Wikimedia Australia and first nations metadata: utilising the ATSILIRN protocols to create culturally appropriate description and access

ABSTRACT

Protocols have been utilised as a tool to build dialogue on the management of Indigenous knowledges in a library and archive context. This paper discusses how the Aboriginal and Torres Strait Islander Protocols for Libraries, Archives and Information Services (ATSILIRN [n.d.] Protocols for Libraries Archives and Information Services. http://atsilirn.aiatsis.gov.au/protocols.php) could guide Wikimedia projects and Wikipedia and Wikidata editors to better describe Aboriginal and Torres Strait Islander content in a self-determined and culturally appropriate manner. Focussing specifically on questions of Description and Classification (Protocol 5), the paper examines the limitations of current metadata and Wikipedia practices and opportunities for embedding and being led by Aboriginal and Torres Strait Islander worldviews and principles. Overall, the article contributes to further understanding the challenges and opportunities of including Indigenous knowledges and perspectives in open access platforms.

KEYWORDS:

• Indigenous knowledges
• first nations metadata
• indigenous protocols
• Wikimedia Australia
• Indigenous librarianship

Introduction

Growing cultural and linguistic diversity of content and participating communities on Wikimedia platforms has become a key strategic focus of the international Wikimedia Foundation over the last decade (Wikimedia Research, 2022). Recognising that the cultural protocols and Indigenous Data Sovereignty principles (Maiam Nayri Wingara & Australian Indigenous Governance Institute, 2018) guiding the management of First Nations knowledge, content and data are disparate from the open access principles that drive the Wikimedia movement, Wikimedia Australia has prioritised deeper engagement with First Nations communities to explore how the development of Aboriginal and Torres Strait Islander content and contributors can be better supported to engage with Wikimedia platforms and activities in a self-determined, culturally appropriate way. This paper outlines and reports on a targeted research initiative undertaken by Jumbunna Research in collaboration with Wikimedia Australia to understand how the Wikimedia movement could better support First Nations content, data and contributors on Wikimedia platforms and beyond. It does this by drawing specifically on protocols for managing Indigenous materials in libraries and archives across Australia.

Research methods

The Wikimedia Australia and First Nations Metadata research project involved desktop research, searching the literature broadly (including across grey literature, websites, and other published sources) to respond to a series of discussion questions. As Williamson (2018) suggests, the literature review plays a crucial role in the research and provides a background and framework to explore the discussion questions in relation to other studies. While Cecez-Kecmanovic and Kennan argue that “Gaps in the literature are rarely obvious or easily identified” (2018, pp. 127–155) it was clear that there was little existing literature on First Nations engagement with Wikimedia, particularly in the context of metadata, that could be drawn upon.

As an exploratory project, the team used researcher reflexivity, informed by Indigenous Standpoint Theory, to reflect on our own professional experiences of working in the Galleries, Libraries, Archives and Museums (GLAM) sector to support Indigenous access to cultural collections and information. According to Behrendt, Indigenous Standpoint Theory supports transparency in the research process:

Indigenous standpoint notes up front that we, as individuals, are shaped by our cultures, cultural values, and experiences within society's institutions […]. In this way, Indigenous standpoint theory plays an important role in challenging the assumptions around neutrality that are actually a way of reinforcing power structures of colonisation and patriarchy. (2019, p. 176)

Several yarning sessions on the discussion questions were undertaken within the research team to share insights on the topic. Yarning, as a research approach, enables scholarship through conversation and storytelling (Bessarab & Ng'Andu, 2018). Collaborative Yarning methods were used to delve deeper into the topics to surface knowledge on our shared understanding of protocols and First Nations metadata. Bessarab and Ng'Andu (2018) describe Collaborative Yarning as an approach where two or more people actively engage in sharing information about a research project and engage in a discussion about ideas. Similarly, Shay (2021) describes Collaborative Yarning as a culturally safe and appropriate method to surface new knowledge. Collaborative yarning and yarning methods can also mitigate settler colonial processes because they focus on relationality (Atkinson et al., 2021). Our approach also focused on strengths as opposed to deficit views and privileges the voices of Indigenous people.

The ATSILIRN Protocols

The Aboriginal and Torres Strait Islander Library, Information and Resource Network (ATSILIRN) Protocols for Libraries, Archives and Information Services (the ATSILIRN Protocols) were used as a framework for examining the topic of First Nations Metadata in the context of Wikimedia Australia. The ATSILIRN Protocols were published by the Australian Library and Information Association (ALIA) in 1995 to guide library and information professional practice in relation to Aboriginal and Torres Strait Islander peoples and materials. The ATSILIRN Protocols were endorsed by the Aboriginal and Torres Strait Islander Library, Information and Resource Network (ATSILIRN). ATSILIRN later updated the Protocols in 2005 and 2012, with the latter update addressing the digital environment. The ATSILIRN Protocols were designed to be broad and to work as guidelines to accommodate and support a diverse range of organisations and the ever-changing library and information context (ATSILIRN, n.d.; Garwood-Houng, 2005; Thorpe, 2013).

