11
Ritual as Interaction with Non-Humans

Present-Day Mountain-Pass Rituals in the South Andean Highlands

Ethnographic and ethnohistorical studies support the notion that Andean peoples share an animistic ontology (cf. Descola 2013), that is, they believe the world is inhabited by countless non-human agents and interpret numerous phenomena that we (Moderns) attribute to natural mechanisms as effects of their intentional actions. These sentient beings can be a variety of things, including plants, celestial bodies, rocks, artifacts, places, or relatively abstract but omnipresent entities, such as Pachamama. Human life unfolds in a web of heterogeneous social relationships; every task, from building a house to healing, from farming to hosting a feast, requires interacting and negotiating with different kinds of non-human agents through proper ritual protocols (Allen 1988; Martínez 1983; Ricard Lanata 2007; Van Kessel 1988, among others).

Long-distance travel is no exception. Depending on the region and period, travel rituals reported in the literature may engage different kinds of non-humans, including ancestors and accompanying spirits, Pachamama, mountains, outcrops, springs, archaeological sites (chullpas), special objects (e.g., figurines, amulets), pack animals (if they are used), and the road itself (Lecoq 1987; Mariscotti de Görlitz 1978; Nielsen 1997; Pimentel 2009; West 1981, among others). Since these agents have their own characters and can affect people in different ways, they are addressed through a variety of ritual actions that have to be performed in specific contexts.

During their caravan journeys to the eastern valleys, for example, the pastoralists (llameros) of Lípez in the southern Bolivian altiplano conduct different rituals at these places: (1) houses, upon departure and arrival; (2) talvarita, a cairn located on high terrain near the house from where the road is clearly visible (upon departure and arrival); (3) mountain passes; (4) dangerous places along the way, such as narrow passages (punkus) and the beginnings of steep slopes; (5) resting areas located every third or fourth day of the march; and (6) when entering the Tarija Valley—their final destination—from the Sama high plateau. The agents they interact with in each case vary from a single being (e.g., a scary-looking rock) to a host of non-human persons, as in the departure ceremony (figure 11.1), where llamas, caravan emblems (e.g., bells worn by leaders of the caravan, ropes), trade goods, San Rafael (the invisible herder), the road (ñan), Pachamama, the local Mallku (Tres Cerrillos), and the main mountains to be sighted along the way are all invoked and engage in different kinds of exchanges. This translates into considerable variation in the ritual actions carried out in each place, which range from a simple gesture—such as pouring alcohol (ch’alla) or naming while marching—to complex ceremonies that last for hours, as in the main kowaco held near the middle of the journey. In 1995, a caravan we traveled with from the southern altiplano to the Tarija Valley carried two bulky bundles containing ritual paraphernalia to perform the necessary rituals along the journey, including eight kichiras (dried llama hearts stuffed with clippings from different parts of the bodies of sacrificed animals), kowa (an aromatic shrub commonly burned in ritual contexts) incense, lard and cornmeal to shape llama figurines (virauñas), miniature bags filled with maize kernels and cornmeal, flamingo feathers, unkuñas (ceremonial textiles for altars), red yarn to “dress the altars,” coca leaves, alcohol, and wine, among other elements (Nielsen 2011).

Figure 11.1. Invoking the assistance of mountains (Mallkus), San Rafael (the invisible drover), and Pachamama during departure ceremony of a llama caravan in Cerrillos, Sud Lípez, Bolivia, 1995

Mountain passes, or abras, are consistently reported as important ritual places in the Andes (Allen 1988:27; Ambrosetti and Debenedetti 1917:111; Galdames 1990:14, among others). There may be several material traits that justify this or, relationally speaking, “performance characteristics” (Schiffer 2011) that passes “afford” (Gibson 1979) travelers, accounting for this phenomenon. First, they open into a new view, frequently the first or last point from where significant landmarks—potentially perceived as animated, like mountains, caves, towns, outcrops, and lakes—can be sighted. They also divide watersheds; given the importance of water in the South Andean deserts, discontinuities in its behavior are powerful landscape signifiers. For those traveling on foot, mountain passes represent the culmination of particularly strenuous stretches of the road, thus marking a sharp inflection in bodily experience. Finally, the topographic constraints that define mountain passes create high redundancy in traffic, making the archaeological traces of previous transit and of ritual actions that may have been carried out at these points particularly noticeable, thus endowing them with particular mnemonic properties. Given these facts, it is hardly surprising that travelers in other parts of the world have also marked mountain passes as shrines (e.g., Muhonen 2012; Valli and Summers 1994).

