Intonation among German-Turkish bilinguals

Robin Queen

Sat. 4:00-5:40 C

While there are many studies examining the influence of cross-linguistic contact on other features of grammar, serious examination of the intonation patterns of bi- and multilingual communities has been rare. Even those studies which do look at intonation in bilinguals (Glempler 1989; Wenk 1986; de Bot 1986) tend to problematize intonation as an effect of second language acquisition in individuals rather than considering intonation from a language contact and language change perspective that relates to communities. In this paper, I present the results of a year-long, descriptive study on the intonation patterns of a bilingual community in Germany. The analysis of these data differs from many other studies of intonation in that it is based entirely on naturally occurring speech and on an inductive rather than deductive method of inquiry.

As there is so little known about the effects of longer term linguistic contact on intonational systems, one of the primary questions for a study such as mine concerns whether or not there are differences in the forms, functions and distribution of intonation patterns in bilingual communities. Thus, in this paper, I examine and compare the forms, functions and distribution of phrase-final tonal contours among Turkish-German bilingual and German monolingual children. The data for the study are drawn from six fourth grade students (2 Germans and 4 Turks), all of whom performed a similar conversational task. The task itself involved a barrier conversation in which one child described a picture made of geometrical shapes with varying colors to a second child.

In comparing the intonation of the two groups of children, I use the basic model for specifying tonal contours outlined by Liberman (1979) and furthered by Pierrehumbert (1980) and Liberman and Pierrehumbert (1984) as well as others. In my study, I focus only on boundary contours, examining the similarities and differences in terms of form, distribution and discourse function. The preliminary results suggest that Turkish bilinguals share most of the forms used by German children; however, the distribution of those forms may vary widely. For example, Turkish children do not implement phrase-final falls as a marker of finality with the same consistency of German children. Furthermore, the Turkish children often align tonal contours in ways not paralleled by their German peers. For instance, German rises typically have a slight dip preceding them (DeLattre et al 1968; Moebius 1993; Uehman 1991). The German children consistently align the dip-rise on the final stressed syllable of an utterance, whereas the Turkish children tend to align the dip-rise over several syllables, ending on the final syllable regardless of whether or not it is stressed. Note the following example (where the numbers represent Hertz values):

285 210 469

| | |

GR child: .... ein gruenes Kreis 'a green circle'

325 250 325

| | |

TR child: .... links ganz unten 'left at the very bottom'

In addition, there are qualitative differences in terms of the use of intonation as a tool for discourse organization, such that the Turkish children actually use changes in pitch range to signal units of discourse whereas the German children use range as a matter of emphasis. Thus, although the forms of the contours among Turkish-German bilinguals are similar to that of their monolingual German counterparts, there are distinct differences in terms of distribution, alignment and function.