The Fortaleza Valley, Peru: Archaeological Investigation of Late Archaic Sites

(3000–1800 BC)
Author(s): Winifred Creamer Alvaro Ruiz Rubio Manuel F. Perales Munguia Jonathan Haas
Source: Fieldiana Anthropology, Number 44:1-108. 2013.
Published By: Field Museum of Natural History
DOI: http://dx.doi.org/10.3158/0071-4739-44.0.1
URL: http://www.bioone.org/doi/full/10.3158/0071-4739-44.0.1

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and
environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published
by nonprofit societies, associations, museums, institutions, and presses.
Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of
BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.
Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries
or rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research
libraries, and research funders in the common goal of maximizing access to critical research.
The Fortaleza Valley, Peru: Archaeological Investigation of Late
Archaic Sites (3000–1800 BC)
Winifred Creamer Alvaro Ruiz Rubio
Adjunct Curator, Anthropology, Field Museum Calle Miguel de Cervantes Saavedra 192
Anthropology Department La Molina
Northern Illinois University Lima, Perú
Dekalb, Illinois 60115 USA

Manuel F. Perales Munguia Jonathan Haas

Instituto Cultural Del Norte Chico MacArthur Curator of the Americas,
Universidad Continental Anthropology Department
Huancayo, Perú Field Museum
1400 South Lakeshore Drive
Chicago, Illinois 60605 USA

Abstract
Monumental architecture, including stepped pyramids, sunken circular plazas, and upright monoliths (huancas), has
been identified at sites throughout the Norte Chico region along the coast of Peru. During 2003 and 2004, test excavations
were conducted at six of these sites in the Fortaleza Valley. Excavation included 1 3 2-m test pits and sections cleared along
road cuts and looters’ pits. Good preservation of plant materials, along with fiber bags, or shicra, used in construction of
monumental architecture yielded ample samples for dating. This report describes and illustrates the precise context of the
samples and provides the radiocarbon dating results for 80 samples. Project results show that the sites tested were occupied
during the Late Archaic Period between 3000 and 1800 BC. The pattern of constructing large mounds paired with sunken
circular plazas clearly begins by the third millennium BC, and a U-shaped layout of sites appears to develop during that
period, as indicated by the regularity of site plans and monuments at sites such as Caballete and Porvenir.

Resumen
La arquitectura monumental, incluyendo pirámides escalonadas, plazas circulares hundidas, y monolitos verticales o
huancas, han sido identificados en sitios en toda la región Norte Chico en la costa del Perú. Durante 2003 y 2004, las
excavaciones de prueba se llevaron a cabo en seis de estos sitios en el Valle de Fortaleza. Las excavaciones incluyeron fosas
midiendo 1m x 2 m y secciones despejadas a lo largo de los cortes de carreteras y las fosas de saqueadores. La buena
preservación de los materiales vegetales, junto con bolsas hechas de fibra o shicras utilizados en la construcción en la
arquitectura monumental dio amplias muestras para poder conseguir la edad por medio de radiocarbono. Este informe
describe e ilustra el contexto preciso de las muestras y proporciona los resultados de datación por radiocarbono de 80
muestras. Los resultados del proyecto indican que los sitios analizados fueron ocupadas durante el perı́odo Arcaico Tardı́o
entre 3000-1800 antes de Cristo. El patrón de construir grandes montı́culos emparejado con plazas circulares hundidas
comienza claramente durante el tercer milenio antes de Cristo y el plan de construcción en forma de "U" parece desarrollarse
durante este periodo, según lo indicado por la regularidad de los planos del sitio y monumentos como Caballete y Porvenir.

Introduction demonstrated by Andean chronology. In Lumbreras’s devel-

opmental sequence (Table 1), the period 3000–1800 BC is the
The four valleys of Huaura, Supe, Pativilca, and Fortaleza Late Archaic, implying the final transition from semisedentary
constitute what is known locally as the Norte Chico (Fig. 1). to settled village life and food production. Rowe characterized
All these valleys are relatively close to each other, with a the period as ‘‘preceramic’’ based on the absence of pottery
distance of less than 60 km separating the southernmost from production in the Peruvian Andes at that time, without
the northernmost. Inland, dry streambeds connect the valleys reference to architecture.
to one another. Although the Norte Chico was on the margins At approximately 3000 BC, however, large residential and
of Peruvian civilization for most of the past 5000 years of ceremonial sites began to appear along the coast, especially in
history and evolution of Andean civilization, it appears to the valleys of the Norte Chico. By 2500 BC, there were 10 or
have played a starring role in the emergence of that more of these sites growing side by side, and by 2200 BC, there
civilization. That archaeologists did not anticipate construc- were more than 30. Among the other large sites occupied in
tion of monumental sites during the third millennium BC is Peru in the third millennium BC were La Galgada (Grieder

FIELDIANA: ANTHROPOLOGY N.S., NO. 44, OCTOBER 1, 2013, PP. 1–108 1

 FIG. 1. Map of Norte Chico region highlighting the Fortaleza Valley and showing sites tested in 2003–2004 and sites mentioned in the text.

et al., 1988), El Paraiso (Engel, 1966; Quilter, 1985), and there was a clear need to determine whether mounds like those
Huaca Prieta (Bird et al., 1985), each an individual isolate. built at Huaca Prieta (Bird et al., 1985), Aspero (Feldman,
From a panregional, Andean perspective, the Norte Chico 1980), and Caral (Shady et al., 2001) were unusual early
stands out as a remarkable example of pristine cultural expressions of complexity or whether the Late Archaic was
florescence at the very beginning of one of the world’s great characterized by many sites with large mounds and associated
civilizations. architecture. We therefore undertook investigation of sites in
In this volume, we document the occupation of six sites in the Fortaleza Valley that lacked ceramics (that appear along
the Fortaleza Valley. When we first visited these sites, each the coast of Peru in the Initial Period, 1800–900 BC) and had
displayed characteristics of Late Archaic sites that had just the large-scale stone platform mounds of Late Archaic Period
been dated to the third millennium BC (Shady et al., 2001). sites in the nearby Supe Valley (Zechenter, 1988; Shady, et al.
Absolute dates for early mound construction were rare, and 2001; Shady, 2004). The testing of these Fortaleza Valley sites

2 FIELDIANA: ANTHROPOLOGY
was preceded by similar testing of mound sites in the Pativilca Pativilca and Fortaleza valleys were recognized as belonging
Valley in 2002 (Creamer et al., 2007; Ruiz et al., 2008). Testing to the Late Archaic. PANC crews conducted informal
of 13 sites in both valleys removed any doubt that reconnaissance in the Huaura, Pativilca, and Fortaleza valleys
construction of large pyramidal structures began during the in 2000, 2001, and 2002 and located a number of sites that
earliest years of the Late Archaic and continued for more than were similar to known Late Archaic sites by the general
a millennium. Presentation of the research carried out absence of ceramics on the surface or in erosional cuts or by
provides context for the chronological data obtained from the presence of pyramidal mounds, sunken circular courts,
each site tested in the Fortaleza Valley and contributes to the upright stone monoliths or huancas, and occasional twined
ongoing discussion of changing complexity in the Andes. textiles, another hallmark of the Late Archaic (Moseley &
The long history of archaeological research in the Norte Barrett, 1969). Systematic and comprehensive surveys of all
Chico began with Max Uhle early in the 20th century. Willey three valleys from the coast to where the valleys narrow at the
and Corbett (1954) brought attention to the region with their start of the Andean foothills were conducted in 2004–2007
excavations at Aspero and other sites in the Supe Valley, (Nelson & Ruiz, 2005; Perales, 2006, 2007). A total of over
though Aspero was considered an Early Horizon site based on 1,500 archaeological sites and over 10,000 separate site
a nearby cemetery (Feldman, 1980). The first awareness that occupations were recorded in the course of these surveys
the region as a whole had a significant role in Andean history (many ‘‘sites’’ in this area were occupied at multiple times in
came with the informal surveys conducted by Paul Kosok and the course of the past 5000 years). Over 30 large sites with
Richard Schaedel in the 1940s and published by Kosok (1965, various combinations of public and residential architecture
p. 223). Kosok recognized the large number of stone platform were identified in the four valleys of the Norte Chico.
mounds in the region and referred to it as the ‘‘Pativilca Survey has inherent limits in clarifying the chronological
Complex’’ of early but unknown age. The importance of the placement of seemingly early sites that lack pottery, as
region as a focus of particularly early cultural development potential variability in settlement patterns is directly relevant
increased when Moseley and Willey (1973) recognized that the to any interpretation of the nature of the Late Archaic
fishing community of Aspero, in the Supe Valley, dated to the occupation (see Creamer et al., 2007). In looking only at
Late Archaic Period (3000–1800 BC, also referred to as the surface indications at various sites identified during the
Late Preceramic or Cotton Preceramic; Table 1) and had informal survey portion of the project, there were clear
communally constructed platform mounds. Subsequently, indications that some had relatively limited and shallow
Engel (1987) and Zechenter (1988), also working in the Supe deposits and were likely occupied for relatively short periods
Valley, identified the cluster of sites including Chupacigarro of time, while others had deeper and much more extensive
Grande, now known as Caral, and dated them to the Late deposits indicating longer occupations. Without pottery, it
Archaic. Following on survey by Williams and Merino (1979) proved difficult to subdivide the 1,200 years of the Late
in the Supe Valley, Shady (1997, 2003a, 2004) recorded the Archaic. We turned to excavation to obtain material for
large early sites and initiated excavations at Caral. Shady’s absolute dating to confirm both the placement of the sites
first dates for Caral (Shady et al., 2001) confirmed the within the Late Archaic and their position within the
observations of Engel and Zechenter indicating that these 1,200 years of that period. For this reason, testing at seven
large inland sites with monumental architecture indeed dated sites in the Pativilca Valley was carried out during 2002
to the third millennium BC. (Creamer et al., 2007; Ruiz et al., 2008), and test excavations
More broadly, the coast of Peru was also used as an were conducted at six sites in the Fortaleza Valley in 2003. As
example in the debate over the origins of complex society from mentioned, all were likely Late Archaic occupations based on
a variety of perspectives. Carneiro (1970) used the coastal the presence of large-scale surface architecture and the absence
valleys of Peru as a case study in his proposal that of surface ceramics.
circumscription, competition, and warfare led to the develop- We present the results of testing in 2003 and 2004. The
ment of hierarchy in society (see also Daggett, 1987; Wilson, Fortaleza is the northernmost of the valleys that make up the
1987). Moseley (1975, 1992) suggested that the coastal waters Norte Chico. The river’s source, high in the Andes, follows a
of Peru, unusually rich in fish and mollusks, provided the steep and narrow canyon until reaching the coastal plain,
surplus production necessary for the development of social where the valley widens, and there are swathes of cultivable
complexity (cf. Raymond, 1981; Wilson, 1981; Quilter, 1992). land on either side. The Fortaleza River usually dries up
Haas (1982) applied a cross-cultural perspective, incorporating during the austral winter months of July and August, similar
data from the complex societies of the north coast of Peru in to the Supe River. The Pativilca River carries a generous
trying to unpack the diverse variables that affect the volume of water year-round (Kosok, 1965). Our area of
development of centralized decision making, hierarchies, and investigation extended from the coast, where the river empties
social inequality (1982). Using a climate change approach, into the Pacific Ocean at Cerro La Horca, to the settlement of
Sandweiss et al. (1996) proposed a model in which the gradual Shaura some 33 km inland. This is the lower valley, the zone
constriction of the coast through rising sea level increased of floodplain, and the slopes of the lowest foothills of the
population density and spurred culture change. Andes. Sites investigated in 2003 included Porvenir, Caballete,
The Proyecto Arqueológico Norte Chico (PANC) was Cerro Blanco 1, Cerro Blanco 2, Huaricanga, and Shaura.
initiated in 2001 to look at the emergence and development of In 2003, a sample of 1 3 2-m test pits was excavated at each
complex societies across the valley boundaries to get a wider site. Walls of existing Looter’s Pits were cleared and the
regional picture. Engel (1987) carried out informal survey in profiles recorded, as were some sections of monumental
the Pativilca Valley, and Vega-Centeno et al. (1998) undertook architecture disturbed by construction. In 2004, one of these
a survey in the Fortaleza Valley associated with the sites, Caballete, was targeted for more extensive testing.
construction of a power line. At the time of these surveys, Caballete is one of the largest and best-preserved Late Archaic
none of the large sites with monumental architecture in the sites in the Norte Chico and merited additional attention. At

CREAMER ET AL: FORTALEZA VALLEY, PERU 3

Caballete, more 1 3 2 test units were excavated as were two the architecture, and exposed profiles. State of preservation
larger units of 5 3 5 m. The latter were designed to expose included a preliminary description of the deterioration of the
samples of architecture and will be discussed in a future site and areas destroyed by human or natural forces. This
volume. The results of mapping, test excavations, and clearing information has proved valuable in providing a base condition
of existing looters’ pits to obtain samples for radiocarbon from which looting, bulldozing, and other destruction can be
dating follow, including the 80 radiocarbon dates obtained documented over time.
from these six sites.
Site Identification and Excavation Strategy

Each site discussed was identified on historic air photos

Field Methods (1940s and 1969–1970) of the Servicio Aereo Nacional.
Architectural remains of archaeological sites, such as mounds,
A pattern of large, Late Archaic sites distributed along the platforms, sunken courts, walls, and looted areas, could be
valley margins was identified in the Pativilca Valley in 2002 identified in areas not covered by agricultural fields, roads, or
(Haas et al., 2004; Creamer et al., 2007) and appeared to structures. Follow-up ground-truth visits recorded mounds
continue in the Fortaleza Valley based on anecdotal informa- and other features identified on air photos (Table 2), sites
tion. The present project was undertaken to determine where archaeological research had been carried out and sites
whether the Late Archaic network of sites extended into the indicated by local residents. Subsequent literature search
Fortaleza Valley or whether the Pativilca Valley was the revealed that the sites having mounds and circular courts
northern margin of the system. Many sites in the Fortaleza had been identified as belonging to either the Initial Period
Valley owe their excellent state of preservation to their (1800–1000 BC) or the Early Horizon (1000–300 BC; Vega-
location above the floodplain on the margins of coastal Centeno et al., 1998). An additional site, Cerro de la Cruz
valleys that cannot be reached with gravity-flow irrigation. (Vega-Centeno et al., 1998), was not identified on the ground
This land, called eriazo, has remained largely uncultivated for in time to be included in the testing program. Detailed survey
the past 5000 years and is covered with many archaeological of the region was not completed until 2007 (Perales, 2007).
sites. For millennia, farmers have lived in this zone, Due to their large size, most sites were subdivided into
maximizing the availability of land that can be irrigated and sectors around specific mounds, though sectors also included
cultivated. Today, it is the place where archaeological sites can areas of obvious occupation, disturbance, or cemetery debris
still be found since most land located on the valley floor has without mounds. These were established before the sites were
been under cultivation for centuries, and ancient sites have mapped, and as a result, lettered sectors vary a great deal in
been gradually plowed under. During the 20th century, much size, especially those not centered on a structure. Sectors
of the floodplain was bulldozed flat and crisscrossed by centered on structures were not set up with boundaries during
irrigation canals to facilitate large-scale cultivation of crops the 2002–2003 field seasons. Starting in 2004, the ad hoc
such as cotton and sugarcane. nature of sectors proved cumbersome and was replaced with
The goal of obtaining samples for radiocarbon dating and site units of 100 3 100 m based on UTM measurements. Site
samples of macrobotanical, pollen, and faunal remains was maps show both the grid and sectors as used during the field
amply met, and the resulting data are subject to ongoing work.
analyses. Subsequent excavations at Caballete and Huari- During 2002–2003, a small number of test units were
canga have yielded data on residential and ceremonial excavated each season, and each of these was numbered
architecture. separately. At Caballete in 2003, for example, only two test
The inventory of archaeological sites identified during the units were excavated: Sector B, Test Unit 1, and Sector C,
project included eight characteristics that were recorded at Test Unit 2. However, in 2004, Test Units A–F were also
each: site name, UTM coordinates, altitude, location, descrip- excavated at Caballete, resulting in two test units:
tion, previous research, time period, and state of preservation. numbers ‘‘1’’ and ‘‘2.’’ In addition to identifying location
Names assigned by Vega-Centeno et al. (1998) were used; in by 100 3 100-m units, subsequent excavations at Caballete
the cases where the site had not been identified previously, we and at Huaricanga identified excavations at ‘‘operations’’ in
used the names given by local residents to the areas or names different parts of the site. Each operation was identified by a
of associated geographic features taken from maps of the roman numeral, and all excavation units were identified
Instituto Geografico Militar. Latitude, longitude, and UTM within the operation. A set of inventory numbers were
coordinates were determined by GPS or from 1:100,000 maps assigned to each operation (e.g., 5000–6000) to avoid overlap
of the Instituto Geografico Militar. Altitude was determined in inventory numbers from different operations. The excava-
by GPS or topographic maps incorporated into a regional GIS tions described in this volume use the ‘‘sector’’ and ‘‘test
based on readings from Trimble differentially corrected GPS unit’’ designations used in the field during the 2003–2004 field
units calibrated to submeter accuracy. Location refers to seasons, even though the sectorization of the site around
geographic location in relation to nearby towns and cities as mounds was not used subsequently.
well as geographic features and political units, such as district At the sites tested in 2003, samples of charcoal and annual
and province. Descriptions of sites were made for each plant material for radiocarbon dating were collected in two
architectural unit, including archaeological material on the ways. In site after site in the Fortaleza Valley, we encountered
surface, site layout, and the number of superimposed exposed profiles remaining from the activities of looters or
components that could be identified. History included all modern construction activity, and we were able to take
known previous investigations at the sites visited. Time period advantage of these areas to collect data. By examining profiles
refers to the chronological position of the site, determined by already exposed by looting or construction, we were able to
the material culture visible on the surface and in looted areas, collect samples from deep within some mounds, places that

4 FIELDIANA: ANTHROPOLOGY
could never have been reached by test pits. Construction in all Fortaleza Valley from Chasquitambo to the Pacific Ocean in a
but one of the sites tested in the Fortaleza Valley used shicra, GIS completed in 2007 (Perales, 2007). The GIS incorporates
woven bags filled with stone and used in mound construction, satellite imagery and historic air photos in a georeferenced
somewhat like a sandbag (Quilter, 1985). The arid climate has database along with locational data from mapping and
preserved these bags and provides an excellent source of regional survey. Thus, all the map information shown here is
annual plant remains that are suitable for radiocarbon dating also part of the GIS.
(Shady et al., 2001; Creamer et al., 2007).
In each of the disturbed areas selected, the loose earth was Radiocarbon Dates: Expressing Calibrated Dates Used in
cleared from the face of the damaged area, and a vertical face This Volume
was created by troweling that exposed an undamaged surface.
After the profile was photographed and drawn, samples for Radiocarbon dates provided by a lab are not exact dates.
radiocarbon dating were extracted from the profile. Care was All dates come with a 6 range that for the present group of
taken to ensure that the samples included organic material samples ranges from a low of 40 to a high of 290. In addition,
that had not been exposed to the elements. In addition to dates within the target range of the project, between 5000 and
shicra bags, plant fiber was often mixed in with clay to create 2500 radiocarbon years before present (RCYBP), have to be
floors. Some samples for dating were extracted from the fiber ‘‘calibrated’’ to adjust for differences in the atmospheric
temper in floor fragments. carbon over the past 5000 years. All of the dates reported here
Test pits were also used to obtain samples in areas that have been calibrated using the Calib 5.0 program. The Cal BC
remain undamaged, situated to sample deposits of stratified dates given here are intended solely to provide a convenient
trash. This strategy was used to maximize the number of average for general comparison. To arrive at these specific
samples that could be collected from each test pit and to dates, we have calculated a weighted average for each. To
sample as much of the occupation sequence of each site as arrive at this weighted calibrated date, we used the ‘‘relative
possible. Test pit locations were selected after examination area under probability distribution’’ at the 95.4% (2s)
of erosion channels, looters’ pits, or other disturbed areas confidence interval. The Calib program generally gives
and through the use of a handheld soil tester to obtain between one and four date ranges, each with a given
samples. Each test pit was excavated using artificial or probability. Thus, for example, an RCYBP date of 3770 6
natural levels depending on the context. When natural layers 70 run through Calib emerges with the following at a 95.4
were identified, they were used to guide excavation. When confidence level: 2458–2417 (4.0), 2410–2022 (95.3), and 1992–
natural layers could not be observed or were more than 1983 (0.7). Thus, the probability that the date falls between
30 cm thick, arbitrary levels of approximately 15 cm were 2458 and 2417 Cal BC is 4.0%, the probability that it falls
used. In general, most test pits were excavated using between 2410 and 2022 Cal BC is 95.3%, and the probability
arbitrary levels, though levels were correlated with natural that it falls between 1992 and 1983 Cal BC is 0.7%. To arrive
features. For example, a level was ended when a use surface at a weighted average, one that includes all the probabilities,
or floor was identified. This method maximized control of an average for each of these ranges is calculated [(2458 +
vertical provenience, valuable in contexts lacking architec- 2417)/2 5 2437.5]. This average is then multiplied by the
ture, such as midden deposits. Most radiocarbon samples probability, in this case 4.0%, to arrive at a weighted
and all of the samples that have been processed for dates contribution of 98 years. The contribution for the range
were collected in situ. All excavated material was screened 2410–2022 5 2112 years, and that for the range 1992–1983 5
through J-inch screen and a portion was also screened 14 years. The weighted contributions are totaled 98 + 2112 +
through 1/8-inch screen. Two-liter soil samples were 14 5 2224, rounded to the nearest 10 for a weighted average
collected for flotation and fine screening (2 mm), and pollen calibrated date of 2220 Cal BC. It must be emphasized that
samples were also collected. Excavation focused on retriev- this Cal BC date is only a comparative figure for purposes of
ing samples of flora, fauna, and artifacts from stratified broad discussion. It summarizes the probability of an interval
deposits along with radiocarbon samples. No effort was of time around the weighted average, and is ultimately more
made to clear architectural features or to identify individual indicative of an approximate age of a sample and chronolog-
structures or burials. ical relationships between samples than RCYBP dates,
unweighted averages date, or uncalibrated dates.
Mapping
Charcoal, Radiocarbon, and Old Wood
Boundaries for each site were established using several
criteria, including the area over which tiny marine shell One of the advantages of working on the Peruvian coast is
fragments could be observed on the surface and the that the dry climate preserves organic material. One problem
surrounding natural features, such as the walls of a dry wash, faced in such environments is that of ‘‘old wood,’’ providing
or quebrada. The location of each identified site was initially misleading early dates (Schiffer, 1986). Residents of the Late
recorded with a handheld GPS, and the information was Archaic sites in the Norte Chico could have picked up wood
transferred to maps at a 1:100,000 scale, obtained from the from dead trees or broken branches in this desert environment
Instituto Geografico Militar. Detailed maps incorporating and used it for building construction such as door lintels, as
5000–25,000 points were compiled for each of the sites tested, firewood, or in toolmaking. Radiocarbon dating identifies the
though not all maps were completed by the end of the field point when carbon is no longer absorbed by the plant, not
testing program in 2003. Maps of the large sites discussed in when the material was employed. Charcoal from such wood
this volume were mapped with a Topcon 210 total station or a yields dates that older than the hearth where they were
Leica TPS 1200 Robotic total station between 2003 and 2007 collected, just as tools made by carving out the hard center of
and were incorporated with field survey data from the lower a tree would yield older dates than wood at the outer edge

