Discourse Strategies in Research Interviews of Gangster Youth

Uncovering the Work of Normative Orders:

A Critical Discourse Analysis of Research Interviews of Gangster Youth

Angel M. Y. Lin, Ph.D., Associate Professor

Department of English and Communication, City University of Hong Kong

83 Tat Chee Ave., Kln., Hong Kong

E-mail:

T. Wing Lo, Ph.D., Associate Professor

Department of Applied Social Studies, City University of Hong Kong

83 Tat Chee Ave., Kln., Hong Kong

E-mail:

Paper presented at the International Conference on Language and Social Psychology, City University of Hong Kong, Hong Kong, July 10-13, 2002.

Abstract

In this paper we present a critical analysis of the normative work that is implicitly accomplished (and sometimes contested) discursively in some social work research interviews of teenagers who had joined gangs and/or gangster activities in Hong Kong. By critically analyzing the deployment of discourse formats and discourse sequencing in these research interviews, we aim at uncovering the ways in which the normative orders that saturate the consciousness of adult social work practitioners might have led them into discourse practices (often pre- or unreflectively) that impose, or coerce interview participants into co-constructing, the adult world’s normative views of what has happened to these teenagers. Rather than providing a discursive space for a genuine dialogue between the interviewees and the interviewers, between the so-called “deviant” teenagers and “corrective” adults, these research interviews might serve merely to reproduce the mainstream adults’ versions, or the social work disciplinary “knowledge”, of what is happening when a teenager engages in gangster activities. The potential consequences of the discourse tools and the potential value of them for use by social work practitioners and researchers in counselling, “corrective” and/or research interactions with gangster teenagers are discussed, with the aim to heighten social work practitioners’ critical awareness of how their own normative worldviews might shape their interactions with teenagers and the possible consequences of such encounters. While recognizing the normative nature of the social work discipline, we believe that a raised critical awareness of their own normative orders will help social work practitioners to be more reflective and perceptive in their interactions with teenagers.

Key Words: critical discourse analysis, social construction of knowledge.

A. The Research Interview: A Special Speech Event

The research interview is not simply a neutral, transparent means of obtaining information from informants, as is sometimes assumed in the positivist sciences. Being a method of research data collection, it is at the same time constitutive of the data it collects (Mishler, 1986, 1999). The research interviewing process is itself a highly interesting interactional phenomenon worthy of detailed analysis in its own right. Epistemologically, understanding the discourse structure and strategies in the research interview will contribute to our understanding of how interview research data are shaped and constituted in particular ways. Ontologically, understanding the interactional process in the research interview will contribute to our understanding of how different subject positions are created or excluded and how particular identities are imposed, resisted or co-constructed in the interviewing process.

The research interview is distinguished from other speech events (e.g., ordinary conversations, debates, lectures) by its typical discourse structure of a series of question-answer pairs. The interviewer occupies the questioner speaking turns and the interviewee the respondent speaking turns most of the time. Conversation analysts have long noted the strength of the question-answer sequence or adjacency pair (Sacks, 1972). The interviewer and interviewee take turns to speak and the interviewer's questioning turn exerts strong interactional pressure on the interviewee both to respond and to respond with material relevant to the question in the immediately preceding questioning turn. No response or irrelevant response will pose high interpersonal pressure on the interviewee. For instance, the interviewee becomes perceived as un-cooperative/unreasonable, the relationship breaks down and the interview cannot continue. Given this special discourse structure of the speech event of the research interview, the interviewer typically possesses much more power than the interviewee in decisions regarding selection, initiation, continuation or change of topics.

In the following sections, we shall present the results of a discourse analysis of 14 audio-recorded research interviews of gangster youth in Hong Kong. The interviews were conducted in community youth centres in the interviewees' neighbourhood by a social worker and/or a college student majoring in social work. Usually the social worker who had befriended the teenager(s) and who introduced the teenager(s) to the interviewers was also present in the interview. The number of interviewees was usually one or two but there were also two interviews, each of four teenagers, and one interview of six teenagers. The interviews ranged from one hour to one and a half hours long. The interviews were conducted as part of a university social work research project with three major aims:

(1) to understand why adolescents join gangs,

(2) to identify the typical activities of gangsters, and

(3) to identify different kinds of effect on adolescents after joining the gangs.

