OUTSIDER AND INSIDER TRADITIONS OF QUALITATIVE INTERPRETATION

OUTSIDER AND INSIDER TRADITIONS OF QUALITATIVE INTERPRETATION

Isabelle Thompson
Auburn University, AL 36849

thompis@mail.auburn.edu

When bystanders observe the same traffic accident, their individual testimonies about what happened and who was at fault may disagree. When two friends read the same novel, their respective retellings may emphasize different events and reflect different impressions of the characters. These differences in interpretation appear in qualitative research as well. Two participant-observers may tell different stories of the same culture. What we as qualitative researchers report from an investigation reflects what we expect to find, what we look for, and how we perceive our role. In other words, our perspective on our research shapes our interpretation and hence what we report as knowledge.

My purpose is to discuss two opposing traditions of qualitative interpretation. One, the outsider perspective, privileges the analysis of data from within the framework of the researcherís existing knowledge and culture. The outsider perspective aims to understand a foreign culture according to the outsiderís own, usually more dominant, cultural language and conventions. The second, the insider perspective, privileges the culture observed. The insider perspective aims to represent the lived experience of the indigenous people through establishing a dialogue rather than analyzing a code. The choice between outsider and insider perspectivesóreferred to by sociolinguists as the "etic" and "emic" distinctionódetermines and represents to some extent the location of authority in a particular investigation. After showing how the outsider and insider traditions lead to quite different interpretations of qualitative research and hence different views of knowledge, I will conclude by drawing some parallels between the reading theory of Louise Rosenblatt and the outsider and insider traditions of qualitative interpretation. Rosenblattís theory of reading as the interpretation of texts from the stance of the reader can provide a framework for situating diverse interpretations of qualitative research.

Outsider Interpretation

Outsider interpretation requires the researcher to willingly take on the authoritarian role in the investigation. She will describe the culture being observed according to the ideology of her own culture and, most importantly, according to the conventions of the scientific community. By using a common code to describe a variety of different cultures, researchers hope to make cross cultural comparisons.

The outsider perspective is parallel to the "etic" distinction in sociolinguistics defined originally by Kenneth Pike (See Headland, Pike, and Harris). Applying the etic distinction to qualitative research in second-language acquisition, Karen Ann Watson-Gegeo says it is "based on the use of frameworks, concepts, and categories from the analytic language of the social sciences" (579). (See Ramanathan and Atkinson for further discussion of the etic/emic distinction as it applies to qualitative research in second-language writing.) Well-known etic interpretations include the many anthropological investigations of kinship terms across cultures. Qualitative researchers in information science, Alan R. Sandstrom and Pamela Effreins Sandstrom strongly favor etic interpretation because it is more scientific, with methods and concepts deriving from the well-established community of empirical researchers.

Matthew B. Miles and A. M. Huberman are perhaps the best known qualitative methodologists to advocate the outsider approach to interpretation. For them, analysis and interpretation are related to coding:

Coding is analysis. To review a set of field notes . . . and to dissect them meaningfully, while keeping the relations between the parts intact, is the stuff of analysis. This part of analysis involves how you differentiate and combine the data you have retrieved and the reflections you make about this information . . . . (56) Miles and Huberman later refer to codes as "efficient data-labeling and data-retrieval devices to empower and speed up analysis" (65). They assume an outsider perspective in part because their analysis is concerned with uncovering the reality represented by the textual information coded. In other words, the field notes, interviews and other verbal data that Miles and Huberman collect are representative of the culture they are investigating. These texts do not constitute the culture itself but instead reflect reality by conveying a meaning separate from language. According to Miles and Huberman, "[f]or our purposes, it is not the words themselves but their meaning that matters" (Miles and Hubermanís italics, 56). As scientific observers, Miles and Huberman are able to maintain objectivity through distancing themselves from the culture they are describing.

