November 2024
Alexandrina Agloro
School for the Future of Innovation in Society
Arizona State University
Abstract: The concepts of digital trust and safety are embedded with Western cultural assumptions that often do not universally translate to the majority world. Digital trust and safety frameworks are built upon the nation-state, where Indigenous people have seen their sovereignty systematically denied. The rise of the algorithmic state is an opportunity to think through how decolonization is not a metaphor[1] and how digital frameworks can offer an opportunity to re-think the rematriation[2] of land, culture, and life. This paper calls for an examination of how trust and safety are culturally specific phenomena. Using the frameworks of Aotearoa/ New Zealand’s data protection and data governance models, I propose examining Indigenous Pasifika practices of traditional midwifery and batok (pre-Philippine tattooing) as ways to understand how culturally specific approaches can build trust and safety online.
We in the Global North take for granted how our digital systems came to be, and the cultural values inherent in our systems. We default to the Western world for standards of digital engagement, but rather than try to fit a Western standard of trust and safety, what could happen if we had to pass through culturally specific nuances that stemmed from regional Indigenous cultures? A working example of this concept is in Aotearoa/ New Zealand where digital infrastructure follows Te Tiriti o Waitangi /Treaty of Waitangi, the agreement from 1840 that promises Māori people and their interests will be honored. At the national level, Aotearoa/ New Zealand operates that data have mana (energy) and whakapapa (genealogy), and much like how the Māori meet mana to mana, data meet mana to mana as well. Māori prayers are coded into national data to give data mana. Sacredness is part of the understanding around data, trust, and safety. These culturally specific expectations for understanding data are the same when foreigners outside Aotearoa/ New Zealand engage with national digital infrastructure.
Why Pasifika midwifery and ancestral marking? These Indigenous practices were designed precisely for trust and safety. Indigenous midwifery is built around trusting the capacity of the human body, and if we are to learn about trust, practices of body autonomy relate to our interactions online. Pasifika tattooing is a visual language and a means of reading each other. In a hardware and software sense, systems have to establish "trust" (the ways computers connect to each other, and software systems connect) but the ways trust is established has been coded by white male westerners whose notions of trust and safety are contextually specific. I'm advocating for localizing digital trust and safety and using local cultural contexts- so the ways we read tattoos as a way of understanding each other- and visual languages can be coded into our systems. A system for Oceania wouldn't be the same for Latin America; but we should be imagining our capabilities as expansive and not boiling it down to the lowest common denominator of the protocols that white western men think of as safety and trustworthiness in our digital systems.
[1] Tuck, E. & Yang, W. 2012. “Decolonization is not a Metaphor.”
[2] “Rematriation is Indigenous women-led work to restore sacred relationships between Indigenous people and our ancestral land, honoring our matrilineal societies, and in opposition of patriarchal violence and dynamics.”- https://sogoreate-landtrust.org/what-is-rematriation/
November 2024
Sandra Céspedes
Computer Science & Software Engineering
Concordia University
sandra.cespedes@concordia.ca
Rob McMahon
Media & Technology Studies and Political Science
University of Alberta
rob.mcmahon@ualberta.ca
Abdelhak Bentaleb
Computer Science & Software Engineering
Concordia University
abdelhak.bentaleb@concordia.ca
Lope Trujillo
Not2Far Technologies
lope.trujillo@lalomaprojects.ca
Claudia Lezama
La Loma Projects
Claudia.lezama@lalomaprojects.ca
Phil Steinhauer-Mozejko
Indigenous Connectivity Institute
Phil@clearskyconnections.ca
Tim Whiteduck
First Mile Connectivity Consortium
First Nations Education Council
TWhiteduck@cepn-fnec.com
Abstract: Our team’s proposal is organized around the early stages of a project involving Indigenous communities in Canada and Colombia. We are excited to hopefully join the symposium to present and discuss these ideas with other participants.
Bridging the digital divide is recognized as crucial for equitable access to the Internet, public services, and economic opportunities, as well as to support Indigenous sovereignty and self determination. However, efforts to improve digital access in rural and Indigenous communities must go beyond a one-dimensional supply-side connectivity solution and directly engage communities in all stages of technical design, deployment, and operations.
Taking a sociotechnical approach grounded in principles of design justice, this project investigates how technological infrastructures intersect with the unique cultural and socioeconomic realities of Indigenous populations. Working directly with underserved Indigenous communities in rural/remote regions of North and South America, we will co-create a comprehensive strategy for digital equity and transformation. Led by Indigenous organizations and facilitated by academic, industrial, and governmental partners, this project explores a multidimensional approach to shape emerging digital technologies to meet the needs and desires of Indigenous communities.
This work is guided by several goals:
Goal 1: Understanding. Previous studies have highlighted the shortcomings of existing strategies in assessing the digital environment and ecosystem of rural/remote Indigenous communities. These gaps include assumptions of adequate access to digital technologies, a lack of specific cultural context and place-based analysis, and low engagement with community members. In this project, Indigenous communities will lead a locally-situated, comprehensive assessment of their communities' digital adoption needs and desires. This sets the foundation to guide the co-design of appropriate digital solutions.
Goal 2: Connecting. Despite significant public investments, rural/remote and Indigenous communities in Canada and abroad continue to face persistent barriers in accessing digital services. This reflects limited progress in reducing digital inequalities through current approaches to Internet connectivity in remote and isolated areas. More challenges arise from the extraction of Indigenous data/knowledge, the dependency on expensive service fees from external service providers, and the lack of user experience information available to policymakers and regulators. This project leverages recent advances in secondary spectrum access and updated regulations in wireless communications to co-design and deploy a cost-effective connectivity solution that includes automated sensors (Internet of Things). Community members will be involved in all aspects of the design and deployment of this solution, ensuring it aligns with their goals and desires. They also gain skills and expertise in operating and maintaining the resulting infrastructure and services.
Goal 3: Interacting. A persistent gap exists between connectivity projects and the cultural contexts of Indigenous communities, often stemming from inadequate consultation, limited community engagement, and a lack of culturally appropriate digital tools. To address this, we will co-design culturally specific digital intervention tools and data collection processes that align with Indigenous ways of knowing and interacting. These tools, such as a conversation agent and customized visual interfaces, will facilitate access to digital data and resources in an engaging and culturally relevant manner. The tools will follow the OCAP™ principles to guarantee First Nations control over data. We anticipate this approach will promote digital inclusion while supporting community goals including sustainable development, the preservation of Indigenous languages and cultural heritage, and environmental stewardship.
Goal 4: Serving. While affordable internet is crucial, meaningful access requires a holistic approach that considers connectivity, quality, and a supportive social environment empowering users to leverage the internet effectively. Guided by the expertise of our Indigenous team members and the UN's Meaningful Connectivity framework, the project will develop and test software tools and methodologies that will enable sustained community engagement and participation in managing digital resources. The aim is to empower communities to control their connectivity and services while aligning them with their cultural values and aspirations.
November 2024
Sean Dorr
Computer Science & Engineering
University of Minnesota
dorr0024@umn.edu
Daniel F. Keefe
Computer Science & Engineering
University of Minnesota, USA
dfk@umn.edu
Abstract: What is an Indigenous iPhone? Is it created with an aesthetic and features that resonate with your Indigenous community's values? Can it withstand the environment and rigor of Indigenous activities? Is it built from an Indigenous ethical center that is timeless, binding users to ancestors, present-day people, and future ancestors? This line of questioning is essential because Indigenous Computing Futures are not about broadening participation in the current paradigm of computing or teaching communities to better use today's computers and data; the future of Indigenous Computing lies in extending Indigenous protocols, technologies, and processes to computing. Indigenous peoples already invent technologies, already have creative practices, and already have protocols to protect their "data"; these do not exist in computer-mediated environments because Indigenous peoples are not the ones authoring computer-mediated environments. While not every Indigenous person needs to be a computer programmer, our position is that Indigenous Computing Futures critically depend on authoring, not just using, computing technologies.
We use the term authoring to refer to Indigenous communities imagining and creating computer-mediated experiences on their own terms. Authoring certainly includes the traditional methods of creating computing tools (e.g., programming, software engineering, user-experience design); however, other creative approaches may also be possible. Recently, we have partnered with two Indigenous communities (Micronesian from Chuuk State and Dakota from Lower and Upper Sioux communities of western Mnisota) to understand how computing technologies might support their goals of knowledge revitalization and environmental sustainability. Barriers encountered with typical computing interfaces (e.g., mouse, keyboard, data charts) led us to explore other styles of computing with tangible user interfaces and first-person, experiential visualizations in virtual and mixed reality, planetariums, and similar lightweight form factors (e.g., projectors, Google Cardboard). We are finding that these alternative computing environments can be quite a good fit–they meet Indigenous knowledges in their natural N-dimensional and spatial form, and there is potential for elders to leverage the power of computers and data while also teaching as they always have, through experiential learning and storytelling. Unfortunately, the (semi-)immersive computing environments can also be some of the most challenging to author, often requiring graduate-level computer graphics programming skill. To help overcome this, we have pivoted our work to creating tools that can extend traditional Indigenous making and teaching practices to computer-mediated environments. One early example is visualizing data and telling stories in an immersive planetarium by inviting community members to author the ecological data and visual encodings by hand (think paper, clay, beadwork) and then translate these into the computer environment with technologies like cameras and green-screen animation stages.
