COMMUNITY VARIATIONS AND NETWORK STRUCTURE IN SOCIAL FUNCTIONS OF COMPADRAZGO IN RURAL TLAXCALA, MEXICO

The functions of compadrazgo (ritual kinship) have been widely discussed in the Mesoamerican literature (Paul Mintz and Wolf Foster Sayres I956; Deshon 1963; Ravicz Thompson In this article we analyze an important aspect of compadrazgo which has not been discussed in the literature: its comparative social functions, as illustrated by data from rural Tlaxcala. The complexity of the compadrazgo system involves many types of relationships and several significant variants in its social functions and discharge. We show how these variations are related to kinship structure, on the one hand, and to the degree of secularization in moral-religious ideology, on the other. In rural Tlaxcala, the shared ideological core of compadrazgo centers on moral-religious obligations that are essentially temporary in nature and in their social manifesta-tions. An important variant in the social functions of compadrazgo is the extent to which these ties and obligations are transformed into relatively permanent relations of social reliance, and the concomitant degree to which compadrazgo networks become the basis for the formation of quasi-corporate social units, which are crucial to an understanding of a wide range of rural communities. But before undertaking this task, it is necessary to provide some background on the general structure of compadrazgo, its syntagmatic sketch, and its temporary and permanent dimensions. aspects of compadrazgo in network formation. Nutini in a four-year project on this important problem. 4. Taking the incidence of compadrazgo types as an index of secularization, we can arrange all rural Tlaxcalan communities on a continuum. Moreover, the twelve pan-Tlaxcalan types represent the most important and traditional aspects of ritual kinship sponsorship in rural Tlaxcala, and the oldest in historical perspective.

and seminucleated communities varying from small hamlets with 400 to 500 people to large villages with 4,000 to 5,000.l The ethnic and somatic composition of rural Tlaxcala is predominantly Indian, and there still persists a strong identification with the Indian past. No more than 15 per cent of the total population can be strictly classified as Indian, yet, culturally speaking, another 20 per cent must be so classified. In such a situation the concepts of Indian and Mestizo have little or no conceptual importance in structuring inter-ethnic relations. In fact, the majority of rural Tlaxcalan communities have remained traditional in the sense that, although there have been many modernizing changes during the past Ioo years or so, the institutional, integrative core of community life-which includes a sacredoriented ideology and world-view, the mayordomia (cargo) system, the ayuntamiento religioso (local religious government), barrio organization, certain elements of kinship and compadrazgo, and a folk religion characterized by Christo-paganism-has proved remarkably resistant to secularization (Nutini and Isaac 1974:366-372). Lingusitically, rural Tlaxcala has approximately 25,000 bilingual speakers of Nahuatl and Spanish, and some 2,000 monolinguals in Nahuatl. In summary, rural Tlaxcala may be characterized as an ethno-cultural continuum in which Indian, transitional, and Mestizo communities share a high common cultural denominator. (2) Casamiento. Marriage (including ten subsidiary compadrazgo relationships, brought about by the following objects and/or events: three padrinos (godfathers) and madrinas (godmothers) of velacion (nuptial benedictions), arras (earnest-money), and anillos (wedding rings); and seven madrinas (only) of lazo (ornamental tie), ramo (flower bouquet), libro (prayer book), rosario (rosary), panuelo (scarf) pastel (wedding cake), and cola (train)). (Io) Apadrinacion de Ornamentos de la Iglesia (AOI). Sponsorship of Church Paraphernalia (including the following objects: palio (pallium), imagenes (images), sagrario (cibary), estandartes (banners), altar (altar), lienzo de Cristo (Christ's tunic), caliz (chalice), ornamentos sacerdotales (priestly ornaments), reclinatorio (pew). (I3) Escapulario. Scapulary.
(i4) Graduacion. Graduation (including graduation from: kinder (nursery school), primaria (grade school), secundaria (junior high school), Each occasion corresponds to a type of compadrazgo, and each type has specific social, economic, religious, and ceremonial aspects which vary in intrinsic and symbolic importance and in the degree of institutionalization of the event and activities that the types require. These 31 types of compadrazgo are broadly divided into two classes, prescriptive and preferential. Prescriptive types require compliance, and individuals incur religious, social, and sometimes economic sanctions if they do not comply. For the preferential types, as the name indicates, there are no social or economic sanctions for noncompliance. Compadrazgo types are further ranked in overall importance, and sacramental types (1-7) do not necessarily rank higher than all nonsacramental types (8-3I). Compadrazgo is a relationship which is established between two individuals or couples and/or a fixed number of related personnel (kinsmen and nonkinsmen) through the link or mediation of a person, image, object, or occasion (the primary actors). The connecting link is the necessary cause of compadrazgo, for without the mediating entity it is impossible to establish a relationship. But central as it is to the system from a structural viewpoint, the mediating entity is not in itself of primary importance, and when all the functional attributes of compadrazgo have been assessed, this aspect shrinks to relative unimportance. Formally speaking, the subject-object (that is, the padr- what is required in every event and action associated with each type of compadrazgo. In this context, the people make a sharp distinction, regarding the functions of compadrazgo, between primary actors (those immediately involved in the contractual arrangement) and secondary actors (the personnel-kinsmen and nonkinsmen-who by extension are affected by the relationship). This is an important dimension in the analysis of the institution, given the fact that rural Tlaxcalans generally cannot ask kinsmen to enter into compadrazgo relationships.
This brings us to a paramount distinction that rural Tlaxcalans make regarding compadrazgo relationships: the difference between asking and being asked. The structural position of an individual asking another to become his compadre is quite different from that of the individual being asked (and by extension all related personnel associated with the main actors). When A asks B, A is requesting that B incur certain economic, social, and religious obligations which will later be reciprocated by A, but which at least initially are often higher for B than for A, and this initial inequality influences the future of the relationship. Further, the person requesting the initiation of the compadrazgo relationship must by definition place himself in the debt of the person being asked, and this too plays a prominent role in molding the future of the relationship. It should be clear that these two aspects are more or less balanced in any individual's total web of compadrazgo, for the average career of compadrazgo relationships shows that in a lifetime an individual is asked to become a compadre approximately as many times as he asks others. Men, women, and children can engage in compadrazgo relationships, both as primary actors or initiators and as mediating entities. The regulation of the system with respect to when, how, and with whom individuals participate in it, and the rules that govern it, are at the core of understanding how the structure of the institution operates in rural Tlaxcala. The different types of compadrazgo are characterized by the structural positions and roles played by men, women, and children.
When compadrazgo is regarded as a system and not simply as a haphazard ensemble of types, another distinction of paramount importance emerges, namely, the distinction between private-individual and public-communal types. The distinction is explicitly recognized by rural Tlaxcalans, and they consciously manipulate it in structuring community action and behavior. Briefly, privateindividual types involve a relatively small number of primary and secondary actors (anywhere from five to 25 individuals) and affect a limited number of people. Public-communal types (numbers 9, Io, 12, I5, 20, and 23), affect settlements as a whole and, albeit temporarily, the community can be said to become a social unit tied by the bonds of ritual kinship. Thus, for periods of time that may vary from three days to two weeks, during the time that the principal rites and ceremonies go on and in their recurrent discharge, the community constitutes symbolically-and to some extent in practice-a global ritual kinship unit with specific behavior patterns, its own terminological usage, and clearly delineated duties and obligations.