CREAMER ET AL: FORTALEZA VALLEY, PERU 5

growing when the tree was cut down or died. Not all charcoal mound closest to the large central open area or plaza (Fig. 3).
samples will yield ‘‘false’’ early dates. Just as the residents can The Sector A circular court is heavily damaged. Later visitors
use old wood for fires, they can also use ‘‘new wood’’ for to Porvenir used part of the sunken court and the low
fires—wood cut from living plants or recently dead plants. platform around it for interments, particularly during the
New wood would yield dates that did reflect the age of the fire Middle Horizon. More than 50% of this zone has been heavily
though not if from interior wood of long-lived species that can disturbed, and many human bones, ceramics, textiles, and
be relatively ‘‘old’’ compared to the exterior of the tree, as other burial items remain on the surface. Looters even
mentioned. To alleviate as much of the ‘‘old wood’’ problem constructed lean-to shelters from discarded textiles. Appar-
as possible, we made an effort to select annual plant fibers ently, there were some higher-status tombs in this area, as the
(leaves, grasses, reeds, and seeds) when available for radiocar- disturbance is more than 2 m deep, and four beams more than
bon dates. These remains survive in the dry environment of the 5 m long that appear to have been removed from the area were
Norte Chico, are abundant at the sites tested, and avoid the piled off to one side of Mound A in 2004. Later burials were
issue of internal/external differences in age of wood from long- interred in the circular courts of other Late Archaic sites as
lived species. However, even annual fibers that present little well (Haas & Creamer, in press). The Sector B circular court,
possibility of being chronologically out of context can yield set in a low rectangular platform and clearly visible on the
stratigraphically inconsistent dates. They may reflect material surface, was not subject to similar reuse. Two other cemetery
that has been moved or contaminated by the exigencies of areas are visible at Porvenir, located in the central open area of
everyday life. They also may fall at one end or the other of their the site. The presence of ceramics in the looted materials in all
respective probability ranges when dated, not conveniently in these cemetery areas indicates that they postdate the Late
the middle. Although it is conceivable that some ‘‘old’’ annual Archaic occupation of Porvenir.
plant fiber material might be incorporated into midden An area to the east of Sector A covered with mollusk valves
material, this is highly unlikely. For making shicra bags and dispersed fragments of charcoal may be an area of short-
specifically, weaving the bags requires fresh reeds. Older fibers term residences. To the east of Sector H are the remains of a
become brittle and difficult or impossible to weave. Woven rock wall that outlines a trapezoidal area about 50 m across,
reeds deteriorate from exposure to light and humidity and are identified as Sector I. An area of intense looting at Porvenir
unlikely to last many years. Shicra bags might be exposed resulted in a large gouge out of the mound in Sector F. This
during remodeling of structures and could possibly have been latter hole was reportedly made with heavy machinery and
reused, but this seems unlikely based on the loss of tensile removed an extensive area within the structure. It is visible in
strength that occurs as reeds dry (Asencios, 2009). air photos taken in early 1970. Considering the length of time
Radiocarbon dating charcoal samples versus annual plant this area has been exposed to the elements, it was a high
samples may yield systematically different results. In three priority for the 2003 field season to clear a portion of this cut
proveniences at the site of Caballete, pairs of samples were and examine the stratigraphy of the structure. Despite the
collected, both charcoal and annual plant fiber. In each case, effects of looting, Porvenir is still among the best-preserved
the resulting calibrated dates for charcoal are 300–400 years Late Archaic sites in the region since the probable residential
older than the annual plant fiber samples (Table 3). These areas around the structures are preserved and the site is not
data are best considered anecdotal but suggest that further under cultivation.
testing is warranted to examine whether annual plant fiber and At Porvenir, three 1 3 2-m test units were excavated, and
charcoal samples yield systematic differences in radiocarbon two exposed profiles were recorded. Test Units 1 and 2 were
dating results. excavated in Sector H in areas that appeared to be residential
trash deposits. Profile 1 was an area cleared in the deep looted
section of Sector F. Test Unit 3 was excavated in Sector A in
an area thought to be residential trash. This unit was
Porvenir specifically targeted to test the hypothesis that scatters of
crushed shell, small charcoal fragments, and lithic fragments
Porvenir is located on the right margin of the Fortaleza visible on the surface represent residential areas (Wulffen,
River, 1.5 km northwest of the village of Lampay and 5 km 2009). A looters’ pit (Sector A, Profile 2) adjacent to the
north of the center of the city of Paramonga. The archaeo- mound in this sector was also recorded, and samples of shicra
logical site fills the mouth of a dry streambed, 50 m above sea were collected.
level. It is bounded on either side by rocky extensions that
extend south from the hills called ‘‘El Porvenir.’’ Adjacent Sector A, Profile 2
fields of sugarcane cover terrain between the site and the
Fortaleza River that was probably both occupied and Looters excavated a hole approximately 4 m in diameter in
cultivated in ancient times. a raised area just off the north corner of the main mound at
Porvenir consists of six mounds distributed around a large Porvenir. While the main mound was built of quarried stone
open space in a U-shaped arrangement of structures that retaining walls filled with shicra bags and loose rock and earth,
opens toward the river valley to the south-southeast (Fig. 2). this area was largely devoid of surface stone. A section of one
The mounds on the east side of the site are widely separated wall of this pit was cleaned and drawn. The three radiocarbon
from the main mound some 100 m distant. The intervening samples were taken from previously unexposed, in situ
area has been heavily washed out, and no indication of deposits with very little chance of contamination (Table 4a–
structures that may have been located in this part of the site c; Table 5; Fig. 4). The profile of the pit revealed relatively
can be seen today. simple stratigraphy with two upper layers of fill separated by a
Two structures at Porvenir are platform mounds with thin clay layer with the outline of a shallow depression. The
associated sunken circular courts located near the side of the lower portion of the profile consisted of very dense

6 FIELDIANA: ANTHROPOLOGY
 FIG. 2. Ground plan of Porvenir showing sectors and location of excavation units and profiles.

concentrations of shicra fibers that may denote an area where unlikely to represent a continuous occupation or use for
shicra bags were prepared for construction and use in the 1000 years. The most that can be inferred from these disparate
adjacent monument. The dates taken from this profile, while dates is that the locality was used during the Late Archaic and
internally consistent from top to bottom (Fig. 4a, c), are may reflect multiple phases of utilization.

CREAMER ET AL: FORTALEZA VALLEY, PERU 7

 FIG. 3. Photographic overview of Porvenir taken from the hills east of the site.

Sector A, Test Unit 3 the Test Unit 3 area for an undefined length of time after the
period of the dated samples. Use eventually ceased in both
This sector includes the main mound located at the northern places, though we cannot determine whether this was a
end of the site. The body of the mound is rectangular, though sitewide abandonment or a shift to a different sector of the
an elevated platform on its east side distorts the overall form, site. There is also evidence of reuse of this area at a later time.
resulting in a kind of L-shaped structure. The surface of the No date is associated with this later period of use in Test Unit
mound is covered with angular rock, with no visible ceramics 3, while later use of the area shown in Profile 2 was dated 1750
and with only a few mollusk fragments. The mound measures Cal BC and is associated with the Phase II fill (Profile 2;
86 3 61 3 10 m. As with other sites in the region, the mound Fig. 4a), suggesting the later occupation dated to the
was built on a slope so that the front—the side facing into the beginning of the Initial Period (1800–900 BC).
center of the site—gives an appearance of being taller, though
the actual cultural volume of the mound is less than is present Sectors B–E
in the mound complex in Sector F (see below). A sunken
circular court is adjacent to the mound, facing the central open Sectors B–E were not subject to testing. Together, these
plaza area of the site. The top of the mound is a mass of structures make up the western arm of the site. Sector D
depressions and raised areas indicating the presence of appears to be aligned with the largest mound, Sector A, while
significant construction of rooms on the top; however, it was Sectors B, C, and E are similarly aligned, forming one ‘‘arm’’ of
not possible to define specific structures. the site. The circular court in Sector B suggests that there was a
The north side of the mound shows evidence of a possible very specific focus in this part of the site, but its use in relation
residential structure or platform. A deep deposit of midden is to the larger Sector A circular court remains to be established.
visible in the sides of a nearby looters’ pit, and while abundant Sector B includes a rectangular mound 55 3 49 3 8 m.
botanical remains can be seen, no pottery was visible. Sector B is adjacent to and south of Sector C. The surface is
Excavation in Sector A was confined to a single unit, Test covered in angular rock with mollusk fragments dispersed
Unit 3, though some comparison can be made with Profile 2. over the entire surface. The central part of the mound shows a
Dates were obtained from the lower levels of both units. In central atrium, and a depression is bordered with rock
Test Unit 3, one sample was obtained from plant fiber indicating its original rectangular shape. A staircase connected
incorporated into a clay use surface, and a second sample the atrium with the clearly visible sunken circular court 23 m
came from fill overlying the clay surface. The dates 2370 and in diameter, set inside a rectangular, raised platform at the
2700 Cal BC (Table 4d, e; Table 6), fall into the middle of the foot of the mound.
third millennium BC. In Test Unit 3 as in Profile 2, an upper Located between Sectors B and D, Sector C includes a
layer of fill consisting of earth, rock, shell, and cultural rectangular mound 51 3 29 3 5 m. The surface of the mound is
materials was identified over a use surface, with a second layer covered with angular rock and a very few river cobbles. On the
of fill below the occupied layer (Fig. 5). Below both layers of mound surface, the upper edges of several rock walls are visible
fill was evidence of construction that yielded the dates in each that are probably the remains of structures built atop the mound.
test unit. In Profile 2, the construction was a solid layer of Sector D is located north of Sector C and west of Sector A
shicra fill associated with construction of a terrace or and includes a rectangular mound 47 3 45 3 7 m, with steeply
platform, while in Test Unit 3, the deposits consist of several sloping sides covered with angular rock and mollusk
layers, including two of clay that are floors or use surfaces. It fragments but with no ceramics on the surface. The alignment
seems that these two parts of Sector A were initially occupied of the structure makes it appear to be more closely associated
about the same time, around 2700 BC, with use continuing in with the main mound in Sector A than with the mounds in

8 FIELDIANA: ANTHROPOLOGY
 FIG. 4. Sector A, Profile 2, at Porvenir, profile of looters’ pit showing depositional layers and the location of dated carbon samples.

CREAMER ET AL: FORTALEZA VALLEY, PERU 9

 FIG. 5. Profile of Test Pit 3, Sector A, Porvenir, showing levels and layers.

Sectors B, C, and E. The Sector D mound shares a similar events undertaken on the mound in Sector F. The five
orientation with the other mounds, however, and suggests a radiocarbon dates from Profile 1 range from 3750 to 1830 Cal
U-shaped layout for the site as a whole. BC (Table 4f–j; Table 7) and suggest a long occupation
At the southeast corner of the site is Sector E, consisting of history at Porvenir throughout the Late Archaic Period.
two low rectangular mounds. The north mound is an irregular Given the association of the very early date of 3750 Cal BC
shape covering 33 3 28 3 4 m and the south mound is with fill below a floor near the summit of the structure, it
rectangular, 43 3 41 3 6 m. The two structures are separated seems likely that this date is not reflective of that construction
by some 12 m. The surface of each is covered with angular event. It is, however, well within the range of dates from other
rock, fragments of mollusk valves, and a few ceramic excavations at other Late Archaic sites (see the discussion of
fragments. The area has been used by herders and is covered radiocarbon dating below). The date of 1830 Cal BC yielded
with caprine feces. Despite this, some wall segments of rock by sample ‘‘f’’ (Fig. 6) may also be from disturbed material,
with clay mortar are visible on the surface. The mounds in given its association with two other samples that are earlier,
Sector E are approximately aligned with those in Sectors B samples ‘‘g’’ (2100 Cal BC) and ‘‘i’’ (2440 Cal BC). The profile
and C, creating a row of structures. was cleared in a disturbed area that despite its size did not
reach the base deposits of the structure. It is likely that at least
Sector F one earlier occupation phase remains to be identified in this
Sector F mound. The most consistent dates from this structure
On the east side of Porvenir, Sector F includes a rectangular range on average from 2440 to 2100 Cal BC, suggesting a
structure 61 3 49 m in size that is 10 m high. All sides of the period when the structure was being built and remodeled.
mound are steeply sloping, covered with angular rock and Clearing in this area revealed several construction episodes
mollusk fragments. The extensive digging through the north that consisted of wall construction, layers of floor plaster, and
side into the center of the mound covers an area 25 m long 3 clay floors (Fig. 6; Table 7, Layers B–J). There are several
5 m deep and is visible on aerial photographs from 1970. The ways to envision the walls and floors in estimating the
sides of this hole are heavily eroded and architectural detail is construction phases for this area. Overall, the portion of the
not visible. No fragments of ceramics could be seen. Profile 1 profile where architectural remains could be detected suggests
explored this deep cut in 2003. a minimum of three construction phases of walls and plastered
floors that were covered by later construction. The uppermost
Profile 1, Sector F deposit that could be identified was a pair of clay floors
(Fig. 6, Floor 1).
Profile 1 covered 8.5 3 3 m and provides a unique view into This profile at Porvenir illustrates the way in which mounds
the interior of a large mound, showing a series of construction increased in size with the addition of retention walls that were

10 FIELDIANA: ANTHROPOLOGY
 FIG. 6. Sector F, Profile 1, at Porvenir.

filled using shicra bags and then covered with a layer of clay Sectors F and G are aligned parallel to Sectors B–D at a
plaster. Mounds were modified to accommodate a range of distance of 230 m. Together these two sets of mounds create
activities that probably changed over time. The changes in two parallel ‘‘arms’’ of the site layout in the form of a U. A
architecture within the mound show a large investment in huaico, or flash flood, removed approximately 25% of the site
labor and complex organization to make it possible to redefine from where Sectors A and I meet on the north, across the site
spaces on different structures. Deposits below Wall 2 in the center and through the south end of Porvenir. This flooding
profile are homogeneous and suggest that there was at least could have removed additional mounds oriented similarly to
one major period of construction that raised the mound the mound in Sector D or in the area noted as Sector I that
substantially in a single episode. The profile does not reach would make the U-shape layout more distinct as well as other
submound deposits, however, and does not show whether this structures in its path. We did not find evidence of such
episode was the initial construction effort. The profile above structures during the limited testing program carried out.
Wall 2 shows several different construction events that show
that the mound was remodeled or expanded on numerous Sector H
occasions, gradually increasing its size. Although every mound
at Porvenir and elsewhere is going to have a unique In the northeast corner of the site, Sector H includes an area
architectural history, the profile of the Sector F mound of approximately 50 3 30 m on the surface where a
demonstrates the variability in deposits that is hidden beneath concentration of mollusk valves and dark-colored earth is
the modern surfaces of these structures. visible, suggesting an area with deposits of botanical material
in midden. The level surface may indicate a residential area,
Sector G though Sector H falls within the washed-out zone. The area
was selected for testing, as it was near the main mound of
Immediately south of the mound in Sector F is Sector G, in Sector A and offered the possibility of high-status residences
which there is a mound similar in orientation and dimensions or trash. Two test units were excavated in this area. In
to that in Sector F, 102 3 59 3 8 m, though the mound in addition, Profile 2 in Sector A mentioned above is located near
Sector G is slightly lower. The surface of the mound in Sector Test Unit 1 in Sector H.
G is covered with angular rock and mollusk fragments but no
ceramics. South-southeast of the mound in Sector G are Sector H, Test Unit 1 and Test Unit 2
structures of adobe bricks, with ceramics visible on the surface
and visible looting disturbance. No testing was conducted in Test Unit 1 revealed a low wall topped with a layer of clay,
Sector G. identified as a retaining wall that was part of a low platform

CREAMER ET AL: FORTALEZA VALLEY, PERU 11

 FIG. 7. Porvenir Sector H, Test Unit 1.

(Fig. 7). Samples yielded two dates (Table 4k, l; Table 8). A III, this area was used as a midden. It was at this time that
shicra fragment from construction fill behind the retaining deposits included ash and food remains, forming the very dark
wall provided a date of 2580 Cal BC (Table 4l) and suggests a layer visible during excavation. The midden was not homo-
Late Archaic date for this feature. The later date, 1250 Cal BC geneous throughout and included a layer of large rock. Trash
(Table 4k), is from plant fibers collected from a similar disposal here eventually ceased, as indicated by the formation
context but not clearly part of a shicra bag. This could indicate of a use surface atop the dark midden at the end of Phase III.
later remodeling or even a later date of use. It is also possible Wall 1 was built on that surface, part of Phase II activities,
that these fibers were out of place, fallen from above or and formed the side of a platform about 50 cm high (Fig. 8).
somehow contaminated. Alternatively, the older date could be The wall and platform were built on top of the midden and did
in error, and the platform may date to the Initial Period. When not intrude into it. As in Test Unit 1, this platform could have
compared to the dated units in Sector A, the older date from been the base for residential structures, or it could be related
Test Unit 1 and the lower deposits from the Profile 2 and Test to ceremonial activities carried out on the nearby mound in
Unit 3 in Sector A all date to the mid-third millennium BC and Sector A.
seem to indicate a period when the site was in use and Wall 1 was excavated as a single provenience, Level 12. A
construction was under way. sample of charcoal from the wall yielded a date of 2840 Cal
Test Unit 2 (Table 9) in Sector H was situated at the east BC (Fig. 8; Table 4n). Since this wall segment crossed the
edge of a road that crosses the site of Porvenir, a locality center of the unit, it is visible only in Figure 9. Subsequent
chosen for testing when an animal burrow nearby revealed Phase II deposits include fill behind the retaining wall and
deep ancient trash, including shell, ash, and lithic fragments deposits after the platform was no longer in use, including
(Fig. 8). The depth of the cultural deposits in Test Unit 2 layers filled with earth and large rock and, above that, shicra.
reached almost 2 m. Unlike most of Sector H, Test Unit 2 is on These may be partly from the collapse of structures built on
a slightly raised area that was not visibly affected by the the low platform, though no indications of structures were
extensive washout across this part of Porvenir. encountered in this test unit. The shicra sample in Level 5
A number of artificial levels were used to partition the dated 2330 Cal BC corresponds to the top of the Phase II
deposits in Test Unit 2, and these were grouped into with four deposits. This shicra bag was distinctive in size and shape but
phases, or periods, of use. During the earliest phase (IV), rock was surrounded by fragments of shicra and loose plant fibers
and earth were deposited in this area, possibly associated with rather than other whole bags, suggesting that it could have
clearing or construction nearby. Subsequently, during Phase fallen from its original context (Table 4m). Test Unit 2 shows

12 FIELDIANA: ANTHROPOLOGY
 FIG. 8. Porvenir Sector H, Test Unit 2, west face.

two phases of use of the area as a midden, once before the Porvenir Discussion
platform was built (Phase III) and later after the platform was
no longer in use (Phase I). The span of dates associated with Three 1 3 2-m test units were excavated at the site and the
Test Unit 2 suggests active use of this area during the early to two looters’ pits cleared, encompassing three different sectors
mid-third millennium BC, similar to the other areas tested. of the site. Together these indicate the location of midden and
Earlier deposits may be present, but a date is not associated areas of buried architecture within the site. Each test unit
with the lowest levels excavated in this unit. identified deposits more than 1 m deep and yielded dates
spanning the Late Archaic Period as well as slightly earlier and
Sector I later periods.
The 13 radiocarbon dates obtained from Porvenir range
Sector I is located east of Sector H and includes a from 3750 to 1250 Cal BC. Comparing the dated phases
rectangular area outlined by a wall of rock and clay. The (Table 4), it is apparent that all the tested areas were
wall is now in fragments, a few of which remain to indicate its occupied during the Late Archaic between 3000 and 1800
orientation. Air photos of the region taken in 1970 show the BC. Eleven of the dates fall within a relatively continuous and
wall in a more complete state. The wall is oriented similarly to overlapping cluster between about 2840 and 1750 Cal BC
the mounds at the site, though its age could not be determined. (Table 4), suggesting principal use of Porvenir during this
No testing was done in Sector I. interval.

CREAMER ET AL: FORTALEZA VALLEY, PERU 13

 later expanded as debris from weaving accumulated. The
platform was later covered by rock and then dense concentra-
tions of partial shicra bags, junco (Graminea Schoenoplectus sp.)
used for making shicra bags, and other plant fibers. These may
indicate its use as a workshop for using wild cane, junco, and
cotton in making string, rope, mats, and shicra bags, all of
which were found in abundance. Employing limited excava-
tions, at Porvenir, it was possible to demonstrate the presence of
mound construction and remodeling, repeated creation of use
surfaces and their occupation accompanied by midden deposits,
and specialized activities, such as weaving, all during the third
millennium BC.