The ATSILIRN Protocols identify twelve areas that are relevant to Aboriginal and Torres Strait Islander peoples and materials and broad principles were developed to highlight these areas and guide practice. For the purposes of this research project, only Protocol 5: Description and Classification was considered, as it was noted by Wikimedia Australia that Wikimedia projects are currently ill-equipped or unsupported in reviewing terminology, subject categories and content for outdated, inaccurate and value-laden terms.

ATSILIRN protocol 5: description and classification

We have been referred to and catalogued as ‘savages’ or ‘primitive’ while Western industrial peoples are referred to as advanced and complex. (Mick Dodson, in ATSILIRN Protocols, Preamble to Protocol 5, 1993)

ATSILIRN Protocol 5 notes:

Indexing terminology, subject headings and classification systems are designed to provide easy access to materials in libraries, archives and information services. However, the use of outdated, inaccurate or value laden terms actually obstructs access. To improve access organisations will:

• 5.1 Use national Indigenous thesauri for describing documentation relating to Aboriginal and Torres Strait Islander peoples and related content.

• 5.2 Promote appropriate changes to standard descriptive tools and metadata. schemas with the aim of retrospectively re-cataloguing items recorded with unsuitable subject headings.

• 5.3 Improve access by the introduction of classificatory systems which describe items by their geographic, language and cultural identifiers.

• 5.4 Consult with Aboriginal and Torres Strait Islander peoples at local, state/territory and national levels in relation to the description, cataloguing and classification of materials in libraries, archives and information services.

• 5.5 Provide opportunities for Aboriginal and Torres Strait Islander people to describe and annotate material that relates to themselves and their communities (ATSILIRN, n.d.).

Research questions

The research project examined the following questions in the context of the ATSILIRN Protocols, and specifically Protocol 5 Description and Access:

1. What should Wikidata editors know about using national Indigenous thesauri in their modelling of items and elements relating to First Nations peoples, languages, place names and bibliographic items in an Australian context?

2. What guidance can Wikimedia Australia give Wikipedia editors to improve the cultural appropriateness of categories, article short descriptions, and content related to First Nations peoples, languages, places and culture?

3. What Wikipedia articles and Wikidata items should be priorities for the community to review and update?

4. What strategies can Wikimedia Australia use to ensure a First Nations perspective in these priorities?’

5. How can Wikimedia Australia work with other metadata organisations to improve access to appropriate classificatory systems and geographic, language and cultural identifiers, ensuring that custodianship of these identifiers remain with the people they describe?

6. How do Wikimedia Australia collaborate with other metadata organisations embed and engage with Indigenous research principles into metadata practices?

7. How do Wikimedia Australia build partnerships with state-based and regional libraries/archives who are able to engage directly with their local communities will benefit all parties?

Understanding indigenous cultural and intellectual property (ICIP) rights and Wikimedia

Wikipedia, the open-source online encyclopaedia, is one of the most popular websites in the world, attracting almost as much traffic as Facebook and Twitter (Statista, 2022). It is contributed to by over 100,000 regular volunteer editors. It forms part of a broader Wikimedia ecosystem that also includes Wikidata, a machine-readable free and open knowledge base and the multimedia database Wikimedia Commons. Wikimedia platforms offer significant visibility of referenced content to a vast global audience however First Nations content from Australia is under-represented and. in some cases, culturally inappropriate or inaccurate. As an open knowledge system, it also enables usage of that content by other third-party platforms (for example, in Google search boxes), amplifying content across the internet, along with any inaccuracies, existing bias and knowledge gaps.

The question of Indigenous Cultural and Intellectual Property (ICIP) rights in connection with Wikimedia is an emerging area of research. Recognising the strong connections between library metadata and Wikidata data modelling as well as Wikipedia citation practices, the impact for this research extends beyond the Wikimedia community to organisations that influence Australian metadata standards and practice. The activation of the ATSILIRN Protocols enables greater visibility on the protection of ICIP rights and the need for First Nations people to have agency and decision-making about the use and re-use of their cultural heritage materials. According to Gilliland, protocols enable co-creator rights in descriptive processes, noting that they “acknowledge Indigenous people as co-creators of these materials and promote their rights” (2012, p. 343).

The ATSILIRN Protocols recognize the moral rights of Aboriginal and Torres Strait Islander peoples as the owners of their knowledge and the importance of their involvement in deciding issues arising from Aboriginal and Torres Strait Islander content and perspectives in documentary materials, media, and traditional cultural property. They also encourage Indigenous participation in the governance and operation of libraries, archives and information services. (Gilliland, 2012, p. 343)

The Protocols encourage dialogue and flexible approaches for responding to Indigenous peoples needs around description and classification. Therefore, while a certain level of guidance can be provided, metadata is expected to be responsive to Indigenous people’s needs locally.

