On spontaneous speech and fluently spoken text: Production differences and perceptual distinctions

Our initial study [J. Acoust. Soc. Soc. Am. Suppl. 1 77, S38 (1985)] of spontaneously produced speech and fluently read speech found that these two kinds of utterance were distinguishable by untrained listeners, and may represent different modes of speech production. The perceptual differentiation of these types of utterance, which we tested using well‐formed sentences free from gross metrical dysfluencies, seemed not to depend on a single acoustic emblem of spontaneity or reading, neither the presence or absence of pitch downdrift, nor average vocal pitch, nor average duration. The simplest acoustic correlate of perceptual success in identifying instances of spontaneous production, and of misidentifying read speech as spontaneous, was the variance of the phonatory frequency. Our present study again employs a perceptual test of the differences between spontaneous and read speech, using partial sentences. Listeners identified a test set composed either of the initial portions of sentence‐size utterances or...