multiculturalism in Chicago

Oral History Association

Transcript Handling: An Ethnographic Strategy Author(s): Michael Agar Source: The Oral History Review, Vol. 15, No. 1, Fieldwork in Oral History (Spring, 1987), pp. 209-219 Published by: Oxford University Press on behalf of the Oral History Association Stable URL: http://www.jstor.org/stable/3674966 . Accessed: 28/10/2013 19:23

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

.

Oxford University Press and Oral History Association are collaborating with JSTOR to digitize, preserve and extend access to The Oral History Review.

http://www.jstor.org

This content downloaded from 129.105.215.146 on Mon, 28 Oct 2013 19:23:13 PM All use subject to JSTOR Terms and Conditions

ORAL HISTORY REVIEW 15 (Spring 1987): 209-219

Transcript Handling: An Ethnographic Strategy

MICHAEL AGAR

Within my discipline, sociocultural anthropology, "ethnography" has had an elaborate history. For many years, ethnographic research was the water in which we swam. Your graduate adviser lifted you by the seat of the pants and tossed you in. Now we have become self- conscious about

ethnography--perhaps to a fault-with meta-

MICHAEL AGAR is a professor and the graduate director in the Department of Anthropology at the University of Maryland, College Park He has worked ethnographically in South India, Austria, and, in an American context, on urban drug use and transportation.-

The present article was originally delivered as a talk to the Oral History in the Mid-Atlantic Region (OHMAR) annual fall meeting held on November 16, 1985, at Catholic University in Washington, D.C., as part of a panel entitled "How Anthropologists and Oral Historians Look and Listen." Data used in it were also presented to a meeting of the Transportation Research Forum. His full ethnography of independent truckers, Independents Declared: The Dilemmas of Independent Trucking, was published in 1986.

209

This content downloaded from 129.105.215.146 on Mon, 28 Oct 2013 19:23:13 PM All use subject to JSTOR Terms and Conditions

210 ORAL HISTORY REVIEW/Spring 1987

languages that run from Descartes to Derrida. Like everyone else in the field, I have opinions about what ethnography is. First of all, three of those opinions are sketched. After that, I demonstrate an approach to transcribed interviews that might be of interest to oral historians. Finally, I consider the approach in light of my general view of ethnography in order to consider the transfer of the method into oral history.

Ethnography Ethnography is the process of showing social action in the

context of one tradition to be coherent from the point of view of another. It is an exercise in translation writ large. In linguistics, translation is based on words and sentences; in ethnography, the raw material for the exercise are "strips," a concept introduced by Erving Goffman,' used by Charles Frake,2 and one that usefully characterizes ethnographic data.

Strips are any bonded phenomenon against which an ethno- grapher tests his/her understanding.3 They might be transcripts, documents of one sort or another, situations participated in and observed, or anything else that an ethnographer regards as "data." "Strip" is used by analogy with "film strip." The term is warranted by the rich variety of materials--consciously elicited or seren- dipitously encountered--required by ethnographic analysis.

Ethnographers usually spend a considerable amount of time with the people studied. One reason is that they want a wide variety of strips-across situations, across persons, across data types. The more the better. First, strip variety enables ethnographers to do the comparing and contrasting that goes on during fieldwork in a manner that systematically constructs understandings--a kind of "natural experiment" approach that organizes the variety that the world naturally presents if one attends to it. Second, multiple strips more adequately constrain the conclusions that emerge. It is one

'Erving Goffman, Frame Analysis: An Essay on the Organization of Experience (Cambridge: Harvard University Press, 1974).

2Charles O. Frake, Language and Cultural Description: Essays (Stanford: Stanford University Press, 1980).

3Michael Agar, Speaking of Ethnography (Beverly Hills: Sage Publications, 1986).

This content downloaded from 129.105.215.146 on Mon, 28 Oct 2013 19:23:13 PM All use subject to JSTOR Terms and Conditions

AGAR/Transcript Handling 211

thing to build a monstrous interpretive edifice on top of an interview fragment; it is quite another to grow a key theme out of a large number of interviews, observations, and documents.

