That summer determined my future as an anthropologist.  On a hot August afternoon in Morelia, Michoacán’s state capital, Foster and I sat waiting in the brown Ford Station Wagon provided to our summer project by the University of California, while Virginia and Lola Pichu set about making their weekly purchases at the “Independencia” wholesale market.  Never a person to avoid an opportunity to move forward, Foster asked me at one point in the conversation, “And what do you plan to do for your dissertation research?”  Having only just completed the first year of graduate studies at the University of California at Berkeley, I hardly had a good answer.  I mumbled something about being interested in the effects of urbanization on pottery-making towns like nearby Capula, Tzintzuntzan on the edge of Lake Pátzcuaro, and Patamban in the distant Sierra Tarasca.  Instead of encouraging me to pursue this variation on Robert Redfield’s (1941) “folk-urban continuum,” Foster asked if I might be interested in following Tzintzuntzan migrants to the national capital.  He knew, as I would come to learn, that such a study had been suggested long ago by Oscar Lewis (1952:41), who had argued that follow-up investigations of migrants from well-studied villages like Tzintzuntzan would yield valuable comparative data on Mexican urbanization.

In 1969, when I began doing my dissertation fieldwork among Tzintzuntzan migrants in Mexico City, most anthropological thinking about Mexican peasant villages included critical assumptions about their “closed corporate” and “limited” characteristics (see Potter, Diaz, and Foster 1967).  My theoretical and methodological frameworks were modeled after the prior studies of rural-urban migrants in Mexico City done by Oscar Lewis (1952) among Tepoztecans and by his student Douglas Butterworth (1962) among Tilantongans.  Thus, from the outset, my objectives in studying migrant adaptation were both comparative (Kemper 1970, 1971a, 1987a) and longitudinal (Kemper and Royce 1979, 1983), framed first within the then prevailing paradigms of acculturation and modernization (Kemper 1977, 1987b), and more recently in terms of broader political-economic relationships between city and countryside (cf. Kemper 1979a, 1994a, 1995a; Kemper and Rollwagen 1996).

This could have been just another one-shot field study leading to a dissertation, a monograph, and some articles.  Instead, it became the basis for a continuing thirty-year-long investigation of the “extended community” of Tzintzuntzan.  In retrospect, it is clear that I entered the field at a critical moment. The late 1960s were not only tumultuous times in Mexico City (e.g., the October 2, 1968 massacre of students at the Plaza de las Tres Culturas took place only six months before I entered the field), but also marked the transition out of the long-standing bilateral agreement known as the Bracero Program, through which a generation of Mexican laborers had been contracted to work in the United States.

The impact of the Bracero Program on towns in Mexico, and especially in Michoacán, was not lost on ethnographers of that era. In fact, in concluding his monograph on the Sierra Tarascan community of Cherán,² Ralph Beals had remarked,

Cherán, like many Indian communities of Mexico, is increasingly influenced by the town and the city. . . . Cherán is probably more influenced by Gary (Indiana, U.S.A.), Mexico City, and Morelia (possibly in diminishing order) than it is by Uruapan and Pátzcuaro” (1946:211).

Foster also had noted the positive experiences of Tzintzuntzeños – almost fifty percent of the village’s adult males – who had traveled north to work in the United States: “Most are very anxious to get back to the United States, either to live permanently or to work for an extended period” (1948:149).  During the 1950s, from 50 to 150 men went to the U.S. each year; at a time when few Tzintzuntzeños ventured to work and live in Mexico City (cf. Foster 1967:275-277).

Because I approached the migration problem from a singular perspective – that of a community already well-known ethnographically through Foster’s work – I could define broadly the people and places relevant for questions such as “Who is migrating?” “Where are they going?” and “From where are they returning?”  Thus, although my first fieldwork was concentrated in Mexico City, for my dissertation I also analyzed Foster’s master data file of all individuals censused in 1945, 1960, and 1970 to determine the destinations of all Tzintzuntzan migrants throughout Mexico and the United States.  In this way, I got a sense of the broader picture within which cityward migration took place – and I opened a Pandora’s box in terms of dealing with the broader migration system in which the community of Tzintzuntzan was participating (Kemper 1996a).

BEING A STUDENT: DISSERTATION FIELDWORK IN 1969-1970

Mexico City was selected as the field site because, at that time, Tzintzuntzan migrants shared with many Mexican peasants, including those of Tepoztlán, a preference for settling in the capital.  Thus, not only would I find enough migrants to make the study worthwhile, I would be able to compare my findings directly with those of Lewis and other scholars, while avoiding the arduous task of locating Tzintzuntzeños in a number of Mexican cities.  Under these conditions, I conducted fieldwork in Mexico City, with several brief trips to the village, from April 1969 to August 1970.

The dissertation fieldwork among the Tzintzuntzan migrants in the capital fell into three phases, punctuated by short trips back to the United States to renew tourist papers every six months.  In the first phase, I faced three main problems: finding accommodations for me and my wife Sandra, locating a relatively small number of migrants in a city of over eight million people, and beginning the process of data collection.  During the second phase, marked by a transition from “passive” to “active” research strategies (Freilich 1970:24), I conducted a detailed ethnographic census of as many migrant households as could be located.  The concluding phase was devoted to gathering life histories, household budgets, and social network data and to administering projective tests and questionnaires about migrant experiences.  Thus, as the fieldwork progressed, the problem orientation became more specific and the data collection procedures more structured.

Selecting a Place to Live

All anthropologists face the problem of finding accommodations compatible with their field situation.  In my case, after two weeks of living in hotels and rooming houses, we finally found an affordable furnished apartment for rent near the center of the city.  Having an apartment independent of all of the migrants had immediate advantages and disadvantages: on the one hand, it gave us a place away from the constant demands of fieldwork in an unfamiliar setting; on the other, it denied us the intimacy of participant observation common to anthropologists who work in small communities.  As it turned out, this separation gave me a neutrality which preserved the opportunity for future research.  If I had made a commitment to reside with a particular family during this first fieldtrip, it might have adversely curtailed my future options (cf. Berreman 1962).

Studying a Dispersed Population

When I arrived in Mexico City, I had a list of about 20 migrants’ names and just two addresses, which I had obtained in a hurried search through Foster’s data files before my departure from Berkeley.  My immediate goal was to locate as many migrants as possible, as quickly as I could, so that I might get on with the “real” fieldwork.  However, it took about a year to develop a network which included nearly all of the migrants.  Eventually, my network became so much wider than that of any of the Tzintzuntzeños that they would ask me where one of their fellows lived or worked.  The migrants were spread among more than forty neighborhoods in the metropolitan area, mainly in the northern periphery where many of the men worked in factories. Although a few people, usually relatives or compadres (co-godparents), lived next door to one another; most were separated by miles of urban traffic.  Nor did the Tzintzuntzeños have a village-based voluntary association to compensate for their geographical dispersion.

Had I known from the outset that the Tzintzuntzan migrants were so numerous (483 persons living in at least 74 households, rather than my original guess of about 100 persons in 20 households), so dispersed (spread over the metropolitan area, whereas I had hoped that they might live in one or two neighborhoods), and that I would log so many miles (9,000) doing the fieldwork, I would have thought about another dissertation topic.  The initial fieldwork in metropolitan Mexico City was grounded in the reality that the migrants were dispersed over some five hundred square miles!  In these circumstances, the field methods associated with nucleated peasant villages seemed completely inadequate.

