Queering archival metadata? The discoverability of queer collections in GLAM institutions and universities in Aotearoa New Zealand

Described as a “larger-than-life personality” Carmen Rupe (1936–2011) was a well-known icon of the Wellington drag scene in the 1970 and 80s. She described herself as a transsexual; however, this term is now considered to be pejorative and has been replaced with the widely accepted transgender, although there are still people who choose to identify as transsexual. Transsexual is also a Library of Congress Subject Heading (LCSH) term (transgender was added in 2018) and is used in current and historical archival descriptions. Nonetheless, sexuality and gender identity language is fluid with new terms introduced and others reclaimed, such as queer, in a rapidly evolving space. Appropriately describing and redescribing materials donated to institutions from queer communities and individuals, by recognising the fluidity of gender identity to continually represent them how they wish to be represented, is critical. This not only retains their trust in the institution's integrity but also expedites effective discoverability and accessibility. Being able to find, access and use queer materials means that LGBTQ + communities are visible in local, regional and national cultural heritage after prolonged exclusion.

This paper explores queer description and metadata in Aotearoa New Zealand's galleries, libraries, archives, museums and universities (GLAMU) by seeking to answer two research questions:

The term “queer” is applied in this paper, in the queer theory context as a disruptor of whatever constitutes the normal, and also as a group identifier. It is acknowledged that queer as an identifying term is contested both as a reclaimed word and when applied generically to include a diverse range of gender identities and sexual orientations. LGBTQ+, the Homosaurus preferred acronym, is also applied and it is recognised that acronym use is also exclusionary.

In this paper, the acronym GLAMU will be used to denote the five institutional types of galleries, libraries, archives, museums and universities.

The authority to include or exclude archival documents resides with GLAMU institutions. This is a formidable power given that “the documentary record provides just a sliver of a window into the event” (Harris, 2002, p. 64) and that having appropriate queer metadata enables effective and efficient discovery. In recent decades, scholars (Carter, 2006; Caswell, 2014) have decried the archive's silence concerning the LGBT + communities, their “symbolic annihilation” from the nation's cultural heritage and the dominance of the cisheteronormative in archival theories and practices.

There have been moves within some GLAMU institutions to find the voices of the marginalised within new and existing collections (Carter, 2006). In the case of the queer communities, this means using appropriate terminology and description during accession to facilitate document discovery or through revisiting existing collections. Users encounter significant problems when attempting to locate queer records in institutional archives, highlighting the need for better queer metadata and description (Freeman, 2023). The terminology used by archivists is crucial to expedite the discovery of queer materials; otherwise, users are unable to access queer materials. This is compounded by the lack of queer community-developed archival descriptors in GLAMU institutions, which further reduces access to queer histories, particularly for more marginalised or newer queer communities (Baucom, 2018). Investigating the situation from a national perspective affords valuable insights and knowledge into the complexities and experiences of using queer metadata, as Freeman (2023) demonstrated by exploring queer access in Australian institutions.

Rawson (2018) highlighted the paradox of working in a sphere that calls for fixity in description and categorisation when queering metadata means pushing boundaries and terminology fluidity. GLAMU classification structures, particularly the LCSH, and evolving language and meaning within queer communities are seemingly incompatible (Johnson, 2010). With its use extending beyond the library to a range of cultural heritage institutions, Watson (2024, p. 249) described the LCSH as a “monolithic, universalising infrastructure”, pointing out that its focus on “aboutness” has serious metadata implications for queer and other marginalised communities. This means that existing terminology and descriptions are often limited and may not identify documents relevant to queer communities (Davison, 2011). Moreover, the slowness of the LCSH to change offensive headings has been well-documented in the literature, without any guarantee that headings will be changed (Bailund et al., 2024; Baron et al., 2024). The identities of genderqueer, agender and pansexual, for example, are still rendered invisible by the lack of subject terms (Angell and Roberto, 2014; Fischer, 2023).

Within the GLAMU sector, there are conceptual shifts to rectify the marginalisation and inequality wrought by these outdated but internationally used systems. Hardesty and Nolan (2021) demonstrated the benefits of linking Homosaurus, an LGBTQ + controlled vocabulary, with the LCSH to offer an improved and equitable search experience for queer communities and researchers. Dobreski et al. (2022) concluded that the range of Homosaurus identity terminology more effectively addresses LGBTQ + communities' requirements than the Library of Congress Demographic Group Terms and the LCSH vocabularies. Institutions that have implemented Homosaurus found it more closely represented user search terminology (Fischer, 2023), with the vocabulary providing a more accurate description of and accessibility to queer materials (Javaid and Parkhurst-Strout, 2024). Furthermore, Homosaurus can be applied across various GLAMU contexts to expedite the discovery of queer materials. The use of appropriate, respectful descriptions and terms is critical to ensuring the representation and visibility of all queer gender identities and sexual orientations in collections, and it remains an under-researched area of scholarship in Aotearoa New Zealand.

