Expressing politeness in Polish and Canadian dialogues

Daniela Hekiert , Magdalena Igras-Cybulska


The aim of this paper is to introduce a novel corpus of speech POLAC - Politeness Across Cultures and to present some preliminary results on acoustic correlates of intended politeness. It was created for cross-cultural comparative studies on Polish and Canadians. It consists of 180 three-minutes long conversations in an arranged situation where the examined person was interviewing a non-native patient with a broken arm. In two experimental conditions, suggestions were made: to be polite, to be unemotional and in the control group there were no suggestions. Recordings were analyzed with voice quality and prosody parameterization methods as well as MFCC-GMM emotion recognition algorithm. The following observations for polite speech were made: in spite of similar F0, Canadians tend to modulate speech significantly more than Polish. The intonation was more flat and monotonous for Polish. No significant differences were found in voice timbre. For both, Polish and Canadians, polite speech took more time than in the no-suggestion condition, and the latter was also longer than unemotional condition. From the point of view of emotions in voice, higher proportion of sadness was detected in Polish than for Canadian group in all conditions, but the tendency is the strongest for polite condition
Author Daniela Hekiert (Wydział Psychologii)
Daniela Hekiert,,
- Wydział Psychologii
, Magdalena Igras-Cybulska
Magdalena Igras-Cybulska,,
Publication size in sheets0.5
Book Setti Gianluca (eds.): 2018 II National Interdisciplinary Scientific Conference TechSpo’18: Power of Algorithms, 2018, Institute of Electrical and Electronics Engineers, ISBN 978-1-5386-8467-2, [978-1-5386-8466-5], DOI:10.1109/TechSpo.2018.8584627
Languageen angielski
Additional file
Oswiadczenie_D.Hekiert_Expressing.pdf 20.93 KB
Score (nominal)20
Score sourcepublisherList
Citation count*
Share Share

Get link to the record

* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.
Are you sure?