Responses to AIDS Awareness Discourse: A Cross-Cultural Frame Analysis

Rodney H. Jones

Research Monograph No. 10 The Department of English City University of Hong Kong
1996

Contents ABSTRACT ....................................................................................... i ACKNOWLEDGMENTS .............................................................. iii INTRODUCTION ...........................................................................1 SCHEMA, SCRIPTS AND FRAMES ..................................................................................... 7 FRAMES AND MEDIATED COMMUNICATION.............................................................. 14 DECODING AIDS......................................................................................................... 18 METHODOLOGY ......................................................................... 22 SUBJECTS AND SITUATION ........................................................................................... 25 CHOICE OF COMMERCIALS ......................................................................................... 29 IDEA UNITS ................................................................................................................... 32 FRAMES .......................................................................................................................... 33 MODES OF RESPONSE................................................................................................... 37 SPECIAL FEATURES IN THE DISCOURSE ....................................................................... 38 RESULTS AND DISCUSSION ................................................... 39 LENGTH AND COMPLEXITY OF RESPONSES AND IDEA UNITS .................................. 39 STORIES VS. LECTURES .................................................................................................. 41 VIEWER FRAME .............................................................................................................. 49 EXPERIMENT FRAME ...................................................................................................... 52 2

PATTERNS OF RESPONSE .............................................................................................. 54 INTERPRETATION .......................................................................................................... 56 EVALUATION ................................................................................................................. 62 CONFUSION AND DISTORTION .................................................................................. 72 EMOTION ...................................................................................................................... 81 CONCLUSION AND IMPLICATIONS ..................................... 83 THE LINGUISTIC LEVEL ................................................................................................. 84 THE SOCIAL LEVEL........................................................................................................ 90 DISCOURSE SYSTEMS AND CULTURAL MODELS OF HEALTH EDUCATION ............. 97

Abstract The way people approach topics in communication depends not just on the ideas, attitudes and beliefs they have about the topics but also on the sets of rules and expectations about language and human relationships they bring to the communication. These sets of rules and expectations, referred to by Gee (1990) as cultural models, and by Goffman (1974) and Tannen (1979, 1980, 1984) as frames, are generated from discourse systems which vary across cultures (Scollon and Scollon 1995). Often misunderstandings in intercultural communication result from the fact that participants are communicating and interpreting messages through the frameworks of different discourse systems. Such difficulties are often particularly salient when topics are emotionally charged, such as those involving issues of sexuality, death or moral conventions. This research report examines the topic of AIDS as a potential site of intercultural miscommunication. It focuses on the frames through which members of different discourse communities decode and recall public health messages about AIDS, and how these frames affect their perception of the topic and its relationship to them. A study was conducted to determine the different frames Western teachers of English and Chinese tertiary students in Hong Kong bring to the viewing of AIDS awareness television commercials using a methodology derived from Tannens (1979, 1980, 1984) comparison of Greek and American narratives of a silent film. The study showed that Western teachers and Chinese students responded to the public health messages in very different ways, presenting conflicting interpretations of actions and situations and focusing on different frames of the message. The Chinese i

students tended to see the commercials as stories, focusing their retellings in the narrative frame and spending most of their time describing, interpreting, and judging the actions and intentions of the characters. Western teachers, on the other hand, approached the commercials as lectures, concentrating on the informational content and technical aspects, as well as on their individual experience as viewers. As a result, commercials with strong narrative content and little information were judged negatively by the teachers, while commercials with ambiguous or incomplete narratives were seen as ineffective by the students. The possible reasons for the different ways teachers and students responded to AIDS awareness commercials is discussed from three different perspectives: a linguistic perspective, which examines such issues as interlanguage and communication strategies; a social perspective, which focuses on the expectations participants bring to interaction regarding participant roles, perceived purpose and intended outcomes; and a cultural perspective, which analyzes the results in terms of the different cultural models for education and communication used by the two groups.

ii

Acknowledgments

The study reported here was part of my MATESL dissertation completed in August 1995. I would like to thank Dr. Ron Scollon, who supervised the dissertation and provided ongoing guidance and support in this project, Dr. Martha Pennington, who gave fruitful suggestions on approaching the data, and The Hong Kong AIDS Foundation, AIDS Concern and the Government Information Service for providing materials and information.

iii

Introduction

Song: ...Okay, Rule One is: Men always believe what they want to hear. David Henry Hwang M. Butterfly (1988) In his play M. Butterfly, David Henry Hwang gives to literature one of its most extreme examples of intercultural miscommunication: a French diplomat, blinded by his fantasies and expectations, falls in love with a Chinese actress, who subsequently turns out to be not only a spy, but a man. In an afterword to the script, Hwang remarks on how preconceptions and stereotypes, the myths though which we view reality, play a powerful role in the way we see and understand people in intercultural and intergender communication: ...the myths of the East, the myths of the West, the myths of men, the myths of women--these have so saturated our consciousness that truthful contact between nations and lovers can only be the result of a heroic effort. Those who prefer to bypass the work involved will remain in a world of surfaces, misperceptions running rampant. This is, to me, the convenient world in which the French diplomat and the Chinese spy lived. This is why, after twenty years, he had learned nothing at all about his lover, not even the truth of his sex. (Hwang, 1988, p.100) When we speak, we never speak just a language. We engage in a discourse system made up of rules and expectations and ways of seeing the 1

world which we take for granted. A herd of cows is perceived in the same way by an Italian and an Indian, writes Umberto Eco (1995, p.124), but for the former it signifies an abundance of food, for the latter an abundance of ritual occasions. Gee (1990) calls these sets of rules, expectations and ways of seeing cultural models. These little simplified worlds we have in our heads, he claims, are the basis on which choices about inclusions and exclusions and assumptions of context are made; every word of the language is tied to a myriad of interconnecting cultural models (p. 90). Goffman (1974) sees cultural models as operating through frames, the alignments people bring to each other in discourse, set by participants expectations about what the procedure and outcome of communication should be. In intercultural communication, the most common mistake we make when we approach topics across discourse systems is assuming that the topic is the same for all participants, that were talking about the same thing. On the contrary, the thing we are talking about may be very different for each participant, seen (or not seen) through the particular set of frames they have erected around the topic and the situation surrounding the communication. Watanabe (1990) writes: If two people from different cultures do not share the same expectations about how to interact, that is, perceive differently the unmarked or marked linguistic features that signal a superordinate definition of what is being done by the talk, it is possible that differing interpretations of the situated meaning of what is said are processed by the interactants. On the surface, it looks as if they understand each other, but it is likely that they will find later that they have misunderstood each other. (p.179) 2

The Scollons (1995) define a discourse system as the rules of communication that groups follow, composed of four primary elements: the dominant ideology or worldview at work within the group, the ways members are socialized into the discourse system, the set of preferred forms of discourse for different situations, and the rules of interaction, or face systems, participants abide by when communicating. When people interpret the discourse of those from other speech communities through the cultural models of their own discourse systems, they may attribute attitudes, aims, and meanings which speakers do not intend. Such interpretations may lead to judgments which obscure understanding. This problem is compounded when topics under discussion are emotionally charged or sensitive, such as those linked to deep-seated moral or religious beliefs, the presentation of the self, or persistent emotional or sexual needs, as in the case of M. Butterflys hapless French diplomat. Without cultural models, living would, of course, prove frustrating and tedious, as we would be forced to construct a new context around each individual act of communication. There is a price we pay, however, for mutually agreed upon meanings, for, as clearly as frames display objects, acts and ideas that are placed inside them, they just as surely hide objects, acts, and ideas that are excluded from them. Lewis (1994) observes: The problem with a society that nurtures and guides its citizens towards common meanings is the tendency to suppress not only the ambiguity of things but the very idea of ambiguity. We behave as if the meaning of things were natural and inevitable. The failure to come up with the socially agreed meaning is often interpreted as stupid or troublesome. In many societies, the very act of digression 3

from this semiological control is seen as subversive because it challenges the fixity of the sign, resistance to cultural diversity and, threatening. Herein lies the by degrees, the breeding

ground for racial and religious intolerance. (p.26) Over the past twenty years, sociological and anthropological linguists have provided numerous examples of how conflicting cultural models can lead to misunderstanding between members of different speech communities. Gumperz and Hymes (1972) and Gumperz (1977) have pointed out how, in interethnic communication, conflicting expectations regarding rules of interaction, prosody, and paralinguistic cues can lead participants to form negative judgements about speakers personalities, abilities and intentions. In their work with Athabaskan Indians, Scollon and Scollon (1981) have shown how the different frameworks of Athabaskan and English speakers in such areas as the presentation of the self, the distribution of talk, information structure, and content organization, result in the formation of ethnic stereotypes. The issue of discrimination they write, to a considerable extent is a problem of communication. Much of the miscommunication between different ethnic groups occurs because of the fundamental differences in the values placed on communication itself, and because of differences in interpretation caused by differences in the values of interpersonal face relationships (Scollon and Scollon, 1981, p.4). In the area of gender and discourse, Tannen (1994) has noted that breakdowns in communication occur when men and women bring different expectations to conversational exchanges, with men emphasizing information over relationship, and women emphasizing relationship over information. 4

Studies in health discourse suggest that miscommunication across discourse systems can affect doctor-patient communication and the transmission of important information regarding disease prevention, sometimes with serious consequences. West (1984), for example, studied how the practices and assumptions underlying the professional discourse system of medical practitioners can contribute to misunderstandings between physicians and their clients, and Witte (1991) reports how an education program for the prevention of dysentery among Bolivian Indians using traditional medical information failed to elicit behavior change, success coming only when the principles of viral causes and prevention of the disease were translated into the form of an Andean myth. Cultural models for health education can also differ across generational and ethnic discourse systems within the same community. A report from the U.S. Centers for Disease Control (1995) suggests that AIDS education and prevention materials designed for older gay men may be ineffective for young gay men under 25, a group which in the U.S. has recently shown increasing rates of infection, because of differing values, ways of speaking and patterns of interaction, and Mays (1992) examines how linguistic differences between how black and white gay men structure talk about sex and sexuality can affect the way they interpret information about AIDS risk reduction. In the field of education, the different frameworks teachers and learners bring to the classroom about the subject being taught and about what constitutes appropriate teaching can also lead to varying degrees of misinterpretation and misunderstanding. Researchers like Kaplan (1966) have found that members of different speech communities use different sets of unique, conventionalised rhetorical structures in academic writing 5

involving different strategies of topic introduction and organization. Cortazzi and Jin (1994), in their studies of divergent expectations held by British tutors and Chinese postgraduate students, observe that there are different cultures of learning, internalized in childhood, which may lead students and tutors from different backgrounds to misunderstand each others social and academic behavior because they bring different frameworks both to the types of discourse appropriate to academia and to notions of what constitutes the proper relationship between teachers and learners. This notion of divergent cultures of learning has been further examined by Flowerdew and Miller (1995), who suggest that Chinese students experience of lectures delivered by Western teachers are affected by potential differences in four distinct areas of culture: ethnic culture, local culture, academic culture, and disciplinary culture. Finally, Scollon (1995d) points out that problems with plagiarism in academic settings arise not just from different attitudes towards academic honesty, but also from different social, political, and cultural frameworks through which people approach texts and understand what constitutes ownership of them. The recent popularity of Communicative Language Teaching, which stresses authentic and socially grounded communication, has in some ways increased the potential for intercultural miscommunication in the language classroom. The more language teaching researchers and practitioners emphasize real-world language use, the more they find their classrooms being invaded by real world issues. When classroom discourse turns away from meta-language and more towards real language, teachers can find themselves in awkward situations, especially when classroom topics are sensitive or controversial or are approached by teachers and learners with different sets of culturally constructed expectations and presuppositions, 6

not only about the facts of the topic, but also about how (or whether) it is proper to discuss the topic in a particular setting. This study examines AIDS discourse as a potential site of intercultural miscommunication both in public health discourse and in language classrooms. It focuses on the different frameworks Chinese tertiary students in Hong Kong and their Western English teachers bring to understanding and responding to AIDS educational materials produced by the Hong Kong government. A better understanding of the ways people from different discourse systems perceive AIDS and communication about it may help health educators deal with the difficulties they might encounter when addressing a multi-cultural public, and may also help language educators deal with problems they might encounter when communicating with their students, not just about AIDS, but about any topic for which teachers and learners may have different, culturally determined expectations.

Schema, Scripts and Frames Various researchers in fields as diverse as language teaching, linguistics, sociology and artificial intelligence have made use of the concept of frames in analysing human behavior, interaction and communication. Alternatively referred to as scripts (Shank & Ableson, 1977), schema (Rumelhart, 1975), structures of expectations (Ross, 1975) and cultural models (Gee, 1990), the notion of framing is grounded in the idea that people organize information and knowledge about new events, objects or situations on the basis of their experience. This experience leads them to build up certain structures of expectation through which they interpret and 7

judge the world. Frames are our plans for understanding, for acting, and for talking about events (Dechert, 1983). Bateson (1972), one of the first to make use of the idea of framing, sees frames as psychological concepts that help people to understand the nature and purpose of communication, much like a picture frame delimits and constrains its contents. In many instances, he writes, the frame is consciously recognized and even represented in vocabulary (play, movie, interview, job, language, etc.). In other cases, there may be no explicit verbal reference to the frame, and the subject may have no consciousness of it (Bateson, 1972, pp. 186-187). Tannen (1990) points out that the concept of frames has been used by researchers in two different senses. First, it has been seen as the set of expectations we bring to situations, what others have called schema or scripts. The second sense, more closely associated with the work of Goffman (1974), sees frames as the alignments people bring to each other in human interaction. Keeney (1990), in his development of what he calls recursive frame analysis, attempts to reconcile these two definitions of frames by emphasizing the reciprocal relationship between frames (the alignments people bring to communicative situations) and their contents (various scripts, routines, and sets of expectations). Content, he claims, is as much determined by the frame that surrounds it as the frame is determined by the content it encloses. This study, following Keeney, uses the term in both senses, examining expectations on several levels and from several perspectives, from the ways individual signs and codes signal to participants the kinds of frames being activated in the communication, to the way the overall framing of an event creates and constrains the kinds of signs constructed within it. 8

One popular way to study the cognitive frames people bring to communication has been to analyze subjects responses to linguistic or visual stimuli, as in Tannens (1979, 1980, 1984) analysis of American and Greek subjects retelling of a silent film as part of the Pear Stories project (Chafe, 1980). By examining how her subjects organized their retellings of the film and by isolating instances of what she calls evidence of framing (in the form of linguistic behavior like generalization, evaluation and interperative naming), Tannen was able to explore the different sets of expectations that Greek and American participants used when they viewed the film and formulated their subsequent comments about it. The Greek participants, she found, structured their responses around the narrative aspects of the film (storytelling frame), whereas American participants were more likely to comment upon the technical aspects of the film (film frame) and their experience as viewers of the film (film-viewer frame). It seems that the Americans, she writes, were concerned with presenting themselves as sophisticated movie viewers, while the Greeks were concerned with presenting themselves as acute judges of human behaviour and good storytellers (Tannen, 1980, p.55). In a replication of Tannens study with speakers of California English and Taipei Mandarin, Erbaugh (1990) found that even though both Chinese and American subjects employed similar strategies of chronological sequencing in their responses, the Chinese subjects stories contained more detail and elaboration as well as social and moral interpretation, while the responses of the American subjects tended to contain more personal comments and observations about the film as a film. Frame analysis has also been used in the study of medical discourse, particularly doctor-patient interaction. Evans, Block, Steinberg and Penrose 9

(1986) claim that the lack of shared frames is one of the main reasons for doctor-patient miscommunication. Keeney (1990), using his recursive frame analysis, found many differences in content to frame and frame to frame relationships between doctors and patients, and Chenail (1991), drawing on Keeneys methodology, documents examples of differences between how physicians, cardiologists and parents organize information surrounding medical referrals of young children with heart problems. He attributes these differences to incongruencies between institutional frames and client frames, noting that the doctor, by virtue of being a member of the medical profession, has a different context than a parent, who has membership in another context, their family (Chenail, 1991, p.92). Language teaching has proven a particularly rich area for the analysis of interpretive frames, providing evidence that culturally determined schemata can play as important a role in communicative competence as lower-level linguistic issues like vocabulary and grammar, and that differences in framing by teachers and learners in language classrooms can sometimes lead to miscommunication or misdiagnosis of language problems. Wanatabe (1990), in his study of classroom discussions, found that Japanese students use framing strategies in small group discussions which are different from those used by American students. When asked to give reasons, for example, the Americans framed theirs as briefing while the Japanese framed theirs as storytelling. In her examination of role-plays in a second-language classroom, Wildner-Bassett (1989) suggests that students performing classroom enactment of real-life communication have to reconcile the demands and restrictions of competing frames in the form of simultaneously existing discourse worlds : the world of the classroom, the world of the performance, and the real world which the performance 10

is meant to imitate. The presence of overlapping and sometimes incompatible discourse worlds in language classrooms, she points out, can affect students language at all levels of comprehension and production. Research into learner interlanguage has also increasingly focused on the supersentential level of framing and how it might influence lower levels of linguistic processing. Dechart (1983) studied the verbalization of a cartoon story by an advanced learner of English, finding that accessibility of and familiarity with narrative schemata aids students in retelling stories. Storytelling, he says, is a multidirected top-down and bottom-up process (Dechart, 1983, p. 194) with performance in one direction affecting performance in another. When the frames of communication are familiar to students, they are much more able to pay attention to complex lower level, processing problems specific to the second-language. However, when the framing of a text or task fails to conform to students expectations, the existence of competing speech plans may contribute to student errors and disfluencies. A significant amount of work with issues of cognitive frames or schema and their relationship to culture has been carried out in the area of second language reading. In a theoretical review of the research, Adams and Collins (1979) state that the fundamental assumption of schema theory is: that spoken and written text does not in itself carry meaning. Rather, a text only provides directions for listeners or readers as to how they should receive or construct the intended meaning words of a text from their own, previously acquired knowledge. The evoke in the reader associated concepts, their past

interrelationships and their potential interrelationships. The organization of the text helps readers select among these conceptual complexes. (p. 3) 11

When a reader or viewer approaches a text which fails to conform to expectations, or when he or she attempts to understand the text using culturally inappropriate associations and interrelationships, the process of constructing meaning may be impeded. Gatboton and Tucker (1971), for example, attribute Filipino high school students misunderstandings of American short stories to a phenomenon of cultural filtering. Steffensen, Joag-dev and Anderson (1979) found that cultural knowledge of text type and topic makes texts based on ones own culture easier to comprehend than syntactically and rhetorically equivalent texts based on a different culture. Johnson (1981) showed that the cultural origin of a text has a more significant effect on reading comprehension than linguistic complexity, and Floyd and Carrell (1987) had a similar finding, noting that whereas level of background knowledge for the content schemata for the text played a significant role in both recall and question answering of information from the original text, level of syntactic complexity had no significant effect of either mode of measuring reading responses (p. 103). Cultural models play a role in comprehension, not just in terms of what happens within the story, but also on how the story is structured. Carrell (1984) calls these two aspects of framing content schemata and formal schemata. In a study which compared students retellings of stories with traditional and altered (interleaved) structures, she found that knowledge of supersentential or rhetorical organization of the text aids readers in both comprehension and recall, with enhanced quality of recall when the story was structured with a rhetorical organization that conformed to the readers schema for simple stories (p. 103). A replication of this experiment with second language learners in Hong Kong (Mahoney, Shillaw, and Hull, 1991) confirmed Carrells observations. Like 12

