6
The Avalanche of Hacienda Expansion

The early decades of the republican period were marked by a relative equilibrium between the hacienda and peasant sectors of the altiplano. This situation began to change during the 1850s. As early as 1867 Colonel Andrés Recharte, subprefect of Azángaro, could write that "most of the damages and abuses suffered by the Indians have stemmed from the mestizos' covetousness for their land."^[1] After the slow decay of the colonial mining-supply economy and the dislocations caused by the Wars of Independence, a newly emerging landholding elite gradually adapted to the wool export economy. By midcentury, as the prospects of wool exports became brighter, they sought to control the important peasant sector more thoroughly in order to capture a greater share of regional income. In this chapter I discuss central aspects of Azángaro's changing property relations between the mid-1850s and 1920.^[2]

Methodological Considerations

I devised a methodology to answer questions as the following: How did the "flows" of land between various social groups evolve over time? How did the size distribution of land parcels offered for sale vary between different social groups? The notarial contracts render bits and pieces of information, permitting the judicious construction of a scheme for categorizing all participants in transactions. They normally list occupation, place of birth and residence, capacity to speak Spanish, and—less dependably—the socioethnic origin and any public offices held by each party. I gathered these data in an index of all participants in notarial transactions, containing some

eight thousand cards, and assigned each party to a notarial transaction to one of the following categories:

1. Indigenous peasant

2. Hispanized large landholder

3. Intermediate group (including persons of indeterminable social status)

4. Church

5. Beneficencia pública

Internal consistency of assignments was maintained through the use of clear criteria in the evaluation of the social indicators. Some values placed the person into category 1 ("indigenous peasants"): for example, the occupations labrador and pastor , the socioethnic labels indígena or indio , lack of knowledge of Spanish, and residence in a parcialidad or ayllu. Other values indicated that a person belonged to category 2 ("hispanized large landholders"): for example, occupations such as hacendado or abogado , knowledge of Spanish, and holding of offices such as gobernador, subprefecto , or juez de primera instancia . Certain terms were used so indiscriminately that I considered them to be neutral. Some notaries applied the occupational labels propietario and comerciante as an insoluble pair to all individuals, the peasant selling a parcel for 50 soles m.n. as well as an hacendado selling an estate for 5,000 soles m.n.

At least two social indicators had to point unequivocally to category 1 or 2 to make an assignemtn. For example, in order to assign a person to category 1 ("indigenous peasant"), he or she had to be described as indígena and labrador, as a non-Spanish speaker and living in a parcialidad. If the social indicators showed any ambivalence, or if there was only scanty information on a party to a contract, I placed the case in category 3. Thus, two subgroups constitute the members of this category: persons for whom information was insufficient, and persons for whom the social indicators seemed ambiguous, truly an intermediate group on the range from indigenous peasant to hispanized large landholder.

This problem brings us to the conception underlying the categorization scheme. Azágaro's social stratification was considerably more complex during the late nineteenth and early twentieth centuries than this scheme implies. However, what mattered here was to choose a scheme that reflected an underlying structural element of Azángaro's society while allowing the categorization of most contract parties despite limited information on their social backgrounds. Thus, the scheme strikes a compromise

between the exigencies of an all-too-imperfect source and the issues that I sought to investigate.

The categories are viewed as lying on a continuous scale in which the various status groups in Azángaro's society grade imperceptibly into each other. The scale is bounded on both ends by ideal types formed by a total clustering of all social indicators capable of defining a category. Categories 1 and 2 are meant as such ideal types at polar ends of Azángaro's social scale: on the one hand, the indigenous peasant who was born and lives in an Indian community, lives poorly from the income derived from herding and crop raising on small parcels of land, is a monolingual Quechua speaker without any schooling, is labeled as "Indian" by members of all other social groups (including notaries), wears rough, self-made baize clothes, and derives supplementary income from working on nearby haciendas, from artisanal home production, or from petty trade activities that remain secondary and do not break through the predominantly agrarian pattern of his life. On the other end of the scale is the hispanized large landholder, the ideal type of category 2, who speaks Spanish, has enjoyed formal education, owns one or more large estates employing dozens of peasant families as colonos, usually lives in a townhouse in the provincial capital, might even own a house in Puno, Arequipa, or Lima, dresses in European garb, consumes imported goods, holds important administrative, legislative, and judicial positions in the province or the department, and maintains close contacts with higher strata of regional and national society. This construction of ideal types describes both the fluidity and the stark polarity in the socioethnic stratification of Peru's sierra.^[3]

Most individuals appearing in notarial transactions were not idealtypical "indigenous peasants" or "hispanized large landholders" but rather approximated one of the two polar constructs. I felt justified in including most merchants and officeholders in category 2, since they were pulled toward the ideal type of the hispanized large landholders. This, after all, is why they appeared so often before the notary as land purchasers. Socially they had more in common with the largest hacendados of the province than with the Indian peasantry. Category 2 thus encompasses all sectors of Azángaro's provincial elite, of which the large landholder was the dominat figure.

Category 3 includes petty traders, owners of minimal estates (perhaps employing two or three colonos) imperceptibly shading into peasant estancias, persons about whose knowledge of Spanish the notarial records waver back and forth, and persons who generally lived in the district capitals but who may also have resided on their modest landholdings. The distinction between this intermediate group and impoverished hacendados

on the one hand, and affluent peasants on the other, is minimal in most cases. For this study the distinction is based on an arbitrary cut in the continuous scale by requiring two social indicators, the minimal sign of "clustering," for assigning an individual to one of the two polar categories.

Although information on the economic situation of the contracting parties (e.g., the amount and type of land property held prior to their participation in notarial transactions) has influenced their assignment to one of the social categories, the transactions themselves were not used for this purpose. To be sure, a person buying or selling a large estate was excluded ipso facto from category 1. But because the accumulation or loss of landed property has to be viewed as one important indicator for social mobility in Azángaro, consideration of the notarial transactions for determining an individual's social category would have resulted in a circular argument. Thus, a person who in early contracts shows all the traits of an indigenous peasant will be treated as such for all subsequent contracts, even though he might purchase such an amount of land that at the end of his "career" he might more meaningfully be placed in category 3 or even category 2.

Hacienda Expansion

The enormous upsurge of sales constituted the single most important factor influencing the provincial landholding pattern between the 1850s and 1920. Sales of rural landholdings represented the overwhelming majority of contracts recorded by notaries. Their annual number and total value grew from a trickle during the 1850s to an avalanche during the first and second decades of this century.^[4] Nearly three-fourths of all sales contracts concluded between 1851 and 1920 were transacted during the last two decades of that period (table 6.1).

This growth of land sales did not proceed in a linear fashion. A first cycle of sales activity reached a peak in 1867 with forty-seven sales contracts evaluated at 35,549 soles m.n.; then their frequency and value declined until about the early 1880s. Only after the end of the War of the Pacific did a recovery begin, and during the early and mid-1890s a level close to the previous peak year of 1867 was reached. A quantitative leap occurred in 1898–99, when within two years the frequency of sales roughly tripled, accompanied by a smaller increase in total value. After 1904 the frequency of sales climbed again to reach 297 sales during 1908, five times more than in the peak year of 1867 during the first cycle. Between 1908 and 1913 the number and value of land transactions reached their peak for the whole period under consideration. In 1914 there followed a brief but sharp slump.

The recovery during the remainder of the decade failed to bring back the frantic rhythm of sales of the prewar years.^[5]

A breakdown of all sales according to price allows a clearer understanding of the types of land property changing hands in Azángaro. In over 88 percent of all sales the price lay below 500 soles m.n., and for nearly two-thirds of all cases the price was below 200 soles m.n. In contrast, only 2.9 percent of all priced transactions concerned pieces of land for which a buyer paid 2,000 soles m.n. or more. But how much land could one buy for 200, 500, or 2,000 soles m.n.? There was no easily discernible correlation between the price of land and its size. Beyond factors such as quality and location of the land, the social positions of buyer and seller often influenced price. Powerful hacendados might force a peasants dependent on them for credit and work to sell their ancestral estancia for a price much below its "market value."^[6]

If we differentiate sales contracts according to property categories, the preponderance of the estancia is overwhelming. But what did contemporaries understand by estancia ? During the colonial period the term had referred to livestock ranches.^[7] For some areas of Spanish America, notably Argentina, estanica even today refers to sizable livestock estates. In the altiplano, however, it ceased to have that meaning by the mid-nineteenth century. Just as David Brading has described for west-central Mexico, the term hacienda began to be used for a particular type of landholding in the

altiplano only during the eighteenth century. It took one hundred years, from the mid-eighteenth to the mid-nineteenth century, for it to replace the term estancia in the semantic field of "livestock estate operating with semi-servile labor."^[8] By the 1850s such properties were usually labeled haciendas or fincas , but on occasion estancia was still used with this meaning.^[9] In 1849 the Puneño Juan Bustamante, while speaking of the region's "haciendas" still referred to their owners as "estancieros," a term soon to be supplanted by "hacendado."^[10]

After the 1860s the term estancia was not used any more when speaking about livestock estates in the Peruvian altiplano. In its new semantic field it referred to the Indian peasant's rural "farm," with its adobe huts, a corral enclosed by a stone wall, pastures, and some parcels planted with potatoes, quinua, and other crops. Estancia became equivalent to "agricultural units of family or sub-family size," in recent social science terminology.^[11] Writing in 1947, Juan Chávez Molina, the heir to Hacienda Churura, district Putina, flatly affirmed that "the estancia today is a small property."^[12] The term had lost the connotation of livestock ranch. In an analysis of Puno's social stratification in 1931–32, Oswaldo Zea asserted that "in the area of crop farming [zona de chacarismo ] small properties are widespread and are called estancias."^[13] The predominance of estancias in Azángaro land transactions thus offers a first indication that land of the peasantry formed the lion's share of the rural property passing hands in the province.^[14]

Haciendas and fincas together accounted for just over 5 percent of the total number of sales.^[15] On average, fewer than three estates were bought and sold each year between the 1850s and 1910. Size was not a sufficient criterion for differentiating fincas and haciendas from lesser landholdings (fig. 6.1). Attempts to rely on quantitative criteria for defining what constitutes a finca or hacienda, such as minimum size, minimum number of livestock, or minimum production of the property, remain unsatisfactory.^[16] By these criteria there was considerable overlap between small fincas and peasant estancias.^[17] What distinguished the hacienda from peasant estancias or other family smallholdings was its internal social organization: a stable, hierarchically ordered population resided on its land, and its members were "directly tied to the owner or his representative through a number of personal obligations, of both a material and symbolic nature."^[18]

Thus, beginning in the 1850s the same paradoxical constellation of continuity and polarity crystallized in the altiplano's land tenure regime that we observed for the region's social structure: on the one hand, a continuous spread in the size of holdings, from large estates to smallholdings; and on the other, a stark juxtaposition between hacienda and Indian

Figure 6.1
Land Sales in Azángaro, by Category of Property and Price Range, 1852–1910

peasant estancia. This dichotomy was based on the necolonial division between hispanized rural elite and an indigenous peasantry, a division more deeply rooted in the perceptions of the two groups than in different rationalities of their economic operations.

Did the accelerating tempo of land transfers alter Azángaro's landholding pattern or merely represent increased exchange of property within the same categories of landholdings and proprietors? To answer this question, we have to differentiate land sales according to the social origin of sellers and buyers. Among the three social categories of the scale, only members of the category of hispanized large landholders appear as net purchasers of rural property in Azángaro. Only for this group do number and value of land purchases exceed sales (fig. 6.2). By value, some 70 percent of all land transfers into the holdings of this group represent land of indigenous peasants; by number, these transfers are nearly 86 percent. The intermediate group of landholders contributed about 27 percent of the value and 13 percent of the number of land parcels purchased by large landholders.

Land purchases by hispanized large landholders were more markedly cyclical than were overall provincial sales transactions. After negligible amounts of land purchases from peasants and members of the intermediate group during the 1850s, hispanized large landholders expanded their landholdings by more than 140 notarized purchases during the 1860s, with transactions peaking in 1867. During the next two quinquennia these purchases steadily declined to a low of 35 transactions for the period 1876–80, roughly 60 percent below the level of 1866–70.

Land purchases from indigenous peasants continued at a low level through the early 1880s and began to pick up again only in the quinquennium 1886–90 (table 6.2). The brief surge of purchases by hispanized large landholders during the early 1880s resulted exclusively from an increase in transactions with persons assigned to the intermediate group. These transactions during the difficult years of the War of the Pacific may point to the problems faced by impoverished owners of marginal fincas. The double crisis of the early 1880s, in which war-related disruptions of livestock production and commercialization of wool were added to the problem of sliding wool prices, perchaps caused some concentration of landholding within the estate sector while expansion of haciendas into peasant land slowed down.^[19] On September 10, 1881, for example, the brothers Felipe and Manuel Figueroa Obando sold Finca Antocollo y Antaña in Putina district without livestock capital for 3,600 pesos (2,800 soles m.n.). At least one of the brothers had been in debt for five years; during the war years they apparently saw no way of extricating themselves from this debt other than selling their land.^[20] After 1881 the Figureroa Obandos disappeared from the ranks of estate owners in Azángaro.

