Journal of Anthropological Archaeology 20, 245–284 (2001) doi:10.1006/jaar.2000.0372, available online at http://www.idealibrary.com on

The Archaeological Study of Empires and Imperialism in Pre-Hispanic Central Mexico

Michael E. Smith and Lisa Montiel

Department of Anthropology, University at Albany, SUNY, Albany, New York 12222

E-mail: [email protected]; [email protected]

Received August 31, 1999; revision received January 6, 2000; accepted August 20, 2000; published online June 1, 2001

The hegemonic-type empires of ancient Mesoamerica are difficult to study archaeologically be- cause they left fewer material traces than more territorially organized empires such as the Inka or Roman cases. We present a material culture model for the identification of such empires using ar- chaeological data. The model, based upon Michael Doyle’s analytical approach to imperialism, is developed from historical and archaeological research on ancient empires from the Old World and South America. Empires can be identified from three types of evidence: characteristics of the capital city, evidence for varying types of political domination of provincial areas, and examples of the pro- jection of influence in a larger, international context. We apply this model to archaeological data on three central Mexican cases—Tenochtitlan, Teotihuacan, and Tula. The results suggest that both Tenochtitlan and Teotihuacan ruled empires, whereas Tula did not. © 2001 Academic Press

The study of empires in pre-Columbian (e.g., the Zapotec or Teotihuacan empires). Mesoamerica has proved to be a difficult Simplistic criteria such as ceramic similar- task for archaeologists and ethnohistorians. ity have too often been used to document In contrast to the Old World, where ancient imperial conquest, a situation recently criti- empires are well attested in the written cized by Zeitlin and Joyce (1999). record, early empires in Mesoamerica arose The subject of ancient empires and impe- among societies with rudimentary writing rialism has seen renewed interest in the past systems that shed little light on political decade. A number of monographs have ap- processes. Furthermore, most Mesoameri- peared recently on individual empires (e.g., can empires were hegemonic in character, Alcock 1993; Berdan et al. 1996; D’Altroy meaning that they relied upon indirect con- 1992; Liverani 1993; Millett 1990; Morrison trol and invested few resources in provin- 1995; Schreiber 1992). Comparative studies cial infrastructure (Hassig 1985). As a re- are becoming more common (e.g., Alcock et sult, they left far fewer durable material al. 2001; Algaze 1993; Blanton 1996; Cherry remains in their provinces than did territor- 1992; Schreiber 1999). As a result of in- ial, or direct-rule, empires such as the Wari creased attention to ancient empires, archae- or Inka empires of Andean South America ologists have made considerable progress in (D’Altroy 1992; Malpass 1993; Schreiber devising methods to analyze various as- 1992). In the absence of rigorous archaeo- pects of the expansion, organization, and logical criteria for the identification of an- collapse of empires (e.g., Adams 1979; cient empires and imperialism, Mesoameri- Cherry 1992; Costin and Earle 1989; D’Al- canists have been free to ascribe empires to troy 1992; Earle 1994; Morrison and Lycett societies that almost certainly lacked them 1994; Schreiber 1999; Sinopoli 1994; Sinopoli (e.g., the Olmecs or the Toltecs), and to and Morrison 1995; Smith and Berdan 1992; deny the existence of empires when there is Stark 1990). One thing is lacking in this grow- in fact strong evidence that they existed ing body of literature, however—a rigorous

245 0278-4165/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved. 246 SMITH AND MONTIEL archaeological method for identifying em- ture of imperialism. Unfortunately, political pires using material remains. In most parts processes are among the most difficult for of the Old World, this is not a pressing need. archaeologists to identify and analyze di- One hardly needs to develop sophisticated rectly. Our approach is to consider some of methods to determine whether the Romans, the social and economic expressions of an- Assyrians, or Achaemenids ruled empires. cient imperialism to identify material evi- But in Mesoamerica, with its hegemonic dence of a type that archaeologists can use. empires, there is a real need for such a We find Doyle’s analytical approach, which method (see Flannery 1998 for a similar ap- incorporates spatial pro-cesses and vari- proach to the archaeological identification ables, useful for organizing our discussion: of pristine or archaic states). Four intersecting sources account for the imper- Our goal in this paper is to present an ex- ial relationship: the metropolitan regime, its ca- plicit material culture model for the identi- pacities and interests; the peripheral political soci- fication of empires and imperialism using ety, its interests and weakness; the transnational archaeological data, and to apply that system and its needs; and the international con- text and the incentives it creates. (Doyle 1986:46) model to three central Mexican test cases, Teotihuacan, Tula, and Tenochtitlan. Our Doyle’s “metropolitan regime” refers to model is based on historical and archaeo- the political, economic, and social dynam- logical evidence from ancient empires ics of the imperial capital and core society. around the world. When the model is ap- We modify this factor to create the first of plied to archaeological evidence from these three general criteria of imperialism, a capi- three Mesoamerican societies, the results tal city, sufficiently large and complex to suggest strongly that both Teotihuacan and rule an empire, that exhibits material evi- Tenochtitlan ruled empires, whereas Tula dence of an imperial ideology. The “periph- did not. The so-called “Toltec empire” eral political society” refers to the condi- based at Tula was an invention of the tions in the provincial areas conquered or Aztecs, and modern scholars have been dominated by an empire. Although crucial misled by an overreliance on Aztec mythi- to the analysis of any specific case of impe- cal history at the expense of archaeological rialism, we do not find this factor useful for data. After discussing the three cases, we the purpose of identifying the existence of briefly explore the relevance of our model ancient imperialism. Empires conquered all to some other likely ancient empires of the sorts of provincial polities, from small, non- New World, including the Zapotec and hierarchical groups to other empires, and Tarascan empires of Mesoamerica and the the archaeological identification of imperi- Wari empire of South America. alism must start with imperial impact on the provinces, not the indigenous situation AN ARCHAEOLOGICAL MODEL OF that preceded incorporation. EMPIRES AND IMPERIALISM Doyle’s third factor, the “transnational system,” refers to the nature of interactions We follow political scientist Michael between the capital and the provinces. We Doyle’s (1986) definition of empires and im- use this element, under the label, “domina- perialism: “I favor the behavioral definition tion of a territory,” as the second compo- of empire as effective control, whether for- nent of our model. We divide transnational mal or informal, of a subordinated society processes into two categories, economic ex- by an imperial society” (Doyle 1986:30). change (between capital and provinces) This definition, like most others in the litera- and political control. Doyle’s fourth factor, ture (e.g., Larsen 1979; Luttwak 1976; Mann the “international context” of an empire, 1986; Sinopoli 1994) stresses the political na- refers to political and economic dynamics ARCHAEOLOGICAL STUDY OF EMPIRES 247 of the larger geopolitical setting in which perial capital, and these cities almost al- empires expand and operate. We narrow ways contained durable displays of imper- this theme to the notion that empires pro- ial ideology, many of which have survived ject various kinds of influence—economic, to the present. These two features—urban political, and cultural—beyond their bor- size and complexity, and material procla- ders. mations of imperial ideology—provide a Table 1 presents a summary of our ar- starting point in the archaeological identifi- chaeological model for the identification of cation of ancient empires. ancient empires. This is a polythetic defini- tion in the sense that most ancient empires A. Large, Complex Urban Center exhibited most of these traits, although not every empire exhibited all of them. In the Imperial capitals were among the most following section we summarize the evi- prominent settlements of the ancient world, dence for these traits among a variety of an- and today the remains of cities such as cient empires, and after that we apply this Rome, Athens, Xian, Persepolis, Vijayana- model to the three case studies. gara, and Cuzco survive as some of the most spectacular archaeological sites 1. The Imperial Capital around the world. These capitals are im- Most ancient empires had a large and pressive not only for their size and complex urban center that served as the im- grandeur, but also for the evidence of social

TABLE 1 Archaeological Criteria for the Identification of Empires

Features Examples 1. The imperial capital A. Large, complex urban center B. Proclamations of imperial ideology 1. Militarism 2. Glorification of king or state

2. Domination of a territory A. Economic exchange between capital and provinces 1. Provincial goods found at capital 2. Imperial goods found in provinces B. Political control of provinces 1. Military conquest 2. Construction of imperial infrastructure 3. Imposition of tribute or taxes 4. Reorganization of settlement systems 5. Imperial co-option of local elites

3. Projection of influence in a larger international context A. Economic influence 1. Trade with extraimperial regions B. Political influence 1. Military engagement and activities along enemy borders 2. Centralization or militarization of extraimperial polities C. Cultural influence 1. Adoption of imperial gods or rituals by distant peoples 2. Emulation of imperial styles and traits by dis- tant peoples 248 SMITH AND MONTIEL complexity revealed by excavations. This et al. 1996), often in a durable form that en- urban complexity typically included the ex- sured its survival from ancient times to the istence of multiple social classes, occupa- present. tional specialization, and ethnic variation, among other types of social variability (e.g., 2. Domination of a Territory Fritz et al. 1984; Hyslop 1990; Osborne 1987; We consider issues in the imperial domi- Owens 1991; Stambaugh 1988; Steinhardt nation or control of its territory under two 1990). Although there are one or two cases headings, exchange and political control. of political empires without major capital All empires exhibited trade between the cities, such as the Carolingian empire capital or core zone and their provinces, al- (Moreland 2001), we suggest that in though the two directions of exchange were Mesoamerica the opposite pattern—large often not equivalent in value (i.e., exchange complex cities that were not imperial capi- was one means of domination). The politi- tals—is far more likely. cal control of provinces forms the core of the imperial relationship, but as mentioned B. Proclamations of Imperial Ideology above this is one of the more difficult processes to monitor with archaeological Almost without exception, the rulers of data. ancient empires invested considerable en- ergy in producing public statements of im- A. Economic Exchange Between Capital and perial ideology. Although the content of Provinces state or imperial ideology varied with indi- vidual cases, all known cases included mili- Written documents and archaeological tarism and/or glorification of the king or excavations reveal the presence of imported state among their prominent themes. These provincial goods in virtually all known an- public symbolic proclamations did not al- cient imperial capitals (e.g., Osborne 1987; ways reflect reality, and they are best Stambaugh 1988). For the present purposes viewed as examples of imperial propa- it is not important whether these goods ganda. Empires relied upon military force, were obtained through coercion (i.e., taxa- and militarism and military glory were tion or tribute) or through commercial ex- prominent themes in imperial capitals. This change; what is important is that they pro- expression took a variety of forms in an- vide clear material evidence for economic cient empires, including public images of interaction between capital and provinces. battle, soldiers, and military themes (e.g., Conversely, imperial goods are commonly Cook 1983; Cotterell 1981), public royal encountered in provincial contexts (Costin proclamations (Larsen 1979), and the con- and Earle 1989; Millett 1990; Woolf 1992). struction of symbolic fortresses in the capi- tal (e.g., the Sacsahuaman fortress in Cuzco B. Political Control of Provinces or the inner compounds of Chinese imper- ial cities; Hyslop 1990, Steinhart 1990). Sim- Territorial empires and hegemonic em- ilarly, imperial capitals typically contained pires employed different forms of provin- public monuments designed to glorify the cial control, and within individual empires ruler or the state, often by establishing a a variety of different means of control were correspondence between the city or empire often present. In this section we review the and the cosmos (Bauer 1998; Briggs 1951; most common forms of provincial control Fritz et al. 1984; Root 1979). These public used by ancient empires. proclamations of state messages served to 1. Military conquest. Most ancient empires materialize imperial ideology (DeMarrais expanded through military conquest (or ARCHAEOLOGICAL STUDY OF EMPIRES 249 threat of conquest), but this process left few mented by archaeologists (Costin et al. direct material remains. Cities and towns 1989). Imperial taxation could lead to a low- were rarely destroyed entirely, since most ering of standards of living in provincial empires were more interested in controlling areas, and again this is a process that ar- provincial populations than in destroying chaeologists can document with excava- them. Damages to buildings and settle- tion. The difficulty with this kind of evi- ments from the process of conquest were dence for imperialism is that all of these probably repaired soon after the fact, leav- changes—agricultural intensification, craft ing few or no archaeological traces. In dis- intensification, and lowered standard of liv- tinction to the low archaeological visibility ing—can also result from processes other of initial military conquests, continuing than imperial taxation (see Smith and military control of provincial areas did Heath-Smith 1994 for discussion). Thus leave archaeological traces in many cases; such changes alone do not constitute clear this is included under our next category. evidence for imperial taxation. 2. Construction of imperial infrastructure. 4. Reorganization of settlement systems. Em- Territorial empires such as the Roman and pires often move people around in order to Inca cases invested considerable resources better control them, to lessen chances of re- in building a provincial infrastructure of volt, or to achieve particular economic towns, garrisons, fortresses, roads, ends. When such changes are systematic bridges, and the like (Alcock 1993; Hasel- and/or large in scale they can be reflected grove 1987; Hopkins 1978, 1984, 1990; Mil- in archaeological settlement patterns (D’Al- lett 1990; Wells 1984). The remains of these troy 1992; Topic and Topic 1993). A common imperial construction projects are typically pattern is forced nucleation, where rural the most obvious and dramatic evidence peoples are moved into towns where they for ancient imperialism in such cases. In are more easily monitored and controlled. hegemonic empires, by contrast, imperial The 16th-century practice of congregación in infrastructure was left to a minimum and Spain’s New World imperial provinces pro- archaeologists must rely upon evidence of vides some dramatic examples of forced a more indirect nature to document politi- nucleation whose results are still visible in cal control. settlement patterns today (e.g., Gerhard 3. Imposition of Tribute or Taxes. Nearly all 1977; Gibson 1966). ancient empires taxed their subjects in 5. Imperial co-option of local elites. Empires, some form, and such taxation often pro- whether territorial or hegemonic in nature, duces material remains that can be recov- typically “buy off” provincial elites with ered archaeologically. Provincial peoples gifts and privileges in order to gain their co- may have had to increase their agricultural operation in administering the provinces. production to meet their taxes, resulting in In hegemonic empires, or on the frontiers of the intensification of farming through the various kinds of empires, this process can adoption of new techniques or the expan- result in the creation of client states that are sion of existing methods. Many such meth- only weakly under the control of the em- ods, including terracing and irrigation, are pire (Braund 1984; Isaac 1990; Morkot 2001; archaeologically visible (Hopkins 1978; Postgate 1992); for example, client states Hopkins 1980; Morrison 1995; Redmond did not pay taxes in the Roman empire. 1983; Sinopoli and Morrison 1995). Simi- This situation can be identified archaeologi- larly, imperial subjects may have had to in- cally by the emulation of imperial styles by crease their production of manufactured provincial elites, or by the presence of high- goods for tax purposes, and changes in the value imperial imports in provincial elite intensity of craft production can be docu- contexts (Postgate 1992:258; Kuhrt 2001). 250 SMITH AND MONTIEL

3. Projection of Influence in a Larger states as they organized themselves to deal International Context with the threat of a nearby expanding em- Empires exist within a larger interna- pire (Edens 1992). This can be determined tional context, and they interact in various archaeologically through evidence for state ways with other independent polities. One formation or centralization concomitant feature that differentiates empires from with the expansion of a nearby empire. smaller states is that empires almost always C. Cultural Influence exert various kinds of identifiable influence on other polities in the larger system. The Because expanding empires are often per- world-systems approach, as modified for ceived by outsiders as prestigious or power- premodern societies, provides a useful ful, distant peoples may adopt imperial gods framework for viewing the role of empires 1, or rituals in the absence of direct conquest or within their larger international context. * incorporation or prior to conquest. Similarly, We discuss the extraimperial influences of distant elites or peoples may emulate imper- empires under the headings of economic, ial styles or adopt imperial practices or traits. political, and cultural influence. Whittaker’s (1994) study of Roman frontiers shows that these processes were quite com- A. Economic Influence mon in extraimperial areas that were in con- tact with the empire or with frontier zones, Most ancient empires traded with exter- and they can leave material traces for ar- nal, independent polities, and these eco- chaeologists to recover. This evidence nomic exchanges formed the backbone of should not be confused with evidence for ancient world systems. This kind of trade political control within an empire’s has been most extensively studied for the provinces, however. We suggest below that Roman empire, whose goods traveled all confusion of this sort has hindered the over the world (Begley and de Puma 1991; proper evaluation of Teotihuacan’s role in Whittaker 1983, 1994). Another example is Classic-period Mesoamerica. Akkadian and Ur III trade with distant areas in the Arabian Gulf and elsewhere INTRODUCTION TO TENOCHTITLAN, (Edens 1992; Edens and Kohl 1993). TEOTIHUACAN, AND TULA

B. Political Influence Teotihuacan, Tula, and Tenochtitlan were all large urban centers whose influence was The expansion of empires often had po- felt throughout many regions of Mesoamer- litical effects on nearby independent poli- ica (Fig. 1), far beyond the extent of their ties. When an empire encountered a power- local polities. In this section we provide an ful enemy along its borders, military introduction to these polities. activities were common occurrences in the frontier zone (Whittaker 1994). If the fron- Tenochtitlan tier was stable for any length of time, fortifi- cations could be built by one or both sides Tenochtitlan is a good test case for our and these are common archaeological fea- model, since documentary sources clearly tures on the edges of empires (Bartel 1980; indicate the existence of an empire centered Hyslop 1990; Millett 1990). Another com- at this city. Some scholars have hesitated to mon effect of imperialism was a process of call the Aztec polity an empire (e.g., “this political centralization among independent was not an empire at all in the very strictest sense” (Davies 1973:110)) because of its lack *See Notes section at end of paper for all footnotes. of a provincial infrastructure and standing ARCHAEOLOGICAL STUDY OF EMPIRES 251

FIG 1. Map of Mesoamerica showing distant cities that interacted with Tenochtitlan, Teotihuacan, and Tula. Key to obsidian sources: (1) Pachuca, (2) Otumba, (3) Ucareo, (4) El Chayal. army. Hassig, however (1985:90), shows documentary record we would never know that such views merely reflect various au- that the Aztec empire existed at all. We offer thors’ inadequate understanding of ancient the present analysis to counter this sugges- empires; the Aztec empire was not a territo- tion. Our map of the central portion of the rial empire like the Late Roman or Inca em- Aztec empire (Fig. 2) is based upon the pires, but was a hegemonic empire whose maps presented in Berdan et al. (1996). organization (and material remains) were According to Aztec native historical tra- quite different. The lack of obvious archae- dition, Tenochtitlan was founded by the ological visibility of the Aztec empire in Mexica people in the early 14th century, comparison with territorial empires has led A.D. on a swampy island in the salt lakes of some scholars to suggest that without the the Basin of Mexico. In 1428 Tenochtitlan

