Hostname: page-component-586b7cd67f-2brh9 Total loading time: 0 Render date: 2024-12-07T21:07:26.690Z Has data issue: false hasContentIssue false

Registering Race: Inclusion and Omission in Bolivian Military-Service Records 1900–1960

Published online by Cambridge University Press:  01 March 2024

Rights & Permissions [Opens in a new window]

Extract

Through presence or absence, indigeneity has long been at the center of Bolivian politics, culture, and nationalism.1 It carries immense social import. Yet what Indigenous means and who gets counted as part of this umbrella category has never been clear or static. Although Bolivia's racial and ethnic categories seem as etched in stone as the Andes themselves, individual classification is fluid and situational based on sociocultural markers that change over time. These markers include occupation, literacy, dress, language, surname, and residence. Individuals may move through these categories (and not only in one direction) over the course of their lives. This fluidity, combined with changing meaning and the valorization of indigeneity at the national and international levels, has led to wild fluctuations in official statistics on race since independence. The social construction of these categories came to the fore in 2012 when the census reported that Bolivians who identified as Indigenous had dropped to 40 percent of the population, down from 62 percent only eleven years earlier. International observers were left wondering, where did all the Indigenous people go?2

Type
Article
Creative Commons
Creative Common License - CCCreative Common License - BYCreative Common License - NC
This is an Open Access article, distributed under the terms of the Creative Commons Attribution-NonCommercial licence (http://creativecommons.org/licenses/by-nc/4.0), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original article is properly cited. The written permission of Cambridge University Press must be obtained prior to any commercial use.
Copyright
© The Author(s), 2024. Published by Cambridge University Press on behalf of Academy of American Franciscan History

Through presence or absence, indigeneity has long been at the center of Bolivian politics, culture, and nationalism.Footnote 1 It carries immense social import. Yet what Indigenous means and who gets counted as part of this umbrella category has never been clear or static. Although Bolivia's racial and ethnic categories seem as etched in stone as the Andes themselves, individual classification is fluid and situational based on sociocultural markers that change over time. These markers include occupation, literacy, dress, language, surname, and residence. Individuals may move through these categories (and not only in one direction) over the course of their lives. This fluidity, combined with changing meaning and the valorization of indigeneity at the national and international levels, has led to wild fluctuations in official statistics on race since independence. The social construction of these categories came to the fore in 2012 when the census reported that Bolivians who identified as Indigenous had dropped to 40 percent of the population, down from 62 percent only eleven years earlier. International observers were left wondering, where did all the Indigenous people go?Footnote 2

As Laurent Dubois has argued, the categories themselves must be at the forefront of any discussion of race; they are “social artifacts that demand” explanations rather than “generate” them.Footnote 3 Investigating a category's construction and deployment is essential to understanding its history and power. Since the 1990s, cultural anthropologists have debated the many factors that influence the changing uses and connotations of Indigenous identity categories in Bolivia.Footnote 4 Historians have more limited sources for understanding the self-identification of people outside of leadership positions.Footnote 5 They have analyzed cultural production, petitions, speeches, laws, and publications to expose the often contradictory ideas about race and ethnicity circulating among Creole and Indigenous elites.Footnote 6 Others have explored Indigenous intellectuals’ construction of race-based social movements.Footnote 7 Historians have also productively studied censuses, education systems, and public health initiatives to understand how different administrations applied these ideologies.Footnote 8

Building on these approaches, this article uncovers the institutional practices that surrounded the assignment and inscription of race in Bolivian military records. As a central institution of the state, the military played an important role in shaping racial categories and ideologies. When officers or clerks assigned each man to a race, they drew on cultural knowledge and personal experience to make abstractions into seemingly fixed and concrete identity categories. Drawing on conscription records, correspondence, and military-justice proceedings, this article examines the way that military officers and conscripts serving as clerks during Bolivia's liberal and revolutionary eras deployed race thinking in the paperwork used to track the men who registered for obligatory military service. Through analysis of official military documents and internal record-keeping, I argue that the institution categorized conscripts by race through the 1950s but omitted this data from public-facing documents starting in the 1920s. The article thus contributes to understandings of how ideology, individual decision making, and bureaucratic structures impact the use of racial categories.

The state shapes identity categories through the paper technologies it produces as part of bureaucratic efforts to understand and control the population.Footnote 9 These technologies are created as its agents develop and use categories to guide their efforts to collect information about the people they administer. State agents then use the data to fashion legal identities and issue demographic statistics through the census and other reports. The process of creating categories and applying them to individuals produces paperwork that certifies identity, nationality, and the completion of requirements, such as paying taxes, voting, serving in the military, and completing roadwork. The very form of these documents determines the results and serves to regularize and formalize categories. What information is collected and what is omitted? In what order is data presented? Are answers free form or are the categories preprinted? Do individuals supply their own information or does a state agent determine some of the answers for them? Through attention to the changing form of the paper technologies related to military service, this article probes the complicated nature of ethnoracial categories and the contradictions between administrations’ ideology and the daily practice of state agents.

Military Manifestations of Liberal Race Thinking

Although racialized hierarchies laid at the heart of their nation-building projects, Latin American elites tended to reject legal discrimination based on race. Many espoused ideas of classical liberalism, which, in acts of willful ignorance, presumed equal and unmarked citizens. They tried to refrain from using race to classify members of a society imagined as being on a trajectory toward becoming a “modern”—understood as homogenous—nation. As Mara Loveman has argued, such states constructed “formally inclusionary nations, but on deeply hierarchical grounds.”Footnote 10 These ideas impacted early twentieth-century Bolivian statecraft, sometimes resulting in explicit prohibitions on recording race. Yet race was so central to daily life that state agents produced racial data in a variety of documents and statistics.

Unlike in parts of Latin America where nationalist efforts led to the celebration of mestizaje, Bolivian intellectuals like Alcides Arguedas adopted prevailing ideas of scientific racism to condemn it as racial degeneration.Footnote 11 In 1899, Creole Liberal elites seized power after winning a civil war with the help of Aymara allies. The new administration moved the capital from Sucre to La Paz. They then worked to legitimize the highland region while simultaneously disavowing the allies who had brought them to power. As E. Gabrielle Kuenzli explains, this resulted in paradoxical efforts to denigrate the contemporary Aymara while making the Indigenous past central to nationalist narratives.Footnote 12 Although more virulent strains of biological determinism were certainly present in Bolivia, most Creole Liberals emphasized environmental determinism, denounced abuse, and expressed hope for cultural assimilation.Footnote 13 The authors of the 1900 census thus confidently predicted the “gradual disappearance of the Indigenous race.”Footnote 14 Yet many Liberal politicians were also rural landowners who depended on Indigenous labor. An important faction soon expressed fear of educational efforts, arguing that the Indigenous population would be easily swayed by caudillos and endanger the republic's democracy.Footnote 15

These contradictory ideas about race, nationalism, and Bolivia's path to modernity also marked rhetoric surrounding the project of universal male military service, which was implemented soon after the Liberal Party gained power.Footnote 16 Liberals frequently insisted that military service would be a sacred duty shared by the male population as whole. A shortened term of service, prohibitions on replacements, and the end of exemptions for tribute-paying Indians theoretically meant that the sons of prominent families, artisans, and Indigenous peasants would meet during service to forge a unified nation in the barracks.Footnote 17 Upon announcing the 1907 conscription law that would stand with only minor modifications until 1963, President Ismael Montes declared that “everyone without distinction of race or class” must be prepared “to give his blood in tribute to the nation.”Footnote 18 He further promised that the barracks would become “a practical training ground for equality where the son of the powerful brushes up against the son of the artisan and the Indian.”Footnote 19

Policies related to military service masked hierarchies of race and social class and presented them as based on personal merit. Stating that “the odious distinction between social classes should not be recognized in the barracks,” President Montes decreed that soldiers would be promoted based on only two factors: their level of education and their good conduct in the ranks.Footnote 20 Unwilling to recognize the institutional racism that prevented most Bolivians from accessing formal education, Montes presented these criteria as merit-based and equally attainable by all.

Documents authored by Bolivian military officers show “Indigenous” to be a marked category in a way that “Mestizo” and “White” were not. When they wrote letters or testified about conscripts who they perceived to be Indigenous, officers consistently identified them as “indios” or used the phrase “the indígena” to precede the person's name.Footnote 21 In similar documents concerning non-Indigenous men, I found them identified by race or social class only once. In that case, Major Alcibiades Antelo identified Alejandro Higuera and Hilarión Sánchez as Mestizos. However, he did so only to explicitly contrast them with two other deserters he described as “pure Indians, ignorant.”Footnote 22 Even more telling is a desertion report from 1911. In it, the prefect of La Paz ordered the capture of four miners who had deserted from the Viacha arsenal. He attached the personal information (filiación) for all four, but only one, an Indigenous man named Agustín Pacaza, was identified by race.Footnote 23 Officials lived in a society defined by the racialized difference between Indigenous and non-Indigenous, with Indigenous perceived to be inferior. They thus explicitly labelled conscripts who fit their definition of Indigenous.