Analysis of 175 compadrazgo careers covering the last 80 years shows that the average couple in rural Tlaxcala contracts a mean average of 92 asking and 90 being-asked compadrazgo relationships throughout its compadrazgo career, that is, from approximately age 25 to 75. Moreover, there is a mean average of 46 couples with whom these 182 relationships are contracted, so that approximately 4.0 compadrazgo relationships are contracted with the same couple. Thus it becomes evident that compadrazgo is far from being simply a dyadic arrangement between two individuals or couples-and perhaps a number of extensions-but a more complex institution endowed with egocentric as well as exocentric dimensions. In fact, we can properly speak in rural Tlaxcala of networks and webs of compadres which clearly transcend egocentric structuring and endure over two or three generations. In such a context, compadrazgo may be viewed as an exocentric system.2 The highest common cultural denominators in the Tlaxcalan area are to be found in social organization and religion and, within these, in the local religious hierarchies, the general organization of ritual and ceremonial, and the compadrazgo system. We can speak of a compadrazgo system for the whole of rural Tlaxcala that is surprisingly uniform in both form and content: the structure of compadrazgo types is basically the same, and what is more important, the ideology governing compadrazgo is virtually identical. The differences are primarily in the incidence of compadrazgo types, which may vary from community to community from a minimum of i6 to a maximum of 34, and in the association of specific types with local social and religious forms, due largely to locally developed traditions and well-established social, religious, and even economic interests.
THE SYNTAGMATIC SKETCH OF COMPADRAZGO Figure i gives a syntagmatic account of events of the formal aspects of compadrazgo types in rural Tlaxcala from inception to final resolution. It indicates the most important ritual and ceremonial nodes of interaction with respect to the events themselves and to the behavioral interaction and participation of primary and secondary actors and related personnel. Thus, the syntagmatic chain of events notes all the occasions with social, religious, and/or symbolic meaning in the formal compadrazgo cycle, while the syntagmatic chain of interaction indicates the personnel associated with each occasion and the direction of the prescribed, binding ties and obligations. The interpretation of Figure i says roughly the following.
The preliminary steps and the previous mode of compadrazgo ties determine the selection ofpadrinos and set off constraints both on those asking and those being asked to enter into a compadrazgo relationship. This node leads directly to thepedimento (formal request for ritual kinship sponsorship), which is always a structural microcosm of the future compadrazgo relationship, once the manner of the pedimento is detailed, and during the verbal exchanges among primary actors. This essentially symbolic and social node is followed by a primarily economic node; that is, the arrangements or preparations made hinge primarily upon the immediate economic obligation which both compadres and padrinos must fulfill in accordance with the arrangement made in the pedimento. This node also has religious and social aspects, but they are usually less important. The arrangements node is in many ways the most important one described in the diagram, given the fact that its economic obligations usually color any future permanent bonds. From the social, religious, and ceremonial viewpoints, however, the compadrazgo rites and ceremonies node is the most important in the cycle. It is during these activities that the mediating object becomes the effective link between the primary actors and that the character of the relationship is imprinted for the future. From the observational and analytical viewpoints, the compadrazgo rites and ceremonies may be regarded as the axial center of the relationship that defines the structural and functional development of the personnel involved. The recurrent rites and ceremonies node must be interpreted as a reinforcement and intensification of the compadrazgo relationship, and this is largely the property of those relationships or types having a primarily symbolic meaning. Recurrent rites and ceremonies may be exercised with or without the intervention of the mediating object, and they may take place from a week to a year after the principal compadrazgo activities. The final node is simply the result of the syntagmatic nodal chain just described, and it was noted in the diagram merely in order to accommodate the interaction of personnel associated with that chain.
Three main nodes characterize the direction of personnel interaction within the syntagmatic chain, and these complete the compadrazgo cycle with its associated, institutionalized patterns. In the top node the personnel involved are mainly the primary actors themselves; that is, the parents, kinsmen, or owners of the mediating object (who for simplicity of description we have called the compadres), on the one hand, and the padrinos or sponsors, on the other. The personnel of this node-what we have termed the compadres-compadres dyad-are the result of the discharge of the prescriptions and constraints associated with the first three nodes of the syntagmatic chain of events; that is, they make possible the proper discharge of social, religious, and economic obligations leading to the main part of the compadrazgo cycle. The second or intermediate node includes all the personnel involved in the discharge of the main compadrazgo rites and ceremonies-the padrinos, compadres, and their respective kinsmen, other compadres, and friends or neighbors, all of whom, in descending degrees of importance, participate in the activities of the fourth node of the event chain. This node includes the largest number of primary and secondary actors and related personnel whose participation in the compadrazgo cycle is structurally binding. The third or bottom node involves mainly the primary and secondary actors already aligned in the relationship, whose main function is to discharge the recurrent rites and ceremonies of the fifth node of the event chain. Finally, it should be noted that the permanent or temporary bonds and obligations denoted by the last node of the syntagmatic chain of events is binding, and forms an intrinsic part of any compadrazgo type, only on the primary and secondary actors specifically denoted by the last personnel node. This is the residue of the compadrazgo cycle, and either it becomes an enduring compadrazgo relationship or, if temporary, is almost invariably transformed into a different and usually more important relationship. the central core of rites and ceremonies, for every compadrazgo type in the system. In all nodes after the central core, the situation becomes more fluid and there may be considerable deviation from ideological prescriptions. The prescription for all compadrazgo types in rural Tlaxcala dictates that once a compadrazgo relationship is established it becomes a permanent, lifelong relationship. In practice, however, this is not the case, and several types come to an end after varying lengths of time, and varying degrees of intensity. We do not mean that temporary relationships are terminated in the sense that primary and secondary actors cease to address each other as compadres, and there is no more interaction among them. Rather the situation grows cold; that is, only the form remains. When temporary relationships grow cold in varying degrees, there remain certain specified patterns of behavior such as use of the prescribed ritual kinship terms, some formal visiting patterns, and the required canons of social and religious respect, but the real social, religious, and economic bonds of help and ayuda (nonreciprocal labor exchange) are no longer active. Thus, we can say that in temporary compadrazgo types the most important aspect of the relationship is the central core of rites and ceremonies and the steps leading to it, while in permanent compadrazgo types the subsequent stages are of equal (in the long run, of greater) importance. Temporary compadrazgo relationships have a primarily symbolic and ritual function; permanent relationships have not only a symbolic function, but just as importantly social, economic, and ritual functions. Permanent types are those in which ideological prescriptions are strictly discharged. In the actual functioning of the relationships, their primary, and to some extent their secondary, actors keep active the right to ask for loans, favors, labor exchange, political support for a variety of purposes, social cooperation on numerous occasions in the life and ceremonial cycles, and, in general, any type of interaction that requires the expenditure of economic, social, and religious resources. The attribute of permanence, of course, comes into existence after the central core of rites and ceremonies node, for prior to this point all compadrazgo types in rural Tlaxcala are the same in ideological-structural articulation. It should be understood, moreover, that we are not saying that permanent compadrazgo types must of necessity keep alive the latent patterns of interaction described above. Rather, we are saying that the personnel of these types feel constrained to comply, thereby giving a high statistical correlation between ideological dictates and structural discharge. In this realm of the compadrazgo system there are no sanctions or coercive mechanisms to enforce compliance with ideological prescriptions, and under these circumstances it is probable that some people choose not to activate the latent bonds inherent in this category. Temporary compadrazgo types are the opposite of permanent types in that the permanence ideologically prescribed is not structurally realized after the central core of rites and ceremonies have ended. Statistically speaking, rural Tlaxcalans do not request economic, social, or religious expenditures in the relationships classified as temporary. Here again, since there are no sanctions to enforce any ideological injunctions, it is possible that particular couples may develop a  Tlaxcalans recognize the affinity of certain compadrazgo relationships associated in various ways with the life and ceremonial cycles, and that certain of the resulting clusters serve as a kind of natural paradigm for the selection of compadres. These natural clusters may affect compadrazgo types regardless of their permanent or temporary classification-for example, the preference for baptism padrinos as PCE compadres when the deceased is a child, or for baptism compadres as marriage padrinos. The same is true of several other compadrazgo types. For example, graduation padrinos are almost invariably related to the arents of the graduate by previous compadrazgo ties, the most common ones eing first communion and confirmation. Thus, virtually all compadrazgo types in rural Tlaxcala can be placed in the position of leading to or resulting from other types. This statement must be qualified, however. The eight types classified as permanent and the four classified as temporary have a statistical incidence of original selection which is much higher than the nineteen types that we classify as leading to or resulting from other compadrazgo relationships. In other words, a relationship that did not involve a previous compadrazgo type with the same couple. We can define compadrazgo types that lead to the formation of subsequent relationships as those which are clustered together in a position of anteriority; conversely, types which result from previous relationships are also clustered but in a position of posteriority. It is beyond the scope of this article to explain how the system works short of analyzing the developmental cycle of compadrazgo, which Nutini (I977) has undertaken elsewhere.