FIG 2. Map of the extent of the Aztec empire in central Mexico (based upon Berdan et al. 1996). 252 SMITH AND MONTIEL achieved political independence from its A number of archaeological studies pro- Tepanec overlords and immediately began duced settlement pattern and artifactual imperialist expansion by conquering data suggesting that Teotihuacan exerted nearby city-states. In the traditional view, strong political and economic control over the Aztec empire was run by a Triple Al- the Basin of Mexico and perhaps other liance of Tenochtitlan, Texcoco, and Tlaco- nearby areas (e.g., Charlton 1991; Hirth pan (Carrasco 1996; Davies 1987), and over 1980; Sanders et al. 1979). Sanders and oth- time Tenochtitlan emerged as the dominant ers have remarked that Teotihuacan ap- partner. In a recent paper, Gillespie (1998) pears to have maintained stronger, more di- has challenged the existence of the Triple rect political control over the Basin of Alliance, arguing that Tenochtitlan was the Mexico than did Tenochtitlan in Aztec sole capital city all along. The conquests of times (Sanders et al. 1979; Sanders and the Mexica kings are recorded in a series of Santley 1983), but none of these authors ex- prose and pictorial documents, as are the plicitly call the Teotihuacan polity an em- tribute payments made by the Aztec pire. Millon (1988) was the first to assemble provinces (Berdan and Anawalt 1992; a range of data on Classic-period central Berdan et al. 1996). Mexico and argue that Teotihuacan was the Prior studies of the material manifesta- capital of an empire in the central Mexican tions of Aztec imperialism have focused on highlands. He later stated that “there was the art and architecture of Tenochtitlan no Teotihuacan empire” (Millon 1994:28), (Broda et al. 1987; Townsend 1979; Um- but we assume that he is referring here to a berger 1996), sculpture and other art in the Mesoamerica-wide empire. New data on provinces (Umberger and Klein 1993), the the Teotihuacan affiliations of early rulers distribution of trade goods (Smith 1990), of the Maya cities of Tikal and Copan have and the effects of Aztec conquest on provin- caused at least some scholars to return to cial society (Silverstein 2000; Smith 1992, the earlier model of Teotihuacan control of 1997b, 2001, 1992). the Maya area (Coe 1999:83–84), but we feel that this interpretation is premature. We Teotihuacan show below that Millon’s (1988) model best fits the archaeological data. The great Classic-period metropolis of Teotihuacan flourished in the northeastern Tula Basin of Mexico from the second through seventh centuries, A.D. It has long been Our discussion of Tula as a possible im- clear that Teotihuacan had important eco- perial capital begins with the Aztec ac- nomic, stylistic, and perhaps political influ- counts of Tula and its Toltec rulers and in- ence throughout Mesoamerica, and the habitants. The Aztecs had a highly inflated study of these influences has been a major view of the accomplishments and grandeur research theme by archaeologists and art of the Toltecs, and the scholars today who historians (e.g., Berlo 1992; Pasztory 1978; accept the existence of a Toltec empire base Santley 1984). Early views of a Mesoamer- their views more on the exaggerated mythi- ica-wide empire centered at Teotihuacan cal accounts of the Aztecs than on empirical (e.g., Bernal 1966; Borhegyi 1971) were archaeological data. The Aztec kings traced abandoned in the 1960s and 1970s in favor the legitimacy of their dynasties to their (al- of interpretations that stressed trade with leged) descent from the Toltec kings. Aztec Oaxaca, the Guatemalan coast and high- accounts of the Toltecs merge myth and his- lands, and other distant regions, and stylis- tory, and the Toltecs were said to have in- tic influence in the Classic Maya lowlands. vented the calendar and the various arts ARCHAEOLOGICAL STUDY OF EMPIRES 253 and crafts, interpretations that are clearly logical project (e.g., Cobean 1990:503–510; false. Here are some of the descriptions of Diehl 1993:283; Healan 1989:4), among the Toltecs given in Sahagún’s Florentine whom the following statements are typical: Codex: “The Toltecs were imperialists motivated by economic goals . . . their brief but spec- The Tolteca were wise. Their works were all tacular career as an imperial power set the good, all perfect, all wonderful, all marvelous . . . They invented the art of medicine . . . And these pattern for later Aztec ventures along the Tolteca were very wise; they were thinkers, for same lines” (Diehl 1983:140); “Tula con- they originated the year count . . . These Tolteca trolled a tributary empire comparable to the were righteous. They were not deceivers. Their later, larger Aztec empire” (Healan 1993:459). words [were] clear words . . . They were tall; they Other scholars who accept uncritically the were larger [than the people today] . . . They were very devout . . . They were rich. (Sahagún, existence of a Toltec empire include Carrasco 1950–1982: book 10, 165–170) (1999:59), Coe (1994:131–132; 1999:167), Fowler (1989:274), Nicholson (2000:155), Most modern interpretations of a Toltec Noguez (1995:206–214), and Weigand et al. empire are based upon a speculative 1961 (1977:22). This trend continues unabated; at a study by Paul Kirchhoff (reprinted as Kirch- recent symposium at the 2000 Annual Meet- hoff 1985). Kirchhoff interpreted some vague ing of the Society for American Archaeology and contradictory accounts in Spanish colo- (“Tula and the Toltec World: Interregional In- nial sources on Aztec native history (princi- teraction During the Early Postclassic,” orga- pally Chimalpahin, the Anales de Cuauhtit- nized by Robert Cobean), at least five partici- lan, and the Historia Tolteca-Chichimeca) as pants mentioned the “Toltec empire” in suggesting the existence of a large empire passing as if it were an established entity. based at Tula that controlled much of north- Some scholars are more critical of the no- ern Mesoamerica. Davies (1977:296–312) re- tion of a Toltec empire, invoking long-dis- views the textual evidence and Kirchhoff’s tance trade to explain contacts between Tula interpretations and points out numerous and distant areas rather than conquest and problems and points of confusion. He then political expansion. Hassig (1992:110–120), proceeds to ignore his own criticisms by pre- for example, notes the lack of contemporary senting an imaginative reconstruction of the written accounts of the Toltec polity and con- fragmentary and contradictory data, includ- cludes that “The available evidence suggests ing a map of the purported Toltec empire not so much a military as a trading empire, (1977:312–328). Davies’s map is shown in one that operated through merchant en- Fig. 4, along with our reconstruction of the claves and settlements rather than military extent of the Toltec regional state (see discus- colonization of outlying areas” (Hassig sion below). Feldman (1974) presents an 1992:117); similar views are expressed by analysis similar to that of Davies. Kelley Lee (1978:293), who suggests that Toltecs (1992) bases a claim for a Toltec empire upon may have controlled the trade of Plumbate tenuous correlations among royal genealo- pottery in Early Postclassic Mesoamerica. gies from the Mixtec region as depicted in Among the few scholars openly critical of pictorial codices. the notion of a Toltec empire are Prem Many scholars apparently accept the in- (1997:142); Weaver (1993:405), who states, terpretations of Kirchhoff, Davies, and “It is best to think of a Toltec sphere of in- Feldman uncritically, speaking of a “Toltec fluence rather than an empire”; Cowgill empire” as if it were an established fact that (2000:295), who suggests that “the area needs no empirical justification. This de- dominated by the Tula state was probably scription applies to a number of members never very extensive”; and López Austin of the University of Missouri Tula archaeo- and López Luján (2000:69), who state that 254 SMITH AND MONTIEL

“Tula was not the capital of a pan- Montiel n.d.). As indicated in Fig. 2–4, this Mesoamerican empire.” Jones (1995:44–60) area was part of the Aztec empire (Berdan provides a useful historical review of schol- et al. 1996); it was part of the Teotihuacan arly attitudes toward the Toltecs, including empire in the models of Millon (1988) and changing opinions on the existence of a Montiel (n.d.); and it was part of the Toltec Toltec empire. empire as reconstructed by Davies (1977). Most scholars accept the identification of The Yautepec Valley therefore provides a the Early Postclassic archaeological site of good test case for the incorporation of pre- Tula, Hidalgo, with the Toltec capital city sumed provincial areas into the three poli- Tollan as described in the Aztec histories,2 ties under consideration. Although many of and the archaeological evidence that we the analyses from these fieldwork projects discuss is from Tula and other Early Post- are not yet complete, we do have enough classic sites in central Mexico. data available to evaluate models of imper- ial incorporation. Because we did not exca- The Yautepec Valley as a Provincial Area vate any Early Postclassic contexts, we will use single-component surface collections We have conducted several fieldwork from that period to compare to our Classic projects in the Yautepec Valley, including and Late Postclassic excavations. excavations of Aztec-period houses at the city of Yautepec (Smith et al., 1999), a full- THE IMPERIAL CAPITAL coverage survey of the Yautepec Valley with intensive surface collections (Cascio et Tenochtitlan al. 1995; Hare et al. n.d.; Hare 2000; Montiel n.d.), and test excavations at a sample of Although Aztec Tenochtitlan today lies Classic and Postclassic sites (Hare 2000; buried under Mexico City, available archae-

FIG 3. Map of the extent of the Teotihuacan empire (based upon Millon 1988 and Montiel n.d.). ARCHAEOLOGICAL STUDY OF EMPIRES 255

FIG 4. Map of the extent of the Tula polity and Davies’s map of the “tentative boundary of the Toltec empire” (Davies 1977; see text for discussion). ological evidence confirms the ancient rial ideology at Tenochtitlan. The existence city’s imperial status (for a description of of an orthogonal grid plan covering the is- Tenochtitlan from historic sources, see land city is also consistent with Tenochtit- Rojas 1986). Salvage excavations along the lan’s imperial status (Calnek 1976; Smith routes of Mexico City’s expanding under- 1997a). ground metro, although poorly published, are sufficiently extensive to demonstrate Teotihuacan the large size of Tenochtitlan. An imperial level of social complexity is suggested by The huge size and high level of social com- the variability in architecture excavated at plexity at Classic-period Teotihuacan are well Tenochtitlan, including simple houses, attested in the abundant record of recent small temples, and massive state pyramids archaeological fieldwork at the site (e.g., and other civic-religious buildings (Matos Cabrera Castro et al. 1991a; Cowgill 1997; Mil- Moctezuma 1979, 1988, 1999). The presence lon 1981). An imperial ideology that included of an explicit imperial ideology is indicated prominent militaristic themes is evident in by massive stone carvings that portray im- many of the polychrome mural paintings, fig- ages of militarism, sacrifice, and the cosmic urines, and sculptures found throughout the glorification of the state and key emperors ancient city (Barbour 1979; Cabrera Castro et (Townsend 1979; Umberger 1996). This ide- al. 1991b; Cowgill 1997; Miller 1967, 1973; ology has been a central focus of investiga- Sugiyama 1998). The military iconography in tion by scholars working on the Aztec Tem- murals includes figures with tasseled head- plo Mayor (Broda et al. 1987; Matos dresses and ringed eyes (Millon 1973), Moctezuma 1988), and Brumfiel’s (1998) bird/owl imagery, and representations of the analysis explores the social context of impe- Storm/War God, felines, spears, and shields 256 SMITH AND MONTIEL

(Berrin 1988; Cabrera Castro 1992). Military out following a grid that covered part of all of orders are associated with owl, eagle, coyote, the urban area, a pattern otherwise quite rare jaguar, and serpent imagery (Miller 1967; in ancient Mesoamerica (most Mesoamerican Von Winning 1948; Von Winning 1987:94). cities exhibited formal planning only in their Unlike in many ancient empires, individual ceremonial cores, and residential zones do leaders were not depicted publicly at Teoti- not show evidence of planning). Compara- huacan, a situation much discussed in the re- tive historical research suggests that the use cent literature (e.g., Blanton et al. 1996; of urban grids is usually related to strongly Cowgill 1997; Pasztory 1997). As at Tenochti- centralized political control (Carter 1983; tlan, the high degree of urban planning sug- Lynch 1981). All three of the cities under con- gested by the extensive orthogonal grid lay- sideration here also contained prominent out of Teotihuacan is consistent with city’s public art with themes of militarism and the status as an imperial capital. glorification of the state. In the case of Tenochtitlan, documentary Tula and art historical evidence shows that its Mexica rulers deliberately imitated the art Although much smaller and less socially and architecture of Teotihuacan, Tula, and complex than Teotihuacan or Tenochtitlan, Xochicalco, a large urban center in Morelos Tula was still an impressive urban center of that flourished between the dominant peri- 11 sq. km, with 30,000 to 40,000 inhabitants ods of Teotihuacan and Tula (Umberger (Diehl 1983:58–60). Excavations have re- 1987, 1996). vealed evidence of social complexity in the The Toltec rulers of Tula may similarly form of varied residential architecture have imitated Teotihuacan in their use of the (Healan 1989) and workshops for obsidian grid pattern and other urban features. These and ceramics (Healan et al. 1983; Hernán- three cities participated in a central Mexican dez et al. 1999). The large central plaza is tradition of imperial city planning in which surrounded by various monumental struc- art and architecture made references to the tures, including two temple-pyramids, a past and were used by rulers to proclaim po- ballcourt, and other civic buildings. Carved litical messages of power, greatness, and cos- stone reliefs with depictions of warfare, sac- mic legitimacy (Smith 1997a). For the Toltecs, rifice, and other militaristic themes are however, we believe that these messages prominent in the urban core of Tula (de la were more empty propaganda than social re- Fuente et al. 1988; Kristan-Graham 1999). ality, since the notion of a Toltec empire fails An orthogonal grid layout was also present on the criterion of domination of a territory. at Tula, but it may not have covered the en- tire extent of the city. The center of the early, DOMINATION OF A TERRITORY: Corral-phase, city at Tula exhibited a grid EXCHANGE BETWEEN CAPITAL AND layout. When the later, Tollan-phase city PROVINCES center was built, a new grid orientation was established, at least for the central city lay- Tenochtitlan out (Mastache and Crespo 1982; Mastache and Cobean 1985:274–285, 1989:62–63). There is overwhelming archaeological evidence for economic exchange between Discussion Tenochtitlan and its provinces, and this is a topic for which archaeology surpasses eth- Teotihuacan, Tula, and Tenochtitlan were nohistory in the quantity and quality of the all sufficiently large and complex to have evidence. The most spectacular archaeolog- served as capitals of empires. Each was laid ical finds are the offerings from the Templo ARCHAEOLOGICAL STUDY OF EMPIRES 257

Mayor, which contained Mezcala-style and not from other places in the Basin of stone sculptures from Guerrero, a variety of Mexico. Several ceramic types (e.g., Texcoco species of marine fauna from the Atlantic Fabric-Marked salt vessels and Xochimilco and Pacific coasts, and ceramics from di- Polychrome jars) probably originated in verse regions (López Luján 1994; Román other parts of the Basin, whereas Aztec-paste Berrelleza and López Luján 1999). The of- ceramic figurines could be from either ferings in the so-called “volador deposit” Tenochtitlan or other places in the Basin. near the Templo Mayor included ceramics Guinda or redware bowls, pitchers, and cups from Morelos and other provincial areas were probably manufactured in both More- (Smith n.d.b; Solís Olguín and Morales los and the Basin of Mexico, and both types Gómez 1991), and sherds from a variety of are found at sites in Morelos. Chalco/ provincial ceramic types have been recov- Cholula Polychrome tripod plates were im- ered in various salvage excavations in Mex- ported from a number of areas, including ico City (seen by Smith in the laboratories Chalco in the Basin of Mexico and various lo- of the Departamento de Salvamiento, Insti- cations in the Puebla/Tlaxcala region (Smith tuto Nacional de Antropología e Historia, et al., 1999). Green obsidian from Pachuca Mexico City). Obsidian from several source predominates at Aztec sites in Morelos areas in the Aztec provinces is abundant in (where it generally comprises over 90% of all Mexico City excavations, and the presence obsidian), and gray obsidian from the of ceramic spindle whorls for spinning cot- Otumba source area in the Basin of Mexico is ton in residential contexts (Cepeda Cárde- also abundant (Smith, unpublished data). nas et al. 1977) points strongly to the impor- The Basin of Mexico was the Aztec core zone, tation of raw cotton from provincial areas and even if a Tenochtitlan origin cannot be like Morelos and the Huaxteca (cotton does established, these varied imports do provide not grow in the Basin of Mexico). evidence for exchange between the core and The most secure evidence for the exporta- the provinces. Although we suggest that this tion of goods from Tenochtitlan to the Aztec commerce in ceramics, obsidian, and other provinces is provided by Aztec III black-on- goods was related to Aztec imperialism, we orange ceramics. Characterization research do not mean to suggest that this exchange by Mary Hodge and colleagues (Hodge et al. was controlled by the state; in fact it was al- 1993) has shown that vessels of this distinc- most certainly conducted through commer- tive ceramic type were produced in several cial channels independent of direct state con- parts of the Basin of Mexico, including trol (Isaac 1986; Smith n.d.b, 2001; Smith and Tenochtitlan. Aztec III sherds from the Berdan 2000). Against this view, Pastrana provincial city of Yautepec include examples (1998) argues that the Aztec empire con- from the Tenochtitlan production area, as trolled the mining, tool production, and ex- well as other areas (Neff and Glascock 1996). change of obsidian, although he cites no evi- This type is found in many parts of the em- dence for this supposed control. pire (Smith 1990), and many of those vessels were produced in the Tenochtitlan area Teotihuacan (Smith, Neff and Fauman-Fichman 1999). Smith’s excavations at Yautepec and other The exchange of goods between Teoti- Aztec provincial sites in Morelos uncovered huacan and its provinces differed from the numerous other artifacts originating in the exchange outside of the empire in quantity Basin of Mexico, but given our uncertainty and context. In the provinces, goods origi- about Aztec exchange mechanisms and net- nating from or controlled by Teotihuacan works it cannot be assumed that all of these include Thin Orange ceramics, censers, fig- were imported directly from Tenochtitlan urines, and Pachuca obsidian (particularly 258 SMITH AND MONTIEL in blade form). Rattray and Harbottle (1992) obsidian became the dominant lithic mate- have demonstrated that the manufacture of rial in the Classic period, a trend that re- Thin Orange took place in Puebla, an area versed in the Epiclassic period (after the fall Rattray claims was outside of the Teotihua- of Teotihuacan). The evidence from More- can empire due to the lack of intrusive los suggests that Pachuca obsidian was im- Teotihuacan cultural elements (Rattray ported into the area already in blade form 1990). The overwhelming monopolization (Montiel n.d.). by Teotihuacan of the distribution of Thin Decorated censers were also imported Orange does, however, strongly suggest into the provinces (Díaz Oyarzábal 1980; that this area was giving Thin Orange as Montiel n.d.). The censers from the Yaute- tribute to Teotihuacan. While Thin Orange pec area are so similar stylistically and tech- was traded outside of the empire in ex- nologically to Teotihuacan censers, that in tremely low quantities as a prestige item, a the absence of characterization analysis it is much higher frequency of this import in difficult to determine if they were locally Teotihuacan’s periphery coincides with in- made imitations or imports. We consider tegration into the empire. Thin Orange these censers as imperial goods based upon comprises 18% of the Classic period ceram- the identification of a large censer work- ics at the provincial administrative site of shop associated with the Ciudadela in Chingu during the Early Classic period Teotihuacan (Sugiyama 1998). Although (Díaz 1980:36) and 2–4% of the ceramics Teotihuacan-type censers are found in dis- from the Classic period in the Yautepec area tant areas outside of the empire (Berlo (Montiel n.d.). Rattray (1981, 1998) reports a 1984), their presence is isolated from other Thin Orange frequency of 3–13% from test indicators of Teotihuacan contact and such excavations of the Teotihuacan Mapping examples are probably local imitations of Project. The distribution of Thin Orange is Teotihuacan censers. Figurines have been also an important indicator of its imperial extensively described for two areas in the nature. At sites within the imperial periph- Teotihuacan empire, Chingu (Díaz ery, Thin Orange is present throughout Oyarzábal 1991) and the Yautepec Valley most residential sites, yet at distant sites (Montiel 1999). These are identical to fig- like Kaminaljuyu, Copan, and Tikal, Thin urines found at Teotihuacan (Barbour Orange is mainly found in isolated (elite) 1979). The presence of molds in these areas funerary contexts. suggests that at least some of these fig- Although the Otumba obsidian source is urines were produced in the provinces. It is only about 15 km from Teotihuacan, the possible that Teotihuacan supplied the Pachuca obsidian source, located 50 km molds. northeast of the city, appears to have been Teotihuacan’s interest in its provinces of more interest to the empire (Charlton varied with the economic and strategic im- 1978). The manufacture of tools from portance of each area. Chingu was probably Pachuca obsidian took place mainly in valued for its lime, which was needed for workshops associated with public build- mural production and plaster manufacture ings in Teotihuacan, suggesting state con- for floors and walls at Teotihuacan. The trol (Spence 1987). This monopoly included Yautepec and Amatzinac valleys were extraction, suggested by Teotihuacan sites probably cotton growing areas as they are such as Huapalcalco and Zazacula located known to have been in the Late Postclassic close to Pachuca (Charlton 1978; Santley period (Maldonado Jiménez 1990; Smith and Pool 1993), production, distribution to 1994). A few possible crude spindle whorls the provinces, and trade with external part- and weaving picks have been found in ners. In the Teotihuacan provinces, Pachuca Teotihuacan although they are not common ARCHAEOLOGICAL STUDY OF EMPIRES 259