The most common ritual presently conducted at abras in the Andes involves adding rocks to the apachetas, cairns developed on the highest point of the trail by the repetition of this gesture over time. Formed in this way, apachetas vary significantly in size and number; passes used intensively or for a long time may have several of these features, which may be up to 8 m high (Boman 1991 [1908]:483). Travelers may also leave wads of coca leaves they have been chewing on the way up (acullico), partake in a short libation (ch’alla), and even “smoke with the apacheta,” often lighting two cigarettes and sticking one of them in the cairn. If time and weather permit, people may stop briefly at this point to catch their breath, enjoy the view, and consume a snack. According to most authors, the deposition of rocks, acullicos, and other objects (e.g., old sandals or clothes) is meant to release the traveler from fatigue, while alcohol and cigarettes are offered to invoke the favors or protection of Pachamama and nearby mountains, or Mallkus (Galdames 1990; Girault 1958). These entities are often addressed through songs and prayers that make explicit the ambiguous character of these agents, which can help or harm people (Mariscotti de Görlitz 1978:69–70), hence the need to negotiate with them socially.

Apachetas are restricted to the Andean highlands, but they are found in other locations in addition to mountain passes, including road intersections, points where two rivers meet (palcas in Quechua), and other places where travelers experience significant changes in the environment—for example, sharp slope changes or locations where a road leaves a plain to enter a narrow valley. This suggests that abras may be pragmatically understood as instances of a more general category of places in which contrasted things are brought together or divided, that is, thresholds or gates—punkus (Cruz 2006)—whose liminality demands similar ritual precaution.

Even when apachetas are invariably associated with roads, they are revered not only while people are traveling. People visit apachetas at other important moments of the annual cycle—toward the end of the rainy season, for carnival (enflorada), and in August, a month devoted to worshipping Pachamama. On these occasions they offer cornmeal, alcoholic beverages, coca leaves, and cigarettes; they “plant” bunchgrasses and branches of different plants (e.g., maize) in the interstices between rocks; and they “dress” the cairns, tying colored yarn around sticks or stones, particularly those that mimic the shape of camelids (figure 11.2).

Figure 11.2. Apacheta in Abra de Sepulturas (Quebrada de Humahuaca) “dressed” during an enflorada ceremony. Note the camelid-shaped rock at the foreground, to the left, which has been ornamented with orange yarn, just like the animals in the corral, which have had their ear tassels (flores) renewed.

The apacheta ritual was described in early colonial sources (De Albornoz 1988 [1584]:168; Polo de Ondegardo 1916 [1571]:189–90). Murúa (2004 [1590]:vol. 2, f. 104v) and Hyslop (1984:311) report Inka ceramics and structures associated with these features. This demonstrates that these practices—or some of them—are of pre-Hispanic origin, at least in the Central Andes. Based on the frequent presence of chronologically diagnostic artifacts next to these features in the Southern Andes, it has been proposed that apachetas date to the time of the Inkas or earlier (Pimentel 2009:15). The lack of conclusive evidence of association between early artifacts and cairns, however, and the absence of these features along important pre-Hispanic roads leave open the possibility that these practices arrived at some regions of the Southern Andes only after the European invasion in the mid-sixteenth century (Berenguer et al. 2005; Nielsen, Berenguer, and Sanhueza 2006).

