You are on page 1of 26

Chronology and Stratigraphy at La Quemada, Zacatecas, Mexico

Author(s): Ben A. Nelson


Source: Journal of Field Archaeology, Vol. 24, No. 1 (Spring, 1997), pp. 85-109
Published by: Boston University
Stable URL: http://www.jstor.org/stable/530563
Accessed: 04/11/2010 00:46

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=boston.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Boston University is collaborating with JSTOR to digitize, preserve and extend access to Journal of Field
Archaeology.

http://www.jstor.org
85

Chronology and Stratigraphyat


La Quemada, Zacatecas, Mexico

Ben A. Nelson
Arizona State University
Tempe, Arizona

La Quemada (A.C.500-900) is a keysitefor understanding theprocessesinvolved in


fluctuations of the northernfrontier of Mesoamerica.Archaeologistshave constructedsce-
narios about itsfounders and political relationshipsin the absenceof adequate informa-
tion about dating. Thispaper summarizes the historicaland theoreticalissuesthat hinge
on the dating of La Quemada, reportsthe stratigraphic contextsof39 radiocarbonsam-
ples obtained in recent excavation, and providesquantitative and qualitative evalu-
ation of the available chronometricdata from thesite and its satellites. Thesedata allow
the testing ofgrowth modelsat severalscales.La Quemada'sgrowth occurredduring the
Epiclassicperiod, whichwas one of decline in the corearea of central Mexico. Thischrono-
logical information clarifiesthe challengeto archaeologistswho might wish to explain the
pattern ofa growing peripheryand declining core,and it also underscoresthe needfor
moredata from satellite sites to understand the organization and developmentof the lo-
cal system.

Introduction tion into the Mesoamerican tradition probably occurred


La Quemada (FIG. 1), located in the Malpaso Valley of approximately A.C.600-900 during the period sometimes
central Mexico, is one of several large settlements compris- called the Epiclassic (TABLE 1). Elaborate writing and calen-
ing the northern frontier of Mesoamerica. It is a ceremo- drical systems such as those found farther south apparently
nial center made up of artificial terraces, platforms, stair- did not develop in this region, but the presence of
cases, causeways, ball courts, sunken patios, temples, and Mesoamerican ideology is expressed in elements of archi-
residential complexes, most of which are enclosed within a tecture, ceramic decorative techniques and iconography,
series of imposing cliffs and massive masonry walls (FIG.2). sacrificialpractices, and the ball game.
The site overlooks the Malpaso Valley (FIG. 3), a locality As one of the principalsettlements in the frontier region,
dotted with villages that were connected to one another, as La Quemada offers an excellent vantage point from which
well as to La Quemada, by an ancient road system. Because to investigate the processes involved in the expansion and
of its visual dominance and physical connections to other contraction of the frontier.Yet, though archaeologists have
settlements, as well as the fact that La Quemada is much offered various interpretations of the site and its relation-
larger than any other settlement in the valley, the other ship to the wider region, they have conducted relatively
sites are assumed to have been political subordinates and little fieldwork at La Quemada. Working from surface data,
are referred to in this paper as "outliers" or "satellites." archaeologists attributed the occupation of the site to such
Prehispanic occupants of the Malpaso Valley may have diverse groups as the Mexica (Batres 1903: 22-24; Clavig-
used the roads in ritual processions, footraces, or military ero 1979 [1787]: 112-117), Tarascans(Batres 1903: 40;
activities that centered on La Quemada. Disarticulated Noguera 1930: 68-71), and Toltecs (Weigand 1977: 23-
human skeletons are a common occurrence at the site, 26, Weigand 1982: 91). A project directed by Pedro
probably as a result of regular sacrificialrites. Armillas (Armillas 1964) in the early 1960s raised intrigu-
Humans were present in the northern Mesoamerican ing questions about environmental change and coloniza-
frontier by ca. 9500 B.c., as suggested by the finding of a tion from the south, but did not produce detailed descrip-
Clovis projectile point in the Huichol region southwest of tions of the excavated remains. The resulting handful of
La Quemada (Weigand 1977). Yet sedentism began no dates from virtually undescribed contexts raise provocative
earlier than A.C.200-400, more than a millennium later questions, but allow a range of interpretations of the
than in the Mesoamerican heartland. Maximal incorpora- dating of the site and, by extension, of the wider processes
86 Chronology and Stratigraphy at La Quemada/Nelson

ACHACO
CANYON

::' -r El Paso
JuArez
? ':-:'. \: :
A CASAS

S"GRANDES
Chihuahua .

"'' " • .:
... :.v

'

*Torre6n
.GULF

. urangg
??0.OF .

. ? IALTA
?? 't / VISTA acatecas ("
.''
..??
SMEXICO
"'" , I "'''

LA UEMADA
I

#'

PACIFIC OCEAN uad-aar pan

~:?'?; ? ? ?~ TULA TEOTIHUACAN*. Ma"ac .


.
.??:
...... Mexico City
I .
1 . ,

Figure 1. Map of northern Mesoamerica and the American Southwest showing sites mentioned in text.

that determined the growth of the settlement. Confusing and sociopoliticalsignificanceto the widerMesoamerican
chronometric data, along with the absence of a ceramic world.
typology that would permit cross-dating, have permitted Recent fieldworkconducted in collaborationwith'the
radicallydivergent views of La Quemada's growth, decline, Gobierno del Estadode Zacatecasand the InstitutoNa-
Journal of Field Archaeology/Vol.
24, 1997 87

0 50 100 METERS
0 300 FEET

HALLOF COLUMNS

BALLCOURT

PYRAMIDOR ALTAR

SUNKEN PATIO
r-
S STAIRCASE

MIDDEN

- TERRACE EDGE
10 METERS
CONTOURINTERVAL

TERRACE 18 "

81

9
S10 17 YAIDO LA

11

Figure 2. Site map of La Quemada showing terraces and middens. Contours after Plano Fotom&trico
de La Ciudadela La Quemada Versi6n Armillas-Weigand. Contour interval 10 m.
88 Chronology and Stratigraphy at La Quemada/Nelson

_--"b
, •-

tsLa Quemada
Adjuntas
,Las Ad..

Los Pilarillos

Presa de
.
Q

I \•l
_111 l,,

* ArchaeologicalSite

SSlope
0 ] 2 3
VILLANUEVA
kilometers

Figure 3. Map of Malpaso Valley showing location of La Quemada and the sattellite villages men-
tioned in text.

cional de Antropologia e Historia (INAH) at La Quemada data. For a variety of technical reasons discussed below,
has resulted in much improved chronological resolution. radiocarbon determinations must sometimes be rejected
This paper presents new radiocarbon determinations in the even though their collection and laboratory analysis are
light of the issues and sampling decisions that led to their faultlessly performed. In documenting decisions about
collection. Because the theoretical and historical implica- dates selected and rejected, the author hopes to provide
tions of the dates are discussed in other papers (Nelson readers with a basis to criticallyevaluate his conclusions.
1990, 1993), emphasis is given here to describing the
contexts from which the dates were obtained and to ac-
counting for why these particular samples were selected Table 1. Chronological periods used in text.
from among the several hundred available for laboratory Period Dates A.C.

analysis.Radiocarbon dating represents a significant invest- Postclassic 900-1519


600-900
ment of scientific resources as well as a set of judgments Epiclassic
150-600
Classic
about the most strategic locations from which to collect
Journal of Field Archaeology/Vol.
24, 1997 89

Interpretive Issues Cuartel area that would permit evaluation of context, still
leaves much latitude for interpretation.
A major issue surrounding La Quemada's dating is its Assigning the occupation to the Classic period tends to
interpretation as an outpost of Mesoamerican empire. This trigger thoughts of Teotihuacan sponsorship; however,
proposition has taken several different forms since archae- archaeologists who work in the Valley of Mexico now see
ologists began to study the site. The most recent, which Teotihuacan's regional influence diminishing significantly
still survives as the main textbook interpretation (Coe ca. A.C.600 (e.g., Garcia 1993: 216-218). Evidence from
1994: 146; Diehl 1983: 48-50, 154, 274; Weaver 1981: elsewhere in Mesoamerica indicates that most of Teotihua-
381-383-but see Weaver 1993: 189-191), is Weigand's can's presence abroad was felt considerably earlier, in the
provocative assertion that La Quemada was a Toltec out- 400s and 500s (Millon 1988: 114-136; Coggins 1979;
post (Weigand 1977: 23-26, Weigand 1982: 91). Wei- Culbert 1988: 135-152) At Matacapan, however, Santley
gand, Harbottle, and Sayres (1977) argue that La Que- (1989, 1994: 248, 261-263) infers extensive interaction
mada was founded during the Early Postclassic (A.C. with Teotihuacan, including a resident Teotihuacan popu-
900-1100) as an intermediate station along a "turquoise lation, from A.C.300-800. Yet at that site the Teotihuacan
trail" that led from Chaco Canyon in the American South- presence is clearly marked by architecture, pottery styles,
west to Tula in central Mexico. This interpretation repre- and figurines; such pronounced stylistic linkages are not
sents a significant shift from earlier readings of the site's present at La Quemada. There is little analogous support
chronology, which had suggested that it dated to the Late for Teotihuacan's involvement in the growth of La Que-
Postclassic period (A.C.1100-1400). mada. Falling in an interval when Teotihuacan's influence
Weigand's model was a reasonable derivation from the in Mesoamerica was waning, the interregnum dating in-
available chronological data, which consisted of a handful deed poses problems for any suggestion that La Quemada
of radiocarbon dates, some of which were analyzed shortly was an outpost of Mesoamerican empire.
after the advent of radiocarbon dating (Crane and Griffin This historical alignment raises a theoretical issue as to
1958: 1100). It was further corroborated by the existence why growth on the Mesoamerican periphery should occur
of a colonnaded hall, a trait commonly connected with in inverse proportion to that of the core. Hers (1989)
Postclassic occupations in central Mexico. Recently archae- explores the thesis that such growth was the product of a
ologists have begun to view the evidence from La Que- colonization by people from central Mexico, and that it
mada as more supportive of a Classic period (i.e., pre-A.c. also set the stage for a migration southward into what
900) alignment. This view is based on 1) the recognition became the Toltec domain. Jimenez (1989) suggests that
that colonnaded halls are a Classic rather than Postclassic the growth was a product of small-scalepeer-polity interac-
trait in Nw Mexico (Holien and Pickering 1978); 2) a new tion; if so, La Quemada and its peers should have arisen
understanding of ceramic cross-ties with the Chalchihuites simultaneously. Nelson (1990, 1993) suggests that the
sequence (Jimenez 1989: 12-20; Nelson 1990: 523-524; periphery might have been in a state of "structuralunder-
Trombold 1990: 316-318); 3) new dates from a satellite development," i.e., subjected to an extractive economic
village (Trombold 1990: 313-316); and 4) re-evaluation relationship that inhibited the accumulation of resources
of the descriptions of the specific contexts from which the by the local population until it was liberated by the disinte-
initial radiocarbon samples were collected (Hers 1989: 42; gration of the core. Recently gathered data on La Que-
Trombold 1990: 311). mada's economy and iconography, however, fail to reveal
The latter two scholars suggest that some of the dated either any resource worthy of long-distance exploitation or
material taken from the Cuartel area of La Quemada the presence of symbols indicating core dominance. Dar-
represents post-occupational reuse of the site as a shrine. ling (1993: 251-252) and Trombold et al. (1993: 268) see
Three of those samples were in fact collected not from little participation by far northern sites such as La Que-
excavation, but from burned material that James B. Griffin mada in the obsidian exchange spheres that encompassed
(personal communication, 1989) observed on the surface much of Mesoamerica. Trombold et al. (1993: 255-256),
of the site. One of the samples, which dated to a.c. 1170 + paralleling Jim~nez (1992: 195), propose a distinction
200 (uncalibrated), consisted of"charcoal selected from a between inner and outer periphery, and suggest that core
concentration of charred wood against a smoke-stained dominance extended only to the inner zone, excluding
wall" (notes archived in the University of Michigan Me- more distant polities such as La Quemada. Jim~nez (1992:
morial-Phoenix Project Laboratory). The fact that those 192-196) and Darling (1993: 252) urge consideration of
early dates carry standard deviations of 200 years, and that the role of local interaction in peripheralpolitical develop-
there is no report on the Armillas excavations in the ment.
90 Chronologyand Stratigraphyat La Quemada/Nelson

