Defying description: searching for queer history in institutional archives

There are numerous obstacles to overcome when conducting queer historical research. While has been a steady increase in work to address the (in)visibility of queer histories and perspectives in institutional heritage collections, this work is often skewed towards outputs such as exhibitions or social media posts. As a result, very little work has been done to interrogate and transform the ways that these materials are managed and made accessible (or not). This paper considers how the lack of rich and explicit descriptive metadata shapes the (in)visibility and (in)accessibility of queer records within institutional archives. It reports the findings of a series of qualitative interviews I conducted with individuals who have searched for queer materials in archival collections in Australia. These interviews explored the practical process of searching for queer records in institutional archives; how researchers identify, locate, and access these records; and the challenges of this process—both in terms of a research methodology, and an experience for queer researchers within potentially hostile or unsafe spaces.


Introduction
"If archivists are authors, then the descriptions they create are narratives; stories told about the record and its meaning…The power of the archivist, then, stems precisely from… [their] position to decide which stories will be told". (Deodato 2006, p. 59).
Where do we find queer 1 history? In Australia, you might look to Trove-the National Library of Australia's database aggregator, and portal to all things historical. Like many people, you may be unfamiliar with the details of how archives are managed and structured, but certainly know how to use a search engine. So, in the search box on Trove, you might enter a commonly used term like 'LGBT'. 'Diaries, Letters & Archives' returns 26 total results. If you refine the date to show records dating between 1900 and 1959, only 1: 'Organization files from the Lesbian Herstory Archives'. Not quite what we're looking for. What about 'Newspapers & Gazettes'? This returns 7,451 total resultsthat's more promising! The earliest result dates from 1832, from an edition of The Sydney Monitor. Have we found a piece of queer history? You look closer, and realise that it was just an error in the OCR. You have found, not an as-yet-unknown piece of queer history, but an advertisement for iron and steel sheet wire.
There are numerous obstacles to overcome when conducting queer/ing historical research. The historic need to be covert in order to remain safe has left us with a limited body of evidence of queer lives. Personal records such as letters, diaries, and photographs-already more likely to be lost over time-were either destroyed, not created to begin with, or were carefully coded and edited to obscure queerness. While institutional and government records paint a history of criminalisation, marginalisation, and oppression-generally from the perspective of the oppressor. As Stone (2018) writes, "[t]he structural forces that marginalized queer communities historically have also hidden and obscured queer history, leading to omissions or obfuscations of historical documents in established archives" (2018, p. 217). These challenges are further compounded if you want to research queer people who were also Indigenous, disabled, or otherwise marginalised.
There has been a steady increase in work to address the (in)visibility of queer histories in heritage collections and institutions. Various institutions across the GLAM sector, both within Australia and internationally, have sought to make the queerness of their collections visible, whether through identification or interpretation. The Victoria & Albert Museum established an LGBTQ Working Group in 2006, which works to " [unearth] previously hidden or unknown LGBTQ histories in the collections". The V&A also permanently hosts LGBTQ Tours, while the New York Public Library created the LGBTQ Initiative to connect queer folks and communities with "the wide array of recourses, services, and expertise" offered by the NYPL, including literature, rare books, and archival collections (NYPL 2022). A 2022 exhibition at the National Gallery of Victoria in Australia, titled Queer, brought together approximately 400 artworks from across the Gallery's collections. The exhibition sought to offer new interpretations and narratives of these artworks, but also to consider absences in the collection "by excavating queer history where it has been omitted or eclipsed, through oversight or intent" (NGV 2022).
However, access via curated exhibitions does little to ameliorate the more entrenched limitations and challenges of accessing queer history. Limited in their scope and duration, exhibitions allow institutions to share the diversity of their collections in a very surface-level way, with no requirement to transform or improve the more fundamental and foundational ways that heritage materials are managed. Sutherland and Purcell (2021) have argued that much reparative redescription work often centres around palatable ideas of diversity and inclusion, and function essentially as 'false generosity', while ultimately failing to disrupt or dismantle the underlying systems of oppression and harm (Harris 2001;Hopkins 2008;Hughes-Watkins 2018). This 'have your cake and eat it too' approach is (ironically) most visible during Pride Month, when institutions plaster their social media with images of queer items from their collections. However, if one searches for 'gay ', 'lesbian', 'transgender', or 'LGBT' in their catalogue, the search will not return these records. Rainbow flags abound, but the catalogue is silent.
This paper considers the infrastructures of history-the systems, processes, standards, norms, and traditions that make the doing of history possible. As Alexandra Juhasz and Theodore Kerr have observed, the question of why some stories, some perspectives, some evidence gets "dragged forward", is.
"not simply a matter of identity or audience. Rather, it is about the infrastructures of history and historiography: who creates and accesses the archive; who has both the time and the desire to consider the past through material artifacts; what urgencies and desires are behind various impulses for an historic turn" (Juhasz and Kerr 2020, para. 16).
The issue of queer representation in the archive is often relegated to one of two issues: the work being done by queer community archives, or the non-existence of records in institutional collections. These two foci work in concert with one another, creating a dialogue from which major archival institutions are largely exempt, and not held responsible for the overt biases in their collections, and in how those collections are managed and made accessible.
In this paper, I will focus on one particular facet of queer historical infrastructure: archival description. Archival description is a process of mapmaking: the creation of a representation of the archive-a tool for orientation and navigation. Andrew Janes' article on description through cartographic metaphor reminds us that.
" [n]o map can tell the whole truth. Even the most accurate and skilfully constructed map presents a simplified and abbreviated view of reality…Being less than life-size, maps must be selective, suppressing some details to show others more clearly. Maps therefore tend to be neater than real places, smoothing over some of the messy complexities. For these reasons, Monmonier refers to maps as 'white lies', i.e. fictions that can illustrate truths. By showing false, outdated, biased or uneven information, maps can also conceal or twist the truth. A blank space on a map is no guarantee that an area is featureless or empty. The need for selective abstraction means that mapmakers must privilege some features and some spatial relationships over others" (Janes 2021, p. 98).
Janes also notes that many archives have a backlog of unprocessed, undescribed, or minimally described records-"the archivists' equivalent of 'Terra Incognita'" (2021, p. 100). Dunely and Pugh (2021, p. 3) have similarly likened archival catalogues to charts of the ocean: some areas are mapped in minute detail, and others "contain little more than the old warning: 'here be dragons'".
These cartographic metaphors offer a compelling view of the ideological construction of archives. Some roads are well-defined-paved, fenced, and well-signed to deal with the traffic. Others are dirt or gravel, maybe a little harder to find on the map, but easy once you know the way, and you can always ask a local for directions. Some aren't roads at all: they are routes you hack through the jungle, clearing the path and finding your way as you go. These are the areas of the map we have not filled in, and ultimately, don't feel a pressing need to. The well-paved highways rarely lead to queer history. This paper reports the findings of a series of qualitative interviews I conducted with individuals who have searched for queer materials in archival collections in Australia. Working within an interpretivist paradigm, these interviews were deeply influenced by my own positionality-not only as a queer person, but also as a queer archivist who has worked, volunteered, and researched in both community and institutional archives. My desire for a queer history, and my knowledge of the structural and practical challenges of queer archival work, often sit at odds with one another. As such, these interviews were designed not only to form part of my broader doctoral research, but also to illuminate the "other side" of the archival systems I work with. I wanted to know: what is it like to walk the archival backroads in search of queer history? The three research questions guiding this paper are: (1) what is the practical process of searching for queer records in institutional archives? (2) how do researchers identify, locate, and access these records, given the general lack of explicit queer descriptive metadata? and (3) what are the challenges of this process-both in terms of a research methodology, and an experience for queer researchers within potentially hostile or unsafe spaces?

