Archaeological Insights into Industrial Ruins at the Foot of a Sacred Mountain

Located at an altitude of about 4,000m at the foot of the southern face of Sajama, this ‘Milling Hill’, or ‘Changamoqo’ in the Aymara language, stands about 150 m tall above a generous wetland with water abundant year-round. Ruins of different kinds of structures with diverse construction techniques are distributed in different groups over a total area of 90 hectares (Figure 4). In this first section, we approach this landscape from an archaeological and ethnographical vantage point, which will form the basis for a further historical examination of these ruins and their history.

Figure 4.

Changamoqo architectural complex: site plan (drawn by Thibault Saintenoy)

The principal feature of the site is the hill itself (Figure 1). Its development dates from the pre-Hispanic period and follows a regional tradition of hills crowned with concentric walls for ceremonial purposes.Footnote ¹¹ To this day, the vast platform at Changamoqo’s summit continues to function as a ritual space for the gatherings and feasting of Caracollo’s ayllus under the benevolent gaze of ‘Doctor Sajama’, the principal sacred mountain of the region. Below, the slopes are covered with abandoned small terraces and plots of pre-Hispanic origin, once used for rainfed cultivation of tubers and quinoa.Footnote ¹² The pre-Hispanic occupation left a substantial density of surface artefacts on the hill, particularly ceramic fragments of plates, bowls and cups decorated with black-painted motifs on a red paste (typical of the Altiplano late pre-Hispanic pottery tradition), along with ceramics of various Inca Imperial styles and agricultural stone tools (mostly digging-stick heads). A 2m² archaeological excavation unit (test pit) in one of the most formal terraces at the foot of the eastern slope yielded only late pre-Hispanic domestic material. In fact, post-Hispanic material culture is virtually absent on the hill.

The best-preserved architectural component of the site is situated at the foot of the hill, adjacent to the bofedal. It corresponds to a large enclosure that the villagers call Kuchutambo, or ‘the corner tambo’ (Figure 5). The name is explained by its position, tucked into a rocky wall, and by the site’s former function as the ‘Cosapa market’, with its designation ‘Mercado de Cosapa’ still visibly engraved on the lintel of the entrance portal. Local informants explain that this market operated at the intersection of major long-distance routes between Chile, Bolivia and Peru, which is why it is called a tambo. The massive architectural complex is organised around two courtyards (Figure 5A and B). The main courtyard, measuring 500 m², formed the inner space of the market, enclosed on all four sides by rectangular buildings, each consisting of wide rooms with 2 m-high walls opening on to the courtyard through doors and windows. The second courtyard, located between the main one and the rocky outcrop, extends over 700 m². It is irregular in layout and comprises a variety of buildings, including rooms, storerooms and small structures (likely domestic ovens). Several structures are carved directly into the rocky outcrop, among them a cave room (roughly 50 m², equipped with a trough and a robust closing system), smaller cavities connected to a vertical chimney opening about 8 m higher on the summit, and multiple niches of different shapes and sizes. One particularly striking feature is a cluster of 50 small square niches arranged in a grid of four rows and 13 columns: associated with the remains of wooden beams, these niches are recalled in local oral memory as part of a dovecote, though they may also have originally belonged to a piece of Inca sculptural art, as suggested by Teresa Gisbert.Footnote ¹³

Figure 5.

Kuchutambo, the European commercial tambo: (A) and (B) virtual reconstruction (drawn by Marta Crespo-Fernández); (C) planimetry with location of archaeological excavation units (drawn by Thibault Saintenoy); (D) and (E) stratigraphy of excavation units (drawn by Felipe Carvajal); (F) compacted earth floor (photo by Thibault Saintenoy); (G) multi-temporal ceramic assemblage from late pre-Hispanic to sub-present (photo by Thibault Saintenoy)

Figure 5 Long description

A multi-panel composite covering seven labeled sections, A through G, related to the archaeological site of Kuchutambo. Panels A and B present virtual reconstructions of the large enclosure, depicting rectangular buildings arranged around two courtyards, with walls, doorways and windows visible. The structure sits against a rocky backdrop with an open landscape surrounding it. Panel C shows a planimetry, a top-down architectural plan of the enclosure, with marked excavation units distributed across the complex. Small labeled squares and rectangles indicate the positions of the units within the courtyard areas. Panel D, labeled Changamoup-Kuchutambo comma Unit 4, shows a stratigraphic cross-section with distinct horizontal layers. The layers are labeled as follows: KB-A superficial sedimentation left parenthesis 30 centimeter right parenthesis, KB-B compact earth floor left parenthesis 5 centimeter right parenthesis, KB-B substratum and Rongo′s burning event. Additional markers indicate Wall and fill plaster. Panel E, labeled Changamoup-Kuchutambo comma Unit 3, shows a similar stratigraphic section. Its layers are labeled: KB-A superficial sedimentation left parenthesis 20 centimeter right parenthesis, KB-B housing left parenthesis 15 centimeter right parenthesis, KB-B post-occupational deposit left parenthesis surface right parenthesis, KB-B transportation fill, foundation footing and KB-C. Markers also indicate Raised findings and Trampled surface. Panel F shows a photograph of a compacted earth floor within an excavation trench, with a rounded stone object visible at one end and clearly defined earthen walls surrounding the floor surface. A scale bar is present at the bottom of the image. Panel G shows a collection of ceramic fragments arranged on a flat surface, including sherds of varying sizes and forms, representing multiple periods from late pre-Hispanic to sub-present. A scale bar is visible at the bottom right.

