Commercial Penetration, the State, and the Shifting Locus of Communal Solidarity

The crisis of common property and joint usufruct of land by descent groups of course did not lead to the disappearance of solidarity within such groups or within the larger parcialidades. Mutual labor exchange, common defense

against outside land invasion, shared rituals and celebrations, and the myriad of everyday activities shared or undertaken jointly—all these continued, contributing to the trust binding such groups together.

Far from being stable, static entities, peasant communities in the altiplano were undergoing a process of complex metamorphosis by the early twentieth century, particularly by branching off or splitting into several new ones.^[53] The dynamic elements in such subdivisions were the descent groups. In the 1940 census the term estancia for the first time denotes rural population centers that had developed out of peasant landholdings divided between a growing number of second- and third-generation heirs.^[54] Often the "locus of solidarity" was shifting from the old parcialidad to smaller groups even while the scope of that solidarity was becoming more limited through the individualization of land-use patterns. The old parcialidades were not affected uniformly by this process, however, and in 1940 many still held considerably more population than did most of the recently constituted ayllus, parcialidades, or estancias. Of Arapa's eight parcialidades from the 1876 census, six continued to thrive in 1940 with populations of between 182 and 588 persons; one had disappeared totally and another one—the important nineteenth-century parcialidad of Yanico—had shriveled to a population of three. The transfer of solidarity from the older parcialidades to the new units based on descent groups probably began at their geographic rims. There the landholdings of descent groups could span several communities, and intermarriage with members of neighboring communities was more likely.^[55] Such marriages contributed to a pattern in which peasants from one parcialidad held land in another parcialidad, weakening solidarity in at least one.^[56]

But between the 1880s and 1920 most of the old parcialidades continued to provide some institutional cohesion to the various descent groups, sectors (barrios ), or estancias that had sprung up inside them. This cohesion was engendered through a curious dialectic of exploitation by outside authorities and defensive assertion of autonomy and solidarity within the community. Along with the growing entrenchment of hispanized local and provincial elites, the extraction of labor and resources from the communities through their hierarchy of offices, not new in itself, reached its greatest scope and elaboration in this era. Yet such extraction also became more volatile and unstable; after 1920 this form of tributary exploitation would enter a phase of decomposition.

The number of offices in the communities grew with the variety and frequency of extraction practiced by the authorities. In some districts the governors, local representatives of the central government, had sixty to eighty communal officers at their disposal, apart from those designated to

serve the parish priest and the justice of the peace. Most communal officers were chosen for a term of one year in elaborate ceremonies held each January 1. Every parcialidad chose one segunda , the highest communal authority, who was "obliged to present himself before the governor every Sunday and on days of festivities together with his subordinates . . . in order to give account of the projects carried out since the last meeting." The segunda coordinated all the work and obligations of the whole parcialidad toward the governor, priest, and justice of the peace. He distributed duties to and collected goods from his immediate subordinates, alcaldes and hilacatas , who represented the various barrios, estancias, and descent groups within the parcialidad. They in turn fulfilled the same function within their sector, supervising the lower officers, alguaciles (guards), propios (messengers), pongos , and mitanis (men and women doing domestic service in the authority's household).

The duties toward the governors were broad in scope, time consuming, often costly, and mostly uncompensated. Either personally or collectively, the officers and their parcialidad planted and harvested fields for the governor's private use; rendered fixed quantities of sheep, wool, and other livestock products; spun and wove wool that the governor had distributed among them; disbursed within the communities the money with which the governor purchased additional wool at a fixed price; transported his goods to and from urban markets with their own animals; served in his household; and collected taxes in the communities.^[57] The authorities used the community officers to recruit laborers for the faenas (public work projects). In 1893 Azángaro's mayor requested the subprefect and district governors to provide Indians from parcialidades to make adobes for the "urgently needed" construction of a new prison; one year later the mayor requested twenty Indians per week to pave the provincial capital's Plaza de Armas, suggesting that they be paid "the same daily subvention as the prison laborers," ten centavos.^[58]

According to Nelson Manrique, this heightened and systematized use of traditional communal authorities meant that in southern Peru "they ended up being reduced to a condition of servants of priests and local functionaries of the central government and unpaid helpmates of the state."^[59] Yet the most affluent and respected families in the parcialidades, from whom the highest communal authorities were recruited, were hardly pliable dupes in this apparent subversion of their venerable offices. They were generally willing to play along, in part because this shabby remnant of the colonial "compact" between crown and community might still afford some protection of the communal domain against outside intrusions. In part they were pursuing their own interests against other peasants.

