CHAPTER 10

Prototyping Personography for The Yellow Nineties Online

Queering and Querying History in the Digital Age

Alison Hedley and Lorraine Janzen Kooistra

Humanities databases such as Orlando Women’s Writing in the British Isles, the Victorian Women Writers Project, Women Writers Online (WWO), and Agents of Change demonstrate that digital tools support feminist efforts to recover historical women’s authorial, artistic, and editorial work. Just as importantly, but perhaps less obviously, the methods and models of these projects also enact a feminist critique of traditional historiography. At Digital Diversity 2015, a gathering that marked Orlando’s twentieth anniversary, Julia Flanders noted that the WWO offers not only an alternative literary-historical narrative but also alternative methods for practicing literary history. Susan Brown addressed the ongoing challenges Orlando faces in balancing historical diversity with the need for standardized metadata, asserting, “We don’t disambiguate social categories, but in a sense, set them at war with each other” (Flanders, Brown, Wernimont, and Smith, “Feminist Literary History”). The reflections offered by Brown, Flanders, and many others at the Digital Diversity conference attest to an increasing theorization and application of feminist praxis in the digital humanities.

Feminism, digital humanities, and periodical studies intersect in our collaborative work on The Yellow Nineties Online, a scholarly site for the study of fin-de-siècle aesthetic magazines through marked-up digital editions, archival paratexts, and critical essays. Our affiliated project, the Yellow Nineties Personography, seeks to investigate the periodicals in our corpora as records of creative collaboration, with each issue representing, in James Mussell’s words, “a particular configuration of contributors attempting to produce an object for the market” at a specific time and place (120). As a contextual dataset about the individual persons encoded in textual markup, a personography enables us to trace the social configurations and networks that produced specific aesthetic magazines in 1890s Britain. Drawing on the markup of four Yellow Nineties Online digital editions—the Evergreen, the Pagan Review, the Savoy, and the Yellow Book—the Personography aims to develop a biographical dataset into manipulatable format, thereby facilitating research on the complex network of editors, authors, illustrators, and readers who participated in fin-de-siècle British print culture.

While our Yellow Nineties biographical dataset is modest by digital standards—it currently includes over a dozen attribute fields, but only 351 persons—our purpose is more ambitious in its theoretical, methodological, and historical aims. Through our personographic work, we aim to make visible the creative productions of women and other historically marginalized persons whose contributions to fin-de-siècle aesthetic magazines have been largely lost to view. Additionally, like other feminist database builders, we also strive to make visible our process for modeling this historical knowledge. Simply recuperating marginalized persons and their work is not a robust response to the cultural and intellectual paradigms that have historically displaced such persons, as Alison Booth and Jacqueline Wernimont point out (Booth, 18; Wernimont, paras. 5–6). Personography has its origins in prosopography—a traditional historiographic method that develops a collective biography of a population by flattening out differences and anomalies to create a typical subject. Along with our recovery of individual persons, we want to challenge the standard classification practices in prosopography, as well as those in computational markup language, that work to make certain persons historical nonentities. We also seek to refute the traditional prosopographic assumption that data are transcendent and unambiguous, rather than contingent and fluid. In other words, we want to queer the data by interrogating the structural formations of the normative, questioning differences within categories, and being self-reflexive about our processes (Giffney, 19). As Noreen Giffney summarizes, “queer thus denotes a resistance to identity categories or easy categorization, marking a disidentification from the rigidity with which identity categories continue to be enforced and from beliefs that such categories are immovable” (21). Given our understanding of identity as fluid and contingent, how might we adequately catalog the human subjects in our dataset, and what are the stakes of such a process? Can we develop a database that allows us to queer, as well as query, the data? This is what the Yellow Nineties Personography aims to model. Our work applies a critical feminist lens to prosopographical and personographical methods and encourages our users to do likewise.