To illustrate outsider interpretation more fully, I searched through the past five years of journals in technical communication and composition to identify an article most clearly representing this analytical perspective. I choose "Writing Together: Genderís Effect on Collaboration" by Louise Rehling published in Journal of Technical Writing and Communication (1996). After a brief review of research and personal teaching experiences justifying her study, Rehling states her purpose and overviews her methods as follows:

The gender differences identified by scholars, along with the cultural stereotypes, obviously might affect how women and men work together in classrooms such as mine. Therefore, I attempted to apply current theories to my students. Using the methods described below, I quantified the experiences of almost 300 students in over sixty groups, reviewing and then coding their collaborative experiences in terms of the assumptions about gender-balanced groups derived from recent studies. I also looked for other gender-related collaboration issues which might apply to my classroom groups, then coded those observations categories as well. (164) The more than 60 groups of students whose experiences Rehling "quantified" were enrolled in five sections of professional writing courses that she taught over a two-year period. Based on the sex of the individual members, Rehling identifies three types of groups: single gender, all the same gender except for one individual, and gender balanced. Data collection consists of Rehlingís observation notes, her record of student comments, her interviews with students, group drafts and final products, studentsí anonymous evaluations of the group work. The classroom-based research project clearly employs qualitative data collection techniques. The fact that Rehling is collecting data from her own students makes her connections with the culture observed even closer than if she had assumed the role of participant-observer in another teacherís classroom. Rehling has insider knowledge of this community.

In establishing her coding system and making her interpretation, however, Rehling attempts to assume an outsider perspective. She derives four of her six coding categories from reviewing recent literature on gender and collaboration. Those categories are stated as assumptions:

Gender-balanced groups are less likely to stereotype women as clerks.
Gender-balanced groups are less likely to use hierarchical collaboration.
Gender-balanced groups are less likely to focus on results over process.
Gender-balanced groups are more likely to experience counter-productive dissent.

The other coding categories are derived inductively, from Rehlingís observations of her students:

Mixed-gender groups are more likely to stereotype men as technical specialists than to stereotype women as clerks.
Gender-balanced groups tended to segregate themselves into working partnerships split down gender lines.

In coding her data, Rehling labels each of the more than 60 student groups as confirming or disconfirming each assumption and observation. Then she conducts a chi-square analysis of the coding tallies to determine if the gender-balanced groups differ significantly in their behavior and attitudes from the single-gender groups or the all-the-same-gender groups. As she says, Rehling finds no significant differences: In sum, gender-balanced groups were not more likely to stereotype women as clerks, to use hierarchical collaboration, to focus on results over process, or to experience counter-productive dissent. (166) Rehling believes that coding categories she developed inductively based on observation are more telling. But it is unclear that they yield significant differences. It is also unclear when and how these categories are developedóa priori based on Rehlingís previous teaching experience or ad hoc based on her review of her data, as Robert C. Bogden and Sari Knopp Biklin suggest. Rehlingís coding process, with its primarily deductive direction, resembles quantitative content analysis (Lauer and Asher 26).

Overall, as Harriet Bjerrum Nielsen points out in her discussion of Miles and Hubermanís scientific tradition in qualitative research, Rehlingís study has many similarities to quantitative research. Using a detached scientific viewpoint, Rehling provides a naturalistic description of a reality represented in the textual data she has collected. Rehlingís outsider perspective requires that she understand a different culture by comparing it with a well-known one. This point of view incorporates the observed into what is already known. Hence, the outsider perspective is by its nature conservative.

It is appropriate that Rehling also projects a conservative ideology in her views of feminism and in her educational philosophy. She ends her short conclusion with advice for instructors of professional writing. First, she suggests that "instructors eschew borrowed commonplaces" (174) such as the four assumptions derived from the research reviewed. Second, in a statement not supported by her findings, she says "instructors can help students watch for gendered differences in unexpected areas of behavior, then control for these differences to better prepare for the transition to future workplaces" (174). Defining instructional goals in terms of the existing ideology which privileges the masculine, Rehling believes that we should prepare students to participate in the workplace according to the current status quo. Hence, her views are conservative and by effect if not by intention hamper societal change because they protect and conserve the dominant ideology. Rehling does not presume that the classroom should become a force to bring about change. Like her research goal to understand her classroom culture, Rehlingís educational goals are functional: to teach students to survive and even triumph in culture as it is. Like her view of language as a representation of reality, her conception of the role of education is also naturalistic: schools represent society. Reflection of the current status quo and incorporation of the dominant ideology are outcomes of interpretation from an outsider perspective. The outsider perspective encourages researchers to describe what is done without considering at the same time how to bring about social change.

