Is Korean Really a Listener-Responsible Language like Japanese?: A Contrastive Discourse Analysis of Korean and Japanese Apologies Sumi YOON Kanazawa University, Japan smy8005@hotmail.com Abstract According to Hinds' typology of languages on discourse level, Japanese and Korean are both considered listener-responsible languages, whereas English is classified as a speaker-responsible language (Hinds, 1987). However, in conversation, Yoon (2009) demonstrated that Korean should be classified as a speaker-responsible language based on her contrastive analysis of daily conversations between married couples in Japanese and Korean, where address terms and fillers are used as contextualization cues (Gumperz, 1982) to convey a speaker's intention to the interlocutor metacommunicatively. The purpose of the present study is to show that Japanese is listener-responsible, while Korean is a speaker-responsible language on the level of conversational communication. In order to test the hypothesis, surveys and recordings of real conversations of Japanese and Korean people were conducted and analyzed. The informants in the present study consisted of four groups: Japanese university students who live in their own country, Japanese university students who live in the U.S., Korean university students who live in their own country and Korean university students who live in the U.S. A Discourse Completion Test (DCT) was completed by Japanese and Korean university students to compare the differences in speaker responsibility in apologies. The results suggest that Korean should be classified as a speaker-responsible language for understanding in conversations, since Korean speakers produce many more utterances and convey more information per utterance to the interlocutor than Japanese speakers. Furthermore, it is found that the responsibility for the understanding of utterances correlate with daily use of American English, especially in the case of Japanese university students. Keywords Speaker responsibility, listener-responsibility, Japanese, Korean, apology Izvleček Po Hindsovi tipologiji jezikov na ravni diskurza naj bi bila oba japonski in korejski jezik v skupini jezikov, pri katerih je sogovorec (oz. bralec) bolj odgovoren pri razumevanju namena diskurza (listener-responsible). Po drugi strani je angleščina uvrščena v skupino jezikov, pri katerih je govorec (oz. pisec) bolj odgovoren (speaker-responsible) (Hinds, 1987). Yoon (2009) je na osnovi protistavne analize vsakodnevnih konverzacij med poročenimi moškimi in ženskami v japonščini in korejščini ugotovila, da je korejščina v slednji skupini, tj. da je govorec/pisec bolj odgovoren za razumevanje namena diskurza: v raziskavi so bili upoštevani nagovori in polnila kot ključi kontekstualizacije (Gumperz, 1982), ki metakomunikativno Acta Lingüistica Asiatica, Vol. 1, No. 3, 2011. ISSN: 2232-3317 http://revije.ff.uni-lj.si/ala/index posredujejo govorcev namen sogovorcem. Cilj pričujočega prispevka je to, da se preveri, da je v japonščini sogovorec oz. poslušalec odgovoren, medtem ko je v korejščini odgovoren govorec na ravni govorne komunikacije. Pogovore med Japonci in Korejci smo posneli in analizirali. Testirani so bili v 4 skupinah: japonski univerzitetni študenti, ki živijo v svoji domovini, japonski univerzitetni študenti, ki živijo v ZDA, korejski univerzitetni študenti, ki živijo doma, in korejski univerzitetni študenti, ki živijo v ZDA. Vsi študenti so izpolnili DCT test (Discourse Completion Test), po katerem bi ugotovili razlike v odgovornosti govorca v opravičevanju. Rezultati kažejo, da so govorci v korejščini, v primerjavi z govorci v japonščini, bolj odgovorni za razumevanje vsebine pogovora, ker izgovorijo veliko več izrazov in tako posreduje več informacij sogovorcem. Dalje smo ugotovili, je odgovornost za razumevanje izgovorjenih izrazov v sorazmerju z vsakodnevno rabo ameriške angleščine, zlasti v primeru japonskih univerzitetnih študentov. Ključne besede odgovornost govorca, odgovornost sogovorca, japonščina, korejščina, opravičila 1. Introduction According to Hinds' typology of languages, Japanese and Korean are classified as reader/listener-responsible languages on the discourse level (Hinds, 1987). However, on the basis of her comparison of address terms in Japanese and Korean conversations between husband and wife, Yoon (2009) criticized his claim and proposed that Korean should be regarded as a speaker-responsible language on the conversational level. In the present paper, apologies in conversations by Japanese and Korean university students are compared in order to support for the claim that Japanese is a listener-responsible language, whereas Korean is a speaker-responsible language on the conversational level, though both are categorized as listener-responsible languages according to Hinds (1987). In addition, the present paper examines whether the daily use of English, which is categorized as a speaker- responsible language according to Hinds (1987), can influence the ways that Japanese and Korean speakers converse with respect to the responsibility for understanding utterances. To examine this, utterances by Japanese university students and Korean university students who live in the United States of America and use English on a daily basis were analyzed. In the current study three research questions arise: (1) Is the total amount of information uttered per turn in Japanese and Korean different? Which language contains a higher amount of information per turn? (2) What semantic formulas are preferred for appropriate communication in apology discourse in Japanese and Korean, respectively? (3) Does the daily use of English influence the use of native Japanese and Korean language with respect to responsibility for the understanding of utterances? 