In documenting the contributors to the Yellow Book (London 1894–1897), the Pagan Review (Sussex 1892), the Savoy (London 1896), and The Evergreen: A Northern Seasonal (Edinburgh 1895–1897), the Yellow Nineties Personography opens up an avant-garde community of writers and artists for discovery and analysis. Despite the brevity of its print run, the Yellow Book is the defining cultural document of the fin de siècle, lending its color to “the yellow nineties” and influencing both artistic practices and social critiques (Jackson, 32). According to Holbrook Jackson, contemporary readers associated the Yellow Book “with all that was bizarre and queer in art and life, with all that was outrageously modern” (45). At the end of the nineteenth century, the “outrageously modern” included decadent and feminist challenges to heteronormative, patriarchal norms in both literature and art. An idiosyncratic forerunner to the Yellow Book, the Pagan Review was published in Sussex by William Sharp, connecting rural neo-paganism and the Celtic Revival to the fin-de-siècle avant-garde. The Savoy, a London-based rival to the Yellow Book, was edited by Arthur Symons and Aubrey Beardsley after the latter was dismissed from the Yellow Book following Oscar Wilde’s arrest for gross indecency in April 1895. This move was intended to reassure both the consuming public and some of its own contributors that publisher John Lane had purged the magazine of its queer associations: at the time of his arrest, Wilde was carrying a “yellow book” (actually, a French novel). Meanwhile, in Edinburgh, the Evergreen countered decadence with regeneration. Its editors designed the four volumes of its print run around the seasons to capture the Celtic politics of the Scottish Renaissance and promote urban renewal and nationalism. Just as the Yellow Book and the Savoy shared Beardsley as an art editor, the Pagan Review and the Evergreen shared literary editor William Sharp, and there are many other contributor crossovers to link the four magazines. Contributors, associations, cultural concerns, and artistic practice interwove an avant-garde network stretching from Scotland in the north to Sussex in the south.

The historical record offers many challenges to documenting the persons in our personography with sensitivity to non-normative identities and relationships. Indeed, the defining category of our work, “person,” is neither self-evident nor transhistorical. More than a quarter of the 351 individuals who contributed art, text, and editorial expertise to the Yellow Book, the Pagan Review, the Savoy, and the Evergreen did not fall within the standard personhood or status classifications of late-Victorian law and society. In other words, these magazine contributors were legal nonpersons and/or socially non-normative individuals. Under British law at the time, women, for example, were nonpersons with regard to fundamental individual rights, and homosexuals were criminals. On the other hand, the reading public conferred personhood (sometimes gendered) on many pseudonymous contributors without being able to verify their legal status or biological identity. In terms of non-normative interpersonal relations, historical evidence suggests that many contributors to these magazines had romantic and familial relationships that were illegal, invisible, or counter to Victorian mores. While some of these unconventional practices were common enough to have designations in cultural discourse, for example, “female marriage,” none were officially recognized as legitimate by legal and social institutions (Marcus, 12).

The limited range of terms and means for recognizing such individuals is particularly notable in a period known for its preoccupation with systemizing and classifying all kinds of information. It was in the nineteenth century, as Michel Foucault observes, that humankind began to be framed as “the human species” and the human individual as “a figure of population” (Security, Territory, Population, 105, 111). This category shift was aided by advances in statistics and demography, which were beginning to be used to document and manage the British population. Constructing the norm as their standard, statistics and other emerging social science practices sought to systemize human bodies and behaviors (Canguilhem, 44). Indeed, it was in the Victorian period that statisticians adopted the practice of normal distribution, otherwise known as the bell curve, as a method for mapping human data. As a result, the word “normal” entered the dominant social discourse to signify the paradoxical concept of the ideal average (Davis, 10).

The Victorian impulse to classify bodies and behaviors extended to fundamental dimensions of personal identity such as gender, sexuality, and interpersonal relationships. In the period during which our magazines were produced and read, intellectuals such as Havelock Ellis and Edward Carpenter, proponents of the new scientific field of sexology, formulated the binary concepts of homosexuality and heterosexuality (Marcus, 22). These concepts quickly gained traction as useful social categories, and a discourse of sexual deviation from the standard, heterosexuality, proliferated terms for different types of “inverted” individuals and relationships. In his History of Sexuality Vol. 1 (1990), Foucault famously contends that late-Victorian society implemented such categories in order to delineate, and therefore control, previously unspoken behavior (34). The Yellow Nineties Personography project sets out to challenge these Victorian mechanisms and binaries. Such categorizations entrenched a division between normalcy and deviance while further marginalizing any individuals and behaviors that did not easily fit into binary categories. Notably, we have found scant evidence that any of the magazine contributors self-identified using terms such as “homosexual” or “invert.” As atypical individuals, their lives were obscured in the historical record.

