Introduction

Starting with the founding of the first local homophile organizations in 1964 and extending to the start of the AIDS crisis in 1981, the LGLC project reconfigures Donald McLeod's monograph series, Lesbian and gay liberation in Canada: A selected annotated chronology, as a TEI-XML encoded resource. The LGLC project extends the codex form by representing the relationships between the people, places, and events in the chronologies in spatial and temporal ways not permitted by print monographs. In addition to broadening the book form, we are investigating the interoperability of the existing modeling and knowledge-capture systems, including TEI, MARC, and graph databases, in the service of prosopography. We are investigating how data exchange formats and database structures shape how we record information and, as a result, what we can know about cultural history and the people who forged that history, when it and they are represented as data.

The project responds to the library and information science community's call for interdisciplinary teams to devise better means to represent the people and events of LGBT history at the level of code. It falls to "librarians, information architects, web taxonomists and other professionals … [to] improve both precision and recall in information retrieval" (Johnson 2008, 25). Despite the increasing diversity of resources online, the Internet's potential for LGBT information retrieval, as Ellen Greenblatt points out, is still developing (2010, 1-3). While improving the discoverability of information about LGBT identities, the design of digital resources ought to engage in "a serious critique of the mechanistic, entity-driven [computational] approach to knowledge" and, where a mechanistic approach may fail to articulate the nuances of humanities knowledge production, improving the approach and the technologies that enable it. (Drucker 2009, 21).

Donald McLeod's original chronology has proved to be both rich and amenable to TEI encoding (see figure 1). We have marked up the titles of published works, locations, personal names, pseudonyms, the names of publications the chronology cites and who wrote them, protests, and legislation. As a supplement to the chronology, Don McLeod included several appendices listing homophile organizations, publications, and bars and clubs that existed in Canada from 1964-1981. These appendices were expanded with additional research during the encoding process. The organizations appendix now includes both the homophile groups working to advance gay liberation across the country as well as the homophobic organizations working against them. The list of gay publications now includes all periodicals cited in the chronology. The bars and clubs appendix has been transformed into a comprehensive "places" file for any item in the chronology with geographic coordinates, from bars to churches, meeting rooms, art galleries, or private homes that were used for political gatherings.

We want to capture information about the liberationists, their supporters, and detractors to find out who was involved in the Canadian gay liberation movement, and how its ideas circulated within Canada. By counting the number of xml:ids on the persname elements, we have uncovered 2596 names listed in the years 1964 to 1975, 1076 of which are unique. The TEI-encoded records that let us discover the number of unique names and resolve pseudonyms to particular people also serve our graph database, which we have produced from the TEI via XSLT transformation.

Figure 1: Lesbian and gay liberation in Canada: A selected annotated chronology page and TEI-encoded event records (Source: LGLC project).

Populating the prosopography, one factoid at a time

We are creating a digital queer Canadian prosopography in the service of historians, literary critics, and the general public. Prosopography, or collective biography, is an important tool that helps historians to build a comprehensive picture of the past. First developed by Ancient historians, and revived by historians in the 1930s, prosopography is a technique for building profiles from fragmentary and partial data—allowing scholars to determine, for example, the average life of a Roman slave from hundreds of incomplete records of slaves' lives. Contemporary scholars have turned to computer-aided prosopography to model patterns even where group members' lives are well documented (Verboven, Carlier, and Dumolyn 2007). Inclusive prosopography, which fully represents previously marginalized groups, can inform and reshape researchers' perceptions of both the past and the present. Addressing the biographies of women that have been generated since the Victorian era, Alison Booth describes biographical data as having been "indispensable aids in the formation of nationhood as well as of social difference" and as having "contributed to each phase of debate about women's roles and rights since early modern times" (2005, 3). Even though the LGLC project focuses on the representation of gender, sex, and sexuality, we expect our findings and methods will be applicable to digital projects that seek to fill gaps in the responsible digital representation of, for example, historically contingent conceptions of class, religion, race, or political affiliation (Balsamo 2011; Chow-White and Nakamura 2012).