Figure 6.2A
Social Categories of Sellers in Azángaro Land Sales,
1852–1910; Left: By value of sales. Right: By number of sales. 
Sources: REPA and REPP, 1852

The net transfer of land from indigenous peasants to hispanized large landholders (purchases minus sales) accentuates the nadir of the curve for hacienda expansions between 1876 and 1885. For the long run, this measure can best depict the cyclical nature of hacienda formation in parts of Latin America or, as Eric Hobsbawm has put it, how "in the course of post-colonial history haciendas have been formed, expanded, split up and reformed, depending on political change and economic conjuncture."^[21] Sales to peasants increased during this decade of severe commercial, fiscal, and political crisis, just when purchases by hispanized large landholders decreased. As in the quarter century after independence, peasants benefited from the crisis of the hispanized landholding elite, at least with regard to retaining their land.^[22]

During the early 1890s land purchases by hispanized large landholders recovered, and from the second half of the 1890s onward their purchases grew at such rapid pace that all previous land transactions are dwarfed by comparison. The value of their acquisitions increased by 160 percent from 1891–95 to 1896–1900, followed by a further jump in 1901–5 of over 50

Figure 6.2B
Social Categories of Purchasers in Azángaro Land Sales,
1852–1910. Left: By value of purchases. Right: By number of purchases.
Sources: REPA and REPP, 1852

percent and another rise of 125 percent from the first to the second quinquennium of the present century. The peak of hacienda expansion was reached between 1908 and 1913. During 1908, with a total of 212 purchases, the frenzy of land purchases by hacendados and other members of the provincial elite approached two transactions every three days.^[23]

How are we to account for the accelerating rhythm of hacienda expansion between the late 1850s and 1913? Both contemporary observers and modern scholars have pointed to the favorable conditions thought to prevail in southern Peru's wool export trade as the most important incentive for many hispanized large landholders to expand their estates and found new ones.^[24] In 1916 José Sebastián Urquiaga—who, through one of Azángaro's most spectacular land-purchasing campaigns, had turned the small maternal finca Sollocota into one of the province's largest estates—explained this context as follows:

About 25 years ago [i.e., 1890–91] the haciendas of the department of Puno passed nearly unnoticed as profitable estates;

their products, such as livestock, wool, dried meat, cheese, butter, etc., were sold at extremely low prices, less than half of what they are today. And as the price improvement has made itself felt year after year, the interest in acquiring fincas in the interior was awakened. But since the owners with very few exceptions did not sell their estates, people thought of buying estancias from the ayllu Indians.^[25]

Land purchases by hacendados followed export market conditions for wool as the single most important product of the altiplano livestock estates closely enough to reflect cyclical swings of wool prices and export volumes (fig. 6.3).^[26] The year 1914 confirms the close correlation between trade conditions and land purchases by hispanized large landholders. The brief dislocation of commercial and credit circuits in southern Peru brought on by the outbreak of World War I coincided with a sharp reduction of land acquisitions by hacendados.^[27]

But by the late 1910s something had changed. In these years the value of wool exports reached unprecedented heights, and hacendados were more prosperous than ever before. yet, although in 1918 and 1919 they again acquired much more land than they had in the crisis year of 1914, acquisitions from peasants lay considerably below the levels of 1908–13. Just when the wool market reached its peak, the correlation with land acquisitions broke down. One mighty reason discouraged many hacendados from attempting to acquire more peasant lands: the Rumi Maqui Rebellion, which had swept through several livestock districts of Azángaro province since mid-1915, lifted the level of peasant resistance to encroachments on their lands to a level not seen in the altiplano since the late colonial period. Resistance by community peasants against hacendados continued for the rest of the decade, becoming broader and more ideologically charged during the early 1920s.^[28] The tide of hacienda expansion thus began to turn even before the wave of prosperity brought on by rising wool exports had crested.

Other authors have explained the drastic and rather sudden increase of land purchases by hispanized landholders since the mid-1890s as the result of political constellations. Karen Spalding suggested that hacendados required strong military backing in order to shift the balance of landholding in their favor. The conditions for extending such military support were created only when the Civilista oligarchy effectively took power in Lima after Piérola's victory over the Cáceres forces in 1895. The growth of altiplano livestock haciendas thus "was primarily the product of the alliance between the serrano [highland] political elite and the nouveaux riches of the coast, who themselves were dependent on their alliance with foreign capital."^[29]

Figure 6.3
Annual Averages of Land Transfers to
Hacienda Sector and of Sheep Wool Exports, 1856–1910, in  Soles m.n.

There are two serious problems with this thesis. (1) There is no evidence for the development of such a general class alliance between altiplano hacendados and the ruling coastal oligarchy. Azángaro's politically active hacendados were split along party lines, and many of them fought the local and regional representatives of Lima's ruling Civilista party. (2) Under normal circumstances hispanized landholders in the altiplano did not require military or police contingents supplied by the central government to expand their estates onto peasant lands.^[30] As early as 1874 Azángaro's subprefect Daniel Rossel y Salas, a native of Putina who later became judge of Puno's Superior Court and, by marriage, co-owner of Hacienda Huasacona in Muñani, had vividly portrayed how the province's large landholders could carry out armed raids totally independent of, and indeed against the will of, the national police.^[31]

In 1931 José Frisancho—who was the offspring of a prominent landholding family from the altiplano town of Pucará and whose judicial career took him from the position of prosecutor in Azángaro's Court of First Instance to the presidency of Peru's Supreme Court—presented an argument similar to Spalding's.^[32] As he saw it, the centralism of Lima's "pseudo-aristocracy" after 1895 subverted the political and moral fiber of serrano society, making it servile and corrupt.^[33] This subversion had a sudden and dramatic impact on landholding in the highlands. Between independence and 1895 the upright, patrician landholders of the sierra were "mere conservers of the colonial haciendas," and "not a single case" existed in which indigenous community lands were incorporated into a latifundium. "After 1895 occurred the rapid transformation of the communities into latifundia, to such a degree that in some provinces the ayllus have disappeared."^[34]

In Frisancho's view the new political regime established by Piérola brought a class of persons to the surface of altiplano society who had no scruples about appropriating peasant lands. In reality, however, there was no sudden qualitative break in the relations between hacendados and Indians and, more specifically, in the methods of acquiring peasant lands. The whole panoply of a paternalistic, neocolonial society—from entrapment through debt to legal ruses and sheer force—became gradually entrenched during the decades before 1895.

But Frisancho's analysis contains one important observation: after the end of the War of the Pacific, and increasingly during the 1890s, many newcomers to the altiplano began to form estates from strings of peasant estancias. These new residents were attracted to the economic opportunities linked to the short-lived mining and rubber collection boom in the ceja de la selva of Carabaya and Sandia but also to the growth of commercial

networks in the altiplano. The military campaigns during the War of the Pacific and the subsequent civil war may also have led some soldiers and officers to settle in the altiplano.^[35] Among the 436 men entered in Azángaro's electoral register for the year 1897, 96 were born outside the province.^[36] The influx of these newcomers during the late 1880s and early 1890s accounts for a good part of the increased land purchases in Azángaro following 1895.

The two decades between Piérola's civil war victory in 1895 and World War I marked the high tide of gamonal power over the Indian peasantry of the altiplano. The consolidation of Peru's oligarchic regime during what has come to be called the Aristocratic Republic set the stage allowing the region's elite of large landholders, traders, and officeholders to take full advantage of the favorable commercial conjuncture that whetted their appetites for pastoral resources and peasant labor. The penetration of a modern network of transportation and communication through the southern sierra, fueled by exaggerated hopes of economic development, and the expansion of the infrastructure and fiscal means of the central state helped to strengthen the altiplano's elite after 1895.

But the path on which this elite had come into being, in an incipient and contradictory fashion between the 1820s and 1850 and in a first moment of ebullient power between the late 1850s and early 1870s, was unique to the south. This newly forming elite was indebted to the central government only through its relative weakness in the southern highlands. The bases of the large landholders' power, independent of any alliance with oligarchic groups on the national level, were their favorable insertion into the expanding commercial circuits centered on Arequipa as well as the consolidation of clientele systems, a process made easier by the commercial expansion. It is true that beginning in the 1890s the provincial elite of the altiplano expected increasing support from the central government to consolidate their gains vis-à-vis the peasantry, both through distribution of funds and offices derived from Lima among their clients and through strengthened police and military contingents in the countryside. Yet the gamonales never saw themselves as junior partners of Lima's oligarchy. They insisted on maintaining independent power in the provinces against what they perceived as the threat of subversive modernization pushed by the central government.

The Value of Land

If indeed an improvement in the economic conjuncture for altiplano livestock products constituted the major motivation for the rapid increase of

land purchases by Azángaro provincial elite, then this improvement should be reflected in rental rates of estates and the price of land itself. Rental rates paid for Azángaro livestock haciendas, in spite of problems with the data base,^[37] show a remarkably clear development.^[38] From just over 8 percent of productive capital in soles m.n. (just over 10 percent in pesos) during the 1850s, the mean rate jumped to just under 10 percent in soles m.n. (just over 12 percent in pesos) in the 1860s. During the following twenty years it stagnated, and only during the 1890s does a new rise become visible. The rate climbed from 12 percent in soles m.n. during the 1890s to 15.6 percent in the following decade and 22 percent between 1910 and 1917, nearly three times as high as during the 1850s. The average annual lease paid on a hacienda fully capitalized at 1,000 OMR jumped from 80 soles m.n. in the 1850s to 220 soles m.n. during the boom years of World War I.^[39] At that time an hacienda produced as much in annual leases without any livestock capital as it had done sixty years earlier when fully capitalized.^[40]

The increased profitability of operating a livestock estate during the 1860s, expressed in rising rental rates, coincided with the first postindependence push toward hacienda expansion in Azángaro. When southern Peru's wool export economy went through its phase of retraction, from about 1873 until the mid-1880s, rental rates stanated and hacendados scaled down land purchases, reaching an extremely low level between the mid-1870s and 1885.^[41] Growing income from wool exports after the mid-1880s—preceding the recovery of the international conjuncture for wool by about ten years because of the devaluation of the Peruvian currency—was reflected in renewed increases in rental rates beginning in the early 1890s. These increases in turn set the stage for the new wave of land purchases by new and old hacendados, which climaxed during the five years preceding World War I.

Continuous sets of rental contracts for one and the same estate spanning the greater part of the period from the 1850s to the 1910s exist only for small, usually church-owned fincas.^[42] Soñata, one such small, church-owned finca, was located in the parcialidad Cacsani in Arapa. Because of its favorable situation on a bend of Río Azángaro, the finca in 1913 could dedicate 100 of its total 668 hectares to crops. On the remaining second- and third-class pastures it could maintain about 1,070 sheep.^[43] Surprisingly, the rent charged by the church on Soñata reflected the general rise of rental rates with a considerable lag. After neglecting even to charge the tenant for excess pastures in 1860, the rate stayed unaltered at 10 percent (in terms of pesos) until the 1890s. Only after 1901 and then again in 1914 did the annual rent paid for Soñata participate in the general upward trend of rental rates.

The explanation for this "unorthodox" finding has to be sought in the slower-than-average rise of minimum rental rates. Small fincas, often undercapitalized and equipped with an insufficient colono labor force, too little water, and inferior pastures, did not partake in the growing profitability of hacienda operations during the 1860s and the years between 1890 and 1920 to the same extent as large ones did. For such estates the church could find renters only at below-average rates.

It proves more difficult to determine the development of land values during this sixty-five-year period. The most obvious problem consists in the scarcity of information on the size of landholdings, mentioned in notarial contracts only since the early twentieth century. The measurement of property, carried out primarily by nonprofessional agrimensores, was required for the inscription of landholdings into the departmental land register, established in 1889. Although these measurements purported to be exact down to the last square meter, their reliability is rather low, and they could vary by 25 percent or more.^[44] Until the 1890s the rare mention of a landholding's size merely gave its approximate perimeter or its length and width by kilometers. Thus, it is not possible to determine the development of the average hectare prices over a longer period of time.

The only way to use the large data base of notarial sales contracts for measuring changes of land prices consists in determining the average price of land per transaction. For most categories of sales the mean purchase price per transaction declined from the 1860s or 1870s to the 1880s and 1890s. In all cases it grew between the 1890s and the first decade of this century. These changing average price levels can be interpreted in two ways: either land prices fell between about 1870 and 1890, or the mean size of land parcels offered for sale went down. Individual property histories suggest that even during the difficult years between the late 1870s and early 1880s land prices did not decline. Thus, it would seem that the mean size of land parcels offered for sale tended to decline between the 1860s and the 1880s or 1890s, an early indication of the impact of population increase on Azángaro's landholding pattern. Of course, we cannot exclude the possibility that land values had already begun to climb in the 1860s or 1870s and that these increases merely failed to show up in mean sales values because of the reduction of the mean size of land parcels. It is equally possible that the size reduction continued after the 1890s but was now covered up by a stronger increase of prices for land. One might interpret the peculiar development of the mean purchase price per transaction as the outcome of a scissorlike configuration between land values and mean size of land parcels offered for sale: until the 1880s or early 1890s decreasing

average size had a stronger impact on mean purchase prices than did the increasing land value; this relation was reversed after 1900.

Prices for land organized in estates underwent sharp increases between the 1850s and 1920, generally by between 100 and 200 percent (table 6.3). Increases occurred both in the early phase, from the late 1850s to the 1880s, and in the second half of the period, from the 1890s. Haciendas whose value temporarily declined (Loquicolla Chico, Huañingora) had experienced a decline in their livestock population.^[45]

The crucial variable in the appraisal of pastureland was its livestock carrying capacity. We can piece together a set of figures covering a longer time span for the value of the unit of land required to feed one sheep year-round plus the value of the sheep itself (table 6.4). This had been the conventional measure in the appraisal of estates since the colonial period. Of course, the actual size of this unit varied according to the carrying capacity of the land. In all, the price of a unit of top-quality pastureland required to feed one sheep plus the sheep itself increased by about 250 to 300 percent between the 1850s and the second decade of the present century.^[46] One can again discern two waves of price increases, one prior to the 1890s, probably during the 1860s, and a second, showing a considerably greater rise, between 1902 and World War I.

This price development for land was closely linked to prices for live sheep, which rose at about the same rhythm as land values (table 6.5). Livestock prices were the key factor directly influencing the value of pastoral land in the altiplano. The values of sheep and cattle, in turn, are directly related to the prices of the products derived from them, and here wool is crucial for the altiplano. The two phases of price spurts for sheep occurred when wool prices rose between the late 1850s and early 1870s and again between the mid-1890s and World War I. Livestock prices did not, however, participate in the slump of the 1870s and 1880s; they merely stagnated. This stagnation probably reflected domestic inflation in Peru but might also indicate that the tradition of price conventions, immune to short-term market forces, still had some influence on livestock prices. The impetus for increases in the price of land, during the 1860s and again between about 1895 and the end of World War I, came, transmitted by livestock prices, from heightened demand for the altiplano's most important commercial product, wool.