In mid-fieldwork, I had the good fortune to be able to attend the 1969 annual meeting of the American Anthropological Association in New Orleans.  At the book exhibit, I happened to lay hands on a new book (Mitchell 1969) dealing with social network approaches to urban populations in central Africa.  Skimming through the pages of the “Introduction” that first time was an epiphany!  Here was a serendipitous solution to my dilemma.  Back in the field, I was able to design a simple social network survey that provided crucial data for understanding how the Tzintzuntzan migrants operated in their dispersed circumstances in Mexico City (Kemper 1973; cf. Foster 1969).

During that first period of some seventeen months of fieldwork in 1969-1970, I occasionally was introduced to migrants who turned out not to be from the village of Tzintzuntzan proper, but instead were from one of the eighteen hamlets located within the boundaries of the municipio (county) or parroquia (parish) headed by Tzintzuntzan.  My initial reaction to discovering such migrants was to record data about them and then to put it into a different file – separate from that on the “real” villagers being studied.  My problem was two-fold: first, that I could not link directly the migrants from these hamlets to the large ethnographic database which Foster had established for the residents of Tzintzuntzan proper; second, that I was uncomfortable about “contaminating” the study of the migrants from Tzintzuntzan proper with information about their friends and relatives who also happened to be living in Mexico City.

In effect, my narrow definition of “community” as a specific place on the ground – rather than as a broader social and economic field – made it hard for me to see what was going on among the migrants in Mexico City.  In the final analysis, I did include data on these migrants from the hamlets around Tzintzuntzan -- but I always felt uneasy about it, because this seemed to violate the prevailing Lewis-Butterworth framework of origin-destination analysis.

Establishing a Satisfactory Role in the Field

My affiliation with Foster provided the credentials necessary for conducting the fieldwork and for participating in professional activities in Mexico City.  Since most of the migrants were aware of his work in Tzintzuntzan, they were willing to accept me as a student interested in their past, present, and future.  Especially in the beginning, Foster’s name (and through his connections, the names of certain well-known villagers) opened doors for me which otherwise might have remained closed.  In terms of my professional relations, Foster gave me letters of introduction to the directors of several anthropological institutions based in Mexico City.  In addition, because the Society for Applied Anthropology was holding its 1969 meeting in Mexico City, he came down and introduced me to a number of leaders in Mexican anthropology.  Later on, he sponsored my membership in the Sociedad Mexicana de Antropología.  At his suggestion, I audited a course at the National School of Anthropology and History during my first months in Mexico City.  Toward the end of the fieldwork, at the invitation of Professor Angel Palerm and with Foster’s encouragement, I taught a graduate seminar on urban anthropology at the Universidad IberoAmericana.³  In these ways, I made contacts with Mexican anthropologists that would be valuable throughout my career.

The implications of my role situation among the Tzintzuntzeños and among Mexican anthropologists are clear.  Under the auspices of a well-known patron, whose guidance placed me on the right path to do the dissertation research, I was able to establish my own contacts among the migrants and with key individuals in the Mexican social science community so that my research could continue.  Following Foster’s lead, I maintained a neutral stance toward the political side of Mexican anthropology; as a result, I have had no difficulty in continuing fieldwork with the people of Tzintzuntzan.

Collecting Field Data

I adopted the widely used HRAF system for coding and filing field notes that Foster had recommended in the “methods” seminar I had taken before going to Michoacán in 1967.  Information was typed onto 5 x 8 inch sheets and filed in two sets: one arranged topically in the HRAF numerical categories; the other by households.  A third copy was mailed periodically to Berkeley where Foster read the notes, wrote comments to me regarding points of special interest, and then filed them for safekeeping.  I was spared a great deal of work, especially in dealing with genealogies, because these data already existed in Foster’s files.  Moreover, we were each sufficiently familiar with the HRAF system (which Foster also used) and with the specific information in the files that it was easy for me to ask for data and for Foster then to retrieve, photocopy, and mail it to me.  In addition to making his data files available, Foster also made a point to follow-up on any leads related to migration when he was in Tzintzuntzan doing his own research.  In recognition of our common interests and in partial reciprocity for his efforts on my behalf, I collected some data on topics of special interest to Foster.  For example, I administered Thematic Apperception Tests (TATS) to fifteen migrants to provide comparative data for his sample of about twenty respondents in the village.  In addition, I also gathered extensive information on compadrazgo relations among the migrants.  I used these comparative data to examine villager-migrant differences and similarities in my dissertation and in subsequent publications (cf. Kemper 1979b, 1982).

The most important common element in our research was the 1970 ethnographic census.  We asked for almost identical information and gave the census at about the same time.  Whereas Foster was able to obtain a 100% response rate in the village – and found that Tzintzuntzan was a community of 2,253 residents – I could census first-hand only about 70% (51 of 74) of the migrant households, although in nearly all of the remaining cases I was able to get reliable second-hand information.  As a complement to my work among the migrants, Foster inquired about the whereabouts of all villagers who had been in the community at the time of the 1960 census but were no longer present.

It was fortuitous that I did my initial fieldwork among the migrants at the same time that Foster was administering a census in the village.  As a result, we now have a good set of demographic and socioeconomic “core” data on the Tzintzuntzeños in the capital comparable to that on their families of origin in Tzintzuntzan.  Moreover, because the Mexican national census was conducted in February 1970, our ethnographic census data for migrants and villagers can be related to broader regional and national situations.

ANALYZING FIELD MATERIALS AND DATA FILES: 1970-1973

Upon returning to Berkeley in September 1970, I began to organize my field materials into a format suitable for a dissertation.  In accord with my original plans, I also examined Foster’s data files for background information on the migrants’ situation before they had gone to Mexico City.  I soon realized that, although Foster had done considerable analysis of his materials, the information I sought was not tabulated according to my needs.  Therefore, we went to considerable effort and expense to convert his 1945, 1960, and 1970 census materials into a format amenable to computer processing techniques.  After several months’ work, I was able to generate a substantial set of cross-tabulations that showed how markedly Tzintzuntzan had changed during the 25 years of Foster’s work there.  This material not only was valuable for the dissertation, but also served as the basis for a joint article (Kemper and Foster 1975) on the impact of Mexican urbanization on Tzintzuntzan.

In addition to devoting considerable time to analyzing the village census data and comparing the results with those of my migrant census in Mexico City, I also counted on the individual data 5” x 8” sheets kept on each person who had been censused by Foster in 1945, 1960, or 1970.  By close examination of this data file, which at that time contained personal information on more than 3,000 individuals, I ascertained that more than 700 villagers had emigrated since the late 1930s.  Furthermore, because of the systematic detail on these data sheets, I could specify the type (and often, the place) of destination for each emigrant.  Then, by correlating these data with the computerized household census files, I was able to show that the migrants to Mexico City were “positively selected” (i.e., came from households with higher living standards, better educational levels, and more innovativeness than found in non-migrant households), but that this selectivity had declined in recent years as the number of emigrants had increased.  Thus, I was in a position to support the demographic hypothesis of migrant “regression toward the mean” (cf. Browning and Feindt 1969) through a combination of my census data on the migrants in Mexico City and Foster’s data on the entire natal community.  This is a use to which Foster would probably not have put his own data, since his interests were far removed from migration theory.