In building our prosopography, we have started with Bradley and Short's relational model and factoid-based approach. Factoids, as envisioned by Bradley and Short, are collections of assertions made by sources about a person. Factoids, they argue, can connect many kinds of structured information together and even profitably contradict each other. By recording factoids, or who said what about whom, we can foreground the historical specificity of the archival sources that underpin the project. In the following two sections we will explore how the liberationist press' assertions about people often stood in direct opposition to the assertions made by the mainstream press. The strength of the factoid-based approach is that it allows the same factoid to provide "more than one perspective on the events it represents when placed in different contexts created by other factoids" (2005, 12).

In designing the project, we wanted to create something that followed Bradley and Short's "heavy use of authority lists to support categorization" with "mechanisms so that new entries can be made in these lists whenever necessary" (2005, 18). Our prosopography is built around MARC Authority records and TEI. Our goal is a prosopography that is interoperable with library catalogs, eventually available via linked open data, thus giving our prosopography a structure that matched authority records was key. Whenever possible, we use existing MARC fields, compatible with the OCLC's Virtual International Authority File (VIAF), to reduce barriers for international library organizations like VIAF, should they want to ingest our data and represent the people of the LGLC online. For example, MARC authority fields allow us to define people by their personal names, associated places, addresses, fields of activity, and associated groups. Field 375, gender and dates applicable, is of particular import to the project—one of our earliest concerns when we started to encode identity in our text was being able to represent people who may have changed gender over time or who once used different terminology than they would use now to represent themselves. Field 375, with its time-bound gender, lets us add time-specific information to our prosopography (Library of Congress 2009). Not all of the MARC fields are time-bound, which initially made it difficult for us to represent real world time-bound factoids, like sex, in prosopographic form. After outlining the character of the liberation movement, and the systems that let us represent it, we will return to the problem of time-specific identification, and how a TEI-derived graph database is helping us overcome that challenge.

Reluctant to reinvent the wheel, we still want to expand on the standards or demonstrate where they are not sufficient so that we can make space for the standards to become more inclusive. Demonstrating the need for a new MARC field is the best way to encourage the Library of Congress to add it. The goal is to make something that is as compatible as possible with existing standards while simultaneously pushing at the boundaries of what current standards can do. For our personal dataset, showing the relationships between people is particularly important. For example, recording the time-bound relationship between sexual partners is important, but is not something that is currently possible in MARC authority records.

Although the source material for the project is indeed a text, we have been heavily influenced by Harvey Quamen's argument that scholars need only markup texts as though they are documents if 1) they are indeed documents and 2) if recording their status as documents is important to the research goals that drive the project (Quamen et al 2013). Although Quamen argues that TEI- XML, with its document-centric focus, may not be the best tool for projects that are not focused on documents per se, we have found that the flexibility offered by the TEI's Names, Dates, People and Places module, which accepts more time data than MARC, lets us readily manipulate the data in the text (Text Encoding Initiative 2014, Quamen et al 2013). Encoding the text in TEI for regularization and discovery (rather than to describe the materiality of Don McLeod's books), followed by transforming the TEI via XSLT into a graph database, has served the project well. By combining text encoding with a graph database, the project benefits from TEI's regularization and disambiguation as well as databases' ability to reveal new patterns in the text (see Figure 2).

We are working to address the "fuzzy" nature of data about people, and avoid either "squeezing… many shades of gray into a limited number of categories" or including so many categories as to make sorting the data in any meaningful way impossible (Bradley and Short 2005, 16). Here is where we believe the factoid model shines. As described by Bradley and Short, the factoid model "allows the prosopographer a place to record in a factoid her/his scholarly assertion that the text she or he is currently reading should be interpreted" in a certain way (Bradley and Short 2005, 17). We have followed their lead in creating a prosopography that allows researchers to capture some of the fuzziness of the data, and to clarify or qualify any assertions the data make.

As we have outlined elsewhere (Crompton and Schwartz 2015), we are also responding to the move to recover lost lesbian and trans histories. At the end of the last century, there was a rush by both the academic and activist community to claim biological females who lived as men^[1] as lesbians who had not had access to "lesbian" as an identity marker. In the 1990s the trans community countered with competing claims that these historical actors were part of trans history (Ross 1995; Warner 2002).^[2] These historically contingent articulations of selfhood are central to identity formation but are generally unrepresented in the models of identity that the digital humanities has inherited from traditional digital prosopographies, such as the Prosopography of Anglo-Saxon England (PASE) and the Prosopography of the Byzantine World (Craig 2012; Pasin and Bradley 2013; Shapiro 2010).