2. Literature review Hinds (1987) claimed that languages can be categorized as reader/listener-responsible languages or writer/speaker-responsible, and pointed out that English is a writer/speaker-responsible language and Japanese and Korean are both reader/listener-responsible languages because of the similarity of writing patterns. He states: In Japan, perhaps in Korea, and certainly in Ancient China, there is a different way of looking at the communication process. In Japan, it is the responsibility of the listener (or reader) to understand what it is that the speaker or author had intended to say. (Hinds, 1987: 144) As he used the word perhaps, it might be assumed that both Japanese and Korean are classified as listener-responsible languages without any analysis of the data on the conversation level, only based on research of an essay written in English and Japanese. He also simply introduced an episode associated with speaker-responsibility and listener-responsibility between an American woman and a Japanese taxi driver by citing Naotsuka & Sakamoto et al. (1981). Hinds (1987) examined an expository essay from the Asahi Shimbun's daily column Tensei Jingo ("Vox Populi, Vox Dei") and its English translation and found two main reasons why Japanese should be categorized as a reader/listener-responsible language. First, in terms of rhetorical pattern Japanese essays, including the one mentioned above, are organized by ki-sho-ten-ketsu. In ten, new subtopics are introduced, but are written in a style which assumes the reader is already familiar with the subtopics. Second, Hinds (1987) noted that there was an absence or attenuation of landmarks that help the reader to understand the relationship between sentences. There are three criticisms of Hinds' theory of reader/listener responsibility vs. writer/speaker responsibility. First, it is not appropriate to compare an original Japanese column and its translated English version. The writer's expectations of the reader may be different. The original column was written for Japanese readers who can understand Japanese and live in Japan, and are familiar with present Japanese society and culture. In contrast, the English version was written for people who are Japanese but find it hard to understand Japanese language, society, or culture because they live in other countries, or are not Japanese but who are interested in Japanese language, society, or culture. Also, the fact that the person who translated the Japanese column into English may not have been a native Japanese speaker or English speaker should have been taken into consideration. If he or she was a native Japanese speaker, it is also possible that his or her Japanese was influenced by English or vice versa. Second, Hinds used the term listener-responsible and speaker-responsible without any experimental consideration on the level of conversation except citing an episode between an American customer and a Japanese taxi driver from Naotsuka & Sakamoto et al. (1981). It is possible that language has different features in regards to responsibility for understanding utterances in written discourse and spoken discourse. Third, Korean is categorized as a reader/listener-responsible language by inference that Korean language and Japanese language share common syntactic features and writing patterns. There are many studies which demonstrate the differences between Japanese and Korean discourse, although the languages are structurally very similar. Thus it is too simplistic to put Japanese and Korean in the same category without any experimental consideration. In regards to conversational level, Takigawa (2006) analyzed a conversation between a Japanese wife and her American husband and found that the Japanese wife stated the point at the end of her story and her American husband had difficulty following her story. His results seem to support Hinds' delayed introduction of purpose (Hinds, 1990: 98) on the conversation level in Japanese. It is important to note that Takigawa (2006) attempted to demonstrate the theory of Hinds (1987; 1990) on the conversation level and found a similar result to Hinds (1990). However, Takigawa (2006) examined the responsibility for understanding utterances with a very specific example: A short Japanese conversation between an American husband who was living in Japan and spoke both Japanese and English, and his Japanese wife who had lived in the U.S. and studied English. Yoon (2009) analyzed samples of conversations which were collected from Japanese married couples and Korean married couples. It was found that Korean married couples not only give more information, but also speak more directly than Japanese married couples to convey their intention to the listener in conversations. However, it is necessary to examine the results in conversations outside married couples. 3. Methodology 3.1 Participants Survey participants of the current study were divided into four groups: Japanese and Korean university students in their countries, and Japanese and Korean university students in the United States of America. Specifically, 101 (male: 55, female: 46) Japanese (Mean Age = 18.7 years; Range = 18-24 years old) and 71 (male: 29, female: 42) Korean (Mean Age = 19.5 years; Range = 18-24 years old) university students who live in the capital spheres of Tokyo and Seoul, respectively. 