State and Society at Teotihuacan Mexico (Cowgill, George)

Annu. Rev. Anthropol. 1997.
26:129–61
Copyright © 1997 by Annual Reviews Inc. All rights reserved
STATE AND SOCIETY AT

TEOTIHUACAN, MEXICO
Annu. Rev. Anthropol. 1997.26:129-161. Downloaded from arjournals.annualreviews.org
George L. Cowgill
Department of Anthropology, Box 872402, Arizona State University, Tempe, Arizona
by University of Montana on 09/12/07. For personal use only.
85287-2402; e-mail: [email protected]
KEY WORDS: archaeology, Mesoamerica, early states, governmentality, ideology
ABSTRACT
Between 100 BCE and 200 CE, the city of Teotihuacan grew rapidly, most of
the Basin of Mexico population was relocated in the city, immense civic-
religious structures were built, and symbolic and material evidence shows the
early importance of war. Rulers were probably able and powerful. Subse-
quently the city did not grow, and government may have become more collec-
tive, with significant constraints on rulers’ powers. A state religion centered
on war and fertility deities presumably served elite interests, but civic con-
sciousness may also have been encouraged. A female goddess was important
but probably not as pervasive as has been suggested. Political control probably
did not extend beyond central Mexico, except perhaps for some outposts, and
the scale and significance of commerce are unclear. Teotihuacan’s prestige,
however, spread widely in Mesoamerica, manifested especially in symbols of
sacred war, used for their own ends by local elites.
INTRODUCTION
Teotihuacan is an immense prehistoric city in the semi-arid highlands of cen-
tral Mexico. It rose in the first or second century BCE and lasted into the 600s
or 700s (Figure 1 outlines the ceramic chronology). Its early growth was rapid,
and by the 100s it covered about 20 km2 with a population estimated to be
around 60,000–80,000 (Cowgill 1979, p. 55; Millon 1992, p. 351). Subse-
quently, there was little change in area, and population grew more slowly, ap-
129
0084-6570/97/1015-0129$08.00
130 COWGILL
parently reaching a plateau of 100,000 or more by the 300s or earlier. No other

Mesoamerican city had such a large and dense urban concentration before Az-
tec Tenochtitlan, in the late 1400s. By the 200s Teo (as I will henceforth call it)
also had the largest integrated complex of monumental structures in Meso-
america: the gigantic Sun Pyramid (with a base area close to that of the largest
Egyptian pyramid), the Moon Pyramid, the 16-ha Ciudadela enclosure with its
Feathered Serpent Pyramid, and the broad 5-km-long Avenue of the Dead,
along whose northern 2 km these and many other pyramids, platforms, and
elite residences are arranged (Figure 2).
Millon et al (1973) published map sheets of the whole city at a scale of

1:2000, based on an intensive surface survey. A 1:40,000 version appears in
Millon (1973, 1974, 1976, 1981, 1988a). Articles by Millon (1981, 1988a,
1992) and Cowgill (1992a,b) include reviews of research and literature on Teo.
I concentrate here on publications since the mid-1960s that bear especially on

state and society. I emphasize relatively accessible sources and do not always
identify earliest publications of specific ideas. Edited volumes with papers on
a wide range of Teo topics include Berlo (1992a), Berrin (1988), Berrin &
Pasztory (1993), Cabrera Castro et al (1982a,b, 1991a), Cardós (1990), de la
Fuente (1995), Diehl & Berlo (1989), McClung de Tapia & Rattray (1987),
Rattray et al (1981), Sanders (1994-1996), Sociedad Mexicana de Antro-
pología (1967, 1972), and special sections of Ancient Mesoamerica [1991,
Vol. 2(1,2)] and Arqueología (1991). The proceedings of a 1993 Instituto Na-
cional de Antropología e Historia (INAH) workshop on Teo chronology, ed-
ited by Cabrera and Brambila, are in press, and a book by Pasztory (1997) has
just appeared. Sugiyama has created a web site: http://archaeology.la.asu.edu/
vm/mesoam/teo.
METHOD AND THEORY

We still only glimpse the outlines of polity and society in the city and the state
it dominated. Surviving inscriptions are few, brief, and hard to read. Teo soci-
ety was destroyed by the 700s or earlier, and to the Aztecs, about whom we
have a wealth of ethnohistoric data, Teo was a place of mysterious ruins; more
mythical than historical. These problems mean that theoretical preconceptions
and methodological assumptions play a large role in determining which inter-
pretations seem intrinsically plausible or even empirically well founded. Sand-
ers et al (1979) and Santley (1983, 1984, 1989) are strongly cultural-
materialist and favor interpretations and explanations in terms of environ-
mental and economic factors, relatively neglecting warfare and nearly exclud-
ing religion and other ideational aspects and the agency of individual actors.
Others give more weight to ideation and individual agency.
STATE AND SOCIETY AT TEOTIHUACAN 131
Figure 1 Teotihuacan ceramic chronology.
Teo is in a challenging twilight zone for direct historical approaches; close

enough to the 1500s to make it wasteful to neglect evidence from later socie-
ties, yet distant enough to make it unsound to project ethnohistoric data uncriti-
cally. Linguistic evidence suggests that Nahua speakers were absent or at least
not influential in the Basin of Mexico before the decline of Teo (Justeson et al
1983). The Aztecs and other Nahua in-migrants adopted much from earlier
central Mexican traditions, but the possibility of significant ethnic discontinu-
ity adds to the uncertainties of direct historical projections. Kubler (1967) went
132
COWGILL
Figure 2 Most of Teotihuacan (Sugiyama 1993). [After Millon (1973). Copyright by René Millon 1972.]
to a skeptical extreme; Pasztory (1992) has returned to this extreme and favors
a “semiotic” approach. López Austin et al (1991) and Coe (1981) are at the op-
posite pole. A more nuanced approach is preferable to either extreme. Using
knowledge from the 1500s to understand Teo is neither impossible nor easy,
and it is best to proceed piecemeal, case by case. Many Teo images have no ob-
vious later counterparts. Others do but must be used cautiously; meanings and
clusters of meanings may have shifted.
GROWTH OF THE CITY

It is notoriously difficult to derive accurate population estimates from archaeo-

logical data. Millon (1973) estimated the Xolalpan Phase population by using
sizes, layouts, and inferred uses of rooms in excavated apartment compounds

to infer that a 60 × 60-m compound would have housed about 60 to 100 people.
His surface survey indicated that over 2000 such compounds were occupied
during Xolalpan times. Making allowances for those larger or smaller than 60
× 60 m, he arrived at an estimate of 100,000 to 200,000 for the whole city, with
125,000 a reasonable middle value (Millon 1992, p. 344). Architectural data
for other phases are less clear, so Cowgill (1974, 1979) extrapolated the Xolal-
pan estimate by comparing quantities of phased sherds collected by the Map-
ping Project, with adjustments for estimated phase durations, assuming that
per capita sherd production remained approximately constant. He did not find
a Xolalpan peak. Instead, early rapid growth was followed by a long plateau.
By ca 1 BCE the city covered about 8 km2 and probably had a population of
20,000 to 40,000 (Cowgill 1979, p. 55). In the century before any known
monumental structures were built, Teo was already a city of exceptional size.
During the Tzacualli phase (ca 1–150 CE) increase continued to around
60,000–80,000, aided by movement into the city of most people in the Basin of
Mexico. After that, growth was much slower. Urban population may have
reached its maximum by the Miccaotli phase, ca 200 CE. Perhaps Teo had
reached a ceiling imposed by difficulties in provisioning a larger city with the
resources and means of transportation available. Most of the farming popula-
tion was concentrated in and near the city, and Teo seems to have underutilized
the southern Basin, including the lands most suited for chinampa cultivation.
It is also possible, if Storey’s (1985, 1992) estimates for one low-status
compound can be generalized, that very high infant and child mortality rates
set a limit to the city’s growth. In any case, Teo’s population seems to have
been fairly stable for several centuries. This suggests that whatever environ-
mental degradation may have occurred must have been gradual.
134 COWGILL
EXTENT OF TEOTIHUACAN RULE

Teo was the capital of an important state, but we know little about it. Teo is in
the northeastern part of the Basin of Mexico, about 45 km from modern Mex-
ico City. The Basin is about 80 × 50 km, ca 5000 km2, ringed on most sides by
volcanoes and high mountains, but more open to the north and northeast, so
Teo was well situated for movement in and out. The Basin is high (ca 2250 m)
but relatively flat and generally suitable for growing maize, beans, and other
food crops, though tracts suitable for canal irrigation are limited and localized,
a major one being a few thousand hectares just west of the city. Teo clearly
dominated the Basin politically, as shown by its drastic interference with the
settlement system.
Very likely Teo’s administrative control extended somewhat beyond the
Basin of Mexico, but perhaps not much beyond. It covered at least 25,000 km2
(Millon 1988a), a radius of about 90 km, and may have reached considerably
farther. Beyond that, Teo probably concentrated on controlling key settle-
ments and routes between them, rather than solid blocks of territory; “he-
gemonic” in Hassig’s (1992, pp. 57–59) terms. Teo’s immense prestige, how-
ever, surely exceeded its political sphere, and we still know little about specific
outposts. Hassig’s lucid account is a fascinating source of conjectures to be
tested, but it presents much as fact that is highly uncertain or sometimes
wrong. Studies such as Kurtz’s (1987) and Algaze’s (1993) also fit ambiguous
or problematic data into preconceived patterns.
Relations between Teo and Cholula, 90 km away, in the next major upland
plain to the southeast, are unclear (McCafferty 1996), though the weight of
evidence suggests it may have been independent. Cantona, further northeast,
on the way to the Gulf lowlands, may also have resisted Teo (García Cook
1994). Teo moved south to control the eastern Valley of Morelos where, unlike
the Basin of Mexico, cotton could be grown, a key resource for a textile indus-
try (Hirth 1978, 1980; Hirth & Angulo Villaseñor 1981).
Northwest, there is a Teo presence in the area around Tula, Hidalgo, notably
at the site of Chingú (Díaz Oyarzábal 1980). It is uncertain how far Teo influ-
ence went west or north of Tula. Aveni et al (1982) argued for Teo presence at
Alta Vista, in Zacatecas. Some features at Alta Vista have astronomical sig-
nificance, and its location on the Tropic of Cancer is probably intentional. We
should not assume, however, that local people were unlikely to make the
needed observations without tutelage from Teo “merchant-scientists-priests.”
Ceramic resemblances suggest only remote, indirect connections. A cross-in-
circle petroglyph motif is shared with Teo, but it is widespread in Mesoamer-
ica; its occurrence need not mean Teo presence. Teo may have received miner-
als from this area, but its impact on local societies is unclear.
In Oaxaca, the Zapotec state was independent and maintained diplomatic

relations with Teo (Marcus & Flannery 1996). Some sort of Teo presence is
known for Matacapan in southern Veracruz (Santley 1989), Mirador in Chia-
pas (Agrinier 1975), Kaminaljuyú, and other sites in highland and Pacific
coastal Guatemala (Berlo 1983, 1984, Demarest & Foias 1993, Kidder et al
1946, Sanders & Michels 1977). Santley’s (1989; Santley et al 1987) exagger-
ated claims about a Teo outpost at Matacapan have created much confusion.
Most of the Matacapan ceramics are strikingly different (Arnold et al 1993),
and the Teo impact seems weaker than at Kaminaljuyú. Matacapan cylinder
tripod vases, for example, show generic resemblances to Teo forms, but very
few, if any, are specific Teo subtypes. Small, relatively crude twin-chambered
incense burners (candeleros) are a stronger point of similarity. A relief from
Soyoltepec in a style closely similar to Teo shows a figure with flaming
torches and a rattlesnake headdress, suggesting military action in the lowlands

