Sacred Uses of AI: Cultural Preservation Through Immersive Virtual Experience Dr. Aria Chen, Theo Vasquez, and Dev Anand HoloDream Research April 2026

Abstract

The accelerating disappearance of the world's cultural traditions represents one of the most significant losses of human knowledge in history. UNESCO estimates that 43 percent of the world's approximately 7,000 languages are endangered, and with each language lost, entire cosmologies, medicinal systems, and philosophical traditions vanish irretrievably. This paper examines the potential of artificial intelligence, combined with immersive virtual environments, to preserve and revitalize endangered cultural traditions not as static museum exhibits but as living, interactive experiences. We explore case studies spanning Indigenous oral traditions, ancient language reconstruction, sacred textual traditions, the African griot lineage, and Japanese cultural arts. We address the profound ethical questions surrounding the digitization of sacred knowledge, propose a framework for respectful AI-mediated cultural preservation, and argue that conversational AI characters enriched with deep cultural knowledge, paired with auto-translation capabilities, represent a fundamentally new paradigm for cross-cultural understanding. We conclude that this technology carries both extraordinary promise and serious risk, demanding a new kind of sacred responsibility from its creators.

Introduction: The Crisis of Vanishing Cultures

We are living through an extinction event that most people cannot see. It is not the loss of species, though that crisis runs parallel. It is the loss of entire worlds of human thought, knowledge, and spiritual practice, disappearing at a pace that would have been unimaginable a century ago. Every two weeks, a language dies. With it goes not merely a collection of words but a unique way of understanding reality, a pharmacopoeia of plant knowledge developed over millennia, a cosmology that maps human experience onto the natural world in ways no other tradition replicates. The scale of this loss is difficult to comprehend. The Endangered Languages Project, maintained by the First Peoples' Cultural Council and the University of Hawaii, catalogs thousands of languages in various stages of decline. UNESCO's Atlas of the World's Languages in Danger classifies nearly 2,500 languages as vulnerable, definitely endangered, severely endangered, critically endangered, or extinct since 1950 (Moseley, 2010). Behind each classification lies a community watching its elders pass away, taking with them songs, stories, rituals, and ways of being that no written record fully captures. But this crisis extends well beyond language. Cultural traditions that have sustained communities for thousands of years are eroding under the pressures of urbanization, economic migration, globalization, and the sheer gravitational pull of dominant media cultures. The Ainu traditions of northern Japan, the songlines of Aboriginal Australians, the elaborate oral histories of West African peoples, the contemplative practices of Tibetan Buddhism, the Sufi poetic traditions of Central Asia: all face varying degrees of existential threat. Not because they lack depth or relevance, but because the economic and social structures that sustained them are being dismantled by the speed of modern change. This paper proposes that artificial intelligence, specifically conversational AI combined with immersive virtual environments, offers a genuinely new approach to this crisis. Not a replacement for living culture, which would be both impossible and undesirable, but a complementary preservation and revitalization tool unlike anything previously available. We argue that the transition from static archival methods to dynamic, interactive, AI-mediated cultural experiences represents a paradigm shift comparable to the transition from oral to written culture, one that, if handled with care and deep respect, could help ensure that humanity's richest traditions survive and remain accessible for generations to come.

The Oral Tradition Problem: When Elders Die, Libraries Burn

The Malian writer Amadou Hampate Ba famously told UNESCO in 1960 that in Africa, when an old person dies, it is as if a library has burned down. This metaphor, widely cited in the decades since, captures a truth that extends far beyond the African continent. In oral cultures worldwide, knowledge is not stored in books or databases. It lives in the memories of specific people, encoded in specific narrative forms, transmitted through specific relationships between teacher and student, elder and youth, master and apprentice. The fragility of this transmission chain is the central problem of cultural preservation. Written and recorded archives can capture the surface of oral traditions: the words of a story, the melody of a song, the steps of a dance. What they cannot capture is the living context, the way a grandmother adjusts a story for a particular grandchild, the way a healer selects from a vast mental library of plant remedies based on subtle diagnostic observations, the way a griot modulates a historical narrative based on who is present and what political circumstances prevail. Conventional preservation methods, while valuable, have well-documented limitations. Audio and video recordings freeze performances in time but strip away the interactive, responsive nature of oral transmission (Hennessy et al., 2013). Written transcriptions impose the logic of literacy onto traditions that were never meant to be fixed in text. Museum collections decontextualize objects, removing them from the living practices that gave them meaning. Digital archives improve accessibility but often replicate the static nature of their analog predecessors. The result is what we might call the museum problem: preserved cultures become artifacts, objects of study rather than living practices. A ceremonial mask behind glass tells us very little about the spiritual world it was meant to invoke. A transcribed myth, stripped of the elder's voice and the firelight and the specific audience gathered to hear it, becomes literature rather than living wisdom. This is not to diminish the critical importance of existing preservation efforts. The work of organizations like Survival International, Cultural Survival, the Endangered Languages Project, and countless local initiatives is essential and must continue. But we suggest that the emergence of sophisticated conversational AI creates the possibility of a fundamentally different kind of preservation, one that captures not just content but something closer to the dynamic, interactive, responsive quality of living cultural transmission.