Caballete

Caballete is located on the right (north) bank of the

Fortaleza River, 9 km north-northeast of the center of the city
FIG. 9. In Sector H Test Unit 2, Porvenir, the corner of a of Paramonga, at approximately 150 m above sea level. The
retaining wall, part of a low platform, was uncovered. Packets of
folded willow branches were associated with the edge of the platform. site fills the mouth of a wide, dry wash bounded by hills that
extend southeast from the chain of hills noted as ‘‘La
Empedrada’’ on national maps. Caballete consists of six
mounds arranged around a large open space (Fig. 10).
The oldest date, 3750 Cal BC from Phase IV of Profile 1
Indications of occupation in the Caballete locality extend
(Fig. 6j), appears to be an outlier but could be a valid date.
over an area of approximately 400 ha. The site also had a very
The sample was annual plant fiber removed from a fragment
long and complex occupation (Haas et al., 2010). A large piece
of clay plaster coating Wall 1 within the mound construction.
of the site was plowed and laced with canals for an irrigation
If accurate, this would suggest that the earliest construction at
project in the 1970s, but the project ultimately failed; no water
Porvenir began well before 3000 Cal BC, though more testing
ever passed through the irrigation system. Plowing uncovered
would be needed to determine whether such an early patches of marine shell, ash, and burned rocks, indicating
construction date is reliable. numerous and extensive areas of pit hearths and cooking.
The most recent date from Porvenir, 1250 Cal BC from Most of the site, including mounds and surface architecture,
mixed plant fiber in Sector H, Test Unit 1 (Fig. 7k), may be remains intact. In the southwest corner of the site is a looted
accurate and indicate use of this area during the late Initial cemetery area where human remains and ceramic fragments
Period, or it could indicate contamination in the fiber sample. are dispersed across the surface. Three of the mounds are
The other date from Sector H, Test Unit 1, is from a shicra associated with sunken circular courts. The space the site
sample, yielding a date of 2580 Cal BC (Fig. 7l), less likely to occupies, its dimensions, and the architecture of the site
inadvertently have been mixed with other fibers. It is suggest that it was among the most important of the
impossible to draw any conclusions about the oldest and preceramic sites in the Norte Chico region. Caballete was
youngest dates based on these two isolated dates alone. The divided into sectors around each of the mounds. Subsequently,
units tested show different aspects of site construction and the site was subdivided into 100 3 100-m units incorporated
occupation. The looters’ pits in Sectors A and F show mound into the GIS. The sectors are described below and are
construction. The profile cleared in the Sector A looters’ pit indicated on the site map along with the 100 3 100-m units
revealed only a single mound construction episode, while the (Fig. 11).
Sector F profile includes evidence of several different
modifications. Although it is now certain that Porvenir dates Sector A
to the Late Archaic Period, it will take additional excavation
to determine when the earliest construction at the site took Sector A includes the principal mound of the site and is
place and the sequence or contemporaneity of the numerous located at the south end of the site, near the edge of an alluvial
construction events identified in the excavations. platform that forms the base of the La Empedrada wash. The
Similarly, Test Unit 1 and Test Unit 2 in Sector H reveal the Sector A mound is rectangular and measures 125 3 54 m and
use of midden to level uneven or sloping ground surfaces. In is approximately 17 m high (Table 2). Adjacent is a circular
each of these places, the test units also uncovered low court 15 m in diameter, outlined by huancas (see Fig. 46). The
platforms that were built on the leveled surfaces. Test Unit 1 axis of the stairway on the face of the mound bisects the court,
showed a single floor layer on the constructed platform while although the court is not visibly sunken like the others at
in Sector A, Test Unit 3 showed repeated construction of use Caballete.
surfaces associated with a low platform. Test Unit 2 shows The surface of the mound includes areas of quarried angular
greater accumulation of debris. The platform was built after rock and others of fine-textured grayish-tan sediment mixed
use of the area as midden to a depth of about a meter. This area with sand and fragments of rock. A small number of marine
may have been residential, though the volume of plant remains, shell fragments are visible on the surface. A broad atrium is
string, rope, and mat fragments suggests a weaving workshop. visible at the top of the mound that faces the circular court of
Alternatively, the platform may have been an activity area that huancas. The northwest portion of the mound looks different

14 FIELDIANA: ANTHROPOLOGY
 FIG. 10. Overview photo of Caballete taken from the hills west of the site.

and appears to have been constructed as an addition at a fine gravel without artifacts of any kind and there were
different time than the bulk of the mound. Less rock is visible irregularities in the surface in addition to the pit that yielded
on the surface in the northwest portion than elsewhere similarly dark-colored material.
(Fig. 10). Flanking platforms at either end of the mound can Test Unit 4 yielded two early dates, 2730 Cal BC from
also be identified, each covered with rock alignments, the Level 10 and 2700 Cal BC from the pit fill at the base of the
foundations of rooms constructed there. Large blocks of stone unit (Table 11b, g); the most recent date from this unit is
outline the base of the mound, most clearly visible on the 1890 Cal BC (Table 11d). These dates show occupation in
south side of the mound facing the valley, away from the this part of the site throughout the Late Archaic Period. The
sunken court. These blocks tend to be naturally rectangular in dates do not fall in chronological order according to
shape, and many have polished faces. There is no evidence that excavated stratigraphy. In every case, it is possible that
these stones were cut to shape or modified beyond polishing. botanical material could have been disturbed in ways that
There are remnants of this stone outline, however, on all sides affected the positioning of the samples of different dates. As
of the mound. discussed previously (see ‘‘Charcoal, Radiocarbon, and Old
Wood’’ above), annual plant fibers made into shicra bags
Sector A, Test Unit 4—2004 are unlikely to demonstrate the ‘‘old wood’’ problem, leaving
the question of positioning and date as areas for further
This unit was located on an area that was slightly elevated examination.
compared to the surrounding terrain, a few meters north of the Test Unit 4 shows that this part the site was in use from
west wing of the main mound in Sector A (Fig. 12). The area 2700 to 1890 Cal BC, and occupation in this locality is divided
appeared to be comparable to the locations of Test Pits 1 and into five phases of activity (Table 11). Two of these, Phases IV
2, excavated in 2003, and which proved to have deep stratified and III, are distinctive. Phase IV includes use of the area for
deposits. trash disposal with a large component of midden debris and
The stratigraphy of this deep test unit reveals a complex dark-colored ash, in use from 2730 to 2200 Cal BC (Fig. 10c–
depositional history (Table 10). Near the surface in Level 4 is a g). A pit excavated into the gravel below the occupied surface
possible flood deposit that would appear to date long after the at the base of the unit contained a large bundle of stems and
abandonment of the site. More than a meter below the surface, leaves of willow twisted into small packets (Fig. 14) that may
Wall 1 marks the corner of a platform (Fig. 13; Fig. 14) and represent an offering.
shows that structures at Caballete may be completely invisible Phase III represents a change in use of this area, indicated
on the surface. Dates obtained from deposits associated with by the clay floors constructed over the midden deposits at the
the floors built against Wall 1 include 2730 and 2390 Cal BC top of Layer E at the end of Phase IV. Wall 1 was then built as
(Table 11a, b). A date of 2220 Cal BC comes from a sample a retaining wall and marks further Phase III use. The
from the base of Level 15, along a former ground surface associated platform and floors make up a separate phase
(Table 11c). Samples from a pit feature filled with midden when this area was an occupied platform and not a trash
yielded dates of 2700 and 2320 Cal BC (Table 11f, g). Two disposal area. Phase III dates are 2730 and 2390 Cal BC,
samples from the base of Level 17 outside the pit feature associated with Wall 1. The date of 2730 Cal BC is from plant
yielded dates of 2300 and 1890 Cal BC (Table 11d, e). The material that could represent older material used as fill. The
dates from the pit fill overlap with one sample from Level 17. date of 2400 Cal BC comes from a shicra bag, probably an
It is possible that those three—samples e, f, and g—come from accurate date of construction of the platform, because shicra
material of similar age since Level 17 was composed largely of bags have to be woven of reeds while they are freshly cut and

CREAMER ET AL: FORTALEZA VALLEY, PERU 15

 FIG. 11. Map of Caballete showing sectors, structures, and excavation units.

pliable, filled immediately as part of the construction process. overlap in dates occurs between approximately 2500 and 2150 Cal
Dried reeds are not pliable enough to weave into the bags, and BC. The dates for this test unit illustrate how difficult it is to use beta-
dry bags filled with stones or gravel would break when lifted. counted (non-AMS) radiocarbon dates to gain detailed information
Platform construction often incorporates layers of shicra about the specific chronology of a stratigraphic sequence since older
alternating with construction debris (see Haas and Creamer, calculated radiocarbon dates may overlie younger results (see
2006; Stanish, 2006). ‘‘Stratification and Radiocarbon Dates’’ below).
The dates of 2220 and 1890 Cal BC (Table 11c, d) The most recent date in Test Unit 4, 1890 Cal BC, may be
demonstrate some of the problems with expressing radiocar- out of place or contaminated. If the extremes, the oldest and
bon dates. In the case of Sample 11c with a weighted most recent dates, are not reliable, then the entire pit
calibration date of 2220 Cal BC, there is a 95.4% probability represents relatively rapid accumulation of debris associated
that the sample dates to the period 2410–2022 Cal BC, while with construction and occupation over the course of about
Sample 10d has a 95.4% probability of dating to the period two centuries from 2400 to 2200 Cal BC. If the outliers do
2036–1738 Cal BC. The calibrated dates give an approxima- reflect extreme ranges of occupation, then construction was
tion of actual age while the date ranges overlap, meaning that spread over about 800 years from 2730 to 1890 Cal BC. The
despite the differences in their weighted calibration dates, they data from this test pit show that for at least two centuries (and
could come from samples of similar age. possibly much longer), there was construction involving walls,
Table 11 ends with a graphic representation of all the dates floors, and platforms as well as periods with frequent
from Test Unit A with their respective ranges. The maximum deposition of trash in this part of Caballete.

16 FIELDIANA: ANTHROPOLOGY
 FIG. 12. Caballete, Sector A, Test Unit 4 (2004), profile view.

Sector B layer of rock. One of the looters’ pits in the mound was
cleaned of debris and profiled to record multiple phases of
Sector B is located at the west side of Caballete and includes building and remodeling along with of mound construction.
a rectangular mound 41 3 39 m at the base and 5 m high. The One test unit was located north of the mound in an area that
surface of the mound in this sector was covered with soil and had what appeared to be deposits of domestic refuse.
angular rock, though there was less surface rock in Sector B
than elsewhere. Only a few mollusk fragments were visible on Sector B, Profile 1—2003
the surface. The upper surface of the mound was perforated by
three looters’ holes, and along the northern margin of the At the top of the mound in Sector B is a pit just over 2 m
mound was a trench made by looters. The deepest of the deep resulting from the work of looters. The walls of this
looters’ holes extended more than 3 m into the mound. In it, a hole reveal multiple phases of building, including the exterior
stone wall finished with a coat of plaster could be seen that wall of a lower room and a portion of an upper room and a
had been covered over by a thick layer of construction fill. The series of floors and use surfaces. The walls of this hole were
deepest level of fill included large angular rocks. This material cleared, and stratigraphic profiles were drawn (Table 12;
in turn was covered by a layer of shicra and finally by another Fig. 15).

CREAMER ET AL: FORTALEZA VALLEY, PERU 17

 FIG. 14. Caballete, bundle of willow branches buried at the base
of Test Unit 4.

bags and a new floor placed on top. Continued occupation

FIG. 13. Caballete, rock wall forming edge of platform in Sector
A, Test Unit 4 (Wall 1).
resulted in the accumulation of at least two more floors with
associated remodeling of each floor. This process of construc-
tion and remodeling continued through the end of the third
The upper portion of this looted area revealed clay floors
millennium BC (3000–2000 BC) and seems to have extended
interspersed with construction fill composed of stone from
even further since a few ceramic fragments were recovered
collapsed room walls, fragments of clay mortar, shell, and
from clearing around the edge of the looted area, implying use
botanical materials. These were not well preserved, though the
of the area or visits to this mound after the introduction of
materials indicate a series of clay floors that were built on top ceramics at approximately 1800 BC.
of each other, examples of remodeling and intensive use. A
These phases show only one part of the construction history
date of 1940 Cal BC was obtained from a shicra bag collected of the Sector B mound. Since the looters’ pit did not extend
just below the series of at least three floors that form the upper down through the entire mound to the sterile soil below, there
surface included in Phase II (Table 11h). Above these floors, are undoubtedly additional occupation layers below the floor
material visible in the profile of the looters’ pit consists of (Phase VI) at the base. Although team members wore hardhats
stone, shicra, and clay from collapsed structures, but no and used ladders to enter and exit the unit, additional
pattern could be detected. Layers A and B likely postdate excavation was not undertaken in this locality because the
occupation of this area. dimensions of the hole restricted excavation and its depth
Two additional dates were obtained from the remains of a made further work hazardous.
structure visible at the base of the looted area. Shicra obtained
from the fill of this lower structure dated 2120 Cal BC (Table 11i), Sector B, Test Unit 1—2003
while plant fiber from plaster on the exterior wall of the structure
dated 3070 Cal BC (Table 11j). The latter date has a large The density of broken shicra bags, braided fragments, plant
standard deviation (6290 years) and spans the time from 3900 to fiber, and reed fragments strongly suggests these materials
2300 Cal BC at the 95.4% confidence interval. It is possible that were worked in this area (Fig. 16; Table 13). The concentra-
this date included older materials or was contaminated with other tion of plant fibers is greater than in any other unit
older inclusions in the plaster. It is unlikely that it represents a investigated during 2003, and this locality may be part of a
significantly earlier construction sequence, given the relative workshop for making shicra, woven mats, or other objects
consistency of 10h and 10i. In this case, the Sector B mound may made of plant fiber. The deposits show leveling, preparation
be considered to have had two main phases of construction and and construction of a low platform, followed by two
use, both in the period between the 22nd and 20th centuries BC. subsequent episodes that deposited a thick layer of plant
One lower structure was constructed and used, then subsequently material, indicating lengthy use of this area with plant
filled with rocks in shicra bags, erecting a mound approximately a materials that accumulated a thick layer of debris. At some
meter higher than the buried structure. Subsequently, more floors point, a floor composed of rock and clay was built on top of
were laid across the structure, gradually raising the height of the the plant debris, and activities continued. A distinctive layer of
mound to at least 5 m. yellow-brown soil capped the plant debris, either marking an
Clearing Profile 1 in Sector B revealed a phase of building end to the activities carried out in this area, sealing the area for
with stone and clay mortar with fine plaster finishing the walls. a subsequent use, or marking the start of a subsequent phase.
This structure was built when the surface of the mound was at A very different activity was then carried out in this zone,
the base of the looted structure, some 2 m lower than at including the excavation of a pit that was subsequently filled
present. The room was later filled in using stone-filled shicra with burned debris. This pattern of ash and rock in a pit

18 FIELDIANA: ANTHROPOLOGY
 FIG. 15. Caballete, Sector B, Profile 1 (2003).

feature may indicate remains of a pachamanca or even multiple indicates extensive working of fiber similar to that used in
pachamancas, a method of food preparation using a pit oven shicra bags.
and heated stones. The dated layers of the unit show long-term use as a work area
Dates of 2280 and 2070 Cal BC (Table 11k, l) were for making fiber items, beginning around 2500 Cal BC and
obtained from plant material in the upper portion of the continuing until around 2000 Cal BC. The workshop was
fiber workshop area. Plant material from the lower portion established on a low platform constructed of shicra over clay
of the fiber workshop, below the interim use surface, yielded used to level the area. During its period of use, this workshop
dates of 2450 and 2350 Cal BC (Table 11m, n). The platform area accumulated more than 50 cm of debris that included a
was constructed over a clay base dated 2620 Cal BC range of midden materials but was mostly junco fiber and strands
(Table 11o), while three samples from the clay layer below of woven junco. Although most of the unit was excavated using
the platform dated 2500, 2340 and 2320 Cal BC (Table 11p– arbitrary levels, samples were recovered from clearly identified
r). The cluster of dates indicates an early phase of contexts including the thick deposit of plant and weaving
construction between about 2500 and 2300 BC and then a materials both above and below the clay floor, from the layer of
later use of the area sometime around 2200 BC, when the fill shicra (Level 8), and from the lowest level of clay floors (Level 9).

CREAMER ET AL: FORTALEZA VALLEY, PERU 19

 FIG. 16. Caballete Sector B, Test Unit 1 (2003), west wall.

Sector B, Test Unit 2—2004 surface suggest that this test pit was in an area of domestic
activity. Excavation suggests that occupation in this area was
This 1 3 2-m pit designated Sector B, Test Unit 2, was brief, and no artifacts were recovered that suggest a specific
excavated in 2004. In Figure 9, this test unit is closer to Sector function. Small charcoal samples were recovered from the
C despite its label. This is among the reasons that regular screened materials in this unit, but no in situ radiocarbon
divisions of the site in units of 100 3 100 m and excavation by samples were collected, and no samples were processed from
operations were initiated in subsequent seasons. Test Unit 2 this unit. This possible residential area is undated but appears
was placed in a zone where small fragments of charcoal and similar to other residential areas at this site.
mollusk shell could be observed on the surface. Patches of In the Sector B excavations—one profiled looters’ hole and
dark sediment were present on the surface, suggesting the two test units—we see that construction was under way
possibility of ancient occupation. The deposits consisted of between 2500 and 2300 Cal BC, as indicated by the dates from
layers of different textures and compositions that included the lowest levels of Test Unit 1, the possible shicra workshop.
cultural material, though it was difficult to determine whether Use could have begun even earlier in this area, as indicated by
these were natural or cultural deposits (Table 14). No evidence the date of 3070 Cal BC from the fiber in plaster at the base of
of structures or features was recovered in this test unit Profile 1. More extensive and deeper excavation would be
(Fig. 17). needed to examine the earliest construction of Mound B to
A fragment of a use surface was identified in Layer F, and compare the earliest construction in this sector with that of
below this layer was a shallow pit. However, the arrangement other mounds at Caballete.
of layers of fine sand and clay cutting across the gravel
deposits in the pit appears to be a natural rather than cultural Sector C
deposit. Although the pit may have had a cultural origin, it
appears to have been filled by natural processes. We can Sector C is centered on a rectangular mound on the western
identify the two episodes of activity in this locality, confined to edge of the space encompassing the principal structures at
the small area of use surface and the pit. Artifacts on the Caballete. Attached to the front, the east face of the mound is

20 FIELDIANA: ANTHROPOLOGY
 FIG. 17. Caballete Sector B, Test Unit 2(2004), north wall.

a low rectangular platform with a sunken circular court in the Sector C, Profile 2—2003
center. The base of the mound is 59 3 55 m; its maximum
height is 7 m. The platform around the circular court is 35 3 Profile 2 was located on the north side of the mound, where
33 m, while the circular court itself is 25 m in diameter. a long trench, apparently from digging with heavy machinery,
The present surface of the mound is covered with fine left a large exposed area. The cleared area extended across an
wind-blown sediment and sand, along with medium-sized area 3 3 12 m (Table 15). The heavy machinery used in this
fragments of angular rock and some mollusk shell. Toward area broke through the walls of rooms, retaining walls, trash
the north corner of the upper surface of the mound are three deposits, and construction levels of shicra. There was a large
looters’ holes. In the sides of the three, layers of trash are quantity of shicra in secondary context, along with mollusk
visible, composed of ash, large quantities of mollusk shell, fragments and lithics.
plant remains, and charcoal. Some wall segments are visible, A date of 2400 Cal BC was obtained from a fragment of
and in one of these holes shicra bags were visible. On the shicra bag in Phase 1, while a date of 2320 Cal BC was
upper part of the mound is a depression that outlines a U- obtained from a sample in the adjacent section of the profile
shaped atrium oriented toward the sunken circular court. The (Table 11s, t). These indicate a period of large-scale construc-
mound is connected with two flanking platforms where the tion, as evidenced by the thick layer of fill in Phase I. A date of
bases of angular stone and clay structures are visible. The 2160 Cal BC (Table 11u) was obtained from the base of the
area surrounding the Sector B mound is part of the plow zone cleared area. The sample comes from material that was used to
described above, and the exposed deposits revealed an fill in the area adjacent to Wall 2, the edge of a platform
extensive area where clusters of burned rock, ash, and (Phase III). The date represents a time after the Phase III
mollusk shells are visible on the ground extending from platform, and the use surfaces or platform surfaces were
northeast to southwest of the mound. These clusters appear identified in use (Fig. 18B, C). The position of date ‘‘u’’ may
to reflect extensive use of this area for pachamancas, or show that the mound was expanded outward from the center
roasting pits. and that the portion of the mound including Wall 2 and Layer

CREAMER ET AL: FORTALEZA VALLEY, PERU 21

 FIG. 18. Caballete, Sector C, Profile 2 (2003).