Graphing gay liberation

Our particular interest is in the historical specificity of the Canadian gay liberation movement, a movement whose identification of inequity—and strategies that address that inequity—differs considerably from the mainstream political strategies of many gay and lesbian communities today. Today's dominant reform-oriented movements, which existed before Stonewall and which surged back to the forefront in the 1980s, emphasized "respectability, entrance into the established institutions of power and assimilation into an expanded conception of the family" (Sears 2005, 96). By contrast, the gay liberation movement post-Stonewall was focused on visibility, militancy, and, as Alan Sears has pointed out, "an end to sexual regulation and the monopoly of the compulsory family system" (2005, 96). The gay liberation organizations represented in the LGLC chronology were characterized by radical activism—through, in the words of Maurice Flood of Vancouver's chapter of GATE (Gay Alliance Toward Equality), the "pressing of gay demands through public action and protest," braving "public ridicule, incidents of physical abuse, and rotten eggs, in order to announce gay rights" (1974, 15). The gay liberation movement represented an "intensely eroticized bawdy politics… a kind of sexual utopianism" (Sears 2005, 96). These bawdy politics can be seen in the publication and ensuing controversy over the "Men loving boys loving men" article by Gerald Hannon, published by the gay liberation periodical, the Body Politic (renamed the "Bawdy Politic" by the Toronto Sun) in 1977 (McLeod 2014, 208).

These bawdy politics also led to schisms in the movement. Privileged gay men focused on their own sexual libertarianism were often blind to power imbalances based in gender, class, and race (Sears 2005, 97). The lesbian feminist movement came head to head with the politics of the gay liberation movement in the fight over age of consent laws. When the National Gay Rights Coalition/Coalition nationale pour les droits des homosexuels (NGRC/CNDH) put forth a resolution to fight for the abolition of all age-of consent laws in 1975, the Rights of Lesbians Subcommittee of the British Columbia Federation of Women (BCFW) were quick to counter the resolution, "arguing that such a move would worsen the sexual exploitation of young women" (McLeod 1996, 245-246).

Eventually, the more radical, collectivist members of the gay liberation movement were forced to the edges by more conservative members. Such was the end of groups like Vancouver's Gay Liberation Front (GLF). Formed in a commune, sharing space with the subversive Yippies, and associated with the Ephemerals, a revolutionary drag group that "specialized in street theatre that satirized drag and mocked traditional male-female roles and stereotypes" (McLeod 1996, 54), the GLF met its demise when the more conservative members split off to form the Canadian Gay Activists Alliance (CGAA). As described by David Myers in the Queer History Project:

I remember having discussions where some argued that the gay and lesbian community should try to reform straight societies' attitudes towards us by emphasizing how 'normal' we really were and by rejecting those persons projecting 'non-normal' images—such as drag queens, transsexuals, leather fetishists and supporters of an S&M life style…I remember arguing that we had no business trying to gain the acceptance of society by practicing discrimination ourselves and trying to draw the line of social approval just below us but above those 'disapproved' groups (Rothon 2008, 24).

Unfortunately for David Myers and groups like the Gay Liberation Front, the reformers have triumphed over the radicals, utopians, and separatists, with the mainstream gay rights movement focus on gay marriage, traditional family units, corporate sponsorship, and commodification. The stakes of prosopographical representation of the gay liberation movement are high; as the section that follows outlines, even in the 1970s mainstream articulations of gay life were starting to take an assimilationist tone. The project recovers liberationist voices to show just how much gay rights discourse has changed in 50 years, in the hopes of assisting scholars and members of the public in imagining a future that is different from the present, a future that cannot, as Jane Rule might say, comfortably dismiss the past (including groups like GLF and the FLH (Front de liberation homosexuelle)).