The preceding discussion of rental rates and prices for land is based on information concerning haciendas or land in the hacienda sector. The development of market prices for peasant lands is much less clear. Moreover, the very notion may be flawed. Consider the constitution of Hacienda

Lourdes in Potoni district between 1880 and 1901.^[47] The eight distinct estancias and fincas making up the new estate had been purchased by Adoraida Gallegos for no more than 4,600 soles m.n. Yet in 1901, less than ten years after most of the parcels of land were purchased, an appraisal gave Hacienda Lourdes' value as 12,156 soles m.n., not at all unlikely for an estate extending over 11,719 hectares. Only a small part of this jump of nearly 200 percent in the value of the integrated landholding over the purchase price of the individual estancias reflects the general price trend. The greater part resulted from the higher unit value of land organized in estates, with a stable resident labor force and its own livestock capital, compared with unincorporated peasant land. The formation of Hacienda Lourdes illustrates the economic advantages enjoyed by Azángaro's owners of estates over peasant landholders.

Repeated sales of peasant lands reveal rather peculiar price developments. On April 24, 1908, for example, the peasant woman Teresa Quispe Huarcaya from parcialidad Curayllo in Arapa sold Estancia Collini Accohuani to José Albino Ruiz, politician and owner of the large Hacienda Checayani in Muñani, for 200 pesos. "More than twenty years" earlier she had bought the estancia from Marta Huarcaya for the same price.^[48] In a

similar case, the Indian peasant Marcelo Sacaca in 1888 had purchased estancia Pachaje Chico in Putina for 1,600 soles en quintos bolivianos (soles q.b.) with money inherited from his mother-in-law Juliana Quenallata. Nearly twenty years later, on June 22, 1907, two granddaughters and coheirs of Juliana Quenallata, Jacoba and María Mejía Machaca and their husbands, sold their share in Pachaje Chico to Manuel Esteban Paredes Urviola, a member of one of the largest landholding clans and the only medical doctor of the province, on the basis of the 1888 transaction, without any price increase.^[49] But in other cases peasant land underwent surprisingly large price increases over short periods. On June 25, 1910, Francisco Adrián Toro Nafria, a merchant and hacendado from Asillo, bought Fundo Quisini, located in parcialidad Jila or Supira (the notarial contracts are contradictory) of the San Antón district, from the peasant Damaso Hispanocca Vilca for 50 soles m.n. Two months later, on August 13, 1910, Toro Nafria resold the land to Pablo Anco Turpo and his wife Victoriana Turpo Ccallasaca, peasants from San Antón, for 150 soles m.n., a 200 percent gain, diminished only slightly by the fees for title deed and sales tax.^[50]

Such contradictory price developments for individual peasant landholdings reflect the social position of both parties to the transactions. Where the landholding was resold without any price increase even after several decades, the seller was an Indian peasant and the purchaser a hispanized large landholder. The speculative gains, by contrast, were realized by hispanized large landholders reselling (or trading) a property to peasants. These price developments, then, have their common explanation in the social dominance of one contracting partner over the other.

The nature of prices for land changed slowly in the altiplano. Around 1850 prices for land, like rental rates and livestock prices, still had a strongly conventional character. Short-term economic fluctuations did not easily disturb this pricing structure. Within the region all people involved in transactions over land or livestock shared a common understanding about these pricing rules. Phrases such as "according to the conventions used in this department" or "customarily fixed at" appeared often in notarial contracts. This is not to say that land values and livestock prices bore no relation to market conditions; rather, major secular economic changes were required to prompt readjustments in those conventional prices. Thus, for example, the conventional price for one adult sheep remained at four reales between the beginning of the nineteenth century and the 1850s.^[51] Rental rates for estates continued to be calculated on the same basis of 10 percent of livestock capital during the 1850s just as they had in the 1770s.^[52] The few examples at our disposal to analyze land values in this early period suggest the same type of stability.^[53]

During the following sixty years the impact of market conditions on land values became more pervasive, and convention gradually lost strength as the basis for determining prices and lease fees on land and livestock. This development was aided by the growing technical capacity to measure the quality and extension of land. But it had its causes in the increasing awareness of the impact of market fluctuations on land values and income levels among some of the region's landholders, itself a consequence of the link to the world market for wool. This process, by which prices for pastureland came to reflect market conditions for the altiplano's livestock products, was far from complete by 1920. Azángaro's pricing structure for land remained ambiguous during the early years of this century, leading at times to seemingly erratic prices for one and the same landholding. This ambiguity allowed members of the socially dominant groups to draw extra profit from land transactions with peasants and tradition-bound landholding institutions such as the church.

To summarize the development of land values in Azángaro: The value of estates generally rose between 100 and 200 percent from the 1850s to about 1920, with prices for top-quality pastureland rising by up to 300 percent. First-class haciendas, endowed with the best pastureland, sufficient resident labor, and, most important, extensive acreage, experienced a steeper rise in their value than did smaller estates with scarcer natural and labor resources. By contrast, the price for peasant land does not reflect market conditions because in most cases it was shaped by the social constellation between buyer and seller rather than general price trends. Still, for estates the link between wool exports, livestock prices, and land values explains why old and new landholders adjusted their purchases of land to the rhythm of the regional wool export economy: earnings from estate operations followed this rhythm.

The Geography and Ecology of Hacienda Expansion

The enormous expansion of estates onto peasant lands between the late 1850s and 1920 affected Azángaro's districts and communities unevenly. Whereas the four districts in which indigenous peasants sold the most land to hispanized large landholders and members of the intermediate group accounted for more than 60 percent of all transactions, the four districts with the fewest sales by peasants accounted for only 8.5 percent.^[54] In Azángaro, by far the district with most sales by peasants to other groups of landholders (nearly 30 percent of the total number), peasants alienated land valued at over 96,000 soles m.n. in 563 transactions until 1910. At

the other extreme, for Caminaca district the notaries registered only 18 such transactions for a total of under 2,700 soles m.n.

Distinct ecological and land-use zones in the province had a bearing on the spatial distribution of land sales. Caminaca and Saman in the south of the province experienced only a limited expansion of haciendas onto peasant lands. On the plains between Lake Titicaca and Lake Arapa agriculture played a relatively more important role than it did in other parts of the province. Until the mid-nineteenth century only a few small estates had been established there, and land was more evenly distributed than it was in the pastoral areas of the province. The comparatively equal distribution of lands and the absence of powerful hacendados inhibited the subsequent advance of estates into peasant communities since large agglomerations of land could not serve as points of departure for acquisition schemes. When a few small estates were finally formed in Saman after 1900—among them Finca Santa Clara, owned by the hacendado and livestock trader Ildefonso González from Arapa, and Finca San Juan, owned by the local merchant and livestock trader Mariano Abarca Dueñas—peasant resistance was particularly adamant.^[55] After 1912 bloody clashes ensued between these gamonales and the peasants.^[56] In the short run gamonales could inflict losses of land, material belongings, and even life on peasants. But in Saman the balance of power between peasants and haciendas was such that after 1940 no hacienda survived.^[57]

The four districts in which hacienda expansion had the greatest impact in relation to population lay in the predominantly pastoral zone of the province to the northeast of Azángaro town. In three of these, San José, San Antón, and Potoni, ecological conditions favored livestock raising, but only a handful of haciendas had existed in the early republican period, perhaps because of their isolation from trade and communication circuits of that time. There hacienda expansion had particularly dramatic consequences for the land tenure pattern. Although by the early independence period a considerable number of estates had existed in the fourth district in this group, Azángaro, it was also the location of the greatest number of peasant parcialidades. The combination of a strong estate base and plentiful peasant lands, as well as the advantages offered by the provincial center with its commercial and administrative opportunities, accounts for Azángaro's predominant share in land acquisitions by established or aspiring large landholders.

Muñani and Putina, with their fertile valley pastures and their arid cordillera slopes unfit for crops, belonged to the areas most suited for livestock raising in Puno department. Yet in proportion to population both districts experienced a low rate of peasant land sales to other groups of

landholders. Here a great part of peasant lands had already been absorbed by haciendas by the early republican period. By the 1870s little peasant land remained in Muñani. In Putina a number of parcialidades survived at least until the early twentieth century, and peasant land sales were rather frequent in absolute terms. Transactions among hacendados and intermediate landholders had a greater weight in Putina than in any other district, corroborating the existence of numerous haciendas and smaller nonpeasant landholdings since before 1850.

In contrast to what happened in the seventeenth and eighteenth centuries, Azángaro's "hacienda frontier" now moved outward from the center of the province. During the 1850s and 1860s more than 80 percent of peasant land sales to hacendados and intermediate landholders concerned land in Azángaro district, with only a sprinkling in all other districts. The bulk of purchases can be attributed to purchases by Juan Paredes and José María Lizares, the most dynamic hacendados in the province between independence and the War of the Pacific. Both men came from families that arrived in Azángaro after independence, and both held administrative offices and participated in trade in the provincial capital. They could build up a clientele system most easily in the immediate vicinity of their residence, seat of business and of official power, and thus began their land purchases in Azángaro district.

Even though the market for livestock products had become favorable, most hacendado families in the rest of the province did not expand their landholdings onto peasant lands during this period. Until the late 1860s it remained relatively easy to find estates for lease, and despite the wool export boom, a sufficient number of hacendados and owners of smaller nonpeasant landholdings were still willing to sell land. Apparently not all established landholders were profiting from the favorable market conditions, and some saw themselves forced to give up their estates. Such sales occurred even during the second boom phase after the 1890s.

Supply and demand for pastoral land were still largely balanced during the 1860s. Aside from the situation in Azángaro and Muñani districts, potential land buyers had no need to resort to peasant estancias. After 1890, by contrast, mushrooming demand for land could not be met any more by nonpeasant landholdings offered for sale. During the first boom period in the 1860s Azángaro's land tenure pattern was still affected as much by changes in the composition of the hacendado class as it was by outright hacienda expansion, whereas after 1890 the weight shifted decidedly to the latter mode.^[58]

From 1870 to 1890 the "hacienda frontier" moved into some of the districts in which relatively many haciendas had existed before, including

Putina, Arapa, and Santiago. Some of the districts that were to become central locations for hacienda expansion between the 1890s and 1920, such as San José, San Antón, and Potoni, still saw no or only minimal land transfers from peasants to hacendados and intermediate landholders. Until 1890 the purchases of peasant lands served mostly to enlarge well-established estates rather than to constitute new ones. Although a few new estates were formed during the 1870s or before, most were founded after 1890. Only the increasing demand for land beginning in the last decade of the past century—coinciding with the entry of a substantial number of newcomers into Azángaro's land market—prompted aspiring estate owners to penetrate hitherto unaffected Indian communities. It was easier to acquire peasant lands in districts with a long-established hacienda-community complex, since there peasants were more closely tied into patterns of dependence and paternalism. By contrast, in areas such as San Antón and Potoni the majority of peasants had traditionally lived and worked outside the sphere of influence of estates.

In these areas the advance of the hacienda frontier was aided after 1890 by the opening of the road from railroad station Tirapata via Asillo, San Antón, through Potoni district to Macusani and from there to the ceja de la selva of Carabaya. The trade connected with the rubber and gold exploitations in Carabaya province, which now passed through these districts, and the growing network of itinerant wool buyers regularly visiting every parcialidad multiplied the opportunities for intricating peasants into clientele systems as the prelude to land acquisition schemes. During the thirty years following 1891 the tide of land transfers from peasants to hacendados and intermediate landholders engulfed nearly the entire province. Only Muñani, with little peasant land remaining, and Caminaca were largely untouched by the avalanche. This period saw particularly rapid growth of old and new estates in the northeastern districts of Potoni, San Antón, and San José. It may not have been mere coincidence that the most serious peasant rebellion, the Rumi Maqui Rebellion of 1915–16, centered on San José district.

Sales of peasant land to hacendados and intermediate landholders varied greatly from parcialidad to parcialidad. The data must be interpreted with great caution, as parcialidades or ayllus could change their boundaries over time.^[59] In seven of Azángaro's ten districts with six or more parcialidades, more than 50 percent of all peasant land sales occurred in just two parcialidades. In five of these districts, the two communities with the heaviest sales activity accounted for more than 70 percent of all such transactions. Of the five parcialidades in Saman, one, Chejachi, accounted for 82 percent of all peasant land sales to other categories of landholders (88.6 percent by

value). Chejachi is located on Saman's northern border, next to Arapa district. Hacendados who owned landholdings in Arapa extended their land purchases into the neighboring district. The two districts in which peasant land sales were least concentrated in a few parcialidades, Azángaro and San José districts, were those with the highest number and value of sales. Here the expansion and new foundation of haciendas spread over most areas of the districts so that no parcialidades stood out.

Different size and population of individual communities may account for some of these variations. In the absence of information on the extension of parcialidades I tested the relevance of population by selecting two districts, Caminaca and San Antón, for which the parcialidades listed in the censuses of 1862, 1876, and 1940 remained the same and were nearly identical to those appearing in the notarial registers. Population does make a difference in the ranking and spread of land sales activity in various parcialidades. But great variations remain unexplained by population. Using the average population from the three censuses, in Hila, San Antón, one sale was transacted by every seven community members, whereas in Sullca one sale was concluded by every thirty-three persons. In terms of value more than ten times as much land per capita was sold in Sillota as in Sullca. In Caminaca, with only a few peasant land sales, per capita sales differed little in three parcialidades, but there were none in a fourth community.^[60]

What accounts for this great variation of per capita land sales in different parcialidades? Just as between districts, the varying weight of agriculture and livestock raising probably played an important role. So did more specific locational conditions, such as the proximity of a parcialidad to established haciendas and the availability of water, good pastures, and collpares. The ten parcialidades with the highest number or value of peasant land sales to hacendados or intermediate landholders (map 6.1) together accounted for about one-third of all such sales. Specific locational factors help explain the heavy losses of land in most of these communities. Parcialidad Jayuraya, for example, was located in the hilly and broken terrain southwest of the pampa of Río Tarucani between Muñani and Putina. For many of the estates clustered around these two towns, and particularly for Hacienda Checayani, Jayuraya formed the only potential area of expansion, since they were hemmed in by other estates on all other sides. Similarly, we can detect locational reasons for limited alienation of lands in some communities. Between the 1850s and 1910 parcialidad Sillota, in Asillo, lost only one-third as many parcels of land to hacendados and intermediate landholders as did Collana (in the same district) between the 1850s and 1910, although Collana had 50 percent fewer inhabitants.

