3.1. Method and procedure
The central question of this study was how to visualize gestures in conversational turn-taking situations.

Firstly, a conversation had to be recorded. The results to be reported were based on transcriptions of speech and gestures during a 21 minutes long discussion of a couple. They were not exactly told about the purpose of the procedure. So, their interaction developed in a relatively natural way, apart from the fact that they were videotaped. Both teachers, the participants were told to discuss the topic of how to make best use of time. Their interaction was concentrated and getting more and more lively.

Secondly, a decision had to be made in the question of visualization of the gestures included in the turn-taking mechanism.

So, two video cameras were used to record the discussion. The cameras were set up in the back of each of the participants who were sitting face-to-face on two chairs. The cameras were at such distance from the participants that both of them were fully visible from head to foot on the tape at all times. As a consequence, more subtle details of facial expressions were not discriminable on the videotape. Less subtle expressions, such as broad smiles and grimaces, were discernable. In contrast, small movements of the hands and fingers, for example, were clearly evident on the tapes, so that fine discriminations of these movements could be made.

No zoom techniques or other special focusing effects were used.

Another decision had to be made in such an analysis is what constitutes a turn at talk. There has been widespread disagreement about this in the literature. According to BEATTIE (1982) in the present study only the vocal identifiers "mm-hmm", "ahh", and brief lexical terms such as "ja" and "okay" with attentional functions are excluded from the class of turns.³⁴ So, there remains a large category of turns.

At last, the videotapes were digitized and processed on a computer which made it possible to produce screenshots in order to discriminate the moments when gestures occured in the turn-taking mechanism. Those screenshots are presented in this paper and in the internet version.