COMPADRAZGO IN CITY AND COUNTRYSIDE: A COMPARISON OF TZINTZUNTZAN MIGRANTS AND VILLAGERS Robert V. Kemper INTRODUCTION As the "first distinctive aspect of Latin American social structure to be discovered by anthropologists" (Foster, 1969:263), the comnpadrazgol has long been an important topic of ethnographic research. Until recently, however, nearly all our knowledge about this widespread form of fictive kinship has come from studies of small rural communitites. Indeed, as Thompson (1970:288) has suggested, the assumed relative unimportance of compadrazgo in urban settings is in some degree an artifact of the discipline's traditional emphasis on fieldwork in villages, hamlets, and dispersed rural populations. As more research on the compadrazgo has been done, and more urban groups studied, anthropologists have revised the initial hypothesis (Redfield, 1934:65) that urbanization operates to simplify and disorganize fictive kinship relationships. For example, rather than arguing for a devolution of the compadrazgo in urban setting, Ravicz observes that it "may withstand the onslaught of modern market and secular conditions, its form signaling the absence of other systems to satisfy needs created by such conditions, as well as mirroring the nature of the needs" (1967:251). Similar views on the compadrazgo's viability and flexibility are expressed by other recent commentators (Berruecos, 1974; Carlos, 1973; Carlos and Sellers, 1972; Foster and Azer, 1970; Thompson, 1973; Wilson, 1969). Though its structure, functions, and symbolism may be modified outside the rural settings in which it was first described by anthropologists, the compadrazgo has proven to be a durable and pervasive social, economic, and religious institution. THE COMPADRAZGO IN MEXICAN CITIES The available ethnographic literature on the compadrazgo in Latin American cities is still conceptually fragmentary and geographically uneven. Nevertheless, the appearance of several recent studies of social organi- zation in Mexican cities makes it possible to assemble a set of fieldwork-based data covering large metropolitan areas, middle-sized cities, and small towns spread throughout the nation.2 Of the groups studied, five are located in Mexico City, two in the city of Oaxaca, and the other eight in Ciudad Juarez (Chihuahua), Uruapan (Michoacan), Queretaro (Queretaro), San Cristobal de las Casas (Chiapas), Juchitan (Oaxaca), Ticul (Yucatan), "Ciudad Industrial" (Hidalgo), and Tonall (Jalisco). The places range from Mexico City, with a 1970 popula- tion of 7,100,000, to Tonala, with a 1960 population of 5,000. The communities are dispersed throughout Mexico and include both mestizo and indio populations. These fifteen case studies emphasize working-class groups, but several of them also treat middle- and upper-class populations. Most of the investigators use neighborhoods as the unit of analysis, while some deal with entire communities, and others examine migrant populations dispersed throughout a metropolitan area. The size of the populations studied ranges from fourteen families to 475 households; the number of compa- drazgo relationships reported ranges from an unspecified small sample to a sample of 1,291. Appendix 1 provides detailed information on the sources used in the comparative analysis. Acknowledgment: Fieldwork in Mexico City and analysis of census data and ethnographic materials have been supported by NIGMS Grant GM-1224 and by grants-in-aid from the Wenner-Gren Foundation for Anthropological Research, Inc. Their assistance is gratefully acknowledged. This article was submitted for publication in November 1976. 25 KEMPER TYPES OF COMPADRAZGO IN MEXICAN CITIES Table 1 shows that the occasions on which people in urban places seek ritual sponsors are far more numerous and diverse than earlier theorists supposed. Of the fifteen ethnographic case studies analyzed, only eleven provide specific information about the range of compadrazgo types. The others merely say that the compadrazgo is "important" or is "widespread". The ceremony of baptism is clearly the most frequent and important occasion for selecting godparents. The other life cycle rituals-confirmation, marriage, and first communion-celebrated within the Catholic Church also represent significant opportunities for establishing compadrazgo bonds. The number of minor "spiritual" compadrazgo types is large, but there is considerable variation among the cases as to the combina- tions in which they occur. In addition to these religious types of fictive kinship, urban residents also practice a variety of "secular" forms of the compadrazgo. The quince an*os party for girls requires a madrina as sponsor; school graduations, both primary and secondary, offer opportunities for children to choose a godparent (madrinas for girls, padnnos for boys). Other, "less serious" forms of compadrazgo include: blessings of businesses, houses, and cars; a boy's first haircut; and even mock weddings at church bazaars. Many of these "secular" compadrazgo ties are ephemeral, but sometimes they may lead to more significant, long-term relationships cemented by baptismal or other sponsorships. The number of occasions on which urban Mexicans may elect compadres is hardly less restricted than that for rural Mexicans (for data on rural compadrazgo in Mexico, see Berruecos, 1974; Carlos, 1973; Davila, 1971; Ravicz, 1967). This observation is valid for large and small localities, as well as for a broad range of socio-economic strata within them. Far from reverting to the limited Spanish form3 (Foster, 1953:26), the compadrazgo appears to be an arena for social innovation in urban life. In addition to the basic religious occasions, urbanites may seek ritual sponsors for a variety of secular and minor spiritual events. At least 19 different types of compadrazgo relationships are reported in the case studies. Clearly, the compadrazgo must be conceived of as an institution which permeates all levels of Mexican society. THE COMPADRAZGO AMONG TZINTZUNTZAN VILLAGERS AND MIGRANTS Tzintzuntzan is a Spanish-speaking mestizo village on the shores of Lake Patzcuaro, 230 miles west of Mexico City. In 1970, the community had a resident population of 2,169 and an emigrant population, spread over much