Both officers and politicians presented conscription as a tool to nationalize and assimilate the Indigenous population. President Montes promised that military service would “incorporate that race [Indigenous] into the political life of the country and civilization in general.”Footnote 24 Three years later, Minister of War Andrés Muñoz assured Congress that conscription was already resulting in “the regeneration of the Indigenous race.”Footnote 25 Officers testified that military training was causing “a complete transformation in the ideas, habits, and customs” of men brought into the institution.Footnote 26 Officials promised that conscripts would learn Spanish, basic literacy skills, and modern hygienic methods during their time in the barracks. Not surprisingly, these statements make clear the assumption that Indigenous men needed to be improved by military service. For example, Minister of War Néstor Gutiérrez noted in 1914, “A few days after his admission to the barracks, the Indian, before retiring and timid, transforms into a soldier, handsome and self-controlled. It is noteworthy how, in the course of a couple months of military education, large modifications in conscripts’ character and temperament come about, equalizing all conditions and producing tenacity even in those least favored by nature, putting them on par with the most capable and intelligent.”Footnote 27 These comments reveal both environmental and biological ideas about race, with the minister of war promising, somewhat paradoxically, that military service would “fix” both cultural traits and those assumed to be inherent.

Leaders thus took additional measures to encourage Indigenous men to register and serve. Decrees in January and December of 1908 and ministerial resolutions in 1913 and 1916 gave Indigenous men more time to register without being pursued as evaders.Footnote 28 They used familiarly paternalistic tropes about Indigenous ignorance to explain these targeted exceptions, citing “the special condition of the Indigenous race” and ordering landowners and civic and religious authorities to carefully explain the rules and penalties to the Indigenous population.Footnote 29

Military conscription took hold during a period marked by widespread uprisings and organized activism from Indigenous communities for land, education, access to markets, and an end to abusive practices.Footnote 30 While state actors hoped military service would help solve these problems by making men less Indigenous, they simultaneously feared the repercussions of arming and training men they did not identify with or trust. They worried that Indigenous conscripts might refuse to repress rural unrest or might use their skills against the state after shedding their uniforms. Oruro Prefect Eduardo Diez de Medina expressed these concerns in 1914: “[The Indian] shows himself to be proud, and in some cases defiant, when he returns to his land, forming perhaps a danger for the security of the internal order.” He suspected that Indigenous men who were eager to enroll had ulterior motives, hypothesizing that they had been “seduced by the prospect of becoming skilled in the use of arms and getting mixed up in aggressive adventures.”Footnote 31 Haunted by the specter of race war, Bolivian elites feared both the disease and their own proposed cure.

Recording Indigeneity in the Early Twentieth Century

Although elites’ liberal ideology discouraged them from collecting data on race, indigeneity appeared frequently in military documents in the early twentieth century.Footnote 32 State agents collected personal and demographic information (filiación, antecedentes, and generales) during the registration and exemption process as part of efforts to construct the bureaucratic apparatus necessary to administer and perhaps someday enforce the conscription system. They also implemented a new paper technology, the military service booklet, which detailed the man's personal information and record of military service.Footnote 33 Because of provisions that demanded it be produced to vote or be employed in the formal sector, this document served as a mechanism to coerce men, especially members of the middle and upper classes, to regularize their status by serving or obtaining an exemption.Footnote 34 The form and content of this paperwork merit detailed analysis because they reveal administrative concerns and the categories that government officials used to understand the population.

Bolivia's 1907 military service law called for military and civilian officials to set up tables every August and September in each departmental capital, provincial capital, and canton to register all eighteen-year-old men.Footnote 35 Registration involved data collection; they were to record the man's first and last names, place of birth, age, marital status, skin color, physical marks, “trade or profession,” residence, and registration date.Footnote 36 The continued mandate to record skin color apparently raised questions because of its association with race. A later clarifying memorandum by Minister of War José S. Quinteros delved further into the question of skin color, instructing state agents to note “the exact color of the skin on the face, that is to say, if he is pale (pálido), olive-skinned (trigueño), dark-skinned (moreno), etc., without taking race into account under any circumstance.”Footnote 37 In addition to showing Liberals’ aversion to generating racial data, his comments reveal the assumption that race existed and could easily be identified (or ignored).

Again prohibiting the inscription of race, Quinteros instructed agents manning registration tables to record conscripts’ “social class,” self-consciously reassuring them that this classification was necessary to avoid confusion between men with the same names and to help the state locate the individual in question.Footnote 38 However, the detailed guidelines and regulations surrounding the 1907 military service law make no mention of social class. Quinteros's instruction to record it suggests the power of paper technologies to shape practice. I suspect that agents were recording this information because they were still using older preprinted forms. As shown in Figure 1, the “social class” category was present in military registration sheets from 1906 but absent by the 1920s.Footnote 39 Until new registration sheets were printed, agents would have continued collecting the information dictated by the older ones.

Figure 1 Military Registration Sheets from 1905 and 1923

Source: September 1905 military conscription sheet with line for social class attached to petition of “indígena Isidro Quispe” to the prefect of La Paz, February 21, 1908, Prefecture-Exped box 164, d. 32, ALP; September 1923 military registration sheet without line for social class attached to petition of Ernesto Bedregal to the prefect of La Paz, September 21, 1923, Prefecture-Exped box 254, d. 11, ALP. Reproduced with permission by the Archive of La Paz.

In the same 1907 document, Minister of War Quinteros suggested that agents choose from “indígena, artisan, cholo, gentleman” to describe social class.Footnote 40 Indígena has a long history as a fiscal, administrative, and legal category during the colonial period, and differentiated duties related to the category persisted well into the Republican era.Footnote 41 The meanings of Cholo have changed depending on time and context. Generally, the term signifies people of Indigenous or mixed descent who do not display some of the sociocultural markers associated with indigeneity.Footnote 42 Quinteros's list of social classes highlights how Bolivia's ethnoracial terms overlapped with class and occupational ones. By making “Indigenous” and “Cholo” into classes rather than races, he implied that the categories were cultural and environmental rather than biological. The minister of war drew on Liberal ideology that rejected racial difference while simultaneously asserting the importance of the categories that shaped his world in early twentieth-century Bolivia. He therefore presented these categories as class-based and mutable rather than race-based and immutable. Yet the categories’ very names revealed them to be profoundly racialized.

On paper, intellectuals might have been able to parse out differences between ethnicity, race, and social class. In practice, however, these categories were profoundly interrelated and imbricated. This divide between theory and practice allowed Quinteros to insist that the state agents manning registration tables ignore race while instructing them to record Indigenous or Cholo as a person's social class. As Rossana Barragán notes: “Even though race was conceived of in biological terms of origin and descendants, in practice the criteria were occupational because occupations were themselves already racialized.”Footnote 43 The profound association between Indigenous status and rural agricultural labor made impossible the idea of an urban Indian or an elite Indian. In the eyes of those doing the categorizing, Indians who left rural areas were no longer Indigenous.

Tellingly, indígena appeared as an occupational category in reports to Congress about the results of conscription efforts. From 1907 to 1912, the War Ministry chose to include in its report data on just three key factors: how many men registered from each department, their marital status, and their occupation.Footnote 44 As reflected in Table 1, these occupational tables were organized hierarchically, beginning with lawyers and ending with artisans and then indígenas. The categories were more specific for prestigious occupations, differentiating between medical, business, accounting, and theology students but lumping together all artisans and all rural workers. The information collected at registration tables was far more granular, including more than fifty different less-prestigious occupations like hatmaker, carpenter, bricklayer, blacksmith, day laborer, and farmworker.Footnote 45 The structuring of these tables reveals that state actors were more concerned with knowing details about the men presenting for service from the upper rungs of society than about the working-class men who constituted most of the ranks.

Table 1: Aggregate Professional Categories Used in Reports on Conscripts, 1907-1912

Data on the 1909 cohort was not presented in either the Boletín Militar or the minister of war's report to Congress. This table includes all conscripts who presented for service, including those exempted through the lottery. The year listed is the year that conscripts presented for service. A dash (-) indicates that the occupational category was not included in that year's report. The order varies slightly between reports, but the order presented is the most common. Because the summary table for the 1907 cohort seems to have a calculating error that misrepresented the number of artisans as 187 and of indígenas as 413, the data included in this table are based on the departmental breakdowns instead. Sources: Boletín Militar 1907, 300-302; Memoria de Guerra 1908, iv-ix; Anexos de Guerra 1910, clvii-clix; Anexos de Guerra 1911, 136-137; Anexos de Guerra 1912, 85-87.

