Wednesday, February 2, 2011

Conversation analysis

On the readings

Atkinson, J. M., & Heritage, J. (Eds.) (1984). Structures of social action: Studies in conversation analysis. Cambridge: Cambridge University Press.

Heritage, J. (1984). Garfinkel and ethnomethodology. Cambridge: Polity Press.

Psathas, G. (1995). Conversation analysis: The study of talk-in-interaction. Qualitative Research Methods Series. Thousand Oaks, CA: Sage Publications.

These are the textbooks for my conversation analysis class. It's always fascinating and scary to delve into a new field. I have read several second language acquisition studies using conversation analysis and tried doing transcriptions myself, just enough to realize how difficult and time consuming it is, but I have never read on the philosophy behind the field. So, it was nice reading on the underpinnings of the field. I have a tendency toward theory and philosophy so unlike some of my classmates I actually enjoyed the readings. Admittedly I do not know much about conversation analysis and I have to say I haven't been converted to it just yet. I cannot help it but I'm inclined to ethnography rather than conversation analysis. Don't get me wrong, I acknowledge the power of the method but still I'd like to use it in addition to other qualitative methods. Perhaps it is a matter of personal taste. I don't know.
Anyway, here is what I understand from the readings so far.

In light of the readings, as I understand it, ethnomethodology is the study of social action.  More specifically researchers who are involved in ethnomethodology are interested in the systematic study of daily interactions. According to Heritage, for Garfinkel—the founder of the term,—ethnomethodology "refers to the study of a particular subject matter: the body of common-sense knowledge and the range of procedures and considerations by means of which the ordinary members of society make sense of, find their way about in, and act on the circumstances in which they find themselves." (p. 4). In other words, ethnomethodology is interested in the meaning making procedures of people as they are engaged in daily interactions.
I see three differences between the two readings. First, perhaps due to different foci the two books have, Psathas gives a broader but briefer historical account whereas Heritage provides a more detailed historical account focusing more on what was going on around Garfinkel when he was a graduate student. Another difference between the two books is related to the conflicts presented in the history of the field. Psathas’s chapter shows a trend in interaction analysis from a more structuralist and analytical approach to a more ethnographic and naturalistic one. The section entitled ‘The critique of category systems’ illustrates the contrast between analytical and ethnographic approaches to the study of social interactions. Overall, I think it is safe to claim that when we look to category systems approach through the lens of today it is seen reductionist since it ignores (1) actions that do not fit into categories, (2) phenomena that cannot easily be quantified, and (3) the role of local context in meaning making. While Psathas seems to focus on a tension between analytical and ethnographic approaches—Bales and Goffman seem to lie on the two ends of the continuum—Heritage sees a different tension defining the beginnings of ethnomethodology, one that focuses on the role to be given to the accounts of the actors.  At least it is my understanding that the tension in Heritage’s book revolves around subjectivity and objectivity, and to some extent, empirical vs. theoretical approaches.   
The third difference between Psathas and Heritage’s chapters is that Heritage tells the history of ethnomethodology whereas Psathas focuses on conversation analysis. I’m not quite sure at this point what the relationship between conversation analysis and ethnomethodology is and to what extent they are similar or different. It is my understanding that ethnomethodology evolved into conversation analysis and conversation analysis has stricter commitments to some basic assumptions, which I try to summarize below as I understand them.
The basic assumption of conversation analysis is that talk-in-interaction is highly organized and systematic. Thus, the main goal of conversation analysis is to discover, describe, and analyze interactions in order to find the mechanism that guides or leads to this organization. Since this machinery is what conversation analysts are interested in, the contexts they choose to study is trivial compared to the mechanism they seek to uncover. As a result of this focus on the machinery of interaction, the particularities of the people or institutions conversation analysts investigate do not matter because elements of that machinery, such as adjacency pairs and turn taking, are at work wherever there is social interaction.


The goal of conversation analysis is to investigate the systematicity in interaction. In order to do that, conversation analysts examine naturally occurring social interactions. In fact, one of the most important distinguishing features of conversation analysis is its commitment to naturally occurring interactions. Considering that conversation analysis has emerged as a reaction to experimental studies, as the readings suggest, it is not surprising that conversation analysts take naturally occurring interactions as their data. Conversation analysts think that researchers do not need to create artificial order in interactions by using experimental designs because the order is already there and, therefore, it deserves to be studied in its own right. From this point of view, experimental reductionism is neither necessary nor justified.
This brings me to another commitment of conversation analysis. Ultimately, conversation analysts are interested in the mechanism that people utilize to make sense of their daily interactions. As a result, not only should one not reduce the task of meaning making that people regularly do—for example, by using experimental designs—but one should not make assumptions about the process of meaning making, for instance, by jumping to conclusions about why a certain individual said what s/he said unless the reasons are made public—and thus, observable—by the participants. The assumption here is that if the goal is to understand what people in interaction do, then it is not justified to impose the categories of the researcher on the participants and not to bring to analysis what is not available to the participants. In sum, conversation analysts are committed to empirical investigation of what is publicly available.
I think it is worth noting that conversation analysis has carved itself a niche as a scientific field by its emphasis on systematicity and its commitment to empirical investigations despite its characteristics—which one might consider “less scientific” from a hard science perspective—such as consideration of the emic perspectives of the participants and the field’s rejection of controlling variables. This is how conversation analysis is situated in relation to hard sciences, however, it is quite interesting that its commitment to uncovering universals and rejection of some ethnographic methods, such as field notes and interviews, are exactly what distinguish conversation analysis from ethnography even though both fields are interested in naturally occurring interactions and reaching an emic understanding (of course, how each would define and describe emic would be quite different). This is the reason why I believe what makes conversation analysis so unique is not so much its distinguishing characteristics and commitments but rather the constellation or combination of these features and commitments. 
I have to say, conversation analysis terribly reminds me of generative grammar. You might be wondering why since conversation analysis is all about what is observable while Chomsky is interested exclusively with the mind. I understand if I am the only one who sees the family resemblance, or perhaps it does not exist, but I think the preoccupation with the mechanism, be it universal grammar or adjacency pairs, makes them so similar. The consequence of this point of view of view is dismissal of the particularity, and maybe to some extent the accounts of the people who speak. Perhaps what I see as family resemblance is just a ghost, an illusion but I cannot shake this feeling that these two so different approaches to language are similar at a fundamental level. This is not to say that they are the same. I know that they are different on so many different issues. To me it seems like they share something like an assumption that I cannot put my finger on. What do you think?
Anyway, a couple of notes I took during class on the methodological commitments of conversation analysis. 
  • not to generalize beyond the instance
  • not to bring preconceived ideas
  • favor naturally occurring talk-in-interaction
  • mechanics and mechanism are always there
  • reflexivity is always there
  • review recorded data--unmotivated looking, collaboration, accessible raw data
  • descriptive as opposed to prescriptive or normative
  • privileges participants' interpretations as reflected in their activities
  • the unit of analysis is the pair not the individual
  • transcription, formalism
One thing that came up during our class discussion is the use of biological approach to science as a metaphor for conversation analytic approach. In this approach generalizability is not one of the main goals. In biology, even if you find one example of a species that is still finding. It doesn't have to be about all the members of a species. For example, if you discover a new species of butterfly it is still an important finding. It doesn't matter if its characteristics are not shared by most butterflies. Can you imagine finding a Big Foot ^_^ It would be a big thing even if you found only one. Anyway, this is all for conversation analysis for now. I will try to keep an open mind about CA. There is much to learn before passing judgment.

No comments:

Post a Comment