Structural and Physical Aspects of Sleep-Utterance
Does sleep-speech tend to be coherent? Correct grammatically.? Wellenunciated and intelligible rather than mumbled? Loud rather than soft? Short rather than long in duration and in word count? That is, what are the characteristics of sleep-utterance from a structural and physical perspective, as opposed to one of mental content? And how do sleep-utterance episodes strike an external observer?
According to MacNish ( 1838), sleep-speech is at one time "rational and coherent and at other times full of absurdity. The voice is seldom the same as in the waking state. This I would attribute to the organs of hearing being mostly dormant, and unable to guide the modulations of sound [p. 181]." It is interesting that this notion foreshadows more recent developments and scholarly thought on the monitoring function of hearing one's own voice ( Holzman & Rousey, 1970; Klein, 1965; Kraepelin, 1906; Mahl, 1972; Skinner, 1957).
In essential agreement, Carpenter ( 1886) remarks that:
Among sleep-talkers there are some who merely utter meaningless sequences of words or strangely jumbled phrases, and are utterly incapable of being influenced by suggested ideas [ideas suggested aloud to the sleeper by an awake observer, but uttered below waking threshold intensity] whilst there are others who give utterance to a coherent train of thought, still without any receptivity of external suggestion; and others again obviously hear what is said to them, and attend to it or not according to the impression it makes on them (and still remain "asleep") [p. 591].