Prosodic Differences in Human- and Alexa-Directed Speech, but Similar Local Intelligibility Adjustments

The current study tests whether individuals ( n = 53) produce distinct speech adaptations during pre-scripted spoken interactions with a voice-AI assistant (Amazon’s Alexa) relative to those human interlocutor. Interactions crossed intelligibility pressures (staged word misrecognitions) and emotionality (hyper-expressive interjections) as conversation-internal factors that might influence participants’ adjustments in Alexa- human-directed (DS). Overall, we find style differences: Alexa-DS has decreased rate, higher mean f0, greater f0 variation than human-DS. In produced toward both interlocutors, response misrecognition were similar: participants more vowel backing (enhancing the contrast between target misrecognition) words louder, slower, at sentence-level. No differences observed human- following displays of emotional expressiveness by interlocutors. Expressiveness, furthermore, did not mediate misrecognition. Taken together, these findings support proposals speakers presume “communicative barrier” (relative interlocutors), but adapt conversational-internal similarly Alexa-DS. This work contributes our understanding human-computer interaction, well theories adaptation.