This position paper posits, first, that authorship is paramount. Then, we describe a technical research agenda to support Indigenous computing authoring in many forms. We suggest this could have great impact in enabling Indigenous communities to see themselves, quite possibly for the first time, in the visual and interactive languages of computing technologies. In this way, we position Indigenous Authoring Tools as a means of decolonizing thought and indigenizing computing.
November 2024
Lisa G. Dirks
University of Washington Information School
lgdirks@uw.edu
Abstract: Emerging technologies such as social media, podcasts, interactive websites, and mobile applications offer pathways to broaden research dissemination beyond traditional academic audiences. These tools hold potential for reaching Indigenous communities, yet few studies have examined their effectiveness in engaging Native audiences with culturally and contextually relevant content. Insufficient empirical evidence exists about the values that Native community stakeholders place on technology-facilitated collaborative dissemination, or what they would want these tools to include. Preliminary studies suggest a divergence between Native and health science ways of knowing, indicating a need for further dialogue to encourage multi-directional learning and enhance partnerships. This could lead to improved communication, trust in research, and increased participation in health research areas where Native representation is currently limited. This paper considers how community-centered co-design using community-engaged research methods could amplify partnership, equity, and dialogue in designing technology tools for collaborative health research results dissemination in Native community contexts.
Conducting research in Indigenous contexts necessitates a community-centered approach that upholds cultural sovereignty, acknowledges diverse knowledge systems, and respects the uniqueness of each Native community. With 578 federally recognized tribes in the United States, as well as non-recognized communities, each with unique protocols, a “one-size-fits-all” approach is unsuitable. Community-centered design must be flexible and responsive to the specific needs, values, and priorities of each Indigenous community, especially when urban environments include individuals from multiple tribes.
Community oversight is integral to Indigenous-centered health research, often managed by tribal health services and community ethical review boards. These boards ensure research aligns with local cultural priorities and may provide guidance at each stage of a project. However, researchers may encounter conflicts between the guidance of these oversight bodies and the input of community members involved in participatory design or co-design research. Transparent communication and respect are essential to reconcile these differences and maintain an ethical and culturally sensitive research process.
Many Indigenous communities are in rural or remote settings. These locations pose logistical challenges for engagement, particularly regarding limited access to technology and high travel costs to reach isolated communities. Addressing these barriers requires researchers to consider the feasibility of providing necessary technology, as well as sustainable training initiatives tailored to the needs of rural settings. Safeguarding Indigenous intellectual property is another critical consideration. Many tribal communities mandate that data gathered within their communities remains under their control. Researchers must navigate data-sharing agreements that honor Indigenous ownership and protect traditional knowledge, underscoring the need for trust and frameworks that respect Indigenous intellectual property rights. Despite these challenges, technology-based methods offer promising tools for improving research dissemination in Indigenous contexts. “Dissemination 2.0” approaches, which leverage social media and other internet-based platforms, promote multidirectional communication aligned with Indigenous preferences for collaborative knowledge-sharing. However, researchers must assess each community’s readiness and technological infrastructure to ensure these tools are effective and culturally appropriate.
Framing productive and meaningful community-centered design in Indigenous contexts requires flexibility, cultural sensitivity, and a commitment to shared ownership of outcomes. Long-term relationship building, cultural competence, and inclusive participation are essential. By building partnerships based on mutual respect, providing training and resources, and adapting dissemination strategies to community needs, researchers can foster trust, uphold Indigenous knowledge sovereignty, and support health equity in Native communities.
December 2024
Marisa Elena Duarte
School of Social Transformation
Arizona State University
Abstract: Indigenous technologists need to collectively develop a critical discourse around the phenomena of information. As a modern scientific construct, the concept and science of information significantly influenced Cold War technologies associated with signal processing, signals intelligence and human intelligence, data-transmission, and media manipulation. In later years, information became the means for formulating ranges of mathematical complexity across stores of metadata, data, and knowledge, an innovation in systems thinking contributing to advances in automation, cybernetics, and machine learning. During the decolonial movements of the 1970s to the 1990s, experts in traditional plant knowledges of Indigenous peoples developed the concept of Indigenous knowledge as part of an effort toward global Indigenous human rights recognitions. Meanwhile, the technologically advanced countries of the world eschewed UN and WIPO calls for Indigenous rights and Indigenous knowledge protections in favor of data-centric economies. Some of these countries—the US, the UK, Canada, Israel among others—developed a sophisticated ecology of intelligence and surveillance services alongside an ecosystem of data-driven industrialists.
For Indigenous peoples, interfacing with technologies such as Internet of Things, social networking sites, DIY robotics, affordable or no-cost enterprise tech, and high speed low-latency Internet to the home has also allowed for the increasing emergence of a global class of Indigenous scientists, social scientists, scholars, professionals, and community advocates and leaders. This class of tech-savvy professionals greatly rely on the technoscience of data and information—including associated infrastructures and systems--for protecting tribal sovereignty and Indigenous rights, supporting Indigenous creativity and technical innovation, and reclaiming Indigenous ways of knowing.
Scientists have meanwhile identified the planetary effect of global technicization as the Anthropocene: a geologic era in which the rhythms of climate, precipitation, and biological diversity are affected by human industrialization more than any other force. In the Anthropocene, Indigenous thinkers’ calls for Land-based knowledges, human rights, and technological regulation are all the more striking, in particular when they gesture toward popular discourses of decolonization, autonomy, and sovereignty. This is due to the inevitable tensions of these sometimes ontologically competing discourses. Increasingly, Indigenous social scientists and humanists are identifying in their investigations the entanglements of such infrastructures as dams, pipelines, and border surveillance assemblages in terms of their effect on ecosystems and human rights. Practitioners who specialize in services to tribal government around matters such as cybersecurity, intellectual property, telecommunications implementations, and economic development are meanwhile identifying the extraordinary need to consistently infuse technical trainings and supplies into Indigenous communities and nations. There is a constant pressure to keep up and protect one’s people amid the ever-advancing ecology of devices and systems that characterize neoliberal profit and governance in the Anthropocene.
Meanwhile, from a scholarly and scientific perspective, the sheer diversity of technical sub-fields and functional terminologies multiplies the scholarly labor of respectfully discerning schools of thought and ethics, especially with due regard for the pluriverse of Indigenous peoples, geopolitics, and philosophies. For this latter reason, striving for a shared and collective discourse around Indigenous information--one that is robust enough to withstand debate, technical specificity, and policy disagreement—is key. Perhaps through emphasizing our strengths across our modes of innovation—regard for Indigenous creativity, the beauty of our knowledges, and the persistence of our justice and rights-based movements—we can arrive at the grounding ethics for this collective discourse.
November 2024
Shristi Gupta
Information Systems and Digital Forensics
University of Albany, SUNY
sgupta4@albany.edu
Chun-Hua Tsai
Information Science and Technology
University of Nebraska, Omaha
chunhuatsai@unomaha.edu
Abstract: The advent of large language models (LLMs) and advancements in artificial intelligence (AI) have introduced powerful tools with the potential to preserve and amplify indigenous knowledge, cultural heritage, and sovereignty. Indigenous communities hold unique worldviews, knowledge systems, and social structures, which contrast significantly with those of dominant cultures and contribute valuable insights that can inform community-centered AI development. However, the current design of LLM systems faces challenges in reflecting this diversity, largely due to a lack of culturally representative datasets and ethical complexities surrounding data collection. Consequently, the development of culturally sensitive AI systems is hindered by inadequate resources, concerns about data ownership, and risks of privacy violations, all of which affect the transparency, inclusivity, and reliability of these technologies.
This study proposes a novel approach using Indigouse-generated data as a solution for addressing these challenges and promoting data sovereignty for indigenous communities. Indigouse-generated data involves community members actively participating in data collection, generation, and analysis, ensuring that the resulting AI applications genuinely serve the communities from which the data originates. This model of data generation aligns well with the democratizing principles of citizen science, wherein community members contribute their knowledge, stories, and experiences to enhance the accuracy and cultural relevance of AI systems. By fostering community involvement, this approach not only respects indigenous sovereignty but also incorporates the values of participatory design, making these communities equal stakeholders in the creation and training of LLMs.
Indigenous knowledge has traditionally been transmitted orally, with cultural and historical narratives shared through storytelling by tribal elders. This practice has led to a scarcity of written records, creating significant gaps in authentic indigenous datasets. Involving indigenous communities in the data generation and annotation processes for LLMs addresses this issue by capturing rich, authentic oral traditions in a respectful and accurate manner. This inclusive data generation method enhances AI transparency and reliability, providing LLMs with culturally meaningful content that reduces algorithmic bias and improves their responsiveness to diverse worldviews.
Moreover, this approach promotes data sovereignty—a core principle for indigenous communities—by ensuring that the data used in AI systems is community-owned, with transparency about its use. As indigenous participants engage in co-designing language models, they not only contribute to preserving their culture but also gain insights into the mechanics of AI systems, fostering informal learning and narrowing the digital divide. This participatory process empowers communities to have greater agency over technology development and builds capacity within communities to manage future AI applications independently.
Hence, Indigouse-generated data for LLMs offers a path forward for culturally sensitive AI that respects indigenous knowledge systems and upholds the democratic, sovereignty-based values of indigenous communities. Through this approach, AI technologies can become more inclusive, transparent, and aligned with the communities they aim to serve, offering a robust model for integrating indigenous perspectives into the broader AI landscape.
