On the 29th October 1935, Lázaro Cárdenas, President of the Mexican Republic, signed a Presidential Resolution granting an ejido to 316 beneficiaries to be drawn from the peones acasillados of the Great Hacienda of Guaracha in the Ciénega de Chapala. Though the President was to sign many more such resolutions in the coming months, and the redistribution of land to contracted wage-laborers represented an important new phase in agrarian policy nationally, this one had a very personal significance for him. The boundaries of the 35,000 hectare Guaracha estate had run up to the outskirts of Jiquilpan, the small provincial administrative town where Don Lázaro had grown up.

This book examines the history and contemporary results of land reform in the backyard of Mexico's greatest reforming President as seen through the eyes of the peasantry who are its supposed beneficiaries. My empirical objectives therefore appear straightforward enough, though the facts of the case turn out to be less so. But there are now innumerable studies of Mexico's ejidos, a vast theoretical literature on the "Agrarian Question" as it pertains to the Mexican case, and an equally extensive corpus of writings on the relationship between Mexican peasants and the Mexican state. Before I make any specific claims for this particular work by sketching out the arguments and lines of analysis it presents, I should begin by addressing the question of why it is worth adding another book, especially one based on the apparently narrow horizons of an anthropological field study.

The Limits and Possibilities of Ethnographic Knowledge

This book is an ethnography. Its historical analysis is based mainly on the oral testimony of local people, and its basic data were collected through structured interviews, supplemented by questionnaires, casual spontaneous conversation and the direct observation of formal and informal occasions of social interaction during fieldwork. It is also an ethnography which focuses primarily on a single community. Yet I am going to use the material I have collected to address and try to advance general arguments, some of which are intended to be relevant not only to other regions of Latin America, but to other parts of the world. The reader is quite entitled to feel that these claims seem contradictory and implausible, for, in a sense, they are.

I do not claim that Guaracha represents rural Mexico in microcosm. Nor do I claim it portrays a "typical" peasant community within its own microregion. It was because this community and its region had certain specific historical and contemporary characteristics that I chose to undertake my research here. All too often, theorizing about the peasantry and its place in developing capitalist societies abstracts from the diversity of regional agrarian structures, a diversity which is particularly marked in the Mexican case, and has been a persistent factor in the social landscape since the dawn of the colonial era. It is equally common for empirical studies of particular regions at particular moments of time to become transformed into ahistorical position markers in polemical debates. This is certainly the fate which has tended to befall Chayanov, and to an even greater extent those Mexican Chayanovians who can less easily be accused of sharing their mentor's essentialism: it is the reward, but also the curse, of original and profound insights into historically bounded reality. Studies conducted in particular places twenty years ago or more can certainly still convey insights of contemporary relevance, but it is necessary to recognize that the world moves on, perhaps at an accelerating rate.

Anthropologists are particularly beset by the problem of the historicity of their endeavors. Fieldwork takes a long time. Analyzing field data takes even longer. And at the end of the day, the individual anthropological researcher can accomplish so little, if he or she persists with the intensive, microscopic examination of human reality characteristic of the so-called method of "participant observation." Anthropologists can, and have, adopted larger frames of reference, and the particular virtues of anthropological research at the regional level as an antidote to the excesses of the "micro" and "macro" poles of analysis have been abundantly demonstrated in terms of practical results as well as justified methodologically at a more abstract level (Wolf 1955, De la Peña 1981, Long and Roberts 1984). I have no quarrel with these arguments. In fact I insist that anthropologists must be prepared to work with national, international and, as this study seeks to demonstrate, transnational, units of analysis. Nor am I at all enthusiastic for the argument that anthropological perspectives can or should be autonomous with respect to work in other disciplines. Little of what any of us has to say has (thankfully) yet become so esoteric as to be unintelligible to scholars in other fields (and vice versa), nor are our professionally inculcated methodologies so arcane as to be beyond at least reasonably effective replication by scholars trained in other fields when analytical need arises. But more importantly, the human problems with which we are all grappling in our several ways are far too serious to be reduced to the objects of professionalization and the commoditization of knowledge, even if such a reduction is to some extent inevitable given the world we live in. Yet we may perhaps make a more modest and legitimate claim for the more detailed kind of knowledge produced by anthropological fieldwork of a traditional kind.

There are genuine virtues in the close analysis of human sociocultural realities through micro-level field studies, and there are ways of looking analytically at those realities which are the collective legacy of the discipline's exploration of the cross-cultural variety of human experience. We can embody those virtues and strengths in our work without self-delusion by systematically addressing the limitations of the different types of knowledge produced by different types of empirical enquiry and levels of analysis. This is what this book sets out to do. Analysis of the specific data on Guaracha will frequently lead me to pose questions which cannot possibly be answered empirically on the basis of the data acquired by this, or any other, case study. My response to this dilemma is to indicate the extent to which collectively accumulated knowledge leads us towards a particular answer at this moment of time, and what further investigation would be desirable in order to pursue the issue. Much of the argument of the book is critical of established ways of looking at "the problem of the peasantry," and I use my particular data to demonstrate the usefulness of other conceptualizations in this specific case. In looking beyond the immediate empirical and historical horizons of my field study, I offer what are simply pointers as to how to proceed with larger arguments which must, of necessity, be conducted on a different basis and at a different level of abstraction. There is a justification for moving towards these issues from an anthropological field study: the local is inevitably conceptualized from the standpoint of prior conceptions of the global, and any limitations in existing higher level abstractions revealed through grappling with the concrete should therefore merit attention.

We can, however, make one final and straightforward claim for anthropological research at the micro-level. Though anthropologists have made all manner of dubious epistemological claims for the profundity of the knowledge of "other cultures" produced by fieldwork, we might do better to tell our paymasters about the more mundane advantages of living in a place and gaining the confidence of those we study. The data on such matters as ejidal land tenure or international migration provided in this book could not have been obtained in any other way. Though there are more "exotic" kinds of data in the book of which the same would be equally true, and much of the data obtained requires a degree of filtering in presentation on my part for ethical reasons, if one wishes to provide an accurate picture of the mundane realities of the social reproduction of a campesino community it is hard to see how such a picture could be obtained without doing anthropological fieldwork. Readers can judge this particular claim for themselves, but there is one type of reader whose judgment is, for me, particularly important. Many campesinos in Guaracha themselves read books (or have them read to them). They gave me information on the understanding that I would try to transmit the reality of their historical experience to the outside world in as accurate a way as I could. Neither they, nor I, are so naive as to imagine that there is a "truth" here which stands above particular points of view and interests, including my own. Since the campesinos are not socially homogeneous, the truth of their history is a contested domain of facts and meanings. But I have tried to write the book in a way which provides a vehicle for these various truths to find expression, without trying to pretend such expression is unmediated by making use of the literary and rhetorical devices which are currently so much in vogue in Anglo-Saxon anthropology or abandoning the objectifying thrust I would maintain is the only honest response on the part of the outside observer of other social realities.

I hope that this volume will convey something of the human experience underlying its more abstract analytical concerns, an experience which is worth documenting in its own right, as a testament to the remarkable nature of "ordinary people." What is at issue here, I hope, is more than a populist enthusiasm for being on the side of the oppressed, or guilty conscience on the part of the privileged bourgeois whose career is advanced by prying into other people's misfortunes. People and their social worlds are multidimensional. Yet much of our academic literature reduces the peasant to one dimension-the occupant of an economic, social or political niche in a larger social world where he (or she, but usually he!) is interesting only as a victim of history and the various structures of domination which define his or her place in the larger totality. It is not that these perspectives are incorrect, since adopting them does raise significant issues. But they are certainly incomplete, and predispose us to lose sight of much that is essential to understanding both the objective and subjective reproduction of the peasantry as a social category.

This, then, is a study which unashamedly embraces the particular facts of a historical social situation, and uses those facts to address generalizations about contemporary agrarian realities in Mexico and their wider determinants and significance. My approach will generally be one of chiselling away at the apparently idiosyncratic and contingent aspects of social existence to identify the social facts which constitute them and the determinants and effects of those social facts in terms of the larger movement of Mexican society since the Revolution. I believe this approach has its virtues. But some might still argue that the entire object of study here is an illusion. What is the place of peasant studies, and especially anthropological peasant studies, in the world of the 1990s? It is well worth posing this question, especially in the Mexican case, since the answer is not at all self-evident, and critics of anthropologists' work in this field muster serious arguments which demand a response.

The Peasantry—A Problem for Historians?

