,.~.Y...v...... .
. I. ‘..

Tn.
Jig...» :

 

‘
b;
g? ihzgz. 2.
‘M- .jxadidfirn “I“
l tun-nary... ”

‘ fig... m
‘ E’I’

up, -
a

flown...) 12’3“... .Iht ..
g. aduJ.)MmH¢.W 9.11.4 It)!
a ;

J l-'

S It!
, 51:53.!
. .laahuu

“Mr; .L...

. . . .. 3y 3.133.291.13 ‘ .,
. . . . ‘ . V. . .E....a..w.a.«w....uwma .Jhwwx&%¢.#mu@

« ..
.. .. :a 1 -

 

::

 

LIBRARY
Michigan State
University

This is to certify that the
thesis entitled

Residential Zoning at Prehispanic Tzintzuntzan, Mexico
Revisited: A Quantitative Analysis

presented by

Christopher James Stawski

has been accepted towards fulfillment
of the requirements for the

MA. degree in Anthropology

 

 

jig/(412$ 4/66/1177 Lu 1/ M

Major Professor’s Signature

We] 127, 9008
fl ' l

Date

MSU is an affinnafive-action. equal-opportunity employer

..._.._--.-

—.-i-.~.-.—.-.

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K‘lProgIAcc8Pres/CIRC/DateDueindd

RESIDENTIAL ZONING AT PREHISPANIC TZINTZUNTZAN, MEXICO
REVISITED: A QUANTITATIVE ANALYSIS

By

Christopher James Stawski

A THESIS
Submitted to
Michigan State University
in partial fiilfillment of the requirements
for the degree of
MASTER OF ARTS
Department of Anthropology

2008

ABSTRACT

RESIDENTIAL ZONING AT PREHISPANIC TZINTZUNTZAN, IVIEXICO
REVISITED: A QUANTITATIVE ANALYSIS

By

Christopher James Stawski

This thesis focuses on the archaeological site of Tzintzuntzan, a Prehispanic site
located in West-Central Mexico. Situated within the Patzcuaro Lake Basin, Tzintzuntzan
was once the capital of the Tarascan Empire, a state-level, Mesoamerican society that
emerged around AD 1350. In 1970, Helen Pollard completed an archaeological survey of
Tzintzuntzan, forming the basis of her PhD dissertation in 1972. Central to Pollard’s
dissertation was the concept of urbanism in the formation of Tzintzuntzan. She tested the
presence of urban planning and zoning, the function of Tzintzuntzan as the major
Tarascan city, and whether it was political, ceremonial, or economic in nature. This
thesis reanalyzes Pollard’s Tzintzuntzan survey with focus on the ceramic artifacts
collected, and provides a quantitative analysis for the classification of sites into zonal
categories.

Through multiple lines of evidence, bridging arguments are made that identify
ceramic variables sensitive to social class and that aid in the classification of sites into
residential and public zones. This allows for the quantitative testing of Pollard’s original
zonal interpretations, and aids in identifying the extent to which urbanism occurred at
Tzintzuntzan. Ultimately, this thesis has created a methodological framework for
reanalyzing archaeological surveyed data from the Late Postclassic period, one that isn’t

limited to Tzintzuntzan but can be applied elsewhere in the Tarascan Empire.

Copyright by
CHRISTOPHER JAMES STAWSKI
2008

ACKNOWLEGDMENTS

First and foremost, I would like to thank my committee chair, Dr. Helen Pollard
for the use of her dissertation data and research used in this thesis. Dr. Pollard has been
unwavering in her support and patience, and I am indebted to her for all of her input and
guidance through this process. I would also like to thank Dr. Bill Lovis and Dr. Alison
Rautman for their helpful comments and assistance while preparing and writing this
thesis, as well as my fellow graduate students for their words of encouragement and
support.

Finally, I would like to thank my family and friends for their support and
encouragement, with a special thanks to my parents, my brother and my sisters for their

continued faith and love.

iv

TABLE OF CONTENTS

LIST OF TABLES .................................................................................. vii
LIST OF FIGURES .................................................................................. ix
CHAPTER 1
INTRODUCTION AND PROBLEM .............................................................. 1
Problem ........................................................................................ 1
Previous Research ............................................................................ 4
Research Goals and Data ................................................................... 5
Review of Literature ........................................................................ 9
Research Methodology .................................................................... 12
Predictions .................................................................................. 14
Significance of Research .................................................................. 15
CHAPTER 2
RESEARCH METHODOLOGY .................................................................. 17
Ceramic Production ........................................................................ 17
Excavations in the Zacapu Basin ......................................................... 21
Excavations and Collections at Tzintzuntzan .......................................... 35
Excavations and Collections at Urichu .................................................. 40
Ethnohistoric Data ......................................................................... 43
Ceramic-Status Associations ............................................................. 45
CHAPTER 3
CERAMIC ANALYSIS ........................................................................... 48
Cluster Analysis ............................................................................ 49
Cluster Analysis — Upper Elite Variables ............................................... 51
Cluster Analysis — Lower Elite Variables ............................................... 53
Cluster Analysis — Commoner Variables ................................................ 55
Cluster Analysis — Public Zones ......................................................... 56
Cluster Analysis — Querenda White Ware .............................................. 58
Interpretations .............................................................................. 60
Statistical Significance of the Cluster Analyses ....................................... 63
Discussion ........................................................... . ....................... 65
CHAPTER 4
URBAN ISM AT TZINTZUNTZAN REVISITED ............................................ 68
Urbanism at Tzintzuntzan ................................................................. 78
Urban Planning ............................................................................. 79
Urban Zoning ............................................................................... 81
Conclusion .................................................................................. 83

APPENDICES ....................................................................................... 88

BIBLIOGRAPHY ................................................................................. l 20

Vi

LIST OF TABLES

Table l — Ceramic Assemblage from Mich. 95, Group B ..................................... 26
Table 2 — Jars: Bowls: Miniature Vessel ratios from Mich. 95, Group B ................... 28
Table 3 — Burials with Age, Gender, and Artifacts from Mich. 95, Group B ............... 32
Table 4 - Ceramics from Offering 1, chata 5 on the Main Platform at Tzintzuntzan
(Season IH) .......................................................................................... 37
Table 5 — Burials with Age, Gender, and Artifacts from Urichu, Patzcuaro Lake Basin,
Area 1 ................................................................................................ 42
Table 6 — Ceramic Variables and Associated Social Classes ................................. 46
Table 7 — Pipe Variables: Stems, Bowls, and Bits ............................................. 58
Table 8 — Zonal Classification of Sites (n=89) Based on Cluster Analysis .................. 62

Table 9 ~Differences between Commoner and Lower Elite Zones — Pollard v. Mine. . . ..67
Table 10 — Survey Sites and Site Size Categorized by Zone .................... . ............. 80
Table 11 — Summary of Spatial Extent of Zoned Sites at Tzintzuntzan ..................... 81

Table 12 — Upper Elite Variables: Cluster Membership — K-Means Cluster Analysis
(11 = 89sites) ......................................................................................... 90

Table 13 - Upper Elite Variables: F inal Cluster Centers — K-Means Cluster Analysis
(n = 89 sites) ......................................................................................... 91

Table 14 — Lower Elite Variables: Cluster Membership - K-Means Cluster Analysis
(11 = 84 sites) ......................................................................................... 94

Table 15 - Lower Elite Variables: Final Cluster Centers -— K-Means Cluster Analysis
(11 = 84 sites) ......................................................................................... 95

Table16 — Commoner Variables: Cluster Membership — K-Means Cluster Analysis
(n=6l sites) ........................................................................................... 97

Table 17 -— Commoner Variables: Final Cluster Centers — K-Means Cluster Analysis
(n=61 sites) .......................................................................................... 97

vii

Table 18 — Public Zone Variables: Cluster Membership — K-Means Cluster Analysis
(n=89 sites) ......................................................................................... 100

Table 19 -— Public Zone Variables: Final Cluster Centers - K-Means Cluster Analysis
(n=89 sites) ......................................................................................... 101

Table 20 — Public Zone Variables: Cluster Membership - K-Means Cluster Analysis
(n=87 sites) ......................................................................................... 102

Table 21 — Public Zone Variables: Final Cluster Centers - K-Means Cluster Analysis
(n=87 sites) ......................................................................................... 103

Table 22 - Querenda White Variables: Cluster Membership - K-Means Cluster Analysis
(n=89 sites) ......................................................................................... 106

Table 23 — Querenda White Variables: Final Cluster Centers - K-Means Cluster Analysis
(n = 89 sites) ....................................................................................... 107

Table 24 - Querenda White Variables: Cluster Membership - K-Means Cluster Analysis
(n=86 sites) ......................................................................................... 108

Table 25 - Querenda White Variables: Final Cluster Centers - K-Means Cluster Analysis
(n=86 sites) ......................................................................................... 109

viii

LIST OF FIGURES

Figure 1 — The State of Michoacan and the Research Area ..................................... 2
Figure 2 — The 1970 Tzintzuntzan Survey ......................................................... 6
Figure 3 — The Site of Las Milpillas .............................................................. 22
Figure 4 - A Tarascan Decorated Polychrome with Basket Handle and Spout ............. 41
Figure 5 — Location of Querenda White Ware Ceramics ...................................... 60
Figure 6 -— Zonal Classification of Tzintzuntzan Sites .......................................... 63
Figure 7 — Public and Residential Zone Classifications from Cluster Analyses ............ 71
Figure 8 — Location of Manufacturing Zones (Li, Lii, Liii) ................................... 73
Figure 9a — Residential Zones ..................................................................... 75
Figure 9b - Public Zones .................................... ' ....................................... 76
Figure 10— The Primary Public Zone and Upper Elite Zone at Tzintzuntzan ............... 77
Figure 11 — Upper Elite Variables: Dendrogram Using Average Linkage (Between
Groups) ............................................................................................... 88
Figure 12 — Lower Elite Variables: Dendrogram Using Average Linkage (Between
Groups) ............................................................................................... 92
Figure 13 — Commoner Variables: Dendrogram Using Average Linkage (Between
Groups) ............................................................................................... 96
Figure 14 — Public Zone Variables: Dendrogram Using Average Linkage (Between
Groups) ............................................................................................... 98
Figure 15 — Querenda White Variables: Dendrogram Using Average Linkage (Between
Groups) ............................................................................................. 104
Figure 16 - Sipiho Gris: Tecuena Group ....................................................... 110
Figure 17 - Tarerio Cream: Jaracuro simple bowl ........................................................... 110
Figure 18 - Tariacuri Cafe: La Vinata Group .................................................................. 110

ix

Figure 19 - Tariacuri Cafe: Cucuhucho (handle) ............................................................. 110

Figure 20 - Tariacuri Café: Restricted Vessel (spout handle) .......................................... 110
Figure 21 - Tariacuri Café var. Sta. Ana .......................................................................... 110
Figure 22 - Tecolote Orange Mini Bowl Fragments ........................................................ 111
Figure 23 - Yaguarato Coarse var. San Bartolo (brasero) ............................................... 111

Figure 24 - Late Urichu Phase: Arocutin Rojo con Negativo ................................. 111
Figure 25 - Late Urichu Phase: Copujo Rojo sobre Crema (small support) ............... 11 1
Figure 26 - Oval Cross Section w/ Decoration ................................................. 110
Figure 27 - Incised Twist Four .................................................................... 112
Figure 28 - Circular Cross Section: Red Slip .................................................... 112
Figure 29 - Oval Cross Section ................................................................... 112
Figure 30 — Tzintzuntzan Survey Map: Northwest Portion .................................. 113
Figure 31 — Tzintzuntzan Survey Map: West Central Portion ............................... 114
Figure 32 — Tzintzuntzan Survey Map: East Central Portion ................................ 1 15
Figure 33 — Tzintzuntzan Survey Map: Lower Northeast Portion ........................... 116
Figure 34 — Tzintzuntzan Survey Map: Upper Northeast Portion ........................... 117
Figure 35 — Tzintzuntzan Survey Map: South Portion ........................................ 118

Chapter One
Introduction and Problem
Problem

This thesis reanalyzes the results of an archaeological survey that was completed
by Helen Pollard at the site of Tzintzuntzan in Michoacan, Mexico. Tzintzuntzan is an
ancient Mesoamerican settlement which became the capital of the Tarascan Empire
around AD 1350 (Pollard 1993: 29). The survey area is located in the northeast portion
of the Patzcuaro Lake Basin on the shores of the lake, and is 901 hectares in size]. The
survey was completed by Pollard in 1970, and formed the basis of a PhD dissertation in
1972. The dissertation and its subsequent publications are an invaluable asset to
archaeology, both within and outside of the Mesoamerican archaeological community.
Since the original survey, the archaeological record has been compromised due to
looting, soil erosion, and construction in or around the site of Tzintzuntzan. Because of
this, it is impossible to recreate Pollard’s original survey to test her interpretations of
Tzintzuntzan.

Central to Pollard’s dissertation was the concept of urbanism in the formation of
Tzintzuntzan. She tested the presence of urban planning and zoning, the function of
Tzintzuntzan as a Tarascan city, and whether it was political, ceremonial, or economic in
nature. These questions are at the center of my thesis, as I reanalyze the artifacts from
the Tzintzuntzan survey to test the internal urban characteristics that Pollard has
concluded were present at Tzintzuntzan. Such questions of urbanism include the

dichotomy that exists between urban planning and urban zoning, ranging from the level

 

' Utilizing GIS technology and methods, the map of the survey area was georeferenced to a rectified aerial
image, and the previous survey size was corrected accordingly. See page 9 for a more detailed
methodological description

Figure 1 — The State of Michoacan and the Research Area

 

 

  
 

Mlchoacén 7
' Lake Pétzcuaro

 

 

Research Area

of the individual to the settlement level (Smith 2007: 3). Michael Smith explains this
dichotomy in his updated approach to ancient urban planning. He discusses urban
planning in terms of a range of variation instead of the commonly held misnomer that an
ancient city was either planned or unplanned. Smith discusses this range in terms of
specific variables that define the level of planning, such as the coordination among
buildings, orthogonality and access and visibility to name a few (2007:7). Furthermore,
both Smith and Pollard acknowledge zoning to be considered a form of urban planning,
where space is patterned upon differential land use, and the land use having different
functions within society. The thesis focuses on the testing of Pollard’s zonal assignments
of the survey sites, primarily with regard to the residential and to a smaller degree the

public zones. This aids in the analysis of the internal structure of urban zoning at

Tzintzuntzan, and helps to test Pollard’s original inferences about the urbanization of
Tzintzuntzan.

My null hypothesis for this thesis is that there are no differences between my
zonal results and Pollard’s original zonal assignments. In order to test this hypothesis, I
have laid out a methodology that builds bridging arguments that tie social status to
ceramic variables, thus allowing for a classification of the survey sites separate from
Pollard’s classification. The two sets of site classifications will then be compared,
revealing any differences between the two analyses. This ultimately tests Pollard’s
interpretation of the urbanism that occurred at Tzintzuntzan, and allows for a different
analysis to be completed with updated theory and techniques.

This thesis works from the assumption that Tzintzuntzan was a socially stratified
city that represents the Tarascan capital. What is being tested for is the extent that
urbani sm occurred at Tzintzuntzan, with a distinction made between zoning and
planning. By testing Pollard’s residential and public zones, I intend to determine what
role zoning and planning had within Tzintzuntzan settlement. I predict that my results
will confirm Pollard’s original conclusion that zoning was the primary form of urbanism,
and that urban planning although present, did not occur at the settlement level (Pollard
20032363)

A major aspect of this thesis is the ability to use different analytical methods to
test Pollard’s data. Pollard’s analysis of the sites and their artifact distributions was done
primarily by intuitive inspection of the artifacts and their attributes, without any
quantitative testing and in a non-systematic way. Through identification and isolation of

specific ceramic variables, Pollard was able to determine survey site fianction and thus

designate the site a zone. This thesis uses quantitative analysis in conjunction with a GIS
database to reanalyze the zonal designations that Pollard had assigned to survey sites.
The ability to apply statistical tests and use a geodatabase for mapping and display allows
for a more detailed inquiry into the artifact distributions and subsequent patterns at
Tzintzuntzan.

Also of importance will be the previous and subsequent excavations done by the
Instituto Nacional de Antropologia e Historia, Mexico (INAH) at Tzintzuntzan. These
excavations were carried out at the main ritual platform at Tzintzuntzan, and are another
important resource that will be analyzed with respect to status and gender associated with
certain artifact variables. The combination of Pollard’s research, the INAH excavations,
and the ethnohistorical evidence provide for a complete study of the spatial distribution

of artifacts and the characteristics of an urban landscape in a Prehispanic state capital.

Previous Research

This thesis is the product of a pilot study done for a graduate level seminar
focused on statistical analysis and research design in archaeology. It was this study
where GIS was first utilized in creating a digital map of the Tzintzuntzan survey and was
used in exploring possible research ideas for the survey data. The project focused on
very specific ceramic variables that were indicative of social status markers, and when
statistically analyzed would help to isolate survey sites that would be indicative of higher
status function. A multivariate cluster analysis was used to analyze the distribution of
these elite variables across sites, and a polythetic classification was undertaken which

assigned the survey sites to clusters that varied based on variable frequency. The results

yielded very interesting patterns, with the survey site clusters falling within the elite
residential zones that Pollard first designated. By using the polythetic classification,
certain sites did stand out from the others based upon the variable’s presence or absence
and the frequency of the different variables. These results allowed for further questions
to be raised for this thesis, and were vital in exploring the data and helping to build a

research framework in which to begin preparing for this thesis.

Research Goals and Data

In the survey of Tzintzuntzan (see Figure 2 for survey map), Pollard collected
artifacts from areas of surface scatter across the survey area, an area that was defined by
both artifact density as well as geomorphology of the land. In most cases, the land was
highly eroded on the outer edges of the survey area creating a void of artifacts and thus
setting the survey boundaries. The collection includes n = 2152 ceramic artifacts
(including sherds, pipes, and other) as well as n = 10471ithic artifacts. During the
survey, Pollard focused on collection of ceramic sherds with rims and decoration, which
allowed for her analysis of the survey sites and the ability to create zonal categories based
on ceramic variables. Plain-ware pottery was collected when it was sensed that the
density of the ceramics was enough to define a site.

The primary unit of analysis is the sites that Pollard designated in her survey. In
her dissertation, Pollard defines the survey area by the artifacts found in the surface
survey along with the associated surface topography (Pollard 1972: 28). Within the
survey boundaries, “concentrations of artifacts were noted and located on the survey

maps”, and “all concentrations that were spatially isolable were given numbers and

considered sites” (Pollard 1973: 28). Because of funding and permit restriction, Pollard’s
survey sampling was non-random, made for the purpose of “isolation of temporal
variability (determination of the contemporaneity or lack of contemporaneity of the sites),
and the isolation of intra-settlement variability that might be reflective of economic,

social, or political differentiation within the settlement” (Pollard 1972: 190).
Figure 2 - The 1970 Tzintzuntzan Survey

 

 

 

 

 

 

4/ ‘ I . /./. . f/ I, - - - — ‘
’/ 7/7 w ., .,;:;-.’/; ;____; Survey Boundary
,’ ' .. "a / .' - Structures
.” 4/ ~ 1' Survey Sites
/.« ,_ , ,_ .,._,
.. ,. .” 59,, w. 1‘ Lake Patzcuaro
. Lil g , ' ".1, .
1 a, ‘1 . , ,jx, ‘. 20 Meter Contour
,, 1._.-.-_.. ' ,1 327,4, Surface Scatter
\j I“. "'7 ‘r
\‘ 'I
‘I I II
I ’I
I ,-‘ -----
0 0.5 1 2 Kilometers
l r r r l 1 r r l

 

From this point, Pollard undertook two classificatory procedures in order to
perform an intra-settlement analysis (Pollard 1972: 191). One procedure, called
analytical classification, involves isolating modes, which are selected attributes or
clusters of attributes that “display significance in their own right” (Pollard 1972: 191).
Examples of modes from Pollard’s work include slip type, paint, decoration and motifs,
and vessel form. The other procedure used by Pollard is the taxonomic classification
termed the type-variety system. This system involves the isolation of a hierarchically
organized series of units which reflect recurrent clustering of attributes (Pollard 1972:
192). Within this system, wares, groups, types and varieties are differentiated by the
sharing or lack of sharing of similar states of particular attributes. For example, wares
are distinguished by the attributes of paste composition and surface finish, types by the
attributes of a vessel from and decorative style, and varieties by minor variations in the
attributes of vessel form, decorative technique, or decorative style (Pollard 1972: 192).

This thesis only examines the ceramic artifact data from Pollard’s original survey.
It utilizes Pollard’s ceramic modes and wares that were derived from the type-variety
classification as variables for this analysis. These variables include slip, vessel form,
decoration, paste (also called ware), paint and supports. These variables have been
chosen because of their ability to isolate social status and function for each of the survey
sites depending on their presence or frequency. The spatial distribution of ceramic pipes
and pipe fragments from the survey are also analyzed. The pipe analysis will provide a
greater insight into the ritual and religious functions of sites and their relation to public

zones, an analogy made by Pollard in her use of the Relacion de Michoacan (1541) to

attribute firnction to pipes as being “used by priests and other members of the nobility in
association with religious ceremonies” (Pollard 1993: 165).

The outcome of these procedures is the ability to isolate ceramic variables that
display some distinguishable pattern that refers to a “zoning” behavior. Zoning,
according to Pollard, refers to the pattern of differential land use within a given area
(1972: 71). This differential land use is based on differential functions of particular
special units (Pollard 1972: 71). For analytic purposes, Pollard defines each site
according to zone characteristics, which are based on historic, ethnohistoric and
archaeological data (Pollard 1972: 73). The main source of this type of data is the
Relacio'n de Michoaccin (1541), which is an Early Colonial document, recorded in
Tzintzuntzan by the Spanish as dictated by Tarascan high priests. Within this, drawn
figures illsutrate food preparation occurring in what seems to be commoner housing by
people in commoner dress. The ceramic material indicative of this type of food
preparation, coarse wares in the form of shallow bowls or everted rim jars, and then
associated with what Pollard refers to as residential type Ri zones (Pollard 1972: 74).
From this, Pollard established three residential zones, ranging from Ri to Riv, Ri being
the low or commoner status, Rii being highest status or upper elite, Riii being an elite
status but not including the king, and Riv being a group defined solely on the presence of
Querenda White ceramics (1972: 80).