The use of indígena as an aggregate professional category supports Barragán's conclusions about the racialization of professions.Footnote 46 Notably, none of the registration lists, individual sheets, or military service booklets I have found list a conscript's profession as “indígena.” Instead, agricultor, labrador, jornalero, colono, and peón commonly appear on the same list, leading me to believe that these categories were self-reported or that different scribes used different terms to signify agricultural labor.Footnote 47 The lack of other aggregate categories related to agriculture suggests that the people generating the statistics for these reports categorized all lower-status conscripts involved in food production as indígena. The sons of large landowners likely appeared as professionals, students, or as secondary graduates. Only in 1912, the last year that the conscription report was published, did it include the category “agriculturalist.” It did not explain how those thirty-three men differed from the two hundred indígenas recorded below them, but the category's placement suggests that it was considered a higher social status than indígena. The explicit inclusion of the descriptor “illiterate” after artisan and indígena in the 1910 report confirms perceptions surrounding the social status of these categories.

The coding of indigeneity as an occupation and Minister of War Quinteros's defensiveness stand in stark contrast to the documentary mandate in the 1900s and 1910s to record race on the most prominent military document of the time. Upon discharge, each conscript received a military service booklet that documented their service record and personal information. The version issued to conscripts between 1909 and 1919 (and perhaps longer), recorded race on the top line of the third page.Footnote 48 The booklet included race under physical appearance (señales exteriores) rather than with place of birth, domicile, and parentage (antecedentes), which suggests that the form's designers understood race to be manifested through physical appearance. The placement of race at the top of the list of physical characteristics implies that race was the primary and most important identifying characteristic. In fact, when race appeared on military documents, it was always the first characteristic, topping descriptors like skin color, hair color, height, and weight.

Paper technologies during the Liberal era consistently included Indigenous status. However, the category to which this status belonged was varied; it appeared as a response to prompts for race, social class, and occupation. This malleability points to the centrality of indigeneity in Bolivia. It also betrays Creole elites’ uncertainty about their nationalist narrative and how to deploy social science categories in an international context dominated by eugenics. In the coming decades, however, Indigenous status both solidified into a racial category and disappeared from public military documents.

Omitting Race from Public-Facing Documents

Despite the documentary prominence of race in the early twentieth century (or perhaps precisely because of it), the designers of the military service booklet eliminated the line for race in the 1920s. This meant the state was no longer racially marking men, at least publicly. When asked to show their military documents by patrols, at the polls, or when called up for war, former conscripts would present a paper technology that no longer categorized them by race. Yet race by no means disappeared from official records. A variety of sources from military and prefecture archives suggest that authorities routinely generated and transmitted racial data at least through the mid-1950s but did not include it in public-facing documents or those designed to be archived.

The form of military service booklets changed over time, with new versions issued in the 1920s, after the Chaco War (1932–1935), and again in the 1940s. None of these later booklets included a place for race to be recorded.Footnote 49 The initial omission of race from military service booklets in the 1920s came as part of a larger paring down of the physical description of conscripts that eliminated reporting on four of nine features.Footnote 50

The military service booklet ventures into the world alongside the former conscript whom it documents; its archival counterpart is the military service sheet, which lists the same personal and service information as the booklet. Literate conscripts or noncommissioned officers serving as clerks in each military unit filled out this form, and then the relevant authorities signed it and sent it to the Territorial Registry of the Ministry of Defense, which houses service sheets dating from demobilization after the Chaco War through the most recent cohort of discharged conscripts. This archive's primary function is to verify men's service or legitimate exemption when they run for office, request documents, or otherwise need to prove compliance with the obligatory military service law.Footnote 51 The service sheets took on a variety of forms and documented different information over the years; however, none of the more than two thousand preprinted service sheets that I sampled from this archive for the years 1936 through 1974 offered a place to classify race.Footnote 52

These public and formally archived documents imply that the military stopped systematically producing racial data on conscripts in the 1920s. Yet isolated sheets from the late 1930s and early 1940s suggest otherwise. Although the form did not solicit conscripts’ physical description, the clerk who filled out service sheets for the Vegara Fifth Artillery Regiment in 1938 used the space designated for medical information to record race along with a description of each conscript's eyes, nose, and hair.Footnote 53 I also found four mimeographed documents from 1938 and 1940 that list both race and skin color as part of conscripts’ physical description.Footnote 54 Are these documents anomalies, the products of errant clerks who recorded extraneous information? Or are they evidence of the continued classification by race during the recruitment process?

Prefecture and military-justice records support the latter explanation. In the 1940s, race persistently appeared on the less formal paper technologies that are in these archives. One example is the desertion notices (parte de deserción) used to locate conscripts who had left the ranks without permission. Rather than professionally printed forms like the military service sheets and booklets, desertion notices were typically mimeographs of typed forms.Footnote 55 This allowed for variability over place and time. An officer or clerk typed or handwrote the conscript's name, address, parentage, physical description, and the circumstances surrounding his desertion. The unit then sent desertion notices to the minister of war, who forwarded them to the departmental prefect, who in turn sent them to the relevant local authority. Reflecting the practice of reporting race in military service booklets at the time, all but two of the sixty-five notices (97 percent) I found from the 1910s included racial data.Footnote 56 More surprising is that all but two of the twenty-nine desertion notices (93 percent) I found from 1940 to 1950 also reported race.Footnote 57 The clerks who filled out victim reports for accidents and deaths also included the conscript's race in addition to other personal information.Footnote 58 Decades after the category's elimination from the booklet and archived service sheets race was still topping the list of conscripts’ physical characteristics in these documents designed for internal use.

The categorization of conscripts by race whenever a form called for it suggests that racial data was still being routinely generated during the recruitment process on a paper technology that moved with the conscript during his service. Before investigating these questions about race, I had assumed that the service sheets served this function, but the appearance of racial information elsewhere led me to examine more closely all the sheets in my sample.Footnote 59 The results suggest that documentary practices varied widely in the late 1930s and the 1940s. Handwriting and typing indicate that most units filled out the sheet in its entirety at the time of discharge whereas some others seem to have filled it out partially upon assignment to the unit, stored it locally, and then completed it upon discharge.Footnote 60 Generally, sheets sampled from same unit in the same year were filled out similarly, but my analysis found no consistency across years within the same unit.Footnote 61 The lack of uniformity in form and procedure is consistent with my previous conclusion that, even after the Chaco War, the Bolivian military lacked the capacity or will to establish an effective bureaucratic regime.Footnote 62

The units that filled out the entire service sheet after completion of service had to use another instrument—that likely included racial data—to keep track of conscripts’ information. This perhaps explains the four mimeographed documents mentioned above. These look like service sheets but differ in three ways: the forms are mimeographed rather than preprinted, they do not solicit discharge information, and they include lines for both race and skin color in the physical description. The first two factors demonstrate the internal purpose of the form. The third suggests that the military was deliberately leaving race out of public-facing documents. These four sheets accidentally ended up archived alongside the same conscripts’ official service sheets likely because these men had been reassigned from their original units to the Ministry of Defense. I suspect that many units used mimeographed sheets like these that included race, transferred the information to the preprinted sheets at the time of discharge, and then destroyed the original. Reflecting the heterogeneity of practice at the time, other units used the service sheet that would eventually be archived.

The Military Register supports the theory that, at least at particular registration sites, the military was systematically creating racial data and then omitting it from the public record. Also archived at the Territorial Registry, this register is a bound volume of the original registration tables from each of the recruitment sites around the country. The size of four standard sheets of paper, these preprinted tables were filled out by hand or typewriter after the close of recruitment. Each man who presented for service, including those medically ineligible or dismissed through the lottery, had his information recorded on a row. The form, in use until 1953, reported each conscript's personal information, medical appraisal, and physical description.Footnote 63 It did not provide a column for race. However, at certain recruitment sites, clerks classified conscripts as M for Mestizo, I for Indigenous, or B for White. Between 1947 and 1952, race was assigned to more than 27,500 conscripts, which was about half of the men whose information made it into the archive.Footnote 64 Clerks in this period bisected the profession column to add an additional column for race. This sandwiched information about the conscript's race between his occupation and literacy status, signaling the continued association of these factors. This placement stands in contrast to the desertion reports and mimeographed registration records, which consistently listed race as part of conscripts’ physical description.

The high percentage of tables with this information and its uniform placement show that recording race in the Military Register was a widespread practice. It was most common in the Cochabamba department, but I found no other consistency in registration sites, handwriting, or the names of the officers who signed for the 561 tables that included race.Footnote 65 Were military personnel only generating racial data at these sites? Or were they generating it at all sites but only entering it on the registration tables at some? The consistent inclusion of race in the desertion, death, and victim reports drawn from registration data leads me to suspect the latter: that military officials classified men by race during the registration process, but only some clerks were told to draw in a column for it when filling out the registration tables. The others did not include it because the preprinted table did not solicit it.

The omission of race from mass-printed forms like the registration tables, service sheets, and military service booklets was likely a conscious decision made near the top of the state hierarchy. This design choice reflected liberal ideals of equality that rejected racial classification. However, evidence from the Military Register, service sheets, and desertion reports makes a strong case that, during the 1930s and 1940s, the standard practice throughout Bolivia was still to assign a race to each man who registered for military service.Footnote 66 Was this decision made at an equally high level? I suspect so, but it also could have resulted from continuance in clerical practice, with recruitment commissions unthinkingly reproducing forms that included race. Either way, the military was producing racial data on conscripts during this period but not including it in public documents.