MAP 6.1.
The Ten Parcialidades with the Greatest Number and/or Value of
Land Sales by Peasants to Hacendados and Intermediate Landholders, Azángaro Province

Much of Sillota's territory, north of Asillo town, offered quite favorable conditions for agriculture in the low foothills on both sides of the Río Grande; only small fincas developed here.

But do locational factors suffice to explain why parcialidades were affected so differently by the process of hacienda expansion? It seems likely that the internal situation of the parcialidades played an important role in determining vulnerability to outside pressure. Factors such as community

solidarity, the affluence or poverty of community members, demographic pressure on the land base, cleavages between rich and poor peasants, and the survival of autonomous cultural traditions could help or hinder access of aspiring outside landholders to the land of community members. Local cultural and ideological traditions and politics thus intervened in blocking or facilitating the avalanche of hacienda expansion.^[61]

Expansion Strategies

Interpretations of the modalities of hacienda expansion in the altiplano cluster around two extremes, which might be called the myths of free will and brutal force. Apologists have stressed the legality of most land acquisitions by hacendados and described the purchases of peasant lands as a contract between free and—as far as the transaction goes—equal parties. This apologetic view was expressed by the first organization of Puno's hacienda owners in a petition sent to President Augusto Leguía in 1921. The hacendados claimed that estates were based either on colonial land grants and compositions or on "just and legal titles for land . . . purchased, without deceit and according to the law, from the Indians since the period in which these were declared owners of the land they had occupied and free to sell."^[62]

The other extreme view regarding hacienda formation became widespread with the rise of indigenismo in the second decade of this century and has remained one of the points of accusation against the agrarian regime of the southern sierra before 1969. As the sociologist Jorge Lora Cam put it in 1976, "Haciendas in Puno were mostly formed by violent usurpation of the land of the community Indians, converting some of them into colonos enserfed to the gamonal power, while others were thrown out if they were not physically liquidated."^[63]

Both characterizations of the process of hacienda formation and expansion fail to grasp the complexity of socioeconomic interaction patterns in the altiplano during the late nineteenth and early twentieth centuries. A sales transaction between free and equal economic agents and a violent usurpation of land have one thing in common: they are punctual actions. Both parties, seller and purchaser in the first case and aggressor and victim in the second, need not have interacted previously. This lack of interaction may be characteristic, on the one hand, of individualistic market societies in which all actors at least ideally have the freedom to make economic choices and, on the other hand, of societies in which social cohesion has utterly broken down, as in conquest or civil war. But in a society such as that of Azángaro, built on long-standing ties of family, religion, and

community and on neocolonial hierarchies, sales transactions between free and equal economic agents and violent usurpations of land were merely the extreme modes of land acquisition on a scale in which most cases were characterized by a mixture of volition and coercion.^[64] The social positions of the contracting parties, their economic resources, the existence or nonexistence of kinship ties between them, and other factors gave special significance to each individual transaction.

Land purchases, the most frequent form of estate formation and expansion, varied greatly as to type and size of land, social status of seller and buyer, and circumstances leading to the transaction. Those with sufficient capital or access to credit facilities tried to purchase small fincas as a central building block in the process of forming a large new estate. The most spectacular example of this strategy was the formation of Hacienda San José in the district of the same name. Sometime during the mid-1880s the half brothers José Sebastian Urquiaga Echenique, born in 1857, and Bernardino Arias Echenique, born in 1860, inherited Hacienda Sollocota, located in the border area of districts Azángaro and San José, from their maternal family.^[65] During the eighteenth century the estate had belonged to Miguel de Echenique, a Spanish miner working veins in Carabaya and an ancestor of President José Rufino Echenique. Although gradually decaying, Sollocota remained in control of the family during the early republican period.^[66] In the decade after Carlota Echenique, mother of José Sebastian and Bernardino, had assumed the administration of the largely decapitalized estate, sometime around 1860, she undertook to restock Sollocota with cattle and sheep.^[67] The family's landholdings expanded only after José Sebastian and Bernardino inherited the finca. Between 1891 and 1905 the two together bought sixty-five separate properties; nearly all were located in the general vicinity of Hacienda Sollocota, some in the northern part of Azángaro district, but most in San José. Among the properties were several small to medium-sized fincas, such as Finca Parcani, with a livestock capital of 1,800 sheep, and Finca Unión. Although Parcani was an older estate, Unión had been recently formed by the previous owner, José Guillermo Riquelme, through agglomerating at least thirteen peasant estancias.^[68] In February 1903 Urquiaga and Arias Echenique bought Finca Quimsacalco, an estate of colonial vintage with a capacity of 4,000 sheep, from three brothers and sisters of José Guillermo Riquelme for 12,000 soles m.n.^[69]

Beginning in 1905 the two half brothers began to purchase land separately, and we may assume that around that year they effectively divided their property.^[70] Urquiaga received the old maternal Finca Sollocota, now greatly expanded by some of the landholdings purchased since the early 1890s. Arias Echenique became exclusive owner of the majority of the

purchased property, including the three Fincas Quimsacalco, Parcani, and Unión. All of this land was incorporated into a new estate called San José. The narrow Río San José now formed the border between the two estates for some fifteen kilometers. Both Urquiaga and Arias Echenique continued to expand their separate haciendas through the 1910s and succeeded in integrating further estates (the colonial Hacienda Puscallani became part of Sollocota) besides dozens of peasant estancias. By 1925 Hacienda San José, according to one measure, extended for 40,000 hectares in the districts of San José, San Antón, and Potoni.^[71] After years of legal fights it passed into possession of the Sociedad Ganadera del Sur in 1944, a corporation owning a dozen large estates in the department of Puno and controlled by the Gibsons from Arequipa, the largest southern Peruvian import, export, and banking house.^[72] Some of the land of San José, within half a century, had been exploited through four different levels of landholdings. Original peasant estancias had been integrated into a small finca (Unión), which became part of the large estate of San José; this hacienda in turn was purchased by the largest corporate landholding company of southern Peru. The formation of the new Hacienda San José, accomplished in thirty years at most, relied on the integration of a number of older estates besides the purchase of dozens of peasant estancias. Without the older estates, in themselves extending over several thousand hectares, San José would have remained a medium-sized finca like most new fincas incorporating only peasant estancias.^[73]

This type of agglomeration was an option only for the wealthiest and most powerful large landholders of the province.^[74] In most cases the expansion or new formation of estates depended entirely on the incorporation of peasant estancias. This was a slow and tedious process, requiring contact with numerous Indian peasants, some of whom might hold rights to only minuscule plots. Given the advancing fragmentation of the original peasant landholdings since the colonial and early republican periods, potential purchasers frequently had to carry out genealogical inquiries to locate all persons with rights to an estancia.^[75] José Albino Ruiz proceeded in this manner in 1908 when acquiring a number of peasant landholdings in the border area between Azángaro and Arapa districts, "with the aim of forming a small finca called Calahuire."^[76] This had been the name of the original landholding owned by the Indian Carlos León Huarcaya in the mid-eighteenth century. Ruiz now was purchasing some ten different shares from third- and fourth-generation descendants who all traced their lineage to that common root, the "stem" of the family.^[77]

Most transactions over land between hacendados and peasants do not mention the circumstances under which they were concluded and, in their

formalistic language, give the impression of a totally free agreement between the two parties. But a sufficient number of contracts grant us glimpses into the great variety of means designed to coax peasants into relinquishing control over their land. Frequently a peasant had received food on credit from a hacendado "for the subsistence of his family" over a period of several years. The creditor would finally present the bill and, if the peasant were incapable of repaying the debt in money or some commodity as wool or livestock, demand payment in form of land. I found 165 cases in which land was explicitly sold as payment of a debt. But in reality the number was higher, as many notarial contracts lacked precision to list such circumstances. Debts leading to the sale of land could arise out of a great variety of situations. The purchase of sheep on credit or monetary obligations in connection with church festivities at times led to land sales by peasants.^[78] More frequent were cases of debts incurred to defray the expenses of a proper funeral.^[79]

Old people often relinquished their land in exchange for future material benefits, including the payment of their funeral, rather than as payment for past debts. Juliana Ramos, a small shopkeeper of peasant origin in Asillo, whose husband and only son had died, in January of 1893 sold her "Hacienda" Viluyo with the adjoining estancias Misquichuno and Hulquicunca and her small store on the plaza of Asillo to Lieutenant Colonel Juan Manuel Sarmiento, a Cuzqueño military officer, for 140 soles. She explained that the properties might be worth a bit more but that she considered the difference a donation to Sarmiento for the services "which I have received from the buyer and for those which I will receive given that Sr. Sarmiento pledges to feed and dress me, to cure me when I am ill, and to bury me when God calls me."^[80]

The sale of land was coupled with the expectation of material aid and protection by a strong hacendado. Often a peasant had no choice but to become integrated into such clientalistic dependencies. For hacendados intent on expanding their landholdings, such transactions brought important advantages. It was easier to persuade Indian smallholders who had come to depend on a hacendado to relinquish legal title to their land than it was to deal with fiercely independent peasants with sufficient economic means to reject any purchase offer or to have to resort to costly legal battles or coercive measures.

Acquiring land from an indebted peasant often reduced the effective purchase price. Although the notarial bills of sale usually stated that the seller had received the full price to his "entire satisfaction," in many cases it was paid only later or not in its entirety at all.^[81] If the seller was a client peasant, the purchaser might pay a considerable share of the total price in

goods, such as foodstuffs, the price of which could be fixed to his or her advantage.^[82] In many cases hacendados withheld a part of the purchase price in order to pay it later to a co-owner of the property not present at the transaction; this money was often retained for good by the purchaser. In other cases hacendados had to pay only part of the purchase price in recompense for services already rendered or to be rendered to the seller in future. In 1867 Juan Paredes acquired an estancia in ayllu Hilahuata from a peasant widow, paying cash for only one-fourth of the property and receiving usufruct of the rest free of charge in recompense for raising the woman's minor child in his household, surely more of a benefit than an expense for the hacendado.^[83]

Cases in which hispanized landholders paid only a small share of the purchase price for a parcel of land after having given legal counsel—commonly without professional qualifications—have become notorious.^[84] In 1899 Lieutenant Colonel Victor Gregorio Rossello, descendant of an old Azangarino landholding family and veteran of the War of the Pacific, was forgiven four-fifths of the purchase price of 500 soles m.n. for Estancia Callapani in district Santiago de Pupuja in recognition of his judicial efforts on behalf of the sellers, a group of peasants and weavers from the parcialidad Mataro in the same district. On top of this, Rossello was forgiven the remaining one-fifth of the purchase price as "indemnification for all the damages which we [the sellers] have caused him through repeated attacks on his properties."^[85] Rossello thus acquired legal title to Callapani without paying one centavo to the previous owners.

This contract demonstrates the ambivalence of paternalism in the context of an expanding hacienda complex and neocolonial social relations. How can we believe that the peasant owners of Estancia Callapani sought out Rossello as their spokesman in court cases when they had had conflicts with him serious enough to lead to recurring attacks on his estate? The notarial contracts cannot directly reflect what the mostly illiterate and Quechua-speaking peasants stated. We hear only what the notary put in their mouths. In the present case direct statements by the peasants probably would have revealed that they were coerced or tricked into the deal with Rossello. The establishment of clientalistic relations here was a one-sided strategy by the hispanized landholder to gain access to the land and labor of Indian peasants.

But such clientalistic links by no means needed to be based on deceit or coercion. Although the dominant partner was likely to benefit most, the reciprocal benefits for the dependent peasant, in the form of protection, credit, and foodstuff in times of need, could constitute significant assets. Some peasants accepted dependence on a paternalistic gamonal as a prudent

course of action without being a victim of specific entrapment beyond that constituted by the structure of society.^[86]

These underlying social relations are indicated by the long delays that often occurred between the conclusion of an informal contract and the entry of the title deed in the notarial register. The original contract might be celebrated before the justice of the peace of the local district or in the residence of the hacendado. It was common knowledge that many justices of the peace were corrupt. An Indian peasant could be induced to sign over land rights without even realizing the impact of the action.^[87] It was simple enough to make up an informal contract through trickery or forgery, since most peasants could not sign their names and one of the witnesses signed for them.^[88] Only years later would the interested party bring the transaction before the notary to elevate it to a fully recognized title document. In the intervening time the peasant might have become a colono on the purchaser's hacienda while continuing to live in his or her old estancia. In other instances the original peasant owner had died in the meantime. Only then did the hacendado approach the heirs to demand the reaffirmation of the original sale, often claiming that part of the purchase price had been paid already to the deceased person as an advance (arrás ). The heirs, unfamiliar with the informal contract as documentary proof of the hacendado's assertions, could often do nothing—short of risking legal actions with all their expenses and delays—but to consent to the notarial reaffirmation of the sale.^[89]

Of 3,060 sales contracts between 1854 and 1910, 278 were protocolizations, with an average lag of nearly eight years between the informal conclusion of the sale and its entry in the notarial register as a title deed.^[90] In another 460 contracts the purchaser was stated to have been in possession of the property for at least four weeks and up to forty years prior to the notarization. In all, there was a lag between the informal conclusion of a sales contract and its notarial registration in a minimum of 24.1 percent of all sales transactions.

The peculiarity of the altiplano land market is illustrated by the importance of a transaction called anticresis or prenda pretorial. It involved the extension of a credit to a landholder, who, rather than mortgaging his or her land as security for the credit, would turn it over in usufruct to the creditor for a period of between five and ten years, at the end of which the loan was to be repaid and the land returned to the owner. The rent of the land and the interest on the loan were considered to be of equal value, so that the debtor did not pay interest and the creditor did not pay rent. An anticresis contract, then, fulfilled two functions, extension of credit and medium-term land transfer, with a minimum of cash transactions.