Although surface artefacts at Kuchutambo include a notable proportion of pre-Hispanic ceramic fragments, they mostly consist of Republican period materials, with metal and tin artefacts, glass (with shards of different nineteenth-century European liquor bottles) and industrial ceramics (printed white earthenware and stoneware), with some fragments of maiolica (Figure 5G). Four excavation units of 2 m² were opened in Kuchutambo, in two rooms in each courtyard (‘archaeological units’ 1–4 in Figure 5C). In the main courtyard, which apparently constituted the mercantile space for the storage and exchange of goods, the two pits showed that the floors of the rooms had been laid with compact, yellowish clay (Figure 5F) that was in both cases possibly destroyed by clandestine excavations or post-occupational agricultural sowing. The absence of activity remains would be consistent with its use as a warehouse (emptied upon its abandonment). The pits opened in the rooms belonging to the other courtyard revealed abundant remains of domestic activity, including hearths and rubbish dumps. One of the two excavation units revealed the footing of a wall foundation that corresponds to a typical construction technique of Republican-period industrial architecture. As a whole, the four excavation units at Kuchutambo show that Republican architecture was established in a place with a dense previous occupation of pre-Hispanic origin (and perhaps also Colonial, although the corresponding chronological evidence is scarce) and whose architectural vestiges were razed by the massive modern constructions.

About 500 m downstream from Kuchutambo, in the bofedal, lie two small neighbouring architectural compounds known to local herders as Taypitambo and Lacatatambo (Figure 6). The first is composed of five large rectangular enclosures arranged orthogonally around a courtyard of about 180 m². Although the layout, elongated form and construction technique (double stone walls with corner bonding) might evoke the kallanka Inca planning model, suggesting a pre-Hispanic origin, archaeological excavations revealed only traces of the recent past with evidence of storage and trade in wood and charcoal, activities remembered by a local informant as having been carried out by her grandmother (Figure 6E and G). Known to local herders as Lacatatambo in reference to its recent fire, the second complex is smaller and partially dismantled. Two excavation units inside the main room revealed its modern construction, including a brick-paved floor and a Bolivian coin dated 1918 (Figure 6F and H).

Figure 6.

Taypitambo and Lacatatambo, the Indigenous tambos: (A) and (B) architectural structures (photos by Thibault Saintenoy); (C) and (D) planimetry with the location of archaeological excavation units (drawn by Thibault Saintenoy): (E) and (F) stratigraphy of excavation units (drawn by Felipe Carvajal); (G) coal sample from a coal deposit layer (photo by Thibault Saintenoy); (H) 1918 Bolivian coin of 10 centavos de peso (photo by Thibault Saintenoy)

Figure 6 Long description

A showing an aerial view of an architectural structure with rectangular enclosures. B showing a ground-level view of a similar structure with a mountain in the background. C showing planimetry with marked excavation units. D showing another planimetry with excavation units highlighted. E showing stratigraphy of excavation unit 1 with layers labeled: surface sedimentation, coal deposit and overburden layer. F showing stratigraphy of excavation unit 2 with layers labeled: surface sedimentation, brick floor and wall plaster. G showing a coal sample with a scale indicating 5 cm. H showing a Bolivian coin with a scale indicating 1 cm.

The existence of these two, likely Indigenous, tambos, run by local inhabitants (as remembered by their descendants), is interesting because it contrasts with the administration of the large market of Kuchutambo by European immigrants, something we will address in greater detail later. It is also interesting to note that the excavations indicate the use of the two Indigenous tambos during the twentieth century, while the written records indicate the abandonment of Changamoqo by its European owners at the end of the nineteenth century.

The last architectural component of Changamoqo is a smelting plant placed in the middle of the bofedal and organised around a large courtyard of about 1,300 m² (Figures 4 and 7). The stone and brick industrial infrastructure was built around a water-wheel system that was fed by channelling the bofedal stream and took advantage of the height difference to generate a waterfall. The movement produced by the force of the water activated milling devices and kiln bellows. Imprints on the walls of the rooms located on both sides of the wheel suggest that they housed milling mechanisms with rammers and furnace bellows. Remains of a three-bodied reverberatory furnace (fire chamber, working chamber and chimney) are still visible. In the courtyard, a set of three pools that were fed by recycled water from the wheel were used for the processes of washing and mechanical concentration of the crushed material. The remains of another kiln are found in a 10 m² enclosure that is isolated in the opposite corner of the courtyard. Its smaller size and more basic configuration suggest that this second combustion infrastructure could have been engaged in complementary roasting and/or refining activities. Several stockpiles of minerals and slags were abandoned in a courtyard in front of the windows of the milling rooms, suggesting that they were brought in from another place for their reprocessing in Changamoqo, in order to recover the metals trapped in them. Stockpiles are organised according to size (some with large slags as direct byproducts of reduction activities, and others with smaller ones produced by grinding), and contain mostly tin ores with only a very small trace of metallic copper.

Figure 7.

Interpretative planimetry of the smelting plant (drawn by Thibault Saintenoy and Pablo Cruz)

Figure 7 Long description

The layout of a smelting plant, highlighting various components. Slag heaps are marked in grey, located near the milling rooms, which are shown in brown. The furnace is indicated in red, adjacent to the chimney. Bellows are marked in purple, situated near the furnace. Water pools are shown in light blue, connected to the water wheel, which is marked in dark blue. A water tank is depicted in blue, positioned near the subterranean canal, marked in dark green. The canal runs alongside the plant, channeling water to the wheel. The map provides a detailed view of the infrastructure, including the arrangement of the milling rooms, furnace and water management systems.

Other structures are more difficult to interpret. This is true of the large and elongated structure located above the industrial plant on a plateau (Figure 3), which was connected to a dug-out shaft halfway down the slope, about 10 m below it, as well as of the row of columns, to the north of the courtyard, which ran parallel to and across the stream (Figure 7). The first was likely a sleeve (or shaft) furnace chimney, analogous to those existing in several contemporary Andean mining centres like Potosí’s Cantumarca, for instance, while the second could have been part of an aerial hydraulic (aqueduct-like) infrastructure. Both could have been part of the same system connected to another furnace, although no remnants of such exist. These unknowns and apparent inconsistencies about the configuration of the industrial plant are undoubtedly explained by the fact that, as we will see, this plant was passed between different entrepreneurs who developed it through different projects and innovations, which were not necessarily successful or completed.