By using these status-sensitive ceramic variables, I more closely analyze the
division that Pollard associates with residential zones. .F or example, the Rii (upper elite)
and Riii (lower elite) zones are both considered elite zones, except that Pollard makes the

distinction that Rii is the highest status zone which includes the king and his family,

while Riii represents the “lower branches of the hi gh—status group in the social structure
of Tzintzuntzan” (Pollard 351: 2003). The difference in these groups is slight, where
residential iii “contains some, but not all of the lithic and ceramic artifacts found at the
high status (Rii) sites” (Pollard 353: 2003). Another zonal question that will be analyzed
was first found in the pilot study described previously. The study isolated what was
considered elite variables, one of them being the Querenda White ware. Analysis showed
the Querenda White ware was limited spatially, occurring only in the northeast portion of
the survey area. In Pollard’s dissertation, she believed this to be an ethnic barrio,
characterized by non-local ceramic production of Querenda White ware ceramics. Both
of these issues present an interesting research question that may be clarified with the
combination of additional ceramic variables in analysis. By utilizing GIS and
quantitative analysis, I analyze these zonal differences more carefully, and provide
statistical evidence that either supports or provides alternate evidence for the placement

of these residential zones.

Review of Literature

The majority of research literature that is used is from Pollard’s work and
publications. The main data resource for analysis is her dissertation. Other literature
includes Pollard’s 1993 book, T ariacuri ’s Legacy: The Prehispanic Tarascan State;
journal articles dealing with urban planning and city-state development both at the capital
of Tzintzuntzan and elsewhere in Mesoamerica; and Amy Hirshman’s 2003 dissertation,
A Case Study in the Relationship between Social Complexity and the Organization of

Ceramic Production from the Lake Patzcuaro Basin, Michoacan, Mexico. This last

resource aids in classifying certain ceramic types based upon variables that designate
them as elite or commoner ceramic types. Hirshman’s main concern was to distinguish
between utilitarian, market-oriented production versus specialized production intended
for elite consumption (2003: 84). Specifically, she discusses decoration, slip, finish, and
vessel form. Vessel form is a very common and wide ranging variable in ceramic
production. According to Hirshman, temporally we see a shift in the percentage of bowls
to jars in burial and non-burial context, with a higher percentage of bowls in burial
contexts, and a higher percent of j ars in a non-burial context (2003: 98). This type of
research sheds light on what the ceramic utility may have been, and with what class
group it is found.

The excavations carried out by INAH at the main platform at Tzintzuntzan are
also an integral part of this research. The excavations are detailed in publications of
individual field seasons as well as summaries of all nine seasons come from Acosta
(1939), Rubin de la Borbolla (1939, 1941, 1944), Castro Leal (1986), and Cabrera
(1987). The use of the INAH excavation reports from the main platform, which includes
elite burials, allow for the selection of specific ceramic variables can be utilized as
markers for elite status.

Gerald Migeon’s and Olivier Puaux’s dissertations and publications are also a
vital asset in providing analogous material in which to bridge the ceramic variables to
specific social status. Mi geon’s dissertation examined excavations within Tarascan
habitations at the site of Las Milpillas on the Malpais of Zacapu, Michoacan during the
Late Postclassic, a site located approximately 37 kilometers to the northwest of the

Patzcuaro Lake Basin. These resources aid in determining the material distinctions

10

between the commoner (Ri) and lower elite (Riii) classes, and also between lower elite
(Riii) and the upper elite (Rii). Of importance is the fact that the excavations by Migeon
have been described as commoner and lower to middle elite residences, lacking the upper
elite which were present at the adjacent site of El Palacio on the southern end of the
Malpais (Migeon 1990).

The dissertation from Olivier Puaux analyzes burials at a main pyramid at the site
of Las Milpillas. These lower elite and commoner burials will be important because they
can be compared to the upper elite burials at Tzintzuntzan. These comparisons will help
to differentiate between the social statuses and define distinct ceramic assemblages to be
used in the zonal classifications. Both Puaux’s and Mi geon’s research provide evidence
of social distinctions from outside of the Patzcuaro Lake Basin, and aid in the discussion
of T arascan urban traditions.

Michael Smith’s work on urban theory is also referred to when defining the urban
extent of the Tarascan capital of Tzintzuntzan. By using his updated concepts of zoning
and planning and setting them into the framework of Mesoamerican cities, Smith
provides the means for describing the range of urbanism that is present at Tzintzuntzan.
Also because of his detailed approach to ancient urbanism, space at Tzintzuntzan is more
closely examined thus providing for a thorough intra-site analysis of urbanism.

A final resource that is utilized is the Relacio’n de Michoaccin (1541), which is an
ethnographic account of the Spanish encounters with the Tarascan people. These Early
Colonial accounts and sketches give an invaluable aspect to understanding the life ways
of the Tarascan peOple in a pro-Hispanic setting. Pollard utilized this by analyzing

specific scenes from drawings that illlustrate specific ceramic usage associated with

11

certain events, such as a commoner meal or an elite burial. By using this resource, I am
able to understand the role of certain socially significant variables seen in the ceramics

collected by Pollard.

Research Methodology

One of the main components of this thesis is the ability to analyze data from a
GIS interactive mapping database. This requires the placement of the data into a
geospatial program that has the ability to spatially map and analyze the sites and their
associated attributes. For this study, I utilize ArcGIS TM (ArcMap TM, ArcCatalog TM,
ArcToolbox TM), which are Geographic Information System (GIS) products created by
the company ESRI. The first step is to convert the original map created by Pollard into a
TIFF file, which is an appropriate file system used in ArcMap. From there, the map is
overlaid upon the aerial photograph of the site of Tzintzuntzan, called a digital
orthophoto quad. Digital orthophoto quads are rectified images from which accurate
maps can be created. The aerial image is tied to specific, real world geographical points
(i.e. lat/long coordinates), and when the map is positioned as closely as possible, the
image is georeferenced, rectified and is then associated with specific geographic
coordinates. At that stage of consolidation, a new map must be made that resembles the
old map, but in this case attribute data can be directly tied to the new digital map and can
then be used in analysis. The map is created from a process called digitization, which
involves tracing over the old sites to create new ones in digital form. After each site is
created, it is given the associated site number, the frequency of artifacts, the types of

artifacts, and any other relevant attribute data that could be used for analysis. In this

12

case, the sites are given the number of observations for each variable that was just
discussed. This allows a query of data that is directly displayed through the map just
created, and is invaluable in data display and spatial analysis.

The unit of analysis is the site and the variables include any ceramic variables that
give further insight into form and function of the site they are contained in. With that
said, the primary statistical approach to be used is a cluster analysis. The cluster analyses
used are a multivariate statistical procedure that starts with a data set containing
information about a sample of entities and attempts to reorganize these entities into
relatively homogeneous groups (Hirshman 2003: 113). The sites are run through an
agglomerative, hierarchical cluster analysis with a polythetic classification, where
“multiple variables are taken into account to form the groups with no one variable
sufficient or necessary to group membership (Hirshman 2003: 113). The hierarchical
cluster used is based on a between group cluster that uses the average linkage between
groups. Average linking is a “space conserving algorithm that defines distance between
groups as the average of the distance between all pairs of individuals in the two groups”
(Hirshman 2003: 117).

The usefulness of the methodology I have laid out is that it provides bridging
arguments that isolate ceramic variables that represent specific status groups, and to then
take these variables and group them accordingly. So for example, variables will be
combined, such as specific slips, decoration, paste, and will all share a common link in
that they are associated with a specific status group. The sites will then be run through an
agglomerative, hierarchical cluster analysis based upon these status variables in order to

aid in placing sites into zonal categories. Alternatively, a monothetic classification may

13

be used with the specific variables, thus creating a much clearer spatial assignment of the
sites based upon variables that must be present in order to be placed into a specific group
based upon zonal characteristics. Depending upon the number of the clusters that the
analysis returns, an additional K-means cluster analysis may be performed to deal with
outlier sites. The K-means cluster analysis is based on the hierarchical cluster, but refines
the results as the user adjusts the subjective number of output clusters required, moving
the cases that were initially miss-assigned in the one-pass hierarchical results (Hirshman
2003:120)

Once the sites have been assigned to zones based upon the clustering of their
ceramic variables, a non-parametric statistical test is carried out that compares my zonal
results to Pollard’s original zonal results. Because the data are ranked, the test will be a
median based statistical test, as opposed to a mean based test. The test will be a
Wilcoxon signed-rank test, which will compare the differences between two
measurements on the same sample. I will also run a sign test, which will analyze the
directionality of the measurements on the sample using a nominal scale. These statistical
tests should display any significant differences in my zonal assignments versus Pollard’s,
thus testing my null hypothesis that there will be no difference between my zonal

classifications and Pollard’s.

Predictions
I predict that my results will not differ from Pollard’s original results. Because of
the quantitative analysis, certain patterns may be shown more clearly when determining

site function and status. I believe that if any sites do not conform to Pollard’s original

14

results, it is because the ability to run statistical tests may reveal site clusters that pattern
in ways that are difficult to determine in the manner that Pollard originally did her
analysis. By being able to run a variety of cluster analyses using both monothetic and
polythetic groupings, it is also hoped that the variables will provide site patterning that
may better define what Pollard calls a Rii, upper elite residential zone. These sites vary
little from the Riii (lower elite), and I believe with the planned analysis the sites will not
differ substantively from Pollard’s original zonal classifications, but still may provide
smaller-scale patterns that better explain the statistical differences between Rii sites

(upper elite) and Riii sites (lower elite).

Significance of Research

One aspect of this research hopes to add to the growing literature on integrating
GIS with archaeological research. Even though a limited analysis is done using GIS in
this paper, it still displays the capability of GIS as an interactive database where
information can be stored, manipulated, analyzed and displayed. GIS is used in this
thesis to create an interactive site map, one where the ceramic variables are tied to the
sites and thus linked to a database capable of running statistical and spatial analysis. The
use of GIS will also bring about specific questions dealing with spatial data that has been
collected through sampling. The statistical and display advantages of a GIS will help in a
better understanding of how a sampling strategy affects the interpretation of the data.

The primary goal of this thesis is to provide a quantitative analysis to the
archaeological survey data from Tzintzuntzan. This will allow for a more comprehensive

view of the role of urbanism at Tzintzuntzan, and will shed light on the internal structure

15

of the social hierarchy of the Tarascan capital. Hopefully, this will convey a better
understanding of the archaeological landscape and the use of space at Tzintzuntzan, and
may aid in the discourse centered on Pre-Hisparric, state-level social theory. The hope is
to provide a methodology that utilizes multiple lines of evidence and statistical analyses
capable of working with small sample sizes, and thus providing results with a higher
degree of confidence than was previously thought. This may result in a framework that
can be tested elsewhere in the basin and aid in the understanding of how we might

interpret concepts of social status from the archaeological record.

16

Chapter 2
Research Methodology

This chapter presents the methodology used in this analysis. Data and results
from previous research that aids in the classification of ceramic variables into distinct
groups based upon social identity within the framework of the socially stratified Tarascan
capital are presented. The data that structures the selection of variables comes from (1)
analyses of ceramic production and the economy of the Tarascan state based on
excavation and survey from the sites of Urichu and Erongaricuaro, (2) mortuary and
household excavations at Tzintzuntzan and Urichu in the Patzcuaro Lake Basin, and the
Tarascan site of Milpillas in the Zacapu Basin, (3) and ethnohistoric data from the
Relacion de Michoacan. All of these provide evidence for social stratification within the
Tarascan Empire, but more importantly provide material evidence of social classes
including ceramic goods. By linking historical and archaeological evidence
representative of material culture to social status, I create bridging arguments for the

association of ceramic variables, such as decoration and vessel form, into social classes.

Ceramic Production

Crucial to any analyses of material goods is the knowledge and understanding of
their contexts for production and utilization. In the case of Tzintzuntzan, ceramic
production is closely aligned with the presence of a socially stratified community, in that
production changes relative to the organization of labor and emerging economies. In her
dissertation, Amy Hirshman looks closely at the relationship between the emergence of

the Tarascan state and the organization of ceramic production. By using the

17

“Standardization Hypothesis” within the context of the social evolutionary model and the
political economy model, Hirshman “assesses changes in the ceramic assemblage
reflective of larger changes in organization of ceramic production” (Hirshman 2003:
abstract). Hirshman’s ceramic assemblage consists of 37,301 sherds from survey and
excavations from the lower southwest portion of the Lake Patzcuaro basin, including the
sites of Urichu, Pareo, and Jaracuaro (2003 :5).

Of importance to this analysis is Hirshman’s interpretation of ceramic
specialization within the Tarascan economy. Craft specialization in terms of the political
economy model could be considered a function of the upper class within the state,
sponsored by the elite for purposes of control within the economic and political realms
(Earle 2002: 145). Specialization may also exist in producing for a general market for
commoners, where specialist producers exist “because of gains in efficiency resulting
from underlying economic conditions” (Earle 2002:145). In either case, specialization is
measured by means of standardization as defined by the “standardization hypothesis”,
which “maintains that the product of specialized producers will exhibit less variation than
the products of non-specialized producers” (Hirshman 2003 :3). To determine the
specialized level of a ceramic product, Hirshman discusses ceramic variables in terms of
efficiency, energy, routinization of task, and labor investment (20033).

As discussed in the introduction I have chosen to focus on ceramic modes rather
than work wholly within the type-variety system of the region. To better explain, a mode
is a standard or concept that evokes a behavior within, for example, a community of
ceramic producers, and displays significance based on its own right (Hirshman 2003:

102). This technique was taken from Hirshman’s analysis, as she too found it easier to

18

use ceramic modes in explaining social status within the organization of ceramic
production. For example, she notes that the modes slip, paint and finish are likely
indicators of labor investment (2003: 135).

Slip is one of the variables I have chosen for this analysis. There are five colors
of slips present within the ceramic assemblage at Tzintzuntzan; red, cream, pink, grey,
and white. Of these Hirshman notes that throughout the temporal phases, red and cream
remain the most important slip colors. This is true in the survey collection from
Tzintzuntzan, with cream slip as 68% and red slip 24% of the total slips present. With
such large percentages, it is clear that both slips range across a variety of ceramic types,
including both coarse and fine wares, and cannot be expected to be limited to any specific
social class. Likewise, we cannot assume that a specific ceramic variable is only found
on elite ceramics based solely on that variable’s low frequency within the assemblage,
although Hirshman does notice that the white slip becomes “more common” over time,
yet still doesn’t become significant within the assemblage (2003:135). At Tzintzuntzan,
white slip is 5% of the total slips from the survey collection. It may be that the increasing
frequency of white slip reflects a need for the growing elite class within the Patzcuaro
Lake Basin. This may explain the slip’s “more common” presence, yet also why it never
is viewed as a common ceramic mode within the assemblage. This of course is mere
speculation, but may provide useful insight when analyzing data from the excavations of
burials and habitations in the Late Postclassic.

Paint may be a more useful variable in determining social status in ceramic
sherds. Hirshman notes the absence of paint dominating in all phases of the temporal

sequence of the Patzcuaro Lake Basin (20032135). Paint colors do occur in all phases,

19

with either one color being used alone, in pairs or in threes on any given sherd, the latter
being rare. Hirshman sees the use of three paint colors as significant, as “three paints
require greater labor, and the only incidence of three paints occurs in the Late Urichu and
Tariacuri Phase” (2003: 137). When discussing paint color, Hirshman notes that the most
common color is red, and the most common two-paint combination is red and white. It is
clear that paint combinations may give more insight into labor investment in ceramic
production, doing so on a continuum of high occurrence to low occurrence in both paint
number and color. Based on the aspects of ceramic production discussed here, we can
begin to suspect that more decorated ceramics, with regard to paint, can be correlated
with higher social classes.

The use of resist, also called negative decoration, is another decorative technique
that Hirshman discusses within Tarascan ceramic production. Resist is a technique where
wax is applied to a previously fired vessel, and then fired again in a reducing atmosphere.
During the firing the wax protects the design underneath and eventually melts away thus
creating a black, smudged pattern where the wax wasn’t applied and leaving intact the
protected original design (Hirshman 2003: 141). Of course this technique is time
consuming as well as requiring a higher cost for labor and a second firing. Hirshman
notes several things about the resist patterns that must be examined more closely within
this analysis. First, she notes that a low percentage of sherds exhibit resist in the
Tariacuri phase. She also notes that resist occurs primarily in areas that have been
designated as commoner contexts (2003: 141). Hirshman also notes that there was no
commoner substitute for the resist technique that produced a similar decorative style,

such as the use of black paint to mimic the resist look (2003: 141). So based upon the

20

labor cost and specialization needed to produce ceramic with resist, one might assume it
is a variable that belongs to an elite ceramic assemblage. On the other hand, I-Iirshman’s
findings also place the resist patterns in commoner contexts, not with elites. These issues
will be kept in mind as the analysis continues within the Tzintzuntzan ceramic

assemblage.

Excavations in the Zacapu Basin

To the immediate northwest of the Patzcuaro Lake Basin are the archaeological
sites of the Zacapu region. These sites represent another social center within the
Tarascan state containing similar organization and social stratification to that of
Tzintzuntzan. The dissertations of Gerald Mi geon and Olivier Puaux describe
excavations of structures and burials at the site of Mich. 95 (Las Milpillas) within the
Malpais of Zacapu. The excavations at Mich. 95 took place in 1984 and 1985, and
focused on the smaller area of Group B of the Milpillas site (Migeon 2003; 104). The
sites range in temporal phase from La Joya, Palacio, and Milpillas. This thesis will be
concerned only with the Palacio and Milpillas phases, or the Postclassic phasesz.

Migeon’s dissertation describes excavations that were carried out within and
around 40 structures3 (see Figure 3) (Migeon 2003: 104). These structures represent both
residential and public buildings, and the in situ excavations will be used to link certain

ceramic types to areas of differential social fiinction and status within the Group B area.

 

2 Migeon doesn’t readily distinguish between phases within the excavations of the structures. The types are
given for the artifacts, but aren’t further categorized into specific time periods. Thus, the artifacts range
from the Palacio and Milpillas phases, the Palacio being the Early Postclassic, the Milpillas being the Late
Postclassic.

3 In this case, “structures” refer to a variety of architectural features, from the level pyramid platforms
(vacatas) to commoner habitation. For the purpose of this paper, I am primarily concerned with habitation
structures, but will also look at ceremonial and communal structures as well.

21

Figure 3 — The Site of Las Milpillas

 

Mich. 95, Group B: --

:1 B38 D837 \s‘\ (”I Group IV ‘\I
B1;J E1515 K“

 

 

 

- Commoner Structures

 

Las Milpillas, Zacapu
Group V B43 \\
.- "l B35 92 ‘1
. 333 ,'
”’1 334 _-/
,1 L, a

 

.Cl . _ ,
\839 Group III' ‘. . [:l. .
‘. . Z i 13 814 3.17 '
'. . E1 ‘. ‘ :
‘ ,' (’1330
. L4 , -
/l “
.' ', _/\.‘329
' x/ .
. B11
. 322. ,' D ‘.
.‘ 820 )1, [1 e12 :
‘, .’ B10 '
. I I
Z - r l
c K, ‘ ‘
, .’ 321 323 'r
’ {I [“1 Group '"9 87 “ D , Group Vl .‘
, X. l 5' ' BE .l' i . 0 20m
Legend D345 ,.
:-_____-_i Groups 628 '
[:l Sites 08?? ’
E ' ' 7.73. Elite Structures I i _______ , Y1
O x”
(x

 

As taken from Michelet, 1998

Also of importance is the fact that during the Late Postclassic, it is known from

ethnohi storic documents that the upper elite were present at the adjacent srte of E1

Palacio as it was the administrative, political, and religious center in the immediate

region during the Late Postclassic (Migeon 1998: 43). Thus, Mich. 95 has been

22

described as an area with a lesser range of social stratification, consisting of the smaller,
plain dwellings of the commoners to the grande maisons (great houses) and centre
civico-ceremoniel (civic ceremonial centers) of the intermediate and administrative elite
of the district (Migeon 1989: 490-91). These will be used to construct material
associations between what Pollard defines as the Riii zones, occupied by the lower elite,
lacking the highest elites defined primarily by the royal family, and the Ri zones, defined
as the commoner residential zones.

Of significance is how Migeon describes the range of variation within the artifact
assemblage of the Group B structures, stating that even though a large range in structure
size exists, the artifact assemblage is fairly homogeneous throughout (19891418). This
gives us some insight into a possible relationship between social classes. The existence
of a relatively homogenous artifact assemblage at the Zacapu site of Milpillas suggests
that the lower elites are actually closer to the commoners, at least in a material sense, than
they are to the upper elites at E1 Palacio. This may provide a useful avenue of analysis to
pursue when looking closer at the class distinction proposed at Tzintzuntzan. One thing
that should be kept in mind are the differences in location between the two Tarascan
centers; where El Palacio is physically removed from Mich. 95 creating somewhat of a
spatial disconnect, the social stratification being analyzed at Tzintzuntzan consists of
shared urban space creating more of a continuum of social classes, both in a physical and
ideological sense.

Migeon describes in detail the size, architecture and surrounding features (such as
proximity to ceremonial centers and pyramids) of each structure. Utilizing the

ethnohistoric information from the Relacion de Michoacan, Mi geon distinguishes

23

between the structures excavated in Group B based upon their size and the interior
arrangement of hearths, walls and altars (1989; 418-21). Structure size and relative
distance from what Mi geon terms the centre civico-ceremoniel (ceremonial/civic center)
is the primary means through which he determines the possible function, and more
importantly social status, pertaining to each structure. He does pair this with the artifacts
excavated in each structure, but his final conclusions for each structure’s social
significance are more inferences, and not based on solid analytical methods. Therefore I
will be looking primarily at what Mi geon confidently asserts are structures of this
intermediate elite, and the commoner residences, with only a few cases in each category.
In certain cases Migeon describes in detail the artifacts found for each structure by lot
(our equivalent of an excavation unit and level), and in other cases they are described in
less detailed terms by summarizing each structure’s artifact assemblage. I will be
combining the two forms in chart format to gain a better understanding of the artifacts
found within each structure.