Insider Interpretation

Insider interpretation requires the researcher to abdicate his authoritarian role in an investigation. He attempts to describe a culture according to the ideology underlying the codes and conventions of the indigenous participants. Instead of cross-cultural comparison, the researcher wants to provide lived experience of the culture being investigated (Van Maanen). Hence, the insider perspective derives primarily from dialogue rather than analysis. The approach is more humanistic than scientific.

The insider perspective is parallel to the emic distinction in sociolinguistics. According to Watson-Gegeo, "emic refers to culturally based perspectives, interpretations, and categories used by members of the group under study to conceptualize and encode knowledge and to guide their own behavior" (580). Emic interpretation is roughly equivalent to Clifford Geertzís notion of thick description. Knowledge is to a large extent constructed through language. Language is not only representative of reality; it constitutes reality. Because they perceive a culture in terms of its texts, researchers favoring insider perspectives use interpretative approaches common in the humanities, for example, hermeneutics and phenomenology. (See Diesling for a discussion of hermeneutics as a method of knowledge-building in social science research.) In his discussion of philosophical hermeneutics, Timothy W. Crusius summarizes the insiderís view of language. He says that it is not possible to have "unmediated" access to the world: "[r]eality is not only language; reality happens through language" (26). For researchers interpreting their investigations from an insider perspective, therefore, knowledge is dialogic (meaning-making) rather than analytical (meaning-capturing).

To illustrate insider interpretation more fully, I will discuss an article "Exploring Agency in Classroom Discourse: or, Should David Have Told His Story?" by Helen Rothschild Ewald and David E. Wallace, published in College Composition and Communication (1994). In the first few paragraphs of their article, Ewald and Wallace discuss their purpose as investigating the merits of teacher-centered vs. student-centered pedagogies and overview their study:

In this article, we do not intend to declare a winner [in the teacher-centered vs. student-centered debate]. Rather, we intend to problematize the notion of agency in the writing classroom. We believe that both teacher and student are constructed agents in the classroom. . . .

Our discussion of agency in classroom discourse will focus on an excerpt from a first-year college writing class taught by David Wallace and four students. Our approach can thus be termed "hermeneutic" . . . . (343)

As Ewald and Wallace say, the episode discussed in this study occurred in a freshman writing class taught by Wallace, the second author of the article, and Wallaceís behavior is the primary focus of the interpretation. During the episode, David (David Wallace refers to himself as "David" in the article) intervenes in a discussion of affirmative action between a white male and a black male by telling the story of two women who were given hiring priority when he interviewed for academic jobs. David claims that the women received this priority not because they were more qualified than their male competitors but because of affirmative action. Two female students also participate in the discussion. After Ewald and Wallace recount the episode, they spend the rest of the article interpreting how Davidís telling of his story affects (or may affect) discourse in his classroom.

The interpretation begins with an initial analysis based on what Ewald and Wallace call "prior theory," a review of research about classroom discourse. They characterize Davidís behavior according to current sociolinguistic research and find that he violated (deliberately, he says) advice about how teachers should direct discussion in a classroom. After this analysis, Ewald and Wallace provide "alternative" interpretations of the episode by interviewing each of the five participants. Ewald and Wallace call this approach "hermeneutic" and define its purpose as follows:

Using a hermeneutic approach commits us to examining the underpinning values, discovering consequences of these values, and looking to the reciprocity achieved between teacher and student, andófor that matteróbetween student and student, not by coding and counting extant patterns in the discourse, but by asking who is making meaning, what meanings are being made, and how meanings are validated. (348) In these alternative interpretations, Ewald and Wallace examine what they call the participantsí "passing theory," "on-the-spot interpretation or ëhermeneutic guessingí regarding the meaning of an utterance" (345). No one participantís "passing theory" is privileged. The effect of the alternative interpretations is a layering of a single lived-through moment, a dialogic meaning-making experience for the reader of the study. Ewald and Wallace say that both "passing theory" and "prior theory" are involved in hermeneutic interpretation.