Describing queer identity can be a challenging aspect of archival description. Baucom (2018) found that when compared to LGBTQ + history websites, archival descriptors in institutional finding aids for gender identity and sexual orientation were outdated. However, for historical personages, this may reflect the decision-making processes of the information professional. Rawson (2018) highlighted the tension between using a historical term, now perceived as offensive, to describe a historical person or employing a current term that may not reflect how the person viewed themselves but is now more acceptable. Wagner et al. (2025) noted that information professionals are inconsistent in their approach. In the finding aid context, some practitioners would respect the language individuals used at the time to identify themselves, while others would apply modern terminology to facilitate discovery by current researchers and users. The varied use of queer terminology, the expansive array of available terms and the lack of a standardised approach will affect users seeking queer materials.

Reparative description is defined by the Society of American Archivists (2024) as the “remediation of practices or data that exclude, silence, harm or mischaracterise marginalised people in the data created or used by archivists to identify or characterise archival resources”. Reparative description must acknowledge the existence of discriminatory and offensive terminology that has been used to describe marginalised communities and rectify the problem. Projects are being implemented at various institutions, including Yale Library (Clemens and Lehman, 2024) and the University of Washington Library (Indemaur, 2024). Notably, much of the literature draws on the experiences of institutions in North America, including public and academic libraries (McCormack, 2024; Gloger and Mikel, 2024), with queer as one facet of a wider marginalised community approach. Where reparative description projects have been queer-specific, they have led to “a meaningful increase in visibility and accessibility for the materials” (Indemaur, 2024, p. 19), although the basis for this assessment was not elaborated on. Olson (2023, p. 3) emphasised the importance of undertaking reparative description, regardless of project size and scope to “continue to strive towards a more inclusive equitable and diverse representation of the populations they document”.

Nonetheless, uncertainty about how to undertake reparative description sustainably, along with varying institutional approaches, means the impact can be difficult to assess (Luke and Mizota, 2024). The slow archives concept, as advocated by Christensen and Anderson (2019), centres on relationships and people rather than faster processing. However, deGraffenreid (2024)found that applying this concept is resource-intensive and thus difficult for institutions to implement. Recognising the inherent difficulties of implementing a slow archives approach, deGraffenreid modified the concept to prioritise reparative and inclusive description for easier implementation within current work practices. While implementing a more formalised metadata process would contribute to the consistency practice, how this can be appropriately evaluated has yet to be determined. If and how reparative description is enacted in GLAMU institutions in Aotearoa New Zealand forms part of this study.

GLAMU information professionals who become queerly informed will have the tools to understand classification and metadata from a queer perspective and how and why classification infrastructures, like the LCSH, can offend queer users (Drabinski, 2013). Expanding on this view, being queerly informed could encompass an empathetic understanding of queer culture and empower queer interventions, such as a queer-controlled vocabulary, like Homosaurus, that works within the system to improve the discoverability of queer materials. In terms of LGBTQ + reparative description, Homosaurus has been viewed as an important and legitimate source for gender-affirming language and as a tool for increasing practitioner understanding (Wagner et al., 2025). Nonetheless, Tanaka et al. (2024), noted issues with scalability because of the labour-intensive nature of supplementing terms as well as the requirement for ongoing monitoring and maintenance. This should not detract from the value of Homosaurus; however, its current application is in small, targeted projects.

Queer community participation in metadata and description shares community understandings of terms that do not cause harm. Shilton and Srinivasan's (2007) participatory archiving model envisaged communities and/or individuals actively involved in the appraisal, arrangement and the description processes of their collection. Knowledge of the collection and community expertise can afford valuable context, producing not only community-driven metadata but building trust-based collaborations for the future. For such participation to be meaningful and dynamic, institutions need to be prepared to fully engage with communities and understand their needs (Rolan, 2017) with the degree of involvement subject to community and/or individual wishes (Huvila, 2015). Such engagement enables creators to use their own terminology to describe themselves (Allgood and Wagner, 2025). Nonetheless, improving queer metadata will not affect discovery if other communities and users remain unaware of the changes (Tanaka et al., 2024). Shifting user perceptions of queer inaccessibility through engagement, education and outreach is therefore vital. These are explored within the Aotearoa New Zealand context to offer fresh perspectives on the potential benefits for queer metadata and description as scholarship in this area remains limited.

Until 1986, homosexuality was illegal in Aotearoa New Zealand. Discrimination was commonplace and being “outed” could mean loss of both family and employment. Queer material, if it was kept at all, was often destroyed by either the person themselves or later by their family, while a pervading “conspiracy of silence” around the existence of queer communities meant they were largely invisible in Aotearoa New Zealand (Parkinson, 1984). The GLAMU institutions, cognisant of these societal attitudes, did not collect queer materials and queer histories remained hidden and inaccessible. Even up until the 21st century when societal attitudes had begun to shift, institutional queer collections were virtually nonexistent because of the negligible research in this space and the resulting lack of demand for queer material. Brickell (2005) asserted “that until fairly recently scholarly investigation of the history of homosexuality was not regarded as a respectable endeavour”, while for transgender, Takatāpui (Māori) and Fa'afafine (Pasifika) this was even later (Hansen, 2020; Kerekere, 2017). These attitudes were by no means confined to Aotearoa New Zealand; indeed, most countries espoused similar views. Downs (2021) pointed to the failure of the American historical profession to embrace gay history, while in Australia, if gay history was referred to at all in official histories, it was only in the context of crime, sin and sickness (Gorman-Murray, 2004).