Literature review
In 1846, Denis Prendergast was due to be executed for his part in the so-called Cooking Pot Riot on Norfolk Island. The night before his death, he wrote a letter: "Dear Lover, I hope you won't forget me when I am far away and all my bones is moldered away. I have not closed an eye since I have lost sight of you … Dear Jack, I value Death nothing but it is in leaving you my dear behind and no one to look after you…The only thing that grieves me love is when I think of the pleasant nights we have had together. I hope you won't fall in love with no other man when I am dead and I remain your True and loving affectionate Lover" (AOT GO 33/57 cited in Niall & Thompson 1999, p. 51).
A magistrate who visited Norfolk Island in the same year of Denis' execution reported that some three-hundred convicts were living as 'man and wife' (Willett 2017). While women interviewed during an inquiry into the Cascades Female Factory in Hobart in 1841 describe "nailing" each other, and the inquiry reports the women using "artificial substances, mechanically secured to the person … [as a] substitute for the male organ" (AOT 62/31/13859 cited in Casella 2000, p. 147). And in a letter written in 1848, Ross Female Factory Superintendent Dr. William Irvine describes "letters intercepted … which will prove the warmth and the impetuosity of the feelings excited in the women towards each other, when allied in such unholy bonds" (AOT MM 62/31/13859 cited in Casella 2000, p. 218).
There is joy and even humour in encountering these records of queerness. The surprise of seeing such explicit (in every sense of the word) evidence of queer people and experiences in the past can overshadow, at least for a moment, the context in which these records were created. Reflecting on Sappho's words "Someone will remember us / I say / even in another time", Heather Love writes that this fragment "… offers a nearly irresistible version of what queer subjects want to hear from their imagined ancestors" (Love 2009, p. 34). What Christopher Nealon has referred to "as the "message in the bottle" dispatched from the queer past (Love 2009, p. 34 citing Nealon 2001). Archives offer not only the kind of material evidence necessary for the doing of (queer) historical work, but also an opportunity to encounter and commune with queer ancestors-what Carolyn Dinshaw (1999, p. 21) describes as a "touch across time". As Gina Watts puts it, "Archives represent material history: the idea that a person can find their families, or those whose lives mirrored theirs, in an acid-free box, and in doing so, find themselves, be recognised by the historical record, and claim their right to take up space in the world" (Watts 2018, p. 104).
Perhaps this is why archives-their incompleteness, their silences-have become so central to not only the practice, but also the conceptualisation of queer history (Lewis 2014;Arondekar 2009). As Love writes, "the longing for community across time is a crucial feature of queer historical experience, one produced by the historical isolation of individual queers as well as by the damaged quality of the historical archive" (Love 2009, p. 37). There is duality in the queer archive: absence magnifies presence, presence magnifies absence. At the crux of this tension, the archive is centred as both bane and benefactor of queer history; depriving with one hand and giving with the other (Arondekar 2009;Cvetkovich 2003;Love 2009). As Lewis observes, the "archive emerges recurrently as an original cause of historical deprivation and the ultimate mechanism by which queer history may be secured or extinguished" (Lewis 2014, p. 16).
However, positioning the archive as the sole source of knowledge for historical enquiry risks submitting to its biases, even as we attempt to critique, challenge, and disrupt them (Arondekar 2009(Arondekar , 2017. Since the emergence of diplomatics in the seventeenth century, archives have been cast as "privileged locations for determining historical truth" (Blouin and Rosenberg 2011, p. 8). As the apparent source of historical understanding, the narratives that were drawn from archives were imbued with this same authority-archives tell the truth, and so whatever is drawn from them must also be the truth. By this logic, if one is searching for the truth, to the archives you must go. This is further entrenched by the preoccupation with presence and absence, discovery and loss, that Arondekar describes: "Implicit in this rethinking [of colonial methodologies] … is the abiding assumption that the archive, in all its multiple articulations, still constitutes the source of knowledge about the colonial past. The inclusion of oral histories, ethnographic data, popular culture, and performances…may fracture traditional definitions of the archive…but these sites still produce a telos of knowledge production propelled forward by what one will find-if only one thinks of more capacious ways to look" (Arondekar 2009, p. 6).
This political and ideological positioning is grounded in the belief that the archive will "somehow yield proper objects of study" (Arondekar 2009, p. 6) and "submit to our engagements in the first place" (Lewis 2014, p. 28). This positioning of the user as the one who must find a way to see through the archive in order to uncover the truth puts the onus of labour, of adaptability, of innovation, onto those working with archives and archival materials, and not the archives themselves. Lewis questions whether the archive will submit to our queer/ing work, but perhaps it is more accurate to say that we are expected to submit to the archive-not only to its presences and absences, but to its structures (taxonomical, epistemological, technological, organisational). This is of particular relevance in places like Australia, where state and national archives are fundamentally repositories of the state: government records, managed in government buildings, by government employees, for government needs. Watts notes that "[i]nstitutional archives tend to be strongest in areas where the records have the highest levels of stability…and in turn, records are most stable when they were created by institutions that have the means and motivation to protect them" (Watts 2018, p. 105). The question of 'where do we find queer history?' is thus not simply a question of which records survive, but also which are in scope to be held by