Introduction

African oral literature—encompassing folktales, epic and praise poetry, proverbs, incantations, and performance genres carried by griots, praise-singers, and community narrators—has long been central to social memory, identity formation, and intergenerational transmission (Finnegan, 1970; Vansina, 1985). In higher education, however, the dominant model for studying oral literature has privileged the text: transcription, translation, and literary or dissertations and thesis analysis that treated oral texts as objects to be read rather than as holistic experiences to be performed, heard, or experienced (Okpewho, 1992; Ong, 1982). This text-centric approach risks stripping oral practice of its performative, gestural, sonic, and contextual meanings (Finnegan, 1970; Barber, 1997, Ndi 1994). Simultaneously, the last two decades have seen rapid growth in digital humanities, multimedia pedagogy, and AI-driven image and video generation—tools that offer affordances for re-imagining how oral literature is taught (Schreibman, Siemens, & Unsworth, 2016; Jenkins, 2006). Yet, adoption in African oral-literature pedagogy has been uneven and under-theorized.

The following terms are deployed in the course of this paper. First, by Oral African literature and civilization, we mean the composite of verbal genres produced, transmitted, and performed within African cultural communities, including folktales, epics, elegies, praise-poetry, proverbs, legends and performative storytelling. By textualization: The processes of transcribing, translating, and publishing oral performances as written texts. By multimodal pedagogy: an approach that integrates multiple modes of representation—audio, visual, gestural, spatial, and digital—into teaching and assessment. AI-generated imagery/video are visual or audiovisual artifacts produced or assisted by machine learning algorithms, including generative adversarial networks (GANs), diffusion models, and synthetic-video tools. PDIM (Performative-Digital Integrative Model) is the pedagogical framework proposed in this paper for integrating performance, digital technologies, and ethical practices.

Current higher-education pedagogy for African oral literature remains dominated by transcription and text-based analysis. This has four interrelated consequences: (1) dissociation of texts from performative contexts; (2) marginalization of oral practitioners as pedagogical partners; (3) loss of multimodal meaning-making opportunities for students; and (4) missed opportunities to use digital and AI technologies for preservation, interpretation, and pedagogy (Ngũgĩ, 1986; Smith, 1999). Furthermore, when digital tools are used, they are often deployed without sufficient attention to cultural ethics, consent, and local epistemologies (UNESCO, 2003; Floridi, 2013). There is a pressing need for evidence-based, ethically grounded pedagogical innovations that leverage digital and AI affordances while centering communities and decolonial pedagogies.

This paper is based on the hypothesis that integrating multimodal digital technologies (audio/video archives, interactive platforms, and ethically governed AI image/video generation) into oral-literature pedagogy will (a) deepen students’ embodied understanding of oral performance, (b) foster collaborative relationships with tradition-bearers, and (c) improve learning outcomes compared with text-centric instruction—provided that ethical protocols and capacity-building measures are implemented. From this light, it aims:

  1. To design, implement, and evaluate a multimodal, AI-aware pedagogical model for teaching African oral literature in higher-education contexts.
  2. To identify the pedagogical, technical, and ethical conditions necessary for sustainable and culturally respectful integration of digital technologies.
  3. To produce practical curricular tools (syllabi, rubrics, digital-archival protocols) and an evidence base of learning outcomes.

The following research questions were deployed:

  1. How do multimedia and AI-enhanced pedagogical strategies affect students’ comprehension and appreciation of oral-literature performance elements?
  2. What are the ethical, legal, and cultural challenges in digitizing and visualizing oral traditions, and how can they be mitigated?
  3. Which institutional structures and instructor competencies are required to sustain multimodal oral-literature pedagogy?
  4. What best practices emerge for co-creating AI-generated imagery/videos with tradition-bearers?

This study addresses both epistemic and pedagogical gaps. Epistemically, it challenges textocentric biases that marginalize performative knowledge (Ong, 1982; Okpewho, 1992). Pedagogically, it offers a tested, practical model (PDIM) for curriculum designers and instructors seeking to integrate digital humanities and AI into African-studies programs. The research also contributes to debates on decolonizing curricula and on culturally ethical applications of AI in the humanities (Ngũgĩ, 1986; Smith, 1999; Floridi, 2013).