Recording Race in the MNR's Bolivia

Many things about Bolivia changed profoundly after April 1952, when the Revolutionary Nationalist Movement's (MNR) coup plot turned into a popular revolution from below. Although led by elite and middle-class reformers, the party gained significant support in the mines, factories, and countryside. After enacting universal suffrage, the MNR responded to pressure to nationalize the largest tin mines and, more reluctantly, enact a widescale agrarian reform that titled land to “those who work it.”Footnote 67 According to popular lore, the MNR destroyed the military and attempted to rebaptize the Indigenous population as peasants. Although the latter was hyperbole, much work remains to understand the meaning and deployment of racialized categories in the MNR's Bolivia.Footnote 68 This section contributes to ongoing scholarly efforts by analyzing data-generation practices in the military, showing that clerks continued to systematically categorize conscripts by race until 1959, when race abruptly disappeared from the Military Register.

Scholars’ understanding of race in revolutionary Bolivia has become increasingly nuanced. Reflecting dominant global trends in the 1950s and 1960s, the first wave of scholarship emphasized the stigma associated with Indigenous status and assumed that assimilation would benefit the individual and nation.Footnote 69 Beginning in the 1970s, as Indigenous movements around the world gained prominence, the coding of this assimilationist vision for Bolivia shifted from celebratory to condemnatory, with scholars detailing the devasting material effects of the MNR's “reformist mestizo project.”Footnote 70 However, recent research is complicating understandings of both the party's ideology and the effects of its policies. Looking at the MNR's cultural project, Matthew Gildner argues that leaders drew on and promoted Indigenous traditions to develop a nationalist cultural project with an “inclusive veneer.”Footnote 71 In terms of effects, Carmen Soliz argues that the MNR's assimilationist discourse contrasted with its agrarian policies, which actually strengthened Indigenous communal structures.Footnote 72

The revolution caused intense upheaval in Bolivia's military, leading to drastic reductions in budget and size. The officer corps also underwent a significant renewal with about half discharged and others returning from exile.Footnote 73 Contrary to popular belief, the new administration did not abolish and rebuild the military.Footnote 74 The events of 1952 did leave a significant mark on conscription records, however. A host of military service sheets from the 1952 cohort are stamped “with insufficient instruction” and show that many conscripts received discharges in late April 1952 instead of serving a full year.Footnote 75 Although a similar number of men registered for service in 1953 as had in previous years, the Military Register reveals the process to have been an uneven affair. In the outlying departments of Beni, Pando, and Santa Cruz, where conscripts primarily worked on road and colonization efforts, men registered and entered the ranks in January, as they had before the revolution. In La Paz, Cochabamba, Chuquisaca, Potosí, and Oruro, conscription did not take place until after the revolution's first anniversary in April.Footnote 76

The revolution also coincided with a major change in one of the paper technologies used for conscription.Footnote 77 Most of the men recruited in 1953 had their information recorded on a new preprinted registration table rather than the one that had been in use since at least 1940, when the Ministry of Defense's records begin.Footnote 78 Omitting columns routinely left blank on the previous version, the new form was smaller and did not need to be folded to fit in the bound volume. It did not have columns for most physical descriptors, only retaining a place to record distinguishing marks. More importantly, registration in 1953 seemed to reflect the MNR's professed goal of eliminating Indigenous status. The clerks who filled out the 431 tables listing the information of more than 12,000 men did not assign a racial identity to any of them.Footnote 79

A homicide case from that year, however, shows that the absence of race from the Military Register does not mean that the postrevolutionary military had ceased producing racial data. When Zenón Vedia Chungara accidentally shot fellow conscript Herminio Pérez Paco while patrolling the town of Corque (Oruro) in July 1953, the subsequent military-justice proceeding included a typed copy of Pérez Paco's service record that listed his race as Indigenous. When Pérez Paco presented for service in the city of Oruro in April 1953, the clerk must have filled out a form that solicited information about his birth, family, address, profession, skills, literacy, level of education, languages spoken, and physical description, which included both race and skin color.Footnote 80 With the glaring exception of race, all this information would need to be transferred to either the Military Register or the service sheet sent to the archive.Footnote 81 This internal form thus adds to evidence suggesting the systematic generation and deliberate suppression of racial data during military conscription.Footnote 82 It also shows that the MNR's revolution effected only the reporting of racial information in 1953, not the collection of it.

Despite the absence of race from the 1953 Military Register, it came roaring back into the records when the military held nation-wide registration in February 1954 and continued to appear in the tables filled out in April 1955, April 1956, April 1957, and January 1958. This trend was particularly pronounced in 1954, when the Military Register listed a race for 51 percent of all registrants. Although the percentage of men identified by race was similar to prerevolutionary levels, the places reporting racial data had changed profoundly. Before the revolution, 72 percent of the men identified by race had presented at registration sites in Cochabamba. In the postrevolutionary period, clerks from at least one registration site in each department, except Beni and Pando, drew in a column for race, and only 20 percent of total racial identifications came from Cochabamba.Footnote 83 Far from eliminating racial identification, the military under the MNR widened the practice across the national territory.

The placement of racial data also changed. Whereas race had earlier been squeezed in alongside profession, after 1952, a column was added under “Observations” on the far-right side of the table. This placed race in a more prominent position on the page and subtly implied that the category could stand on its own as compared to the previous position, which had linked race to occupation and literacy. Out of the 632 tables from the post-1952 period that included race, only one reverted to the prerevolutionary placement. In this table from 1955, the clerk entering data from Aquile (Cochabama) wrote a small “M” or “I” next to the profession of the first twenty-nine men who registered for service. He then omitted race all together on the final four pages of registrants. This suggests that, after completing the first page, he realized his mistake or someone corrected him. The remarkable consistency in placement across hundreds of tables filled out by at least forty-six different clerks indicates that they received specific instructions for recording race.Footnote 84

The inclusion of race in the Military Register declined precipitously after 1954, with race recorded for only 28 percent of registrants in 1955, 15 percent in 1956, and then under 7 percent for the next two years before permanently disappearing in 1959.Footnote 85 This drawn-in column provokes additional questions. Why were some clerks instructed to add race while others were not? Did the military stop producing racial data starting in 1959? Or did it just stop telling clerks to add this information to Military Register? Recruitment manuals unfortunately do not answer these questions; they make no mention of what information to collect or how to record it.Footnote 86 The officers involved in registration or the conscripts who served as clerks might be able to clarify—if they remember such a minor detail this many years on. These questions will likely remain unanswered unless internal communications surrounding recruitment are located.

Even without definitive answers to these questions, the existence and substance of racial data reveals much about Bolivia in the 1950s. Its absence from the 1953 Military Register suggests that the military attempted to eliminate reporting on race to reflect the new administration's ideology. Yet the clerks who registered conscripts clearly classified at least some of them by race that year, and the institution restarted and even widened the reporting of race in 1954. This indicates that the collection and reporting of racial data was a conscious decision rather than a habit that had not yet been broken. Racial labels continued to have tremendous power and meaning in the MNR's Bolivia.

Because racial data was only reported at particular recruitment centers, these records cannot be used to determine the breakdown of assigned race among troops in the military or to see how it changed over time.Footnote 87 Instead, each set of registrants must be analyzed on its own terms, taking into account the demographics and social norms of the surrounding area. The handwriting, cross-outs, and blanks in the Military Register also serve as reminders that people filled out the forms and transferred the information. They made mistakes in the process, and they also made individual judgements about each man's race based on ethnoracial markers that varied depending on region and urbanity.

Rather than providing transparent information about the race of the conscripts, each set can be used to understand how the clerks manning that recruitment center might have assigned racial categories to the men who presented for service. For example, the data collected in 1954 on the sixty-one men who registered in the rural province of Inquisivi (La Paz) shows a strong correlation between race and three other factors: literacy, profession, and surname.Footnote 88 Compared to those identified as Mestizo, men identified as Indigenous were far more likely to be listed as illiterate (47 percent versus 19 percent) and as farmworkers (57 percent versus 94 percent).Footnote 89 Clerks did not identify any registrants with Indigenous surname(s) as White but listed twenty-one of them as Mestizo and fourteen as Indigenous.Footnote 90 Although men classified as Indigenous were more likely than those classified as Mestizo or White to be listed as illiterate, as agricultural workers, and as having at least one indigenous surname, the Military Register also shows that Indigenous status was by no means solely determined by these factors. Many of the men whom clerks identified as Indigenous were literate, had two Hispanic surnames, and listed professions such as office worker, tailor, miner, or student.

Racial data from the Military Register is also useful for understanding the effects of assigned race on military service. Except at few provincial registration sites, clerks identified the vast majority of men who presented for service in the 1950s as Mestizo.Footnote 91 This stands in stark contrast to the 1950 census, which classified 63 percent of the population as Indigenous.Footnote 92 The disproportionate predominance of Mestizos shows that men from Indigenous communities were less likely to register for service, that clerks were registering them as Mestizo, or a combination of the two. Either way, the overrepresentation of men classified as Mestizo supports the idea of the military as an assimilatory institution. The Military Register also shows that registrants designated as White were far more likely to be declared unfit for service or only fit for auxiliary service than those listed as Indigenous or Mestizo.Footnote 93 This suggests that bias(es) correlated with racial classification played a role in the medical examination process that exempted some men from military service.