Mi geon, although somewhat uncertain or unclear about most of his inferences
about social status pertaining to each structure, does distinguish between several
structures that he claims are representative of a specific social class within Group B.
What he classifies as the political elite of the district are present at structures B3, B4, B6-
B9, and B20 (Migeon 1989: 491). According to Migeon, these structures are the primary
elite habitations in the area and are closely tied to ritual and tribute, as they are (l) in a
central location in terms of the site, (2) close to the ritual structures, and (3) in close

proximity to storage structures4 as well as being the largest habitation structures in size5

 

4 What were originally defined as ritual altars and hearths in group B by Migeon and Dominique Michelet
have since been identified as storage structures, called Maritas (Pollard 2005). These storage structures

24

(1989: 491-92). Because Migeon states that structures B3 and B6 are the best examples
of the “big houses” and thus are the best representative of elite habitation in Group B,
only the artifacts from these two structures will be analyzed for the elite category. The
commoner residences are located near the edges of Group B, defined by the smallest
habitation structures. Many of these are referred to as Migeon as “plain dwellings”
(1989: 484). Of the commoner structures, B34, B38, B43, and B44 have been randomly
selected and are those whose artifacts will be analyzed.

Mi geon describes the pottery primarily by vessel form and by ceramic complex
for each structure. The three ceramic complexes, or what are actually groups in the type-
variety system, that are used in Mi geon’s analysis consist of Zacapu Grossier (Zacapu
coarseware), Milpillas Poli (Milpillas polished), and the Malpais a Blane Laiteux Poli
(Malpais with white-milky polish). Each of these groups are determined primarily by slip
and finish; thus the Zacapu group is a coarseware and the primary plainware ceramic; the
Milpillas group is a slipped but non-painted fineware; and the Malpais group is a “milky
white” (white or cream) slipped decorated fineware (Mi geon 1989; 431-432).
Unfortunately, this is the most detailed that Mi geon is when describing the ceramics.
Decoration, paint and slip are variables that are lacking throughout his analysis, except in
the general type descriptions of the three complexes above.

The ceramic artifacts from both the elite and commoner structures may first be
analyzed quantitatively, and are summarized in Table 1 below. Structures B3 and B6

have much higher counts than any combination of commoner structures. The two sites

 

have been depicted in the Relacion de Michoacan as well as the Relacion de Tiripitio (Michelet 2005,
Pollard 2005).

5 In the case of the “Grande Maisons”. which are structures B3 and B6, they are 10 x 10 meters in size with
an interior size of 70 to 90 m2.

25

combined have a total of n=205 sherds of j ars from the Zacapu group, n=24 bowl sherds

from the Zacapu group, and n=21 Zacapu group supports. The Milpillas group yielded

n=471 bowl sherds, the dominant vessel forrrr for this type. The Malpais group yielded

n=3O bowl sherds, and n=34 supports. Also found in all three groups from the elite

structures were miniature vessels (jars, vases, bowls). Their numbers are low, eight in

total from both B3 and B6. The commoner structures present quite a different artifact

assemblage when compared to the two elite structures. The most common artifact found

throughout all four commoner structures is the Zacapu jar. Only eight bowl fragments

were found in the Zacapu group from all four structures. The Milpillas group yielded

Table 1 - Ceramic Assemblage from Mich. 95, Group B

 

 

 

 

 

Elite Structures Commoner Structures
B3 B6 Total B35 BB8 B43 B44 Total
93 jars .
’ 112 jars, .
:znsors 27 221.123,
14 bowi “"5”: sherds 10 . jars . . .
10 bowl ’ jars, l ’ jars, jars, jars, 1
sherds, 1 tecomate, censors, . .
sherds, 4 censor, censor, censor, mini,
°' comale, 4 comales 5 comales, 2 bowl 3 bowl 1 bowl 2 bowl 8 bowl
5 mini,2 14 ’ 21 hrd sherds, hrd h ds sh d
(D mol, 7 rts supports, 4 s e s lmini S e S er er s
:1 suppo , . -
:1. supports, 7 mints, 2
g fewmales tecomates molcajetes
141 bowl 471 bowl
sherds, 7 332 bowl sherds, 9
it jars, 6 sherds, 2 jars, 3 1a: 7 34
g mol,1 jars, 3 tecomates, bciwi 11 bowl 9bowl 7 bowl bowl
3 mini, 3 mol, 2 3 sherds sherds sherds sherds,
__ . . . sherds -
E. supports, mmr, 2 molcajetes, 1 jar
E l tecomates 3 minis, 3
" tecomate supports
3 jars 8 24 bOWl 21:23:13.4
s ,rts sherds, l su fts 5 l bowl 1 bowl
.g uppo ’ mini, 26 . ppo ’ none 2 jars none sherd,
o. g. 6 bowl supports jars, 1 sherd 2jars
"at 8 sherds 2 'ars ’ mini,1
2 0 J tecomate

 

 

 

 

 

 

 

 

 

 

 

26

 

n=34 bowl sherds and one jar, whereas the Malpais group was limited to one bowl sherd
and two jars between the four sites. Also, out of the four commoner structures, only one
miniature vessel fragment was found.

The ceramic assemblage from both the elite and commoner structures exhibit
interesting patterns that may be associated with the artifacts at Tzintzuntzan. The
dichotomies between the two social classes as seen in the ceramics are present both
qualitatively and quantitatively. It is clear that the elite structures had a much higher
frequency of ceramics than the commoner structures, especially of the Milpillas and
Malpais finewares. The presence and consumption of these finewares is an
overwhelming elite phenomenon, as we may have expected. The Malpais group indicates
this fact, as the four commoner structures only had three Malpais group ceramics present,
whereas the two elite structures had n=71 combined ceramic sherds.

The final association that can be made by looking at the ceramics from Group B
at the Milpillas site is the vessel form. The two vessel forms that are most representative
of the commoner structures are the Zacapu jars and the Milpillas bowls, although at a
lower count. The Zacapu jars are the plain ware storage vessels that are most common
throughout the Tarascan state. Used for water and food storage and cooking, it is the
most basic utility vessel and one most common in commoner households. The Milpillas
bowls are a fine ware ceramic that is undecorated and lacks supports. We know this
because both commoner and elite structures had high counts (relative to the overall
artifact assemblage size) and out of n=512 Milpillas bowl sherds from the elite and
commoner structures, only three Milpillas supports were recorded. The Milpillas bowl is

also a utility vessel, used as both cover lids for jars and serving vessels. The reason there

27

is such a high number of these from the elite structures may be due to the fact that they

are being given, along with their contents, as tribute from the commoners to the elites in

Group B. They also may represent multiple family households, with multiple wives and

more surviving children. Also, the Milpillas group is not very representative of jars, a

vessel form that may be most associated with the plain ware ceramic types. The final

type, the Malpais decorated fine ware, is an almost entirely elite ceramic type. Because

of the high support count in the elite structures from this type, we may infer that the

primary vessel for this ceramic type is the decorated tripod bowl, a vessel form not found

in the commoner context. Below, Table 2 shows the vessel forms from the elite and

commoner habitations, exemplifying the phenomenon of open versus closed vessels as

representing a distinction both between the ceramic groups and within the social classes.

Table 2 — Jars: Bowls: Miniature Vessel Ratios from Mich. 95, Group B

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Elite Structures Commoner Structures
B3 B6 Total B35 B38 B43 B44 Total
~ . - . 9 9

93 jars. 112jars. .205. . . ?Jars: 3 . . . ? Jars:
Zacapu 19 17 jars. 36 Jars. bowls: 1 ? Jars. Jars. 8 bowl"
Group bowls:4 bowls: O bowl: 3 2 . . ' lbowl 2 . .'

. . . . . . mm 1 mrnr

mrnr mrnr mrnrs bowl bowl

7 jars: 2 jars: 9 jars: . . . . .
Milpillas 149 336 485 $1”: 0 jar: 11 0 jar: (7)1“ if"
Group bowl: 1 bowl: 2 bowl: 3 bowl bowl 9 bowl bowl bowl

mm mm mm

. . 2 jars: 5 jars: . 2 jars: .
Malpars 3 jars: 6 24 bowl: 30 none 0 jar: 1 0 ' none 2 jars:
Group bowl . . ‘ bowl: 1 bowl 1 bow
l mrnr mini bowls

 

28

 

Although Mi geon (1989) once again is lacking in specific descriptions, he does
provide a count of painted6 sherds found in each structure. The elite structures of B3 and
B6 have the highest total of painted sherds, a combined n=267. The three intermediate
structures only have n=87 painted sherds, and between the nine commoner structures
there are n=74 painted sherds (Migeon 1989; 194-222). The problem is that in relative
percentages, the painted sherds only make up 2% of the total sherds from B3 and B6,
1.3% from the intermediate structures, and 2.7% from the commoner structures (Mi geon
1989; 194-222). Merely explaining the sherds as painted gives no insight into labor
investment, as Hirshman explained, and thus no insight into inferences of status. As the
painted sherds seem homogenous in distribution throughout the structures, a
presence/absence or hi gh/low frequency also doesn’t reveal any patterns into social
status. The one thing that was found was the presence of negative (resist) decoration on
one sherd in structure B38. No such sherds were found in any other context, and
although only one case was found it may aid in directing future analysis.

The dissertation by Olivier Puaux describes the burials excavated within Group B,
and supplements Mi geon’s dissertation by providing more distinct associations between
certain ceramic types and social status. Puaux completed the dissertation on the burials
in Group B of Mich. 95 in 1989, with the work at Group B having been carried out from
1983 to 1987. The dissertation analyzed n=35 burials with a total of n=67 individual
human remains (Puaux 19892110). The important aspects for this thesis are the artifacts
found in association with the burials, and the subsequent interpretations made by Puaux

that may indicate the social status of the buried individuals. Migeon’s work has already

 

6 Paint, in the unclear sense of the term as presented in Migeon’s dissertation, has no relevance when
discussing decoration, except in the one instance of a negative decoration

29

discussed social statuses associated with certain structures, and combining his work with
Puaux’s should give insight into status and ceramic associations.

When discussing the burials, Puaux cites several different variables that may
reflect the individual’s social status. Puaux includes the (1) burial structure, (2) method
of deposit, (3) the orientation and position of the body, (4) the location of the burial in
relation within the cemetery, (5) whether it was an individual, group, primary or
secondary burial, and (6) the nature of the burial goods (Puaux 1989:213-214). He goes
on to state that cremation is the most exclusive category of fimerary methods, one
reserved for the highest social class including the king7 (Puaux 19892213). Since the
Tzintzuntzan survey did not include burials, less attention will be paid to these variables,
while more will be paid to the gender, age and artifacts of the burial.

Puaux also discusses artifact categories that may better elucidate the gender
distinctions and status roles at Las Milpillas, such as the presence of pipes and metal
artifacts among the burial offerings. Puaux states that pipes are found in the context of
adult males, and based on the Relacion de Michoaccin, are tied primarily to ritual and
ceremonial use (1989:177). Metal is the other artifact discussed in detail by Puaux. As
he states, it is known from the ethnohistoric documents that metal was a prized
commodity for the political and religious elites. Within the burials at the Milpillas site,
Puaux notes four instances of metal offerings, primarily related to female burials, with
one instance of an infant burial (1989:189). We can now assume (because of its scarcity
and the ethnohistoric evidence) that burials associated with metal may also be interpreted

as higher status (Pollard 1987; Pollard & Cahue 1999).

 

7 This based on the description of the funeral of the Tarascan king described in the Relacion de Afichoacdn
( 1956).

30

The burials at the B1 pyramid are located on its western facade and are grouped
by Puaux in three sections, the south-west, west, and north-west areas (refer to Figure 1
for a map of the burials). From the burials at structure Bl in Group B, 24 of the 35 burials
were associated with artifacts, and out of the 67 individuals from the 35 burials at BI, 29
had artifacts associated with them. There were also two burials of three individuals each
uncovered near the structure of B14, both of which had artifacts. Of the 32 individuals
(from both B1 and B14) whose burials were associated with artifacts, 13 were male, only
two of which were young males. The remainder were mature, or approximately thirty
years of age or older. There were also ten females associated with artifacts, five of whom
were mature, the other five being young adults. The remaining six individuals were
adolescent or infants, whose gender was undetermined. There were also three
individuals, one young adult and two mature adults whose gender could not be identified.
Table 3 displays each burial, the gender, age, and associated artifacts.

The artifacts from these burials are very sparse, containing a total of only seven
decorated ceramics with the rest being undecorateds. These seven decorated ceramics
occur among five individuals, two female, two male and one undetermined. All except
one, a young adult, are mature adults. The scarcity of artifacts from the burials seems to
convey an understanding of the divide between commoner and lower elite within Group
B at Mich. 95. Even among the burials where artifacts are found, only six were found

with artifact counts higher than threeg.

 

8 Decoration is believed to be paint motifs, plastic manipulation, or negative decoration Puaux is
somewhat vague as to his definition of decoration, thus this analysis must rely on a presence/absence
classification

9 This does not include the large funerary urns, as they are considered a burial container and not an
offering.

31

Table 3 — Burials with Age, Gender, and Artifacts from Mich. 95, Group B

 

 

 

 

 

Burial # n-2 n-4a n-4c n-4e n-4g n-4h
young mature old young mature
Jig; adult adult adult adult adult adult
Gender female female male female male ?
tripod
mini vase
mono- jar large urn, (2),
chrome (2), basin, open decorated
Artifacts minijar mini jar basin mini jar bowls minijar

 

 

 

 

 

 

 

 

 

 

 

 

Burial# n-5a n-5b n-8 n-9 n-12 n-19
mature mature young mature
Age child adult infant adult adult adult
Gender ? male ? male ? male
large
large um,
um, decorated large metal,

Artifacts basin mini jar jar jar, basin basin basin

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Burial # n-ZOa n-21 n-22 n-23 n-24a n-25
mature young old young mature young
_Age adult adult adult adult adult adult
Gender female female male female male female
open
bowl, large decorated
mini urn, tripod,
jar, mini jar, open large urn, mini
Artifacts basin metal (2) bowl jar bowl basin
Burial # n-26a n-26b n-27 n-28 n-29 n-31
old old young mature mature
_Age adult adult adult adult adult adult
Gender female female ? ? female male
mini
basin, decorated vase,
large mini vase, projectile
Artifacts jar jar pipe um large urn points

 

 

 

Burial# n-SZa n-33 n-34a 814 n-1 B14 n-2
mature mature infants mature

 

 

 

 

 

 

 

 

 

 

_Age infant adult adult (2) adult
Gender ? male female ? male

decorated

tetrapod,

large vase, mini

um, jar, mini

copper pipes mini obsidian bowl w/

Artifacts ring (2) vase frags. supports

 

32

It seems reasonable to suggest that these small percentages of burials are those
individuals who would have had an elite status in the community at Group B. Also
associated with this elite status are those burials which included decorated ceramics and
metal artifacts were found. Based upon these lines of evidence, we may conclude that
miniature vessels, and vessels with supports (tripod and tetrapod) are associated with the
middle elite of Group B. The large fimerary urns and the basins10 are somewhat difficult
to comment on in terms of status association. In terms of the burials, both vessel forms
are found in the context of the high status urn burials and are associated with elite
individuals. This fact seems only true in the context of burial and funerary ritual where
they are a mode of burial, with the remains placed inside the um and the basin used as a
cover for that urn. For this reason, the urns and the basins will be disregarded as artifacts
that may be used in association with social status.

The final artifacts to make note of from Puaux’s burial interpretations are the
pipes found in the burials. Three pipes were found in burials 26, 27 and 33, respectively.
Of the three burials, both 26 and 33 are adult male individuals, whereas the remains from
burial 27 are incomplete and the sex cannot be determined. Although this is a small
sample size, we may begin to infer that along with a more elite status symbol pipes are
associated with adult males within the Tarascan culture. The fact that there is such a
small sample of pipes may be indicative of the social class of these individuals, and that
they are in fact of lower elite status.

The burials from Group B in Mich. 95 seem to represent a sample of the artifact

assemblage from the commoner and middle elite within the Tarascan state. Puaux’s

 

‘0 A basin, in this instance, refers to a deep bowl, also used as lids for funerary ums.

33

research deals mostly with vessel form and the presence and absence of decoration“.
The evidence from the burials place miniature vessels and the vessels with supports
(tripod and tetrapod) in contexts of more elite individuals. This is supported by the type
and mode of the burial, where the more elite individuals are those found in primary
burials in separate cavities, or are cremation burials found in urns. In the case of the two
burials found in structure B 14, we can assume an elite status based on the fact that B14 is
described by Migeon as a ritual area, thus associated with an upper class as well as ritual
significance.

Based upon the mortuary and household excavations at Mich. 95 (Milpillas)
within Group B, the following class associations can be established: Within the
commoner assemblage, the most common ceramic artifacts are the Zacapu coarseware,
which is dominated in vessel form by the larger everted rim jar. The commoner ceramics
remain predominantly unslipped and undecorated, with the few exceptions of the I
Milpillas polished ceramics, which are slipped but not decorated. Decorated
polychromes are almost absent from the commoner ceramic assemblage, as are the vessel
forms that usually accompany them including the tripod vessels, miniature vessels, and
plates. Closed vessels, used for storage and cooking, dominate the commoner ceramic
vessel form.

The lower elite at Milpillas are defined by the presence of the decorated
polychromes, but in low frequencies. The burials exhibit miniature and tripod vessel

forms, but once again in low frequencies. The majority of the lower elite ceramic

 

‘1 Decoration, according to Puaux, is plastic manipulation as well as certain paint styles and color
combinations. although decorative motifs are not discussed. For the artifacts found in the burials. specific
description of the decoration is lacking, thus creating a presence/absence analysis rather than a detailed
descriptive analysis.

34

assemblage occurs with the Milpillas polished ceramics, which are slipped and polished
yet not decorated. The main difference between the lower elite and the commoner
ceramics is in quantity, with the lower elite having much higher frequencies of all three
ceramic groups. The vessel forms within the lower elite assemblage are a combination of
open and closed vessels, with the Milpillas polished bowls having the highest frequency.
Miniature vessels are also a distinct form in the burials, along with the presence of pipes
which aren’t found in commoner contexts. It is clear that the lower elite seem to exhibit
aspects of both the commoner and upper elite ceramic assemblages, but with quantitative

differences defining their assemblage.

Excavations and Collections at T zintzuntzan

Beginning in the early 193 0’5, excavations were undertaken at Tzintzuntzan
through the Instituto Nacional de Antropologia e Historia, Mexico (INAH). The work at
Tzintzuntzan spans nine field seasons, consists of ritual, mortuary and habitation
excavations, and ranges in time from the 1930’s to the 1960’s. The area of focus of all
the excavations was the main platform located on the northwest slopes of Cerro
Yaguarato. This platform, called Tz-25 by Pollard, is what constitutes the Zona
Arquelogica as defined by INAH. The main platform consists of five ydcatas (pyramids)
which, according to the RM, were the main ritual zone of the area as well as being an
area of burial for high status individuals (Pollard 1993 :47). Excavations were also
carried out at habitation structures on and adjacent to the platform. These projects and
their subsequent articles and reports are the main focus of this section. Examining the

artifacts found in the context of a highly elite and highly ritual area such as the main

35

platform will allow for bridging arguments to be made as to the nature of the ceramic
artifact assemblage of the upper elite at Tzintzuntzan.

Seasons I and II of excavations at the main platform are described by Acosta and
Rubin de la Borbolla. The first season included habitation and mortuary excavations at
and around yacatas 2 and 5, as well as edificios (habitation structures) A and B. Acosta
goes into great detail about the architecture of the yacatas and ruins on the main platform,
and mentions briefly the ceramic artifacts found. He states that they are of classic
Tarascan typology, and are from the ultima epoca, or what can be translated as the Late
Postclassic (1939:97). The ceramics included decoration of white and black on red,
negative decoration, figurines made from molds, and decorated pipes (Acosta 1939297).
This very brief mention of the ceramics is due in part to the small amount of sherds
found, and also because Rubin de la Borbolla was going to complete a more in depth
analysis later on.

The second aspect of the excavations at the main platform was the burial
excavations described by Rubin de la Borbolla. These excavations took place during the
second season, and are detailed by Rubin de la Borbolla in his 1939 publication. There
were five recorded burial areas, or what Rubin de la Borbolla refers to as entierros. Four
of these were located in pozos12 1 or 2 of edificio B, while the last was defined as the
“ossuary of Tzintzuntzan”, located to the north of edificio B (1939:111). The majority of
Rubin de la Borbolla’s analysis is the mortuary and forensic analysis of the burials, with
brief descriptions and illustrations of the ceramics found in both pozos 1 and 2. The

ceramic assemblage associated with these burials consists of a variety of polychrome

 

‘3 P020, in this instance, refers to a test pit placed in the edificio, or building.

36

pottery, including mini bowls with supports, sherds with decorated motifs, and pipes with
intricate twist stems and incised-decorated stems (1939: 1 16-1 17).

The reports from Season III of excavations at the main platform discuss in more
detail the excavations at yacata 5, and in particular the ceramics found on the interior
stairs of what is called Offering I (Moedano 1941). The excavation was in trench form,
and was placed on the backside of the ydcata through what is described as a primitive
staircase. Moedano summarized the ceramic goods in the offering as many polychromes,
both miniature and regular sized vessels, large red slipped, polished vessels and obsidian
(1941 :21). The more detailed ceramic analysis from Offering I of Season III was
completed by Rubin de la Borbolla, and is summarized in Table 4. This analysis allows
for much greater scrutiny of the ceramic variables than those at the Milpillas site. All
ceramics found typify the Late Postclassic phase.