34 (male: 18, female: 16) Japanese (Mean Age = 24.5 years; Range = 19-32 years old) and 58 (male: 32, female: 26) Korean (Mean Age = 26.3 years; Range = 20-35 years old) university students who were studying at universities which are located in Washington D.C. and Boston, Massachusetts in the United States of America at the time that this research was conducted. The Japanese students' average length of stay in the United States of America is 37.8 months (Range = 6-120 months) and the Korean students' is 33.9 months (Range = 6-84 months). The following abbreviations are used in the present study: JU: Japanese University students KU: Korean University students JMU: Japanese Male University students JFU: Japanese Female University students KMU: Korean Male University students KFU: Korean Female University students JIU: Japanese International University students who live in the U.S. KIU: Korean International University students who live in the U.S. 3.2 Methods A DCT (Discourse Completion Test) was completed by JU and JIU in Japanese, KU and KIU in Korean to compare differences with respect to speaker responsibility. The DCT was conducted in the classroom for JU and KU, while JIU and KIU completed the DCT individually out of the classroom. The DCT was originally developed to compare the speech act realization of native and nonnative Hebrew speakers (Blum-Kulka, 1982, following Levinson, 1975). The test consists of scripted dialogues which are preceded by a short description of the situation specifying the setting and the social distance between the participants and their status relative to each other (Blum-Kulka, 1989). The DCT used for the present study was made by the researcher in light of Blum-Kulka (1989) to compare the way Japanese and Korean speakers use apology expressions. Using the DCT can show not only the total amount of utterances, but also when speakers convey their real intentions in the conversation. The DCT used in the present study has been translated from Japanese and Korean into English and is is shown below. (See appendix 1 for the original Japanese and Korean versions of the DCT.) You arranged to meet a good friend in front of a movie theater to see a movie, but you are about 20 minutes late. You: ( ) Friend: "That's OK. It couldn't be helped because your class finished late." You: ( ) Friend: "Ticket? I have already bought two tickets." You: ( ) Friend: "You don't have to thank me for just getting the movie tickets." 3.3 Data Analysis It is not adequate to only calculate the number of words or sentences used in conversations to compare information in utterances in Japanese and Korean, because there is not a one-to-one correspondence of linguistic items between both languages. Therefore, the data obtained from the participants was analyzed quantitatively by using semantic formulas with respect to information in utterances in corresponding situations between Japanese and Korean speakers (Beebe, Takahashi, & Uliss-Weltz, 1990; Blum-Kulka, House, & Kasper, 1990; Tao, 2007). Semantic formulas are types of semantic units in speech acts, all the utterances involved in completing the dialogue in the DCT were identified as semantic formulas for the purpose of this paper. For example: Hey mate, I am really sorry I am late. The lecture was finished late. (Address term) (Adverb) (Fact) (Reason) ( Apology ) The analysis of semantic formulas can clarify not only the amount of information in the utterances but also the construction patterns of apologies by Japanese and Korean speakers. Table 1, 2, and 3 below show which semantic formulas were used in utterances of the first, the second, and the third turn, respectively. The ADVERB and the INTERJECTION occur before or after the APOLOGY or THANKS expressions in the first and second turn. It is well known that Japanese usually use an apology expression like "sumimasen" (The Hepburn system is used for Japanese and the Yale system is used for Korean in the latin alphabet in the present study.) or "gomen" as a thankful expression (Ide, 1998; Jin, 2002; Kim, 1996; Yamamoto, 2003) and sometimes a Korean apology expression like "mianhay" is also used in appreciations. It is difficult to decide whether "gomen" and "mianhay" are used for apology or appreciation expressions in the DCT. Even though the situation set for the DCT in the present study is an apology for being late for an appointment fundamentally, the participants might say "Thank you" for understanding the reasons for being late or getting a ticket from their friends. Thus, the apology expressions "gomen" or "mianhay" are categorized literally as semantic formula APOLOGY and the thankful expressions "arigato" or "komawe" are categorized superficially as semantic formula THANKS in the second and the third turn. Utterances about buying movie tickets, for example, "Let's go buy tickets" or "Have you bought the tickets yet?" are regarded as semantic formula TICKET in the second turn. Utterances about worrying that the movie has already started are regarded as semantic formula MOVIE in the second turn. (See appendix 2 for the Japanese and Korean versions of examples of semantic formulas.) Is Korean Really a Listener-Responsible Language like Japanese? ... 79 Table 1: Semantic formulas and examples in the first turn in university students' apologies Semantic formulas Examples 1 APOLOGY "I am sorry." 2 FACT "I am late." 3 REASON "The lecture finished late." 4 ADVERB "(I am) really (sorry)." 5 ADDRESS TERM "John!" 6 INTERJECTION "Oh, (sorry)." 7 OTHERS "When did you get here?" Table 2: Semantic formulas and examples in the second turn Semantic formulas Examples 1 APOLOGY "I am sorry." 