The interviews were semi-structured with the interviewers following a schedule of basic items to be elicited from the teenagers but the interviewers could be flexible and ask other open-ended questions as well. The list of items reflected the three major aims of the research project mentioned above. Among the 30 teenagers interviewed, 23 were male, 7 female, aged from 13 to 19. Some were studying in secondary schools, some working and some unemployed. Most of them had joined gangs for more than one year and some for as long as seven or eight years. Most of them started joining gangs young, in primary school or secondary one (i.e., grade 7). In the next section, we shall analyse the recurrent discourse format and strategies used by the participants in the research interviews. In the final section we shall discuss the implications of the analysis for social work practitioners as well as researchers.

B. Discourse Formats and Strategies in the Research Interview

A discourse format is a recurrent discourse structure characterized by a patterned sequence of speaker-turns each with specific discourse functions (Heap, 1988). For instance, a typical discourse format in the school classroom is Initiation-Response-Feedback (IRF) (Sinclair and Coulthard, 1975; Mehan, 1979). The teacher typically does the initiation by asking a question. The response speaker-turn is typically taken by a student or students and the student response is invariably followed by the teacher's feedback turn in which the teacher typically comments on the student response. Most school lessons are organized by the use of a series of IRF formats. The use of the IFR format enables the teacher to have the power of topic selection, initiation, continuation or change. It also enables the teacher to monitor student attention, understanding and performance. It gives the teacher the opportunity to provide feedback that is tailor-made for a specific student response, and typically it allows the teacher to "have the last word" on a lesson topic. It marks the school lesson as a special speech event in which participants do not have equal power regarding decisions on topic selection and change.

The research interviews analysed in this study are characterized by a recurrent discourse format that somewhat resembles the school lesson IRF format. Let us look at an example to see how the format is used in the interviews. The following example is taken from the early part of a research interview of two gangster girls, Candy (C) and Winnie (W)[1], 15 and 16 years old respectively. In this interview, a social work college student was the interviewer (I). The social worker who had introduced Candy and Winnie to the interviewer was not present in the interview.

Excerpt (1):

(From Data Set 261093A)

(Cantonese utterances are transcribed in the Yale system and English translations are given in pointed brackets immediately following the Cantonese utterances)

I: Gum Winnie neih ne? Neih gaa-yahp-jo ji-houh, di friend yauh dim-yeung tai neih aa?

W: Waaih-jo.. gok dak ngoh!

I: Gok-dak neih waaih-jo, heui-deih gok-dak neih waai-jo, daahn gok-dak neih gok-yeuhng-yeh yauh-mouh bin-dou?

W: Yauh!

<Yes!>

I: Dim-yeung?

<How?>

W: Yauh-haih ok-jo! Chyun-jo gum-yeung!

I: Gum heui-deih yauh-mouh beih-hoi neih, dihng yihng-yihn gum jip-juk neih?

In the above example, we see that the interviewer asks a question about how Winnie's friends saw her after she had joined the gang (1). This question is followed by Winnie's answer that her friends felt that she had changed bad (2). The interviewer acknowledges Winnie's answer by restating it two times and then asks a follow-up question which requires Winnie to further elaborate the changes perceived by her friends (3).

We can schematically represent the discourse format in use in the above example as follows:

1. Interviewer: Question

2. Winnie: Answer

3. Interviewer: Acknowledgement of Answer

This cycle is repeated many times in all the 14 research interviews albeit with some variation in the acknowledgement slot. Sometimes, there is a short acknowledgement particle (e.g., Mh. <Yes.>) or a comment in the place of restatement of the answer. Sometimes the acknowledgement is omitted altogether and a new question is asked immediately.

Since the pre-set list of questions referred to by the interviewers during the interview includes a large section on how different kinds of people (e.g., parents, neighbours, friends, classmates, teachers) see the teenagers after they have joined gangs, the teenagers are in effect led to actively co-construct with the interviewer a corpus of more or less negative statements about how other people see them after they have joined gangs. The interviewer’s power to determine/shape the choice of the topic through the use of this discourse format in effect induces the teenagers into active participation in the co-construction of a corpus of negative statements about themselves and their image that comes with gang membership. Examples of this abound in all the 14 research interviews analysed. For instance, in the following excerpt taken from the second half of another interview, Aah-Mouh (A-M), a 19-year-old male who has joined a gang since grade six, is led to state in his own words that in the eyes of his neighbours he and his gangster friends were not good guys and would cause fear in the neighbours:

Excerpt (2):

(From Data Set 201193A)

I: Di gaai-fong gin-dou neih-deih waahk-je heui-deih goh-di yat-kwahn yih-dong yauh di mat-yeh faan-ying aa?

A-M: Sehng daaih baan yahn cho haih-douh, mh-fong hou-yahn lo!

I: Wui geng aah?

A-M: Wui geng ge ngoh lam!