Because it originated from registration sites across the country rather than primarily from Cochabamba, the racial data produced in 1954 offers an opportunity to understand the construction and effects of race. Registrants categorized as White are clearly distinct, listed as having more education and more prestigious professions. More importantly, they were far more likely to be exempted from military service than their nonwhite peers at the same registration site. I found less distinction between those classified as Indigenous versus as Mestizo. Although level of education, literacy, surname, and region correlated strongly with Indigenous status, no one or combination of these factors can be used as a reliable indicator of whether clerks would classify men as Mestizo or as Indigenous. Returning to the example of Herminio Pérez Paco, the clerk listed him as Indigenous despite the fact that he was literate; had completed four years of primary education; had both Hispanic and Aymara surnames; and spoke Spanish in addition to Aymara and Quechua.Footnote 94 The lack of linguistic data in the Military Register prevents wider analysis of this final factor, but limited evidence from other sources suggests that languages spoken was an unreliable indicator of racial classification in the military during the 1950s.Footnote 95

The continued production of racial data during the conscription process points to the enduring power of racial categories and to continuity in the military's bureaucratic practice after the 1952 revolution. Yet the absence of race from the 1953 registration tables and its complete disappearance after 1958 are notable ruptures that likely reflected the MNR's efforts to erase race from the national lexicon and make the population less Indigenous. This ideology may be the reason that most men who registered for military service in this period were classified as Mestizo. The racial data preserved in the Military Register also points to the power of race, particularly whiteness, to shape a man's military service (or lack thereof) in postrevolutionary Bolivia.

Conclusions

The persistent presence of race in military records despite attempts to eliminate it as an acceptable category confirms its importance in Bolivia. Although military service was based on ideals of universality, indigeneity was at the heart of the project. Across party and ideology, conscription policy was always formulated with the explicit goal of incorporating Indigenous people into the nation and reshaping their ways of life. It should thus come as no surprise that the military continued to classify conscripts by race despite anxiety about the practice and the eventual elimination of race from preprinted forms. The evolution of documents related to military service shows the social dimensions and contingency of knowledge production. Many officers directed clerks to draw in a column for race during registration and explicitly marked particular conscripts as Indigenous in correspondence and when testifying. While these officers may have agreed with the principles behind policies that effaced race, they lived in world in which it was practically unthinkable to do so.

Even though people used racial categories as if they were static and easily definable, their content was far more slippery. The ways in which race is present in military records practically screams at researchers to investigate the process that produced these data. Whether and how each man was classified depended on a clerk who drew on his personal experience and societal norms in order to choose whether the man standing in front of him should be counted as Indigenous, Mestizo, or White. The omissions and mistakes apparent in these records should remind researchers of the unreliability of the data that underlies statistics. However, this analysis also points to the benefits of generating racial data. In inscribing a race for the men who registered, the clerks signaled the importance of this category in Bolivian society and provided researchers with a way to document the effects of racialized biases on obligatory military service.

Footnotes

This article had its origins in a 2012 AHA panel, “Racial Silences in the Archive,” organized by Chad Black. My thanks to him, fellow panelists Martine Jean and Roderick Barman, and the audience for helping me start to think through these ideas. John French did me a tremendous service by flagging the few paragraphs related to the omission of race included in the draft of a dissertation chapter and helping me understand that I had not yet figured out my argument about the material. Corinna Zeltsman and Santiago Muñoz Arbeláez's kind invitation to participate in a 2017 conference on Paper Technologies prompted me to rethink these ideas alongside other scholars. My thanks to them and all the other participants for giving me a new vocabulary and new questions to consider. Two summer research fellowships from Oakland University's URC funded the writing of this article and the employment of two undergraduate students, Amber Brown and Shantel Glover, to input data from the Territorial Registry. Finally, I would like to thank Luis Sierra, Jonathan Ablard, and The Americas reviewers and editors for helping me improve and clarify drafts of this work.

References

1. Following Sarah Radcliffe, I do not capitalize indigeneity, which she defines as, “the socio-spatial processes and practices whereby Indigenous people and places are determined as distinct (ontologically, epistemologically, culturally, in sovereignty, etc.) to dominant universals.” Radcliffe, Sarah A., “Geography and Indigeneity I: Indigeneity, Coloniality, and Knowledge,” Progress in Human Geography 41, no. 2 (2017): 221, 226n4CrossRefGoogle Scholar.

2. For a representative headline asking that question, see Sara Shahriari, “Where Have All the Indigenous Gone? Bolivia Sees 20 Percent Drop,” Indian Country Today, August 12, 2013, https://indiancountrytoday.com/archive/where-have-all-the-Indigenous-gone-bolivia-sees-20-percent-drop.

3. Dubois, Laurent, Avengers of the New World: The Story of the Haitian Revolution (Cambridge, MA: The Belknap Press, 2004), 5CrossRefGoogle Scholar.

4. For recent analyses of the literature and major debates as they pertain to Bolivia, see Albro, Robert, “Cholo Politics and Urban Indigenous Self-Fashioning in Bolivia,” Bolivian Studies 25 (2020)Google Scholar; Canessa, Andrew, “Indigenous Conflict in Bolivia Explored through an African Lens: Toward a Comparative Analysis of Indigeneity,” Comparative Studies in Society and History 60, no. 2 (2018)CrossRefGoogle Scholar; Fabricant, Nicole and Postero, Nancy, “Performing Indigeneity in Bolivia: The Struggle Over the TIPNIS,” Anthropological Quarterly 91, no. 3 (2018)CrossRefGoogle Scholar; Weber, Katinka, “Chiquitano and the Multiple Meanings of Being Indigenous in Bolivia,” Bulletin of Latin American Research 32, no. 2 (2013)CrossRefGoogle Scholar.

5. Luis Sierra's recent study of La Paz's Indigenous neighborhoods in the prerevolutionary period represents an important advance in our understandings of the lived experience of racialized categories in Bolivia. Sierra, Luis M., La Paz's Colonial Specters: Urbanization, Migration, and Indigenous Political Participation, 1900-52 (New York: Bloomsbury Academic, 2021)CrossRefGoogle Scholar.

6. Robert Matthew Gildner, “Indomestizo Modernism: National Development and Indigenous Integration in Postrevolutionary Bolivia, 1952-1964” (PhD diss, University of Texas at Austin, 2012); Gotkowitz, Laura, A Revolution for Our Rights: Indigenous Struggles for Land and Justice in Bolivia, 1880-1952 (Durham, NC: Duke University Press, 2007)Google Scholar; Kuenzli, E. Gabrielle, Acting Inca: National Belonging in Early Twentieth-Century Bolivia (Pittsburgh, PA: University of Pittsburgh Press, 2013)CrossRefGoogle Scholar; Brooke Larson, “Redeemed Indians, Barbarized Cholos: Crafting Neocolonial Modernity in Liberal Bolivia, 1900-1910,” in Political Cultures in the Andes 1750-1950, ed. Nils Jacobsen and Cristóbal Aljovín de Losada (Durham, NC: Duke University Press, 2005); Young, Kevin A, Blood of the Earth: Resource Nationalism, Revolution, and Empire in Bolivia (Austin: University of Texas Press, 2017), 50, 136–37Google Scholar.

7. Ari, Waskar, Earth Politics: Religion, Decolonization, and Bolivia's Indigenous Intellectuals (Durham, NC: Duke University Press, 2014)Google Scholar; Khol, James, Indigenous Struggle and the Bolivian National Revolution: Land and Liberty! (New York: Routledge, 2021)Google Scholar; de la Torre, Cecilia Salazar, ed. Intelectuales aymaras y nuevas mayorías mestizas: Una perspectiva post 1952 (La Paz: Programa de Investigación Estratégica en Bolivia [PIEB], 2012)Google Scholar; Alejo, Esteban Ticona, El indianismo de Fausto Reinaga: Orígenes, desarrollo y experiencia en Qullasuyu-Bolivia (La Paz: Producciones CIMA Editores, 2015)Google Scholar.

8. Barragán, Rossana, Asambleas constituyentes: Ciudadanía y elecciones, convenciones y debates (1825-1971) (La Paz: Muela del Diablo Editores, 2006)Google Scholar; Grieshaber, Erwin P., “Fluctuaciones en la definición del indio: Comparación de los censos de 1900 y 1950,” Historia Boliviana 5, no. 1–2 (1985)Google Scholar; Brooke Larson, “Forging the Unlettered Indian: The Pedagogy of Race in the Bolivian Andes,” in Histories of Race and Racism: The Andes and Mesoamerica from Colonial Times to the Present, ed. Laura Gotkowitz (Durham, NC: Duke University Press, 2011); Loveman, Mara, National Colors: Racial Classification and the State in Latin America (New York: Oxford University Press, 2014)CrossRefGoogle Scholar; Pacino, Nicole, “Liberating the People from Their ‘Loathsome Practices:’ Public Health and ‘Silent Racism’ in Post-Revolutionary Bolivia,” História, Ciências, Saúde-Manguinhos 24, no. 4 (2017)CrossRefGoogle ScholarPubMed; Sierra, La Paz's Colonial Specters; Zulawski, Ann, Unequal Cures: Public Health and Political Change in Bolivia, 1900-1950 (Durham, NC: Duke University Press, 2007)Google Scholar.