(Sugiyama 1989b, von Winning 1987).
A Teo connection is manifest at Altun Ha in Belize by the early 200s
(Pendergast 1990), but most Teo influences in the Maya lowlands do not seem
earlier than the late 300s, which suggests that the spread of Teo prestige oc-
curred considerably later than the rapid rise of the city. They are especially
strong at Tikal (Kowalski 1997, Laporte & Fialko 1990) but are unlikely to
represent control by Teo. Many reflect adoption of a limited number of Teo-
related symbols by local elites for their own purposes (Stone 1989, Demarest
& Foias 1993).
WRITING AND LITERACY

Teo had nothing like the writing systems of the contemporary Lowland Maya
or earlier neighbors in Veracruz, though some Teotihuacanos must have been
aware of these systems. There is ample evidence, however, of standardized
signs and a notational system comparable to those of the Aztecs, though few
specific signs are shared (Berlo 1989). Langley (1986) provided an indispen-
sable study and catalog of signs, whereas his later works deal with specific
clusters and compounds (Langley 1991, 1992). Cowgill (1992c) identified a
sign cluster (“red bone-flower”) semantically equivalent to a term used by Az-
tecs. However, no examples of phoneticism have been identified, nor any
grammatical elements, so, even when meanings can be inferred, the signs have
not helped to identify the dominant language or languages of the city.
A remarkable find in recent work in the La Ventilla district directed by
Rubén Cabrera, not far southwest of the Great Compound, consists of over 30
signs and sign clusters painted on the floor of a patio (de la Fuente 1995). They
stand alone, unassociated with representational scenes. They were made
136 COWGILL
quickly and show a control of line that bespeaks an experienced hand. Their
meaning and purpose are obscure, but there can no longer be any doubt that
Teotihuacanos had a notational system adequate for the information-handling
needs of their society. What remains noteworthy is the sparing use of this sys-
tem in sculpture, mural painting, and decorated ceramics; one aspect of the
near lack of public celebration of named specific individuals.
Pasztory (1992) suggested that Teo was almost secretive and made a point of
being different from other Mesoamerican societies. It is easy for us, in our
frustration, to feel they were deliberately being difficult, but I suspect most
meanings of their scenes and signs were intended to be clear to the average
Teotihuacano, and we have trouble only because we still lack many keys. They
seem, however, not to have been very interested in exotic ideas. Some fine for-
eign ceramics were imported, but most exotic goods were raw materials in-
tended for working by Teo artisans. The contrast with the Aztec interest in fin-
ished products from afar, seen in the Templo Mayor offerings (Matos Mocte-
zuma 1988), is striking. Persons in other societies adopted Teo symbols for
their own ends, but there was little flow in the opposite direction; Teotihuaca-
nos seem to have been satisfied with their local style and symbols (Pasztory
1990, p. 187). The great value placed on the exotic in many societies (Helms
1993) is not evident at Teo.
One exception is adoption of interlocking scroll motifs from the lowlands
of central Veracruz (Stark 1995). The earliest cylinder tripod vases were
probably imports from this region (Bennyhoff 1967, p. 26). Teotihuacanos,
however, may have recognized an affinity with Gulf lowland neighbors, and
the adoptions occurred during Miccaotli and/or Early Tlamimilolpa, when Teo
may have been more receptive to new ideas than it was later.
PERSONALITY AND SOCIALIZATION

Approaches that take serious account of individual agency imply that we
should also take account of individual personality. Archaeologists, however,
have done little along this line, and there is almost no explicit discussion of this
topic for Teo. Foucault’s notion of governmentality looks useful but has not
yet been applied. Pasztory (e.g. 1992) characterizes the art as remote and im-
personal, and Cowgill (1993, pp. 564–68) touches on the topic of personality.
No scenes glorify specific individuals, and human beings are shown subordi-
nate only to deities, not to other human beings. This has implications about the
political system, or about how the system was represented, but it also suggests
something about socialization of children and about preferred character traits.
Beginning as early as the repeated images on the Feathered Serpent Pyra-
mid (FSP), identical figures are repeated in numerous copies. Some scenes (es-
pecially of fierce animals) convey tension and vibrant power, as in the

“Mythological Animals” mural (de la Fuente 1995, pp. 93–101), and else-
where (e.g. Berrin 1988, p. 187), but much of the art is stiff, and human faces
look expressionless to modern viewers. Sometimes small human figures are
shown simply clad, in free and playful poses, as in a scene from Tepantitla that
reminds us how small a fraction of Teotihuacano life was depicted in Teo art.
Most scenes show human beings so loaded with clothing and insignia that
faces and other body parts are barely visible. Emphasis is on acts rather than
actors; on offices rather than office-holders. This, together with the multiplic-
ity of identical scenes, suggests an ethos in which individuals were inter-

changeable and replaceable. These properties are found not only in sculpture
and paintings that must have been elite sponsored, but also in ubiquitous ob-
jects such as composite censers and clay figurines. Whether or not impersonal-
ity and multiplicity were deliberately encouraged by state policy, they are
themes that pervaded all classes and social sectors. No evidence of resistance
or dissent has been recognized so far.
NONELITE ELEMENTS OF THE SOCIETY

Households and Apartment Compounds
We know little of Teo housing during the early centuries when the great tem-
ples were being built. In the 200s and 300s, more than 2000 “apartment com-
pounds” were built to house nearly the whole city, of all socioeconomic sta-
tuses. Millon uses “apartment” because each building contains several distinct
suites of rooms, indicating occupation by multiple domestic units, and “com-
pound” rather than “complex” because they are bounded by thick outer walls
with few entrances and are separated from one another, often by narrow
streets. Contrary to widespread belief, in many districts compound sizes are
not highly standardized, and they vary widely around an average of roughly 60
× 60 m. Internal layouts are diverse, although the core of most apartments is a
patio surrounded by rooms and platforms. Yet the facts that compounds are so
substantially built (of rubble walls faced with thick concrete covered with lime
plaster) and approximate the canonical Teo orientation of 15.5° east of true
north, even in outlying areas where they are widely spaced, suggest some sort
of state interest. Possibly the state aided in their construction. Very likely oc-
cupants of a compound formed an important sociopolitical unit, composed of
several households but smaller than a neighborhood (Millon 1976). Societies
with most of the population organized in units of this size are not common.
Construction quality and size of rooms vary considerably, between com-
pounds and within single compounds. Some have spacious rooms and abun-
dant mural paintings [e.g. Zacuala “Palace” (Séjourné 1959)]; others are far
138 COWGILL
more modest [e.g. Tlajinga 33 (Storey 1985, 1991, 1992; Widmer 1991; Wid-
mer & Storey 1993]. When less was known about the compounds it seemed
reasonable to call some of them “palaces,” but over-broad application of the
term has been misleading. Millon (1976) suggests at least six socioeconomic
levels, with uppermost elite residences in the Ciudadela and elsewhere near the
Avenue of the Dead. He would put Zacuala “Palace” in about the third level
from the top. Sempowski (1994) has tabulated information on offerings in Teo
burials, mostly in apartment compounds, and her analyses suggest status dif-
ferences and changes over time in these differences.1 Spence’s (1974, 1994)
studies of nonmetric skeletal traits suggest a preference for patrilocal post-

marital residence in most compounds.
It is unclear whether there was any distinct material gap between the elite
and the merely prosperous, and proportions of residences of varying quality
are also unclear. Further analyses of Mapping Project surface collections may
clarify these questions (Robertson 1997). Even the proportion of compounds
with mural paintings is debatable. Fine murals were not common, and many
compounds had only white-plastered walls, with at most a few borders out-
lined in red. Most floors were plaster over concrete; some were of cobbles or
earth.
We need excavations of many more compounds using the best methods and
concepts of household archaeology. Good examples already exist, in a residen-
tial area in the Oztoyahualco district (Manzanilla 1993, 1996; Manzanilla &
Barba 1990), in the Oaxaca enclave (Spence 1989, 1992), in the “Merchants’
Barrio” (Rattray 1989, 1990), and in the Tlajinga 33 ceramic and lapidary resi-
dential workshop. Manzanilla and her colleagues have used chemical analyses
of residues on plaster floors to infer highly localized activities within rooms.
Barrios, Enclaves, and Districts

Occupants of the city must have recognized distinct neighborhoods. Millon
(1976, p. 225) said that many spatial clusters of apartment compounds can be
identified. Some, such as the block that includes the Tepantitla compound
(Figure 2, NE part of square N4E2), look clear. Craft workshops often form
spatial clusters. Unambiguous distinct small neighborhoods, however, seem
hard to define in much of the city. Most freestanding pyramids are not plausi-
ble barrio temples, and they are absent in large tracts of the city. The excavated
plan of one compound, Yayahuala, suggests that it may have housed a barrio
headman (Millon 1976) but other examples would be hard to recognize with-
out excavation. J Altschul (personal communication) suggests that the impor-
1
1Rattray (1992) also provided data on burials and offerings. See Millon’s corrections in Sempowski
& Spence (1994).
tance of apartment compounds as multihousehold social and political units

may have meant that small neighborhoods were not very important administra-
tive units.
There are at least two enclaves with foreign affiliations. Toward the western
edge of the city, centered in square N1W6, is a cluster of about a dozen com-
pounds with Oaxacan affinities (Rattray 1993; Spence 1989, 1992). Architec-
ture and most ceramics are typically Teo. A small percentage of the ceramics,
however, mostly locally made, are Late Monte Albán II (formerly called
II/IIIA transition) in style, and a few vessels are imports. Stratigraphic evi-
dence suggests that this early style continued to be used from Tlamimilolpa
through Metepec times, which implies either remarkable conservatism
(Spence 1992) or some unresolved problem in ceramic chronology. A further
Oaxacan tie is collective tombs, quite different from the Teo pattern of individ-
ual inhumations beneath floors. Socioeconomic status of the enclave occu-