In the United States, Duarte and Belarde-Lewis (2015) have utilised the decolonising methodology of imagining as a tool to assert Indigenous epistemological views of knowledge organisation. They describe the importance of understanding the distinctiveness of local Indigenous community ways of knowing, being and doing to guide classification and descriptive practices (Duarte & Belarde-Lewis, 2015). Writing later on about the importance of centreing relationality, author Littletree joined Duarte and Belarde-Lewis to discuss how the foundations of Indigenous ways of knowing have to be considered in knowledge production processes associated with classification and description (2020). Their words resonate in an Australian context, as producing incorrect, inaccurate or incomplete metadata can be harmful and reductionist in describing cultural knowledge flows in an Aboriginal and Torres Strait Islander context.

In discussing the use of ATSILIRN Protocol 5: Description and Classification and its implications for Wikimedia and other metadata organisations, it is critical to understand the tensions that exist in the context of Indigenous and western epistemological views of knowledge management. While the discussion that follows aims to highlight areas for collaboration and for building respectful relationships, it is important to continue a dialogue about these areas where different ways of knowing clash. Nakata’s use of the Cultural Interface (2007) provides a framework to understand the intersections between Indigenous and western knowledge systems and the potential spaces for learning across culturally diverse communities.

As discussed earlier, the ATSILIRN Protocols also encourage respectful practice to support First Nations people’s moral rights and associated rights to protect ICIP. ICIP is a term that encompasses a wide range of Indigenous Cultural and Intellectual Property, including both tangible and intangible Indigenous cultural heritage. In a Wikimedia context, it is vital to recognise that ICIP rights encompass areas including Indigenous languages and the documentation of Indigenous heritage and histories (Janke, 2021). The notion that information, records and knowledge are free and open to all through GLAM institutions is not always commensurable with Indigenous worldviews.

Christen (2012) challenged the normative practices of the open access movement, asking the question, “does information really want to be free?” in a First Nations information context, and in doing so, challenged researchers to consider their role in supporting digital rights management for Indigenous knowledge. Because of the nature of First Nations materials being collected as part of colonisation, and ongoing colonial processes, the question of Indigenous people’s moral rights becomes even more pressing. Especially given that some of the historical collections held in GLAM institutions were collected without appropriate ethical considerations, including free prior informed consent. Thorpe and Booker (2022) argue that the lack of informed consent in collecting processes can lead to the misappropriation of Indigenous knowledges, languages, arts and cultural expressions.

International work is underway to develop tools and frameworks to support Indigenous people’s moral and ICIP rights through ethical practices. Across GLAM, the Local Contexts project Traditional Knowledge Labels and Licences are working to ground Indigenous rights across the digital information landscape. For over a decade, Anderson and Christen (2012) were promoting practical strategies for redressing rights to support Indigenous people's ability to control their cultural heritage materials. An early focus was on support for Indigenous people to address issues of ownership, access and control of the material that had been previously documented and recorded and now resides in cultural institutions worldwide. Montenegro (2019) recently described how these tools provide a way for Indigenous people to subvert mainstream metadata standards to support Indigenous people’s priorities for Indigenous Data Sovereignty.

This research project considered what opportunities existed for Indigenous worldviews and perspectives to be incorporated into metadata across classification and description. However, we also recognised the importance of engaging in dialogue about the necessary paradigm shifts that need to take place to make Wikimedia a place where Aboriginal and Torres Strait Islander people want to engage. There are structural and systemic issues relating to privilege, issues of racism and representation, and issues of the lack of control of communities over their cultural heritage materials that must be addressed. Engagement in this topic requires views for deconstruction as well as reconstruction in order to develop initiatives, systems (Thorpe et al., 2021) and methods that support Indigenous people’s cultural information flows.

Discussion

The following themes were identified in the research across the relevant literature.

Understanding first nations information contexts

Methods of organising and disseminating knowledge, such as in Western research, libraries, encyclopaedias, and publishing, have historically been seen as authoritative and objective (Smith, 2012). In terms of Indigenous cultures, these sources have often been created and managed by non-Indigenous people, generally with inherent bias, and have delegitimised Indigenous peoples as experts of themselves and their cultures (Smith, 2012). Kwaymullina (2016) notes that sources built on Western research frame Indigenous peoples as the known but never the knowers. Because of this, there are instances where the inherent bias of those who create and manage this information has led to misrepresentations of Indigenous peoples and cultures (Nakata et al., 2005; Thorpe, 2014). Furthermore, Thorpe has discussed how being positioned as objects of study, and being described as being savages and less than human, has aided the subjugation of Indigenous peoples (2022).

An infamous example of this is Eliza Frazer, a settler woman who in 1836 was shipwrecked on the Butchulla island of K’gari and later recounted her story saying she was captured by Butchulla people and defamed them as savage and cannibalistic. This story became well-known and was the basis for a movie made in 1976. Behrendt (2016) details how Frazer’s embellished narrative was inaccurate, Butchulla people rescued Frazer, and that her falsified violent representations of Indigenous peoples helped promote the British civilising narrative in the local area, leading to massacres and the dispossession of the Butchulla people from their land.