Table 4 -— Ceramics from Offering 1, Yacata 5 on the Main Platform at Tzintzuntzan
(Season HI)

 

 

 

 

 

 

 

 

 

 

 

 

Paste, Slip & Paint Decoration Vessel Form Supports Pipes
barro sandy red-orange white on red, Convex small thin wall,
M paste, non-polish motifs with wall bowl supports - bevel-
(gasteg right angled hollow lipped

lines and cylindrical wall
curves

red-orange paste, red and white: everted rim small bulbous

little sand, polished, motifs include bowl supports - base

no paint vertical dashes, spider

Z band, lines of
dots
barro reddish-brown with red on white, out sloping small bowl with
ca e sand, polished, usually on wall bowl supports - conical
aste painted on two sides handles with conical supports
with red bottom and curved lines
white motifs and/or parallel
lines

brown-orange with red and white Large small zoonrorphic

sand, polist, on cream, large concave jars supports - bowls

painted on one side to red bands and solid foot

bottom with red red and white

(large brush strokes) spirals

and white strips

 

37

Table 4 Continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

barro very sandy brown, red and white Shallow small Anthro-
ca e polished, painted red on polished red Convex supports - pomorphic
aste on one side slip, red on bowls foot-claw bowls
white dots, (thick rim)
Cont. linear fashion
along the
border
sandy brown, uniform 'scarce' tripod bowls regular
polished, painted on white, supports -
one side in red and polychrome, hollow
black. red, black and rattle
white, cylindrical twist stems
triangular motif
with reds,
whites on black
margin
brown brick, slight also present on mini bowls regular circular
sand, no polish, red above style is supports - cross section
painting on both negative hollow w/ incision
sides decoration, rattle w/
both on the wide top
exterior and
interior
dark brown, little applique tripod mini regular
sand, no polish, conical pellets bowls support -
painted red on both (referred to as hollow nub
sides al fresco)
usually with
white na
dark brown, little red on cream Miniature na na
sand, no polish, jars
painted red on one
side
dark brown, great white on black everted rim
thickness, no polish jars
reddish brown, much
sand, very thick, one
side painted red with
medium polish
dark brown little
sand. polished,
painted black on both
sides
dark earthy brown,
polished, painted red

 

 

 

Based on the four seasons of excavations on the main platform and his analysis of
the artifact assemblage, Rubin de la Borbolla summarized the artifacts from the main

platform in his 1948 publication, calling it a uniform material culture. He states that this

38

uniform material culture is based on certain artifact (including metal, ceramic, lithic),
architectural, and mortuary characteristics that define a zone within Tzintzuntzan. He
lists certain prevalent ceramic characteristics that define this culture: spouted vessels with
loop handles, tripods and tetrapods with conical supports, geometric decorations,
negative (resist) decoration, finely polished black ceramics, bowls with diagonal walls,
concave wall bowls, and large convex bowls with handles (1948:29-30). Rubin de la
Borbolla also mentions the use of the pipe as a dominant aspect of this material culture,
with circular and twist stems and incised stems (1948:30). This “uniform material
culture” along with the ceramic analysis from Season III creates strong associations with
these ceramic characteristics and the upper elite that were present at Tzintzuntzan.

The final excavations to be analyzed from the Tzintzuntzan investigations are
from the 10th excavation season from 1977-1978 by Ruben Cabrera. The excavations
center around yacatas 2 and 3, along with several habitation structures on the main
platform. The excavations around yacata 3 yielded several burials to the southwest and
northwest of its front side. The burial goods from these burials included pieces of
copper, obsidian prismatic blades, polychrome vessels including miniature bowls and
vases, plates, pipes, convex wall bowls, and spouted vessels with loop handle.

The excavations at the habitation structures included edificios F, and G, and
structures B and E. Although the descriptions of the habitation area artifacts aren’t as
detailed as the burial goods, we can say a few things about these elite habitations. The
artifacts include copper pieces, pipe fragments, polychrome ceramic fragments, metates,

and obsidian fragments. More specifically, among the ceramics were miniature vases

39

with negative decoration, miniature bowls, everted rim jars, vessels with black paint and

geometric motifs, a jar with applique conical pellet decoration, and tripod vessels.

Excavations and Collections at Urichu

The archaeological project at Urichu was conducted from 1990 to 1996 under
Helen Pollard. The project included surface surveys in the first season, followed by
analyses and block excavations. The reason Urichu was selected for investigation was
because it was one of “ten Protohistoric administrative centers that were ethnohistorically
documented to have been centers of elites prior to the consolidation of the Tarascan state”
(Pollard and Cahue 1999:261). The main focus for this thesis from the Urichu work will
be the Late Postclassic burials that were excavated in Area 1. Within the 19 burials
excavated in Area 1, there were 21 individuals. Nine of the individuals are male, seven
female, and five are unknown. Of these 19 burials, only six were without burials
offerings. Table 5 summarizes the burials and the grave goods associated with each.

The burial goods are representative of elite burials, similar to those found on the
main platform of Tz-25 at Tzintzuntzan. We know that they are associated with the
upper elites and not the lower elites, especially because of the presence of the lip plug
because according to Pollard “only office holders were permitted to wear lip plugs in the

”l3 (1999). Also, the individuals are distinguished by cranial deformation

Tarascan State
and dental mutilation, an elite characteristic that separated these families from the
surrounding populations (Pollard & Cahue 1999: 275). It is clear that these elites are the

local, upper elites of Urichu, yet not of a status as high as the royal family located at

Tzintzuntzan. Regardless, the clear distinction between these elite and the lower elite at

 

13 Pollard directly referenced this fact from the Relacion de .Michoacan, (1956).

40

the Milpillas site is the quantity of artifacts and the presence of certain artifacts that are
seen in the ethnohistoric data as upper elite goods.

It is clear that the elite class is associated with the intricate polychromes of the
Tarascan style. The vessel forms found in the burials include miniature bowls, miniature
vases (no supports), and the spouted vessel with loop handle. Also associated with the
elite burials are the metal artifacts, more specifically bronze. When compared to the
lower elite burials from the Milpillas site, the burials at Urichu contain more burial
goods, and more specialized goods, particularly the metal, polychrome ceramics, beads
and obsidian. The bronze artifacts are similar to those found within the burials at Group
B in the Milpillas site, except that they are more abundant in the Urichu burials.

Figure 4 — A Tarascan Decorated Polychrome with Basket Handle and Spout

  
  
     
  

  

 

_ a3?» - "

 
 

 

100m 15cm

I 10 cm 1
(As Taken from Versluis 1994: 413)

41

Table 5 — Burials with Age, Gender, and Artifacts from Urichu, Patzcuaro Lake Basin,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Area 1
Burial # l 2 3 4 5 6 7
Age adult adult (2) adolescent adult adult adult adult
Gender male males ? male female female female
2
ceramic 3 poly-
beads, 2 chrome
ceramic vessels,
spindle 2 6 spindle
1 ceramic whorls, 5 Tarascan whorls, l
Artifacts none none bead stone poly- none jaderte
polishers, chrome bead, 2
l vessels shell
obsidian beads, 4
blade, 4 bronze
bronze artifacts
artifacts
Burial # 8 9 10 ll 26 27 28
Age adult child (2) adult adult adult adult adult
Gender female ? male male male ? male
7
polychrome
vessels, 3
figurines,
67 beads, 3 l poly- 1 poly- 1 poly-
sprndle
. whorls l chrome chrome chrome
Artifacts stone vessel, 8 vessel, 1 none none none vessel, 1
polisher 2 bronze obsidian splndle
bone ’ artifacts 11p plug whorl
battens, 1
shell bead,
3 bronze
artifacts
Burial # 29 30 31 32 33
Ag adult adult adult adult adult
Gender ? female male female female
3 poly-
chromes,
polychrome bZeaSiienl l poly- 2 b 11181235,];
Artifacts vessel, 9 h 811 chrome 1:: obsidian
bronze S e vessel art1 prismatic
artifacts fiagme’“ blades, 1
bronze
artifact

 

 

 

42

 

Ethnohistoric Data

The final resource that will be used to create bridging arguments for this analysis
is the ethnohi storic document, the Relacion de Michoacan (RM). The RM has been
referenced in this chapter at some length, with both the mortuary and household
excavations at Milpillas and Tzintzuntzan referring to its textual accounts and
illustrations. The RM is a historic document, recorded in 1538-1539 in the Tarascan
capital of Tzintzuntzan, and was given to the Spanish Viceroy in 1540 (Pollard 1993: 17).
The account was recited by the Tarascan high priest, or petamuti, and recorded by a
Franciscan priest. It documents many aspects of Tarascan life before and during the
Spanish conquest. The accounts cover areas such as the Tarascan religion, society, and
history including the Spanish conquest (Pollard 1993 :17). The data used from the RM
for the purposes of this analysis will be taken from the 44 illustrations that accompany
the narrative. These illustrations will help to link the material culture of the Tarascans to
the ritual, social, and economic aspects of their society.

The first illustration to be examined is called “On the marriage of lords”, and is an
account of a marriage of what are the lower elite14 of Tzintzuntzan (1956:207). Depicted
in the illustration are three classes of ceramics, everted rim jars, convex wall bowls, and
spouted vessels with loop handles. There appears to be no distinct markings on the
vessels to determine decoration, paint or slip. A similar illustration, “On the marriage of
commoners”, displays a similar scene to that of the lord’s marriage (1956:213). In the
commoner marriage there are present also the large everted rim jars, and the convex wall

bowls. There is no depiction of the spouted vessels with loop handles in the commoner

 

’4 Lower elite refers to the lords, referred to as senores in this illustration of the RM, are the leaders of
towns or villages and are considered the lowest administrative position of the Tarascans. They are also
referred to as caciques or Angamecha. (Pollard 1983: 125)

43

marriage. The spouted vessels with loop handles show up twice more in the RM, and
both in elite contexts. One illustration, referred to as “How he died and the ceremonies
with which they buried him”, is an account of the king’s funeral and the ritual associated
with it (1956:218). In this case, the spouted vessel is being carried by a woman in the
funeral procession who is walking ahead of the deceased individual. The illustration goes
on to describe sacrifice rituals and funerary rituals, including the cremation of the body.
The other illustration is an account of the king fleeing his household in fear of the
Spanish arrival. In this depiction, servants are witnessed carrying his goods out of his
house. The goods include these spouted vessels, tripod bowls, and everted rim jars.
These illustrations provide an association between the elite and the spouted vessels with
loop handles as well as the tripod bowls, as well as linking the spouted vessel to ritual
and ceremony in the Tarascan society.

The final illustration from the Relacion de Michoacan is called “Concerning the
general administration of justice” (1956:11). This illustration depicts several of the
Tarascan administrators presiding over what seems to be trials, some of which result in
execution. The administrators labeled in the illustration as caziques (caciques, lords) and
are shown smoking pipes. All are male, and all are witnessed to be part of a state
administrative ceremony. This supports the excavation data from both Milpillas and
Tzintzuntzan in confirming the fact that pipes are associated with males and range from
the lower to upper elite status. Furthermore, it demonstrates that pipes are not a basic
household item, but are ritual and ceremonial in nature and are used primarily within

those contexts.

44

Ceramic — Status Associations

The goal of this chapter is to provide bridging arguments for the associations
between certain ceramic variables and specific social classes. To achieve this end I have
reviewed analyses done of ceramic production, excavations in the Zacapu region and at
Tzintzuntzan, and an ethnohistoric analysis of the Relacion de Michoacan. This research
provides useful evidence for what can be defined as a socially stratified society as seen
through the ceramic assemblage. Based on these interpretations, Table 6 summarizes the
ceramic variables associated with the social classes defined by Pollard in her original
analysis (1972).

The vessel form for each class is defined by the vessel’s function. From the
excavations at Milpillas and Tzintzuntzan, the main difference between commoner and
elite when regarding vessel form seems to be open versus closed vessels. The open
vessels, primarily seen in elite contexts, are those vessels used for serving and eating,
whereas the commoner vessels are represented by the closed vessels, used for cooking
and storage. The lower elite, as seen at the site of Milpillas, seem to have evidence of
form, as the smaller supports are from the miniature vessels and are associated with an
elite class. F urtherrnore, the regular, hollow supports are associated with the polychrome
tripods, and are also an elite ceramic. The regular solid supports are associated with

larger plainware vessels, and are therefore a lower elite and commoner variable.

45

Table 6 — Ceramic Variables and Associated Social Classes

 

 

 

 

 

 

 

 

 

Vessel Forms Decoration Paint Sumports Ware Slip
_A convex wall T . ri
g bowls, re ular 82:5:
:5 everted rim Mono- g , ’ .
:- . . supports, Yaguarato unsllpped
c jars, 1ncurved none chrome or . .
E . . solld Coarse, unpainted
rlm bowls, non-palnted . .
g 1ncurved fim conlcal Ichuplo
U jars Coarse
convex wall . .
3 bowls, regular Télrlacurl
a A everted rim . One or two supports, rown,
;- . . all motlfs, . solld Yaguarato .
g; 5 jars, 1ncurved . color palnt . sllpped
B V . no negatlve conical, Coarse,
c, rim bowls, combos 1d 8 'h
.1 spouted so 1 lpl o
vessels spider Grey
mini jars,
mini bowls, small
spouted supports
A vessels, r2231)” Tariacuri
E spouted . Polychrome: _ Brown,
: handles, negfifé’ two and three 5:1; 112:8, Yaguarato
£5 convex wall app .q color combos Cream, .
jg conlcal . . rattle . slipped
bowls, out- mcludlng . . Tarerlo
g; . pellets, all . cyllndncal,
Q sloplng wall . f negatlve as h 11 Cream,
3' bowls, motl S atype O 0W Sipiho
brasero, rattlt: wrde Grey
plates hollpdw
(convex and nub
flat base)

 

 

 

Decoration and paint will be an important factor when determining social status.

In the excavations at Tzintzuntzan, the vast majority of the ceramics were decorated and

polychrome ceramics. At Milpillas, the lower elites were also seen to have these

decorated polychromes, but in much lower frequencies, whereas the commoners have

few to no ceramics of this type. The difficult aspect of the analysis will be defining the

lower elite, as they display ceramic characteristics from both the commoner and elite

assemblages. The key qualifier for the lower elite will be quantitative, as they are

46

 

expected to have a higher frequency of plain ware ceramics, and a lower frequency of the
decorated polychromes.

Ware and slip are the most difficult variables to associate with social class. The
ware of the Tzintzuntzan ceramics at Offering I (see Table 4) was described by Moedano
in detail. This, along with the ceramic groups described by Migeon at the Milpillas
excavations, helped to determine what ware occurred most frequently in commoner, elite,
and lower elite contexts. The slip, though, is much more difficult because as noted by
Hirshman it ranges across a variety of ceramics, including coarse and fine wares and thus
may not be useful in determining social class associations. Instead what will be analyzed
is merely the presence or absence of slip.

These variables have been chosen for their ability to clearly display trends within
the material culture that are representative of class distinctions within Tarascan society.
Emphasis must be made that these trends are just that; tendencies toward a certain aspect
of the material culture. In no way do these categories represent stringent, clear-cut
divisions between classes. The ceramic assemblage as I have presented it exists as a
continuum from commoner to elite, with the ceramics representing aspects of status that
change through time and space. Variables such as access to resources, access to
specialized production, wealth, and location all determine the material culture that
defines that class. This analysis of ceramics has chosen as its variables ceramic
characteristics that best represent social classes, and in no way is complete or all

encompassing in terms of variables that represent each class.

47

Chapter 3

Ceramic Analysis

The methodological framework defined in Chapter Two has created categories of
ceramic variables that define specific social classes within the Tarascan capital of
Tzintzuntzan. These categories must be spatially analyzed to determine the zonal
characteristics of the urban center. This will allow for the survey sites to be classified
into residential zones and/or public zones based on social status. This chapter presents
the statistical analysis of the survey site classifications and the subsequent interpretations
that Will ultimately determine the urban characteristics of Tzintzuntzan.

As discussed previously, the data set consists of 120 survey sites, 89 of which had
artifact collections. The ceramics comprise the bulk of the artifacts, and consist of n =
2152 ceramics sherds, pipes, and other (such as spindle whorls, figurines, or unknown).
The data set is count data, containing many zeros due to the nature of the archaeological
survey process. The unit of analysis will be the survey site. An area of uncertainty will
be the treatment of the n = 31 sites that lack artifacts for analysis. These sites may have
been defined by architectural features, such as Tz-25 that contains that main platform but
where no collections were made by Pollard, or by artifacts noted by Pollard yet not
collected (1993 : 189). To determine the social significance of these sites, they will have
to be interpreted through the field notes that Pollard originally took on the features of the
site15 . This includes any architecture present, descriptions of adjacent sites that may

relate to or overlap the site in question, and artifacts that were noted yet not collected.

 

‘5 The archaeological survey notes can be found in the Appendix 1 of Pollard’s monograph Tariacuri '5
Legacy: The Prehispanic Tarascan State.

48

Cluster Analysis

The primary statistical approach used is a cluster analysis. The cluster analysis
that will be used is a multivariate statistical procedure that starts with a data set
containing information about a sample of entities and attempts to reorganize these entities
into relatively homogeneous groups (Hirshman 2003: 113). The sites will be run through
an agglomerative, hierarchical cluster analysis with a polythetic classification, where
“multiple variables are taken into account to form the groups with no one variable
sufficient or necessary to group membership (Hirshman 2003: 113). The hierarchical
cluster used is based on a between group cluster that uses the average linkage between
groups. Average linking is a “space conserving algorithm that defines distance between
groups as the average of the distance between all pairs of individuals in the two groups”
(Hirshman 2003: 117). This will allow for each variable to help spatially assign the sites
based upon a specific ranking tied to status and function, which will define its
corresponding zone.

Furthermore, K-means cluster analyses will be completed that will show the
membership qualifications for the resulting clusters. A K-means cluster is based on the
hierarchical cluster, but refines the results, moving the cases that were initially miss-
assigned in the one-pass hierarchical results (Hirshman 2003: 120). The one distinction
of the K-means cluster analysis as opposed to the hierarchical analysis is that the number
of clusters in the output is subjective and up to the discretion of the analyst. In this
analysis, the K-means cluster will be done after the hierarchical and the number of output

classes will be defined by results of the previous hierarchical cluster analysis.

49

As previously explained, the variables chosen are in no way complete in the sense
that they include all ceramic characteristics attributed to a certain social class. The reality
of the material culture of the Tarascans is that there is a continuum, and therefore overlap
between social classes as seen in the ceramic assemblage. This will result in an overlap
of variables between social classes, and will also result in an overlap of the sites in the
cluster analysis, meaning the classification of the sites will not be as clear and concise as
the cluster analysis methodology may make it seem. Because of this, the methodology
for the cluster analysis is as follows: All n = 89 sites will be clustered first based upon
the upper elite variables defined at the end of the previous chapter. The resulting clusters
will be analyzed, both quantitatively and qualitatively, on the basis of cluster
membership. The cluster(s) best exemplifying the upper elite ceramic characteristics will
be noted and removed from the analysis. This will effectively be removing the outliers
from the analysis, thus allowing for the remainder of the sites to be clustered further on
the same upper elite variables. This will allow for a more detailed analysis into those
sites that may display both lower elite and upper elite characteristics. As I have stated
before, I believe the primary distinction between the lower elite and upper elite will be
the quantity of the upper elite ceramic variables. Hopefully, by removing the outliers,
these sites may be better analyzed and properly classified.

The following analyses are explained in terms of quality and quantity of the
ceramic variables. To clarify the terms that will be used, I must provide some
background about the data for a better understanding of the proceeding analysis. The
variables are discussed in terms of high, medium and low frequencies when referring to

cluster membership. For instance, the total number of sherds for these n=89 sites is

50

n=1810, thus leaving an average of 20.3 sherds per site. This is a medium-low frequency
for sherds, with a high count ranging from forty to over 100 sherds. The problem is that
for each sherd many variables may exist. To define what a low, medium and high
frequency refers to, the variables from each of the social status categories (elite, lower
elite, commoner) were added across the sites, and then the mean was derived for each
site. For the upper elite variables, the average frequency for each site was n=22.8
variables. For the lower elite variables, the average was n=26.2 variables per site. For
the commoner variables, the average was n=10 variables per site. Thus the range for low
frequencies of variables is n=l-9, a medium frequency is n=lO-20, and a high frequency

isn>20.

Cluster Analysis — Upper Elite Variables

The first analysis to be completed will be on the upper elite variables. These
variables were chosen first due to the evidence seen in previous excavations and surveys,
which were discussed in the previous chapter. It is more likely that the elite ceramics
contains variables that range across all three social class assemblages, exemplifying this
continuum-based material culture. This top-down methodology assumes that the elite
assemblage contains more commoner-defined ceramics (i.e. coarse wares, undecorated
pottery), than the commoner assemblage contains elite ceramics (i.e. polychrome
minibowls, resist decoration).

A hierarchical cluster analysis was completed containing all n = 89 sites using the
upper elite ceramic variables defined in Chapter Two. The result of this first cluster

analysis was two groups, one of which contained four sites. Membership for this cluster

51

was based on extremely high frequencies for the variables, especially the motifs, small
supports, convex bowls, the fine ware pastes, and spouts. In order to take a better look at
the other sites and their distribution of variables, a K-means cluster analysis was done
that separated the clusters further into three groups. In this case, the same four sites
constituted one cluster based once again on the high frequencies of variables. These four
sites, Tz-12, Tz-l6, Tz-17, Tz-91, are the outliers due to their high frequencies of the
upper elite variables. The second cluster’s membership was based on extremely low
frequencies of the variables and in many cases were zero. Most of the sites, 11 = 82 to be
exact, fell into this cluster. The third cluster only contained three sites, and aided in
distinguishing between the final two clusters in the first hierarchical cluster analysis. The
sites Tz-18, Tz-59, and Tz-85 were placed in this cluster with membership being defined
by high frequencies of the variables in convex wall bowls and motifs and medium
frequencies for the remaining elite variables. These sites aren’t as extreme as the four
outlier sites in terms of the frequencies of the variables. These sites require a closer
examination of their variables to determine into what class, lower or upper elite, they fall.
In order to better examine the remaining n=82 sites both a hierarchical cluster
analysis and a K-means cluster analysis were done to determine the nature of the
clustering. The resulting clusters from the hierarchical analysis were much greater in
number and created a much more diverse dendrogram as opposed to the clusters that
formed when the seven outlier sites were included. The outlier sites for this hierarchical
cluster analysis were sites Tz-63, Tz-70, and Tz-99, followed closely by sites Tz-23, T2-
7 8, Tz-9O and Tz-107. When looking at the group membership for all clusters, it is clear

that now we are dealing with a variety of sites that range in variables and frequency from

52

low to medium, with no sites being considered extreme outliers. To better determine how
the cluster analysis was defining its group membership, a K-means cluster was done that
defined four output clusters. Of these four, no one cluster stood out. Cluster one had a
medium frequency of convex bowls and a small to medium frequency of small supports
and motifs. Cluster three was interesting because of its medium high frequency of spout
handles, yet lacked distinct frequencies in any other variables.

It seems that in this stage of the analysis, the upper elite sites have been identified
and we are now dealing with a range of sites that represent the lower elite status. This is
apparent in the low frequencies of elite ceramic variables. Based upon a closer analysis
of these “on-the-bubble” sites, it was determined, based on a low frequency and lack of
the defining upper elite variables, that sites Tz—59, Tz-63, Tz-70, Tz-78, Tz-99, and T2-
107, are most likely lower elite sites, and not upper elite. Sites Tz-18, Tz-85, and Tz-90,
however were seen to contain these defining upper elite variables in medium frequencies,
and even though in lower frequencies than site Tz-12 for example, still exhibit the
characteristics of the upper elite assemblage discussed in Chapter Two. To filrther
examine zonal characteristics, the remaining sites will now be analyzed by cluster

analysis based upon the lower elite variables that were defined in Chapter Two.

