Teens and the Regeneration of Meaning The Use of Peer-Led Discussion Groups in Ethnographic Inquiry

new title:Learning from Teens and the Peer-Led Discussion Group The Journey from Post-Positivist to Constructivist Methods

Lynn Schofield Clark, Ph.D.
Research Associate and Post-Doctoral Fellow
Center for Mass Media Research
University of Colorado
Box 287
Boulder, CO 80309
Lynn.Clark@Colorado.EDU
Presented to the Popular Communication Division, International Communication Association
May, 1999
*Draft*

“But how do you know your informant is typical?” , demanded a renowned qualitative media researcher as I concluded my preliminary analysis of case study research. I struggled for an explanation. I had hoped that my analysis, which centered on my interactions with a single teenage girl over a period of more than a year, would seem as promising and rich to them as it had to me. Yet in that instance I realized -- belatedly -- that my reasons for integrating the insights of recent sociological, cultural anthropological and critical feminist research methods into qualitative media inquiry were far from self-evident. I had borrowed from an emergent research paradigm yet due to the dominance of the paradigms preceding it, I still needed to address the assumptions inherent in those earlier approaches.

The moment of that question turned out to be a pivotal one in the life of the study in which I have been involved for the past three years. It forced me to explore the different vocabularies and related assumptions that guide both research projects and their evaluation within media studies. In this paper, I would like to look back over the preceding years and attempt an answer to that question addressed to me two years ago. I begin by exploring the changing nature of research within the context of contemporary debates over knowledge and its production, a movement I trace in my own work and that of the larger study as one from post-positivist to constructivist research paradigms. Then, I introduce four categories of post-positivist research that have been challenged within constructivist methods, discussing how these emerged in relation to the pragmatics of the research project in which I and the larger team are engaged. I will conclude with some of the methodological issues we face today, and how their resolution relates to the initial findings to be presented here by other members of the research team.


Qualitative Media Research From Symbolic Interaction to Cultural Studies

It could be said that the U.S. qualitative media research tradition predates the "field" of mass communication itself. Models for its current form may be traced in part to the early 20th century Chicago school sociology and the development of symbolic interactionism. This paradigm, associated in its beginnings with John Dewey, William James, George Herbert Mead, Thomas Park, and Charles Horton Cooley, and later with Erving Goffman, Anselm Strauss, and Herbert Blumer, examined the interaction between subject and object relations in conversation and the ways in which people come to know the self through interactions with others. With its emphasis on human relationships, this tradition advocated the full immersion of the researcher into the context of research and in this sense, generally developed its theories of "interaction" to a much greater extent than its theories of the "symbolic." Consistent with the goals of the Enlightenment project and American pragmatism, research was pursued with the intent of finding solutions for the problems that had emerged as a result of industrialization and urbanization. Current debates concerning the “political” agenda and role of research and its practical outcomes, therefore, have deep roots.

While qualitative research fell out of favor for several decades with the emergence of survey methods, the tradition began to thrive again on this side of the Atlantic in the 1970s with the work of the Birmingham School following Stuart Hall. This influence has been central to the reconceptualization of qualitative research into what is more broadly known as cultural studies, foregrounding culture as a site of analysis. The cultural studies tradition of E.P. Thompson, Richard Hoggart and Raymond Williams foregrounded the roles of both historical context and social/structural critiques, introducing new concerns into the research agenda.

Current methodological approaches build upon various combinations of these preceding traditions. Guba and Lincoln organize the field along a continuum that is helpful here they argue that research tends to be conducted within the paradigms of positivism, post-positivism, critical theory, or constructivism, and each offers differing perspectives on the nature of research inquiry itself. Thus while some researchers today aim for descriptive narratives such as those of the symbolic interactionists and E.P. Thompson, for instance, others build upon the categories of social science inherited from the U.S. positivist tradition and are more concerned with developing narratives within certain standards of validity and reliability. Some writings take the form of sweeping historical narratives, while others are narrow stories of the researcher’s own process of knowledge-discovery. It is easy to surmise that the paradigms sparring to replace (or supplement) survey-based empirical research encompass many different traditions and assumptions.