After each alternative is given, Ewald and Wallace discuss the ideology of Davidís behavior. They suggest that in spite of a teacherís authority, students cannot be denied the power to interpret. They further argue that students have agency in the classroom because of "their ability to appropriate stances and make ëresistant readingsí of the class as text" (361). However, the researchers also state that most students are unaware of the ideology that lies behind their resistance. They argue that teaching students to become ideologically aware should become an important educational goal.

Unlike Rehling, Ewald and Wallace are more concerned with bringing about social change than with accurately describing a cultural reality. In the classroom, change would entail teaching that knowledge is symbolic rather than personal, a current premise of student-centered pedagogy. In society, change would derive from an awareness and resistance to the ideology of the status quo. As Ewald and Wallace acknowledge, their educational philosophy is influenced by Paulo Freire and other radical educational theorists.

Insider interpretations, like Ewald and Wallaceís study, are readable and entertaining. However, are they substantive? Do they truly represent the indigenous point of view? Nielsen makes two important criticisms applicable to insider interpretations. First, she claims that to avoid assuming authority over the description some researchers also avoid interpretation and present fragments with no encompassing synthesis. Without a synthesis or interpretation, the research has "no message to convey (so why write at all?)" (10). Second, Nielsen points out that authority is unavoidable. Dialogic texts are as socially constructed as texts reporting scientific studies (10). In "Exploring Agency in Classroom Discourse," by basing their interpretation on "depth hermeneutics," Ewald and Wallace operate from presumptions as strong and clear as those of the scientific method. Paul Diesling defines depth hermeneutics as the "hermeneutics of suspicion" [Dieslingís italics]:

This consists of searching for a covert message that is concealed but also hinted at by the overt message. The overt message is not false; it is the appearance or idealized interpretation of the practice, its outward or public aspect. The hidden message or messages are not the whole message, but only the deeper layers of it. The whole message includes its outward appearance or public face. The evidence for an interpretation includes not only showing how the parts of a message or practice cohere as an intelligible sequence; it also includes showing how the outward layers of meaning conceal but also express the deeper levels. (126) Crusius refers to depth hermeneutics as "negative hermeneutics" and calls it "a continuation of the Enlightenmentís effort to liberate us from the dogma, error, and superstition of the past" (6). Hence, the insider perspective Ewald and Wallace used influenced their interpretation of the classroom as ideological. Whereas Rehling based her analysis on the view that language is a mirror of reality, Ewald and Wallace begin their interpretation with the assumption that language constitutes reality and that researchers are obligated to look beneath this reality at its underlying ideology. Understanding the ideology of the dominant culture should encourage informed resistance. A primary goal of insider interpretation from its inception, therefore, is rapid social change.

Situating Qualitative Interpretation

The purpose of this discussion about traditions of qualitative interpretation has been to make some distinctions and to draw some boundaries. The intention is to use these boundaries to help us become more aware of the assumptions we bring as researchers to qualitative data and to enhance our critical competency as readers or consumers of qualitative research studies. How then can we use this discussion? I think that we can use it to build a broad interpretative framework to situate individual studies.

Such a broad framework for locating interpretation might be at least as usefulóif not more soóas the well-known etic and emic distinction discussed previously. To conclude this article, I will sketch a broad framework for situating qualitative interpretation based on the reading theory of Louise Rosenblatt. According to Rosenblatt, interpretation can be located on a continuum defined by the two most extreme "stances"ósets of expectations and goalsóthat readers can bring to a particular text.

Like most current reading theoristsówhether postmodern (McCormick) or cognitive (Lorch and OíBrien)óRosenblatt assumes that reading is an interpretative act, directed by the readerís current knowledge, attitudes, and reasons for reading. In her scholarship, she intends to describe the nature of that act. I believe that qualitative researchers are also "readers" of texts and that, like readers of novels and maintenance manuals, they bring assumptions to their interpretation of the qualitative data and the stories they collect. Like all readers, researchers define and bring away knowledge from their interpretations based to a large extent on what they start with. Reading a qualitative text from an outsider perspective brings attention to and results in one view of knowledge, while an insider perspective brings attention to and results in another. Rosenblatt discusses the stances that readers bring to texts and their subsequent interpretations on a continuum defined at one end by the efferent stance and the other by the aesthetic stance. The efferent stance is equivalent to the outsider interpretative tradition; the aesthetic stance is equivalent to the insider interpretative tradition.