In short, both archaeological and ethnographic interpretations of Changamoqo’s ruins point to a nineteenth-century industrial establishment that was built in a bofedal at the foot of a pre-Hispanic ceremonial hill. The Republican-period settlement combined a commercial centre for local and imported products with a medium-scale industrial smelting plant (likely operated by fewer than 50 workers), as well as two isolated small complexes built in stone, possibly related with the Indigenous actors of the settlement. While the density of pre-Hispanic artefacts, even in the subsoil of the Republican structures, might suggest the existence of a pre-Hispanic settlement associated with the hill, the archaeological excavations show that nothing was left of possible structures from that period, which would have been completely dismantled by the massive Republican construction. The hypothesis of a possible Colonial occupation is weaker, due to the scarcity of artefacts exclusively affiliated to this period.

Changamoqo in Written Archives and Oral Histories

Written sources also place the commercial and industrial complex of Changamoqo within the Republican era. Administrative archives, travel accounts and maps make no mention of the name ‘Changamoqo’ prior to the second half of the nineteenth century. The earliest reference is found in the Judicial Archives of Corque (the capital of the province of Carangas, which, until the middle of the twentieth century, coincided roughly with the current political federation of Jach’a Karangas, i.e. the entire western part of the department of Oruro) in a public document from 1865 that records the formation of a company ‘to establish … a commercial tambo between the points of Berros and Tambo Quemado’.Footnote ¹⁴ Though not explicitly named, both the date of the document and the absence of other tambos with Republican-era architecture in the region suggest that the document refers to Changamoqo. It details a commercial venture between merchants in the towns of Corque and Turco, Macsimino Zubieta and the brothers Juan de la Cruz López and Gregorio López respectively, and a miner from Corocoro, José María Loayza. It refers to an import–export trade (more specifically in the purchase and sale of liquors, and the collection and distribution of wool and wool-producing livestock), between the city of Oruro and the Pacific coast, with Changamoqo being strategically placed between these destinations.

In a document from the same archive and dated only a few years later, to 1871, Changamoqo was named as the site of commercial and metallurgical activities. The document described an arson case in the establishment, which was operated by a company of foreign merchants for years prior to the event. The owner was identified as Tomás Lyons, an ‘American’ merchant (though described as ‘British’ in another source) and his associates were Juan Jefferson, an English merchant based in the port of Arica near the Pacific coast, and the Spanish merchant Juan Ortega, who were, respectively, partner and chairman of the company. By this point, Changamoqo had a warehouse for merchandise and goods, as well as a plant for smelting tin.Footnote ¹⁵

In his book Reconocimiento militar del Rio Desaguadero y de la altiplanicie andina, Guillermo Enrique Billinghurst mentions that the Changamoqo mining and metallurgical complex had been in operation since about 1870, and that it was at that time in the hands of Blondel y Compañía.Footnote ¹⁶ This Peruvian company, based in the city of Tacna on the Pacific coast, was managed by the Frenchman Luis Armando Blondel, whose main business was tin, mostly obtained in Oruro.Footnote ¹⁷ We also know that Blondel had hired the aforementioned Lyons as a representative at his company branch in Oruro to handle the purchase of tin, its shipment to Tacna, and all the transportation and financial arrangements that those activities involved.Footnote ¹⁸ So it is probably in this capacity that we find Lyons in Changamoqo in 1871.

In 1882, Blondel y Compañía sold the Changamoqo establishment, with all its fixed assets and stock, to the French businessman Pedro (or Pierre) Dupleich,Footnote ¹⁹ who was the eldest of three brothers (Pedro, Casimiro and Arzenio) and owner of P.E. Dupleich y Compañía, a company dedicated to trading between the cities of Arica, Tacna and Oruro, and to mining. Changamoqo was not the first venture of the Dupleichs in the region; together with the aforementioned López brothers, a powerful family from Turco, they ran a charcoal business in the tambo of Titiri, for which they had been renting the Larancagua hill since 1870 from the caciques and jilaqatas (Indigenous authorities) of Turco for the exploitation of queñua trees (see Figure 2).Footnote ²⁰

Control of charcoal production and trade was key to metallurgical activities.Footnote ²¹ As part of the Changamoqo transfer, Blondel signed a contract with Dupleich for monthly deliveries of charcoal to its main mine at Atocha, to a total of 1,500 quintals (approx. 150,000 kg). The same contract also included an arrangement requiring Dupleich to prioritise Blondel’s products in his transportation activities between Oruro, Changamoqo and Arica.Footnote ²² An 1899 letter from the Blondel house confirmed that Dupleich had delivered all the charcoal promised and could therefore constitute himself as the exclusive owner and proprietor of the Changamoqo complex, eventually acquiring possession from the fiscal (judicial official) and the jilaqata of Cosapa in 1890.Footnote ²³

Upon completion of the Changamoqo transfer, Dupleich equipped the industrial complex with an innovative sleeve furnace.Footnote ²⁴ Contemporary technical sources described the sleeve furnace, together with the use of queñua charcoal, as an adaptive solution for copper smelting at this altitude, which was sub-optimal with simple reverberatory furnaces.Footnote ²⁵ We know that, only a few years before, Dupleich had indeed owned concessions for copper ores in the Turaquiri (or Turuquiri) mine, located on the Turco massif only two days’ travel from Changamoqo, through the Compañía Minera Turuquiri that he had founded between 1882 and 1885 in Tacna with the Frenchman Jorge Dupuy.Footnote ²⁶ Today, families in the Turaquiri region recall Republican-era mining activities in the area and its links to Changamoqo. At Palcoma (on the Turco massif, along the route from Changamoqo to Turaquiri), the archaeological remnants of a tambo – constituting, according to local memory, a fundamental infrastructure of the Turaquiri–Changamoqo route – exhibit an architectural style and a material culture analogous to Changamoqo. That said, however, the paucity of data does not permit an assessment of the real scale of copper smelting at Changamoqo,Footnote ²⁷ and the main business of the Changamoqo smelter during the Dupleichs’ time seems to have continued to be tin.Footnote ²⁸ We know that their company was actively linked to mines in the region for the supply of minerals, especially materials transported by llama caravans from Oruro.Footnote ²⁹ Tin production at Changamoqo mainly focused on poor ores, ferruginous slag with 30 per cent to 40 per cent tin, and tailings from silver processing by amalgamation, using improved sleeve furnaces.Footnote ³⁰ It should be noted that the archival sources concur with the archaeological evidence on the use of hydraulic energy fed by the Sajama river to operate the bellows of the furnaces, which had a capacity of 30 tonnes.Footnote ³¹ According to Roberto Arce Álvarez and Fernando Gautier,Footnote ³² Changamoqo was one of the first industrial establishments for tin beneficiation – the process whereby the metal is concentrated – in Bolivia. Most likely, the Dupleichs sourced slag from wherever they could, rather than relying on direct control over a specific supply source.