The Symbolism, Media and the Lifecourse project Original research design issues

It is fair to say that prior to the last decade, the bulk of qualitative research produced in media studies both here and abroad was in the “post-positivist” vein. Thus in the original design of the Symbolism, Media and the Lifecourse project, we operated under many of the assumptions of post-positivism, which I will describe more broadly for background.

Research in the post-positivist tradition critiques positivism’s assumptions of an immutable and objectively apprehendable reality, taking instead a more modest aim. Post-positivists argue that research should aspire to describe a reality that is dynamic and can only be known imperfectly. They note that survey research often assumes a standardized instrument that may be interpreted differently by various subgroups of a society. In response, post-positivist researchers instead seek to get as close as possible to the meanings ascribed to various words and actions by members of the groups in which they are interested. Recognizing that no research can be completely free of biases, post-positivists nevertheless aim for objectivity as an ideal stance toward research, arguing for cross-study comparisons so as to verify findings with qualitative research’s smaller sample sizes.

The Symbolism, Media and the Lifecourse project began with the intent of a multi-methodological qualitative study whose end result would lay the groundwork for a large-scale survey. Such a survey could then add further verification to our findings and also provide the basis for generalization to the population at large. Our goals, consistent with post-positivist qualitative research, were quite general we were interested in exploring the relationship between individual media meaning-constructions and the cultural trends encompassing the decline of both membership in and cultural authority of the historical institutions of religion. Still, we began with a latent hypothesis that major moments of ‘passage’ in the life course are important sites of personal reflection and transformation. As such, they are the key points in life through which the cultural-symbolic realm avails itself to the individual as a resource for identity-construction, transformation, or development.

With this in mind, we developed a research instrument that would query people about their media use during (and surrounding) those moments of significant change. We wondered if media resources might complement or replace the support for such changes that were traditionally provided in religious settings and rituals.

Methodologically, we started with an ambitious goal, even by qualitative measures we would begin with a small number of household-level observations, patterning our efforts after Silverstone and Hirsh and Lull. Next, we would add more households to the sample in order to form “qualitative interview panels.” These panels would consist of several households, each of which would be revisited over a period of six to nine months. We anticipated employing the same interview instrument each time so as to maintain standardization across time, hoping that in this way we could observe the encounters with “life passages” or crises and how these might change the family’s approach to media. In the third and final stage of the research, we would develop a series of focus groups using representative demographic groups.

Like other post-positivist researchers, we sought to distinguish ourselves from the traditions of survey research, and assumed that our findings might contradict those gleaned from the survey-based research conducted in the preceding decades. As the original prospectus stated Most research is discontinuous with daily life, i.e., it attempts a “slice of time” and therefore misses the ways cultural symbols contribute to consciousness and the “flow” of life. People are not passive receivers of knowledge and information. They actively decode and use knowledge and information they encounter in the public, symbolic environment. What we don’t know much about is how they classify and structure those inputs, and where various kinds of messages fit in the ongoing “moral economy of the household.”

This stated, we turned our attention to the design of a research project that would bring us as close as possible to the “life worlds” of those we sought to study.

In the spring of 1996, I began interviewing families of teenagers with two goals first, as the associate director on the Symbolism research project, I wanted to perfect the instrument that Stewart Hoover, director of the project, could then use for the qualitative interview panels. Second, I also wanted to collect data for my own dissertation research on teens, religion and the media. I designed a two-stage interview guide for my initial dissertation-related interviews. We employed a research assistant to help conduct the initial interviews, and by July we had interviewed twelve families and I had done an in-home observation. The learnings from this preliminary work were interesting, but nothing particularly stood out -- except, unfortunately, the fact that we were finding that our concept of the “moments of passage” was proving less fruitful than we’d hoped. This had both theoretical and methodological implications, as we needed to decide whether or not to abandon our original plan, a daunting prospect at the time.