These academic practices and attitudes towards queer histories had a considerable impact on institutions' collecting strategies. Providing resources for researchers is an important part of the work of GLAMU institutions, encouraging acquisitions in that area (Avery, 2013). Excluding queer history and sexual diversity from universities, and by extension academic journals, meant GLAMU institutions were much less likely to accept proffered queer donations or actively seek to develop collections in that area (Downs, 2021; Maynard, 1991). Of those collections that did make their way into an institution, standardised, cisheteronormative classification and description systems, such as the LCSH, were applied.

This research is scaffolded by the queer theory, from queer studies and influenced by Foucault (1978) who contended that sexuality and gender are social and historical constructs. Biological sex functions as the power determinant of what constitutes normal sex and gender identity, resulting in the categorisation of identity instead of behaviour. Queer theory challenges the rigidity of these postmodern constructs of sexuality and gender identity (Carnes, 2019; Sullivan, 2003) advocating for identity fluidity to allow multiple expressions and interpretations. The very nature of constructed concepts invites the creation of alternative narratives of living and thinking that are diverse and inclusive, where difference is the norm, not the exception. Underpinning this is the dynamic concept of “queer” and how it should be interpreted and applied in real-world contexts (Watson, 2005). According to Eichhorn (2018) archival queering has two facets. The first involves applying a queer lens to materials that have not been historically interpreted as queer. The second involves information professionals actively intervening in archival practices to facilitate queer accessibility, visibility and queer collecting. In the context of metadata and description this involves not only employing terminology appropriate for queer materials but also understanding how to queer the description of existing and future materials when the gaze of classification systems and controlled vocabularies routinely used in GLAMU institutions is cisheteronormative.

The research in this paper draws on a larger project, undertaken by the author, that explored the collection, documentation and donation of materials relating to Aotearoa New Zealand's queer communities in GLAMU institutions and queer community archives. The larger project received approval from the Human Ethics Committee of the author's university. Two qualitative methods were employed by the author: survey and interview. For the survey, sampling was purposive to select the most appropriate data sources to answer the research questions (Douglas, 2022). The selected survey sample comprised 29 galleries, libraries, archives and museums located in the top 20 population centres in Aotearoa New Zealand. The archives and/or special collections located in each of Aotearoa New Zealand's eight universities were also included in the sample. Institutions were selected based on three criteria: an online and searchable heritage collection, receiving at least some government or council funding, and accepting donations of archival materials. Of the 37 contacted, 33 institutions responded, including all eight universities.

Qualtrics was used to administer the survey, which featured ten open and closed questions. Of these, only two (one open and one closed) were relevant to this research, as they requested reparative description information about queer materials for existing and future collections. The responses produced by these two survey questions contributed to answering the research questions posed in this paper.

The second method employed was interviews, applying a responsive interview model with semi-structured questions. This model's flexibility allowed participants to contribute their own experiences and views and introduce new lines of inquiry, increasing the value and depth of the data (Rubin and Rubin, 2012). The study conducted 32 interviews with information professionals in roles involving archives, cultural heritage and/or special collections. Each institution typically employed only one or two people in these roles. Three methods were used to identify participants: as survey respondents who consented to be interviewed, through researcher knowledge, or through the institution. Location and/or participant preference determined whether the interview was conducted in person or online. One participant provided a written response (P39). Of the remaining 31 interviews, 24 were individual, while seven were dyadic, increasing the richness of the data from the interactions that occurred. Information professionals who participated in both the survey and the interview, and those who participated only in the interview, are denoted P1–P39. Survey-only respondents are denoted S1–S11. All institutions have been anonymised in this paper. The term “information professional” was employed to encompass the variety of interview participants’ job titles.

Interview questions relating to queer metadata included the fact whether reparative description was occurring and how this was enacted, how queer materials were described and some of the ways this was approached, the use of any queer terminology and queer vocabularies and how these had been selected as well as the participation of queer community donors in the description of their materials. Use of Homosaurus or other vocabularies was also included in an interview question as this had not been covered in the survey but was raised in some responses. Only the data generated from these interview questions were used to answer the research questions posed in this paper.

NVivo was used to analyse responses to the survey's open questions and interview data, using Braun and Clarke's (2006, 2022) reflective thematic analysis. This iterative, reflexive approach involves seeking the best explanations for the data and the best responses to the research questions. Following data familiarisation, initial codes were created which were refined and reviewed during the analysis process as themes emerged. Because applying a queer theoretical lens was integral to this research, a deductive-inductive continuum approach was adopted (Kennedy, 2018). Through continued researcher engagement with the survey and interview data, the data familiarisation process, coding and theme creation became data-driven (inductive) rather than predefined (deductive) along the deductive-inductive continuum (Braun and Clarke, 2022). Applying a queer lens involved interpreting data by looking for evidence of “queerness” (Eichhorn, 2018), identifying correlations between lived experiences and theory and exploring nuances that might account for any disparities.

