Work, Reproduction, and Health in Two Andean Communities

By Anne Larme, 1993.

Chapter 2 - Methodology

(page 20)

Nearly all families actively participate in community life. A high level of committee participation reflects the fact that committee membership is open to all married couples, regardless of wealth, age, or religion. Beyond this, election to cargos appears more limited. Wealthier and older couples such as D, G, O, and P had held the more prestigious political posts. Some younger couples had not as yet held political office, but had sponsored religious fiestas (F, I, L). Finally, some couples appear to have opted out of traditional fiesta sponsorship altogether, choosing instead to hold leadership positions in the Adventist (M) and Catholic churches (B). High levels of community participation are also reflected in the numbers of men (nine out of twenty) belonging to community bands that play for fiestas and national holidays.

The above information refers to the entire sample I used to understand the cultural and health context of Cuyo Cuyo. In the core of my study (Chapter 6, Chapter 7, and Chapter 8), however, I focus on the health of the most productive members of the sample, adults aged twenty and older,⁹ reckoning ages as of March 1988. For this portion of my study I included 38 heads of household¹⁰ and six of their unmarried children aged twenty and above, four from Ura Ayllu (C3, C4, G3, G5; three men and one woman), and two from Puna Ayllu (O3, S6; one woman and one man). Including all individuals aged twenty and above in my sample allowed for a more complete picture of the health of productive adults in Cuyo Cuyo. It also allowed for greater comparability of my analysis with the PSE time allocation analysis (Chapter 5; Winterhalder et al., n.d.).

Data Types and Collections Methods

Data on the research context were obtained primarily through participant observation and key informant interviews. The rapport established through my role as PSE supervisor and through regular interviews with household heads about health matters created numerous opportunities to accompany the research families in their subsistence work and ritual activities, as well as to interview family members and kin about various topics. Fictive kin relations with six of the twenty families greatly enhanced rapport, as well as creating reciprocal obligations. I became comadre (Sp., ritual co-parent) to Families B, H, J, L, M and T, and to several other families in the two communities, through a combination of Andean, Catholic, and secular rituals, including the unuchasqa (Qu., lay baptism), first haircuttings, birthdays, and graduations. Contextual data obtained in these ways were complemented by additional data from my PSE colleagues.

Data on the health context in Chapter 4 were obtained through a demographic analysis of Cuyo Cuyo birth and death records, through epidemiological data obtained from Ministry of Health (MINSA) officials in the city of Puno, and by reviewing records of the local MINSA Health Post. Other sources of data include observations and interviews regarding sanitary conditions in communities and worksites, Graham's nutritional study (1991), observations at the Cuyo Cuyo Health Post and with health promoters in Puna

(page 21)

Ayllu, interviews with Health Post personnel and a French missionary nurse-nun stationed in the District Capital of Cuyo Cuyo, a survey of the types and sources of ethnomedical, popular, and biomedical remedies, reproductive histories of married women in the sample, and participant observation with local healers as they performed divinations and ritual cures.

Ethnographic data on the relationship of work and gender in Chapter 5 were obtained through participant observation and key informant interviews. I participated regularly in days of agricultural work and herding with men, women and children, and in communal work days in Cuyo Cuyo. I also made three week-long trips to the highland mining site of Ancoccala, and one two-week trip to the lowland gold mining site of Huaypetue in the Department of Madre de Dios. Key informant interviews were conducted on various topics, including concepts of work and time, migration to Madre de Dios and to the yunka, and comparisons of life and health in the various worksites.

Time allocation data in Chapter 5, collected by the PSE Project and analyzed by Winterhalder (Winterhalder et al., n.d.), quantify the division of labor and the allocation of time by age and gender. These data -- a total 10,165 individual observations -- were collected in Ura Ayllu and Puna Ayllu through random spot checks carried out between 6:30 A.M. and 5:30 P.M., on a six day cycle, for a period of two years. Limited spot checks in Ancoccala and ethnographic observations provide supplementary information on activities in distant mining sites and in homes and communities during the evening and early morning hours that were not included in the spot checks.

The core of my data, the ethnomedical data in Chapters 6 and Chapter 8, was obtained primarily through semi-structured interviews with household heads over the period of one year (February 1987 to February 1988; see Appendix II for interview questions). I initiated the year of data collection with health and reproductive history interviews in order to provide a context for the symptomatology data to follow. I then began a series of "illness follow-up visits," where I asked about the symptoms or illnesses experienced by each family member since my last visit. Over the year, illness follow-up visits averaged eleven per family in Ura Ayllu, with 4.3 weeks between, and nine per family in Puna Ayllu, with 5 weeks between.¹¹ In all, I collected 1149 symptom/illness reports during these visits, for family members of all ages (Photograph 2.3).