Serendipity plays an often underrated role in qualitative research, however, and several factors worked together in what was to become the first of several shifts in the overall project and goals of the Symbolism project. We met with a number of consultants that summer, one of whom was Judith Stacey, author of the important feminist sociological work, Brave New Families. Her model of in-depth, case study research intrigued me, and after our meeting I was convinced that by getting to know a few teens quite well, my own perspective on the meaning-making and identity-construction projects of teens could be greatly strengthened. It was clear after my initial interviews that two meetings would not be nearly enough to get the kind of detail and discussions I knew I wanted from my teen informants. Working with teens, as I knew from my own pre-graduate school experience, required the development of a relationship based on trust, and that could only be achieved with time and intentionality. My own research goals and those of the larger project diverged somewhat as I sought out and developed ongoing relationships with five teens over the course of the next year.


Teens and the peer-led discussion group

It was in my interactions with my teen case studies that I began to reconsider much of the post-positivist assumptions under which the project had been designed. I had already learned that in talking with me about my research interests in media, religious beliefs, and identity, the teens with whom I worked had become more reflexive about their own media use, which in turn made them better informants. Still, I retained some post-positivist impulses, thinking that I needed to expand my “data-gathering” beyond “just” five teens. So, following advice from Elizabeth Bird at an ICA meeting years earlier, I worked with the first teen, Elizabeth, to develop an interview guide for a focus group that she would organize and lead with her friends.

Elizabeth and I met together several times, discussing various possible topics for a teen focus group, ranging from teen magazines to the telephone to television programs and the Internet. The design that emerged was this I would train Elizabeth in the leading of a focus group, give her money to recruit six of her friends as well as money for pizza, and she would tape-record the conversations. Later, I would transcribe the tapes and we would go over them later for accuracy. I cleared the plan with both Stewart Hoover, the project’s director, and with the university’s Human Research Committee, and off I went.

As Lunt and Livingstone have noted, focus group methods in media audience research have been used for both critical and institutional purposes beginning with Merton’s employment of the method as a supplement to quantitative approaches. In keeping with post-positivism, focus group research has assumed an apprehendable but largely inaccessible “reality” beyond the interview situation, and thus focus groups are designed for “getting in tune with the reality of the interviewee,” as one textbook on focus groups notes. Moreover, some focus group organizers strive for the “objective” distance between themselves and their interviewees, noting that pre-formed social relations may interrupt this process. As one textbook notes Caution should still be used when considering focus groups with close friends, family members o rrelatives, or work groups. People who regularly interact, either socially or at work, present special difficulties for the focus group dsicussion because they may be responding more on past experiences, events, or discussions than on the immediate topic of concern.

My interest in holding the focus groups was already slightly different than that outlined in the post-positivist agenda. I wanted to better understand rather than avoid these contexts of past experiences and other relevant factors and the ways in which they mediate meaning-making practices in the group’s discussion, for I believed that these would shape the narratives that emerge both in the group and beyond it. My own design, therefore, included groups of people who knew each other prior to the group’s gathering, follows the work of Liebes and Katz, who employed this approach to simulate social occasions in which similar discussions about television programs might emerge. Liebes and Katz, as well as Radway in her work on romance readers, were interested in approaching participants as members of a social group, rather than as individual representatives of certain subcultural groups.

I began with a rather instrumentalist approach to my relationship with Elizabeth, viewing her initially as a paid “data-gatherer” for me; our relationship had been defined less by my interest in her own experiences, per se, than by my interest in the work she was doing for me to help me gain access to teen situations and discussions. It was not until I listened to the audiotape of the first group’s gathering that I realized how valuable it had been to know so much about the context of her life before the taping. Gradually, I began to see Elizabeth more as a partner in an ongoing dialogue about both my research interests and her own life more generally. I trusted her enough to tell her about my research project and how its goals related to my own personal life, and she in turn shared reflections on both her own life and its relation to her media practices. In other words, we both “self-revealed,” to use the language of constructivist research. Over the course of the next two years we continued to see each other, sometimes informally for social outings and at other times for more formal interviews. We also communicated frequently e-mail. My relationship with Elizabeth, while occasionally still related to my research, has evolved into a mentorship we now talk much more frequently about her homework assignments and other school challenges than about her media use. This, then, became a model for my relationships with the other four teens with whom I worked most intensively. When I reflected on how much I had learned from the so-called “training” sessions with Elizabeth, for instance, I decided to audiotape and analyze these sessions as well as the group’s conversations involving the other peer-led discussion leaders. These, along with the post-group conversations, became key to my analytical work, again reinforcing the notion that the teens were working with me dialogically. Not only did they answer my preconceived questions, but they discussed my research goals and preliminary findings with me, providing insights ranging from how to ask questions to which categories of interpretation seemed most relevant to them.