2 THANKS "Thank you." 3 TICKET "Let's go buy tickets." 4 MOVIE "Do you think the movie's started already?" 5 COMPENSATION "I will buy you some popcorn." 6 OTHERS "I don't like the professor." Table 3: Semantic formulas and examples in the third turn Semantic formulas Examples 1 APOLOGY "I am sorry." 2 THANKS "Thank you." 3 SURPRISE "Really?" 4 COMPLEMENT "You are a good friend." 5 COMPENSATION "I will buy you some popcorn!" 6 INTERJECTION "Oh, (thanks)." 7 ADVERB "(I) Really (appreciate it)." 8 OTHERS "You shouldn't have bought the ticket for me." 4. Results 4.1 The first turn 4.1.1 JU * KU Table 4 indicates the rates that semantic formulas were used in the first turn for JU and KU. The ANOVA revealed that the amount of semantic formulas used in Korean utterances was significantly higher than Japanese utterances with respect to semantic formulas REASON, ADDRESS TERM, and INTERJECTION. As for the semantic formula ADVERB, a significant difference was found between KMU and JMU. KMU used semantic formula ADVERB significantly more than JMU. Also, there was a significant difference between female and male Japanese students regarding the semantic formula ADVERB. JFU uttered semantic formula ADVERB significantly more than JMU. Table 4: Mention rates of semantic formulas in the first turn of JU and KU in apologies X2ANOVA (two w ay) by arcsin asin Female Male x2 Post hoc analysis 1 JU 100.0 94.5 Nationality 1.63 KU 95.2 96.6 Gender 2.29 Interaction 4.19 2 JU 54.3 50.9 Nationality 0.78 KU 45.2 48.3 Gender 0.00 Interaction 0.24 3 JU 43.5 58.2 Nationality 11.20 * KU>JU KU 78.6 65.5 Gender 0.00 Interaction 4.91 4 JU 19.6 3.6 Nationality 6.64 * KM>JM KU 28.6 17.2 Gender 9.25 * JF>JM Interaction 0.97 5 JU 0.0 0.0 Nationality 36.92 * KU>JU KU 21.4 10.3 Gender 1.35 Interaction 1.35 6 JU 0.0 1.8 Nationality 55.85 * KU>JU KU 26.2 31.0 Gender 1.99 Interaction 0.38 Female Male X2ANOVA (two way) by arcsin asin X2 Post hoc analysis 7 JU 0.0 1.8 Nationality 5.82 * KF>JF KU 7.1 3.4 Gender 0.14 Interaction 2.71 Note: *p<0.05, Unit: %, 1: Apology, 2: Fact, 3: Reason, 4: Adverb, 5: Address term, 6: Interjection, 7: Others, JU: Japanese university students, KU: Korean university students, JM: Japanese male university students, JF: Japanese female university sutdents, KM: Koran male university students, KF: Korean femail university students. 4.1.2 JU/KU * JIU/KIU Table 5 shows the rates that semantic formulas were used in the first turn by JU, KU, JIU, and KIU. Significant differences were found for semantic formulas FACT, REASON, and INTERJECTION because of the main effect of using English in daily conversation. The rates that the semantic formula FACT was used by JIU is significantly higher than JU's. KU uttered the semantic formulas REASON and INTERJECTION significantly more than KIU. Table 5: Mention rates of semantic formulas in the first turn of JU, KU, JIU, and KIU in apologies X2ANOVA (two way) by arcsin asin Japanese Korean x2 Post-hoc analysis 1 JU/KU 97.0 95.8 English 0.03 JIU/KIU 97.1 94.7 Nationality 0.48 Interaction 0.05 2 JU/KU 52.5 46.5 English 7.68 * JU/JIU:JIU>JU JIU/KIU 76.5 57.9 Nationality 3.83 Interaction 1.13 3 JU/KU 51.5 73.2 English 12.20 * KU/KIU:KU>KIU JIU/KIU 50.0 29.8 Nationality 0.02 Interaction 10.67 * JU/KU:KU>JU,JIU/KIU:JIU>KIU 4 JU/KU 10.9 23.9 English 2.70 JIU/KIU 5.9 14.0 Nationality 5.53 * JU/KU:KU>JU Interaction 0.07 5 JU/KU 0.0 16.9 English 0.01 JIU/KIU 2.9 7.0 Nationality 15.32 * JU/KU:KU>JU Interaction 6.05 * JU/JIU : JIU>JU,KU/KIU :KU>KIU X2ANOVA (two way) by arcsin asin Japanese Korean x2 Post-hoc analysis 6 JU/KU 1.0 28.1 English 5.08 * KU/KIU:KU>KIU JIU/KIU 2.9 3.5 Nationality 12.80 * JU/KU:KU>JU Interaction 11.03 * 7 JU/KU 1 4.2 English 13.91 * JIU/KIU>JU/KU JIU/KIU 14.7 15.8 Nationality 0.84 Interaction 0.47 Note:*p<0.05, Unit: %, 1: Apology, 2: Fact, 3: Reason, 4: Adverb, 5: Addressterm, 6 :Interjection, 7: Others, JU: Japanese university students, KU: Korean university students, JIU: Japanese international university students who live in the U.S., KIU: Korean international university students who live in the U.S. 4.2 The second turn 4.2.1 JU * KU Table 6 below shows the rates that semantic formulas were used by JU and KU in the second turn. Table 6: Mention rates of semantic formulas in the second turn of JU and KU in apologies X2ANOVA (two way) by arcsin asin Female Male X2 Post hoc analysis 1 JU 35.6 9.1 Nationality 0.23 KU 26.2 20.7 Gender 8.96 * JU: FU>MU Interaction 4.06 * MU: KU>JU 2 JU 11.1 12.7 Nationality 10.87 * FU: JU>KU KU 0.0 6.9 Gender 4.78 * KU: MU>FU Interaction 3.29 3 JU 97.8 87.3 Nationality 7.02 * JU>KU KU 88.1 75.9 Gender 8.02 * FU>MU Interaction 0.17 4 JU 2.2 1.8 Nationality 22.79 * KU>JU KU 26.2 13.8 Gender 1.66 Interaction 1.15 5 JU 0.0 0.0 Nationality 42.39 * KU>JU KU 11.9 24.1 Gender 1.47 Interaction 1.47 Female Male X2ANOVA (two way) by arcsin asin X2 Post hoc analysis 6 JU 2.2 1.8 Nationality 1.08 KU 2.4 6.9 Gender 0.52 Interaction 0.88 Note: *p<0.05, Unit: %, 1: Apology, 2: Thanks, 3: Ticket, 4: Movie, 5: Compensation, 6: Others, JU: Japanese university students, KU: Korean university students, JM: Japanese male university students, JF: Japanese female university sutdents, KM: Koran male university students, KF: Korean female university students. Regarding nationality, the ANOVA revealed significant differences between Japanese and Korean for all semantic formulas in the second turn except OTHERS. KMU uttered the semantic formula APOLOGY significantly more than JMU. KFU uttered the semantic formula THANKS significantly more than JFU. JU uttered