Tiwanaku urban origins: distributed centers and animate landscapes

JOHN W. JANUSEK

Urban studies tend to emphasize either the centralizing or fluid dynamics of cities. Traditional urban studies, like those of the Chicago school of urban sociology, approached cities as spatially coherent, centralized phenomena or products.

Like all cities, Tiwanaku at any moment was process and product. A long history of incipient urbanism in the southern Andes produced Tiwanaku, and yet, in turn, urban centrality transformed the southern Andes. I tackle two critical aspects of Tiwanaku's emergent centrality. First, I suggest that its early development is best understood as part of a distributed network of Late Formative centers and settlements. Khonkho Wankane and Tiwanaku were sparsely inhabited centers of recurring periodic gathering and ritual activity. Cyclical mobility and regional transaction networks drove the increasing importance of the centers. Furthermore, in their early Late Formative histories, Khonkho and Tiwanaku formed something akin to paired centers on either side of the Kimsachata mountain range, a shared landscape of productive resources. Centers provided recurring gatherings for members of affiliated communities, and themselves comprised multiple geographically separate nodes.

Second, Andean urbanism animated and sanctified specific natural forces, cycles, and features. Centers indexed them spatially, materially, and icono- graphically, animating prominent landscape features and celestial bodies and rendering them critical to the constitution of human and non-human subjects. Political authority crystallized as the strategic mediation of these relations, particularly among those who tended emergent centers, lived near key ceremonial spaces, orchestrated periodic ritual-political events, and “kept time” by tracking complex celestial cycles. To term these foundations of authority “religious” renders them epiphenomenal. Epiphenomenal they were not. Communities and aspiring leaders fostered the animacy of par­ticular forces, cycles, and features, and built them into - embodied them in - their centers precisely because they were considered to constitute the productive bases of life in a challenging high-altitude environment. Tiwa­naku's emergence in the Andean altiplano was founded on an utterly pragmatic, if profoundly ritualized, animistic ecology.

In this chapter I explore the origins of southern Andean urbanism. I sidestep static definitions of what defines a city (minimum size, population, presence or absence of monumental architecture, palaces, corporate art, etc.) by focusing on urbanism as a long-term regional process consisting of multiplex incremental practices. In this perspective, urbanism consisted of the recurring practices that preceded and produced first cities. I begin by summarizing what we know about the city of Tiwanaku. Tiwanaku thrived as an urban center during the Andean Middle Horizon (500-1000 ce). I next explore Tiwanaku's urban origins by explicating the recently investigated Late Formative site of Khonkho Wankane, emphasizing its distributed proto-urbanism as Tiwanaku's precursor and producer. I then explore the reason people came to these centers to begin with, focusing on the import­ance of cyclical social gatherings at built landscapes that facilitated proximity to ancestral monolithic personages.

Tiwanaku: the city (500-1000 ce)

Tiwanaku emerged as a city between 500 and 600 ce in the Andean altiplano or high plateau. For some 500 years hence, it sprawled across the middle of a broad montane valley several kilometers from the southeastern shore of Lake Titicaca (Map ιι.ι). At its greatest extent, in 700-900 ce, Tiwanaku covered 4-6 square kilometers and housed 10,000-20,000 people. Like most

Map ii.i View of the southern Lake Titicaca Basin showing Tiwanaku and Khonkho Wankane in relation to sites and landscape features mentioned in the text.

cities, Tiwanaku comprised built places dedicated to diverse social activities that spanned ritual, residential, and productive concerns. Though certain activities predominated in specific areas and places in Tiwanaku, different classes of activities overlapped in most areas of the city.

Two monumental campuses dominated Tiwanaku's urban core, affording the city a northeast (Kalasasaya-Akapana) to southwest (Pumapunku) ritual axis. Each was a monumental complex comprised of raised platforms, sunken courtyards, and extensive plazas (Figure ii.i).^[CLIV] The northeast campus consisted of two juxtaposed complexes that differ in chronology, spatial organization, and orientation. The Kalasasaya complex includes several structures that were built sequentially over generations: the Sunken Temple - Tiwanaku's earliest extant ceremonial structure - the Kalasasaya platform, and the Putuni complex (Figure 11.2). It manifests paradigmatic spatial patterns of built monumental ritual space, including

JOHN W. JANUSEK

Figure ii.i Plan view of Tiwanaku demonstrating key architectural constructions in relation to key water channels and sunken basins (qochas).

raised platforms with built-in sunken courtyards.

The Akapana and Pumapunku were built just when Tiwanaku was emerging as a primary urban center in the southern Lake Titicaca Basin (500-600 ce). The Akapana was built on the south edge of the Kalasasaya complex, appropriating and aggrandizing its historical legacy. Built in the

Figure 11.2 View of Tiwanaku's northeast monumental complex, facing south. The Kalasasaya complex occupies the foreground, shadowed by the Akapana and Mollo Kontu with its extensive qocha systems (productive sunken basins) further south.

form of a terraced mountain, the structure took the traditional platform- and-sunken-court complex to new heights. Though its upper central area has been gutted, it appears to have once incorporated a high sunken courtyard. Located several hundred meters to the southwest, the Puma- punku was built at approximately the same time. Pumapunku was a lower, more extensive platform that also contained a sunken courtyard. An exten­sive plaza abutted the west side of Akapana and another adjoined the west side of Pumapunku.^[155] These plazas allude to the importance of large-scale social gatherings in association with the new platform structures.