There are several lessons to be learned from my analysis of Foster’s data files in conjunction with the analysis of my own field materials.  First, comparability should be built into a study from the beginning, since it is difficult to agree on definitions of, say, household and family types after investigations are completed.  Second, using another scholar’s data files requires an understanding of what the data “mean;” that is, the ambiguities in the field notes and census materials must be explicable either by the original scholar’s direct explanations or through annotations in the data files.  Third, the most useful data from another anthropologist’s field materials are often those of lesser importance to that person’s current interests.  If this is so, their analysis by a student or colleague may be a valuable service to the original investigator.  Finally, there must be an agreement as to what data may be used and to what purpose they may be put; otherwise, the original fieldworker’s plans for subsequent publications may be compromised.

After about ten months of analyzing my own field materials and those from Foster’s files, I managed to write the dissertation in about ten weeks (Kemper 1971b), so that I could begin an NEH post-doctoral fellowship on Mexican-American studies at Berkeley for the 1971-1972 academic year.  During that year, Foster and I held a seminar with a number of anthropology graduate students writing dissertations based on urban fieldwork.  This seminar led to our collaboration on a volume titled Anthropologists in Cities, from which our introductory chapter (Foster and Kemper 1974) and my chapter on the Tzintzuntzan migrants in Mexico City have been reprinted several times (Kemper 1974a).  In these, and other, articles (Kemper 1973, 1974b, 1975) I realized that there was no “ethnographic present;” instead, constant changes in the villagers’ and migrants’ situations were rendering these publications obsolete even before they appeared.

By late 1973, my sense of frustration had reached a point that getting back to Mexico City and to Tzintzuntzan had become a top priority. To this end, I received a small grant-in-aid from the Wenner-Gren Foundation for Anthropological Research to carry out a ten-week ethnographic survey among the Tzintzuntzan migrants in Mexico City during summer 1974.  This fieldwork provided important mid-decade information on the changing migrant population and on the urban conditions in which it lived.

NO LONGER A STUDENT: FIELDWORK IN 1974 AND AFTER

The 1974 summer fieldwork involved re-establishing contact with those migrants whom I had known in 1969-1970, discovering which migrants had left Mexico City to return to the village or to settle elsewhere, and censusing those newly-arrived in the capital since 1970.  Of course, I did expect to gather additional data, beyond the census itself, on such topics as family and household developmental cycles, socio-economic mobility, and intra-metropolitan geographical movements.  All of these topics needed more study before I could transform the dissertation into a monograph about the migrants’ adaptations to urban life during the early 1970s.

That summer, I accomplished almost as much in ten weeks as I had in as many months during my earlier period in Mexico City.  The “costs” of the research, both in money and personal effort, were substantially reduced during this second field trip.  Not only did I know much more about the population I was studying, I also was a more effective ethnographer.  A number of circumstances made the 1974 research especially productive; I treat these in the same categories as above.

Selecting a place to live

Unlike the situation in 1969-70, when Sandra and I had rented a downtown apartment, on this occasion I went to the field alone.  I had a room to myself in the home of Rafael Campuzano Hinojosa, half-brother of Micaela González Hinojosa, matriarch of the household where Foster – and I, too – resided while in Tzintzuntzan.  Rafael’s family had been our closest friends among the migrants in 1969-1970. His wife, Felipa, took excellent care of all my needs, and was a superb cook as well.  In short, although their house was located somewhat inconveniently in the extreme northeastern section of the Distrito Federal, the situation was ideal for getting maximum work done in a short time.

More important than just having a place to live was the chance to observe first-hand, on a daily basis, interpersonal relationships within a migrant household.  In addition, the family’s entrepreneurship (they had opened two small stores) gave me an insider’s view of how migrants try to improve their economic situation in the city.  And, finally, the family served as willing key informants on a wide range of topics.  Since the family and I both were pleased with the arrangement, we agreed that I should continue to stay there on future trips to Mexico City.

That summer I also went to Tzintzuntzan for my first comprehensive village survey of migration.  Using key informants in different neighborhoods, I was able to check on the current location of all 2,253 persons listed in the 1970 ethnographic census.  For the first time, I began to see the village as the central node of an extended social and economic field which included the hamlets in the municipio, towns throughout the state, large cities like Mexico City, and even places in the United States.

Studying a Dispersed Population

Based on my earlier fieldwork, I knew that it would be impossible, in ten weeks or in a hundred, to locate and survey every Tzintzuntzan migrant in Mexico City.  I did hope to canvass, with some field assistance, about 75% of the population. In fact, we were able to gather firsthand data on more than 70 households and got reliable second-hand data on most of the remainder, for a total of some 600 persons in a total of about 110 households.  I had expected that the older or more affluent migrants who owned homes would still be living where they had in 1969-1970, whereas the younger, poorer, more recent arrivals would be more geographically mobile.  I discovered that, although about half of the migrants had changed residences at least once since 1970, and some were no longer in the capital, this proposition held true.

Since I had no car on this fieldtrip, I depended on the excellent public transportation system of taxis, jitney cabs, buses, and subways.  In addition, Rafael seemed to enjoy taking me to visit other migrant households on weekends and on his days off from his job with the telegraph office.  Even with this assistance, I found that the large number of recent arrivals forced a strict rationing of time and effort.  Although it would have been pleasant (and perhaps ethnographically profitable) to pass time with old friends, I was determined to gather “core” data on people not already in my files.  I also visited Tzintzuntzan to check on the dozen or so “return migrants” and to ascertain who else had left the village since the 1970 census.  In the process, I confirmed the emigration of 205 villagers from Tzintzuntzan between 1970-1974, with 28% going to Mexico City, 29% staying within the state of Michoacán, and 15% traveling north to the United States (Kemper 1977:53).

For studying a dispersed population, a series of visits has many advantages.  This time I knew what faced me and how to plan my available time to see as many migrants as possible. I did not have to build a social network family by family; I already had one that enabled me to cope with the migrants’ geographical distribution in the metropolis.  Moreover, the second visit to the Tzintzuntzeños made me aware of continuities and changes not observed on the first fieldtrip.

Changing Roles in the Field

Previously, I had entered the field as a student; now I was a professor.  Nevertheless, because of my relatively young age (28), I continued to introduce myself as “Roberto” and asked that people so address me. I continued to use the formal usted rather than the informal tú in all conversations except with age mates (and children), whom I knew well and who used it first in speaking with me.

When I had been a student, I could plead poverty when unusual requests for assistance arose.  Now, this was more difficult, and as I continued the research my social obligations to the migrants and villagers have had to be repaid with interest.  I continued to decline requests that I become the godparent of migrants’ children, not only on the grounds that I was not available on a full-time basis but also because I was a Presbyterian instead of a Roman Catholic.  Additional burdens of reciprocity accrued with professional colleagues in Mexico City. As an anthropologist hoping to carry out research in Mexico for the rest of my career, I felt responsible to lecture in local universities, to publish papers and books in Spanish in local journals, and to assist in training local students.  Surely, my greatest coup was arranging to bring Fernando Cámara (then Associate Director of the National Institute of Anthropology and History) to Dallas as a visiting professor for the spring semester of the 1974-1975 academic year.