Offering Pits in Mountain Passes

Over the past twenty-five years we have been able to examine a fair number of mountain passes in a vast area of the Southern Andes, from the Western Cordillera that divides the altiplano from the Atacama Desert to the Eastern Cordillera that dips into the forested valleys of the Yungas region. The archaeological traces found in these small sites show interesting patterns when compared interregionally. Table 11.1 summarizes surface evidence recorded systematically on twenty-two abras spread across the highlands of the “triple frontier” of Bolivia, Chile, and Argentina (figure 11.3). Except for Abra del Altar (3,650 meters above sea level [masl]), all of them are situated between 4,000 m and 5,100 masl. Only passes that present remains other than apachetas—and common refuse generated by their use, such as broken glass or cigarettes—are included to put into focus different ritual practices conducted at these places for which there is little published information. Those shrines located in the Argentine and Bolivian parts of the High Lakes region and along the border between Bolivia and Chile south of Salar de Uyuni were found during systematic survey designed to identify archaeological evidence of trans-Andean traffic (Nielsen 2011). We have also made less systematic observations in other passes, mentioned later in the argument since in some cases they offer important data with which to interpret these contexts.

Figure 11.3. Location of sites mentioned in the text

Note to figure 11.3. Regions: QH = Quebrada de Humahuaca, GC = Doncellas-Guayatayoc, RGSJ = Río Grande de San Juan, SPV = southern Potosí valleys, PNW = northwest Puna, SEL = southeast Lípez, NL = northern Lípez, LR = Loa River, AO = Atacama oases.

One-third of these passes shown on the map have no apachetas; these are mainly on the Western Cordillera or in the Lípez region, and some of them are associated with the Inka road (Qhapaqñan) that links the altiplano with the Atacama Desert. Berenguer and colleagues (2005) also noted the absence of apachetas along significant segments of the Inka road in the Loa River region. Almost two-thirds of the passes in our sample have offering pits (figure 11.4), invariably associated with crushed greenstone, among other items. Offering pits are found in all the cases without apachetas; when both types of features are present, they seem to represent independent contexts and activities. In the wide Abras of Cerro Blanco, Ascotán, and de la Laguna, for example, the offering pits are situated between 40 m and 100 m away from the apachetas and are associated with traces of different trails, possibly used in different periods. The idea of a chronological difference between these practices is supported by at least one observation made at Abra de la Cruz, a pass that connects the headwaters of Quebrada de Humahuaca with the eastern valley of Iruya (see figure 11.3), where road construction sectioned an old apacheta. The profile exposed a layer with greenstone (crushed and worked into beads) beneath the cairn, demonstrating that the apacheta covered an earlier deposit. This case raises the possibility that similar overlaps may have taken place in other passes, where no pits are visible but apachetas occur simultaneous with greenstone offerings.

Figure 11.4. Artificial hollow at Abra del Toro Muerto 1

Offering pits are present all across the Andes at this latitude, a fact that supports their association with long-distance traffic. Pimentel (2009) recently recorded five more cases in the Atacama Desert, including one below 1,000 masl on a route that connects the Middle Loa River with the Pacific Coast. Given the lack of a specifically oriented survey, it is not known at present how far these sites extend to the north and south of our study area. All of the chronologically diagnostic artifacts (decorated pottery, projectile points) recorded in association with these features belong to the Late Intermediate Period (LIP, AD 1000–1450) or the Inka period (AD 1450–1550). This observation is consistent with a radiocarbon date of 800 ± 40 AP (Beta 275739) obtained by Pimentel (ibid.:19) from a wood sample recovered through the excavation of one of these features in the Atacama Desert.

The pits range from 0.5 m to 3 m in diameter and can be up to 1 m deep. We have observed up to thirteen of them aligned on the same pass (Abra del Toro Muerto, Bolivia). They are commonly surrounded by the rocks and dirt extracted from the excavation—often mixed with greenstone and whitestone, attesting to the periodic reactivation of the pits—and they sometimes show a circle made of stones demarcating the depressed interior of the feature. At Abra Lagunita, where the substrate is formed by exposed bedrock, a low (0.3 m) wall was erected to circumscribe an elliptic area, inside of which the offerings were deposited.

Many offering pits show traces of recent excavations, motivated by the widespread belief that these features hide fabulous treasures. According to a legend told frequently by the herders of Lípez, there was an Inka king (Incarrey) who lived on the summit of Licancabur, near San Pedro de Atacama. Since he had no legs, he was carried throughout the region on a litter. This task was so difficult and straining that sometimes, when they reached a mountain pass, his porters would die of exhaustion. The Inka would bury them there with metal treasures as a sign of gratitude for their service. This is why they call these features tapados, or “sepulchers,” although excavations have yet to reveal evidence of the presence of human burials at these sites.