Although we know that the Dupleichs established contracts with European companies for the export of this tin, we have not found any data about the commercial outcomes or successes. In fact, Gautier states that Changamoqo produced tin rods of low purity, reaching values of 10 per cent iron, which were not accepted in the European market.Footnote ³³ Nevertheless, what is certain is that the Dupleichs had woven a wide network linking national and international companies, traders and engineers throughout the central-southern Andes and especially between the coast of Arica–Tacna and the cities of Potosí and Oruro.

Regarding the demography of Changamoqo, both the type of trade and industry developed in the complex, as well as the historical documentation on economic activities in this border area, suggest that its operation involved a diverse and not insignificant number of actors in situ. In particular, in view of the size and configuration of the archaeological remains, the metallurgical plant required a significant (more or less specialised) labour force of possibly as many as 50 people. Written sources describe the mill as having a radius of 600 m, containing rooms for the workers and muleteers, as well as rest stops for the llamas.Footnote ³⁴ However, the absence of archaeological remains of individual housing structures suggests that most of the workers did not live permanently within the complex, and may have been mostly Indigenous people from neighbouring towns, or from other mining sites in the region in the case of the more specialised workers.Footnote ³⁵

Information on the abandonment of the Dupleich industrial project is difficult to interpret. Fourteen years after the Dupleichs’ complete acquisition of the business in 1890, the Diccionario geográfico de Oruro situates Changamoqo in a ‘vast and humid pampa’, where it records the existence of ‘a mill for the processing of metals … owned by French industrialists and found abandoned, in spite of the vast and costly installations’.Footnote ³⁶ In his report on the province of Carangas published in 1910, Zenón Bacarreza mentions it as ‘one of the main limestone deposits … currently being exploited’.Footnote ³⁷ Finally, in a 1921 report, Eduardo A. Lima identifies Changamoqo as an active pumice mining operation.Footnote ³⁸ Whatever the facts of the matter, we know that the property remained until 1925 in the hands of the Dupleich family, who declared that no Indigenous community members had ever opposed their possession of it. After Pedro’s death in 1914, his son ended up buying all his shares and rights to the milling house with its machinery and later sold them to José D. Arando for 2,700 bolivianos in 1925.Footnote ³⁹ Two years later, in 1927, Mercedes Dupleich (probably Pierre’s widow) obtained 1,000 m³ of tin slag heaps in the abandoned Changamoqo mill.Footnote ⁴⁰

Bringing together archival and archaeological data, the hypothesis emerges that the abandonment followed a failed expansion attempt. Indeed, a diachronic archaeological interpretation of the remnants of foundry infrastructure suggests a three-step sequence of industrial development. An isolated reverberatory furnace, associated with supplies of tin for smelting, could correspond to the first step, when the site was originally operated by Lyons under Blondel y Compañía (Figure 7). The main infrastructure built around the water wheel, with the sleeve furnace, would correspond to the second step related to the Dupleichs’ takeover. And finally, the enigmatic, likely unfinished, monumental chimney installation on the top of the plateau above the site hints at an abortive attempt to further expand with a larger furnace which, however, never seems to have been operational. Today, the remains of the discarded stockpiles of materials amidst the ruins of the smelter represent the abandonment of industrial activity. This seems to be the end of the short but impressive history of the industrial project in Changamoqo, although the communal memory of the Caracollo ayllu points to the continued resilience of the site as a place of charcoal production and commercialisation, corroborated by archaeological data and the organisation of fairs, until approximately the middle of the twentieth century.Footnote ⁴¹

In summary, written sources indicate that Changamoqo was a pioneering commercial and industrial complex in the region, but that its activity was short-lived, despite coinciding with very crucial decades of ‘the second conquest of Latin America’.Footnote ⁴² From this site, three generations of European entrepreneurs created new dynamics of commercial and industrial expansion on a macro-regional scale, though gaps and inconsistencies in the sources do not allow us to establish their economic success and the reason for the complex’s abandonment. Fundamentally, Changamoqo operated first and foremost as a tambo – i.e. a way-station on a route – that acquired a multifunctional character due to its insertion in complex commercial networks extending between the Pacific and the Altiplano. Because of this, to fully understand Changamoqo one must identify its nodal nature within the regional territorial dynamics in the Carangas Altiplano.

Carangas: Geopolitical Marginality or Macro-Regional Corridor?

Changamoqo is located in an arid region of the western Altiplano, in the Lauca river basin. With an annual rainfall of about 330 mm, water is scarce though sufficient for the spontaneous formation of scrub and Polylepis forests, and snowmelt from the numerous snow-capped volcanoes contributes to the development of extensive humid zones. In this context, agriculture is marginal, and water resources are mainly used to support wetlands for the grazing and raising of camelids (llamas and alpacas). In fact, the region has a very high animal population, which contrasts with low numbers of humans.Footnote ⁴³ Traditionally, the population tends to have a pluri-residential settlement pattern, dispersed between the marka (the main settlement, religious and political headquarters) and isolated single-family estancias in the rural areas that are generally associated with the bofedales.Footnote ⁴⁴ Since pre-Hispanic times, this pluri-residential pattern has also developed on a macro-regional scale with community members settled on the different ecological levels of the western vertical transect stretching towards the Pacific coast.Footnote ⁴⁵ Today, more and more people from the region are settling in the cities and returning to the village only for collective ceremonial festivities.