AI as Living Archive: Beyond the Museum Model

The core insight driving this paper is simple but far-reaching: artificial intelligence can transform cultural archives from static repositories into interactive experiences. Rather than reading about a tradition, a learner can converse with an AI that has been deeply enriched with that tradition's knowledge. Rather than watching a video of a ceremony, a learner can step into a reconstructed virtual environment and ask questions of AI characters who embody the roles and knowledge of cultural practitioners. This is not a speculative fantasy. The underlying technologies exist today. Large language models can be fine-tuned on specific corpora of cultural knowledge, including oral histories, sacred texts, ethnographic accounts, and linguistic data. These models can then power conversational AI characters capable of discussing cultural topics with nuance and depth. Virtual environment technologies, from VR headsets to browser-based 3D worlds, can create immersive spaces that evoke the physical contexts of cultural practices. And real-time translation systems can make these experiences accessible across language barriers that would otherwise be insurmountable. What makes this approach different from a chatbot trained on Wikipedia is depth and intentionality. A cultural AI character is not a general-purpose assistant that happens to know some facts about Hinduism or the Maori. It is a purpose-built entity enriched with carefully curated knowledge from primary sources, developed in consultation with cultural practitioners, and designed to convey not just information but something of the texture, the values, and the worldview of a particular tradition. Consider the difference between reading an encyclopedia entry about Zen Buddhism and spending an hour in conversation with an AI character who has been enriched with the koans, the dharma talks, the philosophical texts, and the aesthetic sensibility of the Zen tradition. The former gives you facts. The latter gives you a taste of what it might be like to learn within that tradition, to encounter its particular way of questioning, to sit with its characteristic comfort with paradox and silence. We do not claim that this replaces the experience of studying with a living Zen master. Nothing can. But for the vast majority of people on earth who will never have that opportunity, and for traditions where living masters are growing fewer by the year, the difference between a static archive and a living, conversational AI is profound.