Movement into and through built ritual spaces was critical for experi­encing Tiwanaku and maximizing its impact. Stone portals and portal iconography abound at Tiwanaku and its affiliated centers. The so-called “Gateway of the Sun” (here, the “Solar Portal”) is the best known of numerous lithic portals that directed humans into and through Tiwanaku's monumental ritual spaces. Its elaborate frieze depicts a central ancestral deity standing atop a terraced platform or mountain surrounded by human and bird-headed attendants in procession.

Carved stone monoliths were focal objects of Tiwanaku monumental architecture. Portals bounded from sacral space while affording durable pathways that led toward small sunken courtyards featuring stone monoliths. The inner sancta of sunken courtyards afforded potent ritual experiences. Monoliths erected in these spaces depicted anthropomorphic personages wearing elaborate woven garments decorated with zoomorphic and anthropomorphic imagery. Some sunken courtyards housed single monoliths, while others - notably the Sunken Temple - housed monoliths of different carving styles and periods. Iconographic analysis indicates that the monoliths depicted - or more accurately, embodied - focal ancestors that objectified, and condensed in potent material forms, extensive community affiliations.

TiwanaWs emergence also created productive “rural” landscapes within and outside of the city, in the form of extensive raised field systems on low lake shores and networks of sunken basins - or qochas - in relatively high, drier micro-environments.^[156] Within the city, intricate water networks and productive systems surrounded monumental complexes. These included several channels and perimeter canal that isolated the urban core.^[157] Alan Kolata suggests that this canal was a moat that defined the core as a sacred island. The canal and its water network also had productive roles. Mollo Kontu, a sector of the site linked into the network, incorporated multiple extensive qochas that served intra-urban agropastoral activities. The city showcased production and celebrated water. Neither the urban-rural nor the urban-nature dichotomies that characterize Western practical con­sciousness held for Tiwanaku.

The city incorporated several neighborhoods, each comprised of multiple walled compounds.

Nevertheless, relatively few excavated structures at Tiwanaku can be definitely termed “dwellings.” Many of the structures built within bounded compounds housed temporary city dwellers (Figure 11.3).^[159] Analogous to

Tiwanaku urban origins

Figure 11.3 Plan views of two excavated residential compounds at Tiwanaku, Akapana East ι M (A) and Akapana East ι (B). Akapana East ι incorporated several temporarily occupied structures.

contemporary structures constructed to house intimate visitors during key ceremonial occasions, these structures may have housed visiting family, friends, or business partners during important past ritual events. Further­more, at any time, some 40-50 percent of Tiwanaku's urban periphery consisted of abandoned residential structures and secondary deposition.^[160] The extent of refuse at Tiwanaku led the early twentieth-century Swedish archaeologist Stig Ryden to surmise that Tiwanaku archaeology was pre­dominantly that of past “ritual meals.” Secondary deposition included abun­dant “ash pits”: amorphous, often immense pits filled with a distinctive bluish-gray ash, and containing abundant broken pottery, fragmented came- lid remains, and lithic debitage. While some yield evidence for specialized production, most do not. The ubiquity of these pits supports Ryden's point that, at any time, many areas of Tiwanaku were sites of ritual engagement and, more specifically, ritualized commensality.

I argue that Tiwanaku was a cyclically pulsating “ceremonial urban center” to which people from multiple localities came for important ritual

events that centered on intensive Commensality.⁹ Tiwanaku became the center of a vast, distributed urban network that linked multiple centers and regions in the Lake Titicaca Basin and beyond. What else characterized TiwanakTs social organization?¹⁰ In particular, what were its sociospatial origins? Norman Yoffee argues that many early centers were spatio-material symbols for emergent pan-regional communities.¹¹ In the following section, I suggest that in its historical foundation, Tiwanaku was a center of social, political, and ceremonial convergence. I argue that key transformations in the materiality, spatiality, and lithic iconography of Khonkho Wankane and Tiwanaku jointly produced Middle Horizon Tiwanaka as a primary ritual­political center.

Proto-urbanism in the south-central Andes

Until recently we knew little regarding the generations of practices that produced Tiwanaku. Over the past ten years, research has targeted sites dating to the centuries that directly preceded TiwanakTs emergence, what we term the Late Formative (200 BCE-500 ce). Late Formative ritual-political centers in the Lake Titicaca Basin, including Tiwanaku, manifest distinctive regional patterns. These patterns emphasize the importance of cyclical social conver­gence, landscape production, and specific transformations in spatial and material practices in the production of southern Andean urbanism.

The Late Formative was not the first time that centers had emerged in the region. During the preceding Early to Middle Formative (800-200 bce), a cultural complex known as Chiripa predominated in the southern basin. Settlement clustered along lake shores and perennial streams. The site of Chiripa ultimately became (400-200 bce) the most important center. Its primary temple was a sunken court within a raised platform on which a ring of structures were built. In light of evidence for ritual activities in and human burials under them, Hastorf interprets them as focal shrines for the multiple communities that affiliated with and periodically visited Chiripa for ritual and other activities.¹² Chiripa was a place of cyclical ritual convergence.