The fact that the teens led the groups without any adults present is also significant, of course. By having the teens engage in the group without adults, I did not simply want to attempt to replicate an “actual” conversation among teens. Rather, I wanted to turn over some of the authority of ethnographic inquiry to the teens, and to my peer-group discussion leaders in particular. As I believe that teens are most comfortable conversing when they can control the terms of the conversation, I wanted to -- in fact I felt I needed to -- facilitate this control. The results from the group, therefore, were sometimes disappointing to me, as I did not learn exactly what I had anticipated. Upon first listening to the tapes, for example, I felt disappointed with interruptions such as “She already answered that. Next question,” or outbursts of uncontrollable laughter, or even the failure to ask what I believed were “obvious” follow-up questions. Yet I grew to sense that the content of the conversations were guided to a much greater extent by the leaders’ (and group members’) varied levels of comfort with and interest in the topics than would have been the case had I been present to moderate. This encouraged me to pay closer attention to the nuances that were present in the conversations, such as the various cues that caused embarrassment, a quick change of topic, or loud and repeated efforts at being heard by one or more participants.

I should note, too, that each of the peer leaders handed over the tapes with some embarrassment, concerned that I would not be able to “get much out of it” because their group members “didn’t take it very seriously.” Yet I was surprised that this embarrassment was not present at all while we listened to the tapes and read through the transcripts together. I think this is evidence of the fact that the teens sensed the somewhat ambiguous nature of the research endeavor. They correctly surmised that as an adult researcher I would have some specific desires for coherence, clarity, and seriousness, yet they also sensed that the group’s talk did, in fact, represent “how teens talk about these things,” which is exactly what I had expressed an interest in.

Gradually, I came to see that the purpose of my research did not lie in identifying and describing a “typical” or “representative” American teen, but to identify the structures and social relationships that inform and limit the ways in which teens come to construct a narrative of identity -- in the case of my research, focusing on the role of the media in informing notions of religion and religious beliefs.

At that moment, my concerns regarding post-positivism changed my dissertation, but not the overall project. The re-thinking of post-positivism was strengthened with the entrance of three new research assistants into the Symbolism project, however Lee Hood, Joe Champ, and Diane Alters. As each went into the field, they expanded upon the original interview guide, bringing back strikingly different and interesting life stories. Still operating within post-positivism, we nevertheless began referring to the households as case studies, each of interest on its own terms, and we wondered how will we create cross-case comparisons for our analyses? For the moment, we abandoned that concern, as these three researchers instead developed analyses based on individual cases. While we did not realize it then, we were well on our way from a positivist to a constructivist research paradigm. Before proceeding with our own story, however, let me return now to the broader discussion, setting our own experience within the context of contemporary developments in qualitative studies.


Post-Positivism and its Challenges

In the past decade, the post-positivist model of research has increasingly come under criticism. This is the result of various factors, from postmodernism to critical theory and feminism, to the contradictions inherent in the history of qualitative research itself. To explore how these criticisms relate to both the constructivist research paradigm and the work of the research project specifically, I turn now to the description of post-positivism’s writings that Van Maanen refers to as“realist tales” These tales provide a rather direct, matter-of-fact portrait of a studied culture, unclouded by much concern for how the fieldworker produced such a portrait...The representation of social reality [is] seen as technically unproblematic once the facts [have] been unearthed.

Van Maanen notes four characteristics of the “realist tale, which we impliciltly expected to follow in our own research” first, the researcher strives to demonstrate the “typicality” of the persons and situations observed and interviewed, thereby justifying their worth in relation to larger-scale, presumably survey-based studies. Second, in writing up the data the researcher strives to present “the natives’ point of view,” to use the famous phrase articulated by ethnography’s father, anthropologist Bronislaw Malinowski. This has come to include not only what the person says and does, but also, as Geertz has asserted, how the person makes sense of his or her actions and statements and how this fits into the larger sense-making schema of the culture. Third, in the presentation of the final analysis the author is absent; the assumption being that what is ultimately reported represents what any person would have observed or surmised, had they been in the field instead of the researcher. Finally, realist tales, Van Maanen asserts, take for granted that their interpretation is the correct interpretation of the situation observed. These tales usually do not include evidence that might contradict their findings; the data is instead coded and categorized to support the ethnographer’s conclusions.