Collecting Field Data

Since my primary goal was to survey as many migrant households as possible, and since I knew that there would be more than I alone could visit, finding field assistants was one of my first tasks.  One of the sons (Ramon) of Rafael and Felipa served as an interviewer; as did Claudio, the son of Larissa Lomnitz, an anthropologist at the National University.  After several training sessions, both proved to be superb census takers.

The productivity of the 1974 fieldtrip showed me that revisits are not only economical, they also encouraged me to go to the field with a set of specific queries.  Turning over a share of the “core” data collection to assistants gave me time to concentrate on these topics of special interest.  Having my field notes arranged topically by HRAF categories and also by family group meant that I could daily see the depth and breadth of my sample grow.  Let me give a brief example.

During the 1969-1970 field research, I had gathered a substantial body of data on the compadrazgo.  I had discussed these data briefly, in comparison with those gathered by Foster for Tzintzuntzan, in my dissertation, but had not analyzed them fully.  In returning to the field in 1974, the gathering of additional data on compadrazgo was a high priority.  I was particularly eager to find out the variety of occasions (e.g., baptism, marriage) for which compadres could be chosen, especially since the anthropological literature was very sketchy on how the institution of co-godparenthood operated in urban settings. Data collected in the summer of 1974 led to two papers (Kemper 1979b, 1982) in which I showed how urban migrants use the compadrazgo as a mechanism for improving their socioeconomic mobility, for strengthening ties with relatives and friends dispersed throughout the metropolitan area, and for maintaining ties with the village.⁴

THE WENNER-GREN SYMPOSIUM: LEARNING TO TAKE THE LONG VIEW

In summer 1975, I had the opportunity to participate, as the designated “junior” rapporteur and paper presenter, in a Wenner-Gren sponsored symposium on “Long-Term Field Research in Social Anthropology.”  If I had had any doubts about maintaining my work among the Tzintzuntzan migrants, the senior people I met at Burg Wartenstein and the subsequent editing of their papers convinced me of the necessity of longitudinal research for understanding urbanization and migration processes (Kemper 1979c).

In the mid-1970s, I also had to face the academic hurdle of tenure and promotion.  In our department, this meant publishing a monograph in addition to writing articles, doing book reviews, getting grants, receiving excellent teaching evaluations, etc.  I was able to visit the migrants in Mexico City briefly in November 1974, when the annual meeting of the American Anthropological Association took place there, and then again briefly in November 1975, in the midst of preparing two version of a monograph on the Tzintzuntzan migrants. The Spanish-language version (Kemper 1976)was published the Mexican Secretariat of Education, in the affordable SepSetentas series. This edition of 30,000 copies was distributed without charge to the nation’s schools and libraries and, over the years, has been frequently cited by Mexican social scientists.  In lieu of royalties, I was able to acquire 300 copies of the SepSetentas edition, nearly all of which I gave away to interested migrants and villagers during the following years.  The English-language edition (Kemper 1977) appeared a year later, but was declared “out of print” three years later by its trade publisher, and consequently made much less of an impact than did the Mexican edition.

I returned to Tzintzuntzan to repeat the village-level survey of migration and population changes in August 1976 and again in August 1978.  By that time, I realized that these biennial surveys also provided excellent preparation for the upcoming decennial village and migrant censuses.  So, through these rapid surveys, I was not only looking backward to the 1970 census, but also looking forward to the 1980 census and beyond.

SHIFTING PERSPECTIVES ON TZINTZUNTZAN MIGRANTS: THE 1980 CENSUS

In the academic year 1979-80, with a university research leave and a Fulbright research award in hand, Sandra and I spent nearly 15 months in Mexico.  Most of the time, we were in Mexico City, where we continued to live with the same migrant family, although now in a new house which they had built in 1976 on a nearby lot where Felipa’s mother had lived. Onto this new two-story house, I added a roof-top apartment financed with a $6,000 personal loan from the Dallas Teacher’s Credit Union.  The construction of this apartment – to which I had no legal title – reflected the strength of my long-term commitment to the research project and the willingness of Rafael and Felipa to have us living with them for the foreseeable future.⁵

In February 1980, I (and Stanley Brandes, another of Foster’s former students who continues to work in Tzintzuntzan) helped Foster to conduct the 1980 village census.  As we worked on the census, it became clear that the simple division of the population into residents and migrants was inadequate to represent a continually changing social and economic reality. A significant increase in four socio-economic categories – i.e., high school and university students, workers commuting daily to other localities, persons working full-time outside of Tzintzuntzan who returned from time to time to be with their families, and illegal immigrants to the United States – all combined to complicate the definition of who should be counted as “present” in the community.

Our solution to this ethnographic problem was to create an intermediate category which we called the “extended population.”  These people share one important characteristic: they are beyond the village much more than in it, and their eventual ties to the community are uncertain.  Thus, for 1980, we counted a “core” population of 2,506 persons in Tzintzuntzan and an “extended” population of another 143, for a total of 2,649 people.

The villagers of Tzintzuntzan provided us with a dramatic example of how they had “extended” their community when, in 1979, some 45 families invaded a strip of land belonging collectively to their own comunidad indígena (“indigenous community”).  Located about 2 km. south of the village along the road to Pátzcuaro, this new settlement soon took for itself the name “Colonia Lázaro Cárdenas” in honor of a former President of Mexico who, while governor of Michoacán state, had been instrumental in establishing the municipio of Tzintzuntzan in 1930.  Not knowing whether this colonia, composed entirely of people from Tzintzuntzan, would long survive, we chose to count its residents as continuing members of Tzintzuntzan proper.

During the rest of 1980, I worked (again with some assistants, including migrants themselves) with the migrants in Mexico City to carry out a parallel ethnographic census.  We gathered data on around 1,000 persons living in some 200 households dispersed around the metropolitan area (Kemper 1981a).  In addition, Juan Zaldivar, a school-teacher who, in 1975, had created and still served as the leader of the Tzintzuntzan migrant association, was eager to use the census for his purposes as well as for mine.  Therefore, on our behalf, he visited his fellow middle-class migrants not only in Mexico City, but also in several other cities, including Toluca, Morelia, and other smaller cities in the Lake Pátzcuaro region.

I obtained a small Ford Foundation grant so that two students in the graduate anthropology program at the Universidad IberoAmericana could do summer fieldwork with the Tzintzuntzan migrants. One student did some censusing in Mexico City, while the other (Victor Clark Alfararo), returned to his home town, the border city of Tijuana, Baja California, and subsequently accompanied me throughout California in pursuit of Tzintzuntzeños.  In addition, two European students – Beate Engelbrecht and Jorge Mächler – were doing fieldwork in 1980 in Tzintzuntzan as part of their professional training programs.  She did a dissertation (Engelbrecht 1987) comparing the situations of pottery-makers in Tzintzuntzan with those of the Sierra Tarascan village of Patamban, while Mächler did a brief study of the emerging Colonia Lázaro Cárdenas.

I expanded my own fieldwork to the United States, both to southern and northern California and to Illinois, especially to South Chicago and West Chicago.  This additional fieldwork yielded information on about 700 people living in some 150 households beyond what we had found in Mexico City.