The causes of adult symptoms and illnesses, the focus of my study on the interrelationship of work and health, are tabulated and analyzed in Chapter 6. They numbered one or more for each symptom report, and totalled 727 for Ura Ayllu adults, and 414 for Puna Ayllu adults.

To collect data on symptoms and illnesses, I used an abridged version of the Cornell Medical Index adapted for use in the Andes by researchers working in Nuñoa, Puno (see Leatherman 1987; Appendix II), with further advice from Leatherman on how to minimize ethnocentric bias. I also used the questions in an open-ended, rather than questionnaire-type, format. The questions systematically elicit information about symptoms related to specific body parts and organ systems, for example head, chest, abdomen, arms, legs, and skin. As my understanding of Andean ethnomedicine progressed, I adapted the original set of questions further to reflect Cuyo Cuyo illness categories and body concepts. For each symptom elicited, I asked for

(page 22)

a description and for the name of the illness if it had been labeled. I also asked about duration, cause(s), how and by whom the symptom or illness was treated, how the symptom or illness affected the individual's normal activities, and how the symptom could have been prevented.

Regular illness follow-up interviews, in addition to providing the specific data on symptoms and causes in Chapter 4 and Chapter 6, served as elicitation devices to obtain a variety of data on health and illness. They were a major source of information on folk illnesses, body concepts, healing methods and roles, childrearing and developmental concepts, and many related issues. I frequently followed up these visits with in-depth interviews in order to gain a deeper understanding of ethnomedical topics.

Scope and Limitations of the Data

In the following section I discuss the scope and limitations of the data, including the role of field conditions, language, research assistants, and the PSE project.

Adapting to Field Conditions

Anthropological research is necessarily opportunistic: one must adapt to field conditions and revise questions and methods accordingly. For example, after I arrived I found it was much harder to follow the health events of twenty families in two communities than I had initially planned. They worked in geographically dispersed locations throughout the year, making it impossible to interview them as often as I would have liked and to obtain data of equal depth for all. In particular, the migration of Ura Ayllu males to Madre de Dios and of Puna Ayllu families to Ancoccala meant that many of the day-to-day symptoms associated with life and mining in Ancoccala and Madre de Dios had to be missed.

Therefore, although I followed my basic research plan, there was more time between illness follow-up visits, and I found it especially necessary to supplement illness follow-up data with in-depth interview data. Trips to Ancoccala during the mining season, and visits with Ancoccala miners and their families when they were home in Puna Ayllu for agricultural work, meetings, or fiestas during the mining season helped to keep up data collection with Puna Ayllu families. A trip to Madre de Dios, interviews with Madre de Dios miners, and interviews with yunka miners and agriculturalists helped to fill in data on the health of Ura Ayllu men during their lowland migrations.

Another shift in my research plan was my focus on gender. This was partly because of the greater accessibility of females due to seasonal male migration, and partly because of my longstanding interest in gender issues. But more importantly, it arose due to the centrality of gender in Andean culture, and due to the strong sexual division of labor I observed, which is being exaggerated in Cuyo Cuyo's present economic transformation. All of these factors contribute to the differentials I observed in men's and women's health status, and which became increasingly apparent during data analysis.

Working with women had unanticipated benefits during data collection. Women were more reliable informants about ethnomedical concepts and more open to discussing them than were men. Being less educated and less experienced in Western ways, they did not have, nor attempt to impress me with, knowledge of Western biomedicine and Western body concepts. On the other hand, women were sometimes difficult to work with,

(page 23)

especially in the beginning. This was partly due to language barriers and also because they were generally timid with outsiders. But it was mainly, I am convinced, because women had less to "gain" from their contacts with Westerners than men, and because I was somewhat intrusive upon their busy work schedules.

After some initial reluctance to get involved, however, most of the women seemed to learn to benefit in various ways from the interviews. Most seemed to enjoy the attention focused on them, something they had not experienced with most other outsiders who had come to Cuyo Cuyo. Some began to use the interviews as "gripe sessions" about their lives -- probably one of the few places they could express their concerns to a willing listener. Occasionally women asked me for needed information, such as about birth control methods or medicines, although I generally discouraged this behavior for fear it would bias my data. Finally, a few women enlisted my support and that of my assistants for leverage with their spouses, whether through direct requests, or indirectly through compadrazgo invitations. Although I and my assistants tried to stay uninvolved in family matters, the nature of ethnographic research means that this was not always possible.