Local people suspect the existence of hidden treasures in offering pits because of the invariable presence of greenstone on their surface. It appears as raw material (from fist-size rocks to tiny pieces) or worked into beads of various shapes and sizes, discarded whole or as fragments. Most of these stones are copper minerals (e.g., malachite, azuryte, atacamite), but petrographic analysis has shown that in some cases they are chert or other rocks with no copper, suggesting that color rather than metal was the signifying quality of these items. Usually, there is also “whitestone,” mostly volcanic ash, which also appears as beads and what seems to be debris generated during manufacture (e.g., cylindrical pre-forms, pieces broken while drilling the central orifice). Indeed, as table 11.1 shows, 65 percent of the beads in the systematically collected sample (N = 178, both colors) are broken (figure 11.5). These materials appear inside and around the hollows but also as light scatters that fan to the east, as if they had been “sprinkled” from above and dispersed by the dominant west winds before reaching the ground. In passes without pits, greenstone and whitestone offerings are usually found around apachetas, but it is unclear if they are associated with these features or were deposited before the cairns were formed. On a few occasions we have also recorded “yellowstone” in the form of unshaped pieces of sulfur (also found by Pimentel 2009 in some of the hollows he studied) and once—in Abra de Tocorpuri—small gold foil trimmings.

Figure 11.5. Common offerings at artificial hollows: greenstone, green and white beads and yellowstone (sulfur). Surface assemblage from Sepulcros de Chillagüita.

Other features found on occasion are circular stone pavements (1.5–3 m in diameter) and semicircular windbreaks (0.8 m high, 2 m in diameter, maximum) expediently made of dry stones. This suggests that travelers occasionally spent some time at passes, building these precarious refuges to protect themselves from the gelid wind. In Paso del Inka, two of the windbreaks are directly attached to the pavement, as if they were meant to protect people carrying out activities on these prepared surfaces.

Other artifacts commonly associated with pits and windbreaks are ceramic fragments, lithic instruments (e.g., projectile points, retouched flakes, and hoes or hoe fragments), and debitage. In some cases these materials may have been obtained in the region where the site is located, particularly in the Western Cordillera where lithic materials are relatively abundant, but the obsidian and gray chert discarded in the sites around Quebrada de Humahuaca came from a considerable distance, hundreds of kilometers away. By contrast, the diagnostic ceramics are usually from adjacent regions, as is the case with the Mallku (northern Lípez) and Loa (Atacama Desert) groups in the western sites and the Yavi (Río Grande de San Juan) and Humahuaca (Quebrada de Humahuaca) sherds recorded on the eastern passes. As expected, Inka ceramics are found in some shrines associated with the Qhapaqñan. We have observed other perishable materials in recently disturbed sites not listed in table 11.1 (Abra del Toro Muerto 2, Vallecito, and Guacha; see figure 11.3), including tropical bird feathers, llama and vicuña wool, pieces of yarn, rope, and textiles.

Discussion: Gesture, Metaphor, and Agency

We believe archaeological research on ritual should start with a “behavioral” focus, emphasizing patterns of bodily interaction between humans and the material world rather than the hermeneutic or discursive renderings of these activities, pragmatics rather than semiology. This should be seen not as a limitation but as an opportunity for archaeology to bring to the fore a distinctive aspect of religious life that is often missed in the textual evidence privileged by the anthropology of religion and ritual under the influence of the linguistic turn. The patterns that can be recognized in ritual deposition, the gestures involved, and the nature of the non-humans abduced (sensu Gell 1998) become apparent at spatial and temporal scales, which are very different from those revealed by ethnography. Ethnographers and historians working with texts or participant observation may find variations in a ritual (e.g., in the interpretations given by different subjects, in the ways things are done each time), which archaeologists get to know through a single material signature.