July 2024
Shelby Hagemann
School of Informatics, Computing, & Cyber Systems
Northern Arizona University
seh428@nau.edu
Morgan Vigil-Hayes
School of Informatics, Computing, & Cyber Systems
Northern Arizona University
Abstract Despite high incidence of depression, anxiety, and post traumatic stress disorder, stigma and lack of access to culturally responsive behavioral health care resources prevents many Native Americans (NA) from seeking care. However, the rise of culturally-responsive in-person and digital behavioral health resources for NA communities provides new opportunities to address these longstanding health equity issues. The major challenge is helping people in NA communities find these meaningful resources and helping anchor institutions understand how resources are being sought and utilized to support more responsive internal programming. In this context, we have partnered with Hopi Behavioral Health Services (HBHS) to design the Resilience Resource Database to digitally disseminate mental and behavioral health resources. This paper presents initial findings that have resulted from the initial stage of an iterative participatory design process with HBHS.
Note: This is from a paper published as late breaking work at ACM DIS 2024.
November 2024
Tochukwu Ikwunne
School of Informatics, Computing, and Cyber Systems
Northern Arizona University
Jared Duval
School of Informatics, Computing, and Cyber Systems
Northern Arizona University
Jared.Duval@nau.edu
Abstract: Hypertension affects 35.9% of the Navajo population, posing a significant risk for cardiovascular disease, the leading cause of non-accidental death in this community. The root causes of hypertension in the Navajo Nation are multifaceted, encompassing socioeconomic, cultural, and environmental factors. Barriers such as limited health education, lack of healthcare resources in Diné bizaad (Navajo language), rural living challenges, and restricted access to nutritious food and medical care further compound the issue. These obstacles lead to low treatment adherence, with non-adherence rates ranging from 30% to 70%, resulting in elevated rates of clinical inertia. To address these challenges, this project proposes the co-design, development, and evaluation of CardioCare Quest, a culturally adapted telehealth game aimed at supporting Navajo families in managing hypertension. The game provides localized content in Diné bizaad, integrating critical lifestyle factors such as diet, physical activity, medication adherence, and regular blood pressure monitoring. It offers a mobile platform that is accessible offline, thereby addressing the healthcare access gaps experienced by those in rural, underserved regions of the Navajo Nation. CardioCare Quest is structured around modular minigames, each targeting specific behavioral aspects that influence hypertension. These include healthy eating, exercise, and medication routines, while also promoting traditional Navajo activities like farming and dancing. Through a community-based participatory design approach, the minigames will be co-created with Navajo participants, ensuring the content aligns with local cultural practices and knowledge systems. This co-design process, grounded in Research through Design methodologies, will foster stronger community engagement and ensure the intervention resonates with the target population, promoting long-term behavioral changes.
One of the key features of CardioCare Quest is its emphasis on personalized gameplay. By offering multiple minigames tailored to individual preferences, the game allows users to engage in activities that align with their personal interests and treatment goals. This flexibility encourages sustained engagement and allows for the creation of individualized treatment plans.
The project also includes the development of a clinician dashboard, enabling healthcare providers to track patient progress and adjust treatment plans based on real-time, in game data. This HIPAA-compliant platform will facilitate communication between patients and healthcare teams, bridging the gap between clinical care and home-based health management. By providing actionable insights into treatment adherence and health behaviors, the dashboard enhances the overall effectiveness of hypertension management strategies.
The project also includes a cultural probe initiative designed to deepen our understanding of participants' daily activities and routines. This initiative captures pre and post-intervention measures of Motivation, Engagement, and Thriving in User Experience (METUX) related to managing high blood pressure. This information helps tailor the game to support participants' needs, ensuring the intervention is both culturally relevant and personally meaningful.
CardioCare Quest seeks to address the urgent need for culturally relevant interventions in the Navajo Nation, where conventional healthcare approaches often fail to meet the needs of Indigenous communities. By integrating Navajo cultural practices into a digital telehealth platform, this project offers a novel solution for managing hypertension. It emphasizes intergenerational engagement through cooperative gameplay, celebrating everyday practices that promote healthier lifestyles. Ultimately, this project aims to reduce the social determinants of health that exacerbate hypertension in the Navajo community, providing a sustainable, culturally grounded approach to disease management.
November 2024
Lilly Irani
Computational Social Science
UC San Diego
lirani@ucsd.edu
Abstract: Indigenous researchers and nations have been leaders in framing questions of data sovereignty around questions of cultural production, transmission, control, and benefits distribution.1In unceded Kumeyaay territory on the San Diego side of the US Mexico border, I have been involved in coalitions struggling to regulate or refuse surveillance technologies operated by for-profit private companies and deployed primarily by law enforcement where communities have little say in the practice of safety or the production and circulation of data. Smart streetlights record video and license plates of all, including visitors and tribal citizens, moving through the cities. As the US settler colonial government oscillates between political parties, all parties retain a commitment to building border walls out of metal and bits, and the walling of the border extends to wherever data dragnets can capture traces of people as we move. Transparency, accountability, and fairness are not enough when settler institutions evade accountability and are premised on invisibilized unfairnesses of settler colonialism and anti-blackness. In the southern border region, the Kumeyaay nations2 and Tohono O’oodham nations3 have both resisted the intensification of physical border walls; the settler state comes with “smart border” surveillance installations4. My position paper will describe border surveillance technologies through the core values presented by the US Indigenous Data Sovereignty Network5 as an opening to a conversation. I wish to build relationships and so we may more deeply articulate the horizons of data sovereignty, designs, and strategies that do reverse the encroachment of white supremacist capitalist settler colonialism and strategize on how to move towards those horizons.
1 https://www.temanararaunga.maori.nz/;
https://nihb.org/covid-19/upcoming-events/indigenous-data-sovereignty-developing-a-federal-framework-f or-the-handling-and-use-of-tribal-data/; Duarte, Marisa Networked Sovereignty Seattle: University of Washington Press
2 https://atmos.earth/indigenous-border-resistance-the-frontline/
3 http://www.tonation-nsn.gov/nowall/
4 https://inewsource.org/2024/05/15/in-california-border-town-ai-raises-fresh-questions-over-decades-old-s urveillance-tower/
5
https://usindigenousdatanetwork.org/wp-content/uploads/2024/10/Indigenous-Data-Governance-Brief-FIN AL.pdf
November 2024
Dr. Robert Jimerson Jr
Seneca Nation & Rochester Institute of Technology
Alex Jimerson
Seneca Nation
Abstract: Efforts to revitalize Native American languages, such as Seneca, face challenges due to limited resources, expertise, and institutional support, leading to underdeveloped formal learning curricula. Structured curricula are essential for guiding learners through language acquisition, providing a systematic approach that enhances retention and proficiency. This paper discusses the process used in the Seneca-speaking community of Cattaraugus to digitize and organize diverse language sources, including archival materials and recordings from elder speakers, to create comprehensive learning frameworks. The organized data supports both the development of language learning curricula and research projects leveraging deep learning techniques like automatic speech recognition and Word2Vec to further language revitalization. Ultimately, these efforts aim to strengthen Seneca language learning and contribute to the preservation and growth of Seneca culture through innovative technological applications.
November 2024
Nicole S. Kuhn
University of Washington Information School
nskuhn06@uw.edu
Abstract: American Indian and Alaska Native (AIAN) health organizations, communities, and individual content creators are increasingly leveraging the most popular social media platforms to share engaging Indigenous health communication (Calac et al., 2022; Duarte et al., 2021; Kuhn et al., 2020b). However, limited research exists on this critical topic despite its potential to positively impact the health and wellness of AIAN individuals who face some of the most significant health disparities in the United States (US) (Boyd et al., 2023). Human-computer interaction (HCI) social media research has the potential to provide insights that could improve AIAN health communication strategies for social media. Yet, further development of Indigenous research methodologies and ethics in this field is needed at this stage to ensure researchers can most effectively and ethically support this critical work across AIAN communities. HCI and social media research currently provides very limited representation of AIAN experiences (Chordia et al., 2024), along with minimal development of Indigenous research methodologies and ethics needed to initiate and sustain this work. This creates an environment where higher risks and lower benefits may exist for AIAN individuals and communities participating in this type of research. Additionally, there are general concerns about social media research ethics, including inconsistent or weak oversight by universities’ institutional review boards (Lukito, 2024) and additional risks to participants in research related to public health topics (Hunter et al., 2018; Massey et al., 2023). Approaches for expanding Indigenous research in this field must actively work to minimize risks and increase benefits to Tribal communities by respecting Tribal sovereignty in research, understanding that many communities have experienced harmful, unethical research that has caused significant mistrust of researchers and institutions, and acknowledging the diversity among AIAN cultures and communities and their approaches to research (Smith, 2021). Additionally, these approaches must consider the many diverse Tribal research laws and Tribal research review boards (TRRBs) that have been created by AIAN communities across the US (Kuhn et al., 2024). A broad discussion of Indigenous research methodologies and ethics has been developed through an analysis of Tribal research laws and TRRBs’ policies and procedures (Kuhn et al., 2020a; Kuhn et al., 2024). The themes identified in this analysis can be applied to HCI social media research to aid in the development of Indigenous methodologies and ethics for this field while also identifying additional gaps and concerns that need further consideration. Applying this work to community-engaged HCI social media projects can then provide practical insights to further refine these approaches. Developing this foundational knowledge now will help researchers most effectively leverage HCI social media research to positively impact AIAN communities, including supporting community efforts to share health communication across social media. Ultimately, this process can help lay the groundwork for developing an ethical, robust Indigenous HCI social media research agenda that can respond to rapid changes arising in social media use related to the increased integration and impact of artificial intelligence, widespread health mis/disinformation, and responses to global health and climate crises.