Language and Research Assistants

Quechua is the first language of the Cuyo Cuyeños, and most men, younger women, and school age children also know some Spanish. I was already fluent in Spanish and had studied Quechua intensively at Cornell University for six weeks before going to the field. I continued my study of Quechua in Cuyo Cuyo with a retired schoolteacher. Data were collected through a combination of Spanish and Quechua, depending upon the needs of particular topics and informants. Some interviews were carried out alone in Spanish; others through translation with research assistants; others by research assistants alone in Quechua, whose notes were later transmitted to me verbally or in writing. My knowledge of Quechua meant that I had basic conversational skills, could write and understand Quechua words and concepts during interviews, and could check the thoroughness of my assistants' translations.

The need for translators and the mobility of the research families meant that I had to rely on research assistants fairly extensively in my research. I was aided by three able individuals. Two single women, Virginia Ccori Ccapayque and Eufrasina Ccolque Ccori, who had high school educations and who were members of high status families in Ura Ayllu, worked with me in Ura Ayllu from January-July 1987, and from September 1987-March 1988, respectively. I was unable to locate a research assistant for Puna Ayllu until September of 1987, at which time Herberth Andrade Rossell, a university-educated man from a local mestizo family, who had been working with the PSE Project, was able to assist me for the remainder of my work. In the final months of my research, Eufrasina and Herberth were especially helpful in carrying out illness follow-up visits with more frequency than I had been able to before.

Many Andean ethnomedical concepts are difficult to convey in a Western language such as Spanish. Because many of my early interviews were conducted in Spanish with men, there was the added difficulty of getting past men's attempts to impress me with their knowledge of Western culture, to get at the Andean concepts I needed in order to understand the symptomatology data. These problems were greatly ameliorated in the final months of my project by close collaboration with Eufrasina and Herberth. By this time I had established sufficient rapport with the families, and had gained enough knowledge to know the questions I needed to explore intensively in the final critical months. Both Eufrasina and Herberth were skilled at understanding my questions and in obtaining the

(page 24)

information I needed from key informants within and outside of the study sample, whether in Quechua or Spanish. Discussions with Eufrasina and Herberth about their interviews and about life in Cuyo Cuyo helped to verify my understandings of the data and to formulate new questions, and were central to my understanding of Andean ethnomedical and cultural concepts.

One final note about research assistants relates to gender. All three assistants had excellent rapport with Cuyo Cuyo women. Virginia and Eufrasina, as members of the same community, felt comfortable with Ura Ayllu women. Herberth was interested in Quechua culture and sensitive to the concerns of women. Although male and misti (Qu., mestizo), women did not hesitate to talk with him.

I had access to both male and female domains throughout my research due to my ambiguous status as a female carrying out "male" roles, and with the help of both male and female assistants. Working closely with a male research assistant in the latter part of my research also had some unanticipated benefits for me as a lone female ethnographer. Herberth seemed to facilitate my research in several ways. First, he served as an entrée to settings to which I normally would not have had access, including rituals, fiestas, government functions, and the arduous male trip to Madre de Dios. This was because he was a highly respected misti in the District, and undoubtedly because of the machismo of Hispanic Peruvian culture. Another effect was more related to Andean gender concepts, however. Gender complementarity is an important theme in Andean culture; one must be married to be considered a complete person. When I worked closely with Herberth and people viewed me as part of a couple, it seemed that I was considered to be more "normal." I thus had better rapport with community members and could obtain better data.¹² A "team" approach to research, including both male and female researchers, was definitely an asset to my work in Cuyo Cuyo.

The Role of PSE

My relationship with the PSE Project generally aided my fieldwork. In addition to funding, I had a place to live, companionship and "survival" training in my first few months. I quickly decided to work with the same sample of families as PSE in order to more easily coordinate my dual roles of PSE supervisor and health researcher. Although it took me a few months to establish my own identity, I had immediate rapport with most families, following in the footsteps of the first year PSE supervisors. Another advantage was that two of my three research assistants had had prior experience through PSE and came to me partially trained.

Living in PSE quarters created some social distance from community members, but the flexibility and autonomy of this arrangement had many advantages. In contrast to living with one particular family, by living separately I could decide with whom to spend time and when, and I could keep the hours I wanted -- no small matter in a community without electricity, that goes to bed by nine P.M. and is up before dawn. I was able to develop special bonds with several families, especially compadres and especially during the months I was alone. These opportunities provided some of my richest data. While it was sometimes difficult to balance the social and work demands of my PSE role with my own research demands, I tried to blend the two roles to the extent possible. Although my role

(page 25)

as PSE representative never completely ended, staying an additional five months after the Project's close provided sufficient time to carry out my research satisfactorily.

My relationship with PSE has also been beneficial in the analysis stage of my research, where the existence of complementary data sets has broadened and deepened my analysis of work, reproduction, and illness in Cuyo Cuyo.