But an example will suffice to explain the nineteen types in Table i. Let us take first a leading-to compadrazgo type. Scapulary is a good example, in that almost invariably it is an original compadrazgo relationship. While it involves a permanent status, this permanency is not given by the establishment of the compadrazgo itself; rather, it is due to the fact that scapulary leads to the establishment of other compadrazgo types. The types associated with scapulary compadrazgo are primarily three, namely, taking a mother to hear mass on the fortieth day after giving birth, presenting a child in church at the age of three, and the ceremonial cleansing of a sick person. Let us see how this works in practice. Manuel Xolocotzin and his wife wish to give their daughter, an infant of six months, a scapulary medal. As padrinos they choose Pedro Morales and his wife, a couple whom they have known intimately for many years and regard as close friends. After the central core of rites and ceremonies node, the relationship proceeded for more than two years in a temporary fashion. Neither Manuel nor Pedro invited the other to participate in several social and religious sponsorships that they fulfilled in the interim, nor did they ask each other for specific favors involving outlays of money, goods, or other resources, but they complied with the formal compadrazgo etiquette of social and religious respect. When Pedro's ahijada reached the age of three years and two months, Manuel asked him and his wife to become compadres de evangelio by presenting the child in church. After this second compadrazgo, the relationship between Manuel and Pedro became much closer, and they began to ask each other favors involving outlays of money and political resources. When Manuel's daughter was ten, nearly seven years later, she fell ill and a limpia was effected. Manuel asked Pedro and his wife to become padrinos de limpia, thereby establishing with them a third compadrazgo relationship. After this event, the compadrazgo ties binding Manuel's and Pedro's families grew very strong, and the relationship became permanent for life.
Resulting-from compadrazgo types are defined primarily by their position of posteriority, for seldom if ever are they the original relationship binding sets of compadres. They form part of clusters that include an original leading-to compadrazgo type as in the preceding example, in which either scapulary or sacada a misa are leading-to types, while evangelios and cleansing are resultingfrom types. The most important resulting-from compadrazgo types are cleansing, evangelios, quince anos, first communion, graduation, FC, primera piedra, and CABP. When these types are contracted by a given couple, almost invariably they involve at least one and as many as four previous compadrazgo ties binding the set of compadres. The number of previous compadrazgo ties depends, of course, on the nature of the cluster and on contingent variables such as the eventuality of the ritual occasions (for example, cleansings, graduations, acquisitions of important material objects), the position of the ritual occasion in the life or social cycles (such as anniversaries, JJ, and PCE), and previous compadrazgo ties which are contingent upon individual couples by themselves.
However, the nature of compadrazgo clusters in rural Tlaxcala is of great significance because it not only limits the total number of compadrazgo relationships of given couples, but also serves as a mechanism for intensifying previous ties.
We can now analyze the meaning of the permanent-temporary continuum. In the first place, we have seen that the classification of Table i is not fixed but rather contextual, and that the classification is primarily statistical. The eight compadrazgo types classified as permanent and the four classified as temporary do not necessarily mean that when couples establish a compadrazgo relationship of these types it will a priori become permanent or temporary. Rather, it is implied that there is a strong statistical tendency for the types to become, respectively, permanent and temporary relationships. The permanent-temporary dimension is even less mechanical in the case of leading-to and resulting-from compadrazgo types, in which the status of permanent or temporary will depend on the position of the types within the clusters of which they are part. Moreover, even the types categorically classified as permanent and temporary may form part of compadrazgo clusters, and this will be a significant aspect of their final resolution vis-a-vis the total number of compadrazgo relationships of a given couple. Finally, the very terms "permanent" and "temporary" are variable in themselves, in that their structural content may or may not become operative at given times. They might be termed "active" and "nonactive," in order to indicate their variable character. In the light of this discussion, every couple in rural Tlaxcalan communities, where compadrazgo networks are the functional equivalent of kinship networks, has de facto two sets of compadrazgo relationships at any given time: a permanent set with which the couple interacts in terms of the broad spectrum of what the ideology dictates, and engages in behavior patterns involving economic, social, religious, and even political expenditures; and a temporary set with which they interact formally in the restricted meaning of social and religious respect. One of the most striking aspects of compadrazgo in rural Tlaxcala is that it is unquestionably the most pervasive social institution in local culture and society. There is virtually no aspect of community life that is not directly or indirectly associated with compadrazgo, or functionally or structurally involved in its web. This is most noticeable in the discharge of the central node of rites and ceremonies of the various compadrazgo types. But due to the repetition and intensification of individual relationships, there come into existence not only individual, egocentric webs of compadres, but also systemic, exocentric compadrazgo networks with a high degree of permanence. This is translated into a kind of ritual kinship system which is more prominent in the organization of the majority of rural Tlaxcalan communities than is the kinship system itself.
Although defacto the kinship system has rather loose operative units beyond the domestic group, de jure they are not as effective as the web of compadrazgo in fulfilling the ritual, religious, social, economic, or other requirements of community life. This is also facilitated by the fact that the personnel of egocentric relationships and compadrazgo networks overlaps considerably with kinship personnel. In short, compadrazgo and kinship (in its complementary role) constitute the medium of interpersonal relations and are the only meaningful and effective mechanisms of group formation beyond the domestic unit in rural Tlaxcalan culture and society.
From this viewpoint, it seems logical that the compadrazgo system be used conceptually as an important mechanism of description and analysis at the communal level. This is especially advantageous in situations of rather rapid change, when kinship appears to weaken significantly as an organizing principle of community life, and compadrazgo assumes an increasing importance. This is amply substantiated for rural Tlaxcala, and we are reasonably certain that it may be essentially the same in countless comparable areas of Latin America. We suggest, then, that compadrazgo be regarded as the central social institution in the organization of community culture and society, or rather that it becomes the descriptive integrational principle of social life. The particular advantage of such an approach is that compadrazgo, in a variety of forms, is characteristic not only of rural Mestizo and Indian communities throughout most of Latin America, but also of the urban setting and at several integrational levels.
The interlocking of the compadrazgo system within the global context of community culture and society is underlain not only by kinship but by significant socio-religious considerations as well. The great majority of rural Tlaxcalan communities make great demands on the individual in the conduct of socioreligious life, which most often must be met collectively. There are numerous occasions in the life cycle and the religious cycle which demand large numbers of people, and these require mechanisms of permanent or temporary group formation beyond the immediate mechanisms of the institutions involved (for example, the personnel structure of mayordomias, the structure of the ayuntamiento religioso, or even kinship mechanisms of help and ayuda for specific celebrations). The kinship system of the average rural Tlaxcalan community seldom operates effectively beyond the circle of the domestic group. Hence compadrazgo becomes the functional and structural substitute for, or complement to, the kinship system. In the majority of social and religious activities of the communal cycle that are beyond the efficacy of the domestic group, individual, voluntary, or the collective, formal participation of egocentric webs or exocentric networks of compadrazgo relationships are involved. In such a situation, optional decision-making or prescriptive behavior is regulated by the principles of compadrazgo rather than by the inherent mechanisms of kinship, or the mechanisms of ritual and ceremonial participation, cooperation, and ayuda.