pants looks no more than average. Their role in Teo society is unclear, but
similar ceramics in a lime-producing district near Tula (Crespo & Mastache
1981) suggest they may have been masons.
Another enclave with foreign ties is the “Merchants’ Barrio” on the eastern
edge of the city. Most ceramics are Teo, but some are imports from the Gulf
lowlands, and a smaller number are from the Maya lowlands. Some structures
are Teo-like, but others are circular, a form associated with the lowlands (Rat-
tray 1989, 1990). Probably the enclave specialized in lowlands imports, in-
cluding perhaps cotton and other perishable materials.
It is not clear whether Teo was as ethnically diverse as is often suggested.
The early influx from within the Basin would have brought in people with dif-
ferent local affinities, but they may not have differed much in language or cul-
ture. Later, foreigners seem to have been handled by spatial segregation, to
judge from the enclaves described above. Even without ethnic frictions, fac-
tions would have posed sociopolitical management problems.
Aztec Tenochtitlan was divided into four quarters that were important so-
ciopolitical units. Teo is divided into quarters by the Avenue of the Dead and
“East” and “West” Avenues, whose axes pass through the Great Compound
and the Ciudadela. Sugiyama (1993, p. 110) questions the existence of the east
and west avenues, but Millon’s survey found ample evidence of them. They
differ sharply, however, from the northern part of the Avenue of the Dead in
not being lined by pyramids, platforms, or other obviously special structures.
Teo has no long and architecturally prominent east-west alignments compara-
ble to the north-south Avenue of the Dead. Division of the city is more bipartite
than quadripartite.
There is evidence for socially meaningful districts larger than barrios and
enclaves, but smaller than whole quarters (Altschul 1987, Cowgill et al 1984).
140 COWGILL
Freestanding walls bound a number of precincts, especially near the Avenue of

the Dead. They are 1–2 m thick at their bases, but nearly all have been eroded
to ground level. A Mapping Project excavation (TE5) found that the top of one
wall joined the northwest corner of the Moon Pyramid at a height of about 5 m.
Other walls are much farther from the ceremonial center. Starting at the “Plaza
One” three-pyramid complex in square N5W2 (Figure 2), a wall can be traced
eastward for more than 1 km and southward for nearly 1 km. Traces of other
walls several 100 m long are at the east edge of N5W4 and near the south edge
of N6W3. There is no suggestion that the city was enclosed by walls, and walls
may or may not have been militarily significant. They, however, as well as wa-
tercourses, would have strongly affected movement within the city.
Some abrupt changes in density of apartment compounds coincide with walls
outside the central ceremonial district, but changes in socioeconomic status
indicators have not been obvious. A possible exception is the so-called “Old
City,” centered in square N6W3. Millon (1973) recognized during the 1960s
survey that structures in this area were different from most apartment com-
pounds, and excavations by Manzanilla (1993) bear this out. Millon suggested
that the apartment compound innovation and concomitant social changes did
not spread to this area, which preserved an earlier style of housing. I am not
sure whether it was earlier or just different, but it is unlike most of the city. It is
outside the outermost known walls, population density was high, and it has
two of the largest three-pyramid complexes outside the city center. Some arti-
fact categories, such as composite censers, occur in unusually high propor-
tions, but no categories unique to this district have been recognized. A search
for distinctions at the microtradition level might be rewarding.
The Tlajinga district, in S3W2, S3W1, S4W2, and S4W1, is near the south-
ern extreme of the city, separated from the rest by the Río San Lorenzo. The
Mapping Project survey found exceptional proportions of San Martín Orange,
a utility ware, on a number of sites, and excavations at one of these, Tlajinga
33, have confirmed its manufacture there (Sheehy 1992, Storey 1991, Widmer
& Storey 1993). This was probably a district of low-status artisans. Special-
ized ceramic production was plausibly situated in this remote area to shield
higher-status Teotihuacanos from the smoke of pottery firing.
Lineages?
Even modest apartment compounds or room complexes, such as Tlajinga 33
and Tlamimilolpa (Linné 1942) tend to have one or a few relatively richly
stocked graves, which may have been those of founders (e.g. Millon 1976,
Headrick 1996). Millon points out that the inflexible sizes of compounds with
fixed outer walls are ill-suited to the inevitable fluctuations of strictly unilineal
descent groups, and Spence (1974) cites one case in which shared nonmetric
skeletal traits suggest a group of related women who stayed home while their
husbands moved in from elsewhere. Nevertheless, consistent with most of
Spence’s findings, each apartment compound probably was associated with a
core of individuals claiming descent from a common ancestor, plus others
whose rights to residence were based on marriage, some more tenuous kin tie,
or a wide variety of special circumstances, perhaps including servants and ap-
prentices.
Headrick (1996) proposed that descent groups at the apartment compound
level may have been hierarchically organized into much larger groups, whose
heads would have been of elite status. Different apartment compound groups
may have been roughly ranked according to the degree to which they could
claim connection with the apical ancestor through senior links. Many such sys-
tems exist elsewhere in the world. If such higher-level units existed, many of
their head families may have resided on or near the Avenue of the Dead, while
member lineages may have been dispersed in various districts, rather than be-
ing spatially concentrated. A few three-pyramid complexes in the northwest
part of the city, not associated with the Avenue, could be headquarters for
more independent lineages, although other interpretations are possible (Cow-
gill et al 1984).
Household Religion and Ritual
Pasztory (1992) argued that a good deal of village-level religion persisted in
urban Teo. Assuredly, much was distinct from the “state” religion, but it dealt
with domestic and familial concerns that would have persisted no matter how
large and complex the state became. Many such rituals may have been of no in-
terest to the state. Others, however, may reflect the “long arm” of the state im-
posing itself at the household level. I know of nothing that suggests conscious
resistance to the state.
Standardized stone bowls supported on the back of a thin and bent old man,
ubiquitous in apartment compounds of all socioeconomic statuses, probably
belong to a cult of the hearth. They are called “Huehueteotls,” but use of this
Aztec term is problematic; an Aztec revival of the form (López Luján 1989,
Umberger 1987) reinterprets it and shows that Aztecs did not recognize its Teo
meaning. Except that the state had an interest in promoting domestic tranquil-
lity, it is unlikely that these stone carvings had much political significance.
Composite censers are also ubiquitous. These are built from coarse
flowerpot-like bowls, often on a high pedestal base, with a similar inverted
bowl as a lid, from which rises a tubular chimney. Panels, frames, and a profu-
sion of appliqué ornaments largely conceal the chimney. Often the central ele-
ment is a human face. From Late Tlamimilolpa onward, faces and other orna-
ments are moldmade. Some censers are associated with burials, but many are
142 COWGILL
not. Nevertheless, they were probably used in rituals commemorating the

dead. Headrick (1996) and others have argued that Teo stone masks were at-
tached to mortuary bundles containing remains of deceased elites. Composite
censers were probably an equivalent for commemorating honored but less il-
lustrious compound residents. As such, it might seem that the state would have
little interest in them. One of the clearest instances of a state-related workshop
at Teo, however, is in the large enclosure attached to the north side of the Ciu-
dadela, where great numbers of censer ornaments and molds for their manu-
facture were found (Múnera 1985). Múnera and Sugiyama have prepared an
unpublished catalog of these finds. The nature of the state interest in composite
censers will become clearer through further studies of the multitude of stan-
dard signs on their ornaments.
Another indication that beliefs and practices associated with composite
censers may have been connected with the Teo state is that they disappear with
the collapse of that state. Dominant post-Teo censer forms are ladles, much
more portable and adapted for quite different activities. The shift, however,
may simply reflect ethnic discontinuity. This may also be the case with twin-
chambered “candeleros,” small, simple, and often crude incense burners that oc-
cur in great numbers at Teo but that do not survive the city. They are ubiquitous
but are less common in the city center. Probably they were for modest house-
hold or individual rituals. Twin-chambered Teo-like varieties occur rarely in
Maya sites and elsewhere in Mesoamerica, but are not scarce at Matacapan.2
Pasztory (1992) also linked “Tlaloc” jars with popular religion. Some occur
in sites of no obvious prominence, but they are more abundant in high-status
contexts, such as the FSP. Most are well-polished, elaborately hand-modeled,
and represent (by Teo standards) a high level of skill and manufacturing effort.
In murals, the Storm God often carries similar jars, and they are probably asso-
ciated with state religion.
Infants’ burials are often associated with patio altars. It is not clear whether
any were sacrificial victims; high infant mortality probably accounts for the
number observed. In recent Tlaxcala they are seen as especially effective inter-
mediaries between human beings and the supernatural because they have spent
so little time in this world (Headrick 1996, Nutini 1988). This analogy makes
great sense of the Teo data.
Designs pecked on rocks or impressed in plaster floors are common at Teo.
These include rectangles, Maltese crosses, and other forms, but many consist
of a cross and two concentric circles. This motif is widespread, from north of
Alta Vista (Zacatecas) to Uaxactun. Among the first found at Teo were a pair,
2
2F Bove (personal communication) finds them in Pacific coast Guatemalan sites that also have
composite censers.
several kilometers apart, that formed a perpendicular to the Avenue of the

Dead, suggesting that they were important in astronomical alignments and city
planning. So many more have since been found at Teo, often within buildings
where long-distance sightings would have been unfeasible, that their use for
astronomy and surveying has become uncertain. Some consist of approxi-
mately 260 dots, and the case is better for a connection with the 260-day sacred
calendar, perhaps divination, and possibly gaming. Many outside the city are
on prominences suitable for distant sightings, but within the city many are in
seemingly unremarkable buildings. This, plus their abundance and simplicity,
makes state involvement somewhat unlikely. Pecked cross-&-circles and

other motifs are especially profuse in a recently uncovered floor near the Sun
Plaza, but even here they could reflect a popular cult.
The term “temple” has been used at Teo to refer to freestanding pyramids
and also to the platforms that are ubiquitous in apartment compounds. The
former must represent public religion, whether at the state or at some interme-
diate level. The latter are standard Teo “talud-tablero” platforms, which usu-
ally support a room fronted by a portico. Typically there are three such plat-
forms, on the north, east, and south sides of a patio to which they connect by
short stairways and in whose center there is often an altar; the eastern platform
is most prominent. Millon (1976) suggested that a patio-platform group in the
Yayahuala compound, because of its size and accessibility, may have served as
a barrio temple, and the compound may have housed a barrio headman. Many
of these patio-platform groups, however, were more likely used only by com-
pound occupants, and often there is more than one such group in a single com-
pound. They are architecturally substantial, and they were probably used for
some mundane activities as well as for ceremonial occasions.
CRAFT PRODUCTION AND TRADE