As we are entering the stages of analysis and writing in the Symbolism project, it seems particularly important to reflect on these characteristics that Van Maanen has identified. Guiding post-positivism’s data-collection, analysis, and writing strategies, each characteristic is rooted in its notion of an apprehendable but not completely knowable reality. In addition to its pragmatism and its goal of richly descriptive narratives, there is the “typicality” of the research participant, the accurate rendering of the “natives’” view in the data, the role of the objective researcher in the analysis presented, and the representation, finally, of a singular “truth” that is taken as a reflection of an objectively knowable reality. As we will see, however, assumptions underlying these characteristics have come under criticism as a result of the emergence of a constructivist approach to qualitative media research; our own research parallels a trend occurring much more broadly today in media studies. The “typical” respondent

Frustrated with the limitations of standardized survey instruments, post-positivists aim toward a better understanding of the meanings people assign to their responses to a researcher’s questions. In the oft-cited turn of phrase, they sacrifice breadth for depth, choosing to limit the sample size of their inquiry while continually adjusting a relatively standardized interview guide so as to best capture similarities of responses.

Of course, this approach is not without its problems; seeking to meet the quantitative researcher’s standards, the post-positivist is left in the untenable position of having to argue that his or her set of respondents is “typical” and that they therefore serve as a basis for generalization to the larger population (or a subgroup of the population). This is usually done by comparing the demographics of the research participants with those of the broader public. Of course, there is no “scientific” method by which to determine how many people are needed for a generalizable sample. Does sixty sound adequate? What about thirty? What about twelve? Further, how might we boil down the various combinations of race, gender, age, class, education, religious affiliation, or geographic locale so that we might claim that the collection of people we have interviewed comprises a “representative” sample of the population at large? Is one person over the age of fifty enough? What if that person is a wealthy Asian female -- can her experiences be generalized? What, or who, might her comments be “typical” of? The problem is only intensified when one attempts to make in-group comparisons. To make claims that a particular response is “typical” of the group sampled, how many need to voice it - a majority? More than half (which, in a sample of twelve, is only six people)? More than one? If the wealthy Asian American woman and the 15-year-old African American male agree on a certain point, might we then argue that such responses are “typical” of others, as well?

These questions and others beg a reexamination of the notion of the “typical” research participant. The model itself borrows from the work of Talcott Parsons, who at the middle of the century assumed that the empirical world revealed fundamental properties that could be isolated, classified, and studied. In order to study a society, Parsons believed it was most useful to assume that society is stable so that its various functions -- in other words, what worked to hold it together -- might be carefully and independently analyzed. In this model, one might imagine that a “core” of society would include an aggregate of persons with similar attributes and similar affiliations who might act in similar ways, thus contributing to a stable society. The further one was from that core, the less “typical” of the whole he or she was.

In the decades following Parsons’ writings, media studies, along with a host of other fields competing for entrance into the respected realm of scientific inquiry, strove to perfect the description of the role of media in society and in particular how it functioned in a stable society. We tested media use and its effects across the variables of race, gender, age, and other demographic identifiers. Yet with the rise of increased awareness of difference along these lines, media studies, like other fields, began to call into question the assumptions of generalization. How could we generalize across the entire population when members of minority groups comprised such a small number of the sample, we asked? Thus, we began isolating the variables, to use quantitative terms our study would only look at one racial/ethnic group, or one age group, for example, so that we would limit our claims of typicality by particular demographic identifiers. Still there were problems, however, not the least of which was the subtle reinstituting stereotypes through the assignment of an assumed “core” to particular subgroups. Another practical problem involved the writing up of the data aiming for the “typical” responses meant eliminating the unusual and often most rich and illustrative stories, making the potential strength of qualitative inquiry its most serious casualty.

