In 1847, Gaetano Osculati, an Italian adventurer, walked from Quito across the Andes, then headed south to the Napo River, and finally paddled by canoe to Brazil (Map 1). Upon his return to Italy, he published a widely read account of his travels (2003 [1854]). Other narratives soon followed: by Manuel Villavicencio (1858) ; Manuel Almagro (1866) ; James Orton (1871) ; Alfred Simson (1886) ; Marcos Jiménez de la Espada (1998 [1927-1928]) , to name a handful of authors. Almost invariably, they lamented the region’s lack of commerce and modern agriculture, the strangeness of its native people, the incredible riches going to waste, the backwardness of its ruling class. “This portion of the tropics abounds in natural resources which only need the stimulus of capital to draw them forth,” wrote the Rev. J. C. Fletcher in his introduction to Orton’s book (1871: xvii) .

Similar tropes survive today: the Amazon’s marginality, its abandonment by distant governments, the isolation of its indigenous people ( Perreault, 2002: p. 31 ; Sawyer, 2004: p. 40 ; Yashar, 2005: p. 111 ; López & Sierra, 2011 ). “[D]espite 400 years of Spanish incursions and several decades of republican administration under the Ecuadorian government,” writes Perreault (2002: p. 31) , “inhabitants of the Upper Napo remained relatively isolated from the rest of Ecuador―physically and socially―due to scanty communication networks and administrative negligence.” According to this common view, fundamental change began only in the 1960s and 1970s, when government-sponsored land reform, oil extraction and colonization brought about “rapid integration with market forces” and “a disruption of indigenous lifeways” ( Perreault, 2003a: p. 74 ; Perreault, 2003b: p. 104 ; see also Hutchins, 2007: p. 79 ). Yet such tropes seriously distort our view of the region and its inhabitants. They ignore the push and pull of external forces along Ecuador’s Amazonian frontier over the past 150 years. As a result, recent ethnography often fails to explain fully how native identities emerged, what they signify and how they are changing today.

Nineteenth century travelers provided a more nuanced view of Amazonia. They were not simply adventure tourists, plant collectors or gentlemen dilettantes (who came later, in the 1920s). Instead, they frequently told us a great deal about the far-flung trading networks linking native people in the Upper Marañón Valley with Quito and southern Colombia. Such networks became a major organizing principle of indigenous life, because they largely shaped the social, economic and cultural roles available to Amazonian Indians. As they expanded or contracted over the years, native populations, territories and identities merged or differentiated along with them. Zápara, Quichua, Shuar and Achuar, Cofán, Siona and others were defined as much by their place within these networks as they were by their native languages or cultural practices. According to Hill (1999: p. 704) , “there were, and still are, pockets of remotely situated territory where indigenous peoples lived in relative isolation from the independent nation-states in the nineteenth century. However… [they] had all adapted, either directly or through the mediation of other indigenous peoples, to long processes of conquest, missionization, and other forms of colonial domination prior to the rise of independent nation-states.” With the help of newly discovered archival records, it is now possible to reconstruct at least in part these historical processes and evaluate their impact on indigenous populations into the present century.

Most recent anthropological studies in Ecuador’s Oriente have focused on cosmology, worldview and other cultural issues (for examples, see Uzendoski, 2005; Kohn, 2013 ). Here we problematize a different question: the constant reshaping of ethnic boundaries and indigenous populations from Independence in 1822 to the present day. Specifically, we examine long-term economic and demographic changes within what Ferguson and Whitehead (1992: p. xii) call the “tribal zone,” a geographic and conceptual region “continuously affected by the proximity of a state, but not under state administration.” Within this zone, they write (1992: p. 3), “the wider consequence of the presence of the state is the radical transformation of extant sociopolitical formations, often resulting in “tribalization,” the genesis of new tribes.” Ethnic identities “are in reality entirely new constructions, shifting constellations of compound identities and interests that, in their politically potent actuality, never existed before.” The state itself takes multiple and diverse forms, often represented by “proxies” whose identities shift over time: landowners, government officials, rubber collectors, missionaries―outposts of ruling elites who also compete to exercise governance (see Krupa, 2010: pp. 319-320) . These concepts provide a critical starting point in analyzing present-day indigenous populations in the Ecuadorian Amazon.