At the end of a short paper discussing the impact of the urbanization of agricultural land in the state of Morelos, Michael Redclift despairingly remarks:

In the case of Morelos, part of the reason for the demise of "peasant" movements and the restructuring of the peasantry lies in the urbanization of agricultural land, a process which has been ignored by all but a handful of social scientists ... while hundreds of Mexican and foreign scholars still descend on ejidos or communities to record their history and internal dynamics, it is left to a few journalists and students from the Universidad Autónoma Metropolitana Ixtapalapa to study fraccionamiento development. (Redclift 1986:100)

This particular facet of changing "rural" reality is scarcely central to the Ciénega de Chapala at the present time, though much of the capital generated in the agrarian economy does end up being sunk into urban real estate, often in Zamora, Morelia, Guadalajara or Mexico City. But there are a number of other respects in which the entrenched rural-urban divides of so much of the traditional literature are as unsatisfactory even in what is supposedly still a predominantly rural state as they evidently are in contemporary Morelos. Moreover, the short answer to the puzzle posed by Redclift—to what do we owe our ignorance, despite all these armies of rural researchers?—is quite simple. Rural research and peasant studies are an academic specialization.

There is, however, a much longer, more complex, and possibly more meaningful answer. There are intensely practical, political determinants of the various types of literature which have appeared over the decades on the Mexican peasantry, brilliantly surveyed for an English-speaking audience by Cynthia Hewitt (1984). While some, mainly foreign, anthropologists may have dedicated themselves to the study of the peasantry in search of an object of analysis sharing some of the characteristics of the "primitive" small-scale society, or even out of pure romanticism, Mexico's variant on the long-standing debate on "The Agrarian Question" has a genuinely substantial content. The link between agrarian policy and the political strategy of the post-revolutionary elite before Cárdenas becomes somewhat more intelligible if one credits its authors with the conviction that the peasantry was doomed to extinction through the operation of inexorable economic forces. The Mexican state's contribution to the conservation of the peasantry through land reform posed questions which leftward-leaning analysts could hardly have ignored, even in abstraction from the general awakening of interest in the problem of the peasantry's survival throughout the capitalist world during the 1960s. What was the ultimate position of this reconstituted peasantry with respect to the prospects for socialist advance through class-based mass politics? Perhaps land reform is ultimately a conservative force, even if, as in the Mexican case, it originates in a context of genuine popular struggle rather than as a purely manipulative intervention from above.

Mexican neo-populist arguments which denied an ultimate coincidence of interest between "rural" campesinos and "urban" workers at least grappled with some of the evident contradictions between classical Marxist theory and reality on two fronts: the explanation of actual mass politics, and the more normative but politically crucial issue of what constituted an acceptable form of "development" from the point of view of rural people. Though the Marxist-structuralist literature of the late 1960s and 1970s often tended towards economism and frequently degenerated into formalism, the debates on "modes" versus "forms" of production, characterization of peasants as "disguised proletarians," "peasantization" and "depeasantization" and so forth, represented genuine, and in comparison with the absurdities of modernization theories, extremely fruitful advances, advances which to some extent recognized the changing nature of rural reality. Most of these writers also addressed the implications of their attempt at "objective" analysis of class structures for political behavior and the development of mass-based social movements.

It would be an error to dismiss these traditions of politically committed and scholarly work as transcended by the relentless march of a process of social transformation on which they were silent because this would be patently untrue. If, for example, the Morelos studied by Arturo Warman and his associates in the 1960s was a different Morelos from that of the 1980s, Warman's contribution to the Chayanovian tradition was precisely to demonstrate that the logic of peasant economy was to be found in its capacity to adapt and innovate in response to changing circumstances which were changes in urban-rural and peasant-state relationships (Alonso et al. 1974, Warman 1976). It might, of course, seem that one of the fundamental problems with the whole field of peasant studies is that defining peasants as "farmers" (especially "family farmers") is less and less meaningful as more and more of the peasant household reproduction process becomes tied to other activities—including new forms of urban-derived non-agricultural production in rural locations as well as various forms of rural-urban migration—and the residual "peasant" labor process becomes increasingly subsumed by the agro-industrial complex of modern transnational capitalism. But as Shanin points out, the argument may cut both ways:

Peasants still form a major part of mankind, but their numbers are stationary while their share in the population of the "developing societies" is rapidly in decline. They are also being "incorporated" while the livelihoods of those who survive as rural smallholders increasingly include what has been considered as "nonpeasant" characteristics. A decline in the significance and particularity of peasantries leading to a parallel depeasantation [sic] of the social sciences can be predicted, with Chayanov assigned eventually to the archives. Or can it? ...A central element of contemporary global society is the failure of the capitalist economies as well as of state economies to advance unlimitedly and to secure general welfare in ways expected by the nineteenth century theories of progress, liberal and socialist alike ... While in the "developing societies" islands of precapitalism disappear, what comes instead is mostly not the industrial proletariat of Europe's nineteenth century but strata of plebeian survivors ... another extracapitalist pattern of social and economic existence under capitalism and/or third-worldish types of state economy ... Theoretically the analysis of modes of incorporation by a dominant political economy is in increasing need of being supplemented by the parallel study of modes of nonincorporation operating in the worlds we live in. [Shanin 1986: 22-23]

One might quarrel with some of the phrasing and conceptualization of Shanin's argument in these passages, but it is harder to reject the essential logic of what he is arguing. The demise of the "farmer-peasant" is not necessarily equivalent to the demise of the problems posed by this "awkward class." If anything, they have now become the problems posed by the once unproblematic "working classes." It may be helpful to add a quantitative dimension to the picture as it stands for Mexico at this juncture. We should bear in mind here that Mexico is one of the most industrialized countries in the contemporary Third World, and that its modern industrialization process had its beginnings, albeit limited ones, in the Porfirian era, though we should also remember that only a quarter of the Mexican working population is classified as being in industrial employment of any kind in official statistics, and that the real structures of industrial production in all its forms—including unregulated workshops, domestic outwork and so on—are often associated with patterns of household labor mobility from factory work to other forms of activity which make the association of "sectors" with discrete social classes highly problematic (Connolly 1985). Table 1 provides a superficial picture based on aggregate data.

Though the Mexican state's definition of rural population is hardly ideal from the point of view of grasping socio-economic realities, the gross figures have certain advantages over more specific kinds of measures. They underscore the fact that, however they actually earn a living, a very large absolute number of modern Mexicans may still think of themselves as campesinos. From the political point of view, this may be significant, even though only a tiny minority of campesinos is incorporated into the framework of agrarian reform or engaged in family farming outside the land reform sector. It is now a commonplace to observe that there are more Mexicans theoretically entitled to benefit under the agrarian reform legislation alive today than there were at the time of the Mexican revolution, while both land concentration and rural family income distributive inequality measures have actually risen since 1940 (Tirado de Ruíz 1971, Ginneken 1980). How relevant these facts might be politically is something which I will attempt to assess in the conclusions of this book, but it is important to keep absolute numbers of people in mind rather than proportions and percentages when considering "the problem of the countryside."

Again, however, the argument cuts both ways. According to the Fifth Agricultural Census of 1970, more than two million Mexicans held rights to land as ejidatarios under the Land Reform program. Though in theory ejidatarios outnumber "small private farmers with less than 5 hectare" by more than three to one, the 1970 population census only records around 800,000 of the former as "working their own plot of land during most of the year" (Ginneken 1980:63). Even these figures should be taken with a pinch of salt, since the official census can take little effective cognizance of the rental of ejidal land to private entrepreneurs. In a real sense "the problem of the countryside" is not the problem of the ejidatarios, who constitute a relatively privileged minority, but the problems of the landless who remain in the countryside, or move between countryside, city and the United States. These include, of course, a high proportion of the ejidatarios' own children.

The Nature of the Problem Posed by the Ejidos

This study does address itself to the "problem of the countryside" in the larger sense. Nevertheless, the ejidatarios remain a not insignificant problem in their own right. Politically, the land reform peasantry has been a major prop to the regime of the PRI in times past, and while they are becoming an increasingly small percentage of the State's mass clientele in terms of occupational and sectoral divisions, it would perhaps be simplistic to assume that their condition is no longer of any moment. Many of the people who belong to the other, numerically preponderant, "mass" sectors which the PRI seeks to mobilize are, of course, the close kin of ejidatarios and retain regular contact with their rural communities of origin even if they were born in the cities. But more significantly, perhaps, land reform is of enormous ideological significance to the regime. Despite repeated attempts to declare land redistribution a process which has reached its practical historical limits, the Mexican state has been forced to acknowledge its continuing responsibility for generating employment and "development" for the populations of rural areas. When all is said and done, despite all the changes, the ejido, that "peculiarly Mexican institution," with its archaic colonial name, remains at the heart of the state's claims to legitimacy.