Analogous to outsider interpretation, the efferent stance encourages a reader to pay attention to facts and information in the text. According to Rosenblatt, in efferent reading:

the readerís attention is focused primarily on what will remain as residue after the readingóthe information to be acquired, the logical solution to a problem, the actions to be carried out. (23) The analogy between efferent reading and outsider interpretation becomes clearer as Rosenblatt continues her discussion: As the reader responds to the printed words of symbols, his attention is directed outward, so to speak, towards concepts to be retained, ideas to be tested, actions to be performed after the reading. (24). Although Rosenblatt does not assume that language is a transparent medium used to convey meaning, she does suggest that readers can interpret texts from that viewpoint. Like outsider interpretation, efferent reading intends to uncover facts and enhance understanding of new concepts and cultures. Knowledge comes from this understanding as newly discovered ideas are connected with what a reader already knows and currently holds in his memory. Hence, knowledge-building is a slow and conservative process.

Analogous to insider interpretation, the aesthetic stance enables the reader to live through the moment, making new meaning in this process. According to Rosenblatt, in aesthetic reading, "the readerís primary concern is with what happens during the actual event" (24). The goal is not only to understand but to experience the world. Knowledge is built not from facts carried away but from the personal experience of sharing anotherís life. It is defined attitudinally and affectively rather than cognitively. Thick description of an indigenous culture provides the lived through moment for insider interpretation in much the same way that a poem or short story solicits an aesthetic response.

Hence, a framework for situating qualitative interpretation might consist of a continuum similar to the continuum of reading stances proposed by Rosenblatt. One end is outsider interpretationóthe efferent stanceóand the other is insider interpretationóthe aesthetic stance. As discussed previously, the two ends might be characterized as follows:

Outsider/Efferent Stance Insider/Aesthetic Stance

objective subjective

scientific humanistic

naturalistic textual

knowledge as understanding knowledge as meaning-making

analytic dialogic

Outsider and insider interpretations derive from different world viewsóstancesóand consequently search for and find different kinds of knowledge.

However, in the same way that an expert reader is likely to use both stances in interpreting a text, researchers are also likely to incorporate aspects of both traditions in their reading of qualitative research. Hence, it is important that the two ends be seen as idealsópure interpretations that never really come aboutóand that qualitative studies be located on a continuum between the two ends. For example, even as she tries to disengage herself from her community membership in analyzing her collected data, Rehling uses her insider knowledge to identify two new categories for coding. Ewald and Wallace begin their discussion of their episode with an analysis based on an outsider comparison with other classroom discourse research and throughout their interpretation, they bring in other outside authorities for comparison. Therefore, if outsider interpretation constitutes the leftmost end of the continuum, Rehlingís study would be located on that side slightly to the right of the end, and if insider interpretation constitutes the rightmost end, Ewald and Wallaceís study would be located on that side slightly to the left of the end.

My purpose has been to describe and illustrate some differences between the outsider and insider traditions of qualitative research. My intention is to use this discussion of differences to define the ends of a continuum and to use that continuum as a framework for situating qualitative research studies.

Works Cited

Bogdan, Robert C. and Sari Knopp Biklen. (1992). Qualitative research for education: An introduction to theory and methods. 2d ed. Boston: Allyn and Bacon.

Crusius, Timothy W. (1991). A teacherís introduction to philosophical hermeneutics. Urbana: NCTE.

Diesling, Paul. (1991). How does social science work? Reflections on practice. Pittsburgh: U of Pittsburgh.

Ewald, Helen Rothschild and David E. Wallace. (1994). Exploring agency in classroom discourse: Or, should David have told his story? College Composition and Communication 45, 342-68.

Freire, Paulo. (1973). Education for critical consciousness. New York: Seabury.

Geertz, Clifford. (1983). Local knowledge: Further essays in interpretative anthropology. New York: Basic.

Headland, T.N., K.N. Pike, & M. Harris (Eds.) (1991). Emics and Eeics: The insider/outsider debate. Newbury Park, CA: Sage.