One thesis of this paper was that gestures are of considerable importance in regulating the turn-taking mechanism. Though the present study is based on one recorded discussion only, there are some conclusions concerning the role of speaker and listener gestures in conversational turn-taking situations to be made:

• There is evidence that the termination of a hand or arm gesture acts as turn- yielding cue.
  
• There is evidence that gestures act as an attempt-suppression signal.
  
• Interruption seems to be a situation which demands strong gestural behaviours or, the combination of different gestures.
  
• The fact, that there were only four interruptions in 64 exchanges of turn, indicates that the interaction analysed here was well organized and trained, as it was to be supposed for a couple. Generally, p2 made more use of gestures as p1.
  
• The appearance of multiple cues may also be related to the urgency of turn-yielding or turn-requesting.
  
• Auditors diplayed a wider variety of gestural behaviour and displayed those behaviours more frequently than did speakers.

The line of thinking developed here suggests that a "grammar" of gestural interaction exists, the term "grammar" used as implication for the existence of a structure of interaction.

There were parts of the interaction when no gestures could be observed, but in most parts of the conversation gestures were used even in turn-taking situations. Though the turn-taking mechanism may not function without verbal cues, there was a large variety of gestures to be observed that accompanied and reinforced the turn-taking mechanism.