Mr. Tumbles' Maya and the Multiplicity of Personas: Leveraging Illusion and Reality in Educational Game Design for SEN Language Learners

This paper argues that the CBeebies programme Mr. Tumble offers a practical model for designing language-learning games for children with special educational needs (SEN). Its central thesis draws on the Advaita Vedanta concept of Maya, the illusion of multiplicity, to propose that a single non-player character (NPC) embodying many roles can generate believable, emotionally resonant interactions and parasocial relationships. Such relationships, the author suggests, can lower the affective barriers that often impede language learning for SEN children. The work situates this proposal within a literature that has paid limited attention to believability, emotion, and social relationship models in games, and notes that UK language-learning tools for SEN learners remain comparatively underdeveloped.

The main design principles emphasised are a single consistent protagonist, contextually grounded real-world activities, multimodal engagement such as Makaton signing, and emotionally responsive NPCs. These recommendations are supported by reference to research on personas, parasocial relationships, and self-efficacy, though the argument rests largely on theoretical reasoning rather than new empirical evidence. The author acknowledges several risks, including idealised expectations, parasocial confusion, overreliance on NPCs, and the difficulty of gathering the detailed cognitive data needed to personalise designs for the heterogeneous SEN population.

The paper's final contention is that NPC design should incorporate cognitive and knowledge-related profiles rather than relying solely on demographic and behavioural data. It proposes distinct NPC personas tailored to subgroups such as learners with autism, ADHD, dyslexia, and intellectual disabilities. Drawing on make-believe play as an established framework for language development, including perspective-taking, vocabulary expansion, narrative construction, pragmatic skill, and metacognition, the author expects similar benefits from well-designed NPCs. These expectations are presented as plausible rather than demonstrated, as the paper offers no empirical data from implemented NPC systems to test them directly.