Participants 26

Most researches have involved several to thirty subjects per individual grouping and claimed to have significant results (Cohen & Olshtain, 1981; Kasper & Dahl, 1991; Lipson, 1994;

Maeshiba et al., 1996; Olshtain & Cohen, 1983; Suszcynska, 1999). There were 60 Chinese national-university-level students (3^rd grade and up Chinese students at National Sun-Yat-Sen University and National Kaoshiung Normal University in Taiwan)¹ and 60 American university level students (3^rd grade and up Caucasian students at University of Illinois at Urbana-Champaign, Harvard University or MIT) between the ages of 20 to 31 participated in this study. The Chinese students were grouped according to their gender (30 male – 30 female)

and their proficiency and exposure to English (30 English majors and 30 non-English majors) and the 60 American participants were grouped according to their gender (30 males and 30 females). Background questionnaires were given to the participants to make sure the Chinese participants were not with special English exposure and the English speakers were all Caucasians. This study hoped that all the possible factors concerning the participants’ selection were well controlled (See Appendix I & III for the background questionnaires for Chinese L1 and English L1 participants).

3.2 Materials

There were three questionnaires for the Chinese informants and one for the American informants along with a simple background questionnaire. The questionnaires for both ethnic groups were pilot tested and proofread by two native speakers of English who were familiar with both Chinese and American cultures to ensure that the scenarios in the questoinnaires were culturally equivalent. The questionnaires were designed to be equal in contents but different in the task language used in the questionnaires and in the ethnicity of the interlocutors offended for the purpose of cross-cultural and cross-linguistics comparisons and discussions.

There were three sets of questionnaires. The Chinese participants answered all the three sets and the American participants only answered set C, which is in English. Detailed descriptions are as follows:

1. Set A: in Chinese －interlocutors in the scenarios are Chinese; —for Chinese L1 data.

2. Set B: in Chinese －interlocutors in the scenarios are Americans—for comparison of interlocutor ethnicity

3. Set C: in English －for English L1 data and Chinese EFL data.

Each set was made into two different orders to avoid possible test-effects. The questionnaires have been piloted by several students in the same schools of the participants and also by several other students in universities located in other parts of the island. Since the contents of the research design have been carefully designed and carefully examined by members of the NSYSU Center for Speech Act Studies, the main focus of the pilot was to make sure the situations were common, imaginable and the factors were clear, and that the data collection method of the SRQ and DCT were not a burden to the participants.

3.2.1 Scenarios

In the questionnaires, the first part was an offending scenario that required the participants to apologize. The participants answered two SRQ questionas followed by a DCT and then two more SRQs. The following is a sample scenario from the English version of the questionnaires.

Situation 1: Your classmate Laura Johnson is a good student. When you talk about the coming quiz, she invites you to her house to study together. Somehow, you were late for nearly 10 minutes. (Questionnaire Order 2)

severity of the situation is ---high---- 5 4 3 2 1 ---- low--the possibility of you apologizing is ---high---- 5 4 3 2 1 ---- low--I would say:

difficulty of the apology for me is ---high---- 5 4 3 2 1 ---- low--likelihood of the apology accepted is ---high---- 5 4 3 2 1 ----

low--(See Appendix II and III for the complete sets of the questionnaires)

In the study, there were 20 unintentional offending scenarios in each set of the questionnaires. Different offense types were designed into the situations according to common offense situations (Appendix IV). Kasper & Dahl (1991: 16) suggested that 20 scenarios “serve as a rough guide” for decisions on common sociolinguistic factors and more factors may post

common sociolinguistic factors in pragmatic studies and for arousing perceptive ratings and productive remedial exchanges to meet the social norms. They are offense severity / imposition gender, social status and social distance. The organization of the scenarios and factors is presented below in Table 3.1:

Table 3. 1 Contextual factors embedded in the scenarios Scenario

forgetting to save in the computer the important files for international conference presentation of the professor you work for

16 5 + － m Knocking down a vase made in the Ming

Dynasty in a restaurant and it breaks!