(Cabrera 1999). There are many bone nee- tuvo unida a Tula” (p. 218; see also Braniff dles implying that some form of textile was 1975:281). We suggest below that these sim- being produced. Cotton can not grow in the ilarities may have resulted from the incor- Basin of Mexico due to the colder environ- poration of this area into the Tula polity. ment, and Morelos is the closest cotton The Early Postclassic ceramics of the Basin growing area. Morelos was also favored for of Mexico were part of the Mazapan ceramic its strategic location along a trade route sphere (Whalen and Parsons 1982; Cobean with Guerrero. Granular ware, probably 1990), which shares many types (but with from Guerrero, is found in Teotihuacan, but different frequencies) with the local Tollan there are higher frequencies in central and ceramic sphere of the Tula region (Cobean eastern Morelos (Hirth 1980; Montiel 1998, 1990:38–40). Although no chemical charac- n.d.), suggesting that Granular ware moved terization studies have been published, it is through Morelos on its way to Teotihuacan. likely that vessels of some of the common ce- Teotihuacan dominated trade routes to ramic types were exchanged between Tula northern Mexico and Veracruz by control- and settlements in the Basin of Mexico. ling select towns in Puebla, Tlaxcala (García In our survey of the Yautepec Valley we Cook 1981), Hidalgo (García Cook and collected low frequencies of several ceramic Trejo 1977), and Queretaro. In the Valley of types similar to Tollan-phase types at Tula Queretaro, goods that may have been con- such as Macana Red-on-Brown and Proa trolled by the state, such as red-on-buff Polished Cream (Cascio et al. 1995). The bowls in cylindrical and annular-based type Macana Red-on-Brown, a very rare forms, are found in a few large sites in type in Yautepec, appears to be a distinctive strategic locations to control trade routes Tula-based type that was probably traded from Teotihuacan to areas in the north. The from Tula. Cream-slipped ceramics, on the concentration of populations in these settle- other hand, are more common in Morelos ments by the Late Classic period is inter- and also have a much more widespread preted by the investigators as a response to distribution in Early Postclassic central the collapse of Teotihuacan (Brambila and Mexico (Cobean 1990:357–364; Whalen and Castañeda 1993; Crespo 1998; Saint-Charles Parsons 1982:437; Tolstoy 1958). It seems Zetina and Argüelles Gamboa 1991). more likely that they may have been pro- duced in the Basin of Mexico and Morelos Tula in addition to the Tula area. None of the Toltec-like ceramics from Guanajuato, Several areas within the central Mexican Queretaro, the Basin of Mexico, or Morelos highlands exhibit Early Postclassic pottery have been subjected to chemical characteri- similar to that of Tollan-phase Tula (ceram- zation, however, and our suggestions ics at Tula are described by Cobean 1990 should be seen as highly preliminary. So- and Acosta 1956, 1957), and most archaeol- called “Mazapan-style” ceramic figurines ogists suggest that this material indicates are not uncommon in the Yautepec Valley, some form of link with the Toltec city. The and we believe that these are better seen as closest parallels are found in the Bajío area evidence for participation in a central Mex- of Guanajuato and Queretaro, where El ican style zone that included Morelos, the Cerrito, Morales, and other sites share Basin of Mexico, the Tula region, and parts many ceramic types with Tollan-phase Tula of west Mexico (Schondube 2000) than as (Flores and Crespo 1988; Crespo 1991; Bran- evidence for direct exchange between iff 1999). Flores and Crespo (1988) suggest Yautepec and Tula. that these similarities indicate “la existencia In contrast to these likely cases of stylistic de una población que en alguna forma es- similarity is the site of Tlalpizahuac, near 260 SMITH AND MONTIEL

Chalco in the southern Basin of Mexico, tation), providing further evidence for the whose limited Early Postclassic deposits importance of commerce at Tula. Not sur- produced burial vessels and sherds very prisingly, excavations at the site of Teptitlán similar to the ceramics of Tula that may rep- in the immediate vicinity of Tula (Cobean resent imports (Pfannkuch Wachtel et al. and Mastache 1999) reveal a level of ex- 1993). Most of the architecture and deposits change between the city and its inner hin- at this site date to the terminal Classic and terland far higher than between Tula and Epiclassic periods, however (Tovalín 1998), any distant area. with Early Postclassic vessels found pri- marily in some intrusive burials. Tlalpiza- Discussion huac also yielded ceramic imports from more distant areas, including Plumbate Because the Yautepec Valley of Morelos from Pacific coastal Guatemala and Fine was included in the Tenochtitlan and Teoti- Orange and other Gulf Coast types, indicat- huacan empires and in Davies’s model of ing that the inhabitants of Tlalpizahuac the Tula empire (see above), it is instructive were trading with diverse areas. The evi- to compare the frequencies of “imperial” dence does not suggest conquest or domi- imports from these three cities at sites nation by Tula, and Tovalín (1998:177) con- around Yautepec (see Table 2). For Teotihua- cludes that the site was not occupied by can and Tenochtitlan, we use sherd counts Toltecs from Tula. from our excavations; because we did not One of the few clear examples of hinter- excavate any Early Postclassic sites, we use land goods imported into Tula is obsidian, data from single-component surface collec- 3 although the mechanisms are not yet clear. tions for the Tula period. The lower level of The city had numerous workshops produc- probable “imperial” imports at Yautepec ing prismatic blades, bifaces, and other during Tula’s reign compared to the earlier tools from obsidian. The raw material came and later empires is striking. It should be from several source areas, including the noted that if some or all of the cream- nearby Pachuca source and the distant slipped ceramics (comprising the category, Ucareo/Zinapecuaro source area in Mi- “imperial style, uncertain origin”) were im- choacan (Healan 1993; Healan et al. 1983). ported from Tula, this would bring that city Analysis of Early Postclassic ceramics in more in line with the frequencies of the this source area revealed no definite Toltec other two cities. Although we consider this imports but rather an intrusive ceramic unlikely, it cannot be ruled out until chemi- complex (called Cumbres) linked to either cal characterization research is done on the Huamango in the Bajío region (Piña Chán Early Postclassic ceramics. In contrast, 1981; Lagunas 1997) or to sites in the Toluca Valley (Healan and Hernandez 2000). These TABLE 2 authors suggest that populations from Frequencies of Imported Ceramics in the Yautepec these latter areas may have quarried obsid- Valley, Morelos ian at Ucareo, part of which was then deliv- Area of origin Teotihuacan Tula Tenochtitlan ered to Tula as tribute. Alternatively, local groups in either the Ucareo or the Bajío re- Local 70.8 93.3 91.6 Imperial imports 1.8 0.1 2.3 gion could have sent obsidian to Tula Imperial style, 3.9 — 0.1 through commercial exchange networks. local origin Kristan-Graham (1993, 1999) has inter- Imperial style, 9.5 6.4 5.1 preted figures carved in relief on a promi- uncertain origin nent frieze at Tula as merchants rather than Other imports 14.0 — 1.0 Total sherds 14,697 1,684 57,051 as soldiers or chiefs (an alternative interpre- ARCHAEOLOGICAL STUDY OF EMPIRES 261 many of our attributions of Classic and Late 1952) and Oztoma (see below) suggest con- Postclassic types to local vs imported cate- quest, or at least control, of these provincial gories are based upon recent neutron activa- areas, but again the picture is clouded by tion studies (Smith et al. 1999; Montiel n.d.). poor data. The two strongest indicators of The presence of imported ceramics from Aztec imperial domination are the decline areas other than the imperial capital in in standard of living at sites in Morelos fol- Teotihuacan and Tenochtitlan times (Table lowing Aztec conquest and suggestions of 2) points to higher levels of regional ex- the co-opting of provincial elites by the im- change in those periods. perial rulers. The existence of exchange relationships The transition from the Late Postclassic- between a city and provincial areas is a nec- A period (prior to Aztec conquest) to the essary but not sufficient criterion for the Late Postclassic-B period (following Aztec presence of imperialism, and it is difficult to conquest) was marked by a consistent low- imagine an empire having the low level of ering of the standard of living at sites in capital–provinces exchange indicated for Morelos as inferred from a quantitative Tula. The larger sizes of Tenochtitlan and wealth index based upon frequencies of im- Teotihuacan relative to Tula were associ- ported and local decorated ceramics. Al- ated with much higher levels of economic though this trend was probably due in part activity in these cities (including craft pro- to local demographic and agrarian duction and exchange), with greater ex- processes (Smith and Heath-Smith 1994), change between the cities and their hinter- the fact that it was consistent across a num- lands, and with greater exchange between ber of sites in different polities and environ- the provinces and areas other than the im- mental zones (Smith n.d.b) suggests a com- perial capitals. mon origin in Aztec conquest. Current research by Jan Olson is examining issues DOMINATION OF A TERRITORY: of wealth levels and standard of living on POLITICAL CONTROL OF PROVINCES the household level at three of those sites, Yautepec, Cuexcomate, and Capilco. Smith Tenochtitlan and Heath-Smith (1994) argue that al- though the direct effects of the imposition In contrast with the topic of exchange, of imperial tribute at the household level the political control of provinces by were minor, Aztec imperialism had signifi- Tenochtitlan is far more heavily docu- cant indirect effects that combined to pro- mented in written sources than in the ar- duce major economic and social impacts on chaeological record (Berdan et al. 1996; Car- provincial households. One of these indi- rasco 1996). Tenochtitlan relied on pro- rect effects was the support of provincial vincial city-state rulers for local rule in most elites. cases (Smith 2000), and this indirect form of The co-opting of provincial elites, a com- control epitomizes the archaeological diffi- mon process in hegemonic empires, was a culties in studying the empire. The military major component of the imperial strategies conquest of provincial polities is hinted at of the Aztecs (Berdan et al. 1996). Current by the existence of locally built Postclassic research on excavated elite compounds in fortresses in several parts of the Aztec em- Morelos suggest that provincial elites at the pire (Smith n.d.a), but problems with important political center of Yautepec were chronology and the lack of excavations rule better off economically after Aztec con- these out as unequivocal evidence for im- quest, whereas contemporary elites at the perial conquest. The Aztec military settle- rural town of Cuexcomate suffered greatly ments at Quauhtochco (Medillín Zeñil (Jan Olson and M. Smith, unpublished 262 SMITH AND MONTIEL data); this fits what we know of Aztec around imperial regional centers exhibiting strategies in co-opting powerful provincial Teotihuacan state ideology, and state-con- elites. Additional evidence for this process trolled goods present at settlements sur- may be found in the scattered occurrence of rounding these centers. Mexica-style sculptures in the provinces Ocoyoacac in the valley of Toluca has a (Umberger and Klein 1993). For example, north–south orientation, probable grid pat- the presence of the carved name glyph of tern, talud-tablero architecture, as well as the Mexica emperor Ahuitzotl in the an- Teotihuacan ceramics including cylindrical cient provincial temple of Tepozteco in tripod and flat-bottom, flaring wall vessels, Morelos suggests a proclamation of imper- Thin Orange, imitation Thin Orange, ial control and a statement of alliance with censers, and figurines (Díaz Oyarzábal the local elites who probably controlled the 1998). The site was abandoned in the Late temple. Classic with the collapse of the empire. Te- peapulco in southeastern Hidalgo is a re- Teotihuacan gional center that also exhibits Teotihuacan influence in its urban planning and con- A lack of fieldwork at relevant sites tains a high frequency of Teotihuacan style makes it difficult to identify extensive evi- ceramics and Thin Orange (Matos dence for control of a territory by Teotihua- Moctezuma et al. 1981). In Tlaxcala, Teoti- can. There are a number of sites that are huacan ceramics and figurines are found in likely Teotihuacan administrative centers association with the apartment-type com- within its provinces (Table 3 and Fig. 3). pounds and pyramid-plaza groups at the These sites exhibit talud-tablero architec- site of Calpulalpan (Linné 1942:56–89). ture, pyramid-plaza complexes, and or- While Teotihuacan influence in urban plan- thogonal grid layouts, all features associ- ning is absent in Tlaxcala outside of the ated with Teotihuacan (Charlton 1991; Díaz “Teotihuacan Corridor” (García Cook Oyarzábal 1998; Nalda 1997; Torres Cabello 1981), a grave was found in the site of Tete- 1998). The evidence in the provinces con- les de Ocotitla containing a large offering of sists of urban planning influenced by Teoti- Teotihuacan ceramics (Vega Sosa 1981). huacan, the reorganization of settlements From the preliminary work in the valley of

TABLE 3 Urban Characteristics of Administrative Centers in the Teotihuacan Empire

State- Effect of Compact/ Apartment Grid Ceremonial Pyr./plaza Talud- controlled Teo collapse Provincial city Nucleated compounds plan avenues complexes tablero goods present Basin of Mexico: Central Basin X X X X X X X X Southern Basin — — X ? ? X X X Northern Basin X ? — — — ? X X Other areas: Calpulalpan X X ? ? X ? X ? Tepeapulco X X ? X X X X ? Ocoyoacac X — X ? X X X X Chingu X X X X X X X X Hacienda Calderon X — ? ? X X X X San Ignacio X — ? ? X ? X X

Sources. Charlton (1991: Table 15.2); Díaz (1998); Nalda (1997); Torres Cabello (1998). Note. Key: (X) trait is present, (—) trait is absent, (?) presence of trait is unknown. ARCHAEOLOGICAL STUDY OF EMPIRES 263