Which of the two functions was primary for the transaction depended on the social relations between the contracting parties. If the creditor was a hispanized large landholder and the person turning over a plot of land an indigenous peasant, the contract's primary function mostly lay in the appropriation of land by the large landholder. For five to ten years the hacendado was getting a piece of land for extending a sum of money considerably below its sales price. If the peasant failed to return the credit at the appointed time, the creditor could demand the formal sale of the land, for which he or she paid only a minimal additional sum. In this way unfulfilled anticresis contracts represented a variety of the "delay strategy" discussed above.^[91] Adoraida Gallegos, for example, systematically employed anticresis contracts over fifteen peasant estancias to gain control of land at the fringes of her Hacienda Lourdes.^[92]

But anticresis contracts were more important for peasants than for hispanized large landholders.^[93] Nearly one-third of all anticresis contracts were concluded between peasants, a rate much higher than that for sales. The popularity of anticresis contracts with peasants and intermediate landholders reflects two important traits of Azángaro's rural society, the scarcity of cash and the uncertainty of loan repayment.^[94] They affected the poorest people most heavily, so that wherever possible, they recurred to transactions minimizing the flow of cash. For poor landholders in need of a substantial amount of cash, the anticresis contract was preferable to mortgaging their land, since it freed them from regular interest payments. Creditors, especially peasants and other small landholders, preferred to take over the possession of a land parcel as tangible security until repayment of their loan rather than risking the uncertainty and trouble of collecting interest and principle on a loan. Anticresis contracts, then, reflected the fragility of the altiplano's credit market, the weakness of the legal framework for enforcing contracts, and the high level of material protection against losses from transactions demanded especially by the poorest landholders.

The anticresis contract constituted an alternative to renting or buying land. Renting peasant land, although involving smaller cash transactions, was not a real option, as peasants were reluctant to relinquish control of land in exchange for a small rental fee. For the purpose of providing day-to-day subsistence, it was obviously more economical to work the land by oneself rather than to rely on a rental fee, which had to lie considerably below the gross value of goods produced by the land. Only when peasants needed major sums of money in extraordinary or critical situations, such as family celebrations, funerals, crop failures, or livestock epidemics, were they willing to relinquish control over land. From the perspectives of both

the owner and the person with limited cash resources wishing to acquire land, an anticresis contract could be preferable to an outright sales transaction. The owner could hope to regain control of the land on repayment of the loan. For the person with limited cash wanting to acquire land, anticresis was cheaper than outright purchase and also kept that option open.

The contrast between the use of anticresis contracts and rentals in Azángaro is striking. Land rentals from peasants or members of the intermediate group played only a minimal role among the strategies of hispanized large landholders to gain access to more land. In only 7.3 percent of all rentals did indigenous peasants lease land to hispanized landholders. Rental contracts were overwhelmingly concluded over landholdings belonging to hacendados (50.0 percent) and the church (20.1 percent). In most cases these holdings were rented out to hispanized large landholders. The parishes and Cuzco convents never leased their fincas to peasants, an important fact for evaluating the economic role of the church in the province. Intermediate landholders, by contrast, did have some access to land for rent both from hispanized large landholders and from the church.

Newcomers to the province—mestizo and white traders, transport entrepreneurs, administrators, lawyers, and priests—often lacked sufficient capital to purchase an incorporated estate or had not yet established strong clientalistic links that would enable them to form a new hacienda; these newcomers often began their ventures as renters of small or medium-sized fincas. In one-fifth of all rental contracts the renter was born outside Azángaro province. For example, sometime during the mid-1880s Felipe R. López, a trader born in 1851 on the peninsula of Capachica near Puno town, came to Azángaro with his wife, the Arequipeña Petronila Butiler, and established a small dry-goods store.^[95] By 1889 they were renting the small Finca Upaupani, two miles outside of the provincial capital, from a member of the Paredes clan for a yearly rental fee of 32 soles q.b. (26.66 soles m.n.). At the time Upaupani had no livestock capital.^[96] In 1899 López and his wife rented the small Finca Pumire, probably with livestock capital, for five years for the annual rental fee of 100 soles q.b. (63.37 soles m.n.).^[97] Only in the following year, after more than ten years of residence in Azángaro, did López begin to purchase land.

The majority of renters, however, were well-established landholders. Even the greatest hacendados of the province, such as the Paredes clan or the Lizares Quiñones, found it advantageous to rent additional estates, because, as one landholder stated, it could "advance the productivity [industria ] of their estates."^[98] Renting estancias or fincas adjacent to their own

estate could bring about economies of scale, allowing them to run more sheep per flock or to form more flocks for different categories of sheep so as to obtain better breeding results.

But the number of properties offered for rent did not keep up with expanding demand for land; the number of contracts concluded between hispanized large landholders declined slightly, from thirty-four between 1851 and 1870 to thirty-one between 1871 and 1890 and another thirty-one contracts during the following twenty years. As new estates were being formed and old ones expanded at a rather spectacular rate, the possibility of leasing land from hispanized large landholders became more limited. The interest of hacendados in managing their own estates was increasing.

When a hacendado failed to acquire a landholding by one of the contractual procedures outlined above, he or she could resort to the judicial system. In 1916 José Frisancho, Azángaro's state attorney during much of the second decade of this century, published a scathing indictment of the administration of justice in the province.^[99] It climaxed in the oft-cited assertion that "in spite of having been victim of frequent crimes, in not one case has the Indian achieved justice against some hacendado." In Frisancho's view, Azángaro's judicial personnel, from the judge of first instance to the district justices of the peace, the scribes, the court-appointed experts, and the witnesses, had become subservient to the interests of the large landholders, interests that found their strongest expression in the rapid appropriation of the land of the Indian peasantry.^[100] This development largely explained why in Azángaro "latifundism has reached an extreme degree of preponderance, much more so than in any other Peruvian altiplano province." Frisancho resigned in his struggle against the "forensic denaturalization" in Azángaro and found that any reform effort had to collapse "before the impenetrable resistance of the social environment, zealously guarding its vested interests."^[101] Because of such indictments, it is commonplace to characterize the sierra's judicial system during the first century of republican Peru as serving the class interests of the hacendados, a tool for the exploitation of the Indian peasantry.^[102]

In reality, courts fulfilled a more complex role in serrano provincial society. They functioned as an arena for testing the power of the litigants. The outcome of a case did not necessarily rest on which party had the law on its side but rather on who could bring to bear more influence on the court. Such influence could take the form of better legal preparation, more money to spend on the suit—not necessarily for bribes but for lawyers and appeals to higher courts—and more leverage for concluding quid pro quo deals with court personnel, requiring bargaining chips of value outside the court system.^[103] In a legal contest between a hispanized large landholder

and an Indian peasant, the outcome may indeed have been a foregone conclusion. But often peasants were backed by another hacendado whose client they had become. The suit thus became a contest between two hacendados, and the outcome was by no means certain. Litigation over land between hacendados or between peasants occurred just as frequently as legal battles between a hacendado and peasant.

Viewing the judicial system as an arena for power contests between gamonales and their clients helps to explain why many court cases over land were never concluded. Of every hundred cases over land brought before courts in the department of Puno in 1893, only five were concluded within the year.^[104] Many suits dragged on for years or even generations. Often the litigants decided to abandon the judicial struggle and settle out of court. The economically weaker party had been exhausted by the high cost of litigation and was now willing to accept the terms of the stronger party. In the context of hacienda expansion strategies, such cases constituted the continuation of economic means of land acquisition within the judicial arena.^[105] In some cases even the threat of litigation could be sufficient to force poor landholders to relinquish their claims.^[106]

In the absence of title to a claimed property, an interested party could initiate proceedings of formación de titulos supletorios . Witnesses were heard to verify the legal and undisputed possession of the petitioner, and notices calling for any opposition to the granting of judicial title were published and affixed in public places. Of course, neither the witnesses nor the public call for opposition presented an effective palliative against abuse. Supportive witnesses could easily be found among the claimant's clients and friends, and the peasants whose land was mostly at issue in these proceedings could not read the public notices. In 1903 Adoraida Gallegos initiated proceedings on Estancias Ccaramocco, Chuantira, and Pisacani in Potoni district. She claimed to have inherited these lands from her father, Geronimo Gallegos, but no will existed to prove it. Four witnesses, all Indian peasants from Potoni, confirmed her claim, the justice of the peace had the notices affixed in the small population center for thirty days, and they were printed nineteen times in Puno's newspaper, El Ciudadano , which was not likely to be read by anybody in Potoni, 220 kilometers farther north with only a handful of literate residents. When no opposition was voiced, Gallegos was granted judicial title to the estancias by Puno's judge of first instance, and they became part of her Hacienda Lourdes.^[107]

Judicial procedures such as the formación de titulos supletorios, the queja de desahuicio , and the interdicto de adquirir played a notorious role in the surreptitious appropriation of land by members of the provincial

elite.^[108] In 1913, when the wave of hacienda expansion was cresting, the president of Puno's superior court admitted that the "interdicto de adquirir greatly facilitates the usurpation of land." Not infrequently, the Indian owner of an estancia, unaware of any legal proceedings, "sees a judge approach his cottage accompanied by a person whom he does not know and to whom the judicial personage formally hands over the land that he [the real owner] had inherited from his ancestors."^[109]

The departmental property register is filled with examples from Azángaro, in which such judicial procedures led to the desired end without any public opposition.^[110] Cases in which peasants successfully resisted such procedures were rare, although not unknown.^[111] But more than any concrete judicial procedure, the venality and partiality of the judges, scribes, and witnesses made the use of the judicial system appear promising to persons attempting to gain control of some parcel of land, a problem particularly severe at the district level.^[112]

Last, hacienda expansion could proceed through violent usurpation of neighboring landholdings. José Avila, himself a judge in Azángaro during the 1960s and 1970s, has given us an ideal-typical description of sequences of violent actions aimed at land usurpation:

The grabbing of land from Indians begins with the act of daily placing cattle and mules belonging to the usurper in the pastures and cultivated fields of the Indian. In this the colonos and employees of the latifundista use force, and they proceed to kill the few head of livestock of the Indian for their own consumption. Alternatively they drive the Indian's livestock to the latifundia's central building complex, [kill the animals there], and distribute most of the meat and hides among themselves, while reserving the carcasses of the best-fed animals for the patrón or as gifts to the provincial authorities. . . . In the following days the looting of the Indian's hut begins with the object of weakening his economic situation. This goes on until, under the pressure of this display of force, the owner decides to sign the bill of sale. As sales price they receive a small sum in money or kind according to the whim of the land grabber.^[113]

This account includes different violent actions against landholders that were recurrent themes in legal proceedings as well as in indigenista literature of the time.^[114] Avila places the individual episodes of aggression in the framework of escalating use of force in an overall usurpation strategy, but each of these episodes could occur separately. If peasants yielded to a neighboring estate owner after repeated trampling of their planted fields and consumption of their pastures by the gamonal's livestock, an escalation of force would be unnecessary.

Competing claims over the same parcel of land were common in the altiplano during the nineteenth and early twentieth centuries. Labyrinthine title histories and vague border definitions could justify competing claims on innumerable holdings. Border demarcations usually relied on natural landmarks such as hill crests, creeks, and even trees or shrubs and at best on easily changeable ditches or border stones. Many usurpations could be justified as recovery of land properly belonging to the invader.^[115] Even complete knowledge of all relevant title deeds does not always make it clear which landholder held the legally more valid claim since titles themselves could overlap. Violent usurpations occurred not only in the context of hacienda encroachments onto peasant lands but also in interactions among hacendados as well as among peasants.^[116] Display of force was the other side of Azángaro's clientalistic society. It permeated relations between members of all social groups in the altiplano, although peasants became victims most frequently.

It is difficult to estimate how important violent usurpations of peasant lands were for the overall process of hacienda expansion in Azángaro between the 1860s and 1920. Although petty harassment, such as introducing livestock onto the pastures of a neighboring landholder, occurred regularly, the use of more serious levels of force, such as destruction of peasant residences, never lost its character as an extraordinary event. Nevertheless, the huge wave of land purchases by hispanized large landholders between the 1890s and 1920 led to endemic open violence in Azángaro's countryside, an expression of the critical impasse that the altiplano's neocolonial society was approaching.

Hacienda expansion strategies were further differentiated by a number of economic variables, including size of the parcels acquired, capital available for hacienda operations, and the dispersion or concentration of parcels acquired. Land purchases by relatively poor, marginal large landholders differed considerably from those by the wealthiest and most powerful families in the province.^[117] Itinerant traders, small retailers in the district capitals, small-time muleteers, pettifoggers, and other mistis (members of the local elite, usually of mestizo background) desperate to distance themselves from the ranks of the Indian peasantry commanded exiguous funds sufficient only for purchasing a few small parcels with which to enlarge their small fincas, establish themselves as marginal finca owners, or enlarge their scattered estancia holdings.