9. The term “paper technology” comes from Santiago Muñoz Arbeláez. I am indebted to him and Corinna Zeltsman for the opportunity to think through these ideas as part of the “Paper Technologies: The Materiality of Empire and State Formation in Latin America” conference at Wesleyan and Yale in May 2017. The discussion in this paragraph is indebted to their theorization of the term. See also Scott, James C., Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed (New Haven, CT: Yale University Press, 1998)Google Scholar.

10. Emphasis original, Loveman, National Colors, 5.

11. Kuenzli, Acting Inca, 61; Larson, “Redeemed Indians, Barbarized Cholos,” 231–33.

12. Kuenzli, Acting Inca, 8.

13. Kuenzli, Acting Inca, 75; Larson, “Forging the Unlettered Indian,” 139, 143, 146–49.

14. Oficina Nacional de Inmigración, Estadística y Propaganda Geográfica, Censo general de la población de la República de Bolivia según el empadronamiento de 1o. de septiembre de 1900, vol. 2 (Cochabamba: Editorial Canelas S.A., 1973 [1902-1904]), 35. All translations are my own unless otherwise noted.

15. Larson, “Forging the Unlettered Indian,” 145–46.

16. Service was first declared obligatory by law on August 6, 1875, but was reformed and emphasized in the early twentieth century. For more on the project, see Elizabeth Shesko, Conscript Nation: Coercion and Consent in the Bolivian Barracks (Pittsburgh, PA: University of Pittsburgh Press, 2020), chapter 1.

17. Law of January 16, 1907; “Reglamento del servicio militar,” Boletín militar del Ministro de Guerra, vol. 3 (La Paz: Intendencia de Guerra, 1907), 121–48. For commentary, see “La nueva ley militar,” El Comercio, January 20, 1907.

18. Ismael Montes quoted in James Dunkerley, Orígenes del poder militar: Historia política e institucional del Ejército Boliviano hasta 1935, trans. Rose Marie Vargas (La Paz: Quipus, 1987), 86.

19. Ismael Montes quoted in “La nueva ley militar,” El Comercio, January 20, 1907.

20. Presidential decree, January 28, 1907, Boletín Militar 1907, 38. For further discussion, see Shesko, Conscript Nation, chapter 1.

21. See numerous examples in correspondence and testimony between 1912 and 1947. Prefecture-Admin boxes 147, 148, and 190, Archivo de La Paz [hereafter ALP]; ABG-05-010, DES-16-023, DES-16-027, DES-22-003, DES-23-005, Tribunal Permanente de Justicia Militar – Archivo Histórico Militar [hereafter TPJM-AHM].

22. Recommendation of My. Alcibiades Antelo, November 25, 1932, DES-16-023, TPJM-AHM.

23. Prefect of La Paz to Intendente de Policia Seguridad en La Paz, March 4, 1911, Prefecture-Exped, box 178 d. 120, ALP.

24. Ismael Montes quoted in Dunkerley, Orígenes del poder militar, 86.

25. Memoria que presenta el Ministro de Guerra de Bolivia Doctor Andrés S. Muñoz ante el Honorable Congreso Ordinario de 1910 (La Paz: Intendencia de Guerra), 7.

26. Maj. D. Michel. “Por la instrucción del soldado,” Revista militar 37 (1907): 270–72.

27. Memoria de Guerra y Colonización 1914 (La Paz: n.p, 1914), 54.

28. Supreme decrees, January 2, 1908, and December 16, 1908, Boletín militar del Ministro de Guerra, vol. 4 (La Paz: Intendencia de Guerra, 1908), 1–2, 359–60; Ministerial resolution of December 28, 1912, in Anexos de Memoria de Guerra y Colonización presentada por el Dr. Juan María Zalles a la Legislatura Oridinaria de 1913 (La Paz: Intendencia de Guerra, 1913), 20; Memoria de Guerra y Colonización 1916 (La Paz: n.p., 1916), 21.

29. Supreme decree, January 2, 1908, Boletin miltiar 1908, 1–2.

30. Ari, Earth Politics; Roberto Choque Canqui and Cristina Quisbert Quispe, Educación indigenal en Bolivia: Un siglo de ensayos educativos y resistencias patronales (La Paz: Unidad de Investigaciones Históricas Unih-Pakaxa: IBIS, 2006); Roberto Choque Canqui and Esteban Ticona Alejo, Jesús de Machaqa: La marka rebelde – Sublevación y masacre de 1921, 2nd ed. (La Paz: CEDOIN; CIPCA, 1996); Gotkowitz, A Revolution for Our Rights; Brooke Larson, “Capturing Indian Bodies, Hearths and Minds: The Gendered Politics of Rural School Reform in Bolivia, 1910-1952,” in Proclaiming Revolution: Bolivia in Comparative Perspective, ed. Merilee S. Grindle and Pilar Domingo (Cambridge, MA: David Rockefeller Center for Latin American Studies, Harvard University, 2003).

31. Prefect of Oruro, Eduardo Diez de Medina, “Informe Departamental 1914,” “Ministerio de Instrucción y Agricultura, Comensado en 14 de Octubre de 1913, Termina en…,” vol. 191, Archivo de la Prefectura de Oruro [hereafter APDO]. Document and translation courtesy of Robert Smale. See also Dunkerley, Orígenes del poder militar, 86.

32. Luis Sierra notes a similar omission in judicial records. Luis Manuel Sierra, “Indigenous Neighborhood Residents in the Urbanization of La Paz, Bolivia, 1910-1950” (PhD diss, Binghamton University, 2013), 196–97. For a comparative framework on this question, see Loveman, National Colors.

33. Minister of war to prefects, January 21, 1907, Boletín Militar 1907, 25.

34. Law of January 16, 1907; law of February 24, 1908.

35. Law of January 16, 1907.

36. “Reglamento del servicio militar,” 138–42.

37. Memorandum from minister of war to prefects, no. 12 quoted in Prefect of Oruro V.E. Sanjines to subprefects, June 6, 1907, APDO. Document courtesy of E. Gabrielle Kuenzli.

38. Memorandum from minister of war to prefects, no. 12.

39. Military registration sheet attached to petition of indígena Isidro Quispe to prefect of La Paz, February 21, 1908, Prefecture-Exped box 164, d. 32, ALP. Although the sheet appears to record Quispe's entry into the army, the accompanying petition, which requested exemption from service, reveals it to have been a registration sheet. During this period, men registered several months before they were required to present for service in the January after they turned nineteen. I have not yet found archival holdings of these registration sheets. Discharge sheets from Chaco War demobilization through the present are housed at the Ministry of Defense's Territorial Registry.

40. Memorandum from minister of war to prefects, no. 12.

41. Rossana Barragán, “The Census and the Making of a Social ‘Order’ in Nineteenth-Century Bolivia,” in Histories of Race and Racism: The Andes and Mesoamerica from Colonial Times to the Present, ed. Laura Gotkowitz (Durham, NC: Duke University Press, 2011), 117–20; Brooke Larson, Trials of Nation Making: Liberalism, Race, and Ethnicity in the Andes, 1810-1910 (New York: Cambridge University Press, 2004), 40, 212–17.

42. For history of the term, see Brooke Larson, Cochabamba, 1550-1900: Colonialism and Agrarian Transformation in Bolivia, expanded ed. (Durham, NC: Duke University Press, 1998 [1988]), 111. For disputes over its early twentieth-century connotations, see Javier Sanjinés, “Subalternity within the ‘Mestizaje Ideal’: Negotiating the ‘Lettered Project’ with the Visual Arts,” Nepantla: Views from South 1, no. 2 (2000); Larson, “Redeemed Indians, Barbarized Cholos.”

43. Barragán, “The Census and the Making of a Social ‘Order’,” 129.

44. Boletín Militar 1907, 300-302; Memoria presentada al Congreso Ordinario de 1908 por el Ministro de Guerra Doctor José S. Quinteros (La Paz: Talleres Gráficos La Prensa, 1908), iv–ix; Anexos a la Memoria de Guerra de 1910, clvii-clix; Anexos de la Memoria Presentada a la Legislatura de 1911 Servicio de Guerra (La Paz: Talleres Gráficos La Prensa, 1911), 136–37; Anexos de la Memoria de Guerra presentada por el Coronel Julio La Faye a la Legislatura Ordinaria de 1912 (La Paz: Intendencia de Guerra, 1912), 85–87.

45. See the tables from each department listing men who were chosen by the lottery but did not report for service, which included their name, province, city, and profession. Boletín Militar 1907, 169–226.

