Perceiving politeness from speech acoustics alone: A cross-linguistic study on Korean and English

Politeness is a crucial aspect of everyday speech communication; however, there are to date only few acoustic studies on this topic. Winter and Grawunder (2012) showed that for Korean speakers, politeness is reflected in pitch, intensity, voice quality and speaking rate. Here, we extend this production study by investigating whether Korean and English listeners can perceive the intended politeness of short Korean utterances based on speech acoustics alone. In two experiments with a total of 47 English and 30 Korean listeners, we found that both groups can detect the intended politeness purely based on the phonetic qualities of speech. In one experiment, accuracy was low (Korean: 58%, English: 53%) because speakers heard multiple voices in a randomized fashion, not allowing them to familiarize with any particular voice. In a design that was blocked by speaker voice, accuracy was higher (Korean: 70%, English: 58%), showing that vocal politeness can be used as a cue when the voice is known. This shows that politeness is not only expressed by honorific lexical forms commonly employed in Korean, but also by speech acoustics. It is remarkable that English speakers performed above chance at all, pointing to cross-linguistic regularities in how politeness is expressed vocally.