Towards a Framework for Exploring Synthetic Voices in VUI Design

The last decade has seen voice assistants like Apple's Siri and Google's Home become truly ubiquitous. These devices – and many more – rely on Voice-based User Interfaces (VUIs) that specify what the devices can listen for and how they express themselves through synthetic voice design. Today, primarily Western VUI designs often reproduce anthropomorphic and heteronormative vocal stereotypes and interaction scenarios. Instead, we propose an emergent framework for VUI design that opens for a multifaceted exploration of how VUIs sound, what this means for their sociocultural role and how it shapes interactions. The framework extends current VUI design concerns with concepts from sound studies that offer a richer foundation for exploring VUIs’ sonic expressions. We present the theoretical foundations of the framework and how it has developed over three collaborative workshops bringing together sound artists, sound researchers and interaction designers. We outline the framework in detail and point to future experiments and perspectives for expanding the VUI design space.