of Mexico and into the United States, of 695 persons (Foster, 1967; Kemper and Foster, 1975). Its economy, still based largely on pottery production and sales, has expanded sufficiently to provide many villagers with living standards once unattainable. And, perhaps more important, its inhabitants now realize that the world they live in goes far beyond the local community. The data used in this comparison of villagers and migrants are drawn from Foster's analysis of baptismal compadrazgo (1969), from his general observations on the structure and functions of godparenthood (1948, 1961, 1963, 1967), and from my own research among the migrants living in Mexico City. For the quantitative aspects of the comparative analysis, I have used a set of 788 baptismal cases covering the years 1958-1967 for Tzintzuntzan and a set of 106 baptismal and 57 confirmation cases which occurred among the migrants in the capital. A comparison of these two data sets should indicate the extent to which the compadrazgo changes as a result of urbanization. 26 TZINTZUNTZAN COMPADRAZGO 27 TABLE 1 Distribution of Compadrazgo in Mexican Cities CITIES STUDIED1 Ln C4~~ CJ - .0 ON 0 n *.*4 -r4 % 4.1 00 0 en) .94 'P4 r '4 co 0 o o 0 ..0414 .100 0 o u 0. U. -4 U P4 'G - 0 (. "4 'P4 0 co 0 .C S r 0) 0) S.~~ 0 0 CO0 0 - 0 COMPADRAZGO TYPES2 | x | X 0 CY a o| 2| Total baptism + + + + + + + + + + + 11 confirmation + + + + + + + + + + 10 marriage + + + + + + + + + + 10 first coummunion + + + + + + + 7 blessings + + + + 4 Child Jesus + + + + 4 school graduations + + + + 4 de la corona + + + 3 evangelios + + + 3 15th birthday (girls) + + + 3 Saint's day + + 2 burials + + 2 penance + + 2 medidas (blessed ribbon) + 1 tertulias + 1 priesthood ordination + 1 first haircut (boys) + 1 mock weddings 1 /heetzmeek/ Maya (boys) + 1 Total 114 ( 9 [ 8 [ 8 [ 7 | 6 I 4 | 4 | 2 | 2 | 71 Notes 1. The number after each city entry is keyed to Appendix 1. Four ethnographic sources do not provide specific data on the types of compadrazgo and these (4, 6, 7, 14) are thus omitted from the table. 2. The following descriptions of less common compadrazgo types may be useful: blessings or consagraciones may be done for houses, cars, businesses, etc.; Child Jesus (levantar el nifio, del nifio Jesus, arrullo del nifno) involves blessing an image of the Christ child during the Christmas season; de la corona ("crown") involves blessing a child and buying a toy; evangelios involves prayers for a sick child or to prevent a child from becoming ill; medidas involves tying a blessed ribbon on a child to prevent disease; tertulias involves social drinking bonds;/heetzmeek/is a Maya cere- mony which introduces a boy to the tools of his adult life. KEMPER TYPES OF COMPADRAZGO: VILLAGERS AND MIGRANTS In his first discussion of compadrazgo in Tzintzuntzan, Foster stated that there were four types: baptism, confirmation, marriage, and de la corona. He did not mention first communion or any other religious or secular forms of ritual sponsorship (1948:262-264). Subsequently, in his well known paper on the dyadic contract, he discussed the compadrazgo at length but mentions only four occasions on which godparents are selected: baptism, confirmation, first communion, and weddings (1961:1181-1182). More recently, in his monograph Tzintzuntzan: Mexican Peasants in a Changing World, he observes that there are five traditional occasions on which compadrazgo ties are formed: "the baptism of a child, its confirmation, its first commun- ion, its marriage, and the rite known as de la corona. Although the first four are associated with the sacra- ments, and the fifth is validated by a minor religious act, only the first are de grado i.e., of the first importance" (1967:76). In addition to these fundamentally "sacred" types of compadrazgo, some Tzintzuntzeflos also partici- pate in more mundane forms of ritualized social relationships. Since about 1960, when a new school director came to the community, children have selected adults to serve as padrinos for sixth-grade graduation. This is a short-term tie which compels the sponsor to provide a gift and, in more affluent families, a festive meal in honor of the child. More recently, a few young women have taken up the custom of celebrating their fifteenth birthday with a fiesta. The girls' parents are responsible for the fiesta, but a madrina is selected and is expected to provide a nice gift. It appears that the first quince aflos fiesta was held in 1969 in honor of a girl bom to Tzintzuntzefios living in Mexico City. Since then, at least eight others, in nearly all cases from the most affluent families in the village, have had fifteenth birthday parties and chosen madrinas for the occasion. Two other ephemeral types of the compadrazgo also occur in Tzintzuntzan: some persons select compadres for blessings of various kinds (e.g., houses, cars, animals) and, at the time of weddings or similar celebrations, some villagers take compadres de chingere. (This latter form is an emotional, personal tie based on the temporary state of high spirits present at such events and is usually treated as an ongoing "tu-tu" relationship rather than a serious formalized "Usted-Usted" exchange.) Both in the case of blessings and de chingere ties, the persons involved may transform their quasi-bond into a stronger compadrazgo tie by selecting one another on more important occasions (e.g., baptisms). The last decade has seen an elaboration of the wedding compadrazgo to include several minor ties. The major sponsors are, as always, the padrinos de velaci6n; lesser sponsors serve as madrinas de lazo, de arras, de ramo, and de anillos (i.e., for the cord placed over the bride and groom to signify their sacred union; for the token coins given by the bride to the groom so that the couple will never be poor; for the spray of flowers carried by the bride; and for the rings exchanged by the couple). In spite of these recent elaborations in the local compadrazgo system in Tzintzuntzan, the village has a less developed system of ritual kinship than in many other Mexican communities. For example, Nutini (1976) reports 31 different types of compadrazgo for cummunities in rural Tlaxcala. COMPADRAZGO TYPES AMONG THE MIGRANTS The Tzintzuntzefios in Mexico