ELICITING TONAL DATA

Bielefeld, May 2001

Constance Kutsch Lojenga

Leiden University / SIL

Abstract

Tone is an important element in the analysis of African languages. In the past,

this topic has been severely neglected, and only relatively few scholars have

given due attention to this topic, in most cases describing the surface pitches

or tones they heard.

These days, most works pay attention to the analysis of tone in a particular

language in terms of an underlying system and rules to derive the correct

surface output embedded in a modern theoretical framework. Very little has been

said, however, about how to obtain solid data which can serve as basis for such

an analysis. My aim in this paper is therefore to present thoughts on a

methodology for elicitation of tone, thereby hopefully opening up a discussion

on the topic and inviting input from the experience of other colleagues in the

field.

In the introduction, I will briefly present the topic in general, also referring

to the function of tone in African languages. That leads to the need and

importance of reliable data for tone analysis, which leads us to the question

of HOW TO obtain such data, i.e. a methodology for eliciting tonal data.

I will look at the topic of a methodology for obtaining tonal data from two

angles:

1. how to come to a good hearing and transcription of surface pitch, and

2. what data to elicit in order to develop a good hypothesis for the underlying

tonal system of the language.