Spoken chant’ intonation in French

			Szuzsanna Fagyal
	(Universite Paris III, University of Pennsylvania)


This study suggests that the intonational contour referred to as 'spoken
chant' (Pike 1945) or 'vocative chant' (Liberman 1975) in English is
also a widely used tune in French. Although previously illustrated in
familiar calls and chanted listings (Fónagy & al. 1983), the contour's
formal properties and its pragmatic interpretation have not yet been
fully analyzed.
'Spoken chant' intonation in French has been associated with the last
two or three syllables of an utterance: a rise of about eight semitones
on the penultimate syllable followed by a fall of about four semitones
on the final syllable. The lengthened final syllable gives the contour a
chanting air (Fónagy & al. 1983). Although this description accounts
well for the tune, it does not allow to relate the latter to an
underlying phonological unit within the French tonal system. Mertens's
(1987) intonational model for French provides a  l h \HH 
tone-to-syllable representation on the last three syllables of an
Intonational Phrase, but it suggests that besides L and H tones, a third
\HH tone —a lowered-high boundary tone— would be a phonemic tonal height
in French. This study provides a phonological representation based on a
two-tone model (Jun & Fougeron 1995, Fougeron & Jun 1996).
Familiar calls and warnings were elicited from two native speakers in
controlled, read-aloud dialogs. The contexts consisted of voiced target
words repeating the same targets embedded in a previous statement
(fig.1.). French first names Anne, Anna, Joanna, Marie-Anna and
Marie-Joanna were used as targets. Each word corresponds to a one- to
five-syllable Accentual Phrase (AP) and forms a one-word Intermediate
Phrase (ip) and Intonational Phrase (IP). The spoken chant contour —as
it appears in the three-, four- and five-syllable target words (fig.2.)—
can be represented as a sequence of H* H- L% tones, where H* is the
AP-final high tone shifted to the penultimate syllable of the phrase,
and H- and L% are ip and IP tones, respectively, both realized on the
final syllable of the phrase. The sequence of H- and L% tones surfaces
phonetically as a mid tone (following the framework of Pierrehumbert
1980).

underlying AP			L  H     L    H*         
	       ip			H-
	       IP			L%
	 AP()/ip[]/IP{}         {[( L  L     H*) H-] L% }

This tonal representation remains the same when the AP’s total number of
syllables is extended to the presumed maximum (6-7 syllables), as in
Marie-Joanna. (fig.2.e). The tone-to-syllable mapping is difficult to
analyze in the one- and two syllable words (Anne and Anna  fig.2.a,b)
showing a high-low pattern. In IPs containing more than one ip-s, the
contour appears only on the last AP (fig.3. 4.).
Unlike in English, ‘spoken chant' intonation does not seem to occur in
warnings in French, but—just like in English—it appears in familiar
calls. Naturally occurring examples show that two other dialog
contexts—both conveying an implicit, shared convention between the
speakers (Pierrehumbert & Hirschberg 1989)—are characterized by this
contour in French. ‘Chanting’ listings convey that the items enumerated
in the list are not individually informative, but rather intended to
illustrate a tacit agreement between the speakers about the
propositional content of the list. In fig. 3 it is self-evident for both
speakers that the provided list of French writers is 'necessarily'
representative of the most famous French writers of this century.
‘Stylized’ high-rise intonation seem to have a similar listing function
in English (Ladd 1978). In French, spoken chant intonation also occurs
in assertions implying that the content of the utterance has to be taken
by the other speaker as an 'evident truth' deduced of what precedes. In
fig.4s speakers A and B are discussing whether A’s professional career
was influenced by external factors or by some sort of predestination.
The statement of A uttered with ‘spoken chant’ intonation conveys
that—given the preceding elements of the discussion—it is considered
evident by A that her career was mostly influenced by ‘objective
circumstances’. Such ‘implicative’ statements (see Delattre 1966) also
seemed to have a turn-yielding function in the observed dialogs. In
those cases they were often perceived as yes/no questions by the
listener and, therefore, elicited direct ‘yes’ or ‘no’ answers.

References

Delattre, P. (1966). Les Dix Intonations de Base du Français. The French
Review. 40(1):1-14.

Fónagy, I. & Bérard, E. & Fónagy J. (1983). Clichés mélodiques. Folia
Linguistica 17:153-185.

Jun, S-A.& Fougeron, C. (1997). A phonological model of French
intonation. Submitted to publication.

Jun, S-A. & Fougeron, C. (1995). The accentual phrase and the prosodic
structure of French. Proceedings of ICPhS of Stockholm, vol.2, 722-725.

Ladd, D.R. (1978). Stylized intonation. Language 54, 517-540.

Liberman, M. (1975). The intonational system of English. Doctoral
Dissertation, MIT, Cambridge, MA. Published by Garland Press, New York
1979.

Pike, K. (1945). The intonation of American English. Ann Arbor, MI:
University of Michigan Press.

Pierrehumbert, J. (1980). The Phonology and Phonetics of English
Intonation. Ph.D. Dissertation, MIT Cambridge.

Pierrehumbert, J. & Hirshberg, J. (1989). The Meaning of Intonational
Contours in Discourse. in: Cohen, P.R. (ed.), Intentions in
Communication. Bradford, Cambridge (MA) - London (UK), 271-311.