Furthermore, the issue of land redistribution cannot be disposed of too readily. Not only are the gross violations of Mexico's agrarian legislation a feature of the rural social landscape and a symbol of the injustice and duplicity of the country's political class, but the increasing emphasis on land reform elsewhere in the Americas helps to perpetuate the principle of struggle for the land as it is diffused through the mass media. As Mart'nez-Alier (1977) has suggested, it would be wrong to think of the demand for land as a "primordial" peasant demand which requires no explanation since it is universal and natural: land may not necessarily be preferred to employment and a salary, and, as we will see, the major goal of many Guaracheho ejidatarios has been to emancipate their children from the peasant life. On the other hand, the picture from Latin America suggests that some of the most vigorous struggles for the land today are pursued by those who have seen the urban future and have found it wanting. Given the alternatives presented by modern Third World socio-economic systems, it seems unlikely indeed that land reform will cease to be one of the major directions of popular struggle for the foreseeable future. In the time it takes to read this introduction, someone, somewhere will probably have died for the land.

This is not a book about "development policy." There are, of course, powerful arguments concerning the advantages of land reform as a means of securing a form of "development" which achieves progress towards establishing the mimimum levels of general social welfare that were so patently not achieved in the early epoch of post-war economic growth (Griffin 1974). In reviewing the Mexican experience, it is hard not to dwell on its negative aspects. One of my reasons for deciding to work in a region like the Ciénega was to take what would appear to provide a favorable context for ejidal fanning. Although not all the ejidatarios in Guaracha and neighboring communities have access to prime agricultural land, a majority does possess rights to land which is fertile and irrigated. The entire terrain of the ex-hacienda on the plain was distributed among ejidos. Though the region was relatively neglected in terms of infrastructural investments by the central government until the 1960s, in more recent years substantial investments have been made to improve on what was already a comparatively good base. Yet until the 1970s, the region was scarcely a success story for the ejidal system. Vast numbers of peasant plots were rented long-term to a capitalist neolatifundist whom the ejidatarios referred to as their new patrón. If the epoch of neolatifundismo of this kind ended with the introduction of the new state policies to "refunctionalize" the ejidos, and a large proportion of ejidatarios returned to sowing their land in the 1970s, they did so under conditions which set strict limits on the income which could be derived from peasant farming by those dependent on the official credit system, promoted new forms of subsumption of the peasant process by capital, fostered capital accumulation in other sectors, and left the peasants, as most of them see it, where they have always been, "at the bottom of Mexican society."

During my fieldwork, I was able to observe what were probably the last moments of this phase of the region's agrarian history, since the deepening crisis into which the country was sliding was already producing signs of change: on my subsequent visits in 1986 and 1988, rental of ejidal land (and international migration) had shown a further increase. But rental of ejidal land today is a somewhat different phenomenon from that of the earlier period of neolatifundismo. Today's farmers include numerous members of the state's own agrarian bureaucracy or investors from outside the agrarian sector, like restaurant owners from the regional towns, as well as longer established elements of the regional agrarian bourgeoisie. In the 1960s, largescale capitalist tomato growers entered the region from the Bajío, primarily with a view to expanding production for the growing Guadalajara metropolitan market. Today vegetable production on a large scale is increasingly orientated to export. The "modernization" and internationalization of local production systems hands local economic power to truckers and owners of machines, though such people are, in reality, only the immediate local manifestations of much larger and infinitely more powerful forces. As another facet of these larger processes, emigrados with steady jobs in the U.S. acquire titles to ejidal land. The 1980s therefore have brought a multitude of "new men" into the contest to capture control of the patrimony of peasant communities.

These brief observations give an indication of the framework which underpins the analysis of this book: the notion that the post-revolutionary agrarian history of the Ciénega corresponds to a sequence of agrarian cycles characterized by the ebb and flow of peasant commitment to farming the land, and the expansion and contraction of various forms of capitalism and state intervention. The nature of the cycles is more complex than these brief and superficial statements can indicate, but it is important to emphasize from the outset that there are irreversible secular processes of change at work beneath these apparently cyclical phenomena.

From the point of view of the reproduction of the ejidal system, one such secular trend is the rising value of the land. In theory, ejidatarios receive only (inalienable) rights to use the land, which can only be transferred in perpetuity through direct inheritance or the nomination of an eligible successor approved by the agrarian community in the event of no heir being available. In practice, land rights are commoditized. Competition between capitalist renters, and the prospect of securing good returns to investment in production available to those who have the capital, has driven up the going rate for definitive sale of ejidal rights, and acquiring such rights has progressively become beyond the means of the children of peasant families. In past decades, "new ejidatarios" were able to acquire land through savings from seasonal migration to the United States or other forms of work outside the community. Today's "new ejidatarios" are often people earning professional salaries and state employees. If yesterday's purchasers of ejidal titles included such evidently ineligible figures as the ex-neolatifundist, a resident of another community, what is striking about the situation today is that the same entrepreneur has been able to add to his holdings in Guaracha and, worse, has been joined by members of the agrarian bureaucracy. Though the changing land tenure situation is far from being the only secular tendency at work, and is one which could theoretically be reversed at a stroke simply by an agrarian census and purging of the ejidal rolls in conformity with the agrarian laws, the point is, of course, that nothing has ever been done in this region in this sphere. At the present rate, half the ejidatarios may be holders of university degrees by the end of the century.

In the light of what I have already said about the role of "the problem of the countryside" and the place of the ejidal system in Mexico's system of political control, it becomes apparent that we are not simply looking at land reform's "failures" from the point of view of the criteria defined by development policy. We are looking at the wholesale perversion of theory by practice and the incipient total disintegration of the framework of land reform agriculture in a productive and fertile region of the country. In the past, the peasants were exploited by merchant capital while they tried to farm their land, and ceased to be "real ejidatarios" when they rented their land for long periods. But the situation was reversible to a degree: the reexpansion of the state made the ejidatarios "work for capitalism" in new ways, but they were, to an extent, reconstituted as a rural underclass of a specific type, enjoying clear advantages over the landless day-laborers as dueños of their land. It is the possibility of this recurring in the future which is now increasingly in question. The more the ranks of the peasant ejidatarios are eroded, the more dangerous the situation may become politically in the long-term in the absence of an alternative and comprehensive solution to the country's mounting social problems.

In writing about the negative side of Mexico's land reform program, my objective is not really to provoke feelings of righteous indignation but to direct analysis towards a realistic and concrete view of the realities of peasant life. It is pointless to discuss these matters as if they were pathologies with respect to the normal functioning of the system, as many analyses seem to do. If one wants to ask what determined the income from farming of a significant number of ejidatarios in Guaracha in certain years, it would be meaningless to do the calculation in terms of harvest volume multiplied by price minus costs: what in fact determined it was that the inspectors of BANRURAL, the official credit bank, and the ANAGSA, the state agricultural insurance agency, volunteered to record harvests as a total loss, allowing the ejidatarios to sell their crops at the free market price without repaying their debts to the state.

The question immediately at issue here is how much the peasants are willing to pay the inspectors for this service, and how much public money can be wasted or stolen by the bureaucracy without exciting action from a higher level. These questions in turn invite us to explain the social and political conditions which promote such behavior on the part of the servants of the state, on the one hand, and a lack of effective collective resistance on the part of the ejidatarios when it is they who are the chief victims of such abuses—as they often are-on the other. In posing these questions, one begins to see that there is more to these phenomena than a term like "corruption" would imply, and that the nature of bureaucracy in Mexico may be one of several effects of the structures of social power determining the basis and limits of state power in the country. Similarly, the vicissitudes of agrarian policy, with their spectacular effects on agrarian structures, demand an explanation in terms of the complex shifts in national and international social forces which underlie them.

A concern with the scandalous becomes an analytical necessity if one wishes to explain the realities of Mexican social and political life and, by extension, that of many other parts of the world. Nevertheless, it must be conceded that many Mexicans are understandably irritated when foreigners come to their country to do studies which belittle their achievements and dwell only on the negative. Though I could offer an additional defense of such a focus by observing that a majority of the peasantry express themselves dissatisfied with their treatment, I would prefer to emphasize a different point. If one compares the situation I describe for the early 1980s with earlier epochs, especially that of the hacienda, or with the situation obtaining to the south in Central America, then the positive achievements of Mexico's land reform program are apparent enough. Land reform has not solved the problems of rural poverty. It only assists a minority of rural people, and peasants who hold irrigated land in the Ciénega de Chapala are clearly better placed than the many who hold resources of inferior quality. What emerges from this often dismal history is that land reform is capable of improving the lot of rural people. It is the very failures of Mexico's land reform program which underscore what might have been achieved, and what might yet be achieved.