Case Studies and Possibilities

The potential applications of AI-mediated cultural preservation span the full breadth of human cultural diversity. We examine several representative cases. Indigenous oral traditions present perhaps the most urgent case for intervention. The Australian Institute of Aboriginal and Torres Strait Islander Studies (AIATSIS) has documented the rapid decline of Aboriginal languages, with only 13 of approximately 250 languages considered strong (AIATSIS, 2020). Each of these languages carries within it a vast body of knowledge about land, ecology, kinship, and spiritual practice encoded in songlines, dreaming stories, and ceremonial practices that have been maintained for tens of thousands of years. AI systems trained on recorded oral histories, linguistic databases, and ethnographic materials could create conversational experiences that help younger community members reconnect with ancestral knowledge, while also making that knowledge accessible, in appropriate and community-approved ways, to the broader world. Ancient language reconstruction represents another frontier. Projects like the Endangered Languages Project and the work of computational linguists such as those at Google's AI research division have demonstrated that machine learning can assist in deciphering and reconstructing languages with limited surviving texts (Luo et al., 2019). Conversational AI could take this further, creating characters capable of speaking in reconstructed ancient languages, from Sumerian to Classical Nahuatl, allowing learners to practice and engage with these languages in ways that static grammars and dictionaries cannot provide. The world's sacred textual traditions offer an extraordinary opportunity for AI-mediated cultural learning. Hindu philosophical texts, from the Upanishads to the Bhagavad Gita to the vast commentarial traditions of Vedanta, represent thousands of years of sustained intellectual inquiry into the nature of consciousness, reality, and ethical action. Buddhist sutras, Pali canon texts, and the elaborate philosophical systems of Madhyamaka and Yogacara contain insights into the nature of mind that are only now finding resonance in Western cognitive science and psychology (Varela et al., 1991). Sufi poetry, from Rumi and Hafiz to the lesser-known but equally profound traditions of Central Asian and South Asian Sufism, represents one of humanity's great literary and spiritual achievements. These are not curiosities or relics. They are incredibly deep, rich intellectual traditions that have shaped the lives of billions of people. AI characters enriched with these textual traditions could serve as guides and interlocutors, helping learners from any background engage with this wealth of thought in ways that feel alive rather than academic. The West African griot tradition offers a particularly compelling case study. Griots, known as jeli in the Manding languages, are hereditary oral historians, musicians, and cultural advisors whose lineages trace back centuries. They carry within their memories the genealogies of entire communities, the histories of empires, and the moral and philosophical teachings embedded in epic narratives like the Sundiata. As Hale (1998) documented in his foundational study of griots, these practitioners serve as living archives of West African civilization. But the griot tradition, like so many others, faces pressure from urbanization and the decline of the social structures that supported it. An AI system trained on recorded griot performances, historical accounts, and the narrative structures of Manding epic tradition could create a new kind of interactive experience, one that preserves not just the content of griot knowledge but something of the dialogic, responsive quality of griot performance. Japanese cultural arts present yet another dimension of possibility. Traditions such as chado (the tea ceremony), ikebana (flower arrangement), and the martial philosophy of bushido are not merely aesthetic practices. They are embodied philosophical systems that encode specific understandings of time, attention, beauty, and human relationship. The tea ceremony, as Okakura Kakuzo articulated in his influential early twentieth-century work, is a practice of presence and aesthetic sensitivity rooted in Zen Buddhist and Daoist principles. These traditions are maintained by relatively small numbers of dedicated practitioners, and while they are not in immediate danger of extinction, their deeper philosophical dimensions are increasingly inaccessible to both Japanese and global audiences. Virtual environments that reconstruct the physical spaces of these practices, populated by AI characters who can explain and demonstrate their principles, could make these traditions available to a global audience while preserving their essential context.

The Ethics of Sacred Digitization: What Should and Should Not Be Preserved This Way

Any serious discussion of AI-mediated cultural preservation must grapple directly with the ethics of digitizing sacred knowledge. Not all cultural knowledge is meant to be publicly accessible. Many Indigenous traditions include restricted knowledge, ceremonies, songs, and stories that are properly known only to initiated members of specific clans, genders, or age groups. Sacred knowledge in many traditions is not merely private in the Western sense of personal preference. It is understood to carry spiritual power that can be harmful if encountered outside its proper ritual context. The history of Western engagement with Indigenous and non-Western sacred traditions is, to put it plainly, a history of repeated violation. From the theft of ceremonial objects to the publication of secret rituals by early anthropologists, Indigenous peoples have ample reason to distrust outside interest in their sacred knowledge (Smith, 2021). Any AI preservation effort that does not center Indigenous sovereignty over Indigenous knowledge will reproduce these historical harms, regardless of the good intentions behind it. We propose several principles for ethical AI-mediated cultural preservation. First, community consent must be the foundational requirement. No cultural knowledge should be digitized or used to train AI systems without the explicit, informed, and ongoing consent of the communities from which it originates. This consent must be granular: a community might agree to the preservation of certain historical narratives while restricting access to ceremonial knowledge. Second, communities must retain control over how their knowledge is presented, who can access it, and under what conditions. Third, the economic benefits of any AI cultural product must flow back to the source communities. Fourth, cultural preservation AI should be developed in genuine partnership with cultural practitioners, not merely reviewed by them after the fact. These principles are not merely aspirational. They reflect the growing international consensus, expressed in instruments like the United Nations Declaration on the Rights of Indigenous Peoples (UNDRIP) and the work of organizations like the World Intellectual Property Organization (WIPO) on traditional knowledge, that Indigenous and local communities have the right to control access to their cultural heritage (United Nations, 2007).