Similarly, when we come to the economic system per se, or the economic dimensions of religion and the life cycle, we find that many of the traditional functions of ayuda, labor exchange, and other economic transactions are often more effectively carried out in the structural domain of compadrazgo than in that of kinship. This is not entirely uniform throughout rural Tlaxcala, for we can arrange communities in terms of the efficacy of the kinship system, from an operational maximum in which kinship traditionally outshadows compadrazgo in the generation of permanent or transitory social units, to an operationalized minimum in which compadrazgo overshadows kinship almost completely. The average rural Tlaxcalan community, however, represents a situation in which compadrazgo is more important than kinship in the organization of community life and the formation of adequate groups of people for religious, social, and economic purposes. Dynamically speaking, the trend is for compadrazgo to replace kinship as long as the socio-religious ideology remains sacred or traditional, for development of a secular ideology means, among other things, that there is no longer a social need for the unit formation and personnel functions of either kinship or compadrazgo. Analysis of a 35-community sample in rural Tlaxcala indicates that within the context of rapidly increasing secularization and change, in the most traditional, less acculturated communities kinship is more important than compadrazgo in the organization of community life and general group formation. In the least traditional, essentially secular (urbanmodern) communities, neither kinship nor compadrazgo is any longer needed functionally, for organizational and group formation functions are carried out by the domestic group. In this scheme of change from traditional to secular, the various combinations of compadrazgo and kinship represent a transitional stage, which is not adaptive but rather is an inherent part of the changing process. Notice that as the process of secularization is completed, compadrazgo does not necessarily diminish as a global institution. What change are only what we have termed its permanent, group formation attributes, for the contracting of compadrazgo relationships, and concomittantly the discharge of the central node of rites and ceremonies with its primarily magico-symbolic and ritual functions, remains to some extent unaffected. religious life is basically egalitarian; the community must be considered an entity with a high degree of integration and lacking in sectors with totally antagonistic interests that may divide it in one way or another. At the regional level, the situation acquires an even greater fluidity and equanimity. In the urban or industrial context, whatever ethnic and cultural differences may be distinguished at the rural community level tend to become blurred, and the rural population is perceived as a diffuse proletarian mass or simply as the lower class with respect to the city. In this dynamic, rapidly changing view of contemporary rural Tlaxcala, we find individuals, groups of individuals, and entire communities moving along the Indian-Mestizo continuum. Thus, the more than 200 communities in these 21 municipios can be ordered in the following fashion: We have indicated the differential importance of kinship and compadrazgo as we move from traditional, to transitional, and on to essentially secular communities. It is important to determine in some detail the relative strength of these institutions, in order to assess the role they play in the process of network formation. We shall compare the kinship and compadrazgo systems of the communities of Santa Maria Atlihuetzian, Santa Maria Belen Azitzimititlain and San Bernardino Contla, which fall respectively under the categories of Mestizo-secularized, transitional-Mestizo, and Indian-transitional communities. In many ways Contla is quite representative of Indian-traditional communities. In fact, Contla may be classified either as Indian-traditional or Indian-transitional, depending on which social and cultural elements are emphasized. These three communities do not exhaust the full range of variation in economic and kinship organization in rural Tlaxcala, but they are representative of the three main ethno-cultural categories.
All three communities share the same ideology regarding the compadrazgo system. There are the same basic imperatives and constraints, the same general attitudes governing the 2I compadrazgo types that they share in common, and the same potential for expansion. But the compadrazgo systems of these three communities are structurally different, and they occupy different positions within their respective cultures and societies. Perhaps more significant is the fact that these three communities have quite different social (kinship) structures, which we shall outline briefly. (The kinship, community, and economic organization of Atlihuetzian, Belen, and Contla is summarized in Table 2.) San Bernardino Contla (Nutini I968) is a unilineally organized society in which kinship is overwhelmingly the most important organizational and operational principle in community culture and society. There are ten semilocalized, patrilineal exogamous clans, and each clan is subdivided into a number of localized, patrilineal, exogamous lineages, which vary from five to fifteen. The lineage is in turn composed of a number of households varying from three to 52. More than 70 per cent of the households shelter patrilocal extended families, while the remaining include primarily nuclear family households.

The clan in the municipio of Contla corresponds approximately to what
Murdock defines as a compromise kin group. In addition to regulating marriage and having names in common, the clan has important religious and ceremonial society. In addition to the regulation of marriage and the possession of a common name, the lineage has important religious, economic, and social functions, and it represents the widest possible range of kin within which fully effective social and economic cooperation can be sought, since the clan has largely ceased to be an effective social unit and has remained principally an effective religious unit. The structure and functions of the nuclear and extended families in Contla are essentially the same as in Santa Maria Belen-and to some extent in Santa Maria Atlihuetzian-except for implications deriving from the fact that in the former the family is embedded in a unilineal structure and in the latter in a bilateral structure. A brief comparison indicates that Contla has considerably more extended family households than Belen: 75 per cent and 40 per cent, respectively. The extended family household in Contla is perhaps more highly integrated, with a higher incidence of gasto junto (one expense budget) as contrasted with gasto aparte (separate expense budget), and in general with a larger individual and nuclear family composition than in Belen. But in every other respect the structure and functions of the nuclear and extended families in Contla and Belen are similar. It is beyond the domain of family structure that Contla and Belen differ fundamentally. In summary, the social (kinship) structure of Contla is remarkably uniform and homogeneous, and its integrational levels (clan, lineage, and family) are at all times and in every sector regulated by the same principles of patrilineality, reciprocity, hierarchy, and cooperation, and their associated mechanisms. Kinship, then, is the most important overall principle of organization and unit formation in Contla culture and society. From the viewpoint of the individual, the social structure of Contla may be visualized as a series of ascending concentric circles representing the different levels of integration. The individual views his rights, duties, values, and obligations largely as concomitants of these levels of integration-as an expression of membership in the corresponding social units .
By contrast, the communities of Santa Maria Belen and Santa Maria Atlihuetzian are bilaterally organized, but to some extent the former has a strong patrilineal bias. In these communities, the role of kinship within the overall social composition, are to be found in both communities. With respect to kinship behavior and internal nuclear family arrangements, there are again no significant differences except that in Belen the nuclear family household is a more corporate unit in which the division of labor, redistribution of income, and the arrangements of daily tasks are geared not only to the exigencies of daily living, but also to the discharge of ritual and ceremonial obligations outside the family per se. With regard to the extended family, however, there are major differences between Atlihuetzian and Belen. First, the overall incidence of the extended family in Atlihuetzian is roughly half that of Belen, occurring in only about 20 per cent of its households. Second, while in Belen the patrilocal extended family predominates by far, in Atlihuetzian the bilocal extended family is the most common arrangement. With respect to the size of the extended family, however, and its arrangements in terms of component nuclear families and individuals, there are no notable differences except that all extended family households in Atlihuetzian have gasto aparte, while in Belen many have gasto junto. As a corollary, the extended family household in Belen is more highly integrated and has more corporate elements.
In Belen, the nonresidential extended family is still a social unit of some importance, and theparaje (locale)-nonresidential extended family correlation is the most significant integrative aspect of kinship at the community level.
Nothing of the sort can be said for Atlihuetzian, where the structure of nuclear and extended family households is embedded directly in the community as a whole. It is this factor, more than anything else, which makes kinship less important in Atlihuetzian than in Belen. In fact, we can say that the domain of kinship in Atlihuetzian does not go beyond the extended family household, and even at this level it does not approach the integrative, coordinative, and formative importance than it does in Belen. In summary, then, as we move from the level of the nuclear family to the extended family and on to the wider aspects of kinship as exemplified by the nonresidential extended family, the kinship system of Atlihuetzian steadily decreases in structural importance as compared with Belen.