Chronologyis crucialin weighing the variousformula- areawithin the site is unterraced,eitherbecauseof exces-
tions that archaeologistspropose.Accumulatingevidence sive slope or simplybecauseit was left open. The majority
favorsabandoningmodels of intersocietalhegemony,but of the architectureis concentratedin the southernend of
the nature and degree of connectivitybetween the core the site.
and the peripheryremainunclear.Ultimatelythe datamay Trashdepositsaccumulatedin variousplacesalong the
suggestthat sitesin the northernperipherydid not partici- basesof terracesand naturalcliffs.Surfacereconnaissance
pate in large-scalespheresof interactionbecausethey did thus far has detected 25 middens;a few more may exist
not exist when those spheres were active. Also, it may beneath crumbledmasonryat the bases of terraces.The
eventuallybe possible to conclude that the "collapse"of middensrange from thin scattersof artifacts2 x 3 m in
the core set up conditions for the propagationof small- surfaceareato 2 m deep and 30 x 40 m in extent. The
scale local hegemonies.Such conclusionsare only distant horizontaldimensionsof the largestmiddens,however,are
possibilitiesuntil the chronologiesof peripheralsites are exaggeratedby artifactdisplacementon the steep slopes.
controlled. In termsof patternsof trashdeposition,it is significant
The dating of La Quemada also has implicationsfor that only two very large deposits-Middens 6 and 11-
understandinglocal community organization. Did La have been noted. Trashwas not scatteredwidelynor often
Quemadaarise as the most successfulamong severalvil- incorporatedinto architecturalfill. Architecturalareasap-
lages competingfor local power,or did it form by coales- parentlywere sweptfrequently,as remarkably little trashis
cence late in the MalpasoValley'soccupationalhistory? evident on the surfaceof the monumentalcore or in the
Was it, as suggested above, used afterabandonmentas a excavatedarchitecturalareas.These patternssuggest that
shrine?These scenarioshavedifferentchronologicalimpli- trashwas carefullycollected,possiblyin keepingwith be-
cations,some of whichcan be testedby evaluatingpattern- liefs about its pollutingnature(Osborn 1979), and some-
ing amongthe radiocarbondeterminations. times depositedin common dumps.
Decisions about where to excavatewere based on the
Site Structure and Sampling Strategy core-flankand patio complex-middendichotomiesas well
Spaceswithin the site of La Quemada(FIG.2) can be as considerationof previouswork. The latter,mostly un-
stratifiedaccordingto centralityand function. The monu- documentedand concentratedin the monumentalcore, is
mentalcore,markedby a massiveenclosingwall and natu- bettercharacterized as clearingthanexcavation,but reveals
ral cliffs, contains large structuressuch as the Hall of a good dealaboutsite structure.A numberof radiocarbon
Columnsand the VotivePyramid.The flanking areasare dateswereobtainedfromthatwork,although,as discussed
made up of terraceswith less grand but still substantial below, their contextsare far from clear.The carefulwork
structures.In both areas,the slopingnaturalsurfaceswere currentlybeing conducted by the Instituto Nacional de
transformedinto architectural spaceby the constructionof Antropologiae Historia and the Gobierno de Zacatecas
numerousterraces. promisesto providedetailsof depositionand datingin the
Depositsupon and aroundthe terracescan be classified core area.
as belongingto patio complexes, whicharearrangementsof The evident gaps in representationwere in patio com-
structuresaroundsunkenpatios,or middens,consistingof plexesin the flankingareasplus middensin both the core
deliberately discardedtrash. The patio complexappears to and flankingareas.Our excavationshaveconcentratedon
be the fundamental residentialunit. As a rule each patio one largeterraceon the westernflankof the site and the
complex occupies a whole terrace, but some terraces are samplingof 10 middens.The majoreffort has been di-
dedicated to other more specialized structures such as ball rected at structureson the banquettesof Terrace18. On
courts. The distinction between core and flanking areamay the basisof its surfaceareaof approximately3200 sq m,
mark an important social dimension, although the exist- Terrace18 is the largestof the flankingterracesand the
ence of a number of internal causeways and grand stair- second largest in the site. The terrace is located along a
cases linking the flanking areas with the monumental core causeway that must have comprised a significant entryway
demonstrates a fundamental integrity among the social into the site; the causeway passes alongside a temple on
entities represented by the architecture. Terrace 18, proceeds along the north side of its main patio,
To date 56 terraces, with an approximately equal num- and then connects with a grand staircasethat ascends to a
ber of patio complexes, have been identified within the site small pyramid or altarin the monumental core. The terrace
area of approximately 0.5 x 1.0 km. The terraces range has a main sunken patio 20 x 35 m in area dominated by a
from approximately 3 x 10 m in area and 1 m in height to small ball court with walls 11 m long and a floor width of
40 x 80 m in area and 7 m in height. A great deal of the 4 m. There is one clear temple structure, and four plat-
Journal of Field Archaeology/Vol.
24, 1997 91

forms that apparently elevated other buildings, now de- some cases dates can be rejected, either because they are
stroyed, rise above the adjacent architecture. The east, inordinately different from others in the same stratum, or
south, and west banquettes are covered with a variety of because they are contradicted by more plausible dates from
structures, which are themselves arranged around five other strata. Information about each date is summarized in
small patios. Nelson (1995) speculates that these smaller Table 2 in order to facilitate discussion of the context and
patios mark fundamental household units, perhaps the chronology of the samples. The samples from Terrace 18
nuclear units of extended or polygynous families. are discussed first, followed by those from the middens.
Within Terrace 18, potentially datable wood was found The discussion is in depositional order of strata within
in a number of forms, not all of which are represented in contexts, and in chronometric order within strata. Most of
the samples submitted for dating. A few samples of fuel- the rejected dates are mentioned at the end of discussion of
wood were left in hearths after their last uses, though in each context.
general the hearths had been thoroughly cleaned. Other Three points of technical clarification are important to
charred wood came from seeds, roof support posts, roof mention. First, the dates given in discussion are uncali-
beams, and scattered charcoal of undetermined origin. brated, and their calibrated ranges as computed by Stuiver
Some of the charcoal lay in direct association with human and Becker's (1986) CALIB program may be found in the
skulls and long bones, and may represent skull racks (Hers table. Uncalibrated dates are used in the following discus-
1989: 89-93; E. A. Kelley 1978: 114-117; Nelson, Dar- sion because they are required as input for Kintigh's
ling, and Kice 1992: 306). (1994) 14C program, which assumes a normal distribution
The sampled middens represent both the monumental in order to construct the probability plots shown below.
core and the flanking areas. Although all lie outside the Second, the dates are also corrected for isotopic fractiona-
monumental core per se, several are located along the base tion where that information is available.Unfortunately, the
of the natural cliff that serves as the boundary between the laboratory measurements necessary to correct for isotopic
monumental core and the flanking areas. Such locations fractionation were not taken for all samples. While it is a
served as dumps for materials discarded from above, i.e., fairly common practice to apply corrections to one sample
from patio complexes within the monumental core. They based on measurements taken from another in the same
include two of the largest middens, both of which contain context, such adjustments are not appropriate in this case
relatively large amounts of exotic and elaborate artifacts, because of the considerable variety of woods and wide
and these locations make no sense as dumping locations range of resulting correction factors encountered among
for the inhabitants of nearby flanking terraces, who would the measured samples. Third, the term "source" refers to
have had to carry trash uphill in order to dump it there. the botanical and behavioral source of the samples and by
Four of the 10 sampled middens represent the monumen- implication to the probable age of the material at the time
tal core, and the remainder belong to the flanks. of use. For example, the source of one date might be a
The major source of charcoal recovered in the middens construction beam, while that of another might be a
appears to be fuelwood discarded during the cleaning of mesquite bean.
hearths. Invariablythe middens are located in places where
people would have stepped to the edge of a precipice to Terrace 18
discard trash. The gradual accumulation of trash in some Building episodes at Terrace 18 can be distinguished at
middens produced charcoal-laden strata that appear to three levels of detail: 1) substructure expansion; 2) "ar-
represent periods of relatively intensive and regular dump- rangements" of structures constucted either immediately
ing. The very latest material in the middens with well- upon the substructure or following the razing of earlier
developed strata, however, appears to consist not of trash, structures but without change to the substructure; and 3)
but of building materials issuing from crumbling build- maintenance of existing structures and surfaces. Figure 4 is
ings. Roof beams and other building members may be a schematic representation of the architecturalremnants of
included in some of this late material. these episodes as found in excavation. At the most general
The Dates in Their Contexts level, the terrace substructure was built in two major
stages, the first in which the terrace reached about 80%of
In this section the individual dates are discussed in order its current size, and the second when it was expanded
to evaluate their representativenessand reliability.For each vertically and horizontally to take on its current form. The
excavation context, the stratigraphy is discussed and the structures that had been on top of the terrace were com-
dates are examined to determine whether they make sense pletely dismantled during the expansion, leaving only
in view of independent chronological information. In traces where the main patio had been and indications that
92 Chronologyand Stratigraphyat La Quemada/Nelson

Table 2. Radiocarbon dates from SUNY Buffalo excavations 1988-1993.