Carrells multi-cultural group of learners, the Cantonese subjects in this experiment demonstrated better recall for standard than interleaved stories, the only difference being that the overall quality of recall was higher for both versions of the story than it was for Carrells subjects, and quantity of recall (number of story nodes included in retellings) was not significantly different between the standard and the interleaved stories. In research particularly relevant to this study, Pritchard (1990) explored the processing strategies proficient readers of English and Palauan use when confronted with culturally familiar and culturally unfamiliar passages. He found that readers make use of different types of strategies when confronted with the two different types of text, adopting strategies of using background knowledge and establishing intersentential ties with culturally familiar texts, while employing strategies of developing awareness, accepting ambiguity and establishing intrasentential ties when confronted with culturally unfamiliar material. In addition, he observed that Americans and Palauans differed in the variety and flexibility of their strategies; the Americans tended to use more different kinds of strategies often involving questioning, hypothesising, or risk taking when difficulty was encountered, while the Palauans relied more on text based or bottomup strategies when reading both familiar and unfamiliar texts. When reading culturally familiar texts, however, the Palauans produced more detailed and elaborate retellings. As in previous studies, significantly more distortions appeared in subjects retellings of culturally unfamiliar texts. Pritchard (1990) concludes that differences in cultural schemata are a significant source of individual differences in comprehension processing (p. 290). 13

Frames and Mediated Communication One useful way to study framing in discourse is through analysing the dominant forms of communication in a society and how they model and reinforce socially acceptable ways of speaking about particular topics. The frames through which we view reality come from our experience with the world, and in modern technological societies our experience with the world increasingly means our experience with electronic media. Meyrowitz (1985), commenting on the power of media to construct and enforce cultural models, writes: Media, like physical places, include and exclude participants. Media, like walls and windows, can hide and they can reveal. Media can create a sense of sharing and belonging, or a feeling of exclusion and isolation. Media can reinforce a them vs. us feeling or they can undermine it. (p. 7) Media thus function as exponents of cultural models and powerful socializing agents through which viewers are apprenticed in the various frames and codes the society uses to communicate. These codes vary from explicit statements to subtle visual cues which signal to members of the discourse system which cultural models are being activated in the message. Television programs, for example, are coded on many different levels from the dialogue and dress of the characters to the way the camera moves. According to Lewis (1994): Even something as simple and innocent as a close-up shot can operate as a highly coded sign, signifying--to those with access to the appropriate cultural codes--different things in different television contexts. In a TV interview, an extreme close-up may signify that the 14

person being interviewed is probably hiding something...In the TV drama, it may signify depth of feeling...or the point of view of another character who is within an intimate distance. To someone who has never seen a photographic image, it may simply signify that a characters face has, mysteriously, grown extremely large. (p. 26) Just as in face-to-face interaction, the codes used in media messages vary among cultures. Skovmand (1992), for example, shows how conventions of audience and participant interaction vary in U.S., Danish, German and Scandinavian versions of The Wheel of Fortune, and Scollon (1995b) observes that characters in American, Hong Kong, Korean and Japanese situation comedies make use of different cultural codes, in the form of speech, coughs, knocks and shuffles, to signal the start of social interaction. For Carbaugh (1988), such codes contain valuable information about a cultures ideology and social practice; in his analysis of patterns of discourse on the American television program Donahue, he argues that the rules of interaction governing conversation on American talk shows both reveal and reinforce deep-seated cultural notions about the self and its relationship to communication. Television commercials, because of their ritualized quality and their attempt to fit the presumed attitudes and values of the target audience, provide a particularly valuable site at which cultural models can be examined. Vestergaard and Schroder (1985) claim that television commercials assert and presuppose certain rules of behavioral normalcy which arise from and protect the dominant ideology in a society. Furthermore, because of the time constraints of a thirty second commercial, authors have to pay special attention to efficiently marking the codes that signal to viewers the frame and footing they are meant to bring 15

to the message, often employing shorthand strategies like problem simplification (Vestergaard and Schroder, 1985), caricature and stereotyping. Therefore, such signals, and the rules of behavioral normalcy they express, tend to be far more salient in a thirty second spot than they might be in a longer message. Linguists have shown that television commercials across cultures not only favor different linguistic structures but also display different rules for interaction and speech act realization. Schmidt, Shimura, Wang, and Jeong (1990), for example, in their comparison of American, Chinese, Korean, and Japanese television advertising, found American commercials to be much more overtly persuasive, in terms of the frequency of suggestions and the frequency with which imperatives were used, while Asian advertising tended to emphasize other functions like entertainment value or the establishment of positive feelings. Similarly, Kumatoridani (1982) points out that Japanese commercials tend to be less direct, straightforward and argumentative than American commercials. In my examination of language use in Hong Kong AIDS awareness advertisements (Jones, 1995), I found that the language in the commercials varied slightly in both tone and content between English and Cantonese versions, with the English versions more direct and personalized and the Cantonese versions more distant and polite. Miracle (1987) claims that American and Japanese television commercials display contrasting structures in terms of the nature of the response they are designed to elicit. American commercials, he says, are constructed on a Learn-Do-Feel model, in which viewers first find out about product attributes, then experience them through the use of such techniques as mini-drama or demonstration, and finally develop an 16

affective response. Japanese commercials, on the other hand, are structured according to a Feel-Do-Learn model in which the generation of an emotional response precedes the presentation of information, much in the same way Asians prefer to establish relationships in face-to-face communication before introducing topics (Scollon and Scollon, 1995). Cultural models function in media messages not just in the way they are encoded, but also in the way viewers interpret them. People from different cultures bring to messages different notions about desirable forms of behaving and speaking as well as different ideas about the proper form and function of the message. Morley (1980), in a study examining the patterns of perceptions of a single television programme among viewers from different social and cultural backgrounds, found that the sociocultural location of the viewers was an important factor in determining how participants understood and responded to the message. Hall, in his influential essay Encoding/Decoding in the Television Discourse (1973), claims that media messages are made to mean by a two step process of encoding, by which producers shape the message into established textual forms, and decoding, by which viewers bring to bear their interpretative framework on the message. People from different communities may bring to messages different codes of decoding, leading to different readings of the message, from the majority reading, which is closest to what the producers intended, to ideological or normative readings, which alter the message to conform to the viewers ideological presuppositions, and oppositional readings, which are formed in direct resistance to the framers intentions.

17

Decoding AIDS AIDS seems a particularly rich topic for the study of framing in mediated communication, not just because it brings together so many sensitive issues, but also because it has forced what in most cultures are relatively private matters, sexuality and disease, into the sphere of public discourse. This public framing or construction (Berger and Luckmann, 1966) of the topic within such forums as the mass media inevitably turns it to the service of the discourse system of those doing the framing, and the discourse produced inevitably reinforces the cultural models within that discourse system. Cultural scholars and media critics have shown how cultural models mirrored in the discourse of AIDS reveal the assumptions the dominant discourse system holds regarding sexuality, deviance, disease and death (Goldstein, 1991; Lupton, 1994; Patton, 1990; Treicher, 1993; Watney, 1989). Tulloch (1989) points out that AIDS messages undergo a series of transformations as they interact with various aspects of professional television practice such as story conferencing, script writing, lighting and set design and direction. Others have examined instances where the way people talk about AIDS illustrates areas of contention between discourses. Seidel (1990), for example, in an analysis of the discourse of AIDS in Uganda, points out how different verbal strategies used by the government and the church reveal different political agendas in the struggle to control the discourse surrounding the issue. Furthermore, Harper (1993), Treicher (1988), Patton (1990), Stoddard (1989), and others have discussed contestations in the framing of AIDS among different ethnic and sexual minorities in North America. Such work has generated calls in

18

both scholarly and popular literature for the production of more culturally sensitive AIDS education materials. Language teachers have also taken up the issue, exploring how students preconceptions about AIDS, shaped by mass media texts and belief systems, are mirrored in their classroom language about the topic (Patton, 1990; Tulloch, 1992). Bowen (1993), for example, in his analysis of college students compositions about AIDS, found in the way students wrote about it a kind of media ventriloquism in which the content and structure of student discourse reflected the construction of AIDS within public discourses (the media, public health campaigns, political dialogues). In Jones (1995), I analyzed the cultural models encoded in twenty AIDS Awareness Announcements in the Public Interest (APIs) produced by the Hong Kong government between 1987 and 1994 using a variation of Goffmans (1974) frame analysis. I found that the campaign, rather than focusing on disseminating information about AIDS, concentrates on constructing the identities of those at risk from AIDS and, using the systems of inclusion and exclusion implicit in all framing, erects a kind of discursive cordon sanitaire between viewers and victims. This is accomplished by presenting people associated with HIV and AIDS as out of frame. On the physical level, they are often filmed as literally off-center or out of focus, not fitting within the borders created by the camera; in the narratives they perform they are portrayed as violating codes of conduct associated with particular places, situations or social positions; and, on the level of the genre of television advertising, they are denied the opportunity for redemption typically offered to characters in television commercials. In order to further explore how the framing of these messages interacts with social practice, I analyzed classroom discussions about AIDS 19

by four groups of Chinese tertiary students enrolled in a course in English for Professional Communication at the City University of Hong Kong. Like Bowen (1993), I observed that the ways students framed AIDS and their relationship to it mirrored the framework encoded in local public health messages. Students saw HIV infection as outside of the framework of their own experience, restricted to and consequent upon violating prescribed rules of interaction emanating from the authority of society and the family. People with HIV/AIDS were framed as objects of public censure, cultural deviants somehow deserving their fate, as illustrated in this exchange in which students role-play their own brand of AIDS education: (three students sitting around a table eating chocolate biscuits) S1: (to people waiting outside classroom) Were doing a comMERcial. (pause as students adjust camera) S1: Okay^...Stella--..Louisa--..do you know Willie got AIDS. S3 & S4: S3: S4: S1: S3: S1: S3: S1: S4: S3: (burst of laughter) AIDS^ Willie got AIDS^ Yeah. How do you...how does he got AIDS? I dont know..I heard it from Mimi. (laughs) You know..his girlfriend. He deSERVED it...he used to have CASual SEX...with He always proud he had so many sexual partners. 20

others.

S1: S1:

And Mimi said he never used CONDOM-Oh-He DESERVE it--

S3 & S4:

(students resume eating chocolate biscuits) (p.37) Analysing how cultural models are encoded in public health messages and how these models are reflected in cultural practice can only give us half the picture of how these messages operate, since meaning resides not just in the production of the message, but also in how it is received and interpreted by viewers. This is particularly relevant in Hong Kong where public health messages are produced for a multi-cultural audience made up of viewers armed with significantly different cultural models surrounding AIDS and AIDS education. R. Scollon (1995) sees television viewing as a kind of watch within which viewers take up stances of identity and power in an asymmetrical relationship with the producers of the message. Their response to the message, therefore, reflects not just how they frame the topic within a system of beliefs and expectations, but also how they frame themselves as receivers of the message. In an earlier analysis of power and identity in news discourse in Hong Kong, Scollon (1994) suggests that receivers of messages take up roles in regard to the discourse corresponding to Goffmans (1981) production roles of animator, author, and principal. These roles reflect the way they perceive their rights and obligations within the discourse. The role of receptor involves the strictly mechanical aspects of reception; the role of interpreter involves making sense out of what is said, filling in gaps according to contextual clues; and the role of judge is taken up when a viewer makes judgements about the value of what is said and so responds to the responsibility taken by the principal. 21

The present study combines Scollons notion of recipient roles with Tannens notion of frames to explore how members of different discourse communities within Hong Kong perceive public health messages on AIDS produced by the government. It seeks to determine how the framework encoded within the messages (Jones, 1995) interacts with frameworks brought to messages by people socialized into different discourse systems, how viewers knowledge and expectations shape what they see and understand in the messages, and how the viewing can affect the way they perceive and respond to the issue of HIV/AIDS. It further explores issues relevant to second language teaching, in particular the use of communication strategies (Corder, 1983; Tarone, 1983) by subjects when confronted with culturally unfamiliar texts or when confronted with the task of recounting culturally familiar texts in a language not their own.

Methodology The main difficulty in studying the reception of mediated communication linguistically is the fact that reception rarely results in texts (Scollon, 1995c). Most research in this area, therefore, depends on subjects producing some kind of analyzable response. Researchers have studied the effects of television commercials from many different points of view, concentrating on such aspects as recall, cognitive response, involvement, attention, attitude, intention to purchase, and emotional response. These variables have been measured through subjects immediate or delayed response to viewing either in the form of questionnaires or free recall, through data on viewing and purchasing behaviour, and through the electronic measurement of brain waves before, during and after viewing 22

(see Thorson, 1989, for review) . Leavitt (1968) and McConville and Leavitt (1968) developed a psycholinguistic model for the analysis of subjects free recall protocols using as their primary measure the number and type of related references in subjects responses, which they defined as statements about relationships between objects, acts, and ideas in the commercial. They further divided such references into those with only narrative emphasis, those showing product-narrative integration, and those showing product-claim integration. The extent to which subjects integrated the narrative of the commercial with the product being advertised, they claimed, functioned as an accurate predictor of subject recall. Research on narrative memory suggests that recall is a selective process in which events are re-interpreted and re-explained according to viewers and readers preconceptions and expectations (Baddeley, 1990). Emmott (1994) claims that people understand narratives through the use of frames which help them construct fictional worlds. For Emmott, frames are not static mental constructs but dynamic tracking systems through which the elements of fictional context such as time, characters and setting, are monitored. Goffman (1975) sees frames operating in oral narrative on several levels, both in framing the events narrated according to available schemata, and framing of the self in relation to the story and the social occasion for telling it. Research into how narrative recall mirrors both cultural models surrounding topics and situations, and the construction of the self in relation to those topics and situations, has been conducted on groups as diverse as pre-school children (Miller et al., 1992; Scollon and Scollon, 1981; Scollon and Scollon, 1984), primary school students (Michaels, 1981), and university lecturers (Cortazzi and Jin, 1994). It is Tannens (1979, 1980, 1984) examination of narrative retellings in the Pear 23

Stories project which makes most explicit use of Goffmans (1975) theory and terminology. In any speech event, Tannen (1991) notes, many levels of framing exist simultaneously and overlap: ...the speakers expectations about being the subject of an experiment...her feelings about having her voice recorded...the speakers expectations about films as well as her expectations about herself as a film viewer... also come into play. Finally, the events, objects and people depicted in the film trigger expectations about similar events, objects and people in the real world and their interrelationships. All these levels of knowledge structures coexist and must operate in conjunction with each other to determine how the events in the film will be perceived and then verbalized. (p. 22) The way subjects construct their comments about a film, she asserts, indicates the particular levels of framing through which they experience the viewing/telling process. Her analysis focuses on four different levels of framing: 1) the subject-of-the-experiment frame 2) the oral storytelling frame 3) the film frame 4) the film-viewer frame Each of these frames is signalled by particular foci within subjects statements. For example, attention to the storytelling frame is shown though focus on the characters and events in the narrative, whereas the film frame is evidenced by concentration on the film as a film, typically making use of cinematic jargon to comment upon the technical aspects what is seen. 24

Along with these broader frames, Tannen also isolates what she calls evidences of framing in the narratives of her subjects, features in respondents language which reveal their attitudes and expectations about the content of the film, or the context of film viewing. Such features as omission, repetition, false starts, hedges, inexact statements, and generalizations, she says, provide clues to the various scripts or schemata operating in subjects responses.

Subjects and Situation This study examines the frames in which people from different discourse communities operate, and how they operate within these frames, when asked to give a retelling of ten Hong Kong government AIDS Awareness APIs selected from the twenty analyzed in Jones (1995) (Appendix 1a). Since one of the main aims of the research is to understand how language teachers and their students communicate (and miscommunicate) about topics for which they may have different cultural models, the two communities I chose to observe were English lecturers in Hong Kong from a variety of Western cultures, and Chinese tertiary students majoring in Teaching English as a Second Language (Appendix 2). Five males and five females were chosen from each population. The students were an extremely homogenous group generationally, culturally and professionally. The teachers, on the other hand, were more diverse, coming from a variety of Western countries--America, Britain, Canada, Australia, and the Bahamas-- each with its distinct system of cultural models. Their ages varied as well (31-53), as did their length of stay in Hong Kong (less than 1 year to 10 years), and the number of years they had been 25

teaching (3 years to 21 years). I argue, however, that these teachers also constitute a cohesive speech community of sorts. First of all, they are all members of what the Scollons (1995) call the Utilitarian Discourse System, with its focus on individualism, positivism, and the role of communication in efficiently transmitting information. They are, in fact, masters of this discourse system: it is what they teach. Membership in the Utilitarian Discourse System is a major difference between the teachers and the students who, while in the process of apprenticing in the system as university students, are not (and may never be) full members, and whose home discourse system may differ in significant ways from the one they are learning. Furthermore, the teachers shared profession places them in the professional discourse system of ESL teachers, which Scollon and Scollon (1995) describe as characterized discursively by a tendency to pay more attention to the form of messages rather than their function and ideologically by a strong support for the individualism and egalitarianism of the Utilitarian ideology (p. 202). From a larger standpoint, comparing the way these two groups respond to educational materials can shed light on the different ways teachers and learners from different cultures experience texts, and how this can affect the teaching of language and subject areas. The subjects were interviewed individually. Each was shown ten Hong Kong Government AIDS awareness APIs produced from 1987 to 1994 in chronological order by date of production. The Chinese students watched Cantonese versions of the commercials, while the Western teachers viewed them in English. The visual channel in both versions was identical, and the voice-overs and dialogue were equivalent translations with only a few exceptions (see Jones, 1995). 26

After each commercial the tape was stopped, and the subjects were asked to answer the question, what happened in the commercial? They were told that they could say as much or as little as they wished. Occasionally, the researcher asked questions in order to clarify certain aspects of the response, but questioning was used sparingly and questions were formulated to be as unobtrusive as possible, not asked until the subjects were finished giving their initial comments, and not introducing new elements for subjects to respond to but rather getting them to expand or clarify elements already generated in their initial responses. The entire process was videotaped, and careful transcripts were made of subjects responses. Initial responses were divided into idea units which were coded and calculated according to the procedure outlined below. Responses to follow-up questions, however, were not included in the coding and calculation. Subjects were also asked to fill in a form as they viewed the commercials indicating for each commercial whether they had seen it before, had not seen it before, or could not remember. At the end of the interview, they were asked to choose which commercial they found the most effective, which they found the least effective, and to give reasons for their choices. The major differences between the protocol followed in this study and that used by Tannen (1979, 1980, 1984) were the language that subjects used in their responses and the relative power relationships between interviewer and subject. Tannens subjects responded in their native language, Americans using English and Greeks using Greek. Since this study aims to explore how framing affects intercultural communication, all of the subjects were asked to respond in the language with which the two 27

groups normally communicate with each other: English. The students in the study (all in their final year of a degree course in Teaching English as a Second Language) had reached a level of English proficiency at which they were capable of completing the task at hand without difficulty. Furthermore, asking students to respond in their second language, while potentially complicating the interpretation of the results, also gives an added dimension to the study, allowing observation of how native language framing might affect second language production. Liebe-Harkort (1989) calls such a perspective interactive ethonolinguistics. The study of non-native language use, she writes, can give us insight into the original cultural background...It is ethonolinguistic inasmuch as it deals with specific characteristics of language groups, but it is more than that, because it deals with a comparison between two cultures...it is the analysis of cultures in action (p. 100). The second difference between Tannens study and the present one involves the interviewer. In Tannens experiment, as in all data collection in the Pear Stories project, subjects were interviewed by members of their own culture of the same sex and approximately the same age. In this experiment, however, the researcher, himself a Western teacher of English, conducted all of the interviews. Thus, teachers were situated in a symmetrical power relationship with the interviewer, while students were in an asymmetrical power relationship with him. Without a doubt, the relative position of power of a speaker in an interview can affect the carefulness and the content of his or her response. Moreover, since teachers and students traditionally have rather fixed and constrained topics and forms of communication with each other, it is very difficult for studentsubjects and teacher-experimenters to talk to each other without drawing 28

upon the expectations of their professional, extra-experimental relationship. This situation, however, can also be seen as an advantage since it more closely corresponds to the real life circumstances at issue in this study: how cultural models affect how people communicate in language classrooms and in public health education, both situations in which communication takes place to varying degrees within a structure of asymmetrical power relationships.