The trustworthiness and value of this qualitative research was ascertained by applying criteria developed by Lincoln and Guba (1985). The four criteria of credibility, transferability, dependability and confirmability take account of the interpretive nature of qualitative research, recognising that there are multiple accounts of social reality. The criteria are interdependent as some of the ways proposed for one criterion also apply to another. The researcher ensured all four criteria could be corroborated through informal and formal mechanisms, including data triangulation, an audit trail and a transcript approval process.

The extracted findings are grouped under three headings, reflecting the key themes that emerged from the survey and interview data analysis.

The majority of New Zealand GLAMU institutions hold at least a few queer collections or items, although some, including the national institutions, have larger holdings. The queer materials mainly represent the dominant queer communities of gay and lesbian to the detriment of the more marginalised groups, such as transgender and intersex. However, there are considerable backlogs of unaccessioned collections that could potentially contain queer materials, or in existing collections that were not identified as queer materials at the time because descriptive practices were different (P8). “None of the previous generations of archivists had done anything with the collection, except put it in boxes, put it on the shelf and make a catalogue card that said the [name of person] collection” (P22). Often, collections were arranged and described at high-level because of time and resource constraints (P2) and also prevailing archival theory. The more product, less process approach of Greene and Meissner (2005) recommended only high-level description at the collection or series level to facilitate user access by minimising processing. This meant folders, files and items remained undescribed. The general lack of description in many collections led to queer materials being unidentified (P24), even in available queer collections. In the case of one transgender person's collection, which the institution had already accessioned, the participant explained there was so much uncatalogued and unknown material that the staff had to go through it again (P5). The insufficient description of accessioned collections affected the findability of materials, especially for existing collections and impacts the representation and visibility of queer communities in local, regional and national cultural heritage.

Participants also revealed the very limited range of queer keywords; some did not use a general term, such as LGBTQ+, that would facilitate locating material under a broad search (P23). Others had employed “gay” and “lesbian” as descriptors rather than a general nomenclature (P37) but also mentioned that their application had not been consistent, indicating that some materials may not have been identified. The lack of consistency was also evident across institutions that did apply a general term, and it was often the only queer term in their catalogue (P21). One interview participant highlighted the inherent problems of applying and using queer metadata, stating, “you might have something but if you don't put in a subject term, if you don't put in those correct keywords for what people might be looking for, then they just might miss it completely” (P1). Additionally, when queer material was described, the metadata was usually specific, a name or title such as the Homosexual Law Reform Bill, rather than applying queer subject headings or keywords (P1). To locate queer material in this way meant that it was essential for users to possess prior knowledge of the topic in order to search effectively, thus limiting the discoverability of queer information.

The LCSH was seen as outdated (P35) by the institutions that used it, affecting information professionals' ability to describe and classify appropriately and ethically. One participant revealed “there was a subject heading of alternate lifestyles. And when I opened the box, the majority of it was LGBTQ content. And I was like, No, we're not calling it that anymore” (P18). Participants expressed concern that often LGBTQ + related LCSH terms are inappropriate. The rapidity of change means terms that may been acceptable are now inappropriate and harmful for queer communities.

Queer identity metadata, if used, had to be stored in the institution's collection management system (CMS) if it was to function as a searchable keyword. This was revealed to be problematic for several reasons. The existence of a field in institutional CMS to record identity varied; some CMS did not include a gender field (P7). The evidence also showed that having a gender field did not necessarily mean it was populated, and when identity information was recorded, it was often placed in biographical information (P10, P18). However, one information professional noted how easily the term nonbinary was added to their CMS. Gender diverse was also added to the system's gender field of “male, female, other” viewed as most appropriate at the time (P29), demonstrating the potential for CMS gender identification. Even so, the recording of identity information was uncommon; one participant admitted that “everyday social history that isn't screaming rainbow rights, we're not recognising it in our databases and we're not even asking people” (P17). Caswell's (2014) assertion of “symbolic annihilation” appears apt for many institutions in New Zealand, with queer communities rendered invisible through the absence of identifying metadata.

A handful of institutions used a general term, such as rainbow or LGBTQ; most had not agreed on a general term (P29), leaving individual information professionals to make that decision. A form of LGBTQ was the most popular general term, either LGBTQI+ (P21, P14) or LGBTQ+ (P35). Neither acronym is Aotearoa New Zealand culturally friendly as it reduces queer Māori (Takatāpui) and Pasifika (Fa'afafine) to the “plus”. Although it is arguably impossible to represent all queer groups, as Sheffield (2020) has so rightly pointed out, the New Zealand LGBTTFIQ+ (lesbian, gay, bisexual, transgender, Takatāpui, Fa'afafine, intersex, queer/questioning) is more culturally representative, as it includes Māori and Pasifika.