A second opinion I hold is that ethnography is characterized by an "anticoherent" attitude, a kind of symbolic null hypothesis that leads one continually to question the understandings that one thinks are adequate. The attitude is particularly critical when working in one's own society, where all kinds of out-of-awareness understandings are carried into fieldwork It is much too easy to take a single strip from one's own society--an interview fragment, for example-and "fill in the blanks" based on all kinds of unexamined preconceptions about how the world "is." One has constantly to remember that what one thinks is going on probably is not.

A third opinion is that ethnography is a research process best modeled as a "funnel.'"4 You begin with anticoherence--things are not what they seem. You do not yet know what will count as "data" or where those data will lead you. This openness encourages the collection of a variety of unexpected strips to be combined with a skeptical, anticoherent view of them. At the end of the process, you tighten the screws, narrow the focus, and perhaps layer on some traditional hypothesis-testing research to do a final quantitative check of the understandings you have developed.

Ethnography has many more characteristics. But these three- strips, anticoherence, and the funnel-will be useful later in evaluating methods in a different disciplinary context, in this case oral history.

Transcripts One- but only one--type of strip used by ethnographers is the

transcript of a tape-recorded interview. Transcripts can represent minute phonological detail, or they can be edited into clean spoken prose. For the example used here, an intermediate style is used, one developed by the ethnomethodologists in sociology. This style captures many of the devices that English speakers use to foreground and background information--phonological prominence,

4Michael Agar, The Professional Stranger An Informal Introduction to Ethnography (New Yorkl Academic Press, 1980).

This content downloaded from 129.105.215.146 on Mon, 28 Oct 2013 19:23:13 PM All use subject to JSTOR Terms and Conditions

212 ORAL HISTORY REVIEW/Spring 1987

pause, interruptions and overlaps, and basic sentence intonation. The details of the notation will be introduced as the examples are presented.

The analysis of the prepared transcript is one derived from more general developments in discourse analysis and pragmatics. These relatively new branches of linguistics are practiced by individuals from a variety of scientific and humanistic fields who share an interest in elaborate stretches of language that go beyond the traditional single-sentence database of American linguistics. It is beyond the scope of this article to introduce the field, though introductions are available.5

The interview I will use here comes from my recent ethnographic study of American independent truckers. Excerpts from this interview will exemplify the kinds of inferences that one can make. I would like to demonstrate three kinds of inference- methodological inferences concerning the nature of the interview, symbolic inferences concerning critical biographical areas for the interviewee, and cultural inferences concerning the world in which the interviewee lives.

Methodological Inferences Consider the way this segment begins. "Red," as I call him, has

just finished a story of how he got into the trucking industry.

(Agar what was the:-what was the: uh: first of all what-about what year was that? (Red) oh that was in 1964. (3.4) (Colons indicate elongated vowels. Numbers in parentheses indicate pauses to the nearest fifth of a second. A period signals falling intonation, the usual sentence-final signal)

I asked the sort of highly technical, articulate question that years of training and experience make possible. After I finished, Red gave a brief answer, followed by a pause that-by Standard American English conversational rules-ran roughly for an eternity.

'Gillian Brown and George Yule, Discourse Analysis (Cambridge: Cambridge University Press, 1983); Michael Stubbs, Discourse Analysis: The Sociolinguistic Analysis of Natural Language (Chicago: University of Chicago Press, 1983).

This content downloaded from 129.105.215.146 on Mon, 28 Oct 2013 19:23:13 PM All use subject to JSTOR Terms and Conditions

AGAR/Transcript Handling 213

The methodological moral here is that I asked a bad question, a question that did not link into any of the natural organization of experience that would lead Red to talk on and tell a new story. The primary guiding rule for this "career history" interview is chronology. When faced with Red's conclusion to his initial story, I did not know where to go, so I fell back on chronology. Red could have cared less.

a:nd of course (2) you're working on percent-on a percentage of the + total revenue on-for the load. (A "+" shows a longer-than usual pause of less than one second. The "-" marks a continuing intonation, like a fall-rise, signalling "more to come.")

