Conversational intentions as reflected in the breathing pattern Cover Image

Vestluskavatsused Hingamismustris
Conversational intentions as reflected in the breathing pattern

Author(s): Kätlin Aare
Subject(s): Language studies, Phonetics / Phonology, Language acquisition
Published by: SA Kultuurileht
Keywords: phonetics; spontaneous speech; speech breathing; speech planning; turn taking; backchannels;

Summary/Abstract: The main aim of this article was to describe how two functionally different objects in terms of speech planning behave in the respiratory patterns of spontaneous speech. First, the focus was on the timing of utterances and backchannels (short feedback items signalling that the listener understands and follows the speaker) in the respiratory cycle. The second focus was on the parametres of the inhalations preceding utterances and backchannels. The article is an overview of research done in speech breathing using data from spontaneous Estonian conversations. The data is made up of spontaneous dialogues and multiparty conversations held in Estonian and recorded at Stockholm University in 2014 and 2015. Each conversation lasted from 20–30 minutes and was recorded in high quality audio and video. Respiratory data from each participant was collected and recorded using respiratory inductance plethysmography. The results indicate that the difference in the parametres of inhalations preceding utterances and backchannels is functionally motivated and connected to the duration of the speech units. Experiment 1 showed that the starting points of backchannels are more evenly distributed in the breathing cycle than the starting points of utterances. However, utterances as short as the majority of backchannels behave similarly and therefore backchannels cannot be distinguished in the speech flow based on their durational properties alone. Experiment 2 demonstrated that inhalations preceding utterances are considerably larger than those preceding backchannels. Backchannels are usually short and quiet, they can be produced whenever they become relevant in the course of the conversation, and as such are not produced as an attempt to take over the conversational floor. Therefore, they do not need to be planned in the same way as utterances. In addition, the amount of air in the lungs at the end of an inhalation preceding an utterance can function as a sign of taking over the floor.

  • Issue Year: LX/2017
  • Issue No: 05
  • Page Range: 364-380
  • Page Count: 17
  • Language: Estonian