Immersive Cultural Learning: When We Step Inside Another World

There is a qualitative difference between reading about a culture and experiencing even a simulated version of its physical and social environment. Educational research consistently shows that immersive, experiential learning produces deeper understanding and longer retention than passive information consumption (Kolb, 1984). Virtual reality and immersive environments have been shown to increase empathy, cultural sensitivity, and engagement with unfamiliar perspectives (Herrera et al., 2018). Now imagine combining immersive environments with culturally enriched AI characters capable of real-time conversation. A student studying the Silk Road does not merely read about the caravanserais of Samarkand. She walks through a reconstructed virtual caravanserai, hears the sounds of the marketplace, and converses with AI characters who embody the merchants, scholars, and Sufi mystics who traveled those routes. She asks questions and receives answers drawn from historical sources, in her own language, translated in real time. This is not a replacement for travel or for genuine cross-cultural encounter. But it is something categorically different from a textbook, a documentary, or even a well-designed museum exhibit. It is interactive, responsive, and personal. The learner does not observe a culture from the outside. She steps inside it, however partially, and engages with it on its own terms. The implications for education are significant. Cross-cultural understanding is not merely a nice-to-have in an increasingly interconnected world. It is a strategic necessity for peace, cooperation, and the resolution of global challenges from climate change to migration. Yet most educational systems provide students with only the thinnest exposure to non-Western cultural traditions. AI-mediated immersive cultural experiences could democratize access to deep cross-cultural learning in ways that no previous technology has made possible.

Auto-Translation and the Dream of Cross-Cultural Understanding

One of the most transformative aspects of AI-mediated cultural preservation is real-time translation. Language barriers have historically been the most formidable obstacle to cross-cultural understanding. A Portuguese speaker who wants to learn about the philosophical traditions of Tibet faces years of language study before she can engage with primary sources. A Japanese student interested in the oral traditions of the Yoruba people confronts similar barriers. Recent advances in neural machine translation, combined with the multilingual capabilities of large language models, are beginning to dissolve these barriers (Barrault et al., 2023). AI cultural characters can converse in a learner's native language while drawing on source materials in the tradition's original languages. This is not merely a matter of convenience. It is a fundamental expansion of who can access the world's cultural heritage. The dream here is not the flattening of linguistic diversity into a single global monoculture. It is precisely the opposite: making the depth and richness of diverse traditions accessible across linguistic boundaries so that their value can be recognized, appreciated, and sustained. When a farmer in rural Brazil can converse with an AI character enriched with the philosophical insights of the Pali canon, or when a teenager in Lagos can explore the aesthetic principles of Japanese wabi-sabi through an interactive virtual experience, something genuinely new becomes possible in the history of human cultural exchange. We acknowledge the significant challenges of translation, particularly for culturally embedded concepts that resist direct equivalence. No translation of the Sanskrit term dharma or the Japanese concept of mono no aware fully captures the experiential depth of these ideas in their original linguistic contexts. But a conversational AI can do what a dictionary cannot: it can explain, contextualize, offer examples, and engage in the kind of extended dialogue through which understanding of complex cultural concepts actually develops.

From Museum to Living World: Reconstructed Virtual Environments

The museum model of cultural preservation, for all its value, embodies a fundamental limitation. It removes cultural artifacts and practices from the environments that gave them meaning. A totem pole in a museum gallery is separated from the forest, the village, the ceremonies, and the social relationships that constituted its living context. A recorded performance of Balinese gamelan, however faithfully captured, is separated from the temple, the ritual occasion, and the community of participants for whom it was a shared spiritual experience. Virtual environment technology offers the possibility of reconstructing not just individual artifacts or performances but entire cultural contexts. Why have a museum case with a single ceramic vessel when you can reconstruct the workshop where it was made, the market where it was sold, and the household where it was used, all populated by AI characters who can explain the cultural significance of each element? This is not merely a matter of adding visual context. It is about reconstructing the relational, social, and spatial dimensions of cultural practice that static preservation methods inevitably strip away. A virtual reconstruction of an ancient Greek agora, populated by AI characters who can discuss philosophy, politics, commerce, and daily life, conveys something about the integrated nature of Greek civic culture that no collection of artifacts or texts can fully communicate. The technical requirements for such reconstructions are substantial but increasingly feasible. Photogrammetry, LiDAR scanning, and AI-assisted 3D reconstruction can create detailed virtual environments from archaeological sites and surviving structures (Remondino et al., 2011). Historical and ethnographic research provides the basis for populating these environments with accurate cultural detail. And conversational AI provides the crucial element that transforms a virtual environment from a digital diorama into a living world: the ability to talk to someone who is there.