My purpose is not to suggest what would be desirable in the abstract: it is to describe what has been and is and address the problem of causes. The future will be, and must be, determined by what Mexicans choose to do about their situation, though the simplistic voluntarism of the use of the word "choice" here abstracts utterly from the powerful historical forces which are likely to foreclose on many developmental possibilities being realized. Nevertheless, if voluntarism, to some extent a hallmark of western Marxism in the post-war period (Skocpol 1979), is indefensible, so is the idea that history is determined by blind, mechanistic historical forces immune to the impact of human struggles to transform reality. The ideas and data presented by this and other studies may or may not play a minor role in the evolving stream of social and political action which will determine Mexico's future. But what is certain is that such action will shape the future as much as it has shaped the past. Furthermore, as this book will seek to demonstrate, social struggles in which peasants engage cannot be defined in terms of self-conscious collective political action alone, the visible struggles accompanied by slogans and banners. Had the Mexican peasantry's survival been determined solely by its participation in social movements of an overtly "politicized" kind, movements which history demonstrates are so susceptible to the processes of cooptation and incorporation by other strata, today's reality might be very different from what it is. There are other ways in which peasants have contested the forces of capitalist expansion and political domination which can only be discerned in the less visible history of the practice of daily life. They are unromantic forms of social action, and quite ineffective by conventional definitions of what class struggles are supposed to achieve: their effects are neither "conservative" in the strict sense, since survival has been accompanied by transformation, nor are they "revolutionary", either in effect or intent. But they are, I will argue, significant determinants of historical reality.

Agricultural Change and Rural Transformations

The larger problem of which the ejidos constitute a facet, that of Mexican agriculture as a whole, certainly remains of undiminished significance. Though the contribution of agro-pastoral activities to Mexico's GDP fell from 20.6% to 8.4% in the years between 1939 and 1980, while that of manufacturing industry rose from 14.7% to 23.1% over the same period, the agricultural sector made a prominent contribution to the process of industrialization, and the impact of the burdens placed upon the peasant sector in pursuit of industrialization eventuated in a crisis of basic foodstuffs production which, together with mounting rural unrest, forced significant responses from the Mexican state in the late 1960s and 1970s. Today, in another epoch of crisis, the problems of sustaining agricultural productivity and balancing earnings derived from exports of meat, fruit and vegetables against the import and wider social costs entailed in the transfer of resources from basic grains production have scarcely diminished in significance. Now, however, we are in an era where industrialization has transformed agriculture itself.

For the majority of the population of the "developed" world, agriculture has become a profoundly mysterious phenomenon, about which people know little and care less, except in so far as they feel compassion for those they see starving or resentment when politicians draw their attention to the extent to which their taxes are being used to subsidize farmers. Agriculture is also rather boring and "natural." Unfortunately, modern agriculture is a highly unnatural activity, and the functioning of agricultural systems is much more dependent on political economies than the vagaries of the climate, while the man-made ecosystems which such political economies create generate their own share of unnatural disasters. It is vitally important to understand that the political economy which shapes Mexican agriculture is not entirely located within the country's national boundaries, and that Mexican campesinos gain a livelihood from a largar agrarian political economy whose center lies in the United States, whether or not the production of the raw crop is carried out there or to the south.

The impact on agriculture of the internationalization of capital has been subject to some penetrating analyses in the Mexican case (Barkin and Suárez 1985, Sanderson 1986, Durán 1988). This is perhaps to be expected given that while a large part of what has happened in Mexico has also happened in other regions as a consequence of the developmental dynamics of the capitalist world economy, Mexico enjoys the unique privilege of a land frontier with the heartland of northern economic power, whose own "core" is now located to the west, articulated to the new web of relationships which have shifted the center of gravity of the world economy from the Atlantic to the Pacific Rim. Questions pertaining to the internationalization of Mexican agriculture can scarcely be avoided in a study of Michoacán, since the regions of Zamora and Apatzingán have been the sites of particularly significant North American agribusiness developments (Feder 1978). The Zamora strawberry agro-industry's direct influence spread as far west as the Ciénega, from which it has recruited large numbers of young girls and more mature woman workers, though on a declining scale in recent years. But the entire productive structure of Mexican agriculture has been changed through more indirect processes of "internationalization". Direct investment in production, contract farming systems and processing facilities by North American capitalism may wax and than wane, as it has done in Zamora, where the congeladoras have now been sold off to local entrepreneurs. But there are more subtle relations of transformation and domination at work in the evolution of the modern agro-industrial complex, and the nature of the Mexican state's "refunctionalization" of the ejidal system has played an important role in ensuring that these tendencies have impacted with maximum vigor. The initial phases of the decline in the marketed surplus of basic foodstuffs offered by the peasant producer appeared to be the result simply of the domestic policy of squeezing the peasantry to accelerate the pace of industrialization, and therefore correctable by a change of policy. But it is no longer clear whether the improvements the implementation of changes brought about can stem the impact of these deeper forces in the long term. As Barkin and Suárez have argued, the "end of food self-sufficiency" seems to be a general trend throughout the Third World, and a normal consequence of capitalist expansion and the integration of peasants into the world market (1985:240-242). That the situation in Mexico is not worse than it is may owe more to the forms of economic rationality pursued by peasant households under adverse conditions than to official policy. But the real issue is how far the objective social conditions for such behavior are being progressively eroded.

The Ciénaga provides a useful test case for pursuing these questions in a number of respects. As I have already notad, ejidal land includes fertile irrigated terrain, capable of producing large commercial surpluses and susceptible to the unfettered pursuit of the processes of agricultural "modernization." The ejidal farmers who possess such lands have theoretically been the prime beneficiaries of state policies aimed at refunctionalizing the ejidal system. The Guaracha case does not permit an analysis of the alternative policy of ejidal collectivization in its modern forms, since it remains an ejido basad on individual tenure and cultivation. It is, however, important to stress that the ejido did begin as a collective enterprise producing sugar. Furthermore, it was a particularly important case of early ejidal collectivization under Cdrdenas, because the newly emancipated peasants received the ex-hacienda's mill, to be run as a cooperative venture, as well as the land. Since the early and spectacular failure of this experiment, all further attempts to promote cooperative projects have failed in an equally radical fashion in this community. Though this is in itself of more than passing interest, it does foreclose on certain avenues of discussion. What the case does offer is the possibility of a fairly full reflection on the situation with respect to individualized forms of cultivation, since some ejidatarios in Guaracha must dry-farm parcelas located on the stoney lower slopes of the hills surrounding the plain, and cannot in practice opt for the mechanized styles of cultivation which have become dominant on the plain itself. Furthermore, both ejidatarios and non-ejidatarios have access to rain-fed plots higher up in the hills, which ware assigned to the hacienda peons as subsistence plots to supplement their cash wages and maize rations. These plots, still known by their Purhépecha name as ecuaros, have in some cases been made suitable for plow cultivation by animals through clearance of larger stones and terracing, a transformation which represents a considerable investment of labor. From the techno-ecological point of view, therefore, Guaracha and its neighboring communities provide a cross-section of the different conditions of agricultural production faced by Mexico's peasant farmers today.

Non-ecological conditions of production, in particular access to credit, also vary for this aggregate of farmers in a way which reflects the general shifts in financial aid offered to the peasantry at national level under changing state policy orientations and national economic conditions. Where the marginal farming enterprise is located at a particular moment of time is, however, influenced by the conjunction of local conditions: besides the impact of regional price structures, official credit levels have been set in accordance with average regional costs of production, penalizing those farmers with inferior resource endowments given the official bank's assumptions about appropriate technical inputs, degree of mechanization, etc. Furthermore, official policy also influences the location of peasant agriculture as a whole in relation to the margins which can be identified for the agricultural production system as a totality: the real cost of credit determines which, if any, peasant producers are capable of sharing in the appropriation of differential rents and whether they can recover the value of unpaid labor-time which has been embodied in output whose market price is related to average enterprise productivity in a given crop sector—which may be international—and the relative market share of enterprises with different cost structures (Margulis 1979, Gledhill 1981).

Although these matters are usually discussed in terms of a gross division between "capitalist" and "peasant" enterprises, on the assumption that these forms of production are distinguished by their total and secondary dependence on wage labor respectively, peasants in the Ciénega, and by implication elsewhere, are really far too heterogeneous a category for this style of analysis to be totally satisfactory. This point of view will be reinforced when we consider the actual use of wage-labor by peasant farmers in the Ciénega and the extent to which peasant reproduction is commoditized under contemporary conditions. Furthermore, household reproduction is based on the combination of a variety of different income sources, arising from the different contributions of individual family members. Internal household distributive relations, and the constitution of the effective income-sharing consumption unit, vary, as does the combination of activities which together provide overall disposable income. All this leads to varying possibilities for substituting family for paid labor and reducing cash expenditures by household production and pooling arrangements. Thus, beyond the basic differences in possible family economic strategies determined by the unequal distribution of farming resources of varying quality, and the differing possibilities for access to credit determined by these agricultural resource endowments, part of the variability in the overall economic strategies of households is determined by their nonagricultural activities.