F along with Layers E, C, and B are older than the portion of retention walls. This would explain both the presence of cobbles
the mound between Wall 2 and Wall 1. Subsequent in construction fill and their relative scarcity.
construction included a large retaining wall (Wall 1) and a
substantial phase of stone and adobe construction. Phase 1 Sector C, Test Unit 2—2003
raised the level of the mound a meter or more. The dated
portion of the mound ranges from about 2400 to 2160 Cal BC Test Unit 2 was a 1 3 2-m unit located 50 m northeast of the
and shows two phases of construction. Sector C mound. Fragments of shell and lithics were observed
The base of the area cleared as Profile 2 was well above the on the surface, and this locality was chosen because it
base of the mound, suggesting that initial construction took appeared to be a habitation area (Fig. 19).The upper layers
place earlier than any of the radiocarbon dates. Walls of of this unit (Phase I), where a pit was dug and filled in, may
rooms that were present during Phase II or III of mound use postdate the general occupation of Caballete (Table 16). Prior
could be seen in other sections of the profile that were cleared. to this activity, a structure or platform was built and used in
Construction of the rooms identified predates the radiocarbon this area (Phase II), based on Wall 1 and associated use
samples, which were collected from the material that filled in surfaces, as seen in layers of compact soil, evidence of use
the rooms. The dates indicate that construction of rooms surfaces. Below this wall and use surfaces, there appears to
predated 2100 BC and that occupation of the site continued have been a hiatus in occupation, as indicated by deposits of
after that date, based on the occupation layers above the fill fill consisting of compact, dry soil with few rocks that
from which the samples were collected. The mound itself was contained a variety of cultural materials including shell, bone,
comprised of a series of platforms that were subsequently plant remains, lithics, and charcoal. This may also indicate the
covered by more extensive retaining walls and platforms in the area changed use and was a midden for a period of time after
sequence of construction. occupation of Floor 1. That floor was an intentionally
In the profiles of the mounds in Sectors B and C, a common constructed use surface made of clay embedded with small
construction technique is used, employing river cobbles as tools. stones without a smooth clay plastered surface. A thick layer
Although construction is largely of angular rock, probably of plant material accumulated on top of much of the floor,
quarried nearby, the construction fill in both mounds includes suggesting that this may have been a workshop for woven
river cobbles that display use wear. It appears that cobbles were goods for some period of time. This is similar to the workshop
used to trim chunks of angular rock as they were placed in wall in Sector B, Test Unit 1, with less accumulation of plant
construction, presumably to make a flat exterior face for the remains and perhaps in use for a shorter period of time.
wall. After a brief period of use, the cobble, too, appears to have The chart of calibrated dates and ranges (Table 11)
been thrown into the construction fill and another selected to indicates a long period of utilization of this locality, possibly
continue the process of trimming and positioning rock in the as long as from 3620 to 2120 Cal BC. Although there is

22 FIELDIANA: ANTHROPOLOGY
 FIG. 19. Caballete Sector C, Test Unit 2. The top of a plastered wall or bench is visible below a layer of dark midden in the bottom of TU 2.
A shicra bag filled with small stones is adjacent to the wall.

discrepancy in the stratigraphic relationship among the However, charcoal does not always yield older dates older
different dated samples, the calibration ranges indicate a than annual plant fibers. The weighted average date for
consistent period of use between approximately 2660 and 2120 sample ‘‘v,’’ plant fibers from Level 8, is 2490 Cal BC, while
Cal BC, a likely range of dates for this unit. As mentioned the weighted average date for sample ‘‘w’’ is 2330 Cal BC,
above in the section ‘‘Charcoal, Radiocarbon, and Old though it is charcoal from stratigraphically below sample ‘‘v’’
Wood,’’ when there are paired dates, the date from charcoal in Level 9. This problem highlights the challenges of working
is older. The weighted average date for sample ‘‘x’’ from Level in an ancient and preceramic context where the only dating
11 from annual plants is 2120 Cal BC, while that for sample technique available at present is radiocarbon. While it is unlikely
‘‘y,’’ also from Level 11 but from charcoal, is 2640 Cal BC. that annual plant fibers will yield deceptively ‘‘old’’ dates,
Similarly, the weighted average date for sample ‘‘z’’ from charcoal can come from either ‘‘old’’ or ‘‘new’’ wood and either
Level 12 from annual plants is 2570 Cal BC, while the yield dates that are too old—reflecting the death of the tree, not
weighted average date for sample ‘‘aa,’’ also from Level 12 but the use of the wood—or be from wood that is cut and used at the
from charcoal, is 2960 Cal BC. same time—accurately reflecting the age of the deposit.

CREAMER ET AL: FORTALEZA VALLEY, PERU 23

 shows another variation in construction material. The woven
bag was lined with leaves and filled with small rocks that would
have fallen out of the bag without the lining (Fig. 20b). The fact
that this structure appears to date to at least 2660 BC and
perhaps much earlier, coupled with the deep deposits and its
excellent preservation, makes this a promising place for future
explorations of change over time in the occupation of Caballete.

Sector D

Sector D includes the eastern portion of Caballete. This

sector includes a rectangular mound measuring 120 3 47 m at
the base and 7 m high. The mound is covered with a large
quantity of angular rock and a few fragments of mollusk shell.
The central portion of the mound is a raised platform 38 3
45 m on which alignments of stones are visible that formed the
base of structures and where two looters’ holes have been dug.
One of the holes reveals a wall of angular rock set in clay
mortar that retains some of its original clay plaster and white
paint. Stone alignments are present on the upper surface of the
lateral arms of the mound as well.

Sector D, Profile 3—2003

At the rear of the mound on its eastern face, a large

elongated area has been excavated by looters. The profiles
reveal a series of layers of construction fill, including medium
and large angular rocks and shicra bags (Fig. 21; Table 17).
These layers of fill alternate with clay floors, some carefully
prepared. The uppermost section of floor was beside Wall 1.
Plant fiber from the wall plaster yielded a date of 1620 Cal
BC (Table 11ee), while a sample of plant fiber from above
Wall 1 yielded a date of 2150 Cal BC (Table 11ff). These
samples suggest use of the upper part of this structure from
end of the third millennium and beginning of the second
millennium BC.
Below Wall 1, collapsed walls and construction fill were
layered with floor fragments. Shicra fragments were numerous,
and shicra construction appeared to continue below the exposed
FIG. 20a. Caballete Sector C, Test Unit 2, north profile, showing
midden deposits and plastered upper surface of bench. area of the mound. A fragment of a shicra bag from Layer F
FIG. 20b. Shicra bag lined with leaves and filled with small stones dated 2390 Cal BC (Table 11gg). Together, these samples
above plastered bench, Caballete, Sector C, Test Unit 2. suggest the Sector D mound was being built by 2400 Cal BC and
was in use until sometime between 2100 and 1600 Cal BC. The
Dates from Sector C Test Unit 2 (Table 16), range from floor fragments identified suggest that there were a number of
2960 to 2120 Cal BC. Sample ‘‘dd,’’ charcoal from a floor that different construction phases in construction of the Sector D
was associated with the Wall 2 bench, yielded a weighted mound. A massive construction phase using shicra may
average date of 3620 Cal BC. The dates from charcoal samples comprise the lower portion of the mound, where the rocks were
may yield systematically earlier dates than annual plant fiber large and shicra fragments numerous. Most of this material was
(Table 3; Table 16), but this would need further testing. The not exposed and remains to be examined. Evidence of specific
few pairs of samples including annual plant and charcoal activities other than construction was not apparent in this area,
taken from a single provenience are too few in number to tell as only small fragments of floor were uncovered.
whether this is a consistent difference or a largely random
difference based on the use of culturally old wood as a result Sector D, Test Unit 3—2004
of desert preservation. At the same time, a calibrated date of
3620 Cal BC is within the overall range of dates for Caballete. Test Unit 3, a 1 3 2-m test unit, was located in Sector D at
The goal of recovering dated materials from habitation areas the foot of the lateral platform south of the mound
was met in that the Phase V midden produced a variety of food (Table 18; Fig. 22). The platform is outlined by large rocks,
remains even though this area did not produce any habitation the remains of a retaining wall that defined this face of the
features apart from floors and use surfaces. The plastered bench platform. The surface does not show evidence of monumen-
at the base of the unit was not cleared extensively enough to tal architecture, only some mollusk fragments. For this
determine whether it was part of a habitation area or some reason, it was decided to locate the pit here as part of our
administrative feature (Fig. 20a). The shicra bag found adjacent effort to identify domestic activity areas that could date to
to the bench and presumably part of the fill over this feature the Late Archaic Period. Because natural stratigraphy could

24 FIELDIANA: ANTHROPOLOGY
 FIG. 21. Caballete, Sector D, Profile 3 (2003).

be detected, this test unit was excavated in natural layers radiocarbon dates suggest that this is an Initial Period context.
instead of arbitrary levels. However, the use of flat weave does not entirely rule out a
Test Unit 3 included deposits of alluvial gravel under thin Late Archaic date (Doyon-Bernard, 1990, p. 71). Dating of the
layers resulting from human activities. These were associated cloth fragment would be required to follow up on this
with shallow pits and a packed use surface. Unlike some other possibility. At this point, the infant burial may be evidence of
units, no constructed floors, hearths, or specific activity areas a relatively late phase of occupation at Caballete or of use of
were identified, suggesting limited use of this locality. Despite the site after it was abandoned as a community. However,
this general result, a small group of stones was uncovered in the interments made after occupation of the site ended tend to be
center of Layer 5, associated with a fragment of cloth. Buried clustered, such as in Sector F, where most of the sector
20 cm below the surface of Layer E, this proved to be the consists of looted burials interspersed with small architectural
interment of an infant wrapped in cloth in a seated position and units, leaving the temporal position of the infant interment
tied in cord (Fig. 23). The cloth fragment recovered was flat ambiguous. Further discussion and a detailed description of
weave in a two-over/two-under pattern. No pit outline was the burial are available in Bazan (2012).
detected, suggesting the burial may predate the compacted
surface and pit features recorded. This infant burial was not Sector E
associated with any structure or feature and therefore does not
appear to be dedicatory. Sector E is at the southern end of the site, near the edge of
Test Unit 3 provides evidence of occupation on three an alluvial fan within the quebrada, the dry wash, in which the
different occasions that consists of layers of compact soil entire site is constructed. This sector includes two rectangular
forming use surfaces. In the surface layer and the uppermost mounds. The larger mound measures 69 3 53 3 7 m. The
feature in Layer C, some fragments of plain undecorated smaller mound is 56 m to the northeast and measures 53 3 37
ceramics were recovered. At some point, an infant was buried 3 5 m.
below the use surfaces. Neither the surface that was in use The summit of the larger mound clearly displays a
when the infant was buried nor the outline of a burial pit could rectangular sunken atrium flanked by two raised platforms,
be identified. Loom-woven or flat weave cloth is rare during each of which is also covered by a large number of medium-
the Late Archaic Period, and paired threads, as in the sample sized blocks of angular rock that appear to be the remains of
around the burial, are not known from Late Archaic sites. collapsed structures built on the mound. In addition, a scatter
Bird et al. (1985, pp. 53, 191) mention 2 3 2 weaving from the of mollusk shell and a few fragments of heavily eroded ceramics
upper layers of his Test Unit 3 at Huaca Prieta, but the are present. The surface on the plaza side of the mound has a

CREAMER ET AL: FORTALEZA VALLEY, PERU 25

 FIG. 22. Caballete, Sector D, Test Unit 3 (2004).

slight depression that the excavation of Trench 1 suggests is a the edge of the circular feature seen on aerial photographs
sunken circular court that has been covered by alluvial soil. (Table 19).
Excavation in this area (Fig. 24) yielded sloping contours
Sector E, Trench 1—2003 demonstrating a clear depression, as would be expected for a
sunken circular court. No architectural elements, such as
The Sector E mound was constructed of angular rock and stairs, walls, or floors, were found that might confirm the
fits the pattern of preceramic architecture. At the summit of presence and form of a circular court. However, the
the mound, an open rectangular atrium bounded by low walls stratigraphy of the unit does suggest the presence of a sunken
approximately 10 3 10 m is visible. This feature was feature in this area that was excavated more than a meter
connected to the area under investigation at the foot of the below the original ground surface. It would be necessary to
mound by a broad staircase, now collapsed. Beyond the enlarge the trench to confirm the size and configuration of this
staircase, a large rectangular court, 95 3 50 m, is surrounded sunken feature. Excavation of Trench 1 suggests the existence
by a low, thick stone wall. Bounded on one side by the of a sunken circular court at the foot of the mound in Sector
mound in Sector A and on the other by a gully formed by E. An event of flooding may have filled in most of the sunken
flash floods, the level area at the foot of Mound E was court and subsequent erosion and deposition filled the area
thought to hold a circular court that had been filled in by soil even further (Layer A). A charcoal sample was dated to 3120
that on rare occasions washes down the stream bed. Aerial Cal BC, while mixed plant fibers yielded a date of 2520 Cal BC
photographs show a circular feature inside this rectangular (Table 11hh, ii), suggesting that this circular court was
court and centered on the atrium of the mound, although on constructed between 3000 and 2500 Cal BC. Although the
the surface this feature has been obscured by flooding and court may have been a simple circular depression, excavated
alluvial fill. Short rows of large rocks mark the corners of a into a low platform and without slabs of stone lining the
low rectangular platform that extended toward the site center interior, the stratigraphy indicates its shape. This kind of
from the base of the Sector E mound. Such low platforms simple circular court was also found at the site of Cerro
often surround sunken courts, and a test trench 1 3 5 m was Lampay, immediately across the Fortaleza Valley from
excavated in this locality, using the southwest corner of the Caballete (Vega-Centeno, 2005). The presence of a circular
unit as a datum point. Trench 1 was excavated on the east court in this location is also indicated by a clear circular
side of the rectangular court in a location that would transect outline immediately in front of the mound atrium evident on

26 FIELDIANA: ANTHROPOLOGY
 The excavation, though relatively shallow, yielded evidence
of multiple strata, yet none of these provided substantial
cultural remains. Test Unit 5 in Sector E did not yield the
expected results. Despite an adjacent looters’ pit indicating
presence of deep cultural deposits, Test Unit 5 was relatively
shallow. There was evidence of the use of this area for trash
disposal, as indicated by the three pits that were identified
during excavation, one of which was a natural irregularity in
the surface filled with debris. Features 1 and 3, however, have
steep sides and flat bases, though neither was distinguished
during excavation. Two periods of extended use are visible,
including the compact soil Layer B and the features associated
with Layers C and D. Samples for radiocarbon dating were
obtained from this unit, but most were charcoal. A sample of
annual plant fiber was collected during screening and was not
in situ, and no samples from this locality have been processed.
The presence of midden without ceramics, even though
undated, suggests that this locality dates to the Late Archaic
Period, when Caballete was occupied.

Sector F

This sector includes the southwest part of the site, adjacent

to Sectors A and E, between these two sectors and the thick
alluvial deposits that form the base of the La Empedrada wash
where it empties into the valley. This sector does not include
any mounds and is bounded by a wide wall of angular rock set
in clay mortar that averages 1 m wide. The wall runs in a
sinuous curve toward the west, along the edge of the alluvial
deposits. The entire quebrada is composed of thick alluvial
deposits laid down by many years of periodic El Niño episodes
flooding the bed of the wash.
FIG. 23. Caballete, Sector D, Test Unit 3 (2004), photo of infant Looters have heavily disturbed this part of the site, and as a
burial. result there is a significant quantity of human bone on the
surface along with remains of textiles, ceramic, and gourd
bowl fragments that appear to have come from the looted
an aerial photograph taken in the 1940s (Fig. 25). Including graves. This is similar to the condition of the site in the
this feature with the others at the site, Caballete has three northern part of Sector E. However, some looters’ holes have
sunken circular courts, another indication of its important role revealed exposed profiles where wall segments of rock and clay
in the region. Further excavation could clarify the dimensions mortar were plastered with a coat of clay mixed with plant
and nature of this feature. fiber. In some cases, these wall segments are associated with
use surfaces, ash, and trash deposits that suggest activities
Sector E, Test Unit 5—2004 associated with residential occupation.

This test unit was located in an apparent residential area, as Sector F, Test Unit 1
identified by the deposits visible in a profile exposed by
looters. The exposed profile displayed a substantial cultural This 1 3 2-m pit was located just south of the main mound
deposit more than a meter below the surface. A few meters in this sector and within an area where the remains of
south of this profile was an intensively looted cemetery area, rectangular stone structures are visible on the surface
though from the materials on the surface, the cemetery dates (Fig. 27). Excavation revealed a use surface, Layer E,
to a much later period than the main occupation of Caballete. probably associated with activities when Caballete was
Evidence for residential occupation in this area was limited, occupied during the Late Archaic. No dates were obtained
and the deposits were not as deep as they appeared in the from this layer, and it may date to the use of Caballete in the
nearby looters’ pit (Table 20; Fig. 26). A layer of compact soil centuries after its abandonment when the area was used as a
that forms Phase I did not necessarily correspond to a use cemetery. Several different cemetery areas are present at the
surface, as the soil appeared to be compacted by exposure to site, and the remains excavated in Test Unit 1 included two
surface moisture rather than foot traffic. Feature 2, a shallow interments along with midden and construction debris or
pit, held midden remains that did not provide distinct evidence rubble from structures. Other activities were not identified in
of function, nor did the larger and more steep-sided pit, this area.
Feature 3. Evidence of these pit features in the north end of Two individual interments were revealed in Test Unit 1. The
the unit and another in the west side of the unit became visible uppermost of these was an infant, first recognized by cranial
as the profiles were drawn. Cultural materials seem to have fragments associated with textile fragments. When excavated,
come from the pits and the associated Layer D. it was found to be largely intact, an infant or very young

CREAMER ET AL: FORTALEZA VALLEY, PERU 27

 FIG. 24. Caballete, Sector E, Trench 1 (2004).

juvenile lying on its right side in a lightly flexed position. The plant fiber associated with the infant burial was dated 940 Cal
infant appears to have lain on a bed of fibrous plant material, BC (Table 11jj).
wrapped with textiles. Traces of red pigment could be Below the infant was an interment that proved to be an
observed on some textile fragments. No other artifacts were adult female (Fig. 28). The burial was first identified by a
collected with the infant burial (Bazan, 2012). A sample of fragment of wicker basketry that proved to be positioned over
the head. While both individuals were oriented east to west,
the infant’s head was to the east and the adult’s to the west.
The adult was wrapped in textiles, though these were badly
deteriorated, and was seated on or wrapped in a woven mat of
wild cane or similar material. More than 20 fragments of the
textiles and burial wrapping were recovered, although only
two twined fragments are more than ten cm long. The woven
materials include fragments of twined and flat weave fabric,
string, and cane matting. One sample was identified as a ‘‘fiber
mask,’’ though it could not clearly be identified as a mask.
This fragment appeared to be a layer of coarser, more loosely
woven fabric over a more tightly woven flat weave textile. The
top layer is dark in color and mostly deteriorated, while the
flat weave is a lighter color and better preserved.
Three gourd bowls were uncovered by the feet of the
individual, and one additional bowl was beside the head. On
the south side of the individual near the large rocks mentioned
above were other burial goods on reed matting. These included
obsidian blades, shell beads, three small spindle whorls (one
incised with a face on three sides), a mollusk valve, and
implements of bone and wood that appear to be associated with
weaving. Excavation around the head revealed a necklace of 52
disk and tubular pyrite beads or bead fragments. Sizes vary
FIG. 25. Historic air photo of possible circular court in Sector E. somewhat, but each bead measures up to 0.5 cm long and is

28 FIELDIANA: ANTHROPOLOGY
 FIG. 26. Caballete, Sector E, Test Unit 5 (2004).

0.5 cm in diameter with a perforation 0.2 cm in diameter. One were recovered. Initially, the interments were suggested as
hollow gold bead was found, measuring approximately 0.8 3 dating to the Late Archaic based on their association with
0.5 cm. Several of the beads were impressed into hardened earth textile fragments, woven bags, the gourd bowls associated with
around the cranium, indicating that the beads were part of a the adult and the basketry fragment covering the head of the
necklace (Table 21). A radiocarbon date from plant fiber adult. Complete excavation revealed the presence of metal
associated with the Feature 2 burial dated 680 Cal BC beads and obsidian blades, and the radiocarbon dates
(Table 11kk). The dates associated with infant and adult are confirmed the later date of both burials.
not coeval. The period of greatest probability associated with
Feature 2 (adult) is 896–507 Cal BC (98.3%), while the range of Caballete Discussion
probability associated with Feature 1 (infant) is 1270–842 Cal
BC (99.3%). Thus, the two individuals could be from the same Caballete yielded a variety of data during 2003 and 2004
burial episode around the middle of the ninth century BC. that show the antiquity of the site, which may have been
Examination of the pit profiles suggests that the Feature 2 under construction as early as the fourth millennium BC
burial shared the same pit as the infant in Feature 1. Although (Table 11). Four dates fall before 2900 Cal BC. Five of the
initially the burials appeared to date to the Late Archaic six sectors tested yielded dates prior to 2000 Cal BC. Unlike
because of the associated gourd bowls and woven materials, the Porvenir, where occupation of the site appears to fall into
presence of metal ornaments indicates a later date than is two distinct periods, Caballete seems to have been occupied
supported by the radiocarbon results. A detailed discussion of steadily throughout the third millennium BC. Most dates
the burials can be found in Bazan (2012). cluster between 2700 and 1900 Cal BC, with only one Initial
There is precedent for use of Caballete and other Late Period date (Table 11ee). Two later Early Horizon dates, 940
Archaic sites as burial places in the centuries following their and 680 Cal BC, come from what proved to be intrusive
abandonment (Haas & Creamer, in press). Several areas of burials despite the presence of gourd bowls and basketry as
later interments are visible at Caballete, including a portion of grave goods. The site was employed as a burial area for
Sector F. Thus, these individuals excavated from Sector F are centuries after it no longer held a permanent occupation.
probably not associated with the Late Archaic occupation. Radiocarbon dates do not indicate that sectors were
Both date to the first millennium BC, long after the occupied at different times. Dates from Sectors A–D were
monumental mound complex had ceased to function as a most numerous in the period 2600–2100 Cal BC. Sector E
ceremonial center. yielded earlier dates and Sector F later dates. From this, it
Test Unit 1 in Sector F was excavated to examine this appears that many more dates from each sector of the site
portion of the site for remains of residential structures. None will be needed to determine whether these structures were
were identified, though two burials, an adult and an infant, built simultaneously or in sequence. Each pattern has

CREAMER ET AL: FORTALEZA VALLEY, PERU 29

 FIG. 27. Caballete, Sector F, Test Unit 1 (2004), south face.