As these problems were encountered, media research begged for a re-centering for this conundrum. We can trace some of the problems back to the disconnect between theoretical goals and the methods used to explore them in our foundational paradigm of symbolic interactionism. This approach, as stated earlier, aspired to address itself to social concerns while focusing on the interaction between the researcher and the research participant. While the goal of this research was to explore the role and possible solution to social problems, therefore, the method leant itself to an exploration of the meaning systems of individuals. In order for us to return the focus to the social rather than individual level, we can, as noted above, look to cultural studies as it was imported from Great Britain. In our own study, our purpose in constructivist research was then be reframed in this way we are not interested in generalizing our findings across people; instead, we are interested in generalizing to the culture.

What will this shift mean? I believe that it enables us to move away from a position in which we seek to explore how individuals are relatively close or distant from an imagined “core” of society. We can approach each person as a “universal singular,” to use Sartre’s term. Each person’s story becomes important, for it is simultaneously the story of a unique individual and the embodiment of the social world that has produced her or him. Furthermore, in this move we acknowledge an important learning from postcolonial critiques that often it is people who are on the margins of societal groups who are most informative of the boundaries that help to construct what we come to accept as the “typical” or “core” of society. Thus the very notion of the “typical” is problematized and implicitly recognized as a historically created category. Our analysis can then move from needing to claim how “typical” one is to identifying how the informant sees him or herself in relation to some received notion of “typicality.” It also, of course, forces us as researchers to interrogate our own received ideas of “typicality,” as well.


Telling the native’s story

Symbolic interactionism was not the only influence on the current post-positivist approaches to qualitative media studies, of course. With the rise of interdisciplinary inquiry in the 1970s and 80s, media studies increasingly looked to cultural anthropology for methodological clues. Borrowing the term and spirit behind ethnography, media scholars strove to move away from standardization out of a desire to better capture the stories of everyday life. They embraced the tradition of Bronislaw Malinowski, who wrote The goal is, briefly, to grasp the native’s point of view, his relation to life, to realise his vision of his world...To study the institutions, customs, and codes or to study the behaviour and mentality without the subjective desire of feeling by what these people live, of realising the substance of their happiness -- is, in my opinion, to miss the greatest reward which we can hope to obtain from the study of man.

The humanistic and romantic impulse of early 20th century ethnography is evident in Malinowski’s aim for “the subjective desire of feeling” and his desire that ethnography would be about “realising the substance of their happiness,” and indeed this impulse remains largely in tact in contemporary qualitative media research. In her germinal study of women reading romance novels, for example, Radway was particularly interested in “taking the women’s self-understanding seriously,” which she argues was a missing element in the feminist critiques of literary texts up until that point. Lull, another early qualitative media researcher, rather directly pits the benefits of ethnography’s humanism against the presumed anti-humanism of critical (i.e., mass society) approaches Giving the audience a ‘voice’ has produced tangible, positive results. It has tempered the often pretentious and opaque writing associated with cultural studies, the frequent imposition of privileged interpretations of texts, a reliance on fashionable literature, and the dogmatic refusal to cite ‘scientific’ studies. Still, there are problems in the development of audience research in cultural studies. Most troubling is that we often hear too clearly the theorist’s voice presented as if it were the audience’s voice.

Lull, Radway and others have seemed to celebrate “giving voice” to research participants, thus validating the subjective experiences of media use among those studied. This humanism, of course, has been disparaged by those critical theorists convinced that by highlighting subjective experience, we lose our focus on structural constraints. Yet it has also been increasingly denounced by researchers who embrace ethnographic methods, particularly as a result of feminist critiques. Feminists, while wishing to “give voice” to women’s experience, have also noted the need to place analyses of subjective experience within larger frameworks of social constraint (usually those determined by gender). The act of “giving voice” does not necessitate the removal of interpretation, these critics argue; indeed, it is all too easy to cloak one’s own views in the garments of the research participants’ voices.

A constructivist approach retains the political impulse that lies behind the desire to “give voice,” begging an acknowledgment of the history that has made some persons “voiceless” to begin with. It calls into question the notion that the experience of any one group might be assumed to be of universal significance. In writing the findings of inquiry, therefore, it is important to be aware of the constraints under which research participants tell their stories, for this shapes both what is told and how I must contextualize the telling for a larger audience. We must not forget that, as communication scholars Natrajan and Parameswaran point out, the researcher ultimately exercises power simply in the writing of the qualitative research narrative itself.