To make matters more complex still, additional agro-pastoral activities such as goat-herding may be combined with farming, and are occasionally pursued in preference to it. Though one might expect a considerable variety in the economic strategies of the poorer non-ejidatarios, there is, in fact, a very substantial variation among the ejidatarios themselves. Nevertheless, one can still identify a particularly significant margin in the case of the ejidal producers whose land is of sufficient quality to secure a rental income, that at which the ejidatario decides to withdraw from direct involvement in production. Ejidatarios are often unable to secure a family livelihood from the parcela and are constantly faced with choices about how to allocate their labor time between different activities and employment possibilities which have long since become integral to what is generally a semi-proletarian form of peasant existence. One of the most striking features of the current situation in the Ciénega, from the point of view of those peasants whose position was significantly improved under the new state agrarian regime of the 1970s, is the rapid and accelerating increase in the costs of irrigation water and nonlabor inputs: though they may never have been able to secure a family livelihood from their farming, the huge sums to be repaid at the present time have produced a wave of defaults and voluntary withdrawals from the official system, and frequently the total abandonment of farming the parcela. Though short-term expectations shift in accordance with general economic uncertainty, consumer price inflation, and problems with the state guaranteed farm price system, there seems to be little doubt that the margin relevant to levels of peasant participation in production has been shifted in structural terms.

The implications from the point of view of Mexico's agricultural problems are clear: peasant farming of high quality land tends to give way to disguised capitalist farming unless the state intervenes to make it worthwhile for such farmers to farm. The variations in resource endowments, which apparently make the ejidatarios with irrigated land less marginal farmers than those sweating with their horse-drawn plows on unirrigated rocky terrain, do not, in fact, provide a reliable guide to the political economy of peasant agriculture under modern conditions. Of course it might be argued, cynically, that it does not really matter whether a crop is grown by an poor ejidatario, a restaurantowner from the town of Sahuayo, an Inspector from BANRURAL in Jiquilpan, or the village primary school headmaster. Indeed, since the ejidatario is probably in a better position to borrow money to take a trip to California, and may, not need to do so if he uses the money from renting his land, the situation could prove quite satisfactory from the national point of view. The displaced ejidatario exports his unemployment, while the possessor of capital who takes over his agricultural role may have the financial wherewithal to produce a splendid harvest at no risk or cost to the harrassed federal treasury.

The problem is that, aside from any considerations relating to the supposed social policy objectives of agrarian policy, restauranteurs are more likely to be interested in sowing high-value vegetables than basic grains, and there is always the danger that speculative capital entering the ejidos may not husband its ecological resources as it should, as Díaz-Polanco and Montandon (1978) demonstrated for the case of neolatifundist production of sorghum in the Bajío. In practice, this latter problem has not been too serious in the Ciénega. Capitalist entrepreneurs have tended to pursue cultivation practices which would be deemed superior by the standards of modernizers, partly because they have frequently rented long-term, even if their heavy use of chemical fertilizers and pesticides might be subject to a broader kind of environmentalist critique. Renters from the ranks of the agrarian bureaucracy sometimes make improvements to the land (often, of course, without cost to themselves through "creative accounting" with the use of public resources). This was a carrot which was frequently offered to peasants to encourage them to rent in days when land was in shorter supply. From the ecosystemic point of view, the sorghum monocropping regime which a majority of ejidatario farmers chose to pursue in years gone by, in consequence of their undercapitalization, shortage of disposable family labor and the official credit system, has been far more damaging. But suppose we discount the ecological aspect and ignore the possible social and political implications of a renewed disintegration of the ejidal system, given the expectations aroused by the improved situation of the previous decade. It is still hard to conduct an agricultural policy on this laissez-faire basis. Market forces, national and international, will continue to transform the structure of Mexican agricultural production in a way which the state cannot control.

In the longer term, the cost may be high. It would certainly be high if technologically modernized peasant producers can indeed provide supplies of basic grains to the domestic market at lower cost than the alternatives of fully-fledged capitalist production or importation. And if the cost of persuading capitalists to reenter the sector of basic grains production most abandoned a decade ago (Appendini and Salles 1979) were deemed prohibitive, the balance of export earnings over import costs after further reallocation of resources to high value crops for urban and export production would have to become more favorable than it is at present. Furthermore, the agribusiness agro-pastoral sector itself would be affected by the end of peasant production of sorghum for balanced feedstuffs: whether this sector should be promoted at all may be open to doubt, but the impact of a decline of peasant production of raw materials for the agro-industrial system would be unlikely to lead to a retrenchment of that system rather than adjustments to prices and an increase in its already high social cost. The implication of present trends is that any sort of coherent agricultural policy is becoming increasingly problematic. I will seek to demonstrate that official policy has to a degree been premised on a series of misconceptions about the nature of the modern peasantry, partly as a result of the phantasmagorical representations of "the peasant" which have become embedded in the rhetoric of the post-revolutionary regime. But one might still argue that having the possibility of implementing even an imperfect policy is preferable to lacking any possibility of implementing one and letting matters take an uncertain and perilous course by refusing to acknowledge that anything is happening.

But it is already apparent that the analysis of the situation of today's campesinos cannot be identified solely with the analysis of peasant farming. The working ejidatarios' households are seldom sustained exclusively by agro-pastoral production activities. Income remissions from migrant children are frequently crucial to the quality of life enjoyed by mature peasant households and may, indeed, play a key role in financing the farm. In the earlier stages of family development, most ejidatarios must supplement agro-pastoral production with wage-labor for others. The types of wage-labor undertaken by ejidatarios are typically but not exclusively agricultural and the "others" for whom they labor are varied. They may work for other peasants of similar socio-economic status, for capitalist farmers and machine owners who have emerged from the ranks of the ejidatarios, or for rural entrepreneurs of various kinds from outside the ejidos. Some obtain work locally in the public sector, while others seek work in some more distant location—including that offered by the rural and urban patrones north of the border in the United States.

Peasants as Proletarians

Michoacán has been one of the major sources of international migrant labor to the United States in general, and to California in particular. This relationship figures repeatedly in the analysis which follows. A few of the better-off ejidatarios of today are emigrados who invest their earnings from the U.S. in rental of land, purchase of machinery, and family education. From one case, this apparent "success story" should be qualified by the information that the person concerned now faces a premature death as the result of twenty years of inhaling the pesticides used by his unionized, multinational employer in Arizona. Nor should we imagine that all emigrados have prospered to any great extent: those who have worked as casual laborers in agriculture may be little better off than those who have gone illegally, and the journey "al Norte" has not been, as we will see, a generally enriching experience for the majority. We will also see, however, that the flow of labor has not been solely (or in some periods even predominantly) directed towards agriculture, and from the agricultural perspective, it is important to see that there has been an emergent tendency for what is essentially the agricultural production of the United States to be relocated south of the border.

Where this study differs from most accounts of this larger relationship between national entities is in taking seriously the idea that we are dealing with a single social formation. It is not simply economic interdependence across national boundaries which needs to be taken into account, but the cumulative development of a population of Mexican origin in the United States and a persistent structure of social interaction linking families north and south of the border. This transnational community remains such because its articulation to the dominant society militates against the abandonment of its cultural capital and identity, though not without inducing severe contradictions, particularly for young people who are forced to return to Mexico after growing up in the North.

As well as going "al Norte," quite a number of ejidatarios and their children have undertaken seasonal migration to other parts of Mexico, cutting sugar cane in Costa Rica, Sinaloa, for example. This type of migratory movement is also a socially structured one: personal contacts are established with the employer, leading to systematic recruitment of workers from this particular, if distant, locale. Such processes are the essence of the Mexican labor market, which seldom functions on the basis of random, impersonal contact between the buyers and sellers of labor power. But a "semi-proletarian" strategy within the agricultural sector, whether or not it involves internal or international migration, is not the only one to be found in communities like Guaracha. Many of the ejidatarios have, at some time during their lives, experienced work in the cities, either as workers or, in a few cases, as self-employed people. One of the original ejidatarios has spent his entire life working in factories, returning to his land only as a pensioner. Some of the more active farmers in the history of the ejido turned out to have experimented with wage-labor before returning to the land. Although there are non-ejidatarios who have spent their entire lives sowing an ecuaro and working as jornaleros in the region, perhaps occasionally making a foray to cut cane in the Tierra Caliente, the life histories of many landless people in Guaracha also show a marked pattern of past mobility, though there are differences between the migration patterns characteristic of members of ejidatario and non-ejidatario households. But a full understanding of the process of social transformation experienced by the campesinos of Guaracha cannot be obtained simply by taking a snapshot of the community of residence at the time of fieldwork and the working careers of male household heads. We need a larger historical perspective.