IO

II

12

⁹ Janusek, “Residence and Ritual in Tiwanaku.”

Edward Soja, Postmetropolis: Critical Studies of Cities and Regions (Oxford: Blackwell, 2005).

Norman Yoffee, Myths of the Archaic State: Evolution of the Earliest Cities, States, and Civilizations (Cambridge: Cambridge University Press, 2005), p. 16.

Christine A. Hastorf, “Community with the Ancestors: Ceremonies and Social Memory in the Middle Formative at Chiripa,” Journal of Anthropological Archaeology 22 (2003), 305-32.

After 2oo bce, sociopolitical and environmental conditions shifted. During the early part of the Late Formative (LF ι: 200 BCE-250 ce), multiple new centers, each incorporating a distinct type of ceremonial complex, emerged in the southern basin. These include Kala Uyuni on the Taraco Peninsula, Kallamarka and Tiwanaku itself in the Tiwanaku Valley, and Khonkho Wankane in the Upper Desaguadero Basin (Map ιι.ι). While the ceremonial architecture of some centers comprised small ritual chambers - as at Kala Uyuni - others incorporated sunken courts and large open plazas - as at Khonkho Wankane and Tiwanaku.

In the following sections, I focus on Khonkho Wankane and Tiwanaku.

Khonkho Wankane in a multi-centered world

Khonkho Wankane is located in a high (3,900 masl), relatively dry area of the southern edge of the Lake Titicaca Basin. This is a particularly challenging portion of the Andean altiplano. Rather than intensive agriculture, the region lends itself to a greater reliance on camelid pastoralism via domesti­cated llamas and alpacas. There is evidence for (ι) construction of nearby sunken basins (qochas) that served agropastoral practices; (2) exceedingly high frequencies of camelid remains at Khonkho; and (3) a predilection for camelid iconography on stone monoliths at the site.

Khonkho Wankane thrived during the Late Formative period as a ritual­political center. It occupies a portion of the Machaca Plain just below the foothills of the Kimsachata range (Map ιι.ι). It consists of two core mounds - Wankane and Putuni - surrounded by several smaller knolls. Both mounds were artificial platforms raised over natural hillocks during a few massive construction events. Wankane was the primary monumental platform.

Khonkho Wankane's construction demanded tectonic movement and monumental coordination. The platforms incorporated elaborate earthen stratigraphies that simultaneously captured rainfall (via sandy strata) and guided moisture off the mound to prevent erosion (via interlaced clay strata). The platform supported an elaborate ritual-residential complex centered on an extensive Main Plaza for large-scale gatherings. A smaller Sunken Temple on its southwest side provided an intimate space for ritual events, and adjacent ritual-residential compounds demarcated its east and southeast sides (Figure 11.4).

Constructing the two core platforms reciprocally created an anthropo­genic productive landscape. Deep clay quarrying around them created a broad marshy channel fed by two springs in the nearby Kimsachata foothills.

Figure 11.4 Oblique and plan views of Khonkho Wankane's central ritual complex. The image emphasizes key ritual structures, including the Main Plaza, Sunken Temple, and later Dual-Court Complex.

Additionally, activities in the ritual complex themselves supported the platform's surrounding marshy landscape. A centrally located basin in the Main Plaza drained run-off from all of the platform's structures into a massive, hypertrophic subterranean canal that carried water off the mound and into the marsh. The marsh afforded pasture for camelids. Khonkho Wankane incorporated a productive landscape and celebrated water, as did later Tiwanaku. Its encompassing, human-wrought marsh likely inspired the encompassing perimeter canal that surrounded the central portion of Tiwanaku.

Khonkho Wankane was predominantly a place for periodic regional gathering. Deep excavations in the Wankane mound revealed an early, thin occupation dating to ι-ιoo ce. Excavations in the Putuni fill indicated that it covered a contemporaneous occupation consisting of multiple superim­posed, ephemeral surfaces, each covered with a thin layer of erosional sedimentation. Both incorporated short-term hearths and their carbonized cleanings as well as ceramic cookware sherds and splintered faunal remains. Before being covered by thick platforms, early mounds were locales for recurring, periodic gathering and commensality.

Research in the knolls outside of the marshy channel confirms this point. Survey yielded abundant artifacts dating to the Late Formative. We docu­mented a classic two-tier network of inhabited settlements, what we would expect for an incipient ranked society or chiefdom: a multi-community polity. Yet, later excavations in several of the knolls yielded no evidence for permanent in situ occupation, offering suggestive evidence that the thin occupations have been deflated due to ongoing natural (rain, wind) and human (agropastoral) activities. That is, occupations on the peripheral knolls were, like those buried under and preserved by the Wankane and Putuni platforms, ephemeral and cyclical. Like the earliest occupations under Wankane and Putuni, their peripheral knolls housed recurring, temporary encampments.