The role of the researcher

Post-positivism assumes, as noted above, that our best findings will result as we make a close approximation to what would happen in terms of media consumption were a researcher not there to observe these practices. Thus to some extent the researcher is assumed to be interchangeable; someone else entering the field would come to similar conclusions. It is not surprising, then, that those commited to the post-positivist approach are rather resistant to the arguments for reflexivity on the part of the researcher. Expressing the common concern that self-reflexive researchers who are themselves acknowledged fans may not maintain the distance required to “develop any kind of criticism,” Moores, for instance, argues that such researchers risk a “populist acceptance or celebration” of the audience practices they are observing. Yet a critic of this perspective would ask, does “distance” between oneself as a researcher and the practices one observes truly guarantee an objective position from which such criticism can be made? Reseachers of such common practices as media consumption should be particularly wary of this claim, for who among us is not a television or film viewer? And by extension, then, how do we know when we have enough distance from our own experiences and practices to warrant a sufficiently critical interpretation? In our own project, we took a middle course on this issue we sought out as our informants those with whom we had had no previous connection, yet paid attention to the particularities of the relationships that formed between the researcher and those who participated.

Because sense-making on the part of the researcher must always rely upon the knowledge we bring to the research context from either reading or personal experience, a constructivist perspective argues that our awareness of our relation to that which we are observing and analyzing is particularly important. There is a danger that the “story” emerging in constructivist research becomes the story of the researcher’s self-discovery, yet this is not the only alternative to emerge from this paradigm, nor is it the most common. Instead, while the researcher’s acknowledgment of the context and conditions within which the dialogues of research take place may lead to new insights, whether and how they are written into the final writings depend upon other factors, as will be discussed below.

First, I want to extend the argument for a self-reflexive project a bit further in the particular case of our research, which has foregrounded issues of identity. From the writings on representation, critical theory and post-structuralism we learn that identity is never a “fixed” category (or sum of categories) that may be unproblematically interrogated but rather it exists as it is enacted. In the case of qualitative research, identity is enacted in the behaviors observed and the narratives constructed for the audience of the researcher. Constructivists do not approach the research participant’s statements as “raw data” that might be unproblematically analyzed so as to get us close to “the truth,” but as conditional statements made in a particular context and which of necessity require interpretation. Those behaviors and narratives emerging in the research context are contingent upon how the research participant “reads” the situation, the researcher’s intent, and the researcher’s position vis a vis both the subject and the larger culture. In a very important sense, then, any analysis of identity-construction cannot be separated from the specificity of the researcher/researched relationship. The more insight afforded into the factors shaping that situation, the more likely we are to understand the knowledge produced. This does not require an academic’s “life history,” but it does underscore the fact that all knowledge is contingent, that all experience is interpreted from a subject position, and that the specific knowledge constructed in the interview process is, as Visweswaran argues, "itself determined by the relationship of the knower to the known." Reflexivity, therefore, subverts normative understandings of knowledge-gathering and information-processing; as Babcock argues, By confounding subject and object, seer, and seen, self and other, art and life - in short, by playing back and forth across terminal and categorical boundaries and playing with the very nature of human understanding - reflexive processes redirect thoughtful attention to the faulty or limited structures of thought, language, and society."

A constructivist approach that encourages reflexivity on the part of the researcher therefore challenges the "givenness" of data interpretation and the ontological and epistemological assumptions undergirding positivist research.


Telling the “truth”

Assumptions of knowledge and truth, of course, are fundamental to the differences discussed here between post-positivist and constructivist research. Understanding that the goal of reflexivity is to provide the reader with better tools by which to evaluate the “truthfulness” of the presentation, Moores, a post-positivist, claims that reflexivity is not needed due to the “active critical capacities of readers” of audience ethnographies. Yet while critiquing any researcher’s report is a widely-accepted professional practice, this in itself does not address the basic ontological challenge of a constructivist approach. The constructivist critique necessitating reflexivity does not concern the merely the “accuracy” of the particular reported research, as Moores seems to assume, but of the status of that knowledge itself.