In addition to its low population level, the Lauca basin is an Indigenous region that is isolated from geopolitical centres and long considered to be marginal in territorial terms. In the Colonial period, it formed a liminal space between the bishoprics of Arequipa and La Plata (coinciding with the corregimientos of Arica and Charcas), a configuration possibly inherited from the Inca territorial partition of the Colesuyu.Footnote ⁴⁶ However, despite its remoteness and distance from the main geopolitical and population centres of the south-central Andes, the Lauca basin formed an important macro-regional mobility corridor.Footnote ⁴⁷ The Colonial Ruta de La Plata (Silver Route) to Potosí – in fact a complex set of overlapping roads and paths – traversed the region and connected the Altiplano with Arica and Tacna.Footnote ⁴⁸ Cosapa, Changamoqo’s nearest town, is mentioned as a main way-station by Antonio Vázquez de Espinosa,Footnote ⁴⁹ whose description of this route is the oldest on record. According to him, Cosapa was connected via the tambos of Copataya and Guanta to Arica, and via those of Berros, Titiri, Churuma etc., to Potosí. More recent travel chronicles mention its extensive bofedales and availability of pack animals.Footnote ⁵⁰ On a more micro-level, it is interesting to note that this section of the route also coincided with a bifurcation of the Altiplano route towards Arica, on the one hand, and Tacna on the other, making it a strategic crossroads on the Ruta de La Plata (Figure 8).

Figure 8.

Regional map of the south–central Andes focusing on the Arica/Tacna–Potosí area (drawn by Thibault Saintenoy)

Figure 8 Long description

The south-central Andes map focuses on the Arica/Tacna to Potosi area. It displays colonial mines marked with triangles and royal treasuries marked with stars. Routes of silver transport are indicated with colored lines, each representing different historical paths. The legend lists routes such as ′Ruta de Cosapa to Desaguadero′ and ′Ruta de Cosapa to Potosi′, among others. Key locations like Arica, Tacna and Potosi are labeled. The map also includes geographical features like lakes and rivers, with names such as ′Lago Titicaca′ and ′Lago Poopo′. The spatial distribution of routes shows connections between major towns and mining areas, highlighting the historical significance of these paths in silver transportation.

The sources for the Arica/Tacna–Potosí Colonial route are: 1543: Vaca de Castro, Ordenanzas de tambos; 1605 (Camino Real Arica–Potosí): Reginaldo de Lizárraga, Descripción breve de toda la tierra del Perú, Tucumán, Río de La Plata y Chile (Buenos Aires: Librería La Facultad, 1916 [1605]); 1629 (Camino Real Arica–Potosí): Vázquez de Espinosa, Compendio y descripción de las Indias occidentales; 1722 (Tacna–Potosí): ABNB, Fondo Audiencia de La Plata (ALP), Recursos documentales sobre la minería, documento 142/19, 13-04-1722: ‘Real provisión de la Audiencia de La Plata a los Corregidores de Oruro, Carangas y Arica: Para apresar a los que pasan a la costa con piñas de plata sin quintar’ (1722–4); 1729 (Potosí–Arica): ABNB, ALP, Recursos documentales sobre la minería, documento 142/23, 11 March 1729: ‘Expediente seguido por el Real Fisco contra Don Antonio Rodríguez de Guzmán, sobre el trato de géneros prohibidos y extracción ilícita de piñas’; 1795 [1633] (Potosí–Arica): ABNB, ALP, Recursos documentales sobre la minería, documento 143/9, año 1795, ‘Expediente seguido para las adiciones a las cuentas de las haciendas de Francisco y Pedro Escudero’ (1633); 1825 (Potosí–Arica): Joseph Andrews, Journey from Buenos Ayres … in the Years 1825–26 (London: John Murray, 1827); 1833 (Potosí–Tacna): d’Orbigny, Viaje a la América meridional, vol. 3; 1877 (Tacna–La Paz): Ephraim George Squier, Peru: Incidents of Travel and Exploration in the Land of the Incas (London, Macmillan, 1877).

From an archaeological point of view, the location of the Colonial tambo of Cosapa remains unclear. It could coincide with the current-day town’s location, or it could be located nearby. Due to the presence of the bofedal and its connection with the roads, Changamoqo could be considered a legitimate candidate for its location. That said, research at the site has identified only a limited number of artefacts from the Colonial tradition and no architectural remains from this period. The most compelling evidence for Changamoqo’s association with the Ruta de La Plata is a road segment that local populations call the ‘international pathway’ or the ‘road to Potosí’ (see Figure 4). This segment, that passes to the south of the hill, can be identified by the intertwined furrows cut into the ground and extending over rocky outcrops, the products of centuries of mule and llama caravan tracks, as told by the locals (Figure 9). A pedestrian archaeological survey of this segment and its projection in both directions has allowed us to identify multiple instances of materials, not only Colonial but also pre-Hispanic, that indicate the long-term use of this road and its only relatively recent abandonment related to the expansion of motorised transport during the second half of the twentieth century.Footnote ⁵¹ Older locals remember travelling the tracks on foot in their youth, with animals, for interregional trips, particularly to Arica.

Figure 9.

Furrows on the ‘international pathway’ being pointed out by local habitants (photo by Thibault Saintenoy)

By selecting Changamoqo to develop their commercial and industrial project, the European entrepreneurs benefited from the significant history of the place as a node along these ancient routes. They also benefited from a certain legacy in terms of the mining–metallurgical economy since, although the Lauca basin did not form a mining district as powerful as the Potosí and Poopó regions, a mining culture had existed in the region at least since Colonial times.

Mining – Heritage and Scarcity

Vázquez de Espinosa’s testimony about ‘Tulco, where there are very rich silver mines, from which a great quantity of virgin silver has been extracted, and if they had not been flooded, they would have been the richest in the Indies’ is one of the oldest references to Colonial mining in the corregimiento of Carangas.Footnote ⁵² The Los Pobres mine was the most famous, where ultra-high-grade silver ores called ‘machacado’ and ‘tacana’ were extracted, but the Turco massif was not the only mining centre of the Colonial period. Silver ores were exploited in Huachacalla and Curahuara, as well as on the banks of the Todos los Santos river, where about 14 silver processing plants had been set up.Footnote ⁵³ Closer to Changamoqo, the archives mention the Choquelimpie ore discovered in 1643 and the Guallatire refining establishment (see Figure 2), already in operation around 1615.Footnote ⁵⁴ In 1652, the Carangas Cajas Reales (Royal Treasuries) were established, hinting at the importance of the region (Figure 8).