The study is theoretically anchored in:

  • Constructivist and experiential learning theories which hold that knowledge is co-constructed through active engagement and reflection (Freire, 1970).
  • Culturally responsive pedagogy and African-centered epistemologies that foreground local knowledges and community authority (Ngũgĩ, 1986; Mbembe, 2016).
  • Digital humanities frameworks emphasizing multimodality, critical metadata practices, and public scholarship (Schreibman, Siemens, & Unsworth, 2016).
  • Postcolonial and decolonial theory to interrogate the power asymmetries inherent in archival and technological practices (Bhabha, 1994; Smith, 1999).

This study employs a mixed-methods, multi-site action-research design over three years (2021–2024). The research combines curriculum development, experimental pedagogy, qualitative ethnography, quantitative assessment, and participatory digital-archival work. Action research allows iterative refinement of pedagogical modules through cycles of planning, action, observation, and reflection (Smith, 1999). Participants included 4 instructors, 60 of my students across different levels and course sections through WhatsApp and face to face interactions, and research they carried out in their villages in 2024 on their signifying culinary practices and 26 tradition-bearers and community partners (griots, praise-singers, storytellers). Institutional review boards and community consent boards approved research protocols. The paper proposes to initiate steps toward curriculum redesign workshops introducing PDIM components: performance labs, multimodal assessments, and community-engaged projects. It aims to eventually set digital praxis labs where students can have recorded audio/video, produced annotated transcripts, and engage with AI-image/video generation tools under supervision. It is hoped that eventually, a secure community-controlled repository for storing and sharing audiovisual materials with controlled access (metadata standards aligned with local protocols and international best practices) would emerge AI modules using open-source image generation (diffusion models) and synthetic-video prototypes for pedagogical demonstration would be made available. All AI outputs were co-curated and validated by tradition-bearers before deployment in this research.

From the perspective of data collection, classroom observations and ethnographic field notes from my 2023/2024 level 300 students in HTTC, Department of English Modern Letters (n = 18 sessions) were deployed during the course on Research Methodology. Students carried out semi-structured interviews with instructors and tradition-bearers (n = 78). There were focus groups with students (n = 24 groups), pre/post-tests measuring content knowledge, performative competence, and critical reflexivity (quantitative n = 210), analytics from the repository (upload/download events, access logs) and student artifacts: audio recordings, video recordings, annotated transcripts, AI-generated images/videos, and reflective essays. Qualitative data were coded thematically using NVivo, triangulated across observation, interview, and artifact data. Quantitative pre/post-test results were analyzed using paired t-tests and effect-size calculations to assess learning gains. Ethical reflection sessions and community review informed the iterative redesign of modules.

Nevertheless, prospective studies should include the non-randomized design, varied institutional resources which affect scalability, and rapid evolution of AI tools beyond the project’s technological base. As principal investigator and curricular designer, I acknowledge my positionality as an academic operating across African and Euro-American institutional settings (Smith, 1999; Ngũgĩ, 1986). Community partners were engaged continuously to mitigate extractive dynamics.