The shift from "percent" to the phrase that follows is a message that the hearer/interviewer is an outsider. One independent trucker would say to another that he is working "on percentage"--paid a share of the total rate for the shipment-or "on mileage"-paid by miles traveled. Red starts to give an insider's description, then edits it into a more elaborate phrase to spell out how the system works. The ability to expand the background knowledge that an outsider does not have is one attribute of a good ethnographic interviewee.

so: he + gave me an application form.= he said just write your name n (1.6) dress and phone number on it. said I'll take care of the rest of it. (2) so: (4) he said uh: (3.2) (A "=" indicates a run-on with no pause at all)

Red describes his interview with a dispatcher at the new company he is about to go to work for. The dispatcher gives him a form and says he will take care of the rest. There follows a sequence of pauses, fillers, and stretched out vowels that makes the end of this segment stand out like a bottomless conversational pit.

Why is the pit there? As I later learned, Red is at a point in the story at which several regulations are in the process of being broken. DOT regulations require certification of drivers, and insurance company regulations usually require more experience than he has. The gap signals an area in the hiring process that Red chose not to discuss.

A careful analysis of the transcript shows that the method carries within itself information for its own evaluation. The

This content downloaded from 129.105.215.146 on Mon, 28 Oct 2013 19:23:13 PM All use subject to JSTOR Terms and Conditions

214 ORAL HISTORY REVIEW/Spring 1987

examples I used here highlight a poor question that I asked, an on- line editing ability that allows Red to spell out background knowledge for outsiders, and a plausible guess about a topic that he did not want to pursue. The quality of the interview can be documented rather than guessed at.

Symbolic Inferences Red describes how he prepared for the interview with the

dispatcher.

heh + so I went home got all sharped up.= put a suit on. and I get a heh + went down and finally found ABC Freight Line steel division office which is a (1.6) little white bungalow, (1.2) a very sma:ll bungalow, a (1.4) actually a one room sha'ck (1.4)

on a dirt parkin lot down at Sparrow's Point, a: nd I walked in= here's all + half a dozen guys sittin around in work clothes. (Italics indicate phonological prominence, whether due to stress, volume, pitch, or some combination.)

In this segment truth and beauty come together. He begins with "suit" and ends with "work clothes." In between, he shifts from "bungalow" to "one room shack." The beauty comes in the middle- note the patterns of prominence, the elongated vowels, the timing. He speaks poetry without even knowing it, produces spontaneous talk that explains why some linguists are now interested in what they call "conversational aesthetics."

That' s the beauty. The truth comes in when you learn that Red is expressing a contradiction that is an important part of his life. He has a college degree and shifted into trucking from a white-collar job. Even today, he is equally at home hobnobbing with legislators on Capitol Hill or talking shop at the truckstop. His story represents the two worlds and the transition between them; the aesthetics of the transcript help us to comprehend this phenomenon.

Cultural Inferences Red's symbolic contradiction is not completely idiosyncratic,

but still, the ethnographic goal is to make broader inferences about a world he occupies, inferences that generalize to the stories that other independent truckers tell as well. Here's an example from an earlier part of this segment:

This content downloaded from 129.105.215.146 on Mon, 28 Oct 2013 19:23:13 PM All use subject to JSTOR Terms and Conditions

AGAR/Transcript Handling 215

(Agar) how did the deal work in terms of the: lease and (1.8) how did the deal work in terms of the lease and all that kinda stuff + with that particular carrier. (Red) heh + we:ll (2.2) not real well

My question followed the disastrous chronology item already discussed as a methodological issue. The question worked, but for the wrong reason. At this point in the study, I still had a working notion of "lease" as a document, a legal contract that specified the relationship between independent trucker and carrier. When I asked the question, I thought I was asking about the details of a document. As will soon become apparent, Red taught me by his answer that I had actually asked about the general work situation, that independents used "lease" to talk about the carrier in general In ethnography you only learn what you've asked after you get an answer.

I had it all figured out that (2.4) that it was gonna cost me five cents a mile for fuel, and + so much for a license, n - I needed so much a week, n + all a this, n (2.2) a:nd when I got my first + paycheck it was about half of what I really needed.

Red reports what I came to think of as "paycheck shock," a report that I heard over and over again as different independents described their initiation into the trucking business. The carrier deducts costs-fees, payments on permits and insurance, advances provided during the trip, etc.--so that when the check finally appears it is an unpleasant surprise, well below what was expected, based on the rate for the freight.