implications for political organization of the site and the Again, floors were recorded on the upper surface of the
region, but many more dates taken from more individual platform.
phases and construction elements are needed to examine the Apart from the evidence of weaving/shicra workshops,
sequence of construction across Caballete. Any large functions that could be identified were limited by the extent of
complex site from this early time period is going to need excavation and clearing. Elaborate construction is visible at
literally hundreds of radiocarbon dates to assess internal Caballete in two places: the finely plastered bench at the
chronology, construction sequences, and any reasonable bottom of Sector C, Test Unit 2, and the structure with
estimate of site demography. More explicit comparison of plastered walls at the bottom of Sector B, Profile 1. The
charcoal and different kinds of annual plant fiber is needed structure in Profile 1 was not excavated, and though the room
to resolve issues of old wood and reuse of materials. was clearly identified, its purpose was not. The plastered
Testing in 2004 confirmed the presence of domestic and bench is the same. It could be part of a habitation complex,
ritual architecture at Caballete. As at Porvenir, the most perhaps even one of high status. Since this feature was covered
frequently identified architectural features were floors and over without being torn down, it may also be part of a
platforms that varied in size and in finish. The platform in ceremonial area that was covered after use (Vega-Centeno,
Sector A, Test Unit 4, consists of a low retaining wall with 2010) or transformed into a larger structure (Creamer & Haas,
several floors constructed alongside. Platforms in Sector B, 2008). Plastered benches are characteristic of mito temples
Test Unit 1, and Sector C, Test Unit 2, were the base for (Izumi & Terada, 1972) that may have antecedents in the
weaving or shicra workshops. The platform at the base of Norte Chico region (Haas & Creamer, 2010).
Sector C, Profile 2, is similar in construction, with a floor Perhaps more than any other site in the region, the
surface on top. A subsequent and much larger platform excavations at Caballete demonstrated that there is an
was built over the platform at the base of the profile, with abundance of subsurface architecture and occupation that
Wall 1 at one side and a thick rock construction phase. cannot be detected on the surface. Remote sensing technology,

30 FIELDIANA: ANTHROPOLOGY
 FIG. 28. Caballete, Sector D, adult burial below infant in Test Unit 3.

such as ground-penetrating radar, might be effective in rich in botanical remains. Despite this, looting is less intensive
locating such subsurface structures. at Caballete than at other, later sites because of the absence of
ceramics, metal, and elaborate textiles that are often the focus
Conservation and Protection of Caballete of illegal traffic in antiquities.
The greatest danger presently comes from suppliers of
Despite damage caused by an unsuccessful attempt at construction materials who view mounds as suitable quarry
irrigating the Quebrada La Empedrada where Caballete is areas. Mounds are targeted for quarry areas because the rock
located, the site is one of the best-preserved complexes of is already broken and needs only to be loaded and carried
monumental architecture of the Late Archaic Period in the away. On at least two occasions during our 2003 field season,
Norte Chico. This is largely due to its monumentality, its local residents alerted us to the presence of such persons, and
isolation from the main roads such as the highway to Huaraz, we were able to protect the archaeological area from damage,
and its location on the right margin of the valley, where no such as the Sector D looting many years ago.
bridges cross the Fortaleza River. Despite its isolation, there
are two principal threats to Caballete: looters and gravel
quarrying.
The activity of looters has severely damaged the Sector C Cerro Blanco 1
mound at Caballete, where looters have dug a hole more than
2 m deep, destroying monumental architecture. Similar The area around the modern settlement of Cerro Blanco is
damage has occurred in Sector F, where, during the 2004 an extensive archaeological zone that has been intensively
field season, recent looters’ holes disturbed stratified deposits occupied since at least the Late Archaic Period. This long

CREAMER ET AL: FORTALEZA VALLEY, PERU 31

 FIG. 29. Photo of Cerro Blanco 1, looking west over the site.

history has resulted in a large expanse of overlapping Sector A

occupations. Because there are numerous cemeteries in the
area, it has also been subject to extensive looting. Even the Sector A includes two of the mounds on the south side of
areas of occupation that are of interest to the present project, the site (Fig. 30). The mound in the center is rectangular, 34 3
completely lacking in pottery or burials, have been affected by 33 3 5 m, covered by loose earth and small to medium-sized
the search for antiquities. The areas that remain in the best angular rock (Table 2). Some mollusk shells and a very small
condition are those that are covered with substantial quantity of unpainted, nondiagnostic ceramics are present.
architecture since these are slower to be looted or leveled. Two looters’ holes are present on this mound. In the first of
Cerro Blanco 1 is a group of mounds located 1.5 km these, located on the upper part of the mound, a wall of
northeast of the settlement of Cerro Blanco, 110 m above sea angular rock and clay mortar is visible within the looters’ pit.
level, and approximately 10.5 km from the coast. The site is up Parts of the wall retain a coat of plaster. Some mollusk shells
against the slopes of the chain of hills that define the left are also visible. The second area of looting is a trench that cuts
(south) margin of the valley (Fig. 29). Cerro Blanco 1 was first into the mound from the side away from the modern road. In
recorded by Vega-Centeno (Vega Centeno et al., 1998). this area, a wall of large angular rock and clay mortar, also
Construction characteristics, such as the presence of shicra with vestiges of surface plaster is present, along with shicra
in construction fill and scarcity of pottery on the surface or in construction fill using large angular rock. Looting has left the
exposed profiles, suggested that Cerro Blanco 1 dated to the mound in poor condition. Radiocarbon dates ranged from the
Late Archaic and into the Initial Period. Sectors were Late Archaic into the Initial Period (Table 22).
established around mounds A–C, all oriented similarly toward
the northwest, facing out over the valley. Sector A South Sector A, Profile 1
includes an area of artifact scatter that may be a low mound.
Looting in this sector has blurred the outline of the structure, Profile 1 was located on the south side of the central
and a flash flood washed out the area between it and Sector A. mound in Sector A in an area of extensive looting that left a
Further excavation would be needed to determine its shape. trench 5 m long and 2 m deep, exposing several layers of
An irrigation canal runs roughly parallel to Sectors B, A, and construction fill. Among the materials visible were shicra,
A South about 50 m to the northwest, and a road parallels the stone, clay mortar walls, and clay floors. The cleared profile
canal beyond it. included 29 different layers and construction elements

32 FIELDIANA: ANTHROPOLOGY
 FIG. 30. Contour Map of Cerro Blanco 1.

(Table 23; Fig. 31a). The stratigraphy is more difficult to of the mound. The seismic event postdates 1360–1330 Cal BC,
understand than elsewhere because it appears that new layers dates obtained from shicra from Layers F and L (Table 22b, c;
were added to the outside of the mound as well as to the Table 23b, c).
center of the structure. This created some horizontal as well The area was reoccupied though possibly after a hiatus in
as vertical stratigraphy. occupation indicated by the aeolian deposits in Layer N. A new
Two rooms and a rubble filled wall are the principal room was constructed, remodeled by the addition of a wall, and
structures revealed in this profile. Building these was a major then replastered, creating a series of layers of floor and wall
construction effort and changed the entire look of the mound plaster at the top of the mound. A sample of plant fiber from
in Sector A (Fig. 31b). Prior to this, the surface of this area rocky fill above the uppermost floor dated 1690 Cal BC
was a low platform and more than 2 m lower than the final (Table 23a; Table 24a) and may be older fill deposits excavated
height of this structure. During Phase III (Table 23), the and reused. However, the full range of calibrated dates between
double wall and both rooms were completed. Later, during these three samples do overlap at the end of the 15th century BC.
Phase II, both rooms were filled with shicra (Layers F, L), Initial occupation of this area probably dates before 1600
raising the level of the mound substantially. Cal BC, though we cannot be sure when mound construction
The stratigraphy in this profile shows the collapse of a huge, began. When an earthquake damaged the massive structure
double-faced wall. The position of the walls and floors are that comprises the bulk of the mound sometime after 1360 Cal
evidence of a natural disaster, probably an earthquake. At a BC, the mound appears to have been abandoned, although a
single moment during Phase II, the rubble filled double walls, later occupation was established on top of the mound. The
approximately 2 m thick, collapsed to one side (Fig. 32). The duration of that later occupation is indicated by several
dates from shicra used to fill the rooms on either side of the episodes of floor remodeling, but it cannot be determined more
double wall show that it was in place after 1400 Cal BC precisely with the dates we now have for Sector A. Although the
(Table 22b, c). The profile also shows that the spaces on either use of this area does fall squarely in the Initial Period, the
side of the wall had been filled prior to the event that knocked evidence for earthquake damage and subsequent reoccupation
it over because the walls lean on the shicra deposits (Fig. 31b, with a smaller construction is similar to that noted at Late
Layers A–L) that were used to fill the rooms and raise the level Archaic sites in the nearby Supe Valley (Sandweiss et al., 2009).

CREAMER ET AL: FORTALEZA VALLEY, PERU 33

 FIG. 31. (a) Cerro Blanco 1, Sector A, Profile 1, drawing showing excavated deposits. (b) Cerro Blanco 1, Sector A, Profile 1, phases.

34 FIELDIANA: ANTHROPOLOGY
 FIG. 32. Photo of Cerro Blanco 1, Sector A, Profile 1, showing wall collapsed during an earthquake.

Sector A, Profile 2 margins are at nearly right angles and measure 32 and 20 m,
but the rest of the mound has no clearly defined outline. It is
Profile 2 was located at the top of the central mound in possible that this was a rectangular mound and that half of
Sector A. This was another area that had been exposed by the exterior has been destroyed by looting and mining of
looters, leaving a hole 1.5 m in diameter. In the north and material for road building in the area in addition to the
south walls of this hole, it was possible to see the techniques washout already noted. The mound area is about 3 m high
and process of construction (Table 24; Fig 33). The surface of and is oriented 280u (northwest). This sector is covered by
the mound was covered by debris from looting piled around wind-borne soil and angular rock, with a few fragments of
the mouth of the hole, and below this stratigraphic units were ceramics on the surface. Four looters’ holes have been
identified in the profile of the looted pit. excavated in this mound. In these holes were fragments of
The cleared area revealed three episodes of use. The oldest nondiagnostic ceramics and trash with a large quantity of
of these visible, Phase III, consisted of a substantial layer of mollusk shells.
shicra. A sample from the layer of shicra construction yielded The looters’ holes were numbered 1–4, with holes 2 and 3
a date of 1190 Cal BC (Table 24d). Gravel was added (Phase yielding the most information. In hole 2, a wall of large
II) over the shicra, and use surfaces of clay were laid down angular rock without plastered surfaces was visible. In hole 3,
above that layer (Phase I). Thus, Phase II may indicate a break shicra fill was visible along with a wall of medium-sized
in occupation or may be a leveling of the surface for
angular rock in clay mortar with evidence of surface plaster.
construction. Walls 1 and 2, built of rock and clay mortar
Associated with the plastered surface were two floors, and
on the upper use surface, are grouped with the floors as Phase
below the floors was a thick layer of trash with a large
I, as the walls and floor surfaces are likely to be related
number of mollusk fragments and a few fragments of
construction. However, there was no surviving floor or wall
nondiagnostic ceramics. Mound 3 is in a poor state of
plaster that securely demonstrated their connection.
preservation due to the number of looters’ holes that have
Clearing the Sector A looters’ pit, Profile 2, provided
another glimpse of the interior of the Sector A mound. The been made in the mound. A radiocarbon date of 1190 Cal BC
floors and wall segments in this area are stratigraphically (Table 22e) was obtained from a shicra sample in looters’
comparable to the uppermost deposits in Profile 1 (Floors O– hole 2. The date is consistent with the others from Sector A,
S) since in both locations deposits of shicra are covered with a also within the Initial Period. The date suggests that these
layer of gravel and then clay floors and walls. The date of 1190 small mounds were built much later than other sites in the
Cal BC from Layer D comes from a shicra deposit that is region with a multimound layout. In the Norte Chico, Initial
probably comparable in age to the sample from Profile 1 that Period sites are both fewer in number and not as large as Late
yielded a date of 1330 Cal BC (Table 22c). These dates overlap Archaic sites with monumental architecture. In other areas,
in the 1s range, indicating that they may represent contem- Initial Period sites are far larger than anything previously
poraneous deposits. recorded, including the largest sites of the Late Archaic. This
suggests that the balance of power was shifting away from the
Sector A South Norte Chico after the Late Archaic, though the pattern of
mound construction did not disappear entirely, as we see by
This sector is a low irregular-shaped area on the south side the presence of these smaller mounds with an Initial Period
of Sector A (Fig. 29) that may have been a mound. The east date.

CREAMER ET AL: FORTALEZA VALLEY, PERU 35

 FIG. 33. Cerro Blanco 1, Sector A, Profile 2.

Sector B In the process, a variety of cultural material was collected,

including ceramics, lithics, mollusks, textile fragments, cane
Sector B includes the north end of the site and Mound 1, a fragments, and shicra in its original context. Two phases of use
rectangular mound 29 3 20 m with a maximum height of 5 m are visible: an upper floor created over a layer of gravel and a
and an orientation of 320u. The mound is covered by soil and series of floors at a lower level that may include the corner of a
small to medium-sized angular rock. Some mollusk fragments room. The plastered surface turned upward, and although wall
are present, along with a small quantity of ceramics on the material was not identified, this may indicate the connection
lower slopes of the mound. No diagnostic sherds were found between a wall and floor.
during our visit in 2003. The upper part of the mound has been A layer of rock and shicra below the floors recorded
subject to looting. represents an episode of construction, yielding a date of 1340
Cal BC (Table 22f). The looters’ pit did not extend to the base
Sector B, Profile 3 of the mound, so it is not known whether this was the first
phase of mound construction or whether there are earlier
Profile 3 was located in the Sector B mound where we took occupation surfaces and construction phases below this one.
advantage of a looters’ hole, clearing it to record the The construction methods in the Sector B mound are similar
stratigraphy visible in the walls (Fig. 34). Looting had left to those in Sector A, and the single date from this mound falls
exposed stratigraphy that permitted us to record several within the range of those in Sector A. Since the two mounds
construction phases (Table 25). An area 2 3 2.3 m was cleared. are similar in size and placement, they are likely to have been

36 FIELDIANA: ANTHROPOLOGY
 FIG. 34. Cerro Blanco 1, Sector B, Profile 3.

built at approximately the same time during the Initial Period, rock. A few mollusk shells and a few nondiagnostic ceramic
around 1340 Cal BC (Table 22f). fragments were also present. This mound is in a good state of
preservation because looting has not affected it as extensively
Sector C as the other mounds at Cerro Blanco 1. As a result, there was
no visible damaged area that could be used for excavation. A
Sector C is located to the east of Mounds 1–3, separated single test unit was excavated in front of the mound in an
from them by about 60 m, slightly farther up the slope of the area of surface trash.
chain of hills that borders the valley. This is a rectangular
mound 73 3 42 3 5 m, oriented 340u. Construction of the Sector C, Test Unit 1
Sector C mound took advantage of the slope of the hill to
achieve its total overall height. Thus, the labor invested here This unit is located 10 m from the front terrace of the
was substantially less than that used to construct some of the mound, between Sectors B and C. Patches of dark soil and
mounds at other sites that did not use existing hills to boost small shell and lithic fragments were observed on the
their height. On the side of the Sector C mound toward the surface, possibly indicating a habitation zone (Fig. 35).
valley are two terraces in a steplike layout that follow the The excavation revealed a series of irregular deposits of
slope of the hill. Both the mound and the terraces are construction debris or building materials along with fill
covered by soil and some small and medium-sized angular containing a mix of shell, lithics, plant remains, textile

CREAMER ET AL: FORTALEZA VALLEY, PERU 37

 FIG. 35. Cerro Blanco 1, Sector C, Test Unit 1.

fragments, and charcoal in a matrix of ash, rock, and soil later in the Initial Period. Three dates from Sectors A and C
(Table 26). The area seems to have been a midden although range from 1950 to 1690 Cal BC, while five additional dates
not consistent in composition. The shape of the excavated from Sectors A and B range from 1360 to 1190 Cal BC, time
layers suggests that the midden was discarded in a ranges that do not overlap statistically (Table 26). The
depression or a hole completely filled by the end of Phase arrangement of structures at the site is not at all clear since
II. Phase I represents a subsequent episode of use of this the entire zone has been looted for some time. Sector A South
area as a midden. Two samples of plant fibers from Level 3 may have been a mound that has been looted. The Sector C
within the midden yielded dates of 1950 and 1730 Cal B.C mound may be of a different date as well, as no samples for
(Table 22g, h). dating could be obtained. Although Cerro Blanco 1 does not
The excavated levels revealed indirect evidence of occupa- include a circular court (a hallmark of Late Archaic sites), the
tion somewhat earlier than the other sectors of Cerro Blanco rectangular terrace in Sector C holds the position of a circular
1. Occupation may have been nearby, and the midden contents court in a Late Archaic site (Fig. 30). The stepped
included a range of materials associated with habitation, arrangement of terraces against the hillside in Sector C
including subsistence remains, textiles, and a small number of creating the appearance of a mound, along with its central
unslipped and dark red slipped nondiagnostic ceramics. The walled rectangular court, is a layout known from other Late
shallow deposits and lack of architecture suggest that this area Archaic and Initial Period architecture (see the section
was utilized for trash disposal but did not have any other long- ‘‘Cerro Blanco 2’’ below), including Huaynuna, where a
term use. The radiocarbon dates from Level 3 of the mound was built against a hillside (Pozorski & Pozorski,
excavation date to the Terminal Archaic Period or the early 1987). At Tortugas, a nonresidential structure was built
Initial Period, the latter date suggested by the presence of ‘‘taking advantage of the rocky outcrop to enhance the
ceramics in each excavated level. structure’s height’’ (Pozorski & Pozorski, 1987:47), while
Salinas de Chao consists of a series of terraces transforming a
Cerro Blanco 1 Discussion hillside into a mound, along with a sunken court and
rectangular plaza (Moseley, 1992: 111).
Occupation of Cerro Blanco 1 includes two phases: the Sector A South, Sector A, and Sector B may or may not be
transition from the Late Archaic to the Initial Period and related to the Sector C mound. Each of these structures was

38 FIELDIANA: ANTHROPOLOGY
 FIG. 36. Cerro Blanco 2 contour map showing plan of site and excavated and sampled units.

built using shicra bags, a common construction technique Cerro Blanco 2

during the Late Archaic. However, the later dates for these
mounds and the presence of a few ceramics indicate an Initial Cerro Blanco 2 displays characteristics of the Late Archaic
Period occupation as well. The available dates from Cerro Period, including a mound associated with a sunken circular
Blanco 1 indicate that the locality was in use during the court (Fig. 36; Fig. 37). The circular court was used as a
transition between the Late Archaic and the Initial Period cemetery by later occupants in the area after the mound was
and make it an excellent laboratory for studying this no longer in use (Haas & Creamer, in press). These burials and
important period of transition. The extensive profile of the subsequent looting of the burials in historic times have
looted area in Sector A revealed both earthquake damage destroyed more than two-thirds of the circular court. In
and reoccupation at Cerro Blanco during the Initial Period. addition, a rough vehicle and cattle track approximately 3 m
The reoccupation of Cerro Blanco shows continuity in wide crosses the margin of the main mound, damaging part of
occupation in the region, although Initial Period structures the structure. The track appears to have formed during a
were smaller than the preceding Late Archaic mounds even period when rock was removed from the mound, probably for
before the earthquake recorded at Cerro Blanco 1. use in construction. The entire zone between the main mound

CREAMER ET AL: FORTALEZA VALLEY, PERU 39

 FIG. 37. Photo of Cerro Blanco 2 showing mound and circular court from the north.

and the agricultural fields some 100 m distant was used as a this part of the site began before 2100 and ended at some point
cemetery in ancient times, and all this area has been heavily in the Initial Period after 1600 BC, based on the absence of
damaged by looting. walls or floors higher in the profile than Wall 3. Trench 1 is a
At Cerro Blanco 2, a trench 3 3 12 m was cleared. Trench 1 small sample of the entire site, and there is the possibility that
was located on the south side of the mound in the area occupation continued at a later date elsewhere. However, the
damaged by vehicles extracting rock. In addition, the side of a very small quantity of nondiagnostic ceramics identified from
looters’ hole was cleared and a drawing made, designated the trench and the complete absence of ceramics from the site
Profile 1, that revealed a sequence of stratigraphy and yielded surface would indicate that the Initial Period occupation was
samples for dating without the enlarging the area excavated limited.
(Table 27). In the profile drawing (Fig. 38a), the use surfaces in the
western part of the trench are not visible due to the large
Trench 1, Main Mound number of rocks that were disturbed at some point after the
end of the occupation, possibly very recently. In addition, the
The main mound at Cerro Blanco 2 was constructed as an uppermost of the plaster-coated floors is not visible in the
extension from a natural hill facing the river, 120 3 68 3 4 m profile, though the area of dense fiber in Layer I may also
with a circular court 20 m in diameter. The stratigraphy of the indicate a floor (Fig. 38b).
eastern half of the trench shows layers of deposition built up on Radiocarbon dates suggest that Wall 2 was built around
an inclined surface (Fig. 38a; Table 28). The technique of using 2100 Cal BC and that the floors at the base of the terrace date
a natural hill to achieve the appearance of increased mound slightly later, around 2000 Cal BC (Fig. 38a; Table 28).
height was common among Late Archaic builders. The final Ceramics were found in the lower segment of the trench west
form of the mound was achieved by construction of a series of of Wall 2, yet twined textile fragments were recovered from
retaining walls (Fig. 38a, b) that were filled to create terraces. both sections of the trench. Based on this information, the
Plant fibers collected from the clay mortar at the base of initial occupation of Cerro Blanco 2 dates to the Late Archaic
Wall 3 were dated 2130 Cal BC (Table 27c). Two additional Period, and the site continued to be occupied into the Initial
dates were obtained from surfaces below the terrace formed by Period, as indicated by the presence of a small number of plain
Walls 2 and 3, at the base of the mound, 1980 and 1700 Cal red ceramics and the radiocarbon date of 1700 Cal BC from
BC (Table 27a, b). Together, these indicate that occupation in material just under Floor 2.