Modern commercial networks in the Oriente trace their origins to Jesuit missionaries who extended their reach along the Napo and adjacent rivers during the 17^th and 18^th Centuries. Between 1660 and 1750, the Jesuits organized 74 reducciones (settlements) as far east as the Río Negro in Brazil and as far west as Archidona in the Andean foothills (Taylor, 1999: p. 223) . They imposed “a syncretic native culture transmitted by [mission] Quichua,” as well as shared forms of social organization and religious practice ( Taylor, 1999: p. 227 ; see also Reeve, 1994 ).¹ Disease and forced labor took a horrific toll: within mission territory, native population dropped from around 200,000 in 1550 to 20,000 or 30,000 in 1730 ( Taylor, 1999: p. 225 ; Newson, 1995: p. 81 ). Trade with outsiders was prohibited in theory and difficult in practice. At least the missions offered a degree of protection from Portuguese slave raids and provided sporadic access to metal tools ( Golub, 1982 ). As Taylor (1994: p. 18 ) has pointed out, “Many of the traits that are [today] attributed to their primitiveness―small social and domestic groups, egalitarianism, simple technology―in fact represent an adaptation to the colonial world.” In 1767, the Jesuits were expelled from Spain and its colonies. For the next 80 years, native groups were left largely on their own.

Following independence in 1822 and ensuing civil war, Jesuit-dominated commerce in eastern Ecuador was replaced by civil authorities and a few well-connected merchants ( Villavicencio, 1858: p. 392 ). Three major routes crossed the Andes: the main one that Osculati followed, along with two other little-used tracks farther south (Map 2). Most travelers followed the trail from Quito to Baeza, where it split into two branches. Around Archidona, Quichua-speaking Indians (known as “Quijos”) panned for gold in surrounding rivers, as they had in Jesuit times; along the eastern route, they grew agave (for pita fiber) and tobacco ( Osculati, 2003: p. 153 ). These people were isolated, but not free. Passing through Quijos territory in 1857, Jameson (1858: p. 346) was shocked at the rapacity of repartos imposed by local officials.² “Quito nearly monopolizes the trade,” noted another traveler, James Orton (1871: pp. 194-196) . “It is hard to find an Indian whose gold or labor is not claimed by the blancos.” In 1882, the French diplomat Charles Wiener (2011 [1882]: p. 196) called them “prisoners without uniforms.”

As Osculati quickly learned, the Quijos trails formed only a single node within much broader commercial networks. From Ahuano, four separate tracks led southward through Zápara settlements to the Pastaza River.³ Beyond the Pastaza, Zápara war parties raided Shuar and Achuar groups (then called “Jívaro”), who also supplied blowguns, curare, barbasco (fish poison), ornaments and sometimes a little gold.⁴ In effect, the Jívaro occupied one end of an economic network that defined three distinct ethnic roles: “tame” Christian Indians (Quijos), who lived north of the Napo and worked directly for white merchants; “peaceful but uncivilized” Zápara intermediaries (Orton’s words) living on the frontier who traded in Jívaro territory; and infieles (heathens) or indios bravos, who brought curare, slaves and other things from people living farther down the Marañón Basin.

After mid-century, “peaceful but uncivilized” Zápara began to disappear. This transformation occurred in several stages. Around 1850, Shuar families on the Pastaza River moved north into Zápara territory, where they hoped to “extricate themselves from the continuous persecutions of interior [Jívaro] tribes” ( Villavicencio, 1858: p. 50 ). They settled near San José de Canelos, where white merchants sold tools, cloth, needles, beads, thread and fishhooks brought from Riobamba. Within a few years, 165 households―often speaking Shuar, Zápara and Quichua―lived in the Canelos settlements ( Almagro, 1866: pp. 122-125 ). Meanwhile, semi-independent Zápara groups survived only along the margins of Canelos territory, particularly along the lower Curaray River where they were continually harassed by infieles ( Orton, 1871: p. 220 ).⁵ As early as 1845, Fr. Manuel Castrucci (1849: pp. 12-14) estimated the entire Zápara population at 1,000 people, whereas two centuries earlier, it included at least 35,000 ( Newson, 1995: p. 114 ). By 1887, the Dominican pioneer François Pierre (1983) declared that everything west of the Villano River was occupied by Quichua-speaking Canelos Indians―mostly former Zápara, Shuar and Achuar families who had shifted their ethnic affiliation.