Queretaro, there appear to be many sites larities between Early Postclassic ceramics with strong ceramic and architectural simi- in the Bajío region and Tula, discussed larities with Teotihuacan, including San above (see Flores and Crespo 1988), suggest Bartolo Aguascaliente (Brambila and Cas- close interaction between these areas. Both tañeda 1993), El Cerrito (Crespo 1991, 1998), were part of the Tollan ceramic sphere, and and La Negreta and El Rosario (Saint- on this basis Cobean (1990:507) suggests Charles Zetina 1998). that the Bajío was part of the Tula empire. Chingu, near Tula, and San Ignacio in More impressive, however, is the presence Morelos display many of the material corre- of stone sculptures at the site of El Cerrito lates of Teotihuacan regional centers and (Crespo 1991) that are virtually identical to were also abandoned with the collapse of the sculptures at Tula (de la Fuente et al. 1988). empire. Chingu is the only distant provincial These suggest local elite emulation of center that contains all of the evidence for state/imperial styles, probably resulting urban planning illustrated in Table 3 as well from close communication between elites in as a majority of Teotihuacan ceramics and the two areas. The regional polity extend- figurines (Díaz Oyarzábal 1980, 1991). In ing from Tula to the El Cerrito area (Fig. 4) Morelos, the surveys of the Amatzinac (Hirth could be considered a large regional state, 1978, 1980, 1996) and Yautepec valleys (Mon- or perhaps a miniempire similar to the pre- tiel n.d.), and intensive investigations at Las Tenochtitlan states or empires of Texcoco Pilas (Martínez Donjuan 1979) and Hacienda and Azcapotzalco in the Basin of Mexico Calderon (Nalda 1997) provide the most de- (Davies 1973; Carrasco 1996). tailed portrayal of a provincial area of the A few Tula-style stone carvings were also Teotihuacan empire. San Ignacio in the Am- present at Tlalpizahuac in the southern atzinac Valley (Hirth 1978, 1980, 1996) and Basin of Mexico (Granados Vázquez et al. Hacienda Calderon in the Cuautla Valley are 1993), but compared to El Cerrito there are regional centers exhibiting urban planning far fewer examples and they exhibit fewer and a concentration of state goods. All re- similarities with sculptures at Tula itself; in ported Early Classic sites in the eastern half fact these could predate the Early Postclas- of Morelos contain numerous examples of sic period (Tovalín 1998). The largest Early Teotihuacan material culture, both imported Postclassic sites encountered by the Basin and locally made, including Thin Orange of Mexico archaeological survey projects and other ceramic wares, figurines, and are fairly small and unassuming sites classi- Pachuca obsidian blades. Changes in settle- fied as “secondary regional centers” (Par- ment patterns reflect a nucleation of popula- sons 1971:87–88; Sanders et al. 1979). Al- tions in fertile cotton growing areas and/or though these have yet to be investigated in strategic positions, both conditions probably any detail, none of them appear to resemble due to the effect of the empire. The Teotihua- Teotihuacan’s regional administrative cen- can empire appears to have controlled its ters (Table 3) in size, complexity, or likeness provinces outside of the Basin of Mexico to the political capital. We therefore see no using indirect rule, and recent fieldwork in reason to include the Basin of Mexico, these provincial areas points to a diversity of Morelos, or other areas outside of the nar- imperial interactions (Montiel n.d.). row corridor from Tula to El Cerrito in modern reconstructions of the Toltec polity. Tula Discussion Two kinds of evidence suggest that the Tula state may have controlled settlements As noted above, it is very difficult to find in the Bajío region. First, the extensive simi- material evidence of provincial control in 264 SMITH AND MONTIEL empires that were organized in an indirect U.S. Southwest (Weigand et al. 1977; or hegemonic fashion. Such polities did not Weigand and Harbottle 1993) is common at invest much in provincial infrastructure, Tenochtitlan and other Aztec sites (Saville and they relied on the co-option of provin- 1922). Several caches of copper or bronze cial elites for control. The simple presence bells from Michoacan and/or Jalisco have of imperial goods in distant elite contexts been recovered in Tenochtitlan (Seler can indicate either imperial control or im- 1990–1996), and bronze objects are not un- perial interaction with extraempire elites common in other parts of the Aztec empire (Whittaker 1994). The client states of the (Hosler 1994, 1996). In addition to bronze, eastern Roman empire (Braund 1984) pro- obsidian from Michoacan (the Ucareo vide a textually well-documented example and/or Zinapecuaro sources) is not uncom- of indirect rule. Although nominally inde- mon at Aztec sites in Morelos, and Pachuca pendent of Rome, these kingdoms were for obsidian has been found at Tarascan sites in all practical purposes under Rome’s con- the Patzcuaro Basin (Pollard and Vogel trol. Yet archaeological evidence for this sit- 1994). These materials are particularly im- uation is quite elusive (Isaac 1990; Millar portant, for they document the existence of 1995). trade across the Aztec/Tarascan imperial Even with these difficulties, the differ- border (a process not described in docu- ences in the quantity of evidence in provin- mentary sources, which emphasize hostile cial areas between Tenochtitlan and Teoti- activities along the border). Other exotic huacan, on one hand, and Tula, on the goods found at Aztec sites, including jade other, is striking. Our surveys and excava- and marine shell (Thouvenot 1982; López tions in Morelos turn up considerable evi- Luján 1994), provide additional evidence of dence of goods or styles from Tenochtitlan active commercial exchange with extraim- and Teotihuacan, but many fewer examples perial areas. of goods or styles from Tula. Furthermore, Perhaps the clearest case of political influ- our knowledge of the archaeological record ence beyond the Aztec imperial borders is in nearby areas such as southern Puebla the militarization of the Aztec/Tarascan bor- and Toluca suggests that this situation is der. As these two expanding empires came typical. up against one another in the mid- to late 15th century, a number of fortifications were INFLUENCE IN THE WIDER built on each side. The Aztec fortress of Oz- MESOAMERICAN WORLD SYSTEM toma, described in a number of ethnohistoric accounts, is the best known, although to date Tenochtitlan only limited fieldwork has been done at the site (Silverstein 2000). There are many other Mexica trade with places beyond the lim- fortified sites along both sides of this border, its of their empire is well documented ar- some quite large and impressive, and ar- chaeologically. Obsidian from the Pachuca chaeologists are only beginning to study source area, almost certainly carried by the these sites (Gorenstein 1985; Hernández famous pochteca merchants (Isaac 1986), has Rivero 1994b; Silverstein 2000). been found in Late Postclassic contexts in a The cultural influence of the Aztec em- number of distant areas (Smith 1990). Aztec pire beyond its borders is difficult to assess, III black-on-orange ceramics, although since the transmission of many central probably less common than obsidian out- Mexican styles and traits to distant areas side of the empire, were nevertheless began before the formation of the empire traded as far south as El Salvador (Navar- and existing chronologies are not suffi- rete 1996). Turquoise, imported from the ciently refined to pinpoint the dates of indi- ARCHAEOLOGICAL STUDY OF EMPIRES 265 vidual occurrences (Smith 2001). One of the ramics (discussed below). Teotihuacan-type most visible examples of central Mexican decorated censers have been found with iconographic elements in distant areas is local censer types in Escuintla and Lake the Postclassic International Symbol Set Amatitlan, Guatemala (Berlo 1984). Cylin- (Smith and Berdan 2000; Boone and Smith drical tripod vessels, a hallmark of Teoti- n.d.), a group of visual elements found as huacan, appear in sites throughout far away as Maya murals at Tulum and Mesoamerica, although many are locally other sites on the coast of Yucatan (Robert- made variants. son 1970 labeled these elements the “Inter- Surrounding areas responded politically national style” of the Postclassic period). to the rise, domination, and collapse of the The spread of these symbols probably sig- Teotihuacan empire. Monte Alban in Oax- naled the expansion of commercial and cul- aca was a powerful state that had commer- tural networks in the Late Postclassic cial and/or diplomatic relations with Teoti- Mesoamerican world system (prior to as huacan. Marcus (1983) interprets stone well as after the expansion of the Aztec em- reliefs at Monte Alban as records of Teoti- pire), and thus it is only a weak candidate huacan visitors, probably diplomats or for imperial influence in the international traders. The so-called “Oaxaca barrio” at context. Many of the items in Navarrete’s Teotihuacan (Spence 1989, 1992) was a (1996) catalog of presumed central Mexican neighborhood where foreigners, presum- traits found in southern Mesoamerica pre- ably from Monte Alban, were living within date the Aztec empire (and many of them the Teotihuacan urban zone. Monte Alban probably do not even have a central Mexi- declined about the same time as Teotihua- can origin at all). In sum, it is difficult to de- can, which suggests a close connection be- termine whether the apparent lack of Aztec tween the two states (Blanton 1983). Migra- cultural and stylistic influence beyond the tion into areas of Puebla that were empire is due to chronological imprecision independent of Teotihuacan and nucleation or to a simple lack of such influence. This into larger settlement have been interpreted situation contrasts greatly with the earlier as responses to the increasing domination Teotihuacan empire, whose cultural influ- of Teotihuacan (García Cook 1981:263; Gar- ence was widespread and deep throughout cía Cook and Merino Carrión 1997:369). Mesoamerica. Even distant areas in Michoacan were ef- fected by increasing mining of cinnabar and Teotihuacan other pigments used for Teotihuacan mural production (Weigand 1982). Teotihuacan had commercial contacts Matacapan in Veracruz has been sug- throughout most of Mesoamerica. Imports gested as a Teotihuacan enclave due to the found in Teotihuacan include stone masks presence of talud-tablero architecture on one and Granular ware from Guerrero, shell structure and Teotihuacan type ceramics, and cacao from the Pacific coast, Lustrous such as candeleros, tripod vessels, floreros, ware from El Tajin, Polychromes from the censers, and effigy vessels (Ortiz and Santley Peten, and other fine ware ceramics from 1998; Santley 1989; Santley et al. 1987). These the Gulf Coast (Rattray 1979). Pachuca ob- ceramics are locally made, however, and sidian, whose distribution was almost cer- Cowgill (1997:135) has noted that their re- tainly in Teotihuacan hands (Sanders and semblance to Teotihuacan ceramics are not Santley 1983; Santley and Pool 1993; Spence as close as some had previously implied. 1987, 1996), is a rare but consistent com- This pattern may reflect the presence of a modity at Classic Maya sites (Spence 1996; group originating from Teotihuacan, just as Moholy-Nagy 1999) as are Teotihuacan ce- there were residents from the Gulf Coast liv- 266 SMITH AND MONTIEL ing in the “Merchant’s Barrio” of the capital the later cities of El Tajín and Xochicalco (Rattray 1979). The collapse of the Teotihua- than Teotihuacan (Jones 1996:32–35; Schele can empire did not appear to affect the Gulf and Mathews 1998:72). Coast greatly, however (Arnold et al. 1993), The Teotihuacan “influence” at Early and we doubt if this area was ever part of Classic Tikal is restricted to the royal lin- the Teotihuacan empire. eage, and we agree with others (Schele and Kaminaljuyu, near Guatemala City, also Freidel 1990; Schele and Mathews 1998) had links to Teotihuacan, evident in the use that this “influence” was probably not a re- of talud-tablero architecture and goods from sult of conquest by Teotihuacan. Specific Teotihuacan present in some burials rulers of Tikal more likely had alliances (Sanders and Michels 1977). The data sug- with a group from Teotihuacan, perhaps gest that the two cities were related through through elite intermarriage. Tikal certainly ties of elite exchange and interaction rather had its own reasons for seeking powerful than through imperial conquest (see discus- alliances at this time when it was probably sion of these processes in Stark 1990). Teoti- at war with its close neighbor, Uaxactun. In huacan traits are restricted to parts of the fact, the victory of Tikal over Uaxactun is acropolis and elite burial contexts, and they associated with the introduction of Teoti- are found together with the regional Maya huacan weapons, warrior costumes, and traditions. military iconography (Schele and Freidel Recent epigraphic decipherments of 1990: 130–165). monuments from Tikal by David Stuart The founder of the royal lineage in (2000) have brought up the old argument Copan also appears to have had close asso- (again) that Teotihuacan conquered this city ciations with individuals from Teotihuacan. in the Early Classic period (Coe 1999: Altar Q, dedicated 350 years after the royal 90–97). In Tikal, the relevant evidence be- lineage began, depicts the founder with gins with texts associated with the ninth ringed eyes, reminiscent of Teotihuacan ruler. Stuart’s reading indicates that this in- Tlaloc imagery. An Early Classic structure dividual may have been from Teotihuacan, with talud-tablero architecture was found whereas Schele and Freidel (1990:156–159) beneath Structure 16 (Sharer 1995). This claim that the ruler is from the Tikal dy- structure, called “Hunal” by archaeologists, nasty but had close associations with a contains the burial of the dynastic founder group of Teotihuacanos. At Early Classic accompanied by Teotihuacan-style ceram- Tikal, the warrior costume of Teotihuacanos ics. Ongoing stable isotope analysis of the appears on important monuments includ- skeletal remains suggest that he may have ing Stela 31, the accession monument of the had Teotihuacan ancestry. As with Tikal, 11th ruler. Royal tomb offerings include Teotihuacan “influence” at Copan is re- Thin Orange, cylindrical tripod vessels, and stricted to a limited interval during the green obsidian blades from Teotihuacan Early Classic period and occurs within spe- along with Maya ceramics. The “Mer- cific elite contexts. The relationships con- chant’s Barrio” at Teotihuacan contains sisted mainly of elite alliances and the emu- many Maya sherds probably from Tikal lation of Teotihuacan military symbols for (Rattray 1979, 1989), indicating that Teoti- the “cult of war” by local rulers (Schele and huacan may have had a small resident Tikal Freidel 1990:130–165; Schele and Mathews population. At Tikal, a small structure (no. 1998:18). Maya kings aligned themselves 5D-43) connected to the Central Acropolis symbolically with the distant Teotihuacan (the royal palace) exhibits talud-tablero style empire to legitimize or reinforce their au- architecture, although it has been claimed thority. These relationships may also have that the style may more closely resemble been an attempt by individuals from Teoti- ARCHAEOLOGICAL STUDY OF EMPIRES 267 huacan to establish footholds in the Maya prestige of Teotihuacan, some of which area for economic gain, although there is probably derived from its rule of an empire not a lot of evidence from Teotihuacan of a within highland central Mexico. large-scale trade network with the Maya lowlands. Tula By the Late Classic period, there is no ev- idence of direct contact between the Maya Although Tula’s economic and political lowlands and Teotihuacan. However, even importance within central Mexico was far after the imperial collapse, Teotihuacan less than one would expect for an empire, military iconography continues to appear several lines of evidence suggest that the on royal monuments at Maya sites. Yaxchi- city was heavily involved in the larger lan Lintel 24 shows the queen with a Tlaloc Mesoamerican world system during the headdress, and Lintel 25 depicts her having Early Postclassic period. Sherds of Tohil a vision of a serpent with the lineage Plumbate, a widespread Mesoamerican founder dressed as a Teotihuacan warrior trade ware produced in Pacific coastal emerging from its mouth. Piedras Negras, Guatemala, are not uncommon at Tula; Stela 8, and Dos Pilas, Stela 2, both depict Diehl calls this ware “especially abundant the rulers with balloon headdresses, Tlaloc at Tula” (Diehl 1993:268). Strangely, how- and Mexican year sign imagery reminiscent ever, the other major Early Postclassic trade of Teotihuacan warrior costumes (Schele ceramic, Silho Fine Orange from the Gulf and Friedel 1990:Fig. 4.17). Centuries after Coast, is absent from Tula. A cache of Cen- the fall of the Teotihuacan empire, its sym- tral American polychromes excavated at bolic importance in the Mesoamerican Tula (probably the Las Vegas type from world was still strong. Honduras; see Lange 1986) provide further Much of the confusion in the literature evidence for the city’s participation in ex- about Teotihuacan’s imperial status can be change with the distant reaches of attributed to its widespread cultural influ- Mesoamerica, and the prevalence of Zi- ence in Mesoamerica. Traits derived from napecuaro obsidian at Tula shows an addi- or related to Teotihuacan were noted by tional commercial connection. The quantity early scholars, who attributed them to the of exotic imports at Tula is not matched by conquest of much of Mesoamerica by the Toltec exports in more distant areas, how- highland city. This was a time when con- ever. The Pachuca obsidian found at Maya quest and empires were invoked to explain sites in Yucatan in this period (Coe 1999: much of the Mesoamerican past; as an ex- 165–169) was probably obtained through ample, the Classic and Postclassic periods trade with Tula, but Mazapan-style fig- in the Maya lowlands were given the labels urines from Apatzingan, Tizapan, and of “old empire” and “new empire.” Later other sites in west Mexico (Mastache and scholars revised their interpretations of Cobean 1985:295; Cobean 1990:506–508) Teotihuacan elements in distant Mesoamer- may represent stylistic interaction rather ica, rightly arguing that they did not signal than exchange (Schondube 2000). imperial control but rather processes of The Early Postclassic period was a time of trade, stylistic emulation, and political ma- widespread trade and communication nipulation of foreign symbols by local elites throughout Mesoamerica, and the inhabi- (e.g., Stark 1990). In our model, such traits tants of Tula must have participated in these do not signal imperial conquest of areas networks. A number of scholars, however, like Oaxaca or the Maya lowlands, but their have stressed the noncentralized nature of prevalence in these areas does provide sup- Early Postclassic exchange systems. Smith port for our interpretation of the power and and Heath-Smith (1980), for example, show 268 SMITH AND MONTIEL that during the Epiclassic and Early Post- leader of the Itza peoples who conquered classic periods most Mesoamerican long- Chichen Itza and imposed their architec- distance trade systems followed coastal tural styles (Tozzer 1957). An opposing routes, with only limited involvement by model, based upon an earlier dating of the the cities of central Mexico (in strong con- so-called “Toltec” architecture at Chichen trast to the Classic and Late Postclassic peri- Itza, holds that architectural styles either ods, when Teotihuacan and Tenochtitlan spread from Yucatan to Tula or spread to were far more heavily involved in long-dis- both areas from an unidentified origin tance exchange than Tula). Diehl (1993) pro- place (e.g., Andrews 1990). Many scholars poses the existence of a “Toltec horizon” in now date most of the Chichen Itza architec- Early Postclassic Mesoamerica marked by ture prior to A.D. 1000 (e.g, Schele and openwork censers, wheeled toys, and other Mathews 1998:198–201, 357–360; Ringle et ritual objects, and he notes that Tula partici- al. 1998), throwing the former model into pated in this horizon but was probably not doubt. Given the uncertain chronological the place of origin of the objects or the ex- resolution at Chichen Itza, however, this change networks. debate cannot yet be resolved with confi- Given Tula’s quite limited political role dence. Following Kepecs et al. (1994) and within central Mexico, it is not surprising others, we believe that the direction of ar- that there is no clear evidence for political chitectural influence is less important than impact on more distant areas. When we turn the fact that Tula and Chichen Itza were to cultural influence, however, the data tell a clearly interacting with one another both different story. Tula and the Toltecs were commercially and stylistically. These were renowned throughout Mesoamerica, and two of the largest and most influential cities Tula clearly played a major role in the larger in the Early Postclassic world system, and Mesoamerican world system. Diehl’s (1993) this two-way interaction is not surprising. Toltec horizon suggests the inhabitants of Tula were active participants in the interac- Discussion tion networks that created and maintained the horizon style. We would rephrase his All three cities had important economic discussion in world system terms; the wide- and stylistic influence throughout large spread occurrence of this ritual complex parts of Mesoamerica. This kind of long- points to important systems of stylistic in- distance influence is typical of powerful teraction and information exchange. Tula empires, but it can exist in the absence of an was an important participant in these inter- empire. World history has numerous exam- action systems, even if the major traits did ples of large cities that had wide-ranging not diffuse outward from there. influences but were not capitals of empires. The most widely discussed case of stylis- These include religious cities (e.g., Rome in tic interaction involving Tula is the relation- post-imperial times or Mecca) and cities ship between that site and Chichen Itza. that were trade centers (e.g., Venice and Detailed architectural similarities between Genoa in early modern times). This is com- these two Early Postclassic sites have long mon in periods with vigorous world sys- been noted, and a strongly polarized debate tems, defined as systems of widely sepa- has emerged over the nature and direction rated polities that are linked through of influence between the two cities. The tra- commercial exchange and often other ditional argument combines the architec- forms of interaction (Abu-Lughod 1989; tural data with origin myths from the two Chase-Dunn and Hall 1997; Peregrine and areas suggesting that the god-king Quetzal- Feinman 1996). We suggest that all three coatl left Tula and arrived in Yucatan as the major cities participated in world systems ARCHAEOLOGICAL STUDY OF EMPIRES 269

