One of the most spectacular examples of a residence with political functions is without any doubt the palace of Versailles, built in France in the seventeenth century. Commissioned by Louis XIV as a royal residence as well as the seat of government, Versailles materialized Louis XIV's conception of kingship in its architectural layout and luxurious decoration. Building on the new scientific discoveries by Galileo Galilei that indeed the sun and not the earth was the center of the universe, Louis XIV claimed to be the human personification of the sun, the central celestial body. This notion entered into the architectural design of Versailles in that the royal bedroom was the central room of the large palace complex. Other official rooms were associated with the known planets and were arranged around Louis XIV's bedroom, approaching a circular formation that imitated the planets' rotation around the sun. Louis XIV's identification with the sun also entered into court ritual. All members of the royal court were required to gather in the bedroom to watch the daily lever (rising) and coucher (going to bed) rituals of their Sun King, and the whole court could only be active after Louis XIV had risen in the morning and until he had fully withdrawn for the night.
Several other examples of the identification of the ruler and his power with the palace exist outside the Western tradition. The Royal Palace of Benin, for example, expressed through its scale, exclusivity of access, and ornamentation the power of the Oba. The Forbidden City of China displayed royal power through these means and through its cosmic imaging and associations.
The examples from the Americas are different in many ways. The goal of this volume is to present cases of ancient American residential architecture for which there is evidence of a specific political context or propaganda strategy. We selected case studies from North America, Mesoamerica, and South America and from early to close-to-contact chronological periods to show that the relation between political power and architecture is a pervasive theme in the Americas. Nevertheless, the chapters presented should not be considered a complete whole that exhausts the topic. Some areas are noticeably missing, for example, the Mississippian culture in the southeastern United States or the Sican culture on the north coast of Peru. Rather, we hope that this volume will inspire palace studies in these missing areas as well as others outside the Americas and that scholars will be able to compare and contrast their data with the findings presented here. The volume would further be of interest to architectural historians seeking to understand the form of palaces in the Americas.
Most of the existing literature on the subject has been concerned with identifying and classifying residences. Indeed, as noted below, this is the subject of a number of the articles in the present volume. In the Maya area, the scholarly debate focused on the differentiation between temples and palaces in the early twentieth century (Spinden 1913; Tozzer 1911). Temples were understood as tall flat-topped pyramidal platforms supporting single-room structures, whereas palaces were longer multiroom structures standing on low platforms. As the data and number of known examples increased, researchers began to recognize hierarchies among residences (Christie 2003; Willey and Leventhal 1979; Willey et al. 1978). Palaces came to be understood as the royal residence as well as the seat of government and administration, and they were set apart from elite residences and the houses of commoners, which had only domestic functions. The distinguishing criteria were primarily size, quality of construction materials, sculptural decoration, and location with respect to proximity to the ceremonial center.
Similar issues have been discussed in southwestern and Moche archaeology. In Anasazi sites, a clear distinction can be drawn between religious structures or kivas and houses, as the latter are relatively small and simple rectangular rooms. It has not been possible to establish a hierarchy among rooms but perhaps among sites (Lekson, this volume).
The well-known Moche structures of Huaca del Sol and Huaca de la Luna have been variously interpreted as temples or palaces (Benson 1992:303-315; Moseley 1993:166-178). This discussion is particularly complicated by the fact that Huaca del Sol has never been professionally excavated. Claude Chapdelaine (this volume) presents evidence supporting some palace functions at Huaca de la Luna.
Other studies have attempted to reconstruct ancient social organization and political power structures partly from architectural data. The numerous settlement surveys conducted in Mexico and Central America in the second half of the twentieth century fall into this category (Blanton 1978; Kowalewski et al. 1989; Millon 1973; Puleston 1983; Willey and Leventhal 1979; Willey et al. 1978). This kind of research focuses on the density or scarcity of populations and aims to draw conclusions about social organization from these patterns. What emerged from the settlement-pattern data was the realization that most ancient American societies were not simply divided into the rulers and ruled but were a lot more complex. Arlen and Diane Chase discuss these complexities in their edited volume Mesoamerican Elites (1992). They and various contributors analyze what the archaeological evidence reveals about social and political organization in various Mesoamerican societies. Some authors address the question of whether the Maya were organized according to a segmentary or a unitary structure.
The present volume, Palaces and Power in the Americas: From Peru to the Northwest Coast, focuses on specific examples of architecture from various parts of the Americas and how they express aspects of political power. In this manner, it goes beyond two other recent volumes that cover only the Maya: Maya Palaces and Elite Residences: An Interdisciplinary Approach (2003), edited by Jessica Christie, is primarily concerned with identifying palaces and elite residences; Maya Royal Courts (2000), edited by Takeshi Inomata and Stephen Houston, attempts to reconstruct the structure of and life at royal courts in Maya cities, including some comparisons with other types of ancient courts in the Americas. The present volume is also different from the Dumbarton Oaks volume entitled Palaces of the Ancient New World (2004) and edited by Susan Evans and Joanne Pillsbury. Our volume concentrates more specifically on the context and issues of political power and, unlike any of the others mentioned, includes case studies from North America. Although the list of ancient American cultures represented does not in any way pretend to be complete, the goal is that the reader will develop an understanding of how ancient American societies at different times and with different levels of social organization used residential architecture for political purposes. In this light, the value of the volume is twofold: first, it publishes information and often very recent data (Isbell, Demarest) about specific cultures, and second, it compares the different architectural approaches societies take, thereby potentially contributing to architectural theory and history.