Our markup has allowed us to create Gephi-based visualizations and populate a graph database with the text of each event as well as a network of who appeared in events with whom, how gay activists deployed pseudonyms, how queer community groups, publishers, and bars moved around their home cities over time, and which archival sources act as documentary witnesses. The most prodigious writers cited in the bibliographic entries that follow each event in the chronology (see Figure 1), wrote major works on which the chronology is based, like Gary Kinsman's Regulation of Desire. But the network graph of LGLC persons also shows how central Ron Dayman, Ken Popert, and Ian Young, all of whom were major activists at the time, were to the creation of the textual record of the gay liberation movement; not only do they participate in events, but they also created the documentary accounts through which we, from the position of the 21^st century, come to know the gay liberation movement. The figure below shows Ron Dayman, one of the most cited people in the bibliographies, and his connection to everyone else in the text, with the thickness of the line representing how much he was working with that person based on their number of connections in the text.

Figure 2: Ron Dayman Network Graph (Source: LGLC project).

To demonstrate the power of our graph database (which we can query, unlike the Gephi graph in figure 2), let us take as our example the organizations, the people, and the events connected to the John Damien case. John Damien was a Commission Racing Steward for the Ontario Racing Commission, an independent agency of the Ontario government. After being fired for being gay in 1975, Damien, turned to the Gay Alliance Towards Equality (GATE) and the Coalition for Gay Rights in Ontario (CGRO) for help (Rothon, 2007). The activists formed the Committee to Defend John Damien, chaired by Chris Bearchell, a member of CGRO and founder of the Lesbian Organization of Toronto (LOOT) (Canadian Lesbian and Gay Archives 2003). Damien sued for wrongful dismissal and loss of income, and the legal battle dragged on for years, becoming "one of the most prominent gay civil rights cases in Canadian history" (McLeod 1996, 203). The case left Damien unemployed and bankrupt, and took a toll on his health. He died of pancreatic cancer before the case was finally resolved—Damien received a year's worth of wages plus interest for wrongful dismissal, and the Ontario Government agreed to add sexual orientation to the human rights code (McLeod 1996; Rothon 2007). Our prototype Neo4j database has let us link the organizations, the people and the events in the Damien case together. In the graph below the green circles represent people, purple circles represent organizations, and the red circles represent events of the John Damien case.

Figure 3: John Damien Case Network Graph (LGLC project).

Gay liberation's knowing prosopographical resistance

Encoding both the events and the bibliographic sources they rely on has let us plot the relationships between the people, organizations, and authors in Don McLeod's text. Even at this early stage, however, working out how to represent these relationships prosopographically has proven challenging. What defines someone as an activist? How do we represent the relationship between someone writing about an event contemporaneously and the event itself, and how do we distinguish between people who were writing from the vantage point of a homophile organization newsletter or for a mainstream periodical? What is the relationship between the subject of a major liberation campaign and the campaign itself?

John Damien's story is not just the tale of one man battling discrimination in isolation, our markup-derived database shows how his case connected him to prominent gay liberationists. Above, in figure 3, John Damien is the green JDAM node and Chris Bearchell, a leader in the Committee for the Defense of John Damien, is the green CBEA node. The Committee for the Defense of John Damien is the purple CDJD node near her. Chris Bearchell's interest in the case connected Damien to a much larger network of liberationists through GATE Toronto, which spearheaded over twenty other human rights talks, protests, and demonstrations between 1971 and 1976. While graphing the events of the Damien case is fairly straightforward, the politics of the movement, and its participants' print media savvy (the very sources that our chronology depends on) presents a much more difficult modeling challenge. How can we encode the nature of the relationship between the Committee to Defend John Damien and John Damien himself? What about articles about the defense of John Damien published by members of the Committee?

Figure 4: Gay in the seventies (Source: "Gay in the seventies," Weekend Magazine, Globe and Mail, Dec. 17, 1977).