Victorian statistics and demography also engendered prosopography, a methodology for studying historical groups. The development of the prosopographic method is roughly contemporary with the late-nineteenth-century magazine community we study and thus connected to the cultural processes and classification practices that left some historical persons unmarked. Prosopography involves studying the available data of individuals in a particular group and extrapolating a collective biography of that group’s “typical” subject. Historians tend to use prosopography to study marginal populations, rather than dominant ones, because a group study can offer more insights about people for whom few archival records exist (Booth, “Recovery 2.0,” 17). A prosopography’s structure and contents, therefore, are shaped by the primary source texts from which it gathers biographical data on a group of people. The Oxford English Dictionary cites 1896 as the first instance of “prosopography” in this sense. Late-Victorian prosopography focused on normalization, laying unique anomalies aside to establish straightforward, correlative relations between a population group’s identity and its common characteristics.

As practiced from the late nineteenth century onward, prosopography conceptualizes individual people as population units, reflecting a Victorian emphasis on normalization. In their comprehensive overview of this historiographic practice, Koenraad Verboven and colleagues state that prosopography’s goal is to make visible the “particular characteristics of the [identified] population as a whole” (36). Studying a whole population can be a more effective strategy for ascertaining social patterns than generalizing from a few individual cases. Prosopography therefore concerns itself not with “the unique” but with “the average, the general and the ‘commonness’ in the life histories of more or less large numbers of individuals.” Echoing his Victorian predecessors, Verboven states: “The individual and exceptional is important only insofar as it provides information on the collective and the ‘normal’” (37).

Traditional prosopography, as Verboven defines it, assumes that classifying multiple persons as a “more or less homogeneous group” is an unproblematic historical method (39). But by what rubric can prosopographers determine homogeneity? In some ways, the individuals included in the Yellow Nineties Personography constitute a homogeneous group: all of them participated in 1890s journalism and, more specifically, contributed to British avant-garde periodicals. If we were to summarize this group’s characteristics in one typifying profile, we would conclude that the average contributor was a middle-aged, British, Caucasian male who studied at Oxbridge, contributed fiction to the Yellow Book, published under his own name, and regularly socialized with Henry James and Edmund Gosse. This unambiguous profile does not begin to describe many of the individuals in our personography’s dataset. Our biographical research has uncovered ambiguous data that, we argue, deserve more complex documentation and modeling. The following examples of historical individuals who contributed to the magazines in our project illustrate the inadequacy of prosopography’s normative categories of marital status, sex, and occupation.

Like many female Yellow Book contributors, Rosamund Marriott Watson and E[dith] Nesbit veiled their marital status and biological sex with pseudonyms and initial-only signatures. For many years, Marriott Watson was known in personal circles as a woman but in a much broader community of readers by her masculine pseudonym, Graham R. Tomson. Notably, she published under both signatures in the Yellow Book, but her two names—and the coded genders they evoke—remain noncorrelative. There is a cultural story here we do not wish to elide: as Graham R. Tomson, the poet enjoyed significant critical recognition and was even suggested for the laureateship when Tennyson died in 1892. As the scandalous Rosamund Marriott Watson, she left her husband (Arthur Tomson) and child in 1894 to live, unmarried, with H. B. Marriott Watson (who also published in the Yellow Book). As a result, the poet was no longer “known” in some social circles and her poetic profile as Graham Tomson was erased in essays on contemporary women poets by critics such as Richard Le Gallienne and Andrew Lang (Hughes, 222). Her change in signature thus represents both a shift in social standing and an altered authorial status and oeuvre.