Another example described by Sentance (2022) is how a 1921 article in The Australian Museum Magazine, the magazine of the oldest museum in Australia, stated that Aboriginal peoples were “such primitive folk had the minds of children in the bodies of grown men and women”. Sentance (2022) posits that as this was written in a premier institution periodical by a leading scientist, this statement may have helped justify government-sanctioned “protection” policies happening at the time of publication that stole Indigenous children from their families.

Critical information literacy skills

As such, guidance is needed for Wikipedia editors to build their critical information literacy skills. Critical information literacy (CIL) is a theory and practice that aims to understand the social and political context of information and the production of knowledge (Drabinski & Tewell, 2019; Tewell, 2018). This context can help people question “authoritative” sources and, to examine the systems in which that “authority” was established and consider whose voices may be missing and the potential bias of the source creator (Cope, 2010). Moreover, it can get people to reflect on how their own personal values and histories may influence what sources of knowledge they engage with or disregard (Cope, 2010).

One tool that may be useful for Wikipedia editors for this is AIATSIS Guide to Evaluating and Selecting Education Resources (2022). This guide was designed to assist those who use it to choose appropriate resources for teaching Aboriginal and Torres Strait Islander histories, cultures, and languages respectfully and effectively and support those who use it to critically self-reflect on how personal bias may affect the choice of resources. It can be effective in aiding Wikipedia editors to select respectful and appropriate sources for articles related to Aboriginal and Torres Strait Islander peoples and cultures.

Harmful language

It is paramount to draw the line on and prevent racist, offensive and derogatory language from being used to describe Aboriginal and Torres Strait Islander peoples, cultures and histories. Existing articles with this language need to be a priority for editing, editing racist language with appropriate terms or applying sensitivity warnings on the historical context of the language used. To support further understanding regarding what terminology is appropriate, editors can refer to terminology and style guides discussed in this report. Thorpe (2022) has described that further awareness needs to be built in understanding the harms and dangers of racist, derogatory and offensive materials on the lives of Aboriginal and Torres Strait Islander people. Others, such as Wright and Laurent (2021), discuss how in an archival context, liberatory memory work needs to recognise the potential risks for people engaging with archival materials and seek to minimise trauma triggers and harm by developing frameworks of safety, collaboration and empowerment.

Indigenous visibility and reclamation of historical narratives

Potential priorities for Wikimedia to consider could be Country, events and people. Every landmark, town and region in Australia is on Aboriginal and Torres Strait Islander land. Country is present across all lands and waters. Wikimedia has an opportunity to support the enhancement of Wikipedia articles connected to geographic locations in highlighting and referencing the Aboriginal and Torres Strait Islander peoples, histories and cultures of the area. Ideally, this should be done by drawing upon sources created by the Aboriginal and Torres Strait Islander peoples or organisations of the specific area.

Likewise, discussions of historical events often preference settler history and often highlight non-Indigenous peoples as prominent historical figures. The focus on particular historical narratives, for example, those centred around colonial Australia, which ignore or dismiss Indigenous peoples’ presence, work to erase Indigenous histories. To counter historical erasure and provide fuller and richer historical information, Wikipedia editors should focus on identifying and highlighting events of significance to Aboriginal and Torres Strait Islander peoples. Preferably, again, drawing upon sources created by Aboriginal and Torres Strait Islander peoples or organisations relevant to the information being collated.

Furthermore, Wikipedia editors could prioritise reviewing the content of Australian event pages for opportunities to embed Indigenous presence and contribution to that event, such as the sustained Indigenous resistance to colonisation. In addition to this, Wikipedia editors can review language that erases Indigenous histories and consider ways to highlight Indigenous presence. For instance, what O'Brien (2010) refers to as “firsting”, which refers to false claims that a settler is the first to do something on Indigenous lands. Highlighting significant Aboriginal and Torres Strait Islander people throughout history is important in disrupting the invisibility of Aboriginal and Torres Strait Islander sovereignty, self-determination and telling the true history of Australia.

In creating or editing Wikipedia content, there needs to be a consideration of the relationship between who is doing the editing and whose knowledge is being edited. At the 2022 Worlds of Wikimedia conference, Sentance (2022) discussed his experience attempting to create a Wikipedia page about his Great Grandmother. While editors are discouraged from contributing to topics that they have a connection to, Sentance argues his relationship with his Great Grandmother and his community ensured that he was accountable for the knowledge shared. Connection to the knowledge being shared and his responsibility to his family and community created an imperative to ensure the information was correct (Thorpe & Sentance, 2022). At the conference, Thorpe also discussed the lack of representation of Worimi people’s histories on the Worimi Wikipedia page, noting the reliance on settler narratives and anthropological descriptions rather than incorporating resources created by and for Worimi people. Thorpe suggested that if the community was asked, there would be a much more vibrant and representative set of sources used to describe Worimi Country, histories, culture and people (Thorpe & Sentance, 2022). Therefore, mechanisms for building relationships and accountability within content creation and editing could assist in ensuring respectful description and dialogue in regard to Aboriginal and Torres Strait Islander knowledges.

