One Voice Fits All?

Abstract

When a smart device talks, what should its voice sound like? Voice-enabled devices are becoming a ubiquitous presence in our everyday lives. Simultaneously, speech synthesis technology is rapidly improving, making it possible to generate increasingly varied and realistic computerized voices. Despite the flexibility and richness of expression that technology now affords, today's most common voice assistants often have female-sounding, polite, and playful voices by default. In this paper, we examine the social consequences of voice design, and introduce a simple research framework for understanding how voice affects how we perceive and interact with smart devices. Based on the foundational paradigm of computers as social actors, and informed by research in human-robot interaction, this framework demonstrates how voice design depends on a complex interplay between characteristics of the user, device, and context. Through this framework, we propose a set of guiding questions to inform future research in the space of voice design for smart devices. CCS Concepts: • Human-centered computing → HCI theory, concepts and models.

Voice in Human–Agent Interaction

“…From a technical standpoint, voice refers to the manifestation of sound heard by a recipient and vocalized by a sender through some vocal mechanism [29,91,92]. Put simply, it is the "how" of vocalizations [21,171]. The vocal mechanism has traditionally meant the vibration of human or animal vocal tracts, but since the development of computer-based media and related technologies, it can also include machine-based synthesis of sound.…”

“…Speech is the linguistic content of voice, primarily comprising words (vocabulary, including pseudo-vocabularies and slang), grammar and syntax, and phonetics (the physical measurement of speech). It is the structured language of communication [145], or the "what" of vocalizations [21,171]. Voice, then, is a 1 medium of speech.…”

“…In this body of work, voice is not a new topic. Recent reviews on conversational agents [52] and smart devices [21] have revealed that aspects of voice are key in the design of agents. Other surveys have focused on detection and conveyance strategies from a technical angle: algorithms and hardware that "perceive" various aspects of human communication [22,144], such as affect [36,59,111]), non-semantic vocalizations [192], non-verbal cues [52], including laughter [31], and other expressive qualities [64].…”

“…Yet, there is a gap when it comes to the state of the art on "the voice of the machine" as an expressive and influential factor across different kinds of agents. We are motivated by a growing recognition that the creators and gatekeepers of these agents-computer scientists, engineers, designers, data providers, researchers, producers, managers, CEOs, and others-have a key role to play in perpetuating or disrupting social and ethical status quos, as well as in conscientiously designing technologies for the ethical and social good [21,78,178,187].…”

Same benefits, different communication patterns: Comparing Children's reading with a conversational agent vs. a human partner

“…This story was chosen because of the appropriate level of narrative complexity for the age group and potential story interest. To eliminate the confounding effects of the CA with the effects of voice quality (Cambre & Kulkarni, 2019;O'Neal et al, 2019), the CA used a female recorded voice instead of machine synthetic voice.…”