Choice of Commercials Researchers have developed a variety a ways of classifying commercials based on such features as product type, executional style, content attributes, and structure (Thorson, 1989). Puto and Wells (1984) see style and structure of commercials in terms of a continuum rather than distinct categories such as mini-drama, demonstration, and testimonial. They suggest that all commercials contain elements of the transformational, relating to viewers and characters experience with a product, and informational, in which product attributes are specified. Wells (1987) later simplifies this model, characterizing commercials as either lectures, dramas, or a mixture of the two. A lecture attempts to provide the viewer with information, while a drama works by creating a narrative situation through which viewers vicariously experience aspects of the product. The ten ads chosen for this study all contain some narrative elements; they all portray people performing actions. Most also contain an informational channel in the form of a voice-over which gives explicit facts, warnings, or interpretations of the story. Ads of the announcement variety 29

(Pyramid and High Risk Groups), and those containing testimonials of movie stars (AIDS Stuntman and Women Protection) were not included. In Jones (1995) I point out that one characteristic of the Hong Kong governments AIDS Awareness campaign which creates distance between the viewer and the story is a frequent separation of the visual and the audio channel, with the story realized in the visual channel and the audio channel focusing on information. Characters in the stories seldom talk. Rather, they perform a kind of pantomime overlaid by music and a voice-over which presents a message and often a slogan. Seven of the ten commercials chosen for this study were of this type, though the amount of information given in the voice-over varies from the relatively detailed list of facts in MTR to the blunt and simple warning in Youth and Prostitution, Family and Homosexuals. AIDS is a killer disease for which there is no cure. Aids is not restricted to homosexuals. It affects both men and women. Any sexual partner could be an AIDS carrier. Always use a condom. Its not 100% safe, But it is safer. AIDS kills. Use condoms for safer sex. (MTR) It only takes one sexual encounter to pass AIDS onto you. Why risk it? (Youth and Prostitution, Family, Homosexuals) The commercials also vary in regard to the complexity and narrative completeness of the stories they contain. Some, like Homosexuals, render whole stories with rising action, climax, and denouement. Others, like Family and AIDS and Travellers, also contain complete stories, but present them in a series of flashbacks rather than in traditional narrative chronology. Still others, however, contain only episodes or scenes from 30

stories, a couple walking into a mass transit railway station (MTR), a man placing a condom on his bedside table (Safer Sex). The last three commercials (Husband, Girl, Salon) fit more closely with what Wells calls dramas. In these ads the characters actually speak audible dialogue and the voice-over is practically dispensed with, reduced to a slogan at the end of the commercial. Interestingly, these are also the only three commercials written and produced by a professional advertising agency rather than by the Promotions Division of the Government Information Office. The ten commercials used in the study might be placed along Puto and Wellss (1984) informational/transformational continuum as follows: Bar Girls MTR Youth & Prostitution Salon informational------*--------------*------------------------*-------transformational (lectures) Safer Sex AIDS & Travellers Husband (dramas) Family Homosexuals One would expect commercials encoded as primarily informational to elicit a more analytical response from viewers, while more transformational commercials would appeal more to their emotions or intuition. As will be seen, however, this depends heavily on the frameworks of communication and education through which viewers approach the commercials, and how informational and transformational forms of discourse fit into those frameworks.

31

Idea Units In his contribution to the Pear Stories project, Chafe (1980) examines the operation of consciousness in the production of narratives. Consciousness, he observes, seems to work through a series of focuses, the mind resting on one thing and then another in a jerky rather than continuous fashion. This series of focuses constitutes a path, or train of thought, determined by associations between one focus and the next, and by schema--already known paths which have been established for the same or similar areas of information in the thinkers past experience (p. 12). These focuses of consciousness are expressed, Chafe argues, through linguistic structures he calls spurts of language or idea units, each consisting of all or part of a single simple clause. Chafe (1980) sets forth three criteria for identifying idea units: 1) Intonation: most idea units end with a clause-final intonation, usually either a rise or fall in pitch. 2) Pausing: idea units are usually separated by either a slight break in tempo (marked in my transcripts, following Chafe, with two dots) or a longer, measurable pause (marked with three dots). 3) Syntax: idea units tend to consist of a single clause, one verb with accompanying noun phrases. Many begin with conjunctions like and, but, or so. A further criterion implied by Chafe, who sees idea units as embodiments of focuses of consciousness such as temporal orientation, characterization, and evaluation, is that idea units are usually restricted to one type of focus. From the point of view of frame analysis, we can 32

expand this notion to the observation that it is typical for idea units to be restricted to one particular frame and one particular mode of response (see below). Thus, separating subjects responses into idea units provides a way to measure the relative amount of time speakers spend in particular frames, what sorts of language functions are typically performed within those frames, and what sort of paths they follow as they move from idea unit to idea unit, and from frame to frame.

Frames Tannen (1979, 1980, 1984) analyzed subjects responses to the Pear Film in terms of four distinct frames of reference, ranging from the storytelling frame to the subject-of-the-experiment frame. This study, while drawing heavily on Tannens general principles, uses a slightly different collection of frames and criteria for assigning utterances to them. In Jones (1995) I suggested a structure for the analysis of the encoding of television commercials, arguing that media messages function simultaneously within a number of inter-nested frames, corresponding to the physical environment of the television screen (physical frame), the narrative acted out within that environment (narrative frame), the status of the message as a commercial (generic frame), and the position of the message within the wider social discourse (social frame). Tannen combines elements of what I refer to as the physical and generic frames in her concept of the film-frame, in which she places all utterances in which subjects refer to the film as a film, both from the technical point of view of what physically occurrs on the screen, and as a message produced or mediated by various people: actors, director, etc. A 33

silent movie and a public service commercial, however, operate in different ways in terms of aims, content and perceived authorship. Unlike the Pear Film, the AIDS APIs my subjects viewed have clear intentions (to persuade), convey explicit messages (in the form of slogans), and emanate from identifiable principals (Goffman, 1981), in this case, the Hong Kong government. Thus, what Tannen (1979) characterizes as referring to the film as a film functions in this study on two rather distinct levels: there are the physical aspects of the commercial as an audio-visual text, including such aspects as lighting, soundtrack, and the positioning of people and objects within the space of the television screen, which I will continue to label the physical frame; and there is also a level on which the commercials exists as messages, texts whose intent it is to persuade, which I call the generic frame or, more specifically, the commercial frame. For the purposes of this experiment, I also find it necessary to add two further layers from Tannens analysis to my model: the experiment frame, in which subjects attend to the mechanics of the interview itself, and the viewer frame, corresponding to Tannens film-viewer frame, in which respondents focus on themselves as viewers of the commercial and their own processes of understanding, interpreting and interacting with the text (Appendix 4a). The order in which I have arranged these frames is based on the physical and temporal aspects of each: the pictures, or shots that constitute the physical frame exist inside of the plots which make up the narrative frame, which further exist inside of spots, the complete commercial, often including such features as a written slogan at the end of the story, which makes up the commercial frame. This inter-nested organization is not meant to suggest any sort of cognitive hierarchy in which viewers direct their attention first to the innermost frames and then work their way outward. On the 34

contrary, viewers typically move from frame to frame in varied, highly individualistic patterns of processing, sometimes moving inward, sometimes moving outward, and sometimes favoring some interpretative frames while ignoring others altogether. In order to examine how these patterns of response are reflected in subjects comments about the ten commercials, each idea unit in their initial responses was coded according to the level of framing indicated in it using the following criteria: The Physical Frame: Attention to the physical frame is evidenced by statements about the actual pictures appearing on the screen, descriptions of setting, objects, logos and symbols as well as things like music and sound effects. (18) and uh- the lighting is very interesting in this (19) its rather dark and reddish... (T5/Safer Sex) (17) um doom laden music.. (18) and a sort of general atmosphere of gloom.. (T4/MTR) The Narrative Frame: The narrative frame is signalled by descriptions or interpretations of the actions of characters and the relationships between them. (3) and then at first those two men.. (4) go to see some dancing..of the girls (S4/Youth and Prostitution) (4) and then as a mini-bus goes by (breath) (5) the man s-sees a sign of th- the AIDS sign (S3/Homosexuals) 35

The Viewer Frame: Respondents are seen to be operating in the viewer frame when their statements contain information about themselves and their experience with the commercial. Sometimes subjects use generic pronouns like you and we to characterize themselves as viewers. (11) but I was left puzzled by that one.. (T4/Homosexuals) (2) I remember seeing this commercial (3) and it made me laugh.. (T5/AIDS & Travellers) (5) um..(clearing throat)..uh..and then you hear these two guys talking (T6/AIDS and Travellers) (1) were seeing couples (T10/MTR) The Commercial Frame: The commercial frame is characterized by reference to the perceived message of the commercial or the intentions of its makers. (15) and so the message seems to be that.. (16) you die (laugh) (T5/Family) (1) right this one uh..um is obviously aimed at homosexuals..men.. (T8/Homosexuals) The Experiment Frame: The experiment frame includes brackets subjects use to signal either the beginning or end of their response, or comments about the procedure of the interview or their perception of themselves as subjects. (1) (breath) okay..um for this one... 36

(S4/Husband) (12) and...perhaps..thats all for this video (S4/Safer Sex) (22) okay..Im beginning to get the hang of it (T10/Husband) The Social Frame: The social frame is indicated by comments regarding social/cultural knowledge or practice. (17) in Chinese society.. (18) (breath) pe-ple...feels feel shy..to talking about sex (S10/Salon) Modes of Response The modes of response coded onto subjects idea units correspond to what Scollon (1994) sees as the roles viewers take up in relation to the message. Idea units presented as objective observations about, for example, the actions of the characters or the content of the voice-over, were seen as instances of reporting. When subjects made connections or came to conclusions about elements in the commercial based on their own knowledge or experience or the context of the viewing, their statements were coded as interpreting. Interpretations included inferences about the motives or emotions of the characters as well as the intentions of the makers of the commercial. Finally, statements regarding the value, goodness, or effectiveness of elements in the commercial were coded as judging. This is not to suggest that language can be neatly parcelled out into these three rather limited categories; statements of reportage may also include implicit interpretation or evaluation, and interpretative remarks often suggest judgements. Rather, these modes of response seek to 37

describe how subjects package their language within the distinct stances they assume in relation to what they have seen and the act of talking about it. Special Features in the Discourse Tannen (1979, 1980) notes that within broader frames respondents exhibit more subtle language behaviour which reveals presuppositions or expectations about particular aspects of the viewing or the telling about it. She calls these instances of language behaviour evidence of framing, and includes among them such items as omission, repetition, false starts, generalizations, evaluative language, modals, and interpretive naming. Evidence of expectations, what I call, special features in the discourse, often give hints as to respondents preconceived notions or attitudes about such things as settings, character roles, and message content. They also frequently indicate a kind of slippage in modes of response when, for example, a speaker taking up the role of reporter actually performs acts of interpretation or judgement through tone of voice or lexical choice. Several frequently recurring features were identified in transcripts of subjects responses. 1) Distortion (faulty statements about what happened in the commercial) 2) Interpretive Naming (nouns or verbs used to characterize people, objects or actions as being of a special type; describing a man, for example, as a businessman or a woman as a prostitute) 3) Evaluative Language (adjectives or adverbs used which contain implicit or explicit evaluations of people, objects or actions) 38

4) Generalization (single objects, people or actions represented as multiple) 5) Message Signalling (explicitly highlighting statements containing what the subject perceives to be the message of the commercial) 6) Comparison (instances in which subjects compare or contrast commercials or elements within them with other commercials within the experiment) 7) Causal Relationship (explicit statements of cause-effect relationships, often using conjunctions like so and because) 8) Performance (direct quotations of words uttered by the characters or voice-over, often signalled prosodically though alteration of voice quality) 9) Hedging (language indicating a speakers reluctance to commit him/herself to a statement) 10) Emotion (shows of emotion, limited in this study to obvious facial expressions, laughter or sharp intakes or outlets of breath)

Results and Discussion

Length and Complexity of Responses and Idea Units The responses of the Western teachers were generally longer than those of the students, containing an average of 2.5 more idea 39

units, and one commercial, Safer Sex, produced responses from the teachers that were nearly twice as long as those from the students (Appendix 3b). One interesting difference between the length of responses from the two groups was that the length of student responses showed a steady increase as the experiment progressed, growing from an average of 8.1 idea units after the first of the ten commercials to 18.6 (equal to the teachers) after the last one, suggesting that the students were gradually increasing their mastery and/or their involvement with the task, each time able to produce longer and more elaborate responses. Though teacher responses also varied in length (from an average of 11.8 idea units after Bar to 20.9 idea units after Safer Sex), no such pattern of progression is evident; length of teacher responses seemed to have more to do with how much they understood and/or liked a commercial than it did with its sequential position in the experiment. The greater length of the Western teachers responses is in line with Erbaugh (1990), whose American subjects produced longer responses to the Pear Film than the Mandarin speaking Chinese. The greater length, Erbaugh says, comes entirely from the American fondness for adding personal comments such as I liked it, and notes from the film-viewers perspective such as the scene opened or the sound effects were clear (p. 24). A similar penchant for comments in the viewer frame and the commercial frame were observed in the Western teachers in this study (see below). The average length of teachers and students idea units was almost exactly the same (Appendix 3a), though there were some important differences in the fluency and accuracy of the language used by the two groups. Not surprisingly, students responses contained more than twice as many fillers (like um, uh, er), than the teachers, with students averaging 40

one filler every 14 words and the teachers averaging one every 32 words. The teachers responses tended to be linguistically more complex, making more frequent use of such constructions as subordination, subjunctive mood and perfect tenses. Students favored simpler sentence patterns, such as coordination, and made use of more non-standard grammatical and phonological structures.

Stories vs. Lectures In the framing of their responses to the commercials, the Western teachers and Chinese students exhibited extremely disparate patterns, each group focusing on different frames of the message and enacting different receptive roles within those frames (Appendix 4). The students tended to see the commercials as stories, spending more than 60% of their idea units in the narrative frame describing the actions, relationships and motivations of the characters. As expected, all subjects devoted more time to the narrative frame when decoding commercials of a more transformational nature. The students, however, unlike the teachers, even saw commercials closer to the informational side of the continuum in terms of the story they told. (3) a male and a female they walk along the Tsim Sha Tsui eastern coast.. (4) and...I think..uh its night (5) and maybe they have just had a walk...in there.. (6) and they are going home by MTR.. (7) and--..and in the MTR station on the stage.. (8) they see the advertisement box...advertising box... 41

(9) and they can see its a warning about AIDS in the box (10) and so they hold each others hands tightly... (11) and...um...and the speaker in the advertisement always emphasizes AIDS is a (12) ..um...um...untreatable diseases... (S1/MTR) Teachers, on the other hand, spent only a quarter of their time in the narrative frame, choosing instead to focus on the commercials as commercials, describing what they believed the message being conveyed was, discussing techniques used to convey that message, or making conjectures about the framers intentions in nearly 40% of their idea units. So focused were they on the messages contained in the commercials and the devices used to convey those messages that they often they failed to see any story at all. (7) so the theyre targeting it..it seems mostly at heterosexuals (8) which em..is evidenced by that comment (9) and by the fact that all the couples shown are straight... (10) it shows them in sort of familiar setting..very familiar settings (11) like um..eh well at the Arts Centre and the waterfront at Kowloon.. (12) dont know that there was a story there really.. (13) but just the setting is kind of familiar places.. (14) like the MTR (15) so its trying to get you I think to.. (16) feel that you also travel on the MTR (17) you also know..that area of Kowloon.. 42

(18) so it affects you..could affect you um... (T2/MTR) Frequently, teachers described the commercials as collections of separate images unified thematically rather than connected by a narrative: (7) and uh..just a lot of visual images to illustrate.. (8) that uh..you could be the next victim.. (T6/Bar) Sometimes the images were seen as so separate that multiple shots of the same characters or setting were interpreted as shots of multiple characters or settings: (2) they were showing scenes with men and women in bars.. (3) young peoples locations such as discos..um (T2/Bar) The teachers spent a substantial number of idea units, more than twice as many as the students, in the physical frame, describing these images and the music that accompanied them. Like the Americans in Tannens (1979) study, the Western teachers chose images, colors or music rather than the narrative thread characteristic of student responses as the coherence principle (p. 54) of their responses. None of the students mentioned the music in any of their responses. (10) and um..its interesting that the couple is shown mainly in fairly neutral colours.. (11) and all the um..all the ads and warnings that are shown.. (12) the triangle and things like that are in black and white.. (13) which I think adds to the striking terroristic sort of effect.. (T5/MTR) (8) and the music is very interesting here.. 43

(9) because it sounds like Jaws or something like that.. (T5/MTR) (4) the um..this one I like the soundtrack.. (5) sort of seventies porn uh porn funk um.. (T1/Bar) In contrast, the students saw the commercials in terms of actions rather than images, using narrative structure as their coherence principle and formulating their responses as complete stories with setting, complication, climax, and denouement: (6) its in a bar where..where.. most people would find (laugh) a partner... (7) a man..a woman..you know.. um..appraoch a man... (8) and maybe this woman has interest with uh um..in this man..and um--... (9) and--..and then another man..you know they... (10) he stared at another man (laugh).. (11) maybe..well..afterall hes interested in that man..yeah... (12) and then they went away together.. (13) perhaps..uh..uh..they would (breath)..have sex that night (laugh) (S3/Bar) (3) and then first these men go to the bar to see some..dancing of the girls. (4) and then perhaps after they watch such kind of dancing.. (5) perhaps they go to one of the f-fish ball (laugh) fish ball girls.. (6) to some apartment... (7) and then in this apartment there are some hookers.. 44