Sample Excavation '3C Uncalibrated
no. Lab no. area Material Stratigraphic position 14C age 13C/2C ratio adjusted age date (A.C.)
1 B-44790 Terrace 18 Charred beam On temple floor 1260?50 - - 690?50
2 B-44791 Terrace 18 Charred beam On temple floor 1320?60 - - 630?60
3 B-44792 Terrace 18 Charred post in north Main roof support post of 1210?50 - - 740?50
post hole temple, installed as part of
late expansion
4 B-44793 Terrace 18 Charred post in south Main roof support post of 1350?50 - - 600?50
post hole temple
5 B-44794 Terrace 18 Single piece of Fill that accumulated in Patio 1640?60 - - 310?60
charcoal-portion of B after occupation
beam or skull rack
or fuelwood?
6 B-44795 Midden 6 Dispersed pieces of Lowest of three occupational 1270?60 680?60
charcoal strata
7 B-44796 Midden 6 Single piece of Lowest of three occupational 1460?80 490?80
charcoal-firewood? strata, near bedrock
8 B-44797 Midden 11 Charcoal concentration Post-occupationally deposited 1320?60 630?60
material from core area of
site above midden
9 B-44798 Midden 11 Charcoal concentration Post-occupationally deposited 1340?60 610?60
material from core area of
site above midden
10 B-44799 Midden 11 Charcoal concentration Post-occupationally deposited 1660?120 - - 290?120
material from core area of
site above midden
11 B-44800 Midden 11 Charcoal concentration Post-occupationally deposited 1250?60 - - 700?60
material from core area of
site above midden
12 B-62001* Terrace 18 Charcoal concentration West Banquette, early Patio 1320?60 - - 630?60
B, in construction fill
deposited during renovation
13 B-62002 Terrace 18 Charcoal concentration West Banquette, later Patio 1340?60 -24.1 0/00 1350?60 600?60
B, associated with human
crania and long bones
14 B-62003 Terrace 18 Dispersed pieces of West Banquette, accumulated 1220?60 -24.7 0/00 1230?60 720?60
charcoal fill of later Patio B
15 B-62004 Terrace 18 Dispersed pieces of West Banquette between 1250?90 -24.2 0/00 1260?90 690?90
charcoal Floors 2.2 and 2.3
16 B-62005 Terrace 18 Charcoal concentra- North Banquette, late 1120?60 -24.0 0/00 1130?60 820?60
tion-firewood extramural hearth
17 B-62007 Terrace 18 Charcoal concentration On or just above floor of ball 1280?60 -24.8 0/00 1290?60 660a60
court
18 B-62009 Terrace 18 Charcoal concentration West Banquette, associated 1450?80 -25.0 0/00 1450?80 500?80
with human crania and
long bones that fell along
banquette walkway outside
temple
19 B-62010 Terrace 18 Charcoal concentration Main patio, in floor- 1280?120 -25.6 0/00 1280?120 670?120
preparation layer beneath
earliest of 11 floors
20 B-62011 Terrace 18 Dispersed pieces of West Banquette, on or just 1730?100 -25.3 0/00 1720?100 230?100
charcoal above latest floor of
passageway

the main patio floor had been replastered a number of stayedin its previouslocation;its northernwallwas, how-
times before the expansion took place. After the expansion ever, shifted northwardto increasethe interiorspace by
the record becomes more detailed. A new complex of severalsquaremeters.This second arrangementwas fol-
buildings was constructed to fit the expanded terrace;both lowed by a number of maintenanceepisodes involving
the buildings and patio were periodically renovated in a replasteringof floorsand changesin the locationsof a few
series of maintenance episodes. At some point that entire individualwalls.
complex of structures, except for the temple, was again Some examplesof rearrangements and maintenanceepi-
destroyed, and rebuilt in a new arrangement. The temple sodes sufficeto give a senseof the evidenceavailableabout
Journal of Field Archaeology/Vol.
24, 1997 93

Table 2. (cont.)
Sample Excavation 13C Uncalibrated
no. Lab no. area Material Stratigraphic position 14C age '3C/'2C ratio adjusted age date (A.C.)
21 B-62012 Terrace 18 Single piece of East Banquette, lower 1100?70 -24.5 0/00 1110?70 840?70
charcoal-portion of platform, associated with
beam or skull rack human cranial and long
bone fragments and
structural remains, above
walkway floor
22 B-62013 Terrace 18 Charcoal concentration East Banquette, walkway of 1390?60 -23.5 0/00 1420?60 530?60
early interior patio
23 B-62016 Midden 7 Single piece of charcoal Uppermost occuaptional 1490?90 -24.9 0/00 1490?90 460?90
stratum, associated with a
Michilia sherd
24 B-62017 Midden 7 Charcoal concentration Earliest occupational stratum, 1310?60 -25.1 0/00 1300?60 650?60
extends beneath western
talud of Terrace 18
25 B-62018 Midden 7 Charcoal concentration Earliest occupational stratum, 1340?50 -25.1 0/00 1340?60 610?60
extends beneath western
talud of Terrace 18

26 B-62019 Midden 11 Charcoal concentration Latest of three occupational 1280?80 -24.8 0/00 1290?90 660?90
strata
27 B-62020 Midden 11 Charcoal concentration Latest of three occupational 1060?80 -19.4 0/00 1150?80 800?80
strata
28 B-62021 Midden 11 Charcoal concentration Middle of three occupational 1340?80 -22.2 0/00 1340?80 610?80
strata
29 B-62023 Midden 12 Dispersed pieces of Single undifferentiated 134?1 - -
1816+1
charcoal occupational stratum Modern - - Modern
30 B-62025 Midden 15 Dispersed pieces of Single undifferentiated 1180?70 -23.7 0/00 1210?70 740?70
charcoal occupational stratum
31 B-66554 Midden 11 Charcoal concentration Middle of three occupational 1290?80 -23.7 0/00 1310?80 640?80
strata
32 B-66555 Midden 11 Charcoal concentration Earliest of three occupational 1260?90 -21.2 0/00 1320?90 630?90
strata
33 B-66556 Midden 11 Charcoal concentration Earliest of three occupational 1270?100 -24.7 0/00 1280?100 670?100
strata
34 B-66557* Midden 11 Single piece of charcoal Earliest of three occupational 1340?60 -24.7 0/00 1340?60 610?60
strata
35 B-66558 Midden 11 Charcoal concentration Middle of three occupational 1330?80 -20.4 0/00 1410?80 540?80
strata

36 B-66559* Midden 11 Charcoal concentration Middle of three occupational 1550?60 -23.1 0/00 1550?60 400?60
strata
37 B-77238 Terrace 18 Single piece of charcoal Construction fill of early 1580?70 -26.3 0/00 1560?70 390?70
terrace beneath East
Banquette
38 B-77240 Terrace 18 Charcoal concentration Hearth resting on bedrock 1430?80 -26.0 0/00 1410?80 540?80
beneath terrace fill
39 B-77239* Terrace 18 Charred corn cob Hearth resting on bedrock 1110?50 -13.5 0/00 1300?50 650?50
beneath terrace fill
* Date obtained
by AMS (accelerator mass spectrometry) method.

the construction sequence. On the East Banquette, the were in evidence. Stratigraphyshowed that the patio floor
occupants arranged a number of structures around a single had been plastered four times before the expansion of the
small sunken patio after the expansion of the terrace sub- terrace, and thereafterwas replastered seven times more.
structure. Later, all the structures were razed and rebuilt, The building episodes recognized in excavation could
and this arrangement was changed so that there were be grouped in a number of ways to define strata. Ulti-
instead two small sunken patios surrounded by more struc- mately there is no utility in defining strata that are finer
tures. As another example, excavations in the sw corner of than the dating method can distinguish, and the limita-
the main patio revealed a number of maintenance episodes. tions of radiocarbon dating are such that often it cannot
There, both original and expanded versions of the terrace even distinguish adjacent phases of occupation (Dean
94 Chronologyand Stratigraphyat La Quemada/Nelson

1991). On the other hand, in the present context there is 650 ? 50. Both dates are taken from charred material
little independent evidence to suggest how much time found in a hearth built on bedrock, the more circular of
might have elapsed between episodes; therefore it is prefer- the two rock features in Figure 5. Both samples should
able to err on the side of stratigraphic precision. The stem from the same use-episode and should fix the date of
building sequence also has the property of becoming construction of the terrace. Date no. 38 is from fuelwood,
clearer through time, so that the later episodes are more and no. 39 is from a charred corn cob. Because the hearth
easily recognized and more robustly represented with ra- was intact and contained fuel remnants, this materialprob-
diocarbon samples than the earlier ones. ably was deposited immediately before the area was cov-
The major stratigraphic contexts within Terrace 18 are ered by construction of the terrace;otherwise, the charred
thus 1) the ground surface underlying the terrace; 2) the material would have disintegrated and been washed away.
fill deposited in the early episode of terrace construction, It is conceivable that the hearth was actually used by the
which became encapsulated in the later phase; 3) architec- construction crew that built the terrace.
ture of the early terrace; 4) the fill deposited during the The second stratum, fill belonging to the early episode
later expansion of the terrace substructure; 5) the first of terraceconstruction, is also represented by one radiocar-
arrangement of structures after the expansion; 6) the sec- bon sample, no. 37 (a.c. 390 ? 70). This date, which is
ond arrangement of structures after the expansion; 7) corrected, comes from a single piece of charcoalembedded
materials in use at the time of abandonment; and 8) in the construction fill of the early terrace underlying the
post-occupational deposits. Within some of these strataare East Banquette. This date is unreasonablyearly in the light
more subtle substrata, such as the replastering of floors. of date no. 38 and other evidence discussed below, unless
Also within each stratum may be materials that represent there is a small version of the early terracethat has not been
rather different phenomena from a chronometric stand- detected in excavation. Such a small early terrace was
point, e.g., support posts versus seeds, the former poten- anticipated, and evidence of it was sought, but its existence
tially representing an "old wood problem" (Schiffer could not be confirmed. Therefore, this date is rejected.
1986), the latter more likely to reflect the actual time of The third stratum, the architectural components be-
occupation. longing to the early terrace, is represented by two dates.
There are 19 dates from Terrace 18, including the ones Date no. 22, a.c. 530 + 60, comes from a concentration of
from the closely associated Midden 7. The earliest stratum, charcoal that probably derived from a single piece of wood.
the natural ground surface underlying the terrace, is repre- It was resting on the walkwayfloor of a small interior patio
sented by two dates, no. 38, a.c. 540 + 80, and no. 39, a.c. on the East Banquette, and perhaps was part of a beam or

Figure 4. Schematic map of strata in Terrace 18. View is to north. Hatching indicates sunken areas.