5 16 + ＝ ＝ m

Your classmate bought a new digital camera. You ask him to let you try it but you carelessly drop it and it breaks!

11 10 + － m Damaging your best friend’s car while

instructing him with driving skills

6 15 + ＋ m

In a department store, you accidentally bump into a man and crashed a very expensive vase in his shopping bag.

7 14 + ＋ f

Rushing to a class, you bumped into a woman. She was really painful, hurt and mad

17 4 + － f Burned the waitress’ hand with the hot

chocolate you ordered from her

10 11 + ＝ ＝ f Accidentally drop your classmate’s

wedding ring and cannot find it anywhere!

3 18 + － f Breaking your best girl friend’s magical

crystal balls from an exotics country

4 17 + ＋ f Knocking over your coffee and burn a lady

beside you and her white coat is stained

2 19 - ＋ m Forgetting to return the book you borrowed

from your advisor last time in his office

18 3 - － m Realizing you forgot to bring your purse

after you opened a pop in a store

15 6 - ＝ ＝ m Late for lunch with a friend for 10 minutes

13 8 - － m Hiccupping very loud during a chat with

your best friend

14 7 - ＋ m Your stomach rumbles very loud in front of

a deliver person

9 12 - ＋ f Presenting yourself in front of your former

teacher but could not recall her name.

19 2 - － f Changing your order in McDonald’s

20 1 - ＝ ＝ f Late for your classmate’s invitation to help

you with the coming quiz

12 9 - － f Misunderstanding your advise, your best friend’s feelings was hurt

8 13 - ＋ f Waiting in line to get a movie ticket, you

accidentally step on a woman’s foot.

Note: Severity ： ＋high / －low

Status ： ＋Status higher / ＝Status equal / －Status lower Distance ： ＋Strange r/ ＝acquaintance / －close friend Gender m ： male interlocutor / f female interlocutor

The factors in the scenarios are severity (high and low), status (high, equal and low), distance (stranger, acquaintance and close friend) and genders. The study designed and revised the scenarios according to the piloted results and through interviewing experts and the pilot subjects. The severity-high scenarios were situations like breaking valuable properties of someone or physically hurting someone. Severity low scenarios were social gaffe or late for a short period of time. Status high interlocutors were professors. Status equal interlocutors were classmates or acquaintances. Status low interlocutors were waiters. Distance far interlocutors were strangers. Distance equal interlocutors were acquaintances and distance close interlocutors were close friends.

3.2.2 Perception Task: Scaled-response Questions

In order to understand the performance or actual speech of learners, learners’ perception of the

“situation” must be clarified. Bergman & Kasper (1993) have realized there are cultural perception differences between cultures and thus constructed a DCT with four

“context-internal” rating questions – severity of offense, offender’s obligation to apologize, likelihood for the apology to be accepted, and offender’s face loss.

Following this pattern, this study adopted the first three questions to examine learners’

perception of apologies over different “context-external” factors – social distance, social status, gender and severity.

“Face loss” was replaced by the difficulty of the apology for face loss was not the main

the aim of this study to know the difficulty of the apology, because the act is a highly face-threatening act and may post great load on the apologizer.

3.2.3 Production Task: Discourse Completion Tasks

After two context-internal SRQ questions, severity of the offense and obligation to apologize, then a strategy eliciting question was asked, followed by two more context-internal SRQ questions, difficulty and acceptability of the apology. The sequence has been pilot-tested and were found to be more logical and easier in filling in the questionnaires. The aim was to make the sequence of the questions logical and user(filler)-friendly.

The written DCT was used to generate the strategies used by Chinese in L1 and L2 and English in L1. Comparing the results of the questionnaire sets among the groups of participants, the study was able to discuss and answer the research questions. The study was able to probe into cultural differences between Chinese L1s and English L1s by comparing the results from questionnaire set A with the results from English L1s in questionnaire set C. By comparing questionnaire set A and set B, the study could investigate the effects of different ethnicity and through comparing questionnaire set B and set C, the study could discuss the effects of language used in the task.