Given the variety of cultures represented in this volume, their different time periods, and the different aspects emphasized by the authors within the larger theme of political power, we have grouped the articles into four parts according to the authors' approaches.
Part 1. Identification of Palaces
The first section comprises studies that identify palaces in cultures in which palaces have not been identified before. All authors understand a palace as the residence of the highest-ranking member or institution of a polity and as the seat of governmental activities. Hence palaces can be considered an essential trait of complex stratified societies. In the archaeological data of a palace, one would expect to find evidence of domestic life, such as food production and eating, as well as official rooms, perhaps council rooms, audience halls, and state storage rooms for surplus products. Furthermore, palaces often include open spaces for the public display of the ruler. Elite residences, on the other hand, reveal evidence solely of domestic activities. They distinguish themselves from the simpler houses of commoners by means of size, construction materials, quality of construction, and often decoration, but they were usually not locations where political actions were taken (see Barber and Joyce, this volume, for a thorough discussion of the differences between a palace and an elite residence). The examples in which an argument can be made for a political context have to be specially noted, and these constitute the focus of this volume. It follows that palaces are more than just residences of the highest-ranking nobles. What distinguishes palaces from elite residences is the symbolic function of palaces as political statements within their respective societies. Palace architecture, including the layout, position, and decoration of palaces as well as their public (or sometimes exclusive) nature, reflects the structure of the political system in which the palaces exist. Palaces not only respond to the necessities of a society and its highest-ranking institutions to represent status and hierarchy, but they can actually shape the understanding of institutions and take an active role in their respective societies.
Claude Chapdelaine takes up the issue of identifying palaces and distinguishing them from elite residences for the Huacas of Moche Site. Many conjectures have been made and interpretations given in the literature with regard to the nature and function of Huaca del Sol and Huaca de la Luna (see Benson 1992; Moseley 1993). Chapdelaine devises a list of seven criteria that might identify a palace and discusses how they apply to Huaca de la Luna. His discussion includes new data obtained by a long-term Peruvian project codirected by Santiago Uceda and Ricardo Morales of the Universidad of Trujillo, but it does not provide a black-and-white answer. He concludes that Huaca de la Luna was a multifunctional building that included aspects of a palace and of a temple. He stresses the importance of the large northern plaza overlooked by many scholars prior to the Peruvian excavation of 1999. This plaza provides space for a large segment of the population to gather and to participate in religious and/or government rituals and thus supports the palace function of Huaca de la Luna. On the other hand, Chapdelaine admits that "no section has yet been identified as a residential section" and suggests that Complex 8, located immediately to the southwest of Huaca de la Luna, could possibly have been the residence of high priests or even of the ruler.
Of particular importance is Chapdelaine's presentation of new data from the urban sector between Huaca de la Luna and Huaca del Sol. He correlates architectural changes in Compound 9 with the integration of nuclear family households into a single large household unit. This emerging leader of Compound 9 enjoyed the power to modify the urban layout by placing a new street between his household and those of his neighbors, which may demonstrate his high position within the Moche social hierarchy. Compound 5 is of interest because during periods of its occupation it did not have a kitchen. Absence of food-production and cooking facilities may indicate that this compound was, at least during certain times, not an elite residence but a public building. Thus Chapdelaine's data establish evidence that the power of the ruler(s) of Huaca de la Luna and Huaca del Sol was to some degree decentralized.
William Isbell also uses a set of formal characteristics to identify Andean and, in his study, Wari palaces. Since very few data about Wari architecture are so far available, Isbell devises a list of diagnostic architectural features of Inka palaces and of those at the religious center of Pachacamac. Since the Inka succeeded the Wari chronologically, a fair amount of data about Inka architecture exists from Spanish descriptions and archaeological excavations. Isbell tests the presence or absence of such features on the Vegachayoq Moqo Palace at Huari and on the recently excavated Conchopata Palace. His analysis shows that especially the early architectural forms at Vegachayoq Moqo resemble Inka and Pachacamac palaces and that the possible palace at Conchopata shares more of the Inka characteristics.
Isbell then expands his discussion of Wari palaces to the rural hinterland. He examines several small settlements based on architectural form and distinguishes between cumulative residential buildings, which are complexes of small rooms constructed one after another, and planned orthogonal architecture, which consists of a rigidly rectangular compound with a large patio in one sector and orderly room complexes in another. His argument is that the presence of planned orthogonal architecture indicates the concentration of some form of political power, and he suggests that some of these settlements, such as Jargampata and Azángaro, were probably rural estates of a Wari king or nobleman. In addition, Isbell comments on the role of the provincial sites of Pikillacta and Viracochapampa, which exhibit a very rigid version of planned orthogonal architecture that he names the "orthogonal cellular horizon." Here the buildings are characterized by orthogonal enclosures, subdivided into smaller rectangular or square compounds that follow a limited number of standard patterns and create a cellular structure of repeating modular units. Isbell concludes that Pikillacta and Viracochapampa were important Wari administrative centers and that they housed the palaces of powerful governors. On the other hand, Pikillacta and Viracochapampa, as well as Conchopata, seem to lack rural estates, which Isbell understands as significant political power bases. This reasoning leads to the conclusion that Huari was indeed some kind of capital city, because it was the only large settlement ringed by hinterland estates, forming what Isbell calls a "regal landscape."