Platform

PPlatforms
De
Patio B Temple MainPatio(A)

Second Arrangement

First Arrangement

EarlyTerrace
LaterTerrace
Journal of Field Archaeology/Vol.
24, 1997 95

post that was razed during reconstruction. Date no. 19, all of these buildings also were ultimately razed. All of the
a.c. 670 ? 120, comes from the sw corner of the main Stratum 5 dates come from the West Banquette. Date no.
patio. The sample comes from bits of charcoal found in the 4, a.c. 600 ? 50, comes from a main roof support post
underlayment, or floor preparation layer, of the earliest belonging to the temple. The temple was apparently the
floor of the original patio. Although the source of the only building not completely demolished during the final
wood comprising this sample is unknown, the charcoal was rearrangement of structures. The location of the southern
concentrated rather than dispersed, and the relatively high roof support post, from which date no. 4 was taken, did
negative '3C/12C ratio suggests that the wood belongs to not change even though the building was partiallydisman-
a single species rather than a mix. The precision of the tled and enlarged. Conceivably, therefore, the southern
sample is among the lowest in the suite of dates, however, post belongs to the pre-rearrangement stage of construc-
by virtue of its very wide sigma range. tion. The dated material is charred wood from the stand-
Dates representing the early architecture are rare be- ing post. It was not possible to determine whether the
cause, at least in the areas where deep excavations were outer rings were represented, and the sample material
conducted, the earliest buildings were almost totally dis- could represent a mixture of inner and outer rings. The
mantled. This situation is understandable as the builders date has not been '3C/12C-corrected. Date no. 12, a.c.
were placing adobe buildings on top of a rock substruc- 630 ? 60, was obtained from the fill placed and sealed in
ture, so that the outer layer of rock had to be very carefully Patio B, the small sunken patio to the west of the temple,
selected, and was probably reused. Also, deep excavations when it was reconstructed during the rearrangement of
were conducted only in the patio. structures. This date, also not corrected, comes from a
The fourth stratum, fill of the later terrace substructure, concentration of charcoal rather than dispersed pieces.
is not represented by any dates. The fifth stratum, the Date no. 15, a.c. 690 + 90, derives from scattered charcoal
arrangement of structures created immediately after the between the two earliest floors of a room to the west of the
expansion, is more clearly represented even though almost temple; it has been corrected. The source of the charcoal is

Figure 5. Bedrock hearths beneath Terrace 18, each about 50 cm in diameter.


96 Chronologyand Stratigraphyat La Quemada/Nelson

unclear, but its deposition should relate to a fairly narrow insidethe temple,in the mainpatio,in walkways,andin at
period. leastone smallpatio, PatioB. Almostinvariablythe bones
The sixth stratum, which includes structural materials are in some mannerof disarray,restingin a patternsug-
relating to the second and final rearrangementof buildings gesting that they were suspended from walls, roofs, or
after the terrace expansion, is represented by three dates. racks, and the outdoor examplesare all associatedwith
Dates no. 2 and no. 1, a.c. 630 ? 60 and a.c. 690 + 50, are charcoal.The charcoalmay representeitherskullracksor
both uncorrected and are stratigraphically equivalent. firewood that was ritually associatedwith the skeletal
They derive from charred timbers that were resting on the material.
floor of the temple, and probably represent roof beams. The distributionof datesfromsuchspecimens,however,
These beams should date the last major repairs to the suggeststhatthe wood associatedwith the skeletalmaterial
temple structure,1 although they may have been borrowed was kept aroundthe patio complexfor a verylong time or
from an earlier structure. was very old wood when it was cut. One such specimenis
Date no. 3, a.c. 740 + 50, uncorrected, comes from the no. 18, a.c. 500 ? 80 (corrected),which comes from a
northern roof post of the temple. It will be recalled that charcoalconcentrationfound among humanbones along
this post was relocated and possibly replaced when the the eastern wall of the temple on the walkwayof the
temple was expanded northward. The southern post, dis- westernside of the mainpatio.Here the positioningof the
cussed above in relation to the fourth stratum, dates to a.c. skullsand long bones suggestedthat they may have been
600 ? 50. These uncorrected means of a.c. 600 and 740 hung on the exteriorwall of the temple. Date no. 13, a.c.
bracket the construction activity that took place at the 600 ? 60 (corrected),comes from a similardeposit lying
temple, although such an interpretation exceeds the allow- inside Patio B. The skeletalmaterialformeda rough row
able inferential limits. In any case, the dates are extremely along the westernedge of the patio and was restingpar-
important because their behavioral context is clear and the tiallyin contactwith the floor and patiowall and partially
error ranges are as small as the technique allows. in accumulatedfill. The impressionwas that the skeletal
Stratum 7, material in use at the time of abandonment, materialhad been suspendedabove the patio walkwayor
is represented by seven dates. This materialis distinguished floor and had fallen along with whateverstructuresup-
from that of Stratum 6 because of the possibility that ported it. A laterdeterminationmore consistentwith the
significant time may have elapsed between construction of probabletime of abandonmentis date no. 21, a.c. 840 ?
the last buildings and the actual end of the occupation. 70 (corrected),which comes from a single piece of char-
The kinds of materials that belong to this category are not coal associatedwith human cranialand long bone frag-
main posts and roof beams, but smaller elements that may ments. The skeletal materialand charcoalwere resting
have been consumed more rapidly. An example of ideal beneaththe collapsedmasonrywallof the upperplatform,
material is fuelwood from an indoor hearth, but unfortu- which apparentlywas the substructurefor a buildingthat
nately all of the hearths had been thoroughly cleaned. overlookedTerrace18. The originalpositioning of this
Another such category of material, which was availablebut materialwas unclearother than that it appearedto have
unfortunately was not given enough consideration when been outside the building,probablynear the edge of the
the samples were being selected, is the accidentally dis- platform.
carded seeds of annual plants such as corn and beans. Also belongingto Stratum7 aretwo othersamplesfrom
From among the samples that were submitted for radio- contextsthat immediatelyprecedethe abandonment.The
carbon analysis, the best representation of materials in use firstwas taken from a samplethat lay in contact with an
at the time of abandonment apparently comes from char- outdoor floor; unfortunatelythere were no samples in
coal associated with deposits of human skeletal material.As contact with interiorfloors except for the beams in the
noted by Pijoan and Mansilla (1990: 467), displays of temple mentioned above. Sampleno. 17, a.c. 660 ? 60
skulls and long bones are a pervasive feature of La Que- (corrected), was obtained from a charcoal concentration
mada in general and Terrace 18 in particular, occurring resting on the floor of the ball court. The source of this
material is unclear; it could represent construction materi-
1. The Huichol of San Andr's Coahmiata re-roof their main temple als that came adrift in the decomposition of the structures,
every five years in a formally scheduled community activity with religious some small wooden apparatus associated with the ball
significance. In terms of structural renovations, the main purpose is to court, or even a post-occupational fire within the ball court
replace the thick straw thatch of the roof; as the thatch is entirely walls. The second sample potentially representing the
removed at these intervals weak beams are no doubt replaced as well.
Therefore, if similar practices occurred at La Quemada, the dated beams peri-abandonment phase is no. 14, a.c. 720 + 60 (cor-
may have been cut at almost any point in the life of the temple structure. rected), which was obtained from dispersed pieces of
Journal of Field Archaeology/Vol.
24, 1997 97

charcoal in the same layer of accumulated fill in the same the rather steep slope. The dark layers are relatively char-
patio. coal-laden and artifact density is high throughout the
The final date from Stratum 7 and one of the two latest deposit. These soil zones were not, however, clear during
determinations obtained from the terrace is no. 16, a.c. excavation, and so the excavation was done in very small
820 ? 60 (corrected). The dated material, unquestionably provenience units, mostly 0.5 x 2.0 swaths, 0.15 m deep
firewood, was resting in an ephemeral extramural hearth (FIG. 6B), which were later matched with the stratigraphy
about 4.5 m east of the platform overlooking the temple. visible in profile. Radiocarbon samples were also point-
This hearth is believed to have been constructed after that provenienced within the excavation subunits. One of the
portion of the terrace, at least, fell into disuse; an adjacent criteria for selecting radiocarbon samples for submission
extramuralsurface had ceased to be maintained by the time was that they be associated clearly with one stratum or the
the hearth was constructed and had begun to accumulate other.
water-borne sediments. This sample is considered the best Midden 11 can be divided into four strata:early,middle,
indication of the moment of abandonment. late, and post-occupational. The lowest 1 m of deposits,
Stratum 8 is not represented by any dates, although it approximately, are occupational strata that could be di-
could reasonably be argued that date no. 16 was deposited vided in a number of ways; the early-middle-late designa-
post-occupationally. There was one stratigraphic context tions are not intended to imply separateperiods of site use.
that clearly represented post-occupational reuse of the It does appear,however, that these zones were deposited in
terrace; it consisted of a lightly packed surface, a cache of a highly orderly fashion. The higher and later post-occupa-
manos, and a hearth or small windbreak built in the ruins tional deposits, on the other hand, are more difficult to
of the temple. Unfortunately, no charcoalwas encountered interpret in terms of deposition. They are considered post-
among those intriguing features. occupational because the soil is substantially lighter, the
Two other dates from Terrace 18 are rejected because artifact density lower, and the quantity of charcoal smaller
they are clearly contradicted by other dates and associated than in the levels below. Also, they contain chunks of
stratigraphic information. Date no. 20, a.c. 230 ? 100 material that appearto be adobe, suggesting the decompo-
(corrected), comes from dispersed pieces of charcoal on sition of buildings immediately above. The implication of
the floor of an exterior passagewayon the West Banquette. this stratigraphicinterpretation is that while dates from the
This date is unacceptably early and is contradicted by early, middle, and late strata can be considered potentially
nearby readings that are stratigraphically earlier and yet reliable chronometric indications, those from the post-oc-
post-date this sample by centuries. For example, no. 20 cupational zone could represent almost any phase of occu-
contrasts radically with no. 15, which dates to a.c. 690 + pation and are of interest only in terms of dating the site as
90 and yet was sealed between two floors that underlie the a whole.
one from which no. 20 was obtained. Date no. 5, a.c. The earliest date from the early stratum is no. 34, a.c.
310 ? 60 (uncorrected), comes from a single piece of 610 ? 60 (corrected), which comes from a single piece of
charcoal in the aeolian fill of Patio B. The fill accumulated charcoal. It is suspected that the source of this specimen, as
in the patio after the ultimate renovation and presumably well as all the rest from the occupational strata, is fuel-
at the end of the occupation. This is another unacceptably wood. This midden location is a highly logical one for daily
early determination and like no. 20 discussed above, it dumping, and the decision about where to place excava-
accords poorly with material that was sealed beneath it in tion units was made by going to the edge of the cliff above
the same architectural unit-see date no. 12. Date no. 5 and tossing stones down to see where they would tend to
must be discarded; the best explanation for its earliness is fall. The high frequency and scattered distribution of char-
that it represents the inner portion of a very old beam. coal in the midden is best explained by the dumping of
domestic debris on a regular basis. The other dates from
Midden 11
the early stratum are no. 32, a.c. 630 + 90 (corrected), and
The most intensively dated midden at La Quemada is no. 33, a.c. 670 + 100 (corrected), both of which were
Midden 11, one of the dumping locations for the monu- obtained from charcoal concentrations that may or may
mental core located at the base of the natural cliff that not have derived from single pieces of wood.
separates the core from the western flank. A total of 13 The middle stratum is also represented by three dates.
dates were obtained from this context; two are rejected. Date no. 35, a.c. 540 ? 80 (corrected), comes from a
The midden is undisturbed and was formed by gradual charcoal concentration which, based on its low negative
accretion; its profile (FIG.6A) manifests a number of rela- '3C/12C ratio, probably represents a single species of
tively clear soil zones, alternating light and dark, paralleling wood and not a mix. The date represents a minor reversal,
98 Chronology and Stratigraphy at La Quemada/Nelson

Unit 111
A

Z• Humus
Unit 110
1.0 m below , _ _ Brownclayey loam

datum JJJJ
Brownand yellow mottledloam
C---

-
Yellowishbrownclay loam mottled
" withcharcoaland gravel

Excavated1990 Excavated
1990- clayeyloam
Gray
.. .-_- _00-_-_-
Excavated 1992 - lo loam
Gray

,
Yellowishorange loam

Rock

0 .2 .50 1.0 MT.