semantic formula TICKET significantly more than KU. KU uttered the semantic formulas MOVIE and COMPENSATION significantly more than JU. There are three significant differences in regards to gender in the second turn. First, both JFU and KFU uttered the semantic formula TICKET significantly more than JMU and KMU, respectively. Secondly, JFU uttered the semantic formula APOLOGY significantly more than JMU. Finally, KMU uttered the semantic formula THANKS significantly more than KFU. 4.2.2 JU/KU * JIU/KIU Table 7 shows the results of the ANOVA for the rates that semantic formulas were used in the second turn. The JIU uttered the semantic formula APOLOGY significantly more than JU, while KU uttered APOLOGY significantly more than KIU. JIU uttered the semantic formula COMPENSATION significantly more than JU. Table 7: Rates that semantic formulas were used in the second turn of JU, KU, JIU, and KIU in apologies X2ANOVA (two way) by arcsin asin Japanese Korean X2 Post-hoc analysis 1 JU/KU 21.0 23.9 English 0.00 JIU/KIU 38.2 10.3 Nationality 5.26 * JIU/KIU: JIU>KIU Interaction 7.93 * JU/JIU:JIU>JU, KU/KIU:KU>KIU 2 JU/KU 12.0 2.8 English 0.07 JIU/KIU 8.8 3.4 Nationality 5.14 * JU/KU: JU>KU Interaction 0.28 X2ANOVA (two way) by arcsin asin Japanese Korean X2 Post-hoc analysis 3 JU/KU 92.0 83.1 English 3.14 JIU/KIU 82.4 75.9 Nationality 2.67 Interaction 0.18 4 JU/KU 2.0 21.1 English 0.16 JIU/KIU 5.9 17.2 Nationality 15.17 * KU/KIU>JU>JIU Interaction 1.33 5 JU/KU 0.0 16.9 English 5.61 * JU/JIU: JIU>JU JIU/KIU 5.9 22.4 Nationality 25.51 * KU/KIU>JU/JIU Interaction 1.75 6 JU/KU 2.0 4.2 English 9.38 * JIU/KIU>JU/KU JIU/KIU 11.8 15.5 Nationality 0.79 Interaction 0.01 Note: *p<0.05, Unit: %, 1: Apology, 2: Thanks, 3: Ticket, 4: Movie, 5: Compensation, 6: Others, JU: Japanese university students, KU: Korean university students, JIU: Japanese international university students who live in the U.S., KIU: Koran international university students who live in the U.S. 4.3 The third turn 4.3.1 JU * KU Table 8 shows the rates that semantic formulas were used in the third turn by JU and KU in apologies. Table 8: Rates that semantic formulas were used in the third turn by JU and KU in apologies X2ANOVA (two way) by arcsin asin Female Male X2 Post-hoc analysis 1 JU 21.7 7.3 Nationality 5.25 * KM>JM KU 31.0 20.7 Gender 6.13 * JF>JM Interaction 0.49 2 JU 97.8 87.3 Nationality 18.02 * JU>KU KU 71.4 75.9 Gender 1.53 Interaction 4.02 3 JU 43.5 32.7 Nationality 3.20 KU 54.8 44.8 Gender 2.54 Interaction 0.01 X2ANOVA (two way) by arcsin asin Female Male X2 Post-hoc analysis 4 JU 6.5 20.0 Nationality 1.67 KU 16.7 20.7 Gender Interaction 3.74 1.35 5 JU 0.0 0.0 Nationality 54.36 * KU>JU KU 42.9 6.9 Gender Interaction 11.38 11.38 * * KF>KM 6 JU 10.9 7.3 Nationality 124.17 * KU>JU KU 73.8 75.9 Gender Interaction 0.08 0.43 7 JU 21.7 9.1 Nationality 2.92 KU 31.0 17.2 Gender Interaction 6.58 0.01 * JF/KF>JM/KM 8 JU 6.5 9.1 Nationality 21.46 * KU>JU KU 21.4 41.4 Gender Interaction 4.03 1.62 * KM>KF Note: *p<0.05, Unit: %, 1: Apology, 2: Thanks, 3: Suprise, 4: Complement, 5: Compensation, 6: Interjection, 7: Adverb, 8: Others, JU: Japanese university students, KU: Korean university students, JM: Japanese male university students, JF: Japanese female university sutdents, KM: Koran male university students, KF: Korean femail university students. An ANOVA was conducted in order to determine whether the speakers' use of semantic formulas is different depending on the speaker's nationality or gender. The ANOVA revealed that KU's use of semantic formulas is significantly higher than JU's for the semantic formulas COMPENSATION and INTERJECTION. And it was revealed that KMU uttered the semantic formula APOLOGY significantly more than JMU. The ANOVA also revealed that JU's use of APOLOGY is significantly higher than KU's. 4.3.2 JU/JIU * KU/KIU Table 9 shows the use of semantic formulas in the third turn by JU, KU, JIU, and KIU in apologies. Table 9: Rates that semantic formulas were used in the third turn of JU, KU, JIU, and KIU in apologies X2 ANOVA (two way) by arcsin asin Japanese Korean X2 Post-hoc analysis 1 JU/KU 13.9 26.8 English 2.21 JIU/KIU 8.8 17.2 Nationality 4.71 * JU/KU: KU>JU Interaction 0.07 2 JU/KU 92.1 73.2 English 4.22 * KU/KIU: KU>KIU JIU/KIU 88.2 53.4 Nationality 24.65 * JU/JIU>KU/KIU Interaction 1.13 3 JU/KU 37.6 50.7 English 0.92 JIU/KIU 41.2 34.5 Nationality 0.23 Interaction 2.30 4 JU/KU 13.9 18.3 English 5.02 * JU/JIU: JU>JIU JIU/KIU 2.9 12.1 Nationality 3.37 Interaction 0.87 5 JU/KU 0.0 28.2 English 13.00 * JU/JIU: JIU>JU JIU/KIU 14.7 36.2 Nationality 37.32 * JIU/KIU>JU/KU Interaction 5.36 * 6 JU/KU 8.9 74.6 English 0.84 JIU/KIU 50.0 39.7 Nationality 22.89 * JU/KU: KU>JU Interaction 40.26 * JU/JIU : JIU>JU,KU/KIU : KU>KIU 7 JU/KU 14.9 25.4 English 2.17 JIU/KIU 8.8 17.2 Nationality 3.78 Interaction 0.00 8 JU/KU 7.9 29.6 English 0.01 JIU/KIU 14.7 19.0 Nationality 6.85 * JU/KU: KU>JU Interaction 3.07 Note: *p<0.05, Unit: %, 1: Apology, 2: Thanks, 3: Suprise, 4: Complement, 5: Compensation, 6: Interjection, 7: Adverb, 8: Others, JU: Japanese university students, KU: Korean university students, JIU: Japanese international university students who live in the U.S., KIU: Koran international university students who live in the U.S As Table 9 shows, there are four significant differences between JU/KU and JIU/KIU. As for the semantic formulas THANKS and INTERJECTION, KU's rates are significantly higher than KIU's. However, JIU's use of the semantic formula INTERJECTION is higher than JU's. For the semantic formula COMPLIMENT, JU's use is significantly higher