Stephen Lekson takes a rather different approach. He critically examines archaeological attitudes toward palaces in the context of Southwest archaeology and with specific reference to Chaco Canyon. His focus is not so much on the material evidence itself but on the politics of interpretation. The same archaeological evidence may be used for interpreting an ancient society as having a simple unified social organization or a complex stratified one. Southwestern and North American archaeologists have historically chosen the former interpretation and understand Native North American cultures as heterarchies of gentle people working in collectives mostly for ritual purposes. Lekson, on the other hand, prefers the latter and presents the argument that the Great Houses in Chaco Canyon were palaces that included wings of storage rooms, ritual architecture, and public assembly areas. He further argues that Chaco was the ceremonial, political, and economic center in the Southwest from about A.D. 900 to A.D. 1125 and was succeeded by two subsequent capitals—Aztec Ruins and Paquimé—each of which legitimized itself by conscious reference back to Chaco. Future research is needed to confirm Lekson's reasoning, but his ideas certainly "blow a fresh and new wind" into southwestern archaeology and provide "rich food for thought."
Part 2. Palaces as Active Stage Sets of Political Ideology
The second part of the volume includes articles that discuss palaces as "stage sets" for various types of activities that consolidate and express power. In this group fall studies that associate certain pre-Columbian societies with the so-called theater state. A "theater state" was first defined in the traditional society of Bali, Southeast Asia, by Clifford Geertz (1980). In a theater state, cosmic rituals of kingship are performed in such a manner that they do not just mirror reality but actually become it and thus open the potential to construct kingship in the way desired by those who hold it. Stanley Tambiah (1985) criticized Geertz's definition for leaving a disjunction between ideology and practice and proposed an expanded model to bring these two poles closer together. Tambiah (1985:252-259) has coined the label "galactic polity" to conceptualize an arrangement of a center and its surrounding satellites, which may be employed in spatial, social, and political contexts. He emphasizes (1985:324-326) that the center had to continually reaffirm its ideology of control in response to volatile factional struggles mounted by the periphery against the center and that galactic polities were by definition unstable. It follows that in theater states and galactic polities, political power is based on ideology, and architecture is thus designed and cities are laid out to make this ideology visible and public. Such a materialization of political ideology may include, but is not limited to, large open plazas for public assembly, cosmological orientation of buildings, and local cosmograms. In these cases, architecture takes an active role in the reaffirmation of rulership.
Arthur Demarest has elsewhere (1992, 2003) described the Classic Maya as a theater state. In this volume, he distances himself somewhat from the theater-state model and presents a compelling comparison of how architecture was used for differing political purposes by the ancient Maya. The Murciélagos Palace at Dos Pilas exemplifies how architecture would have been designed in a theater state. First, this palace was placed atop a cave system that passes beneath the eastern El Duende complex of Murciélagos and the West Plaza group of Dos Pilas, with the cave emerging in springs at each architectural complex. Thus the location of surface architecture was made sacred through the supernaturally charged underground caves. Second, the Murciélagos Palace was crossed by the east-west processional axis of Dos Pilas between El Duende and the West Plaza group. And third, a shrine in the northern residential compound of the palace did not contain the expected human burial, but instead it entombed the east-west cave system. Thus the symbolic connection between the ruler and his family and the Maya cosmogram was made very explicit: they resided and took political actions upon the sacred mountain above caves and their springs, where the axis mundi was localized by the family shrine. The east-west direction of the cave system was repeated in the public path of processions as well as in the movements of the Sun and Venus, which would have placed the royal family in the Murciélagos Palace quite literally in the center of the world.
One of the smaller elite compounds has been identified as the residence of the "Lady of Cancuén," a principal wife of Dos Pilas Ruler 3, and its primary political function was to proclaim Dos Pilas's alliance with Cancuén. It did so by displaying an architectural style derived from Cancuén and through its carved monuments, one of which shows a scene of bloodletting conducted by a young boy and observed by the Lady of Cancuén and Ruler 3.
By contrast, the Aguateca Palace, which was another seat of the Dos Pilas dynasty, had restricted access and was well defensible. Court artisans who manufactured elite goods also lived in this palace, and it may have been a locale for feasting.
The Cancuén Palace, which has been the focus of Vanderbilt University projects in the western Petén since 1998, presents yet another case. Cancuén is located on the upper Pasión River at a natural bay portage from where the Cancuén dynasty seems to have controlled the river trade between the lowlands and the highlands. The site of Cancuén lacks temples or any large plazas for the performance of public rites, which would have been essential to a theater state. The palace itself consists of a series of restricted courtyards that would have offered an ideal setting for feasting and impressing visiting elites. Workshops of lithic artifacts have been identified nearby. Thus the Cancuén Palace seems to have had primarily economic and political functions.