We have learned much about Teo technology, but scale and organization of
craft production are still poorly understood. Depopulation of the countryside
implies that farming was a major activity of many households. Earlier esti-
mates on the order of 400 obsidian workshops have been revised downward.
Clark (1986) thinks the scale of production and exchange was much smaller.
He overstates his case and perhaps underestimates the immense quantities of
obsidian debitage at Teo. Assessments of the obsidian industry by Spence
(1981, 1984, 1987, 1996) are more reasonable. The Mapping Project made test
pits in or near obsidian and lapidary workshops, but more extensive excava-
tions of obsidian and other workshops are needed.
Obsidian from the nearby Otumba source is of moderate quality and was
used mainly for local consumption. Teo controlled the Pachuca source of supe-
144 COWGILL
rior green obsidian. Much was consumed in the city, and it is found in small
amounts widely in Mesoamerica. There is good evidence, however, that Teo
did not monopolize obsidian production and exchange (Drennan et al 1990).
Sources unlikely to have been controlled by Teo continued to export obsidian
(Stark et al 1992). Teo emissaries could be backed by the city’s prestige and
could carry fine stucco-decorated ceramics and perhaps perishable manufac-
tures, but green obsidian was the only locally obtained valuable raw material
they could offer abroad, which may account for its wide distribution. Com-
merce in obsidian and other materials may have been fairly important, al-
though we have recognized no evidence of a distinct merchant class [Manza-

nilla (1992) doubts that one existed], but the scale of trade postulated by Sant-
ley (1983, 1984; Santley et al 1986) is not supported by evidence. Emphasis on
trade alone underplays military and ideational bases for Teo’s wide influence.
Sheehy (1992) and Hopkins (1995) have studied Teo ceramic production
techniques, and the late Paula Krotser began a review of Mapping Project evi-
dence for production sites, but we still lack a comprehensive picture of the or-
ganization and spatial distribution of pottery making. San Martín Orange util-
ity ware was a specialization in the Tlajinga district. Other utility wares, such
as burnished ollas and cazuelas, may have been made on a smaller scale in less
specialized households. Significant state involvement seems unlikely, except
for the mold-made censer parts noted above.
Turner (1987, 1992) reported on a barrio of lapidary craftsmen on the east-
ern outskirts of the city, and other evidence suggests some lapidary work in
fine stone and marine shell under state sponsorship. Lapidary work at Tlajinga
33 (Widmer 1991) was probably not state directed. An obsidian concentration
in a walled precinct just west of the Moon Pyramid (Spence 1981, 1984, 1987)
implies at least part-time work sponsored by temple or state. This, as well as
the censer ornament workshop in the enclosure attached to the Ciudadela, indi-
cates that some craftsmen worked outside household contexts at least part of
the time. Some may have been attached specialists, but most or all may have
been providing periodic labor services. A great deal of production seems
household based, however, possibly taxed and regulated to some degree by the
state but not state sponsored.
WAR AND THE MILITARY

By the 1960s it was clear that Teo was not a very peaceable society (C Millon
1973, R Millon 1976). Recognized military symbolism, however, was mostly
late, and it seemed that emphasis on military elements increased over time.
New finds and reinterpretations of old data now show that military emphasis
began early. About 200 persons were sacrificed as part of the FSP construction
activities, ca 200 CE (Cabrera Castro et al 1991b; Sugiyama 1989a,b, 1992,

1995; Arqueología 1991, Vol. 6). Many, but not all, were in military garb and
accompanied by weapons. Two large pits, underneath the pyramid and at the
foot of its stairway, may have contained bodies of Teo rulers, but they were
looted anciently, so it is not clear whether the victims accompanied a dead
ruler. In any case, victims and grave goods were arranged in highly structured
patterns, which Sugiyama (1995) argued were related to the calendar and crea-
tion symbolism. The victims may have been enemies or low-status Teotihua-
canos dressed as soldiers and dignitaries, but I suspect they belonged to the
royal household and that the soldiers were elite guardsmen. Anatomical,
chemical, and cultural studies of bones, teeth, and grave goods, now under
way, may resolve these issues.
Many now see the symbolism of the FSP facade as associated with war
(Carlson 1991; López Austin et al 1991; Sugiyama 1989b, 1995; Taube 1992).
For most, the Feathered Serpent itself reflects sacred Venus-related war;
Taube makes the connection by interpreting the figure that alternates with the
Feathered Serpent as a solar fire/war serpent.
“Portrait” clay figurines are abundant at Teo. Their heads, stamped in
molds, are anything but portraits of individuals. Their contorted body posi-
tions have been puzzling (Figure 3). W Barbour (in Berrin & Pasztory 1993, p.
228) suggested convincingly that they are poised to hurl a spear (of perishable
material) in the right hand and held a shield in the left hand. They probably
wore perishable clothing. These figurines also point to the salience of war in
Teo thought.
It is unlikely that Teo could have gained preeminence—however aided by
its sacred significance, location on a strategic trade route, and proximity to
canal-irrigated fields—unless it had been able to overcome armed resistance
from rival centers. Defensible locations of sites of the poorly understood Te-
zoyuca phase (Sanders et al 1979) suggest warring polities in the Basin of
Mexico just before or early in Teo’s rise. Pasztory (1990, 1993, p. 138) thinks
Teo emphasis on war was mainly symbolic, but it was probably very real, at
least initially. Hassig (1992) argued convincingly that Teo armies were highly
effective not only because of their atlatl-propelled darts and other weapons but
because they fought in disciplined masses, using many commoners as well as
elites. This is consonant with everything else we know of Teo. It is less clear
that organization was what Hassig (1992) calls “meritocratic,” i.e. that com-
moners were motivated by the chance for upward mobility if they performed
well; this view underestimates the power of ideology.
Berlo (1983, 1984) emphasizes military symbolism in Teo-derived com-
posite censers in Pacific coastal Guatemala, and war is prominent in Teo sym-
bolism adopted by the lowland Maya (Schele & Freidel 1990, Stone 1989).
146 COWGILL
Figure 3 A “portrait” figurine (after Pasztory 1997).
Conquest of the lowland Maya is unlikely, but the city’s military prestige trav-
eled well. Teo weapons and possibly some tactics seem to have been adopted
by the Maya, at least for a while. But military successes would have been at-
tributed at least as much to potency of the Teo War God as to weapons and tac-
tics, and this would have been a powerful incentive for adopting elements of
Teo religion. Teo’s military prestige may have lasted long after its real military
effectiveness waned.
Teo soldiers were associated with fierce animals, especially rattlesnakes,
jaguars, coyotes, and raptorial birds. There were probably military orders
something like the Aztec eagle and jaguar knights, as suggested by C Millon
(1973, 1988) and argued especially by Headrick (1996). These may have been
sodalities that crosscut kin ties and provided politically important cohesive in-
stitutions. The earliest prominent fierce animal is the Feathered Serpent, de-
picted as a rattlesnake, though with avian feathers and a feline snout. Many
nose pendants of the “butterfly” type (Figure 4b) were in the burial at the cen-
ter of the FSP. Oralia Cabrera Cortés (1995) recognized that they are not but-
terflies (prominent in other Teo war symbolism), but final segments of rattle-
snake rattles. Wearers of this type of nose pendant were identified with the
Feathered Rattlesnake. Whether there was such a sodality is unclear.
Figure 4 Teo nose pendants: (a) “Tlaloc” type from FSP grave 13; (b) a so-called “butterfly”
nose pendant. Actually it represents the final element of a rattlesnake tail (from Cabrera Cortés
1995).
Canid jaws were worn by a few sacrificed soldiers at the FSP, and a few ea-
gle and felid bones were found. Symbolic importance of these animals seems
to have increased over time, since they are shown more centrally and more en-
gaged in activities in later murals, where serpents occur mainly in bordering
frames. Serpents seem to have a (literally) overarching importance, whereas
human beings and symbolic animals carry out the actions. In the West Plaza
Group of the Avenue of the Dead Complex, an earlier balustrade of the central
pyramid stairway has projecting monumental heads that are serpent-like, re-
placed in a later stage by more feline heads (Morelos García 1993). Many war
148 COWGILL
birds have been identified as owls, but most may be eagles. There may have
been an early period, until the mid-200s, dominated by the Feathered Rattle-
snake, followed by growing emphasis on coyote, jaguar, and owl/eagle orders.
This may reflect the rise of military sodalities that limited the power of the
ruler.
C Millon (1973, 1988) has recognized a distinctive tasseled headdress as a
symbol of high war-related office, probably approximately what we would call
“general.” It may represent a level above the postulated sodalities.
STATE/PUBLIC RELIGION
State interests were probably represented by a few major deities. One was the
Feathered Serpent, discussed above. Like all Teo deities, it had multiple as-
pects, and besides its military associations it often occurs associated with
vegetation and the fruitful earth. Another reptilian being was prominent and is
represented by the head-like object that pairs with the Feathered Serpent in the
FSP facades. This is often called “Tlaloc,” but it shows few traits of that god.
Sugiyama (1993, p. 116) and Taube (1992) argued that it is a headdress.
Drucker (1974), Sugiyama (1989b), and López Austin et al (1991) linked it to
Aztec Cipactli, the Primordial Crocodile and the beginning of calendrical
time, whereas Taube linked it with the Xiuhcoatl solar fire serpent.
The Storm/War God
This deity is identifiable by his fangs, distinctive upper lip, receding or absent
lower jaw, and goggles around the eyes. Other attributes, such as aquatic vege-
tation in the mouth, a distinctive headdress, and a lightning bolt in the hand, are
more variable and emphasize different aspects. Pasztory’s (1974) distinction
between “Tlaloc A” and “Tlaloc B” no longer seems clearcut, but she rightly
pointed out a range of contexts and meanings for this god. He is associated
with beneficent rain and fertility, but also with lightning, thunderstorms, and
the crop-devastating hail that often accompanies them. Sometimes weapons
associate him with warfare. The state would have had a profound interest in
maintaining good relations with this god in all his aspects. He may differ in de-
tails but is broadly similar to Aztec Tlaloc and to other Mesoamerican deities
such as Zapotec Cociyo and Maya Chac.
Death and Underworld Gods
Several large skull carvings come from within or near the Sun Plaza (Berrin &
Pasztory 1993, p. 168; Millon 1973). Possibly these and jaguar sculptures from
the Sun Plaza pertain to death, the underworld, and the night sun, and they may
be related to the cave under the Sun Pyramid. The pyramid may have been as-
sociated with a sun god, in day and night aspects.
The “Great Goddess” and Rulership