Known socially as “Mrs. Hubert Bland,” E. Nesbit always published under her maiden name and initial, thus veiling both her sex and her marital status. H. G. Wells gave her the nickname Ernest because when he first read her work, he assumed “E. Nesbit” was a man. Bizarrely, this masculine identification also has an archival history in the Bodleian Library, where she was initially catalogued as “Nesbit, Ernest” (Briggs, 299). Clearly, the stark initial “E” in her signature suggested to some contemporary reviewers and readers that Nesbit was male. Interestingly, in an 1895 Sketch interview, Yellow Book publisher John Lane referred to Nesbit as “Miss E. Nesbit,” while (in the same interview) conferring marital status to “Mrs. Marriott Watson.”[1] Lane knew the actual marital status of both women, so these constructed textual identities may have been marketing strategies by the canny publisher. There are historical nuances of gendered authorial identity and social status here that we would like our personography to highlight.

Ella D’Arcy and Ethel Colburn Mayne are each noncorrelative entities in terms of their dual roles as authors and editors. Both published fiction in the Yellow Book—D’Arcy under her own name, and Mayne under the pseudonym Frances E. Huntley—but they also contributed invisible labor to the magazine, working as subeditors with Henry Harland. Mayne’s role lasted only half a year (January to June 1896), during D’Arcy’s absence on the continent. Notably, D’Arcy was covering for Harland when he and Lane were both away in April 1895, at the famous moment of Oscar Wilde’s arrest. Lane succumbed to the pressure of a group of his Bodley Head authors, who associated Aubrey Beardsley’s decadent art with Wilde’s sexuality, and sent a cable firing Beardsley from the Yellow Book staff. However, it was D’Arcy who was sitting in the editorial chair. As acting editor, she had to recall Volume 5 of the Yellow Book from the press and oversee the removal and replacement of Beardsley’s artwork (Kooistra and Denisoff). Despite D’Arcy’s and Mayne’s editorial contributions to the Yellow Book, however, the magazine itself acknowledges only their authorship, and the historical record likewise reveals a systemic undervaluing of their editorial labor (Windholz).

Male editors, too, sometimes kept their identity deliberately veiled, and disguised their biology with a pseudonym of the opposite sex. According to Elizabeth Sharp’s memoir of her husband, William Sharp edited the Pagan Review (as W. H. Brooks) and coedited the Evergreen, though neither magazine credits this work in print (Sharp, 260).[2] Even more than his invisible editorships, however, we are fascinated by his development of a “second life” with a feminine literary persona. William Sharp’s case differs from the noncorrelative pseudonyms used by women authors such as George Egerton (Mary Chavelita Dunne Bright) or Graham Tomson in that his avatar, Fiona Macleod, had social, legal, and textual life separate from his own.

Sharp began writing Celtic revival works as the reclusive Fiona Macleod in 1894; for the rest of his life, he publicly represented her as a cousin on whose behalf he acted as a kind of literary agent. Sharp sustained a personal correspondence between Macleod and friends such as W. B. Yeats, and even created a legal identity for her by having a power of attorney drawn up so he could act on her behalf (Halloran). In three of the Evergreen’s four numbers, Sharp and Macleod published work as separate entities. Sharp also contributed, under a variety of pseudonyms, to the Pagan Review, and Macleod contributed to the Savoy.

Fiona Macleod occupies a liminal space between history’s usual categories because she has a textual and cultural identity but no discrete biological existence. The Oxford Dictionary of National Biography lists Sharp and Macleod as “William Sharp (Fiona Macleod),” following the pattern established by Elizabeth Sharp’s posthumous memoir of her husband, which revealed his dual identity for the first time. However, bracketing off Macleod as Sharp’s subidentity does not reflect the complex relationship between these two authorial personas in fin-de-siècle culture. Macleod is more than a historical literary hoax; her poetry reveals a spirit and an authorial voice that William Sharp and his contemporaries considered uniquely her own.

The prosopographic impulse to ignore cultural data’s contingent nature and impose a normative hegemony on its documented subjects raises feminist challenges for biographical methods in the digital humanities. The disparity between the individual contributors Rosamund Marriott Watson and Fiona Macleod and the collective contributor profile sketched above (white, upper-class male) underlies our decision to build a personography rather than a prosopography. As we have shown, the traditional prosopographic process of making the average visible tends to make the unusual or noncorrelative invisible. This compounds a difficulty already present in many historical data on persons: individuals and behaviors that fall outside the “average” are screened out. Here, average can denote the dominant sociocultural conception of the subject: traditional prosopography presupposes that a physical body correlates with a cisgender, heteronormative conception of personhood, with male as the assumed standard of measurement. The average can also denote traditional prosopography’s standard personal identity categories—putatively unambiguous classifiers such as status (married or unmarried), role (author or editor), and sex (male or female). The prosopographical average frequently denotes a subject who has been filtered not only by the cultural texts of his or her historical milieu but also by present-day researchers. Scholars can only access historical lives through material artefacts, and our twenty-first-century ways of naming and classifying may not correlate with historical ways.