Of course, the functional extent and importance of the nuclear family, extended family, and nonresidential extended family are in large part a corollary of the structural arrangements described above, but we will add a few remarks. The functions of the nuclear family are essentially the same in both communities, except that in Belen it has greater religious and ceremonial importance in the life cycle. The situation changes a great deal, however, at the extended family level. While in Belen, the extended family is a strongly functional social, religious, and economic unit, in Atlihuetzian it is little more than the structural arrangement forced upon young married sons or daughters for a relatively short time by economic exigencies. In Belen, the extended family is the focus of concerted action in sponsoring all the important occasions from birth to death. Individuals and constituent nuclear families have specific tasks to perform, and the extended family functions as an integrated unit for the provision of services, goods, and cash. The functional structure of the extended family can be seen even more clearly when it operates as a religious and ceremonial unit. There is nothing of this kind in Atlihuetzian, where the extended family may or may not function as an integrated unit, in some limited capacity, in the life cycle and religious sponsorship. Thus, while the extended family in Atlihuetzian is still a significant unit, it has ceased to be an effective, prescriptive functional unitthough now and then it may function preferentially. This again is an expression of the less stringent social and religious requirements in Atlihuetzian, which do not require the generation of the numerous personnel provided by the extended family and nonresidential extended family in Belen.
We have briefly described the kinship structure of three quite different communities and its position in the context of community culture and society.
We should now compare the kinship systems themselves. Obviously, the importance and pervasiveness of kinship in these three communities is almost naturally scaled. At the top is Contla, where kinship pervades every aspect of community culture and society, and there are no effective principles or mechanisms of integration and unit formation except patrilineality. In principle, there is no need in such a situation for nonkinship mechanisms to cope with the structural and functional exigencies of social living, primarily as they are expressed in the life and religious cycles. In a middle position is Belen, where the kinship system is not as pervasive as in Contla but strong enough to make itself felt in most aspects of community culture and society. Kinship in this situation may provide the basis for structuring other principles and mechanisms of integration and unit formation, but it never operates in this way by itself and cannot alone cope with the exigencies of societal living, especially in the life and religious cycles. Thus, in Belen it is the compadrazgo system which complements, expands, and amalgamates the basic kinship elements into an organic whole, thereby achieving what kinship achieves by itself in Contla. Finally, at the bottom of this scale is Atlihuetzian, whose kinship situation is more different from that of Belen than Belen's is from Contla's. Kinship is least important in Atlihuetzian because there is essentially no need for it in order to cope with the societal exigencies, either by itself or as a complement to other organizing principles, and its effective sphere of action is confined to the nuclear and extended families. Not surprisingly, given the adaptability and flexibility of compadrazgo, these three communities, which have different kinship organizations and overall secular development, share the same ritual kinship system. The same imperatives underlie compadrazgo in all three, but the compadrazgo system is discharged differently and is expressed in fundamentally different functions. Since there are no significant ideological differences, Contla, Belen, and Atlihuetzian have exchanged compadres on a horizontal basis for many generations. The differences appear, however, as we begin to enter the structural domain along the syntagmatic chain. In this area there is certainly a gradation of compadrazgo complexity with respect to types: the most complex system is that of Belen with 31 types, followed by Contla with 24, and Atlihuetzian with 21, as shown in Table 3.
A closer look at   Inevitably, differences are bound to arise in the structural discharge of such a vast array of types; but differences among the compadrazgo systems of Belen, Contla, and Atlihuetzian are of the kind which arise in such communities on the basis of the differing kinship structures, size, degree of secularization, economic base, and the like. Thus, Contla, Belen, and Atlihuetzian are variants of what we have called the pan-Tlaxcalan nature of the compadrazgo system.

THE DIFFERENTIAL FUNCTIONS OF COMPADRAZGO IN ATLIHUETZIAN, BELEN, AND CONTLA
We can summarize the functional differences among the three communities by saying that Contla, Belen, and Atlihuetzian share essentially the same compadrazgo system with respect to ideology, form, and structural discharge, but they differ greatly with regard to function and to the way in which the system is embedded in local community culture and society. What does this mean substantively? The answer has two parts. On the one hand we have Belen, in which the compadrazgo system has temporary (the first part of the syntagmatic chain) and permanent (the second part of the chain) dimensions involving significant and clearly configurated magico-symbolic, social, religious, demographic, political, and economic functions. On the other hand we have Contla and Atlihuetzian, in which the compadrazgo system has mainly temporary dimensions involving primarily magico-symbolic functions, with some residual social, demographic, and religious functions. Implicit in this analysis is the fact that the primary magico-symbolic functions of compadrazgo are associated largely with the central core of ceremonies and events leading to it, and the primary social, religious, and even demographic functions largely with the permanent dimensions of compadrazgo.
Again, what does this mean substantively? In Belen there is an unbroken syntagmatic chain in which the temporary and permanent dimensions of the compadrazgo system are blended into a single entity. Contla and Atlihuetzian are fundamentally different. Up to the point where the last event of the central core of ceremonies has been completed, including recurrent rites, (see Figure i), most of what we have said about Belen applies also to Contla and Atlihuetzian. At that point, however, compadrazgo in the latter communities has achieved its functional (magico-symbolic) goals, and henceforth has no social or religious functions: any permanent dimensions that can be attributed to the compadrazgo system are entirely secondary. This, however, needs to be explained in order to understand the differential implications for the global social structure of these communities vis-a-vis that of Belen.
First, what does the absence of permanent primary social and religious functions mean for the individual couple, or in general for egocentrically determined compadrazgo? To be sure, the principle of respeto, the ideological cornerstone of the compadrazgo system, continues to operate. Compadres in Contla and Atlihuetzian continue to address each other by the appropriate kinship terminology, which is essentially that of Belen; they treat each other with careful respect; they may invite each other to their social and religious sponsorships, though not as cooperating members; they may even ask each other favors occasionally. But there are no prescriptive mechanisms governing these kinds of behavior and action, nor are there highly institutionalized principles governing the web of compadrazgo relationships egocentrically as in Belen.
It is clear that any permanent dimensions that can be attributed to the compadrazgo systems of Contla and Atlihuetzian do not involve the principle of reliance, for it is this principle which is most effective in binding egocentrically determined relationships into some sort of a social, religious, and economic complex based upon customary and prescriptive rules. Activities among egocentrically determined compadrazgo dyads are individually contingent and therefore impermanent. If certain optional actions are not carried out, there are no social or religious consequences for the continuation or termination of the relationship; in fact, in Contla and Atlihuetzian a compadrazgo relationship can be terminated in principle without deleterious social or religious consequences, something impossible in Belen.
Second, for our present purposes it is the absence, both structurally and functionally, of permanent primary religious functions from the systemic viewpoint that really makes the compadrazgo systems of Contla and Atlihuetzian differ fundamentally from that of Belen. The principles and mechanisms that generate cooperation and corporate participation in Belen are based primarily on the compadrazgo system. Thus, in its permanent dimensions the compadrazgo system of Belen is a combination of egocentric and exocentric relationships in which the latter often play a prominent role in the organization and discharge of the most varied social and religious activities, thereby analogously expanding and amalgamating the kinship system. Nothing of the sort exists in Contla or Atlihuetzian. Whenever a group of egocentrically determined compadres get together for a certain activity, this is always the result of extemporaneous behavior designed specifically for the occasion and has no continuity or prescription. Occasionally, the individual in Contla or Atlihuetzian may turn to a specific compadre or compadres for social, religious, and even economic cooperation and participation; but the notion of prescriptive mechanisms affording him these benefits from a stable, permanent (or semipermanent), and clearly discernible group of compadres does not exist. Neither does the concept of collective, multiple-dyadic compadrazgo exist in these two communities, and this is exemplified by the absence of the significant public-communal compadrazgo types for which this concept is essential.