For example, Cesar Salas Flores, an orphan born in the capital of the neighboring province of Lampa, came to Azángaro as an unskilled worker with a band of muleteers sometime during the late 1880s or early 1890s.^[118] He married Isabel Mango, a descendant of an impoverished kuraka family who brought only a little property into the marriage. Salas himself had no

capital and did not establish any trade or transport business in Azángaro;^[119] rather, his main occupation consisted in working as a tinterillo , or pettifogger, and acquiring small peasant properties in the process. By 1909 Salas had agglomerated six to eight peasant estancias around the two largest properties. He had acquired one of the two central properties, Vilquecunca, located five kilometers northwest of Azángaro in the plains of the Río Grande close to the road to Asillo, in December 1892 for 520 soles q.b. (433 soles m.n.) from the peasants José María and Juan Quispe.^[120] The newly forming finca lay within the territory of the parcialidades Jallapise and Urinsaya and stretched—as did many estates—from the banks of the river to the hillsides of Cerro Mamanire, the northeastern edge of the gently sloping massif separating the Azángaro and Pucará rivers. Although Vilquecunca was referred to as a "finca" after 1899, effective estate operations took longer to establish. In 1900 Salas gave Vilquecunca and its seven or eight cabañas, the agglomerated estancias, in anticresis to Norberto Vásquez, another newcomer trader busy forming estates, for a credit of 240 soles m.n., obviously without livestock capital.^[121] In late June 1907 he reclaimed Vilquecunca from Vásquez, only to conclude another anticresis contract with him two and one-half months later. This time, however, Salas gave Vásquez only one of the finca's constituent estancias, Cangallo Llinquipata Quilinquilini, for a credit of 200 soles m.n.^[122] Salas apparently now needed the major part of the finca for himself, and we may assume that it had taken him that long to build up a livestock capital commensurate with the pasturelands of Vilquecunca. To operate the estate himself without a minimally sufficient livestock capital would have been both uneconomical and dangerous, as estates with no or too little stock invited usurpations by neighbors. Just one year before resuming direct control over most of Vilquecunca, Salas gave the second finca that he was in the process of constituting, Anccosa, about fifteen kilometers west of Azángaro in Asillo district, in anticresis, also without livestock capital.^[123] He was perhaps pooling all of his livestock capital to use the larger finca's carrying capacity to the highest degree possible. The formation of effectively operating fincas, then, merely began with the agglomeration of a sufficient land base.

Such nominal new estates went through a transitional phase in which their operation differed little from their situation as independent peasant estancias. The former peasant owners, obligated to stay on as colonos of the new estate, now had to plant a few masas (a traditional land measure equal to 760 square meters) of potatoes or quinua for the patrón and render transport and domestic services. But lacking the requisite capital, the marginal estate owner may have been unable to make full use of the labor

and pasture resources for quite some time. This process contrasts with the formation of haciendas by some of the wealthiest families in the province. In the cases of haciendas San José and Sollocota, the owners counted on a preexisting core of one or more older estates and had sufficient capital to stock the expanding pasturelands with additional sheep and cattle. The transition from loosely agglomerated peasant lands to well-integrated, fully stocked estate must have been briefer here.

Marginal large landholders such as César Salas had just sufficient means and influence to build an estate within a narrowly circumscribed area, perhaps only within one parcialidad. But most of the wealthiest hacendado families acquired land throughout the province, in some cases even in several provinces. When José Angelino Lizares was born in the mid-1860s, his family had already achieved affluence and power in Azángaro, since both his father, José María Lizares, and his grandparents, Francisco Lizares and Josefa Quiñones, had persistently increased the family's landholdings.^[124] When José María Lizares Quiñones passed the administration of his estates to his sons José Angelino and Francisco in 1895, the family owned seven fincas in Muñani and Azángaro (Muñani-Chico, Arcopunco, Calla-tomasa, Ticani, Tintire, Quichusa, and Cayacayani) and several smaller estancias in Azángaro and Santiago de Pupuja, with a total livestock capital of 28,000 OMR.^[125]

José Angelino continued the acquisition of landholdings in Santiago, Azángaro, and Muñani and expanded further into San Antón, Arapa, and Chupa. His landholdings now were scattered over some six thousand square kilometers of territory. He succeeded in forming new estates in widely dispersed parts of the province because he not only had sufficient monetary resources but could also call on numerous social ties extending beyond the provincial capital and the family bailiwick in Muñani. These ties derived in part from family connections and in part from broader client networks. His purchase of land in San Antón followed his marriage to Leonor González Terrazas, daughter of the judge of first instance, Federico González Figueroa, who had been putting together Finca Cangalli in San Antón since around 1900. Given as a dowry to Lizares Quiñones's wife, Cangalli became the core of Lizares Quiñones's new Hacienda Esmeralda.^[126] In Chupa, Lizares Quiñones's formation of new estates against the determined resistance of community peasants proceeded with the support of one of the most influential local families, the Salas. Both Lizares Quiñones and Nicomedes Salas had received military titles during the Cacerist administrations between 1886 and 1895.^[127]

Lizares Quiñones's provincewide influence owed much to his political career, which he began as mayor of Azángaro's provincial council during

the early 1890s. He held the congressional representation of the province for many years between 1908 and 1929, followed by a short term as senator for Puno until the overthrow of the Leguía regime in 1930.^[128] These positions allowed Lizares to ingratiate himself with some of the notables of the districts and to influence the distribution of administrative positions on the provincial and district levels to serve his interest. By developing this sociopolitical "infrastructure" throughout the province, he enhanced his capacity to induce local landholders to sell him land and to apply pressure in case they refused.

The majority of large landholders in Azángaro followed Lizares Quiñones's strategy of building up dispersed estates throughout the province, but a few affluent land purchasers whose biographies suggest economic and political resources similar to those of Lizares chose not to form dispersed estates. Some of the most prominent were Arias Echenique and Urquiaga (discussed above) and José Albino Ruiz, the owner of large Hacienda Checayani in Muñani district. They concentrated on developing single, very large estates, whereas most of Lizares Quiñones's fincas, like those of other hispanized landholders owning dispersed properties, such as the Paredes clan, González Figueroa, and the Rossellos, were merely medium-sized. Specific local circumstances may have influenced such divergent patterns: Lizares Quiñones's Hacienda Muñani-Chico, for instance, was totally surrounded by other estates, precluding further expansion. But they may also reflect the first appearance of somewhat divergent goals. For a minority of hacendados, the formation of efficient livestock haciendas took on growing importance, and they emphasized maximum size. Yet most hacendados expanded their landholdings primarily in order to enhance their power and influence in the province. Here a broad geographical spread of the estates could only be advantageous. In other words, Lizares Quiñones's acquisition and expansion of fincas in six different districts not only built on his already considerable sociopolitical power but served to increase and fortify it.

For the department of Huancavelica in the central sierra, Henri Favre has noted that the process of hacienda expansion between the end of the War of the Pacific and 1919 proceeded in three different ways: (1) expansion of existing estates through incorporation of surrounding landholdings, (2) constitution of totally new haciendas by purchases of a large number of small properties, and (3) reconstitution of colonial haciendas that had splintered since the late colonial period.^[129] The first two forms of expansion also characterized the landholding development in Azángaro. Reconstitution of splintered colonial estates, however, occurred only rarely in the altiplano. Not many haciendas had splintered by the late nineteenth cen-

tury. The few reconstitutions in Azángaro province took place in Putina and Muñani, districts whose livestock economy had been closely linked to mining enterprises during the colonial period.^[130] The more pronounced discontinuity of land tenure patterns of some estates in Putina and Muñani, as in Huancavelica, was perhaps a consequence of the integration of mining enterprises with livestock haciendas. Once the mining operations had decayed, some associated haciendas also decayed, and the impoverished owners were unable to prevent their physical disintegration.

But the expansion of Azángaro's estate sector was a much more complex process than suggested by Favre's classification. Variations in the mode of land transactions, the mean value of acquired landholdings, the concentration or dispersion of properties, and the number of sellers per transaction gave each expansion project a different social and economic significance. This complexity contributed to the emergence of a landholding pattern by the second decade of this century that was far from a uniform landscape of large estates.

Inheritance and Sale: The Stability of Landholding Families

Fifty years ago Emilio Romero, the distinguished Peruvian geographer who was born and raised in Puno, wrote that "toward the end of the nineteenth century . . . property in the sierra became somewhat divided."^[131] For Azángaro, Romero's assertion cannot be maintained in this general fashion. Although parcelization began to affect peasant estancias and properties of other small landholders, haciendas survived the crisis of inheritance surprisingly intact. Hacendado families, fully aware of the dangers that any splintering of the family estate entailed, pursued elaborate strategies for countering the centrifugal tendencies of inheritance. But more than any of these strategies, it was the expansionary environment in trade and livestock raising that forestalled the atomization of family estates before 1920.

Property was transferred from one generation to the next in three ways: (1) owners passed on property before their death; (2) wills specified how goods were to be distributed by the executor after the testator's death; (3) the goods of a person dying intestate were distributed strictly in accord with legal prescriptions governing inheritance. The first form of property transfer was known as antícipo de legítimo . Used when the owner could not properly take care of his estate, mostly because of old age, such transfers commonly took the form of a donation. But at times parents also sold landholdings to their children. This form of generational transfer offered the greatest degree of discretion to the owner, who could favor a preferred

heir without coming into conflict with Peru's inheritance laws. Moreover, distributing the estate during the lifetime of the testator allowed continued exercise of parental authority to quell discord between heirs, who often had to promise "to conserve the good family relations."^[132]

The generational transfer of property through wills imposed greater strictures on the testator. Peruvian inheritance law, inscribed in the civil code of 1852, was based on the two Spanish principles of equal inheritance and bilateralism. By law and practice the testator's direct descendants, his or her children and their offspring, enjoyed precedence as heirs. Property that husband and wife brought into matrimony (bienes raices ) always remained the separate estate of each spouse. A married testator could dispose only of the goods that he or she brought into the marriage and 50 percent of the property that the couple had accumulated during matrimony (gananciales ), with the other 50 percent automatically belonging to the spouse.

Given the rapid expansion of many family holdings, the weight of gananciales vis-à-vis bienes raices could be great, and the surviving spouse's share of the total estate might approach 50 percent. Furthermore, it frequently was the wife who brought in the bulk of bienes raices when marrying. Although Peruvian law treated wives as minors who needed the consent of their husbands for legal transactions, after her husband's death the widow often found herself as owner with full control of the major part of the family estate.^[133] A widow thus might come to hold the controlling shares of large estate complexes, particularly when she was considerably younger than her husband (as in second marriages), when the couple had had no children, or when the husband had died young and the children were still minors. In many cases widows were declared executors of their husbands' estates and legal guardians of their children, controlling their property until they came of age. Juana Manuela Choquehuanca, for example, controlled the parental Hacienda Picotani in the cordillera above Muñani for some twenty years after the death of her husband, Mariano Paredes, in the mid-1870s.^[134] Carmen Piérola, a native of Bolivia, managed to be married to three sons of Juan Paredes in a row between the late 1870s and 1904, surviving them all and in the process managing—and expanding—a large part of the clan's landed estates.^[135]

The legal minority of wives and Azángaro's social norms assigning a domestic role to women in hacendado families kept them from managing estates during their married lives. After the husband's death, the widow had to face unaccustomed tasks made difficult by a society in which the exertion of authority was tied to the threat or application of force. Not surprisingly, widows frequently entrusted the management of the family

estate to a friend or relative, left more decisions to the finca's administrator, or even rented it out. All of these options risked the estate's deterioration through neglect or willful overexploitation.^[136]

Family estates in more than one case entered a critical phase when they passed into the control of a widow or a young, unmarried daughter of the deceased head of the family. In contrast to women from peasant and petty shopkeeper families, who were often used to economic decision making, women from the landholding elite had been socialized in such a way as not to be concerned with economic matters.^[137] In several cases widows or young single heiresses could not hold on to family estates.^[138] Contemporary society saw women as unfit to manage livestock estates. The "incapacity in which [female heads of households] find themselves to improve their lot" condemned their families to the fate of impoverishment, in the view of a priest in Saman.^[139] But certainly there were women in Azángaro who had great success as landholders or traders.^[140] Adoraida Gallegos, never having married, formed one of the province's greatest estate complexes between the 1880s and 1920. It is perhaps more than coincidence that she has become a legendary figure in the province, depicted as a fearless and haughty woman, riding around the frontiers of her estate with pistols on her hips, not hesitating to whip any neighbor who tried to appropriate her lands; she was, in short, a woman who acted "as a man."^[141] Her very legend seems to confirm how exceptional such independent women were among the province's hacendado families.

Although Spanish and Peruvian law prescribed equal inheritance among all legitimate children, the testator had the possibility of improving the share of any child through a mejora , usually one-fifth or one-third of all property. Asunción Lavrín and Edith Couturier have suggested that in colonial Mexico mejoras served "to buttress the family's social position and prevent the deterioration of its economic status" or, in other words, to favor the heir who promised to hold the family estate together most efficiently.^[142] In Azángaro the stated purpose of mejoras was to reward a child who had helped the testator during his or her old age with particular dedication or to secure the material situation of the heir facing the greatest economic uncertainties, perhaps a daughter who had married poorly or would probably stay single.^[143] But mejoras appeared seldom in wills of Azángarino landholders.

Family strategic considerations may have played a larger role in the testator's disposition as to quality and type of property rather than its quantity. The heir whom the testator expected to be most capable of preserving and enhancing the family property and social position tended to receive the centerpiece of the landholdings, the best-established, largest,

and most lucrative hacienda. In the settlement concluded by José María Lizares with his estranged wife Dominga Alarcón in 1905, a settlement that was to be his final will, he specified that, whatever the disposition of the remaining family fincas, his son José Angelino should receive for his "loyal services" Haciendas Muñani Chico and Nequeneque, the largest and best established of the family's estates. Notwithstanding an equitable distribution of the family estate in terms of quantity or monetary value among all heirs, the most trusted son was to receive the family's best haciendas.^[144]

Following Spanish legal tradition, most testators attempted an equitable distribution of their property among their legitimate children, regardless of sex or birth order. The goal of providing each heir with equal material belongings took precedence over the goal of assuring the maintenance of the economic and social position of the family through accumulation of property in the hands of one heir. Although contradictory on the surface, these goals in fact were complementary. Testators knew full well that the single greatest danger to the maintenance of the family's economic and social position consisted in interminable legal squabbles. The need to sell land for defraying the costs of litigation damaged the fortunes not only of peasant families but of hacienda owners as well, as in the case of the Choquehuanca family (discussed below). Avoiding costly litigation constituted a precondition for preserving the social and economic position of the family; an equitable distribution of the family estate among all heirs could best assure this goal.^[145]

If someone died intestate, the estate had to be divided according to legal norms between the heirs-at-law. The surviving spouse would receive his or her gananciales, one-fifth of the estate went to the recognized illegitimate children, and the rest would be divided in equal parts among the legitimate children. No mejoras or legacies could be granted, and normally none of the heirs-at-law could be disinherited. The division of the estate could then be carried out either through an out-of-court settlement within the family or by a legal procedure involving its judicial inventory, appraisal, and subsequent division. If the proceedings promised to be protracted, the estate was placed in judicial deposit. The court-appointed trustee was charged with insuring the integrity and continued revenue generation of the properties, which he often chose to lease out.^[146]

On the surface these formal aspects of inheritance suggest that the generational transfer of estates led to dispersion of family landholding complexes, but to what degree this dispersal actually occurred depended on two factors, the number of heirs and the manner in which they chose to operate the estate. The number of children varied widely in hacendado families, from none to eighteen, including illegitimate children. In twenty-

nine wills of Azángarino hispanized large landholders between 1854 and 1909, the mean number of children of the testator, from all marriages and extramarital relations, was 4.89. But on average only 2.17 children were still alive at the time of the father's or mother's last will. The figures were higher for the eight testators between 1854 and 1878 (an average of 8.125 children, with 3.75 surviving at the time of the parental will) than it was for the twenty-one cases between 1892 and 1909 (an average of 3.6 children, with 1.57 surviving at the time of the parental will).^[147] The number of children surviving their parents on average was rather low. To be sure, in a few important hacendado families four, five, or even eight children had to share the parental estate. The large estate of Juan Paredes just sufficed to bequeath one finca to each of his six surviving legitimate children, leave Finca Lacconi to his two illegitimate sons, and set aside the small Finca San Juan de Dios for a beneficent institution.^[148] But in many cases the family estate had to be distributed among only two heirs or passed in its entirety to a single surviving son or daughter.