Verboven indicates that prosopography ultimately serves “to collect data that transcends individual lives” (41). The rhetoric of transcendence gives the prosopographical data agency, as if by virtue of its homogeneity it requires no interpretation or selection and simply presents itself as fact. But transcendent data do not exist. Prosopographical data analysis involves aggregating historical information on a large scale and interpreting patterns that appear through quantitative processing—in other words, it involves distant reading (Schulz; Underwood). Distant reading frequently relies on computer algorithms to process large corpora—a set of operations that N. Katherine Hayles calls machine reading (28). Determining the scope of a dataset, the algorithms for quantification, and the criteria for pattern recognition involve several interpretive steps. So-called transcendent prosopographical data about historical groups emerge through a normative selection process that obscures the data’s actual contingencies.

With the increasing adaptation of electronic scholarship’s critical methods, prosopographers have moved their work out of print narratives and into online databases. Recent prosopographic literature participates in digital humanities discourses that problematize assumptions about data neutrality. Such work recognizes that prosopography’s traditional methods impose, rather than reveal, hegemony.

Developing reliable databases while critiquing traditional prosopography’s normalizing filters is an ongoing challenge for digital humanities prosopographers.[3] Self-identified “new-style prosopographers” such as John Bradley, Harold Short, and Michelle Pasin at King’s College London conceptualize digital prosopography as a “visible record” of scholars’ analytical processes as they sort through source material (Bradley and Short, 5). The King’s College prosopographers therefore record evidence data not as facts, but as factoids. First coined by Norman Mailer in his biography of Marilyn Monroe, a factoid is an assertion made by a particular source about a particular person at a particular moment. As these particularities indicate, a factoid can register many contingencies (Bradley and Short, 8–11). The factoid serves as the basic unit of source-driven prosopography (Pasin and Bradley, 87). As documented utterances, factoids can reveal meaningful matter about historical lives, even if they record bias or contradiction.

By engaging historical data as interpreted factoids, we recognize that what does not leave traces in the historical record can be as important as what does, and the interpretive processes that determine historical data’s visibility matter as much as what the data convey. For example, giving William Sharp and Fiona Macleod separate entries in our personography, rather than listing Macleod as Sharp’s pseudonym, will impact whether users interpret Macleod as an unusually robust pseudonymous persona or as an authorial figure in her own right. Moreover, digital prosopography’s use of relational data structures demonstrates that scholars can only interpret factoids in relation to other factoids. Each node or data point in a relational data model only exists in relation to other nodes, and these nodes appear as a flattened network, rather than as a hierarchy. As a node in a relational network, a factoid’s significance depends on the factoid constellation within which it is mapped (Bradley and Short, 11–12).

While some digital scholars have adopted new-style prosopography, others have developed personography as an alternative method for documenting the biographical data of groups. Flanders defines personography as “the management of the identity of individual people” (“Encoding Textual Information”). As its moniker suggests, personography emphasizes the individuality of persons even though a personographic system, like a prosopographic one, disambiguates biographical data using a classification system. Personography emerged in the twenty-first century as a means of documenting the persons referenced in a TEI-encoded electronic edition. A personography allows a scholar encoding a text to link names that the text references with corresponding biographical data about the referents. Unlike prosopography, personography is usually shaped by a digitally encoded text that is not a primary biographical source, but that otherwise cites those persons. Both new-style prosopography and personography use a relational data model, but a personography’s model documents relationships between persons, not factoids. These persons may include contributors to the text, fictional characters, historical figures, and/or mythical figures. Personographies vary because the kinds of information that scholars encode and their methods for encoding depend on what they wish to emphasize in relating the encoded text to the personography.