References
Boyd, A., Railey, A., Hsu, Y. C., Kirkpatrick, A., Fyfe-Johnson, A., Muller, C., & Buchwald, D.
(2023). Social Media Use among American Indian and Alaska Native People: Implications
for Health Communication Strategies. International Journal of Indigenous Health, 18(1).
Calac, A. J., McMann, T., Cai, M., Li, J., Cuomo, R., & Mackey, T. K. (2022). Exploring
substance use disorder discussions in Native American communities: a retrospective
Twitter infodemiology study. Harm Reduction Journal, 19(1), 1-6.
Chordia, I., Baltaxe-Admony, L. B., Boone, A., Sheehan, A., Dombrowski, L., Le Dantec, C. A.,
... & Smith, A. D. (2024, May). Social Justice in HCI: A Systematic Literature Review. In
Proceedings of the CHI Conference on Human Factors in Computing Systems (pp. 1-33).
Duarte, M. E., George, A., Deschine-Parkhurst, N., & Soto, A. (2021). Caring for Our People:
Indigenous Responses to COVID-19 Era Informatic Colonialism. AoIR Selected Papers of
Internet Research.
Hunter, R. F., Gough, A., O’Kane, N., McKeown, G., Fitzpatrick, A., Walker, T., ... & Kee, F.
(2018). Ethical issues in social media research for public health. American journal of public
health, 108(3), 343-348.
Kuhn, N.S., Kuhn, E. J., Vendiola, M., & Lefthand-Begay, C. (2024). Indigenous research ethics
and Tribal Research Review Boards in the United States: examining online presence and
themes across online documentation. Research Ethics, 17470161241240030.
Kuhn N.S., Parker, M., & Lefthand-Begay, C. (2020a) Indigenous research ethics requirements:
An examination of six Tribal Institutional Review Board applications and processes in the
United States. Journal of Empirical Research on Human Research Ethics 15(4): 279–291.
Kuhn, N., Sarkar, S., White, L. A., Hoy, J., McCray, C., & Lefthand-Begay, C. (2020b).
Decolonizing risk communication: indigenous responses to COVID-19 using social media.
Journal of Indigenous Social Development, 9(3), 193-213.
Lukito, J. (2024). Platform research ethics for academic research. Center for media
engagement. https://mediaengagement. org/research/platform-research-ethics/. Accessed,
30.
Massey, P. M., Murray, R. M., Chiang, S. C., Russell, A. M., & Yudell, M. A. (2023). Social
Media, Public Health Research, and Vulnerability: Considerations to Advance Ethical
Guidelines and Strengthen Future Research. JMIR Public Health and Surveillance, 9,
E49881.
Smith, LT. (2021) Decolonizing Methodologies: Research and Indigenous Peoples. London: Zed
Books.
November 2024
Max Liboiron
Memorial University
Abstract: There are increasing calls from scientists, Indigenous Peoples, and members of the public advocating for co-management of resources, the inclusion of Indigenous Peoples in data analysis and decision making, and other collaborative models where quantitative and calculative literacy and even fluency is required for participation. Yet, those terms cannot be determined solely by Western models for equitable, just, and Indigenous-driven collaboration and decision-making.
This presentation begins from the premise that as Indigenous Peoples, we already have numeric and calculative traditions and capacities. Through the case study of a long-term, community-based plastic pollution monitoring programme in Nunatsiavut (Inuit Land Claim area in northern Labrador, Canada), co-led by one Inuk on her homelands and a diasporic Métis working in a university, we share lessons about how to prioritize local and traditional knowledge in arithmetic methods and statistical models and how these processes change research findings. We start by introducing our methods of participatory statistics (real-time collaborative statistical computation) and other forms of co-analysis that use local knowledge to drive statistical methods. We then outline how we’ve come to understand which metrics and measures matter to community and which do not, and what this does to change Western scientific norms in the monitoring field. We have found, for instance, crucial differences between “population measures” versus measures about those fish over there and that role and status of zeros and nulls differ in community and Western science. For example, environmental monitoring and conservation science are based on population measures to determine ecological health. Yet Nunatsiavut community members didn’t ever ask about our population measures, and articulated groups of animals in other ways—many other ways. This led to genuine (and ongoing) questions about what are valid and meaningful generalized calculative models. A lightening talk will choose one of these examples for illustrative purposes.
Our case study sheds some preliminary insights into what the Indigenization of quantitative methods (and specifically computational methods) might consider more broadly. First, the terms of relationship between local knowledge and Western statistics are not readily apparent and are highly contextual, making oft-used terms for knowledge system integration like braiding, weaving, and co-production misleading. Second, the lines between Indigenizing methods and Indigenous methods are more pronounced and perhaps even crucial than we first anticipated. Third, while understanding and analyzing calculative insights are difficult due to very different worldviews, the methods for collaborative computation are not; participatory statistics and other forms of co-analysis have been extremely successful in terms of providing a wealth of information, context, and meaning. Thus, the issue is not that community co-analysis through calculation and statistics is not possible due to “capacity” issues, but that the knowledge systems at play are so different that their relations are a genuine, difficult, and shifting question.
November 2024
Sandy Littletree
University of Washington
In this short statement written for an ASIS&T panel, I present a vision of information science as part of an imaginary welcome session for new students to the field. This vision, which was paired with five other statements from the past century, represent diverse approaches to information science, ranging from S.R. Ranganathan to Elfreda Chatman. As I wrote this statement, I kept in mind that Indigenous perspectives are very foreign to about 99% of the information science field, including faculty, practitioners, researchers, and students at all levels. I aimed for a gentle introduction, however I recognize that this approach sometimes gives audiences the fuel to take on Indigenous projects despite their lack of familiarity with this work. The extra labor we have as Indigenous information scientists to educate others about our work can be tedious, and I wonder how to better inform students about Indigenous approaches to information science.
November 2024
Richard Ng
Ho’okahauai’ike
richard@designgoodbusiness.com
Abstract: KIAʻI is an AI data governance system designed to protect and promote Indigenous knowledge in the age of artificial intelligence. Through a community-driven validation process, Indigenous communities actively review AI-generated data for accuracy, cultural sensitivity, and ethical representation, ensuring data sovereignty and respecting Indigenous knowledge systems. By leveraging AI algorithms for pre-screening and feedback generation,KIAʻI empowers communities to participate in AI development, fostering responsible AI adoption while safeguarding Indigenous knowledge and cultural practices.
Introduction
KIAʻI, meaning "guardian" in Hawaiian, is an AI data governance system designed to protect and promote Indigenous knowledge in the age of artificial intelligence. The name embodies the system's crucial role in safeguarding Indigenous cultural practices and traditions. Through a community-driven validation process, Indigenous communities actively review AI-generated data for accuracy, cultural sensitivity, and ethical representation, ensuring data sovereignty and respecting Indigenous knowledge systems. By leveraging AI algorithms for pre-screening and feedback generation, KIAʻI empowers communities to participate in AI development, fostering responsible AI adoption while safeguarding Indigenous knowledge and cultural practices.
Background
The historical relationship between AI and Indigenous communities has been fraught with challenges. Indigenous knowledge systems have often been disregarded or appropriated without consent. AI systems are frequently trained on data that is biased or culturally insensitive, leading to inaccurate and harmful representations of Indigenous cultures. Existing approaches to address these issues, such as ethical guidelines and data governance frameworks, have often been insufficient.The more citations you include, the more our workshop participants will be able to follow up and grow the body of work we are collectively advancing.
Proposed Solution: KIAʻI
KIAʻI is an AI data governance system specifically designed to address the unique needs of Indigenous communities. It combines the following key features:
Community-Driven Validation: KIAʻI enables Indigenous communities to actively participate in the validation of AI-generated data related to their knowledge and cultures. Through a user-friendly interface, Indigenous validators can review data for accuracy, cultural sensitivity, and ethical representation. This process involves tagging, annotating, and providing feedback on AI-generated content.
Indigenous Data Sovereignty: KIAʻI prioritizes Indigenous data sovereignty, ensuring that communities have control over their data and knowledge. This includes establishing data governance policies, defining how data is collected, stored, and used, and determining who has access to it.
Data Trusts: KIAʻI supports the establishment and management of Indigenous Data Trusts. Data Trusts allow communities to govern the collection, storage, access, and use of their data, ensuring it is used responsibly and ethically.
AI-Assisted Validation: KIAʻI utilizes AI algorithms to automate tasks like pre-screening data for potential issues (e.g., factual errors, cultural insensitivity) and generating feedback suggestions to validators. This helps to streamline the validation process and improve efficiency.
Cultural Sensitivity: KIAʻI is designed to be culturally sensitive, incorporating principles of Indigenous knowledge systems and cultural values. The system aims to avoid perpetuating biases and to promote respectful and accurate representations of Indigenous cultures.
Workforce Development: KIAʻI promotes the development of a skilled Indigenous workforce capable of participating in AI development and governance. This involves training programs, capacity-building initiatives, and opportunities for Indigenous individuals to contribute to the design, implementation, and evaluation of AI systems.