We can finally ask why the permanent and systemic structure of compadrazgo in Contla and Atlihuetzian does not extend beyond the central ceremonies, while in Belen it does. Given certain socio-cultural constraints and a certain degree of secularization, the formation of social units and the mechanisms of corporateness, participation, and cooperation required in the social and religious domains, are fulfilled by the compadrazgo system-and the most important global function of compadrazgo becomes the expansion and amalgamation of kinship. Thus, the constraints and prescriptions in the global social structure of Belen make it necessary that, in the absence of a truly effective kinship system beyond the extended family or nonresidential extended family, the compadrazgo system comes to function as such. In summary, then, while they are different homologously, the compadrazgo system and the kinship system of Belen are functionally analogous because of the structural overlapping The case of Atlihuetzian is the simplest. This community is the most secularized of the three, hence there is no need for the compadrazgo system to function in forming social units or to be the basic principle of participation and cooperation in social and religious affairs. The structure and functions of compadrazgo are thereby limited to its magico-symbolic dimensions. We do not say that there are no social or religious events in Atlihuetzian that require the operation of principles similar to those in Belen-only that these are temporary, preferential, not wholly institutionalized, and the reflection of the rapidly forming new secular ideology. Let us note also that the limited range of the kinship system of Atlihuetzian cannot permit expansion and amalgamation along the lines of compadrazgo, were there a functional necessity for it. But the limitations of the kinship system are simply a reflection of secularization, for 40 years ago Atlihuetzian was structured basically along the traditional lines of Belen.
Contla, on the other hand, is diametrically opposed to Atlihuetzian. Contla and Belen share approximately the same degree of secularization (although the former is an Indian-transitional community, while the latter is a transitional-Mestizo community), but their compadrazgo systems differ in that the permanent, systemic functions and attributes described for compadrazgo in Belen are discharged entirely by the patrilineal system of Contla. Anything that we can say about the permanent dimensions of the compadrazgo system of Belen has its homologous counterpart in the patrilineal system of Contla. Thus, the compadrazgo system of Contla, like that of Atlihuetzian but for diametrically opposed reasons, has structural and functional efficacy only until the last event of the central core of ceremonies has been completed. Naturally enough, kinship as an ascribed system is inherently more efficient than compadrazgo as an achieved system, hence, the permanence, prescription, and degree of configuration of the principles of unit formation, corporateness, and participation are more efficient in Contla than in Belen. The strictly kinship system of Contla is more efficient than the combined kinship-compadrazgo system of Belen in coping with the essentially similar global, traditional social structures of the two communities. But in the majority of traditional communities and practically all transitional communities in rural Tlaxcala, the situation is similar to that of Belen: a fairly effective bilateral system (often with a significant patrilineal bias) which the systemic aspects of compadrazgo have expanded and amalgamated into a fairly organic and permanent whole.
To conclude, in Contla and Belen the degree of secularization differs from Atlihuetzian, where there is no functional need for a compadrazgo system beyond its unitary ideological, temporary meaning, and where departure from traditionality is shown in the restriction of kinship to family structure. Contla and Belen, on the other hand, share approximately the same degree of secularization, but differ in that patrilineal kinship is the primary principle in the organization of culture and society in Contla, while in Belen the combination of a patrilineally biased, bilateral kinship system and the systemic dimensions of compadrazgo is the primary principle. It is our contention that all rural Belen, and Atlihuetzian. Moreover, they are underlain by the following dynamic process. The two most important variables that determine the degree of permanence of the compadrazgo system are kinship and secularization, and the two are rather intimately interrelated. The general mechanisms in operation are these: when the local kinship system is fully efficacious, the compadrazgo system becomes virtually inoperative beyond the central core of ceremonies and has no permanent dimensions. The degree of secularization is more important than kinship as an exogenous variable, for it determines the need for the permanent dimensions of compadrazgo: the more secularized the community, the less need for permanency.

A NETWORK EXPLANATION FOR VARIATIONS IN THE PERMANENT DIMENSIONS OF COMPADRAZGO
In the preceding sections we have shown that the temporary dimensions of compadrazgo constitute a pan-Tlaxcalan phenomenon, while its permanent dimensions entail different functions in three representative rural Tlaxcalan communities. Thus, structurally and functionally, we explained variations in the incidence of compadrazgo at the regional level. Now we shall try to explain variations in the permanent dimensions of compadrazgo in terms of network structure. But first, we must briefly describe the two basic concepts of compadrazgo in action, which are directly related to the formation of compadrazgo networks.
The concepts of respeto and confianza are universal aspects of the compadrazgo system in rural Tlaxcala. In all 35 communities for which we have gathered data on compadrazgo, these concepts are readily verbalized as a significant component of the structure of compadrazgo in a large variety of contexts. As used by rural Tlaxcalans, the concepts must be characterized as "covering terms"-that is, terms whose semantic payload is much greater and more diverse than their common sense meaning indicates. They stand for a wide variety of behavior extending across the entire spectrum of the compadrazgo system. Thus, when rural Tlaxcalans tell the ethnographer that compadrazgo relationships are marked by variant degrees of respeto and confianza, they are expressing respectively a series of ideological and structural imperatives or constraints that may entail social, economic, religious, and/or symbolic components characterizing specific ways of behaving. From a dynamic viewpoint, respeto and confianza may be regarded as the activating principles that are the most significant ingredients in the formation of compadrazgo networks.
Respeto is the most encompassing of the two principles. It not only unifies compadrazgo into a well-structured system, albeit an open-ended one, but it also delimits the boundaries of actual behavior. When the average rural Tlaxcalan declares that compadres must respect each other or that respeto is the necessary condition for a compadrazgo relationship, these utterances cover a wide range of imperatives and commands, the highest expression of which is embodied in what we have called the sacred nature of compadrazgo. Respeto is not only an imperative to action, but also a command that conditions behavior. The covering properties of respeto, which we cannot detail here, may be regarded as a body of "native categorial imperatives" which in a real sense constitutes a model of the compadrazgo system in rural Tlaxcala. In summary, the principle of respeto embodies the ideology of compadrazgo with respect to prescriptiveness of type, the nature and extent of rites and ceremonies, the formal behavior of compadres, and in general the unfolding of the temporary dimensions of compadrazgo.
The principle of confianza, on the other hand, is much more central than respeto to the permanent dimensions of compadrazgo, for it not only determines the selection of compadres but specifies the conditions for the repetition and intensification of compadrazgo relationships, which are of paramount importance for the formation of compadrazgo networks. We can broadly define confianza as a potential attribute of any compadrazgo relationship which permits its actors to ask each other favors of an economic, social, religious, and even political nature. This does not mean that this potential attribute is always actualized, but rather that it can be actualized given certain specific circumstances. From a slightly different viewpoint, we can say that confianza conditions the allocation of social, economic, and religious resources so as to minimize conflict with the proper discharge of other systems within rural Tlaxcalan culture and society. Thus, the generation of confianza relationships constitutes the single most important factor in the formation of compadrazgo networks. Finally, we must add analogously that the generation of confanza relationships applies equally to kinship, occupational, and several other kinds of networks.
Kapferer (I973: IOI) notes two basic propositions linking exchange theory with the structure of social networks: (i) The greater the extent to which two persons have overlapping social networks with other persons who are known to each other, the more these other relationships are put at risk if one fails to discharge his obligations toward the other.
(2) The greater the concern to protect one's social investments with as many people as possible, the more one will concentrate time and effort on key relationships central to the social network.
