Written Texts
Many linguistic anthropologists look closely at various written texts: historical documents copied from archives, personal letters (such as the love letters I studied – see Ahearn 2001a), newspaper articles, e-mails, or official documents. Researchers who are interested in studying literacy practices – the ways in which people produce, consume, or refer to written texts in their everyday lives – often analyze written texts as mundane as the shopping lists that are pinned on refrigerators – or the billboards that litter urban “linguistic landscapes” (cf. Shohamy et al. 2010; Blackwood et al. 2016), drawing insightful conclusions about important cultural values and social relations. Even linguistic anthropologists who are not primarily concerned with literacy practices often find that paying close attention to the intersections between texts and contexts is not only beneficial but unavoidable in societies that are saturated by the written word.
How Do Linguistic Anthropologists Analyze Their Data?
The logistics involved in conducting research in linguistic anthropology can be very challenging. Many scholars in the field employ a research assistant to help them collect data, whether by conducting a survey, distributing written questionnaires, or translating or transcribing interviews or conversations. Even when a translator is used, however, most linguistic anthropologists emphasize the importance of not relying solely on such an intermediary but instead being fluent enough in the local language(s) to carry out many parts of the research project themselves. Some linguistic anthropologists prefer to conduct every aspect of their research themselves without any help from a translator or research assistant, but this is not possible when the scope of the project is too broad for one person to handle, or when it is not considered culturally acceptable for the researcher to speak alone (or even at all) with a member of the opposite sex or of a different caste or other social group.
No matter which methods are used, it is important to remember that all research, even the most “objective,” number-crunching sort, involves interpretation. From the formulation of a research question through the data collection stage, all the way through the data analysis process, all scholars, including linguistic anthropologists, knowingly or unknowingly engage in interpretation. If they are not careful, this interpretation process can involve the imposition of the researcher’s own culturally specific categories, which can prevent the researcher from gaining a deep understanding of the topic being studied. Therefore, the research process requires constant reassessment by the scholar.
A painful but extremely instructive example of the unwitting and inappropriate imposition of a researcher’s own categories comes directly out of my own fieldwork. At the outset of my dissertation research, which was about the shift away from arranged marriage toward self-initiated, or “love marriage” in Junigau, Nepal, I decided to conduct a quick survey of all the adults in the village. I asked each person whether his or her marriage was either self-initiated or arranged by parents – assuming that these were the only two types of marriage. As I went along, I would occasionally receive a response that confused me, as it sounded to my ears as though the person were saying the equivalent of, “My marriage was just like jabar.” Now, jabar by itself is not a word in Nepali, so when people mentioned the actual word jabarjasti (violently, by force), I heard jabar (which I did not understand) and jastai (“just like”). I had spent years in Nepal by that point and was fairly fluent in Nepali after spending three years in the country as a Peace Corps volunteer, but I had never heard the word jabarjasti. This was partly because my fieldwork occurred before the violent Maoist insurgency that began in 1996 and partly because forced capture marriage (jabarjasti chhopeko) was a stigmatized practice in the village, spoken of openly by very few people. So, whenever I heard what I thought was “just like jabar” as an answer to what kind of marriage a person had had, I asked, “Just like what – self-initiated or arranged?” The person would then usually choose either the self-initiated or the arranged category. One day, however, in discussing the village’s marriage practices with my Nepali sister-in-law, I mentioned the problem I had been having understanding jabarjasti, and she explained to me that there was a third category of marriage in Junigau: capture marriage, a forceful kidnapping of the bride and sometimes of the groom as well that was common in the village decades before. Since all of the responses to my initial question about whether people’s marriages had been self-initiated or arranged were potentially tainted by my incorrect assumption that there were only these two types of marriage in the village, I had to start all over again with the survey. As a result of this eye-opening mishap, I eventually came to reformulate my research as a study of marriage narratives in which Nepalis talked about their involvement (or lack thereof) in the decision-making processes surrounding various types of marriage.
Once linguistic anthropologists have all of their data, interpretation becomes a process of searching for patterns in order to find answers to the research questions that inspired the project – or to answer questions that emerge during analysis of the data. For many linguistic anthropologists, this involves reading and rereading fieldnotes and other documents, transcribing interviews and naturally occurring conversations, and statistically analyzing survey responses. Some scholars then go on to conduct a micro-level analysis of conversational data, while others focus on data concerning language policies or ideologies at a broader scale.
One approach to the micro-level analysis of linguistic data is known as Conversation Analysis (CA), which we discussed at greater length in Chapter 2. Developed in the 1960s and 1970s, partly as an outgrowth of ethnomethodology, a school within sociology that seeks to uncover the ways in which people work to establish and maintain taken-for-granted social structures in their everyday activities (Garfinkel 1967), CA is both a tool within disciplines such as linguistic anthropology and a discipline in its own right. Most linguistic anthropologists apply CA as one approach alongside others, using it to conduct detailed analyses of talk, then complementing those analyses with insights gleaned from other methods.
Whether a linguistic anthropologist collects recorded or transcribed conversations, written fieldnotes, survey results, or other types of information, once it has been collected, the next step facing the researcher is to analyze what usually amounts to mountains of data. Some scholars use index cards or sketch out flow charts to organize their thoughts and their data, while others turn to computer software to help them do this. There are many computer programs available to help researchers sort through their data, whatever form the information takes. Some programs allow the user to identify themes or codes in text files, graphics files, sound files, or even video files, then organize those themes in whatever ways make most sense to the researcher. Other programs help the investigator conduct a statistical analysis of quantitative data or analyze aspects of speech such as pitch or pronunciation. Even with the help of the most sophisticated computer programs, however, researchers ultimately have to discover patterns, make connections, and draw conclusions themselves.
What