than JIU's. 5. Discussion 5.1 Explanations for apology preferred by Korean As shown clearly in the results for the rates of use the semantic formulas above, Korean speakers tend to explain their situation positively, while Japanese speakers leave the understanding of the situation to the interlocutors. As for the semantic formula REASON in the first turn (KF: 78.6%, KM: 65.5%, JF: 43.5%, JM: 58.2%; %2= 11.20, p<0.05), Korean students tried to explain the reasons why they were late for the appointment to the listeners actively. In contrast, only about 50% of Japanese students referred the reasons despite having a particular reason that the lecture finished late in the DCT. Surprisingly, little attention has been given to the strategy of apology in the field of comparison study between Japanese and Korean, since it is easy to assume that there are few differences between Japanese and Korean communication because of cultural, linguistic, and geographic similarities. Most of the studies that are associated with apology have focused mainly on the expressions of apology in Japanese and Korean (Ogoshi, 1993; Kim, 1996; Jin, 2004; Hong, 2006;). However, there are some contrastive studies of strategies for apology between Japanese and American speakers. Kondo & Taniguchi (2007) compared the apology strategies between Japanese and American English speakers and found that the impressions of giving reasons for their apology are different. Japanese listeners take the reason as a "defense", while American speakers, who speak English as a speaker-responsible language according to Hinds (1987), regard it as a "polite explanation" in apologies. In her comparison of strategies of apology between Japanese and Americans, Ikeda (1993) also described that American speakers use the explanation strategy more than Japanese. The results of Ikeda (1993) are similar to those of the present study. Ikeda (1993) also pointed out that Japanese speakers tend to utter only apology expressions without any other strategic utterances, while American speakers utter not only explanation but also compensation with apology expressions. The ANOVA revealed that Korean speakers utter compensation significantly more than Japanese speakers in the present study. In the third turn, Japanese students utter just expressions of apology or appreciation, while Korean students added compensation when they knew that their friends had bought two tickets for themselves already (KF: 42.9%, KM: 6.9%, JF: 0.0%, JM: 0.0%; x =54.36, p < 0.05), as shown: Friend: "Tickets? I have bought two tickets already." Japanese: "Arigato." (Thank you) Korean: "Komawe. Nayka phapkhon salkey." (Thank you. Let me buy some popcorn.) Korean speakers used semantic formula COMPENSATION in the second turn (KF: 11.9%, KM: 24.1%, JF: 0.0%, JM: 0.0%; =42.39, p < 0.05) to compensate their friends for being late by offering that they want to buy their friends a ticket, while no Japanese speakers uttered COMPENSATION and they only asked their friends to go buy tickets, or asked whether the friends had bought the tickets already or not, as shown in the conversation below. Friend: "That's OK. It couldn't be helped because the class finished late." Japanese: "Gomen. Tokorode chikettowa mo katta?" (I am sorry. By the way did you buy your ticket already?) Korean: "Mian. Kutaysin thikheysun nayka salkey." (I am sorry. I will buy a ticket for you then.) According to Ikeda (1993), the reason for such differences between Japanese and American speakers is caused by the difference in attitudes to face (Brown & Levinson, 1987) in both countries. In other words, the face to Japanese speakers is deeply related to be admitted and accepted by their interlocutors because of the vertical social-cultural structure (tate shakai) in Japan. It is for that reason that Japanese speakers tend to avoid using the explanation strategy, especially when they talk to a person who is older or of higher status than themselves, and they try to put weight on conveying an apology expression itself and apologize efficiently. Japanese and Korean social structures are fairly similar compared to that of the U.S., thus it is hard to explain the phenomena by just the difference of social structures, since the tendency of Korean speakers to give a reason for their apology in the results of the current paper is similar to that of American speakers. It is suggested that Korean is a speaker-responsible language similar to American English, since Korean and American English share some features with respect to giving extra explanations to help listeners' understanding of utterances, even though Korean was categorized as a listener-responsible language by Hinds (1987). There is no significant difference between Japanese and Korean for the utterance rate of the semantic formula COMPLIMENT in the third turn (KF: 16.7%, KM: 20.7%, JF: 6.5%, JM: 20.0%). However, it was found that the ways of uttering COMPLIMENT were different in Japanese and Korean. Korean speakers tend to utter a complete sentence in order to comment about and convey their thankful emotions for receiving a movie ticket, while Japanese speakers tend to use only a typical word for compliment like "sasuga" (just as one thought) or "yahari" (as expected) as the example below: Friend: "Tickets? I have bought two tickets already." Japanese: "Arigato. Sasuga." (Thank you. It's just (like you to have bought a ticket already)). Korean: "Komawe. Yeksi netapta." (Thank you. It's just like you (to have bought a ticket already)). 