The publication context of primary source data shapes the assertions in the prosopography and calls for Bradley and Short's factoid method. For example, activists' self-definition varies widely according to publication venue, and may reveal the roots of contemporary assimilationist politics. The Globe and Mail published figure 4 in their Weekend magazine in 1977. They entitled it "Gay in the seventies." The Globe and Mail wanted "average gay people," no one who worked in an exclusively gay business or who was a "professional homosexual," and no one who was unemployed (Body Politic 18).^[3] The lists suggests that everyone was very "respectable" – working in jobs ranging from engineer at Ontario Hydro to psychologist at Toronto General Hospital:

1. Ian Young, author, Toronto
2. Trevor Mountford-Smith, engineer, Ontario Hydro
3. Michael Lynch, assistant professor, University of Toronto
4. Bill Lewis, microbiologist, University of Toronto
5. Dr. Rosemary Barnes, psychologist, Toronto General Hospital
6. John Alan Lee: author and sociologist, University of Toronto
7. Ron Shearer: businessman, Toronto
8. Marie Robertson: federal civil servant, Ottawa
9. Mark Whitehead; university student, Toronto
10. Charles Hill: assistant curator, National Gallery, Ottawa
11. Edgar Z. Friedenberg: author and professor, Dalhousie University, Halifax
12. Jim Quixley, head librarian, Glendon College, York University, Toronto
13. David Gibson: graphic artist and civil servant, Toronto
14. David Garmaise: division manager, Canada Post, Ottawa
15. Clarence Barnes: chemical engineer, University of Toronto
16. Ed Jackson: researcher, Ontario Institute for Studies in Education, Toronto
17. Therese Faubert (with her baby Jodie), teacher, Toronto
18. Konnie Reich, photolab technician, Toronto
19. Stuart Russell: typesetter, Montreal
20. Debbie Parent, clerk, Bell Canada, Ottawa
21. Christine Bearchell, office worker, Toronto

The Globe and Mail sought "respectable" people who would be willing to be out in a major national magazine. The resulting image, however, is exclusively of gay activists in the radical wings of the gay liberation movement, who, in jest, we speculate, pretended to espouse the Globe and Mail's assimilationist goals. The engineer at Ontario Hydro was also the secretary of the Committee to Defend John Damien. In fact, at least four of those pictured belonged to the Committee to Defend John Damien. Ed Jackson, described as a "researcher, Ontario Institute for Studies in Education" was, at the time of the Weekend photo, being charged by the police in the obscenity case resulting from the "Men loving boys loving men" article. And, as noted by Rick Bébout, "everyone in that picture (even the baby) had appeared, or later would, in the pages of The Body Politic" (Bébout 2007). The view of the lives of "average gay people" presented by Weekend magazine was very different from the lives they were living as members of the gay liberation movement.

The photo was published in Weekend in December 1977. The February 1978 issue of the Body Politic that followed featured stories on the John Damien case dragging into its fourth year, the police raids on the offices of the Body Politic following the obscenity charges (8), and the Emanuel Jacques murder trial, which had resulted in a great deal of anti-gay sentiment roiling the city of Toronto (Hardy 11). The February 1978 issue of the Body Politic also republished the image from Weekend, mocking the descriptions that Weekend had given, and provided alternate descriptions of the same people (18):