40 FIELDIANA: ANTHROPOLOGY
 FIG. 38. (a) Cerro Blanco 2, Trench 1, profile. (b) Cerro Blanco 2, Trench 1, plan view.

CREAMER ET AL: FORTALEZA VALLEY, PERU 41

 FIG. 39. Cerro Blanco 2, Profile 1, drawing.

Cerro Blanco 2, Profile 1 is also unlikely because the circular court is still visible, and
flooding would have affected the entire area, filling the
Profile 1 was located 150 m north of the main mound in one circular court to some degree.
of the largest and deepest looters’ holes we identified, more It appears that after the occupation of Cerro Blanco 2, the
than 2 m deep. This looters’ hole cut through cultural levels, circular court was a favored spot for burials. The use of
including layers of trash. The east side of the hole was cleared circular courts at Late Archaic sites for cemeteries in later
and drawn to verify whether a level of cultural deposition generations has been observed at other sites, including
dating to the Late Archaic Period could be identified (Fig. 39). Porvenir and Upaca (Creamer et al., 2007; Haas & Creamer,
Four different layers were identified that may indicate in press) One scenario for this area is that at different times,
occupation (D, G, I, J), each consisting of compact materials. groups of people visited Cerro Blanco 2 to bury an individual
The layers were discontinuous, however, not extending all the in the circular court, perhaps as recognition of an ancestral
way across the cleared profile. Since the surrounding area is a connection to the site. On the occasion of a burial, in addition
cemetery, these layers may possibly be associated with to the digging of the pit, interment may have been
interment ritual, each layer associated with a particular accompanied by ritual, food preparation, and feasting. Each
interment episode, forming a packed surface around the burial event would have resulted in the creation of a compact
interment spot that did not extend more widely (Table 29). use surface in the immediate area surrounding the interment.
The presence of a small quantity of ceramics throughout the Two such events may be visible in Profile 1 in Layers D and G,
cultural remains of this profile points to an Initial Period or while smaller and less compact deposits, related to events that
later occupation. Either the Late Archaic component was less took place farther away, are marked by deposits that are
extensive around the mound than expected, or it was smaller, such as the fire indicated by Layer I, or by more
concentrated in another part of the site. The profile was diffuse deposits, such as Layer E.
excavated well below the base of the cultural materials to see In the case of Layer E, it is not as clear whether this deposit,
whether there could have been deep flood deposits covering thicker than the others and with only a small quantity of
earlier occupation, and none were encountered. Such flooding cultural material, accumulated from the excavation of a burial

42 FIELDIANA: ANTHROPOLOGY
pit at a distance or whether this layer shows wind, water, and in Sector B, and a second small circular court is associated with
foot traffic buildup of soil over the centuries between a low mound complex on the south end of the site. There are
occupation events at the site. some surface indications of a third, large circular court on the
north side of the main mound.
Cerro Blanco 2 Discussion Sector A

Radiocarbon dates from Cerro Blanco 2 come from three Sector A, adjacent to the Late Archaic complex, is the
contexts. Plaster covering Wall 2, a retaining wall, provided a Initial Period/Early Horizon site of El Castillo de Huaricanga
date for construction of the mound, 2130 Cal BC (Table 27c), (Authier, 2012). The three principal mounds at El Castillo de
while the other two samples yielded dates for occupation on Huaricanga are arranged in a U shape facing north, parallel to
the terrace below the retaining wall, 1980 and 1700 Cal BC the channel of the Fortaleza River. The area surrounding the
(Table 27b, a). These dates suggest that the mound at Cerro principal mounds is under cultivation, though the rectangular
Blanco 2 was built around 2100 Cal BC and may have court retains its overall form. This area was visited in 1958 by
continued to be occupied for more than 400 years. Even after a Japanese research team from the University of Tokyo
the mound was no longer used, people came to the site to bury (Ishida, 1960). A radiocarbon date of 1396–1208 Cal BC (GaK
their dead within the circular court. 109, 1s range) was obtained from El Castillo de Huaricanga at
Cerro Blanco 2 is only about 2 km from Cerro Blanco 1. that time (Kigoshi et al., 1962). This date and the presence of
The sites are similar in overall size and in their later reuse as ceramics on the surface suggest that Sector A is a later
cemeteries. The two sites differ, however, in the number and occupation. No testing was carried out in Sector A in 2003.
arrangement of structures and the courts. Cerro Blanco 1
consists of at least four mounds, three small mounds
Sector B
(including one that is completely destroyed), and a larger
rectangular structure built out of the hillside. There is evidence
Sector B at Huaricanga encompasses the portion of the
for a small court at the base of the rectangular mound at Cerro
site west of the Paramonga-Huaraz highway from the
Blanco 1, though it does not appear to be either sunken or
highway to El Castillo de Huaricanga (Fig. 40). Testing in
circular. Cerro Blanco 2 also consists of a large rectangular
2003 focused on three structures. Profile 2 took advantage of
structure built out of the hillside and a distinct circular court,
an irrigation canal that cut through a mound in Sector B,
though it has been largely destroyed by later burials and their
while Test Unit 1 in Sector B recovered samples from a
subsequent looting. Smaller structures like those identified at
deposit atop a natural hill. Trench 1 in Sector B cleared part
Cerro Blanco 1 are absent. There is considerable similarity
between the structures and layout of Cerro Blanco 1 and other of the surface of a low mound associated with a circle of
stones. Each of these samples yielded materials dating to the
Late Archaic sites, such as Cerro Lampay (Vega-Centeno,
1995) several kilometers to the west in the Fortaleza Valley and Late Archaic Period, though some samples dated later and
Cerro Colorado in the Supe Valley to the south (Shady, 2004). may be associated with the occupation of El Castillo de
Huaricanga (Authier, 2012).
Cerro Blanco 1 and Cerro Blanco 2 appear to be sequential
occupations according to the radiocarbon dates obtained from
Sector B, Profile 2
each: 2120–1680 Cal BC at Cerro Blanco 2 and 1950–1160 Cal
BC at Cerro Blanco 1. The date ranges overlap at the 2s
Profile 2 is located in Sector B2 (Fig. 40; Fig. 42), some
range, 2210–1520 Cal BC at Cerro Blanco 2 and 1890–970 Cal
20 m from the highway. This is one of the smaller mounds
BC at Cerro Blanco 1, giving the overall impression that these
at the site, approximately 29 3 29 3 3 m (Table 2). Con-
occupations were sequential. It is even possible that Cerro
struction of a modern irrigation canal through the side of the
Blanco 1 replaced Cerro Blanco 2 during the transition away
from the Late Archaic pattern after 2000 BC. Both sites are mound exposed a long profile that provides a complete view
closely spaced and show occupation extending across the of the construction strata within the mound. The profile is
transition from the Late Archaic to Initial Period and 40 m long and ranges from 1.5 to 3 m in height (Fig. 43). The
reinforce the fact that the area was not completely abandoned upper surfaces of the mound, covered with modern debris,
at the end of the Late Archaic but continued to be occupied on have been heavily disturbed by looting and years of use as a
a small scale during the Initial Period. work/play area for local residents. At the top of the mound,
below the disturbed upper surface (Phase I), was a large-scale
episode of deliberate fill over a meter thick (Phase II) that
created a formal mound out of what had been a stratified
Huaricanga series of structures. Below this cap of construction fill is
a clear sequence of superimposed buildings and floors (Phase
Huaricanga is a large Late Archaic site covering an area of III).
100 ha and located 25 km from the coast. The main Late Among the distinctive features of this profile was what
Archaic occupation, Sector C, is divided from the rest of the site appeared to be the inclusion of adobe bricks in the fill of the
by the Pativilca-Huaraz highway and consists of one central walls in this mound. Distinctive blocks of adobe were also
mound complex (C1), 285 3 100 3 15 m, the largest Late visible in the profile, though no ceramics of any kind were
Archaic construction in the Fortaleza Valley. Smaller auxiliary identified during the clearing and recording. Rather than
mounds parallel the main mound, including C2, 150 3 413 7 m, being early examples of adobe bricks, however, these blocks
and a small mound (C3) 533 38 3 4 m with a circular court 15 m were later shown to be chunks of naturally occurring clay
in diameter (Fig. 40; Fig. 41). A small circular court is located quarried from thick deposits of laminated layers of clay along

CREAMER ET AL: FORTALEZA VALLEY, PERU 43

 FIG. 40. Huaricanga contour map.

the edges of the wash north of the site. Locally, these blocks numbers in Table 31 suggest. A question that arises from the
are called yapana. stratigraphy of this unit is whether the change in construction
Phase III includes a series of distinctive room remodeling and layout from Phase III to Phase II represents a change of
episodes that can be seen in the exposed profile (Fig. 44; political or ideological outlook that resulted in demolition of
Table 30), where a room with a low bench around the sides of structures and piling up of fill or whether the Phase II deposit
a sunken central space was built and then remodeled. Over represents a lapse in occupation and a later reoccupation
time, a series of alterations were carried out that retained the during Phase I.
basic configuration while the central area became progres- Profile 2 provided the opportunity to record a complete
sively smaller. This sequence of structures bears a resem- sequence of construction of a small mound in Sector B,
blance to the basic plan of Mito-style temples found in the providing examples of both horizontal and vertical stratig-
highlands immediately to the east of Huaricanga (Izumi & raphy. The profile reveals an opportunistic approach to
Terada, 1972; Burger & Salazar-Burger, 1986; Bonnier, 1997; building materials in the use of yapana from a local deposit
Haas et al., 2010). Mito temples are marked by exterior walls, along with locally available stone and clay. The profile shows
a low bench around a square-shaped, sunken area within that the area was repeatedly remodeled in ways that
which there is a central hearth that is often ventilated from gradually raised the level of the mound and includes a series
below. Confirming the nature of these structures will require of structures similar to Mito temples that date to 2500–2300
further excavation. Cal BC.
A date of 2360 Cal BC was obtained from Layer G fill
between floors of the two of the upper Mito-style temples Sector B, Test Unit 1
(Table 31a), and a second date of 2510 Cal BC from charcoal
was obtained from the base of Phase III deposits (Layer NN, This unit was located on the north slope of a small mound
Table 31b), indicating that the period of transformation from or natural hill in Sector B3, also cut by the highway (Fig. 40).
a level, sandy area to a low mound with repeatedly remodeled The unit was positioned to be aligned with a trash layer visible
structures took place between 2500 and 2300 Cal BC. The in the road cut. Excavation revealed layers of deposition tilted
two radiocarbon dates obtained from Profile 2 are statisti- toward the northeast (Fig. 45). There appears to have been an
cally contemporaneous, though the charcoal sample has a effort to level the area by depositing trash fill consisting of
large standard deviation (110 years). The actual difference shell, bone, lithics, charcoal, and other botanical remains
between the two dates is unlikely to be as great, as the along the slope of the hill (Phase III). The use of trash for

44 FIELDIANA: ANTHROPOLOGY
 FIG. 41. Photo overview of Huaricanga from hills ESE of the site. The inset is a reconstruction of the main Late Archaic mound at the site
(courtesy M. Authier).

FIG. 42. Photo of Sector B, Profile 2, beside irrigation canal.

CREAMER ET AL: FORTALEZA VALLEY, PERU 45

 FIG. 43. Huaricanga, Sector B, Profile 2, along the irrigation canal with close-up view of a sequence of remodeled floors.

leveling suggests that the area was already occupied when the cultural deposits lie below or whether the remainder of the
leveling was undertaken. Two dates obtained from charcoal in mound is a natural formation, as it appears to be.
this layer yielded dates of 3550 and 2700 Cal BC (Table 31e, f;
Table 32). After the leveling effort, additional construction Sector B, Trench 1
using gravel and compact earth formed the base of a platform
that was completed by a retaining wall (Wall 1) and the rocky Trench 1 was located near the northeast corner of Sector B (B1)
fill behind it (Phase II). During a subsequent period, the on a mound 47 3 41 3 4 m high. The trench measured 3 3 12 m,
platform was no longer in use and was covered with more fill extending from the base to the middle of the mound. The mound’s
consisting of cultural debris, soil, and ash. The consistent form and the associated circle of stones with an upright monolith
composition of the later covering and the absence of floors or in the center approximate a preceramic site layout. The court is not
packed surfaces within the Phase I material suggest that after sunken but consists of a circle delimited by large stones in a
the platform was no longer in use, the area was used for trash position similar to that of mound/circular court combinations at
disposal. There is no indication that there was a specific act other Late Archaic sites, especially Caballete. A single boulder over
leading to the platform being covered. Two dates of 2570 and 1 m high is located in the center of the circle (Table 2; Fig. 46).
2260 Cal BC were obtained from charcoal in the fill covering The trench was intended only to clear the surface of the
the platform and suggest that the period of use of the structure mound and identify the edge of the mound and the uppermost
was a minimum of 150 years and possibly much longer materials. The excavation consisted of removing the loose
(Table 31c, d). disturbed surface material and a few centimeters of in situ fill
Four radiocarbon dates were obtained from Test Unit 1 in to allow for identification of the uppermost features. Organic
Sector B, ranging from 3550 to 2260 Cal BC. This reinforces the materials were very scarce in these upper deposits, and we had
Late Archaic date of Sector B noted in the Profile 2 dates. Wall 1 only two in situ samples of organic material that could be used
appears to have been built after some trash and earth were for radiocarbon dating. A series of shallow layers were
deposited to level the area, creating a low platform that was the removed in an effort to distinguish building stones that had
focus of occupation in this area. The test unit reached a depth of fallen down from stones still in their original positions
138–162 cm below surface and ended at culturally sterile deposits, (Fig. 47a). Remains of low retaining walls were uncovered
yet the mound cut by the highway was more than 3 m high, that showed a shallow stepped structure, perhaps more
showing that the mound into which this unit was excavated was accurately termed a platform than a mound (Table 33). A
constructed on top of a slight natural rise. Additional excavation date of 2430 Cal BC obtained from Layer C (Table 31g)
would be necessary to determine whether the sterile layer reached suggests the platform was in use by that date. The later date of
in Test Unit 1 could be flood or alluvial deposits and additional 1040 Cal BC from plant fiber in Wall 4 appears to reflect much

46 FIELDIANA: ANTHROPOLOGY
 FIG. 44. Huaricanga, close-up of remodeled floors (Mito temple) in Phase II.

later use and remodeling (Table 31h). The stepped form of the The construction included a mix of angular rock, probably
structure and the numerous walls show many episodes of quarried from the hillsides nearby, and cobbles from the
remodeling and reconstruction, though these did not add riverbed, also nearby. This range of materials is notable for
much height to the structure (Fig. 47b). the mix of nearly equal parts cobbles and angular rock.
We were unable to establish a correlation between the Elsewhere, sites tend to be constructed of predominantly one
profiles of the north and south sides of the trench, largely due to type of rock or the other, depending largely on the distance
the small area cleared. The radiocarbon sample from above from the river to the site. At Huaricanga, located as it was on
Wall 4 is more recent than the Layer C date, suggesting later the first terrace above the river in an open dry wash, both
reuse of this area or that debris from a subsequent period river cobbles from the banks of the river and angular rock
accumulated here. At the same time, no ceramics were found in quarried from the sides of the wash were available for mound
this area, making it unlikely that a substantial later occupation construction.
was present. Rather than considering the Wall 4 date This test trench revealed a complex sequence of construc-
anomalous, it may be associated with the period of occupation tions. While specific functions could not be determined from
of the adjacent El Castillo de Huaricanga, occupied during the the limited testing, there are no indications that it was
Initial and Early Horizon periods (Authier, 2012). residential architecture. There were no associated deposits of
A section of Floor 1 had been removed, marked by a domestic trash, no hearths, and no heat-altered rocks that are
distinct edge of clay floor material and a gap between the floor associated with heated rock or pachamanca-style cooking.
and the nearest wall, Wall 1, where the plaster curved upward, Given the association of these structures with a stone-rimmed
showing a former floor/wall connection (Fig. 47b). Wall 1 had circular court, it is most likely that this construction was
been the side of a broad plastered step bounded by Wall 4. A associated with communal or ceremonial activities.
lower step was bordered by Wall 2. Remodeling and
rebuilding are characteristic activities of the Late Archaic Sector C
Period, part of the long-term process of mound building. Wall
4 was replaced by Wall 5, broadening the step, and Wall 6, Two profiles were recorded on the large mound in Sector
introduced to create an additional step in the platform, was C at Huaricanga. Profile 1 included a north-facing portion of
replaced with Wall 3. Eventually, Wall 1 was replaced by walls the large mound, Sector C1. Another section of the same
that seem to have decreased the size of the upper step, where mound was recorded in Profile 3, where a long profile on the
the corner of two walls was identified as a platform. Although west end of the mound was exposed by the construction of
the Sector B mound at Huaricanga did not reach more than the adjacent highway. One test unit was excavated in Sector
3 m at its highest point, numerous low retaining walls were C2 to provide a sample of midden associated with the
built and rebuilt. mound.

CREAMER ET AL: FORTALEZA VALLEY, PERU 47

 FIG. 45. Huaricanga, Sector B, Test Unit 1.

Sector C, Profile 1 repeated construction phases, while the dates suggest a long
period of use as well. The oldest date comes from the base of
Profile 1 was located on the north side of the main mound Wall 2 from a sample of plant fiber and the youngest date from
in Sector C (C1), east of the highway to Huaraz (Fig. 48; plant fiber in floors below Wall 2. It is difficult to assess the
Fig. 49). Unlike other Late Archaic mound sites in the Norte possible sources of error apart from the possible mixing of older
Chico, the main mound at Huaricanga is a long series of and younger materials in samples of plant fiber compared to
connected structures that extends almost 300 m from the samples of fiber bags. Despite the issues, the range of dates
highway (Fig. 41, inset). The profile was recorded along a obtained is entirely within the Late Archaic Period.
section of the mound where a portion of the structure was A number of unbaked pottery tokens were found during the
taken for fill and gravel during the 1970s when the highway clearing of Profile 1 just above Wall 3 (Fig. 49; Fig. 50). Their
was widened. The exposed profile is 6–7 m wide and 7 m high, location, tucked along the base of a wall, would indicate that
displaying a series of floors, retaining walls, coats of plaster, they were intentionally placed, but it could not be determined if
and layers of construction fill that show the construction their position was related to some specific function or if they
techniques employed and the construction phases at this end were some kind of offering. The group includes eight tablets and
of the mound (Table 31; Table 34). Four radiocarbon dates two fragments made of untempered clay, ranging in shape from
were obtained from this profile, ranging from 2800 to 2340 rectangular to trapezoidal. One is triangular. The unbroken
Cal BC. Two radiocarbon dates, 2620 and 2440 Cal BC tokens were 6–8 cm long, 3–7 cm wide, and about 1 cm thick.
(Table 31j, k), were obtained from materials in Phase III, a Two surface finds in the disturbed fill at the top of this profile
major construction episode using shicra bags to fill retaining may also have been offerings, but their surface context makes
walls, such as Wall 2. A sample of plant fibers from the base this difficult to determine. A slate fragment engraved with
of Wall 2 at the top of the Phase IV materials dated 2800 Cal hatched lines was found (Fig. 51), similar to hatching on a small
BC (Table 31l), and plant fibers from a floor below Wall 2, one polished stone from a Late Archaic level at Huaca Prieta (Bird
of a series of floor surfaces that make up Phase IV at et al., 1985, p. 89, fig. 4) and to an ax with incised hatching from
Huaricanga (Fig. 48; Table 31i) dated 2340 Cal BC, indicating Kotosh from a ceramic bearing level (Izumi & Sono, 1963, pl.
an earlier construction phase, probably when Wall 3 was filled 110). Several perforated land snail shells were also found,
in and a series of floors built above it. The stratigraphy shows including some still strung together on a spun string. Although

48 FIELDIANA: ANTHROPOLOGY
 FIG. 46. Upper: Photo of rock circle at Huaricanga located on the SW side of the low mound in Sector B and extending onto the low slope of
the mound. Lower: Circle of upright stones, or huancas, forming a circular court at Caballete.

individual disc-shaped shell beads have been recovered from raised the height of the mound; Phase V (Wall 3), Phase III
excavations at Huaricanga and other sites, this was the only (Wall 2), and Phase II (Wall 1) include retaining walls. These
find of a necklace of whole land snail shells. were punctuated by two phases of mound occupation that
Profile 1 at Huaricanga reveals that mound construction included the creation of successive floors (Phase IV and Phase
using retaining walls and fill to form successive platforms II). The dates obtained show that construction and occupation

CREAMER ET AL: FORTALEZA VALLEY, PERU 49

 FIG. 47. (a) Huaricanga, Sector B, Trench 1, map. (b) Huaricanga, Sector B Trench showing the construction of walls over time. 1. Oldest
construction. Walls 1, 2, and 4 create two stepped layers. 2. The addition of Wall 6 adds a third step. 3. The middle step is made wider with the
replacement of Wall 6 with Wall 3. 4. The middle step is made wider with the replacement of Wall 4 with Wall 6. 5. The highest step is made
smaller by the replacement of Wall 1 with the platform.