B 1 0-4- 1
110-5-4
-110-5-3110-5-1
110-6-3 110-6-4110-6-3110-6-2
-7-0-7-4 - 0-7-7-5
-7-4 110-7-3 110-12-8
111-5-141111-4
l-5-1 - 111-4-2 110-8-5 110-8-4
110-23-12 110-8-3 110-13-9
4-- 111-5-I
111-6-14 111-6-4
111.-5-12111-6-3
111-6-22 1 2-5 I
1-1-77 110-10-6
111-7-5 111-7-4 111-7-3
11-8-14111-12-8 110-6-11

Excavated1990 110-17-1111
111-4-1 11 -18 -11
11,-13-10
1 -13
. v1-14
E d 9. 111- 2 111-1-12
111-1- . 110-19-11

111-3-14 111-3-13 111-3-12, 111-3-11 110-21-12


111-4-14 111-4-13 111-4-12 111-4-11 110-22-12

111-5-14 111-5-13 111-5-12 111-5-11 110-23-12

111-7-14 111-7-13
111-8-14

Figure 6. Profiles of Midden 11. A) Soil zones of varying charcoal content. B) Schematic of prove-
nience units.

and its earliness relative to the rest of those from the determination for date no. 26 is a.c. 660 ? 90 (corrected),
occupational strata is probably attributable to sampling and that of no. 27 is a.c. 800 ? 80 (corrected).
error. Date no. 28, a.c. 610 ? 80 (corrected) comes from a The post-occupational stratum, as noted above, was not
charcoal concentration, as does no. 31, a.c. 640 ? 80 recognized as such until after excavation was done and a
(corrected). number of radiocarbon samples had been submitted;
The late occupational stratum of Midden 11 is deline- hence it is overrepresented. While dates from this stratum
ated by two dates, both from charcoal concentrations. The are not readily interpretable from the standpoint of mid-
Journal of Field Archaeology/Vol.
24, 1997 99

den deposition, they are useful for considering the overall associated with and partiallycovered by Terrace 18. These
span of site occupation as well as that of the architectural dates are particularly important because they represent
units above from which they probably originated. All three materials discarded from the same residential and ceremo-
dates are from charcoal concentrations rather than dis- nial contexts that were intensively excavated on Terrace 18.
persed materials, and all are uncorrected. They are no. 9, Like two of the other middens discussed above, this mid-
a.c. 610 ? 60, no. 8, a.c. 630 ? 60, and no. 11, a.c. 700 + den can be divided into early,middle, late, and post-occu-
60. The source of these dates is believed to be construction pational strata. Two of the dates come from the early
materials from decomposing buildings on the terrace stratum and one from the late stratum. Those from the
above the midden in the monumental core, although it is early stratum belong to a deposit that extends beneath the
possible that they represent redeposited trash and hence western exterior wall of Terrace 18, apparently having
possibly fuelwood. accumulated there before the terrace expanded to its final
Two dates from Midden 11 are rejected for lack of fit size. These samples should therefore be relativelyearly and
with the stratigraphicordering of the other dates. Date no. comparable to Stratum 2 or 3 from within the terrace itself
10, a.c. 290 ? 120 (uncorrected), comes from the post- (see discussion above). The dates, both taken from char-
occupational stratum. This sample consisted of a very small coal concentrations, are no. 25, a.c. 610 ? 60 (corrected)
charcoal concentration that required extended counting and no. 24, a.c. 650 ? 60 (corrected). The date from the
time in the laboratory. Its standard deviation is the highest late stratum of Midden 7 is no. 23, a.c. 460 ? 90 (cor-
among those of recently submitted samples. If the deter- rected), which rather clearly represents a reversal,not only
mination is accurate it most likely represents the inner because it is contradicted by the aforementioned dates
rings of a beam that was quite old when cut. Date no. 36, from the early stratum but because the date was taken from
a.c. 400 ? 60 (corrected), comes from the middle stratum a piece of charcoal lying in association with a Michilia
but cannot be accurate since there are five dates in the same sherd from Chalchihuites, which should date ca. a.c. 750-
stratum and the one below that are two to three centuries 900 (Kelley 1985). Date no. 23 is therefore rejected.
later. This date is also likely to be the product of old wood From Midden 12 came a radiocarbon sample that appar-
or a determination error. ently represents a recent brush fire. The date is no. 29, a.c.
1816 ? 1 (corrected). This midden had a single, undiffer-
OtherMiddens entiated occupational stratum, and there appears to be
From the remaining nine middens, dates are available little hope of obtaining a better determination from it.
from Middens 6, 7, 12, and 15. One of these dates is The final date is taken from dispersed pieces of charcoal
rejected. These middens were less intensively dated and in Midden 15. The date is no. 30, a.c. 740 ? 70 (cor-
also were not dug in quite the same highly controlled rected), and represents the single occupational stratum in
fashion as Midden 11. For example, in Midden 6, from that midden, which is associated with one of the terraces
which two dates were obtained, the arbitrary levels were below Terrace 18. If it is imagined that the string of
1 x 2 m in area and 0.15 m deep instead of the finer- terraces of which Terrace 18 is a part grew gradually
grained 0.5 x 2 x 0.15 m provenience units used in Mid- downslope then the mean date of occupation for the
den 11. Most of the other middens were dug in sloping 15 terrace from which this date was obtained should be some-
cm levels in an attempt to capture the sloping stratigraphy, what later than that of Terrace 18. Insofar as such matters
an effort that was partiallysuccessful. can be judged from a single date, this sample seems to
Two dates were obtained from Midden 6. Date no. 7, agree with that proposition.
a.c. 490 ? 90 (uncorrected), comes from a single piece of
charcoal from the earliest of three occupational strata in Quantitative and Qualitative Evaluation
the midden. While this date is significantly earlier than Archaeologists tend to depict radiocarbon dates either as
most others that have been obtained, and is particularly means or as ranges; neither representation is satisfactory.
surprising in view of the peripheralposition of the midden, The radiocarbon mean is important because it points to
there is little basis for rejecting it except that a considerably the greatest probability of actual dating, yet misleading
later date comes from the same stratum. That date is no. 6, because there is a significant chance that the true date is
a.c. 680 + 80 (uncorrected). The context of these two some rather different value. The most widely adopted
samples appeared to be well sealed, but the material for the solution to this problem is to consider the date as a range
latter date was collected from dispersed bits of charcoal bracketed by one or two standard deviations above and
rather than from a single piece. below the mean. While useful in calling attention to the
Three dates were obtained from Midden 7, which is breadth of possibilities, this solution is undesirable because
100 Chronologyand Stratigraphyat La Quemada/Nelson

it implies that the occurrence of any date along the contin- summing the probabilities for each interval, allows the
uum is equally probable. In adopting this solution, the creation of an aggregated probability distribution that
archaeologist loses information and accepts diminished gives appropriate weight to each date and its respective
analyticalpower. precision. A date with a "large sigma" will be spread over a
Rather than treating radiocarbon dates as means or large number of intervals, with proportionally less weight
ranges, it is preferable to think of them as probability given to each interval than would be contributed by a
distributions. Before calibration, 14C determinations rep- more precise date. Also, if one wishes to make the assump-
resent samples drawn from normally distributed popula- tions that the dated materials are a random sample of
tions. One of the properties of the standard normal curve wooden objects discarded or abandoned in the dated con-
is that the area beneath it, which is 1, is equal to the text, and that such materials were deposited at a regular
probability that the parameterof interest occurs within the rate, then it is possible to quantitativelyevaluate the overall
range encompassed by the curve's range. The area under distribution as if it represented an actual population of
the curve diminishes to zero with increasing distance from dates. By assuming, for example, that the outer 12.5% on
the mean. The probability that the parameteroccurs within each end of the distribution represents sampling error, one
a specified interval is given by the area under the curve may ask what interval contains 75% of the "probabilistic
corresponding to that interval. dates" in the distribution, and obtain a reasonablyaccurate
Kintigh (1994) resourcefully takes advantage of these estimate of the span of occupation.
properties in creating an algorithm that calculates the A disadvantage to this approach is that it cannot be
probabilities that a true date belongs to a series of intervals, applied to calibrated dates because they are not normally
the width of which is set by the analyst. Instead of charac- distributed. Also, calibration recognizes the problem of
terizing the distribution in terms of the standarddeviation, multiple intercepts between the curves representing radio-
the archaeologist is able to suggest intervals that are easier carbon and sidereal ages; the probability distribution rep-
to cognize, such as 10-, 25-, 50-, or 100-year intervals. resenting calibrated ages would have to incorporate the
Thus, for a single date, e.g., no. 4 from the south post of resulting multimodality. Although there are no doubt
the temple on Terrace 18 which has a radiocarbon mean of curve-fitting solutions to these problems, the present study
a.c. 600 and standard deviation of 50 years, one can deals with them by creating probability distributions from
inquire as to the probabilities that the true date lies within uncalibrated dates and then judgmentally considering the
the intervals a.c. 550-559, 560-569, 570-579, etc. The effects of calibration. Given the amount of noise in the
sum of the answers to these questions is a probability data, this line of reasoning probably produces adequately
distribution that represents both the range of possible accurate answers to the chronological questions.
dates and the greatest likely age of the post. The data supplied by the probability distributions ulti-
It might appear that this method would smooth the mately can be used to test growth models at four scales-
picture of occupation, creating an illusion of continuity. In macroregional, regional, community, and intrasite. At pre-
fact, it has less of that tendency than the standard ap- sent, definitive conclusions can be reached only at the
proach, which forces the archaeologist to assume that each macroregional scale, yet at the finer scales there is im-
date represents either a moment of occupation or, more proved resolution and the impetus for further inquiry. The
plausibly but still unacceptably, a span during which occu- specific questions that can be addressed are 1) whether the
pation was continuous and equally intense. Kintigh's prob- growth of La Quemada as a whole coincides with that of
ability-distribution approach strikes a balance, creating a the Mesoamerican core; 2) whether La Quemada was
picture of greatest occupational intensity around the most coeval with other centers that constitute the archaeolo-
likely date, but acknowledging the outwardly diminishing gist's "Mesoamerican frontier;" 3) whether La Quemada
probability of continuity. Using this method of repre- grew at the same time as its satellites; and 4) whether the
sentation, a gap of occupation is more likely to be correctly various parts of the site grew simultaneously or according
inferred than by visual examination of one- or two-sigma to an order that reflects expansion and contraction within
ranges. Similarly, it permits dates with different standard the settlement itself.
errors to be compared more realistically,portraying the less The macroregional relationship of La Quemada's
precise readings as flatter distributions. growth to that of the Mesoamerican core is now clear. La
A further advantage of this approach is that the prob- Quemada is an Epiclassic site in the truest sense of the
ability distributions of different dates can be combined. word; its growth occurred within the Classic period but
This additive property, which is achieved by standardizing coincided with and followed upon the decline of Teotihua-
the distributions associated with individual dates and then can. The probability distribution in Figure 7 represents the
Journal of Field Archaeology/Vol.
24, 1997 101