Cluster Analysis — Lower Elite Variables

The five definite upper elite sites (Tz-12, Tz-l6, Tz-l7 and Tz-91) were removed
from the next cluster analysis, leaving n = 84 sites to analyze. A hierarchical cluster
analysis was performed for these sites, using the lower elite variables defined in Chapter

Two. The resulting dendrogram looking very similar to the last hierarchical cluster

53

analysis performed for the upper elite variables. To better understand the clusters created
by the hierarchical analysis, a K—means cluster analysis was done with the number of
output classes set at four. The result was one cluster defined solely by site Tz-18, two
clusters defined by medium frequencies for the majority of the variables, and a final
cluster containing none of the variables. It is clear that Tz—l 8 may be considered both an
upper elite and lower elite site. Due to the fact that the cluster analysis shows it as an
outlier, distinct from the other sites, makes me think it belongs in the upper elite category.
Tz-18 is defined by a medium frequency in the motif, a high frequency of the red on
cream paint scheme and a medium frequency of the all over red on cream paint scheme.
It is clear that polychrome pottery is significant at this site, thus placing it into the upper
elite category. The remaining n = 26 sites in clusters 1 and 2 are representative of the
lower elite category as defined in Chapter 2. Both clusters base membership on medium
frequencies across all variables, with focus on the paint variables, motifs, jars and wares.
These sites are indicative of this lower elite class, in that they do contain
polychrome pottery but lack the minibowls and tripod vessels that define the upper elite
assemblage. They also tend to have a higher amount of j ars than bowls, though both
vessel forms are represented in average quantities. The coarse wares are also more
prevalent within the site’s assemblages than at the upper elite level, showing a tendency
in the lower elite assemblage to contain aspects of both the commoner and upper elite
ceramic variables, but never in great numbers. To test this fact, I added miniature bowls
as a variable to the lower elite cluster analysis to see if this upper elite variable played a
significant role in the clustering. As it turns out, the n= 84 sites clustered in exactly the

same fashion with the minibowls than they did without the minibowls. This shows that

54

these sites are now definitely out of the range of the upper elite category, as the presence
of an upper elite variable had no statistical significance in the remaining site’s

classification.

Cluster Analysis — Commoner Variables

The sites not classified into the upper or lower elite categories may ultimately be
considered the commoner sites. To be sure that these sites fit the commoner category,
they will be put through a final cluster analysis with the lower elite variables being used
for classification. It is possible that some of these n = 61 sites may not exhibit any of the
ceramic characteristics of the commoner zones, thus relegating them to zones such as
lithic manufacturing or public zones. To clarify once again, the commoner sites will be
defined by the presence of coarse ware ceramics, the higher frequency of j ars to bowls,
and the absence of decoration, paint, miniature vessels, and overall lower counts for all
variables.

The first hierarchical cluster analysis that was done included n=6l sites, excluding
the definite upper elite sites (n=7), and the lower elite sites defined by the previous lower
elite cluster analysis. The resulting dendrogram was more complex than the previous
upper and lower elite clusters, and showed no definite outliers. The sites seem to be more
associated based on the variables chosen, as would be expected now that the elite classes
have been selected for. A K-means cluster was then done and three output classes were
chosen. Only Tz-70 stood out as being different from the remaining sites. Because of the
higher than normal frequencies expected for a commoner site, Tz—70 will be examined

more closely to determine whether it belongs in the commoner category or the lower elite

55

category. The remaining sites were placed into two clusters. The first cluster contained
n=43 sites, and membership was defined based on a frequency of one or absence of the
variables. The second cluster contained n=17 sites, and was defined by low frequencies
in bowls, jars, and wares. These sixty sites will most likely constitute the commoner sites

for the survey area.

Cluster Analysis — Public Zones

The public zone cluster analysis is based on the presence of pipes and pipe
fragments at the survey sites. The research in Chapter Two equates pipes with males,
ritual, and a tendency toward elite use. Based on these notions then, the distribution of
pipes will aid the classification of the sites in two ways: (1) It will provide an alternative
analysis for those sites that remain in the grey area as to their classification into lower
elite or upper elite zones, and (2) will locate the survey sites whose zonal function may
not only be a residence but also a public zone. According to Pollard, the public zones are
defined by ritual and religious use, and are the local religious centers of the settlement
(1987: 49). She goes on further to state that the primary public zone for Tzintzuntzan
was the main platform of Tz-25, and is described as such in the Relacion de Michoacan
(1987: 47). The problem with the secondary public zones is that they are also located
within residential areas within the settlement. Pollard’s research led her to conclude that
the public zones also contain middle-status (lower elite) habitation assemblages, and are
all associated with looted burials (1987: 49). This possible overlap inthe zones will help

in providing insight into the urban structure of Tzintzuntzan, and may provide further

56

insight into the social status of the residences that have already been classified in the
cluster analysis.

To determine how the sites were classified into public zones, variables were
chosen that best represented social differences in the pipes. The burials and excavations
at the main platform at Tzintzuntzan yielded pipe fragments displaying a range of
decorative characteristics, including: incision on the stems, painted stems and bowls, and
elaborate bowls in decorative styles such as anamorphic and zoomorphic figures (See
Table 7 below for a detailed description of the pipe variables used in the following cluster
analysis) The decoration-based variables will be applied to the pipe cluster analysis thus
allowing for a possible identification of elite sites versus commoner sites. This will aid in
defining elite zones and public ritual zones, but will remain a secondary factor as the
quantity of pipes will be the primary designator for public zones.

The hierarchical cluster analysis clearly showed site Tz-12 as an outlier, because
of its extremely high frequencies of pipes with circular cross section pipe stems. A
second cluster included seven sites, with the third cluster containing the remaining n=81
sites. To determine the reasons for cluster membership, a K-means cluster was
performed with an output of four classes so as to try to limit the effect of Tz-12 on the
results. Once again Tz-12 was given its own cluster, cluster 1, because of its high
frequencies of pipes. Cluster 2 had 21 members, membership defined by a medium
frequency in the circular and oval pipe stems, no incised stems, and one twist-four pipe
stem. Cluster 3 contained n=66 sites, and was based on solely on the presence of one
circular cross section pipe. Finally cluster 4 contained one site, Tz-112, because of its

high frequency in the circular cross section.

57

Table 7 - Pipe Variables: Stems, Bowls, and Bits

 

 

 

 

 

Bowls Stems Bits

thin wall beveled lip simple twist ground around
thin wall conical

bowl incised twist four beveled

bulbous base twisted and incised pattern burnished
conical srmports twisted and scalloped incised

 

oval cross section
circular cross section
incised twist three
scalloped

corn cob

 

 

 

 

 

 

 

 

 

To determine if any other sites were distinct from the n=21 or n=66 sites, Tz-12
and Tz-112 were removed and a K means was run again. This time the bits and bowls
were removed, leaving only the stems as variables in the cluster analysis. It was clear in
the first two cluster analysis, that due to the extremely low frequencies of both bowls and
bits, they were playing no part in the classification of the sites in the analysis. The stems
proved valuable because of the range in slip, decoration, and form, thus providing a good
reflection in social status. The second cluster analysis was run with an output of three
classes. The fist cluster contained nine sites, with membership based on a medium
frequency in circular cross sections. Cluster 2 contained n=22 sites, and also was based
on circular cross section stems but in a lower frequency. Cluster 3 contained the

remaining sites, and was based on an absence of pipe variables.

Cluster Analysis — Querenda White Ware
The final cluster analysis is solely based on the presence of the Querenda White
ware. Previous research has shown this ware to be spatially isolated to the northeast

section of the settlement at Tzintzuntzan. Unlike the other wares, the Querenda White

58

was thought to be different based on the distinct paste that constitutes it. Pollard
hypothesized that this ware is indicative of an ethnic variation within Tzintzuntzan, one
defined by a distinct ceramic tradition characterized by the Querenda White ware.

The resulting cluster analysis displayed a limited number of sites with a frequency
of Querenda White ware ceramics. The first hierarchical cluster analysis returned three
sites as outliers, Tz-59, Tz-63, and Tz-70. The following three class K-means also
supported this pattern, and placed Tz-59 in its own cluster due to an extremely high
frequency (n=31) of Querenda White ware ceramics, with Tz-63 and Tz-70 as a second
cluster with a medium frequency of the ware. The remaining sites formed a cluster due
to a low or zero frequency. The three outlier sites were removed, and the K-means was
run again with an output of three classes. The result was one cluster with a frequency of
zero which contained n=67 sites, one cluster with n=l6 sites containing a frequency of
n=1, and a final cluster of three sites that had a frequency of n=3. Figure 4 contains a
map locating the sites that contained the Querenda White ware ceramics.

After mapping the distribution of the Querenda White ware, it is obvious that they
are in fact concentrated in the northeast area of the survey site. The high frequency of the
ware in this area and extremely low frequency in other areas seems to form a distinct
area. Originally this northeast area was thought to contain an ethnic barrio, or
community, whose origins derive from outside the basin”. More recent research on the
paste and firing techniques have shown that this ware is most likely from an earlier
occupation at Tzintzuntzan, probably before the emergence of the state and Tzintzuntzan

as the capital (Pollard, personal communication, March 2008).

 

'6 Pollard located sherds of identical paste in such areas as Zacapu, Guanajuato and others elsewhere
outside the Patzcuaro Lake Basin, leading her to believe the ware’s presence was due to an ethnic variation
within Tzintzuntzan (Pollard 1993:215-216).

59

Figure 5 — Location of Querenda White Ware Ceramics

 

 

\ I ~
\ ’ — - — ’ ' t -
\ _ __ _ — — r a ’ ' - ' I ’ ‘ .- " ’
- r 2‘ ’ _ — - .. ’
x - ' v-
------ t I
‘ I -1
I ‘a' i 1
\. \
-~\\‘ ‘ ‘. '. C’ 9
x '. ../. Q " O—__
‘ g 9. ’ r """""
. 193‘,
1 .. ‘V “ H; .'
\ ‘2‘." " I ‘5‘]
\_ \m .- r
\ I; O~-' ' , v ’
l \ ,‘l 'I
. c J ,
5 \
, ~ ,. i’I . I
I. \, “w“ k i ’
I / ‘. I
i x - ' T ‘ \ l‘ ' / , /
—-' ' 2. .p‘, I
’ “V \ \\
I .
I \ Jill ‘
1 €- ‘
‘ \
\
N ‘ \ \ x l
\ \ ,4 4,
l ‘ ' ..
\ l
\ . . 1'
s .. .
‘ 1 i; ‘ V - . ,. " . .' I
\ ‘ . . .‘ ‘ ’
‘ ‘ ._ _ _ .. > ’ ‘ . . r .
. . . » - ,' Querenda White Ware
1
I
, 1. .. _ -,_, ; Frequency
I ' ‘ \ .
- ~ . . 1 .WW.
1’ F- ii ' fl
, i3: ’
, r. ,’
1
I
l . ' l
’ a.,_ l
I . ..._. I
l l
l l
I r i
l i
I ‘ . . . .
| r j "x . '1‘ "’- l ’. ‘
. . ._._ \_ . t : : Survey Boundary
. » l — - — -
1' g 1 ,- ,- . . .
F: , l , ‘ .. Surface Scatter
\.
Vi. _- ‘5. .- \ ’ ‘ '
2% ‘
' 1 \
\' ‘
\
x " I
I
I
\ I
‘ I
‘ I
\ I
\ i I
. ' .. v _ -
I ’ ‘ , ,
I I I d
0 0 5 ‘ 1 2 Kllometers
l l 1 l l l 1 l l
Interpre tatrons

The purpose of these cluster analyses was to place the sites into zones that
represent certain social statuses within Tzintzuntzan society during the Late Postclassic
period. The variables that represented each zone were chosen through a carefirlly

researched methodology that included analyzing excavations and survey data from

60

Tzintzuntzan, sites immediately within the Patzcuaro Lake Basin, and Tarascan sites
outside of the lake basin. The cluster analyses provided a statistical foundation for the
classification of sites into zones, a factor that was absent in Pollard’s original analysis of
the Tzintzuntzan survey data.

To summarize the results, the main hypotheses for the expected social distinctions
within the ceramic assemblages were accurate. The cluster analysis for the upper elite
showed that the outlier sites were based on both high frequencies and the most distinctive
of elite ceramic variables, such as the motifs and paint schemes as seen in the polychrome
pottery, the resist decoration and the small supports. As expected, the primary distinction
between these sites and the sites identified as the lower elite sites was measured
quantitatively. The lower elite sites did contain aspects of the upper elite ceramic
assemblage, but in much lower frequencies. The lower elite sites also contained aspects
of the commoner ceramic assemblage, with more plain ware and non-decorated ceramics.
Finally, the commoner sites saw a dramatic decline in the elite ceramic variables, with the
cluster memberships being defined by medium to low frequencies in jars and the coarse
wares. The overall variable counts were low across the board for the commoner sites.

The results of the cluster analysis are displayed in Table 8 below, as well as in the
Figure 5 map of the zonal assignments. The sites that are classified into zonal categories
are the n=89 sites that were analyzed in the cluster analysis. The only change made was
Tz-ll due to the combination of a medium pipe frequency with a low frequency in lower
elite variables. In this case of Tz-l 1, the qualitative distinctions override the quantitative
factors, and it was moved from a commoner zone to a lower elite zone. Also, because

from the ethnohistoric evidence we know that Tz-25 was the main ritual platform, sites

61

Tz-l, 2, 3 and 5 which are all located either on or directly adjacent to Tz-25 are

considered Primary Public zones (Pi). As noted in Chapter Two, these sites have been

the focus of many previous excavations and surveys, leaving the main platform and its

nearby sites devoid of any artifacts. Even though the artifacts found are typical of the

quantity and quality of commoner zones, the fact that they are related to the primary

public zone overrides this fact. The remaining sites are all classified based upon their

zonal assignments from the cluster analysis. The remaining n=31 sites will be looked at

further and assigned a zonal category based on Pollard’s interpretations and the variables
and methodology from chapter two of this thesis. These sites will not be used in testing a
possible association between Pollard’s zonal classifications and my own, but will be used

in discussing the urban characteristics of Tzintzuntzan.

Table 8 — Zonal Classification of Sites (n=89) Based on Cluster Analysis

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Primary
Lower Elite Public Secondary
Upper Elite Zones (Riii) (Pi) Public (Pii)
(Rii) n=7 n=24 Commoner Zones (Ri) n=58 n=7 n= 11
Tz-12 Tz-8 Tz-63 Tz-l Tz-35 Tz-55 Tz-93 Tz-l Tz-l 1
T2- 1 6 Tz-l 0 Tz-70 Tz-2 Tz-36 Tz-56 Tz-95 Tz-2 Tz- l 2
Tz-l 7 Tz-l l Tz-75 Tz-3 Tz-37 Tz-58 Tz-97 Tz-3 Tz—70
T2- 1 8 Tz- 1 5 Tz-76 Tz-4 Tz-3 8 Tz-60 Tz-99 Tz-6 Tz-75
Tz-85 Tz-l9 Tz—77 Tz-5 Tz-40 Tz-62 Tz-102 Tz-7 Tz-76
Tz-9O Tz-20 Tz-78 Tz-6 Tz-4l Tz-65 Tz-104 Tz-8 Tz-87
Tz-9 1 Tz-22 Tz-84 Tz-7 Tz-42 Tz-67 Tz- 1 05 Tz-9 Tz-90
Tz-23 Tz-107 Tz-9 Tz-43 Tz-68 Tz-106 Tz-104
Tz-39 Tz-110 Tz-13 Tz-46 Tz-73 Tz-108 Tz-112
Tz-53 Tz-l 12 Tz- l 4 Tz-47 Tz-74 Tz-109 Tz-l l4
Tz-59 Tz-l 14 Tz-21 Tz-49 Tz-8 l Tz-l 10a Tz-l 1 7
Tz-61 Tz—l 17 Tz-24 Tz-50 Tz-86 Tz-l ll
Tz-32 Tz-51 Tz-87 Tz-l l3
Tz-34 Tz-54 Tz-89 Tz-l 15
Tz-l l9
Tz-120

 

62

 

Figure 6 — Zonal Classification of Tzintzuntzan Sites

 

‘ i Public Zones
A \\ ' , I I,» /l m PI (Prlmary)
I %§Z Pil (Secondary)
Residentlal Zones

, ‘ ‘ r . l Rl (Commoner)
l I - Rii (Upper Ellte)
. 6‘ .7 l - Rlll (Lower Elite)

1. C : j Tzln_survey_boundary

0 0.5 1 2 Kilometers
1

Statistical Significance of the Cluster Analyses
The intention of the cluster analysis is to place the sites into zonal categories
based on statistical methods, methods not used by Pollard in her original analysis of the

Tzintzuntzan survey. By creating these zonal categories, I am now able to statistically

63

compare Pollard’s original classifications to my own. The reanalysis will inherently test
a two-part question, both parts reliant upon each other. Pollard’s original ceramic
analysis and zonal interpretations will be tested, along with the conclusions Pollard made
about the extent to which urbanism was present at Tzintzuntzan.

To test whether my zonal interpretations are different from Pollard’s a Wilcoxon
signed-ranks test will be carried out. This test is a nonparametric analysis designed for
ordinal data that is not dependent on a given distribution (Sokal and Rohlf 19952444).
Thus, the test is not concerned with specific parameters but instead with the distribution
of the variates. The Wilcoxon signed-ranks test is an analysis of two dependent
observations taken from the same population. In the current case, the two observations
are my zonal classifications of the survey sites and Pollard’s zonal classifications. This
test determines the magnitude of difi‘erence between the two observations as well as the
directionality of the difference. For this test, my null hypothesis is that there will be no
difference in zonal classification between the zones determined by Pollard and myself.

The zonal data are ranked and coded based upon social status. The zones were
coded as follows; Ri (commoner residential) = 1, Pii (secondary public)= 2, Riii (lower
elite residential) = 3, Pi (primary public)= 4, Rii (upper elite residential)"; 5. The
differences between my zones and Pollard’s are calculated and ranked without regard to
sign. The ranked sums are then calculated and interpreted as to whether the difference is
significant and in what direction the difference occurs. The lowest, absolute numerical
ranked sum was taken, in this case was T = 81. Because the sample size contained so
many ties, the sample size was dropped to n=21. The 2 score was calculated in both

SPSS and by using 2 = T - p. / o. In this case, p = n(n +1)/4, and o = ‘ln(n+1)(2n+1)/24.

64

The score for z is calculated to be -1.255, which is within the 0.209 critical level of a two
tailed significance test. This means that approximately 80% of the time, differences seen
between the two paired observations (Pollard v mine) are not due to chance but instead
have different medians.

The problem with a test such as the Wilcoxon rank is that it doesn’t take into
consideration the ties that occurred, and only looks at the differences between ranks. In
this case, there were n=58 ties between Pollard’s zonal assignments and my own, which
out of n=79 total cases is approximately 73%. It is clear that Pollard’s assignments and
my own are highly associated, yet in areas of disagreement there seems to be significant
differences. These differences are a key component of this analysis, as they will give an
insight into details of the artifact assemblages that may have been overlooked by Pollard

but are in fact of statistical significance.

Discussion

A closer look at the n=21 differences in zonal classifications between Pollard and
myself reveals striking patterns. Of the multiple combinations that are possible for zonal
assignment, the differences occurred within six combinations. In terms of social ranking,
only one drastic difference stands out, where at Tz-13, I classified the site as a Ri
(commoner) and Pollard an Rii (upper elite). The following differences in classification
are minor in terms of frequency and social status ranks:

o Tz-41, Tz-56, Tz-l 11 - Pollard classified as Pii (secondary public), I classified as
Ri (commoner residential)

o Tz-S3 — Pollard classified as Pii (secondary public), I classified as Riii (lower elite
residential)

65

o Tz-8 - Pollard classified as Pi (primary public), I classified as Riii (lower elite
residential)

The interesting aspect of this analysis is the remaining n=15 cases where Pollard
and I differed. In n=12 instances, the differences were between Ri zones (commoner)
and Riii zones (lower elite). In these circumstances, half of the time I classified the site
as a lower elite site and Pollard classified the site as commoner, and vice versa. In the
remaining n=3 cases, the differences were between Rii (upper elite) zones and Riii zones
(lower elite), where in both cases I classified the site as being a lower elite and Pollard
classified as being an upper elite. The main reasons for the classification into a lower
elite zone rather than an upper elite zone is not the lack of presence of the upper elite
variables, but rather the lower frequencies with which they occur. The latter distinction is
a very intriguing aspect of the analysis, as previously it was thought the main differences
between Pollard and my classifications were to be between the upper and lower elite
zones, and not between the commoner and lower elite. Below in Table 9 are the differing
classified sites and possible reasons why they were classified as they were.

As Table 9 shows, the trend of classification into lower elite zones seems to be the
low frequency of the upper elite variables, such as resist decoration, motifs, minibowls,
small supports and spouted vessels and spouted handles, and paint combinations. All of
these variables are deemed significant due to the ethnohistoric and archaeological
evidence as described in Chapter Two. Due to their low frequency in the above sites (Tz-
19, Tz-39, Tz-70, Tz-76, Tz-107, Tz-l 10), they may have been have been overlooked by
Pollard in her original analysis but yet were statistically significant in the cluster

analyses.

66

Table 9 —Differences between Commoner and Lower Elite Zones — Pollard v. Mine

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Reason for my zonal classification ( f =
Sites My Zone Pollard Zone fremrency)
Tz-14 Ri Riii low gr nfl ] of all elite variables

medium f of slip/painted sherds; low f of spouts,
Tz-l9 Riii Ri spouted handles, resist and motifs

low f of spouted handles, motif, supports;
Tz-39 Riii Ri medium of paint and slipped sherds
Tz-65 Ri Riii no ceramic artifacts associated

medium I of painted sherds; low f of supports
Tz-70 Riii Ri and motifs; high f of slipped sherds

medium f of painted/slipped sherds; low f of
Tz-76 Riii Ri resist and motifs

low f of spouts, supports, resist; medium f of
Tz-86 Ri Riii paint, sliL

low f slip, paint, motif(n=l), and low jars with
Tz-89 Ri Riii n=1 bowl
Tz-93 Ri Riii low frequencies throughout, low slip and paint

low f minis, plates, supports, spouted; high f of
Tz-107 Riii Ri slip/paint sherds; medium f of motifs

low f minis, supports, resist, motifs; medium f
Tz-l 10 Riii Ri painted sherds; higlLf slipped sherds
Tz-119 Ri Riii only ceramic consisted of one grater sherd

 

The original hypothesis about any possible distinctions between Pollard’s analysis

and my own were thought to be between the categories of upper and lower elite. Yet the

cluster analysis shows the differences between these two groups seems markedly more

distinct than between the lower elite and the commoner. The lower elite seem to share a

more related assemblage with the commoner class than they do with the upper elite.