(8) and then perhaps the manager of this apartment he-... (9) he takes one girl for those men to choose.. (10) and then perhaps they will have sex with her in the rooms later.. (S4/Youth and Prostitution) When they did refer to the commercial, it was usually by way of either introduction or conclusion, like Erbaughs (1990) Chinese students, who rarely referred to the film as a film except at the beginnings and ends of their responses. (1) about the third commercial um-.. (2) the commercial start..begins with a sexy lady... (3) dancing in a theater (S10/Youth and Prostitution) The students did offer social interpretations to the stories they told, but when they explicitly stated what they took to be the message, their interpretations were tacked onto the ends of the stories like codas, in the same way storytellers attach morals at the end of their tales. (12) so..um..I think this advertisement can show me that.. (13) of you..have s-sex..uh..too casually with any girl (14) maybe just for money or just for fun.. (15) then it is dangerous... (16) uh..but..I think the advertisement can..(breath).. (17) give uh one more message is that eh if you... (18) if you get AIDS from the prostitute (19) then you may in you may in turn..transmit the AIDS to your wife (20) then it is more harmful.. (S10/Family) 45

(30) so in this advertisement..s-seems (breath) to arouse..uh..publics awareness... (31) um it told the audience not to avoid..talking about AIDS.. (32) but had to have..um..get in depth discussion (laugh).. (33) and know what AIDS is.. (S1/Salon) Analysis of pronoun usage also confirms the contrast in focus between the students and the teachers. When the students used the pronouns they, them and their, they were four times more likely than the teachers to be referring to the characters in the commercial. 55% of the teachers use of the third person plural pronoun was in reference to the authors of the commercial (e.g. theyre targeting..it seems.. heterosexuals, theyre saying use a condom, they have one scene of some billboard, they fade out, etc.), whereas less than 4% of the students third person plural pronouns were used in this way (Appendix 5a). The difference in the patterns of framing between teachers and students, students operating chiefly in the narrative frame, while teachers concentrated on the commercial and physical frames, is reminiscent of Tannens (1979, 1980, 1984) finding that Greeks and Americans approach the retelling of a film in different ways, the Greeks presenting it as a story while the Americans talked more about the technical aspects of the film and their experience as viewers. It is also in line with Erbaughs (1990) finding that Chinese subjects, closer to the Greeks (in Tannens study) (p. 27) told more detailed and elaborate stories in response to the Pear Film than their American counterparts. In fact, one possible reason for the increase in the length of student responses from the first to the last commercial is that 46

as the experiment progressed the commercials became increasingly more transformational , containing more and more of a story and so giving the students more to talk about. The results also parallel Wanatabes (1990) observation regarding framing differences between Japanese and American students. Similar to the Chinese students in this study, Wanatabes Japanese subjects seemed to have the expectation that they should present details as fully as possible, in chronological order, while the American participants framed their presentation as briefing or reporting (p. 192). This contrast suggests that the students and teachers had different sets of expectations about how a public service commercial should present information, and what the content and tone of that information should be. Asian students seem to expect more narrative from the commercials, treating the plot and characters as more salient than the message or technique. as public we saw the..we saw the ad..we..we will pay attention to the people in the ad..um..so..uh..if the people in the ads can do something or even they say something it is better. (Student 7) Non-Asian teachers, on the other hand, seemed to expect information from the commercials. They appeared significantly more concerned with the message the commercial had to offer, and the techniques the makers of the commercial used to get the message across, as the following illustrative remarks testify: (14) so in cinematic terms.. (15) this is quite sophisticated (16) but in terms of message 47

(17) its stone-aged Im afraid.. (T4/Family) In Wellss (1987) terms, the teachers wanted lectures; the students wanted drama. Further indication that the students fit the commercials into a narrative framework while the teachers brought to their viewing and responses a more expository framework can be seen in the varying functions the two groups ascribed to particular elements in the commercials. Responses to MTR provide a good example: in the commercial, a young couple walks into a Mass Transit Railway station and notices a government billboard about AIDS. The camera flashes two closeups of the billboard listing facts about AIDS in English and Chinese and then moves to a close-up of the couple clutching hands. Half of the Western viewers commented that the information on the billboard had been presented too fast: (11) and there were short flashes of this strange black pyramid.. (12) and there was facts about AIDS (13) but they came on the screen so very quickly.. (14) okay both in English and Chinese.. (15) but I dont think anybody could read them (16) in the time that was given.. (T4/MTR) (7) they showed a few facts about AIDS.. (8) um- written on a billboard or something.. (9) but I really didnt have time to read it.. (T6/MTR) 48

None of the Chinese students, on the other hand, were bothered by the speed in which the information on the billboard was presented. Instead, they tended to mention the billboard as an element in the narrative. (6) and- at the end of the advertisement well..um.. (7) the couple hold well hold their hands quite tight (breath).. (8) maybe they are quite nervous.. (9) they are aware of that (10) and they seen well they saw the billboard in the MTR.. (11) um they are alert.. (12) theyre afraid they might get AIDS. (S3/MTR) Seeing the commercial as a story, the students assumed that the information on the billboard was intended to be read by the characters, not the viewer. The important message for viewers was in the characters reaction to the information. In contrast, the teachers, bringing a lecture framework to the commercials, expected that when information was presented in point form on the screen, it was meant to be read.

Viewer Frame Like the Americans in the Tannen (1979) and Erbaugh (1990) studies, Western teachers in this experiment were much more likely to talk about themselves as viewers of the commercial , spending 13.8% of their idea units in the viewer frame, as opposed to only 4.6% by the Chinese students. The teachers used significantly more pronouns to refer to themselves, saying the words, I, me, and my, almost twice as often as the students, and other times using pronouns like you and we to signal the viewer frame 49

(Appendix 5a). Sometimes teachers regarded the viewing as an intensely personal experience. What happened in the commercial? what happened..I felt angry basically.. is basically what happened (laugh) (T2/Family) Although comments in this frame were often connected to people or objects in the narrative, the way the statements were structured reminds the listener that these people and objects are being filtered through the perception of the viewer. (3) I didnt get the impression that it was necessarily a bar (T3/Youth and Prostitution) (1) were seeing couples.. (2) different venues.. (3) holding hands and uh walking along.. (4) thats what were seeing.. (T10/MTR) Sometimes teachers spoke in this frame as if they were actually experiencing the commercial from the inside: (10) and then we shift from the airplane...suddenly.. (11) zoom into the hospital... (12) and you see somebody lying in a bed (T10/AIDS & Travellers) There were also places in the teachers recounting where the commercial frame seemed to invade the narrative, the voice-over giving its message to the characters themselves: 50

you mean the end of the scene where the men went out^..um...they followed the advice of the voice-over (laugh)..avoided casual relationships..so they were listening to the voice-over.. (T7/Bar) (7) and the uh..voice-over sort of admonishes the couple.. (8) it tells them that they must use condoms. (T5/MTR) Often when teachers did perceive a story, they chose to tell it within the viewer frame as in the following example in which the subject situates the narrative primarily within the viewer and commercial frames and describes it chiefly in visual terms: (1) um..this one visually is quite subtle.. (2) it shows um a happy healthy looking young guy (3) setting off on holiday.. (4) um and you see him sitting in the plane.. (5) and then its cut to a scene of him in hospital (6) hooked up to a drip.. (7) we dont see the face (8) but you presume its the same guy.. (T7/AIDS & Travellers) [emphasis mine] Teachers comments in the viewer frame are not just a matter of perspective. The viewers experience in watching the commercial becomes the real narrative in the teachers responses, supplanting the narrative of the characters. It is as if the teachers interpreted the question What happened in the commercial? as What happened to you when you were watching the commercial? 51

Tannen (1979) observes a similar tendency among her American subjects: the interplay between (the viewers) expectations and the events of the film are part of her narrative content. Her experience as a viewer is part of her story which therefore becomes a story not only of the movie but of her viewing of it as well. (p. 154)

Experiment Frame Both students and teachers spent relatively little time in the experiment frame, and when they did their statements were mostly in the form of what Goffman (1986) calls brackets, defined as boundary markers to mark off a social activity from other ongoing events (p. 251). Utterances in this frame also included brief comments on the subjects own performance or requests for clarification regarding the procedure of the experiment, but these were few and far between. Interestingly, none of the students requested feedback or confirmation regarding their responses. The only instances of such requests were found in the responses of the teachers. The relative lack of focus on the experiment frame not only suggests that the subjects felt relatively unselfconscious in the role of subject-in-alinguistic-experiment (a role not unfamiliar to most of them), but also may show shared, tacit understanding of the etiquette of such experiments which deems it inappropriate for subjects to make explicit interpretations or judgments about the procedure. There is, however, some indication that teachers saw the commercials chosen for the experiment in a more integrated and 52

analytical way in the fact that they tended far more than students to compare and contrast them, frequently remarking about commercials other than the one they were responding to and interpreting them as a connected text. Teachers made nearly twice as many statements of comparison than students. For example: (12) so it seems that those three in sequence.. (13) you cant have sex with your regular girlfriend.. (14) you cant have sex with someone you pick up in a bar.. (15) and if you go to a brothel.. (16) thats not safe either.. (17) (laugh) so you know theres not very much left it seems... (18) I think the message is dont have sex full stop (19) and all the messages are directed very much I think to men.. (20) and theres the idea that men are at risk from women (21) rather than the other way around.. (T4/Youth and Prostitution) The teachers also occasionally referred to the context of the experiment in order to assist them in forming interpretations. The students, on the other hand, seemed to regard the retelling of the commercials as a more natural task, hardly ever calling attention to it. This is perhaps not surprising considering that students are accustomed to being asked to perform tasks like this in which they discuss or display their comprehension of texts, and are unaccustomed to questioning or commenting upon the reason or purpose behind such tasks. The different expectations the teachers and students had about what was expected of them in the context of the experiment can also be seen in the comments they made in the viewer frame, particularly those comments 53

containing negation and thus indicating failure to fulfil their expectations regarding what constituted appropriate performance (Appendix 5d). Tannen (1979) claims that the use of a negative statement is one of the clearest and most frequent indications that an expectation is not being met (p. 47). The statement, I dont remember, for example, indicates the subjects expectation that s/he should have remembered details in the commercial like actions and characters. For the teachers, negative statements in the viewer frame were most often associated with words like understand, figure out, think, know and see. The students made more negative statements regarding their ability to remember or be sure about what they witnessed. Thus, it seems that the teachers interpreted their task as understanding the commercials, whereas the students were more concerned with remembering and being sure about details in the commercials.

Patterns of Response The patterns in which students and teachers moved from frame to frame was highly individualistic. There were, however, certain sequences that seemed particularly popular (Table 1, Appendix 4e). Table 1 Preferred Patterns of Response (in order of frequency) Teachers CM--PHY/NAR--CM VW--PHY/NAR/CM--VW CM 54 Students PHY--NAR--CM CM--PHY/NAR--CM NAR

PHY--NAR--CM

NAR--CM

(see Appendix 10) More than 10% of the students responses were completely in the narrative frame. When they talked about the ad as a commercial, they tended to leave these statements for the end of their responses, though sometimes they began in the commercial frame by way of introducing the story. The teachers, on the other hand, were much more likely to bracket their recollection of the narrative (if they included it at all) with statements from either the commercial frame or the viewer frame. In other words, the order of students remarks fit more closely with the way the information was presented in the commercials themselves: setting--story--message, a more inductive (Scollon and Scollon, 1995, Young, 1982) and text-based (Pritchard, 1990) pattern, similar to Erbaughs (1990, p. 26) Mandarin speakers who followed chronological order...more strictly than the Americans. In contrast, the teachers seemed to favor a more deductive, meta-linguistic pattern: message--setting/story--message. One might say that the students used a more Feel-Do-Learn strategy when recalling the commercials, while the teachers used more of a Learn-Do-Feel strategy (Miracle, 1987). Student responses were more likely to follow a narrative schema, including such elements as setting, beginning, complication, outcome, and ending (Dechert, 1983; Freedle and Hale, 1979). The way teachers patterned their responses tended to conform more to an expository schema, beginning with a proposition, offering reasons and evidence to support it, and then restating the proposition in a conclusion. The teachers also made greater use of the subjunctive mood (Appendix 5b), which, according to Freedle and Hale (1979) is more characteristic of expository prose.

55

Interpretation Students and teachers spent approximately the same amount of time in their responses explicitly interpreting what they saw, with students interpreting slightly more than teachers (S=35.6%, T=30.2%). The main difference between the groups was the frames within which these statements of interpretation occurred. Students made most of their interpretative statements in the narrative frame, more than twice as many as the teachers, assigning emotions or motivations to characters, commenting upon the significance of their actions or making conjectures about their past or future behaviour. Teachers reserved most of their interpretative statements for the commercial frame, where they concentrated primarily on finding a message in what they saw and describing what they believed the authors of the commercials were trying to do and how they were trying to do it. Even within the same frames, teachers and students seemed to focus their interpretations on different elements. In the narrative frame, teachers tended to restrict their interpretative statements to the actions or identities of the characters, while students often commented upon the characters internal states, how they felt:1 (6) they are afraid (7) and thats why they go together.. (S5/Youth and Prostitution) (15) and suddenly (laugh)..the girl um..working the hair salon (16) uh feel so embarrassing.. (17) because they are Chinese (18) and so he she think that its quite an embarrassing..(breath) topic 56

(19) to talk about AIDS eh--..in the public place... (S1/Salon) In subjects responses to Family, for example, which shows a man returning home to his wife after a sexual encounter with another woman, half the students mentioned that they thought the wife was suspicious or worried, while none of the teachers chose to comment on the wifes internal state. Similarly, in Husband, which portrays a man learning that his old girlfriend has got AIDS, half of the students noted that the man was frightened or shocked, whereas none of the teachers did. (15) theres a dead silence.. (16) maybe the man..well..was really afraid (S3/Husband) Students were also more likely to attribute personality traits or habitual actions to characters based on the story: (4) and then perhaps the girl was very sociable.. (5) sociable..sociable girl (6) and then perhaps she would like to make friends.. (7) a lot of male friends (S4/Girls) (8) perhaps he always like to go to some of the bar.. (9) and then to ask some of the hookers to be with..with him.. (S4/Family) They were also more likely to extrapolate, telling what they thought had happened or would happen three times more than the teachers. (8) and I think she probably had sexual relationship with that boy.. (S8/Girls) 57

(10) and maybe later they..may have sex with each other.. (S5/Homosexuals) In most of the commercials shown to the subjects, the outcomes or endings are left out, leaving the viewer to guess what is going to happen to the characters. The students were much more willing to do this, using future tense forms nearly twice as much as the teachers did (Appendix 5b), transforming the incomplete narratives in the commercials into full stories in their retellings. Related to this use of future tense forms is an air of inevitability associated with the subjects expectations about HIV infection (the belief that HIV infection is an inevitable consequence of casual sex, that AIDS is an certain consequence of HIV infection, and that death is an inevitable consequence of AIDS). This orientation can be seen in the following remarks by the students: (12) all the things are having sex with other women.. (13) and thats why he died. (S5/Family) (4) and it implies that uh.. not only will the man get AIDS... (5) but also his family..his wife,,,will also get AIDS (6) and the last part theyre uh..I think..uh..er..they want to show that (7) the ending of such kind of relationship is death. (S7/Family) (12) because he..and he decided not to go to that mans [hou] home... (13) because he knows that..uh..it will lead to um..(laugh) horrible diseases. (S1/Homosexuals) 58

This element of inevitability is, in fact, characteristic of many of the Hong Kong Governments AIDS education materials. In Sharing Needle, a commercial not used for the study, for example, the outcome of a scene involving intravenous drug users sharing a syringe is presented as certain and unavoidable, ignoring the presence of variables such as dose, frequency of exposure and individual susceptibility to infection (National Academy of Sciences, 1988): 1) This group is about to share a needle to inject drugs. 2) One of them is an AIDS carrier but he doesnt know it. Neither do the others. 3) Now they will all be infected with AIDS. (Sharing Needle) The prevalence of future extrapolations in the students responses my also be a reflection of students tendency to focus more on the endings of the stories. Mandler (1978), in his comparison of narrative recall by native language and second language speakers found that the native English speaking subjects tended to recall endings less well than ESL subjects. Students also made more statements explicitly pointing out causal relationships in the stories, more than eight times as many as teachers made, and used the word because more that twice as much as teachers. That is not to say that causal relationships were not implicit in the teachers retellings, but that the students chose to make them more explicit in order to link up elements in the narrative, (8) and why why why why why did he die^... (9) well just because um.... (10) he well he has..he had sex with somebody... 59

(11) and so he was infected with AIDS (12) and therefore..so um..he died.. (13) and then his family (breath) w-as very sad..yeah . (S3/Family)

Part of the reason for prevalence of future tense forms and certain connectives may be transfer from the students native language. Young (1982), for example, points out that Chinese speakers of English use connectives like because and so not just to show casual relationships but also to initiate discussion and mark transitions to summary statements. Tannen (1984) attributes the tendency to interpret in the storytelling frame to the desire to interest the reader (or listener) in a good story (p. 33), and she found this tendency to be relatively stronger in Greek than American narratives: another striking difference between the Greek and American narratives (was) the tendency for the Greeks to interpret and make judgements about the events and the people portrayed, While a number of Americans develop their narratives into extensions of the theme that the film had a strange soundtrack, a number of Greeks develop their narratives into extensions of some theme about the significance of the events in it. (Tannen 1979, p. 155) Interpreting the commercial as a technical product by focusing on the execution of the message, as the teachers in this study did, may also be a way of involving the listener, but one, as Tannen points out, much more closely associated with literate or school related discourses. While the students seemed to approach the task of responding to the commercials

60

from the framework of oral narrative, the teachers used a more schoolbased, analytical framework. Often interpretation is not explicit, but rather manifests itself in more subtle exercises of lexical choice. Implicit interpretation typically occurs in the reporting mode when a speaker takes up an ostensibly objective stance while expressing interpretation covertly though the way s/he names objects and actions. Again, students were more likely to engage in interpretative naming, especially in the narrative frame. More students than teachers, for example, referred to the main character in Girls as a student, and to the figures in the photographs in AIDS & Travellers as prostitutes. Erbaugh (1990) found that her Chinese subjects used more explicit terms to describe characters (e.g. fruit picker rather than man), prefer(ing) nouns which define social status and relations (p.31). Similarly, the Chinese students in this study refer to characters in terms of their social roles and relationships (husband, wife, friend) more than the Western teachers (Appendix 5e). In terms of the way students and teachers interpreted the message in the commercials, particularly important to note is that the two groups sometimes had very different ideas about what the ads were trying to say. In responses to Bar, for example, five teachers mentioned the advocation of condom use as the ads primary message, as opposed to only one student. Meanwhile, five of the students reported that the main message conveyed was that condom use is unsafe, because it is not 100% effective.