The conclusion of Demarest's comparison is a warning against quick labeling and an observation that the power of Classic Maya kings was not based on one single element, such as agricultural management, control of long-distance trade, or theater-state performance. Instead, elites in different regions and in different circumstances relied to varying degrees on alternate sources of power, which are reflected in the location and layout of palace architecture.
Colin Grier's work on Coast Salish households is included in this part of the volume even though Northwest Coast cultures have never been described as theater states. As indicated in the discussion of Lekson's article above, Northwest Coast cultures are widely understood as non-state-level societies.
Grier discusses the class system of household chief, nobility, commoners, and slaves and how individuals of these classes interacted and related within the social unit of the household itself. He then analyzes the wooden plankhouse as the material expression of such power relations. His particular example comes from the precontact Coast Salish Dionisio Point village site located on one of the Gulf Islands of southern British Columbia, Canada. Although no standing architecture remains, the types and distribution of artifacts in House 2 provide insights into the power relations of its residents. In general, the typical Coast Salish dwelling was the shed-roof house consisting of a series of compartments, each of which provided a domestic family space with a small hearth. This type of building structure suggests egalitarian relations between families and provides little opportunity for social differentiation. However, in Dionisio Point House 2, status-indicating artifacts, such as carved stone bowls, labrets, and beads, were all found associated with the center of the house. Furthermore, the central hearth exhibited burning patterns pointing to ceremonial rather than cooking functions. Grier concludes that the economic, social, and ritual-symbolic sources of power were concentrated in the center of House 2 and were probably in the hands of the elite family residing in the nearest compartment. In this manner, the center of the house acted as a "stage set" within the context of the household and certainly on a very small scale when contrasted with other cultures, such as the Maya. The types of performances acted out in the center of House 2 were likely dances owned by the chief and elites as well as meetings and the reception of elite visitors from other households. Grier's study is the only example in the volume in which power relations are reconstructed within one residence.
In her study of Tajín Chico, the elite residential zone of El Tajín, Patricia Sarro argues that this elevated grouping of buildings, plazas, and ritual structures functioned as a palace, much like those of the lowland Maya. Within it were buildings in which the various rituals of rulership were carried out. In some, benches and a possible throne imply places of privileged gathering of noble counsels. Others appear to have been residences, but even these have open porticoes that overlook the areas with multiple courtyards and plazas. Sarro focuses on Building A, whose unique design constitutes a cosmogram in which the earthly zone takes the form of miniature ballcourts. Whatever this building's other functions—and these may include residence—Sarro believes that it served as the stage for rituals that connected the ruling elite with the power of the game.
Part 3. Correspondences between Material Aspects of Elite Residences and Social Status
The third part of the volume presents cases in which differences in elite residential architecture imply differences in the form of rulership or social status. The material differences may lie in the form of residences or in their size, location, or associated artifacts.
Michael Blake, Richard G. Lesure, Warren D. Hill, Luis Barba, and John E. Clark take the reader to the Early Formative village site of Paso de la Amada in Chiapas, Mexico. Paso de la Amada was founded around 1500 B.C., grew to its largest size in the Locona phase (1400-1250 B.C.), and began to lose population after 1000 B.C. Blake et al. present archaeological data from Mounds 6 and 12, which were obtained during major excavations carried out at the site between 1985 and 1995. Their data focus specifically on mound platform volumes; building structures, shapes, alignments, and superpositioning; features such as floors, hearths, pits, and burials; distribution of micro-artifacts on the floors, including obsidian, ceramics, charcoal, and bone; and chemical traces on structure floors, such as pH levels, phosphates, fatty acids, and albumen.
A comparison of these elements for both Mounds 6 and 12 shows that the differences were primarily architectural. For example, in Mound 6, before a new building was constructed, a layer of earthen fill was laid down to cover and ultimately protect the old floor and features. This practice resulted in the gradual raising and extending of the platform on which the buildings stood over a period of 200 to 250 years. In Mound 12, on the other hand, there was no platform construction until about 1100 B.C. As a consequence, the earlier floors were not protected by fill and are poorly preserved. But the analysis of the features, distribution of micro-artifacts, and chemical patterns shows no significant differences or changes in the floor sequence of Mound 6 or between Mound 6 and Mound 12. This would indicate that similar activities that left micro-artifactual or chemical traces were carried out over time and in both mounds.
Blake et al. use the results of their comparison to support the notion that the people of Paso de la Amada were organized in a transegalitarian society. In egalitarian societies, corporate groups may construct large public buildings, but transegalitarian societies exhibit not only public facilities but also the residences of emerging leaders, which distinguish them from the commoners. Blake interprets the platform construction in Mound 6 as just this kind of residential elaboration that sets the inhabitants of Mound 6 apart from the commoners in Mound 12. The observation that the activities carried out in both mounds were largely similar shows that this emerging elite continued to perform domestic tasks.