Francisco Mostajo, the eminent progressive liberal intellectual from Arequipa, observed in 1923 that

the Indians of Huancané province gladly work in the construction of roads and the repair of churches, the only public work projects in those parts. They have internalized the social concepts of the Inca Empire or of still earlier societies so much, that they believe that these collective labor projects give them title to the property and usufruct of the land on which they live and plant crops. I build roads, I repair the church—the Indian will tell you when he is drunk or when he perceives a threat of being imprisoned—and this is my land: nobody can take it from me. When Protestant Indians refused to participate in the repair of the church, the other Indians said: you should not have any land then, you may not use "our" roads.^[60]

In part this view reflected the indigenista reinterpretation of Andean collective work customs that Mostajo hoped to harness for a "wise social legislation." But the notion that a link existed between fulfilling one's public duties in the parcialidad and the claim to land was alive among some altiplano peasants during the early twentieth century. In a will of 1910 Tiburcio Choquehuanca, a seventy-five-year-old peasant from parcialidad Jallapise with sufficient means to have purchased a dwelling in Azángaro town, espoused this view. He thought that his title to Estancia Parajaya, acquired in part through inheritance and in part through purchase, was reaffirmed because he and his wife had "rendered for this parcel the communal services or offices of mitani, pongo, alquacil, alcalde, and segunda, services that my children have also rendered for the custom that is common in these districts and that gives title to assure the possession of property."^[61]

Peasants thus appealed both to notions of private property and to the older notions of reciprocal obligations and rights as legitimizing their possession of land. It was a kind of insurance policy in the face of threats against their land resources arising from various quarters of the local hispanized elite. As long as private property rights to peasant estancias continued to be threatened by neighboring estate owners through their predominance in the courts and use of force, appeal to the legitimacy of possession of land through the "compact" with the authorities of state and church might provide a modicum of protection. The segunda, alcaldes, and hilacatas could rally the whole parcialidad to defend land and to reject demands by outsiders that went beyond customary obligations.

At the same time, however, the communal authorities were involved in the complex contests over political power between the various gamonales,

the provincial and local bosses using patronage and clientalism to advance their family's own interests. Depending on specific local power constellations, segunda and alcaldes in some cases were allies of the governor, receiving a share of the taxes they collected, entering into commercial deals with him, or pursuing common strategies to appropriate municipal or communal land.^[62] Such alliances heightened centrifugal forces on the cohesion of old parcialidades, persuading peasants in its various sectors to transfer the "locus of solidarity" increasingly to the descent group or to strengthen the private property claims to the family's land.

In 1900 the contribución personal, the disguised Indian head tax that President Castilla had abolished in 1854 but that continued to lead a spurious life during the second half of the nineteenth century, was abrogated for good in the department of Puno, five years after its abolition by Peru's national congress. This ended the long history of tribute and its various successor taxes, which had formed one of the key links between the state and the Indian communities.

For the national congress and the executive in Lima, increasingly under the sway of coastal financiers, merchants, and agro-exporters after the revolution of 1895, the tax had become a dispensable embarrassment, too reminiscent of the crude extractions by the Spanish crown. The Indian community had ceased to be important for the financial well-being of the central government. For the notable citizens and authorities in the sierra, however, the abrogation of the contribución personal was a matter of concern, since it was the major source of income of the departmental councils. Established in 1886 by President Andrés Cáceres to oversee fiscal decentralization, these bodies demonstrated the political power of the serrano gamonales and hacendados during the decade of Peru's "New Militarism" after the War of the Pacific. The tax's abolition weakened their political autonomy, making them more dependent on the treasury in Lima. Puno's Prefect Manuel Eleuterio Ponce, himself a notorious land grabber in Arapa, lamented that departmental tax income had plummeted by two-thirds in 1901. But far from helplessly accepting this shrinkage of regionally controlled funds, authorities in the altiplano reacted by shifting the tax burden on the Indians to increased collection of property tax (contribución de predios rústicos ) on their estancias.^[63] During the following decade the mean tax assessment grew more rapidly in districts with relatively few estates and a predominant peasant sector, such as Saman and Achaya, than in those with high concentrations of estates (table 7.3).