The first point to be emphasized is that the Guaracheños did not begin the twentieth century as "peasants" in the conventional sense. Although the hacienda's agrarian regime allowed the limited reconstitution of the peasant form of production within it after achieving an effective monopoly of land based on the complete expropriation of the peasant communities it encapsulated, the peones acasillados were essentially proletarianized, albeit in a somewhat contradictory way. Land reform did not, at the outset, reconstitute them as peasants, since the state capitalist organization of the collective ejido did not fundamentally transform the hacienda regime, it eliminated some of its barbarities and exacerbated existing forms of spontaneous resistance to the direct subsumption of their labor power on the part of the ex-peones. Nevertheless, while one of the problems with the Collective was that Don Lázaro had originally promised the peones individual land rights and the life of a peasant farmer, many had difficulties in making the practical adjustment to the new regime when it was finally introduced in 1940. Though the ex-peons had been imperfect proletarians, many became imperfect peasants, since they had become habituated to certain conditions of proletarian existence of which they were suddenly, and unexpectedly, deprived at the end of the Cárdenas era. More significantly, in the immediate term, the creation of an individual ejido and the progressive collapse of the sugar industry in Guaracha created an employment crisis for the majority who did not obtain land. The closure of the mill was accompanied by substantial rural-urban migration.

But it was not simply those who did not receive land who left for the cities. So did a significant proportion of those who became ejidatarios. The post-war period was one of substantial industrial growth and growing demand for labor, but the emigration was not simply a spontaneous response to the effects of a state policy which increasingly favored the urban over the peasant sector. To an extent, Cárdenas himself promoted it, by arranging for the creation of various employment-generating schemes oriented towards satisfying those who had not benefited from land reform. Political patronage relationships also led to people finding work in the public sector. Later on, such patronage created further opportunities indirectly: Cárdenas personally assisted in the education of some of the children of his former peasant allies and clients, but he also offered opportunities to the community at large through the establishment of an Escuela Práctica de Agricultura in the casco of the former hacienda. It produced more urban professionals than farmers, and they found their less fortunate paisanos manual jobs in both the public and private sectors in the cities. The family of the ex-hacendado also offered Guaracheños positions in its houses in Mexico. Given that even relatively menial occupations were often "opportunities," relative to what the countryside offered the majority in this period, and that many of them were allocated by particularistic social networks in which ejidatarios were firmly inserted, it is not surprising that they were taken up by families who had land as well as those who lacked it.

From the outset, then, land reform in Guaracha did not only constitute a process of "repeasantization". A form of agrarian capitalism characterized by a high demand for labor was replaced by a peasant agricultural regime which absorbed much less of the disposable labor-power of the community, promoting rural-urban migration of relative surplus population. This was then reinforced by the abandonment of land by a number of ejidatarios who opted for the urban alternatives, generally another form of proletarianization. But the relative surplus population continued to grow as a result of "natural increase" (despite the fact that the intensified production of the neolatifundio increased local demand for labor, at least in its early stages). Although mechanization was very limited until the late 1950s and early 1960s, and the bracero program encouraged people to attempt the combination of seasonal international migration and continued residence in the village, the rural-urban drift accelerated as ejidatarios' children grew up in a climate unfavorable to farming prosperity and holders of land titles continued to rent their land. The landless had poorer prospects of going "al Norte" than landholders, who found it much easier to raise the loans necessary to pay the bribes needed to receive a place on the official lists. This factor favored their transition to the cities, but the movement was not restricted to the landless, and cannot be explained simply in terms of "unemployment" initially, though it is certainly related to the hardships of rural life in this period and lack of expectations of improvement.

Over the last two decades relative surplus population has been increased by technical transformation: mechanization, the increasing use of labor saving chemical inputs such as herbicides, and the increasing substitution of the industrialized crops like sorghum and safflower for the more labor intensive wheat, maize and beans. The trend was countered to only a marginal extent by government programs to promote cultivation of basic foodstuffs. Declining local demand for labor has been offset somewhat more by the penetration of labor intensive capitalist vegetable production into the ejidal system of the Ciénega. But the trend in adult male employment remains a downward one. In contrast to earlier decades, the increasing relative surplus population can no longer be absorbed by expanding domestic urban demand for labor.

Though the roots of this problem are structural, and scarcely confined to Mexico, the end of the bracero program in 1964 marked the beginning of a new epoch, as millions of mestizo rural Mexicans flooded back into the domestic labor market, and the crisis of the 1980s has created pressures of a new order of magnitude as existing jobs have disappeared throughout the economy. Since during the 1970s the ejidatarios themselves reduced their absorption of labor, most now see the solution to their dilemma as a need to create "rural industries": some even focus their attention on the idea of returning to a regime of sugar cultivation and processing, an ironic demand in historical terms. International migration has continued to offer the region a vent for its surplus population, despite the increasing problems inherent in "crossing the line," technological change in North American agriculture, and the increasing number of people competing for jobs. Although agricultural demand for labor in the North is falling, making life hard for older people, younger men and women have found new niches in the American labor market, in restaurants, catering, supermarket checkouts and the expanding service sectors of this ever more affluent, and ever more unequal, society. Factory or workshop jobs in manufacturing are also to be obtained: in one case, a Guaracheño ejidatario in his sixties was found work by his sons in a factory making furniture in Bemais. But even if it is simply a fantasy to imagine that there is a "solution" to the problem of "undocumented" migration in the sense that effective action could be taken from the American side to stop it, there are increasingly disturbing signs that the U.S. economy is now finally reaching a point of saturation in its demand for what monthly becomes an increasing number of Mexican migrants.

It is important to see what these various different routes out of peasant society amount to as a macro-social process. campesino communities like Guaracha are precisely the kind of communities which historically produced the modern Mexican urban working class. While it is true that the bulk of urban population growth since the 1960s has come from natural increase among urban resident families rather than from rural-urban migration, there is scarcely a household in Guaracha which does not have children working in the cities, and most of the older people have siblings or other close relatives who are deeply entrenched in the urban milieu. Many of the ejidatarios who gave up their land definitively did so because relatives or children offered them a niche in the urban economy. One of the implications of this simple fact about most mestizo rural communities is that urban crisis and rural crisis may not be as discrete as those who argue that the peasants will remain contented while they have beans and tortillas imagine (leaving aside the fact that many rural people today, even the farmers, buy their beans and tortillas). Rural households feel the impact of declining urban living standards when supplementary income derived from urban resident members of the family declines. Similarly, these same communities are locked into economic networks which involve relatives working in non-agricultural activities north of the border. What is different about the modern rural exodus from that of years past, to return to Shanin's point, is that it is increasingly a matter of the expansion of membership of social underclasses whose incorporation into classical forms of capitalist wage-labor relationships is at most transitory.

Studies of the contemporary working class of core countries have demonstrated an increasing polarization between households which are stably incorporated into the ranks of salaried labor, generally with multiple wage-earners of both sexes, and those which are not (Pahl 1984). Though such studies suggest that classical models of the proletariat may be becoming increasingly irrelevant, their inadequacy for countries like Mexico may not stem from any convergence between the different patterns of capitalist development. A number of recent studies suggest that the typical urban working class household moves over its development cycle from a greater to a lesser involvement in wage-labor (Connolly 1985, Gonzdlez de la Rocha 1986). What are generally low absolute levels of remuneration for all forms of labor power generate a working class which is relatively homogeneous in living standards in comparison with the situation envisaged by advocates of the "labor aristocracy" versus "marginalized mass/informal sector" distinction, though it is fragmented in terms of the relations of production and inter- and intea-household relations which organize its social existence.

I contend that the reproduction of the peasantry is an aspect of a much larger process of class formation, which is, in this particular case, not even restricted to a single national territorial space. The dominant tendency is rural-urban outmigration: most of the Ciénega peasantry of 1940 ultimately left for the cities. Period. But they were, of course, demographically replaced. In terms of the processes which occur within the rural milieu itself, there is an apparent cycle of recomposition and decomposition at work, but underneath this cycle is a secular process in which the nature of peasant life changes profoundly. It is not simply that an original population is redistributed, as its numbers grow, into a series of differentiated socio-economic niches corresponding to transformations in the larger economic system. Not only individual households, but single human individuals, move between niches and create a process in which they are linked together, not only as material modes of reproduction of labor power, but as alternative conditions of social existence which impact on the consciousness of the human actors.

Thus, on the one hand, we can study the "reproduction of the peasantry" as an objectified process which is a condition for various types of capitalist and state capitalist accumulation: peasant households are reproduced over time through combinations of urban and rural wage labor and production and in their turn supply products and labor power to various branches of capitalist enterprise. The complex combination of rural and urban activity sustains a laboring population exploited in a number of different ways at different times and in different places, and is not the unique responsibility of any particular faction or sector of the international capitalism which draws upon its labors. Traditional concepts, in particular the notion of a "reserve army of labor," do not fully capture the complexity of the processes involved. A son's wagelabor in a factory in Los Angeles may be a condition for a father's production of sorghum for processing by a balanced-feedstuffs plant in Jalisco controlled by a multinational company. The cultivation of sorghum in the Ciénega provides possibilities for capital accumulation by local machine-owners, who buy machines whose components are fabricated in Mexico City and Seattle. And so on. There is no "grand design" here on the part of the capitalist system, but a complex series of adaptations to evolving circumstances on the part of different sectors or fractions of a heterogeneous and to some extent conflictive capitalism, national state apparatuses and the rural underclasses themselves.