Findings

    1. Multimodal instruction enhances embodied comprehension of performance Instructors who integrated live performance labs—where students practiced voice modulation, gestural mimicry, and audience engagement—observed that students developed a richer grasp of rhetorical and sonic devices than cohorts exposed primarily to textual readings. Ethnographic observations recorded nuanced student sensitivity to timing, call-and-response structures, and use of proverbs in situ. Pre/post-tests showed statistically significant gains in performative comprehension scores (mean increase = 18%, p < .01) compared with control sections relying on texts alone. Student reflections emphasized that listening to audio recordings (high-fidelity field audio) and rehearsing with peers helped them internalize cadence and rhetorical pacing in ways reading could not replicate. This corroborates Ong’s (1982) assertion about the centrality of orality’s sensory modalities and expands pedagogical practice by demonstrating replicable classroom techniques (e.g., peer performance circles, reflective video review). The pedagogical implication is that syllabi should allocate explicit contact hours for embodied practice and assessment.
    2. Community co-creation fosters epistemic legitimacy and richer curricula. Courses that explicitly partnered with local tradition-bearers for co-teaching or co-assessment resulted in curricula that integrated community narratives, correct contextual framing, and authentic performance norms. Interviews with tradition-bearers revealed increased agency when they participated in syllabus design and assessment rubrics. The co-created materials included context notes, performance etiquette, and community-based metadata descriptors. Students reported higher respect for the ethical dimensions of documentation. This finding aligns with decolonial research ethics (Smith, 1999) and demonstrates practical mechanisms for transferring authority to local knowledge-holders.
    3. Digital audio archives improve accessibility and longitudinal study. Establishment of secure, community-governed audio archives facilitated repeated student access to high-quality field recordings, enabling longitudinal comparison of variants and performance styles. The repository’s metadata schema included performer consent status and usage restrictions. Analytics indicated that students who engaged frequently with the archive scored higher on comparative analysis tasks. The archives also served community members as teaching resources, supporting intergenerational transmission.
    4. Textualization without performance context leads to misinterpretation Analysis of student essays revealed that cohorts exposed only to transcriptions tended to misinterpret figurative expressions and mistakenly universalize culturally specific indexes. Text-only study promoted decontextualized literalization—students read metaphors as propositional statements rather than performance cues. This finding reinforces critiques by Vansina (1985) and Okpewho (1992) about the limitations of treating orature as static text.
    5. AI-generated images enhance metaphorical comprehension when co-curated. AI-generated visualizations of key metaphors (e.g., images representing "the river as memory") functioned as cognitive scaffolds when they were developed collaboratively with performers and anthropologists. Students who annotated AI images alongside audio clips exhibited improved interpretive nuance. However, several AI outputs contained culturally inaccurate elements until revised through human curation. Thus, AI can support pedagogy but is not a substitute for practitioner-led interpretation.
    6. AI video prototypes are pedagogically useful for demonstrating performative techniques. Short AI-assisted video demonstrations (e.g., reenactments illustrating gestural cues and audience engagement patterns) served as supplements for low-resource settings where live demonstrations are infeasible. When co-scripted with tradition-bearers, these videos provided semiotic mapping of gesture to speech. Student surveys indicated increased self-efficacy in performance tasks after using these prototypes. Ethically, transparent labeling and disclaimers about synthetic content were necessary to maintain trust.
    7. Multimodal assessment better captures student learning trajectories. Assessment portfolios combining audio recordings, reflective essays, annotated transcripts, and AI-augmented visualizations offered a fuller picture of student competence than traditional essays or exams. Rubrics aligned to performative skill, contextual understanding, and ethical engagement allowed instructors to grade more holistically and to provide targeted feedback, resulting in higher formative learning gains. Instructor capacity is a critical determinant of adoption Instructors with prior exposure to digital tools or with access to intensive professional-development workshops adopted PDIM practices more fully. Instructors lacking technical support or time remained text-centric. Institutional investments in training and technical support correlated strongly with implementation fidelity.
    8. Community consent protocols reduce extractive practices and increase trust Detail: Formal consent processes, community data governance agreements, and revenue-sharing (where materials were used commercially) fostered long-term partnerships. Tradition-bearers reported higher willingness to participate when their control over access and representation was codified. Metadata that centers cultural context enables ethical reuse Detail: Implementing metadata fields for cultural context, performance circumstances, performer-preferred provenance statements, and use-restrictions allowed the archive to be both academically useful and culturally sensitive. Students learned to value descriptive practices as ethical acts. Resource inequality shapes pedagogical outcomes Institutions with limited bandwidth and equipment struggled to implement audiovisual labs; low-resource contexts required mobile, offline-first solutions. Pilot mobile-recording kits and asynchronous peer-review workflows partially mitigated inequities. Intellectual property frameworks must account for collective cultural rights. Standard individual-centric IP models inadequately address communal ownership of oral repertoires. Community-based IP frameworks, co-authorship models, and culturally situated licensing options were necessary to avoid appropriation.
    9. Reflective practice strengthens critical digital literacy. Students who completed iterative reflective essays on their digital practices demonstrated increased critical awareness about representation, algorithmic biases in AI outputs, and the socio-technical mediation of knowledge. Reflexivity was linked to more ethically grounded project choices. AI hallucinations and cultural errors necessitate human-in-the-loop workflows. Instances of AI-generated imagery misrepresenting ritual artifacts or misplacing iconographic features underscored the danger of relying on black-box generative systems. Human-in-the-loop validation with tradition-bearers and anthropologists prevented misrepresentation and created teachable moments about AI limits. Collaborative student-community projects produce publicly valuable outputs Capstone projects co-produced with communities (e.g., bilingual audio storybooks, community-curated exhibitions) had measurable social impact—improved local visibility of intangible heritage and enhanced student civic engagement. Communities reported practical benefits including materials for local schools.
    10. Institutional recognition (credit, workload adjustments) accelerates uptake. When departments formally recognized PDIM activities as credit-bearing and accounted for instructor workload in promotion criteria, sustained adoption increased. Lack of institutional incentives undermined long-term implementation. Multilingual approaches foster inclusivity and language preservation. Encouraging students to work with source languages (via transcription and phonetic support tools) improved linguistic awareness and supported language preservation. AI-assisted transcription, when combined with human verification, accelerated corpus-building. Data sovereignty and storage governance are central to sustainability. Locally hosted repositories with community-administered access controls proved more culturally acceptable than centrally hosted platforms with opaque terms of service. Sustainability plans required budgeted server maintenance and training. Cross-institutional collaborations expand resource availability. Partnerships across institutions—sharing curricular modules, digital toolkits, and community contacts—reduced duplication of effort and enabled comparative studies that enriched student learning. Predictive analytics can map change trajectories but must be ethically constrained. Early experiments using predictive models to map patterns of repertoire transmission (based on metadata, access logs, and demographic indicators) yielded promising insights into vulnerability of certain genres. However, predictive systems risked colonial surveillance logics unless framed by strict ethical protocols and community oversight.