For Puno's peasantry, the meaning of the shift from contribución personal to the property tax was ambiguous. During the 1880s and 1890s communities in Bolivia's altiplano had successfully resisted this shift,

designed by that nation's liberal oligarchy as the cornerstone of a radical policy of breaking up communal property by administrative fiat. The communitarian ideology of Bolivia's altiplano peasantry saw "both the 'servicios forzosos' and the payment of tributo, in accordance with traditionally accepted norms, . . . [as] the communal counterpart of a pact of reciprocity with the state."^[64] But in Puno no tear was shed by community peasants when the contribución personal was finally abolished, and by 1900 they had accepted payment of property tax in principle . Here the tax was applied to peasants gradually, with a few appearing in matriculas as early as 1850. Instead of imposing a completely alien notion of property, the slowly advancing incorporation of the department's peasantry into the property tax rolls paralleled rather than preceded the advance of notions of private property. The connection of this shift with changes in the institutional corporate character of Puno's parcialidades is crucial. The abolition of the contribución personal weakened the position of the tra-

ditional communal officers, who had collected the tax from every adult male in their jurisdiction for the district authorities.

The increasing collection of the contribución de predios rústicos among the communal peasantry coincided with the establishment of a new tax collection agency, the Compañía Nacional de Recaudación, independent of district and provincial authorities. Because many families in the communities did not have to pay property tax, the Compañía's agents may have dealt directly with the heads of those households that did.^[65] They may also have approached the hilacatas of relatively affluent descent groups or barrios, where many families were entered in the tax rolls. In any case, tax collection ceased to be an issue for the parcialidad as a whole. Descent groups opposed to the impositions by communal authorities now grasped the opportunity to strengthen title to their own land by willingly cooperating with the collection of property taxes, in the process furthering the transfer of solidarity from the old parcialidad to their own smaller groups.^[66]

This is not to say that Puno's peasantry did not resist the increased collection of property tax. The massive revamping of Peru's tax structure under Presidents Nicolás Piérola, Eduardo López de Romaña, Manuel Candamo, and José Pardo in the decade after the revolution of 1895 coincided with a new wave of peasant mobilization in the altiplano. The salt tax and monopoly introduced in 1896, the new excise taxes on sugar and alcohol of 1904, and the increases in property taxes amounted to a completely new program of fiscal extraction from the country's still mostly rural population. The tax system was now based on a notion of individual consumers rather than on corporate groups of producers. But in contrast to Bolivia, in the Peruvian altiplano property taxes, which strengthened a landholding regime based on private title by individuals and their families, never became the dominant issue of peasant resistance.^[67]

At least two-thirds of all landholdings enrolled in the property tax registers between 1897 and 1912 belonged to peasants. It was public knowledge that assessed peasants paid a proportionally higher rate than did most hacendados. Not only did estate owners frequently underreport their livestock capital, resulting in a low assessed tax rate,^[68] but tax commissioners also liberally estimated stocks for many landholdings of illiterate peasants, who had little chance to do anything against this abuse. The 600 to 1,200 peasants enrolled in the tax registers between 1897 and 1912 represented only a small share of all peasants owning land in the province. By legislative resolution of October 30, 1893, all landholding units producing income below 100 soles m.n.—equivalent to a livestock capital of up to 500 head of sheep—were exempted from paying property tax.^[69] Most peasants owned smaller livestock herds. Nevertheless, until 1907 tax com-

missioners entered many peasants with annual incomes of below 100 soles m.n. in the rolls; in the 1897 rolls more than 80 percent of all assessed peasants fell into this category. Even so, most peasants were not assessed for the property tax. Nobody paid who had an estimated annual income below 20 soles m.n., equivalent to a livestock capital of 100 OMR; in some communities, and even entire districts, hardly any peasants were entered.