The presence of pre-Hispanic remains other than apachetas on mountain passes of the Southern Andes has been identified recently, and their variability is just beginning to be explored, so any interpretation of them can only be preliminary at this point. The data presented in the previous section demonstrate that many passes across the Andes between 20° and 24° S Latitude (at least), from the Atacama Desert to the last mountain ranges above the eastern valleys, show traces of ritual deposition—mainly of green and white beads, together with what seems to be the debris produced by their manufacture. This material was frequently sprinkled over, placed in, or buried in pits, often demarcated by stone circles. Other materials that frequently accompanied these offerings were maize, Geophrea and Prosopis fruits, wool (spun or not) and textiles, feathers, marine shell, ceramics, and chipped stone. On occasion, people spent some time at these places carrying out activities that involved flintknapping and the manipulation of ceramic containers, for example, preparing and consuming food and beverages. Stone-paved circles protected by windbreaks in some passes suggest that other activities took place on these features, but there are no elements to define them at this point. These activities took place during the LIP and the Inka period, but they may have started during the first millennium AD and could have persisted into colonial times.

In this part of the Andes, apachetas seem to have been later than offering pits, perhaps introduced from the Central Andes in early colonial times, but the possibility of both ritual practices being contemporaneous should not be ruled out. Moreover, although in some passes apachetas may have covered offering pits—as observed in Abra de la Cruz—the frequent presence of greenstone and beads next to cairns raises the possibility that these were among the offerings made to apachetas as well, at least in some cases or periods. The persistence of similar practices among travelers until colonial times is revealed by blue/green glass beads found in at least two caravan campsites in the area (Patilla and Vega de Tres Cruces; see figure 11.3).

The connection between offering pits and long-distance caravans indicated by the association of these features with important interregional routes is also supported by the invariable presence of greenstone and beads at almost every caravan campsite with pre-Hispanic occupation that has been investigated in this part of the Andes (Berenguer 2004; Nielsen 1997, 2013; Pimentel et al. 2011). Independent proof of the high value travelers assigned to these colored stone offerings comes from a burial excavated at Los Amarillos, one of the largest LIP residential sites known in Quebrada de Humahuaca. The body of an adult male had been tied with ropes and a wooden toggle (distinctive drovers’ artifacts) and interred with a half-gourd containing minuscule fragments of greenstone (atacamite in this case) and white volcanic ash (Angiorama 2007).

Moving away from mountain passes, pits with greenstone and whitestone offerings (i.e., raw mineral and beads) have been found in three other kinds of contexts: caravan overnight stops, water springs, and mid-altitude mountaintops. Offering pits have been recorded at four way-sites associated with interregional routes of the triple frontier. Two of them (Huayllajara and Peñas Negras) are consecutive campsites along a llama caravan trail that connects San Pedro de Atacama with the southeastern corner of the Uyuni salt flat. The other two are next to Inka way stations associated with the Qhapaqñan. One of them is Tambo Chasquillas (east of Quebrada de Humahuaca), located next to a mountain pass; the other is Aguas Calientes (west of San Pedro de Atacama), where the hollow is 100 m away from the buildings, near a water spring.

This brings us to the second kind of location where ritualized pits occur: water springs. This association is also found in Silala 3, a campsite with traces of use since the late Formative period (500 BC–AD 500) located next to a spring on a natural route that connects the southern altiplano with the ancient caravan hub of Santa Barbara (Berenguer 2004). Even the rocks at the place where the water surges have been painted with a green substance or thick paint probably made with copper mineral. Volcanic ash beads have also been found at Baños de Turi, in the Upper Loa region (Gallardo, personal communication, 2009). A variation on this pattern is found at Chillagüita, near the west shore of Laguna Colorada (southwest Lípez), where four large pits with abundant greenstone, whitestone, and yellowstone (sulfur) offerings crown a small hill overlooking the lake.