One new component to the village work in 1980 was the creation of an album of several hundred black and white photographs which contained shots of every street and virtually every building in Tzintzuntzan, as well as those in Colonia Lázaro Cárdenas. This album has proven useful as an historical benchmark for measuring changes in housing and commercial activities, especially when used in conjunction with the detailed maps drawn of every block in the town’s grid plan.  I have continued to take photographs on each visit to Tzintzuntzan to document changes in housing and commercial enterprises (cf. Royce, in this volume).

In August of 1982, 1984, 1986, and 1988, I continued the biennial village-based survey of local demographic changes.  During the middle 1980s I also worked with Ben Wallace, a colleague at SMU, on an agricultural research project in Bangladesh.  There, I had the opportunity to study village-level migration in a quite different social and economic setting.  In fact, it was the research in Bangladesh that led me to develop the concept of the “extended community” (Kemper, Wallace, and Wilson-Moore 1989).⁶

By the end of the 1980s, my perspective on Tzintzuntzan migrants had shifted from Mexico City back to Tzintzuntzan, but also had expanded from rural-urban migration to international migration.  My initial short-term preoccupation with dissertation fieldwork (and then dealing with tenure and promotion issues) had been transformed into a long-term dedication to understanding migration as a major theme in the lives of the people of Tzintzuntzan.  I now conceived of the village as the central node (what some Tzintzuntzeños called el meollo, a word that literally refers to the brain, but which figuratively means the “essence” or “principal component” of something – in this case, a community) in a spatially and temporally complex social field and as a dependent node in an even more complex political-economic field.  I felt, and Foster agreed, that the three categories – resident, extended, and migrant – were simply three points along a continuum through which individuals and family units might pass during their life careers.

TAKING ON MORE RESPONSIBILITY: THE 1990 CENSUS

The municipal elections in late 1989 left Tzintzuntzan and many other communities in turmoil.  Although no one was killed in Tzintzuntzan, the representatives of the Partido Revolucionario Democrático (PRD) laid siege to the municipal offices to protest the ruling PRI party’s declaration of victory.  When Foster and I arrived in town in early January 1990, followed soon thereafter by Stanley Brandes and one of his graduate students (Sarah Miller), there was no competent municipal authority whom we could inform about our plans for carrying out another ethnographic census, as Foster had always done in the past.  Eventually, a “mixed commission” (composed of PRI and PRD members) was established to govern the Tzintzuntzan municipio, with the job of Presidente Municipal falling to a well-regarded Tzintzuntzeño who long had served as head of the local civil registry.  When, eventually, he was replaced in the civil registry by an outsider, we discovered that we no longer had direct access to birth, death, and marriage records in the civil registry.  Thus, since the early 1990s, we have had to rely on the records maintained in the parish church.  Unfortunately, however, these archives do not record information on deaths, this being a civil and medical matter.

Beginning in January 1990, and continuing for two months, I worked with Foster and the rest of our team to carry out the fifth ethnographic census of Tzintzuntzan.  Then, I continued with my own work on migrants in Mexico City, in southern and northern California, in Illinois, and – for the first time – in Washington state, an area with no Tzintzuntzeños in 1980 and more than 60 households in 1990.  In the absence of Juan Zaldivar (who had been killed in a car crash in the mid-1980s), I organized and trained several local people to pursue migrants in the municipio and throughout the rest of the state of Michoacán; in other Mexican cities such as Toluca, Guadalajara, and Tijuana; and in southern California as well.  At one time or another, three professional anthropologists (Foster, Brandes, and Kemper), six student anthropologists (Marianka Nathanson, from the University of Toronto; Sarah Miller and Matthew Gutmann, both from Berkeley; Socorro Torres Sarmiento, from UC-Irvine, plus Claudia Caro and David Diaz, both undergraduates from UC-Irvine), three non-anthropologist social scientists (arranged through a colleague at the Instituto de Investigaciones Antropológicas of the National University in Mexico City), and eight Tzintzuntzeños – a total of twenty individuals – helped with the census in the village and among the migrants.

Our collective efforts revealed that, in 1990, Tzintzuntzan had a “core” population of residents of 2,999 persons living in some 580 households, but also had 350 individuals (from 178 different households; i.e., 31% of the total) classified as “extended” population, for a total local population of 3,349.  As in 1980, we counted the population of Colonia Lázaro Cárdenas as a component part of Tzintzuntzan, although its more than 200 residents might better have been thought of as part of the “extended” community.  In 1991, another invasion occurred – across the road from the Colonia – involving perhaps another 20 families (some of whom are the married children of colonia residents).  By adding a second colonia, this one called Tzintzuntzita (“little Tzintzuntzan”), the villagers themselves continue to redefine the spatial boundaries of their community, in ways that would be difficult to comprehend through a one-shot, short-term investigation.

Living Arrangements: Problems and Resolutions

Unlike 1969-1970 and 1979-1980, my first priority for the 1990 census was Tzintzuntzan itself rather than the migrants in Mexico City.  So, I spent more time in the village than in the capital.  As luck would have it, in the late 1980s, Ramon Campuzano built a new house down the street from his parents’ house.  This permitted me to move back up to the roof-top apartment for the time of the decennial migrant census in 1990.  It also meant that there would be space available to bring Sandra for the first time since 1980 and our six-year old son John for the first time ever.  They came during the summer of 1990 and again in the summer of 1991, staying at the Campuzano home in Mexico City and with Micaela Gonzalez’s family in Tzintzuntzan.

I continued to have the use of the roof-top apartment at the Campuzano home in Mexico City until January 1996, even though Rafael had died in December 1991 and Felipa had been taking care of two of her grandchildren, while their parents – her oldest son (Roberto) and his wife (Lupe) – had been in Santa Barbara, California, for several years as undocumented workers.  Now, Lupe was returning to live at Felipa’s house and to care for her children.  This meant that there would no longer be a place for me, not even in the downstairs room – since Felipa’s college-age grandson would need that room, while Lupe and her daughter would require all of the roof-top apartment.  So, I left behind the furniture, transported clothes and supplies to Tzintzuntzan, and took the computer system back home to Dallas.

I was greatly saddened to think that, after more than a quarter of a century, I was being separated from one of my two field “families.”  Fortunately, this separation lasted less than two years.  Felipa was diagnosed with cancer in September 1997, and soon afterward we had a tearful reconciliation when we both happened to be in Tzintzuntzan to celebrate Micaela’s Saint’s Day later in that month.  As a result, I once again had a room downstairs, which I was able to use on my occasional visits to the capital throughout the year 2000 migrant census, in which Ramon and his two teenage boys and Ramon’s sister Edelmira and her teenage children all played significant roles as census takers.

Unfortunately, we all suffered a great loss when Felipa died on February 9, 2001, just four days after her 66^th birthday, from the effects of the cancer she had suffered for the past several years.  Her children are preparing to sell the house and divide the proceeds according to the terms of her last will and testament.  So, in March 2001, I traveled for one last time to the house that had become another home, once more gathering up all my clothes, supplies, and computer equipment and transporting them to Tzintzuntzan.  If the children of Ramon and Edelmira grow up, get married, and move out over the coming decade, perhaps a space will open up for me in one of their homes in time to deal with the 2010 migrant census.  But I know, as they too know, it will never be quite the same.

