- NAJAKS – Nordic Association of Japanese and Korean Studies - http://www.najaks.org -

“Funny talk” corpus and speaking style variation in spoken Japanese

Posted By admin On May 31, 2013 @ 9:02 am In Abstracts for 2013, Linguistics and Language Teaching | Comments Disabled

Toshiyuki Sadanobu, Professor

Graduate School of Intercultural Studies

Kobe University, Japan


This paper discusses a problem borne by the present corpora of spoken Japanese language and demonstrates that our new corpus Omoshiroi hanashi (funny talk) is effective for breaking through the problem and for teaching spoken Japanese as L2.

The prior status of spoken language over written language has been recognized traditionally (e.g. Hockett, C. 1960 “The origin of speech” and Lyons, J. 1981 Language and Linguistics). However, actual research on language has focused on written language, and language education also has followed this trend. Such discrepancy between the ideal and the reality of spoken language is being improved by the development and spread of technology. Technology for recording, editing, and releasing audio-visual information sheds light on spoken language, and now we can utilize various corpora for research and education on spoken Japanese. However, these corpora have in common a problem of balance. None of the present corpora is well-balanced in a strict sense. This is because our knowledge of the variation of spoken Japanese still is very limited and premature. How many variants of speaking style does Japanese language have? In what situation does each of them appear, and to what degree? We do not have enough knowledge to answer these questions, which means that the effectiveness of quantitative research is limited (as was pointed out long ago by Schegloff, E. 1968 “Sequencing in conversational openings”).

In order to deepen our understanding of variation of speaking style, we have begun to construct a new type of corpus, Omoshiroi hanashi (funny talk). This corpus is an audio-visual collection of talks about three minutes in length entered in a funny-talk tournament, with Japanese subtitles and English/French/Chinese translations. Although we cannot say that the talks entered in a tournament are as natural as everyday conversation in all aspects, this corpus abounds in speaking styles barely addressed in hundreds of hours of standard spoken-language corpora. By using this corpus as a complement to previous corpora, we can learn the actual variation of Japanese speaking style.

In this paper, we demonstrate such complementary efficiency of our corpus for a better understanding of the real variety of speaking style with a special focus on prosody (intonation patterns) and phonation (voice qualities). For example, in our corpus lexical accent often is affected by intonation to such a degree that it has completely lost its original shape. This contradicts sharply with the previous observation (e.g. Amanuma, Otsubo, and Mizutani 1978) that there is no competition between lexical accent and phrasal intonation for realization and that Japanese lexical accent is not subservient to intonation. Another example is rikimi (pressed voice), a subtype of creaky voice. In our corpus pressed voice often conveys the attitudinal meaning of kyoshuku (a mixed feeling of apology and embarrassment), which cannot be accounted for by the traditional simplistic view of voice quality as a mere means of emphasis.

* This work was partially supported by the Ministry of Education, Science, Sport, and Culture, Grant-in-Aid for Scientific Research (A), 23242023.


Article printed from NAJAKS – Nordic Association of Japanese and Korean Studies: http://www.najaks.org

URL to article: http://www.najaks.org/?p=1048