Among the various languages of the ten families of the Tupi trunk, the Mundurukú language (Mundurukú family) is the only one with distinctive laryngalization. The facts verified in the preliminary research show some features of this laryngeal trait as well as its close relationship with the glottal occlusive and pitch. The results show some relevant points: (i) laryingalized syllables always occur with the low pitch of the two existing ones (High and Low) in the tonal system; (ii) glottal occlusive and laryingalization do not occur simultaneously in the same syllable; (iii) the glottal occlusive appears on the surface preceding a syllable with High pitch (in this case the vowel followed by glottal can, more or less freely, manifest itself as a laryingalized vowel); (iv) the third person singular prefix causes the pitch of the initial laryngalized syllable of the verb radical to change to High (when this happens, the laryngalization of the initial syllable disappears). Such facts lead us to consider the hypothesis that the variable laringalization is due to a glottal occlusive at the end of the syllable in the underlying form. This hypothesis is reinforced when Mundurukú data are compared with data from other languages of the same stem. In certain lexical items, the glottal occlusive, retained in those languages, corresponds to a vowelaringalized vowel in Mundurukú. The evidence that a glottalic segment generates different types of phonation and consequently affects pitch has so far only been mentioned for the Athabaskan and Tibeto-Burmese families, and a similar process is now noted in Mundurukú. Such evidence contributes considerably to the reconstruction of Proto-Tupi which, according to Rodrigues & Dietrich (1997), presented a glottalized series of consonants. But the relationship between glottalic segments, phonation types and pitch leads us to think of another hypothesis, that there was a glottalic occlusive in the final position of the syllable, as occurs in Karo (Ramarama family).
|
