A digital language archive (DLA) is an archive that includes born-digital and/or digitized language resources in any language(s) or dialect(s), a searchable repository in which resource discovery is powered by metadata. The collections comprising DLAs are often created with the goal of preserving, or documenting, a language, especially an endangered language. Once documented, these materials assist the revitalization efforts of the language(s) in question (Hinton, 2001). Thus, items selected for these collections demonstrate language use in various settings and through different media and file formats (Chelliah, 2021). Some of the common item types in DLAs are also collected by broader-scope digital repositories stewarded by cultural heritage institutions, and include audio-visual recordings of oral histories, performances, interviews, text resources: books, articles, letters, diaries, etc. However, DLA also include a significant proportion of resource types that are not found in other digital repositories: wordlists, field notes, AI-generated and Large Language Models related content, corpus publications essential for computational linguistics research; linguistic annotations, transcriptions, and software (Bird & Simons, 2003).

In the past, language communities, despite often being the source of language archive materials, were typically not involved in the creation or maintenance of archives (Henke & Berez-Kroeker, 2016). This practice privileged linguistics and anthropology researchers, missionaries, and colonial administrators as creators of archival records (i.e., items and collections), ascribing provenance to them over the individuals and communities depicted in the archival records (Jones, 2019). However, like other archives, DLAs undergo significant changes and evolve to meet the requirements of the archival paradigm shift toward community-oriented ways of archival thinking and participatory methodologies (Cook, 2013; Genovese, 2016; Rolan, 2017). For DLAs, that means that source language communities of the archival records take an increasingly active part in decision-making regarding donation, appraisal, preservation, metadata, access, dissemination, etc. (Burke, 2021; Dale et al., 2022; Harris et al., 2019; Olson, 2023; Pugh & Christen, 2011; Roeschley & Kim, 2019).

Information science and linguistics researchers explore the archive user needs of language community representatives and scholars. Developing understanding of these needs is important for improving access as they guide user interactions with and expectations towards DLA. Some of the user needs tend to focus on the services offered by the DLA and include stream-able audio and video, user interface accessibility on mobile devices, as well as machine-readable text files and bulk download availability necessary for computational linguistics research (e.g., Burke et al., 2022). Endangered language community members also report needs – related to supporting the revitalization of their languages – that inform collection development decisions as language communities representatives call for inclusion of specific kinds of resources beyond the obvious choice of dictionaries, audio and video files: textbooks and teaching aids covering different disciplines, folktales and storybooks for children, resources on religion and culture, and overall attractive items that capture attention (Burke, 2023). Research demonstrates that information needs of DLA users are often not met, and that one of the key reasons for this is the lack of academic instruction for information professionals on supporting the stewardship and digital curation of these collections (Al Smadi et al., 2016; Wasson, Holton, & Ross, 2016; Zavalina & Chelliah, 2021).

While there is an overlap in the principles and practices of broader digital archiving, digital archiving of linguistics research data, and community digital archiving focusing on language, there are substantial differences in the goals, scale, content, challenges, ethical considerations, and impact. One example is that types of items unique to language archives and specific attributes of these resources, as well as complex relations between these items must be represented differently in bibliographic metadata, archive’s navigation, and controlled vocabularies to meet DLA user information needs (e.g., Aljalahmah & Zavalina, 2023; Paterson, 2024; Zavalin, 2023). Greater focus is needed on authorship and intellectual rights (e.g., with AI-generated data), as well as on ownership, archiving spaces, and mutually beneficial relations between communities and partnering researchers and institutions, identifying, and disseminating best practices, and developing collaborations to support community archiving and to engage communities in stewardship of their DLA collections housed by LAMs (Zavalina & Chelliah, 2021; Chelliah, 2023).

University of North Texas (UNT) educators experimented with adding DLA training for LAM students as an individual DLA-focused module in the advanced metadata course. Preliminary results of DLA-focused learning module effectiveness indicated that during 2021-2023 testing, 75% to 90% students attained the target cumulative score of 85% or more on module assignments and activities (Zavalina, 2023). Student survey data on teaching effectiveness however were collected and analysed only at the course-level, not for the DLA-focused module. Realization of the need for a more robust in-depth and open-access DLA training for LAM students led to successful proposal to the US Institute or Museum and Library Services (IMLS) for addressing this need.

To meet this demand, our project’s interdisciplinary team develops curriculum with a strong experiential component to educate information professionals in the archiving and curation of resources that provide the means to revitalize community memory and language. The project is producing learning materials intended to be useable in courses offered in LIS and Archival university programs, LAM on-the-job trainings, and archiving workshops for members of indigenous, immigrant & refugee communities who are documenting their community heritage.