A main reason for the scarcity of queer metadata was the fear of “outing” historical figures or using new terminology that they may not have been comfortable with (P19, P24). The prevailing view of information professionals was that “we are looking at it through today's lens, we can't make an assumption about what that person's experience was and what they would have wanted put out into the world” (P21). Information professionals explained that they would consult with the LGBTQ + person about terminology use, especially if the person's identity was fluid and subject to change or if the most current term was being applied (P14, P29). However, for personages who could no longer be consulted, representing those life experiences accurately and respectfully, raised considerable ethical issues for information professionals and their institutions regarding appropriate descriptors. Some, like Carmen Rupe, who was discussed in the introduction, used terms like transexual to describe themselves during their lifetime. Superseded by transgender and now regarded as derogatory, an information professional would not apply transexual unless clear supporting evidence existed, as in Carmen's case. In other cases, historical relationships seen through today's lens could be perceived as queer. As one information professional commented, “in the first half of the 20th century, there are quite a lot of artists in my collection who could well have been gay but were never out or were never explicit” (P24). They went on to explain that rather than definitively label these artists, adding context to the description about the content of artworks, or about the artist's living situation, were ways to increase the visibility of queer histories in their collections.

Caution underpins metadata use by information professionals and institutions, meaning queer identifiers would not be applied if there were any concerns or unease. There had to be publicly available evidence and even then, the situation might be such that no descriptor to that effect would be used, nor would assumptions be made. As one information professional put it, “you don't want to be the one to put a label on them, even though there's something going on” (P29). Families of historical personages could also voice concerns; one information professional noted a label had contained a reference to a man living with his male partner in the 1930s and that “The family of the partner, they didn't get cross about it, but they noted it and it was definitely a point of slight tension, I think for them” (P24). Navigating family concerns could be challenging and information professionals had to balance queer visibility with the wishes of historical personages' families.

To work around this issue, information professionals would aim to provide enough information to alert people to the possibility but allow them to make their own interpretations without having to use the queer or LGBTQ + label.

Areas of descriptive tension for information professionals involved deciding whether to identify and then how and where to identify using appropriate metadata. If a person was definitely gay but never out, then “that's the kind of thing that I might put into a catalogue record, somewhere that's not visible to the public, because I don't think that's public information. But I also don't want that information to be lost” (P24). This documentation practice meant privileged gender identity information was restricted to the information professional community, the only people with access to the system.

Using an LGBTQ + vocabulary, like Homosaurus, could mitigate some of the inconsistencies in terminology, and can also be used in conjunction with the LCSH and for archival description. In New Zealand, Homosaurus has been introduced at three GLAMU institutions. Homosaurus use was suggested during a staff meeting at one institution to make material more keyword-accessible in their archives management system for queer audiences. “So, we decided we'd start using some keywords like queer, rainbow and that's when the Homosaurus discussion came up and we decided we'd like to use it” (P19).

Despite this, most information professionals reported being unaware of Homosaurus, as they had not needed to describe queer materials (P2, P13, P14). One institution used the Getty Thesaurus and applied to have queer New Zealand Māori and Pasifika terminology accepted. They found this process time-consuming and only partially successful, with some requests refused (P14, P15). Conversely, Homosaurus responded positively to the addition of Aotearoa New Zealand-specific terms to the vocabulary, such as Takatāpui (Māori). As a result of the interviews and increased awareness, several participants requested metadata and description resource links, including Homosaurus.

All participants were aware of the deficiencies in queer metadata and description and many expressed a desire to engage in reparative description. Some commented they had not yet had the opportunity (P1, P13), others admitted it was not a priority because of resource constraints (P20) while other institutions had no plans to instigate reparative description (P8, P31). One institution would review metadata only if someone raised an issue of misdescription (P32) while another intended “to improve subject classifications in our database as more options and fields become available to us but this is usually on a case-by-case basis” (P23). The majority of reparative description was therefore handled on a case-by-case basis with information professionals fixing inappropriate metadata only when a problem was brought to their attention (P5, P19, P30).

A few institutions intended to be more proactive in this area to varying degrees, often as part of larger projects such as a collection audit, or as part of a digitisation project for some of the collections (P19). One institution undertaking an audit acknowledged the challenges in revisiting queer description, not only because of the lack of time and resources, but also “due to their descriptions, these items can be difficult to identify in our database and online” (S3). Other institutions reported similar problems. Many of their collections contained limited descriptions as they had been arranged and described only at the top level (S3, S5, P2, P18, P23). Some of these institutions were investigating the evaluation and subsequent expansion of their archival descriptions to increase discoverability, which would potentially include identifying and describing queer materials (S3, S5, P18). Another institution was in discussions about a queer tagging project to make queer artworks, archives and narratives more visible on the institution's website (P11).