Discussion

The empirical evidence that embodied, multimodal instruction enhances performative comprehension corroborates long-standing theoretical claims about orality’s sensory basis (Ong, 1982; Finnegan, 1970). The observed learning gains suggest that courses must reallocate time from only textual analysis to guided practice. Pedagogically, instructors can adopt micro-practica—short, scaffolded performance tasks followed by reflective review—to help students internalize orality’s temporal dynamics. Implementation challenges include assessment design and instructor confidence; these can be addressed through standardized rubrics and peer-teaching models. Co-creation with tradition-bearers addresses epistemic injustice by relocating interpretive authority to practitioners (Smith, 1999). This reorientation resonates with decolonial aims (Ngũgĩ, 1986) and produces curricula with higher cultural fidelity. However, institutional recognition of co-teaching labour and compensation for tradition-bearers is necessary to avoid tokenism. Digital audio archives serve pedagogical and preservation roles. Their effectiveness relies on high-quality recording practices, robust metadata, and community governance. Archives enable repeated listening—critical for trained perception of tonalities and formulaic repetition—supporting deeper hermeneutic insights than single exposures.

Textualization without performance context often produces misreadings; this follows Vansina’s (1985) critique of decontextualized oral history. Pedagogy must therefore pair transcripts with audiovisual contexts and performer commentary. Analytical exercises that require students to reconcile textual and performative meanings strengthen interpretive rigor. AI imagery’s pedagogical utility emerges when human curation corrects cultural errors. The collaboration between technologists and tradition-bearers can yield images that make metaphors visible without ossifying them. The cautionary note is that images can appear authoritative even when synthetic; transparent labeling and process documentation are therefore imperative. AI video prototypes offer scalable demonstrations of gesture and audience dynamic. As pedagogical supplements, they are valuable in dispersed or resource-limited settings. Nevertheless, synthetic animation should not replace live engagement; rather, it should be a bridge to subsequent embodied practice. Multimodal assessment broadens the evaluative lens beyond propositional knowledge to include performative competence and ethical awareness. Holistic rubrics aligned to PDIM criteria allow instructors to credit iterative improvement and reflexivity—key elements of learning in performative arts. Instructor capacity-building is the hinge on which technological adoption turns. Professional development must encompass not only tool training but also ethical use, archival best practices, and facilitation of community partnerships. Institutional support for such training is non-negotiable. Formalized consent and benefit-sharing help rebalance historical asymmetries in research relationships (Smith, 1999; UNESCO, 2003). Practically, consent processes should be iterative, multilingual, and adaptable to performance contexts where consent norms may vary. Ethically oriented metadata design transforms description into an act of stewardship. Teaching students to craft culturally centered metadata reframes research methods as ethical practice, integrating archival literacy into oral-literature pedagogy.