61

Evaluation Teachers were much more evaluative in their responses to the ten commercials, making explicit judgments in 16% of their idea units, compared to only 2.3% by the students. Furthermore, most of the teachers judgements were negative. They often began their comments with a judgement: (1) Ah..I thought that one was much more to the point..uh..uh (T6/Bar) (1) oh dear..even worse (T4/Youth and Prostitution) And sometimes they framed their entire response as an evaluation: (1) (breath)...I think thats an appalling public announcement... (2) I can see it being a drama.. (3) I could see it being..um..a television play or something like that.. (4) (breath)...where its function would be quite different.. (5) it might be to um..for sort of social comment or(6) or um..a play about modern society... (7) but that one seems designed.. (8) seems targeted at making people feel scared after the fact... (9) and I think thats thats appalling.. (T2/Husband) Occasionally judgments arose out of their own inability to recall or understand what had happened. Rather than owning their confusion by simply reporting it, they attributed it to the commercials; it was the message that was confused, not them, confirming Scollons (1993) observation that when communication fails Westerners often put the 62

blame on the sender for failing to package his thoughts or feelings properly (p. 42): (14) so I think the message is somehow confused (T4/Bar) and he sort of thought about it and then thought twice..and left on his own^ did he^ leave on his own^ (laugh)..I cant remember..forgettable..I think that one..forgettable (laugh). (T8/Bar) Teachers judgments concentrated on ways in which the commercials and their makers had failed to conform to their framework of what AIDS education should be and do, a framework arising from the discourse system of which they are members. The Scollons (1995) point out that the Utilitarian Discourse System prefers certain forms of discourse modelled around the C-B-S (clarity, brevity and sincerity) style. Members of the discourse system are educated in valuing communication which resists the use of rhetoric, and presents information plainly without appearing to be making an attempt to influence the listener except though his or her exercise of rational judgement (Scollon and Scollon, 1995, p.108). Further, ideal communication should support the egalitarian and individualistic ideology of the discourse system by displaying a stance of symmetrical solidarity and not appearing intrusive or infringing on the rights of individuals to make their own choices. Not only had the teachers in this study been thoroughly educated in these tenets of communication, they were also actively involved in teaching them to members of other discourse systems. It is not, then, surprising that their evaluations of the ads clearly reflected these principles. 63

The teachers saw the primary function of the ads as providing information, and often complained when they felt they had not been given enough: (1) Well this one is high on moralizing and low on information.. (T4/Bar) They also demanded that this information be presented relatively free of ambiguity, with all important terms clearly defined. (17) theres no definition of casual sex..whats casual.. (18) well you know who with, when..its not indicated.. (19) so basically its you know dont pick people up in bars..I suppose.. (T4/Bar) (7) it seems theyre not making clear.. (8) what they mean by sexual encounter (T2/Family) And they felt the ads should offer practical solutions: (21) but again where are the condoms..where..where? (T4/Salon) Particularly prominent was the idea that AIDS education should be nonthreatening, that it should not be designed to elicit fear or anxiety from the viewer. Instead, they believed the ads should take a positive approach. (35) so I think that one is really..chilling.. (36) its not supposed to be chilling (T2/Husband) (11) this is just scare-mongering..I think... (12) not helpful (T4/Girls) 64

Some of the teachers also criticized what they perceived to be an authoritarian stance in the ads, which violated the principles of egalitarianism and symmetrical solidarity they perhaps expected from the messages. (10) they give a very strong message in that authoritative voice.. (11) thats booming down like its coming down from God.. (T7/MTR) Many also demonstrated a resistance to what they perceived to be to be the promotion of stereotypes in the ads, indicating adherence to the utilitarian principles which in interaction promote the uniqueness of each individual and in communication value specificity and originality: I think this one does it more effectively..well it seems and more..well perhaps less stereotypical..character. (Teacher 1) (36) its it does for me it doesnt push too many eh-.. (37) push too many stereotypes.. (37) some of the other ones... (38) I felt like they were trying to show like a stereotypical person.. (T6/Salon) Finally, teachers objected to ads that seemed to be intruding on the individuals right to make his or her own moral choices. In fact, morality and education are seen by them as two separate things: (10) theyre sort of focusing on a campaign of.. (11) dont be too naughty... (T1/Homosexuals) (25) and there are some quite useful artistic intentions involved 65

(26) but that the educational intentions are subverted (27) by the need to um..to make a bow.. (28) to conventionally accepted morality in Hong Kong.. (T7/AIDS & Travellers) When asked which of the ads was most effective (Appendix 6b), the majority of teachers chose either Safer Sex or Salon. What they liked about these commercials were those elements in them which promoted the utilitarian sanctity of information and the individual. Safer Sex portrays a man, practically alone (as the figure in the bed behind him appears to be asleep), exercising his individual freedom to make choices, and displaying a practical measure to avoid HIV infection--a condom. I thought that one was by far the most effective..and most positive..well thats I mean because its the most positive..there they are..theyre getting on with their lives as people do..theyre having they can sex..they may be screwing around..but theres still something do about it right there.. (Teacher 2) Salon, while it does not actually give much information, does present information as the solution to the problem of HIV/AIDS in its slogan, AIDS, The More You Know, The Less the Risk. which one do I like the best^..ah..well then number ten..that ones the best one for me..because as I said before..its just sort of encourages you to find out things..and it doesnt try to threaten you. (Teacher 5) The students, approaching the ads from a different discourse system, were far less likely to make explicit judgements about the commercials, and when they did, they were more often favorable. Furthermore, their 66

comments suggest that the ways they formulated their judgements were quite different from the utilitarian criteria applied by the teachers. Rather than information presented in a clear and systematic way, the students seemed to be looking for good stories that affected them emotionally rather than cognitively. They often used words like effect and impact when evaluating the commercials: it brought me great impact..and we we can know the..effect of making love..with some strange women..and uh..yeah..it brought me impact..yeah so I like this one. (Student 9) Whereas the teachers criticized commercials for not saying enough, students criticized commercials in which not enough happened. I wont know what would happen or what they are talking about..um until the very last moment..the man put the condom on idea is the table..I I I will guess uh whats going to happen..what what I wont know..the not very clear. (Student 5) the one I like most..family..or uh the also two..no the eight and the ninth one..that means the girl and the husband that kind of thing..that means I will know the effect immediately..but..from the advertisement of safer sex I just..uh..huh..he will use a condom..and um..and uh..he will make love..uh..with a prostitute..so..whats the effect..Im not sure. (Student 9) The students also occasionally criticized what they saw as inconsistencies or unrealistic elements in the stories or messages, usually drawing upon 67 what they are doing what is the man trying to do..

their background as members of the society for which the commercials were made. if they are close friends I think, uh..as Chinese..they seldom ask such uh..and embarrassing question..yeah..even if Sally had (laugh) had sexual behavior with Raymond I think..she would say no..no..of course not (laugh) to her friend. (S1/Girls) (14) um..I think this statement may not may not be true.. (15) eh because in a in our Chinese society.. (16) uh..if you have a..fixed sexual partner.. (17) and and you know that..they..he..the sexual history of your partner is (18) ..um..(breath)..um..is uh..what should I say.. (19) is clear and without such AIDS background then... (20) then you can safely assume that you will have you will have nothing..um (S10/Youth and Prostitution) The commercial deemed most effective by the majority of the students was Family, in which scenes of a man establishing intimacy with a woman in a night-club are intercut with scenes of the mans wife and daughter attending his funeral. They preferred this ad not just because it conformed more closely to the transformational framework they brought to the viewing, but also because it more closely exemplified the ideology of their discourse system in which an individuals actions are seen not just in terms of how they affect the actor, but also in terms of the impact they have on the group of which the actor is a member, in particular the family. The importance of the family unit in Chinese society has been pointed out 68

by many researchers (see for example Hsu 1985; Scollon and Scollon, 1995). Just as the individual making choices seems to operate as a kind of icon for the teachers, the family unit functions as an icon for the students: because in Hong Kong..not not in Hong Kong..perhaps I will say in Chinese culture..perhaps it is that we are quite emphasize the importance of family..and then you must have a husband..a wife..and then you have a lot of children..if you have got such kind of things..and then you and then all the people will say that you have a happy family..and then..but if there is some kinds of troubles in this family for example if the husband have..eh..som-- have contracted from some kind of disease..and which is incurable..and then it will become a great threat to the family..because if you do something wrong it will destroy your family..your wife..your children or grandmother grandfather etcetera etcetera..so..become a great trouble. (Student 4) Which commercial did you find the most effective? I think the fifth one..the family..because uh-...I think..everyone.. um..regardless of sex their religions..um..have the have the love relationship..with their family members..um.. I think all the people love their family members..um..it is very sad when I um..saw about the commercial.. that the husband of the family..uh..was died..uh mh..was died because of the AIDS.. uh hm..um especially when..he is a man..the husband who..could possibly earn the living for his family..now the family will be in a in a poor maybe in troubles because of the man who earn the money has died..and it is very poor 69

for the young

woman to continue..uh..of her fam- her living also for the daughters or the children. (Student 10)

Teachers objections to moralizing in the ads may come from the utilitarian separation of the concepts of education and socialization (Scollon and Scollon, 1995), according to which education is meant to inform, whereas moral decisions are seen as private matters. In Chinese culture, however, there is no such separation. Moral education is an intricate part of public school in all Chinese societies (Wilson, 1980), and such training often takes the form of stories...consciously constructed to elucidate moral rules (p. 123). The idea that education could be subverted by the expression of moral rules and standards would doubtless be incomprehensible to many Chinese. Thus, rather than objecting to the commercials for being threatening, students seemed to accept this tone as an inevitable part of the genre, evidenced in this exchange from Jones (1995) in which students wonder if the ads are threatening enough. S6: S7: S6: Imagine we WERE gay...We would remind this commercial Right...friends^ BOYfriends. when we met new friends.

S10: But do you think its threatening enough? I doubt it..even though it uses the technique that is threatening, I dont think thats so strong. S7: Yes. S10: Remember some advertising from...throw the litter in our harbor.. remember one that is VERY threatening..step on the nail^...Do you remember? 70

S6:

Yes, I remember.

S10: I think thats very threatening...but maybe Im not the target. I dont feel it. Do you think thats effective? (p.33) The different patterns of explicit evaluation between teacher and student responses also reveal different aspects of the interpersonal face systems in their respective discourse systems, and their perception of power relationships in both the viewing of the commercials and the telling about them. The students, situated in a clearly asymmetrical relationship with both the people who made the commercials (the government) and the person they were talking to about the commercials (the interviewer), chose more of a strategy of indirectness (Scollon and Scollon, 1995) by avoiding explicit judgements and offering up a dominant reading. The teachers, on the other hand, perceived their relationship to the makers of the commercials in a very different way. The utilitarian ideology promotes egalitarian relationships between people and their governments. When teachers perceived the commercials to be violating this principle by taking up an authoritarian and intrusive tone, they responded with scepticism, fashioning more oppositional responses. This should not be interpreted as suggesting that the Chinese students were less judgmental than the teachers. Actually, students displayed a significant amount of evaluation, but rather than presenting it as judgement, they preferred to use the more subtle tool of evaluative language within the reporting mode. Overall, students exhibited slightly more instances of evaluative language, and, in the narrative frame, they used evaluative verbs, adjectives and adverbs eight times more than the teachers did. But again, most of their evaluation was directed not at the commercials 71

or the people who had made them, but at the characters within the stories and their actions: (5) that means that when some people go uh go to the bar..so often.. (6) maybe..er-..they are not so good... (7) especially girls (S9/Bar) (4) and then a girl tried to trap him.. (5) and have some sexual experience with them... (S10/Family) (12) and he has a wife and also a...a little girl (13) an then eh..have a contrast to show that.. (14) in the past he had done something very..eh..bad.. (S4/Family) Erbaugh (1990) observes that in her study The Chinese telling the pear/guava stories relished discussions of theft and thanks, while the Americans shied away from moral judgment (p. 31). This study reveals a similar pattern: evaluations by the Chinese students were focused more consistently on the moral behavior of the characters, while the teachers concentrated more on rhetorical judgments directed towards the makers of the commercials.

Confusion and Distortion Confusion and distortions are particularly strong evidence of framing, demonstrating, among other things, disattention to a particular frame or especially resistant sets of expectations which do not allow conflicting 72

information to be processed. Sometimes distortion or confusion can result when the elements encoded in the message are different from those of the viewers system of coding. Steffenson, Joag-dev and Anderson (1979) and Pritchard (1990), for instance, found more distortions in readers recollections of stories from a different culture, and Carrell (1987) observed that when texts violate expected story schemata both quantity and temporal sequencing of recall can be affected. The teachers produced more than twice as many distortions per idea unit in the narrative frame than the students. These distortions ranged from misrepresenting settings to falsely reporting about characters and their actions. Teachers, for example, mistook men for women (T1/2, T2/2, T3/2, T6/2, T10/2), a brothel for a doctors surgery (T2/3), a disco for a church (T6/5), and a father hugging his daughter for a social worker taking a child away from her parents (T10/5). As might be expected, the students, members of the discourse community in which the ads were produced , seldom reported difficulty in understanding what they had seen in the story. In contrast, teachers often admitted being mystified by the narratives, unable to see the logical connections between people and events. In all, the teachers reported confusion nearly four times more than the students: (1) well, I dont understand that one at all.. (2) I think if I didnt know it were about AIDS uh.. (3) it would be a very strange advertisement.. (T10/Family) (1) I have to admit, (2) this one I didnt get at all.. (3) I have no clue whats going on... 73

(T5/Homosexuals) It might be argued that the confusion and distortions reported by the teachers, as well as the corresponding tendency for students to retell the stories in more detail, resulted from the fact that more of the students had seen the ads before on local television, and so were already more familiar with them (Appendix 6a). While this is certainly a factor, many of the teachers, having resided in Hong Kong for a number of years, were also familiar with the ads. Furthermore, patterns of confusion and distortion, as well as broader patterns of framing and modes of response, were consistent across ads, even for those commercials which more teachers than students reported having seen. In responses to Salon, for instance, which seven teachers reported having seen before as opposed to six students, teachers reported confusion seven times in their responses, while none of the students reported any confusion at all. Sometimes, in fact, having seen the ad before only increased teachers reports of confusion, as they added memories of past confusion to their reports of present confusion. Another possibility is that teachers, in their relatively more symmetrical power relationship with the interviewer, felt more secure about admitting uncertainty when they felt it. However, judging from the accuracy and richness of detail of students responses in the narrative frame, it seems unlikely that they were experiencing confusion but not admitting it. In fact, when true instances of ambiguity arose in the commercials, students seemed more likely to report the possibility of multiple interpretations. In Safer Sex, for example, in which one characters gender is obscured by the fact that s/he is facing away from the camera and covered 74

with a sheet, nearly half of the students pointed out that they could not be sure whether it was a man or a woman, as opposed to only one teacher. Eight of the teachers chose to assign a gender to the character, five describing the character as a woman, and three identifying the character as a man. One teacher failed to see the figure at all, insisting that the main character was going to bed alone, a fact that figured prominently in his negative evaluation of the ad: oh I didnt think that was very effective.. for one thing why would a guy be wearing a condom and going to bed alone..you know (Teacher 6) Nowhere were the differences in patterns of confusion and distortion between teachers and students more apparent than in the two commercials featuring representations of gay interaction, Homosexuals and Bar. Bar focuses on several couples in a bar. The first few shots show the growing intimacy between a man and woman, signalled by such tie-signs (Goffman, 1971) as leg touching and food play. Meanwhile, a man behind the couple rebuffs advances from another woman by neither responding to her tie-signs nor establishing any with her. Suddenly something seems to catch his eye. The camera cuts to a young man in a football jacket sitting at a table alone. The man who failed to respond to the woman walks, instead, over to the young man in the jacket and whispers something in his ear, and the two of them leave together. As they exit, the older man puts his arm around the younger mans shoulder while the voice-over cautions against casual sex. In Jones (1995), I argued that this ad operates by constructing a framework of a straight bar, then violating it with a portrayal of same sex interaction, thus increasing the out-of-frame status of the gay characters. 75

Apparently, in viewing this commercial, many of the subjects were unable to integrate information that did not conform to the cultural models signalled by the setting and the initial interaction. While half of the students interpreted it as a portrayal of same-sex coupling, only two of the teachers did. The students who saw the gay story were, for the most part, able to produce fairly detailed renderings of the characters actions and intentions: (5) but it seems like the first man who appear in the advertisement is a.. ah-- homosexual guy.. (6) so even that a sexy or beautiful girl sitting beside him.. (7) and maybe asks him to..uh..yeah-(8) but the man havent pay any attention to the girl.. (9) but when theres a smart..smart man..(laugh).. (10) he says something to that man... (11) and so they may be (laugh) going to do such thing (breath).. (S1/Bar) Most of the teachers failed to see the homosexual content in the ad, even when prompted: Were there any gay couples? I didnt notice any gay couples..there were people there who might have been gay but they were..ah..sort of unattached..so I didnt notice any particular sort of gay theme about it at all. (T2/Bar) Some, while admitting the possibility of a gay narrative, dismissed it and instead attempted to fit what they had seen into a heterosexual framework: Two men left the bar together..having before that one man..had kind of whispered in his one..his friends ear about something..I 76

presume about wearing a condom..and they had gone off together (breath) um..it could have implied that there was a homosexual relationship between them..but I didnt feel that..I felt that um the emphasis was still on heterosexual sex. (T3/Bar) Most of the teachers were not even able to accurately report what had occurred in the commercial. In a demonstration of gender confusion reminiscent of M. Butterfly, four of the teachers reported that a man and a woman had left together, and two of them said one man had left alone. Subjects responses to Homosexuals also suggests that Chinese students and Western teachers brought to the commercials different sets of expectations regarding representations of gay interaction. In the commercial, two male characters meet at an outdoor cafe. Their growing intimacy is signalled by increasingly animated talk, and an episode in which one character reaches for the bill, the other places his hand on top of the firsts to prevent him from paying, and their eyes meet. Then one character leaves the table, and the other follows. The final shot shows one character beginning to get into a car which the other has already entered when he sees a Hong Kong government AIDS Awareness poster on the side of a public light bus and suddenly and resoundingly slams the car door. Nine of the ten students reported this as a gay narrative, while only four of the teachers did. Most of the teachers admitted that they had trouble making sense of the story at all: (1) I have to admit this one I didnt get at all.. (2) I have no clue whats going on.. (3) as far as I could tell (4) these two guys..are having a drink.. 77

(5) first I thought it was some kind of foreign locale.. (6) because I remember there was an AIDS commercial (7) about picking up AIDS in Thailand or something (breath).. (8) that theyre having a drink.. (9) and then one guy seems to say no to..the bill (laugh)..something like that.. (10) and then he goes off and gets on a Hong Kong public light bus.. (11) and sees an ad.. (12) and..I have no clue as to whats going on in this story.. (T5/Homosexuals) Even teachers who recalled a gay plot sometimes hedged or reported resorting to the experiment frame to aid their interpretation: (11) its not clear.. (12) I mean..you think they are together.. (13) I thought they were together.. (T9/Homosexuals) (9) it looked all very respectable..um.. (10) I..didnt get the feeling that there was.. (11) that they were trying to pick each other up (12) or that there would be any kind of..sexual innuendoes there.. (13) until..well because of this experiment (14) I knew it was about AIDS.. (15) so I ..knew that..it was trying to portray (16) two gay men picking.. (17) you know cruising each other..um.. (T7/Homosexuals) 78