5.2 Expressions of apology and thanks Even though the situation set by the researcher in the present study is basically an apology, many Japanese and Korean participants mixed apologetic expressions and thankful expressions in use, especially in the second (Apology; KF: 26.2%, KM: 20.7%, JF: 35.6%, JM: 9.1%, Thanks; KF: 0.0%, KM: 6.9%, JF: 11.1%, JM: 12.7%) and the third turn (Apology; KF: 31.0%, KM: 20.7%, JF: 21.7%, JM: 7.3%, Thanks; KF: 71.4%, KM: 75.9%, JF: 97.8%, JM: 87.3%). Two reasons could account for this. First, the participants who were late reacted differently when their friend arrived on time and bought them a ticket. Some participants felt sorry and some participants felt appreciation in the same situation. Second, it is hard to distinguish between the expressions for apology and thanks clearly in Japanese and Korean compared to "I am sorry" and "Thank you" in English. It is well known that the Japanese expression for apology "sumimasen" is also used as an expression of appreciation in Japanese discourse, as described above (Kim, 1996; Ide, 1998; Yamamoto, 2003). For example, on someone's birthday the following kind of conversations could happen in both Japan and Korea, but not in the U.S: A: "Happy birthday! Here you are. I hope you like this present." B: "I am sorry." (Japanese: Sumimasen.) (Korean: Mianhakey...) In the example of receiving a birthday present above, a Korean speaker might not say as conclusively "I am sorry" as a Japanese speaker, that is why little attention has been paid to the use of Korean apology expressions as thankful expressions in past research. It is interesting that Korean speakers say "Mianhakey" which means "I feel sorry I made you to do this for me" while they don't use "I am sorry" in situations showing appreciation. The results of the third turn in the present study show that Korean male speakers uttered apologetic expressions when they found that their friend had bought a movie ticket for them significantly more than Japanese male speakers (KM: 20.7%, JM: 7.3%; x2 =5.25, p < 0.05). Korean female speakers uttered apologetic expressions more than Japanese female speakers (KF: 31.0%, JF: 21.7%) though there was no significant difference between them (about 9.3 percent). Also, Korean male speakers uttered apologetic expressions significantly more than Japanese male speakers (KM: 20.7%, JM: 9.1%; %2 =4.06, p < 0.05) in the second turn when their friend said he or she could understand why they had been late for the appointment. It is reasonable to suppose that Korean speakers try to convey the emotion that they feel sorry for being late for the appointment to the interlocutor actively by uttering apologetic expressions through all the turns. Also, the results of the present study are different from those of Kim (1996) which claimed that Korean speakers barely use apologetic expressions in thanks. Thus, it will be necessary to examine what kinds of expressions are used for apology and thanks in Japanese and Korean in other situations. Korean speakers also used the semantic formulas ADVERB (Turn 1; KF: 28.6%, KM: 17.2%, JF: 19.6%, JM: 3.6%, Turn 3; KF: 31.0%, KM: 17.2%, JF: 21.7%, JM: 9.1%) and INTERJECTION (Turn 1; KF: 26.2%, KM: 31.0%, JF: 0.0%, JM: 0.0%, Turn 3; KF: 73.8%, KM: 75.9%, JF: 10.9%, JM: 7.3%) which modify the apologetic and thankful expressions in order to convey their feeling actively. It has been pointed out that both Japanese and Korean speakers use adverbs more in apologetic expressions than thankful ones if the feelings of apology are stronger or the interlocutor is older than the speakers (Kim, 1996). The results of the current study clearly show that Korean speakers use more semantic formulas ADVERB and INTERJECTION right before or after expressions of apology or thanks compared to Japanese speakers in corresponding situations. It was found that Korean speakers used the semantic formula ADDRESS TERM in the first turn while no Japanese speakers used ADDRESS TERM (KF: 21.4%, KM: 10.3%, JF: 0.0%, JM: 0.0%). The results of this study confirm the claim of Yoon (2008) that Korean speakers use address terms as contextualization cues (Gumperz, 1982) more frequently in conversations compared to Japanese speakers. In the present study, many Korean speakers uttered "Chinguya" which means "Hey, friend" in order to attract the listener's interest and to make sure that the relationship between the speaker and the listener (good friends) stays unchanged and established before or after saying "I am sorry." 5.3 Influence of the daily use of English in the responsibility for understanding utterances In regards to whether the daily use of English, which is classified as a speaker-responsible language, can influence the ways Japanese and Korean speak their native languages, the results show that JIU used more semantic formulas than JU generally. The ANOVA revealed that JIU uttered the semantic formulas FACT (JU: 52.5%, JIU: 76.5%) and ADDRESS TERM (JU: 0.0%, JIU: 2.9%) in the first turn, APOLOGY (JU: 21.0%, JIU: 38.2%) and COMPENSATION (JU: 0.0%, JIU: 5.9%) in the second turn, and COMPENSATION (JU: 0.0%, JIU: 14.7%) and INTERJECTION (JU: 8.9%, JIU: 50.0%) in the third turn significantly more than JU. It is interesting that JIU uttered FACT significantly more than JU in apologies and FACT is the only semantic formula that JIU and KIU uttered more than JU and KU, respectively, though there is no significant difference between KU and KIU (KU: 46.5%, KIU: 57.9%) in the first turn. Regarding the semantic formulas ADDRESS TERM in the first turn and COMPENSATION in the second turn, nobody in JU uttered either semantic formula. The data suggests that Japanese, which is classified as a listener-responsible language, could be influenced by English, which is regarded as a speaker-responsible language. 