A law of December 4, 1908, reaffirmed the exemption of property producing less than 100 soles m.n. in annual income, and this time it stuck in Azángaro. It had the curious effect that the presumed livestock capital of many peasant holdings, from which the estimate of annual income was calculated, doubled or tripled in subsequent tax rolls, from 200 or 300 OMR to 500, 800, or even 1,000 OMR. At the same time the number of peasants assessed declined by half from the 1902 rolls. Although tax assessments clearly reflected widely different levels of income derived from livestock operations among the peasants, who paid and how much he or she paid evidently depended in good measure on power constellations and on the client networks of the tax commissioners.

Social differentation in Andean communities is not a new phenomenon of the past century. Even the often repeated claim that it has increased immensely with advancing commercial penetration or the "transition to capitalism" is difficult to prove because of the lack of comparable income and property statistics for the nineteenth and earlier centuries. Migration and the need to pursue income-earning activities outside of agriculture have characterized generations of Andean peasants in previous centuries as well. All we can say is that these phenomena became more massive (and thus more visible) during the twentieth century and that the criteria and mechanisms for social differentiation gradually underwent important changes.

During the colonial period, and to a certain degree even during the decades immediately following independence, wealth and prestige in the communities had been closely tied to lineage. The families of kurakas and many originarios maintained a higher level of access to community resources by birthright, even if they increasingly used the privileges of their offices and the opportunities in the marketplace to enhance their position vis-à-vis commoners. Although they had lost their privileges during the agrarian reforms of the 1820s, the descendants of minor kurakas remained among the wealthiest peasant families as late as the 1850s. Some owned estancias with a thousand head of sheep or more, not unlike small to mid-sized fincas. This was the case with the Puraca family and their Estancia Buenavista de Conguyo in parcialidad Moroorcco, or the Zecenarro Mamanis' landholding San Antonio de Lacconi.^[70] Without relin-

quishing their Indian identity, affluent community peasants during the middle decades of the past century still had enough social standing to be called on as guarantors in contracts between hispanized landholders, pledging their estancias and livestock capital as security.^[71]

In the more complex and fragmented economic and political environment of the altiplano around 1900, social differentiation within the peasant communities also became more complex, and families relied on a broader range of strategies to ensure reproduction of their household economies. A cautious interpretation of the 1897 property tax rolls, raised before political and fiscal pressures introduced the massive distortions of the subsequent decade, allows us a glimpse of the more affluent peasant families of that time. Perhaps as many as 8 to 10 percent of the families in the communities, between 600 and 800 families, owned livestock herds of 100 or more animals.^[72] The wealthiest among them, possibly 2 percent of Azángaro's community peasantry, had herds of 500, 1,000, or even 2,000 animals. They were prevalent especially in cordillera communities on the northeastern rim of the province, between Putina and San Antón, where few crops could be grown. Population pressure tended to be lower there, and perhaps communal institutions remained stronger, blocking competition by new families for scarce resources such as moyas.^[73] The wealthier descent groups thus could fully benefit from increased demand for sheep and especially alpaca wools.

In communities with stronger pressures from neighboring landholders or from internal population growth, especially in the plains and valleys and on lakeshores, accumulation of wealth required additional strategies. In some cases prominent families in the parcialidades relied on political assets to get more land, pay less taxes, and avoid being hindered in their commercial activity. Such ends could be achieved through patronage with the local authorities or a powerful hacendado. In districts with a weak elite of hispanized large landholders and traders, the dividing line between mistis and affluent community peasants was quite fluid, and here prominent comuneros might become justices of the peace or serve on the town council.^[74]

Purchase of additional land from other peasants or, less frequently, from hispanized large landholders became increasingly important for families with large herds. Whereas for the whole period 1852–1910 purchases by peasants accounted only for 14.4 percent of all sales transactions by number and 5.5 percent by value, between 1913 and 1919 they constituted 24.9 percent by number and 6.9 percent by value. The boom of the wool export economy propelled affluent peasants in the altiplano to expand their landholdings, secured by notarial title deed.

[Full Size]

Figure 7.1.
Annual Income of Peasants in Chupa in 1897, by Age.
Left: 49 years or younger. Right: 50 years or older.  Source: Matrículas de contribuyentes, 1897, BMP.