Sarah Barber and Arthur Joyce present a detailed overview of elite residential architecture in the Valley of Oaxaca from the Early Formative (1800-850 B.C.) to the Postclassic (A.D. 1200-1522) period. They evaluate what the architectural data reveal about forms of government and changes over time. The earliest differentiation in houses and evidence for economic inequality have been noted in the San José (1150-850 B.C.) and Guadalupe (850-700 B.C.) phases. At this time, some residences at San José Mogote, Tomaltepec, and Fábrica San José exhibited walls with whitewash, larger posts, drains, stone foundations, more participation in craft activities, and greater access to exotic items. In the Rosario phase (700-500 B.C.), some household units at Tomaltepec and San José Mogote were built in close physical proximity to public buildings or temples. Barber and Joyce interpret these and other data as efforts by an elite to define their identity as fundamentally separate from that of commoners.
Monte Albán was founded as a new settlement about 500 B.C. and grew rapidly in the Late Formative (Monte Albán I, 500-100 B.C.) and Terminal Formative (Monte Albán II, 100 B.C.-A.D. 200). Iconographic references indicate that the layout of Monte Albán reflected the cosmos, and high-status residences were part of this cosmogram. The latter were concentrated around the northern end of the Main Plaza, implying a connection between nobles, the celestial realm, and noble ancestors whose tombs often contained effigy vessels depicting the Zapotec lightning or sky deity. More high-status residences were found at other sites in the Valley of Oaxaca as well as in some of Monte Albán's barrios beyond the Main Plaza. This increase in numbers of elite residences is seen as the spread of a separate noble identity and as the development of a more decentralized administrative organization.
During Monte Albán IIIa (A.D. 200-500), elite residences continued to be built close to the North Platform. Three residential "types" may be distinguished, mostly by size, since all consist of enclosed patio groups. Elite status was primarily expressed through the ornamentation of tombs with elaborate polychrome murals. From a political point of view, tombs can only be viewed by small audiences and never provide a public forum for the expression of power.
The spatial layout and the architectural elaboration of elite residences in Periods IIIb-IV (A.D. 500-1200) follow the Period IIIa examples, at least in Monte Albán itself, which loses its position as the major center in the Valley of Oaxaca by about A.D. 800. Other sites on the valley floor develop new ways of displaying power. For example, in Lambityeco Structure 195, stucco panels depict named individuals holding human femurs on the walls of a semipublic patio.
By Monte Albán V, or the Late Postclassic (A.D. 1200-1521), the distinction between private and public spaces had become blurred, and monumental architecture clearly combined residential and public functions, as, for example, in the Palace of the Six Patios at Yagul or the Group of Columns at Mitla.
Barber and Joyce's chronological discussion emphasizes that although elites began to separate themselves from commoners as early as 1150 B.C., Zapotec elite residences did not organize space in a way that allowed for the public expression of individual power until after the collapse of Monte Albán in about A.D. 800, as in Lambityeco Structure 195. Only the Late Postclassic examples from Yagul and Mitla were palaces in the sense that they not only combined but fully merged open and public spaces with residential spaces. It follows that in the Valley of Oaxaca, architecture expressed political power in ways that sharply contrast with the theater-state examples of the Maya or El Tajín. Thus Barber and Joyce offer an important alternative perspective.
William T. Sanders and Susan Toby Evans trace the development of Teotihuacan rulership through different stages of palace architecture. They postulate that the earliest palace was possibly the Xalla compound, located between the Pyramids of the Sun and Moon in the northeastern sector of the city. In the third century A.D., it was replaced by the Ciudadela as the political administrative center. Sanders and Evans suggest that the Ciudadela functioned as the possible residence and seat of government of one or two powerful rulers whose lineages were associated with the Feathered Serpent and with the Fire Serpent, as depicted on the Temple-Pyramid of the Feathered Serpent. Around 300 A.D., a reaction against the political dynasts set in, and this Temple-Pyramid was concealed by the Adosado. At approximately the same time, many of the city's apartment compounds were constructed, including the Street of the Dead Complex that straddles the Street of the Dead between the Pyramid of the Sun and the Ciudadela. According to Sanders and Evans, the Street of the Dead Complex was the palace of a new type of government that consisted of a hierarchy of relatively depersonalized bureaucrats whose identity as individual power holders was not emphasized in the visual arts.
Sanders and Evans continue to discuss the power and influence Teotihuacan had exerted abroad, and they relate the changes in government to the contemporary developments at Tikal. They show that the much-debated Early Classic Teotihuacan presence at Tikal occurred in the late fourth century, based on hieroglyphic as well as archaeological evidence that corresponds to the proposed change in Teotihuacan government from the powerful dynast(s) to less individualized bureaucrats. Sanders and Evans pose the question of whether a ruling Teotihuacan dynasty associated with the cult of the Feathered Serpent traveled to the Maya region, possibly into exile, which led to the outlined changes at home. They further ask whether Teotihuacan's political advance into the Maya area may have been an Early Classic actualization of the Early Postclassic story of Quetzalcoatl. Quetzalcoatl, or the Feathered Serpent, was said to have been exiled from Tula and subsequently became an influential figure among the Maya in Yucatán. These ideas will contribute to the current scholarly debate about the relationship between Central Mexico and the Maya.
