In this article, we address to what extent the proverb "the sound makes the music" also applies to humanrobot interaction, and whether robots could profit from using speech characteristics similar to those used by charismatic speakers like Steve Jobs. In three empirical studies, we investigate the effects of using Steve Jobs' and Mark Zuckerberg's speech characteristics during the generation of robot speech on the robot's persuasiveness and its impressionistic evaluation. The three studies address different human-robot interaction situations, which range from online questionnaires to real-time interactions with a large service robot, yet all involve both behavioral measures and users' assessments. The results clearly show that robots can profit from using charismatic speech. 4:2 K. Fischer et al.or another. For this reason, speech melody is necessarily present in robot speech as much as in human speech.In robot speech, usually off-the-shelf speech synthesizers are used (e.g., Mary TTS, Festival, eSpeak, FreeTTS, Vocoder, HTS), which come with speech melodies that depend on the sentence structure, for instance, by placing stress on content words [cf. Taylor 2009;Dutoit 2013]. These speech synthesizers are not implemented with persuasiveness in mind, but target as many possible uses as possible: "commercially developed TTS system have emphasized coverage rather than linguistic sophistication, by concentrating their efforts on text analysis strategies" [cf. Dutoit 2013:148]. Thus, speech melody is not adapted to different purposes, and most systems do not even allow the manipulation of prosodic features, such as speech melody, duration, spectral distribution, formant dynamics, and so on. Some studies in HRI have therefore used human recorded speech [e.g., Nishio et al. 2012;Kanda et al. 2008], and several studies show that people prefer human speech over presynthesized speech [e.g., Nass and Brave 2005]. Nevertheless, in the long run, with robots becoming more flexible and situations of use becoming more varied and unpredictable, robots will have to synthesize their own speech and cannot rely on pre-recorded utterances alone.In this article, we present three studies that investigate the influence of speech melody and other prosodic features in robot speech both on users' impressionist evaluations of the respective robots and on the robots' persuasiveness. The results show that speech melody, and prosody in general, has a considerable effect on how robots are evaluated and how persuasive they may be; this finding proved to be robust in that it holds across HRI scenarios and robots. These results indicate that robot designers should pay careful attention to how robots say what they have to say if we want robots to function seamlessly in social interactions.
PREVIOUS WORKPrevious work indicates that various characteristics of robot speech may play a role in humancomputer and human-robot interaction [cf. Nass and Brave 2005]. Especially whether the robot's voice is human or whether it is synthesized seems to make a cons...