Chapter 4 will discuss the possible reasons for the trends seen in this analysis chapter,

and the implications of these analyses on the nature of urbanism at Tzintzuntzan.

67

 

Chapter 4

Urbanism at Tzintzuntzan Revisited

The cluster analyses in Chapter 3 resulted in several patterns that must be
discussed within the larger scope of the urban characteristics of Tzintzuntzan. These
patterns can be summarized into three themes: (1) the dichotomy between the commoner
and lower elite assemblages that arises when comparing Pollard’s zonal interpretations to
mine, (2) the theme of zonal overlap that occurs between the public and residential zones,
and (3) the larger issue of zoning and planning as expressed by the ceramic assemblages
at Tzintzuntzan. Also, a brief discussion concerning the spatial distribution of
manufacturing zones in relation to the residential and public zones may add further
insight into the social and economic behavior of the Tarascans. By examining these
themes in relation to Pollard’s original analysis of Tzintzuntzan and by positing them
within current urban theory in Mesoamerica, I hope to shed some light on the urban
nature of the Tarascan capital as it existed in the Late Postclassic period.

Before examining the zonal classifications in detail, I must make explicit the
methods of the original analysis completed by Pollard. In her original analysis of the
Tzintzuntzan survey, Pollard used the full range of artifacts found in the survey as well as
evidence of architecture, proximity to areas such as the main platform, and the
ethnohistoric data to classify the survey sites into zones. Using an alternative line of
evidence, the lithic artifacts for these sites are examined (with focus on the obsidian) to
determine if any other artifact type was used by Pollard in these instances to assign a
specific zone. It may be that in certain cases where I have classified the site as a

commoner zone, Pollard looked at the frequency of obsidian and thus classified the site

68

as a lower elite zone. Obsidian is considered both an elite and ritual item as evidenced in
the Relacio’n as well as in relation to high status burials, and therefore can be associated
with the elite social classes (Pollard 19992278).

Of the n=21 differences mentioned in Chapter Three, only four seem to have been
given a different zone due to the presence of lithic artifacts. Tz-8 was classified by
Pollard as a primary public zone (Pi) instead of a lower elite zone (Riii) because of a high
frequency of grey obsidian and prismatic blades. Tz-lO was categorized as an upper elite
site because of the presence of red and green obsidian, and an obsidian ear/lip plug, an
artifact described in the Relacio'n as an adornment only permitted to be worn by office
holders in the Tarascan state (RM 1956:201-202, Pollard 1999:265). Likewise, Tz-65
was also classified as a lower elite zone (Riii) because of the presence of an obsidian
lip/ear plug. Finally, Tz-93 was classified as a lower elite zone (Riii) by Pollard due to
its very high (n=76) frequency of grey obsidian as well as the presence of the more rare
red obsidian. The remaining sites that saw differing classifications from Pollard and me
were not affected by the presence/absence of lithic artifacts.

This leads me to believe that the cause of the differing classifications between
the lower elite and commoner zones is due to either the presence or absence of a low
frequency of elite ceramic variables. In each case, as presented in Table 9, those sites
classified into a lower elite zone were due to the presence of elite variables at lower
frequencies. The variables, such as the small supports, the resist and motif decoration,
spouts and spouted handles and miniature vessels, all have been defined in Chapter Two

as upper elite variables. I propose that the presence of these variables in low frequencies

69

is something that is statistically significant in terms of the cluster analyses performed, yet
is small enough to have been possibly overlooked by Pollard in her original analysis.

From this we can begin to evoke issues of behavior within the social classes
present at Tzintzuntzan. It seems that the lower elite ceramic assemblage is in fact more
closely related to the commoner (Ri) assemblage than it is to the upper elite (Rii) ceramic
assemblage, a zone whose sites were very clearly defined in the cluster analysis. The
distinctions between the lower elite and commoner classes though were much harder to
discern, even when working within the results of the cluster analysis. The distinctions
were slight yet significant and helped to define the difficult lower elite zones. From a
spatial standpoint, a closer look at Figure 6 shows that the lower elite sites form a buffer
between the upper elite sites and the commoner sites. This is the physical portrayal of the
continuum of the social classes as can be seen in the artifacts. The lower elite represent
an amalgamation of both extremes of the continuum, and it is my belief that due to the
nature of the spatial orientation of the sites, they are more closely related to the
(commoner in terms of ceramic characteristics than they are the upper elite. The spatial
orientation shows that the lower elite sites are in closer proximity to commoner areas than
they are the one upper elite area in the west-central portion of the settlement, especially in
the southern and north-east areas (see Figure 6). This speaks to the nature of the lower
elite’s role in Tzintzuntzan, and reinforces concepts of authority and the social hierarchy
within the Tarascan state. We know from the Relacion as well as Pollard’s research that
these lower elite were office holders and held the administrative positions that carried out
the will of the king in a “flow of authority from the center to the village” (Pollard

1993 : 124). These positions included overseeing tribute storage, payment and

70

redistribution, the military, and the secondary public zones dedicated to the state religion,
all of which included dealing and living directly among the larger commoner class. This
concept, which has been applied to the Tarascan state, also applies to an intra—site
analysis, explaining the close relationship between the lower elite and commoner artifact
assemblages at Tzintzuntzan.

Figure 7 — Public and Residential Zone Classifications from Cluster Analyses

Public Zones

 

\\ ll \1 ‘ it "a I, Pi (PrimarY)
I) X y , , % Pii (Secondary)

J, . , L9 7 i .’ Resldentlal Zones
/ o .. ' .. I ,
I; ‘ 1 ' Ri (Commoner)
| . |\ - Rii (Upper Elite)
i -. . . ‘ . x - Riii (Lower Elite)
l
\ . \ l : ] Tzln_survey_boundary
\ . \
\ l
'. l
\ /
\ /
l _ f , " f
o 0.5

2 Kilometers
J

71

The second issue that explains economic behavior within Tzintzuntzan is the
location of manufacturing zones. Pollard distinguishes between three types of lithic
workshops that are characterized as manufacturing zones. The primary function for the
type 1 lithic workshop (Li) is the production of lithic tools, primarily prismatic blades
(1993 :43). Type 2 lithic workshops produced other various artifacts, and are
characterized by a high frequency of red and green worked obsidian (1993143). Type 3
workshops included what Pollard thinks were the use and not the production of large
scrapers (1993:43-45). Although I am limited in my analysis of these zones, the spatial
distribution can be seen in relation to the public and residential zones and some
inferences may be made about the economic nature of these sites as well as the larger
economic structure of Tzintzuntzan.

As Figure 7 reveals, the manufacturing zones are located in close proximity to the
residential and public zones assigned by the cluster analysis. Some patterns that can be
seen from the spatial orientation of the workshops is that the Li workshops where the
blades are manufactured are located either within or in close proximity (<3 00 m) to a
public zone. The Lii workshops are more sporadically distributed spatially throughout
the settlement, and of the four sites designated as Lii workshops, three occur in
commoner zones. The final workshop, Liii, also occurs in the context of commoner
zones. It seems that in the primary zonal context for the Li workshops occur in or nearby
public zones and lower elite residential zones, whereas the Lii and Liii zones occur
entirely in commoner residential zones. The fact that the workshops that manufacture the
prismatic blades (Li) are closely related to public and lower elite zones tells us two things

about the possible nature of the Tarascan economy. First, the lithics manufactured at

72

these Li sites are being used directly by the public zones, explaining their close proximity
to each other. Second, it reinforces the concept that the obsidian prismatic blades are a
ritual item, used primarily by the elite classes. And third, it suggests that once again due
to the fact that these workshops are occurring in or near lower elite sites that the

producers are either attached specialists, distinct from the commoner manufacturers or

that they are being produced by the lower elite themselves.

Figure 8 — Location of Manufacturing Zones (Li, Lii, Liii)

a ‘ A,"
. I
\\\ ' ___/’,’
.1 .. ”g N

Public Zones
Pi (Primary)
" I 21: Pii (Secondary)
«I 9 A , Resrdenual Zones
," . i _ , RI (Commoner)
I . I
‘ I - Rii (Upper Elite)
1 - Riii (Lower Elite)
\
i .. . \ , » \ .12“. Survey Boundary
I i . i l A Li (Blade Workshops)
0 Lil (Non-Utllltarian)
"._ . x . Lill (Scraper)

 

2 Kilometers
4|

73

This leads to the third theme of this analysis, which is the zonal overlap of the
public and residential zones. For both the primary (Pi) and secondary (Pii) zones, there
are n=23 sites classified. Of these n=23, there are eleven which were classified by the
pipe cluster analysis. The remaining sites are either those that had no collected artifacts
in the original survey and the classifications were made by Pollard, or they are sites
located on or directly next to the main platform (Tz-25). Of the n=1] sites classified as
both residential and public by the cluster analysis, n=2 occurred in sites also classified as
upper elite residential sites, n=7 occurred in sites also classified as lower elite sites, and
the remaining n=2 sites occurred in sites also classified as commoner. In Figures 8a and
8b, I have created larger zones, distinct from the zonal categories assigned to each
individual site that the previous maps have displayed. The illustrations display the spatial
orientation of the public sites in relation to residential zones. The zones, as defined by
Pollard, are grouping of sites that are defined by zonal characteristics, and a cluster of
similarly categorized sites represent a zone (Pollard 1972271).

Thus, analyzing the sites defined by public zonal characteristics in terms of the
larger residential zones gives insight into the level of overlap of these two firnctional
zones. Of the n=23 sites defined by their public zonal characteristics and assigned by
either myself or Pollard, n=12 occur in lower elite residential zones, n=2 occur in upper
elite residential zones, and n=1 falls into a commoner residential zone. There is one site
that falls outside of any definable residential zone (Tz-27), because it lacked any
collected artifacts to be analyzed and the sole classification it was given by Pollard was

as a secondary public zone. The other remaining n=7 sites are those that are

74

Figure 9a — Residential Zones

    
 

. - Primary Public(P.i)
- Secondary Public (Pii)
Commoner (Ri)

. Mower Elite"(Ril)
-' Lower_Elite_i(Riii) '

4 :W'SufiaceScattér '7

survey "Sitesee- ._ .

  

1' Tzi n_'survey_boundary

r /
\ ./ '1 ,
\ ix g , r
\ \ffl /
\ /
\ _ _ _ ./
0 0.5 .1 2 Kilometers
l r r r l I r J J

 

75

Figure 9b —— Public Zones

. ,/,
-———\ ,’ l

 

I Public Zones
I: Pi (Primary)
@ Pii (Secondary)
Public Site Categories
‘\ - Pi (Primary)
‘ - Pii (Secondary)
\ C Survey Sites

 

x“ \ f_ :7 Survey Boundary

2 Kilometers
1

located in or around Tz-25, which is the main platform. These seven sites constitute the
only primary public zone at Tzintzuntzan (see Figures 8b and 9). Unfortunately due to
looting, extensive excavation and erosion, this area is practically devoid of artifacts, and

the cluster analysis shows no significant residential sites related to this primary zone.

76

The edificios which were discussed in Chapter Two and are located on the main platform
are evidence of habitations occupied by priests of the state religion. This type of
residency further demonstrates the distinct firnctionality of the primary public zone.

Figure 10 — The Primary Public Zone and Upper Elite Zone at Tzintzuntzan

  
 
 
 
 
 
 
 
 
   
 

Public Zones
Pi (Primary) ‘ x) _
m Pii (Secondary) 4‘ * '
Residential Zones _’
C}: Ri (Commoner) 1'
- Ril (Upper Elite) ‘
- Rlii (Lower Elite)

,__

L ~ _ Survey Boundary

- Structures

 

 

‘ Tz-91. .T1-'

Upper> Elite Residential Zonefl‘j;

 

 

 

 

0 175 350 700 Meters
L 1 m

It is clear that the overlap between elite residential zones and these public zones
are very closely related, and signify an important aspect to an urban tradition at
Tzintzuntzan. Pollard addresses this aspect of organization by stating “While the
religious subsystem was clearly distinct in the zoning of firnctions as revealed in
structures, features, and artifacts, the political subsystem was heavily embedded in the

social structure of the upper class” (1999:67). While I agree with Pollard that this is true

77

for the primary public zone as defined by the main platform and its immediate
surrounding sites, it is less clear when discussing the secondary public zones. As I have
stated earlier, the secondary zones seem heavily embedded into the residential zones,
especially in the case of the lower elite residential zones where the majority of the
secondary zonal characteristics may be witnessed. One very interesting example of this
is site Tz-12, which contains both a high count in upper elite ceramic variables as well as
an extremely high count for ritual items, such as obsidian prismatic blades and pipes.
Although I have classified the site based on primary public zonal characteristics, it is not
close enough to the main platform to place it into the primary public zone and is thus will

be considered a secondary public zone, along with site Tz-112 (see Figure 6).

Urbanism at T zintzuntzan

The last theme that emerges from the zonal analysis is the issue of urbanism at
Tzintzuntzan. The main concepts employed in the current urban theory on ancient cities
include issues of urban zoning and planning. Zoning, as defined by Pollard, is the
patterning of differential land use within a given area, with that differential land use
being based on differing functions of particular spatial units (Pollard 1999248). For this
analysis these spatial units are the survey sites. Planning is described by Michael Smith
in two new components of urban planning theory: (1) coordination among buildings and
spaces in a city. This includes the arrangement of buildings, formality and
monumentality of layout, orthogonality, other forms of geometric order, and access and
visibility (Smith 2007z7). The second component is (2) standardization among cities in

terms of urban architectural inventories, spatial layouts, orientation, and metrology.

78

Smith describes this theory in detail, and explains that the past literature focuses on ideals
of simplistic urban planning when referring to ancient cities”. In simple terms, the past
theory refers to cities as planned versus unplanned, without addressing the variations
described above (Smith 200725). This discussion will utilize the updated urban theory
that Smith proposes in conjunction with the conclusions made by Pollard in her previous

research on a Tarascan urban tradition.

Urban Planning

Using the concepts laid out by Smith, we find that in fact Tzintzuntzan contains a
lesser level of urban planning. Both Pollard’s and my analysis show that Tzintzuntzan
exhibits a relatively low degree of standardization and layout. However, the surviving
buildings do exhibit components of what Smith discussed as being planned. There is
definite evidence that the construction of the two central platforms represent a planned
attempt at coordination of a space. The main platform (Tz—25) consists of the five
yacatas which are standardized both between each other as well as represent an
architectural form present elsewhere in the basin”. They are laid at an alignment 45
degrees east of north, and they are most definitely monumental in their construction and
are visible within the known settlement. The second platform consists of the Santa Ana
platform (Tz-29), which is the center of the upper elite residential zone. It is also an

elevated platform that was constructed 45 degrees west of north and contains the remains

 

‘7 Smith defines an ancient city to “designate preindustrial urban settlements outside of the Classical world
of Greece and Rome.” He goes on to state that this includes Mediterranean cities before the sixth century
BC, and cities in other parts of the world prior to European conquest or industrialization (Smith 2007z4).
‘8 Ya’catas that exhibit identical structural form and architecture are also present at Ihuatzio and
Erongaricuaro (Pollard 1980: 685-690).

79

of structures. It is thought that the royal family including the king lived at this site. (See
Figure 8 for a map of both platfonns)
Table 10 — Survey Sites and Site Size Categorized by Zone

*(Ri = Conunoner, Rii = Upper Elite, Riii = Lower Elite, Pi = Primary Public, Pii= Secondary
Public)

 

 

Rii“ (ha) Riii“ (ha) Ri* (ha) Ri (com a) Pi* (ha)

Tz-12 1.24 Tz-8 0.47 Tz-l 1.04 Tz-56 0.15 Tz-l 1.04
Tz-16 0.21 Tz-10 0.57 Tz-2 1.36 Tz—58 0.09 Tz-2 1.36
Tz-17 0.58 Tz-ll 0.58 Tz-3 0.15 Tz-6O 0.13 Tz-3 0.15
Tz-18 0.35 Tz-15 0.54 Tz~4 0.03 Tz-62 0.18 Tz-6 0.2
Tz-85 0.45 Tz-l9 0.58 Tz-5 0.46 Tz-65 0.13 Tz-7 0.15
Tz-90 0.39 Tz-20 0.93 Tz-6 0.2 Tz-67 0.06 Tz-8 0.47
Tz-9l 0.35 Tz-22 0.92 Tz-7 0.15 Tz-68 0.35 Tz-9 0.38
Tz-23 0.87 Tz-9 0.38 Tz-73 0.1

Tz-39 0.16 Tz-13 0.12 Tz-74 0.1 Pii“ (ha)

Tz-53 0.28 Tz-14 0.96 Tz-8l 0.29 Tz-ll 0.58
Tz-59 0.99 Tz-21 0.34 Tz-86 0.16 Tz-12 1.24
Tz-61 0.78 Tz-24 0.12 Tz-87 0.15 Tz-70 1.21
Tz-63 0.28 Tz-32 0.1 Tz-89 0.27 Tz-75 0.43
Tz-70 1.21 Tz-34 0.28 Tz-93 2.07 Tz-76 0.41
Tz-75 0.43 Tz-35 0.21 Tz-95 0.15 Tz-87 0.15
Tz-76 0.41 Tz-36 0.12 Tz-97 1.19 Tz-90 0.4
Tz-77 5.8 Tz-37 0.16 Tz-99 0.3 Tz-104 0.57
Tz-78 6.04 Tz-38 0.19 Tz-102 0.39 Tz-112 2.9
Tz-84 0.34 Tz-40 0.01 Tz-104 0.57 Tz-114 0.39
Tz-107 0.84 Tz-41 0.08 Tz-105 0.8 Tz-117 6.1
Tz-l 10 0.12 Tz-42 0.06 Tz-106 0.17
Tz-112 2.85 Tz-43 0.05 Tz-108 0.36
Tz-114 0.39 Tz-46 0.11 Tz-109 1.74
Tz-l 17 6.09 Tz-47 0.11 Tz—110a 0.07
Tz-49 0.49 Tz-l 11 0.71
Tz-50 0.14 Tz-l 13 1.99
Tz-51 0.14 Tz-115 0.06
Tz-54 0.4 Tz-l 19 0.93
Tz-55 0.29 Tz-120 0.35

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The extent of the planning at Tzintzuntzan coincides with what, according to
Smith, occurs in the earliest cities. He states that in most cases the central district, also
termed the urban epicenter, exhibit planning whereas the surrounding residential zones do

not (200727). The epicenter at Tzintzuntzan is clearly defined as the central area

80

extending from the upper elite residential zone on the west-central area to the primary
public zone in the east-central area of the settlement. It is clear that the zones, both
residential and public, and the architecture represent a planned center, whereas the
surrounding zones are not formally arranged. Regardless of the informal arrangement of
the zones, they still exhibit planning to the extent that they are zoned, and represent areas

of differential land use defined by the function of the sites that constitute them.

Table 11 — Summary of Spatial Extent of Zoned Sites at Tzintzuntzan

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Zone # Sites Total Hectares
Upper Elite (Rii) 7 8.83
Lower Elite (Riii) 24 32.5
Commoner (Ri) 58 22.26
Primary Public (Pi)19 14 14.98
Secondary Public (Pii) 11 14.4
Sites + Surface Scatter 145.80 ha

Residential Zone Sites 63.59 ha 43.6% of total
Public Zone Sites 29.38 ha 20.2% of total

 

 

Urban Zoning

The concept of urban evolution theorizes that urban centers become increasingly
larger and complex through time (Pollard 1999268). This is clear in the zoning
characteristics of an urban settlement, and as Pollard points out “can be taken as a
postulate that as urban settlements evolve, there is an increase in the number and kind of
zones, and in the degree of specialization” (1999268). This analysis has discussed three

types of zones, residential, public, and manufacturing. Due to the nature of Pollard’s

 

‘9 Primary Public refers to the sites analyzed and classified in Chapter 2, as well as the remaining sites that
compose the Primary Public Zone (see Figure 9 for map of the Primary Public Zone)

81

analysis and the use of more variables than just the ceramics to identify zones, she has
identified two more zones (agricultural and commercial) in addition to the previous three.
These five zones constitute the urban settlement of Tzintzuntzan.

This analysis has shown that of the three zones examined, all three show a high
level of overlap and embeddedness within each other. The most embedded can be seen in
the residential zones, as they are multipurpose in terms of the political and social
structure of Tzintzuntzan. Pollard makes this point, as she states “From the Relacion it is
clear that the king and lords, the upper class, occupied the dominant positions within the
political administration as their primary status” (1999:67). My analysis supports this
statement, as the analysis in no way provides any separate zone that is defined by any
political function, whether it is administrative, judicial, or legislative.

The second aspect of zonal overlap is one that exists between the residential and
public zones. It is clear that due to the nature of the zonal classifications for the survey
sites, that the majority of the lower and upper elite classified sites also contained ceramic
variables characteristic of a religious and/or ritual function. The religious and ideological
systems can be removed, unlike the political system, from the residential zone and
identified as a distinct zone and are defined as public zones. This is very evident in the
primary public zone, which includes the main platform at Tzintzuntzan (Tz-25), and is
almost totally removed from any residential zone. Pollard attributes the zonal overlap to
fact that the “political system dominated the religious hierarchy, utilizing the ideological
system to reinforce political power and legitimacy” (1999:67).

My analysis does support this fact, but also shows a more embedded role between

the secondary public zones and the lower elite residential zones at Tzintzuntzan than may

82

have been previously thought. It may be that in these lower elite residential areas all
functions of the state, including the political, economic and religious aspects, are being
combined and carried out at a more local level by the lower elite class. The location of
the lithic workshops, especially the blade workshops (Li), support this fact, as they are
either in very close proximity or are contained within lower elite sites. It seems that the
lower elite are in charge of the more localized state functions, ranging from production of
ritual items to the rituals themselves, of religious function through their residential zones.
These zones seem to have more embedded functions than perhaps was previously
thought, as they envelop all aspects of what is defined as a public space in terms of the

state .