+ a:nd uh: (1.8) back in those days uh: + steel haulers particularly were (2.4) treated pretty shabbily + really. it was nothin uncommon ta go to a steel mill ta pick up a load n + sit there fer + four six eight twelve hours.

Now Red shifts from his opening statement about the cost/ revenue squeeze to how they treat you. The carrier dispatches you to a steel mill and then you wind up just waiting around. Once it is on your truck, on-time delivery is the name of the game; but until they are ready to load you, why, you can just cool your heels, or maybe "tires" would be the more appropriate term. I frequently heard that

This content downloaded from 129.105.215.146 on Mon, 28 Oct 2013 19:23:13 PM All use subject to JSTOR Terms and Conditions

216 ORAL HISTORY REVIEW/Spring 1987

complaint. Just after this excerpt, Red spells out the fact that you are not getting paid while you wait around; that passage was already cited in the methodology section to show how he shifted from "percent" to "percentage of the total revenue."

a:n uh (2.4) this carrier particularly + uh: (1) very cheap rates. a:nd I guess that's how they got all their business. + I know it is as a matter of fact. (2)

Back to economics. Red adds another undesirable attribute of this carrier-their rates are too low.

The fascinating thing about this complaint is that it cannot be true. In the time of the story, 1964, the Interstate Commerce Commission regulated rates. Now, in the 1980s, deregulation is the trend and truck rates are dropping like mad. Red takes a current trucker complaint and works it into the story. It is not obvious in the transcript, though in retrospect I noticed: (1) The chunk is bounded with pauses, discourse particles, and falling intonation. The topic is. not elaborated or recycled like most of the others. (2) The prominence on "very cheap rates" and the shift from "I guess" to "I know" is a signal of "he doth protest too much."

so: anyhow fer (1.6) oh about two weeks they ha-they ran me back and forth to Pittsburgh.= rId take a load up one night.= bring another load back the next night. (2) time you made two rounds like this n + sit around the steel mill fer (1.4) half-half the-half the afternoon n (1.8) waiting ta-waiting for your load, + time ya put in about four days like this why you're pretty well + whipped out. (My "backchannel signals" like "uh huh" are left out.)

Red moves back from the economic topic of the previous segment to the human qualities of the work again.

a:nd uh: + but I wasn't really making enough-enough money.

Now he returns to the cost/revenue squeeze with which he began.

n then they (2.2) they started runnin me up into New Jersey, an uh Eastern Pennsylvania, an uh + most the time deadheadin back, so I + really in bad shape moneywise. (2)

This content downloaded from 129.105.215.146 on Mon, 28 Oct 2013 19:23:13 PM All use subject to JSTOR Terms and Conditions

AGAR/Transcript Handling 217

"Deadheading" is the term for running empty. Sometimes a carrier pays a mileage rate if they dispatch you from one point to another without freight, but usually when your trailer is empty you are not getting paid. The deadhead runs Red reports add to his costs and contribute nothing to his revenue.

so I went-after r d been there about oh I guess about six weeks, (1.4) I went and asked em ta + put me back to running the Pittsburgh + thing. (2.2) a:nd I was (3) informed in no uncertain terms that they would make the decision as to where I was to run, an uh: (2) if I didn't like it we:ll at's too ba:d you know.

Here Red ties cost/revenue and human problems together. He is going broke; he asks them if he cannot get back the run that, though not great, is at least better than what he is doing; the carrier responds in an authoritarian style.

Red makes a story that elaborates upon his original answer to my question about "the lease"-"not real well." As he makes it, he draws on his experience with different carriers and his knowledge of the trucking industry. And that is what I am after as an ethnographer. I am less concerned with this particular carrier and this particular independent and more interested in how the description patterns with other descriptions by other independents.

Red draws on two major themes to build his story. The first is profitability. He is not making enough. His costs are higher than he had thought they would be. He has to sit around and wait for loads without reimbursement, the famous issue of "detention time." He runs empty-" deadheads"- again without reimbursement. The carrier will not dispatch him to more profitable runs. The second theme is the poor human quality of the workplace-he is tired and overworked and has to deal with an authoritarian carrier.