50 FIELDIANA: ANTHROPOLOGY
 FIG. 48. Huaricanga, Sector C, Profile 1.

of the Sector C mound took place between 2800 and 2300 Cal Sector C, Profile 3
BC. This profile did not show any of the overlapping of
structures that can be seen in the contour map and aerial views Part of C1, Profile 3 was exposed when the western end of
of the site. Future research will be needed to determine the the mound was removed by the Huaraz highway construction.
form of the initial structures in Sector C and whether they Profile 3 is about 60 m long, exposed by heavy machinery
represent an innovation or experiment in mound construction. during roadwork. The profile was cleaned and straightened,
The long line of aggregated mounds is different from most but no new excavation into the hillside was conducted. The
other Late Archaic mounds in the Norte Chico. Long mounds profile ranges from 1.5 to 2.2 m high (Fig. 52; Fig. 53). We
like this one at Huaricanga characterize the sites of Huacache were not able to identify shicra bags in Profile 3, and there was
and Peñico in the Supe Valley, but the construction sequence no other fiber visible in the exposed stretch. Two samples of
remains unexplored for either of those sites. By itself, Profile 1 charcoal were radiocarbon dated, from Layer C and Layer G
illustrates a distinct point within the range of acceptable (Fig. 52).
variability in the architectural canon of Late Archaic The date of 2220 Cal BC from Layer C (Table 31m) was
monument construction, and it speaks to a lack of a regional from a sample of charcoal from Feature C1, a lens of
centralization, as local leaders were exercising independence in compact clay mortar within construction fill of medium-sized
determining the shape of monuments at the site level. rocks and clay mortar. The date falls into Phase III

CREAMER ET AL: FORTALEZA VALLEY, PERU 51

 FIG. 49. Huaricanga, schematic of Sector C, Profile 1.

(Table 35) and indicates a period of time when Wall 1 was fine gravel. This is the oldest deposit at the west end of the Sector C
being built, the period of large construction phases in this mound, more than 3 m below the top of the mound and situated
part of the site. on a layer of soil lacking artifacts. The Layer G ashy deposit may
The date of 2390 Cal BC comes from a sample of charcoal reflect a special activity carried out to inaugurate the construction
in Layer G, the earliest occupation of the area, visible at the that involved burning some materials, or it may be associated with
base of the roadway. In this layer are a series of overlapping residential activities early in the sequence of occupation.
deposits that suggest habitation. The layers are immediately
above a sandy deposit that appears to represent the original Sector C, Test Unit 2
ground surface in this locality. It appears that there was
occupation in the form of floors, fires, and some trash A test unit was opened in Sector C (C2), south of the main
deposition early in the history of this portion of Huaricanga mound at Huaricanga, beside a looters’ hole where a layer of
that was replaced by construction on a much larger scale. trash could be observed. The unit was 1 3 2 m and was
This profile clearly shows a sequence of construction episodes, excavated in artificial levels (Fig. 54). Two distinct periods of
each represented by construction fill. The date of 2390 Cal BC use were identified during the occupation of this part of the
comes from the lowest stratum visible, a distinct layer of ash and site. A borrow pit (Layer E) (Table 36) was excavated into the

52 FIELDIANA: ANTHROPOLOGY
 FIG. 50. Photo of clay tablets from Sector C, Profile 1, at Huaricanga.

original ground surface. This was later filled in with a mix of

cultural debris, including ash, shell, and botanical remains.
There appears to have been a hiatus in occupation at this point
and then later use of this area as a midden. The trash
deposited during the renewed period of use (Phase II) included
lithics and shell in a dense matrix of charcoal and plant
material. A sample of charcoal from this level dated to 2400
Cal BC (Table 31o), indicating a relatively early period of use
for this locality since the date is from the later dense midden
and not from the filling of the borrow pit.
Test Unit 2 confirms the presence of domestic trash in close
proximity to the main mound of Huaricanga. This unit
reached a depth of 97 cm below surface, a substantial deposit.
Similar deposits of midden near large mounds were identified
at Caballete (Sector A, Test Unit 4, and Sector B, Test Unit
1) and Porvenir (Sector A, Test Unit 3). However, more
excavation would be needed to determine the actual use of
the area during the period when the surface formed over
Layer E or earlier, when the first borrow pit was excavated
here.

Huaricanga Discussion

Testing at Huaricanga revealed a long period of Late

FIG. 51. Incised rock from Huaricanga, Sector C, surface above Archaic occupation, beginning toward the end of the fourth
the profiled area. millennium or beginning of the third millennium BC,

CREAMER ET AL: FORTALEZA VALLEY, PERU 53

 FIG. 52. Huaricanga, Sector C, Profile 3, along highway.

FIG. 53. Huaricanga, photograph of profile along highway, with Guillermo Galvez G.

54 FIELDIANA: ANTHROPOLOGY
 FIG. 54. Huaricanga, Sector C, Test Unit 2.

continuing for 800 years or more. Construction identified to examine habitation. However, in three places–the base of
during testing included three mounds built of retaining walls Sector C, Profile 3 (2390 Cal BC); the base of Sector B, Profile
filled with rock in Sector B and Sector C; a low platform 2 (2510 Cal BC); and the trash deposit above the borrow pit in
exposed in Sector B, Test Unit 1; and a borrow pit that was Sector C, Test Unit 2 (2400 Cal BC)—evidence of habitation
later filled in and subsequently used for trash disposal in was found in the form of trash deposits, lenses of ash and clay,
Sector C. Areas of overlapping floors indicating habitation or packed earth floors. Dates for these features cluster
were identified at the base of the Sector C and Sector B between 2510 and 2390 Cal BC (Table 31; Table 40). These
mounds. may indicate an active period of mound construction, with
The earliest date for construction at Huaricanga is 3550 Cal groups of construction workers living close to the terraces and
BC, from Test Unit 1 in Sector B, but this is an outlier, and platforms being built to raise the height of the Sector B and
other dates from the same unit date 2700–2260 Cal BC. Earlier Sector C mounds. Occupation of the area continued into the
construction probably took place among the mounds that Initial Period, as indicated by the adjacent Castillo de
make up Sector C, dated 2800–2340 Cal BC (Table 32). Much Huaricanga and the date of 1040 Cal BC from Sector B,
of the construction at Huaricanga overlaps in date (Table 31), Trench 1 (Table 31h), and which may date to use of Sector B
and a particular structure cannot be singled out as the oldest during occupation of El Castillo de Huaricanga.
of the structures present at the site. There is some suggestion
that the northern end of the Sector C mound is older than the
southern end near the highway. In this case, construction of
the series of overlapping mound structures begun around 2800 Shaura
Cal BC in Sector C proceeded from the base of the hills
toward the edge of the river terrace, continuing over the In air photos of the Fortaleza Valley taken in 1969, the site
following 500 years. of Shaura is clearly visible and appears to have been one of
The goal of testing at Huaricanga was to obtain radiocar- the most impressive of the sites in the upper section of the
bon dates for different portions of the site and not specifically valley, with a main mound (67 3 52 3 10 m), an associated

CREAMER ET AL: FORTALEZA VALLEY, PERU 55

 rock used in wall construction, clearly an intentionally
placed object (Fig. 58).
The area cleared in Trench 1 at Shaura revealed a single
phase of occupation on the main mound. This included a
segment of wall with a broad, plastered surface extending
out from it in Layer E and Layer F. This segment appears to
be the base of a step in the overall construction of a step-
sided structure. The trench excavation at Shaura yielded
evidence that a portion of the mound structure is still present
at the site. Clearing the trench revealed that the mound was
constructed almost entirely of rounded river cobbles rather
than quarried rock. A large in situ wall segment was cleared,
and a compact surface above the wall was revealed. Thus,
despite the cobble construction, Shaura appears to have been
FIG. 55. 1969 air photo of Shaura.
constructed in a manner similar to other sites where
retaining walls of large rock were filled with shicra bags
and other fill; leveled with small rocks, gravel, and sand; and
sunken circular court 24 m in diameter, and two secondary
then covered with a thick layer of clay mortar (Table 38).
mounds (Fig. 55). Since that date, however, Shaura has
The polished boulder has a different shape than the upright
served as a quarry from which rock was removed for the
stones, or huancas, seen elsewhere at Late Archaic sites. At
construction and maintenance of the highway between
the same time, the stone is polished in patches on the surface,
Pativilca and Huaraz, built in the 1970s. The main mound
similar to some huancas, and it was found in the rubble of
at Shaura has been virtually destroyed by heavy machinery
the main mound, suggesting that it was once placed on the
employed in building the highway that mined the mound for
structure.
stone, and only a small fragment of the mound base remains
Two charcoal fragments from Shaura were collected from a
intact. Review of the area and the air photos suggested that a
row of cobbles that form the wall at the upper edge of the
small portion of one side of the main mound at Shaura was
preserved step. These appear to be part of the original mound
still preserved, and it was in this area where we carried out
construction. A date of 1330 Cal BC (Table 37a) was
limited testing. Shaura was not divided into sectors since so
obtained on a sample of charcoal from the fill behind Wall
little of it remained (Fig. 56).
1. A second date of 2050 Cal BC (Table 37b) came from
Excavations at Shaura included two units. Trench 1 was a
charcoal within the stones of Wall 1. The dates are divergent,
3 3 9-m unit cleared to assess the nature of construction of
though both indicate the relatively early date of the site. The
the main mound. Samples for radiocarbon dating would
earlier date does overlap statistically with the date obtained
help us locate Shaura within a regional chronology. A 1 3 2-
from Test Unit 1 and suggests that a Late Archaic date for
m area was designated Test Unit 1, with the goal of
occupation of the site is reasonable. Since Trench 1 was an
determining the presence of lower levels of the main mound
effort to clear the surface and record surviving features,
or midden possibly dating to the preceramic period. The
additional evidence of construction is likely to be preserved
bulk of the mound was no longer present, but construction
beneath the cleared surfaces.
and occupation layers dating to early phases might be
preserved.
Test Unit 1
Trench 1, Main Mound
Test Unit 1 was located in the center of what had been the
Trench 1 was located in the preserved corner of the main main mound at Shaura. At the time of testing, the area
mound at Shaura, though initially it was difficult to determine appeared to be a low platform crossed by a dirt road. The
whether the inclined area where the trench was placed was the objective was to determine whether any portion of the main
exterior of the mound or merely a slope created by gravel mound was preserved and whether the platformlike area was
mining. What remains of the main mound at Shaura is an area part of the mound or a remnant of earthmoving activities.
of river cobbles, sloping steeply toward the southwest. From Test Unit 1 proved to be relatively shallow, though there
the air photos and the orientation of the site in this area, it were some archaeological features present (Table 39; Fig. 59).
appears that this was originally the western corner of the main Two wall segments, one with an irregular base layer of clay
mound, and our clearing in Trench 1 established that this was and a burned area or patch of burned fill that had been
a preserved portion of the exterior surface of the structure covered with earth, shows one period of use, while the ashy
along its northwest margin (Fig. 57). surface below the walls may indicate another period of use.
Clearing a section of the slope revealed two sections of Charcoal from the base of the ash layer dated to 1880 Cal BC
wall parallel to one another and forming a step of the (Table 37c).
overall step-sided mound structure. A packed clay floor Although it was possible to identify a surviving portion of
covered the cleared area at the base of Wall 2 as well as part the main mound at Shaura, the area excavated in Test Unit 1
of the space between the top of Wall 2 and the base of Wall was small, and it could not be determined whether the walls
1. Sitting on that step (Layer D) was a highly polished were part of a structure or perhaps the lowest courses of
boulder, though it is not certain whether this was the stone’s retaining walls for subsequent platform construction. It would
original position. The black, fine-grained volcanic rock is be possible to examine this possibility further with a series of
distinctive in form and different in composition than the additional test units or auger holes that could show whether

56 FIELDIANA: ANTHROPOLOGY
 FIG. 56. Shaura contour map showing location of test units.

any cultural deposits lie farther below the surface in the construction dating to the Late Archaic Period is still
general area of Test Unit 1. preserved at the site, while the polished boulder indicates the
ceremonial function of the mound.
Shaura Discussion The date of 1880 Cal BC from Test Unit 1 falls within the
range of dates from the trench and is consistent with Late
Trench 1 confirms that a corner of the main mound still Archaic occupation. The Test Unit 1 excavation suggests that
exists at Shaura. Although the two dates from the trench come there may be material from the lowest levels of the mound at
from charcoal samples (Table 37), the dates suggest occupa- Shaura that are still preserved that could provide an
tion of this mound prior to 1300 BC. The presence of two opportunity to examine layers that are ordinarily deeply
walls and a use surface in Trench 1 indicates that permanent buried under mound structures. Whether this type of

CREAMER ET AL: FORTALEZA VALLEY, PERU 57

 FIG. 57. Shaura, Trench 1, main mound showing the location of the polished boulder nicknamed the ‘‘piedra voluptuosa.’’

excavation will take place before the remainder of Shaura is 2003 and 2004 extend across much of this distance, from
mined for gravel is the more critical question. Porvenir within 5 km of the coast to Shaura, 33 km inland.
The number of sites and their size can be considered in terms
of the duration of occupation at each site and the sequence of
Discussion: Fortaleza Valley occupation across the region. If large sites were established
sequentially during the Late Archaic, each in turn acting as a
The Lower Fortaleza Valley spans some 46 km from the central place for the entire valley, we would expect to find a
ocean to the foothills of the Andes, and the six sites tested in cluster of radiocarbon dates indicating peak occupation at a
different date for each site. An alternative to sequential
construction and occupation of sites would be construction of
each site separately depending on local conditions. This would
result in a series of centers overlapping in time. Sites may have
come into competition with one another as their area of
influence expanded into that of neighboring groups. Archae-
ologically, we would expect to find sites similar in size and
date, though there could also be evidence of consolidation
over time with one site increasing in size and another
decreasing.
Systematic and comprehensive surveys of all three valleys
from the coast to where the valleys narrow at the start of the
Andean foothills were conducted in 2004–2007 (Nelson &
Ruiz, 2005; Perales, 2006, 2007). A total of over 1500
archaeological sites and over 10,000 separate site occupations
were recorded in the course of these surveys (many ‘‘sites’’ in
this area were occupied at multiple times in the course of the
past 5000 years). Over 30 large sites with various combinations
FIG. 58. Shaura, polished boulder on main mound. of public and residential architecture were identified as

58 FIELDIANA: ANTHROPOLOGY
 FIG. 59. Shaura, Test Unit 1.

probably Late Archaic occupations in the four valleys of the there was most likely an occupation in the Fortaleza Valley
Norte Chico. These survey data strongly indicated that the that cut across the transition between the Late Archaic and
cluster of sites first identified in the Supe Valley and now the Initial Period. The occupation of the valley in this later
shown to extend across four valleys is part of an emergent period, however, was shifting away from the large sites with
regional system of cultural complexity. the largest mounds toward smaller sites with shorter
Comparing the radiocarbon dates among the six sites tested occupations. Definitive chronology for Shaura, unfortunate-
in 2003 (Table 40), some patterns in occupation can be ly, will forever remain an unknown, as it has been so
identified. The largest sites, Porvenir, Caballete, and Huar- thoroughly destroyed. Based on the radiocarbon dates
icanga, all have the earliest and longest occupations. Although reported here as well as on direct surface observations, all
this might be due to obtaining more dates from these sites, three of the large early sites as well as Cerro Blanco 2 were
looking at their respective distributions shows that while there revisited and reused in later time periods. All have
are fewer dates from Cerro Blanco 1, Cerro Blanco 2, and significantly later (Early Horizon to Middle Horizon)
Shaura, they all cluster at the end of the Late Archaic and cemetery areas that have been heavily looted. It is significant
beginning of the Initial Period. Looking at the specific to note that in both Porvenir and Cerro Blanco 2, the sunken
chronologies for the three largest sites, we find continuous circular courts and surrounding architectural features were
blocks of dates, with reasonable confidence ranges, that show specifically targeted as burial areas in later times (Haas &
long-term occupation of each. In all three of these figures, the Creamer, in press). These later intrusions may indicate later
dark vertical lines mark the weighted average calibration of groups claiming a hereditary relationship with the ancient
the earliest and latest dates that bracket the continuous population or using interment in an abandoned site as a way
clusters of dates. Dates for Porvenir cluster between 2840 and to legitimize political acts. The revisiting and reoccupation of
1750 Cal BC (Table 4); for Caballete, dates cluster between sites is a common phenomenon with a variety of purposes
2730 and 1890 Cal BC (Table 11); and for Huaricanga, dates and has ancient roots at these sites.
cluster between 2800 and 2220 Cal BC (Table 31). There are Comparing dates from all six sites, it is apparent that the
earlier and later dates at all three sites that fall outside the three largest sites were all constructed and utilized over a
continuous clusters. These charts do not confirm one way or span of centuries in the middle of the third millennium BC
another whether these ‘‘outlier’’ dates reflect earlier or later and that all were contemporaneous with each other. The idea
activities at these sites. The breaks in the charts may be of a sequence of dominant centers, with regional leadership
artifacts of gaps in the excavation record, indications of passing from one site to the next, seems unlikely in the
contaminated or orphan samples, or evidence of some limited present context. Competition and negotiation among similar-
level of early/late activity. The dates at Shaura, Cerro Blanco sized centers seems likely, a possible model of peer–polity
1, and Cerro Blanco 2, taken together with the various early model interaction (Renfrew, 1986; Vega-Centeno, 2005;
dates from Porvenir, Caballete, and Huaricanga, indicate that Creamer et al., 2006).

CREAMER ET AL: FORTALEZA VALLEY, PERU 59

 In addition to the six sites tested as part of the present Archaic. These new data yield broader vision of the
project, Cerro Lampay, a small mound/sunken circular court development and extent of emergent polities in the Norte
site across the valley from Caballete, was extensively Chico region and the Andes as a whole. The Norte Chico was
excavated by Vega-Centeno in the early 2000s (Vega-Centeno, something of an enigma for archaeologists for much of the
2005, 2007). The cluster of calibrated dates from Cerro 20th century. Investigators recognized the existence of a
Lampay ranges from 2480 to 2130 Cal BC. This would make it complex of apparently early sites in the Norte Chico region as
fully contemporaneous with the latter half of occupation at the early as the 1940s, when Kosok and Schaedel (Kosok, 1965)
three large sites in the valley. Distribution patterns based on reported the presence of sites in this area that lacked ceramics
GIS modeling raise the possibility that Cerro Lampay may on the surface. Kosok called this large group of seemingly
have been an offshoot or daughter community from one of the anomalous sites the ‘‘Pativilca Complex.’’ Willey and Corbett
larger sites, probably Caballete (Rutherford, 2008). One (1954) conducted excavations at Aspero in the Supe Valley in
additional site, Cerro de La Cruz, situated between Huar- the late 1940s but did not recognize it as early. It was on
icanga and Shaura, was constructed with shicra bags visible in Willey’s return to Aspero with Moseley that the site was
looted areas and lacked ceramics on the surface and in recognized as ‘‘preceramic,’’ although there were no radiocar-
exposed profiles. While we were unable to test this site, it bon dates available at that time (Moseley & Willey, 1973).
appears likely to have been occupied during the Late Archaic Feldman (1980) provided the first radiocarbon dates for
as well. In total, then, there are seven confirmed and probably Aspero, dating to the third millennium BC. While working at
eight Late Archaic mound sites identified in the lower valley of this coastal maritime site Feldman (1980, 2009) also conducted
the Fortaleza River. informal surveys in surrounding areas and noted the presence
of what appeared to be more early sites in the region. Pozorski
Stratification and Radiocarbon Dates and Pozorski (1990) made similar observations. Engel (1987)
recorded the early sites in the Pativilca Valley but did not
The majority of the dates reported here involved traditional immediately recognize them as unusual or early. Zechenter
radiocarbon counting techniques that require 20 g or more of (1988) conducted sample excavations at several sites in the
sample material and generally give a wide standard deviation Supe Valley and recovered radiocarbon samples showing that
range of 660–90 years. This means that the date is likely to fall at least two of the big sites were occupied in the period
within a span of 120–180 years. Accelerator mass spectrometry between 3000 and 1800 BC. Vega-Centeno et al. (1998)
(AMS) technology requires less material and generally gives a recorded all the major early sites in the Fortaleza Valley but
narrower range of dates, (generally between 615 and 50 years), identified them as belonging to the Initial Period. It was not
though again it does not yield an ‘‘exact’’ date. All but two of until the work of Shady et al. (2001) at Caral, however, that
the samples in the present report are traditional, non-AMS chronological data were acquired and the Late Archaic date
dates, as the broader range of possibilities was considered and anthropological significance of the Norte Chico region
acceptable given the research questions of the project. The was more widely recognized and accepted.
broad range produced by traditional radiocarbon dates makes Shady (2004, 2009) identified a number of sites in the Supe
it difficult to assign a specific date to any given deposit. As a Valley as dating to the period from 3000 to 1800 BC. In 2002,
result, the average calibrated dates do not always correspond the Proyecto Arqueologico Norte Chico conducted test
to stratigraphic superpositioning of deposits. Thus, we found excavations in the Pativilca Valley (Creamer et al., 2007)
that the average calibrated dates from upper deposits could be and identified six large sites dating to the Late Archaic. The
older than the average calibrated dates from stratigraphically present project identified six more sites in the Fortaleza Valley
lower deposits. The significance of this pattern is twofold. as Late Archaic. The excavations by Vega-Centeno at Cerro
First, average calibrated dates can give a false impression of Lampay identified it as a Late Archaic site as well, and
precision. They function for broad comparative purposes but systematic survey of the lower valley has identified at least two
not to provide precise dates for any given association. Second, more sites that have the characteristics of the Late Archaic
in many deposits, the broad range of results provided by (Perales, 2007). Cardenas (1979) conducted a survey in the
standard count radiocarbon dates can often exceed the entire Huaura Valley in the 1970s that included Late Archaic sites.
period of deposition or use of the location being studied. Fung (1988, 2004) tested the site of Bandurria, a site on the
Accordingly, an accurate absolute chronology for any large littoral in the Huaura Valley, the southernmost valley in the
preceramic site requires analysis of a substantial number of Norte Chico region, and obtained radiocarbon dates indicat-
radiocarbon dates, all of which should be dated with samples ing that there were burials at the site dating to the Middle
of annual plant fibers analyzed with AMS. It may be possible Archaic as well as the Late Archaic. Chu’s subsequent
to construct relative chronologies for the Late Archaic sites in excavations at Bandurria have demonstrated a major Late
the Norte Chico region based on characteristics such as Archaic occupation at the site, with monumental architecture
building techniques, weaving styles, or stone tool types; and at least one large sunken circular court. Survey in the
however, such relative chronologies will require suites of Huaura Valley revealed yet more sites that lack ceramics and
AMS radiocarbon dates from multiple excavation units have terraced platform mounds, sunken circular courts, and
distributed across any large site. Efforts to build relative shicra bag construction (Nelson & Ruiz, 2005). Altogether,
chronologies without large suites of radiocarbon dates are at today over 30 major sites in the four valleys of the Norte
best a rough estimation (cf. Shady, 2004, 2006). Chico are identified as having been occupied during the period
from 3000 to 1800 BC.
The Late Archaic Period in the Fortaleza Valley From the radiocarbon dates obtained in 2003 and 2004
discussed here, there is now evidence that the U-shape site
The radiocarbon dating for the Fortaleza sites has broad layout pattern originated during the Late Archaic. Porvenir
ramifications for interpreting the chronology of the Late and Caballete, both of which were occupied by the mid-third