City represent 35 percent of all emigrants from the village; they are by far the largest concentration in any one locale. In 1970, when the village had 360 households, there were at least 74 migrant households in the capital with a total population of 483 persons. This migrant group may be divided into the following categories: 246 "full-time" Tzintzuntzan migrants, 38 "part-time" labor migrants, 39 spouses bom outside Tzintzuntzan, 116 children bom in the capital, 24 children bom elsewhere, 10 non-Tzintzuntzan in-laws, and 10 non-Tzintzuntzan non-kinsmen. Within this diverse migrant population, the types of compadrazgo may be more numerous than in the village. Baptism, confirmation, first communion, and marriage are standard among the migrants. Girls' fif- 28 TZINTZUNTZAN COMPADRAZGO teenth birthday parties, school graduations, the Child Jesus, blessing of an image at church, and blessing of a new home also provide occasions for establishing compadrazgo ties. The explanation for this expansion of fictive kinship types seems to lie in the increased contacts of Tzintzuntzan migrants with other subcultural traditions. When the migrants move to the capital they do not reside in tight clusters but spread themselves among more than forty neighborhoods. This geographical dispersion gives them an exposure to other uses for the compadrazgo beyond the basic types they knew in Tzintzuntzan. As they are asked to serve in these ritual roles, they, in turn, begin to use these sponsorships for building affective and instrumental ties with migrants and native alike. Tzintzuntzan migrants are not just receptors of new compadrazgo forms; they also serve to disseminate them to the village. As we pointed out above, in recent years, it has become fashionable for Tzintzuntzan children to select padrinos for primary school graduation and for girls to celebrate their fifteenth birthdays with parties sponsored by a madrina. While these forms of compadrazgo are not de grado, they do show the influence of urban forces on an institution once conceived of by anthropologists as preeminently rural and traditional. THE CHOICE OF COMPADRES: RELATIVES VERSUS FRIENDS In Tzintzuntzan, according to the statistical analysis of 788 cases of baptismal compadrazgo for 1958-1967, 67 percent of compadres are friends and only 33 percent are relatives. The choice of paternal or maternal lines is about even: 17.4 percent for the former and 15.6 percent for the latter. When a relative is selected as a compadre, there is a "greater than 50 percent chance that he or she will be a sibling of one of the parents" (Foster, 1969:270). The next most common choice (16 percent) is a parent's first cousin. Foster notes that, in comparison with compadrazgo choices made in the 1920's, the villagers show an increasing preference for approximate age mates as compadres. Within the godparent pairs, the great proportion (78 percent) are married couples. Brother-sister pairs are next most common (12 percent), followed by god- mothers alone (8 percent). Other combinations (fathers/daughters, mothers/sons) represent the remaining cases (Foster, 1969:27 1). When birth order is taken into account, the proportion of friends versus relatives selected as compadres declines from 81 percent for the first child to 57 percent for the fifth child. By the ninth child, the percentage has risen to the 67 percent average for the entire sample, and for subsequent children the curve rises to the 81 percent level (Foster 1969:272-273). Foster's explanation of this pattern of compadre choice emphasizes the life cycle and post-marital residential patterns of patrineolocality: "By usually selecting friends for compadres for the first several children, social networks are rapidly expanded, and at the same time major threats to the sanctity of the institution from joint family friction are avoided. After several children most young couples live alone. Their social networks are approaching optimum size, and they may wish to reduce baptismal expenses. They can now meet both situations by selecting compadres from among close relatives, including parents and siblings in parental households, since the tension-producing intimacy of the first several years of married life no longer exists" (1969:277). COMPADRE CHOICES AMONG THE MIGRANTS Tzintzuntzan migrants in Mexico City select relatives as godparents less frequently than do the villagers. For baptisms, they prefer friends 77 percent to 23 percent; for confirmations, the figures are 70 percent to 30 percent. If we combine the baptismal and confirmation data sets, we find that the migrants in the capital select friends three times as often as relatives (75 percent to 25 percent). Table 2 shows the differences between migrants and villagers in the selection of compadres. 29 KEMPER TABLE 2 Choice of Friends or Relatives as Compadres Population Compadre Choice Villagers' Friends 524 122 646 Relatives 264 41 305 Total 788 163 951 X = 4.338, P < .05 (1 df) Notes 1. baptismal compadres 1969:269) only, for 1958-1967 (Foster 2. baptismal and confirmation compadres; baptismal data set (n = 106 cases, of friends and 24 relatives), x2= 5.047, using only the which 82 are P < .05(1 df) The increased proportion of friends among the migrants is due to the relatively small number of kinsmen in the capital and the geographic distribution of those who are available as potential compadres. Moreover, since heavy emigration from Tzintzuntzan to Mexico City is fairly recent, few migrants have members of the first ascending generation (i.e., parents, uncles, aunts) with them in the capital. Since this set of kinsmen represents 18 percent of all baptismal godparents in the village, it is clear that the migrants have fewer chances to select kinsmen. When relatives are chosen as compadres, they are evenly divided among the husband's and wife's families. This is similar to the pattern in Tzintzuntzan, and presumably reflects the essential bilaterality of the Mexican kinship system. As in the village, the most common choice among relatives is siblings of the child's parents (20 of 41 cases), the next most frequent are first cousins (7 cases) and the category "other distant Idnsmen" (7 cases), followed by parents (4 cases), nephews/nieces (2 