In the early 1860s, international quinine prices rose and a Colombian company sent 1,000 collectors into Ecuador’s eastern forests ( Esvertit, 2008: p. 121 ). By 1875, however, quinine production had shifted to Asia and economic activity in the Oriente largely collapsed. Meanwhile, the Jesuits returned and forced white merchants to leave San José de Canelos, the area’s major commercial center ( Simson, 1886: p. 56 ; Wiener, 2011 [1882]: p. 195 ).⁶ When the English traveler Alfred Simson arrived at Canelos that year, he found its inhabitants eager to trade for his fishhooks, needles and beads, but with little to give him in return (Simson, 1886: p. 100) . He attributed this to laziness; more likely, it reflected the enforced isolation of Jesuit rule and a generalized collapse of commerce. Simson himself (1886: p. 102) noted that Indian men occasionally walked ten days over a nearly impassable mountain trail to Riobamba, where they bought steel spearheads, although he failed to understand why they would do so.

Nonetheless, his account remains useful for us because it provides an important picture of events along Ecuador’s eastern margins just before the Rubber Boom. On the Upper Napo, Simson took refuge with Antonio Llori, a merchant who lived in Ahuano with “a couple of traders [called Quintero] and their wives, with the mother-in-law of one of them (a Záparo woman)…” (Simson, 1886: p. 238). This woman had become an important trader in her own right, especially among her Supuni Zápara kinsmen on the middle Curaray (Simson, 1886: p. 177) . Not long before, he wrote, the Supini had attacked another Zápara group, the Nushinus, “killing many of the men and robbing the women, children and their chattels, the second either for use as servants or for sale.” Children were sold for “a hatchet, a knife, a couple of yards of coarse lienzo [cloth], and a few fish-hooks, needles and thread, or any special article they may most stand in need of…” Evidently, Jesuit-induced scarcity had revived older patterns of warfare among Zápara groups in the tribal zone.⁷

Within ten years of Simson’s travels, the Ecuadorian frontier underwent profound economic and social transformation. During the 1870s, rubber harvesting extended from the lower Marañón into the Napo Basin.⁸ At first, only four white merchants traded for rubber there. But between 1880 and 1890, rubber collection grew from a dozen small posts to 35 large ones (Gamarra, 1996: p. 47) . Most of the remaining Zápara families along the Upper Curaray were soon absorbed into these new enterprises. As caucho collection spread, Quijos and Canelos people were also swept into “the vortex” and dispersed throughout western Amazonia.⁹ Blocked by Peruvians from further expansion toward the Marañón after 1895, many Ecuadorian caucheros shifted their operations to the Napo and Aguarico Rivers (Bravo, 1907: p. 63; Bravo, 1920: p. 90; Alomía, 1936: pp. 307-308) . There they became known as señores ribereños, “lords of the river banks.” By 1900, 72 fundos (estates) were recorded along the lower Napo―double the number of the 1880s (Gamarra, 1996: p. 47) . In 1905, according to Fuentes (1908) , señores ribereños on the Napo sent nearly 35,000 kg of rubber to Iquitos and brought back supplies worth 150,000 soles ($75,000, worth roughly $1.5 million today).¹⁰

Indigenous people on the lower Curaray bore the early impact of rubber collection.¹¹ They included Zápara and Canelos people from Villano and Bobonaza, along with highland Indians, Cofán, Siona and Quichua- speaking people (who usually called themselves runa) recruited from the Upper Napo or fleeing from other caucheros. (Trujillo, 2001: pp. 204-207 ; Porras, 1973 [1905] ; Cabodevilla, 1994: p. 142 ). As rubber collection expanded, indigenous work gangs penetrated further into the forest, where they often fought surviving bands of indios bravos. The consequences for unconquered bands were inevitable: “death or capture for many; fragmentation or flight by survivors into areas held by others, with whom they mixed or fought or finally displaced (Cabodevilla, 1994: p. 160) .”¹²

Once rubber shifted to the Napo, Quijos people around Tena and Archidona were quickly drawn into debt peonage on the fundos. In 1890, 2,000 indigenous families still lived near these towns (Oberem, 1980: p. 48) . But by 1924, Fr. Emilio Gianotti (1997) counted only around 50 runa families there, while another 60 inhabitants were scattered among former Jesuit settlements. Native communities in Loreto, Ávila, Coca and Concepción effectively ceased to exist as their inhabitants moved downriver or onto the San Miguel River.¹³ Meanwhile, Gianotti wrote, 4,700 mostly Quichua peons worked on 55 estates on the lower Napo River with (Map 3).¹⁴