Relatively low mountains (3,300–4,500 masl) in areas densely settled during the LIP and the Inka period served as stages for similar rituals on their summits. A pit with the usual colored offerings is located on Cerro Agua Colorada, across the river from the town of Rodero (Quebrada de Humahuaca). Systematic survey recently conducted in the southern part of Laguna Pozuelos (Jujuy, Argentina) identified shrines with greenstone offerings on the tops of six highly visible mountains in the area (table 11.2, figure 11.6). These sites are not directly associated with interregional routes but are close to settled areas, so it is likely that the offerings were made by the local population. No chronologically diagnostic artifacts were found in the shrines, but the survey revealed that the main occupation of the region took place in the period AD 1200–1600 (Angiorama 2012). Nearby sites include rock art representations of llama caravans, so it is reasonable to think the local pastoral communities were actively involved in long-distance traffic and may have been among those making offerings in different mountain passes.

Another region where homologous contexts have been found in association with residential settlements is the Middle and Upper Loa River. The features known as “walls and boxes” (muros y cajas) are shrines formed by rows of slab-lined square pits (“boxes”) protected by windbreaks (“walls”) and facing the main surrounding mountain peaks (Berenguer 2004). Some of the offerings found in the boxes resemble those recorded in offering pits—for example, greenstone fragments and beads, whitestone beads, marine shell, feathers, wool, ceramics, lithic debris, and fragments of sandals, among others. Interestingly, these shrines are sometimes found on the outskirts of LIP villages (e.g., Lasana, Santa Bárbara) in direct association with caravan trails that connect the main settlements (Calama, Chiu-Chiu, Lasana) or lead to the altiplano—one of the reasons they have been associated with interregional caravan traffic (Berenguer 1994; Pimentel 2013; Sinclaire 1994). This location is analogous to present-day talvaritas, rock cairns placed on high places near the herders’ houses where family members see the caravan off on the day of departure and come to meet the travelers two or three months later, when they return from the valleys (Nielsen 2001). If the analogy is valid, both talvaritas and walls-and-boxes shrines would operate as thresholds between the domestic space and the road.

What can we learn from these contexts about the entities travelers were addressing in mountain passes and related places and about their relationships with humans? We propose three, non-exclusive interpretations at this point. First, they could mean or anticipate through ritual the union of resources from multiple productive and cultural areas that long-distance traffic intends to achieve (Nielsen 1997:355). This would account for the fact that in the walls and boxes and offering pits with the best preservation of organic remains, the offerings include items that reference every eco-zone of the Southern Andes: marine shell, copper minerals, camelids (dung, fiber, bones, textiles), maize, Prosopis, Geoffrea, feathers of flamingoes and tropical birds, among others.

Figure 11.6. Artificial hollow with greenstone offerings on the top of Cerro Pan de Azúcar, Pozuelos Basin, Jujuy, Argentina

Second, the repetition of the gesture of sprinkling and introducing into the earth greenstone and whitestone beads and the debris produced during their manufacture suggests more broadly that passes, water springs, mountain summits, and walls and boxes had something in common, since they shared the same form of “structured deposition” (Joyce and Pollard 2010; Richards and Thomas 1984). Perhaps they were all conceived as animated entities in themselves, sentient gates or thresholds, punkus between reality domains—that is, different views or “watchers” (Allen 1988:25), divergent water flows, underworld and surface, earth and sky, house and trail. The “ceremonial trash” (Walker 1995) found there would be the material traces of transactions travelers made with these “place spirits,” securing their permission to continue through the gift of goods that were highly valued by humans and non-humans alike.

Another possibility is that these “openings” put people in contact with other beings who were the actual recipients of the gifts or the partners in these exchanges. Focusing on the walls and boxes of the Loa River, for example, Berenguer (2004) interprets copper minerals (greenstone) and other colored offerings as South Andean versions of mullu that people fed to the local mountains (Mallkus) and to Pachamama during ritual meals, committing their reciprocity in the form of health and fertility in general and particularly good fortune for their journeys. But perhaps these liminal places exposed travelers to other, less “Christianized,” whimsical spirits of the underworld, or ukhupacha (see Van Gijseghem and Whalen, this volume), which demanded special ritual protection. Pimentel (2013 :328) notes that the Quechua word for bead (huallca) is related to the term shield (huallccacancca [González Holguín 1952 [1608], libro 4, 172]), both sharing a semantic field articulated around the concept of protection. The only pre-Hispanic shield known from northwest Argentina, found at the late pre-Hispanic site Angualasto (San Juan), is made of Prosopis sp. wood, and its front surface is covered with 3,811 malachite and turquoise beads that contrast sharply with a series of red feldspar plaques that sketch a curvilinear design against the light-green background of the beads, a “ceremonial shield” according to González (1967). The piece was presumably associated with—among other things—wooden snuff trays and related paraphernalia. These trays—perhaps as portals or punkus into other states of consciousness—frequently have green (copper minerals), white (shell), and red (Spondylus) beads as inlays surrounding the cavity in which the hallucinogenic substances were placed.