At this point, post-positivists are to be excused if they wish to throw up their hands and sigh, “if all is relative and we can’t say anything definitive, then why do the research to begin with?” Yet constructivists believe that this pessimism is unwarranted. The point of research in our pragmatist American tradition, a constructivist would argue, is not to reveal “the truth,” after all, but to provide insight that can possibly change understandings in emancipatory ways. While the goals might be shared, the constructivist’s approach to analysis and writing is quite different than that of the post-positivist. Rather than seeking to establish objectively-defined categories that “emerge” from the “raw data,” the constructivist evaluates what has been learned from research participants in light of other insights he or she brings to the field. The constructivist realizes that the “truth” that emerges is indeed a “truth” for him or her; the challenge, then, becomes how to present this viewpoint in a way that is enlightening and convincing to others in other words, one must establish “ethnographic authority,” to use the words of anthropologist James Clifford. Erasing the researcher from the writing as post-positivists do assumes rather than establishes such authority; the absent writer stands behind the veil of “objectivity” and scientific inquiry. Ang makes a similar critique The comfortable assumption that it is the reliability and accuracy of the methodologies being used that will ascertain the validity of the outcomes of research, thereby reducing the researcher’s responsibility to a technical matter, is rejected...the empirical is not the privileged domain where the answers should be sought. Answers - partial ones, to be sure, that is, both provisional and committed - are to be constructed, in the form of interpretation.

I would go further than Ang, noting that the interpretation itself might take the form of an argument that establishes the researcher’s authority through the presentation of a credible interpretation.

This is our challenge at this point in our research project we are working to refine an organization of our data into an argument. The argument’s form turns out to be quite different from that which we expected at the project’s outset due to our move from post-positivism to constructivism. Still, it retains an interest in providing a richer analysis of the intersections between religion and the media in the context of people’s home lives. I believe that the change in our research paradigm has enabled us to reframe our analytic goals, so that now rather than focusing only on “meanings made,” we are able to see those meanings within larger contexts of what we call the “public scripts,” or discourses, of the media. We then can situate the conversations of our research within broader frameworks of “dominant memory,” or those important resources for conversations that are drawn from public and popular culture. These “dominant memories,” we argue, serve as important resources in the construction of private memory. Oral accounts given to us in our research are therefore less unique than we might wish in a more positivist framework. Yet acknowledging their contingency on other cultural narratives is, I believe an important step toward what a constructivist might define as “theoretical rigor.”


Conclusion

Critiques of post-positivism are nothing new to our field. Unfortunately, as Bird notes, media studies has recently devoted much more time and space to such critiques than we have to actual ethnographic investigation. Murphy illustrates the problem by noting that in a 1989 volume of the Journal of Communication Inquiry titled “Cultural studies, ethnography,” each article is concerned with the politics of ethnography, yet not one actually illustrates the practice! From these conundrums, one might assume that ethnographic research is impossible or unfruitful in media studies. Yet I agree with the anthropologist Ortner, who argues It is our capacity, largely developed in fieldwork, to take the persepctive of the folks on the shore, that allows us to learn anything at all - even in our own culture - beyond what we already know. Further, it is our location 'on the ground' (i.e., studying everyday lives) that puts us in a position to see people not simply as passive reactors to and enactors of some 'system,' but as active agents and subjects in their own history.

A constructivist perspective, I propose, is able to explore these agentive practices as they are articulated discursively and materially, for as Martin-Barbero has argued, it is in this way that we may truly learn of the relationship between ideology and experience.

In the Symbolism, Media and the Lifecourse project, we therefore continue our efforts in analysis and writing of the research. Of course, if knowledge is something that must be produced, as I have argued within a constructivist paradigm, it makes sense that it evolves gradually first within the context of interviews, then as our entire group mulls things over in our weekly analysis meetings, and then finally as we corral our conversations into a particular form for writing. In this final stage, our writings become an argument rather than a set of findings; an argument that must both establish our ethnographic authority and present our explanations for why we believe the world is a certain way. It is our hope that we can do this latter task in a way that is convincing and, perhaps, enlightening.


BACK TO MAIN ANALYSIS PAGE