A couple of institutions used student interns to undertake short-term projects to improve the level of queer description. However, these projects depended on access to students interested in queer archiving and able to undertake an internship at a GLAMU institution (P22, P29). One institution is beginning to undertake ongoing proactive reparative description (Roache, 2024), albeit not retrospectively; this is one of three in Aotearoa New Zealand, the other institution is still in discussions (R31). This is a collection-wide initiative, encompassing queer materials within its scope. The small team at this institution focuses on identifying and remedying outdated terminology and improving metadata in under-described collections. For LGBTQ + metadata, this involves replacing inappropriate gender and sexuality-related LCSH with terms from Homosaurus. Nonetheless, this manual process was described as “very time-consuming”, an issue for a team operating under budgetary and time constraints (P38).

The literature (Baucom, 2018; Olson, 2023) recommends queer community participation and expertise in describing their materials and providing community-driven metadata. One information professional who planned to “completely redescribe and potentially digitise some collection items” had “met with the president of the University student Queer Soc to discuss whether that would be meaningful for them and what level of involvement they would like” (P19). This has led to ongoing institutional discussions about carrying out reparative description to make collections more accessible. At another GLAMU institution, the aim was to “make description a more participatory practice and process by consulting the LGBTQ + communities and respecting their wishes in terms of access, description, and use of the collections” (P25).

Most GLAMU institutions in New Zealand have no connections with queer communities; however, the benefits from a queer metadata perspective of building trusting relationships are considerable. Access to specialist knowledge is a notable advantage, with one institution's information professionals noting that their transgender donor's expertise led them to use current, community-accepted terminology instead of outdated nomenclature (P14, P15). Another participant outlined the beneficial metadata outcomes of a substantial and ongoing donation of queer digital materials to their institution's archival collections. The digital materials were all community tagged and became the catalyst for discussions about applying outdated LCSH to queer materials (P35). As a consequence, Homosaurus was introduced for use alongside the LCSH “so that way we can use those more appropriate subject headings on the collection description” (P35) and to include the community tags through enriched descriptions.

One information professional explained how their local LGBTQ + group, with whom they have recently developed a connection, is now undertaking an oral history and object collection project supported by the GLAMU institution. The descriptions will be community-derived in collaboration with the information professional (P9) as the institution will host the digital materials on its website. From a metadata perspective, these queer histories and narratives become visible and accessible through collaboration between institutions and communities, using community metadata to improve discoverability while utilising institutional systems.

According to the International Council of Archives (2023, p. 6), one of the three essential purposes of description is the “ongoing use and reuse of records”. Without appropriate metadata and description records become inaccessible and unusable to their communities. The lack of queer terminology means users may require search abilities to find materials, while the inconsistent keyword usage across institutions reduces findability.

In Aotearoa New Zealand's GLAMU institutions, current and historical descriptive practices meant queer materials were often unidentified, under-described or assigned topic keywords that required prior queer knowledge for effective searching. Consequently, users had to develop search skills and knowledge, for instance, knowing that “Homosexual Law Reform” would locate queer materials. Freeman (2023) identified similar issues faced by LGBTQ + users in Australian institutions, highlighting that the lack of queer metadata affected search and discovery. Applying general terminology, such as LGBTQ+, was typically practiced by institutions in Aotearoa New Zealand, rather than using specific terminology to differentiate gender identity (Baucom, 2018). Moreover, the research showed that if more specific gender identity groups were distinguished, these still tended to be gay and lesbian, rather than representing the range of sexual orientations and gender identities. This is to the detriment of these communities, as these practices have significant implications for the discovery of queer materials.

Discovery was affected by previous descriptive practices, which had involved collections receiving only top-level description, as well as institutional backlogs of unaccessioned collections. Stemming from resource constraints, a lack of queer awareness among information professionals and social attitudes at the time were also contributing factors. These practices meant potential queer materials remained hidden from view.

The findings also revealed that while it is evident that some information professionals are making changes, these are often introduced on an individual, ad hoc basis rather than as the result of a coordinated and consistent approach, even within the same institution. There is also no evidence of coordination between institutions regarding queer metadata application, meaning different terminology is frequently deployed, including various acronyms, as well as the terms rainbow and queer. This affects the institutional cross-discoverability of queer materials, as well as their discoverability within the institution's own collections.

The research revealed that evolving queer identities made description complex and problematic. This specific queer context complements Kaplan's (2000) less nuanced contention about identity inconsistencies. The research revealed that queer description is also problematic because of the shifting consistencies in queer identities. The research also emphasised the influential role of information professionals in determining and creating identities. The type of metadata and description used can improve or reduce the discoverability and accessibility of materials. This furnishes information professionals with considerable power, often on an individual basis. Rawson (2018) noted that materials are described based on the information professional's interpretation, and, with a cisheteronormative lens dominating Aotearoa New Zealand's institutions, queer nuances would likely remain undetected.

