F0 Declination Line: More Evidence from Infants' and Children's Speech (9-48 Months)
Résumé
In this study, we examine the Fo declination line (FoDL) i n full term and preterm babies, between the age of 9 months, when a child still babbles, and the age of 48 months when a language with longer and complex sentences is already acquired. Two main situations of communication were chosen: solitary babbling/ interaction with adult. An instrumental analysis (Pitch detection, Fo curves) was conducted, a linear regression on all detected Fo values was computed and its significance tested statistically. The first results show that both situation of communication (solitary vs interaction with an adult) and age influence the presence and use of FoDL. Also, full term and premature babies seem to behave differently. The physiological and phonological aspects of the FoDL i n communication and langage development are discussed. 1. INTRODUCTION Is the Fo declination line (FoDL) constrained physiologically as an intrinsic feature of pitch or does it depend on linguistic constraints [1]? Despite many studies on this problem, the answer is still controversial. The existence of the declination line and its origin are still regularly put to question. Our purpose is not to examine the different answers or t o discuss them, but to provide some new elements taken from our long ongoing studies on the acquisition of language. These could throw some light on this problem. To the best of our knowledge, the existence of the FoDL has never been studied at the beginning of language, particularly in very young children's speech. Thus, it would be useful to know if i t is a physically innate phenomenon or if it is acquired and if so, when. It should be recalled that FoDL has been found even in the productions of non human primates [2]. In a preliminary case study [3] the FoDL was examined in one baby between the age of 9 months, when the child still babbled, and the age of 36 months when a first language, with some longer and complex sentences, was already acquired. To the variable ÒageÓ (V1), was added the variable Òsituation of communicationÓ (V2). Two main situations were chosen: non communicative solitary gibberish and productions emitted i n interaction with an adult. Two groups of hypotheses were put forward: a) in the gibberish style FoDL should not be present because there is absolutely no constraint on these productions, not even on voice quality, going from (low) creak to (high) screaming; in interactive productions, on the contrary, there should be at least some evidence of FoDL; b) FoDL should be linked with age; it should appear more and more clearly as the child gets older. The first results seemed t o show that our hypotheses were for the most part upheld. The percentage of items showing the presence of a slope (with a p-value <0.01) was directly correlated to the FoDL phenomenon. The "situation of communication" variable was found to be a significant factor. The same conclusion hold for the "age"