Following the appeal of so many parents not to divide the family fincas, heirs frequently sought to operate the estates jointly.^[149] In September 1893, for example, the four surviving children of Mariano Solorzano and Augustina Terroba, traders from Putina who had shifted their operation to the capital of the department, formally established a company to manage the undivided estate left by their parents. The capital of the company consisted of the family's real estate, including Haciendas Collpani and Loquicolla Chico in Putina and two large houses in Puno town, as well as livestock capital, household furnishings, and business credits. The object of the company was to "administer, improve, and increase the property that constitutes its capital" and to "dedicate itself to the customary businesses of the family, such as trade in wool, planting of potatoes, barley, and quinua, slaughter, trade in alcohol, and other transactions up to the value of 1,000 soles monthly."^[150] The management of the company was to rotate between the four fraternal partners on a quarterly basis. The managing partner was to receive 55 percent of profits or losses on new business deals concluded under his management. Fifty soles monthly were to be distributed to each partner from company earnings for subsistence, and remaining profits would be reinvested. Company real estate could not be sold, and proceeds from the sale of other company property were to be reinvested into the purchase of further real estate.

The company did not last long. Ten months after its founding it was dissolved and the property divided among the five heirs, including the widow of the deceased brother Mariano Casto Solorzano. Each heir received a portion worth about 13,000 soles m.n. To effect an equitable

division, three fincas were carved out of the two existing estates. The portion with the greater part of Hacienda Collpani fell to Adrian, and that containing most of Hacienda Loquicolla Chico's lands to Julio. The newly created Finca Pampa Grande, mostly on the lands of Loquicolla Chico, fell to Emilia Toro, the widow of Mariano Casto Solorzano, in representation of her daughter Natividad's rights. The two remaining heirs, Maria Manuela and Natalia, received the houses in Puno and several small chacras located in the outskirts of that town.^[151]

In subsequent years only two of the five brothers and sisters continued the same type of economic operation the parents had exercised, ranching and trade. The other three heirs or their descendants shifted the source of their livelihood to Puno town. Adrian Solorzano, who had received the greater part of Hacienda Collpani, studied law in Lima and Arequipa and in July 1900 was accredited as lawyer in Puno.^[152] He took out loans on Collpani, leased it to other family members, and in December 1905 consented to sell the estate for 10,500 soles m.n. without livestock to his brother Julio.^[153] He effectively withdrew from landholding and trade and dedicated himself to work as lawyer, teacher, and journalist in Puno, becoming a prominent and trusted public figure in the capital of the department.^[154]

Between 1896 or 1897 and 1904, Finca Pampa Grande, property of the deceased Mariano Casto Solorzano's minor daughter Natividad, was administered by Alberto Gadea, second husband of Natividad's mother Emilia Toro and director of Puno's Colégio Nacional San Carlos during the decade following Nicolás Piérola's civil war victory.^[155] After Emilia Toro's death sometime before April 1904, Pampa Grande was leased jointly to Natalia Solorzano de Zaa and her brother Julio, who was still renting the finca during the 1920s. Natividad and her maternal family became urban rentiers.^[156]

María Manuela Solorzano had married the Puno notary San Martín even before the establishment of the company in 1893. In the subsequent division of the paternal estate she received one of the large houses in Puno town. In the notarial registers she does not appear again as a landholder in Azángaro, and we may assume that she and her family lived off the income from her husband's business and any property he may have owned.

Natalia Solorzano had received the other urban property in the 1894 division, but she and her husband, Arturo Zaa, remained active traders in Azángaro province. By 1900 the couple was again acquiring landholdings by lease or anticresis in the same area where the family haciendas were located. In 1913 Zaa and his wife acquired Hacienda Loquicolla Grande, adjacent to the Solorzano family estate Loquicolla Chico, from the Puno diocese in exchange for urban property.^[157] Natalia Solorzano and her husband thus were reestablishing their position as large landholders and traders.

Julio Solorzano, the fifth heir who had received most of Hacienda Loquicolla Chico in the property division, was soon able to expand his landholdings in Putina through leases and purchases from family members as well as other landholders.^[158] After the dissolution of the family company, and probably taking up old family business connections, he established himself in trade selling wool to Arequipa export houses such as Stafford and Ricketts.^[159] By 1902 he owned the most important dry-goods store in Putina,^[160] and within a decade his position compared favorably with that of his parents in spite of the estate's division.

Thus, of the five heirs to the estate of Mariano Solorzano and Augustina Terroba, only two continued to own livestock haciendas and trade in a wide range of goods. Some ten years after the division of 1894 the rural landholdings were reconcentrated in the hands of one heir through Julio's purchase of Collpani and his long-term lease of Pampa Grande. Two heirs, Julio and Natalia, achieved an expansion of family holdings into adjacent estates and estancias. Heirs who relinquished control over their portion of the family landholdings preferred family members in their sales, as stipulated in many divisions of estates.^[161] In the case of the Solorzano family, the three heirs who by choice or circumstances lost control over family estates did not suffer social descent. Through marriage or occupation they managed to transfer their social status to an urban base.

The majority of Azángaro's multi-estate complexes underwent a similar partition shortly after the death of the patriarch. These partitions rarely led to a long-term splintering of family estates prior to 1920. This was a period during which old haciendas expanded rapidly and new ones were formed. The paternal social status and client networks could usually be carried over to the next generation. This situation helped heirs in expanding their portion of the family estate through purchase or lease of adjoining landholdings, just as Julio Solorzano was doing. In a recurring pattern, some heirs withdrew from landholding altogether, while one or two heirs aggressively participated in the wave of expansions and new formations of estates. Children or grandchildren of the founders who had established or consolidated the family estate between the 1840s and early 1870s at times purchased as much or more land than their parents or grandparents had. Besides the Lizares family, we may cite the descendants of Juan Paredes, whose granddaughter Sabina, together with her husband Carlos A. Sarmiento, belonged to the five greatest land purchasers between the 1850s and 1920.^[162]

Heirs tried to insure the territorial integrity of a single parental finca by any of three techniques: (1) operating it jointly (pro indiviso ); (2) rotating the lease of the estate to each of the heirs while the other heirs drew rent corresponding to their share of the estate; or (3) leasing the whole

estate to a third person and dividing the rent among all heirs. Joint operation of an estate tended to be adopted particularly by heirs of rather poor families who had always lived on the finca itself or in the district capital and had no other source of income. But these arrangements were fragile and did not resolve the principal problem of having a growing number of family members depend on the income from one estate. In most instances where such schemes were applied to medium-sized or large estates, they did not last longer than ten years. Heirs setting up a business, building a new house in town, or having some other urgent need for cash became interested in selling their shares. Again and again one heir succeeded in reuniting all shares of the estate, thus ending the danger of property splintering.^[163]

A fascinating example concerns Hacienda Checayani in Muñani, where the estate was reunited only in the third generation. The hacienda had belonged to the Choquehuanca family during the late eighteenth century. By 1892 it had become divided among seven grandchildren of the last owner of the whole estate, Mariano Riquelme, who had acquired it sometime before 1844. By 1906, after fifteen years of intrafamily squabbles, a contest between two in-laws for control of the hacienda, a momentary decision to relinquish family control over it altogether, and more than twenty-five notarial contracts between family members, Natalia Riquelme and her husband, José Albino Ruiz, again were sole owners of Checayani, a condition last enjoyed more than fifty years earlier by her grandfather.^[164] When Natalia Riquelme dictated her fifth will shortly before her death in childbed in 1908, she reflected on the tortuous property history of Checayani and admonished her heirs not to give up what had been reached:

I declare . . . so that my children may know that the whole [title] documentation of Finca Checayani until its definite registration and the judicial possession has been a very difficult undertaking which cost many thousand soles and innumerable privations. All of this is owed to my husband, don José Albino Ruiz, to whom my children should be eternally grateful. Thus I recommend to them and their heirs never to sell Checayani to an outside person nor to divide it . . . in order to conserve the integrity of the finca and so that this estate which was acquired by my grandfather Mariano Riquelme and his wife Juliana Aragón may not disappear.^[165]

In the 1980s the core part of Checayani, which was not affected by the agrarian reform during the early 1970s, was still owned and operated by Natalia's grandson, Martín Humfredo Macedo Ruiz, a fifth-generation owner from the same family. Indeed, none of Azángaro's large estates and relatively few mid-sized estates became divided through inheritance prior to 1920.^[166]

The outstanding exception to this stability of family estate complexes concerned the Choquehuanca family. Legal quarrels occupied the numerous descendants of Cacique Diego Choquehuanca for a whole century following his death in 1796. Of the eleven estates that the family owned in Azángaro province at the time of the Túpac Amaru Rebellion, only five remained family property by the 1840s: Haciendas Catacora, Ccalla, and Puscallani in Azángaro district and Haciendas Picotani and Nequeneque in Muñani. By 1910 the family had definitely lost three of these (Catacora, Puscallani, and Picotani) and held on only to shares of the other two estates.^[167] Each of four lines of Choquehuanca descendants with some claim to family property fought legal battles about all of these five estates against all other members of the family. Several of the cases were decided by the Supreme Court in Lima only after decades of litigation. Every generation brought new dissensions between heirs.^[168] The endless and costly lawsuits must be viewed as the major cause of the ultimate loss of most Choquehuanca landholdings and the ensuing impoverishment of most family members. Successive generations of Choquehancas took out large loans to defray court costs, promised their lawyers shares of estates in lieu of honoraria, and proceeded to sell portions of or complete estates, all to defray fees for lawyers and the courts.^[169] By the first decade of the present century the family had lost its preeminence among Azángaro's landholders.^[170] The Choquehuancas, who in 1780 had been the greatest landowners in the province, some 120 years later had become marginalized.

Intrafamily litigation about land was common among Azángaro's hacendado families, but the Choquehuancas hold the record for duration, frequency, and ubiquity of such fights, with about every branch of the family at some point suing every other line. To a degree, the lack of family cohesion, which produced such devastating effects, must be viewed as the consequence of losing the office of cacique. As long as one family member held the cacicazgo of Azángaro's parcialidad Anansaya, the Choquehuanca estate had survived intact.^[171] The loss of the cacicazgo appears to have brought with it the atrophy of the intrafamilial structure of authority: no heir could claim any primacy over his or her brothers and sisters or cousins. As the Choquehuancas lost the economic benefits associated with the office of cacicazgo, such as the cacique's salary and the opportunity to exploit the Indian peasants, a chaotic scramble for the family's resources commenced.^[172] What at first glance appears as the peculiar decline of a single family of large landholders may have been symptomatic for the decline of a social group, the affluent elite of colonial cacique families descended mostly from the prehispanic Andean nobility.^[173]

Estates faced acute dangers of atomization among poor hacendado families when none of the heirs succeeded in acquiring other property or

establishing themselves in some trade or business of equal prestige and income-earning potential. During a period when commercial opportunities blossomed and it was fairly easy to constitute new estates, this problem did not arise often. Not surprisingly, the few cases of long-term atomization of Azángaro estates had originated during the economically difficult period before the 1850s.

Fincas underwent a process of splintering in two different ways. The parcels of all heirs could be divided, resulting in independent estancias of diminishing size. This had been the fate of Finca Nuestra Señora de las Nieves de Chocallaca in Putina, last owned and operated as a whole by one Juan Ortíz in the late eighteenth century. One hundred years later, grandchildren, great-grandchildren, and great-great-grandchildren of Ortíz all owned independent estancias or shares of such, connected by nothing but family ties among the owners and their memory of an "ancient and extinguished finca."^[174]

In the other mode of estate splintering, heirs never formally divided the estate and continued to own and operate it pro indiviso. Despite the fragility of this construction, under certain conditions a finca remained common property of all heirs even over several generations. Although the integrity of the estate was preserved, in the long run its internal structure changed profoundly. In effect, the growing number of co-owners lost the characteristics of hacendados and gradually approached the status of members of a peasant community. This process occurred in Finca Carasupo Chico in Muñani. The estate had belonged to one Diego Vargas probably before the end of the colonial period. Around 1900 some twenty-five heirs—at least great-grandchildren of the common ancestor—lived on the estate with their families. Although all heirs continued to own the property jointly, each family individually worked a small segment, a cabaña. In negotiations with outsiders—for example, with neighboring landholders about border disputes—they acted as a group representing the whole estate. By 1900 the status of Carasupo Chico was becoming confused. Sometimes still referred to as a finca, in other documents it began to be called a parcialidad or comunidad. In an 1897 property tax register the owners of Carasupo Chico appeared as follows: "José S. Endara, Mariano Arizales, Felipe Serna, and the other comunarios ." In July 1907 all the co-owners leased Carasupo Chico to a neighboring hacendado, Federico Gonzales Figueroa. Because they lived on the estate, we may assume that they became the tenant's shepherds. Most of the generation of owners in possession of Carasupo Chico around 1900 no longer spoke Spanish. Although some worked as rural weavers or cobblers, others were referred to as agricultores , an occupational label nearly always applied to peasants in

the notarial contracts.^[175] At the end of a process that spanned at least three generations, the heirs of hacendado Diego Vargas were becoming peasants.

