually covered in reference books or in scholarship on women's authorship, it is simultaneously a limited selection that perhaps reflects the particular knowledge and interests of the editors who created it.
Probing the data reveals gaps at both the micro- and macro-level. For the former, within individual author listings, I grappled with constraints in collecting writer connectivity data created by what page editors opted to include in the "Life and Writing" summaries, as well as under the "Friends", "Associates", and "Family" tags. In their introduction, the editors describe their methodology and reasoning: [End Page 53]
Entry length is governed by a range of factors. The first is the historical importance, as we see it, of the writer. Authors with full entries have been picked for historical or literary interest (or both); a few treated only briefly in timeline material are candidates for full entries in some approaching update; a few (mostly from early periods) have no "Life" screen because information about them is so sparse.("Literary History with a Difference")
Michel Rolph-Troulliot describes this process as "the moment of fact assembly," during which archival silence becomes encoded (Klein 663). The privileging of certain authors over others, as reflected in the degrees of completeness for individual entries, clearly shows that not only information availability but also commercial concerns, scholarly interest, and marketability were taken into account. Given that the Orlando Project relies on paid subscriptions for financial viability, its selection bias is understandable, yet this selectivity also makes it difficult to judge the influential status of writers within print culture in an unbiased and unfiltered way. On a macro-level, the difficulties I encountered when collecting my data on women writers' relationships reinforce what Emily Midorikawa and Emma Claire Sweeney term the "mythologizing" of female authors "as solitary eccentrics or isolated geniuses" (13). Archives necessarily isolate and detach a subject or individual component in order to present information, but this treatment simultaneously removes individuals or texts from their connections and contexts. It is the work of the scholar, then, to act as the mediator who mines the subject/component of a database or archive and makes the connections visible to readers in order to highlight the importance of a specific author or text. But the question remains, how do we perform this work without privileging canonical authors over others? How (if at all) can we situate women writers within (cross-)temporal communities in an unbiased way?
Approaching archival data and its practical usages from a macroanalytic standpoint allows scholars to pose academic questions in a way that at first seems relatively free from the conventional biases that inform literary study. Yet, if the information included in the database is itself informed by what we already know about authors from conventional research—with canonical women writers' lives and works being much better known than those of their non-canonical sisters—then how can we ever hope to achieve an unbiased vision? "If we mine only for 'x' … we are getting a very partial intellectual picture" (Onslow 3). Perhaps we can hope for only partial truths when utilizing any data set. Nevertheless, it is a valuable exercise to try to gain a sense of distance from our usual objects of analysis and experiment with methods that defamiliarize them in engaging ways. [End Page 54]
By utilizing network analysis tools and data visualization, we are able to see not only what is present but also what is not. Visualization transforms data so that we can examine it in context with other sets of data points and thus consider it in relative terms. In analyzing connectivity data for women writers of the long nineteenth century and discovering surprisingly important nodes within these networks—such as those of Joanna Baillie, Geraldine Jewsbury, and Margaret Storm Jameson—we can begin to redefine what constitutes a writer's influence in literary history. We are invited to confront our own culpability in overlooking certain writers in favour of others. As Lauren Klein argues, the realization of "absence challenges us as critics to make the unrecorded stories that we detect—those that we might otherwise consign to the past—instead expand with motion and meaning" (675). Repeated scholarly attention to a select group of women writers has not only solidified their place within the canon but also reinforced "readers' familiarity with these authors," underscoring "their perceived worth and significance" (Murphy and Gaipa 42). By engaging with network analysis on a macroanalytic scale rather than taking the canonical individual or coterie as a starting point, I hope to raise fresh questions about women authors' experience in the literary marketplace—and about the archives that claim to represent them. A distant reading approach, followed by the close reading of biographical data, promises to dislodge our preconceptions and stereotypes about women's writing, canon formation, and the idea of influence and importance in literary history.
ANDREA STEWART recently received her MA from the University of St. Thomas, where she also works as an editorial assistant for Victorian Periodicals Review. Her research areas of interest centre on Victorian literature and its intersections with modern media culture studies, as well as quantitative analytic approaches to mapping biographies and networks of British women writers of the same period.
1. For example, Natalie Houston's ongoing map of the textual relationships between Victorian poets and publishers ("Toward a Computational Analysis of Victorian Poetics") or Anne DeWitt's data visualization study concerning reviews and the Victorian theological novel ("Advances in the Visualization of Data: The Network of Genre in the Victorian Periodical Press").
3. The first component is, of course, intrinsic to the scope of my analysis. The second was included for assurance of mapping connections between women writers specifically. Texts published pseudonymously or anonymously could not be attributed with absolute certainty to a female author, so certain authors could not be entertained for this study.
4. Examples of previous studies that focused on group networks include Patrick Leary's examination of the Punch brotherhood, and P.D. Edwards's exploration of Dickens and his circle of writers.
5. Indeed, as Barbara Onslow laments, the names of writers such as Geraldine Jewsbury are familiar only to "scholars specializing in nineteenth-century feminism" and not to general readers (1).
6. One useful feature offered by Gephi is the ability to create a personalized 3-D graph for each author showing the author's connections on an individual scale to other writers in the larger whole.
7. The exact number of Jameson's reviews for the Athenaeum is unknown since many of the periodical's reviews remain unattributed, "Anonymity [having] remained entrenched in the reviewing of the 1850s" (Shattock "'Orbit' of the Feminine Critic").
8. Brown, Clements, and Grundy serve as project editors for the site and oversee content development, each taking a section of history. Brown oversees women's writing from 1820–90, Clements covers writing from 1880–present, and Grundy manages writing from the "beginning" to around 1830. The archive also lists seven co-investigators, fourteen technical personnel, eight post-doctoral fellows, one hundred and six research assistants, seven administrators, and two external consultants, all of whom have worked in various capacities to build and launch the Orlando Project into its current digital form.
9. The database updates its records on a biannual basis, but additions are fairly minor (ten to twenty-five new writers may be added), and existing writers' content also undergoes smaller revisions.