Output details
34 - Art and Design: History, Practice and Theory
University of Dundee
The value of visualising tone of voice
This paper highlights the importance of tone of voice, yet how difficult it is to describe––even for phoneticians. It proposes a role for design in provoking and supporting wider and deeper discussion about the lack of tone of voice in speech synthesis, for example in communication devices for people who cannot speak. It is 'design for conversation' in both senses.
As an output of the EPSRC-funded Creative Speech Technology (CreST) network, a special edition of the journal Logopedics Phoniatrics Vocology (LPV) has been published. LPV has a 2011 Impact Factor of 0.836 and addresses a completely different disciplinary audience to previous outputs of this author's research.
Its originality includes proposing a role for design in the shaping of speech technology––perhaps a surprising role, given design's traditional association with all things visual, and given the invisibility of speech. The two research projects, Six Speaking Chairs and Speech Hedge are each original, as are the corresponding exercises in eliciting and analysing audience responses.
Its rigour comes partly from including descriptive research with 90 respondents. In the methodological context of Daniel Fallman's Interaction Design Research Triangle, the notion of rigour has more to do with the fact that this research has demonstrably seeded new conversations (Fallman, Daniel, and Erik Stolterman. 2010. Establishing criteria of rigour and relevance in interaction design research, Digital Creativity, 21:4, 265–272) within the CreST network.
Its significance has been to challenge the dominant 'emotional' approach to expressive speech technology, whereas when respondents were free to write their own descriptors of tones of voice, less than a quarter of these were described in emotional terms. Its impact could transcend AAC: telephone services, screen readers, digital audiobooks and video games are each significant mainstream markets in which people other than speech technology experts might interact with synthetic speech.