cases), and uncles/aunts (1 case). Also as in Tzintzuntzan, within godparent pairs (both relatives and friends), the great proportion (78 percent) are married couples. In these three domains, migrants and villages use the compadrazgo in almost identical fashion. Migrants2 Total 30 TZINTZUNTZAN COMPADRAZGO Although the migrant sample is too small to make statistical comparisons, it is possible to observe that birth order of children does influence the choice of friends and relatives as godparents. Indeed, the pattern is much the same as in the village. Among the migrants, for a sample of 73 cases for "own" children the proportion of friends declines from about 80 percent for the first child to 50 percent for the sixth and seventh children. I cannot say if the proportion of friends rises thereafter, since none of the families for whom I have data have more than seven children. By moving to Mexico City, the migrants are adopting an extreme form of neolocality, which may follow upon a patrilocal and/or neolocal period of residence in the village. Single migrants (both men and women) who marry after leaving Tzintzuntzan proceed directly into neolocal residence unless they elect to live next door to (or nearby) one of their few relatives in the city. (It should be noted, in passing, that as the current generation of migrant children grows up, the pattem of patrineolocality may reemerge as a strategic adaptation to the availability of acceptable housing in the capital). Under prevailing circumstances, the choice of friends and relatives cannot be satisfactorily explained in the case of the migrants by post-marital residence rules. Instead, we must consider the strategy of building social networks in an urban area which contains few persons from one's home community. Whereas in Tzintzuntzan 87 percent of all baptismal godparents are fellow villagers, among the migrants in Mexico City only 34 percent of baptismal and confirmation godparents are from Tzintzuntzan. From the villagers' view- point, the compadrazgo serves to intensify social bonds with kinsmen (and persons who are already compa- dres); it extends ties to unrelated villagers and to persons outside Tzintzuntzan. In contrast, the migrants use the compadrazgo to intensify relations with the village and with fellow migrants. This intensification includes Idnsmen, compadres, and friends. Extension of social relationships means establishing ties beyond the migrant group and thus brings the Tzintzuntzefios into the social orbits defined by neighborhoods, jobs, associations, and other urban socio-economic categories. The extent to which the migrants use the compadrazgo mechanism to build bridges to non-migrants is shown in Tables 3 and 4. THE RELATIVE STATUS OF COMPADRES: VILLAGERS AND MIGRANTS In Tzintzuntzan baptismal godparents usually have the same socio-economic status as the child's parents (75 percent). If unequal, they are likely to be of higher status than the parents (21 percent). Among the migrants in Mexico City, only 56 percent of the baptismal and confirmation godparents are of the same socio-economic status. Thirty-one percent are of higher status and 12 percent are lower. Both of these vertical dimensions of the compadrazgo are consequently accentuated in the urban setting. As Tzintzuntzan develops a more elaborate class structure (Kemper and Foster, 1975:74), the vertical aspects of the compadrazgo within the community will also become more important. For the present, it is clear that the migrants in Mexico City use the compadrazgo (both in asking others to serve and in accepting the requests of others to serve) as a device for establishing patron-client ties. Usually, the Tzintzuntzefilos in the capital try to improve their opportunities for upward mobility by asking non-Tzintzuntzan urbanites, whereas most ties among the mi- grants are between persons of similar status. In this respect, they behave like their village counterparts, who emphasize same-status ties within the community and higher-status ties with outsiders. Table 5 provides statistical data on the choices made by migrants and villagers for higher versus same/lower status compadres. INTERPERSONAL RELATIONS AMONG COMPADRES In addition to providing quantitative data on the types and choices of compadrazgo in Tzintzuntzan, Foster offers valuable observations on the difference between ideal and real behavior between compadres and -on the functions of the compadrazgo within and beyond the community. From the data gathered during more than a 30-year span of fieldwork, he concludes that "far from always being a close and sacred relationship, compadre ties often are routine in the extreme. They may become tenuous, and sometimes they are broken to 31 32 KEMPER TABLE 3 Provenience of Compadres: Villagers and Migrants Population Provenience Villagers' Migrants2 Total Tzintzuntzan born 683 55 738 Born elsewhere3 105 108 213 Total 788 163 951 X2= 187.710, P<.001 (1 df) Notes 1. baptismal compadres only, for 1958-1967 (Foster 1969:268) 2. baptismal and confirmation compadres; using only the bap- tismal data set (n = 106 cases, of which 36 are Tzintzuntzan born and 70 born elsewhere), x* 164.906, P<.001 (1 df). 3. provenience of compadres in the village sample is defined as "within 10 mile radius" (n = 57) and "beyond 10 mile radius" (n = 48); for the migrant sample, provenience is defined as "Mexico City born" (n = 50); and "born elsewhere" (n = 58). The latter group is composed of compadres from Durango, Guanajuato, Guerrero, Hidalgo, Jalisco, Mexico, Michoacan, Morelos, Oaxaca, Puebla, Queretaro, Tlaxcala, and Zacatecas -- i.e., all states in the central region of the nation. the point where compadres do not even speak to each other" (1967:80). He notes that perhaps 15 or 20 percent of compadres continue to use the informal second person personal pronoun tu in addressing one another, despite the theoretical rule which enjoins compadres to formalize their relationships through use of the third person pronoun Usted. This familiarity is especially common among siblings, who will usually continue to call each other by their names and use the informal mode of personal address, and among compadres who differ greatly in age. If a much older person has always called