46. Barragán, “The Census and the Making of a Social ‘Order’,” 129.

47. This is not to suggest that these terms are synonyms. Agricultor and labrador are more general terms that I understand to be basically synonymous. Colono, peón, and jornalero are more specific, indicating a more precarious or servile status, but agricultor or labrador could be used in their stead.

48. Military-service booklets of Eloy Suaznábar (1909), SUP-97-001; Aurelio Aramayo (1910), INC-60-011; Antonio Chavez (1914), FAD-35-013; Miguel Contreras (1914), INC-60-004; Andrés Calderón (1914), FAD-35-014; military exemption booklet of Máximo Murillo (1919), FAD-35-005, TPJM-AHM.

49. I have not been able to determine precisely when these new booklets were first issued. The last booklet of the older style that I found was from the 1919 cohort and the first new one was from 1927. For examples of the three new styles, see 1927 military service booklet of Félix Tila Tintaya, DES-16-013; 1935 demobilization booklet of Enrique Angles Peñaloza, DEL-14-004; and 1942 military service booklet of Félix Vásquez P., ACC-07-006, TPJM-AHM.

50. The new booklets from the 1920s eliminated space for recording race, face shape, mouth shape, and facial hair. They retained descriptions of conscripts’ skin color, eye color and shape, nose size and shape, hair color and type, and identifying marks. The method of recording also changed. Whereas the earlier booklets provided a blank line for agents to fill in a free-form answer, the ones from the 1920s provided a short list of options to be circled for everything except identifying marks. Like the booklets from the 1900s and 1910s, the new booklets contained information about height, weight, and thorax measurement but included these data on a medical examination page that added respiratory capacity and charted the difference between incorporation and discharge from the military. Demobilization booklets designed for veterans at the end of the Chaco War recorded the same physical descriptors but reverted to free-form answers. The version of the booklet issued in the 1940s restored description of conscripts’ lips but still did not classify by race. Like in the 1920s, the 1940s booklet had a printed list of descriptors rather than a blank line. Height, weight, and thorax information returned to the physical description page. The biggest change was that the new booklet contained twenty pages of patriotic and legal material before listing the conscript's personal and service-related data.

51. Personal observation from conducting research in this archive.

52. Sample of 2,038 sheets from 1936–1974, Registro Territorial – Ministerio de Defensa Nacional [hereafter RT-MDN]. My sample consisted of approximately fifty sheets from each year. I chose three to six sheets from each of the sampled units for that year. Three sheets were chosen at random, and then I also sampled sheets with a different form than others. This sample was not meant to be representative or used for statistical purposes but rather to gain a sense of the variety of practices and forms used over the years.

53. Service sheets of Miguel Zurita Saravia and Luis Rojas Velarde from the Vegara Artillery Regiment. Both conscripts were incorporated in September 1938 and were discharged in April 1940. Their service information was recorded on demobilization forms, which were in widespread use as service sheets from 1935 to 1940. RT-MDN.

54. Service sheets recording the assignment of César Soliz Lugones (1938), Fernando Calderón Arce (1940), Jaime Guzmán Ituralde (1940), and Eduardo Zuazo Cuenca (1940) to the Ministry of Defense. These sheets were interleaved with those from Regimiento Castrillo, Regimiento Escolta, and Compañía de Comunicaciones, RT-MDN. My methodology led to the inclusion of these four mimeographed documents in my sample. The service sheets are loosely bound or foldered by year and unit served. Therefore, most sheets in each group have the same format. When I sampled these sheets, my research goals included understanding the different paper technologies used and the data collected. I thus randomly sampled sheets that had the same format but also specifically included documents that used a different form.

55. I found two desertion notices (both from 1951) on preprinted forms. The two forms solicited the same information in the same order but had different fonts and designs. Like the mimeographed, typed, and handwritten notices, the preprinted ones included a line for race at the top of the physical description. Desertion notices for José Quispe Huanca, DES-25-006, and for Urbano Olmos Torrico, DES-25-009, TPJM-AHM.

56. Prefecture-Exped box 178, d. 20; Prefecture-Admin boxes 147-149, ALP.

57. DES-21-012, DES-21-015, DES-21-021, DES-22-006, DES-22-007, DES-22-014, DES-22-016, DES-23-007, DES-23-010, DES-23-013, DES-24-001, DES-24-003, DES-24-004, DES-24-005, DES-24-007, DES-24-012, TPJM-AHM.

58. Prefecture-Exped box 190, d. 17, ALP; ACC-07-004, ACC-07-010, HOM-50-002, HOM-50-004, HOM-50-006, MUE-69-010, MUE-69-014, MUE-71-001, MUE-71-006, MUE-71-007, TPJM-AHM.

59. This examination was also informed by the insights of Kathryn Burns, Into the Archive: Writing and Power in Colonial Peru (Durham, NC: Duke University Press, 2010).

60. Some sheets have the medical information in one handwriting, suggesting that the medical professional doing the initial examination filled out that section during the recruitment process. Others have the training and discharge information filled in separately, suggesting that the form was partially filled out upon assignment to the unit, stored locally, and completed upon discharge. However, most of the sheets included in my sample had the same handwriting, ink, or typing throughout. The number of sheets in this category is likely artificially inflated by my methods for data collection. Because I was looking for sheets that had a different preprinted form from those of the rest of the cohort in the unit, such sheets are overrepresented in my sample. I suspect that some of these sheets were filled out later than the other sheets because of damage or loss and that is why they were on a different preprinted form. Sample of 890 service sheets from 1936-1952, RT-MDN.

61. An example may help clarify: All sections of each 1941 sheet from the Campos Regiment have the same handwriting and ink, but those from the same regiment in 1942 have personal information in one handwriting, medical information in another, and service record in a third, suggesting that it was filled out on three different occasions and/or by three different people. Sample of service sheets, RT-MDN.

62. Shesko, Conscript Nation, 130–31, 162–63.

63. Clerks filling out the registration tables in this era commonly filled out the columns for the following data (in order): name, age, marital status, profession, reads?, writes?, height, weight, chest measurement, names and residence of parents or closest relatives, residence, first fitness diagnosis, skin color, hair, eyes, nose, mouth, and identifying marks. The tables also had columns for a second fitness diagnosis, military assignment, military training, mobilization information, pre-military instruction, change of address, and observations. These were almost never filled out. 1947–1952 Military Registers, RT-MDN.

64. Approximately 57,600 names are recorded on 1,200 tables. 1947–1952 Military Registers, RT-MDN.

65. Of the thirty-six sets that included race in this period, twenty-seven came from the Cochabamba department (75 percent). A set is a group of tables recording the men who presented at a particular registration site during the registration period. RT-MDN.

66. If racial data was only being collected at the sites where it was recorded in the military register, then Cochabamba would be overrepresented in the desertion reports that include race. Instead, almost all the desertion reports from this period include race. Most (63%) are from La Paz, but they also report conscripts who registered in Oruro, Cochabamba, Potosí, Chuquisaca, Tarija, and Santa Cruz (listed in descending order by number of reports). Boxes 21 through 25, TPJM-AHM.

67. Slogan explained in Carmen Soliz, Fields of Revolution: Agrarian Reform and Rural State Formation in Bolivia, 1935-1964 (Pittsburgh, PA: University of Pittsburgh Press, 2021), 260n3. For recent works that revise understandings of the 1952 revolution, see Sarah T. Hines, Water for All: Community, Property, and Revolution in Modern Bolivia (Berkeley: University of California Press, 2021); Khol, Indigenous Struggle and the Bolivian National Revolution; Nicole Pacino, “Constructing a New Bolivian Society: Public Health Reforms and the Consolidation of the Bolivian National Revolution,” The Latin Americanist 57, no. 4 (2013); Soliz, Fields of Revolution; Young, Blood of the Earth.

68. More work is needed to understand the overlap and differences between the terms Indian and campesino. Despite the MNR's stated goal of privileging class terms, even a cursory glance at MNR discourse reveals the consistent use of indígena or indio as synonyms for campesino. See Pacino, “Liberating the People,” 1112; Soliz, Fields of Revolution, 183n2. The historical importance of occupation to Indigenous categories, detailed above, suggests that racialized connotations also drove the term campesino. This long-standing link supports Dwight Hahn's assertion that the MNR sought to erase the negative connotations of indio but that campesino retained an ethnic meaning and signaled an Indigenous way of life. Dwight R. Hahn, The Divided World of the Bolivian Andes: A Structural View of Domination and Resistance (New York: Crane Russak, 1992), 3–4. More recently, Matthew Gildner has argued that that the MNR fostered both the identities of campesino and indio as “two mutually-reinforcing images of Indigenous Bolivians,” one coded positively and linked to progress and the other negatively as resisting and holding the nation back. Gildner, “Indomestizo Modernism,” quotation from 25.