In a separate article, Susan Toby Evans brings to life the Aztec palace and then goes on to compare its form to what is known about Teotihuacan's and Tula's palaces and to deliberate what the observed similarities and differences might suggest about state administration. The Aztec palace, called a tecpan in the Nahuatl language, is well known from several Spanish eyewitness accounts. They all agree that the tecpan had three identifying elements: (1) a dais room reserved for the ruler and located on the short side of (2) a courtyard in which probably the noble class gathered. This courtyard, in turn, faced (3) a main plaza in which the common people assembled. The archaeological evidence of tecpans is scarce because very few examples have been fully excavated, but Evans illustrates a city-state tecpan and a village tecpan that seem to share the layout described in the ethnohistorical sources.
She then poses the question of whether the tecpan had antecedents in earlier cultures of Central Mexico, such as Teotihuacan and Tula. As Evans and William T. Sanders outline in their chapter on Teotihuacan (this volume), there may have been a chronological sequence of three palaces, all located adjacent to the Street of the Dead: the earliest was possibly the Xalla compound in the northeastern sector; the second was the Ciudadela; and finally, Teotihuacan's rulers may have occupied the Street of the Dead Compound in the center of the city. A comparison of the plans of these Teotihuacan palaces with that of the Aztec tecpan shows that they do not clearly contain the package of dais/entry courtyard/plaza. Therefore, Evans concludes, Teotihuacan's government system must have been unlike that of the Aztec.
At the site of Tula, many large buildings have been called palaces since the nineteenth century. Evans observes that the Palacio Tolteca was most similar to the Aztec tecpan because it includes the characteristic package: it is next to the main plaza, it has a huge main courtyard, and the entryway faces a raised dais room. She argues that the Aztec copied and appropriated elements of Toltec elite architecture as well as economic and spiritual power and that the function of the tecpan and its Toltec predecessor was largely secular and administrative.
Part 4. Comparison of Palaces across Cultures
The fourth part of the volume consists of two articles that compare the architectural strategies of two cultures and how their leaders made political statements in the design of buildings. In this approach, it is significant whether the social organization of the cultures under consideration is similar—as it is in Carol Mackey's study, which compares two centralized militaristic states, the Chimú and the Inka—or different—as in Jessica Christie's report, which mostly contrasts the Inka with the Maya, who, unlike the Chimú and the Inka, were organized in independent city-states and were less expansionistic.
Carol Mackey's essay is a comparison of the elite architecture at Farfán, a provincial center on the north coast of Peru, during the occupation of two empires—the Chimú and the Inka. As the Chimú expanded their territory from their capital, Chan Chan, to the north, they confronted polities affiliated with the Lambayeque culture. The political and ceremonial center in the Jequetepeque Valley was Pacatnamu. The Chimú subdued Pacatnamu and built Farfán as an intrusive center in the middle of this valley that functioned as the northern frontier of their empire from about A.D. 1200 to A.D. 1350. Intrusive centers typically use an imperial architectural style, which Mackey demonstrates in her analysis of Farfán Compounds II and VI. Both compounds contain U-shaped audiencias, storage rooms, enclosed plazas, and a burial platform, all of which are "loan" elements from Chan Chan. Mackey emphasizes that throughout the history of Farfán, an imperial style was used in architecture as well as in ceramics. These stylistic findings may be an indicator that it was not important for the Chimú to incorporate indigenous lords into the highest echelons of their political system. Instead, the integration of local lords seems to have occurred at lower-ranked sites, many of which have been identified throughout the valley. Given Farfán's size and importance as the seat of the provincial government and the residence of the regional governor, it is striking that it appears to have been staffed by only a small number of high-ranking elites at any given time.
The Inka conquered the Chimú Empire in approximately the 1470s and established their presence as well at Farfán. However, the Inka chose to co-opt Farfán as their settlement. Mackey describes how they changed and altered Compound VI to their interests and liking; for example, they allocated more space for storage, increased security by further restricting access to stored goods, and added a large area of well-planned tapia structures outside and to the west of Compound VI. Altogether, Compound VI and the tapia units housed a larger number of Inka officials than they had of the previous Chimú elite. Mackey stresses that it was not always Inka policy to co-opt existing settlements as their own regional centers, noting that they built intrusive centers on the southern coast. On the north coast, however, the Inka co-opted centers, probably because Chimú sites such as Farfán, Manchan, and Tucume already existed, and it became a matter of convenience to reuse them.
Mackey's comparison clearly illustrates that though the Chimú and Inka were both hierarchically organized, expansionistic states, their methods of establishing control in conquered areas differed, as did some of their political priorities. One could also conclude that there was more than one successful way of administering political control in the Jequetepeque Valley from a faraway capital.
Jessica Christie discusses ways political strategies of the Maya and Inka were expressed in residential architecture. The formal comparison shows that the general layout of houses around courtyards was similar, but that the similarities ended there. Each Inka emperor had to construct his own palace, whereas Maya rulers continued to occupy the palace complexes of their predecessors. From this perspective, the Inka palace may be understood as a monument to a particular ruler, whereas Maya palaces functioned as lineage compounds.
Christie continues to address the ranking of residential architecture in both societies. In Maya studies, various models have been developed that attempt to classify and group houses and sites in a hierarchical order with regard to the social status of their inhabitants (for example, Gordon Willey's model for Barton Ramie and Copán). Other attempts have been made to reconstruct ancient Maya social organization based on the number of high-status residences and their location and distribution. The debate targets the issue of whether the Maya had a segmentary or a unitary state.