Pasztory (1977) identified certain images, characterized by a nose pendant and
a bird in the headdress, as a goddess. Others tend to agree (C Millon 1988;
Berlo 1983, 1992b; Taube 1983; von Winning 1987). Pasztory (1992, p. 281)
says there is a near consensus that there was a single Great Goddess with sev-
eral aspects, including a military persona, whose image became progressively
more important from about 200 CE, who is shown superior to the Storm God,
and who was apparently the major deity of Teo.
Among the multiplicity of Teo images of deities it has not been possible to
decide how many distinct individuals there are, and how many iconographic
complexes represent aspects of a single deity. Teotihuacanos may not have felt
a need to settle this question. Gender identification is also a problem. Most Teo
figures are too heavily clad for biological sex to be inferred from physical fea-
tures, so usually we must rely on costume to infer socially constructed gender.
One deity with female dress is the “Diosa de Agua” found near the Moon Pyra-
mid (Pasztory 1992), which is 3.9 m high and weighs 22 tons. Some much
Figure 5 Frieze from the West Plaza Group of the Avenue of the Dead Complex. It has been
identified as the Great Goddess by some scholars, but it probably represents rulership (from Berlo
1992a:282).
150 COWGILL
smaller stone figures wearing female dress may also be deities (Berlo 1992b,
p. 138, Figure 11, p. 144, Figure 18; Pasztory 1992, p. 309, Figure 23). None of
these figures in female dress has nose pendants or other supposed Great God-
dess diagnostics. The unfinished and somewhat damaged Colossus of Coatlin-
chan (moved in the 1960s to the front of the Museo Nacional), which weighs
180 tons and is over 7 m tall (Berlo 1992b, p. 138, Figure 10) is said to wear fe-
male clothing, but I find its dress ambiguous.
A female goddess with multiple aspects was certainly important in the state
religion, but I am not convinced that she was as important or pervasive as Pasz-
tory and others argue. The difficulty is that attributes that may be only diacriti-
cal are treated as diagnostic. For example, a goddess is shown in a mural at
Tepantitla, wearing a distinctive nose pendant consisting of a bar in which
there are three circles, and from which fangs depend (Berlo 1992b, p. 130, Fig-
ure 1; Langley 1986, p. 277, No. 153, type E nose pendant). Rather than treat-
ing the pendant as a diacritical element that emphasizes some aspect of the
goddess, it is treated as a diagnostic that marks any other figure that bears it as
a manifestation of the Great Goddess. This, as well as a headdress with birds, is
what led Pasztory (1992) to identify a frieze from the West Plaza Group in the
Avenue of the Dead Complex as probably a depiction of the Goddess (Figure
5). If one does not take the nose pendant or the birds in the headdress as diag-
nostic of the Goddess, however, nothing else in the figure proves it to be fe-
male. Further doubts are raised by the discovery of very similar nose pendants
of green stone in Burial 13 of the FSP (Figure 4a), one of them associated with
an unusually robust male.
The figure in this frieze holds a torch in each hand, from which flames and
smoke emerge, together with budding plant stalks. The torches are wrapped
rods with “year-sign” variants on their fronts. Torches were symbols of ruler-
ship in Preclassic Mesoamerica (Grove 1987). The frieze probably symbolizes
rulership rather than the Great Goddess (Cowgill 1992a,b). Linda Schele (per-
sonal communication) has independently reached a similar conclusion.
One could read the frieze as an example of the Great Goddess’s identifica-
tion with rulership. Pasztory (1992) (in one of the few explicit considerations
of gender ideology at Teo) suggested that a female was chosen as the supreme
deity because a female could be seen as benevolent, maternal, and impartially
transcending factions associated with male heroes. This makes assumptions
about Teo social construction of gender that are plausible but need further test-
ing. I think any connection between this frieze and the Great Goddess is ques-
tionable.
This is not to say that the frieze represents a specific individual. Morelos
García (1993, Figure F.2) illustrated additional fragments of two more figures
from the same context, apparently identical to the relatively complete one. It
seems that the idea or office of rulership, rather than any specific rulers, is rep-
resented. It is interesting that this frieze comes from the West Plaza Group of
the Avenue of the Dead Complex, which I think may have been a setting for
top-level government activities (Cowgill 1983).
The colossal figure in female dress that was found near the Moon Pyramid
might be connected with the moon. This would be consonant with the idea, de-
veloped most fully by Sugiyama (1993), that the Sun and Moon Pyramids and
the Venus-related Ciudadela represent a triad of astral deities, a concept wide-
spread in Mesoamerica.
THE NATURE OF TEOTIHUACAN RULERSHIP

Palaces
The Ciudadela apartment compounds flanking the FSP have been interpreted
as residences of the heads of the Teo state (Armillas 1964, Coe 1981, Cowgill
1983, Millon 1973, Taube 1992). Sugiyama (1993, pp. 110, 123) appears to be
skeptical, perhaps because he believes the Ciudadela was associated with the
underworld, while administrative centers for the “present” world would more
likely have been near the Sun and Moon and/or along the Avenue of the Dead.
The Ciudadela and the Great Compound, however, are very different from any
other Teo complexes, which suggests that they served unique purposes. Never-
theless, the Ciudadela is unlike many better-known royal palaces, such as
those in Tenochtitlan when the Spaniards arrived. The Templo Mayor shows
the practice of rebuilding a temple on the same spot, each new structure enlarg-
ing on and covering its predecessors. In sharp contrast, Aztec rulers tended to
build new palaces; in 1519 Axayacatl’s was still standing, not far from that of
Motecuhzoma II. Aztec palaces were luxurious, with numerous facilities for
large staffs and a wide range of civic and private activities.
The Aztec pattern is similar to many palaces in European and other tradi-
tions. Frequent major changes are most likely when rulers have relatively un-
restricted control over a large fraction of state resources and can command the
construction of residences as much for their personal glorification as for the
state. When heads of state directly control fewer resources, and especially if
their residence cannot be viewed as family property, as, for example, the US
White House, major changes are less likely.
The Ciudadela fits this second pattern better (Cowgill 1983). It is about
three fourths the volume of the Sun Pyramid, but most of the mass is in the
great outer platforms and the FSP. The total area of apartments would only ac-
commodate a few hundred occupants. To begin with, there seems to have been
little differentiation among apartments. Probably it was designed to serve a
152 COWGILL
ruler who could command great resources but was accustomed to operating
with a small staff. The South Palace remained nearly unchanged, but some
doorways in the North Palace were blocked to make it less public, it was en-
larged by another room complex that projected into the great plaza of the Ciu-
dadela, and it had relatively good access to the large walled compound imme-
diately to the north, where specialists in moldmade censer adornos and perhaps
other artisans worked. These look like features intended to facilitate adminis-
trative purposes at relatively low cost and without too drastically changing
outward appearances.
Another macrocomplex that may have been used for high-level government
and may possibly have housed rulers is the “Avenue of the Dead Complex”
(Cowgill 1983). It straddles the Avenue of the Dead, is partially enclosed by
large walls, extends about 350 × 350 m, and has many groups of rooms, apart-
ment compounds, pyramids, platforms, and plazas. Their number and variety
would have provided for more administrative activities than could easily be ac-
commodated by the Ciudadela. It includes the luxurious Viking Group com-
pound, the Superposed Buildings group, and, in its west center, the “West
Plaza Group” (Morelos García 1993). Much remains unexcavated. Earliest
major structures are probably a little later than the Ciudadela, although we
need greater chronological precision. Major rebuildings followed. Perhaps the
rulers’ residence shifted to the Avenue of the Dead Complex during the politi-
cal changes that may have occurred somewhere between 250 and 350 CE.
An Oligarchic Republic?
Supreme Teo political authority may not always have been strongly concen-
trated in a single person or lineage. R Millon (1976) suggested that Teo might
have been an oligarchic republic. The case now seems stronger, though not yet
overwhelming. Rulership in early states was not always monarchic. It is no
longer widely thought that states arose as responses to social and/or environ-
mental problems and benefited commoners as well as elites. The main expla-
nations, however, of how elites could exploit the rest are that they could
threaten force and that they promoted ideologies in which the gods and the
very nature of the universe made inequality right, or at least unavoidable. Un-
doubtedly force and religious ideology were used by the Teo state. However, a
more mundane civic consciousness, a sense of the virtue of “good citizenship,”
may also have been a factor. Given the prominence of this theme in modern so-
cieties, it is surprising that ancient Mesoamericans are not often credited with
such perceptions and sentiments.
Pasztory’s (1992) concept of a “utopian” society touches on similar issues,
but I do not think Teo was utopian in any reasonable sense. Nevertheless, civic
pride and a sense of citizenship, and not just submission to overawing deities and
overpowering rulers, may explain much about Teo’s stability and why there
was no abundance of self-glorifying rulers. Blanton et al (1996) are probably
correct in considering the Teo state “corporate,”3 at least in its later stages.
An oligarchic republic is not necessarily democratic or egalitarian. There
are many Old World examples. Romila Thapar (1966) described republics in
early India. Other cases are the city states of Classical Greece, Rome before the
Empire, and some of the small states of Medieval and Renaissance northern It-
aly. Venice is a notable example.
Venice
Venice differed greatly from Teo (Muir 1981), yet it is interesting to compare.
By the 800s this island city was ruled by a duke subject to the Byzantine em-
peror in Constantinople. Over time, Venice gained its independence and the
duke (called “doge”; “ducal” is the adjective) was chosen by popular acclama-
tion. By around 1300 this was formalized in a system that lasted until the Na-
poleonic conquest of 1797, wherein the adult males of specified elite families
comprised a “Great Council,” whose members were the only ones eligible to
elect the doge and to hold that and a number of other high offices. The doge
was elected for life by an intricate system of balloting and lottery intended as
much to counteract factionalism (a recognized problem in other Italian states)
as to ensure representation of elite majority will. The office circulated widely
among leading families. Venetians thought themselves remarkably free of fac-
tions, and many outsiders saw them that way. Factions were probably more im-
portant than Venetians liked to admit, but they seem to have been less divisive
than in other Italian states.
Many restrictions were imposed on doges’ use of public funds or their pri-
vate resources; for example, gift giving was sharply restricted. In Eisenstadt’s
(1969) terms, doges commanded limited “free-floating resources.” Some do-
ges tried to subvert the system and gain more personal power, but with little
success. The elite were also relatively successful in keeping popular resistance
under control; some disturbances occurred but the masses never overthrew re-
gimes, as they sometimes did in other Italian states.
Individual doges were more celebrated than Teo rulers appear to us, but
their pedigrees seem to have been unimportant, as long as they belonged to
elite Great Council families. There are numerous portraits of doges, and many
3
3Millon (1992) noted that “corporate state” commonly refers to systems and ideologies that glorify
personal rule and the cult of the leader—the opposite of how Mesoamericanists have used the term.
“Collective” would be preferable, but “corporate,” in the sense of collective, may be too entrenched
to be changed easily.
154 COWGILL
had fine tombs (at least one was criticized for living and dying too simply), but
these seem by way of keeping up a certain dignity for the prestige of the state,
and I do not believe any ducal tombs or images became important in state ritual
and myth (as did the relics of various saints, especially St. Mark). If we had as
little data on Venice as we do on Teo, visibility of the doges would probably be
low.
The doge’s palace immediately adjoins the basilica of St. Mark, the princi-
pal religious structure of the state. An earlier ducal palace on this site was de-
stroyed by fire, but the present one has persisted for many centuries. Occa-
sional efforts to move the ducal residence to another site were successfully re-
sisted. Each new doge would move his immediate family and household fur-
nishings into the palace, but upon his death the survivors had only a few days
to remove themselves and their goods. Although various doges renovated or
modified the palace, its location and basic structure remained unchanged for a
long time. It sounds something like the Ciudadela palaces.
Early Autocracy?
How much did the Teo political system change over time? Teo probably never
emphasized inheritance and validation of rulership through pedigree as much
as the Classic Maya, yet early rulers may have been powerful and self-
glorifying. All the awe-inspiring monuments are early, and they represent an
audacious plan imposed on several square kilometers of landscape. Millon
(1992) thinks the layout developed over time in several stages, beginning
shortly after the concentration of most of the population of the Basin of Mex-
ico in the city around 1 CE. Sugiyama (1993) argued that all major elements of
the layout were probably envisioned as an integrated plan from the beginning,
although it may have taken some time to complete the construction project. He
relies most strongly on key linear dimensions of structures and distances be-
tween them, which he feels translated key calendrical numbers into a unified
spatial pattern, and he downplays ceramic evidence for the length of time that
elapsed between the earliest Sun Pyramid and the Ciudadela.
Whoever is more nearly correct about how much of the present pattern was
fully conceived from its inception, the layout of the monumental part of Teo
was created in two centuries or less. Teo began its urban growth in the last cen-
turies BCE and already covered about 8 km2, with a population of
20,000–40,000, before anything very monumental was constructed. The great
surge of building does not seem to represent the thought of weak rulers or of
persons strongly beholden to advisory councils. Moreover, it is just at the be-
ginning of this interval that virtually the entire population disappeared else-
where in the Basin of Mexico (Sanders et al 1979). People were evidently re-
settled in Teo. The official ideology may or may not have been collective, but
in any case it looks as if there were a few very powerful, very able, and very
imaginative rulers, who were probably not self-effacing persons. The immense
structures were probably seen as lasting monuments to these rulers, who
needed no inscriptions and no statues to reinforce the messages of the build-
ings. Sugiyama (1993) and Millon (1992) suspect that a royal tomb is associ-
ated with the Sun Pyramid; Millon suspects one also at the Moon Pyramid.
Absence of different plans for different city districts contrasts significantly
with many other Mesoamerican centers, where there is coordinated planning
within large segments but no single plan that encompasses all segments. Impo-
sition of one plan for almost the whole city is another sign of early strength of
the central authority at Teo and suggests relative weakness of intermediate so-
cial units, such as large lineages.
A Shift to More Collective Governance?
In the ensuing centuries, from about the middle of the 200s to the 600s or 700s,
the city’s population remained high and the total volume of monumental con-
struction was quite large. It consisted, however, of enlargements and modifica-
tions of existing complexes. It was also at this time that architecturally sub-
stantial apartment compounds were built. These soon housed nearly all resi-
dents, of low as well as intermediate and high status. Emphasis on building
apartment compounds rather than new pyramids may have been part of a con-
scious shift to greater concern for general well-being than for individual glory.
There is evidence that this change began violently. Our 1988–1989 excava-
tions revealed that the FSP and the temple atop it were burned in a hot fire, and
large fragments of modeled clay walls and other debris from the temple were
used as part of the fill for the stepped platform (“Plataforma Adosada”) that
covered (and preserved) most of the front of the FSP. Instead of being buried
by some grander pyramid, most of the ruined FSP was left exposed, perhaps a
reminder to any future ruler tempted to overstep, and it suffered further dam-
age4. It was probably at this time that looters tunneled into the FSP and re-
moved most contents of the largest pits. If these events happened soon after the
FSP was built, and if FSP victims were in fact loyal high-status Teotihuacanos,
it may be that elites saw the sacrifices as excessive and reacted strongly (Mil-
lon 1988b, Pasztory 1988). Identities of the victims, however, are not yet es-
tablished, and incomplete ceramic analyses suggest that a century or more may
have elapsed before the Plataforma Adosada was built. Perhaps several auto-
cratic rulers succeeded the one responsible for the FSP and the sacrifices, and
perhaps it was some time before a less able ruler made revolt possible.
4
4Conceivably this was when work on the idol of Coatlinchan halted; it may have been a ruler’s try at
personal glorification.
156 COWGILL
Whether reaction was swift or delayed, it seems to have initiated a period of