someone tu, as one would to a child, this pattem may continue even after both have become adults and have established a mutual compa- drazgo bond. Foster also observes that "people who feel themselves less Indian and more modem are more relaxed about compadrazgo terminology" (1967:81). TZINTZUNTZAN COMPADRAZGO 33 TABLE 4 Distribution of Compadrazgo Choices among Tzintzuntzan Migrants in Mexico City Compadrazgo Occasion Tie with Baptism Confirmation Total % Village a. relatives 8 4 12 b. friends 8 6 14 sub-total 16 10 26 16% Fellow Migrants a. relatives 6 4 10 b. friends 15 3 18 sub-total 21 7 28 17% Other Mexico City Residents a. in-laws 9 7 16 b. neighbors 28 15 43 c. work companions 18 5 23 d. miscellaneous/in- sufficient data 13 11 24 sub-total 68 38 106 65% Non-Mexico City In-Laws 1 2 3 2% Total 106 57 163 100% In Tzintzuntzan, the compadrazgo functions to cement commercial ties as well as to build a cohesive social structure. Commercial ties are manipulated with outsiders as well as with fellow villagers through the compadrazgo system. In the past, muleteers from the village established compadrazgo ties with people along the routes they traveled to and from the tierra caliente in southern Michoacan. Today, it is common for middlemen in the pottery industry to set up ties with distributors and retailers in other communities, perhaps as far north as the U.S. border region, to which they drive their half-ton trucks with loads of pottery and other locally-made crafts. At the same time, these middlemen are likely to have compadrazgo ties with the artisans in Tzintzuntzan who produce the best wares. This ensures them a good supply of product while it guarantees a steady market for the potters. KEMPER TABLE 5 Relative Status of Compadres among Villagers and Migrants Godparents' Relative Status Population Villagers Higher 164 51 215 Same/Lower 624 112 736 Total 788 163 951 X2= 8.472, P < .01 (1 df) Notes 1. baptismal compadres only, for 1958-1967 (Foster 1969:266). 2. baptismal and confirmation compadres; using only the bap- tismal data set (n - 106 cases, of which 35 are of higher status and 71 of same/lower status), X2 = 8.045, P < .01 (1 df). Because Tzintzuntzan is still a small community, compadres are in frequent, even daily contact with each other. The opportunities for mutual assistance and potential conflict are almost always present. This is, of course, less so where villagers have outsiders as compadres. The balance between daily contacts, usually with one's peers, and frequent contacts, such as with patrons in other communities or in government agencies, is not always easy to maintain. Villagers able to maintain local and extralocal compadrazgo ties appear to be among the most successful of the people in Tzintzuntzan, INTERPERSONAL RELATIONS AMONG COMPADRES: THE MIGRANTS As a dispersed group within a great metropolitan area, the Tzintzuntzan migrants depend heavily on the compadrazgo to strengthen social and commercial bonds with other urbanites. When I asked informants why they established compadrazgo ties with other individuals, the most common response was "friendship" (52 percent). The other, less frequent responses were: "kinship ties" (9 percent), "previous compadrazgo ties" (9 percent), and "economic factors" (8 percent). (The remaining 22 percent could not specify their reasons.) Since friendship and kinship within the migrant population are at once minimal and highly specific (Kemper, 1973), it is understandable that other urban dwellers are a favored group as compadres. Migrants2 Total 34 TZINTZUN7ZAN COMPADRAZGO Because of the geographical dispersion of the migrants and the wide distances between their homes and jobs, the visiting patterns are rather different from the daily interaction possible in the village. Nevertheless, compadres are much more prone to visit one another than are friends or even relatives, although the distances to be traveled influence all types of visiting. Compadres have closer relationships than do other Tzintzuntzan nigrants generally. During a period of twelve months prior to being interviewed, only 26 percent of the compadres surveyed had not been visited by their respective compadres, 48 percent had been seen occasion- ally, 9 percent had been visited frequently, and 17 percent had received regular visits. However, in 4 percent of the "no-visit" cases the compadrazgo tie had dissolved because of the death of the godchild, in accord with the common attitude that "muerto el ahijado, acab6 el compadrazgo" ("when the godchild has died, the god- parenthood is finished"). Table 6 provides detailed quantitative data on visiting patterns among baptismal and confirmation compadres. It shows that the baptismal compadres do have stronger ties, if we use visits as a measure of affect and instrumentality in social relationships. TABLE 6 Visiting Patterns among Migrant Compadres Visiting Frequency1 Compadrazgo Type Baptism Confirmation None (0) 22 21 43 26.4 Occasional (1-6) 53 25 78 47.8 Frequent (7-24) 9 5 14 8.6 Regular (25+) 22 6 28 17.2 Total 106 57 163 100.0 Note 1. Visiting period of twelve months prior to time of interview. Visiting patterns are the social manifestation of the strength of compadrazgo ties. Modes of address provide clues to the institution's symbolic qualities. Although the cultural ideals indicate that compadres use formal types of address (e.g., Usted compadre), a substantial minority (38 percent) continue to use "mixed" or informal modes of address after the compadrazgo ties are established. Table 7 makes clear that there is an essential reversal in the proportion of formal and informal relationships before and after persons become compadres. This applies to both baptismal and confirmation types of godparenthood. Total a/ ib 35 KEMPER TABLE 7 Forms of Address among Migrant Compadres Compadrazgo Type Form of Baptism Confirmation Total Address Before After Before After Before After Formall 36 70 18 31 54 101 Mixed2 14 16 5 5 19 21 Informal3 56 20 34 21 90 41 Total 106 106 57 57 163 163 Notes 1. "formal" address is characterized by the reciprocal use of Usted by both pairs of godparents, avoidance of given names, and consistent use of compadre or comadre as a form of direct address and form of reference. 