The database of Inka residences remains small when compared to that of the Maya. Yet the urban design of many Inka sites, house compounds, masonry, and rock art displays a style that favors geometric elements. Christie discusses numerous examples and concludes that this geometric style often functioned like a signature of the Inka state. Although its origins predate the Inka, they elaborated it into an aesthetic system derived from geometry, which seems to have been employed as a symbol of control and political power over the landscape and subjected peoples.
Christie applies a model developed by Olivier de Montmollin that analyzes the degrees of a segmentary or a unitary organization within a given society but does not endeavor to classify an entire culture in terms of absolute typologies. The comparison of the architectural data suggests that Maya political organization had more elements of a segmentary structure, whereas the Inka system was more unitary.
Thus our volume presents a great variety and complexity of palace structures and cultural scenarios, as well as differing levels of our knowledge and understanding of these cultures. I would conclude that three important variables should be analyzed when studying palaces and power in the Americas: (1) form, (2) a thorough comparison with elite residences, and (3) the culture- and case-specific sources of power.
Form is certainly the starting point for any investigation of palaces. Such a formal analysis involves the tracing of existing walls, determining their relationship to open spaces, and a reconstruction of original walls and possible earlier substructures. In my previous volume (Christie 2003:316-322), I defined four different Maya palace types. In this volume, Chapdelaine (Chapter 1) offers a list of formal elements that might characterize a Moche palace, and Isbell (Chapter 2) presents related but not identical lists of features that would be present in Inka and Pachacamac palaces. He further examines Wari sites with planned orthogonal architecture that consists of rigidly rectangular compounds often resulting in a cellular structure of repeating modular units. Isbell interprets this form of architecture as an indicator of political power at, for example, Pikillacta. A version of planned orthogonal architecture is present at selected Inka sites, such as at Ollantaytambo or Raqchi. However, I think the Inka examples exhibit less formal rigidity and that other factors influenced the design of elite architecture as well. Indeed the Inka generally sought a harmony with nature when they laid out their settlements and designed their buildings. Therefore, planned orthogonal examples are all the more significant. The arguments by Sanders and Evans (Chapter 9) about changes in the government structure of Teotihuacan and Evans's discussion of the Aztec tecpan (Chapter 10) are heavily based on form. Evans investigates whether the basic structure of the tecpan—plaza, courtyard, dais room—occurs in the architecture of earlier cultures in Central Mexico (Tula and Teotihuacan), and based on this comparison, she discusses whether their government systems were similar to that of the Aztec.
All the palaces discussed in this volume seem to have one formal element in common: at least one open plaza space—with one exception. The foregrounded space at Dionisio Point lies within the shed-roof house; it is a central area where status items were found concentrated and where larger, perhaps ceremonial fires were lit (Grier, Chapter 5). The confinement of activities to the interior of communal houses is due to the climatic conditions on the Northwest Coast. Winter was the traditional ceremonial season because temperatures were low and rainfall was high. Thus climate should be considered another palace variable!
The form of the Inka palace remains somewhat elusive to researchers. Some of the chroniclers have described and even drawn examples of their vision of what these palaces may have looked like, and Isbell (Chapter 2) offers a list of features that help define an Inka palace. The problem is that very few cases are fully known archaeologically. A general consensus is that the Inka palace was a large, expanded, and extravagant version of a kancha—which is a courtyard surrounded by four single-room structures and often enclosed by walls—and that elites and commoners lived in more modest versions of the kancha.
Comparisons between Palaces and Elite Residences
Any palace has to be understood in relation to the form and numbers of elite residences. Chapdelaine (Chapter 1) distinguishes between Huaca de la Luna, which clearly exhibits aspects of a palace, and the elite residential compounds located between Huaca de la Luna and Huaca del Sol. In Maya architecture, the formal differences between palaces and elite residences are minor, and one has to focus on size, ornamentation, and location—the latter because palaces typically sit in the ceremonial core of Maya cities (for an in-depth discussion, see Christie 2003). In Chaco Canyon, the roads connecting the Great Houses to the smaller outliers are fundamental to conceptualizing the power of the former. Yet at the same time, it would be too speculative to call the outliers "elite residences," since we know little about social stratification in Anasazi societies (Lekson, Chapter 3; Lekson 1999). Barber and Joyce (Chapter 8) carefully trace the development of elite residences in the Valley of Oaxaca from the Early Formative to the Late Postclassic period. They clearly separate high-status houses from the dwellings of commoners, but they are unable to pinpoint palaces that would incorporate larger open public spaces until the Postclassic. In their view, some of the first and few examples of fully developed Zapotec palaces are the Postclassic Palace of the Six Patios at Yagul and the Group of Columns at Mitla because they clearly combine residential functions and public spaces. Thus the Zapotec scenario paints a picture of power relations quite unlike that of the Maya, who elevated one-man rulership in each city. The Maya system of kingship is mirrored in the identification of one dominant palace in most cities.
Teotihuacan had many higher and lower elite residences, represented by the large number of apartment compounds,1 but palaces are harder to identify. Sanders and Evans (Chapter 9) propose that the Xalla compound, the Ciudadela, and the Street of the Dead Complex were successive palaces. One important aspect that sets these palaces apart from the elite residences is location: they line or straddle the Street of the Dead, whereas most apartment compounds are located farther away from the city's axis. Based on the form and layout of the palace complexes, Sanders and Evans argue for changes in the government structure over time.