National Indigenous thesauri

During a collaborative yarning session (yarning session 2022), the limitations of using national Indigenous thesauri were discussed, including raising questions about the status of maintenance of these tools, which is sometimes unclear. The Australian Institute of Aboriginal and Torres Strait Islander Studies (AIATSIS) Subject Thesaurus was highlighted as a well-known national thesaurus. The subject thesaurus (Australian Institute of Aboriginal and Torres Strait Islander Studies, 2007) which contains terms that AIATSIS uses to describe the items in the AIATSIS Collections, has both an online web-based version and a downloadable version. However, it is unclear whether updates across these two versions are synchronised. The Governance of the AIATSIS Thesauri document (2021–2024) states that “Additional terms and changes to existing terms in the Subject and Placename thesauri are reflected on the AIATSIS Pathways website on a monthly basis” (Australian Institute of Aboriginal and Torres Strait Islander Studies, 2021). Discussion on the AIATSIS Indigenous thesauri (yarning session 2022) identified a shared uncertainty regarding whether the thesauri was active or relevant and a resulting reluctance to therefore use the thesauri. However, it was also discussed that increasing visibility of communication and avenues for user feedback would provide further clarity on the issues raised.

Fifteen years ago, Stroud reported on the development of digital pathways for access to the AIATSIS collections and how the organisation was leading efforts to create “culturally sensitive and appropriate descriptions of our collections”, also describing that a more appropriate way of describing the AIATSIS collections “covering the rich diversity of Aboriginal and Torres Strait Islander cultures, languages and lifestyles is required” (2009). Not only is the subject thesaurus used for the description of the AIATSIS collections, but it is also a recognised tool for consulting the use of terms, including preferred and non-preferred terms relating to Indigenous topics. Further, the subject thesaurus is considered to be the authoritative source for subjects relating to Aboriginal and Torres Strait Islander people and was recognised as such in its approval by the Library of Congress for use in library catalogue records internationally (Stroud, 2008). Koch (2020) refers to the AIATSIS Thesauri as a tool for cataloguing and setting guidelines for creating catalogue records, including addressing terminology that may be outdated, inaccurate or offensive. Stoker et al. also refer to the AIATSIS thesauri as a “vocabulary of culturally appropriate terms to describe material relating to Aboriginal or Torres Strait Islander peoples” (2019, p.174).

However, there is a lack of clear guidance on using the AIATSIS thesauri, as well as visible information on updates or processes for users to provide feedback and suggest changes. It also requires critique to understand metadata in the context of developing methods for decolonial and anti-racist descriptive practices. The lack of current and dynamic national Indigenous thesauri tools is problematic. As Tai (2022) discusses, for anti-oppressive description to be effective, it needs to be iterative and flexible, changing to match the changing needs of the communities represented. Tai’s views correspond with those of Gilliland (2012), who argued that in an archival context,

Description that is responsive to the needs, epistemologies, practices, and collective identities of communities of record, whether employing a protocols-based approach or some other strategy, is going to have to be nimble and open enough to address such diversity and dynamism. (Gilliland, 2012, p. 345)

In comparison to this, AustLang, the Aboriginal and Torres Strait Islander language thesaurus and database also maintained by AIATSIS (n.d.) is flexible, as it includes detailed citations of the source material. It also enables responses to feedback from Aboriginal and Torres Strait Islander community representatives, as Sentance’s experience in requesting updates demonstrated (yarning session 2022). AustLang is recognised as an authoritative national thesaurus for providing information on Aboriginal and Torres Strait Islander languages. It has been developed through an iterative process to include multiple resources. It has been recognised by the Library of Congress Machine Readable Cataloguing (MARC) language code list as an authoritative “controlled vocabulary of persistent identifiers, a thesaurus of languages and peoples and information about Aboriginal and Torres Strait Islander languages” (Australian Institute of Aboriginal and Torres Strait Isalnder Studies, n.d.).

In addition, it was also discussed that some community members contest AustLang due to citations of record and research predominantly being created by non-Indigenous anthropologists and linguists, and as such, they may contain inaccurate information and misrepresentations. Nevertheless, AustLang is an important starting point in understanding the diversity of Aboriginal and Torres Strait Islander languages and identifying the specific languages connected to particular geographic areas. In 2021, Miniter and Vo-Tran described the benefits of AustLang language codes being incorporated as MARC bibliographic records, suggesting that AustLang can support the classification of Indigenous knowledges. However, while AustLang’s use as a library tool for metadata creation was approved by the Library of Congress for use in 2018, there remain gaps in skills to implement and use the tool and training is required across the sector (Miniter & Vo-Tran, 2021).

