What do we expect spontaneous speech to sound like?

Listeners have been shown to distinguish text read aloud from spontaneous speech, with a range of prosodic features suggested as cues to speech style. However, significant variation is seen across studies, both in speech elicitation methods and in the nature of listeners’ orientation to prosodic cues. We asked whether listeners could distinguish spontaneous ‘map task’ speech from lexically identical read utterances. Experiment 1 found that, although our spontaneous speech differed prosodically from read speech, listeners did not appear to use available cues to distinguish styles. Experiment 2 found that, even when matched spontaneous and read utterances were presented consecutively, listeners still did not reliably discriminate between styles despite available cues. We suggest that listeners’ ability to distinguish between speech styles derives from the interaction of expected and available cues, including prosody, mediated by listeners’ interpretation of such cues as being representative of speech context and the intentions of the speaker.