% Probabilistic Dates
18

16
14

12

10

4 -

O 0 LO LO0 LO 0 LO 0 0 0 LO 0 O0O 0
LO
COI'Cl C) O ( CO
-w- CO 0) 0- 0

Figure 7. Probability distribution of dates from La Quemada and its satellites combined.

whole range of dated contexts from La Quemada and its modern sample from Midden 12, the four dates from the
satellites. It represents all dates that have been obtained, Cuartel area that recently have been identified as post-oc-
minus those rejected. The distribution includes the 38 cupational (Hers 1989: 42; Trombold 1990: 310-313),
dates from our excavations plus dates submitted earlier by and one early date from the Cuartel that Trombold (1990:
other investigators (TABLE3). These include three from 313) regards as inexplicably early.
Griffin's 1956 collection (Crane and Griffin 1958), nine The peak of this distribution is at a.c. 650, and 75% of
from Armillas's 1963 season (notes from the University of the probabilistic dates fall into intervals with midpoints
Michigan Memorial-Phoenix Project Laboratory), and between a.c. 550 and 800. There is no evidence of multi-
four from Trombold's 1986 excavations at Site MV 138 modality, although there is a slight suggestion, not statisti-
(Trombold 1990), minus those that archaeologists have cally significant, of skewness toward the early end of the
rejected on stratigraphicgrounds. The excluded dates are occupation. Even though approximately one-third (14) of
the five early ones from Terrace 18 and Midden 11 that are the actual dates are from Terrace 18, the distribution
rejected on the basis of the above discussion, the one should be quite different if the overall span of occupation

Table 3. Radiocarbon samples submitted by other investigators.


Sample '3C adjusted Uncalibrated
no. Lab no. Site Submitted by 4Cage age date (A.c.)
'3C/2C ratio
40 M-430 La Quemada Griffin 885?200 - - 1065?200
41 M-431 La Quemada Griffin 775?200 - - 1175?200
42 M-432 La Quemada Griffin 1205?200 - - 745?200
43 M-1651 La Quemada Armillas 1230?120 - - 720?120
44 M-1652 La Quemada Armillas 1540?120 - - 410?120
45 M-1653 La Quemada Armillas 1230?120 - - 720?120
46 M-1654 La Quemada Armillas 1080?120 - - 870?120
47 M-1655 La Quemada Armillas 1180?120 - - 770?120
48 M-1656 La Quemada Armillas 770?110 - - 1180?110
49 M-1658 La Quemada Armillas 1020?120 - - 930?120
50 M-1659 Presa de AmboscoArmillas 1100?120* - - 850?120
51 M-1660 Presa de AmboscoArmillas 960?120 00-10.11t 1190?110 730?110
52 B-18194 Las Adjuntas Trombold - 00-10.11 - 620?70
53 B-18195 Las Adjuntas Trombold - - - 750?100
54 B-18196 Las Adjuntas Trombold - - - 780?70
55 B-28036 Las Adjuntas Trombold - - - 500?60
* The
University of Michigan archives contain a note about this sample stating: "The 1100?120 is correct. Published date at 950 is wrong.
Should be A.D. 850. JBG [James B. Griffin]."
f Correction for '3C/'2C ratio applied as suggested by Trombold 1990: 313.
102 Chronology and Stratigraphy at La Quemada/Nelson

were much unlike that of Terrace 18. The refined data set and through the 500s, one in the late 600s to the late
also includes six dates that are from the satellite sites of 700s, and a finalone in the earlyto mid 800s. Onlyone of
Presa de Ambosco and Las Adjuntas, and not from La these periodsmatchesLa Quemada,which shows strong
Quemada itself. The justification for including these dates growth in the 600s and 700s. This patternmight imply
is that the road network implies a highly integrated set of thatthe growthepisodesat the two centersweresomewhat
sites that is best regarded as a single community. Also, the independent.On the other hand,Schiavitti(1995) argues
few dates from the outlying settlements fall in the interval that Kelley'searlygrowth episode is illusory,and in the
defined by La Quemada's own occupation. meantime25 new, unpublishedradiocarbondates from
One concern in evaluating this overall distribution is AltaVistaremainto be factoredinto the evaluation(J. C.
that the monumental core, which makes up considerably Kelley,personalcommunication,1994). Importantpro-
more than half of the surface area of the site, is represented gress can thereforebe expected on the issue of regional
by dates from only two contexts-Midden 11 and the growthpatternsin the nearfuture.
Cuartel-whereas the flanking areas are represented by At the communitylevel, it is possibleto askwhetherLa
Terrace 18 and a number of middens. Also, there are no Quemadagrew before, after,or duringthe growth of its
dates taken directly from architectural contexts repre- spatialsatellites.The high degree of connectivitybetween
senting early strata in the monumental core; the core is La Quemadaand the other sites in the MalpasoValley,as
represented only by dates from the "outer shell" of occu- expressedin the road system, suggests contemporaneity,
pation in the Cuartel area and the sequence of deposits in but that connectivitycould maska sequenceof develop-
Midden 11. Although the latter should represent the ments. La Quemadacould have been founded consider-
whole spectrum of dates, it would be comforting to have a ablyearlieror laterthan most of the surroundingcommu-
broader sample. It is hoped that dates obtained from the nities, yet still have been connected to them by road at
work by the Gobierno de Zacatecas and INAH will address some point in the courseof events.
this gap. Figure8 givesthe probabilitydistributionfor datesfrom
More dates from deeper strata in the monumental core La Quemada contrastedwith those the from Presa de
are not likely to change the inference about overall span of Ambosco (notes from the Universityof MichiganMemo-
occupation very much, and almost certainly will not make rial-PhoenixProjectLaboratory)andLasAdjuntas(Trom-
it seem significantly later. The late dates from the Cuartel bold 1990) sites. The sampleof dates from the satellite
area are similar to the late-stratum dates from Terrace 18, sites is farfromadequate,especiallysinceboth sitesbelong
though perhaps enough later to allow the suggestion that to the clusterthat Trombold(1991: 151) labelsthe "Pila-
a remnant group made use of standing architecture in the rillos aggregate,"one of three groups that surroundLa
core for a century or so after the main occupation faded. It Quemada.Despite the potentialbiases,one can still ask
is reasonable to think of Terrace 18 as representativeof the whetherthese datescontradictthe presumedcontempora-
apogee of La Quemada's growth and the Cuartel as its late neityof LaQuemadaandthe smallersurroundingvillages.
end. If this inference is correct, the site may have been The distributioncenterson A.(C. 750, and the removalof
founded around a.c. 500 and occupied until around 900, the satellitesfrom the La Quemadadistributionpushesits
with the peak of use occurring at ca. a.c. 600-750. How- own apparentpeak back to A.C.650. If there were no
ever the data are viewed, there seems to be no support for questionsabout the representativeness of the samples,this
a significant Postclassic occupation. patternwould implya differenceof timing. In view of the
At the regional scale, the question is one of timing verylimitedrepresentationof the smallsites, however,it is
between La Quemada's growth and that of other centers. better to conclude that the model of contemporaneityis
If a peer-polity model (Renfrew 1986) accounts for the challengedby these data but not refuted.Additionalre-
growth of centers in northern Mexico (Jimenez 1992: searchis neededon the satellitepopulation.
192-196; Minnis 1989: 301-305), then La Quemada The intrasite scale is the smallest one at which questions
should be contemporary with other regional centers. Alta of contemporaneity can be addressed; here the data are
Vista, the only other well-dated center in far northern used to evaluate an expansion-contraction model. This
Mesoamerica, appears to have a span of occupation that is model predicts that the founding population of the site
similar to La Quemada's, ca. a.c. 450-900 (Aveni, Har- occupied what eventually became the core, that the site
tung, and Kelley 1982: 331-334; Kelley 1985). On the grew outwards in various directions from that core over
face of it this span would suggest support for a peer-polity time, and then eventually shrank back into the core at the
model. On the basis of a detailed analysis of construction end of the occupation. Why should this have been the case?
dates, however, Kelley (1985: 274) suggests that there are The core area is the most visible from the populated
three periods of growth at Alta Vista, one in the late 400s portions of the valley below, and it is clear that creating an
Journal of Field Archaeology/Vol.
24, 1997 103

% Probabilistic Dates
18 T

16
16 La Quemada
14
14 Sattelites
12

10

6
4

o LO O L)
O LO O LO O LO O LO O LO O LO O LO O LO
Ct CV) C O t O LO CD C N-
O - tO O) 0) O O

Figure 8. Probability distributions contrasting La Quemada with its satellites.

imposing monument was part of the occupants' design. this part of the site and reused elsewhere, reducing the
The core is surrounded by cliffs and an expansive enclosing representation of the latest occupation. Given these com-
wall, and many terraces are outside of that enclosed area, plexities it is impossible to know how representative the
suggesting that as the site grew, new terraces eventually dates are, but it seems fair to conclude that each part of the
had to spill outside the protected area. Finally,the greatest occupation has some chance of being represented. The
degree of architectural superposition appears to occur in probability distribution, therefore, probably provides a
the core, based on glimpses that can be gained from roughly accurate depiction of dates.
eroded areas. Midden 11, the other intensively dated deposit, affords
The probability distributions of three areasof the site are a different view of the occupational history of the site (FIG.
given in Figure 9. The distribution for Terrace 18 (FIG.9A) 9B). Materials in this midden should contrast with those in
centers on A.C.650 and appears to be a smooth unimodal Terrace 18 in their social contexts, their behavioralsources,
curve. Seventy-five percent of the probabilistic dates fall and possibly in the span over which they were deposited.
within the intervals bounded by the midpoints of A.C.550 Materials in Midden 11 were discarded from above, within
and 800. All else being equal, the interval A.c. 600-649 the monumental core of the site. Instead of consisting of
would be considered the one during which occupational construction materials, most of the dated specimens from
intensity was at its peak and the maximal number of Midden 11 probably represent fuelwood. There should be
datable specimens was generated. Depositional history, of a close relationship between death of the specimen and its
course, is not that simple; datable materials generated at period of use, and the problem of recycling should be
different times in the course of the occupation may have minimal.
unequal probabilities of being selected. Materials depos- If the site grew outward from a founding core, reached
ited early in the occupation may be less likely to have been an apogee, and then shrank back into that core prior to
recovered than later materialsbecause of the bias in excava- abandonment, then Midden 11 should have a substantially
tion toward the latest arrangement of structures. longer span of use than Terrace 18. There should be
Nine of the 15 dates are from the late arrangement of significantly greater proportion of dates on the early and
structures. On the other hand, all or almost all of these late ends of the spectrum, assuming that the differences in
dates are from structural members as opposed to fuelwood behavioral sources of dated wood do not distort the pic-
or other short-lived objects. With proper maintenance, ture. If, on the other hand, growth of Terrace 18 was
structural beams and posts may last a number of centuries. coeval with that of the core, then the span of dates as well
They were probably recycled when buildings were razed as the form of the curve represented in Midden 11 and
and rebuilt, so that some of the dates from the latest Terrace 18 should be very similar.
arrangement of structures could in fact represent the earli- The probability distribution of dates from Midden 11 is
est occupation. There is also the possibility that many of strikingly like that of Terrace 18 and has exactly the same
the latest beams were scavenged upon abandonment of peak. Seventy-five percent of the probabilistic dates fall in
104 Chronology and Stratzgraphy at La Quemada/Nelson