these repositories, and which are deemed worthy of management and preservation. The records of queer history held in government archives have been retained and carefully managed over time not because they evidence the lives and experiences of queer people, but because they evidence the activities of government. As a result, we are left with archives that evidence the policing instead of the policed, criminalisation instead of the criminalised. From an archival standpoint, this has substantial practical impacts on the management, access, and use of these records. Arrangement and description are done according to provenance. Going back to our cartographic metaphor: both the map of the path (description), and the path itself (arrangement), are constructed to reflect and accommodate government perspectives and requirements. If you happen to come across evidence of queerness on your journey, it was not intentional on the part of the record creator, or the archivist. You journeyed beyond the lines of the map and stumbled across a dragon.
In considering the mechanisms through which queer historical 'recovery' is made possible (or, indeed, more difficult), we must move beyond the immovable focus on absence and presence (Arondekar 2009). This is true of archival scholarship as much as historical or sociological. When considering the visibility and accessibly of queer histories, the only issue we acknowledge is that of presence or absence: we have things, or we don't. In doing so, we as archival scholars and practitioners attribute blame for the limitations of our collections to recordkeepers of the past and place the onus of ameliorating this archival lack onto those communities who most acutely experience it. We must move beyond the archive of absence and confront the non-neutrality of our interventions today. We must consciously and critically engage with the archive (or rather, with archives) as something not already become, but rather something that is constantly becoming (McKemmish 1994, p. 200). How do we manage what is there?
While there is a strong body of work exploring queer/ing archives (Cifor 2016;Ajamu et al 2009) it is dominated by discussions of archives as sites of historical absence (Arondekar 2009;Lewis 2014;Love 2009;Watts 2018), or on the numerous queer community archives that emerged in the wake of gay liberation in the 19760s and 1970s. Very little has been done on the specific, systemic, and material ways in which archival thinking and practice continue to influence the visibility and accessibility of queer records outside of community archive settings (Lee 2016;The Trans Metadata Collective et al. 2022). This stands in contrast to the extensive critical work on the representation of queer materials in library catalogues, which has been evolving continuously for decades (Olson 2001;Watson 2020;Edge 2019). Initially dominated by critiques and advocacy surrounding homophobia and transphobia in subject headings (Greenblatt 1990;Olson 2000Olson , 2001Christensen 2008;Adler 2015), it has since expanded to consider the potential incompatibility of established classificatory systems (or, potentially, the entire endeavour of classification) with queer perspectives, sensibilities, and ways of being (Adler 2017;Drabinski 2013;Keilty 2009). A more detailed review of this literature is far beyond the scope of this paper, but I do want to emphasise one key insight from queer/ing library scholarship that we can apply to our exploration of queer/ing archival description: we must make our materials visible and accessible to users in a way that is intelligible and obvious to them (Adler 2017).
This leads us to the fact that very few user studies on the experiences, challenges, and requirements of queer researchers in institutional archives have been undertaken. The general lack of research into how users identify, locate, and access archival materials has been noted for many years (Beattie 1989-90;Chassanoff 2013;Duff and Johnson 2002;Lytle 1980;Yakel 2004). However, this lack of understanding is further compounded by the distinct needs and challenges of queer archival research, most of which have not been adequately explored in the archival literature. For instance, multiple studies have found that names are an extremely common way to locate records (Bearman 1989-90;Berner 1971;Cole 1998;Collins 1998;Conway 1994;Duff & Johnson 2002;Gagnon-Arguin 1998;Stevens 1977). Duff and Johnson have noted that the use of names "may be a fundamental practice in historical research since the past is often interpreted through the activities of individuals or organisations" (2002, p. 493). However, as I will discuss in my findings, the use of names to identify records of queer individuals has very limited applications and is much less "fundamental" than in general archival searching.
Social history is an area which has been addressed (although in a limited capacity) in existing user studies, and many of these challenges have substantial overlaps with queer history. Several studies (Beattie 1989;Duff and Johnson 2002) have reported the limitations of provenance searching (either individual or organisational) for particular kinds of historical enquiry-in this case, social history-and the potential usefulness of subject access for collections. As with social history, queer history is rarely neatly organised by provenance, and researchers may sometimes need to locate the single queer record within an entire collection, often without knowing an individual's name. However, while the insights provided by these user studies are useful and often relevant, the specific and intersectional sociocultural factors that influence the visibility and accessibility of queer records still remain underexplored and largely unaddressed.
This study responds to these gaps in the literature and aims to illuminate the various challenges of identifying and accessing queer materials in institutional archives. The three key research questions guiding this paper are: (1) what is the practical process of searching for queer records in institutional archives?; (2) how do researchers identify, locate, and access these records, given the general lack of explicit queer descriptive metadata?; and (3) what are the challenges of this process-both in terms of a research methodology, and an experience for queer researchers within potentially hostile or unsafe spaces?