2. "mixed" address is characterized by combinations of tu and Usted within the compadre pairs, usually along kinship lines; e.g., siblings will continue to use tu-tu while the in-laws use Usted-Usted. In patron-client compadrazgo ties, the senior partner is likely to address the junior by tu, while the junior uses Usted toward him. 3. "informal" address is characterized by consistent, reciprocal use of tu by both pairs of compadres. The ritual formality of the compadrazgo is often observed only by inference. Several times I have observed compadres who began conversations and visits with the proper, formal modes of address only to shift later into informal, highly personalized discussions after the initial bow to tradition. Then, when the compa- dres parted company, they returned to the ritualized forms of behavior, presumably in order to reestablish the correct relationship between themselves. Informality is most often found among siblings or other close kin, 36 TZINTZUNTZAN COMPADRAZGO but their spouses (connected only by in-law bonds) would still follow the strictly formal modes of address. I interpret "mixed" behavior within the same pairs of compadres to mean that the compadrazgo validates and strengthens in-law ties and, at the same time, leaves kinship bonds unaffected. Another symbolic aspect of the compadrazgo is the ritual of kissing the padrino's hand. This is not common in the migrant group, but occasionally children are urged to do so by parents who think it an obligation according to a strict interpretation of the compadrazgo. One afternoon, at the home of an inform- ant's compadre, the ahijados were told by their father to kiss their padrino's hand, first upon our arrival and then again at our departure. Afterward, my informant said that this was a village custom now rare in the city; furthermore, he personally does not like to have his hand kissed but tolerates it in the hope that as the boys (respectively about 8 and 10 years old), grow up, they will abandon the practice. Another of his ahijados in the same family is 16, and still kisses his hand. In contrast, an oldei son (age 21) of his compadre is not an ahijado. This young man is very "fashionable"; he greeted his father's compadre with the "thumb-wrist" handshake popularized by Blacks and Chicanos in the United States, and now frequently seen among young Mexicans. Thus, one ritual merges with other symbols in the transformation of social interaction. CONCLUSION We have examined the compadrazgo through quantitative and qualitative data available for the village of Tzintzuntzan and for its migrants living in Mexico City. The villagers and the migrants show expected varia- tions in their compadrazgo systems. The proportion of relatives chosen by migrants is lower than that chosen by villagers; the migrants also display greater verticality in their use of the compadrazgo in the urban setting; but the migrants are similar to the villagers in their balancing of maternal and paternal lines, in their choice of "zero" generation godparents, and in their preferences for married couples as baptismal compadres. Based on this village-migrant comparison, and taking into account the literature surveyed on the com- padrazgo in Latin America, it is possible to suggest several hypotheses which may be tested through further research: 1) The elaboration of compadrazgo among urban migrants depends on the level of the compadrazgo system's development in their communities of origin and destination. Specifically, their use of the compa- drazgo will approach the mean of the systems of the two localities. In the present case, this explains why the migrants have a more complex system of compadrazgo than do the villagers, but lack a system as elaborate as some found in the metropolis. 2) Among urban migrants, intensification and extension within the compadrazgo system are not defined just in terms of ties with kinsmen (and previous compadres) versus ties with friends. Fellow migrants, whether kinsmen or just friends, are equal targets for intensifying the affiliation with the community of origin. The same principle applies to slum-dwellers whose social universes are narrowly defined in residential terms, regard- less of kinship affiliations. 3) The proliferation of the compadrazgo in urban settings emphasizes secular rather than religious themes. This enhances, rather than hinders, the institution's flexibility and viability as a mechanism for building socio-economic alliances. 4) In urban settings, the critical social dimension of the compadrazgo is continuity of contact. The traditional commitment to lifelong obligations is converted into a series of shorter, more instrumental and functional relationships whose lifespan will depend on the type of compadrazgo contracted and on the godparents' spatial and socio-economic mobility. 5) The compadrazgo is more sensitive to social stratification than to rural-urban distinctions (which in themselves may contain inherent status differences). Its use among the urban poor is more likely to correspond to that among the rural poor than to that among the urban middle- and upper-sectors. It emphasizes the 37 KEMPER egalitarian ties within the urban lower class and among most rural populations, and is an effect of massive cityward migration of these groups. The compadrazgo is a social institution whose flexibility is especially compatible with the diversity of Mexican urbanization. It can help bind neighbors, kinsmen, work companions, employers and employees, and fellow migrants into instrumental and affective alliances. These alliances may be both vertical and horizontal, both intensive and extensive. The compadrazgo provides many opportunities for mutual assistance and for mutual "defensive" protection against quarrels and illicit affairs. The compadrazgo offers urban residents a range of social options, from the most reserved formal posture of respeto to the most informal personal stance of confianza. Moreover, because many types of ceremonial sponsorship involve no (or few) permanent obliga- tions, a large number of a person's ties remain latent. Only a few of one's compadres during the life cycle receive special attention; the rest are recognized as of less (or temporary) importance within an