1. Ian Young, co-founder of Canada's first gay lib group, the University of Toronto Homophile Association; founder of the gay publishing house, Catalyst Press; regular writer for The Body Politic
2. Trevor Mountford-Smith: member of the Gay Alliance Toward Equality; secretary of the Committee to Defend John Damien
3. Michael Lynch: regular contributor to, and former collective member of, The Body Politic, presently chairperson of the Committee to Defend John Damien
4. Bill Lewis: early member of Gays for Equality in Winnipeg, currently part of the news staff for The Body Politic
5. Rosemary Barnes: an active member of the Lesbian Organization of Toronto (LOOT)
6. John Lee: head of the Toronto Gay Academic Union
7. Ron Shearer: an early and very active member of the Community Homophile Association of Toronto (CHAT), the city's biggest and most important gay group in the early seventies
8. Marie Robertson: a former member of Gays of Ottawa, presently active with Lesbians of Ottawa Now (LOON)
9. Mark Whitehead: co-founder and still co-ordinator of Gay Youth Toronto, and a former member of the Committee to Defend John Damien
10. Charlie Hill: another co-founder of the University of Toronto Homophile Association, and one of the people who helped start Gays of Ottawa
11. Edgar Friedenberg: writer on the sociology of adolescence
12. Jim Quixley: active in the Gay Academic Union; an early volunteer with the Canadian Gay Archives
13. David Gibson, past secretary of the Committee to Defend John Damien; news editor and collective member of the Body Politic
14. David Garmaise; active in Gays of Ottawa; former co-ordinator of the National Gay Rights Coalition; The Body Politic's Ottawa news correspondent
15. Clarence Barnes: member of the Gay Academic Union and an occasional news writer for The Body Politic
16. Ed Jackson: collective member of The Body Politic since 1972, the person who has built the "Our Image" section of the paper. Also one of those charged by the police in the current insanity.
17. Therese Faubert: active in the Revolutionary Workers' League, a pro-gay leftist group, she was the RWL's candidate in the last provincial election, running directly against the Premier.
18. Konnie Reich: active with the Lesbian Organization of Toronto (she's wearing a LOOT t-shirt) and a member of the Gay Offensive Collective — they're putting together a cable TV program called "This Program may be offensive to heterosexuals"
19. Stuart Russell: very active with the Association pour les droits des gai(e)s du Quebec, and The Body Politic's very diligent Montreal correspondent
20. Debbie Parent: a member of Lesbians of Ottawa Now
21. Chris Bearchell: dyke dynamo, busy day and night as a member of the Lesbian Organization of Toronto, the Gay Alliance Toward Equality, the Gay Offensive Collective, and most lately, the BODY POLITIC FREE THE PRESS FUND. She usually writes the "Dykes" column in The Body Politic (emphasis added).

This spurs us on to investigate how people are represented, how those representations change over time, and how people describe themselves versus how they are described by others. For instance, Weekend described "Christine" Bearchell as an office worker and Konnie Reich as a lab technician, but in the Body Politic version Chris Bearchell became "dyke dynamo" and "member of The Body Politic Free the Press fund" who "writes the dyke column for The Body Politic." Konnie Reich, who had managed to sneak a LOOT t-shirt into the Weekend photo shoot became "active with LOOT and the Gay Offensive Collective, putting together a TV program called This Program May be Offensive to Heterosexuals."

There is another factoid about Bearchell and Reich, beating its wings in the background of this image. In 2003, reflecting on the Weekend photo, Rick Bébout, another prominent liberationist, gave additional descriptions for these people, and he noted that Konnie Reich was a "lover and later longtime group housemate" of Chris Bearchell. He also quoted Chris Bearchell, described as an "office worker" by the Globe and Mail, on her work with the Body Politic:"We were doing exactly what we'd hope to be doing if we didn't have to work for a living" (Bébout 2007). Chris Bearchell was strategic in her deployment of identity in the context of two different news sources. While "office worker" was not Chris Bearchell's primary identity, it is one whose contextual use is worth recording. It is just the type of strategic self-naming that contemporary factoid-based prosopography can capture.

Just as we were glad to work with MARC's 375 field, so too are we pleased by MARC's more recent 386 demographic group field (gender with optional dates applicable), which was just added in 2013 (Library of Congress 2013). Field 386 allows us to add demographic groups to people, for example something like "sexual orientation group – lesbians" and "gender group – women" to our MARC¬-informed prosopography. 386 does not have a date applicable option, so we cannot construct a factoid that specifically states that someone defined themselves as a dyke in 1977 but did not identify as a dyke in the 1980s using MARC alone. However, by using demographic groups, we improve the search value of the LGLC graph database, which, unlike MARC, can take date information, tying a person to a particular identity only for as long as we have factoids to support the assertion. If a person in our database self-defined as a woman in the seventies and then later identified as a man, researchers who wanted to find people who were active in the period in question as women could still use the demographics group to find people who identified at some point as women, without keeping people in the event records from being defined as the sex they are now as opposed to the sex they were then. We hope to do justice to Jane Rule's call not to reduced human experience, by engaging in responsible digital representation and preservation of cultural heritage through prosopography that supports a nuanced conception of personhood, enhances public discourse in Canadian society, and avoids tendencies to narrow human experience in an attempt to understand it.