Methodology
To answer these questions, I conducted interviews with individuals who have searched for queer records in institutional archival collections in Australia-meaning state archives, the National Archives of Australia, and the archives of major heritage institutions. By 'queer records' I mean records that are used in the doing of queer history-they might document a queer person, a queer action, or an authority's suspicion or 'accusation' that someone was queer. The study includes participants who exclusively relied on archival collections, and those who use archives in tandem with private collections, community archives, oral history methodologies, and other research practices. As my participants were located across the country, all interviews except one were conducted remotely via video.
Ranging from 90 to 120 min in duration, the interviews were conducted in an adaptive, collaborative, and conversational way (Neuman 2014; Williamson 2013), with planned questions being used as prompts more than a rigid structure. Verbatim transcripts were automatically generated and then corrected using manual transcription, which also served as a first round of analysis. Then, using software to highlight and tag the transcripts, I produced a codebook iteratively, by analysing and re-analysing the transcripts in concert with one another, adding both degrees of granularity and new topics throughout this process of analysis. This analysis was then further refined and enriched through the process of writing this paper.
Participants were recruited through a combination of personal contacts and snowball sampling. Due to the highly specific nature of my target participants, a high degree of homogeneity was unavoidable (Biernacki and Waldorf 1981;Olsen 2012). Saturation has been considered a problematic term, given the high degree of variability and subjectivity in determining when "enough" data has been gathered (Guest et al. 2013). After conducting the initial analysis, I determined that the data was both rich and extensive, and that further participants were not necessary. This sample was not designed to be representative. However, I did utilise a degree of criterion sampling (Patton 2015), as I wanted to interview people who had searched for records of queer history prior to the 1970s (although this did not need to be exclusive). I chose this date for a few key reasons. The advent of gay liberation in the 1960s, the commencement of decriminalisation in 1975, and the formation of the Australian Queer Archives in 1978, all changed the queer historical and archival landscape substantially. Records created and described after the mid-twentieth century tend to contain clear and specific (if not accurate or respectful) language to refer to queerness, making these records much more findable, particularly in contrast to records created before the twentieth century.
The dominance of white cisgender homosexual men in queer history and culture is something I was consciously considering throughout this research-particularly in my participant selection. Multiple participants had experience in transgender and gender-diverse historical research. However, I did not interview any participants with expertise in Aboriginal and/or Torres Strait Islander gender and sexuality history. The scope of this study, and my knowledge and experiences as a white settler, are not appropriate to meaningfully address the intersections of homophobia, transphobia, white supremacy, and colonialism that operate within institutional government archives, as well as within Australian society and culture. I acknowledge that this is a substantial limitation to my research, and one that perpetuates the privileging of settler perspectives within both queer history, and archives more broadly. I hope to address this limitation through future collaborative work. I refer readers to the work of Barrowcliffe (2021), Barrowcliffe et al. (2021), Faulkhead et al. (2010), O'Sullivan (2013, and Thorpe and Galassi (2014).
The field of queer history in Australia is relatively small. To ensure anonymity, I will be using they/them pronouns to refer to all participants throughout this paper and will refer to each participant using a number. Additionally, I will be restricting and generalising the demographic information about the participants that I include in this paper, even when it directly influences the research practices and foci of participants. Out of seven participants, the majority were historians who were currently working on, or had completed, a doctoral degree in their field. Two of the participants had experience working in libraries and archives at some point in their careers. Many of the participants have written and published journal articles, books, or blog posts about their work researching queer history; however, I have removed any specific details about their work. The findings of this paper are predominantly informed by participants' experiences in Australian archives, but many also spoke of their experiences in the United States and the United Kingdom. While I did not explicitly request demographic information about sexuality or gender identity as part of the recruitment or interview process, multiple participants shared their experiences as LGBTQ + people as part of our conversation. These experiences sometimes related to their research work, but also sometimes to their own journeys as queer people, and how these influenced their thinking about queerness as both identity and action. All of these factors came together to shape and inform our conversations, and the findings which I derived from them.