Other than working on the fundos, their prospects were grim. They might become indebted laborers on the ten or so haciendas around Tena that produced aguardiente (rum), mainly for sale to rubber workers (García, 1909) . Or they might become “government people,” often (and misleadingly) called tamberos or indios libres. After 1895, President Eloy Alfaro’s Liberal Revolution brought a new wave of ambitious white officials and traders, eager to exploit the region’s most valuable natural resource: native labor. Under the Ley Especial de Oriente (1900), they commandeered all free Indian men for so-called public works projects.¹⁵ In theory, these men received a daily wage equivalent to US $.40; in practice, they were simply rented to the highest bidder and paid in lienzo (cloth) at inflated prices (if at all).¹⁶ Overwhelmingly, they chose the fundos. Economic life in the Napo river valleylargely became a contest over native workers that pitted señores ribereños against government authorities in Quito and their patronage appointees in Archidona or Tena.

Along the Pastaza and Curaray, Canelos Indians faced similar choices. At first, the Dominicans waged a successful campaign to exclude white merchants and rubber collectors. After 1895, however, Alfaro’s ministers revoked the missionaries’ civil authority and appointed their own officials in Sarayacu, Andoas and San Antonio on the Curaray. In 1906, the district commissioner reported that Canelos and Zápara men spent six to eight months a year along remote rivers harvesting caucho, while their wives and children tended cattle, sugarcane other crops for the ribereños (Bravo, 1907: pp. 61-63) . When rubber supplies in one place failed, indigenous tappers simply moved farther into the forest (Rice, 1903: p. 406) .¹⁷ Meanwhile, mission settlements at Sarayacu and Canelos remained virtually deserted, although Indians working in San Antonio occasionally made the eight-day journey upriver to bury their dead. By the late 1920s, ethnic identities in the Curaray Valley had converged (Cabodevilla, 1994: p. 235 ; Cabodevilla, 2010: p. 38 ; Hill, 1999: p. 741 ). After more than a century, “peaceful but uncivilized” Indian traders―a role filled earlier by Zápara and a few Jivaroans―had become Quichua-speaking peones on rubber estates.

And then, in 1930, international markets for Ecuadorian rubber collapsed. Señores ribereños in western Amazonia begged their buyers in Iquitos and Manaus to keep them afloat. But the markets in New York and London had dried up forever. Among native people, another “radical transformation” of ethnicity and economic roles was about to occur.

In 1924, Fr. Gianotti (1997) counted around 700 Indians living at Armenia, the largest fundo on the Napo, while another 200 collected rubber for its owner on the Ucayali in Peru.¹⁸ By the 1930s, however, travelers and military officers regularly reported that the fundos had largely fallen into ruins (Holloway, 1932; Loch, 1938: p. 96 ; Samaniego & Toro, 1939) . Almost 5,000 ex-peones moved to the Upper Napo, where they settled in areas that had been depopulated a generation or two earlier: Coca, Loreto, Ávila, Tena, and the narrow valleys between Archidona and Baeza. Around 800 of them became conciertos (indebted workers) on sugar and cotton haciendas near Tena and Puerto Napo.¹⁹ But that left nearly 4,000 others with no way to support themselves except subsistence hunting and farming or panning for gold, while they bought supplies on credit from local patrones.²⁰ Pre-existing forms of social and territorial organization had been largely destroyed by a half-century of “forced labor, fractured clans, ruptured families and communities, dispersion to unknown parts of the forest” (Cabodevilla, 1994: p. 143) . To mobilize this new workforce, white officials, merchants and landowners in the foothills set about modifying and reintroducing “traditional” institutions that undergirded an extensive system of labor control.²¹

Most often, Indians were compelled to work by need rather than terror. “Axes, machetes and shotguns had become essential for their own subsistence,” writes Muratorio (1991: p. 78) , “and textiles had become socially indispensable…The systematic use of terror was not required to make them work, although they were never free from abuses and systematic violence.” Violence included public whipping and the stocks, which were still found (half hidden) in Archidona and other places through the 1940s. More often, less overt coercion was used: for example, the Road Conscription Law (1944) and a system of internal passports that restricted Indian movements unless they were travelling on their employers’ behalf.²² Men who resisted found themselves in jail and were forced to work off their fines in projects assigned by the jefe político or his tenientes.²³ “There is not one single free Indian living between the Napo and the Marañón,” wrote Gianotti in 1938 (quoted in Cabodevilla, 1994: p. 148 ). By the late 1940s, according to Hudelson (1981: p. 218) , Quichuas in Loreto and Ávila identified all Indians as gobierno runa (government people) or deudores (debtors), who worked on the haciendas.

