An earlier study showed that listeners in conversations insert brief responses ("mm-hmm," "I see," and the like) almost exclusively at the ends of rhythmical units in the talker's speech (Dittmann & Llewellyn, 1967). In this study these vocal responses were compared with a visible one, the head nod, and it was found that the 2 occurred together more often than chance would predict. Content analysis showed that these co-occurrences usually serve an interpersonal function: the wish of the listener to speak or the wish of the talker for feedback. When they did occur together, nods were found to precede the vocal response slightly. Apparently the listener must hold a vocal response politely until the speaker has finished a unit, but may nod before then.

No abstract

The relationship between body movement and speech rhythm was newly formulated following Boomer's work on hesitations in speech: movements were predicted to occur early in phonemic clauses and at points following nonfluencies within clauses. A preliminary study of old data for which the movements were located by watching motion pictures bore out the prediction, and led to a more intensive study using more representative speech samples, and recording techniques designed to eliminate possible artifacts. The results were highly significant, but the amount of movement variance accounted for was small. The data collected by this method allowed direct test of statements by Pittenger, Hockett, and Danehy, and by Scheflen, whose claims of very close speech-movement relationships were found to be exaggerated. The linkage found between hesitations and movements was interpreted in terms of speech encoding processes.1 This paper is an extension of one given at a symposium chaired by Paul Ekman, titled "New Approaches to the Study of Facial Expression and Body Movement," at the American Psychological Association meetings in New York, 1966.2 Requests for reprints should be sent to Allen T.

Listener responses ( LRs) have been found to appear far less frequently among children than among adolescents and adults, and the initial explorations of this finding, both empirical and theoretical, are reported in this paper. Observations of children in fint, third, and fifth grade in school (but not in school-like situations) showed LRs to be almost absent except under the strongest social "pull," and to be very variable from child to child. Video taped sessions in the laboratory \bowed older children to make small movements which may not be communicative, to make the more definite ones a bit late compared with adults, and confirmed the wide individual variability. Discussion focuses on two functions of LRs, the linguistic and the social.In this paper I shall be discussing a number of topics which have been related to nonverbal communication over the years : encoding and decoding speech, feedback in social interaction, the development of communication skills, and the usefulness of coded information sources for efficiency in communication. The reason this particular set of topics is gathered together hcre stems from research begun in an apparently new direction, that of children's behavior in conversational situations. I shall present some preliminary findings in this area, then discuss their implications for the different topics I have listed above. To anticipate briefly, we have found that small children while conversing do not produce what we have called listener responses (best exemplified by head nods and brief vocalizations like "Mm-hmm" ) anywhere nearly as frequently as do adolescents and adults. BACKGROUND

No abstract