In many cases peasants were buying from relatives parcels of lands that had formed part of the estancia of the descent group.^[75] Thus, the pattern so frequent among hacendado families repeated itself among a minority of affluent peasant families. The most entrepreneurial son or grandson attempted to reunite the original holding, increasingly splintered through equal inheritance. Possession of land was also linked to the life cycle of peasant families in the altiplano, something first suggested by A. Y. Chayanov for Russia. In 1897, in three districts with numerous affluent peasants, Arapa, Asillo, and San José, nearly 60 percent of peasant landholders paying property tax were in their fifties, sixties, or seventies, a much higher proportion than in the population at large. In Chupa the proportion of peasants with a relatively high annual income, an expression of the amount of livestock and ultimately of the land they possessed, was considerably higher among older peasants than among the younger ones (fig. 7.1). In other words, over the years well-to-do peasant families were often able to expand their landholdings as the labor power of the grown children allowed them to keep larger herds. But in contrast to Chayanov's model, no automatic shrinkage of the family domain occurred during the parents' old

age. To be sure, old age presented a crisis for many peasant families. Adult children established their own households based on parcels of the family estancia. Aging parents were often swayed to accept dependence on a neighboring hacendado in the hope of receiving protection and material benefits, from food supplements to payment of funeral expenses. But among affluent peasants in the tax rolls, many were in their seventies, veritable patriarchs controlling the family domain. After their death the heirs continued to operate the estancia jointly (pro indiviso), manifested in the tax rolls by dozens of entries such as "heirs of Juan Quispe" or "Andrés Maldonado and co-owners." Where descent groups continued joint ownership of the family estancias, the Chayanovian down phase of the landholding cycle of peasant families could be avoided.

But such traditional family strategies do not suffice to explain growing land purchases by peasants. In many communities transfers of land among peasants remained negligible as late as 1910. For the entire sixty preceding years a total of nine land sales between the peasants of Saman's five parcialidades were recorded by the notaries of Azángaro and Puno. By contrast, the ten parcialidades with the greatest sales activity between peasants accounted for nearly half of all such land transfers in Azángaro. Most of these parcialidades were not identical with those that lost a great deal of land to hispanized large landholders. For some parcialidades the correlation between strong peasant purchase activity and numerous affluent peasants in the tax rolls is striking. Families such as the Callohuancas, the Sucaris, and the Huaricachas in Asillo's parcialidades of Sillota, Anoravi, and Hila or the Sacacas and Quenallatas in Putina's Huayllapata appeared as owners of five hundred head of livestock or more in the tax rolls and as active land purchasers in the notarial registers. There were, then, a few communities in which the land market was considerably more active than in the rest and which had a particularly vigorous group of kulak peasants.

Commercial activities played an important role for many affluent peasants. The sale of rural crafts and of livestock products from their own estancias were part of the survival strategies of most peasant households, an expression of their precarious subsistence. But real trade, with goods acquired from neighboring peasants, from other traders, or from producers in distinct regions, could be more than that. It allowed dozens of affluent peasants to accumulate funds for the purchase of additional lands or livestock or for social investments in outfitting festivities in the communities. Although rural crafts largely remained an expression of peasant poverty, trade often was associated with affluence. Peasants were especially active

as traders of livestock, coca leaves, and alcohol; some worked as itinerant wool buyers. Just as with land purchases, peasant traders were concentrated in a few districts and in specific communities.^[76]

In the absence of detailed ethnohistorical information, we can only conjecture about what distinguished communities with an active land market and numerous peasant traders. No single factor can account for all communities in which market forces seem especially strong. Some, such as those in Asillo, Putina, Chupa, and Arapa, benefited from a broad mix of resources, with relatively plentiful croplands, riverine, or lakeshore pastures propitious for cattle grazing and sufficient higher pasture grounds to maintain numerous sheep and alpacas. The social differentiation in these communities was more complex than it was in the cordillera or on the broad plains in the center of the province. The intensity of peasant trading owed much to the location of communities. For example, many families were active as coca leaf traders in Chupa and Putina communities, from where peasants regularly undertook barter trips across the nearby cordillera to Larecaja province in Bolivia, an old center of coca estates. Livestock trade, by peasants as well as by hispanized traders, was strong in Saman and Arapa, strategically located for driving herds to market in Juliaca and Puno.