(see note 1) that linked most or all of trol over a large hinterland in central Mex- Mesoamerica through trade and stylistic ico. Tula does appear to have maintained communication. The relevance of the world close economic interactions with the Bajío re- systems approach for present purposes is gion northwest of Tula, and these interac- that it provides a framework for under- tions may have included political control. standing the ways in which long-distance We have deliberately avoided a precise defi- stylistic and commercial influences can nition of the lower limits of the category of exist independently of the presence of em- empire, since subjective evaluations come pires. into play (just how much territory must be controlled? at what distance from the center? CONCLUSIONS do small-scale regional states based upon the conquest of nearby neighbors qualify?). The Nonexistent Toltec Empire But even if we take an inclusive approach and call the Tula/Bajío polity an empire, it The material culture indicators of ancient was an order of magnitude smaller than the imperialism reviewed above suggest that Teotihuacan polity (Figs. 3, 4), which was Tenochtitlan and Teotihuacan ruled em- not a particularly large empire compared to pires covering large areas of highland cen- examples like the Aztec and Tarascan em- tral Mexico, but Tula did not. Our interpre- pires. On the other hand, the city of Tula was tation of Tenochtitlan is hardly surprising, certainly large and complex enough to have since the documentary record is clear about been an imperial capital, and it contained the existence of an empire centered at this public monuments with imperial-type ideo- city (e.g., Berdan et al. 1996; Carrasco 1996). logical messages. Also, Tula was an impor- This case lends support to our model by tant part of long-distance commercial and showing that a historically known Meso- stylistic exchange networks in Early Post- american empire can indeed be identified classic Mesoamerica. In sum, we interpret from its material remains alone. The exis- Tula as the capital of a regional state that tence of a central Mexican empire based at some might consider a miniempire. Teotihuacan has been less widely accepted, We believe that scholars have been led to but there have been few explicit arguments misinterpret the Toltec polity as a large and against such an interpretation. Most schol- powerful empire for three reasons. First, ars have been too preoccupied with Teoti- there is a tendency to give too much weight huacan’s interaction with the Maya states to the native historical record in central to worry about the Teotihuacan polity in Mexico, in spite of its obviously propagan- central Mexico. We feel that the data pre- distic nature (see Marcus 1992 or Smith sented above provide strong support for 1986 for comment). Thus many scholars Millon’s (1988) model of a Teotihuacan em- have been tempted to believe Aztec ac- pire, and Montiel’s (n.d.) dissertation elabo- counts of Toltec greatness, even when they rates on the nature and local effects of Teoti- are contradicted by the archaeological huacan imperialism. record. Second, there has been very little in- Our interpretation of the Toltec case de- tensive archaeological research on Early parts most strongly from currently accepted Postclassic central Mexico (outside of Tula, interpretations. As mentioned above, most the Bajío, and a few other sites), making it scholars who have dealt with the issue of the difficult to disprove the common view of Toltec polity classify it as an empire. The ar- great similarities among the roles of Teoti- chaeological evidence reviewed here, how- huacan, Tula, and Tenochtitlan (e.g., ever, provides little support for the notion Sanders and Santley 1983). Third, there has that Tula exerted political or economic con- been a general lack of models in archaeol- 270 SMITH AND MONTIEL ogy and anthropology for the large-scale in- been the topic of recent debate. Monte Alban fluence and role of large cities beyond the was a large and complex urban center at that imperialism model (see Feinman and time (Blanton 1978), and public reliefs in the Nicholas 1991 or Stein 1999 for examples main plaza have been interpreted as con- and discussion). We feel that the world sys- quest monuments (Marcus 1992). In the tems perspective provides an alternative Cuicatlan Valley to the north, archaeological view that helps explain these seemingly evidence points strongly to conquest and anomalous features of Tula. A city did not imperial administration by Monte Alban have to rule an empire to be an important (Redmond 1983; Spencer and Redmond force at the level of Mesoamerica as a 1997). Settlement patterns were reorganized, whole. We can acknowledge Tula’s consid- fortresses built, and economic ties with erable cultural, ideological, and economic Monte Alban were strong. The presence of influence throughout Mesoamerica without the earliest known skull rack in Mesoamerica having to call it the capital of an empire. at the provincial site of La Coyotera suggests the imposition of imperial ideology in this Other New World Empires provincial area. Other nearby regions under- went political centralization at this time, The material culture model presented probably in response to the expansion of the here can be applied to other ancient polities Monte Alban polity (Balkansky 1998). of the New World to assess the likelihood Although the imperial conquest of the that they were empires. Although there is Cuicatlan Valley is accepted by most schol- not space to review these cases in detail ars, the expansion of the Monte Alban em- here, we can offer a few preliminary sug- pire to the south and southeast, toward the gestions about several well-known exam- Pacific Ocean, is more controversial. Mar- ples. The status of the Tarascan polity of Mi- cus and Flannery (1996:201–202) suggest choacan as an empire is well documented that the lower Rio Verde Valley was con- in the documentary record, and archaeolog- quered by Monte Alban on the basis of a ical evidence largely conforms to our model shift to grayware ceramics at that time. (e.g., Gorenstein 1985; Hernández Rivero Scholars working in the Rio Verde Valley 1994a; Pollard 1977, 1993). The capital and the Tehuantepec area, however, have Tzintzuntzan was a large and complex urban not found the kinds of evidence for imper- center, material evidence for exchange be- ial conquest described in our model (Joyce tween the capital and the provinces is abun- 1993; Zeitlin 1990, 1993; Zeitlin and Joyce dant, and there is evidence for several types 1999). These areas engaged in trade with of political control of provinces, including Monte Alban, but imports from the Valley the construction of imperial infrastructure of Oaxaca at sites in the lower Rio Verde (fortresses) and co-option of local elites. Valley declined at the time of Monte Alban Tzintzintzan traded with extraimperial expansion, and the graywares, which re- areas, both to the west and across the hostile sembled wares at Monte Alban, were lo- Aztec/Tarascan border to the east (see Fig. cally produced (Joyce n.d.). Although more 2), and the Tarascan center exerted political data are needed before definitive conclu- influence in the form of militarization of the sions can be drawn, our reading of the evi- Aztec border. The available evidence sug- dence suggests that the Monte Alban em- gests that the Tarascan empire employed pire in period II controlled the Cuicatlan more direct forms of territorial control than area but did not conquer distant areas to did most other Mesoamerican empires. the south and southeast. The imperial status of Monte Alban in the The Postclassic Mixtec polity of Tututepec in Monte Alban II phase (100 B.C.–A.D. 200) has Pacific coastal Oaxaca has been interpreted as ARCHAEOLOGICAL STUDY OF EMPIRES 271 an empire by Spores (1993) and Davies and complex urban center with various man- (1968:181–213) based upon ethnohistoric evi- ifestations of imperial ideology. Exchange be- dence. Mixtec pictorial histories talk of con- tween capital and provinces has not been quests and marriage alliances by the famed extensively researched, but chemical char- Mixtec king 8 Deer Tiger Claw from his capi- acterization of Wari pottery from several tal of Tututepec, and local administrative doc- provincial areas reveals it to have come uments from hinterland towns mention being from the Wari area. conquered and paying tribute to Tututepec. Political control of the Wari provinces is This is a problematic case. The archaeological amply documented in most of our cate- remains clearly do not fit our model, perhaps gories. There are military themes in the due to a lack of relevant fieldwork, although provinces suggesting Wari conquest; the the documentary evidence does suggest that construction of imperial infrastructure is Tututepec was a large and powerful polity, clear in the trademark “Wari compounds” perhaps on an imperial scale. The Postclassic and D-shaped structures; settlement pat- settlement at Tututepec was quite small, terns were reorganized in several areas, in- hardly qualifying as an urban center, much cluding the Mantaro Valley, Carahuarazo, less as an imperial capital (O’Mack 1986; and Nazca valleys; and Wari iconography Spores 1993). Documents mentioning tribute and elite burials in the provinces suggest payments (Spores 1993:171) favor the inter- the co-option of conquered elites. Projection pretation of empire, although capital–hinter- of influence beyond the empire is evident in land trade has yet to be documented. There is the complex economic and political rela- modest evidence for political influence be- tionships between the Wari and Tiwanaku yond Tututepec’s borders in the care taken by polities, particularly at the Wari outpost of the Aztec empire in approaching and sur- Cerro Baul within Tiwanaku territory. rounding this area (Berdan et al. 1996:141; These features are documented in a number Davies 1968). Tututepec appears to have been of publications (e.g., Czwarno et al. 1989; Is- a powerful and expanding polity, and per- bell and McEwan 1991; McEwan 1987; haps should be classified as a miniempire or a Schreiber 1992, 1999). In sum, the Wari conquest-state, similar to the polities of Az- polity clearly fits our criteria for an empire, capotzalco, Texcoco, and Cuauhnahuac in a finding in agreement with current inter- central Mexico prior to the founding of the pretations as presented at a recent sympo- Tenochtitlan empire. sium (Malpass and Jennings 1999). Other As a final example, we mention the Middle Andean polities—the Moche, Tiwanaku, Horizon Wari polity of the Andes. In contrast and Chimor states—have been classified by to the hegemonic empires of Mesoamerica, some as empires, and it might be instruc- Andean empires such as the Inka and Wari tive to apply our model to their archaeolog- exhibited a greater degree of direct territorial ical manifestations; such efforts lie beyond control, and consequently the durable evi- the scope of this paper, however. dence of their expansion is quite extensive and visible. The combination of documen- The Archaeology of Imperialism tary and archaeological evidence on the Inka polity leave no doubt that this was a power- We view our material culture model for ful empire (e.g., Bauer 1998; D’Altroy 1992; the identification of ancient empires not as Earle 1994; Hyslop 1984; Malpass 1993; a mere typological exercise but rather as an Murra 1980). There is no documentary evi- organizing principle for the archaeolog- dence for the earlier Wari polity, but its ar- ical analysis of these polities. Ancient em- chaeological remains clearly fit our model of pires were large and complex entities, and ancient empires. The city of Wari was a large archaeological fieldwork can only illumi- 272 SMITH AND MONTIEL nate bits and pieces of them. Our cate- cials of the Instituto Nacional de Antropología e Histo- gories can help organize the available data ria, especially Norberto González Crespo, Hortensia on individual ancient empires, and they de Vega Nova, Victor Hugo Valencia Valera, and Joaquín García-Bárcena. We would like to acknowl- can help guide continued fieldwork and edge the assistance and insights of our colleagues on archival research. This scheme can also the Yautepec project, especially Timothy S. Hare and help identify areas where evidence is weak Cynthia Heath-Smith. Smith would also like to thank and more data are needed. For example, the participants in two recent conferences for stimulat- the clarification of Monte Alban’s role out- ing discussion on issues relating to our theme, “Imper- ial Designs: The Comparative Dynamics of Early Em- side of the Valley of Oaxaca in period II pires,” organized by Susan Alcock, Terence D’Altroy, will require additional data that bears ex- Kathleen Morrison, and Carla Sinopoli in Mijas, Spain, plicitly on our second category, the domi- in October, 1997, and “The Postclassic Mesoamerican nation of a territory. The mere presence of World System,” organized by Smith and Frances F. grayware ceramics, taken by some as evi- Berdan in East Lansing in April, 1999 (both confer- ences were funded by the Wenner-Gren Foundation dence of imperial conquest, is insufficient for Anthropological Research). A preliminary version by itself to resolve the issue. Without addi- of some of the ideas discussed here was presented by tional information the “Zapotec-imperial- Smith at the Texas Maya Symposium in Austin, TX, in ism argument” (Zeitlin and Joyce 1999) March, 1998. We thank George Cowgill, Timothy cannot be fully resolved. Once it is ac- Earle, Arthur Joyce, Jay Silverstein, Katharina Schreiber, and an anonymous reviewer for their help- knowledged that Tula most likely did not ful comments on earlier drafts of this paper. We also rule an empire in central Mexico, scholars thank Cynthia Kristan-Graham for pointing out the El can work on developing more appropriate Cerrito sculptures to us and Timothy S. Hare for pro- models for Tula’s widespread influence in viding the Early Postclassic type counts from the Early Postclassic Mesoamerica. Inversely, Yautepec survey and for providing digital base maps of central Mexico. now that we can speak safely of a Teoti- huacan empire, it is time to move on to more refined analyses of that polity in the NOTES light of our theoretical and comparative understanding of ancient empires. Thus 1We follow a modified world-systems approach sim- our interpretation that Tenochtitlan, Teoti- ilar to that currently advocated by many archaeolo- huacan, and several other New World gists (e.g., Blanton and Feinman 1984; Chase-Dunn and Hall 1997; Peregrine and Feinman 1996; Edens polities were most likely empires is less a and Kohl 1993). As described elsewhere (Smith and conclusion than a starting point for the Berdan 2000), our world systems approach to ancient continuing analysis of these polities by ar- societies emphasizes the systemic role of exchange chaeologists. that crossed political borders and provides insights into large-scale economic and political dynamics in early complex societies. This “world systems perspec- tive” (Peregrine 1996) differs considerably from the ACKNOWLEDGMENTS original formulation of “world systems theory” by Wallerstein (1974), which has been criticized by Stein Much of our interest in the identification of central (1999) and others for its inapplicability to precapitalist Mexican empires comes from our work with the Al- economies. More detailed discussions of our world bany Yautepec project. The Yautepec excavations were systems approach can be found in forthcoming case supported by grants from the National Science Foun- studies (Montiel n.d.; Smith and Berdan n.d.). We real- dation, the National Endowment for the Humanities, ize that the phrase “world system” often has a polariz- and the Heinz Foundation. The Yautepec survey was ing effect among archaeologists (e.g., Stein 1999), and supported by the National Science Foundation (disser- we wish to emphasize here that our material culture tation grants to Montiel and Timothy S. Hare), the model of imperialism does not depend upon the world Wenner-Gren Foundation for Anthropological Re- system approach. search, the National Geographic Society, the Institute 2Prior to 1940, most scholars associated the ruins of for Mesoamerican Studies, and the University at Al- Teotihuacan with the semilegendary Toltec capital city bany. We are grateful for help and support from offi- Tollan as described in the Aztec native historical ac- ARCHAEOLOGICAL STUDY OF EMPIRES 273 counts (e.g., Vaillant 1938). In a highly influential ramics, are included as “Imperial style, uncertain ori- paper, Wigberto Jímenez Moreno (1941) changed the gin,” since their origin in Morelos or the Basin of minds of most archaeologists by using toponyms to Mexico is uncertain. “Other imports” includes a vari- show that the city Tollan in the Aztec migration histo- ety of types from other areas of Morelos, from Guer- ries corresponds to the Early Postclassic ruins at Tula, rero, the Toluca Valley, the Tarascan borderlands, and Hidalgo. More recently, however, a number of authors Puebla. have shown that Tollan was an important ideological concept in Mesoamerica, where it meant a sacred and powerful city of great symbolic importance. The label Tollan was probably applied by Mesoamerican peo- ples to a number of ancient settlements, including REFERENCES CITED Teotihuacan, Tula, Cholula, Tenochtitlan, and proba- bly others as well (Carrasco 1992:126; Boone 2000; Abu-Lughod, Janet L. Schele and Mathews 1998:200–201). Our concern here 1989 Before European hegemony: The world system, is with the Early Postclassic city Tula and its polity, re- A.D. 1250–1350. Oxford Univ. Press, New gardless of what other cities or peoples may have been York. called Tollan or Toltecs. Acosta, Jorge R. 3The data in Table 2 are of diverse origin and they 1956 Resumen de los informes de las exploraciones are only comparable in a rough sense. For Teotihua- arqueológicas en Tula, Hidalgo, durante las can, the data are mean frequencies from test excava- VI, VII, y VIII temporadas, 1946–1950. Anales tions in 12 residential midden deposits at seven sites del INAH 8:37–115. in the Yautepec Valley. The middens date to the Teoti- 1957 Resumen de los informes de las exploraciones huacan imperial phase, covering the Terminal Forma- arqueológicas en Tula, Hidalgo, durante las IX tive through Middle Classic periods at Teotihuacan y X temporadas, 1953–1954. Anales del INAH (Millon 1973, 1992). “Imperial imports” refers to Thin 9:119–169 Orange, whose distribution was controlled by Teoti- Adams, Robert McC. huacan. “Imperial style, local origin” includes two 1979 Late prehispanic empires of the New World. In imitation Thin Orange types, censers, floreros, and Power and propaganda: A symposium on ancient candeleros; this interpretation is based upon macro- empires, edited by Mogens T. Larsen, pp. 59–73. scopic analysis of ceramic pastes. Large basins Akademisk Forlag, Copenhagen. (“craters”) and red-on-buff bowls are stylistically Alcock, Susan very similar to those from Teotihuacan, but our pre- 1993 Graecia Capta: The landscapes of Roman Greece. liminary neutron activation analyses exclude the Cambridge Univ. Press, New York. Teotihuacan area as the place of manufacture (Mon- tiel n.d.). Until we conduct further characterization Alcock, Susan E., Terence N. D’Altroy, Kathleen D. analyses on sources in Morelos, we include them in Morrison, and Carla M. Sinopoli (Editors) the category “Imperial style, uncertain origin.” As 2001 Empires: Comparative perspectives from archaeol- discussed in the text, Granular ware is probably from ogy and history. Cambridge Univ. Press, New Guerrero, so we classify it in the category “Other im- York. ports” for Classic-period Yautepec. Fuller presenta- Algaze, Guillermo tion of these data will be given in Montiel (n.d.). For 1993 Expansionary dynamics of some early pristine Tula, the data are mean frequencies for all 11 sites states. American Anthropologist 95:304–333. with single-component, Early Postclassic, surface col- Andrews, Anthony P. lections from the Albany Yautepec Valley survey. We 1990 The fall of Chichen Itza: A preliminary hypothe- classify Macana Red-on-Brown as an “imperial im- sis. Latin American Antiquity 1:258–267. port,” and cream-slipped sherds (including some or- Arnold, Phillip J., III, Christopher A. Pool, Ronald R. ange on cream) as “imperial style, unertain origin”; Kneebone, and Robert S. Santley see text for discussion. For Tenochtitlan, the data are 1993 Intensive ceramic production and Classic-pe- mean frequencies for eight residential middens from riod political economy in the Sierra de los ϩ Molotla-phase (A.D. 1440–1520 ) houses excavated Tuxtlas, Veracruz, Mexico. Ancient Mesoamerica at Yautepec (Smith et al. 1999). “Imperial imports” 4:175–191. refers to ceramics from the Basin of Mexico, including Balkansky, Andrew K. Aztec III black-on-orange, Texcoco Fabric-Marked 1998 Urbanism and early state formation in the salt vessels, and several additional low-frequency Huamelulpan Valley of Southern Mexico. types (the Basin origin of the first two types has been Latin American Antiquity 9:37–67. confirmed through neutron activation). “Imperial style, local origin” describes a local imitation of Aztec Barbour, Warren III black-on-orange. Polished redwares, or guinda ce- 1979 The figurines and figurine chronology of ancient 274 SMITH AND MONTIEL