For most native people, avoiding forced labor―for example, by fleeing into remote areas―was not an easy option. Within each district (tenencia política), white commissioners named one or two police agents (celadores or guardas) who were responsible for assembling native work gangs and overseeing their tasks: delivering supplies to remote outposts, building roads and bridges, or doing field labor. Generally speaking, celadores depended on indigenous officials known as justicias, capitanes or guaynaros to provide the workers.²⁴ In theory, workers received a fixed wage that was deposited ahead of time in the district office. In practice, white officials usually paid the justicias or capitanes in cheap cloth, beads, needles or other goods for distribution among the work crew. It is difficult to say how much they received or what it was really worth.

In 1937, Shell Oil arrived to explore for oil in the Oriente. At first, it met the governor’s demand to hire manual laborers through their patrones, not on the open market.²⁵ But the landowners often cheated, so Shell hired its own contractor and told him to recruit laborers among “free Indians” in the foothills. Indignant landowners filed a complaint with the district commissioner―who was, after all, also a landowner.²⁶ It is difficult to know how this conflict would have been resolved because Shell ended its operations in 1947 without finding much oil. Yet 30 years later, both Whitten (1976: p. 254) and Muratorio (1991: p. 167) interviewed Canelos and Quijos Indians who recalled being paid the legal minimum wage in cash for the first time while working for Shell or its contractors.²⁷ Among Quijos people in Loreto y Ávila, Hudelson (1981: p. 219) recorded similar memories: “in the early 1940s, many Quichua Indians…enjoyed the novel experience of travelling freely” in the Oriente.

Once Shell departed, white landowners and political officials quickly reasserted their control over the company’s Quichua laborers. Yet Shell had unintentionally set in motion a chain of events that within ten years largely ended concertaje (debt labor) throughout the Andean foothills. Before it left, it was required by the government to build a road from Ambato in the highlands to Puyo, and then onward to the Napo itself.²⁸ Now landowners along the road could raise cattle, tea and other commercial crops for sale in Quito or Guayaquil. As the unpaved roadway advanced toward Tena and Archidona, white residents north of the Napo filed homesteading claims along its projected route; runa families soon followed suit.²⁹

By the early 1960s, white landowners no longer measured their wealth in Indians; instead, they calculated it in pastureland. Far better to hire a few men from “free” indigenous communities when they were needed than to maintain a large indebted labor force.³⁰ Through the 1930s, for example, nearly 1,500 peones worked on Hacienda Ila, which grew sugarcane, cotton and many other products. By the mid-1960s, however, only seven native families still tended 350 head of cattle there (Angel Misueta in Dall’Alba, 1992: p. 166 ).

In turn, indigenous households staked out unclaimed territory and also raised livestock or commercial crops. Legal title to their land was often held by the leaders of each extended family and close relatives, who competed intensely with neighboring groups for favorable boundaries. In each community, one or two family names― marking territorial boundaries―predominated (Dall’Alba, 1992: p. 54) . Rights to gold panning, fishing, gardening and hunting were strictly enforced (Macdonald, 1997: pp. 47-50) . A new phase of “retribalization” had begun, organized around newly empowered elders with access to land and the government officials who controlled property rights (cf. Whitten, 1976: p. 125 ; Macdonald, 1999: p. 54 ).