This finding transforms our understanding of Khonkho in its regional setting and the sociospatial specificity of Tiwanaku urban origins. It indicates substantial cyclical mobility in the region. Foreshadowing later Tiwanaku, not all of those who affiliated with Khonkho or participated in its construc­tion projects and recurring rituals lived at the center. Most did not. People came periodically, perhaps in seasonal or annual ritual cycles, to briefly inhabit the early core mounds and, later, the knolls that surrounded the two emergent platforms. Considering Khonkho was one of multiple centers in the region, Khonkho was less the center of a multi-community polity than it was one of many in a multi-centered macro-community. It was one in a distributed network of interlinked centers in the southern Lake Titicaca Basin.

Evident mobility during the Late Formative suggests some regional coordination among the communities affiliated with these centers. Perhaps each center offered particular times of ceremonial engagement that rotated over repeating temporal cycles and a ritual focus that served as a recurring matrix for communal social, economic, and political activity. Highly mobile networks linked centers in the southern Lake Titicaca Basin. These net­works shaped succeeding generations of Tiwanaku urban hegemony.

Tiwanaku and Khonkho Wankane as paired centers during the Late Formative

Located just across the Kimsachata mountain range, Tiwanaku was an important Late Formative center. Little Late Formative research has been conducted at Tiwanaku, and most early occupations were buried or des­troyed during the Middle Horizon. We know even less about local Late Formative regional settlement, except that it consisted of a number of smaller sites.^[161] Whether they were temporary encampments like those that surrounded Khonkho is a task for future research. Still, the research that has been conducted in and around Tiwanaku indicates that it thrived as a ritual center parallel to Khonkho Wankane during the Late Formative.

There are remarkable parallels between the sites. First, TiwanaWs earli­est known ceremonial space was a Sunken Temple similar in size, form, and orientation to that at Khonkho. At each site, the court maintained a primary stairway entrance in its south wall prioritizing a primary north-south axis and a southward-facing visual path. The primary entrance to both sunken temples is offset from the center of the wall, at Khonkho to facilitate a southward view of Mount Sajama and at Tiwanaku to Mount Kimsachata. The parallel orientation may have simultaneously provided nighttime celes­tial observations. Each courtyard was relatively intimate and connected with an extensive plaza for larger-scale gatherings and events.

Second, stylistically similar monolithic sculptures were focal ritual objects at both centers during the Late Formative, and they collectively differ from those at other centers in the Lake Titicaca Basin. Four are known from Khonkho and four from Tiwanaku. Hewn of bright red sandstone, each monolith depicts an anthropomorphic being - I argue a mythical ancestral personage - with impassive face and arms placed over the chest. Though most were found ex situ, three in situ monoliths indicate that they originally occupied key sunken ceremonial spaces and plazas at each site.

Most telling is the precisely parallel location of the centers on either side of the Kimsachata range. The source of the springs that fed Khonkho Wankane, this range also produced springs and an extensive subterranean aquifer that sustained Tiwanaku. Kimsachata was the source of the red sandstone employed to craft temples and monolithic ancestral personages at both sites.^[162] The shared range was a critical source of vitality for the two centers, a theme that was narrated on their personified monoliths (see below). Most remarkable, the two centers were settled on a nearly precise north-south alignment on either side of Kimsachata.^[163] Their sunken temples share the same longitude (68° 40' 21”, +/- 1”). Thus, the north-south geographical axis of the centers was mirrored in their ritual inner sancta.

Khonkho and Tiwanaku were paired centers during much of the Late Formative. This was another key element of their distributed character and of Tiwanaku's urban foundations. During Late Formative 2, 300-500 ce, that paired relation began to change.

Tiwanaku as emergent city in the south-central Andes

Tiwanaku's emergent urban centrality involved critical transformations in its spatiality, materiality, and iconography. These transformations created novel relations to landscape features and celestial - or “skyscape” - cycles. They were undoubtedly the work of new regimes. At Khonkho, a Dual­Court complex was built on the west side of the early Main Plaza (Figure 11.4). Marking a shift from the early Sunken Temple, which then deteriorated, the complex created a new axis of sight and movement oriented east-west. At Tiwanaku (Figure 11.2), the Kalasasaya was built on the west side of the early Sunken Temple and its adjacent plaza. Like Khonkho's new Dual-Court complex, Kalasasaya created a new axis of sight and movement oriented east-west. By the end of Late Formative 2, the Kalasasaya was an enormous platform construction with megalithic revet­ments. This was an important transformation that helped produce Tiwa- naku as a primary urban center.

Spatial transformations at Tiwanaku

In the later part of Late Formative 2 the Kalasasaya platform was expanded (Figure 11.5). Platform expansion aggrandized old walls, built predominantly of roughly hewn sandstone blocks, and created a westward “balcony wall” girded by eleven impeccably crafted volcanic stone (andesite) pillars. This wall, we now know, served as a celestial observatory. A central stone platform stands some 58 meters to the east. Standing on the platform and facing the wall in the evening, a person during the Middle Horizon wit­nessed the sun setting on the western horizon. From the platform, one could witness the sunset behind each of the pillars at critical times of the annual cycle. The two outer pillars demarcated the sunset on each of the two solstices: the north pillar demarcating the austral winter solstice and the south pillar the austral summer solstice.^[164] Nearby stands the well-known Solar Portal. Its iconography mirrored a temporal sequence of visual paths from Kalasasaya's balcony observatory. Eleven rayed faces on the Portal's lowest band depict the eleven recurring sunset points on the western horizon as marked by Kalasasaya's pillars. The portal likely once stood over the viewing platform, serving as both an entrance into the observatory, perhaps limited to a priestly caste, and a ready guide for the specialists who produced Tiwanaku's calendars.