Teotihuacán Mexico. Ph.D. dissertation, Univ. of Boone, Elizabeth H. Rochester, Rochester. 2000 Venerable place of beginnings: The Aztec un- Bartel, Brad derstanding of Teotihuacan. In Mesoamerica’s 1980 Colonialism and cultural responses: Problems classic heritage: From Teotihuacan to the Aztecs, related to Roman provincial analysis. World edited by Davíd Carrasco, Lindsay Jones, and Archaeology 12:11–26. Scott Sessions, pp. 371–396. Univ. Press of Col- Bauer, Brian S. orado, Niwot. 1998 The sacred landscape of the Inca: The Cusco ceque Boone, Elizabeth H., and Michael E. Smith system. Univ. of Texas Press, Austin. in Postclassic international styles and symbol Begley, V., and R. D. de Puma (Editors) press systems. In The postclassic Mesoamerican world, 1991 Rome and India. Univ. of Wisconsin Press, edited by Michael E. Smith and Frances F. Madison. Berdan. Univ. of Utah Press, Salt lake City. Berdan, Frances F., and Patricia R. Anawalt (Editors) Borhegyi, Stephan F. 1992 The codex Mendoza, 4 vols. Univ. of California 1971 Pre-Columbian contacts—The dryland ap- Press, Berkeley. proach: The impact and influence of Teotihua- Berdan, Frances F., Richard E. Blanton, Elizabeth H. can culture on the pre-Columbian civilizations Boone, Mary G. Hodge, Michael E. Smith, and Emily of Mesoamerica. In Man across the sea: Problems Umberger of pre-Columbian contacts, edited by Carroll L. 1996 Aztec imperial strategies. Dumbarton Oaks, Riley, J. Charles Kelley, C. W. Pennington, Washington, DC. and Robert L. Rands, pp. 79–105. Univ. of Texas Press, Austin. Berlo, Janet Catherine 1984 Teotihuacan art abroad: A study of metropolitan style Brambila, Rosa, and Carlos Castañeda and provincial transformation in incensario work- 1993 Los basamentos con espacios hundidos. shops. BAR International Series, no. 199. British Cuadernos de Arquitectura Mesoamericana Archaeological Reports, Oxford. 25:73–78. Berlo, Janet C. (Editor) Braniff C, Beatriz 1992 Art, ideology, and the city of Teotihuacan. Dum- 1975 Arqueología del norte de México. In Los pueb- barton Oaks, Washington, DC. los y señoríos teocráticos: El periodo de las ciudades Bernal, Ignacio urbanas, primera parte, edited by Eduardo Matos 1966 Teotihuacan: ¿Capital de imperio? Revista Moctezuma et al., pp. 217–272. Instituto Na- Mexicana de Estudios Antropológicos 20:95– 110. cional de Antropología e Historia, Mexico City. Berrin, Kathleen (Ed.) 1999 Morales, Guanajuato y la tradición Tolteca. Insti- 1988 Feathered serpents and flowering trees: Recon- tuto Nacional de Antropología e Historia, structing the murals of Teotihuacan. Fine Arts Mexico City. Museums, San Francisco. Braund, David Blanton, Richard E. 1984 Rome and the friendly king: The character of the 1978 Monte Alban: Settlement patterns at the ancient client kingship. St. Martin’s, New York. Zapotec capitol. Academic Press, New York. Briggs, Lawrence P. 1983 The urban decline of Monte Alban. In The cloud 1951 The ancient Khmer empire. American Philos. people: Divergent evolution of the Zapotec and Soc., Philadelphia. Mixtec civilizations, edited by Kent V. Flannery Broda, Johanna, David Carrasco, and Eduardo Matos and Joyce Marcus, p. 186. Academic Press, New York. Moctezuma 1987 The great temple of Tenochtitlan: Center and pe- 1996 A consideration of causality in the growth of riphery in the Aztec world. Univ. of California empire: A comparative perspective. In Aztec Imperial strategies, pp. 219–225. Dumbarton Press, Berkeley. Oaks, Washington, DC. Brumfiel, Elizabeth M. Blanton, Richard E., and Gary M. Feinman 1998 Huitzilopochtli’s conquest: Aztec ideology in 1984 The Mesoamerican world system. American the archaeological record. Cambridge archaeo- Anthropologist 86:673–682. logical journal 8:3–14. Blanton, Richard E., Gary M. Feinman, Stephen A. Cabrera, Oralia Kowalewski, and Peter N. Peregrine 1999 Textile production and the political economy of the 1996 A dual-processual theory for the evolution of Teotihuacan state. Paper presented at the 1999 Mesoamerican civilization. Current Anthropol- Annual Meeting, Society for American Ar- ogy 37:1–14. chaeology, Chicago. ARCHAEOLOGICAL STUDY OF EMPIRES 275

Cabrera Castro, Rubén Chase-Dunn, Christopher, and Thomas D. Hall 1992 A survey of recently excavated murals at Teoti- 1997 Rise and demise: Comparing world-systems. West- huacan. In Art, ideology, and the city of Teotihua- view, Boulder. can, edited by Janet Catherine Berlo, pp. Cherry, John F. (Ed.) 113–128. Dumbarton Oaks, Washington, DC. 1992 Archaeology of empires. World Archaeology Cabrera Castro, Rubén, Ignacio Rodríguez García, and 23(2). Noel Morelos García Cobean, Robert H. 1991a Teotihuacan, 1980–1982: Nuevas interpreta- 1990 La cerámica de Tula, Hidalgo. Instituto Nacional ciones. Instituto Nacional de Antropología e de Antropología e Historia, Mexico City. Historia, Mexico City. Cobean, Robert H., and Alba Guadalupe Mastache Cabrera Castro, Rubén, Saburo Sugiyama, and George 1999 Tepetitlán: A rural household in the Toltec heart- L. Cowgill land/Tepetitlán: Un espacio doméstico rural en el 1991b The Templo de Quetzalcoatl project at Teoti- area de Tula. University of Pittsburgh Latin huacan: A preliminary report. Ancient Meso- American Archaeology Publications, and In- america 2:77–92. stituto Nacional de Antropología e Historia, Calnek, Edward E. Pittsburgh/Mexico City. 1976 The internal structure of Tenochtitlan. In The Coe, Michael D. Valley of Mexico: Studies of pre-Hispanic ecology 1994 Mexico: From the Olmecs to the Aztecs, 4th ed. and society, edited by Eric R. Wolf, pp. 287–302. Thames and Hudson, New York. Univ. of New Mexico Press, Albuquerque. 1999 The Maya, 6th ed. Thames and Hudson, New Carrasco, Davíd York. 1982 Quetzalcoatl and the irony of empire: Myths and Cook, J. M. prophecies in the Aztec tradition. Univ. of 1983 The Persian empire. Schocken, New York. Chicago Press, Chicago. Costin, Cathy, Timothy Earle, Bruce Owen, and Glenn 1999 City of sacrifice: The Aztec empire and the role of Russel violence in civilization. Beacon Press, Boston. 1989 Impact of Inka conquest on local technology in Carrasco, Pedro the upper Mantaro Valley, Peru. In What’s new, 1996 Estructura político-territorial del imperio tenochca: edited by S. E. van der Leeuw and R. Torrence, La triple alianza de Tenochtitlan, Tetzcoco y Tlaco- pp. 107–139. Unwin Hyman, London. pan. Fondo de Cultura Económica y El Colegio Costin, Cathy Lynne, and Timothy K. Earle de México, Mexico City. 1989 Status distinction and legitimation of power as Carter, Harold reflected in changing patterns of consumption 1983 An introduction to urban historical geography. in late prehispanic Peru. American Antiquity Arnold, Baltimore. 54:691–714. Cascio (Montiel), Lisa M., Timothy Hare, and Michael Cotterell, Arthur E. Smith 1981 The first emperor of China. Penguin, New York. 1995 Archaeological survey of the Yautepec Valley, Cowgill, George L. Morelos, Mexico. Paper presented at the 1995 1997 State and society at Teotihuacan, Mexico. An- Annual Meeting, Society for American Ar- nual Review of Anthropology 26:129–161. chaeology, Minneapolis. 2000 The central Mexican highlands from the rise of Cepeda Cárdenas, Gerardo, Ernesto González Licón, Teotihuacan to the decline of Tula. In The and Guillermo Ahuja Ormaechea Cambridge history of the native peoples of the 1977 Rescate arqueológico en el barrio de Tepito, Americas, volume II, Mesoamerica, part 1, edited México, D.F. In Los procesos de cambio en by Richard E. W. Adams and Murdo J. Mesoamérica y áreas circunvecinas. XV mesa re- MacLeod, pp. 250–317. Cambridge Univ. donda, pp. 397–402. Sociedad Mexicana de Press, New York. Antropología, Mexico City. Crespo, Ana María Charlton, Thomas H. 1991 El recinto ceremonial de El Cerrito. In Queré- 1978 Teotihuacan, Tepeapulco, and obsidian ex- taro prehispánico, edited by Ana María Crespo ploitation. Science 200:1227–1235. and Rosa Brambila, pp. 163–224. Instituto Na- 1991 The influence and legacy of Teotihuacan on re- cional de Antropología e Historia, Mexico gional routes and urban planning. In Ancient City. road networds and settlement hierarchies in the 1998 La expansión de la frontera norte (y la new world, edited by Charles D. Trombold, pp. cronología oficial para Teotihuacán). In Los rit- 186–197. Cambridge Univ. Press, New York. mos de cambio en Teotihuacán: Reflexiones y discu- 276 SMITH AND MONTIEL

siones de su cronología, edited by Rosa Brambila Earle, Timothy and Rubén Cabrera, pp. 323–334. Instituto Na- 1994 Wealth finance in the Inka empire: Evidence cional de Antropología e Historia, Mexico City. from the Calchaquí Valley, Argentina. Ameri- Czwarno, R. M., F. M. Meddens, and A. Morgan (Edi- can Antiquity 59:443–460. tors) Edens, Christopher 1989 The Nature of Wari. BAR International Series, 1992 Dynamics of trade in the ancient Meso- no. 525. British Archaeological Reports, Ox- potamian “world system.” American Anthro- ford. pologist 94:118–139. D’Altroy, Terence N. Edens, Christopher, and Philip Kohl 1992 Provincial power in the Inka empire. Smithsonian 1993 Trade and world systems in early Bronze Age Institution Press, Washington, DC. western Asia. In Trade and exchange in prehis- Davies, Nigel toric Europe, edited by Christopher Scarre and 1968 Los señoríos independientes del imperio Azteca. In- F. Healy, pp. 17–34. Oxbow Books, Oxford. stituto Nacional de Antropología e Histo- Feinman, Gary M., and Linda M. Nicholas ria, Mexico City. 1991 New perspectives on prehispanic highland Mesoamerica: A macroregional approach. 1973 The Aztecs: A history. Univ. of Oklahoma Press, Comparative Civilizations Review 24:13–33. Norman. Feldman, Lawrence H. 1977 The Toltecs until the fall of Tula. Univ. of Okla- 1974 Tollan in Hidalgo: Native accounts of the cen- homa Press, Norman. tral Mexican Tolteca. In Studies of ancient Tollan: 1987 The Aztec empire: The Toltec resurgence. Univ. of A report of the University of Missouri Tula Archae- Oklahoma Press, Norman. ological Project, edited by Richard A. Diehl, pp. de la Fuente, Beatriz, Silvia Trejo, and Gutiérrez 130–149. Univ. of Missouri, Columbia, MO. Solana, Nelly Flannery, Kent V. 1988 Escultura en piedra de Tula: Catálogo. Instituto de 1998 The ground plans of archaic states. In Archaic Investigaciones Estéticas, Universidad Na- States, edited by Gary M. Feinman and Joyce cional Autónoma de México, Mexico City. Marcus, pp. 15–57. School of American Re- DeMarrais, Elizabeth, Luis Jaime Castillo, and Timo- search Press, Santa Fe, NM. thy Earle Flores Morales, Luz María, and Ana María Crespo 1996 Ideology, materialization, and power strate- Oviedo gies. Current Anthropology 37:15–31. 1988 Elementos cerámicos de asentamientos Tolte- Díaz Oyarzábal, Clara Luz cas en Guanajuato y Querétaro. In Ensayos de alfarería prehispánica e histórica de Mesoamérica: 1980 Chingú: Un sitio clásico del área de Tula, Hidalgo. Homenaje a Eduardo Noguera Auza, edited by Instituto Nacional de Antropología e Historia, Mari Carmen Serra Puche and Carlos Navar- Mexico City. rete Cáceres, pp. 205–220. Universidad Na- 1991 Las figurillas teotihuacanas en Chingú, Hi- cional Autónoma de México, Mexico City. dalgo. In Teotihuacan 1980–1982: Nuevas inter- Fowler, William R., Jr. pretaciones, edited by Rubén Cabrera Castro, 1989 The cultural evolution of ancient Nahua civiliza- Ignacio Rodrígez García, and Noel Morelos tions: The Pipil-Nacarao of Central America. Univ. García, pp. 259–272. Instituto Nacional de of Oklahoma Press, Norman. Antropología e Historia, Mexico City. Fritz, John M., G. Michell, and M. S. Negaraja Rao 1998 Ocoyoacac: Un sitio con influencia Teotihua- 1984 Where kings and gods meet: The royal center at Vi- cana en el valle de Toluca. In Los ritmos de cam- jayanagara, India. Univ. of Arizona Press, Tucson. bio en Teotihuacán: Reflexiones y discusiones de su García Cook, Ángel cronología, edited by Rosa Brambila and Rubén 1981 The historical importance of Tlaxcala in the Cabrera, pp. 353–375. Instituto Nacional de cultural development of the central highlands. Antropología e Historia, Mexico City. In Supplement to the Handbook of Middle Ameri- Diehl, Richard A. can Indians, Volume 1, Archaeology, edited by 1983 Tula: The Toltec capital of ancient Mexico. Thames Victoria R. Bricker and Jeremy A. Sabloff, pp. and Hudson, New York. 244–276. Univ. of Texas Press, Austin. 1993 The Toltec horizon in Mesoamerica: New per- García Cook, Ángel, and B. Leonor Merino Carrión spectives on an old issue. In Latin American 1997 Notas sobre asentamientos prehispánicos for- horizons, edited by Don S. Rice, pp. 263–294. tificados o en posición estratégica del suroeste Dumbarton Oaks, Washington, DC. de Puebla. In Homenaje al profesor César A. Doyle, Michael W. Sáenz, edited by Ángel García Cook, Alba 1986 Empires. Cornell Univ. Press, Ithaca. Guadalupe Mastache, Leonor Merino, and ARCHAEOLOGICAL STUDY OF EMPIRES 277