Implementation and Evaluation
KIAʻI is currently under development, with a strong focus on community engagement. Indigenous communities are actively involved in shaping the system's design, functionality, and evaluation process. Evaluation methods include:
Accuracy: Evaluating the system's ability to accurately identify culturally appropriate and insensitive content.
Fairness: Assessing the system's fairness across different cultural groups, ensuring it doesn't systematically favor one group over another.
Cultural Sensitivity: Gathering feedback from Indigenous communities on the system's outputs and its alignment with cultural values.
Discussion and Conclusion
KIAʻI offers a promising approach to addressing the challenges of AI and Indigenous knowledge. By empowering Indigenous communities to participate in AI governance, by prioritizing data sovereignty and cultural sensitivity, and by fostering workforce development, KIAʻI has the potential to:
Protect Indigenous knowledge and cultural practices from misuse and exploitation.
Promote responsible AI development and deployment within Indigenous contexts.
Facilitate the equitable use of AI for the benefit of Indigenous communities.
KIAʻI represents a critical step toward ensuring that AI technology serves the interests of Indigenous communities. We encourage the adoption of this approach and the development of similar tools to safeguard Indigenous knowledge and cultures in the digital age.
November 2024
Xavier Nokes
The Design School & School of Arts, Media and Engineering
Arizona State University
xavier.nokes@asu.edu & luxnaut@gmail.com
Abstract: Indigenous knowledge (IK), regarded as such both by its originators and keepers (globally situated Indigenous peoples) and in its status as “alternative” or “other” (largely due to its position within the context of Western academic spaces), is a term referencing any number of ways of understanding and ways of being as delineated and upheld by Indigenous communities. Many cite the misconception of IK as distinctly in opposition to its Western counterparts, explicitly weaponizing the nature of IK’s derivation from place via the native communities that inhabit them (or have inhabited) and by extension collecting a number of disingenuous interpretations. This unfair ascription lends itself to the divisiveness of IK as valuable, while additionally bolstering the extractive practices performed by Western researchers and practitioners that deem such practices as the method by which IK might be recognized as “valuable”. This bastardized transcription (i.e. contextual cleansing and reassemblage) bolsters the notion of IK’s validity only when presented through the discerning slant of Western perspective. Such processes are by no means new, and indeed can be recognized within many marginalized communities that have had their knowledge made “palatable”.
Instead, we might propose the utilization of new and emerging (or perhaps, re-emerging) technologies geared towards experiencing knowledge as a platform by which to exemplify the very power that sets IK apart. The work presented in this paper aim to both reflect on work done by the author in addressing the potential towards the experience of IK, along with other forms of relative/cohabitating knowledge, specifically in the way of immersive and/or responsive audio and/or visual work, digitally aided creation, and adaptation of existing systems specifically in the way of exemplifying the understanding of their existence in and of themselves. Additionally this paper will touch on the power of gate/keeping, specifically in reflecting on these topics individually and in correlation with each other. Finally, we will examine the nature of knowledge procurement as it relates to IK and advancing technologies, specifically with respect to the often open nature of knowledge as it exists within Western academic spaces.
This writing is composed of components of the author’s larger body of work as a doctoral candidate in the School of Arts, Media and Engineering as part of their dissertation in the Media Arts and Sciences PhD program. Additionally, works discussed include intrapersonally-informed projects, community-led and/or aided projects, and prospective projects and ideas. In opposition to a strictly “solutions-based” perspective, this work hopes to exist in the way that so-called “living” documents do, with an emphasis on adaptability and intentionality.
November 2024
Temi Odumosu
Information School
University of Washington
todumosu@uw.edu
Miranda Belarde-Lewis
Information School
University of Washington
mhbl@uw.edu
Abstract: In her short experimental film These Networks in Our Skin (2021), artist Mimi Ọnụọha shows four Black women quietly brading and infusing hair and other natural substances into fiber optic cables. Drawing on traditional Igbo cosmology, their quiet ritual becomes a meditation on the remaking of technological power and the very act of systems creation. In this paper we describe and consider this and other experimental methods being used by artists working in collaboration with technology. We show how their theorisations on networks, cultural algorithms, digital sovereignty, and data in all its manifestations are investments in narratives of repair, that ask: How can ancestral knowledge be embedded within digital and artificial systems in ways that are beneficial and generative to Black and Indigenous communities?
As curators of museum exhibitions, we regularly initiate conversations about complex colonial histories, and are now considering what it means to work in networked information ecosystems that bridge the material and immaterial. In our approach to this paper, we use a curatorial analytic of juxtaposition to converse about multidisciplinary practices that interrogate how culturally attuned computing can be conscious in its values, be sensitive to multiplicities, and transformative in its corrective potential. We also identify how such practices intervene and/or break the rules and default settings that have so far produced biased technological outcomes; the loop of ongoing colonial extractions.
This paper introduces two conceptual threads that weave together, as we look to the artists for theoretical support in both research and practice. First, we engage the ethical “sixth sense” proposed by Eve Tuck and K.Wayne Yang, which asserts that “the [technological] pursuit of objectivity, always defined by those in power to protect their power, occludes the intuition of the observer—the sixth sense that could be his or her ethical radar and moral compass viewed as an ethical awareness” (Tuck and Yang 2014, 814). In this paper we see this intuitive radar as an essential aspect of ancestral knowledge held within indigenous communities, and thus a necessary way to orient research and practice.
Secondly, thinking alongside artists such as Susanne Kite, Stephanie Didkins, Wally Dion, sidony o’neal, and Ọnụọha, we explore the concept of “making kin with machines” and follow
their methods for human-computer interaction – centered on connection, reclamation, repair, and interdependency. Through their work, we ask: How might algorithms be informed by cultural practices and narratives that better serve diverse communities? What kinds of speculative possibilities do artful technological interactions offer? How are artist-machine collaborations imagining the future?
November 2024
Aspen K.B. Omapang
Information Science
Cornell University
ar2465@cornell.edu
Abstract: In this position paper, I make two claims. First, accounting for whiteness in tech development is an ongoing practice and commitment, and two, white people have a role to play in re-imagining and co-creating anti-colonial technologies. The outcome is guidance informed by relational epistemologies that can be used in current efforts to develop anti-colonial technology. Whiteness is not isolated to interactions with, and within, white people. Whiteness is baked into infrastructure and the logics that dictate what technology is and the tools that allow us to create (Harris, 1993; Star, 1999). Whiteness also permeates our cosmologies, the way we come to understand the world and all that inhabits it (Hickman, 2017). In our current landscape, we are subject to both the “masters’ tools” and masters' logics (Lorde, 2012). Whiteness is also reactive (McWhorter, 2009). Therefore, rooting out whiteness must address whiteness’ multiplicity. Secondly, while whiteness has been thoroughly critiqued, the role of white people in anti-colonial technology development has largely gone unaddressed. Whiteness is our enemy, but white people have a necessary role to play, as both victims and beneficiaries of white supremacy, in co-creating an anti-colonial future (Mamdani, 2020).
Our traditional governance and relational epistemologies, our practices of Indigeneity, resist the core tenets of whiteness. Our identification of human and non-human kin, and the land, sky, and waters as that which nourishes us is antithetical to notions of property, human-centricity, and individualism. Such as hāloa (the long breath) to maintain pono (just, balanced) relationships and wahkohtawin (relatives and relations) (Lewis, et al., 2018). The commitment to relations can result in “settler ontocide” (TallBear, 2019). In TallBear’s interpretation of Nick Estes' term, “It means ridding ourselves of the category of the settler along with its discourse of white supremacy and assertions of an inherent right to these lands and waters.” I posit that when we bring technology development into our traditional governance and relations, so too can our practices keep whiteness in check and offer a series of ongoing commitments to white co-developers, which prioritizes anti-coloniality.
This position paper attempts to apply relational epistemologies in order to root out whiteness in anti-colonial technology development and provide guidance on how to place white people as co-conspirators in dismantling white supremacy. In the full paper, I will use two technologies as case studies to apply the theoretical and relational frameworks referenced above. First, a Kānaka Maoli-developed (Native Hawaiian) software application1 to aid in data collection through the practice of kilo (sense and place-based observation). Second, I bring in the speculative technology of a large-language model (LLM) designed to aid in Native language learning. The goal is not to advocate for a particular technology, but to reinforce the ongoing effort needed to Indigenize development.
Developing anti-colonial technologies is no small feat. Our traditional ways of knowing and relational practices reveal new futures, and also have to contend with the material reality of the present. This position paper addresses the reality of technology development by addressing whiteness and orienting white people to anti-colonial ontologies.
Notes:
References:
1. Harris, C. I. (1993). Whiteness as property. Harvard law review, 1707-1791.
2. Hickman, J. (2017). Black Prometheus: Race and radicalism in the Age of atlantic
slavery. Oxford University Press.
3. Lewis, J. E., Arista, N., Pechawis, A., & Kite, S. (2018). Making kin with the machines.
Journal of Design and Science.
4. Lorde, A. (2012). Sister outsider: Essays and speeches. Crossing Press.
5. Mamdani, M. (2020). Neither settler nor native. Harvard University Press.
6. McWhorter, L. (2009). Racism and sexual oppression in Anglo-America: A genealogy.
Indiana University Press.
7. Star, S. L. (1999). The ethnography of infrastructure. American behavioral scientist,
43(3), 377-391.