In rural Tlaxcala, where compadrazgo types differ in importance with regard to respeto and prescriptiveness (the more important the type, the more prescriptive and the more respeto), the actual selection of compadres typically works on the basis of confianza, that is, confidence in fulfillment of support, granting favors, and maintaining trust in the relationship, symmetrically, between two parties. Compadrazgo thus represents a social investment. There is a strong concern in the performance of compadrazgo duties and obligationsgenerally of a temporary, ritual and ceremonial nature-for protecting one's social investment in compadrazgo and in other confianza relationships as widely as possible.
In this context, then, it follows from propositions (I) and (2) that the compadre with whom it is most likely that one will fulfill confianza expectations is that compadre with whom one has many common confianza relationships among whom there is also confianza. Thus, for existing compadrazgo ties, confianza relationships tend to be activated in such a way as to form clusters or cliques of mutual support (reliance groups) based on reciprocal expectations. It also follows from proposition (2) that since members of a cluster with whom one has one's primary confianza relationships are more central in one's social network, one will invest in these relationships by contracting new compadrazgo Our theoretical explanation of the formation of reliance groups within compadrazgo networks is incomplete, however, for it suggests that a person would tend to contract compadrazgo ties within a single cluster or clique rather than in several. The explanation for multiple-cluster affiliations may be sought in a further extension of Kapferer's propositions: (3) A person will invest heavily not only in persons most central to one network, but in persons most central to different sectors of several networks. The focus of these networks might be defined by type of activity: family, kinship, occupation, common interests, exchange, and so on.
The most important consequence of the tendency to form compadrazgo clusters based on confianza is that as soon as a compadrazgo bond to a member of a cluster expires, given the inherently temporary nature of compadrazgo ritual and ceremonialism, the centrality of this relationship in the confianza network requires further social investment or intensification. This is achieved either by contracting another (resulting-from) compadrazgo relationship, or by converting an important, already existing compadrazgo relationship into permanent social links. The pressure to create permanency in compadrazgo linkages, then, derives from the expectations and risks associated with the structure of the social network which results from the dynamics of confianza. For most Indian-transitional and all transitional-Mestizo communities, then, where maximizing confianza.is the prime consideration in compadrazgo selection, our exchange-theory propositions offer a relatively full explanation of why compadrazgo acquires permanent dimensions associated with enduring reliance groups, and a model for the process of formation of such groups. The structure of confianza reliance groups formed through permanent compadrazgo relationships in the type of system we have described acquires many of the functional characteristics of extended family or kinship groups: quasi-corporate mutual interests, mutual support, well-defined expectations and sanctions, and multiple allegiences-as in consanguineal versus affinal kin group affiliations. There are specific conditions, however, under which compadrazgo links do not tend to become permanent nor to form reliance groups. This is the case in all Indian-traditional and some Indian-transitional communities, where kinship is the primary basis of confianza, and compadrazgo selection is based on respeto relationships with individuals who are members of different kin groups who cannot be expected to share confianza. Throughout rural Tlaxcala there is a widespread proscription against contracting compadrazgo relationships with kinsmen. The reason given for this practice is that kinsmen are too liable to disagree, especially over land and inheritance, or the allocation of resources, which is inimical to the sacred nature of egalitarian, horizontal compadrazgo  (Forbes I97I: 24). In other words, in all Indian-traditional and some Indiantransitional communities, the formation of confianza reliance groups is overwhelmingly achieved on the basis of kinship, in spite of the potentialities for conflict, particularly in communities where agricultural or craft occupations require extensive extended family cooperation. Thus, in the most traditional communities in rural Tlaxcala, compadres cannot be selected primarily on the basis of confianza, as outlined above. Respeto, however, remains a key feature of compadrazgo selection. Respeto, like confianza, is ideally symmetrical, but unlike confianza, differences in age and status often make compadrazgo relationships asymmetrical. If one fails to comply with the expectations of a respeto relationship of a given degree, there is not sanction of symmetrical reciprocity operating to restore the balance of equality or to maintain respeto at a given level: the relationship merely changes to reflect either greater asymmetry or a lessening of the intensity of mutual respect. Thus, proposition (I) does not apply: failure to fulfill respeto expectations in one relationship does not put other relationships at risk. Moreover, social investment in actors central to one's respeto compadrazgo network does not occur, and as a consequence, there is no tendency for respeto relationships to form clusters or reliance groups.
The dynamic of respeto relationships is individualized or insulated in its network ramifications because of the possibility of asymmetry, which is absent in confianza relationships. In the most traditional rural Tlaxcalan communities, then, compadrazgo selection based solely on respeto does not have network consequences leading to the formation of clusters or reliance groups. The resultant social structure is atomistic and, by nature, temporary. Compadrazgo may serve to link individuals or families on a temporary basis, either by patching up conflicts or building temporary alliances, but such alliances lack the network sanctions that would make them permanent. Clearly, also, under such conditions, failure to comply with compadrazgo obligations does not put one's kinship relationships at risk. The obligations of compadrazgo, as a temporary link between kin groups, may be fulfilled or not, depending on reciprocity with that particular group.
Finally, in the most acculturated, exclusively Mestizo-secularized communities, with massive labor migration and an essentially urban orientation, the permanent dimensions of compadrazgo and kinship cooperation and exchange are greatly reduced. However, there are still alternate social and economic bases for the development of confianza relationships outside compadrazgo and kinship networks-for example, occupational, common-interest and exchange networks.
Thus, in Mestizo-secularized communities, compadres are selected individualistically on the basis of respeto. The consequences for compadrazgo are essentially the same as in Indian-traditional communities, except that at this extreme of the ethno-cultural continuum compadrazgo acquires a significant vertical component. Moreover, the asymmetrical aspects of respeto are here utilized to the fullest extent in contracting extracommunal compadrazgo relationships which are basically of the patron-client type. Individual maximization of economic and social behavior acquires under these circumstances its maximum expression, and reciprocity is the main determinant of the greater or lesser degree of permanence in these relationships.
One general point which emerges from our network and exchange-theory analysis is that some form of symmetrical reciprocity must be present in the dynamics of compadrazgo selection and maintenance in order for permanency and clustering to emerge in the network structure. This condition is satisfied only in intermediate (most Indian-transitional and all transitional-Mestizo) communities, where confianza (symmetrical reciprocity) is the basis of compadre selection. At the extremes of the continuum (all Indian-traditional and Mestizo-secularized and some Indian-transitional communities), compadre selection is based on respeto (which allows asymmetry), thereby precluding the formation of permanent reliance groups. Thus, in the most traditional communities kinship is the sole basis for generating reliance groups, while in the most secularized communities they develop, albeit in a more attenuated form, from confianza relationships on the basis of occupational, common-interest, and other kinds of networks. In the structural-functional analysis we discussed kinship and secularization as the main variables conditioning compadrazgo as a mechanism for unit formation and the structuring of reliance groups. But in a more precise explanation of the differential functions of compadrazgo, there are economic and demographic factors which must be taken into consideration as communities move along the continuum.

CONCLUSIONS
Our structural-functional and network analyses of compadrazgo are based on two fundamental premises: the predominance of the compadres-compadres link in the organization of compadrazgo behavior, and the centrality of compadrazgo as a social institution in regions undergoing rapid change, where analogously it comes to replace kinship as the main organizing principle of community life.
These especially are the dynamic underpinnings of our network study of compadrazgo now in progress. Inasmuch as we cannot yet give definitive answers to the many problems involved in the network structure of compadrazgo, in the remainder of this article we shall present our general theoretical approach in the form of two related models with their corresponding hypotheses. It goes without saying that the replicability of the models is the justification for these remarks.