69. Robert Alexander, a political scientist whose scholarship and activism shaped US opinion on the revolution, amplified a celebratory narrative that the MNR had successfully turned Indians into peasants. Quoting MNR journalist and historian Luis Antezana Ergueta, Alexander informed readers that the word Indian had “disappeared and was wiped from the language to become a relic in the dictionary.” Luis Antezana Ergueta, Resultados de la Reforma Agraria en Bolivia (Cochabamba: Imp. F.O. Guenca, 1955) quoted in Robert J. Alexander, The Bolivian National Revolution (New Brunswick, NJ: Rutgers University Press, 1958), 75. This narrative was so pervasive that Albert Hirschmann drew on it in his famous 1970 work, which used Bolivia's revolution as an example of successful collective mobility. Albert O. Hirschmann, Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States (Cambridge, MA: Harvard University Press, 1970), 110.

70. Quotation from Javier Sanjinés C, Mestizaje Upside-Down: Aesthetic Politics in Modern Bolivia (Pittsburgh, PA: University of Pittsburgh Press, 2004), 17. Silvia Rivera Cusicanqui's groundbreaking 1984 work forcefully condemned MNR's strategies of rural “co-optation” and “civilizing proposals” in service of a Hispanicized “culturally unified nation.” Silvia Rivera Cusicanqui, Oppressed but Not Defeated: Peasant Struggles among the Aymara and Qhechwa in Bolivia, 1900-1980, [translated from the Spanish] (Geneva: United Nations Research Institute for Social Development, 1987), 60. Waskar Ari fleshed out the effects of MNR policy by documenting the repression of religious practice, men forced to cut their braids and change their dress, and union members rejecting Andean textiles. Ari, Earth Politics, 137–38, 153. In her research on public health initiatives, Nicole Pacino argues that the MNR targeted Indigenous populations for hygienic improvement as part of a move from overt to “silent racism” based on factors that correlated to race. Pacino, “Liberating the People,” 1109, 1112, 1118, 1121.

71. Gildner, “Indomestizo Modernism,” quotation from xiii.

72. Soliz highlights the 1954 decree that Indigenous communities used to win restitution of communal lands. While still emphasizing the MNR's racism and privileging of class identity, she also argues that it facilitated the creation of a powerful rural sector by seceding control of land expropriations to peasant unions that were structured according to Andean worldviews. Soliz, Fields of Revolution, 120–40, 164.

73. Jerry W. Knudson, Bolivia: Press and Revolution, 1932-1964 (Lanham, MD: University Press of America, 1986), 320.

74. For more on the changes in the military and conscription during the MNR period, see Shesko, Conscript Nation, chapter 7.

75. Sample of 1952 service sheets from Andino Detachment, 2nd Military Shipyard, Riberalta Battalion, and Bolívar, Camacho, Campos, and Pérez Regiments. The conscripts in my sample of service sheets from the Montes, Sucre, and Warnes Regiments and the Vergara Battery served the full year. RT-MDN.

76. The archived tables for 1953 reflect the registration of approximately 12,400 men, which was more than in 1952 but fewer than in 1951. I do not know when registration took place in Tarija because the tables from that department were of the old style, which did not provide a place to enter the date of conscription. 1953 Registro Militar, RT-MDN. Many of the men who entered the army in April served only six months while others served up to a year. Sample of seventy service sheets from 1953, RT-MDN.

77. The archived service sheets did not change after the revolution; officials continued to fill out the same instrument that had been in use since 1943 and would continue to be used until 1964. Sample of 2,038 sheets from 1936-1974, RT-MDN.

78. A few copies of the new table appear in the registers for 1945, 1946, 1949, and 1951, but all are from relatively remote sites near the southern border (Choreti, Cuevo, Camiri, Charagua, and Carandaiti) and appear to have been filled out by the same typewriter. I suspect that these tables were inserted later due to loss, damage, or delays in sending to the archive. A few sites continued to use the old table after 1953. The last old tables I found were in the bound volumes for 1955. 1940–1974 Military Registers, RT-MDN.

79. 1953 Military Register, RT-MDN.

80. HOM-50-006, TPJM-AHM.

81. Level of education was recorded in the Military Register but not the archived service sheet.

82. Pérez Paco's form differed significantly from the ones from the late 1930s and early 1940s, discussed above. This makes continuance of clerical practice a less plausible explanation. I suspect the form was developed in 1953 to reflect the new information being included in the Military Register tables introduced that year and the service sheet that been in use since 1943.

83. 1947–1958 Military Registers, RT-MDN.

84. Conscription took place simultaneously throughout the national territory over the course of three days and thus had to be staffed by different people. This is supported by the handwriting and other documentary differences in the Military Register. While almost all tables with race placed it on the far right of the table, they differed in other ways, such as whether to label the column, whether to draw a vertical line to set off the column, and whether to use the word or initial for racial labels. Some clerks used a red pencil to draw in the race column and modified their handwriting for the label so that the column matched the others on the table. Some drew in three columns, one for each of the three races they were recording. This leads me to conclude that they were told to record race under observations but were not given an example or specific instructions.

85. 1955–1974 Military Registers, RT-MDN. Registration in 1959 was scheduled to begin on April 27 but was delayed until mid-May due to a coup attempt by the Bolivian Socialist Falange (FSB) that centered on the Sucre Barracks and Plaza San Francisco. See “El 27 próximo se dará comienzo en toda la rep. al reclutamiento militar,” Última Hora, April 13, 1959, 5; “Se libraron sangrientos combates en varios sectores de la ciudad durante más de siete horas consecutivas,” El Diario, April 20, 1959, 5; date of registration from 1959 Military Register, RT-MDN.

86. Bolivia, Ministerio de Defensa Nacional, Prescripciones para el reclutamiento del año 1943 (La Paz: Intendencia Central del Ejército, 1942). Pamphlet consulted at AHM.

87. I have digitized all available tables of the Military Register for 1940–1974 and would be happy to share these sources with scholars interested in analyzing them.

88. I chose Inquisivi for detailed analysis because it was one of the few registrations sites that had significant diversity in the racial classification of registrants.

89. My sample of 20 percent of the names from each of the registration sites also shows an increased likelihood for registrants classified as Indigenous to be listed as illiterate and/or as agriculturalists, as compared to those classified as Mestizo or White. Literacy by racial classification: White – 94 percent, Mestizo – 85 percent, Indigenous – 72 percent. Agriculturalists by racial classification: White – 15 percent, Mestizo – 25 percent, Indigenous – 55 percent. 1954–1958 Military Registers, RT-MDN.

90. This was primarily due to the larger number of registrants listed as Mestizo. The data still shows a strong correlation between Indigenous status and traditionally Indigenous surnames. Of the seventeen men listed as Indigenous, 82 percent had at least one Indigenous surname as compared to 57 percent of those listed as Mestizo. 1954 Military Register, RT-MDN.

91. In my 20 percent sample of each registration site, 84 percent of registrants identified by race were listed as Mestizo. Although I did not enter data for all 19,769 registrants identified by race between 1954–1958, I did look at all the sets and note the general racial breakdown. The only sites where Mestizo was not the race assigned to the majority of registrants were Quime (La Paz), Aquile (Cochabamba), Salinas de Garci Mendoza (Oruro), Corque (Oruro), Sacaca (Potosí), and Charagua (Santa Cruz). The registrants on these sets represented only 3.7 percent of the total identified by race. 1954–1958 Military Registers, RT-MDN.

92. Dirección General de Estadística y Censos, Censo demográfico, 1950 (La Paz: Editorial “Argote” [1955?]), 117–28.

93. This was particularly striking in Inquisivi's 1954 registration, where only one out of six men classified as White was listed as fit for service (17 percent) as compared to 68 percent of Mestizo registrants and 88 percent of Indigenous registrants. In my larger sample of 2,119 registrants with racial information from 25 sites from 1954-1958, this pattern holds in a less extreme way: 65 percent of White registrants, 81 percent of Mestizo registrants, and 86 percent of Indigenous registrants were declared fit for service. 1954–1958 Military Registers, RT-MDN.

94. HOM-50-006, TPJM-AHM. Paco is an Aymara surname that means alpaca. My thanks to Waskar Ari for help in identifying Indigenous surnames.

95. Very few of the military service sheets in my sample from this period report that the conscript spoke only one or more Indigenous languages. The vast majority of service sheets present conscripts as bilingual or trilingual. Many conscripts who likely would have been categorized as White were listed as speaking Indigenous languages. Of the 316 service sheets in my sample from 1953–1957, only twenty-seven recorded the conscript as having white skin. Of those, 52 percent were reported as speaking at least one Indigenous language in addition to Spanish. RT-MDN.

Figure 0

Figure 1 Military Registration Sheets from 1905 and 1923Source: September 1905 military conscription sheet with line for social class attached to petition of “indígena Isidro Quispe” to the prefect of La Paz, February 21, 1908, Prefecture-Exped box 164, d. 32, ALP; September 1923 military registration sheet without line for social class attached to petition of Ernesto Bedregal to the prefect of La Paz, September 21, 1923, Prefecture-Exped box 254, d. 11, ALP. Reproduced with permission by the Archive of La Paz.

Figure 1

Table 1: Aggregate Professional Categories Used in Reports on Conscripts, 1907-1912