individual's social and economic spheres. And because the compadrazgo is a dynamic social phenomenon, it continues to operate as a pervasive institution throughout rural and urban sectors of Mexico and Latin American countries. NOTES The compadrazgo is a system of ritualized personal relations established between two sets of individuals: the child (ahijado) and his godparents (padrnos) and the parents and godparents (compadres), with the latter ties taking precedence over those between child and godparents. The compadrazgo is usually associated with life cycle rites within the Catholic church, especially those of baptism, confirmation, first communion, and marriage, although compadrazgo ties are also established at other occasions, such as school graduation. Ideally, compadrazgo ties are highly formal, consecrated by use of compadre or comadre (for females) as the mandatory form of address (regardless of pre-existing kinship or friendship ties), and the use of the formal Usted rather than the informal tu. In actual behavior, compadres may vary widely from the ideal model, so that different degrees of informality may exist between sets of compadres according to the degree of daily social contact, kinship ties, socio-economic status, and other variables. 2Only a few studies of the urban compadrazgo are available for the rest of Latin America. Certainly the best of these to date is Middleton's (1975) work in the intermediate-sized town (35,00040,000 population range) of Manta, Ecuador. His mixture of qualitative and quantitative data on the compadrazgo is precisely what we need for the other communities, both large and small, in other Latin American countries. 'According to Foster (1953:7), the compadrazgo in Spain is much less complex than in Latin America: "Primary bonds are established between godparents and godchildren. Only in the case of baptism is the compadre relationship ever recognized as establishing a bond which would not otherwise exist. The average individual acquires relatively few compadres, padrinos, and ahijados during life, and numbers are not felt to constitute an advantage. The institution of the compadrazgo appears to have relatively little importance in stabilizing and integrating communal life within a village or town, to have few significant economic aspects, and to play no really important role in the whole picture of Spanish social structure. This appears to have been the case for a number of centuries." 38 TZINTZUNTZAN COMPADRAZGO APPENDIX 1 CHARACTERISTICS OF THE CASE STUDIES For each of the fifteen ethnographic sources used in the comparative analysis of the compadrazgo in Mexican cities, the following information is presented: a. Characteristics of group studied b. Compadrazgo sample size and type c. Research methods and techniques d. Period of fieldwork MEXICO CITY (1970 population = 7,100,000) 1. Lewis (1952) a. Tepoztlan migrants living in 22 different neighborhoods; working class; 100 families surveyed b. 100 families c. survey interviews; a student team study d. summer 1951 2. Lewis (1969) a. central-city slum vecindad; "culture of poverty" b. 14 families c. in-depth interviews; survey d. several visits during the 1960s 3. Butterworth (1962) a. Tilantongo migrants living in a few squatter settlement neighborhoods; working class; 31 families b. 31 families c. interviews and participation observation; single fieldworker d. nine months, 1961-1962 4. Wilson (1969) a. workers in four manufacturing plants of varying type and size b. total of 152 workers in three of these plants (small clothes factory, middle-sized foundry, a subsidiary of a U.S. chemical plant) c. interviews with workers and key informants in plants d. ten months, 1966 5. Lomnitz (1974, 1975) a. squatter settlement; about 200 households; 70 percent migrants; lower "marginal" class b. survey of 142 households; total of 1,291 compadrazgo ties, of which 1,038 were contracted in the city; 575 baptism cases, 291 confirmation cases, 79 first communion cases, 31 marriage cases, and 62 miscellaneous cases (covering ten other minor types of compadrazgo) c. survey questionnnaires; interviews; mapping d. nine months, 1969; additional visits through 1971 39 KEMPER CIUDAD JUAREZ, Chihuahua (1970 population = 424,000) 6. Ugalde et al. (1974) a. peripheral neighborhood covering 42.5 acres; 1,000 residents; working class b. 125 households c. sample survey; team of students; 111 item questionnaire d. November-December, 1969 OAXACA, Oaxaca (1970 population = 116,000) 7. Chance (1973) a. peripheral neighborhood; 3,500 persons; lower class b. survey of 40 households c. questionnaire; observations d. 1969 (no months specified) 8. Higgins (1974) a. peripheral neighborhood; 3,000 persons; lower class b. no specifics c. participant observation; interviews; life histories; genealogies; spouse was co-fieldworker d. 22 months, 1968-1972 URUAPAN, Michoacan (1970 population = 80,000) 9. Hubbell (1971) a. middle-class women b. 19 middle-class female informants; 380 cases of compadrazgo for their own and others' children c. participant observation; in-depth interviews; single female fieldworker d. 12 months, 1969-1970 QUERETARO, Queretaro (1960 population = 50,000) 10. Whiteford (1964) a. community study of the whole city; emphasis on social class structure b. not specified; a student did a project on compadrazgo c. participant observation; interviews; mapping team study d. 1955, brief survey; 1958, January-April restudy SAN CRISTOBAL DE LAS CASAS, Chiapas (1960 population = 30,000) 11. van den Berghe and van den Berghe (1966) a. community study of the whole city; emphasis on social class structure and ethnic groups b. no specifics c. informants; no specifics d. summers of 1959 and 1965 40 TZINTZUNTZAN COMPADRAGZO JUCHITAN, Oaxaca (1970 population = 30,000) 12. Royce (1975) a. community study of the whole city; emphasis on social classes, associations, and ethnicity b. no specifics c. participant observation; interviews; spouse as co-fieldworker d. 12 months, 1971-1972 TICUL, Yucatan (1970 population = 13,000) 13. Thompson (1971, 1973) a. community study of whole town; a bi-cultural population of Maya- and Spanish-speakers; a cottage-industrialized economy b. random, representative sample survey of 123 