In consequence of these adaptations, which shift over time, capitalism and a peasantry reproduce themselves through a process of transformation. The peasants reproduce themselves through an existence which is generally semiproletarian, and capitalism extracts surplus-value from them in any way it can get it, but with an increasing emphasis on means which are characteristic of the agro-industrial era rather than merchant capitalist subsumption of a more autonomous peasant production process. In the last analysis the peasantry of the Ciénega de Chapala has been reproduced through the allocation of its labor power to activities outside the farm. In some cases, these activities have involved social mobility for household members, giving those who remain access to capital otherwise unobtainable. In other cases, peasant households have transformed themselves totally by wholesale migration. There have, however, been some moves in the opposite direction, and more significantly, polarization of life chances in the countryside today may be increasing as the lowest strata of the rural underclasses find themselves facing increasingly restricted migratory opportunities both domestically and internationally. But beyond the effects of peasants' own economic strategies, significant though they are, in the long term the intervention of the state has clearly played a central role in the process of peasant "conservation" in the Ciénega. This intervention has been a response both to economic pressures caused by the withdrawal of capitalist production from certain key sectors and to the political pressures created by peasant unrest outside the region.

This last observation brings us on to the second dimension of "peasant reproduction," that which involves peasant action which places demands on "the system": the conscious striving after the goal of survival as "peasants" rather than as people. There is a social category "campesinos" which exists, not only in the minds of academics and politicians, but in the minds of real people who use it to define their social identity. These people are not socially homogeneous: some of them could, in a meaningful way, be described as "rural capitalists" and in some contexts some of these people define themselves as having other, alternative, identities. They paint the word agricultor on the side of their pickups, and feel different from the campesinos who are not, as they are, "businessmen." Yet there are other contexts in which these same people also see themselves as part of a campesino community. What is interesting to investigate is why, despite the profound social cleavages and antagonisms in communities like Guaracha, the campesino identity retains its force. There is, of course, no guarantee that to be a campesino today means the same thing to the people who see themselves as having such an identity as it meant twenty or forty years ago, or even that people who occupy different places in the local system understand being a campesino in the same way even if they acknowledge each other's claims to the status. The socio-economic conditions of peasant reproduction have clearly had a substantial impact on the consciousness of the less well-off peasants, landed and landless.

In a region like the Ciénega we must consider processes of relatively long duration. Besides the experience of what might well be accurately defined by Wallerstein's phrase "coerced cash-crop labor" during the epoch of the hacienda, a significant number of local people in the Ciénega experienced industrial wage-labor of the classical kind in the foundries of Chicago and other American cities during the first great wave of emigration terminated by the mass deportations of 1929. The conditions of bracero labor were more cognate with peonage, though they generally permitted far higher standards of male personal consumption and a very different quality of life in the dance halls, bars and whorehouses. For many, the experience, relatively speaking, was liberating. But this was only one facet of a wider experience which also involved very substantial movement between countryside and city within Mexico. As the years have gone by, and participation in non-agricultural urban activities has increased, ample scope for conscious reflection on the basis of experience has been created. Today, people spontaneously discuss the pros and cons of the different conditions of laboring existence notionally available to them. No ejidatario thinks it odd to be asked whether they would rather have the land or work as they did in the U.S.: opinions differ, but the most frequent response is to give a precise definition of the sort of stable wage labor which would be more satisfactory than sowing with BANRURAL. Casual speculations on the nature of Castro's socialism over a beer are not uncommon.

Ultimately, of course, behavior is determined by what people see as the realistic possibilities before them, and it is in this context that we should understand their desire to remain campesinos, given that they cannot become doctors, teachers or lawyers. Being a campesino may mean going al alambre to the North, or it may mean being relatively poor but not having to do much work by tending the parcela, but either of these things may be preferable to being another member of the urban underclasses.

I have argued elsewhere that the generic term campesino cannot be reduced to a simple socio-economic category (Gledhill 1985). To be a campesino is to be politically dominated. In an important sense, it is this which unites rich and poor peasants: both are subordinated within larger structures of social power and see themselves as subordinated in that way. So the argument I have just advanced about the impact of occupational experience on forms of consciousness cannot be complete. I have not outlined the broader web of significations attached to the different modes of the peasantworker's experience. The U.S. is not simply a place where one has to work hard but gets better wages. It is a place which peasants see as having a different system of law and government, again with its pros and cons. campesino consciousness is a product of the synthesis of different experiences, in which the contradictions of one social situation are weighed against the contradictions of another. The result is that the different constructions of reality form in opposition to each other, in their mutual contrasts.

Since the experiences of different generations are not identical, it is difficult to believe that forms of consciousness can remain stable. Equally obviously, the political socialization of the younger generation in an epoch of extended schooling and mass media has clearly been very different from that of the older generation. In the end, therefore, the oscillatory and cyclical dimensions of peasant reproduction as part-farmer, part-proletarians, may be overridden by the effects of cumulative social transformation, as new forms of consciousness develop. Simply for pragmatic reasons, anything this book has to say about attitudes is focussed on older men: those born in the hacienda era and their mature children. It is possible to describe the objective and subjective effects of structures of political control. We can produce explanations which account for the views people express directly and for the kinds of actions and responses they are observed to display when confronted with particular kinds of situation, which are often discrepant with what they say they think, of course. This sort of analysis provides a "theory" of peasant identity, and of peasant action and its limitations in the present. What it cannot do is predict how the next generation will react to the circumstances they find themselves in, since this would have to rely totally on their expressed attitudes and aspirations. Mature people do tend to see the world in a different way. To this extent, even the contemporary analysis of this book is, in a sense, historical, and its discussion of the determinants of social and political action will be inevitably more effective when we deal with the past.

By emphasizing the extent to which cyclical processes have also been ones of secular transformation, we can perhaps get a clearer idea of the possible future shape of peasant communities in Mexico. Unless the future is mechanistically predestined by the past, there must remain a vast number of uncertainties, the greatest of which is how the young people of today will react to the future which larger social forces seem to be creating for them. History does not really repeat itself, even as farce.

The Mexican Peasantry and the Power Structure

It must be conceded that the picture of peasant reproduction sketched in the previous section, corresponds to a process of social disintegration. As a result of regional developments and the extension of international migration, mechanisms of reciprocal aid between households and non-commodity allocation of resources progressively broke down in Guaracha from the 1950s onwards. Though there are still some individuals who practice these forms of economic behavior, they have become a minority in a community in which commodity relations dominate all social intercourse.

Many have, of course, viewed such practices as central to the resilience of the peasant community, and the means by which reproduction of the campesino population at large is secured despite unequal distribution of land and substantial wealth differentiation (Scott 1976, Warman 1976). In the case of Guaracha, commodity relations have penetrated deep into the heart of the domestic unit itself, causing an increasing atomization of units of consumption. There is sometimes severe inter-generational conflict over the right of the parental generation to share in the income acquired by their unmarried children or to have access to their labor power directly. As Arizpe (1985) has observed, though primarily with reference to small-holding agriculture in rain-fed areas, transformation of the social structures of rural communities has reached a stage where simply pumping funds into rural areas cannot be sufficient to revive the social structural conditions of viable small family farming. She suggests "giving back the social and political intiative to the communities" so that they can "rebuild their social networks" and "regain their bargaining power" in the political sphere (1985:219). Although I have every sympathy with the sentiments expressed here, they remain strikingly voluntaristic and idealistic from the standpoint of the historical experience discussed in this book. The lesson of the Guaracha case is that the essence of the post-revolutionary experience is the removal of initiative and bargaining power from the base: even developments in state policy towards the ejidos which might be considered "improvements" in a narrowly material sense are increasingly negative from this point of view.

The singular interest of the Guaracha case lies in its association with the phenomenon of Cardenismo. The literature on the Cárdenas era as it is conventionally understood in Mexican historiography has now become so voluminous that it virtually defies summary. I have argued elsewhere that the Mexican revolution must be seen as a process of long duration, its roots deep in the nineteenth century (Gledhill 1987). It was not the Obregón years, nor the epoch of Calles, which saw the crystallization of the post-revolutionary state, but the Cárdenas presidency. This is not to say that Cardenismo established a fixed structural pattern for Mexico's future: it is simply to agree with Theda Skocpol's contention that social revolutions of the Mexican type are consummated with the establishment of an effective mass-incorporating state. By describing the Cardenista achievement in these terms, I in no way imply that "mass incorporation" rests on the system's achieving high levels of legitimacy. Indeed, I shall argue that the regime now possesses very little legitimacy in the conventional sense of the term. Nor will I ignore the arguments which have been advanced against seeing the Cardenista state as a "relatively autonomous" moment in the history of national state formation in Mexico: this case will in large measure reinforce the general arguments of Nora Hamilton (1982) on the extent to which the social power of an existing economic oligarchy was conserved through this epoch of apparent radicalism. Throughout my discussion, the emphasis will be on the ultimate weakness and compromised nature of Mexico's national state structure. It is, in retrospect, rather ironic that the region where Cárdenas himself grew up has proved particularly recalcitrant to the processes of incorporation into the national state, and provides an exemplary demonstration of the way the social projects of the Cardenista state could be thwarted by both old and new regional bourgeoisies, which the center could only subsume through accommodation.