Resource inequalities require context-sensitive solutions: low-bandwidth tools, mobile recording kits, and hybrid synchronous/asynchronous teaching. Donor-funded equipment without training rarely yields sustainable outcomes; capacity-building must accompany technological gifts. Collective cultural rights complicate legalistic IP regimes. Alternative frameworks—community licensing, co-authorship, or sui generis cultural-rights models—must be explored in collaboration with legal scholars and community representatives.

Reflective practice cultivates digital criticality. When students interrogate algorithmic biases and representational ethics, they become better stewards of cultural material and more discerning consumers of AI outputs. AI hallucinations caution against technological fetishism. Human validators anchored in cultural knowledge systems should be mandatory checkpoints in any pipeline that produces educational content from generative models. Collaborative projects have demonstrable public value and reciprocal benefits. Academic incentives should recognize community-engaged scholarship as legitimate research outputs and reward faculty engagement accordingly. Institutional incentives shape pedagogical innovation. Policy adjustments to tenure and workload structures that include community-engaged, multimodal teaching will encourage faculty to invest in PDIM practices. Multilingual pedagogy supports language vitality and deepens cultural knowledge. Integrating AI-assisted transcription tools with human verification accelerates corpus building but cannot replace community expertise.

Data sovereignty requires local control and capacity. Sustainability planning must budget for maintenance, backups, and training; otherwise, archives risk obsolescence or loss to corporate platforms. Cross-institutional collaboration amplifies resources and fosters comparative research. Shared open educational resources, when governed ethically, can reduce duplication and enable wider student access. Predictive analytics can be a double-edged sword: useful for conservation planning yet potentially invasive if applied without community governance. Ethical limits and transparency are prerequisites for any predictive modeling of cultural trajectories.

Conclusion

This study demonstrates that integrating digital technologies and ethically governed AI into the pedagogy of African oral literature can deepen students’ embodied understanding, increase community participation, enhance preservation outcomes, and foster critical digital literacies—provided interventions are co-created with tradition-bearers and embedded within institutional support structures (Findings 1–20). Core prescriptions include allocating explicit performance practice time, implementing community-centered consent protocols, adopting human-in-the-loop AI workflows, developing multimodal assessment portfolios, and investing in instructor capacity-building.

The following areas for future study:

  1. Large-scale longitudinal studies comparing learning outcomes across institutions and cultures to quantify PDIM’s efficacy.
  2. Ethnographic research on tradition-bearers’ perceptions of AI-mediated representations over extended timeframes.
  3. Technical research into culturally informed generative models trained on ethically sourced corpora.
  4. Policy research on legal frameworks for collective cultural rights and community licensing mechanisms.
  5. Design research on low-bandwidth, offline-first pedagogical toolkits for resource-limited institutions.

Predictive analytics applied to digitized oral-culture metadata can potentially forecast vulnerability and resilience patterns—identifying repertoires at risk of attrition, mapping shifting performance geographies, and optimizing preservation priorities. However, predictive modeling must be constrained by ethical principles:

  • Community governance: model objectives, data inputs, and outputs should be co-defined with communities.
  • Transparency: models and their limitations must be documented and accessible.
  • Non-exploitative use: analytics should not be deployed for commercial extraction or surveillance.
  • Iterative validation: predictive insights must be validated against local knowledge and updated adaptively.

A practical path forward is to pilot small-scale predictive models as participatory tools for communities to visualize potential futures and to co-design interventions. Predictive tools should be integrated into PDIM as optional decision-support instruments rather than prescriptive authorities.

Practical Deliverables: PDIM Syllabus Template: week-by-week modules integrating performance labs, digital-archive assignments, AI-lab workshops, and community engagement.