The themes he uses are also found in the management literature. The first, profitability, is the traditional American guideline. The second, personal qualities of the workplace, is the "new" discovery of Japanese management. Red, however, goes beyond a simple statement of theme to teach us some of the consequences of the theme for independent truckers- consequences for profitability, detention, and deadheading; for personalism, dispatch--and how they intertwine to produce a description of a carrier.

The narrative accomplishes what Red set out to do-to show

This content downloaded from 129.105.215.146 on Mon, 28 Oct 2013 19:23:13 PM All use subject to JSTOR Terms and Conditions

218 ORAL HISTORY REVIEW/Spring 1987

why "the lease" was such a bad deal. It sets up the next section of the transcript, the story of going to work for a new carrier, by showing that there were no reasons to stay at the old one. It is a good creative piece of story-making, a piece that teaches the outsider something about the working world of independent truckers from the inside out.

Ethnography and Transcripts The analytic perspective of discourse analysis, applied to a

fragment of an interview transcript, shows how the transcripted talk can be organized to support methodological, symbolic/poetic, and cultural inferences. The perspective helps us see what was foregrounded, what was backgrounded, and what was chunked into a discourse unit with what else--when the speaker was on a roll and when he tripped and stumbled.

Once in a while, Red actually laid out some inferences for us. After he talked about detention time, he added that you are paid on percentage; therefore, you are not paid for waiting around. But the more typical case in transcript analysis is that inferences are constructed by the analyst, anchored in and motivated by the talk, but arcing away from it toward more general ethnographic truths.

To build those inferences in the examples, I smuggled in information not present in the transcripts themselves. In the methodological examples, I had to tell you what kind of interview this was, explain different methods of payment, and fill you in on the application process and the regulations relevant to it. In the symbolic/poetic case, you needed to hear more about Red, his past life, and his current activities outside of driving a truck. In the presentation of the cultural material, paycheck shock was shown to be typical, deadheading and detention time were explained, the recent history of deregulation was outlined, and the role of this fragment in the interview was described--Red is using characteristics of the working world to build a case that he should leave the carrier.

Where did all this new information come from? From other strips. In ethnography, the emphasis is on multiple strips, as noted in the earlier part of this article: strips from other moments with an interviewee, from other interviewees, from other situations besides interviews, and from documents or any other sources. The business of ethnographic analysis is the construction of understandings that

This content downloaded from 129.105.215.146 on Mon, 28 Oct 2013 19:23:13 PM All use subject to JSTOR Terms and Conditions

AGAR/Transcript Handling 219

account for the patterns across different strips. Inferences are constructed from and evaluated against strips--transcriptions or any other type-with an emphasis on their convergence into core themes of group life.

I do not think I could have taken only this fragment and said the things I did. Instead, I took a strip that is part of an already accomplished ethnography and showed you how some key inferences that I have learned are important in many other strips are useful in understanding elements here. In other words, given the outline of ethnography presented in the introduction to this paper, I cheated. The actual study of independent truckers started out anticoherent, at the broad end of the funnel This demonstration is procoherent, composed after the study had fallen out of the narrow end of the funnel at the end of a couple of years of work.

The moral of the story is this. Discourse analysis is a powerful analytic tool for the transcribed interview, as useful to oral historians and ethnographers as it is to any other group with an interest in texts. However, the inferences that ethnographers construct, using transcripts as the occasion, presuppose a broad range of strips besides the one of interest at a particular analytic moment. Part of the analysis exemplified here relied on the organization of the data provided by discourse analysis, but another part relied on strips pertaining to the person interviewed, his occupational role, the industry in which that role was located, and the place of that industry in the American political economy.

Transfer of the method into oral history, then, would not be a simple matter of transcript notation, although my guess is that use of the notation alone would be informative. But to produce full benefit from its analytic power, the method should be applied with a variety of other strips from the person, his/her community, and the broader political economy in which s/he participates. Whether or not such an approach is useful and practical within the oral historian's research tradition is, of course, a matter for members of that tradition to evaluate. But I hope that this outsider's presentation of one ethnographic strategy is at least, from the oral historian's point of view, an interesting mistake.

This content downloaded from 129.105.215.146 on Mon, 28 Oct 2013 19:23:13 PM All use subject to JSTOR Terms and Conditions