The students perceived the narrative in terms of a seduction and rejection. Elements prevalent in their recollections included (1) the length or manner of the characters conversation as a sign of growing intimacy, (2) the episode involving the bill showing confirmation of intimacy, (3) the eye contact between characters acting as an invitation, and (4) the slamming of the car door to signal a refusal of the invitation. Few of the teachers mentioned any of these elements, and when they did it was often to report their inability to decode them: (22) there was a lot in that commercial I didnt understand.. (23) when they showed the bill and one went to pay for it (24) and the other one..reached over and stopped him.. (25) I didnt understand what that was all about.. (26) and then it seemed like he one of them went away (27) and then got on the minibus..or something.. (28) I couldnt get what it was all about there. (T7/Homosexuals) Students, on the other hand, confidently, and sometimes in great detail, offered up explanations of the codes employed in the narrative: (10) because when the man intended to pay the bill... (11) the other man um paid it for the man.. (12) and use (laugh) his eyes to say something to the other man (13) and..and then...it intended that the other man...um.. (14) he want to go to the character ones home..or..yeah.. (S1/Homosexuals) What happens with the bill? The bill..and then when one of the waiter take the bill for the man..and then when the man wants to pay for the bill and then 79

the the

other one refuses...and and perhaps he asked them that IlI pay

bill for you..you dont have to pay it..because it seems that..you have already..we are already friends..and then we are already have some such kind of relationship later perhaps..I must be a good..manner (laugh)..and then to to give you some preference at first (laugh). (S4/Homosexuals) The failure of teachers to recognize the gay narrative in these two

commercials does not, however, indicate a reluctance to suggest gay interaction. In another commercial, Youth and Prostitution, portraying two men visiting a brothel, three of the teachers saw the possibility of a homosexual plot, while none of the students did: (3) and then we saw two guys walking up a stairwell together.. (4) at first I thought maybe they were homosexuals or something.. (5) they were walking closely (T10/Youth and Prostitution) The contrast between Western teachers and Chinese students interpretations of these commercials may partly be a consequence of the overall tendency of the students to attend more to the narrative frame. It may also be an indication that members of these two groups have different frameworks through which to interpret gay interaction, and so tie-signs encoded by the authors and animators with reference to one cultural model were difficult to interpret for those decoding the commercial with reference to another cultural model. Another factor has to do with respondents overall framework for AIDS and the degree to which they associate it with homosexual behavior. Notably, students spent more time in the social frame when discussing Homosexuals than any other commercial (13.3%), and statements made in 80

that frame when discussing Homosexuals and Bar often posited a link between homosexuality and AIDS: (3) because this one is about..er.. (4) I think the main cause of AIDS.. (5) the homosexuality... (6) that means if a man..er..um..uh.. (7) has a sex with er um another man.. (8) that means some kinds of homo relationship (9) I think uh..they will get AIDS.. (S9/Homosexuals) ..and even warned those girls or female that...um..men....man or (laugh)...men..uh....is not...not...dont dont easily making love with men..easily, because they may have uh...(breath) (laugh)..they may have special hobbies. Special hobbies? Special habits? Habits...and they may...they may...the one who having AIDS. What do you mean by special habits? Oh...just...(laugh)..(pointing to TV) homosexual (laugh). (S1/Bar)

Emotion Often shows of emotion operate differently in different discourse systems, signalling different internal states and intended to elicit different kinds of responses from other people. Scollon and Scollon (1995) note: ...from one cultural group to another there is a great deal of variability about when one smiles or laughs and what it should be 81

taken to mean...it has been widely observed that Asians in general tend to smile or laugh more easily than Westerners when they feel difficulty or embarrassment in the discourse. This is, then, misinterpreted by Westerners as normal pleasure or agreement, and the sources of difficulty are obscured or missed. (143) In light of this, it is not surprising that the Chinese students in this study displayed emotion three times more than the teachers, mostly in the form of laughter. Specifically, students laughed during their response more than three times more than the teachers (Appendix 6c). Frequent bursts of laughter most probably reflected feelings of nervousness or embarrassment associated with either the content of the commercials or the context of the interview in which they were required to address an authority figure from another culture in a language not their own in an artificial and unfamiliar situation. Often corresponding to the introduction of such topics as sexual behaviour and death in their responses, student laughter also indicates a certain degree of discomfort with these topics, or at least with the introduction of these topics within this context. Sometimes, however, laughter seemed to fulfil a rhetorical function as well, coming in many instances at both the beginning and the end of the students responses and thus operating as brackets, indicating the start and end of a stretch of speech. This observation is particularly important in regard to intercultural communication about sensitive topics such as AIDS in which shows of emotion may give rise to misunderstanding such as one party interpreting the others embarrassment as not taking the topic seriously.

82

Conclusion and Implications The differences in framing noted in the responses of Western teachers of English and their Cantonese speaking students to AIDS awareness advertisements point to several sites of potential intercultural miscommunication between health educators and their clients, and between language teachers and their students. In order to account for these differences and identify these potential trouble spots, the data must be viewed from several different perspectives: a linguistic perspective, a social perspective, and a cultural perspective. When seen as a language task, the viewing and retelling of public health commercials presented separate but equal challenges to the students and the teachers. For the teachers, though the messages were in their native language, they may have featured relatively unfamiliar settings, relationships and values. For the students, though they were viewing messages from their own culture, they had to speak about them in a second language. The demands challenges like these may have placed on processing, and the communication strategies the subjects may have used to meet these demands both affected and were affected by the frames subjects brought to the task. On the social level, frames are brought to interaction in response to such factors as participant identities and roles, context, and the perceived purpose of the interaction. On this level, writes Gumperz (1992) frames operate to convey information on what is to transpire, what role relations and attitudes are involved, what verbal strategies are expected, and what the potential outcomes are (p. 307). Participant identities and roles in this experiment differed dramatically for the two groups of subjects, both in 83

terms of the relationship (power/distance) between subjects and experimenter and between subjects and the government which produced the messages. Furthermore, because of constraints and expectations arising from these roles and from the context, subjects in the two groups were likely to perceive the purpose of the interaction and what was expected of them quite differently. Finally, the results highlight particular cultural differences between the two groups regarding expectations about the appropriate form and purpose of public health education. These differences are consequences of the various factors that go into forming the discourse systems of the two groups: ideology, socialization, rules of interaction, and notions about proper forms and uses of discourse (Scollon and Scollon, 1995).

The Linguistic Level One possible reason for the differences in framing and modes of response in the teachers and students comments about the commercials is the difference in language proficiency between the two groups. Although the students were advanced learners and soon to become English teachers themselves, the prevalence and placement of pauses and fillers, and the frequency of false starts and non-standard grammatical and phonological constructions reminds us of the linguistic obstacles faced by the students in this task when compared with the native speaking teachers. Research into learner interlanguage has increasingly focused on the notion of communication strategies (Corder, 1983, Tarone, 1983), compensative behaviors employed by second language speakers in response to obstacles to communication. Corder defines communication strategies 84

as systematic technique(s) employed by the speaker to express his meaning when faced with some difficulty (p. 16). Communication strategies are employed by both language learners and native speakers, and they operate on both the sentential and supersentential levels. Common communication strategies include: Evasion (including message abandonment and change of focus) Paraphrase ( including approximation, circumlocution, description and word coinage) Transfer (including literal translation and language switch) Appeals (for assistance, clarification or feedback) Non-verbal (gesture, mime, facial expression) Checks (giving further information, examples, contextualizing, comparing) Simplification (or grammar or vocabulary) (Riley, 1989) Some researchers, like Dechert (1983), have explored ways in which schemata or frames can affect communication strategies and interlanguage fluency and accuracy. L2 and L1 knowledge sources quite often compete with each other, Dechert notes, not only in the language of beginners, and certainly not only on lexical and syntactic, but also on strategic levels. Often, he goes on, the competition of speech plans is responsible for the various disruptions, disfluencies and errors which occur not only in the speech of second language learners but of any speaker whatsoever (p. 185). On the sentential level, the language in the responses of both the teachers and the students in this study exhibited the use of nearly all of the 85

above strategies, though they were more evident in the language of the students who obviously faced more difficulty on this level. Strategies for dealing with difficulties in communication, however, also occur on the level of framing and may partially account for the prevalence of students idea units in the narrative frame. Freedle and Hale (1979) posit that narrative schemata are developmentally prior to expository schemata, and research as early as Piaget (1955) has shown that young L1 learners do better at recalling and answering questions about stories than they do about non-narrative prose. Thus, one reason for the students preference for the narrative frame and avoidance of the more expository commercial frame may be due to the employment of the strategy of evasion on the supersentential level. Faced with the already potentially difficult task of formulating utterances in the second language, it would be natural for the students to opt for a more familiar pattern of framing and avoid more cognitively demanding schema. The fact that the students language in the narrative frame was particularly detailed and elaborate can also be seen as a function of the subjects relative familiarity with the different schema; operating within a more familiar framework of a story, students were probably able to devote more concentration to linguistic processing. The relatively greater variety in the frames used by teachers could also have been affected by proficiency problems and evasion strategies. Carrell (1987) found that in ESL learners retrieval processes when confronted with unfamiliar schema showed less flexibility when compared to those of native speakers. The risk involved in producing language within the possibly more demanding and certainly less familiar commercial frame, and the students willingness or lack of willingness to take this risk might also be a 86

consequence of previous educational experience. Pritchard (1990) found that American speakers of English showed more flexibility and risk-taking behavior than Palauans in interpreting funeral stories, even when the text was in their own language, though their responses when interpreting a story from their own culture were more detailed and elaborate. He suggests that these differences might be related to the rote memorization and oral recitation which characterize Palauan schools (Pritchard, 1990, p. 289). A comparable observation might be made about the education system in Hong Kong, particularly in the area of language education in which strong and early emphasis is placed on memorization and preparation for public examinations (Lin, 1996). It is difficult, however, to measure the extent to which subjects employed evasion strategies, since such analysis involves speculation about what the subjects might have said but didnt. Although evasion strategies might explain some of the difference in framing between the teachers and the students, there is also evidence to suggest that it was not the only, nor even the main, factor at work. First of all, when the students did communicate in the commercial frame they did not exhibit any more errors or disfluencies on the sentential level than they did in the narrative frame. Furthermore, their language in the narrative frame often involved acts of interpretation more typically associated with expository frames. Students stories were not merely text-based descriptions of what they had seen, but rather more sophisticated musings on the meaning, significance or motivation behind characters actions and relationships. Finally, the notion that the students chose to speak in the narrative frame rather than simply having been forced into that frame by their relative lack of proficiency is born out in the comments they make regarding the effectiveness of the 87

commercials. Their fondness for the narrative frame is evident not just in their framing but also in what they say in these frames. Obstacles to the task of viewing and responding to the AIDS awareness APIs in this study occurred not just for the students. The teachers also faced different but comparable difficulties, causing them to produce different but comparable errors in their retellings. Whereas the students errors occurred on the level of language, the teachers errors occurred on the level of content (lapses in memory and distortions). Communication strategies employed by teachers to deal with the difficulty of decoding more culturally unfamiliar texts also had a potential effect on the ways they framed their responses. Pritchard (1990), in his examination of the different strategies used by proficient readers in processing culturally unfamiliar material, describes four types of comprehension strategies: 1. Developing Awareness (including referring to the experimental task, recognizing loss of concentration, and stating a failure to understand) 2. Accepting Ambiguity (including formulating questions, considering alternative inferences, and suspending judgment) 3. Establishing Intrasentential Ties (including gathering information, using contextual cues, and reacting to style and text surface structure) 4. Establishing Intersentential Ties (including relating information to previous portions of the text, and extrapolating from information presented in the text) 5. Using Background Knowledge (including using background knowledge of the discourse format, responding affectively to text 88

content and speculating beyond the information presented in the text) (Pritchard, 1990, p. 280) He found that subjects, when faced with a culturally unfamiliar text were much more likely to employ strategies from the first three groups, like stating an inability to understand, considering alternative inferences, suspending judgments, and seeking to establish ties above the sentence level. When responding to culturally familiar material, subjects favored strategies from the last two groups, making use of background knowledge and focusing on more text-based elements on the sentential level. Subjects in this study displayed similar difference in strategy use. The teachers, confronted with commercials produced by a culture different from their own, were often compelled to express their difficulty in understanding them, which helps to explain the frequency of teacher responses in the viewer frame. Moreover, they were more likely to perceive and express ambiguity and to entertain alternative inferences, strategies more naturally associated with the commercial frame. Finally, teachers seemed to employ more top-down comprehension strategies, preferring to focus on more global aspects of the texts rather than on small details. Students, on the other hand, were able to draw upon their background knowledge for understanding the stories in the commercials, and more often employed more bottom-up strategies, concentrating on more specific aspects of the texts and attempting to establish links between the details as they were presented. These differences in strategy use can also help to account for students relative involvement with and favorable attitude towards the commercials. Pritchard (1990) observes that subjects using strategies from 89

the last two groups seemed more actively engaged in the text and appeared to be understanding it more fully (p. 281).

The Social Level Language behavior is largely conditioned by the social roles we play in interaction. Goffman (1974) has observed that the assumptions we associate with different events or situations act as a filter through which we sift our knowledge in order to retrieve that which is most useful for the situation at hand. Our social roles condition our communication by determining what kinds of talk is appropriate and imposing constraints not just on the content of what is said but also on such interactional aspects like turn-taking and questioning. The framework we bring to communication contexts is affected by a variety of factors, including beliefs and values we associate with the topic and the occasion for interaction, what we perceive to be the social import of the transaction, and the expectations we have about participants roles and power. Ethnographers must be continually aware of how such factors may affect subjects behavior in studies like the one reported here. In any human experiment, observes Pinxten (1991), Not only is the observer continually observed as well, but the very act of observation is constrained, guided and even structured by a most often implicit system of mutual control between the informant and the ethnographer (p. 131). Certainly one of the major factors affecting the response of subjects in this study was the topic they were asked to talk about. In every culture, talk about of HIV and AIDS is influenced by deep seated social and psychological associations with sex, disease, and death as well as social and 90

individual experience with and knowledge about these topics (Jones, 1995). Such associations, experience and knowledge result in culturally determined notions regarding individuals perceptions of risk, possible measures for prevention, and the appropriateness of discussing the topic in particular social situations with particular interlocutors. The teachers in this study were likely the have had more exposure to information about HIV and AIDS because they were older, for the most part widely traveled, and came from places where HIV infection and public discourse about it are more prevalent than in Hong Kong. Furthermore, the types of AIDS discourse the teachers would have been exposed to are considerably more varied, from the health education provided by their own governments and a burgeoning number of AIDS service organizations, to more oppositional discourses from sources as divergent as Jesse Helms and ACT-UP. Therefore, when faced with the task of responding to AIDS awareness discourse, they were armed not just with a particularly sophisticated vocabulary for discussing the topic, but also with notions about the politically correct way of speaking about it. The students, because of their age and environment, have had considerably less exposure to information about AIDS, as well as less information about and experience with domains often associated with it: sexual behavior and IV drug use. Most of the information they have had about the topic has come from the media (Hong Kong AIDS Foundation, 1993), a combination of government produced posters, pamphlets and television commercials and lurid tales in comic books and a press which often sensationalizes the issue. Moreover, the governments attempts at education have often proved to be poor sources of information, frequently focusing more on constructing the identity of potential AIDS carriers 91

than giving facts about modes of transmission and methods of prevention, and emphasizing the relationship between AIDS and such socially proscribed behavior as prostitution, adultery, drug addiction and homosexuality (Jones, 1995). Questionnaire studies measuring knowledge about HIV and AIDS done with both adults (Hong Kong AIDS Foundation, 1993) and secondary school students (Benson, 1993) reveal both a high level of ignorance regarding routes of transmission and means of prevention and a low level of tolerance for people living with HIV or AIDS. These differences in knowledge and attitude about the topic undoubtedly had some effect on the way the subjects framed their responses to the commercials. The teachers, drawing on more extensive background knowledge and exposure to more critical readings of the topic, were more able to compare the commercials with similar discourse from their own or other cultures, more apt to approach them skeptically, and more equipped to offer evaluations or suggestions as to how they might have been done differently. The commercial frame is the natural setting for such comparisons, evaluations and suggestions. Teachers were also apt to have more experience in critical meta-discourse in the fields of media studies and film criticism which would provide them with the vocabulary to operate in this frame. The students, more accustomed to seeing AIDS in terms of the identity and behavior (guilt) of the infected (or potentially infected), and less prepared to make comparisons or judgments about the information being presented, were likely more comfortable focusing on the aspects of the commercials about which they were more able to respond: characters and their actions and relationships. Participant roles and expectations about the purpose and intended outcomes of the interaction also play a potential role in the way speakers 92

fashion their responses. In the context of this study, subjects played several simultaneous and overlapping roles based on their different relationships with the government that produced the ads and the teacher who conducted the experiment, each role engendering its own set of face relationships and expectations about appropriate ways of speaking. Wildner-Basset (1989) calls these contextually determined collections of roles, face systems and expectations discourse worlds, and discusses how in language classrooms the possible disparity between co-existing situational frames can effect the ways participants manage their talk and the kinds of constraints they place on the interaction. In the context of this experiment, subjects also operated in coexisting and overlapping discourse worlds with different and sometimes conflicting socio-cultural conventions for managing communication. The students simultaneously played the roles of subject, student, and citizen of the society which produced the messages (Table 2). The teachers found themselves in the co-existing roles of subject, colleague, and guest in the society which produced the ads (Table 3). Table 2 Students World I Participants: Addresser is the government of the addressee; Addressee is a citizen of that government Context: Social/ Political Purpose: Unknown Purpose:Health Education, Inculcation of

World II Participants: Addresser is a subject in an experiment; Addressee is an experimenter Context: Research

World III Participants: Addresser is a student; Addressee is a teacher Context:Academic Purpose: Language Education (Possibly Test of Proficiency)

Restrictions: Turn-taking, Questioning, Restrictions: Turn-taking, 93

Social/Cultural Values Restrictions: Uni-directional communication, Hierarchical Face System, Comparatively Less Exposure to Topic Table 3 Teachers World I Participants: Addresser is a foreign government to the addressee; Addressee is an expatriate Context: Social/ Political Purpose: Health Education Restrictions: Uni-directional Communication, Symmetrical Face System

Hierarchical Face Questioning, System, Comparatively Hierarchical Face Less Exposure to Topic System, Comparatively Less Exposure to Topic

World II Participants: Addresser is a subject in an experiment; Addressee is an experimenter Context: Research Purpose: Unknown (But because subject is also a researcher in same field, speculation is possible) Restrictions: Turn-taking, Questioning, Symmetrical Face System

World III Participants: Addresser is a teacher; Addressee is his/her professional colleague Context: Academic Purpose: Sharing Observations about Linguistic Data Restrictions: Symmetrical Face System

In the first discourse world, participant roles varied chiefly due to the different relationships the two groups had with the government which made the commercials and different attitudes regarding appropriate responses of governments to public health issues. In viewing the ads, the students were situated in an asymmetrical face system in relation to their 94

government and more likely to expect the government to tell them what to do. According to Lau and Kwan (1988), Hong Kongers maintain to some degree the traditional Chinese belief that the government, along with its administrative duties, also has important moral and pedagogical duties. 93% of the respondents in their survey agreed or strongly agreed that the government should set examples and teach the people correct morals and behavior (Lau and Kwan, 1988, p. 90). The hierarchical face relationship and the students expectations about the role of the government in inculcating morality doubtless played a part in making their responses considerably less critical of the makers of the ads than the teachers responses. The teachers, in a comparatively more symmetrical relationship with the authors and with considerably more cynical attitudes towards the place of government in the moral lives of the populace, were more able to approach the ads with a more judgmental stance. In the role of subjects, the two groups also may have had different expectations regarding the purpose of the interaction and appropriate ways of communicating within it. According to Pinxten (1991) informants in research typically attempt to please the researcher, measur(ing) the intentions or expectations of the ethnographer and try(ing) to comply with them (p. 142). Although most of the students had participated in linguistic experiments before, for the most part they had only played the role of subjects. The teachers, on the other hand, researchers themselves, were more wise to the possible purpose of the interaction and expectations of the experimenter. Less able to draw inference about how they were supposed to act in the context of the experiment, it is very likely that students frequently resorted to the more familiar sets of roles and constrains contained in World Three, the world of the language classroom. 95

Being asked to respond verbally to a text was, for the students, not a particularly unfamiliar task, but one they were quite accustomed to in their roles as students. Much in their comments and patterns of framing suggests that it was this identity that influenced them the most in formulating their comments. It is natural that the students, given the assignment of responding to the commercials by a teacher in an academic setting, either consciously or unconsciously interpreted the task as a test of comprehension or proficiency. Furthermore, in their past experiences with language learning students were likely to have been rewarded for use of memory strategies and attention to detail. The Chinese school system, writes Erbaugh (1990), provides a Confucian equivalent to years of memory experiments which reward meticulous recitations and punish outbreaks of personal opinion (p. 30). Chinese students, she says, are taught to give unadorned recounting(s of texts) which begin with time, place and actors, then follow chronologically to a clear conclusion. Whimsical, critical, or editorial comments are most convincing when succinct and saved for the end (Erbaugh, 1990, p. 31). Approaching the experiment as a language test, students were more likely to infer that what was expected from them was accuracy and detail, and might have believed that too much interpretation or judgment would have been inappropriate within this context. The teachers and students may also have interpreted the experimental question differently. For the students, what happened in the commercial? might have seemed like a display question, a request for proof that they had comprehended the text and were able to summarize it. The teachers, less accustomed to people asking them questions to which the asker already knows the answer, were more likely to interpret the 96

question as a real request for information, and so, fully cognizant that the experimenter had himself watched the commercial many times before, searched for things to tell him that he did not already know--specifically, their personal interpretation of and experience with the commercials.