A
% ProbabilisticDates
20
18
16
14
n = 19
12
10
8
6

B
% ProbabilisticDates
25

20

15n=11

10

C
%Probabilistic Dates
16
14

12
10 n=5

Figure 9. Probability distribution of dates from various parts of La Quemada. A) Terrace 18. B) Mid-
den 11. C) The Cuartel Area.
Journal of Field Archaeology/Vol.
24, 1997 105

intervals bounded by the midpoints of A.C.550 and 750. Figure 10 is a chronologically ordered plot of individual,
This distribution suggests that the growth pattern at Ter- uncalibrated radiocarbon means against their one-sigma
race 18 was very similar to that in the monumental core. and two-sigma calibrated ranges as calculated by the
An alternative interpretation is that the areas of the monu- CALIB program of Stuiver and Becker (1986). An impor-
mental core that were "feeding" Midden 11 were some- tant pattern is visible here: most of the uncalibrated radio-
what marginal within the core itself, and thus the midden carbon means fall at the lower end of their one-sigma
reflects only a part of the core's occupational history. calibrated ranges. This implies that if an algorithm were
Arguing against this alternative is the fact that Midden 11 developed to produce a probability distribution from the
is the largest and deepest midden known at La Quemada, calibrated data (which are non-normal and in many cases
which suggests that it probably was one of the longest and multimodal), it would look very much like the one pro-
perhaps most widely used. Also, Midden 11 is located in a duced from the calibrated dates except that it would be
place that would have been conducive to trash deposition somewhat flatter because of the multiple intercepts and
from above even when the nearby terraces had not yet would have a later mean, perhaps by about 78 years, which
been constructed. It is situated below the edge of a cliff is the mean of the standard deviations in the sample. Thus
adjacent to one of the flattest parts of the core. The the calibration problem is not a cause for alarm or confu-
comparison between Terrace 18 and Midden 11 seems to sion; it is addressed relatively easily by the imprecise but
suggest, therefore, that the core and flank areas had quite sufficiently accurate judgment that the "true" peak of the
similar occupational histories. occupation is about 78 years later than the uncalibrated
Adding weight to the expansion-contraction model is probability distributions suggest.
the probability distribution from the Cuartel area (FIG. 9C). Another kind of refinement can be made by removing
The five acceptable dates from that portion of the site the dates that have wide standard deviations and multiple
center on an interval that is a century later than the peak of intercepts-although the effects of this latter selection
occupation for Terrace 18. If any of the even later dates must be considered carefully.Removing dates with multi-
that have been rejected in fact pertain to the site's occupa- ple intercepts essentially involves eliminating repre-
tion and not to post-occupational visits, then this tendency sentation of the periods when the calibration curve is
would be much stronger. Also, it must be noted that the jagged, meaning that a value on the axis of radiocarbon
excavation sample only encompasses the latest phases of ages can match more than one point on the axis of sidereal
construction in the Cuartel area. With an excavation strat- years. If the sample of acceptable dates is reduced by
egy like that used on Terrace 18, it is conceivable that one eliminating all dates with standard deviations greater than
would recover materials dating not only to the period of 80 years and all dates with multiple intercepts, then 100%
Terrace 18's occupation, but to an earlier phase as well. of the remaining calibrated intercepts fall within the inter-
Relying strictly upon the data available, however, the best val A.C 600-750. This pattern seems to provide fairly
interpretation seems to be that the early end of the expan- strong confirmation of the inferences already made about
sion-contraction model cannot be tested with the radiocar- the peak and span of La Quemada's occupation.
bon dates, although other data make it seem reasonable,
and that the late end of the model is confirmed. Conclusion
The discussion thus far has dealt with uncalibrated dates The history of interpretation surrounding La Quemada
in order to take advantage of some convenient properties provides a good example of the importance of detailed
of the normal curve. As noted above, this approach fails to analysis of chronology. Wishing to make the most of avail-
account for the fact that radiocarbon years are not quite able data, archaeologists may accept unfounded assump-
equal to sidereal years, and also ignores the problem of tions about dating that in turn affect their interpretations
multiple intercepts between the radiocarbon and sidereal of broader issues. In the case of La Quemada, scenarios of
curves. To some extent these concerns are minimized by conquest, colonization, and exploitation by distant peoples
the fact that the entire chronology of Mesoamerica has such as the Toltecs were plausible because the inadequate
been constructed on the basis of uncalibrated dates. The dating of the site. While many finer-grainedchronological
beginning date of A.C.750 for the Alta Vista phase, for issues remain, the new data permit alignment of La Que-
example, is uncalibrated. Considering the effects of calibra- mada's growth and decline with events and processes in
tion is important, however, for comparison of future deter- the wider Mesoamerican world.
minations with these data, and can also help to refine La Quemada probably was founded in the early A.C.
current inferences. Particularlygermane is consideration of 500s. Although a number of earlier dates have been dis-
the multiple-intercept problem. cussed, virtually all are anomalous with respect to their
106 Chronology and Stratigraphy at La Quemada/Nelson

1200-

1100-

1000.

900
48

16
800 2
54
447

24 1933

9 25 28 34
60 0 ....
......... ....4..
13. 4 ...... I .. .. . ...

35 38

500 8............... ... .... .................

400

300

200

orderedplot of uncalibrated
Figure 10. Chronologically radiocarbonmeansagainsttheirone-sigma
and two-sigmacalibratedranges.Y-axisrepresentsyearsA.C.Note: date no. 39, a.c. 650 + 50, is not in-
cluded in the plot.
Journal of Field Archaeology/Vol.
24, 1997 107

stratigraphiccontexts.The existenceof materialdatingto cadencewas coincidental,and if not, why politieson the


the 4th or 5th centuriesdoes not demonstratethat the site peripheryflourishedwhile those in the core disintegrated.
was occupied at those times since the dates could come
fromwood that was alreadyold at the time of use. On the Acknowledgments
other hand, some of the earlycontexts at the site may be The work summarizedin this paper was conducted
unrepresentedbecausethey are buriedbeneathlaterarchi- underpermitfromthe InstitutoNacionalde Antropologia
tecture.La Quemadagrew into its main occupationfrom e Historia. The field project was made possible by the
600-750 and by the late 800s had shrunkback into its support and cooperationof Peter Jimenez Betts, Gerala
monumentalcore. Full abandonment,except for periodic FelixCherit,BaudelinaGarciaUranga,Ratl ToledoFarias,
visits,probablyoccurredin the early900s. Archaeologists AlejandroMartinezMuriel,LorenaMirambellSilva,and
suggesting that dates from the 9th and 10th century Jose FranciscoRomfinGutierrez.Much of the field work
representpost-occupationalreuse of the site as a shrine wasablysupervisedby AndrewDarling,VincentSchiavitti,
appearto be on solid ground. Terrace18 representsthe Nicola Strazicich,and PaulaTurkon.The authoris grateful
apogee of the site'sgrowth and probablyspansmost of its to Robert Santleyand two anonymousreviewersfor their
occupation.One of the mainentrancesto the westernside insightfulcriticismsof an earlierdraftof the paper.Figures
of the site passes through Terrace 18, suggesting that 1 and 3 were drawn by Victoria Vargas;Figure 2 was
Terrace18 was integralto the site duringthe height of its illustratedby MichaelKagelmacher and is reproducedwith
development.Thereis little or no Postclassicoccupationat permissionof the Society for AmericanArchaeology.Fi-
La Quemada;despite some of its seemingly Postclassic nancial support was provided by the National Science
traitssuch as the largecolonnadedhallit cannothavebeen Foundation(GrantsBNS-8806238 and DBS-9211681),
founded,and possiblywas not even occupied,duringthat the Facultyof Social Sciencesof the State Universityof
period. New Yorkat Buffalo,andthe NationalEndowmentfor the
The historical and processual implications of these Humanities,an independentfederalagency.
findingsare outlinedin the introductionto this paperbut
are worth reiteratingin the light of the above discussion.
The appealingidea that La Quemadawas a Toltecoutpost Ben A. Nelson (Ph.D. Southern Illinois University at Car-
establishedto facilitatethe acquisitionof turquoisefrom bondale, 1980) is Assistant Professorat Arizona State
the AmericanSouthwestis unsupported.La Quemada's University. His researchinterestsinclude the archaeology
of northern Mexico and the American Southwest,prehis-
occupation was not contemporarywith that of Tula toric political organization, ethnoarchaeology,and ce-
Grandenor with that of ChacoCanyon,New Mexico,the ramic analysis. Nelson is currently directing the Malpaso
proposedterminusof the turquoisetrail.Also, the growth Valley-LaQuemada Archaeological Project in Zacatecas,
of La Quemadaprobablycannotbe attributedto anyother Mexico. Mailing address:Department of Anthropology,
core state, since the growth of the site falls between the Arizona State University, Box 772402, Tempe,AZ 85287-
apogees of "mega-polities"in the Mesoamericancore re- 2402.
gion.
The notion that La Quemadawas an outpost of empire,
whetherof Teotihuacan,the Toltecs,the Tarascans,or the Armillas,Pedro
1964 "CondicionesAmbientalesy Movimientos de Pueblos en
Aztecs, should be abandonedin favor of more dynamic
models of mutualismor autonomy.Peripheralpopulations la FronteraSeptentrionalde Mesoamerica,"in Homenaje
a Fernando Marquez-Miranda.Publicacionesdel Semi-
were not simplyreactingto processesin the core, but were nario de EstudiosAmericanistasy Seminario de Antro-
constructingsocial power in accordancewith their own pologia Americana. Madrid: Universidadesde Madrid y
local needs and resources. New local power structures Sevilla,62-82.
developed within conditions created by core disintegra- Aveni, Anthony F., Horst Hartung, and J. CharlesKelley
tion. La Quemada is only one of a number of peripheral 1982 "AltaVista (Chalchihuites),AstronomicalImplicationsof
a MesoamericanCeremonial Outpost at the Tropic of
polities that flourished at about the same time on the Cancer,"AmericanAntiquity47: 326-335.
northern periphery of Mesoamerica (Cabrero 1989, 1991;
Batres, Leopoldo
Jimdnez 1989; Trombold 1990). It is now clear that the 1903 Visita a los Monumentos Arqueologicos de "La Que-
growth of La Quemada occurred not in concert with that mada,"Zacatecas.Mexico City:Imprde la vda. de F. Diaz
of the core, nor independently of it, but apparently in de Leon.
inverse proportion to it. One theoretical challenge before 1971 "Visit to the ArchaeologicalRemains of La Quemada,
archaeologists of the region is to determine whether this Zacatecas, Mexico," in Basic C. Hedrick, J. Charles
108 Chronology and Stratigraphy at La Quemada/Nelson