MAINTAINING THE “MASTER” DATA FILE: CHALLENGES AND OPPORTUNITIES

During the 1960s, Foster realized that the data in the 1945 and 1960 ethnographic censuses could be combined with information from other sources (e.g., the civil registry, the parish registry, genealogies) to develop a master data file – which came to be known as the fichero, the Spanish word for file or archive.  By the time I went into the field in 1969, the fichero was already a wonderful tool for checking all kinds of individual and familial information.  After I returned to Berkeley in 1970, Foster manually updated the fichero to include the 1970 village census data. For my part, I worked to create simple codes that could be used with the available mainframe computer programs of that pre-SPSS era.  I spent many late-night hours punching Hollerith cards at the campus computer center and generating vast quantities of green-bar printouts.

As a result of that work on the 1970, 1960, and 1945 village censuses, I began to see that managing the data for the Tzintzuntzan migrants had to be linked to the much larger issue of managing the huge quantities of information Foster was continuing to gather in the village.  In the run-up to the 1980 census, we were able to make use of Foster’s hand-written data extracted from the civil and church registers as well as biennial population surveys I did in 1974, 1976, and 1978.  The 1980 census went well, but the problem of data coding and analysis remained to be faced.  Foster continued with his manual updating of the fichero, while I used the mainframe computer at SMU to carry out basic statistical analysis of the household and individual data in the census.  By the mid-1980s, the first personal computers with enough memory and hard disk storage to handle database and statistical programs began to come onto the market.  Not long afterward, the first PC version of SPSS became available and I began to migrate the data files from the university mainframe to personal computers.

In preparation for doing the 1990 village census, I had a DOS version of the recently introduced Paradox database program loaded with the 4,000+ names in the master fichero as well as their basic socio-demographic data.  Similar database files were created for all individuals and households censused in 1945, 1960, 1970, and 1980.  After the 1990 census was completed, an updated Paradox version of the master fichero was created, now containing more than 5,200 individuals.  Foster and I worked together for a week at his home in Berkeley to manually update the individual data sheets and to photocopy all of these into a bound 8.5” x 11” format more suitable for carrying around in the field than the legal-sized post binders that contain the original household census data sheets.  Similar files were created for all persons and households in the migrant censuses.

Collecting Field Data: New Initiatives

During the 1990s, I returned to Tzintzuntzan on a regular basis and conducted the biennial census survey in August of 1992, 1994, 1996, and 1998, just as I had been doing since 1974, but with a new twist.  I found that I could use printed lists from the Paradox database to do the biennial surveys much more easily than in earlier decades – when I had read name after name, household after household, to my key informants who listened patiently for hours on end as we worked together through the entire census volume.  Now, I could leave a set of printed lists with spaces for my informants to fill in any changes that had taken place during the past two years.  Then, we could work over all of the “changes” to see if I understood them completely.  Then, back in Dallas, I had an assistant enter all of the changes into a census update file.  So, when I was preparing for the 2000 village census, I had an update file with all of the survey results for 1992, 1994, 1996, and 1998 to supplement the previous decennial census.

With the assistance of a few talented local residents, I also developed a project in the parish church to enter all of the baptismal, confirmation, marriage, and death records into laptop computer databases (using Microsoft Works in Spanish).⁷  This archival work, started in 1991 with the cooperation of the local priest, has been sustained until the present through continuing agreements with his three replacements. At this point, we are working to convert and combine a large number of Works databases into a handful of much larger Microsoft Access databases, including more than 40,000 baptismal records dating back to the late 1700s, plus confirmation records, marriage records, and death records (up until around 1960).

I plan to do a monograph on the historical changes in the local community, analyzing the parish records in the light of our ethnographic censuses, genealogies, and related information.  At an earlier time, Foster discussed doing such a work himself, but the amount of manual data processing required then would have been enormously burdensome. As the capabilities of personal computers and database programs improved steadily from the mid-1980s through the 1990s, the analysis of the parish archives now seems much more feasible.

COLLABORATIVE PROJECTS AND SECONDARY DATA ANALYSIS

One small-scale collaborative project took place in the 1980s, when Gabriel Lasker requested some data sets from Tzintzuntzan in order to compare names used in this lakeside community with names used in the Sierra Tarascan community of Paracho (Lasker et al. 1984).

During the 1990s, several anthropologists and journalists worked with the community and people of Tzintzuntzan.  Two dissertations were completed based on fieldwork in Tzintzuntzan or its migrants: the first, by Scott Anderson (1998) of the University of California (San Francisco) dealt with the problem of growing old in Tzintzuntzan; the second, by Socorro Torres Sarmiento (1999) of the University of California (Irvine) focused on the economic life of Tzintzuntzan migrants in southern California and in Washington.  Michael Shott (Shott and Williams 1997, 1999), an ethnoarchaeologist, included Tzintzuntzan in a regional project on pottery-making and pottery use.  Each of these anthropologists requested and received special data sets related to their particular interests.

In addition, anthropologist-filmmaker Jack Rollwagen came to Tzintzuntzan in 1991 to do videotaped interviews with George Foster about the history of his research and transformations in Tzintzuntzan, and with me about tourism.  Later, we did videotaping of the Campuzano family in Mexico City, among migrants in Chicago, and with Stanley Brandes at his Berkeley office. All of these taped interviews were assembled into “modules” for a videotape called Tzintzuntzan in the 1990s (Rollwagen 1992).  In 1993, Rollwagen returned to shoot additional footage on the February Señor del Rescate festival and then traveled with me to Tacoma where he videotaped migrants at work, at leisure, and in church.

For more than a decade, I have been collaborating with Professor Luis Alberto Vargas, a medical anthropologist at the National University in Mexico City.  To date, only one paper (Vargas, Kemper, and Casillas 1993) has resulted, but we are optimistic that we will be able to carry out a major research project in Tzintzuntzan related to medical anthropology.  At the same time, I have maintained an affiliation with several anthropologists at El Colegio de Michoacán, located two hours to the west of Tzintzuntzan in the city of Zamora.  I serve as a member of the advisory board for their journal Relaciones, have given seminars and taught courses in the graduate anthropology program when I am in the area, have been invited to participate in workshops and symposia (Kemper 1994b, 1995b) have sought out colleagues there when publishing opportunities arise (e.g., Roth-Seneff and Kemper 1995), and have established a personal book exchange with their library so that, eventually, they will receive upwards of one thousand volumes of books and journals from my collection.

In 1998, a special opportunity came to deal with some of the challenges inherent in the massive Tzintzuntzan data sets.  Doug White, who previously had worked with Scudder and Colson on their Gwembe data, wanted me to collaborate with him and a Swiss demographer colleague (Eric Widmer) in applying White’s social network analysis tools to the Tzintzuntzan data files.  We have spent some two years working together and look forward to continuing with the collaboration in years to come (see the Tzintzuntzan website, currently available at http://www.santafe.edu/tarasco/Mexican.html).⁸

BECOMING STEWARD OF THE TZINTZUNTZAN PROJECT: THE 2000 CENSUS

To become a steward is not the same as being the master of one’s own house.  In taking over the full responsibility for doing the 2000 village and migrant censuses, I was always aware that Tzintzuntzan is still the place made famous among anthropologists by my mentor-colleague George Foster.  Even though ill health prevented Foster from coming to the village, and Brandes’s teaching responsibilities kept him away as well, the sixth ethnographic census of Tzintzuntzan went forward as scheduled.