males, ages 20-63; 252 baptismal compadrazgo cases c. participant observation; series of interviews on the compadrazgo d. 12 months, summer 1968-summer 1969 "CIUDAD INDUSTRIAL", Hidalgo (1970 population of town = 12,000) 14. Miller (1973) a. new industrial town and its region b. survey of 475 households in seven towns and villages in the region; 784 cases of compadrazgo c. social survey; informants; participation by a team of multi-disciplinary students and professionals d. 1966-1970 TONALA, Jalisco (1960 population = 5,000) 15. Diaz (1966) a. suburan town near Guadalajara; pottery industry b. analysis of 130 compadrazgo cases in church register; participant observations as part of the community study c. participant observation; key informants; census d. August 1959-August 1960; summer 1962 41 KEMPER REFERENCES Berruecos, Luis, 1974. "El Compadrazgo en America Latina." Paper presented at the XLI Congreso Intema- cional de Americanistas, Mexico, D.F. Butterworth, Douglas S., 1962. "A Study of the Urbanization Process Among Mixtec Migrants from Tilan- tongo in Mexico City." Amenrca Indigena 22(3):257-274. Carlos, Manuel L., 1973. "Fictive Kinship and Modernization in Mexico: A Comparative Analysis." Anthro- pological Quarterly 46(2):75-91. and Lois Sellers, 1972. "Family, Kinship Structure and Modernization in Latin America." Latin American Research Review 7(2):95-124. Chance, John K., 1973. "Parentesco y residencia urbana: grupo familiar y su organizaci6n en su suburbio de Oaxaca, Mexico." America Indigena 33(1):187-212. Davila, Mario, 1971. "Compadrazgo: Fictive Kinship in Latin America." In Nelson Graburn (ed.), Readings in Kinship and Social Structure, pp. 396406. New York: Harper & Row. Diaz, May N., 1966. Tonald: Conservatism, Responsibility and Authority in a Mexican Town. Berkeley and Los Angeles: University of California Press. Foster, Donald, and John Azer, 1970. "Research Strategies and the Compadrazgo." Paper presented at the Annual Meeting of the American Anthropological Association, San Diego, California, 19-22 November, 1970. Foster, George M. (Assisted by Gabriel Ospina), 1948. Empire's Children: The People of Tzintzuntzan. Smithsonian Institution, Institute of Social Anthropology, Publication No. 6. Mexico, D.F. - 1953. "Cofradfa and Compadrazgo in Spain and Spanish America." Southwestern Journal of Anthropology 9(1):1-28. - , 1961. "The Dyadic Contract: A Model for the Social Structure of a Mexican Peasant Village." American Anthropologist 63(6):1 173-1192. , 1963. "The Dyadic Contract in Tzintzuntzan, II: Patron-Client Relationship." American Anthro- pologist 65(6): 1280-1294. ,1967. Tzintzuntzan: Mexican Peasants in a Changing World. Boston: Little, Brown & Co. 1969. "Godparents and Social Networks in Tzintzuntzan." Southwestern Journal of Anthropology 25(3): 261-279. Higgins, Michael J., 1974. Somos Gente Humilde: Etnografta de una colonia urbana pobre de Oaxaca. (serie de Antropologia Social, no. 35). Mexico, D.F.: Instituto Nacional Indigenista. 42 TZINTZUNTZAN COMPADRAZGO Hubbell, Linda J., 1971. "The Network of Compadrazgo Among Middle-Class Mexican Women." Paper pre- sented at the Annual Meeting of American Anthropological Association, New York, 21 November, 1971. Kemper, Robert V., 1973. "Factores sociales en la migraci6n: el caso de los Tzintzuntzenios en la Ciudad de Mexico." America Indfgena 33(4):1095-1118. ,and George M. Foster, 1975. "Urbanization in Mexico: The View from Tzintzuntzan." In Wayne A. Cornelius and Felicity M. Trueblood, (eds.), Urbanization and Inequality: The Political Economy of Urban and Rural Development in Latin America, (Latin American Urban Research, vol. 5), pp. 53-75. Beverly Hills: Sage Publications. Lewis, Oscar, 1952. "Urbanization without Breakdown: A Case Study." The Scientific Monthly 75(l):31-41. ,1969. A Death in the Sanchez Family. New York: Random House. Lomnitz, Larissa, 1974. "The Social and Economic Organization of a Mexican Shanty Town." In Wayne A. Cornelius and Felicity M. Trueblood, (eds.), Anthropological Perspectives on Latin American Urbanization (Latin American Urban Research, vol. 4), pp. 135-255. Beverly Hills: Sage Publications. , 1975. Como Sobreviven Los Marginados. Mexico, D.F.: Siglo vientiuno Editores. Middleton, DeWight R., 1975. "Choice and Strategy in an Urban Compadrazgo." American Ethnologist 2(3):461-475. Miller, Frank C., 1973. Old Villages and a New Town: Industrialization in Mexico. Menlo Park, CA: Cummings Publishing Co. Nutini, Hugo G., 1976. Ritual Kinship: The Structure and Historical Development of the Compadrazgo System in Rural Tlaxcala, and Its Comparative Ideological Implications for Latin America (2 vols.) Austin: University of Texas Press. Ravicz, Robert, 1967. "Compadrinazgo." In M. Nash (ed.), Handbook of Middle American Indians, vol. 6: Social Anthropology, pp. 238-252. Austin: University of Texas Press. Redfield, Robert, 1934. "Culture Change in the Yucatan." American Anthropologist 36(1):57-69. Royce, Anya Peterson, 1975. Prestigio y Afiliacion en una Communidad Urbana: Juchitan, Oaxaca. (Serie de Anthropologia Social, no. 37) Mexico, D.F.: Instituto Nacional Indigenista. Thompson, Richard A., 1970. "Status, Ethnicity, and Mobility in a Yucatec Town." Ph.D. dissertation, University of Texas at Austin (Xerox University Microfilms, order no. 71-11,614). 1971. "Structural Statistics and Structural Mechanics: The Analysis of Compadrazgo." Southwest- em Journal of Anthropology 27(4):381403. 1973. "A Theory of Instrumental Social Networks." Journal of Anthropological Research 29(4):244-265. 43 44 KEMPER Ugalde, Antonio (with Leslie Olson; David Schers; Miguel Van Holgen), 1974. The Urbanization Process of a Poor Mexican Neighborhood. Austin: Institute of Latin American Studies, The University of Texas at Austin. Van den Berghe, Gwendoline, and Pierre L. Van de Berghe, 1966. "Compadrazgo and Class in Southeastem Mexico." American Anthropologist 68(5): 1236-1244. Whiteford, Andrew H., 1964. Two Cities of Latin America: A Comparative Description of Social Classes. Garden City, NY: Anchor Books, Doubleday & Co. Wilson, Charles J., 1969. "The Social Organization of the Mexican Factory." Ph.D. dissertation, Columbia University. (Xerox University Microfilms, order no. 69-20, 199).