Varenne: Ethnography and Participant Observation -- XI

Once one is, however temporarily, satisfied with one's corpus of texts forming the "data" for the project, the next step is to develop an approach to get the data to speak to one's overall interest while respecting them.

This is the beginning of the shaping of the final text, when matters of "interpretation" become preeminent.

There are many things one can do:

A most common step is to look for "themes" or "patterns"
1. This is a common way to bring out something from a set of (relatively) structured interviews when the researcher strongly framed the interaction and did not leave much room for the participant to improvize away from the topic-as-set-up.
2. This is often "under-theorized" and appears to consist of coming up with a few overall codes under which some of the data is presented. It may be a step towards a "grounded theory."
One may also push further to bring out what had not be planned but may very well be in the corpus that might be discoverable even though it could not have been imagined.

In Chapter 7, Garfinkel is making one fundamental methodological (in the philosophical sense) point:

That which is most constraining on individuals in the actual course of their life is available to us (researchers) through the same means that it is made available to co-participants.
This requires very detailed analysis of what it is that the participants actually do:
1. Note how stringent are Garfinkel's requirements for an adequate statement about this: ethnography, from his point of view, can only give "a sense of the order" to the extent that it is writing about accounts that cannot be systematically linked to the original event.
2. While the participants show in their work (not simply their "behavior") the order that they are constructing (e.g. "a lecture"), the properties of this work is not directly available by them for detailed accounting in the course of other work (e.g. during interviews). In other words the awareness that participants have of the conditions of their work does not take the form of the story that they may tell later about this work. The ethnographer is in a closely related position.

Garfinkel then illustrates what this can yield in his listing of various performances that are (necessarily?) aspects of the work of lecturing. They are both made necessarily by the organization of lecturing and constitutive of lecturing:

Some instances:
1. =(The lecture hasn't begun)=
2. =(Taking and holding places)=
3. =(Seeing the room starting to fill up)=
4. =(Late)=
5. =(Interruption)=
Focus on what is to qualify as an =(interruption)= (p. 230-231)

Alton Becker (1983) speaks of this process as a "philological" one.

Becker is invoking the task of European scholars in the 19th century attempting to decipher and then translate texts in long dead languages). He demonstrates how this might be done when the data is a Burmese proverb and what a Burmese language teacher told him.

What he does is move:

from the literal message as transcribed in a preliminary or naive way
to that which makes it something to say/do at the time of the utterance.
This what is implied in Becker's talk about the "past anchoring" of utterances and their constraining power on future utterances.
this requires a determined focus on "how-in-context" leading to a search to what where the resources for, and constraints on, what was said and what the saying imply about possibilities.
all language can thus be approached as indexical, that is pointing to, the conditions of the saying and that which it makes for the recipients and overhearers of the message.

Note that Becker, like Garfinkel, is making both a theoretical and methodological argument.

An example from sec. 26-32 in minute 1 from my research (Ambiguous Harmony). In this example every word is taken as sometimes multiply indexical

I	? overmarking of speaker (root metaphor); individualism;
had	? marking of tense
the people	? anonymous mass (the counter-root metaphor) possibly further specified by the "check the sink" phrase (they are people who check sinks)
come	? deixis inscribing that marked movement is toward the speaker (see work by Clifford Hill)
up	? not "over" thereby marking that the speaker is pointing to her living conditions in an apartment rather than a house.
to check the sink	? agency of objects acting an requiring further action (checking)

Thus, the methodological issue is more encompassing than might be suggested by thinking of Becker as only talking about translation or cross-cultural understanding.

The roots of Becker's stance in anthropology are partially in Malinowski's understanding of all meaning being produced by the "context of situation" of any symbol or utterance ([1935] 1965)
This methodological stance is extremely useful in any research using natural language when the goal is to gain a new (different from common sense) understanding of what we know very well as intimate participants in the same kinds of settings where we have found our informants.
1. the last sentence is an attempt to be more precise about the more common sense statement "as member of our own culture." The point is that, of course, our "knowledge," as participant, is not the same kind of knowledge that we hopefully gain by conducting this kind of analysis.
We can then listen to our informants for what they
1. affirm when they speak;
2. what they say with difficulty;
3. what they leave in the shadows (hide?); (see Kilkelly 1992)
Things get tricky when one notices, either on the basis of theory or comparison (and maybe even wishful thinking)

that participants have not said something;
1. Of course, what has not been said is infinite, but an analyst may be willing to offer that given that these speakers (those observed or interviewed) have said in other settings, (or that speakers have been reported as saying in other research), the not saying of those particular matters may be significant for analytic purposes.
  1. Justifying these sometimes permissible "interpretations" must be done with great caution and be received with great skepticism.
what they cannot say;
1. This, of course, cannot be documented through observation, though certain forms of theorizing (e.g. in psychoanalysis, marxism, etc.) may encourage or allow one to make statements about that which cannot be said by these people. Justifying these sometimes permissible statement must be done with great caution and be received with great skepticism.

Above all, one must respect the text, not only as "meaning" but in its form-in-context: "meaning-for-ethnography" is in this total text ("total social fact" to paraphrase Mauss), not elsewhere. Meaning is not "attached" to behavior in the now cliched rendering of the human condition. Meaning is immanent to the joint behavior of the significant group (that is the group for whom the behavior makes a difference).
1. This fundamental point is made even more complex if one takes seriously, as one must, Merleau-Ponty on the "prose of the world" and (1973 [1969])

I have talked about all this as "deconstruction" not only as play on a recent philosophical movement but also as a good account of an analytic enterprise that is fully consonant with the "constructed" aspect of cultural (social, arbitrary) orders.

In this perspective the social scientific task, in its descriptive state (both before and after theory), consists in the labor of discovering that which was needed for social life to be constructed.
This is a general task that can only be accomplished through painstaking discovery of what was involved in the construction of the most local of sequences (events).

"Immortal, ordinary society is strange... It is only discoverable. It is not imaginable." (Garfinkel 2003: 96)

"Immortal, ordinary society is strange... It is only discoverable. It is not imaginable."
(Garfinkel 2003: 96)