Findings
Due to the conversational and semi-structured nature of these interviews, many themes emerged. The findings in this article focus on the practicalities of searching for queer records in institutional archives. How do researchers identify, locate, and access relevant materials? How useful (or not) is descriptive metadata to this process? What impact does language have on this usefulness? What is the experience of a queer researcher working with materials that document their own oppression? How do they navigate institutional (and often government) contexts in doing this work?

Finding aids and research guides
Consistent with some existing studies (Conway 1994;Gordon 1992;Steig 1981) finding aids and research guides were consistently described as being of limited help in locating queer records in institutional archives, with researchers generally relying on their own systematic searching, or knowledge of the organisation/group being researched. It was more common for participants to report consulting archivists directly, or asking other researchers. In contrast, finding aids at queer community archives (such as the Australian Queer Archives) were described as more or very helpful because they explicitly describe records using known terms such as homosexual, lesbian, or transgender. Several participants noted that this greatly improves the speed and efficiency of their research, as well as general findability.

Collection searching
Because finding aids were of such limited usefulness, most of the processes and techniques described by participants were around collection searching through online interfaces, and two main methods emerged: item-level and provenance searching.

Item searching
Most participants not researching specific organisations described working at the level of individual records, conducting keyword searches, both in online catalogues and within records themselves (via OCR). Participants described developing a lexicon of terms and working through them very systematically, often with extensive truncation variations (e.g., homo*, homosex*, homosexual*, homosexuality, homosexuals, homosexualist). They also described adding to the lexicon as the search revealed terms they weren't aware of previously. The bulk of the searching work takes place within the catalogue and/or digital repository, and was the primary method used by many participants when researching specific individuals (that were not attached to an organisation). This method of searching is a version of 'some records may exist in some form, and may be in any number of locations, so I will see if I can find anything'.
Participants described numerous reasons why keyword searching was often the most effective means of identifying relevant records. While one participant described services such as Ancestry.com as being "incredibly important" (P3) when researching individuals, this requires the user to know key information such as full name and date of birth-information which is not always known when researching queer individuals. Another participant observed that while marriage records are often a great way to trace individuals.
"… when it comes to trans people they just disappear, because to legally change your name in the nineteenth century, you just had to reintroduce your-self…there was no process, there was no record, it was just like "Hi, I'm Jim" … there's nothing, people just disappear. So we have to rely on…either them disclosing, or other people mentioning it, [such as] cases where they get exposed to the press" (P5).
Because these tried-and-true archival search methods do not always work for queer histories, multiple participants cited extensive use of digital catalogues and OCR-enabled platforms such as Trove, describing them as either very useful, or absolutely vital for queer historical research.
"thank f**k for Trove, I would be so incredibly f****d if it weren't for Trove and digitised records, to be honest … digitisation has changed the landscape … particularly for [queer and transgender] history because having searchable records really, really changes the way that you can go about finding different cases" (P5). However, the methods described by participants to find records in these systems were complex, time-consuming, and required a high level of specialised knowledge-both of historical terminology and contexts, and archival systems.
While participants described their detailed lexicon spreadsheets with varying degrees of enthusiasm and pride, they also began to list many and varied challenges associated with this method. First was the question of what to search for when conducting a keyword search. There are numerous terms that were used to describe queer and gender-diverse behaviours and identities in the past, and these differed across temporal, geographic, cultural, and linguistic contexts, with the additional complication of spelling variations. Some of the core terms identified by participants were 'pervert', 'invert', 'deviant', and 'sodomist'. These terms also reflect the strong skew of extant records that represent queer men, and the overall lack of sources that document queer women.
Specific challenges emerged for those researching trans and gender non-conforming histories. Medicalisation of trans identities did not filter into the popular press until the 1920s and 1930s, and so earlier reporting used sensationalising language that is often very euphemistic and non-specific. For example, terms like 'masquerade' or 'masquerading as', 'impersonation', 'believed to be', or even 'freak' (as in 'freak of nature' or 'freak weather event-meaning a strange occurrence). Knowing to search these terms requires specialised knowledge, "you need to learn the language they were using at the time in order to do keyword searches, but to know the language you have to find them in the first place" (P5).
Participants also noted that many common terms tend to bring up a large number of unrelated records.
"All of those words come up so much [through OCR] and…it brings up so much that it's really hard to identify what's relevant…I'll just be scrolling through this list of…thousands that have come up" (P2). This is because of the specificity of the euphemistic terms (P5) "you get a lot of masquerade balls, and you get a lot of electoral impersonation", but can also be because of OCR errors, or a consequence of truncation: one participant (P4) described searching for transgender-related records using the truncated term 'trans*' and getting results relating to transmission or transport.
Perhaps due in part to the limitations of keyword searching, several participants described a particular kind of archival breadcrumb trail. Instead of consulting records related by provenance, or linked through formal citation, several participants described tracing explicit references to other cases of gender nonconformity mentioned in media reports: "… when a new case came up the newspapers would be like "this is so weird, we've never seen anything like this before. By the way, do you remember this one, this one, this is the names [sic]" and then a list of incidents to go and search for. Or sometimes people weren't named, but there would be details like he worked at this dairy or this orchard, or there was this dispute with an employer, and then I search for a phrase around that or like search for that employee around that date …" (P5).
The particular value of these kinds of media reports are their explicitness. Rarely for queer history, a sort of lineage is created (albeit through sensationalising media, and discourses of 'outing'). This kind of explicitness is extremely valuable to researchers in the archive, specifically because it is so rare.
Many participants explained that the 'queerness' of a record is often not explicitly identified. Records that document policing (such as arrest records and media reports) are often only identifiable by the specific charges that were commonly laid against queer people. These charges were rarely a charge for 'homosexuality' for example, but instead for offensive behaviour, public indecency, or vagrancy (these charges in particular were often laid against people dressing in a gender non-conforming way).

Provenance searching
A few participants navigated archives at the level of provenance, collection, series, or container. This approach is most common when the researcher has a specific organisation or group in mind and knows where relevant records are most likely to be. The bulk of the searching work takes place after the container has been requested for access (e.g., manually checking each item in the container for relevance). This method of searching is a version of 'if this record exists, it will most likely be in this location. If I cannot find it, it likely doesn't exist'.
Participants who worked at this higher level of navigation encountered many of the same challenges as those working at item-level: a lack of explicit descriptive metadata, vague or archaic language, and having to work through large quantities of unrelated records to find a few items of relevance. One participant explained how legal clerks sometimes used outdated terms in official registers, using their own language and understanding of what an offense was to indicate the nature of those files in an official record. For example, the formal charge may be 'unnatural offence' but the register would list it as 'sodomy or 'buggery'. Another participant who had worked with the records of a specific organisation and observed how insular and specific the language could be. One organisation, or even one subset of individuals within that organisation, may use the same terms or euphemisms to describe male homosexuality, for instance. The participant described collating meeting minutes that used a certain term, only to later realise it was about homosexuality: "So I had inadvertently collected an archive about how [organisation] was dealing with [homosexuality] without really knowing what it was" (P6).

Language
Across both methods of archival searching, participants noted that they rarely or never rely on item-or box-level descriptions, primarily because they rarely explicitly describe records as queer: "obviously I'll use the boxes to like work out exactly what year…when I'm looking at lots of things, but beyond that [I don't use them] a whole lot… say the [state] archives…those descriptions never have anything in them that tells me it's going to be relevant …saying, you know, this is related to 'gender diversity'…so what would be the point [of referring to them]" (P2).
Instead, as I've described, participants often rely on keyword searching within the records themselves (via OCR). However, this raises the issue of the often-pejorative language used within these records to describe queer people and queerness, either through the terms and phrases themselves ('pervert', 'sodomite', 'invert', 'heinous act'), or through the context of their use (legal charges or highly medicalised terms).
While participants had encountered numerous terms that would not be used today-and if encountered outside of a historical research context would be much more confronting (and potentially harmful)-their personal reactions to these terms in an archival setting was surprising. One participant felt affirmed in their research project when encountering homophobic terms in a mid-twentieth century police record: "I'd been researching for so long and I'd had this doubt … So actually, finding that note around 'perversion' was quite a revelatory moment and I wasn't particularly offended by that term, and it was actually quite exciting that I'd found a primary resource that suggested it" (P3).
Another participant found humour in the terms used to describe queer people, while still being conscious of how they might impact non-researchers: "… it makes me feel like I need to be very cautious about my browser history. Like when I'm teaching and I'm bringing up a web page or something, I don't want it in the drop down … But I often find that, like some of them so far removed that I can find them a little funny" (P5).
However, while sometimes they found the terms funny, this same participant did emphasise that if a non-expert encountered these terms, the potential harm is much greater because they lack knowledge and context: "…they wouldn't know what they were looking at, and it would be awful… sometimes history is bad, and [includes] terminology you shouldn't repro-duce…it can be useful, but it needs to be handled carefully … there's a way to recontextualise the past without reproducing bigotry" (P5).
This last point-that we don't need to reproduce the past in order to study it-is one that several participants emphasised.