After 500 ce, Tiwanaku builders initiated Akapana, Pumapunku, and several other monumental constructions. Built near the older Kalasasaya complex,^[165] Akapana blocked the southward visual path from the interior of the Sunken Temple toward the peak of Mount Kimsachata, resolutely directing visual attention toward the Akapana itself. The Akapana covered the plaza associated with the early Sunken Temple, and included the construction of a new one in front of its principal west stairway. Puma- punku, built at the west edge of Tiwanaku, may have served as a place where visitors, diplomats, and pilgrims first entered the center. In both Akapana and Pumapunku, a primary west stairway led ritual participants to the summit, where they were afforded an impressive glimpse of Mount Illimani, a glaciated peak east of the center in the Eastern Andean Cordillera, before being led into a sunken inner sanctum.

Figure 11.5 Kalasasaya, demonstrating the location of its west balcony wall (top image). The Sun Portal (middle image) occupies the west side of Kalasasaya. Its “serpent band” weaves together eleven solar faces, which match the eleven massive andesite pilasters (one has since been removed) that gird the west balcony. The northern and southern pilasters (lower image) establish solar setting places on the western horizon during austral winter and summer solstices. The central pilaster marks the sun's setting place on the horizon during each annual equinox.

The terraced platforms of Akapana and Pumapunku were human- wrought, “perfected” material icons of mountains.^[166] Their terraced form figuratively domesticated natural mountains while ritual processions into and out of them made visual references to the powerful, ancestral, distant peaks that they manifested. Like the vital streams that flowed from moun­tain springs, Akapana and Pumapunku - like Khonkho Wankane before them - incorporated elaborate drainage canals. The drains carried seasonal rainwater from ceremonial spaces down onto and through lower terraces. Water movement was not only visible but also distinctly audible to ritual participants.

Stone portals like the Sun Portal led ritual specialists and participants into increasingly sacred, potent ritual spaces. A key innovation during the Tiwa- naku period, they are known exclusively from Tiwanaku itself. Most formed narrow openings between rigorously bounded ritual spaces. Most also opened into narrow chambers,^[167] instilling a sense of mystery, disorientation, and esoteric power as a person entered increasingly interior inner sancta.

The nested architectural plans of the Kalasasaya, Akapana, and Puma- punku platforms aggrandized the nested moldings of such portals. The temples themselves served as “portals” that facilitated rapport with powerful animate forces immanent in the environmental features and elements that human groups sought to appropriate to Tiwanaku. Ritual movement through the temples - entered via stairways that facilitated views of a majestic peak, punctuated by monolithic portals, and culminating in ritual performance within intimate inner sancta modeled on ancient Sunken Temples - facilitated ritual experience in a new key. Sprawled over massive landscapes like the mountains they indexed, the temples gathered the forces of earth and sky.

Material transformations at Tiwanaku: sandstone and andesite

Stone construction of monumental structures indexed potent natural fea­tures and embodied their sacral forces. The use of stone revetments, stairways, and pavements afforded ceremonial complexes mass and permanence. Megalithic stones were quarried from nearby mountains, the very natural features the temples symbolically domesticated and whose immanent productive power they sought to appropriate. Temples were not simply icons of mountains; by incorporating the material essence of those mountains, they embodied them. In consecrating temples of stone, Tiwanaku builders sought to capture the generative forces of the earth. Yet stone quarrying, carving, and iconography changed significantly in relation to other monumental transformations. Most notable was a shift in emphasis from sedimentary sandstone to volcanic andesite, and throughout Tiwana- ku's apogee, the strategic integration of these two lithic materials.^[168]

During the Late Formative, monumental construction at Khonkho Wan- kane and Tiwanaku consisted of red sandstone and smaller quantities of other rock. Their early Sunken Temples are key examples. The structures and the monolithic personages housed within them consisted largely of sandstone. Red sandstone derives from nearby quarries in the Kimsachata range between Khonkho and Tiwanaku. The color indexed visible bedrock in the mountains, and may well have invoked blood, the fluid that affords life for llamas and humans. Red was also the primary decorative color for most Late Formative serving-ceremonial wares, potent material vehicles for the fermented liquids and foods that fueled recurring commensal events.

At Tiwanaku after 500 ce, craftsmen learned to quarry and hew massive blocks of andesite into exquisitely carved pilasters, portals, and monoliths. Quarrying and working andesite required an entirely new corpus of tech­nical expertise. Andesite derived from distant quarries across the southern portion of Lake Titicaca, and largely from the foot of Mount Ccapia, a volcano west of Tiwanaku. Kalasasaya's andesite west balcony and Sun Portal were early Ccapia andesite constructions. If they manifested a critical new importance for solar observations, they also emphasized the increasing importance of volcanic stone for Tiwanaku's monumental construction. A visual path through Kalasasaya's east doorway offers a clear view of Mount Ccapia, the source of the massive pilasters that supported its west balcony.