The research also found that identity description is governed by the proviso of not causing harm, which encourages risk-averse practices. Applying a contemporary lens meant that a person in a collection or collection creator might be perceived as queer. However, the fear of outing someone is a significant concern, and a lack of terminology to describe undetermined identity meant information professionals would not identify queerness without indisputable evidence. If and how to appropriately describe the identities of historical personages is another area of tension, as this had to be balanced with discoverability and current queer terminology. As the findings indicated, information professionals tended to carefully consider how historical personages would have wanted to be described and to respect their self-identity, as this is usually very personal. Consequently, there is a reticence to ascribe a current term the person may not have used, aligning with Rawson (2018). Information professional reticence was also consistent across institutions and influenced practice, in contrast to Wagner et al. (2025), who noted varying approaches to current terminology use. Only a few information professionals employed practices aligned with the recommendations of deGraffenreid and Goodrich (2023). This involved using context in descriptive notes for historical personages with complex gender identities and relationships when gender identities were indeterminate, allowing readers to draw their own conclusions. Furthermore, Wagner's (2022) descriptive approach to visual materials, which aimed to avoid misgendering when corroboration was not possible, was similarly underutilised. Current descriptive practices tended to forgo a reference to gender identity and sexual orientation if the evidence for queerness was not clearly articulated and/or if families were involved.

Some information professionals recorded problematic identity metadata in hidden fields in the CMS, rendering these materials LGBTQ + invisible by not making the information public and searchable. While this practice retains the information, conversely, it can be argued that such practices could cause harm by hiding queer identities and making the records inaccessible to LGBTQ + communities that are then unable to see themselves represented in cultural heritage institutions. Furthermore, identity practices are inconsistently applied both within and across institutions as they depend on the individual interpretation and decision-making processes of the information professional. This is influenced by the information professional's positionality and whether they are able to view the information from a queer perspective.

The inconsistency of approach and the under-recording of possible queer identities in institutions affect the visibility of queer materials. Using resources developed for this purpose, for example, by The Queer Metadata Collective (2024) and The Trans Metadata Collective (2023), could mitigate some of the tensions for information professionals working in these areas and develop consistency in practice. GLAM institutions could also ask about identity description in institutional donation or gift agreements, including preferred pronouns, and respect the right to silence if people choose not to make a disclosure.

Metadata challenges also arose with the LCSH, used in some Aotearoa New Zealand GLAMU institutions for archival, cultural heritage and for special collections materials as well as for library-related items. The evidence of this research revealed that current practice was to describe queer materials with LCSH's outdated terminology, which not only impacts accessibility but also causes offence to these communities. The LCSH has been the subject of scholarly criticism for its inappropriate metadata and descriptions for LGBTQ + communities (Johnson, 2010; Drabinski, 2013). Another challenge facing research participants was the LCSH's slow adoption of updated LGBTQ + terminology, supporting other research in this area (Bailund et al., 2024; Baron et al., 2024). This affected work practices around applying appropriate metadata, increasing the marginalisation of these communities. Similarly, describing contemporary gender identities was also challenging for Aotearoa's information professionals because of a lack of appropriate and respectful LCSH terms, augmenting existing scholarship (Angell and Roberto, 2014; Fischer, 2023).

There is a critical need for metadata to represent the array of queer communities using community-appropriate terminology. With three institutions (including one national) at various stages of implementing Homosaurus, other GLAMUs might consider introducing this vocabulary given its ready acceptance of country-specific terminology and its compatibility with the LCSH. Moreover, in Aotearoa New Zealand many GLAMU institutions already implement Ngā Upoko Tukutuku (Māori subject headings) alongside the LCSH and are thus familiar with using additional vocabularies to address some of the LCSH's metadata inadequacies. Scholarship advocates implementing an LGBTQ + controlled vocabulary in conjunction with LCSH to appropriately describe queer materials. Linking Homosaurus with the LCSH would not only offer an improved and equitable search experience for LGBTQ + communities and researchers (Hardesty and Nolan, 2021) but would also enable the use of appropriate metadata and description by GLAMU institutions (Dobreski et al., 2022; Javid and Parkhurst-Strout, 2024). Thus, implementing Homosaurus more widely in Aotearoa New Zealand GLAMU institutions could increase access to queer materials by applying appropriate queer metadata and ensuring consistent use of terminology to support queer discovery.

However, for institutions that possess queer materials, the shift towards using an LGBTQ + vocabulary to supplement the LCSH has been slow. Many research participants indicated they were unaware of the existence of a queer vocabulary. Moreover, institutions that had begun implementation reported there were challenges. While scholarship has addressed those relating to scaling (Tanaka et al., 2024), this research highlighted issues in applying Homosaurus in day-to-day operations. One information professional acknowledged the labour-intensive processes that are required while operating with limited resources, as well as the need to collaborate with the IT team on implementation. Such challenges could affect other institutions' adoption of Homosaurus.

One benefit of applying a queer-controlled vocabulary would be to cultivate a queerly informed perspective among information professionals unfamiliar with queer materials as most research participants fell into this category. Applying a queer lens for descriptive purposes should form part of the way materials are viewed, to recognise that they are “telling us a story, a certain something that is unsaid” (P9). Queering material does not require information professionals to be queer themselves, but it does require an understanding of queer culture to effectively apply a queer lens as a component of queering (Eichhorn, 2018). Queering means information professionals are able to critique, challenge and reflect on cisheteronormative systems dominated by a specific epistemic approach. This involves having the tools and skills to comprehend classification and metadata from a queer perspective, and understanding why and how classification systems, such as LCSH, can offend and harm LGBTQ + communities (Drabinski, 2013). The extent to which research participants were informed about queer cultural issues also varied considerably, with the majority unaware. Information professionals becoming queerly informed would not only increase the visibility and representation of queer materials in the institution, but also help recognise possible identity and description issues that may not be apparent through a cisheteronormative institutional lens.