Though recent prosopographic and personographic works challenge their predecessors’ assumptions, scholarship in these fields has not given adequate attention to the politics of gender and sexuality. We consider Booth’s Collective Biographies of Women project a bellwether of feminist prosopography. Like other feminist digital humanities databases, such as Orlando and Women Writers Online, Collective Biographies increases historical women’s visibility (Booth, Collective, 18). Crucially, Booth’s project uses prosopography to develop a feminist analysis of gendered biographical conventions and the changing life histories of women. As of her project’s 2009 inauguration, such an approach had “never been attempted in digital projects” (33–34). Undertaking the Collective Biographies was necessary, Booth asserts, because “it is not enough to recover knowledge of as many women of the past as possible; we should re-examine the texts in which their narratives and images circulate for different constituencies and interests and claims.” Booth notes that digital technology uniquely enables such prosopography “to collect, interpret through encoding, interweave, sort, compare, and visualize large bodies of discourse” (18). As Booth’s reflections on the Collective Biographies attest, using computational technology amounts to more than simply adding complexity to a conventional system of classification: through digital tools, scholars can provide new perspectives on historical data that prompt us to reconceptualize our historical practices.

The Yellow Nineties Personography encourages its users to interrogate classificatory logic, and the research process more broadly, by acknowledging historical data’s contingency. Like Michelle Schwartz and Constance Crompton, the creators of Lesbian and Gay Liberation in Canada, we have found rigid taxonomy inadequate for our database, if we truly want to “mark,” in the sense of making visible, the 27 percent of magazine contributors whose biographical data are ambiguous, fluid, or noncorrelative (Schwartz and Crompton). Our project is inherently nonnormative, given that the personography’s primary goal is to document the complex cultural and social networks out of which the Yellow Book, the Pagan Review, the Savoy, and the Evergreen emerged at the fin de siècle. We aim to build a reliable, manipulatable dataset while expressing personographic data in ways that, rather than flattening ambiguities, highlight the noncorrelative by showing the politics of gender and sex at work.

In the case studies mentioned above, the imperative to engage historical data as interpreted factoids gives rise to practical challenges in digital humanities practice. Do we encode Rosamund Marriott Watson’s biological sex (F) and/or Graham R. Tomson’s publicly performed gender (M)? What does each choice imply and make visible? Should E. Nesbit’s sex be coded as F (female), M (male), O (other), or U (unknown)? How can we document the invisible character of the editorial labor that Ella D’Arcy and Ethel Colburn Mayne contributed to the Yellow Book? How should we represent the conjoined identities of Sharp and Macleod in our database? We have decided to list Sharp and Macleod as two (related) contributors so that Fiona Macleod’s historical existence is visible, but how can we describe her relationship to William Sharp? Does assigning the same unique identification code to both persons in our dataset adequately convey that one physical body correlates with at least two distinct authorial and social identities?

We have not yet answered these questions to our complete satisfaction. However, in response to the challenge of parsing the complexities of such data, we have conceptualized the Yellow Nineties Personography as a prototype for interpreting historical persons within their contexts by making noncorrelatives—and the processes by which we curate them—visible. In order to do this, we have built the Yellow Nineties Personography in three iterations. Each iteration serves as an interface that visibly mediates biographical information with a different value hierarchy.

The first iteration is a spreadsheet-based biographical dataset that represents all personographic <persons> as equal. The dataset gives equal representational value to contributors who lacked social and/or legal status at the time of magazine publication and contributors who had a high social profile in the 1890s, and later became canonical in literary and/or art historical studies. In this iteration, then, Aubrey Beardsley, the internationally acclaimed art nouveau artist of the Yellow Book and the Savoy, Effie Ramsay, a forgotten Celtic designer for the Evergreen, and Charles Hare, the engraver whose miniature signature can occasionally be glimpsed on an illustration but is not otherwise credited, each have the same value. The spreadsheet records all persons as named entities simply by virtue of their contributions to the four avant-garde magazines in our digital corpora. These named entities include women, who were legal nonpersons at the time; unknown or ambiguously identified contributors, who became cultural nonentities; and engravers and editors, who were textual nonentities. Established writers and artists, such as Henry James and Paul Sérusier, whose names and biographies have enjoyed a textual afterlife are also, of course, <persons> in the dataset.