The Vision: AI Characters as Cultural Ambassadors

At the heart of this paper is a vision of AI characters that serve as cultural ambassadors, entities enriched with deep knowledge of specific cultural traditions, capable of engaging learners in meaningful conversation, and situated within immersive environments that evoke the physical and social contexts of those traditions. These are not chatbots that recite facts. They are carefully crafted characters with distinct perspectives, personalities, and areas of expertise, designed to convey not just the content of a cultural tradition but something of its texture, its values, and its way of seeing the world. A Sufi poet character does not merely summarize Rumi. It speaks in a way that reflects the Sufi understanding of the relationship between the human soul and the divine. A character embodying the Confucian scholarly tradition does not merely list the Analects. It engages with questions of ethics, governance, and human relationship in a way that reflects the Confucian emphasis on cultivated moral character and social harmony. The development of such characters requires a fundamentally interdisciplinary approach. It demands collaboration between AI engineers, cultural anthropologists, historians, linguists, religious studies scholars, and most importantly, living practitioners of the traditions being represented. It requires humility about the limitations of any digital representation and honesty about what is gained and lost in the translation from living tradition to AI-mediated experience. But if done well, such characters could play a role in cultural preservation and cross-cultural understanding that no previous technology has been able to fill. They could bring the world's deepest traditions to life for audiences who would otherwise never encounter them. They could help younger members of endangered cultural communities reconnect with ancestral knowledge. And they could foster the kind of deep, empathetic cross-cultural understanding that the twenty-first century desperately needs.

Risks and Concerns: Appropriation, Flattening, Loss of Context

We would be irresponsible to present this vision without a frank accounting of the risks. The history of technology's engagement with non-Western cultures is not encouraging. From the exoticization of Indigenous peoples in early cinema to the reduction of complex spiritual traditions to New Age consumer products, the pattern of appropriation, simplification, and decontextualization is well established (Deloria, 1998). AI-mediated cultural preservation faces several specific risks. The first is flattening: the reduction of complex, internally diverse traditions to simplified, easily consumable versions. Hinduism is not one thing. Buddhism is not one thing. Indigenous Australian culture is not one thing. Any AI representation must resist the temptation to present a single authoritative version of traditions that are, in reality, characterized by profound internal diversity and ongoing debate. The second risk is decontextualization. Cultural practices derive their meaning from specific social, ecological, and historical contexts. A tea ceremony AI character that teaches the physical movements of chado without conveying the decades of disciplined practice, the specific teacher-student relationship, and the Zen philosophical framework that gives those movements meaning is not preserving a tradition. It is creating a simulacrum. The third risk is commodification. There is a real danger that AI cultural characters become products to be consumed rather than traditions to be respected. The line between cultural education and cultural tourism is not always clear, and the commercial incentives of the technology industry do not naturally favor the kind of slow, deep, respectful engagement that genuine cross-cultural understanding requires. The fourth risk is the displacement of living practitioners. If people can learn about a tradition from an AI, will they still seek out human teachers? Will funding for living cultural programs be redirected to cheaper digital alternatives? These are not hypothetical concerns, and the technology community has a responsibility to ensure that AI preservation efforts complement rather than replace support for living cultural practitioners. Finally, there is the risk of error propagation. AI systems can generate plausible-sounding but inaccurate information, and when that information concerns sacred traditions, the consequences of error are not merely academic. Misrepresenting a spiritual practice or a community's history is a form of harm that must be taken seriously in the design and deployment of cultural AI systems (Bender et al., 2021). Addressing these risks requires ongoing engagement with source communities, transparent processes for identifying and correcting errors, clear communication to users about the limitations of AI-mediated cultural experiences, and a genuine commitment to the principle that technology serves culture, not the other way around.