Conclusion

The ceramic analysis of the Tzintzuntzan survey sites was done to test the original
interpretations made by Helen Pollard about the nature of the settlement and its urban
characteristics. The null hypothesis for my analysis was that there would be no
difference between Pollard’s original zonal classifications and my zonal classifications.
The analysis showed that 73% of the time, Pollard and I agreed as to the zonal
classification given to the individual sites. The remaining instances of disagreement
could be explained in part by the larger number of variables that Pollard used to define a
site’s zonal category, as well as the discovery of a distinct patterning between the ceramic
assemblages of the lower elite class and the commoner class. Overall, I have failed to
accept my hypothesis but I believe that the major components of Pollard’s interpretations

have been supported by my analysis of the Tzintzuntzan ceramics.

83

Speaking to the urban nature of Tzintzuntzan, my analysis parallels Pollards in
that it shows Tzintzuntzan to display a low to moderate level of urban planning overall,
with an urban center displaying a higher level of urban planning and the surrounding
zones displaying a low level of planning. To this point, I agree with Pollard in that “the
implication is that Tzintzuntzan was a small, relatively less complex urban form at an
early phase in urban evolution” (1999268). This is also discussed by Smith, as he found
that “in a comparison of the sizes of Postclassic Mesoamerican cities, ...the absolute area
of the urban epicenter (i.e., the planned proportions of cities) was strongly associated
with the administrative level of cities; the most powerful capitals had the largest planned
areas” (2007230). Though I cannot speak to the effect of Tzintzuntzan’s overall impact
on the larger landscape, this point does explain the highly embedded use of space by the
upper elite for state functions that encompassed the economic, political, and religious
aspects of Tarascan society. Because of this overlap and the lack of space defined by
distinct functional characteristics, it is evident that Tzintzuntzan was still in an early
phase of state and urban evolution at the time of the Spanish contact.

The research provided in this thesis has created a methodological framework for
the reanalysis of an archaeological survey. Because it is now impossible to recreate the
original survey completed by Pollard in 1970, the reanalysis of the survey data must be
approached with a range of analytical tools and multiple lines of evidence in order to
provide the most thorough results. This resulted in a methodology that utilizes multiple
lines of evidence to create bridging arguments that relates social status to the artifact
assemblage. The methodology also utilizes statistical analyses that aid in site

classification based on the status sensitive assemblages as well as testing the results from

84

the site classification to Pollard’s original results. Finally, the analysis worked within a
geospatial database and GIS framework that helped to display, analyze, and map the
Tzintzuntzan archaeological survey data.

The goal of the bridging arguments was to link the concepts of a hierarchical
social system in a state-level society to specific ceramic variables that represent the
distinct social classes. To do so, multiple lines of evidence were needed to create these
associations. They included excavations at Tzintzuntzan within the INAH official
archaeological zone, which includes the main platform (Tz-25), excavations of burials at
the major center of Urichu within the Patzcuaro Lake Basin, and the habitation and burial
excavations at the site of Las Milpillas at Zacapu outside of the lake basin. Also,
research of the ceramic production and specialization within the lake basin added the
economic and manufacturing aspects of the association between social class and artifact.
And finally, the ethnohistoric evidence as seen in the Relacion de Michoacan (1956)
provided firsthand accounts of Tarascan life that associated aspects of Tarascan society to
social class and ceramic artifacts. These areas of evidence helped to create specific
categories that were able to detect status-sensitive distinctions in the ceramic assemblage.

The second aspect of the methodology provided in this thesis is the use of both
hierarchical and K-means cluster analysis and non-parametric statistical tests to analyze
the sites in terms of the status-defined ceramic variables. In Pollard’s original analyses,
statistical and systematic tests were not done to derive her zonal classifications. The
statistical tests used in this analysis were important because they provided a thorough
method with a higher degree of confidence when working with small sample sizes in

count data form.

85,

The combination of these multiple lines of evidence and statistical analyses
provide a methodology that isn’t limited to the site of Tzintzuntzan. The method may
now be applied elsewhere in the Patzcuaro Lake Basin to test the range and variability of
Tarascan social hierarchy in terms of the archaeological record. By utilizing this
methodology at such sites as Urichu, Erongaricuaro, Pareo, J aracuaro, and others within
the lake basin, the artifact-status associations can be tested and may provide uniform
artifact assemblages that display homogeneity between social classes across Late
Postclassic period sites. The application of this methodology may also show distinctions
between sites, and we may find that, for example, the lower elite are represented
differently in the archaeological record at the capital of Tzintzuntzan than they are the
center of Erongaricuaro. Evidence such as this may provide a closer look at social
distinctions within Tarascan society, and help us to better understand the dynamic aspects

of the Tarascans both at the emergence of the state and at the time of Spanish contact.

86

APPENDICES

87

Upper Elite Variables: Cluster Analysis
-Hierarchical Cluster Analysis (n=89 sites)

Figure 11 — Upper Elite Variables: Dendrogram Using Average Linkage (Between
Groups)

Rescaled Distance Cluster Combine

C A S E 0 5 10 15 20 :5
Label (Site) Num + ————————— + ————————— + --------- + ————————— + --------- +
47 86 —4
49 87 .3
60 11 -fi
115 7: .—
40 79 __
113 64 —-
119 65 -—
1 27 —-
13 31 —-
4 2 ——
7 2 ——
35 76 —~
6 22 —~
81 36 —+
14 '2 -—
65 15 .3
9 20 ——
68 18 ——
132 48 ——
2 28 ——
36 77 __
97 54 “—1
58 9 J
104 57 -4
108 61 —J
43 80 —4
2'4 73 —i
34 75 -+
56 7 A
54 6 —e
42 81 .3
10 19 —e
11 29 ——
89 52 ——
41 84 ——
110a 71 -
50 88 ——
109 2 ——
21 8 ——
120 66 .3
117 68 -
3 ’P -—
105 58 -
55 5 ——
5 25 -m
73 4 ——
46 85 -—
106 60 —-
51 89 -—
39 82 -
37 78 —‘
67 17 —J
93 51 -e
38 83 -4
74 3 —-
32 74 -
61 12 ——

 

88

l.) O‘l
. O) C) [‘3

*3 0" C!)
J (N

78
90
107

1

l;

O
O

l
l

lpl—*(Ol—‘U‘Chxl
‘DUJO

OerH ~41an

HP»

)_i
U)

L_|llllllllllllllllllLllll
J
*I

no)
i)\l

O) m LA.‘
J (‘1 to ([1

l O

46
45
49
30
44

 

 

 

 

 

 

 

 

 

89

 

Table 12 — Upper Elite Variables: Cluster Membership — K-Means Cluster Analysis (11 =

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

89 sites)

Case # Site Cluster Distance
1 75 1 7.824
2 76 1 5.831
3 74 1 4.575
4 73 1 3.513
5 55 1 3.905
6 54 1 2.136
7 56 l 4.04
8 53 1 8.05
9 58 1 3.116

10 59 3 8.192
11 60 1 3.908
12 61 1 6.204
13 62 1 5.877
14 70 1 11.007
15 65 1 3.732
16 63 1 13.691
17 67 l 3.31
18 68 1 3.819
19 10 1 2.475
20 9 1 3.732
21 8 1 5.292
22 6 1 3.555
23 4 1 3.908
24 7 1 3.908
25 5 1 3.555
26 3 1 3.534
27 l 1 3.908
28 2 1 3.021
29 ll 1 2.055
30 12 2 15.906
31 13 1 3.908
32 14 1 3.585
33 77 1 7.112
34 78 1 10.467
35 19 1 5.486
36 81 1 3.208
37 20 1 6.269
38 21 1 2.692
39 22 1 6.956
40 23 1 8.712
41 18 3 8.913
42 86 1 7.617
43 15 1 7.856
44 16 2 17.564
45 17 2 12.728

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Case # Site Cluster Distance
46 85 3 10.96
47 84 1 8.541
48 102 1 3.468
49 91 2 12.021
50 90 1 10.905
51 93 1 2.772
52 89 1 2.618
53 87 1 4.949
54 97 1 3.235
55 95 1 4.68
56 99 1 9.648
57 104 1 2.91
58 105 1 3.778
59 107 1 8.847
60 106 1 3.415
61 108 1 3.461
62 109 1 2.253
63 111 1 4.243
64 113 1 3.908
65 119 1 3.908
66 120 1 3.623
67 114 1 6.2
68 117 1 2.811
69 112 1 8.191
70 110 1 3.189
7] 110a 1 2.552
72 115 1 3.908
73 24 1 3.538
74 32 l 4.68
75 34 1 3.22
76 35 1 3.963
77 36 1 3.021
78 37 1 2.803
79 40 1 3.908
80 43 1 4.249
81 42 1 2.296
82 39 1 3.492
83 38 1 3.372
84 41 1 1.754
85 46 1 2.772
86 47 1 3.908
87 49 1 3.908
88 50 1 2.061
89 51 1 3.716

 

90

 

Table 13 — Upper Elite Variables: Final Cluster Centers — K-Means Cluster Analysis (11 =
89 sites)

Final Cluster Centers Cluster
2

 

 

Tarerio Cream 27

 

16

H

Yaguarato Cream Number of Cases in each Cluster

 

 

Siphio Grey 18 LCluster l 82

 

 

Bowls: convex wall 19 1 2 4

 

 

out-sloping wall 3 3

 

 

minibowls: total Valid 89

 

 

 

 

Plates: convex wall Missing 0

 

 

 

flat base

 

spouted vessel: 2.7.1,5,4

 

smut handle 2.7.2
small-spider

 

 

conical

 

flat-oblong

 

foot—claw

 

flared

 

hollow cylinder

 

solid foot shape

 

small supports

 

hollow rattle cylind.

 

" " wider top

 

" nub

 

Motifs: 3.3.2

 

3.3.8

 

3.3.7

 

3.3.9

 

3.3.1

 

3.3.3

 

3.3.4

 

3.3.5

 

3.3.6

 

3.3.10

 

3.3.11

 

3.3.12

 

3.3.13

 

3.3.14

 

Motifs

U)

 

Applique conical Pellet

 

fl

negativgtotal)

 

ngg/red

 

negfied/white

 

red,white&black/cream

 

leg/cream

 

 

 

 

 

 

OOOOOOONOOOOOOOOOOOOHOOOOi—tOOOOOOOOOOOOOv—I-iNNH
OOOOOI-‘OOOOOOOOr-‘OOOh-‘Or—IUJNOOHNOOOOOH—OWOHONNO\A0003

H
UIOOWt—‘OOQOOOOi—‘ONNONNOOOOOOOQOOOOOv—ANI-‘ON-ho

neg./red&white/cream

 

91

Lower Elite Variables: Cluster Analysis
-Hierarchical Cluster Analysis (n=84 sites)

Figure 12 -- Lower Elite Variables: Dendrogram Using Average Linkage (Between
Groups)

 

 

Rescaled Distance Cluster Combine

C A S E 0 5 10 15 20 25
Label Num + ~~~~~~~~~ + ————————— + --------- + --------- + ————————— +
13 3O 1
40 74 —
7 24 —-
97 49 —
65 15 —
4 23 --
6 22 —
6O 11 —
113 59 —
1 27 —
119 60 —
67 17 —
108 56 —
102 44 —
58 9 —
-1 37 ~—
104 52 ._
3b 72 a
56 7 —
68 re 4
2 28 —
14 31 —
5C) 83 -—
37 73 —
34 so _
81 35 —
47 81 —
9 20 —-
110a 66 —
35 71 _
49 82 -—
109 57 —
43 75 —
106 5 ——
42 76 —
55 5 —
41 79 —-->—-«
11 29 —
89 47 -
120 61 —l
5 25 4
3 26 ~
86 41 -—~
93 46 -—
46 80 -—
74 3 —-
10 19
105 53 T
62 13
32 69 —]
73 4
70 14 3‘ -
111 58
95 50
115 67

 

 

22 38
87 48

3“—
76 3
112 64 :i—_i—l I

 

92

54
19
C39
51
117
110
114

i-—' -\l ('1) H (T) *9 \3 ON 1” 0) ()1 0‘ C0 \1 LA) (.0 lo ix)
UlU‘OCOl'"C*U‘LMKO-D-\lCOlprh-C;
\l

00 U1

0x

(I: Ln ()3
>2. F—J ..Ls

ONO) 0)

(Jo
«1000:3an

anssmgaasim
llllllllll
llLill n

|>—-'J‘a
O\U1

iJ-b-
i—‘I‘O

.2.
OH

 

 

 

 

 

 

 

 

 

 

 

93

 

 

 

Table 14 —Lower Elite Variables: Cluster Membership - K-Means Cluster Analysis (11 =
84 sites)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Case# Site Cluster Distance Case # l Site Cluster Distance
1 75 1 9.975 43 84 2 5.254
2 76 1 5.008 44 102 4 2.843
3 74 4 3.941 45 90 2 8.316
4 73 4 5.737 46 93 4 3.78
5 55 4 3.093 47 89 4 2.609
6 54 4 7.562 48 87 2 6.583
7 56 4 2.285 49 97 4 2.897
8 53 2 3.433 50 95 4 5.482
9 58 4 2.381 51 99 4 6.016

10 59 1 6.376 52 104 4 1.727
11 60 4 2.623 53 105 4 3.976
12 61 2 3.845 54 107 2 4.428
13 62 4 5.26 55 106 4 1.776
14 70 4 5.806 56 108 4 2.629
15 65 4 3.093 57 109 4 2.03
16 63 1 7.815 58 111 4 4.798
17 67 4 1.88 59 113 4 2.801
18 68 4 2.33 60 119 4 2.862
19 10 4 5.264 61 120 4 2.974
20 9 4 2.063 62 114 1 5.959
21 8 1 5.406 63 117 1 4.591
22 6 4 2.707 64 112 1 4.936
23 4 4 2.862 65 110 1 3.496
24 7 4 3.176 66 110a 4 2.713
25 5 4 2.474 67 115 4 5.978
26 3 4 3.471 68 24 1 4.335
27 1 4 2.481 69 32 4 5.114
28 2 4 3.003 70 34 4 2.662
29 11 4 2.308 71 35 4 3.031
30 13 4 3.176 72 36 4 2.088
31 14 4 2.763 73 37 4 2.367
32 77 1 4.01 74 40 4 3.176
33 78 2 7.209 75 43 4 2.389
34 19 1 7.335 76 42 4 1.592
35 81 4 2.675 77 39 2 4.687
36 20 2 5.718 78 38 1 4.893
37 21 4 2.794 79 41 4 2.255
38 22 2 6.264 80 46 4 4.435
39 23 2 3.774 81 47 4 3.126
40 18 3 0 82 49 4 3.381
41 86 4 4.493 83 50 4 2.323
42 15 1 7.714 84 51 4 4.819

 

 

 

 

 

 

 

 

 

 

 

94

Table 15 — Lower Elite Variables: Final Cluster Centers — K-Means Cluster Analysis (n =
84 sites)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Final Cluster Centers Cluster

1 2 3 4
Tariacuri Brown 6 6 6 1
Siphio Grey 2 2 3 1
convex wall 0 0 0 0
spouted vessel: 2.7. 1,5,4 0 0 1 0
Yaguarato Coarse 2 1 1 1
Jars - everted rim 7 5 12 2
regular — solid conical 0 0 0 0
Motifs 3 6 6 1
solid spider 0 0 0 0
incurved rim 0 0 3 0
white/red 0 0 7 0
red/white 0 l 2 0
red&white/white 0 0 0 0
red/cream 8 3 21 1
white/cream l l 2 0
red&white/cream 4 9 6 1
white all over red/cream 0 0 1 0
all over red/cream 1 0 16 0

 

 

 

 

 

 

Number of Cases in each Cluster

 

 

 

 

 

 

 

 

 

 

Cluster 1 l4
2 1 l

3 l

4 58

Valid 84

Missing 0

 

95

Commoner Variables: Cluster Analysis
-I-Iierarchical Cluster Analysis (n=61 sites)

Figure 13 — Commoner Variables: Dendrogram Using Average Linkage (Between

 

 

 

 

 

 

 

 

 

 

 

 

 

Groups) Rescaled Distance Cluster Combine
C A S E 0 5 10 1 20 25
Label Num + ————————— + --------- + ————————— + --------- + ————————— +
95 33 —
5O 60 —
2 21 -
97 32 —
93 29 —
110a 44 1
104 35 —l
36 50 —
5 6 —
5 18 —
37 51 —
56 5 —
3 19 --
113 41 —
119 42 -—
l 20 -—
115 45 —
40 52 —
49 59 —
65 10 —--——e
34 48 —-l
35 49 —(
13 23 4
l4 :4 —(
4 l6 4
7 7 —
6 15 -—
6O 7 —
120 43 —
42 54 —
9 14 -——1
106 37 — "—1
a9 30 4
68 1.:
81 2'5 3—
74 l
21 26 j—
102 28
105 30 j—
73 2 —-
55 3 —
46 57 -
51 61 — —-
43 53 ~—
47 58 -4
10 13 --
41 56 -—
109 39 —
99 34 -'
62 8 —-1
ll 22
108 39 T—
111 40 —
67 11 j—
86 27
54 4
24 46
87 31 :1——
32 47
7O 9

 

 

96

Table16 — Commoner Variables: Cluster Membership — K-Means Cluster Analysis (n=61
sites)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Case # Site Cluster Distance Case # Site Cluster Distance
1 74 1 1.938 32 97 1 0.971
2 73 1 1.997 33 95 1 1.491
3 55 2 1.685 34 99 2 1.883
4 54 2 4.002 35 104 1 0.471
5 56 1 1.253 36 105 2 2.38
6 58 1 0.471 37 106 1 1.646
7 60 1 1.681 38 108 l 2.28
8 62 2 1.441 39 109 2 1.668
9 70 3 0 40 111 1 2.798

10 65 l 1.29 41 113 1 1.196
11 67 1 3.439 42 119 1 1.467
12 68 1 2.099 43 120 1 1.491
13 10 2 1.771 44 110a 1 1.225
14 9 1 1.646 45 115 1 0.971
15 6 l 1.29 46 24 2 3.525
16 4 1 1.29 47 32 2 3.467
17 7 1 1.29 48 34 1 1.29
18 5 1 1.574 49 35 1 1.29
19 3 1 0.843 50 36 1 0.471
20 l 1 0.971 51 37 1 1.41
21 2 1 0.971 52 40 1 1.29
22 ll 2 3.099 53 43 2 2.048
23 13 l 1.29 54 42 1 1.491
24 14 1 1.29 55 38 2 3.003
25 81 1 2.099 56 41 2 1.771
26 21 l 2.39 57 46 2 0.698
27 86 1 5.293 58 47 2 1.787
28 102 1 2.38 59 49 1 1.29
29 93 1 1.225 60 50 1 1.491
30 89 1 2.032 61 51 2 1.668
31 87 2 3.432

 

 

 

 

 

 

Table 17 — Commoner Variables: Final Cluster Centers — K-Means Cluster Analysis
(n=61 sites)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Final Cluster Centers Cluster Number of Cases in each Cluster
1 2 3

Taruacuri Coarse 0 1 l 1 Cluster 1 43

Yaguarato Coarse 0 2 2 17

lchupio Coarse 0 1 l 3 1

Bowls - convex wall 1 1 ll Valid 61

incurved rim 0 0 0 Missing 0

Jars - everted rim 1 4 2

incurved 0 0 1

regular - solid conical 0 0 0

unslipped-unpainted 0 0 2

 

 

 

 

 

 

97

Public Zone Variables: Cluster Analysis
-Hierarchical Cluster Analysis (n=89 sites)

Figure 14 — Public Zone Variables: Dendrogram Using Average Linkage (Between
Groups)

Rescaled Distance Cluster Combine

C A S E 0 5 10 15 20 25

Label Num + ————————— + --------- + --------- + --------- + --------- +
109 62 —4
50 88 —4
7 3 -e
21 38 ——
93 51 ——
62 13 -
37 78 -—
105 58 -
42 81 ——
65 15 ——
4O 79 -
47 86 —-
54 6 ~—
3 76 ——
36 77 —-
119 65 ——
110a 7- ——
15 43 -—
113 64 ——
13 31 ——
81 36 ——
1 :7 _
2 28 .3
5 :5 __
3 26 __
4 23 __
7 24 -—
9 20 -—
6 :3 __
67 17 ——(
68 18 -—
58 9 ——
60 11 -
106 60 —
84 47 __
55 5 .3
115 72 ——
22 39 -—
111 63 ——
63 16 -e
39 82 -—
108 61 —~

7 45 —
91 49 .3
107 59 -—
46 85 —-
73 4 -—
23 4O -
59 10 -
61 12 —i
19 35 -n
10 19 —~
16 44 -
99 56 -—
43 80 -
97 54 -
14 32 —-
102 48 -
53 8 -‘

 

98

3.2
86
38
89
24
49
110
120
76
87
20
18
34

O
kl

95

51
56
'77
78
41
90
117
104
114
'70
75
11
12

117’

74
4C
8'

52
73
87
7O

66

53

30

L111

 

1111

111111

1

 

 

 

 

Ll

 

 

99

Table 18 — Public Zone Variables: Cluster Membership — K-Means Cluster Analysis

(n=89 sites)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Case # Sites Cluster Distance
1 75 2 4.347
2 76 2 1.132
3 74 3 0.605
4 73 3 1.195
5 55 3 1.016
6 54 3 0.72
7 56 3 3.21
8 53 2 2.802
9 58 3 0.72

10 59 3 1.031
11 60 3 0.72
12 61 3 1.268
13 62 3 0.605
14 70 2 2.369
15 65 3 1.031
16 63 3 1.447
17 67 3 0.72
18 68 3 0.72
19 10 3 1.706
20 9 3 0.72
21 8 3 2.264
22 6 3 0.72
23 4 3 0.72
24 7 3 0.72
25 5 3 0.72
26 3 3 0.72
27 l 3 0.72
28 2 3 0.72
29 ll 2 2.919
30 12 1 0
31 13 3 0.72
32 14 3 1.892
33 77 3 2.834
34 78 3 3.317
35 19 3 2.168
36 81 3 0.72
37 20 2 2.126
38 21 3 0.605
39 22 3 1.088
40 23 3 1.661
41 18 2 1.938
42 86 2 1.361
43 15 3 0.72
44 16 3 1.661
45 17 3 1.361

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Case # Sites Cluster Distance
46 85 3 1.733
47 84 3 1.207
48 102 3 2.658
49 91 3 1.361
50 90 2 3.162
51 93 3 0.605
52 89 2 1.661
53 87 2 1.526
54 97 3 1.9
55 95 3 1.733
56 99 3 1.596
57 104 2 2.17
58 105 3 1.031
59 107 3 0.955
60 106 3 1.195
61 108 3 1.488
62 109 3 0.605
63 111 3 1.457
64 113 3 0.72
65 119 3 0.72
66 120 2 2.192
67 114 2 2.59
68 117 2 2.69
69 112 4 0
70 110 2 1.975
71 110a 3 0.72
72 115 3 1.315
73 24 2 1.067
74 32 2 2.192
75 34 3 2.303
76 35 3 0.72
77 36 3 0.72
78 37 3 1.143
79 40 3 0.72
80 43 3 2.075
81 42 3 1.031
82 39 3 1.394
83 38 2 1.361
84 41 2 3.192
85 46 3 1.859
86 47 3 0.72
87 49 2 1.838
88 50 3 0.605
89 51 3 2.413