Despite the need for labour to exploit its minerals, the corregimiento of Carangas experienced significant depopulation during Colonial times, largely due to the obligation to send forced labourers as tribute to the mines and mills of Potosí.Footnote ⁵⁵ In this regard, Raquel Gil Montero writes that ‘the encomenderos required the service of the Carangas people in the mining companies they had in Porco and Potosí, they were also asked to tend the farms in La Plata, to provide them with domestic labour in the same city, and they were also employed in various jobs’.Footnote ⁵⁶ She even adds that ‘by 1683 there were more Carangas living in Potosí than in the more numerous towns of their home territory’.

Following a boom during the first half of the sixteenth century, mining activity in Carangas was mostly small-scale, intermittent and informal. In fact, the Carangas Caja Real focused principally on mineral production at the Huantajaya site in Tarapacá.Footnote ⁵⁷ Sources record that in the region, by the end of the eighteenth century, most of the mining sites had been abandoned and their infrastructure was in ruins.Footnote ⁵⁸

During the nineteenth century, in postcolonial times, mining activity remained scarce and intermittent. The written sources mention the exploitation of copper mines in Huayllamarca and Choquelimpie by a company with British and French capital.Footnote ⁵⁹ They also refer to initiatives aimed at rehabilitating and using old minerals that had not been massively exploited due to lack of capital or profitability.Footnote ⁶⁰ This was probably true of the Dupleich project at the Turaquiri and Carangualla mines (in the old mining district of Los Pobres). Likewise, the Changamoqo industrial plant was established in a context of infrequent extractive mining activity in the Carangas region, which undoubtedly explains why it is often associated, in the archives, with technological innovations for the recycling of slag.

An Ambiguous Borderland

Towards the end of the nineteenth century, the saltpetre boom in Peru attracted the interest of Chile, which after four years of armed conflict ended up annexing the Peruvian provinces of Tarapacá and Tacna, and the Bolivian province of Antofagasta. The War of the Pacific, which ended in 1883, had the effect of redefining the international border through the bilateral treaties of 1904 (Bolivia–Chile) and 1883 and 1929 (Peru–Chile), and which divided the Lauca basin between two antagonistic nations. When the Dupleichs began the process of buying Changamoqo in 1882, Arica and Tacna were under Chilean occupation and about to formally pass to Chile after the Treaty of Ancón in 1883. For Bolivia, the international conflict meant the loss of its access to the sea and resulted in disputes over land ownership and economic assets between Indigenous settlers, investors and foreign owners – as the Dupleichs also experienced regarding their ownership of Choquelimpie and the Chilcaya borax mines (now in Chile) – and state officials.Footnote ⁶¹ However, historical sources do not provide details about the effects of the war and the reconfiguration of the border on Changamoqo. The geopolitical and economic context in which Changamoqo operated impacted the border in an ambiguous way, simultaneously constraining and encouraging cross-border mobility and exchange.Footnote ⁶² Bolivia and Chile both sought to implement greater border control but lacked the necessary capacity. Meanwhile, several initiatives emerged from economic elites to connect the region to the transportation revolution, though with little success due to the lack of state support and capital.Footnote ⁶³

‘Tambo Capitalism’

The local and regional history of Changamoqo is inscribed within the history of the global development of modern capitalism in Indigenous territories.Footnote ⁶⁴ Its study contributes to the classic, albeit reinvigorated (since the 2008 global financial crisis) academic discussion on capitalism that is conceptualised as a historical system oriented to the endless generation of surplus value, and in particular on its multilinear and hybrid manifestations outside the global North.Footnote ⁶⁵ The visible protagonists of this case study – those whose traces remain in written archives – may well have been mainly European entrepreneurs, yet Changamoqo’s history involved complex and extensive social and economic networks. They connected Indigenous localities with Western projects of exploration and trade, and articulated different sociocultural realities whose interactions contributed significantly to the formation of contemporary Andean society. In particular, the history and archaeology of Changamoqo shines a light on an interesting manifestation of transnational mercantile and industrial capitalist development in the nineteenth-century Andes, and on Latin America’s export age more broadly, roughly 1870–1930.Footnote ⁶⁶ In this specific context, Noa Emrys Corcoran Tadd observes a process of ‘reinvention’ of the tambo as an institution operating within ‘new forms of individual ownership and commercial circulation’.Footnote ⁶⁷ In order to contribute to this discussion, we tentatively propose the concept of ‘tambo capitalism’, which can be defined according to the following characteristics:

• • the adaptation and transformation of the Andean institution of the tambo,

• • the strategic control of transportation and trade routes,

• • the control of trade in both manufactured and handcrafted goods, connecting the global and local economic spheres, and

• • the takeover of mining concessions and industrial innovation for the recycling of abandoned Colonial sites, with the following results:

  • - the development of multisite business networks on a macro-regional scale, hand in hand with

  • - the development of extensive social capital involving and articulating a diversity of local and international stakeholders.

To this day, Changamoqo is often referred to as a ‘tambo’, a designation that alludes mainly to its character as a station along a route, undoubtedly the most generic of the various functions to which the term refers. The tambo is an institution whose function and operation evolved over the centuries, in accordance with historical junctures. Luis Miguel Glave summarises this history by starting with the Inca tambo, an imperial administrative centre whose centralised public management involved the work of state officials and local corvée labourers for the production, storage and transport of goods.Footnote ⁶⁸ In the Colonial period, the tambos were first appropriated by the encomenderos who exploited the institution for their personal benefit,Footnote ⁶⁹ before the crown ordered and promoted their conversion into nodes of a global trade and transport system, as in the time of the Incas, albeit exploited through a mercantile economy as were the Iberian ventas and mesones.Footnote ⁷⁰ Colonial tambos served a variety of different functions, in a more or less centralised way depending on local contexts, and with the generally strong participation of Indigenous people, especially in local supply and manufacturing services, in addition to transportation.Footnote ⁷¹

In this centuries-old trajectory of transformations of tambos and the practices of transport, Changamoqo represents a last stage: that of the appropriation of the institution of the tambo – its ‘reinvention’ – by foreign immigrant entrepreneurs in the Republican context for capitalist ends.Footnote ⁷² In Changamoqo, we can see how the Andean institution was adapted in concrete ways, as a strategy to control commercial networks, both local and global, in which international entrepreneurs, provincial elites, muleteers and mining workers of the region participated and collaborated, along with local Indigenous communities. Rodolfo Boock’s photograph of Pulacayo in the Poopó region show clearly what industrial activity at Changamoqo may have looked like (Figure 10).