Chimú palaces can be securely identified by the ciudadelas at Chan Chan. Mackey (Chapter 11) reports on two compounds that are formally very similar to the Chan Chan examples at Farfán, an intrusive center the Chimú established in the Jequetepeque Valley at the northern border of their empire. At Chan Chan, elite residences accumulated outside the ciudadela walls. Mackey investigates how the Inka reused Compound VI and shows that they added an immense area of tapia (battered mud wall) structures to the west of this compound. These tapia architectural units constituted elite residences, which in this case are neatly separated—spatially and formally—from the palace.
Palaces should be compared to elite residences not only with regard to form but also with regard to numbers. The ratio between the numbers of palaces and elite residences is an indicator of social organization. Thus, in most Maya cities, the number of elite residences increased dramatically during the Late Classic, showing that the absolute power of the ruler declined and that elites probably played an expanding role in the political decision-making process. Barber and Joyce (Chapter 8) report similar data from Monte Albán: during Monte Albán II (100 B.C.-A.D. 200), large numbers of high-status residences were built, which they interpret as the emergence of a decentralized administrative structure.
Finally, the very important issue must be raised that elite residences cannot be lumped into one general category. Already in 1992, Diane and Arlen Chase (1992:7-11) questioned the traditional two-tiered model of grouping Mesoamerican societies into rulers and the ruled or elites and commoners. In their edited volume, they present case studies from throughout Mesoamerica as well as theoretical discussions emphasizing that there was a third intermediate group, or middlemen, and that Mesoamerican political and social organization was a lot more complex than previously assumed. Their findings are relevant to this volume as well and probably also apply to South American and North American societies. The problem is that the data needed to fine-tune elite residences are often lacking and may appear lumped together in scholarly texts. The reader should keep in mind that elites were stratified probably in all precontact and postcontact American cultures.
Sources of Power
Demarest (Chapter 4) emphasizes the important point that Maya rulers relied on various sources of power, which are reflected in the location and the layout of palace architecture. Depending on the site-specific circumstances of different Maya cities, individual rulers might base their power on economic aspects such as agricultural management or long-distance trade, on military planning, or primarily on ideology and theater-state performance. His examples from Dos Pilas, Aguateca, and Cancuén clearly illustrate differing settings of power.
Thus the concept of power should be further analyzed and defined as to what it meant in particular political systems. In the Inka state, one fundamental source of power for individual emperors was private property. Land belonged to the panaca (royal lineage) of an emperor, and each new, succeeding ruler had to acquire his own properties. These private properties were often graced with royal estates to which the ruler would withdraw to relax as well as to carry out more private administrative and political functions (see Niles 1999). The study of such estates and how they articulate power poses an ongoing and fascinating challenge (Burger and Salazar 2004; Christie in preparation). Isbell (Chapter 2) argues that the city of Huari was surrounded by similar rural estates that formed significant political power bases with regard to agricultural wealth produced and the large group of loyal retainers who maintained them. These estates may rank the city of Huari as a type of capital over the administrative centers of Pikillacta and Viracochapampa, which lacked rural estates. Most Inka private properties were located in the Urubamba Valley and stood in a similar relationship to Cusco as the rural estates did to Huari in that example.
Alternate sources of power could be a road system between the center and the periphery (as in the Inka and Chaco Canyon cases) or a deliberate referral back to ancestral models on the part of rulers or elites. Lekson (Chapter 3) presents such a case in which the elites of Aztec Ruins and Paquimé consciously recalled Chaco. Sarro (Chapter 6) shows that one source of power for the elite ruling El Tajín was a connection with the ballgame of their ancestors.
An additional source of power is undoubtedly military might and conquest. Mackey (Chapter 11) demonstrates how two conquest states established different kinds of settlements at Farfán: The Chimú built Farfán as an intrusive center, but the Inka co-opted the existing Chimú settlement. The data highlight opposing strategies employed by two conquest states to establish their presence in newly conquered territories. The case of Farfán speaks to the shortcomings of models and the pitfalls of quick labeling. It shows that conquest states act in individual ways and cannot be lumped into one generic category. For example, Mackey points out that the Inka erected intrusive centers on the south coast, but on the north coast, it was probably more convenient for them to reuse existing Chimú settlements.
To conclude, this brief introduction has made it clear that "palaces and power in the Americas" is still a gigantic mosaic puzzle. Though pieces of data are constantly being added, many empty sections remain, and the picture is far from complete. Lack of information limits the validity of comparisons between many palaces, cultures, and regions. It is also imperative to include variables outside strict architectural parameters, such as sources of power, burials, and others. Gair Tourtellot, Jeremy Sabloff, and Kelli Carmean (1992) reinforce this issue in their analysis of elite architecture at the Maya cities of Sayil and Seibal. They conclude that architectural construction per se is not always a conclusive indicator of status and that archaeologists must search for "still finer discriminations within construction types at similar levels of apparent expenditure" (98). I think the excitement of future research is in doing just that and in filling in diverse pieces of the puzzle and watching a more complete picture of palaces and elite residences emerge.