Serendipity and doubt
In their 2002 user study, Duff and Johnson reported that.
" [a]lthough historians often speak about the role of serendipity in their discovery of relevant material, there is strong evidence to suggest that this process is influenced less by serendipity and more by the deliberate tactics of the expert researcher. In other words-what appears to be accidental discovery is accidentally found on purpose" (Duff and Johnson 2002, p. 495).
This study came to a similar conclusion. Participants described methods of searching that were painstaking and methodical, time-consuming, highly detailed, laborious, and extremely dependant on expert knowledge of both queer history and the structure of archives. However, participants who worked at both item-and provenance-level had stories about "serendipitously" finding relevant records. The serendipity was that they managed to find these items while working either against, or without, descriptive metadata. I've already noted one participant who described "inadvertently" collecting an archive. Another described quickly looking through a box of records to see if they were relevant, looking for 'S for sodomist' and 'H for homosexuals'. They found nothing, and almost sent the box back, before realising that they were interview transcripts, alphabetised by surname. The records were not clearly described, there was no catalogue record of there being interviews at all, and no one (to the participant's knowledge) had found records on this topic before: "my breath quickened, I started shaking" (P1). The participant had requested this box as one of several hundred during an international research trip, simply hoping to find something of relevance. Had this box been properly described, it would likely have been the first thing they requested.
The systems and structures, as they currently exist, do not make finding queer records a simple or straightforward process. This is possibly why the feeling of serendipitous discovery is so potent, and so memorable. The lack of adequate archival processes to support their discoverability gives the impression that queer records miraculously emerge from the ether. Of course, serendipity is Janus-faced, and many participants described feelings of uncertainty, doubt, and anxiety about their ability to find any (let alone all) relevant records within current systems. Multiple participants who relied on item-level searching expressed uncertainty about whether this method allowed them to find everything that might be relevant: "…it works, I guess. I wouldn't say it's comprehensive…If you don't know what to search for, you won't find it. If you assume that what's in your search results is everything that's there, and not just what the OCR picked, you're pushing yourself into a corner" (P5). "…the assumptions that I made when I was tackling my research about what language…would bring up results…I'm still very like mindful of the fact that, like, there could be a magic term that I haven't stumbled across that would bring up a wealth of sources…I'm in this gap, I always find, between the terms that we think that will bring up sources and the terms that do. And then, how do we capture things that we might have missed?" (P2).
One participant (P3) noted how disheartening it can be to find nothing, despite laborious searching, and stated that they have been forced to abandon projects because the materials either do not exist or cannot be located-and they weren't sure which was the case, which was particularly frustrating. Another wondered aloud how "we can find those things that might not be occurring to me [to search for] … in terms of how we bring everything together" (P5).
In contrast, participants who navigated at provenance level, expressed less concern about whether they had looked at "enough", or whether their search was comprehensive.
"I mean, in almost no context have I been able to look at everything that would be relevant. But in looking at one institution or doing biographical work, you can get a sense of a [sic] relatively complete…I'm sure there would be more relevant stuff that I haven't found…but I found quite enough…" (P6). This difference in confidence between item-and provenance-level searching methodologies perhaps stems from the broader view that provenance-level searching affords. A provenance-level researcher surveys logically grouped records that are directly tied to their subject of interest. There is no question of whether things haven't been found, only the acknowledgment that they don't exist in that collection-something that is out of the researcher's hands. Conversely, for item-level searching, the onus to locate all relevant records is placed onto the researcher, something that can become very personal to researchers who are working in the area of their own identities and communities: "… it would be easy and irresponsible to just stick with what was easily findable…that's not the kind of history I want to write…The assumption of non-existence is part of the problem" (P5).