From the 1930s onward, settlement on the Amazonian frontier closely followed―or rather, anticipated―the expanding road system ( Casagrande et al., 1964: p. 295 ; see also Whitten, 1976: pp. 205-264 ; Macdonald, 1999: p. 57 ). In 1963, Ecuadorian officials drew up ambitious plans to transform the Andean foothills into a prosperous agricultural zone (JNPC, 1963) . By this time, the road from Puyo to Tena―muddy but generally passable― extended toward Baeza and Quito (JNPC, 1963) . Within four years, a consortium of petroleum companies led by Texaco announced that it had found oil near the Colombian border and―responding to pressure from the Ecuadorian government―began building a major highway between Baeza and Lago Agrio. Even before the highway was completed in 1972, it provided access to new areas of settlement in the northern Oriente and allowed the government to implement part of its 1963 plan ( Wasserstrom & Southgate, 2013 ; see also Bromley, 1972: p. 288 ; Robinson, 1971 ). According to Hiraoka and Yamamoto (1980: p. 427) , the highway eventually opened 1.5 - 2.0 million hectares to colonization extending as far south as Méndez in Morona Santiago Province. By the time colonization ended in 1994, around 110,000 people had moved into the rainforest from other parts of Ecuador. In 2010, total population there reached nearly one million (Figure 1).

New roads also brought access to antibiotics, occasional wage labor, canned food, and in some cases, schools and clinics. As a result, Indians in Ecuador, like other native groups throughout the Amazon Basin, began to recover from their long demographic decline (Grenard & Grenard, 2000 ; McSweeney & Arps, 2005 ). Between 1950 and 2010, the indigenous population there grew at an annual rate of 3.5%, from 24,300 to 196,000 people― far higher than the country’s overall increase of 2.5% (Table 1). Even smaller, more isolated groups―the Cofán, the Siona-Secoya and the Waorani―expanded significantly as contact with missionaries and other outsiders became more frequent (Figure 2).

Sources: INEC (1950-2010) ; Cooper in Robinson (1979) ; Fitton (1999: p. 39) ; Robinson (1979: p. 22) ; Ruíz (1997) ; Vickers (1981: pp. 51-61) ; Uquillas (1985: p. 92) ; Vickers (2003: p. 48) ; Summer Institute of Linguistics in Rival (1996: p. 16) ; CONFENIAE (1995) ; CODENPE (2003) .

Throughout the 1960s, competition for land rights and road access among Indians in the Andean foothills intensified. Even before the road to Baeza was completed, migrants from elsewhere began to settle around Tena and Archidona (Macdonald, 1999: p. 87 ; Perreault, 2002: pp. 67-68) . By 1968, indigenous families with few prospects at home left for the northeastern forests and hoped that a road would follow them soon. Yet moving to the rainforest offered only short-term relief. “By the time the road arrives…” write Rudel and Horowitz (1993: p. 133) , quoting an extension agent, “the lands are tired.” Within a few years, Quichua migrants on the frontier were again running out of space. By 1998, nearly two-thirds of migrant households had at least one son who was farming elsewhere (Barbieri & Carr, 2005: p. 100) .

Population growth, competition for land, agricultural expansion and oil development brought new “sociopolitical formations” and ethnic divisions to the Andean foothills. Beginning in the 1960s, native communities formed regional and provincial federations to petition for land or protect it from expropriation (Rogers, 1996: p. 81; Perreault, 2003b: pp. 592-593 ; Erazo, 2013 ). Through the 1970s and 1980s, these communities obtained legal recognition as communes, cooperatives or centros and often received financial support from international donors (Bebbington & Ramón, 1992 ; Perreault, 2003a: pp. 70-71 ; Perreault, 2003c: pp. 340-341 ; Erazo, 2013: pp. 60-96) . Novel lines of authority were drawn as older family heads contested power with younger functionaries holding local or federation offices (Macdonald, 2002 ; Hutchins, 2007 ; Hutchins, 2010 ; Erazo, 2013 ). Previously unfamiliar forms of inequality and factional politics subsumed earlier competition among extended families ( Rogers, 1996: pp. 83-85 ; Perreault, 2003d: p. 105; Luque, 2008: p. 62 ; Wilson, 2002 ; Wilson, 2010: pp. 234-242) .

By the 1990s, communal identities based on family ties had mostly given way to membership in organizations affiliated with the Confederation of Indigenous Nationalities of the Ecuadorian Amazon ( CONFENIAE ), a major political group. Since then, younger Indian leaders have broadened their focus from simply protecting community land to questioning the role of government in modernizing inequality and perpetuating subordination. As Macdonald observes (2002: p. 177) , “They thus shifted course away from requests for government favors, limited participation, and individual land titles, toward a more expanded playing field in which entire groups engaged the state on basic priorities and practices.” Increasingly, ethnic identity has been redrawn along federation and regional lines, not local territories.