Figure 10.

Pulacayo region foundry from the period of 1870–1900 that is contemporaneous and likely comparable with Changamoqo; the picture shows the congregation of muleteers, the mules and llamas used for transportation, the bundles of queñua firewood used in the ovens and the vibrating and washing machines

Source: Rodolfo Boock, Bolivia. Photo album, vol. 2, fo. 15r. Bibliothèque Nationale de France, Département Estampes et Photographie, PET FOL-VH-572.

The history of the actors involved in the industrial project in Changamoqo, in particular that of the Dupleich brothers, sheds light on the dynamics of a business model aimed at a diversification of entrepreneurial activities. This model was developed through the opportunistic exploration and undertaking of endeavours along established trade routes, facilitating the formation of a business network that accompanied a new phase of macro-regional capitalist expansion, and exposes the characteristics we identify under ‘tambo capitalism’.

First, it is noteworthy that the Dupleichs’ initial investment in South America was a transport business between Arica–Tarapacá and Oruro.Footnote ⁷³ Beyond commercial profitability (in terms of access to and the distribution of goods), this activity undoubtedly also made available to them valuable information about the resources and business opportunities scattered over a vast region, particularly its mining and metallurgical potential. Beyond transportation, the Dupleichs sought to revitalise former Colonial-era macro-regional trade routes (such as those to Potosí) through infrastructure investment proposals. For example, in 1887, Pedro Dupleich offered to build a road between Changamoqo and the city of Oruro, in exchange for his exclusive concession for 30 years, ‘with no enterprise that is not authorised by me being allowed to circulate on this road, except for mules, sheep, cattle, horses, coaches and private vehicles, as long as they do not transport merchandise or commercial or industrial products’.Footnote ⁷⁴ The offer was made to the office of the president of Bolivia. Although other actors proposed analogous projects – such as a highway between Arica and Oruro via Chilcaya or an ‘international road’ through Carangas, put forward by the famous ‘tin baron’ Moritz (or Mauricio) Hochschild – no project of this nature came to fruition.Footnote ⁷⁵

Second, both the archaeological evidence found at Changamoqo and the written sources indicate that European entrepreneurs were involved in the trade of both manufactured goods and handicrafts. Hardware and liquor bottle fragments abound in the archaeological record at Changamoqo: these are the same types of imported products as those mentioned in the written sources documenting the principal wares of the trading houses of the time.Footnote ⁷⁶ Here we should note that the Dupleichs were not the only merchants in Changamoqo. In 1891, Eloísa Renjel managed a trading house, hiring freight forwarders to transport products such as alcohol, soap and rice to different points in the province of Carangas, a business on a more local scale.Footnote ⁷⁷ Likewise, Taypitambo and/or Lacatatambo could have belonged to this trader, if she did not use part of Kuchutambo (Figures 5 and 6). Understanding the articulation between the business dealings of European entrepreneurs and other local actors would require further archaeological research, but it is clear that Europeans traded a variety of goods and probably controlled a significant proportion of the market. Indeed, in addition to manufactured goods imported from the ports and trading posts of the Pacific coast, European entrepreneurs also traded local products such as wool and charcoal. As we have seen, the Dupleichs established a major business in queñua charcoal, which was destined both to feed their own industrial activities and to be sold to other entrepreneurs.Footnote ⁷⁸

Third, the sources clearly indicate that mining–metallurgical activity was a central pillar of tambo capitalism, marked by singular features distinctive to the period of industrial innovation that characterised the late nineteenth-century mining boom. The first is the constant search for mining opportunities, even in peripheral or marginal regions, which resulted in numerous speculative concessions that did not necessarily translate into exploitation, such as: the activities of the Dupleichs in Turaquiri, their participation in the Compañía de Choquelimpie, and their borax mines in Chilcaya. That said, we found that the Dupleichs’ main mining–metallurgical activity was the recycling of abandoned Colonial exploitations, particularly from the Poopó region.Footnote ⁷⁹ The sources seem to indicate that most of the materials smelted at Changamoqo were slag and other residues abandoned by the Colonial industry, newly feasible thanks to engineering innovations imported from Europe. This type of operation was undoubtedly profitable, considering that the cost of acquiring these wastes must have been almost nil and that the transportation companies also belonged to the foundry owners. A good example of this is the French engineer Louis Soux, who became associated with the Dupleichs after being lured to Bolivia by the wealthy creole landowner (and later President) Aniceto Arce, famous for having made his fortune from the tailings and slag heaps of the Colonial exploitation of Potosí.Footnote ⁸⁰

Finally, the microhistory of Changamoqo, despite being centred on one place, clearly shows that a central dimension of European industrial capitalism in the nineteenth-century Andes was the development of macro-regional business networks for the integration of marginal areas into new ‘commodity frontiers’ involving extensive networks of actors.Footnote ⁸¹ European entrepreneurs were great travellers who knew how to take advantage of the international borders and, as foreigners, could easily organise activities on both sides of borders and respond to the opportunities that presented themselves; furthermore, they were unafraid of running multisite businesses over long distances. While navigating economic networks, obtaining concessions, establishing partnerships, managing muleteers and llama caravans, and buying and selling interests, the Dupleichs followed a pragmatic ‘poly-business’ strategy.