Resisting queer/ing
In addition to describing their own search processes and techniques, participants reported that once they had identified potentially relevant records, they would often encounter additional barriers to access. Multiple participants described very extensive redaction ("basically the whole document") or the redaction of details such as place names (e.g. parks that were known beats in the 1960s and 1970s) that are already very common knowledge. Alternatively, they would only be given access to a small percentage of a larger document, file, or collection-often for unclear reasons. Participants who experienced this said that they would understand and respect a logical or legal explanation, but often none was given. Several participants described challenges surrounding embargo periods. They would request a certain file (because it related to a known individual, event, or location) but the embargo period would be much longer (by years or decades) than other, comparable records. Often no explanation would be given for this, and so they described having "suspicions" that it was because of the queer content of the records. One participant had tried to help someone access their own records through a government-adjacent repository. They were initially denied access for "privacy reasons", and then eventually given a heavily redacted copy-of their own records.
One participant (P6) described an experience of archivists at one institution being "visibly uncomfortable" with "salacious" material and suggesting that they "wouldn't want to look at that". The participant did note that the material was not simply about queer people, but about instances of sexual violence, and that this might have been the source of their discomfort. However, they felt that the archivists' job was to provide access to records, and not to "pass judgement" on which materials they were requesting access to. One participant compared the moral "policing" of queer materials in archives to that of pornographic materials, which are often kept off public display, and under heavy access restrictions.
Multiple participants had experiences with active and ongoing mechanisms of erasure and described a systemic and deeply embedded resistance to queer readings of records. One participant was researching an individual who may have been gay, and had experiences with external parties who viewed his research as being an insult to the individual: "I was very conscious of, "you can't accuse someone of being gay"" (P3). Several other participants expressed a similar sentiment: that queer readings were viewed as an insult against someone who could not "defend themselves". Government and government-adjacent repositories were particularly acute sites of this resistance. One participant, who studied an organisation's response to homosexuality, noted that while the organisation was "happy to name and shame" people in positions of authority that enacted homophobic policies, "the [queer people] themselves are silent, they have been erased from the record … a deliberate archival policy to silence the subjectivities of these indi-viduals…archivists are considering themselves to be the gatekeepers of morality, which they are not. I wasn't explicitly saying these [people] were queer, but I was still being denied a queer reading of these records" (P1).
And another participant stated, "… I think a lot of it comes from…not wanting to like overreach and make decisions on behalf of people in the past, but I think it's not more objective to presume people were cisgender or heterosexual by default; it's still a decision that you're making to prioritize that particular way of understanding historical truth" (P5).
The challenge of describing queerness in the past is certainly complicated by the evolution of language over time and across contexts, and the reality is that assigning modern terminology to actors and actions in the past is inherently anachronistic. However, concerns around anachronism, and the question of precisely what terms we should use to describe records when identifying them as 'queer' (in whatever sense) are distinct from an embedded and institutional resistance to queer readings overall. The experiences of these participants are particularly notable because the records in question were often explicitly and solely created to document homosexuality. Ironically, while archivists are concerned about 'accusing' those documented in the records of being gay, that was often precisely why the records were created. Lytle (1980) has suggested that user studies are potentially more challenging in archives than in libraries, as "research needs are difficult to assess; the needs are diffuse and the users are unaccustomed to articulating their information needs" (p. 70). I won't comment on the general accuracy of this statement. However, in my experience of conducting these interviews, I found queer researchers, and researchers of queer history, to have specific, explicit, and well-articulated information needs. Participants were extremely adept at describing their information-seeking behaviours, explaining the challenges they faced, and articulating both the potential benefits and limitations of potential solutions. In conducting these interviews, I sought to answer several key questions surrounding the visibility and accessibility of queer records in institutional archives. What is the practical process of searching for queer records? How do researchers identify, locate, and access relevant materials, given the general 1 3 lack of explicit queer description in archival catalogues? And what are the challenges of this process-both in terms of a research methodology, and an experience for queer researchers within potentially hostile or unsafe spaces? The conversations I had with my participants extended far beyond these questions, reaching from the challenges of managing multiple large spreadsheets, to their experiences of queer identity, and how they felt dealing with archives and archivists as someone doing queer historical work.