8. TallBear, K. (2019). Caretaking relations, not American dreaming. Kalfou, 6(1), 24-41.
November 2024
Indulata Prasad
School of Social Transformation
Arizona State University
iprasad@asu.edu
ABSTRACT: The Indian education system, long regarded as the main vehicle for social mobility, has failed miserably for Dalit-Bahujans (Guru 2002; Paik 2009). Little education occurs in public schools that serve Bahujan (literally the “majority”) groups, that is, those who come from oppressed classes or castes, including Dalits (formerly referred to as “untouchables”), Adivasis (Indigenous “tribes”), and religious minorities, due to a deeply rooted caste bias that views them as existing solely to serve privileged castes (Singh 2023; Nambissan 2020). Although the progressive Right to Education (RTE) Act in 2009 legislated that all children had a right to free, compulsory elementary school education from ages 6 to 14, and led to a 40% increase in student enrollment in government-run schools throughout India, only 20% of children attend more than a few years of primary school in the state of Bihar, which has a significant Dalit population (Jan Jagran Shakti Sangathan 2023; also see, Drèze, J. 2005; Jain & Dholakia 2009; Malvankar 2018; Sadgopal 2010). The defunding of the education sector, coupled with rampant government corruption, makes attending private schools the only viable educational option, but they remain beyond the reach of most Dalits in Bihar (Kalapura & Vishal 2019; Bihar Dalit Vikas Samiti 2022). Attending school also competes with basic survival needs as Dalit and other Bahujan families often rely on their children’s household and agricultural labor to get by. Girls, who are customarily viewed as burdens to the family, are disproportionately impacted by structural inequalities (Gordon 2023). Unlike boys, girls are not sent to private schools or provided tutors because their families do not consider it worth investing scarce resources in their education since it is assumed they will be married off in their teens and sent to live with another family.
Structural factors alone do not account for the low educational levels of Dalits in rural Bihar. Pursuing an education means being confronted by societal biases that alienate them from their educational experiences. When children from Dalit communities attend school, they are taught by teachers who belong to the privileged castes and both public and private schools elide references to caste and class inequalities; their curricula conform to societal schisms rather than challenging them (Sharma 2018; Goel & Husain 2018; Ingole 2020; Saandilya 2023). The erasure of Dalit histories and perspectives further contributes to children’s sense of alienation and high dropout rates. Even those youth who complete their primary education and go on to obtain a high school diploma lack the skill sets normally associated with that level of formal education. Finally, the few who do manage to complete their studies are faced with the harsh reality of having to take up caste-prescribed, low-paid labor that defeats the claim that “education leads to social mobility.”
Transforming Dalit-Bahujan Education: To mitigate these challenges to education in rural Bihar, we propose to develop and test a curriculum for training and supporting Dalit and other Bahujan youth to generate Community-Driven Archives (CDAs) in the Gaya region. CDAs disrupt top-down models of history writing and archives by training communities to be curators of their own histories, making them available for present and future generations. CDA theory posits that all individuals possess valuable experiential knowledge and should not be treated merely as passive recipients of education (Godoy 2017; Smith 2022). The CDA as a pedagogical tool thus has the potential to undo what Paulo Freire referred to as the “culture of silence” (2014: 30) promoted by conventional education methods that kill innovation, critical thinking, and experimentation by enforcing conformity, often through harsh discipline. We envision that involving youth in the creation of CDAs that document their community histories, cultural practices, and lived realities will be a more potent educational tool than the conventional banking model of education (Freire 2014, hooks 1994), which devalues Dalit and other Bahujan youth. Drawing on eclectic theories from critical caste and gender studies (Guru 2009; Paik 2016), community memory studies (Cotera 2015; Godoy & Dunhan 2017; Lozano et. al. 2023), feminist and critical pedagogies (Freire 2014, hooks 1994), among others, we seek to develop and test a CDA training curriculum centered on the experiences of marginalized caste groups (Dalit/Bahujan). We expect that providing youth with the intellectual and technical skills (Vaidehi, Reddy & Banerjee 2021) for collating stories from their communities will not only improve their educational outcomes but equip them to examine and document real-life issues and seek their own answers to society’s challenges.
This expectation is based on the results of a pilot research project, supported by a seed grant from Arizona State University (ASU), that trained Dalit-Bahujan girls and boys between the ages of 16 to 25 to use smartphones to interview elderly activists in three villages and document their oral history of the Bodhgaya Land Movement (BGLM). In the 1970s and 1980s, thousands of Dalits and other landless laborers, including women, mobilized to obtain legal title to and control over agricultural land. The movement resulted in the dismantling of an ancient feudal system of enslaved labor, but it has largely been forgotten by the younger generation (Prasad 2021b). Ever since PI Prasad began conducting ethnographic fieldwork on the long-term impacts of the BGLM, Dalit elders have lamented that their memories and history of effective activism are being lost with their passing.
Workshops conducted for the pilot project yielded preliminary insights on the transformative potential of developing a CDA-centered education in historically marginalized communities in India. Youth mentioned feeling empowered from learning about their silenced histories and agency. Some youth began questioning the erasures of Dalit/Bahujan struggles from dominant memory and school curricula. They expressed pride after learning about their parents’ or grandparents’ struggles, which paved the way for the creation of an inter-generational safe space for community learning. As one girl exclaimed, “I did not know I could ask questions!” Four of the youths (two girls and two boys) became committed to developing lifelong learning skills by documenting community histories for posterity. These youths will join the research team in developing a scalable CDA curriculum appropriate for marginalized communities in rural Bihar.
November 2024
Robert Pugh
Computational Linguistics
Indiana University
pughrob@iu.edu
Erika Sosa
Hispanic Linguistics
Indiana University
ersosa@iu.edu
Francis Morton Tyers
Computational Linguistics
Indiana University
ftyers@iu.edu
Abstract: We examine and critique a recent trend in the field of natural language processing of publishing "massive multilingual" solutions which leverage data from hundreds or thousands of languages, and typically claim that "[task X] is now supported for N,NNN languages!", where such tasks have included ASR (Pratap et al. 2023), morphological analysis (Nicolai et al., 2020), and others. While such efforts may seem to promote inclusivity in language technology, there are important problems with this popular approach that call into question the validity of claims of support, particularly for marginalized and Indigenous language communities.
Given the scale, researchers cannot feasibly collaborate with impacted language communities. This leads researchers to rely on "found data" (online data or previously published work), which for Indigenous languages means Bible translations and religious texts. Resulting models built using this data typically have a tenuous relationship with real-world language use.
Because massive multilingual NLP systems typically rely on found data and lack input from Indigenous language community members, any evaluation that might be done for such languages cannot be expected to be reflective of reality. We describe such evaluation attempts as performative rather than substantive.
Finally, we suggest that dubious claims of "supporting" a given marginalized language community with respect to some fundamental language technology may result in deprioritization of community-based data collection and technology development. If researchers, or large tech companies, for example, claim that ASR has been developed for a given Indigenous language,
other researchers in the NLP community, and potentially funding bodies, may decide to focus their efforts on the "remaining" languages for which such support does not exist. Since, as we have discussed, the claims of support are only true using a very narrow definition of "support," the end result is that a large number of language communities are left with poorly-performing solutions that do not serve any purpose connected with real world language use.
To illustrate this final point concretely, we present a case study evaluating the recent "Massively Multilingual Speech" ASR model from Meta AI1 on a selection of Indigenous languages from Mexico. We assess the model’s performance using "real-world" speech recordings that reflect contextually-appropriate language use – grounded in day-to-day realities. This allows for a more genuine assessment of the model’s effectiveness for Indigenous languages, highlighting gaps in performance and underscoring the limitations of claims of “supports".
Without meaningful community involvement, the creation of high-quality, contextually appropriate data, and transparent evaluation practices, claims of linguistic "support" by massive multilingual models are fundamentally misleading. Such claims may inadvertently harm Indigenous language communities by overshadowing or undercutting community-led efforts to develop technology in alignment with local needs. In order for language technology to genuinely benefit Indigenous communities, collaboration and transparency must be prioritized over scale and unsubstantiated claims of support.
1 This is but one of many examples of a "massively multilingual" approach in NLP. Our critique is not leveled at individual researchers but at the system that proposes and rewards the unethical framing of these lines of investigation.
References
Pratap, V., Tjandra, A., Shi, B., Tomasello, P., Babu, A., Kundu, S., ... & Auli, M. (2024). Scaling speech technology to 1,000+ languages. Journal of Machine Learning Research, 25(97), 1-52.
Nicolai, G., Lewis, D., McCarthy, A. D., Mueller, A., Wu, W., & Yarowsky, D. (2020, May). Fine-grained morphosyntactic analysis and generation tools for more than one thousand languages. In Proceedings of the Twelfth Language Resources and Evaluation Conference (pp. 3963-3972).
November 2024
Phil Steinhauer-Mozejko
Indigenous Connectivity Institute
University of Alberta
Phil@clearskyconnections.ca
Abstract: Our proposal would explore the intersection, interplay and tension between Indigenous Peoples and digital connectivity, as observed by citizens of Saddle Lake Cree Nation. I offer “Indigenous” in the working title for the sake of flexibility. But this contribution—at least preliminarily—would be based on my graduate work completed in 2024 during my time in the Master of Arts in Communications and Technology program at the University of Alberta (Canada). I worked and conducted primary research in my home nation of Saddle Lake, located in Treaty Six Territory, with relatives and family members, a few of whom are Elders, Knowledge Keepers and ceremonialists.