In many ways, this iteration of the Yellow Nineties Personography is the most accessible for users who are not familiar with computational methods for processing information. The table parses biographical data in human-readable categories. It also gives equal visual-spatial weight to each person’s entry, thereby foregrounding a key aspect of our personography’s argument. However, this spreadsheet is visually unwieldly precisely because it presents all persons and associated data simultaneously; the columns display too much information to study at once, and are not conducive to pattern recognition.

The personography’s second iteration is a data visualization series that introduces temporal, spatial, and relational hierarchies. Highlighting patterns of biographical disparity, the visualizations compel us to investigate the politics and cultural contexts that have shaped these disparities. Through filters of data and visualization choice, particular inequities between contributor data become prominent. In this iteration, our <persons> are no longer equal units, and the revealed asymmetries become prompts for investigation, discovery, and analysis. For example, a visualization of magazine contributors by gender (M, F, U, or O) displays an unequal proportion of F-designated contributors in any role—author, artist, editor, or engraver. More (biologically and legally documented) males than females contributed to the four magazine titles in our corpora. Unsurprisingly, this offers the personography’s most obvious entry point into late-Victorian print culture’s gender politics. A visualization of personography factoid frequencies displays a high proportion of factoids available for contributors designated M, in comparison to factoids about contributors designated F, U, or O. This pattern points to a hegemony at work in archival documentation itself, where records are preserved based on perceived historical value, and value itself is historically contingent on prevailing norms.

Users with little knowledge of data visualization may have difficulty engaging this iteration of the personography. Visualization can make a dataset’s patterns more readily visible and can provide users with a sensorial interface for interpretation. However, advanced engagement with visualizations, including both the manipulation of a data display’s output and the critical assessment of the methodology used to produce it, requires some technical familiarity.

The Yellow Nineties Personography’s third iteration is the TEI encoding of <persons>. The TEI P5 Guidelines acknowledge that “the meanings of concepts such as sex, nationality, or age are highly culturally dependent, and the encoder should take particular care to be explicit about any assumptions underlying their usage” (Burnard and Bauman, 13.3.2.1). However, we have found that in practice, markup methods tend to instantiate, rather than critique, cultural norms. In developing our biographical markup schema, we intend to draw on, and develop, all the flexibility that TEI can provide in describing the variable traits and states of historical persons.

TEI requires us to make unambiguous declarations about each individual contributor, but allows us to tailor the extensible markup language (XML) and structure. For example, the <addName> element is vital to the personography. We primarily describe contributors by the name under which they contributed to our four magazines, but we also document other names by which they were known. Such <addName> components may point to layers of social history at odds with a contributor’s textual history, as is the case for William Sharp and Fiona Macleod. They may also point to a textual-cultural identity expanding beyond our periodicals: this is the case for Ethel Colburn Mayne, who began publishing under her own name after her Yellow Book stint as Frances E. Huntley. The series of <persName> elements that describe a given entity foregrounds our interest in documenting each contributor’s social and textual presence.

The TEI P5 Guidelines offer some invaluable features for our encoding process, such as the “One Document Does It All,” or ODD file. This formal mechanism for explaining how we define our fields allows us to make our interpretive choices visible to users who read the extensible markup language, or XML. The ODD is a file that explains how an encoded document uses its TEI (Burnard and Bauman, 1). The TEI document includes a pointer or reference to this ODD in its header. Our personography’s ODD includes descriptions that detail our use of each element and attribute in the personographic entries. Because some personography users may not be TEI-literate, we also plan to make our intervention visible through an editorial introduction to the personography website that will take the familiar online form of an “About” page.

Some limitations to TEI’s personographic conventions require us to innovate. Because of prosopography’s historical orientation, it temporally hierarchizes biographical data; prosopographies are often structured genealogically or are otherwise event based. Much personographic work has adopted this practice and organizes data by event. We have considered organizing personographic data by event (birth, death, marriage, magazine contribution), but a precondition of our personography’s scope is that its persons existed during the same historical moment, the 1890s. Moreover, using temporal characteristics to organize the data would undermine the purpose of our project, which is to curate data about the interpersonal patterns that collaboratively create a magazine and its networks. At its core, the Yellow Nineties Personography is community based, rather than event based. We have therefore adopted a relational network model akin to that of new-style prosopographers.

