Abstract:

In this paper we trace tonal displacement in one group, Anyi and Nzema and another group, Central Baule and Kode Baule that represent a few of the principal languages of the Tano group within New Kwa. Both language groups are intimately related and are classified as southern Bia, a subcluster of Central Tano.

From the methodological point of view, we assess tonal developments by comparing regular tonal correspondences among cognates in closely related languages. Our intention is primarily to investigate tonal patterns using comparative paradigms for the two respective groups mentioned above. Our objective is to formulate explicitly regular rules of correspondences that could be the proper base for a reconstruction of the tonal patterns in Tano in general. Currently most current work on tonal variations in languages of the same Bia family has focussed on tonal contrasts to posit prototonal patterns. We suggest focussing on tonal alternations as a source of evidence.

Bia languages substantiate proposals on tonal developments based on rightward tone shift in Bantu as documented by Clements and Ford (1979), Hyman (1974, 1983) and Stewart (1983). Stewart’s account of Asante deals with a rightward shift that must have been independent, since Asante is a separate group from the Bia group that our Ivoirian cases belong to. Strikingly, Hyman (1983) explicitly argues that two distinct rules of rightward spreading apply in in Dagbani, a Gur language of northern Ghana. If that is correct, it is not surprising that spreading can enter a language group, or even a language, at two separate stages in its history.

Thus, typologically and historically the paper makes an empirical case for the rightward displacement of tone in Tano languages. The evidence comes from the comparison of lexical constrasts and syntagmatic alternations with numerals. Specifically a careful investigation of a comparative word list of Anyi and Nzema suggests that Nzema has innovated an initial Low tone that links to the first syllable at the left edge of the phonological word, the consequence of which is the displacement of the entire sequence of tone. So, for example, the Anyi sequence LHL corresponds to LHL8 in Nzema (cf. also Leben 1979, 1984), that is, the mapping of a tone to a syllable from left to right starting with an initial one is shifted away from the first initial syllable to the second one. Such a displacement of tones leads to the fact that the last tone becomes unlinked and floating. In most cases, the initial Low tone can be reconstructed as a reflex of a Low tone prefix that is typical to most Central Tano languages (Kouadio 1983).

However, a comparison of Central Baule and Kode Baule reveals that the latter has innovated an initial High tone, the origin of which is less obvious as synchronically no prefix with High tone is attested, unless we assume that the prefixes in Northern Bia had a different tone from those in Southern Bia. We consider also alternative explanations such as flip-flop rules of a High and Low tone as well as possible phonetic effects of tonal phonologization of L L, realized phonetically as ML, to H L. Interestingly, in such patterns, the initial L in LL and the initial H in HL in Central Baule are realized phonetically at approximately the same mid tone level (cf. Hyman and Schuh 1974). we intend to provide accurate laryngograph and acoustic measurements to illustrate that neutralization which suggests a possible development of an initial Low tone to High, which has been phonologized in Kode Baule.

Despite the apparent differences in tonal patterns in the comparative paradigms, the two language groups under comparison are similar in that rightward tonal shift has given rise to floating tones due to the displacement of following tones, adding a toneless syllable on the right.

We conclude that the tonal changes in Tano tend to preserve a core melody and simply modify a link at the right end and add a default tone at the left.