The modifiability of the temporal structure of spontaneous speech

The notion that time is an integral dimension of speech is not especially new or surprising. Speech is emitted in linear strings, or sequences, of sounds and silences. Each "sound" is not necessarily a word; it may be a vocal segregate ("mm-hmm"), or it may in fact be part of a word. More commonly, a sound will comprise several words that are not separated by any silence that can be detected by the human ear. In any case, each sound follows a silence and each silence a sound, and it is this succession of events that, in part, allows for a perception of the sequence as "occupying" or "taking up" time (Fraisse, 1963). Moreover, each sound and each silence, as well as the entire sequence, has a certain measurable duration, which is the amount of time it "takes up."

What is perhaps surprising is that the durations of these events, particularly (although not only) the silences and particularly within the context of a two- person exchange, seem to be related systematically to a number of interesting psychological states and processes. Some silences, of course, are a consequence of the structure of the organs used for speech production and of the type of language spoken. For example, since in occidental languages the inspiration of air is not used for the production of sounds, some of the silences can be shown to be breathing pauses. Other silences are related at least in part to the syntactic structure of the language. However, part of these silences and certainly other nonsyntactic silences are thought to reflect cognitive processes and/ or affective states of the speaker. Moreover, the durations of the silences, as well as of the sounds and combinations of the two events, appear to be associated with certain personality characteristics attributed to speakers by listeners and by the speakers themselves.