Thus, Changamoqo was central to diverse operations, between places of extraction and processing of raw materials, trading houses, marketplaces and sources of credit, between Tacna, Arica, Oruro, Potosí, and even France. Integration into these networks required establishing and maintaining strategic personal relationships and increasing social capital. These connections included influential international figures as well as members of the national and provincial elite among whom asymmetrical interdependencies crystallised due to inequalities in political power, cultural status and socio-economic debts. In the first category, large merchants with international capital stand out, like Messrs. Devès frères, who ran a significant French mercantile trading operation, and Blondel y Compañía, as well as European partners such as Jorge Dupuis (or Dupuy; Dupleich’s partner in the Turuquiri mining company) and the afore-mentioned Soux, into whose family Casimiro Dupleich’s daughter married.Footnote ⁸² It was important for the Dupleichs to maintain relations with France, where they obtained credit.Footnote ⁸³ In the second category, it is worth mentioning the López brothers from Turco: this family of vecinos (non-Indigenous provincial elites) participated in the establishment of the commercial tambo at Changamoqo and in Pedro Dupleich’s charcoal business at the Titiri tambo.Footnote ⁸⁴ In other words, these European entrepreneurs did not operate in a vacuum. Just like vecinos in rural villages,Footnote ⁸⁵ the Dupleichs skilfully navigated and wove together the worlds of the Andean community and international capital as a strategy to consolidate their economic and political power in the region, whilst generating asymmetrical power relations. In this sense, these businesses contributed to the formation of Bolivian Republican society of the twentieth century.

However, despite the fact that the labour, knowledge and skills of Indigenous actors were essential to these enterprises, their invisibility in the sources is worth noting: of all the archives we have consulted in this work, only one records Indigenous individuals, specifically as miners, smelters and a store owner who appear in judicial documentation as witnesses before the provincial court.Footnote ⁸⁶ This undoubtedly testifies to the racial segregation of postcolonial society at this time, even in marginal rural regions where the Indigenous population formed a huge majority. At Changamoqo, although Indigenous labour was likely essential to both the operation of the foundry and the transport of commercial and industrial goods, Indigenous agency remains, as explained above, almost invisible in the written sources. Apart from recognition of their possession of the site by the jilaqata of Cosapa,Footnote ⁸⁷ we do not know what kinds of political relationships the European entrepreneurs established with the Indigenous communities living near Changamoqo, nor the basic details of Indigenous workers’ participation in its enterprises. The strongest evidence rather comes from archaeology and ethnography, particularly the existence of Taypitambo and Lacatatambo, both identified by present-day local inhabitants as Indigenous as well as being consistent with Indigenous architectural patterns.Footnote ⁸⁸ Even so, it remains unclear whether these two tambos operated contemporaneously with the European tambo or were founded only after its abandonment. As an aside, the case of Changamoqo’s three tambos also clearly illustrates the semantic shift in the Andean term ‘tambo’, which, by the twentieth century, had come to refer generically to commercial establishments associated with trade routes.

Abandonment and Post-History of Changamoqo: The Industrial History of a National Park

After half a century of documented commercial speculation that was substantiated in massive structures, the Changamoqo industrial project came to an end during the first decade of the twentieth century. It was survived by the charcoal business managed by Indigenous actors, documented in community memories and archaeological remains, until 1939, when Bolivia’s first national park was created around Sajama.

As we have seen, there are no clear explanations or reasons given in the archives for the abandonment of the industrial project nor, in fact, is there any concrete evidence of its success while in operation. The potential reasons are multiple. First, it is likely that the creation of the Arica–La Paz railway during the first decade of the twentieth century had a significant impact on the strategic situation that Changamoqo had held up to that point along the mobility corridor that linked the cities of Arica and Tacna to the Altiplano.Footnote ⁸⁹ Second, the absence of clear evidence of commercial success of the mining–metallurgical operations in the historical sources should be noted as significant, in addition to the low grade of the metals produced in Changamoqo for sale in Europe.Footnote ⁹⁰ Also, the fact that Pedro Dupleich had to mortgage real estate in France to fuel his Bolivian investments could effectively indicate low profitability.Footnote ⁹¹ If so, the Dupleichs could also have favoured other more profitable businesses, such as the mining–metallurgical companies in the Poopó region, where they had invested in several industrial plants (Figure 10). Third, an environmental reason could be invoked that involves the question of how resilient the queñua forests were to intense exploitation. As we have seen, queñua was an essential resource for metallurgical activities, and depletion of the forests could have restricted the processing potential of the plant. This last factor also illustrates how the expansionist impulse of capitalism winds up exhausting its own ecological foundations.Footnote ⁹²

The reasons for the creation of Bolivia’s first national park in the region undoubtedly had different dimensions. From an environmental perspective, the park’s official history mentions the need to conserve the high Andean biotope and especially to protect the ‘highest forest in the world’ of queñua, whose exploitation had become very intensive since the Changamoqo industrial episode.Footnote ⁹³ From a geopolitical perspective, the creation of the national park in Sajama could have been related to the border conflict in existence since the War of the Pacific. According to Bianca De Marchi Moyano, the establishment of the park was a Bolivian nationalist territorialisation initiative in a marginal Indigenous region whose territories and landscapes lacked border infrastructure and national landmarks.Footnote ⁹⁴ From a geo-symbolic perspective, the creation of the park contributed significantly to the conversion of the Sajama volcano into an icon of nature and a national monument. On a local level, although the establishment of the national park may have created certain tensions with the Indigenous communities, particularly when Jach’a Karangas attempted to obtain a single collective land title (the Tierra Comunitaria de Origen or TCO) in 2001,Footnote ⁹⁵ its management combined nature conservation and livestock raising and resulted in a unique model that is valid to this day.Footnote ⁹⁶ Indeed, camelid raising, which constitutes the region’s main economic activity (from pre-Hispanic times to the present), relies on the sustainable reproduction of the bofedales, and favours environmental protection.Footnote ⁹⁷ Furthermore, the Sajama volcano provides a charismatic identity for the national park because it is significant not only in the naturalistic ontological regime,Footnote ⁹⁸ but also in the Indigenous cosmologies that have been transmitted since ancestral times.Footnote ⁹⁹