Conversing Using WhatsHap: a Phoneme Based Vibrotactile Messaging Platform

We demonstrate the feasibility and experience of having a haptic conversation using WhatsHap: an instant messaging system that delivers speech or text as a sequence of vibrotactile representations of English phonemes to the arm. Previous haptic speech communication studies established feasibility in single-phoneme or word-level encodings, but did not investigate how such communication functions in practice with real-time remote conversation between two individuals. Participants used WhatsHap through the framework of a joint communication task, where they had to converse to achieve a goal, with 88% of all tasks successfully completed. We analyze conversations and user interviews both qualitatively and quantitatively, describing considerations when building a system to mediate conversation haptically, exploring influences on user conversational experience, and offering an account of how linguistic structure changes to accommodate such a mode of communication. In this regard, phoneme-based haptic conversation led to linguistic forms distinct from written and spoken English. Additionally, participants felt that haptic conversation was best suited for information-centered communication in contexts where there is shared knowledge between users.