Sonia Rivero Torres, pp. 367–388. Instituto Na- Healan, Dan M., and Christine E. Hernandez cional de Antropología e Historia, Mexico City. 2000 The Toltec world as seen from northeast Michoa- García Cook, Ángel, and Elia del Carmen Trejo can. Annual Conference, Society for American 1977 Lo teotihuacano en Tlaxcala. Comunicaciones Archaeology, Philadelphia. 14: 57–70. Healan, Dan M., Janet M. Kerley, and George J. Bey III Gerhard, Peter 1983 Excavation and preliminary analysis of an ob- 1977 Congregaciones de indios en la Nueva España sidian workshop in Tula, Hidalgo, Mexico. ántes de 1570. Historia Mexicana 26:347–395. Journal of Field Archaeology 10:127–146. Gibson, Charles Hernández, Carlos, Robert H. Cobean, Alba 1966 Spain in America. Harper & Row, New York. Guadalupe Mastache, and Maria Elena Suárez Gillespie, Susan D. 1999 Un taller de alfareros en la antigua ciudad de 1998 The Aztec triple alliance: A postconquest tradi- Tula. Arqueología 22:69–88. tion. In Native traditions in the postconquest Hernández Rivero, José world, edited by Elizabeth Hill Boone and Tom 1994a Arqueología de la frontera Tarasco–Mexica: Con- Cummins, pp. 233–263. Dumbarton Oaks, formación, estrategía y practicas de control. Licen- Washington, DC. ciatura, Escuela Nacional de Antropología e Gorenstein, Shirley Historia, Mexico City. 1985 Acambaro: Frontier settlement on the 1994b La arqueología de la frontera Tarasco–Mex- Tarascan–Aztec border. Vanderbilt University ica: Arquitectura bélica. In Contribuciones a Publications in Anthropology, no. 32. Depart- la arqueología y etnohistoria del Occidente de ment of Anthropology, Vanderbilt Univ., México, edited by Eduardo Williams, pp. Nashville. 115–155. El Colegio de Michoacan, Zamora. Granados Vázquez, Daniel, Jorge Carrandi Ríos, Hirth, Kenneth G. Garbiel Lalo Jacinto, and Belem Zúñiga Arellano 1978 Teotihuacan regional population administra- 1993 Apuntes sobre otros materiales arqueológicos tion in eastern Morelos. World Archaeology procedentes de Tlalpizáhuac. Expresión 9:320–333. Antropológica 3:117–132. 1980 Eastern Morelos and Teotihuacan: A settlement Hare, Timothy S. survey. Vanderbilt University Publications in 2000 Political economy, spatial analysis, and postclassic Anthropology, no. 25. Department of Anthro- states in the yautepec Valley, Mexico. Ph.D. dis- pology, Vanderbilt Univ., Nashville. sertation, Department of Anthropology, Univ. at Albany, SUNY. 1996 Teotihuacán clásico: Una perspectiva regional sobre el valle oriental de Morelos. Cuadernos de Hare, Timothy S., Lisa Montiel, and Michael E. Smith n.d. Prehispanic settlement patterns in the Yautepec Arquitectura Mesoamericana 30:13–44. Valley, Morelos, Mexico. Manuscript in prepa- Hodge, Mary G., Hector Neff, M. James Blackman, ration. and Leah D. Minc Haselgrove, Colin 1993 Black-on-orange ceramic production in the 1987 Culture process on the periphery: Belgic Gaul Aztec empire’s heartland. Latin American An- and Rome during the late Republic and early tiquity 4:130–157. Empire. In Centre and periphery in the ancient Hopkins, Keith world, edited by Michael Rowlands, Mogens 1978 Economic growth and towns in classical antiq- Larsen, and Kristian Kristiansen, pp. 104–124. uity. In Towns in societies, edited by Philip Cambridge Univ. Press, Cambridge. Abrams and E. A. Wrigley, pp. 35–77. Univer- Hassig, Ross sity Press, Cambridge. 1985 Trade, tribute, and transportation: The Sixteenth 1980 Taxes and trade in the Roman Empire (200 century political economy of the Valley of Mexico. B.C.–A.D. 400). Journal of Roman Studies Univ. of Oklahoma Press, Norman. 70:101–125. 1992 War and society in ancient Mesoamerica. Univ. of Hosler, Dorothy California Press, Berkeley. 1994 The sounds and colors of power: The sacred metal- Healan, Dan M. lurgical technology of ancient west Mexico. MIT 1989 Tula of the Toltecs: Excavations and survey. Univ. Press, Cambridge. of Iowa Press, Iowa City. Hosler, Dorothy, and Andrew Macfarlane 1993 Local versus non-local obsidian exchange at 1996 Copper sources, metal production and metals Tula and its implications for post-formative trade in late Postclassic Mesoamerica. Sci- Mesoamerica. World Archaeology 24:449–466. ence 273:1819–1824. 278 SMITH AND MONTIEL

Hyslop, John 1999 The architecture of the Tula body politic. In 1984 The Inka road system. Academic Press, New Mesoamerican Architecture as a Cultural Symbol, York. edited by Jeff Karl Kowalski, pp. 162–175. Ox- 1990 Inka settlement planning. Univ. of Texas Press, ford Univ. Press, New York. Austin. Kuhrt, Amélie Isaac, Barry L. 2001 The Araemenid empire: Continuities, adapta- 1986 Notes on obsidian, the Pochteca, and the posi- tions, transformations. In Empires: perspectives tion of Tlatelolco in the Aztec Empire. In Eco- from archaeology and history, edited by Susan E. nomic aspects of highland central Mexico, edited Alcock, Terence N. D’Altroy, Kathleen D. Mor- by Barry L. Isaac, pp. 319–343. JAI Press, rison, and Carla M. Sinopoli. pp. 93–123. Cam- Greenwich, CT. bridge Univ. Press, New York. Isaac, Benjamin Lagunas R., Zaid 1990 The limits of empire: The Roman army in the east. 1997 Costumbres funerarias y caracteristicas biocul- Clarendon, Oxford. turales de la población prehispánica de Hua- Isbell, William, and Gordon McEwan (Editors) mango. Expresión Antropológica (Instituto Mex- 1991 Huari administrative structures. Dumbarton iquense de Cultura) 6:7–28. Oaks, Washington, DC. Lange, Frederick W. Jiménez Moreno, Wigberto 1986 Central America and the Southwest: A com- 1941 Tula y los Toltecas según las fuentes históricas. parison of Mesoamerica’s two peripheries. Revista Mexicana de Estudios Antropológicos In Research and reflections in archaeology and his- 5:79–83. tory: Essays in honor of Doris Stone, edited by E. Jones, Christopher Wyllys Andrews V, pp. 159–177. Middle Amer- 1996 Excavations in the east plaza of Tikal. University ican Research Institute, Tulane Univ., New Or- Museum, Univ. of Pennsylvania, Philadelphia. leans. Jones, Lindsay Larsen, Mogens T. 1995 Twin city tales: A hermeneutical reassessment of 1979 The tradition of empire in Mesopotamia. In Tula and Chichen Itza. Univ. Press of Colorado, Power and propaganda: A symposium on ancient Niwot. empires, edited by Mogens T. Larsen, pp. Joyce, Arthur A. 75–106. Akademisk Forlag, Copenhagen. 1993 Interregional interaction and social develop- Lee, Thomas A., Jr. ment on the Oaxaca coast. Ancient Mesoamer- 1978 The origin and developent of plumbate pot- ica 4:67–84. tery. Revista Mexicana de Estudios Antropológicos n.d. Imperialism in pre-Aztec Mesoamerica: Monte 24:287–300. Albán, Teotihuacan, and the lower Río Verde Linné, Sigvald Valley. In Warfare and conflict in ancient 1942 Mexican highland cultures: Archaeological re- Mesoamerica, edited by M. Kathryn Brown and searches at Teotihuacan, Calpulalpan and Chal- Travis M. Stanton. [In preparation] chicomula in 1934/35. Ethnographical Museum Kelley, David H. of Sweden, Stockholm. 1992 Yucatán y el imperio tolteca. Arqueología Liverani, Mario (Editor) 8:113–119. 1993 Akkad: The first world empire. Sargon, Padua. Kepecs, Susan, Gary M. Feinman, and Sylviane Boucher López Austin, Alfredo, and Leonardo López Luján 1994 Chichen Itza and its hinterland: A world- 2000 The myth and reality of Zuyuá: The feathered systems perspective. Ancient Mesoamerica 5: serpent and Mesoamerican transformations 141–158. from the Classic to the Postclassic. In Mesoamerica’s classic heritage: From Teotihuacan Kirchhoff, Paul 1985 El imperio tolteca y su caída. In Mesoamérica y to the Aztecs, edited by Davíd Carrasco, Lind- el centro de México: Una antología, edited say Jones, and Scott Sessions, pp. 21–84. Univ. by Jesús Monjarás-Ruiz, Rosa Brambila, and Press of Colorado, Niwot. Emma Pérez-Rocha, pp. 249–272. Instituo Na- López Luján, Leonardo cional de Antropología e Historia, Mexico 1994 The offerings of the Templo Mayor of Tenochtitlan. City. Univ. Press of Colorado, Niwot. Kristan-Graham, Cynthia B. Luttwak, Edward N. 1993 The business of narrative at Tula: An analysis 1976 The grand strategy of the Roman Empire from the of the vestibute frieze, trade, and ritual. Latin first century A.D. to the third. Johns Hopkins American Antiquity 4:3–21. Univ. Press, Baltimore. ARCHAEOLOGICAL STUDY OF EMPIRES 279

Lynch, Kevin Matos Moctezuma, Eduardo (Editor) 1981 A theory of good city form. MIT Press, Cam- 1979 Trabajos arqueológicos en el centro de la ciudad de bridge. México. Instituto Nacional de Antropología e Maldonado Jiménez, Druzo Historia, Mexico City. 1990 Cuauhnahuac y Huaxtepec: Tlalhuicas y Xochimil- 1988 The great temple of the Aztecs. Thames and Hud- cas en el Morelos prehispánico. Centro Regional son, New York. de Investigaciones Multidisciplinarias, Uni- 1999 Excavaciones en la catedral y el sagrario metropoli- versidad Nacional Autónoma de México, tanos: Programa de arqueología urbana. Instituto Cuernavaca. Nacional de Antropología e Historia, Mexico Malpass, Michael A. (Editor) City. 1993 Provincial Inca: Archaeological and ethnohistorical Matos Moctezuma, Eduardo, María Teresa García assessment of the impact of the Inca state. Univ. of García, Fernando López Aguilar, and Ignacio Iowa Press, Iowa City. Rodríguez García Malpass, Michael A., and Justin Jennings 1981 Proyecto Tepeapulco: Resumen preliminar de 1999 Recent research on the Wari polity of the cen- las actividades realizadas en la primera tem- tral Andes. In Symposium at the 1999 Annual porada de trabajo. In Interacción cultural en Conference, Society for American Archaeology, México central, edited by Evelyn Childs Rat- Chicago. tray, Jaime Litvak King, and Clara Díaz Mann, Michael Oyarzábal, pp. 113–148. Universidad Nacional 1986 The sources of social power, volume 1: A history of Autónoma de México, Mexico City. power from the beginning to A.D. 1760. Cam- McEwan, Gordon F. bridge Univ. Press, New York. 1987 The middle horizon in the Valley of Cuzco, Peru: Marcus, Joyce The impact of the Wari occupation of the Lucre 1983 Teotihuacan visitors on Monte Alban monu- Basin. BAR International Series, no. 372. British ments and murals. In The cloud people: Diver- Archaeological Reports, Oxford. gent evolution of the Zapotec and Mixtec civiliza- Medellín Zeñil, Alfonso tions, edited by Kent V. Flannery and Joyce 1952 Exploraciones de Quauhtochco. Gobierno del Es- Marcus, pp. 175–181. Academic Press, New tado de Veracruz, Jalapa. York. Millar, Fergus G. 1992 Mesoamerican writing systems: Propaganda, 1995 The Roman Near East: 31 BC to AD 337. Harvard myth, and history in four ancient civilizations. Univ. Press, Cambridge. Princeton Univ. Press, Princeton. Miller, Arthur G. Marcus, Joyce, and Kent V. Flannery 1967 The birds of Quetzalpapalotl. Ethnos 32:5–17. 1996 Zapotec civilization: How urban society evolved in 1973 The mural painting of Teotihuacan, Dumbarton Mexico’s Oaxaca Valley. Thames and Hudson, Oaks, Washington, DC. New York. Millett, Martin Martínez Donjuan, Guadalupe 1990 The Romanisation of Britain. Cambridge Univ. 1979 Las Pilas, Morelos. Instituto Nacional de Press, New York. Antropología e Historia, Mexico City. Millon, Clara Mastache, Alba Guadelupe, and Robert H. Cobean 1973 Painting, writing, and polity in Teotihuacan, 1985 Tula. In Mesoamérica y el centro de México: Una Mexico. American Antiquity 38:294– 314. antología, edited by Jesús Monjarás-Ruiz, Rosa Millon, René Brambila, and Emma Pérez-Rocha, pp. 1973 Urbanization at Teotihuacan, Mexico, volume one, 273–307. Instituo Nacional de Antropología e The Teotihuacan map. Univ. of Texas Press, Historia, Mexico City. Austin. 1989 The Coyotlatelco culture and the origins of the 1981 Teotihuacan: City, state, and civilization. In Ar- Toltec state. In Mesoamerica after the decline of chaeology, edited by Jeremy Sabloff, pp. Teotihuacan, AD 700–900, edited by Richard A. 198–243. Univ. of Texas Press, Austin. Diehl and Janet C. Berlo, pp. 49–68. Dumbar- 1988 The last years of Teotihuacan dominance. In ton Oaks, Washington DC. The collapse of ancient states and civilizations, Mastache, Alba Guadalupe, and Ana María Crespo edited by Norman Yoffee and George L. 1982 Análisis sobre la traza general de Tula, Hgo. In Cowgill, pp. 102–164. Univ. of Arizona Press, Estudios sobre la antigua ciudad de Tula, pp. Tucson. 11–36. Instituto Nacional de Antropología e 1992 Teotihuacan studies: From 1950 to 1990 and Historia, Mexico City. beyond. In Art, ideology, and the city of Teotihua- 280 SMITH AND MONTIEL

can, edited by Janet C. Berlo, pp. 339–429. Neff, Hector, and Michael D. Glascock Dumbarton Oaks, Washington, DC. 1996 Compositional analysis of Aztec black-on-orange 1994 The place where time began: An archaeolo- pottery from Yautepec, Morelos, Mexico. Unpub- gist’s interpretation of what happened in lished report. Missouri University Research Teotihuacan history. In Teotihuacan: Art from the Reactor, Univ. of Missouri. City of the Gods, edited by Kathleen Berrin and Nicholson, H. B. Esther Pasztory, pp. 16–43. The Fine Arts Mu- 2000 The iconography of the feathered serpent in seums, San Francisco. late Postclassic central Mexico. In Mesoamer- Moholy-Nagy, Hattula ica’s classic heritage: From Teotihuacan to the 1999 Mexican obsidian at Tikal, Guatemala. Latin Aztecs, edited by Davíd Carrasco, Lindsay American Antiquity 10:300–313. Jones, and Scott Sessions, pp. 145–164. Univ. Montiel, Lisa Press of Colorado, Niwot. 1998 On the border of the Teotihuacan empire: Archaeo- Noguez, Xavier logical investigations in the Yautepec Valley. 1995 La zona del Altiplano central en el Posclásico: Paper presented at the 1998 Annual Meeting, La etapa tolteca. In Historia antigua de México. Society for American Archaeology, Seattle. vol. 3, el horizonte Posclásico y algunos aspectos 1999 Figurines and Teotihuacan imperialism in the intelectuales de las culturas mesoamericanas, Yautepec Valley, Morelos. Paper presented at the edited by Linda Manzanilla and Leonardo 1999 Annual Meeting, Society for American López Luján, pp. 189–224. Instituto Nacional Archaeology, Chicago. Antropología e Historia, Mexico City. n.d. Teotihuacan imperialism in the Yautepec Valley, O’Mack, Scott Morelos, Mexico. Ph.D. dissertation, Depart- 1986 Reconocimiento arqueológico en el pueblo de Tu- ment of Anthropology, Univ. at Albany, SUNY. tutepec, Oaxaca. Unpublished report. Centro Moreland, John Regional de Oaxaca, Instituto Nacional de 2001 The Carolingian Empire: Rome reborn? in Em- Antropología e Historia, Oaxaca. pires: perspectives from archaeology and history, Ortiz, Ponciano, and Robert Santley edited by Susan E. Alcock, Terence N. D’Al- 1998 Matacapan: Un ejemplo de enclave teotihua- troy, Kathleen D. Morrison, and Carla M. Si- cano en la costa del Golfo. In Los ritmos de nopoli, pp. 392–418. Cambridge Univ. Press, cambio en Teotihuacán: Reflexiones y discusiones New York. de su cronología, edited by Rosa Brambila and Morkot, Robert Rubén Cabrera, pp. 377–460. Instituto Na- 2001 Egypt and Nubia. In Empires: Perspectives from cional de Antropología e Historia, Mexico City. archaeology and History, edited by Susan E. Al- Osborne, Robin cock, Terence N. D’Altroy, Kathleen D. Morri- 1987 Classical landscape with figures: The ancient Greek son, and Carla M. Sinopoli, pp. 227–251. Cam- bridge Univ. Press, New York. city and its countryside. George Philip, London. Morrison, Kathleen D. Owens, E. J. 1995 Fields of victory: Vijayanagara and the course of in- 1991 The city in the Greek and Roman world. Rout- tensification. University of California Archaeo- ledge, New York. logical Research Facility, Berkeley. Parsons, Jeffrey R. Morrison, Kathleen D., and Mark T. Lycett 1971 Prehistoric settlement patterns in the Texcoco re- 1994 Centralized power, centralized authority? Ide- gion, Mexico. Museum of Anthropology, Univ. ological claims and archaeological patterns. of Michigan, Ann Arbor. Asian Perspectives 33:327–350. Pastrana, Alejandro Murra, John V. 1998 La explotación Azteca de la obsidiana en la Sierra 1980 The economic organization of the Inka state. JAI de las Navajas. Instituto Nacional de Press, Greenwich, CT. Antropología e Historia, Mexico City. Nalda, Enrique Pasztory, Esther 1997 El noreste de Morelos y la desestabilización 1997 Teotihuacan: An experiment in living. Univ. of teotihuacana. Arqueología 18:103–117. Oklahoma Press, Norman. Navarrete, Carlos Pasztory, Esther A. (Editor) 1996 Elementos arqueológicos de mexicanización en 1978 Middle classic Mesoamerica: AD 400–700. Co- las tierras altas mayas. In Temas mesoamericanos, lumbia Univ. Press, New York. edited by Sonia Lombardo and Enrique Nalda, Peregrine, Peter N. pp. 305–342. Instituto Nacional de Antropología 1996 Introduction: World-systems theory and ar- e Historia, Mexico City. chaeology. In Pre-Columbian world systems, ARCHAEOLOGICAL STUDY OF EMPIRES 281