The majority of families managed to hold on to their haciendas over two or three generations. Probably two factors contributed more to this stability than anything else: the replacement rate remained, on the average, relatively low during this period, as child mortality was still quite high, and the expansionary tendencies of southern Peru's wool export economy between the late 1850s and 1920 created income-earning alternatives in the regional economy for "excess heirs." Such heirs could form new estates, establish themselves in trade, or seek administrative positions. Expansion thus helped to keep down the burden of supporting growing numbers of descendants who encumbered family estates. Although the passage of landed property from one generation to the next often entailed a crisis, family estates were generally kept together, later to expand through reuniting most portions in the hands of one or two heirs and the acquisition of further landholdings.

The stability of estate ownership is confirmed by a sample of thirty-two fincas and haciendas with substantially complete information on title transfers between the 1850s and 1920 (tables 6.6, 6.7). On average there was nearly one sales transaction per estate during the seventy-year period from 1851 to 1920, with more than one-third (37.5 percent) not entering the market at all. In order to measure the continuity of estate possession by Azángaro's hacendado families, we need to exclude sales transactions within families. Half of the thirty-two estates underwent no extrafamily sales transaction between 1851 and 1920. Thirteen estates were sold once

outside the families of the owner, but only three passed to different families two or three times.

These figures indicate a remarkable degree of stability of landholding in Azángaro's estate sector. For a comparable seventy-year period, from 1690 to 1760, seven of eleven haciendas in the area of Huancavelica were sold more than three times, and only one remained property of the same family throughout.^[176] But the overall stability of landholding in Azángaro's estate sector masks rather significant differences between large haciendas and small and medium-sized estates: although seven of the ten large estates in this sample never were sold outside the family between the 1850s and 1920, the same held true for only nine of the twenty-two other estates, that is, just over 40 percent as compared with 70 percent for large haciendas. At the same time, only a small fraction of the small and medium-sized fincas were ever transferred by sale within the same family, whereas such sales occurred with half of the large haciendas. Large estates were sold much less often than were small and medium-sized fincas and were more likely to be purchased by another family member, perhaps an heir attempting to reunite all the shares. Small and medium-sized fincas showed a higher propensity to be sold and were less likely to be purchased by other family members. Put differently, heirs to shares of small or medium-sized fincas found it more difficult to reunite the parental property than did families owning large haciendas. This confirms that the generational transfer of

landholdings presented a more critical situation for poor hacendados than it did for Azángaro's landholding elite.

The Lands of the Church

In spite of repeated attempts by liberals and military caudillos to expropriate church estates in Azángaro, they survived nearly intact until the early twentieth century.^[177] None of the estates was operated directly by church organizations. Except for two or three estates—Purina and Posoconi in Asillo and Potoni in the district of the same name—church fincas were relatively small, with a mean size slightly below the average for all estates in the province. With the establishment of new haciendas, the church's share of estates declined from close to a third in the early 1830s to between 10 and 15 percent by the second decade of this century.

Some church estates were rented by newcomers as an affordable first step toward becoming hacendados, but most were rented or held in emphyteusis by established families of large landholders from the province. Often a hacendado leased a church estate immediately adjacent to a finca of his or her own and integrated both into one livestock operation. In the case of Hacienda Posoconi in Asillo, around 1840 Coronel Rufino Macedo sold the emphyteutic rights, acquired in 1829 for 150 years, to José Mariano Escobedo, the Azangarino merchant and politician residing in Arequipa. Posoconi, with 1,568 hectares, had a livestock capital of 4,080 head of sheep and brought rental fees of 400 pesos annually for the church of Asillo.^[178] Because the estate had no adequate water supply, it made eminent sense to integrate it with some neighboring landholdings. Since 1805, through a censo , Escobedo's family had been in possession of the adjacent Finca Payamarca, property of the Indian community of Asillo, and, since the early 1830s, of the Sociedad de Beneficencia Pública in Puno.^[179] Between the mid-1830s and 1857 Escobedo bought another three substantial landholdings, all bordering on Payamarca and Posoconi. Although the church continued to hold the property title to Posoconi and its livestock capital, the whole complex by the late 1850s was operated as one single estate with the considerable livestock capital of 16,600 sheep. Escobedo bequeathed all these landholdings to his illegitimate daughter Teresa O'Phelan,^[180] whose husband, Arequipeño Manuel Velando, purchased further adjacent peasants estancias and one finca during the 1870s. Another adjoining finca, Rosaspata, was added to the complex during the early 1910s.^[181]

By the 1910s Posoconi had become the center of a sprawling private estate complex of some 4,673 hectares, which surrounded it on all sides.

In 1915 the Velando O'Phelans finally consolidated the property title to the hacienda when the church sold Posoconi in fee simple to the family for 4,800 soles m.n., roughly one-fourth of its appraised value.^[182] Unable to repay massive debts contracted during the boom years of World War I, the Velando O'Phelans in 1923 had to sell Posoconi, with more than 36,338 head of sheep, for 13,500 libras peruanas (135,000 soles m.n.) to the Arequipeño merchant house Enrique W. Gibson. In 1926 the estate became part of the newly founded Sociedad Ganadera del Sur.^[183]

Throughout the first century after independence, liberal critics continued to blame the church's mortmain property for its retrogressive influence on Peruvian agriculture, much as Choquehuanca had done in 1830. Failing to see the effect of changed commodity markets, the French agriculturalist J. B. Martinet erroneously blamed mortmain holdings for increased land prices and rental rates during the 1860s.^[184] As late as 1930 Julio Delgado, a Cuzqueño social scientist, leveled the same charge against the church properties as Choquehuanca had done one hundred years earlier, claiming that in short-term leases of church fincas their productivity declined since "the tenant does not concern himself with improvements, but only with drawing the greatest possible gain by extracting the maximum from the land."^[185]

For the approximately twenty church haciendas in Azángaro that were actually operated under short-term leases, Delgado's charge is correct. But the overall importance of these estates was limited, as they were mostly small, comprising a few hundred hectares and with a livestock capacity rarely exceeding 1,000 sheep. Tenants often overexploited these fincas during their five- to nine-year leases. Again and again the estates had less livestock at the end of the lease than at the beginning.^[186] For the parishes in Azángaro and the Puno diocese that shared the income from the estates, the recurring losses of capital translated into a diminution of their revenue. Because of undercapitalization and the insecurity of their borders, these church estates regularly were rented out at lower rates than were privately owned haciendas, and the church saw itself forced to cede part of the annual rental fee to the tenant for restocking. Hacienda Ocra in Muñani lost three-fourths of its livestock capital between 1870 and 1890 during its lease to Luis Paredes; the church had to accept a corresponding decline in its effective lease fee when it handed the estate over to a new tenant, José Angelino Lizares Quiñones, in 1890.^[187] But low rental rates were not limited to the church estates leased for short terms. As annual fees for emphyteutic haciendas were fixed for the 150 years' duration of the contract, after about 1860 they lay considerably below rising lease fees paid for private estates.

The church undertook only feeble efforts to recover losses from over-exploitation of its estates. Rather than itself taking legal steps against the responsible tenant, it merely obligated the following tenant to pursue the recovery of embezzled livestock. Judging from the chronic undercapitalization of many small church fincas, these endeavors generally met with little success. It proved difficult for the church to press charges against former tenants, who blamed the shepherds for the decline of the estate's livestock capital, left Azángaro after terminating the lease, or pleaded poverty and incapacity to pay the debt.^[188] The only real alternative consisted in leasing all estates by long-term emphyteusis contracts. But few tenants could be found willing to take over small, undercapitalized, and unstable fincas for more than a few years.

But why could the church not avoid such despoliations of its estates to begin with? It was improbable that a priest would mobilize the colonos of one of the parish estates in order to stop invasions by neighboring hacendados by force, an action routinely taken by private landowners. More important, the parish priests were firmly tied to the local society of their parishes through commercial contracts, friendship, and family relations. Because the tenants commonly belonged to the same small group of district notables as did the priest, the latter usually pursued the repayment of lost or embezzled church livestock capital rather half-heartedly and recommended to the diocese to lease estates at low rental rates.^[189] It thus appears that the church could neither avoid despoliations of its small fincas nor hope to recover most of its losses from short-term leases.

The decapitalization of the small church fincas allowed the tenants, often owners of haciendas themselves, to improve their income, increase their own livestock capital, and lease pastures at low rates in order to supplement fodder for their animals. Contrary to the church's critics, church estates in themselves cannot be seen as a major factor impeding economic progress in the Peruvian sierra. They were part and parcel of an agrarian system that for a number of reasons failed to stimulate the development of the rural economy. In brief, until the early twentieth century the landholding regime of the church contributed to the economic stability of the altiplano's seignorial livestock regime. It subsidized the interests of those new and old hacendado families who continued to rely on capital-extensive relations of production for sustaining their privileged socioeconomic status.

On November 7, 1911, law number 1447 went into effect, obligating the owners of all real estate held in emphyteusis to sell their title (dominio directo ) to the persons enjoying the usufruct right of such property, the enfiteuta . Coming half a century after the peak of anticlerical land legislation throughout Latin America, the law was a major step in reducing

mortmain landholdings; it seems not to have led to heated public debates about the role of the church in civil society. According to a complicated formula, it prescribed the share of the property's assessed value that the enfiteuta had to pay, a share that declined the longer the emphyteusis contract had been running. Puno's Bishop Valentín Ampuero immediately began to implement the law, ordering assessments of all church estates in question. In Azángaro the first emphyteutic finca, Hacienda Cancata in Santiago de Pupuja, was consolidated on September 11, 1912. By the end of 1918 eleven emphyteutic church fincas in Azángaro, among them such valuable large estates as Purina and Posoconi, had been alienated through consolidation (table 6.8). But Bishop Ampuero, for reasons not entirely clear, went one step further and, without legal necessity, initiated the outright sale of church fincas operated under short-term leases, a decision that earned him severe criticism within the church.^[190] Between April 1912 and March 1914 the diocese sold six fincas in Azángaro operated under short-term leases. They were mostly acquired by the tenants actually in possession. Altogether the church alienated seventeen estates in Azángaro in the seven years following the consolidation law, half of all its landholdings in the province. By value and acreage the reduction of church property was considerably steeper, since most of the large haciendas were consolidated or sold. Thereafter, the role of the church as landholder, with 5 to 7 percent of all estates, became insignificant in Azángaro.

The holders of emphyteutic haciendas profited greatly from consolidation. After having enjoyed bargain lease fees for decades, they now had to pay only a fraction of the estate's assessed value. In May 1913 Elena Landaeta, widow of José Luis Quiñones, paid 3,786.42 soles m.n. for the consolidation of Hacienda Parpuma, assessed at 9,351 soles m.n. Two months later she sold the finca and a few small adjacent estancias to Pio León Cabrera, a notorious land grabber from Sandia province, for 18,000 soles m.n.^[191] With the law of consolidation the landholding elite in the altiplano—and presumably elsewhere in Peru—benefited one last time from the church as a major player in the region's agrarian ancien régime.

The church's importance for the agrarian structure of Spanish America is thought to have rested on three factors: (1) its role as landholder; (2) the "huge amount of encumbrances"—such as chaplaincies and pious works—that weighed on privately owned estates; and (3) its role as creditor to private estate owners. In Arnold Bauer's view, church influence significantly declined throughout Spanish America during the century from 1750 to 1850 because of anticlerical action by the Bourbon reformers, by political and military leaders in the era of the Wars of Independence, and again by mid-nineteenth century liberal politicians.^[192] In the Peruvian

altiplano, however, church influence on the agrarian structure declined much more slowly during the first century after independence, perhaps because the region was distant from the centers of civil and ecclesiastic power. Church encumbrances on land had never been extensive there, and by the nineteenth century there is not much evidence for the church as a source of credit either. Legislation to abolish chaplaincies and censos before the War of the Pacific did nothing more than prevent new donations or encumbrances. As late as 1877 Martinet found that "the chaplaincies in Peru still exist nearly in their entirety."^[193] In Azángaro the few chaplaincies finally ceased to exist by the early twentieth century; church credit to private landholders remained as rare then as it had been in 1850 or even 1820.^[194] But the role of the church as a landholder remained strong for nearly a century after independence, at last virtually disappearing with the execution of the law of consolidation.

Communal group preparing fallow land for sowing, Melgar province.
Photograph by Pedro Condori, Talleres de Fotografia Social (TAFOS), 1989.

Members of a cooperative (ex-hacienda) near Ayaviri sack and weigh
alpaca wool for dispatch to wholesalers. Photograph by TAFOS, 1989.

Central building complex of Hacienda Muñani Chico of the Lizares
Quiñones family. Note the watch tower at the right. Photograph by the author, 1976.

The end of an era: the cooperative (ex-hacienda) Quisuni in Orurillo,
Melgar province, after it was burnt and destroyed by a Sendero Luminoso
detachment. Photography by Damaso Quispe, TAFOS, August 16, 1989.

Adoraida Gallegos, owner of Hacienda
Lourdes, circa 1910. Photograph from
the private archive of Mauro Paredes, Azángaro.

José Angelino Lizares Quiñones, one of
the most powerful hacendados and politicians
in Azángaro during the early twentieth century.
Photograph from an electoral pamphlet, 1924.

Electoral rally in Azángaro town for José Angelino Lizares Quiñones,
1924. Photograph from the private archive of Mauro Paredes, Azángaro.

Founding meeting of Azángaro's Fraternal Society of Workers in 1929.
Some hacendados were members. Seated, second from the right, is the
subprefect of the province. Photograph from the private archive of
Mauro Paredes, Azángaro.