more collective rule. There may have been a consciously new theory of gov-
ernance. Possibly, however, Teo political theory always favored collective
rule, and the time of powerful rulers may have been seen as a tyrannical aberra-
tion. If so, the reaction would have been perceived as the restoration of tradi-
tional government.
Teotihuacanos living from infancy in sight of the great pyramids may no
longer have been overawed by them, but they must have taken immense pride
in them; one likely reason for their disinterest in things foreign.
The scarcity of obvious boasting by Teo rulers has prompted comparisons

with the Harappan civilization of the Indus Valley. Teo differs markedly, how-
ever. Harappan sites do not have the monumental civic-religious structures of
Teo nor the wealth of pictorial art, and settlement patterns are very different.
DECLINE AND COLLAPSE

Metepec-Period Decline?
During Teo’s last century the city’s population may have declined signifi-
cantly. The extent of decline, however, will not be clear until there is more
agreement on Metepec period ceramic diagnostics. Some households re-
mained quite prosperous, but disparities among households may have been in-
creasing (Sempowski 1994). Centers such as Xochicalco and Cacaxtla possi-
bly developed only after the fall of Teo but may have begun earlier, as a declin-
ing Teo lost its ability to punish upstarts. Better control of the chronology of
Teo’s decline and the rise of other central Mexican sites is crucial. It is easy to
imagine ways in which Teo government and society might have been in trou-
ble, through some combination of bureaucratic proliferation; failure to adapt to
“Epiclassic” styles of government, commerce, and religion that were develop-
ing elsewhere; and possibly environmental problems. Without new income
from new conquests and without crises posed by outside threats, rulers may
have found it hard to break free of increasingly stultifying constraints and un-
able to adjust to changes even if they had the will and wisdom (Cowgill 1992a,
Millon 1988b).
Fiery (But Selective) Destruction
The Teo state was physically destroyed by the burning of temples and elite
residences and the smashing of idols, especially in the central part of the city.
Millon emphasizes how selective the destruction was. It was intended to de-
stroy the artifacts and physical facilities of the Teo state. Millon (1988a) be-
lieves it could only have been done by insiders, but I think surrounding socie-
ties may have gained power and numbers to the point where they, or some
combination, perhaps including dissident insiders, could have defeated a

weakened and no longer well-led city. A sizable population, perhaps 40,000,
survived or resettled the city, which has remained a town of some importance
ever since. But Teo was never again the capital of a regional state.
ACKNOWLEDGMENTS
Preparation of this article was aided by work on a manuscript in press, during a
1992–1993 Fellowship at the Center for Advanced Study in the Behavioral
Sciences, Stanford, California. I am grateful for financial support provided by
NSF grant SES-9022192. Oralia Cabrera Cortés, John E Clark, René Millon,
Ian Robertson, Barbara L Stark, and Saburo Sugiyama made useful sugges-
tions on an earlier draft, but I am solely responsible for this version.
Visit the Annual Reviews home page at http://www.AnnualReviews.org.
Literature Cited
Agrinier P. 1975. Mounds 9 and 10 at Mira- Berlo JC. 1984. Teotihuacan Art Abroad. Br.
dor, Chiapas, Mexico. New World Ar- Archaeol. Rep. Int. Ser. 199, Oxford
chaeol. Found., Pap. 39, Provo, UT Berlo JC. 1989. Early writing in central Mex-
Algaze G. 1993. Expansionary dynamics of ico: in tlilli, in tlapalli before A. D. 1000.
some early pristine states. Am. Anthropol. See Diehl & Berlo 1989, pp. 19–47
95:304–33 Berlo JC, ed. 1992a. Art, Ideology, and the
Altschul JH. 1987. Social districts of Teoti- City of Teotihuacan. Washington, DC:
huacan. See McClung de Tapia & Rattray Dumbarton Oaks
1987, pp. 191–217 Berlo JC. 1992b. Icons and ideology at Teoti-
Armillas P. 1964. Northern Mesoamerica. In huacan: the Great Goddess reconsidered.
Prehistoric Man in the New World, ed. JD See Berlo 1992a, pp. 129–68
Jennings, E Norbeck, pp. 291–321. Chi- Berrin K, ed. 1988. Feathered Serpents and
cago: Univ. Chicago Press Flowering Trees: Reconstructing the Mu-
Arnold PJ, Pool CA, Kneebone RR, Santley rals of Teotihuacan. San Francisco: The
RS. 1993. Intensive ceramic production Fine Arts Museums of San Francisco
and Classic-period political economy in Berrin K, Pasztory E, eds. 1993. Teotihuacan:
the sierra de los Tuxtlas, Veracruz, Mex- Art from the City of the Gods. New York:
ico. Anc. Mesoam. 4(2):175–91 Thames & Hudson
Aveni AF, Hartung H, Kelley JC. 1982. Alta Blanton RE, Feinman GM, Kowalewski SA,
Vista (Chalchihuites), astronomical impli- Peregine PN. 1996. A dual-processual the-
cations of a mesoamerican ceremonial out- ory for the evolution of Mesoamerican
post at the tropic of Cancer. Am. Antiq. civilization. Curr. Anthropol. 37(1):
47(2):316–35 1–14
Bennyhoff JA. 1967. Chronology and change Cabrera Castro R, Rodríguez I, Morelos N,
in the Teotihuacan ceramic tradition. See eds. 1982a. Teotihuacan 80–82: Primeros
Sociedad Mexicana de Antropología 1967, Resultados. Mexico City: Inst. Nac. Antro-
pp. 19–29 pol. Hist.
Berlo JC. 1983. The warrior and the butterfly: Cabrera Castro R, Rodríguez I, Morelos N,
central Mexican ideologies of sacred war- eds. 1982b. Memoria del Proyecto Ar-
fare and Teotihuacan iconography. In Text queológico Teotihuacan 80–82. Mexico
and Image in Pre-Columbian Art, ed. JC City: Inst. Nac. Antropol. Hist.
Berlo, pp. 79–117. Br. Archaeol. Rep. Int. Cabrera Castro R, Rodríguez I, Morelos N,
Ser. 180, Oxford eds. 1991a. Teotihuacan 1980–1982: Nue-
158 COWGILL
vas Interpretaciones. Mexico City: Inst. Analysis in Archaeology, ed. HJ Hietala,