Conclusion: A New Kind of Sacred Responsibility

The traditions that are disappearing from the world today are not curiosities of the premodern past. They are living repositories of human wisdom, developed over centuries and millennia, that address the deepest questions of human existence: how to live, how to die, how to relate to one another and to the natural world, how to find meaning in suffering, how to cultivate joy. The Western intellectual tradition, for all its extraordinary achievements in science and technology, represents one way of answering these questions. The traditions of India, China, Japan, Africa, the Indigenous peoples of the Americas and Oceania, the Islamic world, and countless other cultures represent others, no less rigorous, no less profound, and in many cases addressing dimensions of human experience that Western thought has systematically neglected. We have, in AI and immersive technology, tools that could help preserve these traditions as living, accessible, interactive bodies of knowledge rather than dusty archives. We could build virtual worlds where anyone, regardless of geographic location or linguistic background, can step into the environment of another culture and engage with its deepest ideas through conversation with AI characters enriched with centuries of accumulated wisdom. We could use auto-translation to dissolve the language barriers that have kept the world's great traditions siloed from one another. We could give younger generations of endangered cultural communities new ways to connect with ancestral knowledge. But the possibility of doing these things does not mean we will do them well. The gap between the vision we have outlined and the commercial reality of the technology industry is vast. Doing this work properly requires patience, humility, deep collaboration with cultural communities, and a willingness to let those communities lead. It requires investment in the hard, slow work of building culturally accurate AI systems rather than the quick deployment of superficially impressive but ultimately hollow products. It requires treating cultural knowledge not as data to be harvested but as heritage to be honored. This is, we believe, a sacred responsibility in the most literal sense. Not sacred in the colloquial sense of merely important, but sacred in the sense that the knowledge we are entrusted with has been held sacred by the communities that created and maintained it. To handle it carelessly would be a betrayal not just of those communities but of the human capacity for depth, diversity, and meaning that their traditions represent. The technology is here. The need is urgent. The question is whether we will rise to the responsibility.

References

AIATSIS. (2020). National Indigenous Languages Report. Australian Institute of Aboriginal and Torres Strait Islander Studies. Canberra. Austin, P. K., & Sallabank, J. (2011). The Cambridge Handbook of Endangered Languages. Cambridge University Press. Barrault, L., et al. (2023). SeamlessM4T: Massively Multilingual and Multimodal Machine Translation. Meta AI Research. Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021). On the dangers of stochastic parrots: Can language models be too big? Proceedings of FAccT 2021. Deloria, P. J. (1998). Playing Indian. Yale University Press. Hale, T. A. (1998). Griots and Griottes: Masters of Words and Music. Indiana University Press. Hennessy, K., Lyons, N., Loring, S., Arnold, C., Joe, M., Elias, A., & Pokiak, J. (2013). The Inuvialuit Living History Project: Digital return as the circulation of knowledge. Museum Management and Curatorship, 28(3), 283-299. Herrera, F., Bailenson, J., Weisz, E., Ogle, E., & Zaki, J. (2018). Building long-term empathy: A large-scale comparison of traditional and virtual reality perspective-taking. PLOS ONE, 13(10), e0204494. Kolb, D. A. (1984). Experiential Learning: Experience as the Source of Learning and Development. Prentice Hall. Luo, J., Cao, Y., & Barzilay, R. (2019). Neural decipherment via minimum-cost flow: From Ugaritic to Linear B. Proceedings of the 57th Annual Meeting of the ACL. Moseley, C. (Ed.). (2010). Atlas of the World's Languages in Danger (3rd ed.). UNESCO Publishing. Okakura, K. (1906). The Book of Tea. Putnam. Remondino, F., Barazzetti, L., Nex, F., Scaioni, M., & Sarazzi, D. (2011). UAV photogrammetry for mapping and 3D modeling. International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences, 38(1/C22). Smith, L. T. (2021). Decolonizing Methodologies: Research and Indigenous Peoples (3rd ed.). Zed Books. United Nations. (2007). United Nations Declaration on the Rights of Indigenous Peoples (UNDRIP). UN General Assembly Resolution 61/295. Varela, F. J., Thompson, E., & Rosch, E. (1991). The Embodied Mind: Cognitive Science and Human Experience. MIT Press. World Intellectual Property Organization. (2022). Intellectual Property and Genetic Resources, Traditional Knowledge and Traditional Cultural Expressions. WIPO. Zuckermann, G. (2020). Revivalistics: From the Genesis of Israeli to Language Reclamation in Australia and Beyond. Oxford University Press.