Discourse Systems and Cultural Models of Health Education The way individuals respond to health education messages depends largely on the frameworks of expectations of the discourse systems of which they are members. These frameworks include preconceptions not just about the topic of the message but also about the proper form and function of the message and the appropriate ways to respond to it. Thus, how health educators in different cultures approach their job depends not just on how particular cultures view the issues of health, morality and education, but also on how they see communication. In Western societies is has long been assumed that the proper function of public health discourse is to inform, and that is proper form is the straightforward and objective reporting of facts. Sontag (1991) in her classics, Illness as Metaphor and AIDS and Its Metaphors, argues that the only way to efficiently deal with disease and protect its victims is to cleanse our discourse about it of any trace of metaphor, an extreme statement of the utilitarian anti-rhetorical principle. Furthermore, cultural critics (Alcorn, 1989; Bolton, 1992; Goldstein, 1991) have often criticized governments, the media and individuals for constructing a moral model around AIDS rather than promoting a more medical model. Scholars like these argue, along with the teachers in this study, that appropriate AIDS education materials should present individuals with enough information to make reasoned 97

decisions about such behaviour as sexual activity and drug use without infringing on their individual rights to make such decisions and without appealing to emotions like guilt, shame, and fear. Educational materials employing dramatization, excessive rhetoric, and appeals to the emotions are seen as both intrusive and ineffective. Baddaley (1990), in his examination of audience reaction to different styles of media AIDS education in Britain, Canada, Denmark, Norway, Sweden, and the United States, supports this non-judgmental, anti-rhetorical approach, finding that the campaign materials that were perceived by the viewers as the most effective from an educational standpoint were those which presented hard facts in a simple and straightforward manner. Complex and emotional campaign techniques drew negative responses (p. 116). Such research, however, has thus far been restricted to Western countries where most of those receiving the messages are either adherents to or at least familiar with the basic philosophies of the Utilitarian Discourse System. Outside of the Utilitarian Discourse System, however, this neat moral/medical dichotomy sometimes becomes muddled, because it is based on beliefs about science, education and socialization that may not be shared by other discourse systems. The different frameworks teachers and students brought to the AIDS education materials in this study, and the different ways they evaluated the materials, are, to a large degree, reflections of the different principles of their respective discourse systems. The teachers, applying the rules and expectations of the Utilitarian Discourse System, brought to their viewing of the ads the following principles: 1) education should be morally neutral 98

2) education should respect the rights of the individual 3) health decisions are individual decisions 4) education operates best through a straightforward presentation of facts The students, on the other hand, appeared to bring a different set of principles to their viewing of the ads, principles similarly derived from the rules and expectations of their discourse system: 1) morality is an intrinsic part of education 2) education should train the individual in how to fulfil his or her role in the family and in the society 3) health decisions have consequences beyond the individual 4) education operates best through a process of direct transmission (Scollon, 1993) or empathy The separation of education and socialization has been a powerful concept within the Utilitarian Discourse System since the establishment of formal government-controlled public schooling (Scollon and Scollon, 1995). While not without detractors (present, for example, in the on-going debate on school prayer in the United States), the pretence, if not the practice, of morally neutral education remains central to most formal schooling in the West, and to modern Western conventions of health education. Though epidemics and plagues are often occasions for the violation of this principle, consistently precipitating moral crusades whose leaders attempt to link 99

the notions of sickness and sin (see Fee and Fox, 1988), most mainstream health educators in the West approach disease as the result of amoral physical factors and attempt to inform people of how these factors operate. Doctors and researchers in Europe and North America, for example, criticized the use of the word promiscuous in early medical writing about AIDS because it was seen as both judgmental and inexact, and therefore inappropriate for medical discourse (Bolton, 1992). Furthermore, whether or not people follow the advice of health professionals is seen as an individual decision. In Chinese society, however, moral education is seen as much a responsibility of the school as it is of the home. Wilson (1980) points out that formal schooling in Taiwan, Hong Kong, and the Peoples Republic of China lays heavy emphasis on training students in their obligations and responsibilities as members of families and of society, and on the consequences of not fulfilling these obligations and responsibilities. Illness is is also seen in moral terms primarily because it interferes with a persons ability to discharge his or her duties as a member of the group. Health decisions are family decisions. Kleinman (1980) writes: Their tenure in the family places upon them privileges and obligations, the chief of which is to improve the family fortune, while not bringing shame on it. Related to this obligation is another, requiring them to treat their own person and body as if they are as inviolable as the family. Since neither belongs to them alone, injury to the person is in part injury to the family. (p. 134)

100

Different discourse systems also have different notions about communication and its role in either exchanging information or ratifying relationships. Scollon (1993) has pointed out that Chinese and Americans use different metaphors to describe communication, Americans seeing language as a container for ideas, and Chinese seeing language in terms of bodily secretions. Such metaphors provide insight into the ideas and attitudes behind the preferred forms of discourse within the two discourse systems. For Westerners communication is regarded as an efficient vehicle for the transmission of ideas, as long as the ideas are packaged in a logical and straightforward way. Chinese, however, see communication as, at best, an imperfect means of understanding others thoughts and intentions. Real knowledge can only come from direct experience. The consequences of differing philosophies of language and communication can be seen in the different forms of discourse prevalent in socialization practices in Western and eastern cultures. Socialization involves not just apprenticeship in rules of behaviour, but also apprenticeship in the patterns of discourse through which such rules are normally expressed within a given culture. Shweder and Much (1987), in their comparison of methods of socialization between Indians and middle class Americans found that in India moral arguments are often presented in the form of a narrative account, whereas socialization of American children is based on rational analysis of rights and obligations. Metzger (1980) observes that moral ideas in Chinese society are derived from and exemplified in narratives of the struggles between heroes a villains. These stories, he says, provide frameworks within which new problems and situations are judged. Fung (1994) shows how stories told by and about 101

children in Chinese families operate as an intricate part of their moral education. It is therefore not surprising that the students in this study not only saw the function of the AIDS awareness APIs they watched as moral education, but also brought to them the essentially narrative framework which their culture associates with this task. And, for the most part, the framers of the messages fulfilled their expectations, presenting strong stories with identifiable heroes and villains and clear morals. This is not to dismiss the criticism of the campaign lodged by the teachers in this study or by myself in Jones (1995). It is only a reminder that such criticism itself comes from cultural frameworks, and that the design of culturally appropriate health education materials depends not just on an understanding of a cultures beliefs about health and disease, but also on an understanding of its beliefs about language. The suggestion that people from different cultures read public health messages in different ways has important implications, not just for health educators but for language educators as well, helping us to remember that the texts we bring into our classrooms may not be experienced in the same way by teachers and learners. Not only may participants bring different attitudes to different topics, they may also bring different expectations about the forms of discourse and rules of interaction appropriate to these topics. This fact should not be seen as a barrier to education, but as an opportunity to focus students awareness on relevant aspects of cultural models (Gee, 1989), for what we teach when we teach language is not just the mechanical encoding and decoding of linguistic information, but also the way speakers of different languages fit this information into different social and ideological frameworks. 102

When my students start talking about things like this, I just turn off, one teacher in this study admitted. I just cant stand their attitudes. The results of this research suggest that what this teacher perceives as her students inappropriate attitudes arise from a complex system of framing involving ideas not just about AIDS but about language itself. It is, therefore, at such moments that we as language teachers must resolutely turn on, must look behind what is being said and, together with our students, begin to unravel the web of cultural models that enclose us and the topics that we talk about. Indeed, it is at such moments that we can see our vocation grow to the task of teaching not just the practical rules of language use, but also how to recognize the borders that we place around ourselves and others in all the discourse that we create, and hopefully how to eventually transcend them.

103

104

Notes

1. Ho (1993) in an investigation into oral and written narrative recall strategies of Hong Kong secondary school students, found the opposite. Subjects in the study showed a marked reluctance to comment upon the internal states of the characters in the stories they read. This difference can perhaps be partly attributed to the different genres used in the two studies. Television commercials tend to present characters emotional reactions in much broader and more obvious ways than literary fiction. The transcription conventions used in this paper are chiefly based on those used by Tannen (1980) (and other linguists working on the Pear Stories project) with certain minor alterations. ... .. . , -^ indicates a measurable pause (more than 0.1 second) indicates a slight break in timing indicates sentence-final intonation indicates clause final intonation (more to come) indicates lengthening of preceding phoneme or syllable for words or syllables spoken with a heightened pitch (often as a phonetic signalling of a question) ? for questions (where the grammatical structure signals an interrogative regardless of intonation) CAPS for syllables spoken with heightened stress or loudness BOLD for syllables spoken with heighten stress or loudness which are already conventionally written in capital letters (such as AIDS) () for non-verbal utterances or behaviour such as laughter or gestures // to enclose transcriptions which are not certain |____ indicates overlapping speech | References

Adams, M.J., & Collins, A. (1979). A schema-theoretic view of reading. In Freedle, R.O. (ed.) New Directions in Discourse Processing, Norwood, N.J.: Ablex. pp. 1-22. Alcorn, K. (1989). AIDS in the public sphere. In E. Carter & S. Watney (eds.) Taking liberties: AIDS and cultural politics, London: Serpents Tail. pp. 157. Baddeley, A. (1990). Human memory: Theory and practice, Hove: Lawrence Erlbaum Associates. Benson, C. (1993). What do adolescents in Hong Kong know and think about AIDS and sex? Hong Kong Practitioner 15(3). pp. 2495-2501. Baggaley, J. (1993). Media health campaigning: not just what you say but the way that you say it! In A. Klusacek & K. Morrison (eds.) A leap in the dark: AIDS, art and contemporary cultures, Montreal: Vehicule Press. pp. 109119. Bateson, G. (1972). Steps to an ecology of mind, New York: Ballentine Books. Berger, P.L., & Luckmann, T. (1966). The social construction of reality, New York: Doubleday. Bolton, R. (1992). AIDS and promiscuity: Muddles in the models of HIV prevention. In R. Bolton & M. Singer (eds.) Rethinking AIDS prevention: cultural approaches, Langhorne, Pennsylvania: Gordon and Breach Science Publishers. pp. 7-85. Carbaugh, D. (1988). Talking American: Cultural discourses on Donahue, Norwood, NJ: Ablex. Carbaugh, D. (1990). Cultural communication and intercultural contact. Semiotica 80(1/2), pp. 15-35. Carrell, P.L. (1984). Evidence of a formal schema in second language comprehension. Reading in a Foreign Language 1(2). pp. 87-112.

Centers for Disease Control (1995). Information for HIV prevention community planning groups (#2): HIV/AIDS among young gay and bisexual men. Fact Sheet. September 1995. Chafe, W.L. (ed.) (1980). The pear stories: Cognitive, cultural and linguistic aspects of narrative production. Norwood, NJ: Ablex. Chafe, W.L. (1980). The deployment of consciousness in the production of a narrative. In W.L. Chafe (ed.) The pear stories: Cognitive, cultural and linguistic aspects of narrative production, Norwood, NJ: Ablex. pp. 9-50. Chenail, R.J. (1991). Medical discourse and systemic frames of comprehension, Norwood, N.J.: Ablex. Corder, S. Pit (1983). Strategies of communication. In Faerch, C. & Kasper, G. (eds.) Strategies in interlanguage communication, London: Longman. pp. 15-19. Cortazzi, M., & Jin, L.X. (1994). Narrative analysis: Applying linguistics to cultural models of learning. In D. Graddol & J. Swann (eds.) Evaluating language, Clevedon: Multilingual Matters. pp. 75-90. Dechert, H.W. (1983). How a story is done in a second language. In Faerch, C. & Kasper, G. (eds.) Strategies in interlanguage communication. London: Longman. pp. 175-195. Eco, U. (1995). Apocalypse Postponed. London: Flamingo. Emmott. C. (1994). Frames of reference: Contextual monitoring and the interpretation of narrative discourse. In M. Coulthard (ed.) Advances in written text analysis, London: Routledge. Fee, E., & Fox, D.M. (eds.) (1988). AIDS: The burdens of history, Berkeley: University of California Press. Erbaugh, M.S. (1990). Mandarin oral narratives compared with English: the pear/guava stories. Journal of the Chinese Language Teachers Association 25(2). pp. 21-42.

Evans, D.A., Block, M.R., Steinberg, E.R., & Penrose, A.M. (1986). Frames and heuristics in doctor-patient discourse. Social Science and Medicine 22. pp.1027-1024. Floyd, P., & Carrell, P.L. (1987). Effects on ESL reading of teaching cultural content schemata. Language Learning 37(1). 89-108. Flowerdew, J., & Miller, L. (1995). On the notion of culture in L2 lectures. TESOL Quarterly 29(2), pp. 345-373. Freedle, R.O., & Hale, G. (1979). Acquisition of new comprehension schemata for expository prose by transfer of a narrative schemata. In Freedle, R.O. (ed.) New directions in discourse processing, Norwood, N.J.: Ablex. Fung, H. (1994). The socialization of shame in young Chinese children. Unpublished Dissertation. Chicago: University of Chicago. Gatboton, E.C., & Tucker, G.R. (1971). Cultural orientation in the study of foreign literature. TESOL Quarterly 5. pp. 137-143. Gee, J.P. (1989) What do English teachers teach? (or Why isnt the Pope a bachelor?). Journal of Education, 171(1). pp. 135-146. Gee, J.P. (1990). Social linguistics and literacies: Ideology in discourses. London: Falmer Press. Goffman, E. (1971). The presentation of the self in everyday life. Harmondsworth: Penguin. Goffman, E. (1974). Frame analysis. New York: Harper & Row. Goffman, E. (1981). Forms of talk. Philadelphia: University of Pennsylvania Press. Goldstein, R. (1991). The implicated and the immune: Responses to AIDS in the arts and popular culture. In D. Nelkin, D.P. Willis, & S.V. Parris (eds.) A disease of society: cultural and institutional responses to AIDS, Cambridge: Cambridge University Press. pp. 17-42.

Gumperz, J. (1977). Sociocultural knowledge in conversational inference. In Saville-Troike, M. (ed.) 28th Annual Round Table Monograph Series on Language and Linguistics. Washington, DC: Georgetown University Press. Gumperz, J. (1992). Contextualization and understanding. In Duranti, A. & Goodwin, G. (eds.) Rethinking context, Cambridge: Cambridge University Press. Gumperz, J., & Hymes, D. (1972). Directions in sociolingusitics: The ethnography of communication, New York: Holt, Reinhart and Winston. Hall, S. (1973). Encoding and decoding in the television discourse. CCCS Stencilled Paper 7. Birmingham: University of Birmingham. Ho, C.F. (1993). An Investigation into narrative comprehension and recall of Cantonese-speaking students. Unpublished Masters Thesis. Hong Kong: City University of Hong Kong. Hong Kong AIDS Foundation (1993). Annual report 1992-1993, Hong Kong: Government Printing Office. Hsu, F. (1981). Americans and Chinese: Passage to differences. Honolulu: University of Hawaii. Hsu, F. (1985). Self in cross cultural perspective. In A.J. Marsella, G. Devos & F. Hsu (eds.) Culture and self: Asian and Western perspectives, New York: Tavistock. pp. 24-55. Hwang, D.H. (1988). M. Butterfly. New York: Penguin. Jin L. X., & M. Cortazzi (1993). Cultural orientation and academic language use. In D. Graddol & L. Thompson (eds.) Language and Culture, Clevedon: BAAL & Multilingual Matters. pp. 84-97. Johnson, P. (1981). Effects on reading comprehension of building background knowledge of a text. TESOL Quarterly 16. pp. 503-516. Jones, R. H. (1995). Talking about AIDS in Hong Kong: Cultural models in public health discourse. Paper Presented at the SEAMEO Regional Seminar on Language in Education, April 17-19, Singapore.

Kaplan, R. (1966). Cultural thought patterns in inter-cultural education. Language Learning, 16. Keeney, B.P. (1990). Recursive frame analysis: A method for organizing therapeutic discourse. Therapia Familiare 33. pp. 25-39. Kleinman, A. (1980). Patients and healers in the context of culture, Berkeley: University of California Press. Kumatoridani, T. (1982). The structure of persuasive discourse: A crosscultural analysis of the language in American and Japanese television commercials, Unpublished Dissertation. Washington, D.C.: Georgetown University. Lau, S.K., & Kwan, H.C. (1988). The ethos of the Hong Kong Chinese, Hong Kong: The Chinese University of Hong Kong. Leavitt, C. (1968). Response structure: A determinant of recall. Journal of Advertising Research, 8. pp. 3-6. Lewis, J. (1994). The meaning of things: Audiences, ambiguity and power. In J. Cruz & J. Lewis (eds.) Viewing, reading and listening, Boulder, CO: Westview Press. pp. 19-32. Liebe-Harkort, M.L. (1989). Interactional ethnolinguistics. In Oleksy, W. (ed.). Contrastive pragmatics, Amsterdam: John Benjamins. pp. 101-111. Lin, A (1996). Doing English lessons in secondary school: A sociocultural and discourse analytic Study. Unpublished Doctoral Dissertation. Tornoto: University of Toronto. Lupton, D. (1994). Moral threats and dangerous desires: AIDS in the news media, London: Taylor and Francis. Mahoney, D., Shillaw, J., & Hull, J. (1991). Sorting simple stories: Narrative recall and the Cantonese speaker. Perspectives: Working papers of the department of English, City University of Hong Kong 3(2). pp. 57-80. Mandler, J.M. (1978) A code in the node: the use of story schema in retrieval. Discourse Processes 1. pp. 14-35.

