Band 26, Heft 3

This paper reviews the inherent time variables of speech and then discusses the importance of these variables for spontaneous speech activities in French. A comparison with English shows that the French speak faster than the English, this being mainly due to the smaller number of pauses introduced into their discourse, not to the rate of articulation, nor to the length of pauses. Moreover, it follows from the analysis of the syntactic distribution of unfilled pauses that 75 % occur at the end of sense groups (60 % at the end of clauses) and that pauses inside clauses are shorter (median 0.40 sec) than final ones (median 0.50 sec). Finally, the investigation of secondary variables (other hesitation pauses) shows the importance of ‘articulation’ pauses (which consist of filled pauses and of drawls) and the prime role played by grammatical words in the categories: drawls, repetitions and false starts.

A trivial utterance was pronounced in 26 different ways by an actress according to 26 contexts. 80 subjects were asked to guess these contexts and to place the sentences on semantic scales (cf. OSGOOD and E. ULDALL). The success of message transmission varies between complete failure and almost total equivalence. Adequate vocal interpretations of suggested situations do not necessarily conjure up these situations when they are separated from their contexts. There is a kind of homonymy in the domain of vocal stylistic messages. The encoding and decoding of secondary messages, far from being a ‘natural’, ‘instinctive’ phenomenon, is a very complex process which presupposes coordination, the integration of a series of partial messages and the reconstruction of the total message from the partial ones. The most elementary and the most specific messages were transmitted with the least distortion. Vocal performance – intonation, stress, articulation – does not contain a direct reference to a concrete situation, it hints at the situation indirectly by the successive and/or simultaneous expression of different simple attitudes. The secondary vocal messages are yielded by an archaic code which works only with motivated signs. These signs are, at the same time, integrated, more or less perfectly into the linguistic code.