6. Conclusion The present study set out to investigate the differences between Japanese and Korean speakers' responsibility for the understanding of utterances in a conversation. The experiment was carried out to compare how semantic formulas are used in utterances as helpful information for the listeners in corresponding situations in both Japanese and Korean. To answer the first research question, the results of the current study suggest that Korean should be classified as a speaker-responsible language for the understanding of utterances in conversations, since Korean speakers uttered more semantic formulas than Japanese speakers. To answer the second research question, Korean speakers uttered semantic formulas REASON, ADDRESS TERM, ADVERB, INTERJECTION, and COMPENSATION in apologies significantly more than Japanese speakers. Finally, to answer the third research question, it was found that the daily use of American English influenced Japanese and Korean speakers' use of their native language, especially in the case of Japanese speakers. References Beebe, L. M., Takahashi, T., & Uliss-Weltz, R. (1990). Pragmatic Transfer in ESL Refusals. In R. C. Scarcella, E. Anderson, & S. C. Krashen (Eds.), On the Development of Communicative Comoetence in a Second Language (pp. 55-73). New York: Newbury House. Blum-Kulka, S., House, J. & Kasper, G. (1989). Cross-Cultural Pragmatics: Request and Apologies. Norwood, NJ: Ables. Brown, P. & Levinson, S. (1987). Politeness: Some universals in language usage. Cambridge: Cambridge University Press. Gumperz, J. J. (1982). Discourse Strategies. Cambridge: Cambridge University Press. Hinds, J. (1987). Reader Versus Writer Responsibility: A New typology. In U. Connor & R. B. Kaplan (Eds.), Writing Across Languages: Analysis of L2 Text (pp. 141-152). MA: Addison-Wesley. Hinds, J. (1990). Inductive, Deductive, Quasi-inductive: Expository Writing in Japanese, Korean, Chinese, And Thai. In U. Connor and A. M. Johns (Eds.), Coherence in Writing: Research and Pedagogical Perspectives Alexandria (pp. 89-109), VA: Teachers of English to speaker of Other Languages Inc. Hong, M. (2006). Kansyato syazaihyogenno nikkanhikaku [A contrastive study of expressions of thank and aplogy between Japanese and Korean]. Nihongogaku [Japanese linguistics], 25(6), 84-89. Ide, R. (1998). 'Sorry for your kindness': Japanese interactional ritual in public discourse. Journal of Pragmatic, 29, 509-529. Ikeda, R. (1993). Syazaino taishyo kenkyu: Nichibe taisyokenkyu [A contrastive Study of apology: A contrastive study between Japanese and American]. Nihongogaku [Japanese linguistics], 12(12), 13-21. Jin, S. (2004). Nikkanniokeru syazaihyogenno ikkosatsu [A study of apology expressions in Japanese and Korean]. Nihongo nihonbunka kenkyu [Studies of Japanese Language and culture], 14, 23-35. Kim, Y. (1996). Ilponekyoyukuy kwancemeyse pon hankwukinuy kamsawa sacoyphyohyeney kwanhan kochal [The expressions of apology and Thank in Korean the view from Japanese education]. Ileilmwunhak [Japanese language and literature], 6, 213-227. Kondo, F., & Taniguchi, H. (2007). A Comparative Study of Percetions of Apology Stategies between Japanese and Americans. Gendai Syakaigaku [Modern Sociology], 9, 131-154. Naotsuka, R., & Sakamoto, N. (1981). Mutual understanding of different cultures. Tokyo: Taishukan. Ogoshi, M. (1993). Shazaino taisho kenkyu: Nikan taisho kenkyu [A contrastive study of apology: A contrastive study between Japanese and Korean]. Nihongogaku [Japanese linguistics], 12, 29-38. Takigawa, Y. (2006). Misunderstandings in Intercultural Communication: Different Strategy in Making the point in a story Between Japanese and Americans. Osakajogakuindaigakukiyo [Memoirs of Osaka-Jogakuin-University], 3, 15-23. Tao, L. (2007). The Construction of Semantic Formulas on Apologizing Expressions n the Chinese. Ningensyakaikankyookenkyuu [Human-Socio Environment Studies], 14, 19-38. Yamamoto, M. (2003). Kanshanoshazaihyogen 'sumimasen'; 'sumimasen'ga kanshato shazaino ryohono imio motsuwak [Two meaning of "Sumimasen": Apologies and Thankse. Shinshudaigaku ryugakusenta kiyo [Journal of International Student Center, Shinshu University], 4, 1-13. Yoon, S. (2008). Comparison between Korean and Japanese address terms as a contextualization husband-wife's dialogue. Inquiries into Korean Linguistics, 3, 377-387. Yoon, S. (2009). A Contrastive Study of Metacommunicative Functions of Address Terms between Husband and Wife in Korea and Japan. Current Issues in Unity and Diversity of Languages: Collection of the Papers Selected from the International Congress of Linguists (CIL), 18, 3819-3831. Appendix 1 The DCT written in Japanese fcftf : ( ) fcftf : ( ) fcftf : ( ) The DCT written in Korean ( ) r ( ) ^ 2 ^ 4 ( ) ^ s Appendix 2 The semantic formulas written in Japanese and Korean Table 1: Semantic formulas and examples in the first turn in university students' apologies Semantic formulas Examples 1 APOLOGY 2 FACT "mfc" 3 REASON 4 ADVERB 5 ADDRESS TERM "oo^v " o o °>" 6 INTERJECTION "fc, (r^) " 7 OTHERS 'Vo^fc?" Table 2: Semantic formulas and examples in the second turn Semantic formulas Examples 1 APOLOGY "r'^fc" 2 THANKS "fc^^ 5" 3 TICKET 44" 4 MOVIE 5 COMPENSATION 6 OTHERS Table 3: Semantic formulas and examples in the third turn Semantic formulas Examples 1 APOLOGY "r'^fc" 2 THANKS "fc^^ 5" 3 SURPRISE 4 COMPLEMENT 5 COMPENSATION "44 w 4^1" 6 INTERJECTION "fc, (fc^£5)" 7 ADVERB 8 OTHERS 4 4^