edited by Peter N. Peregrine and Gary M. Fein- onomy of thin orange ceramics from the man- man, pp. 1–10. Prehistory Press, Madison, WI. ufacturing sites of Rio Carnero, Puebla, Mex- Peregrine, Peter N., and Gary M. Feinman (Editors) ico. In Chemical characterization of ceramic pastes 1996 Pre-Columbian world systems. Prehistory Press, in archaeology, edited by Hector Neff, pp. Madison, WI. 221–232. Prehistory Press, Madison, WI. Pfannkuch Wachtel, Tilman, Soledad García Sánchez, Redmond, Elsa M. and Alejandro Tovalín Ahumada 1983 A fuego y sangre: Early Zapotec imperialism in the 1993 La cerámica arqueológica de Tlalpizáhuac. Ex- Cuicatlán Cañada, Oaxaca. Univ. of Michigan, presión Antropológica 3:43–73. Museum Anthropology, Ann Arbor. Piña Chán, Román Ringle, William M., Tomás Gallareta Negrón, and 1981 Investigaciones sobre Huamango y región vecina George J. Bey III (memoria del proyecto). Dirección de Turiso del 1998 The return of Quetzalcoatl: Evidence for the Gobierno del Estado de México, Toluca. spread of a world Religion during the epiclas- Pollard, Helen P. sic period. Ancient Mesoamerica 9:183–232. 1977 An analysis of urban zoning and planning in Robertson, Donald prehispanic Tzintzuntzan. Proceedings of the 1970 The Tulum murals: The international style of the American Philosophical Society 121:46–69. late post-Classic. In Verhandlunger del XXXVIII 1993 Tariacuri’s legacy: The prehispanic Tarascan state. Internationalen Amerikanisten-Kongres, Stuttgart- Univ. of Oklahoma Press, Norman. Munchen, 1968, Vol. 2, pp. 77–88. Kommis- Pollard, Helen P., and Thomas A. Vogel sionsverlag Klaus Renner, Stuttgart. 1994 Late postclassic imperial expansion and eco- Rojas, José Luis de nomic exchange within the Tarascan domain. 1986 México Tenochtitlan: Economía e sociedad en el In Economies and polities in the Aztec realm, siglo XVI. Fondo de Cultura Económica, Mex- edited by Mary G. Hodge and Michael E. ico City. Smith, pp. 447–470. Institute for Mesoameri- Román Berrelleza, Juan Alberto, and Leonardo López can Studies, Albany. Luján Postgate, J. N. 1999 El funeral de un dignatorio mexica. Arque- 1992 The land of Assur and the yoke of Assur. World ología Mexicana 7:36–39. Archaeology 23:247–263. Root, Margaret Cool Prem, Hanns J. 1979 The king and kingship in Achaemenid art: Essays 1997 The ancient Americas: A brief history and guide to on the creation of an iconography of empire. E. J. research. Univ. of Utah Press, Salt Lake City. Brill, Leiden. Rattray, Evelyn C. Sahagún, Fray Bernardino de 1979 La cerámica de Teotihuacan: Relaciones exter- 1950–1982 Florentine Codex, General history of the nas y cronología. Anales de antropología things of New Spain, 12 books. Translated 16:51–70. and edited by Arthur J. O. Anderson and 1981 Anaranjado delgado: Cerámica de comercio Charles E. Dibble. School of American Re- de Teotihuacan. In Interacción cultural en Méx- search and the Univ. of Utah Press, Santa ico central, edited by Evelyn C. Rattray, Jaime Fe/Salt Lake City. Litvak King, and Clara Díaz Oyarz´abal, pp. Saint-Charles Zetina, Juan Carlos 55–80. Universidad Nacional Autónoma de 1998 Los problemas para la periodificación del Bajío México, Mexico City. y la secuencia de San Juan del Río. In Los ritmos 1989 El barrio de los comerciantes y el conjunto de cambio en Teotihuacán: Reflexiones y discu- Tlamimilolpa: Un estudio comparativo. Arque- siones de su cronología, edited by Rosa Brambila ología 5:105–129. and Rubén Cabrera, pp. 335–345. Instituto Na- 1990 New findings on the origins of thin orange ce- cional de Antropología e Historia, Mexico City. ramics. Ancient Mesoamerica 1:181–196. Saint-Charles Zetina, Juan Carlos, and Miguel 1998 Resumen de las tendencias cronológicas en la Argüelles Gamboa cerámica y panorama general de Teotihuacán. 1991 Cerro de La Cruz: Persistencia de un centro In Los ritmos de cambio en Teotihuacan: Reflex- ceremonial. In Querétaro prehispánico, edited by iones y discusiones de su cronología, edited by Ana Ma. Crespo and Rosa Brambila, pp. 57–97. Rosa Brambila and Rubén Cabrera, pp. Instituto Nacional de Antropología e Historia, 255–281. Instituto Nacional de Antropología e Mexico City. Historia, Mexico City. Sanders, William T., and Joseph W. Michels (Editors) Rattray, Evelyn C., and Garman Harbottle 1977 Teotihuacan and Kaminaljuyu. Pennsylvania 1992 Neutron activation analysis and numerical tax- State Univ. University Park. 282 SMITH AND MONTIEL

Sanders, William T., Jeffrey R. Parsons, and Robert S. seum of Anthropology, University of Michi- Santley gan, Ann Arbor. 1979 The Basin of Mexico: Ecological processes in the 1999 Regional approaches to the study of prehistoric evolution of a civilization. Academic Press, New empires: Examples from Ayacucho and Nasca, York. Peru. In Settlement pattern studies in the Americas: Sanders, William T., and Robert S. Santley Fifty years since Virú, edited by Brian R. Billman 1983 A tale of three cities: Energetics and Urbaniza- and Gary M. Feinman, pp. 160–171. Smithson- tion in pre-Hispanic central Mexico. In Prehis- ian Institution Press, Washington, DC. toric settlement patterns: Essays in honor of Gor- Seler, Eduard don R. Willey, edited by Evon Z. Vogt and 1990–96 Excavations at the site of the principal tem- Richard Leventhal, pp. 243–291. Univ. of New ple in Mexico. In Collected works in Mesoameri- Mexico Press, Albuquerque. can linguistics and archaeoalogy, Vol. 3, pp. Santley, Robert S. 114–193. Labyrinthos, Culver City. 1984 Obsidian exchange, economic stratification, and Sharer, Robert the evolution of complex society in the Basin of 1995 Latest findings at Copán, Honduras. Final Report Mexico. In Trade and exchange in early Mesoamer- submitted to the Foundation for the Advance- ica, edited by Kenneth G. Hirth, pp. 43–86. Uni- ment of Mesoamerican Studies, Inc. versity of New Mexico Press, Albuquerque. Silverstein, Jay E. 1989 Obsidian working, long-distance exchange, 2000 A study of the Late postclassic Aztec–Tarascan and the Teotihuacan presence on the south frontier in northern Guerrero, Mexico: The Oz- Gulf Coast. In Mesoamerica after the decline of tuma–Cutzamala project. Ph.D. dissertation, De- Teotihuacan, AD 700–900, edited by Richard. A. partment of Anthropology, Pennsylvania State Diehl and Janet C. Berlo, pp. 131–151. Dum- University. barton Oaks, Washington, DC. Sinopoli, Carla M. Santley, Robert S., and Christopher A. Pool 1994 The archaeology of empires. Annual Review of 1993 Prehispanic exchange relationships among Anthropology 23:159–180. Central Mexico, the Valley of Oaxaca, and the Sinopoli, Carla M., and Kathleen D. Morrison Gulf Coast of Mexico. In The American South- 1995 Dimensions of imperial control: The Vijayana- west and Mesoamerica: Systems of prehistoric ex- gara capital. American Anthropologist 97:83–96. change, edited by Jonathon Ericson and Timo- Smith, Michael E. thy G. Baugh, pp. 179–211. Plenum, New York. 1986 The role of social stratification in the Aztec em- Santley, Robert S., Clare Yarborough, and Barbara pire: A view from the provinces. American An- Hall thropologist 88:70–91. 1987 Enclaves, ethnicity, and the archaeological 1990 Long-distance trade under the Aztec empire: record at Matacapan. In Ethnicity and culture, The archaeological evidence. Ancient Meso- edited by Reginald Auger, Margaret F. Glass, america 1:153–169. Scott MacEachern, and Peter H. McCartney, 1992 Archaeological research at Aztec-period rural sites pp. 85–100. Archaeological Association, Univ. in Morelos, Mexico, Volume 1, excavations and ar- of Calgary, Calgary. chitecture/Investigaciones arqueológicas en sitios Saville, Marshall H. rurales de la epoca Azteca en Morelos, tomo 1, ex- 1922 Turquoise mosaic art in ancient Mexico. Museum cavaciones y arquitectura. Monographs in Latin of the American Indian, Heye Foundation, American Archaeology, no. 4. Univ. of Pitts- New York. burgh, Pittsburgh. Schele, Linda, and David Freidel 1994 Economies and polities in Aztec-period More- 1990 A forest of kings: The untold story of the ancient los: Ethnohistoric introduction. In Economies Maya. William Morrow, New York. and polities in the Aztec realm, edited by Mary G. Schele, Linda, and Peter Mathews Hodge and Michael E. Smith, pp. 313–348. In- 1998 The code of kings: The language of seven sacred stitute for Mesoamerican Studies, Albany. Maya temples and tombs. Simon and Schuster, 1997a City planning: Aztec city planning. In Ency- New York. clopaedia of the history of non-Western science, Schondube, Otto technology, and medicine, edited by Helaine Selin, 2000 The nature of “Toltec-ness”: A view from west pp. 200–202. Kluwer Academic, Dordrecht. Mexico. Annual Conference, Society for Ameri- 1997b Life in the provinces of the Aztec empire. Sci- can Archaeology, Philadelphia. entific American 277(3):56–63. Schreiber, Katharina J. 2000 Aztec city-states. In A comparative study of 1992 Wari imperialism in middle horizon Peru. Mu- thirty city–state cultures, edited by Mogens ARCHAEOLOGICAL STUDY OF EMPIRES 283

Herman Hansen, pp. 581–595. The Royal Dan- 1989 Excavaciones recientes en Tlailotlaca, el barrio ish Academy of Sciences and Letters, Copen- Oaxaqueno de Teotihuacan. Arqueología hagen. 5:81–104. 2001 The Aztec empire and the Mesoamerican 1992 Tlailotlacan, a Zapotec enclave in Teotihuacan. world system. In Empires: Comparative perspec- In Art, ideology, and the city of Teotihuacan, tives from archaeology and history, edited by edited by Janet C. Berlo, pp. 59–88. Dumbar- Susan E. Alcock, Terence N. D’Altroy, Kath- ton Oaks, Washington, DC. leen D. Morrison, and Carla M. Sinopoli. pp. 1996 Commodity or gift: Teotihuacan obsidian in the 128–154. Cambridge Univ. Press, New York. Maya region. Latin American Antiquity 7:21–39. n.d.a Aztec military settlements. In Enciclopedia Spencer, Charles S., and Elsa M. Redmond Archeologica, edited by Marco Curatola. Inti- 1997 Archaeology of the Cañada de Cuicatlán, Oaxaca. tuto Della Enciclopedia Italiana, Rome, in American Museum of Natural History, New press. York. n.d.b Tlahuica ceramics: The Aztec-period ceramics of Spores, Ronald Morelos, Mexico. Institute for Mesoamerican 1993 Tututepec: A Mixtec conquest state. Ancient Studies, Albany, in press. Mesoamerica 4:167–174. Smith, Michael E., and Frances F. Berdan Stambaugh, John E. 1992 Archaeology and the Aztec empire. World Ar- 1988 The ancient Roman city. Johns Hopkins Univ. chaeology 23:353–367. Press, Baltimore. 2000 The postclassic Mesoamerican world system. Stark, Barbara L. Current Anthropology 41:283–286. 1990 The Gulf Coast and the central highlands of n.d. (Eds.) The postclassic Mesoamerican world. Univ. Mexico: Alternative models for interaction. Re- of Utah Press, Salt Lake City, in press. search in Economic Anthropology 12:243–285. Smith, Michael E., and Cynthia Heath-Smith Stein, Gil J. 1980 Waves of influence in postclassic Mesoamer- 1999 Rethinking world-systems: Diasporas, colonies, ica? A critique of the Mixteca– Puebla concept. and interaction in Uruk Mesopotamia. Univ. of Anthropology 4:15–50. Arizona Press, Tucson. 1994 Rural economy in late postclassic Morelos: An Steinhardt, Nancy S. archaeological study. In Economies and polities in 1990 Chinese imperial city planning. Univ. of Hawaii the Aztec realm, edited by Mary G. Hodge and Press, Honolulu. Michael E. Smith, pp. 349–376. Institute for Stuart, David Mesoamerican Studies, Albany. 2000 “The Arrival of Strangers”: Teotihuacan and Tol- Smith, Michael E., Cynthia Heath-Smith, and Lisa lan in classic Maya history. In Mesoamerica’s clas- Montiel sic heritage: From Teotihuacan to the Aztecs, edited 1999 Excavations of Aztec urban houses at Yaute- by Davíd Carrasco, Lindsay Jones, and Scott pec, Mexico. Latin American Antiquity 10:133– Sessions, pp. 465–514. University Press of Col- 150. orado, Niwot. Smith, Michael E., Hector Neff, and Ruth Fauman- Sugiyama, Saburo Fichman 1998 Archaeology and iconography of Teotihuacan 1999 Ceramic imports at Yautepec and their implications censers: Official military emblems originated for Aztec exchange systems. Paper presented at from the Cuidadela? Teotihuacan Notes: Internet the 1999 Annual Meeting, Society for Ameri- Journal for Teotihuacan Archaeology and Iconogra- can Archaeology, Chicago. phy. Solís Olguín, Felipe R., and Davíd A. Morales Gómez Thouvenot, Marc 1991 Rescate de un rescate: Colección de objetos arque- 1982 Chalchihuitl: Le jade chez les Aztèques. Institut ológicos de el Volador, Ciudad de México. Instituto d’Ethnologie, Paris. Nacional de Antropología e Historia, Mexico Tolstoy, Paul City. 1958 Surface survey of the northern Valley of Mexico: Spence, Michael W. The classic and post-classic periods. Transactions 1987 The scale and structure of obsidian production of the American Philosophical Society, vol. 48, in Teotihuacan. In Teotihuacán: Nuevos datos, no. 5. American Philos. Soc., Philadelphia. nuevas síntesis, Nuevos problemas, edited by Topic, John R., and Theresa Lange Topic Emily McClung de Tapia and Evelyn C. Rattray, 1993 A summary of the Inca occupation of Hua- pp. 439–450. Universidad Nacional Autónoma machuco. In Provincial Inca: Archaeological and de México, Mexico City. ethnohistorical assessment of the impact of the Inca 284 SMITH AND MONTIEL

state, edited by Michael A. Malpass, pp. 17–43. Weaver, Muriel Porter Univ. of Iowa Press, Iowa City. 1993 The Aztecs, Maya, and their predecessors: Archae- Torres Cabello, Ma . Olivia ology of Mesoamerica, 3rd ed. Academic Press, 1998 La cuenca de Chalco y su relación con Teoti- San Diego. huacán. In Los ritmos de cambio en Teotihuacán: Re- Weigand, Phil C. flexiones y discusiones de su cronología, edited by 1982 Mining and mineral trade in Prehispanic Za- Rosa Brambila and Rubén Cabrera, pp. 347–352. catecas. Anthropology 6:175–188. Instituto Nacional de Antropología e Historia, Weigand, Phil C., and Garman Harbottle Mexico City. 1993 The role of turquoises in the ancient Tovalín Ahumada, Alejandro Mesoamerican trade structure. In The American 1998 Desarrollo arquitectónico del sitio arqueológico de Southwest and Mesoamerica: Systems of prehis- Tlalpizáhuac. Instituto Nacional de Antropología toric exchange, edited by Jonathon Ericson and e Historia, Mexico City. Timothy G. Baugh, pp. 159–177. Plenum, New Townsend, Richard F. York. 1979 State and cosmos in the art of Tenochtitlan. Dum- Weigand, Phil C., Garman Harbottle, and Edward V. barton Oaks, Washington, DC. Sayre Tozzer, Alfred M. 1977 Turquoise sources and source analysis: 1957 Chichen Itza and its cenote of sacrifice: A compara- Mesoamerican and the Southwestern USA. In tive study of contemporaneous Maya and Toltec. Exchange systems in prehistory, edited by Timo- Peabody Museum, Harvard Univ., Cambridge. thy K. Earle and Jonathon E. Ericson, pp. 15–34. Academic Press, New York. Umberger, Emily 1987 Antiques, revivals, and references to the past Wells, Colin in Aztec art. Res 13:62–105. 1984 The Roman empire. Stanford Univ. Press, Stan- 1996 Art and imperial strategy in Tenochtitlan. In ford. Aztec Imperial Strategies, by Frances F. Berdan Whalen, Michael E., and Jeffrey R. Parsons et al., pp. 85–106. Dumbarton Oaks, Washing- 1982 Ceramic markers used for period designa- ton, DC. tions. In Prehispanic settlement patterns in the Umberger, Emily, and Cecilia Klein southern Valley of Mexico: The Chalco-Xochimilco 1993 Aztec art and imperial expansion. In Latin Amer- region, edited by Jeffrey R. Parsons et al., pp. ican Horizons, edited by Don S. Rice, pp. 385–459. Museum of Anthropology, Univ. of 295–336. Dumbarton Oaks, Washington, DC. Michigan, Ann Arbor. Vaillant, George C. Whittaker, C. R. 1938 A correlation of archaeological and historical 1983 Trade and frontiers of the Roman empire. In sequences in the Valley of Mexico. American Trade and famine in classical antiquity, edited by Anthropologist 40:535–573. Peter Garnsey and C. R. Whittaker, pp. 110–127. Vega Sosa, Constanza Cambridge Philological Soc., Cambridge. 1981 Comparaciones entre Los Teteles de Ocotitla, 1994 Frontiers of the Roman empire: A social and eco- Tlaxcala y Teotihuacan a través de materiales nomic study. Johns Hopkins Univ. Press, Balti- cerámicos. In Interacción cultural en México cen- more. tral, edited by Evelyn C. Rattray, Jaime Litvak Woolf, Greg King, and Clara Díaz Oyarza´bal, pp. 43–54. 1992 Imperialism, empire and the integration of the Universidad Nacional Autónoma de México, Roman economy. World Archaeology 23: Mexico City. 283–293. Von Winning, Hasso Zeitlin, Robert N. 1948 The Teotihuacan owl-and-weapon symbol and 1990 The Isthmus and Valley of Oaxaca: Questions its association with “Serpent Head X” at Kami- about Zapotec imperialism in formative period naljuyu. American Antiquity 14:129–132. Mesoamerica. American Antiquity 55:250–261. 1987 La iconografía de Teotihuacan: los dioses y los sig- 1993 Pacific coastal Laguna Zope: A regional center nos. Universidad Nacional Autónoma de Méx- in the terminal formative hinterlands of ico, Mexico City. Monte Alban. Ancient Mesoamerica 4:85–102. Wallerstein, Immanuel Zeitlin, Robert N., and Arthur A. Joyce 1974 The modern world-system: Capitalist agriculture 1999 The Zapotec-imperialism argument: Insights and the origins of the european world-economy in the from the Oaxaca coast. Current Anthropology sixteenth century. Academic Press, New York. 40:383–392.