Description and comparative analysis of the forms of social organization in Mesoamerica and much of South America have been complicated by the manystranded character of social relationships such as compadrazgo, bilateral kinship, effective ties in economic and political alliances, and by the lack of clear-cut structural units resulting from such open-ended types of relationships. Foster's (I96I) characterization of the "dyadic-contract" identified the process by which many social relationships are formed in peasant communities, but did not clarify the types of social configurations which may result from such open-ended contracts. Ethnographic work is badly needed on the structure of social networks in societies which emphasize poly-stranded relationships, and on the consequences of different types of network structure for role behavior.
Recent developments in the understanding of compadrazgo links (Nutini I977)  stranded social organization. We are not using here the concept of social network in the loose or metaphorical sense. Barnes (I975), in tracing the development of metaphorical usage of the network concept, notes that for most problems of social organization, a detailed network analysis of social relationships may be redundant in that such analytical concepts as social roles or corporate groups could serve equally well in the exposition of social organizational data. Here we take the position that network analysis is meaningful in its own right only where the behavior of actors in the social setting is related not to roles or characteristics of individuals and groups, but rather to the particular way in which relationships are configurated in a network. Thus, network analysis, if applicable to a problem, should be able to elucidate (and predict) certain types of individual and group behavior that are not functions of individual attributes or even dyadic relationships per se.
The data presented in this paper and Nutini's (I977) recent study make a strong case for the applicability of network analysis to the compadrazgo system of rural Tlaxcala. While the various types of compadrazgo are either dyadically or communally contracted, and related to the life, ceremonial, and economic cycles of individuals or the yearly religious cycle of the community, the resultant exocentric patterns of social relationships tend over time to form structures which acquire a dynamic of their own. Clusters of individuals with a high density of compadrazgo ties emerge, and for certain purposes they show behavioral regularities at the community level, sometimes in relation to concerted economic, religious, or social action, sometimes in relation to conflict resolution or the avoidance of direct confrontation. Without refined tools for analysis of clustering, decomposition of a finite population according to network linkages, and the compounding of role relationships (H. White I974), it is impossible to define and specify the structure of social networks beyond their merely descriptive integration as Nutini has already done. The aim of our research is to transcend descriptive integration and generate more powerful explanations which will specify the structure and content of compadrazgo networks as an expression of social, economic, religious, and political action. Given the nature of rural Tlaxcalan culture and society, our basic assumption is that social networks are used to generate resources to accomplish social ends. In giving our theoretical approach to the functioning of social network linkages in the form of hypotheses, it should be noted that both the nature of the resources and the conception of relevant social ends vary according to the internal resource structure or the external orientations of community members.
In the community setting, the first set of independent variables that are relevant to network formation are found in the differential utilization of resources: (I) Involvement in agriculture. The more closely an indivi group is tied to the productive resources of the land, the g that network linkages will be established that are social propinquitous. Compadrazgo will tend to reinforce kinship ties, and at the community level, compadrazgo will tend to the community.
(2) Involvement in extracommunal resource systems. individual invests in extracommunal exploitation of reso migration, the more likely it is that a significant number of network linkages will be formed based on factors of continuity and propinquity in work groups.
Thus, relative to the home community of labor migrants, the number of external network relationships will increase in relation to the investment of time outside the community and the continuity of those outside contacts with persons engaged in the same work.
(3) Involvement in diversified occupations. With occupational diversification, kinship can no longer generate sufficient resources to accomplish social ends, because kin linkages give no assurance of tapping a variety of specialized occupations.
Thus the more open-ended and poly-stranded network linkages, such as compadrazgo, will tend to be utilized to establish a broader set of linkages. This accounts for the much greater importance of compadrazgo in transitional communities, as compared with traditional communities in rural Tlaxcala. Also in the community setting, national integration and secularization constitute a second set of independent variables relevant to network formation: (4) Integration with the state and national politico-economic system. The greater an individual's or community's orientation toward the national media, politics, or market, the greater the likelihood that network linkages will be generated to serve the individual's social mobility, job security, and political security. Under these circumstances, network linkages will be less widely extended horizontally, and more vertically structured to incorporate persons who can be used to serve the individual's particular ends.
(5) Secularization of the content of social relationships and community institutions. As the proportion of individuals with a national orientation increases in the community, the effect is to undercut traditional, communally based socio-religious institutions; that is, to secularize the content of compadrazgo ties, involvement in the ayuntamiento religioso and the mayordomia system, and the moral-religious obligations of some aspects of kinship ties.
These hypotheses indicate that kinship will offer the most important network linkages in communities with greater dependency on agriculture (traditional communities), and that compadrazgo ties will be discharged in accordance with traditional religious ideology. However, as the economy diversifies in terms of labor migration and cottage industries (transitional communities), kinship will no longer suffice to generate resources even for communally linked social ends, and horizontal compadrazgo ties will take over as the primary axes of network formation. Finally, as national orientations become dominant among individuals in the community (secularized communities), the structure of compadrazgo will collapse into smaller horizontal clusters but expand to serve vertical linkages which individuals can manipulate to their own ends, while the content of compadrazgo links will become largely symbolic and secularized as opposed to moral and sacred.
Outside the community setting, compadrazgo networks will develop in work groups where there is a continuity of personnel, and compadrazgo links of a vertical nature will often develop between individuals in different status positions in different locales. In addition, it is possible that compadrazgo networks may form around other extracommunal activities such as cyclical markets or political organizations. Nutini's (I977) description and analysis of community variation in compadrazgo structure and the foregoing hypotheses give rise to the following predictions about the relationship between community type and network structure: (a) In traditional communities, compadrazgo linkages will tend to be highly endogamous and follow kinship and territorial lines of organization.
Compadrazgo links will reinforce kinship networks, but have only a moderate importance for community integration. (b) In transitional communities, compadrazgo linkages will tend to develop into exocentric network structures independent of kinship. Compadrazgo acquires major importance as the primary means of social integration in the community, with kinship in a subsidiary position. (c) In secularized communities, traditional compadrazgo types diminish in importance, while remaining largely endogamous and reinforcing kinship linkages, but only within the restricted domain of nuclear and extended family organization. Secondary and more secularized compadrazgo types (for example, graduation, quince anos, and CEON) become more important in building networks along independent lines, often through vertical and extracommunity links.
Throughout this discussion we have assumed that positions in compadrazgo networks are not simply indicators of social position, but rather that compadrazgo positions (defined structurally within a network) are the basis on which power, cleavage, or solidarity are built at the exocentric or group level. That is, an individual may have the resources to participate in the ayuntamiento religioso, but his actual date of entry will depend upon the existence of required social linkages within the community. Similarly, among several children within a family, an individual's decision to migrate for wage labor or to stay and work the land may depend, not on his relative age or other characteristics, but partially on the extent to which a strong or weak social network has been created for him by his parents.
Throughout this paper, we have analyzed the most salient structural and functional attributes of compadrazgo in rural Tlaxcala, which bear directly on the network formation of the institution and its ethnographic pervasiveness at the community and regional levels. There are probably countless regions in Latin America in which compadrazgo exhibits basically the same characteristics that we have described here, but anthropologists have tended to see compadrazgo strictly in terms of dyadic relationships without any awareness that such relationships can be structured into poly-stranded, multidyadic networks.
Thus, given the ideal conditions of compadrazgo as a natural system, the combined analytical approach of this paper may merit replication in other regions of Latin America. NOTES i. The population of these 21 municipios is roughly 315,000, of which slightly more than comprise the four small cities of the region: Tlaxcala, the state capital (Io,ooo), A (20,000), Chiautempan (i8,000), and Zacatelco (7,500). These are not only small citie socially, religiously, and economically intimately tied to the rural environment. 2. The subject matter of this short outline of compadrazgo in rural Tlaxcala is discussed at length in Part II of Ritual Kinship (Nutini I977). The reader may wish to consult the b