Further examples of attempts to build more inclusive language in metadata standards demonstrate the need in the sector. Recent advocacy work was undertaken in 2022 by the Health Libraries Australia to develop more appropriate metadata practices with Medical Subject Headings (MeSH) (Siemensma & McCulloch, 2022), and the recent announcement from the National, State Libraries of Australasia (NSLA) of a First Nations Collection Description project. The project, which includes broad sector representation of a Working party and Advisory group, will aim to

(1) Provide a sector-wide picture of current practice in First Nations collection description by collating guidelines currently in use by the partner organisations and/or their members; (2) Develop and disseminate sector-wide guidelines that will result in greater confidence and consistency in description of First Nations collections by libraries and publishers across Australia. (NSLA, 2022)

This work will fill a major gap that currently exists in providing national guidance on Indigenous descriptive practices. However, at this stage of the project, it is unclear how local communities and ways of knowing will be incorporated into institutional guidance. Further work may be required to bring participatory practices into the thesaurus, as was the case in ensuring First Nations leadership and governance in the development of the Māori subject heading project Ngā Upoko Tukutuku (National Library of New Zealand, n.d.). It is important that Wikimedia Australia are aware of and connected with these emerging projects.

Style and terminology guides

Different style and terminology guides, such as Reconciliation Australia’s Demonstrating inclusive and respectful language, University of Sydney Library’s Aboriginal and Torres Strait Islander Cultural Protocols, University of Technology Sydney’s Top 10 Tips for Terminology, and Chew’s Inclusive Terminology Project can be useful in understanding what terminology can be more empowering and less likely to propagate racist descriptions that undermine First Nations self-representation. Other resources, including the chapter developed by Andersen (2022) on Appropriate Terminology and Inclusive Metadata, can provide guidance on terminology and assist people in considering issues of positionality. The development of style and terminology guides can also aid descriptive work to counter colonial biases and support a right of reply (Indigenous Archives Collective, 2021) and a pathway for redressing racist, offensive and derogatory descriptive practices.

However, when using these guidelines and tools, it needs to be made clear that they are regularly reviewed and are up to date and have been created either by or in collaboration with Aboriginal and Torres Strait Islander peoples. Furthermore, the creation of terminology by or collaboration with Aboriginal and Torres Strait Islander peoples or organisations who are directly connected to a subject is always preferable alongside a national guide or thesaurus.

A potential issue arising from the development and use of national guides is the formation of a pan-Indigenous approach, which in turn may erase the diversity of Aboriginal and Torres Strait Islander peoples, cultures, languages, histories and knowledges. Using thesauri such as Austlang can address this, and so can the development and use of localised resources which can assist in avoiding misrepresentation of dynamic and diverse communities. An example of a localised resource is the Galiwin’ku Classification System from the East Arnhem Land community of Galiwin'ku which reorganised books according to local Aboriginal concepts (Thompson & Trevaskis, 2018). Another international example is the X̱wi7x̱wa Library, which is unceded territory of the hən̓q̓əmin̓əm̓ speaking xʷməθkʷəy̓əm (Musqueam) people which locally developed classification and subject heading system to organise material in their collection (The University of British Columbia, n.d.). Resources and initiatives such as this are preferred and should be supported.

Referring to guides may assist non-Indigenous editors in understanding how the use of terminology can undermine Aboriginal and Torres Strait Islander peoples’ self-representation. For example, the term “Australian Aboriginal” is rejected by many Aboriginal and Torres Strait Islander people as it supports the nation-state’s ownership of Indigenous people as “Australian”. Similarly, Pearson notes some reject the term First Australians as “we are actually the last people who were allowed to have citizenship status which would make us more accurately the Last Australians” (2021). Outside of racist terminology, there can be different preferences between individual Aboriginal and Torres Strait Islander people or Aboriginal and Torres Strait Islander communities regarding terminology around self-representation. Therefore, editors should consider the relationship between the sources they use, and the community or person being represented.

Style and Terminology guides offer tools for considering appropriate and inclusive use of language. However, it is essential that these guides are up-to-date, and that information is shared on the level of Indigenous participation in producing them. Ensuring a balance between the national tools and the incorporation of local First Nations ways of knowing is critical. These kinds of thesauri resources function within a context that is dynamic, fast-moving and highly politicised, and therefore require clear pathways for contestation and feedback and potential to be updated. It is vital that these tools are understood as representing a diversity of views—including contested material—and not fixed. A list of recommended style and terminology guides would be a worthwhile activity for Wikimedia Australia to produce.

Indigenous research principles

Much of Western research on Indigenous peoples and cultures has been noted as a process that has extracted, stored, and controlled Indigenous knowledges by outsiders (Gaudry, 2011; Rigney, 1999; Smith, 2012). This has led to its misappropriation, secret and sacred information being freely available despite cultural protocols, and Indigenous communities and community members not being properly attributed (Anderson & Christen, 2019; Janke, 2021). To support Indigenous rights and prevent this extraction of Indigenous knowledges continuing, different research guidelines and protocols have been developed, such as: Code of Ethics for Aboriginal and Torres Strait Islander Research by AIATSIS; Ethical conduct in research with Aboriginal and Torres Strait Islander Peoples and communities: Guidelines for researchers and stakeholders by the National Health and Medical Research Council (NHMRC); Documenting Traditional Knowledge toolkit by World Intellectual Property Organization (WIPO); and the CARE Principles for Indigenous Data Governance by Research Data Alliance International Indigenous Data Sovereignty Interest Group.