Andesite's bluish-gray color likely indexed the color of Lake Titicaca, near the shores of which volcanic stone was quarried. “Titi Kaka” refers to the gray-haired local mountain feline frequently depicted on Tiwanaku drinking vessels. Just as sandstone indexed the bedrock of local sedimentary formations, andesite indexed more distant volcanic sources and invoked the complementary life-principle of Lake Titicaca's water. During the Tiwanaku period, vast lacustrine floodplains were transformed into anthropogenic landscapes focused on raised field farming systems, which depended on a critical range of lake and water table levels. In this light, andesite invoked and propagandized Tiwanaku's new political horizons and productive power.

Monoliths and monolithic iconography

Located at the ends of concatenated pathways that brought ritual partici­pants to the centers and wove them through labyrinthine spaces, the stone sculptures punctuating Khonkho's and Tiwanaku's temple complexes pro­vide an intimate perspective on the changes elemental to Tiwanaku's urban ascendance (Figure ιι.6). Quarried pieces of mountains, stone personages constituted corporeal landscapes. Late Formative monolithic personages depict mythical ancestors or their living representatives (Figure ιι.6a). They make a distinctive arm gesture in which one arm is placed above the other across the torso. Their bodies were decorated with terrestrial zoomorphic creatures. The monoliths themselves were considered “persons” much as were totem poles among many native North American societies. They idealized the corporeal forms, gestures, and iconography of didactically narrated and collectively remembered ancestral personages. In this sense they did not simply depict past persons or actions. They presented ideal persons and ritual attitudes that living people could strive for in their lives and after-lives. Furthermore, the monoliths did not just represent ancestral persons; crafted of sandstone quarried from Kimsachata, monoliths embodied those persons as metonymic instantiations of the powerful gen­erative landscape that Khonkho and Tiwanaku shared.

Monolithic personages remained the focus of dramatic cyclical rituals during the early Tiwanaku period, but their forms, gestures, iconography, and overall meaning changed dramatically (Figure ιι.6b). Perhaps by way of sumptuary law, large monolithic personages were now restricted to Tiwanaku itself. Many were now crafted of volcanic andesite. This may have occurred just as Tiwanaku monopolized the volcanic stone quarries of Ccapia. In place of crossed arms, each personage now made a dual presentation. In one hand it held a ceremonial drinking kero and in the other a tablet for ingesting psychotropic substances. These were vehicles for mind-altering substances that facilitated two dimensions of religious experience - one relatively communal and one relatively personal.

Figure ii.6 Monoliths. The top image, (A), presents three Late Formative monolithic stelae, the first two (drawn) from Khonkho and the third (decapitated monolith) from Tiwanaku. The bottom image, (B), depicts two representative Tiwanaku-style monoliths: the sandstone Bennett and andesite Ponce stelae.

They were, I argue, complementary ritual attitudes that came to consti­tute the ideal Tiwanaku subject.

Tiwanaku period monoliths differed from earlier personages in that they did not simply depict deified ancestors. The impassive faces still denote deified status, but bodily decoration now emphasized elaborate clothing; specifically the tunic, sash, and headgear of an elite person. These person­ages now depicted either ancestral deities bedecked as elite persons or elite persons dressed as ancestral deities, and it is likely that iconography deliber­ately promoted ambiguity. What these icons reveal is a lithic presentation of social status that, through recurring ritual practice and ongoing appeal to their didactic properties and telluric productive powers, legitimized the crystallization of class differences after 500 ce.

Unlike Late Formative monoliths, Tiwanaku monoliths depicted imagery that indexed the sky. Iconography included personages wearing headdresses with solar designs and a new emphasis on predatory avian imagery. For example, the braided tresses hanging from the back of the Bennett mono­lithic personage end in avian heads. The posterior iconography of the personage unfolds around a central figure standing on a terraced temple­mountain wearing a radiating solar headdress. Above the figure are disem­bodied portraits of the same face wearing solar headdresses, intermixed with attendants wearing the beaked masks of predatory birds and facing upward toward the sky.

Kalasasaya's Sun Portal, as noted above, indexed the sun and solar cycles. Partially covered in gold lamina, it dramatically performed this indexical relation by reflecting solar light across the space that it faced, capturing and aggrandizing the material power of the sun. Tiwanaku's emergent urban centrality was tied to a novel attention to the productive power of the sun and its recurring cycles, and an intensified attention to skyscape overall. Emphasis on solar cycles allowed emergent leaders to coordinate and integrate the multiple ritual cycles and productive rhythms of the diverse social and productive communities - fishers, farmers, and herders - they sought to integrate.²¹ It was an astute strategy of cosmic integration.

Tiwanaku residential expansion and elite distinction Transformations in cosmology and ritual practice were central to Tiwana- ku's success. This is demonstrated in Tiwanaku's dramatic growth as a densely populated urban center after 6oo ce. The Late Formative centers of Khonkho Wankane and Tiwanaku were lightly occupied relative to the size of their monumental cores. This is clearest at Khonkho, where occu­pants of the main platform were directly involved with the supervision and maintenance of its ceremonial spaces and the orchestration of rituals held therein. This appears to have been the case for Tiwanaku as well, at least until Late Formative 2. Both were centers dedicated to periodic rituals.