We would not necessarily position this paper to include/provide a solution or answers to the problem of rapid technological integration with our (Indigenous) ontological and epistemological orientations. Rather, we seek to challenge (Indigenous and non-Indigenous) conceptions and/or interpretations around “digital equity” or “digital sovereignty,” and examine the notion (or supposition) that digital connectivity, including AI, requires ethical integration with longstanding Indigenous traditions, knowledge, and culture. In addition, our position will be grounded in discussion of Indigenous (presumably nêhiyaw) traditions around relationship. Digital connectivity and, more specifically, the internet, as a technological medium, have played a significant part in exporting Western media and culture (Euro-American hegemony) around the world; it has penetrated deeply into communities and the daily lives of people everywhere.
They have also been conducive to the development of an all-encompassing market ontology. It has been said that the architecture of the internet is “inescapably capitalist.” We are concerned about nêhiyawak (Plains Cree) and other Indigenous communities being able to maintain/sustain their languages, knowledge systems, and traditional ways in a time of rapid globalization via digitalization.
We do not necessarily believe that the problem of (post)modern digitization can be solved via digital or data equity strategies or imperatives, or that inclusion and diversity initiatives which incorporate a variety of Indigenous perspectives, stories and datasets could lead to more representative/better (“people-centred”) sociotechnical infrastructure or digital technology being created that would holistically benefit or work for the people, or that they would even lead to more positive social and/or cultural “outcomes” for our communities.
Through a combination of autoethnography, Indigenous research methodology (IRM), and a braided framework that integrates critical theory and a postmodernist orientation with a nêhiyâw epistemology, this position paper would interrogate and highlight the concurrent challenges, opportunities, ambiguities and complexities which digital connectivity uniquely present for Indigenous communities. Social and cultural impacts and implications of digital connectivity vis-à-vis Indigenous Peoples, and vice versa, are investigated. The research unpacks the concepts of digital equity, digital inclusion, and digital sovereignty via a critique of the neoliberal language and ideology of human rights and recognition-based frameworks undergirding them. While digital connectivity offers potential benefits for Indigenous cultures, economies, community welfare, and political mobilization, it also presents significant challenges related to maintaining sovereignty and distinct cultural integrities in a rapidly globalizing world. This paper would contribute to ongoing discourses concerning the holistic impacts and implications of digital connectivity with Indigenous Peoples. It would offer a nuanced perspective on the ways Indigenous communities are leveraging digital technologies while grappling with real and potential threats to autonomy, culture, and self-determination. In a sense, this position paper would function as a provocation. If we seek to create more equitable, culturally harmonious, restorative and responsive digital technologies that may truly represent/reflect our values systems or ways of knowing and being, and which serve our diverse interests, then we must look at some of our core philosophical underpinnings and theorize/conceptualize how they might interrelate with current digital systems undergirded by (and designed to promote) notions of progress, and liberal democratic (colonial) thought of the Enlightenment era. I lean on the nêhiyaw concept of “wahkotowin” — translated to “relationship” or “the natural laws/rules that govern all relations” — and theorize how it interacts with digital connectivity. We also would explore where the commonalities, congruences, reciprocities exist. Conversely, we also theorize on cultural compatibility and/or commensurability. For lack of better descriptors, we are interested in whether the internet—in whole or in part—can be “Indigenized” and/or decolonized.
Our organization, the Indigenous Connectivity Institute, is interested in promoting true digital equity among First Nation(s), Inuit, Métis, American Indian, Native American, Alaska Native, Native Hawaiian, Aboriginal/Tribal/Indigenous communities across Turtle Island and Inuit Nunangat. However, we do so with caution, in close collaboration, and only at a community’s request/choice.
November 2024
Cindy Tekobbe
Gender and Women’s Studies
Communication
University of Illinois Chicago
ctek@uic.edu
Abstract: When discussing generative AI, we talk about the data sets it is trained on and the logical protocols underpinning its functions. We talk about racial and gender bias baked into the data, but we do not disrupt these biases at their roots of settler colonialism and patriarchy. We do not talk about culturally situated ontologies as an approach to AI training or kin protocols as network theory. For this paper, I want to think through the potential for indigenizing AI and the kinds of non-Western traditions that imagine artificial intelligence in multiplicities not fixed taxonomic points. I want to think through employing non-western knowledge-making to create unique data paradigms. I want to ask what it would change to include the voices of Indigenous designers. I want to consider notions of access in a world where 700 million people don’t have reliable electricity. I want to imagine land and water as sacred and collaborative stakeholders and not as resources for extraction. I want to resist all extractive models. I want to treat our work in AI design as a commitment to the future-thinking, meaning-making communities in which we are all situated.
I am Cindy Tekobbe, and I am of Choctaw and Irish descent. I live in Chicago, Illinois. I am enrolled with the Choctaw Nation of Oklahoma. When I introduce myself to other Natives, I also name my mother, Helen Jane, and my grandmother, Esther Belle, and that my people are in Phoenix. This is my cultural identity that informs my ontologies and situates me in my kin networks, my communities, and greater Creation.
There is a lot to unpack in the protocol of an Indigenous introduction, but primarily introductions are used to make connections between epistemologies and kin networks and to locate people in their communities (Kovach, 2010). Indigenous people’s lifeways and ontologies are not monolithic, but we can make some generalizations about our ethics in this world and our relationships with its occupants and how these may apply to AI design. We do not locate humans at the center of creation or its pinnacle. We do not revere the individual before the community or sort beings into western hierarchal taxonomies. We extend rights to humans and non-humans alike - water, land, creatures. We co-exist in the “creatureness” (Deloria, 1972) of Creation. We are all related. We do not make knowledge in isolation. Rather, when we make meaning, we locate that meaning in the context of its community, its time, and its ecology (L. Smith, 2012). Our ways of arriving at collaborative and collective meaning are sacred because they are situated within these creations, protocols, and lifeways.
Coding computers involves normalizing and standardizing data into homogenous sets. It involves taxonomies of data types and hierarchies of instruction. It invokes a kind of top-down logic that even object-oriented paradigms do not supersede. Recently, when Sam Altman was asked if students still need to code, he said yes because it teaches you how to think (B. Smith, 2024). I argue that it teaches you one western way of thinking. We can resist the tendency to colonize data and settle meaning into rigid types. AI mistakes are called hallucinations, but I think culturally and ecologically sensitive AI could be called a dream. I want to dream of an AI that connects rather than extracts.
November 2024
Suanmuanlian Tonsing
School of Information
University of Michigan-Ann Arbor
tonsing@umich.edu
In this paper, I propose two strategies to address the harmful colonial terms embedded in our daily digital interactions, often overlooked yet deeply harmful. The first is a global Indigenous digital repository of colonial terminology historically used to harm Indigenous communities, serving as a resource on their historical weight and impact. The second involves a tooltip feature on major social media platforms, linking the platforms to the global repository, providing insight to users into the harmful language encountered online.
My drive to pursue this research grew after our community’s experience of ethnic violence in May 2023, in Manipur, India. Twitter (X) flooded with misinformation labeling us as “illegal immigrants” and a “threat” to India. Forced to counteract these narratives, our community migrated from WhatsApp to X, though unfamiliar with the platform. Through May and June, I operated on little sleep, countering misinformation, supporting survivors, and writing articles. Eventually, I became a target of online violence myself, prompting my return to the United States.
Back in the U.S., I developed a framework I call “critical indige-autoethnography” to reflect on these experiences. Reflecting on my experiences and the stories from my community, I began to see how our optimism toward technology– our belief that it could amplify our counter-narratives was shaped by What Maggie Walters and Tahu Kukutai call our corrupted “Indigenous lifeworld.” Even as we engaged with digital tools hoping to share our perspectives, we often relied on content sourced from colonial archives, unknowingly leaving digital footprints that reinforced the very colonial structures we sought to challenge. This highlighted what I term the “techo-decolonial gap”: when Indigenous communities co-opt technology for survival but inadvertently reinforce colonial histories rather than dismantling them.
The question haunts me: “How many harmful colonial terms and identities of Indigenous communities do we encounter online daily, unaware of their harm to these communities?” As part of a minority Indigenous tribe, lacking the numbers and the networked relationships to shape digital policy, it’s disheartening to witness these harmful colonial terms continue to persist, especially when the urgent task of decolonizing the colonial harms embedded in archives feels like an overwhelming challenge on its own.
In my paper for Indigenous Computational Futures, I aim to explore with Indigenous communities what it might truly look like to decolonize the harmful colonial language that permeates digital spaces like Twitter, Instagram, and Youtube, among others. I put forward two ambitious proposals. That Indigenous peoples create a global Indigenous digital repository- a resource that documents harmful colonial terminology historically weaponized against Indigenous groups, capturing each term’s context and historical weight. This reservoir would link with major social media platforms, integrating into their frameworks. Through a simple “tooltip” feature on these social media platforms linked to the global Indigenous digital repository, users would be alerted when they encounter a colonially harmful term and could access the reservoir for historical insight.
While this proposal may seem “straightforward”, its success depends on a coalition of Indigenous technologists and policy makers united across the globe, able to collaborate with both governments and industry leaders. Only through this collective effort can we think and reshape digital spaces to promote respect, awareness, and truth about Indigenous peoples.