The paucity of queer metadata and the need to improve descriptions in existing records were common themes with profound implications for the discovery of queer materials. Part of a reparative archive approach advocated by Hughes-Watkins (2018), repairing the harm caused by embedded structures and systems, including queer description, is extent, becoming part of current practice in some of Aotearoa New Zealand's GLAMU institutions. However, the research indicated that implementing reparative description is slow and variable.

Olson (2023) and McCormack (2024) have pointed to the growing need for GLAMU institutions to implement reparative description to fix offensive and outdated metadata to prevent queer and other communities from harm. For LGBTQ + projects, this is important because reparative descriptions improve the discoverability of materials (Indemaur, 2024). Conversely, the findings revealed that reparative description is an ongoing practice in only one institution; queer communities are a designated priority of this wider initiative. This highlights Olson's (2023) view that for reparative description initiatives to be valuable, they should be institutionally supported to ensure consistency and continuity. The “in response to feedback” approach employed by many Aotearoa New Zealand GLAMU institutions is unlikely to have a significant overall effect, although its impact is difficult to assess (Luke and Mizota, 2024). Another issue was resourcing, as many smaller, and some larger, institutions in Aotearoa New Zealand face constraints. Several participants explained that any reparative description undertaken was informal because of competing priorities and a shortage of resources, including staff time.

A few information professionals discussed reparative descriptive projects at their institutions, including employing a student intern to work on a specific collection and revisiting existing descriptions as part of a broader collection audit, particularly if some materials were to be digitised. Predominantly, queer materials were a project component alongside other marginalised communities rather than the specific focus. Nonetheless, participants were aware of the issues with queer description and metadata and intended to rectify these. These approaches, including student internships or collection digitisation, were designed to improve description and the application of appropriate metadata and could be adapted to suit other Aotearoa institutions looking to revisit description and metadata practices, particularly for queer communities.

Working with queer collections can bring considerable benefits through building relationships built on trust and mutual respect. Moreover, it can prove a catalyst for transformative change regarding metadata and description as the importance of describing queer materials appropriately and aligning with queer community terminology is recognised and validated by the institution (Caswell et al., 2017). However, many institutions lacked connections with their queer communities. Developing queer connections not only builds trusted relationships but also helps ensure the metadata is queer-community appropriate and compatible with institutional cataloguing systems and controlled vocabularies. This was evident in the donation of a large queer collection to a national institution, which instigated not only discussions about LCSH appropriateness and the use of queer community-derived metadata but also provided the catalyst for introducing Homosaurus into institutional practice.

As the findings from two institutions indicated, collaborating with local queer communities on projects such as oral history or object collection can build sustainable relationships. Such collaborations not only raise queer cultural understandings (and preserve queer stories), but applying queer community-based metadata in descriptions would improve the discoverability of queer materials.

While many institutions hold some queer materials the use of appropriate queer description and metadata is inadequate. Limited keywords, use of outdated subject headings and lack of consistency means queer materials are difficult to search and find in GLAMU institutions in Aotearoa New Zealand. Inadequate levels of description in existing collections also reduces the discoverability of queer material, while tensions exist around identity-related decision-making, affecting queer visibility.

This research demonstrated that there are several approaches institutions could adopt to improve queer metadata and description and enhance the discoverability of queer materials. Implementing a queer-controlled vocabulary, like Homosaurus, would help to alleviate outdated and harmful terminology and inconsistencies currently present, offering some uniformity in a fluid queer terminology space. While characterising queer identity enables discovery, appropriate description can be problematic. Information professionals often exhibited considerable caution and were concerned about queer misrepresentation. To support information professionals' decision-making processes, community-devised resources can aid informed decision-making (Watson et al., 2023) and help them develop their own institutional guidelines.

Reparative description occurs on a predominantly case-by-case basis in many institutions thereby continuing not only to cause harm with outdated queer metadata but also decreases the discovery of materials. However, this research also highlighted that many institutions in Aotearoa New Zealand, undertake reparative description in response to public feedback or when they identify a problem during daily operations. While resource constraints limit many smaller institutions in this area, some of the larger institutions could build reparative description into their work processes more consistently, or implement a reparative description internship project to help effect more meaningful change.

Engaging with queer communities and queer donors can produce not only greater knowledge and understanding of queer culture but can also offer benefits if communities can incorporate their own descriptors and metadata in consultation with the institution, either through a project or donation. This not only helps build trusting relationships but also produces meaningful and queer community-appropriate metadata, increasing the findability of these collections.

Queer materials need to be discoverable for users seeking such information with appropriate metadata and levels of description. The experience of Aotearoa New Zealand indicates that it is through wider and more consistent approaches that some of the identified metadata and description issues can be mitigated.