In Europe and the People without History, Eric Wolf (1982: p. 353) provided a useful starting point to resolve one of historical anthropology’s key paradoxes: Why did the expansion of European capitalism since the late 18^th century generate seemingly pre-capitalist forms of labor relations―for example, plantation slavery and debt servitude―in other places? The case presented here allows us to reexamine this issue and explore its relationship to identity formation and population growth in northeastern Ecuador.

As Brown and Fernández (1992: p. 176) have written, “…the interest of the state in frontier regions such as the Upper Amazon is typically spasmodic, growing and ebbing in response to large-scale economic processes in the metropole…” Beginning in 1750, disintegration of the Jesuit missions brought about extensive “tribalization” of their former residents (Taylor, 1999: pp. 230-246) . More than a century later, Jiménez de la Espada (1998 [1927-1928]: p. 110) noted the sharp ethnic boundaries that marked his route from Baeza to the Napo; and like most observers, he assumed that they had existed since time immemorial. Quichua-speaking Indians would travel only within their own territories and always carried their own provisions. They neither gave nor expected food from other Quijos, unless they were directly related by kinship.

Such boundaries were swept away by the quinine and rubber booms 30 years later. From 1875 onward, Ecuador’s Amazon region produced quinine, rubber, cotton, sugar, cattle and eventually oil for national and international trade (Cuvi, 2011) . Detribalized Quichua-speakers worked as rubber tappers alongside Zápara, Cofán, Siona, and others throughout the western Amazon. Rubber was a risky business, subject to price fluctuations, high transportation costs, shipwrecks, oligopolistic buyers and other challenges (Barham & Coomes, 1996: p. 68) . With a little bad luck, even white caucheros might end up as indebted peons (Woodroffe, 1914; Yungjohann, 1989 [1916]) . Seemingly archaic modes of production and social relations based on debt (repartos) or unfree labor did not precede commodity booms; they were created by them. In effect, debt peonage spread the risks of rubber collection to indigenous workers―just the way that sharecropping in the Southern U.S. spread the risks of cotton production to poor farmers in the same years (Foner, 1988) .

With little power to bargain or resist, indigenous peones in the western Amazon were especially vulnerable. Their reemergence as distinct ethnic groups after 1930 represents a remarkable example of cultural regeneration. “In each of these cases,” Whitehead (1992: pp. 134-135) has written, “specific tribal identities have been shaped by the slow and tenuous expansion and contraction of …states in the region, notwithstanding their geographic distance or relative isolation from these states.” Eventually, such arrangements underwent further change, as new roads were built and commercial agriculture expanded in the 1950s and 1960s. Even so, debt labor persisted far from the roads until oil companies began to pay the legal minimum wage in 1964 (Beghin, 1963 ; Dall’Alba, 1992: pp. 178-179) .

Ethnogenesis, writes Jonathan Hill (1996: p. 1) , arises from “an ongoing process of conflict and struggle over a people’s existence and their positioning within and against a general history of domination…” As Whitten (2011: p. 9893) points out, similar trends continue today. “In the late twentieth to early twenty-first centuries processes of ethnogenesis are well under way in this region, as myriad people once submerged in the realm of ‘Runa’ cultural orientations again emerge as distinct, although usually Quichua-speaking, peoples...” In lowland Ecuador, ethnic identities have enabled indigenous people to rebuild once the storm tide of gold or rubber or perhaps oil has receded. In more recent times, they have also strengthened native claims to land, political participation and citizenship (cf. Reeve, 2014) . Yet we cannot understand this process without also understanding the competing elites who have dominated Amazonian society: landowners, caucheros, merchants, government officials, missionaries, soldiers, oil men.

Anthropologists have tended to view elites as relatively remote agents of the state―but elite interests were seldom aligned and they often competed fiercely for control of Indian labor. Conflict among elites often shaped the Oriente’s economic and social relations. It is incorrect to attribute such relations simply as the on-again, off-again tug of formal government institutions. Unravelling this complex puzzle remains one of the most difficult and significant challenges of historical anthropology in the western Amazon.

This article is dedicated to the memory of Blanca Muratorio. The authors would especially like to thank Marc Becker, Brian Ferguson, Jonathan Hill and Douglas Southgate, who commented on earlier versions.

Map 1. Ecuador’s Amazon region and contiguous areas.

Map 2. Major trade routes. 1840-1875.

Map 3. Rubber Fundos and Haciendas, ca. 1925.