Tiwanaku's urban expansion began during late Formative 2, coincident with notable changes in its monumental spatiality and ritual practices. From this period through Tiwanaku collapse people lived in bounded residential compounds, yet most dwellings were now rectangular rather than circular.

Clusters of adjacent compounds formed tightly connected barrios, some of which were dedicated to specialized activities such as the production of ceramic vessels. Others, in particular those located closest to the monumen­tal core, supported ceremonial activities associated within Tiwanaku's increasingly extensive and complex temples. It is significant that all monu­mental and residential construction at Tiwanaku from Late Formative 2 through the Middle Horizon followed the axial orientation established during its early history, tied to proxemic and visual relations to specific celestial movements and terrestrial features.

Sociopolitical hierarchy crystallized just as Tiwanaku emerged as one of the most influential centers of the Middle Horizon Andes. At the Late Formative centers of Khonkho and Tiwanaku, status resided in proximity to key ceremonial spaces and in curating the ancestral monolithic person­ages that stood inside of them. It was tied to proximity with and purchase over these spaces and objects. Status continued to be spatially and materially linked to monumental space as Tiwanaku became a primary center. Elite activity was located adjacent to the Putuni and on top of the Kalasasaya and Akapana platforms. Human burials near the Putuni contained elaborate vessels and sumptuous bodily adornments crafted of rare materials such as coastal shell, sodalite, silver, and gold. Excavations in Kalasasaya recovered several diadems of hammered laminar gold encrusted with malachite and azurite. Thus, as Tiwanaku expanded, relative status, constituted as prox­imal relation to monumental places and practices, became more intensely polarized as elite “distinction.”

What became of Khonkho Wankane? If Khonkho and Tiwanaku were closely interconnected during the Late Formative, it is odd that its monu­mental core was abandoned by 500 ce. I have suggested that the two centers had always been in competition and reached an inevitable showdown. Yet there is no evidence for violence or destruction at Khonkho. Rather, Khonkho's stylistically latest monolith appears to have stood within the main plaza for centuries. Although slumped over and eroded, it remains in its plaza today as one of the most powerful animate objects in the local region. Although speculative, I suggest that many of Khonkho's residents may have transferred to Tiwanaku. This idea meshes better with Tiwana- ku's predominant incorporative political strategies. In one scenario, Khon- kho's original inhabitants came to Tiwanaku and occupied the area that produced the Pumapunku campus. In fact, recent research indicates that Pumapunku was first occupied during Late Formative 2, when Khonkho was abandoned.

Kolata and Ponce^[169] first suggested that Akapana and Pumapunku formed paired monumental ceremonial spaces in Tiwanaku. New evidence for the historical relation between Khonkho and Tiwanaku supports this idea. An early relation between powerful Late Formative centers on either side of the Kimsachata range, I hypothesize, may have been reconstituted as a paired relationship within Tiwanaku itself. This act would have recreated an enduring political aesthetic of dual centers within the emergent urban center of Tiwanaku. Resolving this hypothesis rests on the balance of future research.

Conclusions

Reciprocally process and product, Tiwanaku emerged over a long, complex history in the southern Lake Titicaca Basin. It originated as one in a network of interacting early Late Formative centers. Later, urban process found an axis in the pairing of Khonkho Wankane and Tiwanaku across the Kimsa- chata range. Khonkho and Tiwanaku formed paired centers of social gathering and cyclical commensal ritual for several generations. After 500 ce, Tiwanaku emerged as the primary center in the region. Urbanism was always distributed throughout the region, whether among multiple or between paired centers, but always focused in one or more centers that cyclically gathered distant populations for key ritual events. In each config­uration, urbanism was distributed across a regional landscape of centers and settlements.

Astute agropastoral productive strategies allowed human groups and increasingly complex social institutions to thrive at one of the highest liveable altitudes on earth. Productive strategies, in turn, depended on pragmatic relations to the world that emphasized knowledge of the vitality inherent in terrestrial forces and recurring seasonal and celestial cycles. Political authority at Khonkho Wankane and Tiwanaku thrived not on shuttering attention to natural processes as in contemporary cities, but on rendering some of those processes central to the well-being of humans, their crops, and their herds. This was done by way of ritual practices linked to an ecology in which key landscape phenomena were deemed animate, ances­tral forces that humans could tap into and influence for their own good. The groups who successfully positioned themselves at the hub of social, spatial, and cosmic networks that linked humans with these forces accrued inordin­ate status. After 500 ce, these were the elites who lived on or near Tiwana- ku's primary monumental campuses.

Urbanism in the southern Andes was an ongoing project that continually produced a highly ritualized yet profoundly pragmatic cosmology that I term an animistic ecology. Ancestral mountains, lakes and rivers, celestial bodies, and other key Andean landscape and celestial features were treated as animate persons who act in the world, ideally on behalf of self-identified descendants. Tiwanaku's terraced temples, laced with labyrinthine subterranean canals, were animate, human-wrought, life-affording moun­tains that, by way of ritual acts, people entreated to influence the greater ancestral mountains they indexed and, constructed of their lithic essence, embodied. Archaeologists have treated monuments and their personified monoliths as iconic representations of natural phenomena. Here, monu­ments and monoliths embodied and, thus, materially condensed and mani­fested those phenomena as agentive ancestral personages.