Nac. Antropol. Hist. pp. 154–95. Cambridge: Cambridge Univ.
Cabrera Castro R, Sugiyama S, Cowgill GL. Press
1991b. The Templo de Quetzalcoatl proj- Crespo AM, Mastache AG. 1981. La presencia
ect at Teotihuacan. Anc. Mesoam. 2(1): en el área de Tula, Hidalgo, de grupos rela-
77–92 cionados con el barrio de Oaxaca en
Cabrera Cortés MO. 1995. La lapidaria del Teotihuacan. See Rattray et al 1981, pp.
Proyecto Templo de Quetzalcoatl 1988– 99–106
1989. Licentiate thesis. Esc. Nac. Antrop. de la Fuente B, ed. 1995. La Pintura Mural
Hist., Mexico City Prehispánica en México I: Teotihuacan.
Cardós A, ed. 1990. La Epoca Clásica. Mex- Tomo 1: Catálogo. Mexico City: Inst. In-
ico City: Inst. Nac. Antropol. Hist. vestigaciones Estéticas, Univ. Nac. Autón.
Carlson JB. 1991. Venus-Regulated Warfare Méx.
and Ritual Sacrifice in Mesoamerica: Teo- Demarest AA, Foias AE. 1993. Mesoamerican
tihuacan and the Cacaxtla “Star Wars” horizons and the cultural transformation of
Connection. College Park, MD: Cent. Ar- Maya civilization. In Latin American Ho-
chaeoastron., Tech. Publ. 7 rizons, ed. DS Rice, pp. 147–91. Washing-
Clark JE. 1986. From mountains to molehills: ton, DC: Dumbarton Oaks
a critical review of Teotihuacan’s obsidian Díaz Oyarzábal CL. 1980. Chingú: un sitio
industry. See Isaac 1986, pp. 23–74 clásico del área de Tula, Hgo. Mexico
Coe MD. 1981. Religion and the rise of meso- City: Inst. Nac. Antropol. Hist.
american states. In The Transition to State- Diehl RA, Berlo JC, eds. 1989. Mesoamerica
hood in the New World, ed. GD Jones, RR After the Decline of Teotihuacan. Wash-
Kautz, pp. 157–71. Cambridge: Cam- ington, DC: Dumbarton Oaks
bridge Univ. Press Drennan RD, Fitzgibbons PT, Dehn H. 1990.
Cowgill GL. 1974. Quantitative studies of ur- Imports and exports in Classic Mesoameri-
banization at Teotihuacan. See Hammond can political economy: the Tehuacan Val-
1974, pp. 363–96 ley and the Teotihuacan obsidian industry.
Cowgill GL. 1979. Teotihuacan, internal Res. Econ. Anthropol. 12:177–99
militaristic competition, and the fall of the Drucker RD. 1974. Renovating a reconstruc-
Classic Maya. In Maya Archaeology and tion: The Ciudadela at Teotihuacan, Mex-
Ethnohistory, ed. N Hammond, GR ico. PhD thesis. Univ. Rochester
Willey, pp. 51–62. Austin: Univ. Tex. Eisenstadt SN. 1969. The Political Systems of
Press Empires. New York: Free Press
Cowgill GL. 1983. Rulership and the Ciu- García Cook A. 1994. Cantona. Mexico City:
dadela: political inferences from Teotihua- Salvat
can architecture. In Civilization in the An- Grove DC. 1987. Torches, “knuckle dusters”
cient Americas, ed. RM Leventhal, AL Ko- and the legitimization of Formative Period
lata, pp. 313–43. Cambridge, MA: Univ. rulership. Mexicon 9(3):60–65
N. M. Press & Peabody Mus., Harvard Hammond N, ed. 1974. Mesoamerican Ar-
Univ. chaeology: New Approaches. London:
Cowgill GL. 1992a. Toward a political history Duckworth
of Teotihuacan. In Ideology and Pre- Hassig R. 1992. War and Society in Ancient
Columbian Civilizations, ed. AA Demar- Mesoamerica. Berkeley: Univ. Calif.
est, GW Conrad, pp. 87–114. Santa Fe, Press
NM: Sch. Am. Res. Press Headrick A. 1996. The Teotihuacan Trinity:
Cowgill GL. 1992b. Social differentiation at unMASKing the political structure. PhD
Teotihuacan. In Mesoamerican Elites: An thesis. Univ. Tex., Austin
Archaeological Assessment, ed. DZ Helms M. 1993. Craft and the Kingly Ideal.
Chase, AF Chase, pp. 206–20. Norman: Austin: Univ. Tex. Press
Univ. Okla. Press Hirth KG. 1978. Teotihuacan regional popula-
Cowgill GL. 1992c. Teotihuacan glyphs and tion administration in eastern Morelos.
imagery in the light of some early colonial World Archaeol. 9(3):320–33
texts. See Berlo 1992a, pp. 231–46 Hirth KG. 1980. Eastern Morelos and Teoti-
Cowgill GL. 1993. Distinguished lecture in ar- huacan. Nashville, TN: Vanderbilt Univ.
cheology: beyond criticizing New Arche- Publ. Anthropol. 25
ology. Am. Anthropol. 95(3):551–73 Hirth KG, ed. 1984. Trade and Exchange in
Cowgill GL, Altschul JH, Sload RS. 1984. Early Mesoamerica. Albuquerque: Univ.
Spatial analysis of Teotihuacan: a meso- N. M. Press
american metropolis. In Intrasite Spatial Hirth KG, Angulo Villaseñor J. 1981. Early
state expansion in central Mexico: Teoti- vestigaciones Antropol., Univ. Nac.

huacan in Morelos. J. Field Archaeol. Autón. Méx.
8(2):135–50 Manzanilla L. 1996. Corporate groups and do-
Hopkins MR. 1995. Teotihuacan cooking mestic activities at Teotihuacan. Lat. Am.
pots: scale of production and product vari- Antiq. 7(3):228–46
ability. PhD thesis. Brandeis Univ., Manzanilla L, Barba L. 1990. The study of ac-
Waltham, MA tivities in classic households. Anc.
Isaac BL, ed. 1986. Economic Aspects of Pre- Mesoam. 1:41–49
hispanic Highland Mexico. Greenwich, Marcus J, Flannery KV. 1996. Zapotec Civili-
CT: JAI Press zation. London: Thames & Hudson
Justeson JS, Norman WM, Campbell L, Kauf- Matos Moctezuma E. 1988. The Great Temple
man T. 1983. The foreign impact on low- of the Aztecs. London: Thames & Hudson
land Mayan language and script: a sum- McCafferty GG. 1996. Reinterpreting the
mary. See Miller 1983, pp. 147–58 great pyramid of Cholula, Mexico. Anc.
Kidder AV, Jennings JD, Shook EM. 1946. Mesoam. 7(1):1–17
Excavations at Kaminaljuyu, Guatemala. McClung de Tapia E, Rattray EC, eds. 1987.
Washington, DC: Carnegie Inst. Wash. Teotihuacan: Nuevos Datos, Nuevas Sínte-
Kowalski JK. 1997. Natural order, social or- sis, Nuevos Problemas. Mexico City: Inst.
der, political legitimacy and the sacred Investigaciones Antropol., Univ. Nac.
city: the architecture of Teotihuacan. In Autón. Méx.

Mesoamerican Architecture as a Sacred Miller AG, ed. 1983. Highland-Lowland In-
Symbol, ed. JK Kowalski. New York/Ox- teraction in Mesoamerica. Washington,
ford: Oxford Univ. Press DC: Dumbarton Oaks
Kubler G. 1967. The Iconography of the Art of Millon C. 1973. Painting, writing, and polity
Teotihuacan. Washington, DC: Dumbar- in Teotihuacan, Mexico. Am. Antiq. 38(3):
ton Oaks 294–314
Kurtz DV. 1987. The economics of urbaniza- Millon C. 1988. A reexamination of the Teoti-
tion and state formation at Teotihuacan. huacan tassel headdress. See Berrin 1988,
Curr. Anthropol. 28(3):329–53 pp. 114–34
Langley JC. 1986. Symbolic Notation of Teoti- Millon R. 1973. Urbanization at Teotihuacan,
huacan. Br. Archaeol. Rep. Int. Ser. 313, Mexico, Vol. 1, The Teotihuacan Map.
Oxford Part 1: Text. Austin: Univ. Tex. Press
Langley JC. 1991. The forms and usage of no- Millon R. 1974. The study of urbanism at Teo-
tation at Teotihuacan. Anc. Mesoam. 2(2): tihuacán, Mexico. See Hammond 1974,
285–98 pp. 335–62
Langley JC. 1992. Teotihuacan sign clusters: Millon R. 1976. Social relations in ancient
emblem or articulation? See Berlo 1992a, Teotihuacan. In The Valley of Mexico, ed.
pp. 247–80 ER Wolf, pp. 205–48. Albuquerque: Univ.
Laporte JP, Fialko V. 1990. New perspectives N. M. Press
on old problems: dynastic references for Millon R. 1981. Teotihuacan: city, state, and
the Early Classic at Tikal. In Vision and civilization. In Supplement to the Hand-
Revision in Maya Studies, ed. FS Clancy, book of Middle American Indians, Vol. 1,
PD Harrison, pp. 33–66. Albuquerque: Archaeology, ed. JA Sabloff, pp. 198–243.
Univ. N. M. Press Austin: Univ. Tex. Press
Linné S. 1942. Mexican Highland Cultures. Millon R. 1988a. The last years of Teotihua-
Stockholm: Ethnog. Mus. Sweden can dominance. In The Collapse of Ancient
López Austin A, López Luján L, Sugiyama S. States and Civilizations, ed. N Yoffee, GL
1991. The Temple of Quetzalcoatl at Teo- Cowgill, pp. 102–64. Tucson: Univ. Ariz.
tihuacan: its possible ideological signifi- Press
cance. Anc. Mesoam. 2(1):93–105 Millon R. 1988b. Where do they all come
López Luján L. 1989. La Recuperación from? The provenance of the Wagner mu-
Mexica del Pasado Teotihuacano. Mexico rals from Teotihuacan. See Berrin 1988,
City: Inst. Nac. Antropol. Hist. pp. 78–113
Manzanilla L. 1992. The economic organiza- Millon R. 1992. Teotihuacan studies: from
tion of the Teotihuacan priesthood: hy- 1950 to 1990 and beyond. See Berlo
potheses and considerations. See Berlo 1992a, pp. 339–429
1992a, pp. 321–38 Millon R, Drewitt RB, Cowgill GL. 1973. Ur-
Manzanilla L, ed. 1993. Anatomía de un Con- banization at Teotihuacan, Mexico, Vol. 1,
junto Residencial Teotihuacano en Oz- The Teotihuacan Map. Part 2: Maps. Aus-
toyahualco, 2 Vols. Mexico City: Inst. In- tin: Univ. Tex. Press
160 COWGILL
Morelos García N. 1993. Proceso de Produc- ico City: Inst. Investigaciones Antropol.,
ción de Espacios y Estructuras en Teoti- Univ. Nac. Autón. Méx.
huacán. Mexico City: Inst. Nac. Antropol. Robertson I. 1997. Thematic mapping of em-
Hist. pirical Bayesian parameter estimates at
Muir E. 1981. Civic Ritual in Renaissance Teotihuacan, Mexico. J. Quant. Anthro-
Venice. Princeton, NJ: Princeton Univ. pol. In press
Press Sanders WT, ed. 1994–1996. The Teotihuacan
Múnera LC. 1985. Un Taller de Cerámica Rit- Valley Project Final Report, Vol. 3, The
ual en la Ciudadela, Teotihuacan. Licenti- Teotihuacan Period Occupation of the
ate thesis. Esc. Nac. Antrop. Hist., Mexico Valley. 4 Parts. University Park: Matson
City Mus. Anthropol., Pa. State Univ. Press
Nutini HG. 1988. Todos Santos in Rural Sanders WT, Michels JW, eds. 1977. Teoti-
Tlaxcala. Princeton, NJ: Princteon Univ. huacan and Kaminaljuyu. University
Press Park: Pa. State Univ. Press

Pasztory E. 1974. The Iconography of the Teo- Sanders WT, Parsons JR, Santley RS. 1979.
tihuacan Tlaloc. Washington, DC: Dum- The Basin of Mexico. New York: Aca-
barton Oaks demic
Pasztory E. 1977. (1973). The gods of Teoti- Santley RS. 1983. Obsidian trade and Teoti-
huacan. In Int. Congr. Am., 40th, Rome- huacan influence in Mesoamerica. See
Genoa, 1972. 1:147–59. Reprinted in Pre- Miller 1983, pp. 69–124