November 2024
Andrew Wiebe
Faculty of Information
University of Toronto
andrewj.wiebe@mail.utoronto.ca
My work with Indian Residential School Survivors and 2-Spirit and Indigi-Queer peoples is deeply rooted in Indigenous stories of beavers as the builders of bridges between worlds and sustainable environments. This underlying metaphor draws heavily from Leanne Betasamosake Simpson and Audra Simpson, as well as other forms of storytelling that envision forms of inter-species diplomacy. However, this can be difficult to conceptualize in online spaces, specifically digital infrastructure, because we are implicated in technological infrastructure designed to be invisible yet enmeshing us in its relations.
I want to reconfigure Andrew Wilford’s use of the term settler colonial mesh, which describes the specific entanglements of Indian Residential Schools. Settler colonial mesh is the layers of entanglements that reinforce forms of oppression against Indigenous peoples that will tighten on people depending on their specific relations to both Land and the settler colonial state. These layers are the technologies of law, social services, police, and education. To examine these different state formations and the technologies they employ, records become a key site of violence against Indigenous peoples. Specifically, looking at the intersection of law and social services, there has been an effort to erase human accountability and Land relations through adoption records.
As a case study, the Truth and Reconciliation Commission in Canada resulted in many records entering digital space in one form or another out of an obligation to honour the gathering and sharing of “truth.” Moreover, many of the provinces in Canada have opened the ability to access adoption records for people who have left care. On the one hand, entering these records into digital infrastructure has enabled community access to family history and community resources. However, this digital infrastructure has also amplified a set of settler colonial enmeshments with technology that must be addressed. Namely, the interoperability of this infrastructure has been fragmented by the settler colonial state, resulting in long processes such as unsealing adoption records to find no useable information or permanently sealed records before being able to even apply for genealogy or Treaty Status. Each province in Canada has different rules and regulations, resulting in an inconsistency between access points, where unsealing records is seen as “simply filling out a form online,” but this results in the absence of humanity during a sensitive and deeply personal process. Moreover, many reserve communities in Canada do not have reliable internet access at home (about 42.8%, according to Stats Canada).
As a Beaver, I want to propose solutions by dramatically reimagining the settler colonial mesh in which adoption records sit. I suggest reimagining horizontal and vertical accountability, in which records that the government created through any of the elements of the settler colonial mesh are in relation. Moreover, I propose that there needs to be more attentiveness to Land and how that informs the keeping and sharing of these records. Ultimately, I believe we must first map the infrastructure and then choose which areas to cut down to build a bridge between worlds.
November 2024
Lili Yan
Michigan State University
yanlili@msu.edu
Caitlin Kirby
Michigan State University
kirbycai@msu.edu
Jennifer Gauthier
College of Menominee Nation
jgauthier@menominee.edu
Marci Hawpetoss
Menominee Indian Tribe of Wisconsin
mhawpetoss@menominee.edu
Julie Libarkin
Michigan State University libarkin@msu.edu
Abstract: Applying emerging technologies in Indigenous contexts is often accompanied with multi-layered tensions, which pose challenges to achieve culturally sustaining goals in a profound way. The design of technology, in fact, is all rooted in particular cultural, historical, and political knowledges and values (e.g., Litts et al., 2021). Technological advances have historically alienated or been used against marginalized communities, including Indigenous peoples (e.g., Swanson & Loft, 2014). Currently, artificial intelligence (AI) is in unprecedented expansion and development, in some cases involving the exploitation of resources and communities. Amid the unprecedented expansion and development of AI, we, as a multi-sector team and multicultural, argue that it is a particularly important time for us to pause and (re)engage with the cultural nature of the fundamental concepts of AI and what it means to be AI literate. In this position paper, we share a culturally centered approach for working in partnership across industry, academia, and Indigenous communities to advance guidelines for productive and meaningful community-centered AI design in Indigenous contexts.
AI is often assumed to be acultural. However, current forms of emerging technology, including AI, perpetuate systemic injustices across multiple settings (e.g., Eubanks, 2018; Noble, 2018), which privileges certain knowledge systems over the others. To date, discourses of AI are mostly centered on anglophone Euro-Western portrayals and expectations, and effectively marginalize their non-Western counterparts (e.g., Cave et al., 2020). Therefore, the design, development, and deployment of AI in Indigenous contexts requires continuous (re)engagement with AI as cultural and intentional efforts to make culture visible across the technology lifecycles of AI tools. To date, noble efforts have been made to examine AI imaginaries in different cultures that will impact AI futures profoundly (Cave & Dihal, 2023). Furthermore, Lewis and colleagues (2020) propose centering Indigenous knowledge systems in the conversation around AI and society to disrupt fundamental anthropocentrism in Euro-Western science and technology. Building on these existing works, we argue a culturally centered approach is needed for joint work on AI in Indigenous context.
Our exploration of a culturally centered approach has been shaping from a broader project working with multiple communities to construct a culturally centered AI literacy framework. Specifically, our team includes learning scientists, education technologists, tribal community development leaders, AI industry workers, and community-engaged STEM education researchers; we are initiating a community engagement process to redefine AI literacy. In this position paper, we first synthesize relevant discourses on AI to articulate foundations of a culturally centered approach. Guided by the question of what it means to create a framework across communities, we provide key tensions to illustrate our proposed culturally centered approach, which we identified from our discussions over the past ten months. These tensions generate a series of questions we would like to invite the broad community of interests to engage with. Our work reimagines a roadmap of repurposing AI to center communities, which allows for co learning and co-developing across diverse ways of knowing and being and therefore has implications for designing AI for diverse futurities.
References
Cave, S., & Dihal, K. (Eds.). (2023). Imagining AI: How the World Sees Intelligent Machines. Oxford University Press.
Cave, S., Dihal, K., & Dillon, S. (Eds.). (2020). AI narratives: A history of imaginative thinking about intelligent machines. Oxford University Press.
Eubanks, V. (2018). Automating inequality: How high-tech tools profile, police, and punish the poor. St. Martin's Press.
Lewis, J. E., Abdilla, A., Arista, N., Baker, K., Benesiinaabandan, S., Brown, M., ... & Whaanga, H. (2020). Indigenous protocol and artificial intelligence position paper. Retrieved from https://spectrum.library.concordia.ca/id/eprint/986506/
Litts, B. K., Searle, K. A., Brayboy, B. M., & Kafai, Y. B. (2021). Computing for all?: Examining critical biases in computational tools for learning. British Journal of Educational Technology, 52(2), 842-857.
Noble, S. U. (2018). Algorithms of oppression: How search engines reinforce racism. New York University Press.
Swanson, K. & Loft,S. (Eds.). (2014). Coded territories: Tracing Indigenous pathways in new media art. University of Calgary Press, 2014.
November 2024
Marina Johnson-Zafiris
Kanien’kéha Ratiwennahní:rats
Cornell University
mlz32@cornell.edu
Abstract: Fifteen years ago, big data was announced as the “next frontier” for innovation, competition and productivity. Couldry and Mejias introduce the term “data colonialism,” a topic describing the “process by which governments, non-governmental organizations and corporations claim ownership of and privatize the data that is produced by their users and citizens.” However, for Indigenous peoples data colonialism is not a parallel to “historic” frontier colonialism, but rather concurrent to the active and ongoing colonization of Turtle Island. Extractive and extortive computing relations of settler-colonial entities impose themselves onto Indigenous communities, both infrastructurally (where the data is held) and procedurally (how data is operationalized).
At this intersection of people, contention (or conflict around power dynamics), information and technology, is the field of data activism. Though often overlooked and undiscussed in Human Computer Interaction literature, there is a growing body of Indigenous data activism and action research dedicated to creating and cultivating insubordinate (and unapologetically Native) technologies that counter entrenched powers. This prompts Indigenous and allied computational designers to decenter the end product or material artifact as the main objective, but rather place the social function of designing and co-creating as a core tenant of the process. Historically, designers and researchers have approached this through various methods of participatory design and design justice, where there is an attempt to collapse or push at the normative borders between the designer and user.
However, to work within an anti-colonial framework, we must disestablish the above methodologies as the start of community-based research and design; participatory research is ancient and exists prior to the context given to it by the academy. This is to say, Indigenous people’s have been conducting participatory research and design since time immemorial. Gaudry calls us to enact “insurgent research,” ie. explicitly employing Indigenous worldviews and orienting knowledge creation towards our own communities of belonging.
As a Rotinonshonni researcher, I argue we do not exclusively carry a responsibility toward the academy, but rather, and more importantly, a responsibility to employ Kanonsonionwe (the ways of the longhouse) again as our original research paradigm. This means not only employing our wampums as frameworks or metaphors of understanding but as embodiment.
This paper focuses on Rotinonshonni, specifically Kanien’kehá:ka (Mohawk People) operationalization of the Kaianerekowa (The Great Good Path, Great Law of Peace), which established and set protocols for our clan-based consensus model of decentralized, non-hierarchical information gathering and decision making. Deliberation, discussion, analysis all of these are built into the structures of the which we have has operated for centuries pre-contact to present. I elaborate on recent efforts to to move towards a clan-based consensus model as a form of participatory design for community based computational technologies. To look to the future and imagine the information gathering, the prototyping, the iterative redesigning within our Kawatires (our clan-families) as a crafting of computational systems we can live alongside and know how to sustain.