Mays, V.M. (1992). The language of black gay mens sexual behavior: Implications for AIDS risk reduction. The Journal of Sex Research 29. pp. 425434. McConville, M.N., & Leavitt, C. (1968). Predicting product related recall from verbal response. Proceedings of the American Psychological Association Annual Convention, Washington D.C.: American Psychological Association. pp. 677-678. Metzger, T. A. (1980). Selfhood and authority in neo-Confucian political culture. In A. Kleinman and T.Y. Lin (eds.) Normal and abnormal behaviour in Chinese culture, London: D. Reidel Publishing Company. pp. 7-27. Meyrowitz, J. (1985). No sense of place: The impact of electronic media on social behavior, Oxford: Oxford University Press. Michaels, S. (1981). "Sharing time": Childrens narrative styles and differential access to literacy. Language and Society 10. pp. 423-442. Miller, P.J., Mintz, J., Hoogstra, L., Fung, H., & Potts, R. (1992). The narrated self: Young childrens construction of self in relation to others in conversational stories of personal experience. Merrill-Palmer Quarterly 38(1). pp. 45-67. Miracle, G.E. (1987). Feel-do-learn: an alternative sequence underlying Japanese commercials. In F. Feasley (ed.) Proceedings of the 1987 Conference of the American Academy of Advertising. Morley, D. (1980). The nationwide audience. London: British Film Institute. National Academy of Sciences (1988). HIV infection and its epidimiology. In Confronting AIDS: Update 1988, Washington, D.C.: National Academy of Sciences Press. pp. 33-55. Patton, C. (1990). Inventing AIDS, New York: Routledge. Philipsen, G. (1987). The prospect for cultural communication. In Kincaid, L. (ed.) Communication Theory: Eastern and Western Perspectives, New York: Academic Press. 245-254.

Piaget, J. (1955). The Language and Thought of the Child, New York: Meridian Books. Pinxten, R. (1991). Fieldwork as a form of intercultural communication. In Blommaert J. & Verschueren, J. (eds.) The pragmatics of intercultural and international communication, Amsterdam: John Benjamins. pp. 131-143. Pritchard, R. (1990). The effects of cultural schemata on reading processing strategies. Reading Research Quarterly 25(4). pp. 273-275. Puto, C.P., & Wells, W.D. (1984). Informational and transformational advertising: The differential effects of time. In T.C. Kinnear (ed.) Advances in Consumer Research, Vol 11, Ann Arbor, MI: Association for Consumer Research. pp. 638-643. Riley, P. (1989). Well dont blame me! On the interpretation of pragmatic errors. In Oleksy, W. (ed.) Contrastive Pragmatics, Amsterdam: John Benjamins. pp. 231-249. Ross, R.N. (1975). Elipsis and the structure of expectation. San Jose Occasional Papers in Linguistics 1. pp. 183-191. Rumelhart, D.E. (1975). Notes on a schema for stories. In Bobrow, D.G. & Collins, A.M. (eds.) Representation and understanding, New York: Academic Press. pp. 211-236. Schmidt, R., Shimura, A., Wang, Z.G., & Jeong, H.S. (1990). Suggestions to buy: Television commercials from the U.S., Japan, China and Korea. Paper presented at the World Congress of Applied Linguistics. Thessaloniki, Greece. April 15-21. Scollon, R. (1994). Discourse identity: Person and power in business and news discourse. Hong Kong: City University of Hong Kong, ms. Scollon, R. (1995a). Matching discourse identity to social identity. Paper delivered at the 5th International Conference on Cross-Cultural Communication: East and West. Heilongjiang University, Harbin, China. August 15-19, 1995.

Scollon, R. (1995b). Cultural codes for calls: Knocks, coughs, shuffles, and other ways of opening up social interaction on Asian and American television. Hong Kong: City University of Hong Kong, ms. Scollon, R. (1995c). The ethnographic study of receptive communication: Reading as a kind of watch. Hong Kong: City University of Hong Kong, ms. Scollon, R. (1995d). Plagiarism and ideology: Identity in cultural discourse. Language in Society 24, pp. 1-28. Scollon, R., & Scollon, S.W. (1981). Narrative, literacy and face in interethnic communication, Norwood, NJ: Ablex. Scollon, R., & Scollon, S.W. (1984). Cooking it up and boiling it down: abstracts in Athabaskan childrens story retellings. In D. Tannen (ed.) Coherence in spoken and written Discourse, Norwood, NJ: Ablex. Scollon, R., & Scollon, S.W. (1995). Intercultural communication, London: Basil Blackwell. Scollon, S.W. (1993). Metaphors of self and communication: English and Cantonese. Perspectives: Working papers of the department of English 5(2). Hong Kong: City University of Hong Kong. pp. 41-62. Seidel, G. (1990). Thank God I said no to AIDS: On changing discourses of AIDS in Uganda. Discourse and Society 1(1), pp. 61-84. Shank, R.C. & Ableson, R.P. (1977). Scripts, plans, goals and understanding, Hillsdale, N.J.: Erlbaum. Shweder, R.A., & Much, N.C. (1987) Determinations of meaning: Discourse and moral socialization. In Kurtines, W.M., & J.L. Gewirtz (eds.) Moral development through social interaction, New York: John Wiley & Sons. pp. 197-244. Skovmand, M. (1992). Barbarous TV international. In M. Skovmand & K.C. Schroder (eds.) Media cultures, London: Routledge. pp. 84-103. Sontag, S. (1991). Illness as metaphor & AIDS and its metaphors, London: Penguin.

Steffensen, M.S., Joag-dev, C.,& Anderson, R.C. (1979). A cross cultural perspective on reading comprehension. Reading Research Quarterly 15, pp. 10-29. Stoddard, T. (1989). Paradox and paralysis: An overview of the American response to AIDS. In Carter, E. & S. Watney (eds.) Taking liberties: AIDS and cultural politics. London: Serpents Tail. pp. 95-106. Tannen, D. (1979). Whats in a frame? In Freedle, R.O. (ed.) New directions in discourse processing, Norwood, N.J.: Ablex. pp. 137-181. Tannen, D. (1980). A comparative analysis of oral narrative strategies: Athenian Greek and American English. In W. L. Chafe (ed.) The pear stories: cognitive, cultural, and linguistic aspects of narrative production. Norwood, NJ: Ablex. pp. 51-87. Tannen, D. (1984). Spoken and written narrative in English and Greek. In D. Tannen (ed.) Coherence in spoken and written discourse, Norwood, NJ: Ablex. pp. 21-41. Tannen, D. (1990). Whats in a frame? In Tannen, D. (ed.) Framing in discourse, New York: Oxford University Press. pp. 14-56. Tannen, D. (ed.) (1990). Framing in discourse. New York: Oxford University Press. Tannen, D. (1994). Gender in discourse, Oxford: Oxford University Press. Tannen, D., & Wallat, C. (1983) Doctor/mother/child communication: Linguistic analysis of a paediatric interaction. In Fisher, S. & Todd, A.D. (eds.) The social organization of doctor-patient communication, Washington D.C.: Center for Applied Linguistics. pp. 203-219. Tarone, E. (1983). Some thoughts on the notion of communication strategy. In Faerch, C. & Kasper, G. (eds.) Strategies in interlanguage communication, London: Longman. pp. 61-74.

Thorson, E. (1989). Television commercials as mass media messages. In J.J. Bradac (ed.) Message effects in communication science, Newbury Park: Sage Publications. pp. 195-230. Treicher, P.A. (1988). AIDS, gender and biomedical discourse. In E. Fee & D.M. Fox (eds.) AIDS: The burdens of history, Berkeley: The University of California Press. pp. 190-266. Treicher, P. A. (1993). Seduced and terrorized: AIDS and network television. In A. Klusacek & K. Morrison (eds.) A leap in the dark: AIDS, art and contemporary cultures. Montreal: Vehicule Press. pp. 136-151. Tulloch, J. (1989). Australian television and the representation of AIDS. Australian Journal of Communication. 16, pp. 101-124. Tulloch, J. (1992). Using TV in HIV/AIDS education: production and audience cultures. Media Information Australia. 65, 28-35. Vestergaard, T., & Schroder, K. (1985). The language of television advertising, Oxford: Basil Blackwell. Watanabe, S. (1990). Cultural differences in framing: American and Japanese group discussions. In Tannen D. (ed.) Framing in discourse, New York: Oxford University Press. pp. 176-209. Watney, S. (1989). Taking liberties. In E. Carter & S. Watney (eds.) Taking liberties: AIDS and cultural politics. London: Serpents Tail. pp. 1-57. Wells, W.D. (1987). Lectures and drama. Paper presented at the Association for Consumer Research, Boston. October. West, C. (1984). Medical misfires: Mishearings, misgivings and misunderstandings in the physician-patient dialogues. Discourse Processing 7. pp. 107-134. Wilson, R.W. (1980). Conformity and deviance regarding moral rules in Chinese society: A socialization perspective. In A. Kleinman and T.Y. Lin (eds.) Normal and abnormal behaviour in Chinese culture, London: D. Reidel Publishing Company. pp. 117-136.

Wildner-Bassett, M.E. (1989). Coexisting discourse worlds and the study of pragmatic aspects of learners interlanguage. In Olesky, W. (ed.) Contrastive Pragmatics, Amsterdam: John Benjamins. pp. 255-275. Witte, K. (1991). The role of culture in health and disease. In L.A. Samovar & R.E. Porter (eds.) Intercultural communication, 6th edition, Belmont, CA: Wadsworth Publishing Company. Young, L.W.L. (1982). Inscrutability revisited. In J. Gumperz (ed.) Language and social identity. Cambridge: Cambridge University Press.

Appendix 1 AIDS Awareness APIs Used in the Study Produced by the Hong Kong Government 1987-1994 Title Use Condom (MTR) Bar Launching Date 01/07/87 Description Man and woman walking into Mass Transit Rail station. Two men meet at a straight bar and leave together. Slogan Use condoms for safer sex.

01/04/88

You may be the next victim... You never know who could be an AIDS

Youth and Prostitution Homosex uals

10/11/88

Portrayal of two young men going to a prostitute. Two men meet in outdoor cafe. They begin to leave together, but one changes his mind when he sees AIDS poster on public light bus. Portrays businessmans encounter with a prostitute interspersed with scenes of his funeral.

carrier. It only takes one sexual encounter to pass AIDS on to you. Why risk it? It only takes one sexual encounter to pass AIDS on to you. Why risk it?

10/11/88

Infection of Ordinary People (Family)

10/11/88

It only takes one sexual encounter to pass AIDS on to you. Why risk it?

Title Safer Sex/ Condom

Launching Date 01/01/90

Description Portrays man in cheap hotel room with companion (gender indistinguishable) in bed behind him. Portrays man returning from a trip abroad while his friends speak in the voice-over about his sexual adventures. Teenage girl finds out that a boy she knows

Slogan AIDS kills. Use a condom.

AIDS & Travellers

01/12/90

When travelling abroad you need to take extra precautions to avoid exposure to AIDS. Think about AIDS. It could happen to

Girls

01/08/93

Husband

01/08/93

has AIDS. A man who is just about to become a father finds out an old girlfriend has AIDS.

you. Think about AIDS. It could happen to you.

Salon

23/06/94

The more you Woman who is expecting a child talks know, the less the risk. to her friends in the beauty salon about having an AIDS test.

Appendix 2 Subject Profile 1. Students Place of Origin Hong Kong Education Age Sex

Numbe r All

BA TESL (3rd year)

21(5), 22(4), 24(1)

M(5), F(5)

2. Teachers Number 1 2 3 Place of Origin Canada Britain Britain Education MA PhD BEd Years Teaching 5 1/2 21 15 Years in Hong Kong 5 1/2 3 9 Age 33 45 43 Sex M M F

4 5 6 7 8 9 10

Britain U.S.A. U.S.A. U.S.A. Britain Australia Bahamas

MA MA BA MA Dip. TESL MA PhD

10 9 7 15 3 8 4

10 8 less than 1 5 1 7 4

53 31 32 43 34 44 37

F F M M M F F

Appendix 3a Average Number of Words per Thought Unit Teachers 8.4 Appendix 3b Average Number of Thought Units Per Response Commercia l MTR Bar Youth & Pros Homosexuals Family Safe Sex Travellers Husband Girls Teachers # of I.U. 13.8 11.8 13.7 12.9 15.2 20.9 17.4 17.2 17.5 Students # of I.U. 8.1 9.1 11.7 11.6 13 12.1 16.6 16.4 16.1 Students 8.5

Salon Student Avg. 13.35 Teacher Avg. 15.42

18.5

18.6

Appendix 5: Frames and Modes of Response Appendix 4a Frames Experiment Frame Social Frame Viewer Frame Commercial Frame Narrative Frame Physical Frame

Appendix 4b Frames in Subjects Responses to AIDS APIs Summary Frame Experiment Frame Social Frame Commercial Frame Viewer Frame Narrative Frame Physical Frame Teachers 2.7% 1.8% 39.4% 13.8% 26.5% 15.8% Students 2.7% 2.6% 24.2% 4.6% 61.2% 6.5%

Appendix 4c Modes of Response Summary Mode of Response Reporting Interpreting Judging Misc. (Brackets, etc.) Teachers 51% 30.2% 16% 2.8% Students 59.4% 35.6% 2.3% 2.7%

Appendix 4d Frames/Modes of Response in Subjects Responses to AIDS APIs All Commercials Frame Percent of Idea Units Teachers EXP SOC--JUD SOC--INT SOC--REP CM--JUD CM--INT CM--REP VW--JUD VW--INT VW--REP NAR--JUD NAR--INT NAR--REP PHY--JUD PHY--INT PHY--REP 2.6% 0 .9% .9% 12.2% 17.6% 9.4% 1.7% 1.9% 10.2% 1.4% 7.8% 17.3% .7% 1.9% 13.2% Students 2.7% .1% 2.2% 1.3% 1.7% 11.7% 11% .3% 1% 3.6% .2% 19.5% 37% 0 1.2% 6.3%

Appendix 4e Patterns of Response Pattern NAR CM NAR--CM PHY--NAR--CM CM--PHY/NAR--CM VW--PHY/NAR/CM--VW Other (2 or fewer responses) Students 12 1 7 22 17 1 40 7 2 7 33 15 36 Teachers 0

Appendix 5: Information and Lexical Choice

Appendix 5a Total Occurrence of Pronouns Word I, me, my you (meaning one or viewer) we they,them,their (meaning characters) they,them,their (meaning authors) they,them, their (other) Appendix 5b Grammatical Markers Marker future tense perfect tenses subjunctive mood Appendix 5c Total Occurrence of Selected Connectives Word Teachers Students Teachers 26 73 51 Students 49 8 26 Teachers 394 160 41 58 85 11 Students 200 120 16 237 9 10

because then and Appendix 5d

32 66 423

79 178 572

Negation in Viewer Frame (words collocating with not/*nt) Word Teachers understand, get, figure 15 out know 31 remember 5 sure 9 think 11 see 16 Appendix 5e Total Occurrence of Selected Lexical Items Word AIDS death/ dead/ die* sex* homosexual gay condom prostitute government husband/ wife friend/s businessman message Teachers 117 8 76 5 12 34 10 0 27 23 3 55 Students 160 23 136 5 2 33 21 7 62 60 10 19 Students 2 26 10 14 2 2

warn* Chinese interesting effective

5 21 21 12

19 8 4 4

Appendix 5f Elements in Subjects Responses 1) MTR object/event/message Students Billboard as Narrative Device Billboard as Informational Device Use Condoms Condoms Not Safe 5 0 10 3 Teachers 2 5 4 0

2) Bar Object/Event/Message Students 2 men leave (gay relationship) 2 men leave (straight/friends) 5 2 Teachers 2 2

man and woman leave (heterosexual) 1 man leaves alone 2 women leave dont remember/ not mentioned use condoms condoms not 100% effective get AIDS easily 3) Youth & Prostitution object/event/message

1 0 1 1 1 5 4

4 2 0 0 5 0 0

Students One Inexperienced (innocent) Character Both Characters Inexperienced (innocent) Possible Gay Narrative Sex Can Lead to Death 6 2 0 5

Teachers 3 2 3 0

4) Homosexuals object/event/message Students Gay Narrative Unsure/Unclear Eye Contact 9 1 5 Teachers 4 5 0

Bill Ad on Bus (Narrative Device) Ad on Bus (Informational Device) Door Slam

4 6 0 4

4 3 2 0

5) Family object/event/message Students Family Wife Suspicious Man Died of AIDS 7 5 6 Teachers 4 0 2

6) Safer Sex object/event/message Students Man and Woman Man and Man Cant Tell Man Alone 7) AIDS & Travellers Object/Event/Message Students Going Abroad 5 Teachers 1 5 1 4 0 Teachers 5 3 1 1

Returning from Abroad Photos = Prostitutes Thailand/Southeast Asia Man Got AIDS Nurse IV

1 4 5 6 3 0

4 1 2 0 0 3

8) Husband Object/Event/Message Students Character Frightened/Shocked 5 Teachers 0

9) Girls Object/Event/Message Students Caller = Friend Student Dropping Pencil Callers Question (Did you sleepwith him?) Sally Slept with Raymond Sally Afraid/Shocked/Astonished Sally Casual/Sociable/Open to Sex Avoid sex 9 5 4 7 6 6 2 3 Teachers 5 3 2 2 6 2 0 1

10) Salon Object/Event/Message Students Suspicions of Husband Reason for Test Hairdresser Embarrassed/Shocked Hairdresser Changes Attitude Talk about AIDS 6 6 4 7 Teachers 3 3 1 4

Appendix 6: Miscellaneous Appendix 6a Number of Subjects Reporting Having Seen Commercials

Commercial 1. MTR 2. Bar 3. Y & P 4. Homosexuals 5. Family 6. Safer Sex 7. Travellers 8. Husband 9. Girls 10. Salon

Students 9 8 7 6 5 7 9 8 10 6

Teachers 2 2 3 3 3 6 5 3 4 7

Appendix 6b Ads Judged Most and Least Effective

Commercial

Most Effective Students Teachers

Least Effective Students 2 Teachers 2 1

MTR Bar Youth & Prostitution Homosexuals Family Safer Sex AID & Travellers Husband Girls Salon 2 2 2 1 6 4 2

2 2 4 1 1 1 1 1

Appendix 6c Total Occurrence of Laughter During responses Teachers 31 Students 110