 

100

 

Table 19 — Public Zone Variables: Final Cluster Centers - K-Means Cluster Analysis
(n=89 sites)

Final Cluster Centers Cluster

 

Number of Cases in each
Cluster

 

 

Bowlsz-thin wall bevel liJ) '- LCluster

 

 

" " conical bowl

 

 

thick modeled

 

 

AVON—-

bulbous base

 

 

conical supmrts Valid

 

 

 

 

Bits: -Plain Missing

 

 

 

ground around
beveled
pattern bumish
incised
Circular Cross-section 2
Simple Twist

Incised twist four
Incised Twist Three
Twisted & Incised
Twisted & Scalloped
Oval Cross-section
unslipped: pattern bur.
stamp incised

incised str. Line
incised wavy line
Scalloped

"corn cob"

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

OOOOOOO—‘OOOHOU‘OOOOOOOOOON

 

 

OOOOOOOHbNU‘Nm—‘OHNI—‘OHONNH
OOOOOOOOOOOOt—OOOOOOOOOOW
OO~OOOWOAONUJ¢NOHOOOOHO~15

 

 

101

Table 20 — Public Zone Variables: Cluster Membership — K-Means Cluster Analysis
(n=87 sites) (Tz-12 and Tz-112 removed and variables pipe bits and bowls removed)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Case # Sites Cluster Distance Case # Sites Cluster Distance
1 75 1 3.051 45 85 2 1.338
2 76 1 1.444 46 84 3 1.063
3 74 3 0.747 47 102 2 2.29
4 73 3 1.384 48 91 3 1.159
5 55 3 0.993 49 90 1 1.281
6 54 3 0.448 50 93 3 0.747
7 56 2 2.499 51 89 2 1.587
8 53 2 2.038 52 87 1 1.757
9 58 3 0.448 53 97 3 1.234

10 59 3 1.128 54 95 2 1.338
11 6O 3 0.448 55 99 3 1.219
12 61 3 1.358 56 104 1 1.482
13 62 3 0.747 57 105 3 0.879
14 70 1 1.659 58 107 3 1.063
15 65 3 0.879 59 106 3 1.063
16 63 3 0.975 60 108 3 1.219
17 67 3 0.448 61 109 3 0.747
18 68 3 0.448 62 111 3 0.975
19 10 3 1.371 63 113 3 0.448
20 9 3 0.448 64 1 19 3 0.448
21 8 2 1.304 65 120 2 1.032
22 6 3 0.448 66 114 1 2.102
23 4 3 0.448 67 117 1 1.556
24 7 3 0.448 68 110 2 1.075
25 5 3 0.448 69 1 10a 3 0.448
26 3 3 0.448 70 115 3 1.304
27 1 3 0.448 71 24 2 1.338
28 2 3 0.448 72 32 2 1.776
29 11 1 2.882 73 34 2 1.405
30 13 3 0.448 74 35 3 0.448
31 14 3 1.304 75 36 3 0.448
32 77 2 1.75 76 37 3 0.747
33 78 2 2.407 77 40 3 0.448
34 19 3 1.997 78 43 3 2.014
35 81 3 0.448 79 42 3 0.879
36 20 2 1.304 80 39 3 1.495
37 21 3 0.747 81 38 2 1.587
38 22 3 0.975 82 41 2 2.658
39 23 2 1.558 83 46 3 1.924
40 18 2 1.075 84 47 3 0.448
41 86 2 1.587 85 49 2 0.654
42 15 3 0.448 86 50 3 0.747
43 16 3 1.495 87 51 2 1.529
44 17 3 1.159

 

 

 

 

 

102

Table 21 — Public Zone Variables: Final Cluster Centers - K-Means Cluster Analysis
(n=87 sites)
(Tz-12 and Tz-112 removed and variables pipe bits and bowls removed)

Final Cluster Centers Cluster

 

 

conical supports
Circular Cross-section
Simple Twist

Incised twist four
Incised Twist Three
Twisted & Incised
Twisted & Scalloped
Oval Cross-section
unslipped: pattern bur.
stamp incised

incised straight line
incised wavy line
Scalloped

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

OOOOOOOOOOOOOM

 

OOOOOt—OOOHOQOH
OOOOOOOOOv—‘OWON

 

 

Number of Cases in each Cluster

 

 

 

 

 

 

 

 

l Cluster 1 9
2 22

3 56

Valid 87

Missing 0

 

103

Querenda White Ware: Cluster Analysis
-Hierarchical Cluster Analysis (n=89 sites)

Figure 15 — Querenda White Variables: Dendrogram Using Average Linkage (Between
Groups)

Rescaled Distance Cluster Combine

C A S E 0 5 10 15 20 25

Site Num + --------- + --------- + ————————— + --------- + --------- +
50 88 —1
51 89 —-
75 1 —

7 86 —
49 .7 _
38 83 —
46 85 -—
42 81 —
39 82 -—
40 79 —
43 80 —
36 77 —
37 78 —
34 75 —
3 76 —-
24 73 _
32 7 _
110 70 —
11.0a 71 _
114 67 —
112 69 —
119 65 —
120 66 —
111 63 -—
113 64 -
108 61 —
109 62 —
107 59 —-
106 60 —
99 56 -—
105 58 —
97 54 —
95 55 —
89 52 —
87 53
93 51 —
102 48 —
91 49 —
85 46 —
84 47 —
16 44 —-
17 45 —
86 42 -l

5 43
23 40 j
18 41 —-
21 38 —-
22 39 —
81 36 —
20 37 —
78 34 —
19 35 —
13 31 ——
14 32 —
11 29 —
12 30 -—
4 Z" —n
7 24 —(

 

104

G\\IG\UWO\O\U’1CCir-‘U1\lro
OUJ‘Jl)LuIDCDCDP—'

U70)-JU7(I\

LD (1.)

Q

t—‘ to to

}_l

\l

J:- l-_" U1 1“.) \J 0)") U7 P.)

CO

 

 

 

 

LlLlllllll LllllllJLLlllllll;

 

105

 

Table 22 — Querenda White Variables: Cluster Membership - K-Means Cluster Analysis
(n=89 sites)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Case # Sites Cluster Distance Case # 1 Sites Cluster Distance
1 75 1 0.291 45 17 1 0.291
2 76 1 0.291 46 85 1 0.291
3 74 1 0.709 47 84 1 0.291
4 73 1 0.709 48 102 1 0.291
5 55 1 0.291 49 91 1 0.291
6 54 1 0.291 50 90 1 0.291
7 56 1 0.291 51 93 1 0.291
8 53 1 2.709 52 89 1 0.291
9 58 1 0.709 53 87 1 0.291

10 59 3 0 54 97 1 0.291
11 60 1 0.709 55 95 1 0.291
12 61 1 0.709 56 99 1 0.291
13 62 1 2.709 57 104 1 0.709
14 70 2 3.5 58 105 1 0.291
15 65 1 0.291 59 107 1 0.291
16 63 2 3.5 60 106 1 0.291
17 67 1 2.709 61 108 1 0.291
18 68 1 0.709 62 109 1 0.291
19 10 1 0.291 63 111 1 0.291
20 9 1 0.291 64 113 1 0.291
21 8 1 0.709 65 119 1 0.291
22 6 1 0.291 66 120 1 0.291
23 4 1 0.291 67 114 1 0.291
24 7 1 0.291 68 117 1 0.709
25 5 1 0.709 69 112 1 0.291
26 3 1 0.709 70 110 1 0.291
27 1 1 0.709 71 110a 1 0.291
28 2 1 0.709 72 115 1 0.709
29 11 1 0.291 73 24 1 0.291
30 12 1 0.291 74 32 1 0.291
31 13 1 0.291 75 34 1 0.291
32 14 1 0.291 76 35 1 0.291
33 77 1 0.709 77 36 1 0.291
34 78 1 0.291 78 37 1 0.291
35 19 1 0.291 79 40 1 0.291
36 81 1 0.291 80 43 1 0.291
37 20 1 0.291 81 42 1 0.291
38 21 1 0.291 82 39 1 0.291
39 22 1 0.291 83 38 1 0.291
40 23 1 0.291 84 41 1 0.709
41 18 1 0.291 85 46 1 0.291
42 86 1 0.291 86 47 1 0.291
43 15 1 0.291 87 49 1 0.291
44 16 1 0.291 88 50 1 0.291

89 51 1 0.291

 

 

 

 

 

 

106

Table 23 — Querenda White Variables: Final Cluster Centers - K-Means Cluster Analysis
(11 = 89 sites)

Final Cluster Centers
Cluster

1 2
[Querenda White 0 12 31

 

 

 

 

 

 

 

Number of Cases in each Cluster

 

 

 

 

 

 

 

 

I Cluster 1 86
2 2

3 1

Valid 89

Missing 0

 

107

Table 24 - Querenda White Variables: Cluster Membership - K-Means Cluster Analysis
(n=86 sites)
(Tz-59, Tz-63, Tz-70 removed)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Case # Sites Cluster Distance Case # Site Cluster Distance
1 75 l 0 44 84 l 0
2 76 l 0 45 102 l 0
3 74 3 0 46 91 1 0
4 73 3 0 47 90 1 0
5 55 1 0 48 93 1 0
6 54 1 0 49 89 1 0
7 56 l 0 50 87 1 0
8 53 2 0 51 97 1 0
9 58 3 O 52 95 l 0

10 60 3 0 53 99 l 0
11 61 3 0 54 104 3 0
12 62 2 0 55 105 1 0
13 65 1 0 56 107 1 0
14 67 2 0 57 106 1 0
15 68 3 0 58 108 1 0
16 10 l 0 59 109 1 0
17 9 1 0 60 111 1 0
18 8 3 0 61 113 1 0
19 6 1 0 62 119 l 0
20 4 l 0 63 120 l 0
21 7 l 0 64 114 1 0
22 5 3 0 65 117 3 0
23 3 3 0 66 112 1 0
24 l 3 0 67 110 1 0
25 2 3 0 68 110a 1 0
26 11 l 0 69 115 3 O
27 12 l 0 70 24 l 0
28 13 l 0 71 32 1 O
29 14 l 0 72 34 1 0
30 77 3 0 73 35 l 0
31 78 l 0 74 36 1 O
32 19 l 0 75 37 1 0
33 81 l 0 76 40 l 0
34 20 1 0 77 43 l 0
35 21 1 0 78 42 1 0
36 22 l 0 79 39 l 0
37 23 1 0 80 38 1 0
38 18 l 0 81 41 3 0
39 86 1 0 82 46 1 0
40 15 1 0 83 47 1 0
4 l 1 6 l 0 84 49 l 0
42 17 1 0 85 50 1 0
43 85 1 0 86 51 1 0

 

 

 

 

 

 

 

 

 

108

Table 25 — Querenda White Variables: Final Cluster Centers - K-Means Cluster Analysis
(n=86 sites)
(Tz-59, Tz-63, Tz-70 removed)

 

 

 

 

 

 

 

 

 

 

 

 

Number of Cases in each Cluster Final Cluster Centers
I Cluster 1 67 Cluster
2 3 l 2 3
3 16 [Querenda White 0 3 1
Valid 86
Missing 0

 

 

 

 

109

Tarascan Ceramics

 

Figure 20- Tariacuri Cafe: Restricted Vcsscl ». Figure 21 - Tariacuri Cafe var. Santa Ana
(spout handle)

110

Tarascan Ceramics (cont)

  
    

 
    

- ' » cu:

tar-.4. 1.1.x. irra- ._ 7 is”: ..~ , .

Figure 22 - Tecolote Orange Mini Bowl Fragments Figure 23 - Yaguarato Coarse var. San
Bartolo (brasero)

 

Cm' CM CM .. . '
Figure 24 - Late Urichu Phase: Figure 25 - Late Urichu'Phase: Copujo Rojo
Arocutin Rojo con Negativo sobre Crema (small support)

111

Tarascan Pipe Stems

 

             

43“;

. e
n .h. 11 I V 7 5;. VP" »
Figure 28 - Circular Cross Section: Red Slip

igure 29 - Oval CrossiSection

112

Figure 30 — Tzintzuntzan Survey Map: Northwest Portion

 

 

 

 

 

 

 

 

o 0.125 0.25 0.5 Kilometols x‘ - ‘ ’
l 1 l l L I l | - ’
r232“ ,
T249“ \..
Tz-51 "7 Tz-so _ ,
I ~. ‘“ ”T248
N ‘ N \ Tz4i“
\ § \‘ T2.“ V
\ ‘ a
I \-.. .
: .Tz-45
r A ‘12? T2740: _ 12.33 ). ,.
I i" T all: \ 1‘ * '- _ r231
i T244 , 1 T2.” “
\ /,.., 1242
T243 Tz.34
‘ (v, . 2 1‘
- _ _ _ “‘ “g E M “(Tl-32
L : Survey Boundary ,‘ ‘ L K
- - - I 5 ”"7295 " Tz-33
- Structures .'
. ,. . 1.1.37
1 .. Survey Srtes ,'
___. [I 1'. 1.1.36
” “i ~H-V‘\_
’I
I, Tz'z‘}...
I! l .1 a
I

 

 

113

Figure 31 — Tzintzuntzan Survey Map: West Central Portion

 

 

 

 

 

' : Survey Boundary ‘\

— Structures

:: Survey Sites

 

 

Tz-16 1. .
Tz-91l; ,rz-so f .
‘ +2-92
Tz-iriz‘z.’
0 0.125 0.25 0.5 Kilometers
l I I l l

 

 

114

 

Figure 32 -— Tzintzuntzan Survey Map: East Central Portion

 

 
  
   

- Structures

I'II: Survey Boundary

- Survey Sites

 

 

 

 

 

 

 

0 0.125 0.25 0.5 Kilometers
l l L 1‘1

115

Figure 33 — Tzintzuntzan Survey Map: Lower Northeast Portion

 

 

 

 

 

- Structures

:_ :1 Survey Boundary

Fl Survey Sites

  

 

 

~ _ (1:12-57
‘ ‘r-\> 12.62“" ‘4. '
1“ \ /»... " ------
\ r‘v >, ’
“ , Tz-70 - ‘3 ..5" - r'
\.... Tz-71 ' x
._ ‘ I
N -, ,
z I
‘ __\' T2151 ”V I
'. ~ I
:~. I
( I
'_: .. l
I
l
'
— ’ ’
d ‘ — ' - —
O ‘ ‘ ‘
I
I
l
|
I
|
\
\
.

0.5 Kilometers

 

 

116

Figure 34 — Tzintzuntzan Survey Map: Upper Northeast Portion

 

 

0.5 Kilometers
J

 

 

 

 

 

0 0.125 0.25
g r_ 4 4 1
- Structures
5: - - -: Survey Boundary
{5 m Survey Sites ,1»
rife
_) Tz-54
Viz-53' -1 j
‘ frz-ss
72-55 - " .1.
4‘ Tz'73
., L
Tz-58
, — Tz-57 TZ-74 ’ x

 

, t
. .
. t
t
.
\
\
\
\l
I
I
I
l
.. '
I
i
Tz-75
I.
I
I
I
I
l
I
I
I
I
I
I
f
I
I
I
I
I
I
I
I

 

 

117

Figure 35 — Tzintzuntzan Survey Map: South Portion

- Structures

I. _ _
Survey Sites

_ ‘1 Survey Boundary

0.5
L

1 Kilometers
l

 

 

 

L 4L L L
'1 Tz-1033-. :
I ‘1 . ‘ "I I
I ‘5 ex ,. 4 I
.’ '- 72.101 m .'
,’ i e v-u~--— 1 33Tz-99 '
' . K; I’
I Tz-104 ,'
I A l
T241“ 12411 ~ ,1 .'
,’ Tz-110‘ f, fin-105 T2 106 :
,' Tz-109 \---'Tz-108 “\C ,‘I
1‘ i: K.“ h "“— -._:\ 1
. Tz-107 .
I I
: I
I l‘
: ,————. ._ __. W 1,. ~ 1" r3 ‘
. I . . Tz-113 " '1 - ‘-
: T2414; v ".~ i TZ-112 ’ '1 ”TI-11.8 “
| .. ‘ ‘ ‘ I ‘ \ H , |
. : “2‘ ~ ‘
I Tz-114a _ Tz-119"\M,' ‘-
\ '\ ‘

 

 

 

 

 

 

 

 

 

118

 

 

BIBLIOGRAPHY

119

Acosta, J.
1939

BIBLIOGRAPHY

Exploraciones arqueolégicas realizadas en el estado de Michoacan durante
los afios de 1937 y 1938. Revista Mexicana de Estudios Antropolo'gicos,
3(2): 85-99.

Cabrera Castro, R

1977-1978

1987

Exploraciones en T zintzuntzan, Michoacan: Decima temporada de
excavaciones (notas de campo).

Tzintzuntzan: Decima temporada de excavaciones. In Homenaje a Roman
Pifia Chan. Serie Antrpolo’gica 79, pp. 531-565. Instituto de
Investigaciones Antropologicas, Universidad Nacional Autonoma de
México, Mexico D.F.

Castro Leal, M.

1986
Earle, T.
2002

Gali, R.
1946

Hirshman, A.
2003

Lec'm, N.
1888

T zintzuntzan, capital de los tarascos. Gobiemo del Estado de Michoacan,
Morelia, Mexico.

Bronze Age Economics. Westview Press, Boulder, CO.

Arqueologia de Tzintzuntzan. Anales de Museo Michoacano 2(4): 50-62.

A Case Study in the Relationship between Social Complexity and the
Organization of Ceramic Production fiom the Lake Patzcuaro Basin,
Michoacan, Mexico. PhD dissertation, Michigan State University.

Las Yacatas de Tzintzuntzan. Anales del Museo Michoacano, Morelia
1:65-70.

Michelet, D., A. Ichon and G. Migeon (ed.)

1988

1998

Residencias, barrios y sitios postclasicos en el Malpais de Zacapu. 1.
Instituto Nacional de Antropologia e Historia, Querétaro.

Topografia y Prospeccion Sistemati ca de Los Grandes Asentamientos del
Malpais de Zacapu: Claves Para Un Acercamiento a Las Realidades
Sociopoliticas. Genesis, culturas y espacios en Michoacan, ed. Veronique
Darras.

120

Michoacan, R. (1. (RM)

1956(1541)

Migeon, G.
1 990

2003

Moedano, H.
1 94 1

Noguera, E.
193 1

Pollard, H.
1972

1977

1980

1993

1999

2003

Relacion de la ceremonias y rictos y poblacio'n y gobernacio'n de los
indios de la provincial de Michoacan. Reproducci on Facsimilar del Ms Iv
de El Escorial, Madrid. Aguilar Publiscistas, Madrid.

Archéologie en pays tarasque. Structure de l’habitat et ethnopréhistorie
des habitations tarasques de la région de Zacapu Michoacan, Mexique).
PhD. Dissertation, Universidad de Paris I, Pan’s, France.

Abandonos planificados, rituales de vasij as matados 0 de clausura y
ocupaciones posteriores: Los sitios del cerro Baraj as, Guanajuato y de
Milpillas, en el Malpais de Zacapu, Michoacan. Trace, 43.

Estudio Preliminar de la Ceramica de Tzintzuntzan. Temporada HI. 1939-
1940. Revista Mexicana de Estudios Antropologicos.

Exploraciones Arqueologicas; En Las Regi ones de Zamora y Patzcuaro,
Estado de Michoacan. Anales, T VII, 42(12).

Prehispanic Urbanism at T zintzuntzan, a T arascan Settlement in Western
Mexico. Volume 1: Text and Volume I]: Appendices. PhD dissertation,
Columbia University, New York.

An Analysis of Urban Zoning and Planning at Prehispanic Tzintzuntzan.
Proceedings of the American Philosophical Society, 121(1).

Central Places and Cities: A Consideration of the Protohistoric Tarascan
State. American Antiquity, 45(4).

T ariacuri ’s Legacy: The Prehispanic T arascan State. University of
Oklahoma Press, Norman.

Mortuary Patterns of Regional Elites in the Lake Patzcuaro Basin of
Western Mexico. Latin American Antiquity, 10 (3).

Central places and the cities in the Core of the Tarascan State. Lugares
Centrales y Ciudades en el Nucleo del Estado Tarasco. In Urbanization in
Mesoamerica. El Urbanismo en Mesoamerica, edited by W.T. Sanders
and AG. Mastache, pp. 345-390. vol. 1. Pennsylvania University Press,

121

Instituto Nacional de Antropologia e Historia, Mexico, University Park,
PA, Mexico, DP.

2007 Manual Visual de la C eramica Prehispanica: C uenca de Patzcuaro,
Michoacan. Michigan State University, unpublished manuscript available
from author.

Puaux, O.

1989 Les Pratiques F uneraires T arasques (approche ethnohistorique et
archeologique). PhD. Dissertation, Universite de Paris I, Panthenon
Sarbone, France.

Rubin de la Borbolla, DP.

1939 Antropologia Tzintzuntzan-Ihuatzio: Temporadas I y II. Revista
Mexicana de Estudios Anthropologicas 3(2):99-121.

1941 Expolarciones arqueologicas en Michoacan: Tzintzuntzan Temporada III.
Revista Mexicana de Estudios Anthropologicas 5(1): 5-20.

1944 Orfebreria tarasca. C uadernos Americanos 3: 125-38.

Smith, M. _

2007 Form and Meaning in the Earliest Cities: A New Approach to Ancrent

Urban Planning. Journal of Planning History, 6(1).

Sokal, R. and Rohlf, F. 3
1995 Biometery: The Princzples and Practice of Statistics in Biological
Research. W.H. Freeman and Co., New York, NY.

Thomas, D.H. . .
1976 Figuring Anthropology: First Principles of Probability and S tattstzcs.
Holt, Rinehart and Winston, New York, NY.

Versluis, V.A. .
1994 The Iconography of the Protohistoric T arascan State of Western Mexico:
The Material Expression of the State Ideology. Masters thesrs, M1 chi gan

State University.

122

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

lll/llI/l/IllI/l/lll/Illllltill/lll/l/ll

3 1293 02956 8429