All these established guidelines and protocols for research involving Indigenous peoples, heritage and data are aligned with United Nations Declaration on the Rights of Indigenous Peoples (UNDRIP) and the foundational requirement of free, prior and informed consent (FPIC). FPIC is a right of Indigenous peoples to give, refuse or revoke their consent and that “agreement must be voluntarily given, free from coercion, duress or pressure” (AIATSIS, 2020). FPIC may also be referred to as “informed consent” and is an imperative for the undertaking of ethical research involving Indigenous peoples.

As part of this research, we began to consider how Indigenous research methodologies might be considered in a Wikimedia context. What structural reconsiderations would need to be made to engage with the principles of the AIATSIS Research Code? Or Archibald’s Storywork principles relating to (1) respect, (2) reverence, (3) responsibility, (4) reciprocity, (5) holism, (6) interrelatedness, and (7) synergy (2008). We spent time questioning what kind of systemic changes would be needed to reconsider Wikimedia platforms if they were to respect principles of relationality and accountability in terms of the management of Indigenous knowledges?

The research project found that a key area would be ensuring that processes were inbuilt where planning was led by First Nations interest groups and processes were built into flag or proactively manage content with respectful practices. The Local Contexts Traditional Knowledge Labels and Licences are an example of tools being used to support ICIP rights and the management of metadata about Indigenous attribution, protocols for access, and to alert people about culturally sensitive materials (n.d.). This additional layering of metadata using labels could assist Wikimedia platforms in creating respectful processes, for example, in editorial practices to ensure that historical systems of privilege are not perpetuated on Wikipedia today. The principle of do no harm requires that mechanisms are created for participation and a right of reply to information sources.

Enhancing support for First Nations Wikipedia editors and communities of interest should be a priority to create relevant governance structures and seek guidance on how Indigenous content, particularly First Nations metadata, is created and managed. A priority in all this work is to increase dialogue and discussion about how Wikimedia can build relationships and trust to support First Nations priorities. This means identifying the major pain points and sites of risk, for example, understanding description and classification in relation to power, within the Wikimedia ecosystem and in a wider political context.

Strategies to ensure first nations perspectives

The research project identified that more questions are needed to be asked to create the necessary strategies to ensure First Nations perspectives. A key point that was raised was that these strategies to increase Indigenous representation were not just a matter of “filling the gaps”. In an article on Wikipedia and gender gaps, Ford and Wajcman discussed the disproportionate rates of engagement of women with Wikipedia and the structural issues relating to Wikipedia’s infrastructure that contributes to this disparity. The authors argue:

… that Wikipedia's origins and the infrastructures on which it relies are based on foundational epistemologies that exclude women, in addition to other groups of knowers whose knowledge does not accord with the standards and models established through this infrastructure. (2017)

Ford, in response to the Wikimedia Movement’s 2030 strategic direction, also authored a blog post on a critique of the proposals identified to support a “knowledge gap index” (2021). Within this discussion, Ford called attention to the lack of discussion on issues of power, stating,

Framing Wikipedia’s problems as gaps that need to be filled is a mistake because it doesn’t enable us to see how Wikipedia is a system governed by unequal power dynamics that determine who is able to be an effective contributor. (Ford, 2021)

Strategies to ensure First Nations perspectives also need to consider structural and systemic issues related to power imbalances. Internationally other First Nations communities have warned about the challenges of Indigenous knowledges being brought into Wikipedia. For example, Van der Velden (2013) challenges Wikipedia to de-center design in the context of managing Indigenous knowledges. The author points to the need for information systems to incorporate Indigenous protocols for knowledge management. Similarly, Mushiba and colleagues have written about processes of engaging the OvaHerero of eastern-central Namibia and the need for the adaptation of Wikipedia to become an appropriate technology for Indigenous knowledge holders (Mushiba et al., 2016).

In an Australian context, these protocols need further consideration across Wiki platforms, particularly to ensure that consent is considered in the context of ICIP rights. Janke (2021) details how a Wikipedia article on the Aboriginal language palawa kani was published without the consent of the palawa community. In wanting to ensure they could project their language and prevent it from being misused, The Tasmanian Aboriginal Centre (TAC) requested to have the page removed which Wikimedia Foundation refused to as languages cannot be copyrighted under Australian or international copyright laws (Janke, 2021). This breaches Indigenous peoples right to protect, control and maintain their knowledges as stated in Article 31 of the UNDRIP.