But the nature of commercial penetration, the relation between communal resources and household economy, and the prevalence of a stratum of kulak peasants also depended on politics and on the effectiveness of solidarity in each and every parcialidad. The correlation between incipient market-based stratification in the communities and ecological, demographic, and commercial factors was never perfect. It should also be stressed that Azángaro's kulaks largely continued to be well-integrated members of their respective communities, drawing advantages from the exercise of high office, relying on relations of asymmetrical reciprocity and clientelage to receive cheap labor services, and prepared to rally the whole parcialidad in defense of "communal" property.^[77]

Most peasant families faced a more precarious existence. Perhaps as many as 80 percent of Azángaro's peasants owned between twenty and a hundred sheep, a cow with its calf, and perhaps two or three transport animals. They could sell only a few livestock products before diminishing their herds, with devastating consequences over several difficult seasons. Any crisis could seriously upset their livelihood: droughts or livestock epidemics, increasing the mortality of the animals; crop failures; a protracted fight over land; theft of animals; onerous impositions by the governor or the priest; recruitment into the army; and the expenses and

difficult adjustments required by major events in the life cycle, especially marriages and deaths. The limited amount of surplus livestock products that they could sell made other sources of income more important: crops; rural crafts, such as spinning and weaving, pottery making, cheese making, and rope making; occasional labor for neighboring hacendados or for hispanized residents in the district and provincial capital; seasonal work in the mines in the Cordillera de Carabaya or further down the eastern slope of the Andes in gold-washing camps, coca plantations, or rubber tree forests, or in construction projects and as day laborers in cities such as Puno and Arequipa.

This is an unexceptional picture of a traditional middling and poor peasantry, facing the cyclical crises in their precarious family subsistence with a great variety of strategies. Perhaps "proletarian" forms of seasonal supplementary labor had increased somewhat by 1900 from fifty years earlier, with growing needs for workers during the brief mining boom in the Cordillera de Carabaya as well as the beginning expansion of employment opportunities in southern Peru's cities. But such types of seasonal wage work for altiplano peasants were not totally absent during the early years after independence. In any case, they remained the income sources of last resort; peasants preferred to supplement their farm income through peddling and barter, rural crafts, and occasional labor in nearby towns and estates.

Indeed, it seems unlikely that the average income of Azángaro's peasants declined significantly between the early decades after independence and World War I. A rough estimate suggests that the number of livestock held by every man, woman, and child in the province's communities stayed roughly the same between 1825–29 and 1920 at ten OMR.^[78] With prices for animals and livestock products during the 1860s and perhaps again between the mid-1890s and 1917 outpacing the price of comestibles, stimulants, and manufactures typically purchased by peasants, it cannot be ruled out that their real income rose during those years.

Yet the old and familiar cyclical crises of peasant households could take on new and drastic consequences. The heightened volatility of prices for pastoral goods, demographic pressure, and a consolidated hispanized elite with more power and money all contributed to force thousands of poor and middling peasants to relinquish control over their lands and become colonos on estates. In most cases the sale of land was not the consequence of long-term impoverishment but of a relative loss of autonomy. The capacity of many peasants to minimize the effects of the cyclical crises on their household economy through the solidarity of the community was

diminished. An increasing share of their exchange relations were with the hierarchy of hispanized traders, shopkeepers, and commercial agents. The value of their marketable surplus of livestock goods underwent sharper cyclical swings than had been common until midcentury.