I had been planning for this occasion since 1990-1991, when I had taken my last university research leave, and so had saved two semesters’ research leave for the time of the 2000 ethnographic census.  Actually, I had known since September 30, 1994, that the responsibility for the 2000 census would fall on my shoulders – for that was the day when Foster handed over to me the metal file cases containing the original set of the master fichero.  Perhaps the moment was not as dramatic as when the Old Testament prophet Elijah passed his mantle to Elisha, as the story is told in 2 Kings 2:13, but I surely felt the “spirit” pass into my hands.

Taking up that calling in January 2000, I assembled a team of some thirty local residents, whose work with the census I would coordinate, with help from Peter Cahn (see his chapter in this volume).  In the next few months, our team visited more than 700 households, inhabited by 3,346 residents and another 264 “extended” individuals, yielding a total population of 3,610 for Tzintzuntzan proper and its colonias (Lázaro Cárdenas, Tzintzuntzita, and the newest one, called San Juan).  After censusing the nearby hamlets of Ojo de Agua and Ichupio for the first time since 1945, team members moved on to find migrants throughout the Lake Pátzucaro area, in Morelia, and in other towns and cities in Michoacán. We also gathered data on migrants in Mexico City and in Guadalajara, as well as in California, Illinois, and the state of Washington, which now claims the largest enclave outside of Mexico – with perhaps as many as 200 families!

With respect to people living beyond the boundaries of Tzintzuntzan, we again have censused more than 3,000 migrants (including spouses, children, grandchildren, and great-grandchildren) in dozens of different localities.  Simply put, about half of the people of Tzintzuntzan are now living in and around the village and the other half are living temporarily or permanently removed from their natal community.  This is a dramatic shift from the situation in 1970, and even more striking when compared with the circumstances in 1945, when Foster first did fieldwork in Tzintzuntzan.

TZINTZUNTZAN AS EXTENDED COMMUNITY

There are certainly hundreds, if not thousands, of community studies in the anthropological literature.  Most of these works, especially those from earlier times, are focused tightly on a single place in time and space.  Our continuing long-term fieldwork among the people of Tzintzuntzan suggests that we need to rethink our notion of community (cf. Foster, in this volume, for a similar perspective).

First, I am struck particularly by how hard it has become to define who is and who is not a resident of Tzintzuntzan. As we have seen, a  growing proportion -- currently more than 30% -- of the local households have one or more “extended” members away from the village at any given moment, yet the families tend not to think of these persons as being migrants. In addition, given the flow of individuals and families from and back to the village in accord with academic schedules, vacations, and the planting-harvesting cycles in US fields, the population is always in flux.

Second, distance is no longer the significant factor that it used to be.  The transportation infrastructure permits relatively inexpensive travel by bus, private car, or even airlines to all parts of Mexico and to the United States. In recent years, several young adults who died from car accidents in the United States have had their bodies sent back by air to Guadalajara, thence by bus or family-owned truck to Tzintzuntzan for funeral and wake services.

Third, the limited access to telephone service has been upgraded to fully automatic service (now that Teléfonos de México has been privatized) so that rapid voice communication is replacing the slower and somewhat unreliable postal and telegraph systems as a preferred means of contact between the village and elsewhere.  A recent effort to sign up new subscribers for  the automatic direct dialing system has resulted in more than 100 households being subscribers – although they are listed in the phone book of the state’s capital city as if they resided in nearby Quiroga.

Fourth, the increased participation of the young people of Tzintzuntzan in the higher educational system of Mexico means that many of them are away from the village for several years, except for vacation periods or weekends.  Still, they continue to return to the community even after settling in urban areas where their hard-won job skills can be applied.  For instance, one medical doctor has his practice in the state capital of Morelia but insists (even over his non-Tzintzuntzan wife’s objections) on maintaining a weekend practice in Tzintzuntzan.  More than 150 persons from the village have become school teachers, which is remarkable enough, but more interesting is that most desire to practice their profession with Tzintzuntzan as a base of operation.  Thus, they live in the village and commute daily to other localities or even come back just on weekends if the distance is too great for commuting.  The local primary and secondary schools are currently staffed almost entirely by teachers living in Tzintzuntzan.

Fifth, earnings and remittances from persons beyond the immediate community have become vital to the continuing economic survival of Tzintzuntzan and hundreds of similar Mexican villages.  Since 1970, the Mexican peso has gone from a fixed value of 12.5 to the dollar to about 3,000 to the dollar by 1990, then suffered a 1000 to 1 “adjustment,” and has continued to decline to more than 9.0 new pesos to the dollar at present.  As a result, the shift of the extended community toward the United States – and away from Mexico City – has become marked.  When men fail to send money or do not return after the fall harvest season in the United States, their spouses and children suffer serious deprivations.  On the other hand, temporary movements away from the village can be used to supplement inadequate local wages.  Even some school teachers travel to California or Washington to work in the fields during the summer vacation period.  There they can earn the equivalent of their entire year’s salary in a couple of months of hard labor.

Sixth, inheritance is an important issue for residents and their relatives living away from Tzintzuntzan.  A number of older migrants – some away from the village for more than forty years – still own houses or land there.  Will they leave this property to their children, some of whom have rarely traveled to Tzintzuntzan, or will it pass to family members still living in the village?  On the other side of the equation, many younger migrants are concerned to establish their claims to a share of the family lands by continuing to visit Tzintzuntzan on a regular basis.

Finally, Tzintzuntzan is well-known for fiestas and religious celebrations (Brandes 1988). Many migrants – as well as thousands of other tourists – come back to the community to participate in these festivals.  Especially at the February festival, in honor of the Señor del Rescate, many migrants return to baptize or confirm their children in the parroquial church.  Often, they select compadres (co-godparents) from village residents or from other returned migrants, thus emphasizing their solidarity with their natal community.  This sense of membership and affiliation also applies to the people of the hamlets surrounding the village proper.  Conveniently, the main elements of the annual fiesta cycle occur at times when Tzintzuntzeños away from the village can return as participants or observers.  For example, one of the villagers responsible for the Holy Week celebrations is currently living in Orange County in southern California.  He takes his two-week vacation to coincide with his duties in Tzintzuntzan. But since so many migrants now live in the Tacoma area, and a considerable number cannot return yearly to Tzintzuntzan, these migrants have begun to celebrate their own version of the Señor del Rescate fiesta each February!

The concept of the “extended community” has emerged over several decades during our work with the people of Tzintzuntzan. When Foster began work in Tzintzuntzan in 1945, it had only 1,231 residents, nearly all of whom were born in and lived most of their lives in what then appeared to be a “closed corporate community.”  By comparison, since I began work there in 1969 more than 4,000 individuals have been added to the master data file.  Time has transformed who “counts” in the village population, just as it has changed the make-up of the research team involved with Tzintzuntzan and its “extended community” – all the people who consider themselves to be Tzintzuntzeños wherever they may be, whether in the town itself or as far away as Tacoma, Washington.