Kelley, and Carroll L. Riley, eds., The North Mexican Jimenez Betts, Peter
Frontier.Carbondale:Southern Illinois UniversityPress, 1989 "PerspectivasSobre la Arqueologia de Zacatecas,"Ar-
1-20. queologia5: 7-50.
CabreroG., MariaTeresa 1992 "Una Red de Interacci6ndel Noroeste de Mesoamerica:
1989 Civilizaciin en el Norte de Mexico:Arqueologiade la Una Interpretaci6n,"in BrigitteBoehm de Lameirasand
Canada del Rio Bolanos.Mexico City: UniversidadNa- Phil C. Weigand,eds., Origeny Desarolloen el Occidentede
cional Aut6noma de Mexico. Mtxico. Zamora, Michoacan, Mexico: Colegio de
Michoacan, 177-204.
1991 "CulturaArqueol6gicade Bolafios (Zacatecasy Jalisco):
Una Frontera Cultural,"Ancient Mesoamerica2: 193- Kelley,Ellen A.
204. 1978 "TheTempleof the Skullsat AltaVista, Chalchihuites,"in
Carroll L. Riley and Basil C. Hendrick, eds., Acrossthe
Clavigero,FrancescoSaverio ChichimecSea:Papersin Honor of J. CharlesKelley.Car-
1979 TheHistoryofMexico,Vol. 1. New York:GarlandPublish- bondale:SouthernIllinois UniversityPress, 102-126.
ing. [Translationof the StoriaAntica del Messico, reprint
of the the 1787 edition reprinted by G. G. J. and Kelley,J. Charles
J. Robinson, London.] 1985 "The Chronology of the ChalchihuitesCulture,"in Mi-
chael S. Fosterand Phil C. Weigand,eds., TheArchaeology
Coe, Michael D. of Westand NorthwestMesoamerica.Boulder, Colorado:
1994 Mexico.London: Thames and Hudson. WestviewPress,269-288.
Coggins, Clemency Kintigh, Keith W.
1979 "Teotihuacin at Tikal in the EarlyClassicPeriod," Actes 1994 Toolsfor Quantitative Archaeology. Privatelypublishedby
du 42e CongresInternational de Amaricanistes8: 251- the author,2014 E. AlamedaDr., Tempe, AZ 85282.
269. Paris:InternationalCongressof Americanists.
Millon, Rend
Crane, H. R., and James B. Griffin 1988 "The LastYearsof TeotihuacanDominance,"in Norman
1958 "Universityof Michigan RadiocarbonDates II," Science Yoffee and George L. Cowgill, eds., The Collapseof An-
127: 1098-1105. cient Statesand Civilizations.Tucson: Universityof Ari-
zona Press, 102-164.
Culbert, T. Patrick
1988 "Political History and the Decipherment of Maya Minnis, Paul E.
Glyphs,"Antiquity 62: 135-152. 1989 "The CasasGrandesPolity in the InternationalFour Cor-
Darling, J. Andrew ners," in Steadman Upham, Kent G. Lightfoot, and
1993 "Notes on Obsidian Sources of the Southern SierraMa- Roberta A. Jewett, eds., The SociopoliticalStructureof
dre Occidental,"Ancient Mesoamerica4: 245-253. Prehistoric SouthwesternSocieties. Boulder, Colorado:
WestviewPress:269-306.
Dean, JeffreyS.
1991 "Thoughts on Hohokam Chronology,"in George J. Gu- Nelson, Ben A.
1990 "ObservacionesAcerca de la Presencia Tolteca en La
merman, ed., Exploringthe Hohokam:PrehistoricDesert
Peoplesof the American Southwest.Albuquerque:Univer- Quemada, Zacatecas," in Federica Sodi Miranda, ed.,
Mesoamericay Norte de Mexico Siglos IX-XII, Vol. 2.
sity of New Mexico Press, 61-150. Mexico City: InstitutoNacionalde Antropologiae Histo-
Diehl, RichardA. ria, 521-540.
1983 Tula: The Toltec Capital of Ancient Mexico. London:
Thames and Hudson.
1993 "Outposts of Mesoamerican Empire and Architectural
Patterningat La Quemada,Zacatecas,"in Anne I. Woose-
GarciaChavez, Raiil ley and John C. Ravesloot, eds., Culture and Contact:
1993 "EvidenciasTeotihuacanasen Mesoamericay su Posible Papersin Honorof CharlesC. Di Peso.Albuquerque:Uni-
Significadoparala Cronologia de Teotihuacan,"in Taller versityof New Mexico Press, 173-190.
de Discusionde la Cronologiade Teotihuacan.Fondo Na- 1995 "Complexity,Hierarchy,and Scale:Chaco Canyon, New
cional Arqueol6gico, Centro de Estudios Teotihuacanos, Mexico, and La Quemada, Zactecas."American Antiq-
Instituto Nacional de Antropologia e Historia, Mexico uity 50: 597-618.
City,207-228.
Nelson, Ben A., J. Andrew Darling, and David A. Kice
Hers, Marie-Areti 1992 "Mortuary Practices and the Social Order at La Que-
1989 LosToltecasen TierrasChichimecas.Cuadernosde Investi- mada,Zacatecas,"Latin AmericanAntiquity3: 298-315.
gacionesEstiticasNo. 35. Mexico City:Instituto de Inves-
tigaciones Est&ticas,UniversidadNacional Aut6noma de Nelson, Ben A., and Vincent W. Schiavitti
Mexico. 1992 Trabajosconducidospor la State Universityof New Yorkat
BuffalodentrodelProyectoLa Quemada1989-90. Buffalo:
Holien, T., and R. B. Pickering Department of Anthropology, State University of New
1978 "Analoguesin Classic Period ChalchihuitesCultures to Yorkat Buffalo.
Late MesoamericanCeremonialism,"in Esther Pasztory,
ed., Middle Classic Mesoamerica:A.D. 400-700. New Nelson, Ben A., PaulaW. Weintraub,and Vincent W. Schiavitti
York:ColumbiaUniversityPress, 145-157. 1993 Informe Parcial del ProyectoValle de Malpaso-LaQue-
Journal of Field Archaeology/Vol. 24, 1997 109

mada, Temporada1992. Buffalo:Departmentof Anthro- Within the Northern MesoamericanPeriphery,"Ancient


pology, State Universityof New Yorkat Buffalo. Mesoamerica4: 255-270.

Noguera, Eduardo Weaver,Muriel Porter


1930 Ruinas Arqueoldgicasdel Norte de Mixico:CasasGrandes 1981 TheAztecs, Maya, and Their Predecessors: Archaeologyof
(Chihuahua), La Quemada, Chalchihuites(Zacatecas). Mesoamerica.(2nd edn.) New York:AcademicPress.
Mexico City: Secretariade Educaci6nPiblica. 1993 TheAztecs, Maya, and Their Predecessors: Archaeologyof
Mesoamerica.(3rd edn.) San Diego: AcademicPress.
Osborn, Ann
1979 La Ceramicade los Tunebos:Un EstudioEtnografico.Bo- Weigand, Phil C.
goti: Fundaci6n de InvestigacionesArqueol6gicos Na- 1970 "Ceremonial Reuse of a Fluted Point by the Huichol
cionales. Indians,"AmericanAntiquity 35: 365-367.
Pijoan, C., and J. Mansilla 1977 "The Prehistoryof the State of Zacatecas:An Interpreta-
1990 "Evidencias rituales en restos humanos del norte de tion," in Cuauhtemoc EsparzaSanchez, ed., Anuario de
Mesoamerica,"in F. Sodi Miranda, ed., Mesoamiricay Historia Zacatecana. Zacatecas, Mexico: Universidad
Norte de Mixico SsiglosIX-XII, Vol. 2. Mexico City: Aut6noma de Zacatecas,1-39.
Instituto Nacionalde Antropologiae Historia,467-478. 1982 "Mining and Mineral Trade in PrehispanicZacatecas,"
Renfrew,Colin Anthropology6: 87-134.
1986 "Introduction:Peer Polity Interactionand Sociopolitical Weigand, Phil C., Garmen Harbottle, and EdwardV. Sayre
Change," in C. Renfrew and J. Cherry,eds., Peer Polity 1977 "TurquoiseSources and Source Analysis:Mesoamerica
Interactionand Socio-PoliticalChange.Cambridge:Cam- and the SouthwesternU.S.A.," in Timothy K. Earleand
bridge UniversityPress, 109- 116. JonathonE. Ericson,eds., ExchangeSystemsin Prehistory.
New York:AcademicPress, 15-34.
Santley,Robert S.
1989 "Obsidian Working, Long-Distance Exchange, and the
TeotihuacanPresenceon the South Gulf Coast," in Rich-
ard A. Diehl and Janet C. Berlo, eds., MesoamericaAfter
the Decline of Teotihuacan,A.D. 700-900. Washington,
D.C.: Dumbarton Oaks ResearchLibraryand Collection,
131-151.
1994 "The Economy of Ancient Matacapan," Ancient
Mesoamerica5: 243-266.

Schiavitti,Vincent W.
1995 Organizationof the PrehispanicSuchilMining District of
Chalchihuites,Mexico,A.D. 400-950. Ph.D. Dissertation,
State University of New York at Buffalo. Ann Arbor:
UniversityMicrofilmsInternational.
Schiffer,Michael B.
1986 "RadiocarbonDating and the 'Old Wood' Problem:The
Case of Hohokam Chronology,"JournalofArchaeological
Science13: 13-30.

Stuiver, Minze, and Bernd Becker


1986 "High-PrecisionDecadal Calibrationof the Radiocarbon
Time Scale, A.D. 1950-2500 B.C." Radiocarbon28:
805-838.

Trombold, CharlesD.
1990 "A Reconsiderationof the Chronology for the La Que-
mada Portion of the Northern MesoamericanFrontier,"
AmericanAntiquity 55: 308-323.
1991 "Causewaysin the Context of StrategicPlanningin the La
Quemada Region, Zacatecas, Mexico," in Charles D.
Trombold, ed., Ancient Road Networksand Settlement
Hierarchiesin the New World.Cambridge: Cambridge
UniversityPress, 145-168.
Trombold, CharlesD., James F. Luhr, ToshiakiHasenaka,and
Michael D. Glascock
1993 "ChemicalCharacteristicsof ObsidianFromArchaeologi-
cal Sites in WesternMexico and the TequilaSourceArea:
Implicationsfor Regional and Pan-Regional Interaction