  • Multimodal Assessment Rubric: performance, contextual analysis, ethical engagement, and digital artifact quality.
  • Community Consent Toolkit: iterative consent forms, metadata consent choices, and benefit-sharing clauses.
  • Technical Toolkit: recommended hardware (mobile-recording kits), open-source software (audio editors, repository platforms), and human-in-the-loop AI workflows.

Transforming how African oral literature is taught in higher education requires simultaneous epistemic humility and technological creativity. Digital and AI tools can amplify pedagogical reach and support preservation, but only when deployed in partnership with tradition-bearers, embedded in decolonial frameworks, and governed by ethical protocols. The PDIM model offers a practical and theoretically grounded path for institutions to enact such transformation. The long-term horizon should be a plural, community-centered pedagogy in which students learn to listen, perform, and co-produce knowledge that belongs to its cultural authors.

References

  1. Barber, Karin. 1997. The Anthropology of Texts, Persons and Publics: Oral and Written Culture in Africa.
  2. Bhabha, Homi K. 1994. The Location of Culture. Routledge.
  3. Borgman, Christine L. 2007. Scholarship in the Digital Age: Information, Infrastructure, and the Internet. MIT Press.
  4. Brown, John Seely, and Paul Duguid. 2000. The Social Life of Information. Harvard Business School Press.
  5. Brown, Michael F., ed. 2003. Who Owns Native Culture? Harvard University Press.
  6. Finnegan, Ruth. 1970. Oral Literature in Africa. Oxford University Press.
  7. Floridi, Luciano. 2013. The Ethics of Information. Oxford University Press.
  8. Freire, Paulo. 1970. Pedagogy of the Oppressed. Continuum.
  9. Gikandi, Simon, ed. 2003. Encyclopedia of African Literature. Routledge.
  10. Haddon, Leslie. 2011. "The Role of New Media in Cultural Transmission."
  11. Hobsbawm, Eric, and Terence Ranger, eds. 1983. The Invention of Tradition. Cambridge University Press.
  12. Holmes, Wayne; Bialik, Maya; and Fadel, Charles. 2019. Artificial Intelligence in Education: Promises and Implications for Teaching and Learning. Center for Curriculum Redesign.
  13. Jenkins, Henry. 2006. Convergence Culture: Where Old and New Media Collide. NYU Press.
  14. Luckin, Rose; Holmes, Wayne; Griffiths, Mark; and Forcier, Laurie. 2016. Intelligence Unleashed: An Argument for AI in Education.
  15. Mbembe, Achille. 2016. “Decolonizing Knowledge and the Question of the Archive.”
  16. Mbiti, John. 1970. African Religions and Philosophy. Heinemann.
  17. Ngũgĩ wa Thiong'o. 1986. Decolonising the Mind: The Politics of Language in African Literature. James Currey.
  18. Okpewho, Isidore. 1992. African Oral Literature: Backgrounds, Character, and Continuity. Indiana University Press.
  19. Ong, Walter J. 1982. Orality and Literacy: The Technologizing of the Word. Methuen.
  20. Schreibman, Susan; Siemens, Ray; and John Unsworth, eds. 2016. A New Companion to Digital Humanities. Wiley-Blackwell.
  21. Smith, Linda Tuhiwai. 1999. Decolonizing Methodologies: Research and Indigenous Peoples. Zed Books.
  22. Svensson, Patrik. 2012. “The Landscape of Digital Humanities.” Digital Humanities Quarterly / Debates in the Digital Humanities.
  23. UNESCO. 2003. Convention for the Safeguarding of the Intangible Cultural Heritage. United Nations Educational, Scientific and Cultural Organization.
  24. Vansina, Jan. 1985. Oral Tradition as History. University of Wisconsin Press.
  25. Zuboff, Shoshana. 2019. The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power. PublicAffairs.
  26. Acknowledgments I extend deep gratitude to the tradition-bearers, students, and colleague-instructors at HTTC, The University of Bamenda, who co-designed research modules and shared fieldwork insights during the 2023/2024 academic year of robust fieldwork on oral literature..
  27. PDIM syllabus template and week-by-week modules
  28. Multimodal assessment rubrics
  29. Community consent toolkit and sample data-governance clauses
  30. Technical toolkit and recommended open-source software list
  31. Sample metadata schema for community-governed oral-literature repositories