The land resources in the communities were often becoming too scarce to provide the buffer that would allow amiable settlements of conflicts arising out of greater pressure of livestock populations on available pastures within the descent groups. In such situations, blurred pro indiviso title rights to the estancia of an extended family became the wedge by which hispanized landholders broke into the land of community peasants. Those avid to expand their estates now had sufficient cash or credit, and often the political backing, to exploit any crisis of a peasant household's subsistence economy. After gradually tightening peasant's dependence through loans of money or foodstuffs, through labor exactions, or through protection before the courts, the police, or the local administration, an estate owner would finally force the peasant to give up title to his or her parcel of land and incorporate it, together with the former owner's household, into the hacendado's estate. This process of increasing dependence often had little to do with impoverishment.^[79]

The multiple pressures on the community peasantry thus resulted in a rapid and vast growth of the ranks of colonos in the estates. This status, in which the household economy of the former community peasant remained intact, was an ambiguous and ill-defined halfway point between peasant autonomy and the complete loss of "peasantness" experienced by landless peasants who left their native communities. The hacienda thus functioned as a kind of forced catchment basin, a depository for a peasant reserve army. With the disintegration of some newly formed estates during the decades after 1920, numerous colonos would revert back to a fully autonomous status as community peasants.

But how many peasants lived in the communities without land or even without livestock? Such poor peasants appear in the manuscript census as early as 1862, living as dependents in the households of more affluent peasants, who had often taken them in as young children.^[80] Their situation approached that of colonos on estates. Their main task consisted in herding a flock of their patrón. In return they were allowed to keep animals of their own and possibly build a residence in the center of a sector of the patrón's pastures.^[81] In such cases landless dependents could form households and families. This may also have been the situation of landless or near-landless rural artisans earning the overwhelming share of their livelihood through their craft. In communities where families relied primarily on craftwork, such as the potter communities in Santiago de Pupuja, the availability of

this alternative helped to keep on the land dozens of poor families who might otherwise have migrated or sought employment in a nearby hacienda.

More frequently, however, a landless peasant would stay as a single retainer in the household of his or her patrón, assigned whatever jobs needed to be done—from herding to spinning or weaving, to construction and repair of the patrón's cabaña and transport of some products to market—against the sole recompense of food and lodging. In this position he or she could not accumulate resources sufficient to form an independent household. This type of retainership probably corresponded to specific phases in the life cycle of poor peasants. As children they were given as dependents into the households of more affluent peasants to strengthen clientalistic ties. When they reached adulthood they attempted to form independent households, either as shepherds on an estate, as artisans, or as muleteers. But some—especially women, for whom the lack of a dowry formed a serious obstacle—might be forced to remain as dependents in their patróns' households throughout their lives. Widowed older peasants without descendants, unable to maintain their own household economy, also attached themselves to more affluent peasant households. The social structure within communities was characterized by clientalistic relations similar to those between peasants and hispanized large landholders, albeit on a smaller scale of exchange between economic resources and labor services.^[82]

Landless peasants had a few avenues to establish independent households tied into the agrarian economy. If they owned some animals, they could turn them over to one or more peasants with sufficient pastures. Either the owner of the land would pay rent on the animals and keep all of their products, or owners of land and animals would share in the products, a contract known as waqui in some parts of southern Peru.^[83] In 1910 Casimira Mamani, a sixty-year-old widow, owned and lived in half a house in Azángaro town. She possessed no land but had eight cows, two young bulls, one riding horse with two colts, one donkey with its colt, and thirty sheep, maintained on the pastures of six different peasants.^[84]

In rare cases landless peasants were granted vacant plots in the communities by a process of petitioning district authorities with the concurrence of communal officers.^[85] Some were assigned plots of families extinguished in one of the severe epidemics, such as the typhoid fever epidemic of 1856–58.^[86] Of course, theoretically landless peasants could lease, take in anticresis, or purchase land from other landholders. But it is my impression that most peasants buying land or taking up a plot in anticresis were affluent peasants.^[87]

Despite growing pressure on land resources landless peasants seem not to have formed an increasing share of community population in the early years of this century.^[88] There are two reasons for this situation. Many landless peasants living as retainers in households of other peasants did not found families and had fewer children than did peasants with sufficient resources or income for sustaining households. Consequently, their status tended not to be inherited; rather, this social stratum was renewed from generation to generation. Second, with growing demographic pressures in the communities it became preferable for impoverished peasants to become colonos on estates, where they could maintain their own household economies. Others emigrated to the ceja de la selva, the Cuzco valley system, or Arequipa. But before 1920 the number of families permanently leaving the countryside was dwarfed by those who continued to work as peasants in the estates.