In the above excerpt, the discourse format of question-answer (with the optional acknowledgement slot omitted) is repeated in the co-construction of statements about the negative, unwelcome image that Ah-Mouh and his gangster friends gave to their neighbours.

In many similar examples, the question-answer(-acknowledgement)[2] discourse format is used quite skilfully by the interviewers (perhaps not consciously) to lead the teenagers into expressing in their own words the negative effect and image that come with gang membership. However, some defiant teenagers might contest or resist collusion in co-constructing negative statements about themselves and gang membership, especially when the interviewer does not ask apparently factual questions but uses conspicuously value-laden words in the questions. For instance, in the following excerpt, Aah-Huhng (A-H), a 19-year-old male who has joined a gang since twelve, refuses to directly answer the interviewer's questions:

Excerpt (3):

(From Data Set 271093A)

I: Gum yauh-mouh gok-dak gan-jo yahn go-di pahng-yauh hou-chih waaih-jo di?

A-H: Gong gwaai neih dou mh-seuin laa!

I: Mh mh... heui-deih haih dim-yeung waaih faat aa?

A-H: Dim-yeung waaih faat aa?

I: Je-haih heui-deih gan-jo yahn ji-hauh wui heui si-haah

A-H: Do-jo yi-di yeh cheuit aa-maa, neih pihng-sih bin wuih waah yauh ho-yih hai sau je!

I: Mh mh.

A-H: Taam dak-yi, maih si leuhng haah.

I: Mh mh... gum-yeung, je-haih yauh-mouh waah, gan-jo yahn ji-hauh, tuhng uk-kei yahn ge gwaan-haih, wui-mh-wui chaa-jo?

A-H: Go-biht gwaa.

I: Go-biht. Gum neih ji-gei ne? ....

In the above excerpt, we can see that Aah-Hung does not directly answer the interviewer's obviously value-laden, leading question (1: "Then do you think those friends who have joined gangs seem to have become a bit bad?"). Instead of saying "yes" or "no", Aah-Hung exposes the hidden assumption that seems to be already firmly held by the interviewer by saying that if he says they are good the interviewer will not believe it (2). This statement of Aah-Hung is remarkable because it seems to have achieved multiple functions:

(1) It enables him to avoid giving a direct answer of either "yes" or "no" to the interviewer's immediately preceding question. If he says "yes", he's condemning his friends, which most probably he does not feel comfortable doing. If he says "no", he is likely to be pestered further with questions about why and how he thinks these friends have not become bad and probably he knows he cannot gather enough warrants acceptable to this mainstream authority figure (the interviewer from the university).

(2) It exposes the assumption hidden in the interviewer's leading question.

(3) It asserts his observation that the real point of the interviewer's question is not one of obtaining information but one of forcing Aah-Hung to take a stance that aligns with the mainstream societal assumption that gang membership causes one to become bad.

By subverting the normal question-answer cycle, Aah-Hung seems to have succeeded in resisting to participate in co-constructing a negative statement about his gangster friends. In turns 3 to 8, he further subverts the question-answer cycle by changing the question-answer sequence into a question-question sequence. When the interviewer asks in what ways his friends are bad (turn 3), Aah-Hung simply asks the same question back (turn 4: Bad in what ways?). The interviewer answers by suggesting that after joining gangs they would try drugs (turn 5). Notice that this is no longer formulated as a question but as a statement or a claim. Aah-Hung then offers another statement/claim about how this is facilitated by the greater degree of availability of drugs (turn 6). Notice also that he does not directly link gang membership to drug-taking. In turn 7, the interviewer acknowledges Aah-Hung's statement and in turn 8, Aah-Hung offers another statement claiming drugs are tried for fun. Turns 6 to 8 are rare instances of utterances that do not fit the question-answer format recurring in the research interviews. These two turns resemble more ordinary conversation between equal interactional partners than research interview conversation.

Aah-Hung's substitution of a question-answer sequence with a question-question sequence (turns 3-4) is quite remarkable as it rarely occurs in research interviews. In fact the interviewer resumes his questioner role soon. In turn 9, she asks another obviously value-laden, leading question about the effect of gang membership on one's relationship with family members. In turn 10, Aah-Hung answers by saying it varies (across people), thereby avoiding giving a definite answer. However, the interviewer zooms in by forcing Aah-Hung to answer the question about his own relationship with his family members. What follows (not shown in Excerpt 3 above) is another lengthy negotiation process between Aah-Hung and the interviewer with Aah-Hung refusing to collude with the interviewer to cast himself and his gangster friends in a negative light.