Could we account historically for aspects of these rituals in terms of a “genealogy of practices” (Pauketat 2001)? Recall that the use of the green-white color contrast goes back—at least—to the early Formative period (1500–500 BC), when communities of pastoralists-hunters like those at Tulán 54 (Núñez et al. 2006) manufactured large quantities of greenstone (copper minerals) and whitestone (marine shell) beads for interregional trade, probably already using llamas as pack animals. Although there is no evidence to sustain the idea that mountain passes were ritualized at this time, it is clear that many of the routes where the late pre-Hispanic shrines described in this chapter are located had been regularly traveled since the late Archaic period—or earlier—by hunter-gatherers who articulated the valleys/oases on both sides of the Andes with the altiplano through seasonal movements (“transhumance” sensu Núñez 1989).

By the end of the first millennium AD, this chromatic structure was transposed to new practices: offerings of green and white huallcas (beads/shields) “to” or “through” artificial hollows placed on mountain passes, summits, springs, and other punkus, perhaps with an ambiguous propitiatory and protective intent. This would be more in accordance with the whimsical nature of pre-Hispanic non-human agents, who were probably less predictable than those who inhabited heaven/hananpacha, earth/kaypacha, and hell/ukhupacha, a cosmological structure that seems strongly influenced by Christianity (Bouysee-Cassagne and Harris 1987). Following this line of reasoning, it seems hardly accidental that this change in the form of deposition of green and white beads took place at approximately the same time agro-pastoral communities were moving from vulnerable to fortified villages and pukaras across the south Andean highlands. The defensive attitude implied by the repetition of this “protective” ritual gesture in every punku along the way or wherever unpredictable evil forces could be stalking people seems attuned to the climate of social unrest that characterized the LIP in the area.

This understanding of things apparently persisted in some parts of the Southern Andes during the Inka period—as attested by the association between offering pits and segments of the Qhapaqñan—and into the colonial era, as revealed by the presence of blue/green glass beads in caravan campsites. It may have even more recent echoes in the aesthetics of two curious offerings we observed during our fieldwork. One of them was found in Pozuelos (San José 155, table 11.2), where among the greenstone spread around an apacheta on the top of a hill, someone had deposited fragments of an unidentified plastic object of the same color. The second was a set of nine plastic beads—two orange, seven green—attached to a pin left on a large apacheta in Abra del Cóndor in the Eastern Cordillera (figure 11.7). What did these gestures mean to those who performed them? What do they tell us about the way such “diachronic homologies” or ritual traditions are reproduced in practice?

Figure 11.7. Offering of plastic beads at the apacheta of Abra del Cóndor, Jujuy, Argentina

The myths that associate offering pits with the Inka—the leg-less Incarrey who honored the service of his porters or the one who hid his treasures from the Spaniards (Pimentel 2009:20)—are good examples of the unique ways these features entangle the practices of current Andean highland travelers. Like other saqra places, these “sepulchers” mobilize the imagination of local peasants, who can spend hours remembering—telling “urban stories” about—old acquaintances who left the community because they contracted a mysterious disease while excavating tapados or found fabulous treasures that allowed them to buy a taxi or set up some other business in the city, where they now lived enjoying their riches.

The archaeological study of shrines and ceremonial refuse can help us explore the heterogeneous sociality that involves a past and present “other” life. We believe that in this way, from the archaeology of ritual, we can contribute significant insights about the nature of non-human agencies in the ancient Andean world. This knowledge would be as important as that provided by ethnography and anthropological history, but it is unique because of its focus on the practical entanglements of cosmology with matter.