While never leaving the incubator state and currently inactive, the Noongarpedia project demonstrates the opportunities and exemplifies a Wikipedia project that ensures First Nations perspectives. While active, it included First Nations leadership, governance and self-determination as key project drivers. Starting as a joint Australian Research Council project through the University of Western Australia and Curtin University, the project also included a partnership with Wikimedia Australia. Reflecting on the lessons learned from the project, Buchanan et al. (2016) have described the advantages of using Wikipedia to support a “Noongar knowledge network” that enables both the management of knowledge and a means for sharing it.

Wikipedia may be seen to provide not only a technological means to record ‘knowledge artefacts’ of the past, but also a social means to activate Noongar and others in the performance of that culture as future-facing civic action. (Buchanan et al., 2016)

The questions below were asked to create critical conversations about how descriptive practices and knowledge sharing of Wikipedia could be shaped by respect, reciprocity and relationality.

These questions can also serve as a prompt to pause to reflect before engaging in any new Wiki projects related to First Nations peoples and knowledges. Understanding the question of “Who are the editors”? provides an opportunity for editors to reflect on their own positionality, knowledge and bias.

Wikimedia Australia partnerships

The work described in this discussion paper requires partnerships between GLAM institutions across state-based and regional areas and with First Nations communities. Participatory approaches, based on the earlier described principles of the AIATSIS Code of (1) Indigenous self-determination, (2) Indigenous leadership, (3) Impact and value, and (4) Sustainability and accountability, could guide the work of Wikimedia Australia in building relationships with First Nations communities.

In the context of metadata and First Nations Description, relationships with peak bodies such as AIATSIS are vital in understanding how emerging thesauri, guides and tools can be disseminated for use within the WMAU community. The research has highlighted the importance of tools, workflows and metadata being informed through partnerships with communities, including, for example, peak representative bodies and community organisations. Future projects for WMAU might include considering approaches where expressions of interest are made for areas such as: seeking input from First Nations communities and representatives for prioritising Wikimedia pages; identifying workflows and governance structures for material that contains ICIP or offensive materials to be flagged; identifying approaches for decolonising Wiki projects including education and awareness raising; planning a programme of work to support Indigenous self-determination across Wiki platforms as a commitment to the principles of mutual-benefit so that WMAU projects are facilitative and not extractive. The example of the partnership for the Noongarpedia project highlights the potential for collaboration across community, industry and university partnerships. By integrating these established protocols and fostering meaningful partnerships, Wikimedia Australia can ensure that its initiatives are both respectful and representative of Aboriginal and Torres Strait Islander peoples.

Reflections and recommendations

The following reflections were made as part of the research project as recommendations for Wikimedia Australia to consider:

• Develop specific guidelines and tools to help assist Wikipedia editors in assessing appropriate resources, prioritise resources developed by First Nations peoples, and improve the appropriateness of articles content and descriptions relating to First Nations topics.

• Develop strategies to respond to the prevalence of settler narratives and seek to rebalance and inject recognition of, and representation of First Nations sovereignty.

• Build a greater understanding of the harm of description and classification on the wellbeing of First Nations people and ensure that the site is aware of what information can be a trauma trigger.

• Review and engage with established toolkits and guidelines for decolonial and anti-racist descriptive practices. Prioritise engaging with tools and metadata initiatives that are relevant to working with Aboriginal and Torres Strait Islander communities, cultural heritage and data, for example, draw from the leadership of AIATSIS.

• Consider approaching future projects collaboratively by calling out for “expressions of interest” to enable engagement with First Nations communities and representatives.

• Build awareness of how First Nations people can become Wikipedia editors through targeted efforts and support communities of interest of First Nations editors.

• Consider how Indigenous research principles, and tools, such as the Traditional Knowledge Labels, might be utilised in Wikipedia contexts to support ICIP rights through metadata.

Conclusion

This paper discussed an exploratory project that sought to learn more about First Nations Metadata in the context of the ATSILIRN Protocols and their use on Wikimedia projects. Several themes were identified as areas of priority for Wikimedia to investigate, and recommendations were made as reflections for future action. In addition, the paper identified the need for further discussion on the existing tensions regarding the inclusion of Indigenous knowledges in open-access platforms and the potential for ICIP rights needing to be supported. The theoretical and philosophical considerations underpinning this article emphasise the importance of Indigenous self-determination, and the need for decolonial and anti-racist practices. By reflecting on questions raised in this paper, editors can perhaps better understand their positionality, knowledge, and biases before engaging in new Wiki projects related to First Nations peoples and knowledges. The protocols developed in GLAM contexts, and the emerging Indigenous Data Sovereignty movement provide pathways to develop respectful practices across Wikimedia platforms.

Acknowledgements

This research project was enabled by funding from WMUA in collaboration with the Jumbunna Institute for Indigenous Education & Research, University of Technology Sydney. The article draws significantly from the Wikimedia Australia and First Nations Metadata: Discussion Paper (2023) developed by the authors with input and guidance from WMAU. The authors of this report acknowledge the Traditional Owners on whose Country this report was written. We acknowledge their continuing knowledge authority and pay our respects to Elders, and leaders, past and present.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Wikimedia Australia.