Our relational emphasis complicates the process of encoding the personography, as TEI documents hierarchize well, but accommodate relational networks less easily. The TEI Guidelines organize the information about people into three broad groups: traits, states, and events. In order to capture the complexity of our dataset, we tag most biographical data as traits, rather than states. By documenting relationships, occupations, education, and affiliations as traits that do not have a beginning or an end, we emphasize that all categorized aspects of contributor lives were part of a network of potentially simultaneous influences. This network was temporally situated at the fin de siècle, but the dataset demonstrates that its influence extended backward (for example, W. B. Yeats critically rehabilitated the work of William Blake in the Savoy) and forward (for example, Edmund Gosse and other frequent Yellow Book contributors were colleagues and mentors to modernist writers such as Ford Maddox Ford and Edith Wharton).

Because the Yellow Nineties Personography is community based, rather than event based, we anticipate that it will be most productively visualized as a network, rather than a timeline. Aside from birth and death dates, the personography includes few data items that a user could map chronologically. We recognize that privileging a relational structure obscures the temporal conditions that influenced contributors’ lives and their participation in the transatlantic artistic community that produced our periodicals. This is an interpretive choice that we will foreground in our ODD and in our editorial introduction to the personography. We see the personography’s relational emphasis as self-reflexively highlighting the contingent nature of the data we deploy, as well as supporting feminist inclusivity in our digital humanities praxis. The persons in our dataset exist in relation to their periodical contributions, and to their personal and professional associations. Similarly, their visibility within in the 1890s’ artistic community was tied to their publicly available works, their social connections, and their participation in British and transnational cultural networks. Given that in our dataset all contributors have value by virtue of their periodical contributions, we hope that users will investigate cases in which a person’s limited role in the personographic network suggests that the person was less visible within the artistic community and/or is less visible in the archives.

We continue to hone a personographic vocabulary that expresses classifications as interpretive choices and foregrounds layers of historical information screened out by the normative impulses of generalized biographies. In building our prototype, we are curating the personal landscapes of the Evergreen, the Pagan Review, the Savoy, and the Yellow Book. We recognize our markup and database choices are interpretive and will condition the possibilities for meaning-making by users. Nevertheless, we hope users can create new knowledge about the individuals and relationships in our dataset. Visualizing the data will give users a sense of some broad patterns in contributor identities and textual lives, illuminating historical ambiguities and obscurities that warrant further investigation. We want users to query the personographic data in meaningful ways and critique our strategies for queering the data—the ways in which we have modeled our interrogative engagement with the factoids of historical lives. By transparently documenting our processes and inviting users to critique and reinterpret our presentation of the historical record, we insist that how sources present biographical traces and how we interpret them are as important as what the traces themselves express.

Our attention to making our own interpretive practices visible attests to the importance we place on how our dataset models knowledge. Like the King’s College prosopographers, we are building a biographical database that models how we think about our materials and our task (Bradley and Short, 3). According to Willard McCarty, a model of knowledge is “a representation of something for the purposes of study,” and a model for knowledge is “a design for realizing something new” (24). Our personography is not a model of the contributors to the Yellow Book, the Evergreen, the Pagan Review, and the Savoy. It is, instead, a model for interpreting the magazines’ contributors within the context of their sociocultural network—for interrogating the research process and acknowledging that historical data are always contingent. Moreover, our database’s relational form will ensure that it remains open and responsive, accommodating user-added data and facilitating collaborative knowledge-building in ways that we cannot anticipate, but welcome.

If prototypes can argue, as Alan Galey and Stan Ruecker contend, then the Yellow Nineties Personography instantiates a claim that digital humanities strategies can be used to mobilize feminist and queer approaches to periodical history and the politics of prosopography (405). Such approaches not only make historical nonpersons and noncorrelatives visible but also enable us to continue improving our use of digital tools and computational methods to grapple with subjective, contingent, humanist knowledge. In resisting the disambiguating and normalizing impulses of traditional prosopography, we also celebrate the subversive spirit in which the project’s fin-de-siècle contributors produced these avant-garde periodicals.

Notes

Bibliography