This project, implemented by the interdisciplinary team of information scientists and linguists, also has the goal of collaborating directly with language communities to offer training on language archiving and obtain community feedback. The training consists of short workshops in which participants learn about the history, purpose, principles, and practicalities of digital archiving, and take part in hands-on activities designed to familiarize them with digital archives and provide experience with the process of adding an item to a DLA. While community workshops are still in progress, the planning of these collaborations in and of itself has introduced multiple insights that are relevant for future researchers prioritizing community-oriented participatory methods. While the needs of a language community differ in each community, we identify four areas that should be considered in the development of workshop materials for language communities: digital literacy, inter-generational collaboration, technology availability, and language accessibility.

Multiple factors such as socioeconomic status, proximity to urban areas, and community perception of technology, affect the acquisition of digital literacy – skills required to use digital technologies – in communities. Language documentation projects have noted the lack of digital literacy as a barrier for community members to participate in digital resource creation (Frawley, Larkin, & Smith, 2017; Hausknecht et al., 2021; Mwanza, Molepo, & Goduka, 2018). This lack severely impacts the ability of a language community member to contribute to the creation of material in their language through digital archiving, which is detrimental to the language preservation (Domeji et al., 2019). When contributing to a DLA (by submitting an item or a collection), a depositor needs a level of digital literacy that allows performing the associated actions, such as uploading content from a separate device to a computer, file management, creating metadata, and filling out intake information on a web page.

In scenarios where community members lack sufficient digital literacy, DLA community workshops need to cover foundational digital literacy. However, when the digital literacy gap is prominent among older community members, but less pronounced in younger generations, inter- generational collaboration can alleviate the necessity of digital literacy training while simultaneously strengthening relationships between community members. One workshop in our project’s series is designed for members of a language community that are university students. While the students have varying levels of proficiency in the language, partnering with older speakers in the community is mutually beneficial as they can support the older speakers navigate the technology used to contribute to archives and learn more about their language and culture from working with the older speaker.

The technology availability affects what happens both during and after the DLA community workshop. The workshops include hands-on activities, such as exploring a digital archive and preparing a file for deposit. To facilitate these activities, the participants must use a computer, that many of them don’t have, so it is important to schedule the community workshop in a computer lab or bring laptops for the participants to use. In the non-university settings, access to a computer lab is fairly simple, but in other settings, this can present a substantial coordination challenge. It is also essential to note the technology the participant will be using after the workshop, including but not limited to the general access to computational technology as other factors can cause confusion: for example, differences between the operating systems or text / audio-visual editing software available for participants after the workshop versus what community participants used in training.

Building on ideas of general accessibility, the language used to present DLA to language community members is influential in facilitating understanding and garnering interest in digital archiving. For this reason, the way the content is presented to information science students or linguistics students must differ from the way it is presented to language community members. While the content remains largely unchanged, field-specific terminology should be modified to be intelligible to people without information science or linguistics background. For example, mentioning that the ability to query the metadata associated with items is one of the major advantages of DLA is not a persuasive point when the person has not been exposed to the concepts of metadata and querying. Similarly, expressing the importance of file naming conventions and organization of files must be substantiated in its practical utility: saving time and not losing things. The language used to discuss DLA must be accessible, both in the terminology that it employs and in the practical grounding of concepts.

The first offering of the project-developed LAM graduate course demonstrates overall success in curriculum development, as well as substantial demand among current students for coursework that is designed to develop the knowledge and skills necessary for successful stewardship of DLA, including community archives. Now that the course is officially added to UNT graduate catalogue as INFO5385, and students who completed it are sharing their experiences with fellow students, the awareness grows among our LAM student population. With 14 additional IMLS-sponsored

tuition stipends for the course, another class of 25 is expected to take INFO 5385 after revisions based on results of student evaluations completed so far and on the upcoming results of language community evaluation of project-developed learning materials.

For the longer (10-week) semester of 2025, we will add to INFO5385 the content covering two course-level objectives that did not reach the 70% student confidence target, break four instructor presentations into 2-3 shorter videos per module, add two live online meetings to the asynchronous course. The changes also include augmenting the data collection by adding the pre- semester survey for course-level objectives. The student feedback will be comparatively analysed and used in further refinements of learning materials.

Our team is also working on integrating relevant content from project-developed course modules into other UNT LAM courses and Indiana University Bloomington linguistics courses. We are disseminating preliminary results to obtain feedback from LAM and Linguistics educators and practitioners through presenting at professional meetings (Coronado & Zavalina, 2024; Zavalina & Paterson, 2024, etc.) and raise awareness of the project’s products and opportunities it provides among the library and archival communities of practice, as well as among LAM faculty and students. Once the project is completed, project-developed learning materials will be made available as open access resources and are expected to make a solid contribution to addressing the needs of language communities through the extension of LAM professionals’ preparation to stewarding DLAs.