Columbian Art History, ed. A Cordy- Santley RS. 1984. Obsidian exchange, eco-
Collins, J Stern, pp. 81–95. Palo Alto, CA: nomic stratification, and the evolution of
Peek complex society in the Basin of Mexico.
Pasztory E. 1988. A reinterpretation of Teoti- See Hirth 1984, pp. 43–86
huacan and its mural painting tradition. Santley RS. 1989. Obsidian working, long-
See Berrin 1988, pp. 45–77 distance exchange, and the Teotihuacan
Pasztory E. 1990. El poder militar como reali- presence on the south Gulf Coast. See
dad y metáfora en Teotihuacan. See Diehl & Berlo 1989, pp. 131–51
Cardós 1990, pp. 181–204 Santley RS, Kerley JM, Kneebone RR. 1986.
Pasztory E. 1992. Abstraction and the rise of a Obsidian working, long-distance ex-
utopian state at Teotihuacan. See Berlo change, and the politico-economic organi-
1992a, pp. 281–320 zation of early states in central Mexico.
Pasztory E. 1993. An image is worth a thou- See Isaac 1986, pp. 101–32
sand words: Teotihuacan and the meanings Santley RS, Yarborough C, Hall B. 1987. En-
of style in Classic Mesoamerica. In Latin claves, ethnicity, and the archaeological
American Horizons, ed. DS Rice, pp. record at Matacapan. In Ethnicity and Cul-
113–45. Washington, DC: Dumbarton ture, ed. R Auger, M Glass, S Mac-
Oaks Eachern, P McCartney, pp. 85–100. Cal-
Pasztory E. 1997. Teotihuacan: An Experi- gary: Univ. Calgary
ment in Living. Norman: Univ. Okla. Press Schele L, Freidel D. 1990. A Forest of Kings.
Pendergast DM. 1990. Excavations at Altun New York: Morrow
Ha, Belize, 1964–1970, Vol. 3. Toronto: R. Séjourné L. 1959. Un Palacio en la Ciudad de
Ont. Mus. los Dioses. Mexico City: Inst. Nac. Antro-
Rattray EC. 1989. El barrio de los comerci- pol. Hist.
antes y el conjunto Tlamimilolpa. Ar- Sempowski M. 1994. Mortuary practices at
queología 5:105–29 Teotihuacan. See Sempowski & Spence
Rattray EC. 1990. The identification of ethnic 1994, pp. 1–314
affiliation at the Merchants’ barrio, Teoti- Sempowski ML, Spence MW, eds. 1994. Mor-
huacan. In Etnoarqueología: Primer Colo- tuary Practices and Skeletal Remains at
quio Bosch-Gimpera, ed. Y Sugiura, MC Teotihuacan. Salt Lake City: Univ. Utah
Serra, pp. 113–38. Inst. Investigaciones Press
Antropol., Univ. Nac. Autón. Méx. Sheehy JJ. 1992. Ceramic production in an-
Rattray EC. 1992. The Teotihuacan Burials cient Teotihuacan, Mexico: a case sudy of
and Offerings. Nashville: Vanderbilt Univ. Tlajinga 33. PhD thesis. Pa. State Univ.
Publ. Anthropol. 42 Sociedad Mexicana de Antropología. 1967.
Rattray EC. 1993. The Oaxaca Barrio at Teo- Teotihuacan, Onceava Mesa Redonda.
tihuacan. Puebla: Univ. Américas, Mexico City
Monogr. Mesoam. 1 Sociedad Mexicana de Antropología. 1972.
Rattray EC, Litvak J, Díaz C, eds. 1981. Inter- Teotihuacan, XI Mesa Redonda. Mexico
acción Cultural en México Central. Mex- City
Spence MW. 1974. Residential practices and City of Teotihuacan. Tuscaloosa: Univ.
the distribution of skeletal traits in Teoti- Ala. Press
huacan, Mexico. Man 9(2):262–73 Sugiyama S. 1989a. Burials dedicated to the
Spence MW. 1981. Obsidian production and Old Temple of Quetzalcoatl at Teotihua-
the state in Teotihuacan. Am. Antiq. 46(4): can, Mexico. Am. Antiq. 54(1):85–106
769–88 Sugiyama S. 1989b. Iconographic interpreta-
Spence MW. 1984. Craft production and pol- tion of the Temple of Quetzalcoatl at Teo-
ity in early Teotihuacan. See Hirth 1984, tihuacan. Mexicon 11(4):68–74
pp. 87–114 Sugiyama S. 1993. Worldview materialized in
Spence MW. 1987. The scale and structure of Teotihuacan, Mexico. Lat. Am. Antiq.
obsidian production in Teotihuacan. See 4(2):103–29
McClung de Tapia & Rattray 1987, pp. Sugiyama S. 1995. Mass Human Sacrifice and
429–50 Symbolism of the Feathered Serpent Pyra-
Spence MW. 1989. Excavaciones recientes en mid in Teotihuacan, Mexico. PhD thesis.
Tlailotlacan, el barrio oaxaqueño de Teoti- Ariz. State Univ., Tempe
huacan. Arqueología 5:82–104 Taube KA. 1983. The Teotihuacan Spider
Spence MW. 1992. Tlailotlacan, a Zapotec en- Woman. J. Lat. Am. Lore 9(2):107–89
clave in Teotihuacan. See Berlo 1992a, pp. Taube KA. 1992. The Temple of Quetzalcoatl
59–88 and the cult of sacred war at Teotihuacan.
Spence MW. 1994. Human skeletal material Res 21:53–87

from Teotihuacan. See Sempowski & Thapar R. 1966. A History of India, Vol. 1.
Spence 1994, pp. 315–445 New York: Penguin
Spence MW. 1996. Commodity or gift: Teoti- Turner MH. 1987. The lapidary industry of
huacan obsidian in the Maya region. Lat. Teotihuacan, Mexico. PhD thesis. Univ.
Am. Antiq. 7(3):21–39 Rochester
Stark BL. 1995. Estilos de volutas en el Turner MH. 1992. Style in lapidary technol-
período clásico. In Rutas de Intercambio ogy. See Berlo 1992a, pp. 89–112
en Mesoamérica, ed. EC Rattray. Mexico Umberger E. 1987. Antiques, revivals, and
City: Univ. Nac. Autón. Méx. In press references to the past in Aztec art. Res
Stark BL, Heller L, Glascock MD, Elam JM, 13:61–105
Neff H. 1992. Obsidian-artifact source von Winning H. 1987. La Iconografía de Teo-
analysis for the Mixtequilla region, south- tihuacan. Mexico City: Univ. Nac. Autón.
central Veracruz, Mexico. Lat. Am. Antiq. Méx.
3(3):221–39 Widmer RJ. 1991. Lapidary craft specializa-
Stone A. 1989. Disconnection, foreign insig- tion at Teotihuacan: implications for com-
nia, and political expansion: Teotihuacan munity structure at 33:S3W1 and eco-
and the warrior stelae of Piedras Negras. nomic organization in the city. Anc.
See Diehl & Berlo 1989, pp. 153–73 Mesoam. 2(1):131–41
Storey R. 1985. An estimate of mortality in a Widmer RJ, Storey R. 1993. Social organiza-
pre-Columbian urban population. Am. An- tion and household structure of a Teotihua-
thropol. 87:519–35 can apartment compound: S3W1:33 of the
Storey R. 1991. Residential compound organi- Tlajinga barrio. In Prehispanic Domestic
zation and the evolution of the Teotihua- Units in Western Mesoamerica, ed. RS
can state. Anc. Mesoam. 2(1):107–18 Santley, KG Hirth, pp. 87–104. Boca Ra-
Storey R. 1992. Life and Death in the Ancient ton: CRC Press

State and Society at Teotihuacan Mexico (Cowgill, George)

Uploaded by

Copyright:

Available Formats

State and Society at Teotihuacan Mexico (Cowgill, George)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

State and Society at Teotihuacan Mexico (Cowgill, George)

Uploaded by

Copyright:

Available Formats

What was the population size and growth of Teotihuacan?

What was the population size and growth of Teotihuacan?

What major structures were built in Teotihuacan?

What major structures were built in Teotihuacan?

Annu. Rev. Anthropol. 1997.

STATE AND SOCIETY AT

85287-2402; e-mail: [email protected]

KEY WORDS: archaeology, Mesoamerica, early states, governmentality, ideology

parently reaching a plateau of 100,000 or more by the 300s or earlier. No other

Millon et al (1973) published map sheets of the whole city at a scale of

I concentrate here on publications since the mid-1960s that bear especially on

METHOD AND THEORY

Figure 1 Teotihuacan ceramic chronology.

Teo is in a challenging twilight zone for direct historical approaches; close

GROWTH OF THE CITY

It is notoriously difficult to derive accurate population estimates from archaeo-

sizes, layouts, and inferred uses of rooms in excavated apartment compounds

EXTENT OF TEOTIHUACAN RULE

In Oaxaca, the Zapotec state was independent and maintained diplomatic

torches and a rattlesnake headdress, suggesting military action in the lowlands

WRITING AND LITERACY

PERSONALITY AND SOCIALIZATION

pecially of fierce animals) convey tension and vibrant power, as in the

ity of identical scenes, suggests an ethos in which individuals were inter-

NONELITE ELEMENTS OF THE SOCIETY

studies of nonmetric skeletal traits suggest a preference for patrilocal post-

Barrios, Enclaves, and Districts

tance of apartment compounds as multihousehold social and political units

ual inhumations beneath floors. Socioeconomic status of the enclave occu-

Freestanding walls bound a number of precincts, especially near the Avenue of

not. Nevertheless, they were probably used in rituals commemorating the

several kilometers apart, that formed a perpendicular to the Avenue of the

makes state involvement somewhat unlikely. Pecked cross-&-circles and

CRAFT PRODUCTION AND TRADE

though we have recognized no evidence of a distinct merchant class [Manza-

WAR AND THE MILITARY

activities, ca 200 CE (Cabrera Castro et al 1991b; Sugiyama 1989a,b, 1992,

Figure 3 A “portrait” figurine (after Pasztory 1997).

The “Great Goddess” and Rulership

THE NATURE OF TEOTIHUACAN RULERSHIP

Whether reaction was swift or delayed, it seems to have initiated a period of

The scarcity of obvious boasting by Teo rulers has prompted comparisons

DECLINE AND COLLAPSE

combination, perhaps including dissident insiders, could have defeated a

Visit the Annual Reviews home page at http://www.AnnualReviews.org.

vas Interpretaciones. Mexico City: Inst. Analysis in Archaeology, ed. HJ Hietala,

state expansion in central Mexico: Teoti- vestigaciones Antropol., Univ. Nac.

city: the architecture of Teotihuacan. In Autón. Méx.

Press Park: Pa. State Univ. Press

Genoa, 1972. 1:147–59. Reprinted in Pre- Miller 1983, pp. 69–124

Spence MW. 1994. Human skeletal material Res 21:53–87

You might also like