As we will see, the reality of Cardenismo viewed from this particular regional context, is a morass of contradictions. Cardenismo promised the peasants liberation but it brought new forms of arbitrary power and caciquismo to the communities it emancipated. It would be tempting to say that the experience of Guaracha reflected special features of the case. One might be the role of Don Lázaro's brothers, who acted as the region's political bosses during and after his presidency. Then there is the crucial fact that there was no large-scale spontaneous rural mass-movement here: the peones of Guaracha, along with a majority of their neighbors, rejected the Cardenista program. There was substantial support for the Cristero cause in the Ciénega, and opposition to the Cardenista program was aggravated by the excesses of some of Cárdenas's followers in pursuit of the anti-clerical cause. But I will argue that these "special" factors are not really special at all, but merely symptoms of the underlying limitations of Cardenismo as a revolutionary movement.

This is, in my view, not a matter of its ideological orientation in the programmatic sense—the fact that it was not a socialist movement. Some of the limitations of Cardenismo were also limitations of the so-called "proletarian revolutionary" movements. It is clear that neither the Cardenistas nor the Bolsheviks were from the social underclasses they claimed to represent, and that neither the Mexican nor the Russian revolutions produced outcomes which accorded with the objectives and world-views which can be imputed to the spontaneous agrarian movements which underpinned the radical turn taken by what began as political revolutions and ended as social revolutions from above. I take no real issue here with conventional Marxist accounts of the historical limitations of agrarian movements, but I take strong issue with the naivety of traditional Marxist perceptions of the nature of proletarian revolutions and the largely negative view of the objectives pursued by agrarian social movements which are entailed in the fantastic vision of the "world-historical" roles of social classes this delusion promotes. The Zapatistas may have "failed" because they were not state builders, but they could not have been state builders without failing to be a popular agrarian movement. In the case of the Ciénega, we need to do some yet more radical thinking about the nature of class struggle and popular reactions to the experiences of oppression and dispossession if we are to produce a positive account of the behavior of the rural underclasses rather than simply another negative appraisal of the reasons for the absence of large-scale popular social movement dedicated to agrarian issues. But I will argue that it is necessary to make such an attempt, for two reasons.

Firstly, it is simply not the case that even the peones of Guaracha were "apathetic," "passive" or "fatalistic" in the face of exploitation and oppression: though the majority were against land reform, they were for religion in a way which cannot be explained simply in terms of the ideological hegemony of the clergy and "false consciousness." Secondly, it is vital to see that both the Cardenismo of Don Lázaro and the Cardenismo of his egregiously corrupt brothers alike was ultimately, and from the beginning, antagonistic to genuine self-organization on the part of its peasant clients. What popular leadership and genuinely popular mobilization did emerge from among unrevolutionary peasants of the Ciénega was suppressed, and at times quite ruthlessly suppressed. This is not, I contend, a matter of bad faith, duplicity, or the perversion of principle by the exigencies of practice, though Cárdenas was certainly a consummate political realist. And my intention is certainly not to slight the personal reputation of Lázaro Cárdenas: no one who interviews peasants who knew the man personally could fail to appreciate his quite remarkable personal qualities, which amounted to much more than mere personal charisma. What is at issue here is the nature of the revolutionary process, and the vision of social transformation which dominated it. Cárdenas was ultimately just as much of a modernizer as Calles had been, and Cardenismo represented simply another (though perhaps more attractive) variant of the repeated revolts of the provincial urban periphery against the monopolistic structures of economic, social and political power which had characterized Mexico from the colonial epoch to the Porfirian era. Although the resurgence of such movements in the Mexican revolution was, in a sense, a rerun of earlier conflicts of the nineteenth century, what differentiated the developments of the twentieth century from their predecessors was precisely the fact that the building of a viable national state was increasingly seen as dependent on the incorporation of the masses into the process. Cardenismo's peculiarity was its vision of the possibility of a novel and humane solution to the Agrarian Question. But this elevated objective did not eliminate its authoritarian outlook on the way the peasantry should be trained to fulfill their new historical destiny, or entail any fundamental disagreement with other viewpoints on the relationship between "modernization," technical rationality, and urban-industrial development. In fact, despite the peculiarly Mexican facet of Cárdenas's agrarian policy, it had much more in common with the Soviet perspective on modernization than most analysts seem willing to acknowledge.

At another level of abstraction, it becomes clear that Cardenismo was characterized by a complex of contradictions which reflected the way in which modernizing, state-building revolutionary elites have to grope their way through short-term crises and unexpected difficulties as best they can, whatever paradigmatic precepts guide their actions. Cardenismo was never equipped with a hugely coherent ideological program in the first place, and had to deal with entrenched power structures which the revolutionary process had only disrupted in a limited way. In the end, in the face of renewed international and internal pressures, muddling through was not sufficient to consolidate the main Cardenista social project in the countryside. The Ciénega was allowed to go its own way at a relatively early stage, and more or less ignored by the center for the next twenty-five years. Nevetheless, we will find Lázaro Cárdenas reentering our local history in person in the final years of his life, and once again striving to reshape the agrarian order. Ironically, Cárdenas's prestige in peasant circles is higher today than it was at the time of land reform, and even in 1940. The significance of Cardenismo gradually changed after Don Lázaro left office and began to dedicate himself to new projects in the Tierra Caliente. It has endured in Michoacán, undergoing further transformations, after the General's death, and taken new directions under the leadership of his son Cuauhtémoc, who was state governor from 1980 to 1986, prior to his break with the PRI and ascent to national leadership of the left-leaning opposition.

It is not the purpose of this book to provide the more specific, complex and multi-faceted history of Cardenismo in Michoacán which would be required if we were to analyze the way myth, representation and practice enter into a dialectic which has reconstituted the significance of Cardenismo for later generations and given it a renewed vigor as the source of legitimacy and inspiration for genuinely oppositional practices with spontaneous popular roots. Nor do I dispose of more than fragments of the data required to undertake such a study. But it is important to recognize that the myth has been reconstituted in the years since 1940, simply because it does provide the basis for a continuing reflection of the gulf between what is and what ought to be in the minds of those involved with the agrarian sector, an effect which becomes all the more powerful the more detached it becomes from the reality of the Cárdenas years. I pronounce no final judgment at this stage on whether the possession of this peculiar political culture may provide a vehicle for challenging the status quo or ultimately serves to defuse such challenges. The latter day interventions of Cardenismo in the life of the region remain as pregnant with ambiguities as the earlier ones, but the belated transformation of the ex-peons of Guaracha into Cardenistas can scarcely be dismissed as insignificant in the light of their past history and present place in Mexico's agrarian structure. I think that this local history, for all its specificities and idiosyncracies, does reveal something of the mechanics of the power structures which have enabled the Mexican "system" to stagger successfully through crisis after crisis since the 1940s—and it may also provide some insights into the ultimate limits of adaptation of that system.

This final issue returns us from the terrain of the cyclical to that of the long-term, secular tendency. If the Mexican social system has changed as radically over time as this book argues that it has, to what extent can the political system, originally adapted to a very different social world, endure without itself undergoing more radical structural change than it has thus far experienced? It would be naive to imagine that we are talking here of the possibility of a "classical" type of social revolution. But it may well be that we lack imagination when it comes to the kind of scenarios which will ultimately prove relevant to the Mexican case: there appear to be no real historical analogies available for the type of social formation which confronts us today. That, in the last analysis, is why it is worth preoccupying ourselves with the details of concrete histories and concrete social situations. Though they can never supply complete answers to the "big" issues, they can at least help us to avoid the grosser errors of conceptualization which emerge from a lack of detailed knowledge of the nature of contemporary social reality.

Although the community of Guaracha is now named "Emiliano Zapata," I have generally used the archaic name in the text. It is impossible to disguise the community's identity effectively. "Emiliano Zapatas" are legion in Mexico, and the name would have afforded a convenient enough disguise. But many local people still prefer the old name, and no harm can be done by leaving the village as a whole with its specific historical identity. The names of individuals, living and dead, have, however, often been changed in order to preserve their anonymity.