Emergent urbanism in the southern Andes has much to offer studies of past urbanism globally. Its sociospatial specificity was distributed and ani­mating in its high-altitude landscape. It provides an alternative to many political economy models. According to the latter, past cities and political centralization were largely the work of a few naturally ambitious aggrand- izers and their presumed lust for “wealth” - a laden term that flattens all kinds of practices and values to a single utilitarian function: profit. In the south-central Andes, emergent authority was at least as much about medi­ation as it was about aggrandizing. Further, it was made possible due to the production of intricate relations among diverse communities and built ritual-political centers, and of intimate relations between humans and worldly animate forces, which, together, through periodic ritual events, were rendered vital to human well-being.

FURTHER READINGS

Abercrombie, Thomas A., Pathways of Memory and Power: Ethnography and History among an Andean People, Madison: University of Wisconsin Press, 1998.

Albarracin-Jordan, Juan V., Carlos Lemuz Aguirre, andJose Luis Paz Soria, "Investigaciones en Kallamarka: primer informe de prospeccion,” Textos Antropologicos 6 (1993), 11-123. Allen, Catherine, The Hold Life Has: Coca and Cultural Identity in an Andean Community, Washington, D.C.: Smithsonian Institution Press, 1986.

Couture, Nicole C., “Ritual, Monumentalism, and Residence at Mollo Kontu, Tiwa­naku,” in Alan L. Kolata (ed.), Tiwanaku and Its Hinterland: Archaeology and Paleoecol­ogy of an Andean Civilization, Washington, D.C.: Smithsonian Institution Press, 2003, Vol. ιι, pp. 202-25.

Janusek, John Wayne, “El surgumiento del urbanismo en Tiwanaku y del poder politico en el altiplano andino,” in Krzysztof Makowski (ed.), Sehores del Imperio del Sol, Lima: Banco del Credito del Peru, 2010, pp. 39-56.

Janusek, John Wayne, and Victor Plaza Martinez, “Khonkho e Iruhito: tercer informe preliminar del Proyecto Arqueologico Jach'a Machaca,” Research report submitted to the Bolivian Viceministerio de Cultura and the Unidad Nacional de Arqueologia, La Paz, Bolivia (2007).

“Khonkho Wankane: Segundo informe preliminar del Proyecto Arqueologico Jach'a Machaca,” Research report submitted to the Bolivian Viceministerio de Cultura and the Unidad Nacional de Arqueologia, La Paz, Bolivia (2006).

Kolata, Alan L., “Tiwanaku Ceremonial Architecture and Urban Organization,” in Alan L. Kolata (ed.), Tiwanaku and Its Hinterland: Archaeology and Paleoecology of an Andean Civilization, Washington, D.C.: Smithsonian Institution Press, 2003, Vol. ιι, pp. 175-201.

Ohnstad, Arik T., “La escultura de piedra de Khonkho Wankane,” in John W. Janusek (ed.), “Khonkho Wankane: Primer Informe Preliminar del Proyecto ArqueoliSgico Jach'a Machaca,” Research Report submitted to the Bolivian Viceministerio de Cultura and the Unidad Nacional de Arqueologia, La Paz, Bolivia (2005), pp. 52-68.

Ponce Sangines, Carlos, Descripcion sumaria del templete Semisubterraneo de Tiwanaku, La Paz: Juventud, 1990.

Tiwanaku: Espacio, Tiempo, Cultura: Ensayo de sintesis arqueologica, La Paz: Los Amigos del Libro, 1981.

Portugal Ortiz, Max, and Maks Portugal Zamora, “Investigaciones arqueol(5gicas en el valle de Tiwanaku,” in Arqueologia en Bolivia y Perui, La Paz: Instituto Nacional de Arqueologia, 1975, Vol. ιι, pp. 243-83.

Portugal Zamora, Maks, “Las ruinas de Jesus de Machaca,” Revista Geografica Americana 16 (1941), 291-300.

Posnansky, Arthur, Tihuanacu: The Cradle of American Man, New York: J. J. Augustin, 1945, Vols. i and ιι.

Protzen, Jean-Pierre, and Stella Nair, “On Reconstructing Tiwanaku Architecture,” Journal of the Society of Architectural Historians 59 (2000), 358-71.

Rivera Casanovas, Claudia S., “Ch'iji Jawira: A Case of Ceramic Specialization in the Tiwankau Urban Periphery,” in Alan L. Kolata (ed.), Tiwanaku and Its Hinterland: Archaeology and Paleoecology of an Andean Civilization, Washington, D.C.: Smithso­nian Institution Press, 2003, Vol. ιι, pp. 296-315.

Smith, Scott Cameron, “Venerable Geographies: Spatial Dynamics, Religion, and Polit­ical Economy in the Prehistoric Lake Titicaca Basin, Bolivia” Ph.D. dissertation, University of California, Riverside, 2009.

Vranich, Alexei, “Interpreting the Meaning of Ritual Spaces: The Temple Complex of Pumapunku, Tiwanaku, Bolivia,” unpublished Ph.D. dissertation, University of Pennsylvania, 1999.