Discussion
A key finding to emerge from these conversations is that current institutional descriptive practices do not make queer records visible and accessible. I began this study with the knowledge that the records of queer history held in institutional archives would be highly individualised-that is, they would be records of individuals' interactions with the state and the media. As a result, keyword searching across collections is (in many instances) more effective than provenance-level searching. However, keyword searching for queer records is a particularly challenging task. Not only are the terms used in most records vague, euphemistic, or pejorative (or some combination thereof), they are also highly context-dependant, and therefore require a high degree of specialised knowledge. Additionally, keyword searching relies on a range of other technologies and systems. This study revealed that digitisation-and particularly OCR-are vital to the doing of queer history in institutional archives. In one participant's words, researchers "would be so incredibly f****d" without it. The tensions around digitisation are now well-established. While user-demand for digitised collections continues to grow, while the challenges of resource availability (particularly funding), technical capability, and the long-term infrastructural demands of digital content preservation and accessibility only seem to become more complex. For clarity, I am not suggesting that digitisation would be a panacea to the challenges of queer archiving. In fact, I want to emphasise that this reliance on OCR to compensate for the overall lack of accessibility to queer records is not sustainable-not only in terms of time and labour, but also the environmental cost. While digitisation and OCR are certainly useful and valuable tools to make archives accessible, they cannot compensate for inadequately managed collections.
This relates to another key finding of this study: there is a pressing need for rich, specific, and explicit metadata to identify queer records in institutional collections. The fact that clear and explicit metadata makes records more accessible should be self-evident. However, the nuances of how this importance manifests for marginalised histories were surprising. Contrary to my expectations, participants were not particularly distressed or off-put by encountering pejorative language, either in records themselves, or in descriptions. As I described in the Findings, some participants even described being excited or amused by the language. However, in this "all you can do is laugh" response, there is a clear and obvious problem to which we, as archivists, can respond: the visibility of queerness in institutional archives is so limited that the use of pejorative and discriminatory language is the best-case scenario, because it allows queer records to be (relatively) easily searched for and identified. Explicit metadata, in every sense.
The fact that participants found the slurs contained in records to be a more effective way of finding relevant items than consulting archival descriptions and finding aids should be a bucket of cold water to the archival profession. There is a "second order of violence" enacted when people encounter pejorative archival description (Hartman 2008, p. 5). They can be (re)harmed through this interaction, and we should not diminish the potential reach or impact of this harm, particularly in the digital context, where increased accessed to both collections and finding aids can magnify the impact of harmful description (Sutherland and Purcell 2021). Moreover, as this study has demonstrated, these terms are not effective metadata. As Chilcott has articulated in relation to racist language, "[t]he problems are twofold: discoverability may be reduced if records are described using terms that users are unlikely to search and ethical concerns around the perpetuation of historic racism arise" (Chilcott 2019, p. 360). Participants described searching for the archaic, euphemistic, or pejorative language used to describe queer people as difficult, timeconsuming, and ultimately not one-hundred-percent effective. However, it was still the best tool they had at their disposal.
The findings of this study clearly demonstrate that the lack of explicit queer (re) description using contemporary and inclusive language actively reduces the accessibility of queer records. Public archives should be accessible to the entire population, and marginalised communities especially should be able to access their own histories. However, even these expert participants, who have the required time, knowledge, and resources to search archives, reported a lack of confidence in their ability to identify relevant materials. The lack of visibility for trans and gender-diverse histories was even more acute, with sensationalised media reports forming a substantial amount of the available evidence. Given all these challenges, how is the broader queer community meant to access their own history? I am not advocating for universal item-level (re)description. This would not only be unfeasibly time-consuming and labour-intensive but would in many instances be unnecessary. However, there is a fundamental and important difference between equality and equity. Marginalised groups may need additional mechanisms to be represented accurately, respectfully, and fully within archival collections. This may be (re)description to manage pejorative language; metadata enrichment to improve visibility or accessibility; or the creation of detailed finding aids or research guides to bring dispersed and disparate queer records together. The value of these targeted access mechanisms is clearly visible in community archive spaces, as participants noted that one reason queer community archives were so easy to navigate was the explicit and clear language used in description and finding aids.
This study has identified multiple challenges to implementing explicitly queer descriptive metadata. A core one being the question of what terminology to use is a complex one. How can we know how an individual would have identified, particularly when the records were created about and not by them? There is also the issue of context, and the fact that our sociocultural understandings of gender and sexuality have evolved, and continue to evolve, substantially over time (Baucom 2018;Roberto 2011). Terms that were used by individuals or communities in the past are sometimes considered offensive or inappropriate now (Wood 2013), while the terms we might apply to individuals today are often "out of time and out of place" in archives (Rawson 2018, p. 333). These questions have been, and continue to be, deeply and thoughtfully considered within queer community archives. More than a problem to solve, language has been utilised as a powerful tool within the context of queer archives and archiving. As Poole notes, "…terminology constitutes a foundation for the development and affirmation of identity. Community archives revamp arrangement and description both to represent their materials and to provide for findability, access, and (re)use on-and often literally in-their own terms" [emphasis added] (Poole 2020, p. 11). Cifor provides a potent example of this, describing how archivists at the Sexual Minorities Archive have "a self-created system of subject classification…that embraces affectively and politically charged language. It responds to hate, as found in those materials that negatively describe sexual minorities, as "bullshit" (2016, p. 764). Along similar lines, Rawson explains how the Digital Transgender Archive conceptualises and uses 'transgender' "to refer to a broad and inclusive range of non-normative gender practices" (Rawson 2018, p. 333).
While community archives are generally more interested in how we represent queer identities through description (Baucom 2018;Cifor 2016;Cifor and Rawson 2022;Rawson 2018), institutional archives remain stuck on the question of whether we should. Participants described an embedded resistance to queer readings and uses of records-and by extension, to queer historical work generally. This resistance to, and erasure of, queer readings was manifested and enacted in multiple ways-access restrictions, extensive redaction, and even outright and explicit "moral gatekeeping" (as one participant described it) on the part of archivists. These findings suggest that the lack of queer visibility in institutional archives is not simply a matter of benign neglect. Not only are archives and archivists not doing work to improve the visibility and accessibility of queer history and perspectives in institutional collections, but in many instances, they are actively working to resist them. Entrenched homophobic and transphobic ideas mean that queer readings are still often viewed as being an accusation, against which people cannot defend themselves. Even when records held by institutions are recognised as being definitely 'queer' (generally records documenting the prosecution or institutionalisation of individuals), some participants described feeling that institutions are more interested in performatively vilifying homophobic policies and officials than giving dignity and voice to queer people themselves. This not only made their historical work more practically challenging to do, but also required queer people to enter and navigate spaces that were potentially hostile to both their work, and them as individuals.

Conclusion
Through conducting interviews with queer archival users, this study illuminates some of the substantive issues and challenges of locating queer records in institutional archives. It highlights the central role in rich, clear, relevant, and respectful metadata in making records visible and accessible, and reveals the limitations of current descriptive practices for queer archival users and uses. Finding aids and research guides are generally of limited value to queer archival users and uses, and participants noted that they rarely or never rely on item-or box-level descriptions, primarily because they rarely explicitly describe records as queer. This was placed in direct contrast with queer community archives, where the use of explicit and community-generated language made relevant items easy to identify. The methods described by participants to find records in institutional archives were complex, time-consuming, and required a high level of specialised knowledge-both of historical terminology and contexts, and archival systems. The lack of adequate descriptive metadata to support discoverability of queer records meant that many participants lacked confidence in their ability to find any (let alone all) relevant records within current systems. While several participants stated that they felt either affirmed or amused by pejorative language, they also expressed a desire for clear and respectful metadata, emphasising that that we don't need to reproduce the past in order to study it. At the same time, multiple participants had experiences with active and ongoing mechanisms of erasure, and described a systemic and deeply embedded resistance to queer readings of records, and a broad persistence in the idea that queer history involved "accusing" people of being queer. These two findings in combination paint a picture of archival practice, and archival institutions, as being hostile and potentially harmful spaces to queer users, and queer communities more broadly.
This article does not propose any specific or immediate solutions to the challenges of queer (in)visibility and (in)accessibility in institutional archives. Indeed, this is only the first of two papers reporting on the findings of this study. However, in establishing the existing challenges of queer/ing history in institutional archives and painting a clear and holistic picture of the experiences of queer archival users, it demonstrates the need for transformative change in the way that we think about and do archival description. As I've noted, we must move beyond our preoccupation with the archive of absence and confront the non-neutrality of our interventions today. If the archive is constantly becoming (McKemmish 1994, p. 20), then how can we shape this becoming? How can we make extant queer records visible and accessible not only to queer researchers, but to the broader queer community? How can we ensure that these (often forgotten) histories become part of our broader collective memory? How can transformative approaches to arrangement and description increase the visibility and accessibility of the records we do have? What role do emergent digital technologies play in this transformation? What benefit can participatory approaches bring? And ultimately, how can institutional archives and mainstream archival practice evolve alongside and in collaboration with transformative community approaches to do this work? At the beginning of this paper, I asked: where do we find queer history? But perhaps the better question is how do we find queer history? And how do we leave a map for those following in our footsteps?

Conflicts of interest The author declares that they have no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.