IN PLACE OF A CONCLUSION

This chapter is no more than an “interim report” on my role in the Tzintzuntzan project.  What will the situation be like in the year 2010 and beyond?  By then, we will be working with the grandchildren and great-grandchildren of adults known to Foster in 1945.  At this point, Foster is no longer able – at age 87 – to travel to the village, but Brandes and I maintain frequent contacts with the community, and a third generation of fieldworkers is represented by Peter Cahn, who has completed a dissertation (Cahn 2001) on religious change based on his recent year-long study in Tzintzuntzan.  His chapter follows this one and provides an appropriate continuation (and new beginning) to the story of Tzintzuntzan and its anthropologists.

If Tzintzuntzan continues to develop as it did during the 1970s, the 1980s, and the 1990s, the challenge to our theories and methodologies will be significant.  What once was treated -- by villagers and anthropologists alike -- as if it were a “closed” system has become a spatially and temporally “extended community” whose changing characteristics cannot be ignored.  Indeed, for the migrants and their children returning to what some of them have jokingly identified as “Tzintzun-landia” (a play on “Disneyland,” of course), the community is both “home” and “re-creation” site where they, like other tourists, can encounter Purépecha [now the preferred term for Tarascan] and Spanish colonial culture in twenty-first century dress.  Even more important, they encounter their own cultural and social circumstances being transformed before their very eyes.⁹  Through our long-term field research, we have seen how the concept of “community” involves not only the sense of physical place (pueblo) but also the commitment to common identity and values (communidad indígena) regardless of whether the people of Tzintzuntzan are physically resident in the town or living elsewhere.  This is a community still tied to pre-Conquest legends  and more than 400 years of colonial and post-colonial history, caught up in government-sponsorsed tourist campaigns to retain its traditional facade, but also striving to cope with modern life in a global system represented by satellite dishes, border crossings, and transnationalism – a world in which the community of Tzintzuntzan find itself on the front page of The Washington Post.

ACKNOWLEDGEMENTS

The initial fieldwork in Mexico City during 1969-1970 was funded through NIGMS Training Grant GM-1224, which also covered the earlier summer field training in Pátzcuaro as well as the analysis and write-up of field materials for the dissertation.  Additional financial support was received from the Center for Latin American Studies (University of California at Berkeley) for dissertation write-up during the months of July-September 1971.  The Wenner-Gren Foundation for Anthropological Research provided a Grant-in-Aid (No. 3027) for the summers of 1974 and 1976 and for preparation of the Spanish- and English-language monographs based on that fieldwork.  In the periods July 1979-June 1980 and July-December 1991, fieldwork was supported by Fulbright-Hays Advanced Research Awards (American Republics Program).  The Ford Foundation (Office for Mexico and Central America) provided a small grant to support two Mexican graduate students on the project in 1980.  I also have received continuing support from Southern Methodist University through Faculty Research Fellowships and Awards for the periods August-December 1979, June 1986-July 1987, January-June 1990, and August 1999-May 2000. In addition to these “standard” sources, a few private individuals have provided the project with significant funding, which I have matched through my own personal financial contributions.  To all of these institutions and individuals go my most sincere thanks for their continuing support.

Of course, there would be no Tzintzuntzan project at all if George Foster had not blazed the trail that I and others have traveled more easily because of his decades of dedicated work.  He and his wife Mickie, their children, and grandchildren have been gracious, generous, and welcoming for more than thirty years.  In fact, since 1967, I have spent more time with the Fosters in the home of our extended Tzintzuntzan “family” – with Doña Micaela González, her daughters Virginia and Lola, and their friend María Flores – than with the families of my parents, siblings, and in-laws!  Perhaps, the people of Tzintzuntzan who ask me if I am “un hijo de Foster” (“Foster’s son”) understand something profound about our relationship, and the sense of gratitude I owe to the only anthropologist in Tzintzuntzan who ever will be known, respected, and loved as “el Doctor.

NOTES

1.  As I was completing my summer’s research in Pátzcuaro, John Durston arrived on the scene with the intention of studying the regional economy.  I shared with him some of my field notes and sketch maps.  His monograph (Durston 1976) stands as the last point in the systematic study of the Pátzcuaro marketplace.

2.  It is a curious coincidence that, at the time when I was studying the Tzintzuntzan migrants in Mexico City, the community of Cherán was the focus of an intentional restudy designed to examine acculturation and change since Beals’s original study in 1940. In discussing the “adaptation” of this “traditional” community, Castile (1974:139-141) commented on the continuing importance of temporary labor migration to the United States.

3.  Among the students in the urban anthropology seminar was Larissa Lomnitz (who earlier had taken undergraduate work in anthropology with Foster at Berkeley), who not only gave me valuable assistance with the Tzintzuntzan migrant project, but also provided me with numerous opportunities to meet the many Mexican and international anthropologists who also enjoyed her gracious hospitality in Mexico City.

4.  These findings on compadrazgo have been widely disseminated through the Annual Editions volumes, Global Studies: Latin America, edited by Paul Goodwin (Ninth edition, 2000; Guilford, CT: McGraw-Hill/Dushkin).

5.  By 1983, Ramon Campuzano had married and needed a place to live.  We worked out an arrangement by which I “sold” him my interest in the roof-top apartment.  Thereafter, I was given the use of a room downstairs, whenever I happened to be in Mexico City.  As a result, I took some furniture and equipment to Tzintzuntzan, sold Ramon other pieces, and spent increasingly more time in Tzintzuntzan than in Mexico City during the rest of the 1980s.

6.  It is surely a curious coincidence that Foster recognizes the “trigger effect” of having read Honigmann’s article about a village in West Pakistan in the formulation of the “Image of Limited Good” and that my ideas about the “extended community” were similarly inspired by fieldwork in Bangladesh (formerly East Pakistan) and the reading of relevant literature, including Morris Opler’s classic article on “The Extensions of an Indian Village” (1956).

7.  Back in the late 1960s and early 1970s, Foster had copied many of the church (and civil) records by hand or by carrying the large record books to Ajijic, located on Lake Chapala, in Jalisco state.  There he had access to a photocopy machine.  Unfortunately, the cost of processing these handwritten copies or photocopies into a computer database was prohibitive.  It was much less expensive simply to enter all of the church records into a computer database program.  After ten years of data-entry, virtually all of the two hundred years’ of more than 50,000+ existing records have been logged into a computer by Lucila Marín (who also is an assistant to the priest in charge of the parish).

8.  In addition, I have acquired and registered the domain name tzintzuntzan.org as the permanent home of the Tzintzuntzan Project on the Internet.

9.  In addition to works focused on migration, I also have examined Tzintzuntzan’s transformations in other domains, including tourism (Kemper 1979d), urbanization and development (1981b), the production, distribution, and consumption of foodstuffs (1996b), changes in the use of the Tarascan language (1998), and manipulation of ethnic identity (Royce and Kemper n.d.).

REFERENCES CITED

Anderson, Scott T. (1998) “Growing old in Tzintzuntzan.” Ph.D. dissertation. University of California at San Francisco.

Beals, Ralph L. (1946) Cherán: A Sierra Tarascan Village. Smithsonian Institution, Institute of Social Anthropology, Publication No. 2. Washington, D.C.: U. S. Government Printing Office.

Berreman, Gerald (1962) Behind Many Masks: Ethnography and Impression Management in a Himalayan Village. Society for Applied Anthropology, Monograph No. 4. Ithaca, NY: Cornell U