60 FIELDIANA: ANTHROPOLOGY
millennium BC and possibly earlier (Table 3; Table 11), have florescence in this region during the period from 3000 to
a marked U-shaped layout that is partially though not 1800 BC Although there are somewhat similar, single, large
completely symmetrical (Fig. 2; Fig. 11). It appears that sites located in other parts of Peru, such as Sechin Bajo in the
sections of Late Archaic sites were built at different points Casma Valley (Fuchs et al., 2008), El Paraiso in the Chillon
in time based on the repeated construction of mound/circular Valley (Engel, 1966), and La Galgada in the upper reaches of
court units that, in addition to serving a ceremonial purpose the Santa Valley (Grieder et al., 1988), the concentration of
on their own, were situated to maintain and extend the U- sites in the Norte Chico is unparalleled anywhere in South
shaped layout. Not all Late Archaic sites were arranged in this America—or indeed the world—at that time period. The
way, as demonstrated by Huaricanga (Table 31), which dates social, political, and economic relationship among these sites
to the same period and consists of a linear series of remains enigmatic and will take years more excavation by
overlapping structures (Fig. 40; Fig. 41). The major site of investigations in all four valleys. Although there has been an
Caral in the Supe Valley also does not have a U-shaped argument presented that the site of Caral in the Supe Valley
arrangement of its mounds (Shady, 2004). Thus, a pattern was the capital of a broad regional state (Shady, 2006; Shady
appears to have been developing that was not adopted by all & Leyva, 2003), empirical evidence to support this theory has
the major centers in the region, suggesting a degree of not yet been presented. At none of the sites tested in either the
independence among different sites in the planning and Pativilca Valley or the Fortaleza Valley is there any indication
location of structures. that they were under the political hegemony of Caral or any
The Late Archaic sites in the Norte Chico region other site.
demonstrate a major investment of labor in construction of There are several challenges to establishing a more
pyramids and sunken courts. Dedication of labor to construc- comprehensive understanding of the nature of the relationship
tion is shown by the number of sites with monumental among sites within and between the valleys of the Norte
architecture and by the labor involved, including the weaving Chico. Fine-grained absolute dating is probably the first and
of shicra bags that were used to haul rock, and the volume of foremost of these challenges. A number of sites, including
construction. Building efforts included retaining walls subse- Caral (Shady et al., 2001; Shady, 2004, 2009), Porvenir,
quently filled with shicra to create platforms that were then Caballete, and Huaricanga, all have suites of radiocarbon
topped with clay floors. The excavation of circular courts was dates that span a millennium or more. In addition to
a further investment of labor that took place at sites across the exceptionally long occupations, these sites are quite large
region, with up to three circular courts identified at a single and have virtually continuous remodeling, expansion, and
site. Walls and floors were finished with thick layers of clay
architectural reorganization. As a result, thus far, it is
plaster tempered with plant fiber. Finer plaster was sometimes
impossible to firmly establish periods of absolute contempo-
tinted with colored clay. Surfaces were frequently remodeled
raneity or particular periods of building or growth at
with additional layers of plaster or by altering the features of a
individual sites. Such analyses will take considerably more
particular space, such as the series of remodeled structures
work and many more radiocarbon dates at all these sites. A
identified at Huaricanga (Fig. 44).
second challenge is the lack of pottery and other specialized
At the sites tested, ample evidence of residence was
craft goods that might be useful in determining centers of
identified by the presence of small-scale architecture, cooking
production, distribution of luxury goods, or control of specific
facilities, and domestic trash. Trash deposits varied in
resources. Late Archaic sites, compared to any subsequent
thickness from a few centimeters to almost a meter deep,
and blackened rocks from cooking with heated stones, or time period in Andean prehistory, have a very limited and
pachamanca cooking, were recovered from midden deposits. comparatively simple material culture. In other time periods
Small shell fragments, charcoal, and lithics dispersed on the and other areas, patterns in the manufacture and distribution
surface proved to indicate areas of short-term residential of material culture are primary vehicles for determining
structures, such as windbreaks, lean-tos, and post and brush patterns of intersite interaction and organization. Finally,
structures associated by hearths and midden deposits despite there has been significant destruction of many of the Late
the absence of surface irregularities or remains of structures Archaic sites. In the Fortaleza Valley, this is most notable in
(Wulffen 2009). the site of Shaura, which was almost completely destroyed in
In contrast to the evidence of labor investment in the latter decades of the 20th century. Unfortunately, almost
construction, little evidence of craft specialization was all the sites in the Pativilca Valley have been subjected to
recovered apart from weaving of a variety of wild plant fibers major damage (Creamer et al., 2007), as have most of the Late
into shicra bags, mats, nets, string, and rope. Specialized Archaic sites recorded in the Huaura Valley (Nelson & Ruiz,
production areas for making woven goods were identified in at 2005). This tragic loss of resources vital to telling the story of
least two places at Caballete, and a thick layer of fiber was the early emergence of a distinctive Andean society means that
identified at Cerro Blanco 2. Small lengths of cloth, bags, and there will always be some gaps in the story; however, careful
belts were created by looping and twining plant fibers and, to and judicious excavation of the sites that have not been
a much lesser extent, cotton, with very little weaving that destroyed will allow us to fill in the primary elements of the
involved separate warp and weft threads. Even that was finger cultural development.
weaving, with no evidence of the use of looms until the Initial The testing in the Fortaleza Valley further extends our
Period (Doyon-Bernard, 1990). understanding of the Norte Chico region as a major center for
The testing done in the Fortaleza Valley in 2003 and 2004 the ‘‘pristine’’ emergence of complex polities in the Andean
expands on the broad base of data provided by other region (Fried, 1967; Haas, 1982; Shady & Leyva, 2003).
researchers working in the Norte Chico region over the past Pristine in this sense is not used to mean development without
50 years or more. Collectively, archaeological work in the outside contacts but rather indicates the development of
region now paints an extraordinary picture of cultural complexity without the influence of already existing complex

CREAMER ET AL: FORTALEZA VALLEY, PERU 61

polities. In this one small cluster of coastal valleys at the Field Crew for 2003 and 2004 Field Seasons
beginning of the third millennium BC several sites appeared Mario Advincula
both on the coast and inland. Although we know little of the Carmela Alarcon Micaela Alvarez
founding of any of these sites, we do know that they began Carlos Escobar Jesus Holguin
building large-scale communal architecture very early in the Felipe Libora
millennium. By the end of the millennium, there were six to Kit Nelson
eight sites in each of the four valleys. These sites provide a James H. Duncan Sr.
window to understanding how the first leaders emerged Rebecca Osborn
endogenously in incipient centralized populations and how Rebecca Bria Dan Corkill
leaders began to exercise power over their respective popula- Lillian Haas Lyra Haas
tions (Haas & Creamer, 2004, 2006). For this, we need to Guillermo Galvez Eugenio Galvez
understand the role of economics, religion, and intersite Santos Bardales Maximino Bardales
relationships in the origin and development of power relations Hernan Gamarra
in this extraordinary early political system. Luis Verastegui
The comparatively dense cluster of large sites in the Norte Flash
Chico, all with large-scale communal architecture, requires a
serious reconsideration of the early cultural history of the Barranca Crew for 2003 and 2004 Field Seasons
Andean region. In a region where there were ancient patterns Gaim Byrne Esperanza Natividad Carpio
of mobility between highlands and coasts over the course of Miguel Loo Lili
1000 years, it seems highly likely that knowledge of the Norte Fernando Gamarra Fiorela Byrne
Chico cultural phenomenon would have been widespread. Meri
What effect or influence this ‘‘precocious’’ (Shady, 2003b) cultural
development would have had on subsequent evolution of the
Andean political landscape still remains to be defined. At the
same time, however, the Late Archaic Norte Chico development Literature Cited
must be taken into consideration in assessing the beginnings of
cultural complexity in all parts of the Andean region after about ASENCIOS, G. 2009. Investigaciones de las shicras en el sitio
2500 BC. This is to argue not that the Norte Chico represented a preceramico de Cerro Lampay. Licenciatura Thesis, Faculty of
kind of ‘‘Mother Culture’’ for the Andes but rather that the Social Science, Universidad Nacional Mayor de San Marcos, Lima.
precocious development in this area was an integral part of the AUTHIER, M. 2012. Monument, memory, and exchange at Huar-
ancient or deep history of the region as a whole. icanga, Peru. Paper presented at the 77th annual meeting of the
Society for American Archaeology, Nashville, Tennessee.
BAZAN PEREZ, A. 2012. Reocupaciones de complejos monumentales
del period Arcaico Tardio en el valle de Fortaleza, Costa Nor
Central del Peru. Licenciatura Thesis, Faculty of Social Science,
Acknowledgments Universidad Nacional Mayor de San Marcos, Lima.
BIRD, J. B., J. HYSLOP, AND M. D. SKINNER. 1985. The preceramic
We thank the directors of the community organizations and excavations at the Huaca Prieta, Chicama Valley, Peru. Anthro-
pological Papers of the American Museum of Natural History, 62:
the civil authorities of the communities that neighbor the sites 1–294.
where we worked in 2003, including Cerro Blanco, Huar- BONNIER, E. 1997. Prehispanic architecture and civilization in the
icanga, and Shaura. We recognize the support of the mayor of Andes, pp. 120–144. In Bonnier, E. and H. Bischof, eds.,
the Municipal District of Paramonga, Ing. Enrique Respicio Preceramic Architecture in the Andes: The Mito Tradition. Reiss
Lopez; the mayor of the Municipality of Pativilca, Dr. Arturo Museum, Mannheim, Germany.
Espinosa; and the town council member responsible for BURGER, R., AND L. SALAZAR-BURGER. 1986. Early organizational
cultural events in Pativilca, Katy Ly Albujar, all of whom diversity in the Peruvian highlands: Huaricoto and Kotosh, pp. 65–
82. In Matos, R. and S. Turpin, eds., Andean Archaeology. UCLA
displayed their affection for the region through their helpful Institute of Archaeology, Los Angeles.
collaboration. CARDENAS, M. 1979. A chronology of the use of marine resources in
Crucial to the success of the project was the participation of ancient Peru. Boletin Arqueologica, Pontifica Universidad Catolica
all the members of PANC who so ably forged an excellent del Peru, 104: 1–30.
working group and showed their commitment to the objectives CARNEIRO, R. 1970. A theory of the origin of the state. Science, 189:
of the project. Each and every person worked their hardest to 733–738.
allow us to reach the goals set for this investigation. Two CREAMER, W. 2007. Archaeological investigation of Late Archaic sites
(3000–1800 BC) in the Pativilca Valley, Peru. Fieldiana: Anthro-
anonymous reviewers made detailed comments that helped us
pology, n.s., 40: 1–79.
improve the manuscript in many ways. We take full
CREAMER, W., AND J. HAAS. 2008. Monumental architecture
responsibility, however, for any errors. as a process of emergence: An example from the Norte Chico,
We gratefully acknowledge the support of the National Peru. Paper presented at the 73rd annual meeting of the Society for
Science Foundation Collaborative Research Grants to Haas American Archaeology, Vancouver, B.C.
and Creamer, the Field Museum, Northern Illinois University CREAMER, W., A. RUIZ, AND J. HAAS. 2006. Late archaic regional
(NIU), and especially the NIU Center for Latino and Latin organization in the Norte Chico, Peru. Paper presented at the 71st annual
meeting of the Society for American Archaeology, San Juan, Puerto Rico.
American Studies. Friends who have supported our research
DAGGETT, R. 1987. Toward the development of the state on the north
include James H. Duncan Sr., James and Susan Hannigan, Jeri central coast of Peru, pp. 70–82. In Haas, J., T. Pozorski and S.
Alles and David Hanley, Donald and Marianne Tadish- Pozorski, eds., The origins and development of the Andean state.
Marzec, and others. Thank you all. Cambridge University Press, Cambridge.

62 FIELDIANA: ANTHROPOLOGY
DOYON-BERNARD, S. J. 1990. From twining to triple cloth: experi- NELSON, K., AND A. RUIZ. 2005. Proyecto de investigación arqueo-
mentation and innovation in ancient Peruvian weaving (ca. 5000– lógica: Valle de Huaura, Perú. Informe final. Instituto Nacional de
400 B.C.). American Antiquity, 55: 68–87. Cultura, Lima.
ENGEL, F. 1966. Le complexe précéramique d’el Paraiso (Pérou). PERALES, M. F. 2006. Proyecto de investigacion: Reconocimiento
Journal de la Société des Américanistes, LV-1: 43–95. arqueologico en el valle bajo de Pativilca, Lima-Peru. Informe final
———. 1987. De las begonias al maı́z: Vida y producción en le Perú 2006. Instituto Nacional de Cultura, Lima.
antiguo. Universidad Agraria La Molina, Ediagraria, Lima, Peru. ———. 2007. Proyecto de investigacion: Reconocimiento arqueolo-
FELDMAN, R. 1980. Aspero, Peru: Architecture, subsistence economy gico en el Valle de Fortaleza, Lima-Ancash, Peru. Informe final
and other artifacts of a Preceramic maritime chiefdom. Unpublished 2007. Instituto Nacional de Cultura, Lima.
PhD dissertation, Harvard University, Cambridge, Massachusetts. POZORSKI, S. 1990. Reexamining the critical Preceramic/Ceramic
———. 2009. Talking dogs and new clothes, or the maritime Period transition: New data from coastal Peru. American
foundations hypothesis revisited, pp. 89–98. In Marcus, J. and P. Anthropologist, 92: 481–491.
R. Williams, eds., Andean Civilization. Cotsen Institute of POZORSKI, S., AND T. POZORSKI. 1987. Early Settlement and Subsistence
Archaeology, Los Angeles. in the Casma Valley, Peru. University of Iowa Press, Iowa City.
FRIED, M. 1967. The Evolution of Political Society: An Essay in QUILTER, J. 1985. Architecture and chronology at El Paraiso, Peru.
Political Anthropology. Random House, New York. Journal of Field Archaeology, 12: 179–197.
FUCHS, P., R. PATZSCHKE, G. YENQUE, AND J. BRICENO. 2008. Del ———. 1992. To fish in the afternoon: Beyond subsistence economies
Arcaico Tardio al Formativo Temprano: Las investigaciones en in the study of early Andean civilization. Andean Past, 3: 111–125.
Sechin Bajo, valle de Casma. In Kaulicke, P. and Y. Onuki, eds., El RAYMOND, J. S. 1981. The maritime foundations of Andean
Periodo Formativo: Enfoques y evidencias recientes. Segunda civilization: A reconsideration of the evidence. American Antiquity,
Parte. Boletin de Arqueologia PUCP, 10: 111–166. 46: 806–821.
FUNG PINEDA, R. 1988. The Late Preceramic and Initial Period, pp. RENFREW, C. 1986. Peer-polity interaction and sociopolitical change,
67–96. In Keatinge, R., ed., Peruvian Prehistory. Cambridge pp. 1–18. In Renfrew, C. and J. Cherry, eds., Peer-Polity
University Press, Cambridge. Interaction and Sociopolitical Change. Cambridge University
———. 2004. Reseña del salvamento arqueológico en Bandurria. Press, Cambridge.
Perı́odo Precerámico Tardı́o, pp. 325–336. In Quehaceres de la RUIZ, A., W. C REAMER, AND J. HAAS. 2008. Investigaciones
arqueologı́a Peruana—Rosa Fung, compilación de escritos. Edi- arqueológicos de sitios Preceramicos en el Valle de Pativilca, Norte
ción del Museo de Arqueologı́a y Antropologı́a, Lima. Chico, Peru. Instituto Cultural del Norte Chico, Barranca, Peru.
GRIEDER, T., A. BUENO MENDOZA., C. E. SMITH, JR. AND R. M. RUTHERFORD, A. 2008. Space and Landscape in the Norte Chico
MALINA., EDS. 1988. La Galgada Peru: A preceramic culture in Region, Peru: An Analysis of Socio-Political Organization through
transition. University of Texas Press, Austin. Monumental Architecture. M.A. thesis, Anthropology Depart-
HAAS, J. 1982. The Evolution of the Prehistoric State. Columbia ment, Northern Illinois University
University Press, New York. SANDWEISS, D., J. RICHARDSON, III., E. REITZ, H. ROLLINS, AND K.
HAAS, J., AND W. CREAMER. 2004. The Late Archaic in Andean MAASCH. 1996. Geoarchaeological evidence from Peru for a
5000 years BP onset of El Niño. Science, 273: 1531–1533.
Prehistory: 3000–1800 BC. pp. 35–50. In Silverman H, ed., Andean
Archaeology, Blackwell Global Archaeology Series. Blackwell SANDWEISS, D., R. SHADY SOLIS., M. E. MOSELEY, D. K. KEEFER, AND
Publishers, Oxford. C. R. ORTLOFF. 2009. Environmental change and economic
development in coastal Peru between 5800 and 3600 years ago.
———. 2006. The crucible of Andean civilization. Current Anthro-
Proceedings of the National Academy of Sciences, 106: 1359–
pology, 47: 745–775. 1363.
———. 2010. Late archaic temples in the Norte Chico region of Peru. SCHIFFER, M. 1986. Radiocarbon dating and the ‘‘old wood’’
Paper presented at the 38th annual meeting of the Midwest problem: The case of the Hohokam chronology. Journal of
Conference on Andean and Amazonian Archaeology and Ethno- Archaeological Science, 13: 13–30.
history, IPFW, Fort Wayne, Indiana.
SHADY, R. 1997. La ciudad Sagrada de Caral-Supe en los albores de la
———. In press. Intrusiveness at Late Archaic (3000–1800 BCE) sites in civilizacı́on en el Perú. Universidad Nacional Mayor de San
the Norte Chico region of Peru. In, Chicoine, D. and R. Mann, eds., Marcos, Lima.
Archaeologies of Intrusiveness. University of Colorado Press, Boulder.
———. 2003a. Reprinted. Caral-Supe: La civilización más antigua
HAAS, J., W. CREAMER, AND A. RUIZ. 2004. Dating the Late Archaic del Perú y América,, 335–340, In Shady, R. and C. Leyva, eds., La
occupation of the Norte Chico region in Peru. Nature, 432: 1020– ciudad sagrada del Caral-Supe: Los orı́genes de la civilización
1023. andina y la formación del estado prı́stino en el antiguo Perú.
ISHIDA, E. 1960. Andes: The report of the University of Tokyo Instituto Nacional de Cultura, Lima, Peru. Originally printed 2002.
Scientific Expedition to the Andes in 1958. Andean Institute, Cultura Peruana, 3: 60–68.
University of Tokyo, Tokyo. ———. 2003b. Reprinted. Caral-Supe y la costa norcentral del Perú:
IZUMI, S., AND T. SONO. 1963. Andes 2. Excavations at Kotosh, Peru, La cuna de la civilización y la formación del estado prı́stino,
1960. University of Tokyo Press, Tokyo. 139–146, In Shady, R. and C. Leyva, eds., La ciudad sagrada del
IZUMI, S., AND K. TERADA. 1972. Andes 4: Excavations at Kotosh, Caral-Supe: Los orı́genes de la civilización andina y la formación
Peru, 1963 and 1966. University of Tokyo Press, Tokyo. del estado prı́stino en el antiguo Perú. Instituto Nacional de
Cultura, Lima, Peru. Originally printed 2000. Historia de la
KIGOSHI, K., D. LIN, AND K. ENDO. 1962. Gakushuin natural cultural peruana, 1: 45–87, Fondo Editorial del Congreso del Perú,
radiocarbon measurements I. Radiocarbon, 4: 84–94. Lima.
KOSOK, P. 1965. Land, Life and Water in ancient Peru. Long Island ———. 2004. Caral: La ciudad del fuego sagrado. Centura SAB,
University Press, New York. Lima.
MOSELEY, M. E. 1975. Maritime Foundations of Andean Civilization. ———. 2006. La civilización Caral: sistema social y manejo del
Cummings, Menlo Park, California. territorio y sus recursos: su trascendencia en el proceso cultural
———. 1992. Maritime Foundations and Multilinear Evolution: andino. Boletı́n de Arqueologia PUCP, 10: 59–89.
Retrospect and Prospect. Andean Past, 3: 5–42. ———. 2009. Caral-Supe y su entorno natural y social en los orı́genes
MOSELEY, M. E., AND L. K. BARRETT. 1969. Change in Preceramic twined de la civilización, pp. 99–120. In Marcus, J. and P. R. Williams,
textiles from the central Peruvian coast. American Antiquity, 34: eds., Andean Civilization. Cotsen Institute of Archaeology, Los
162–165. Angeles.
MOSELEY, M. E., AND G. R. WILLEY. 1973. Aspero, Peru: A SHADY, R., J. HAAS, AND W. CREAMER. 2001. Dating Caral, a
reexamination of the site and its implications. American Antiquity, Preceramic urban center in the Supe Valley on the central coast of
38: 452–468. Peru. Science, 292: 723–726.

CREAMER ET AL: FORTALEZA VALLEY, PERU 63

SHADY, R. AND C. LEYVA., EDS. 2003. La ciudad sagrada de Caral- WILLEY, G. R., AND J. M. CORBETT. 1954. Early Ancon and Early
Supe: Los orı́genes de la civilización andina y la formación del Supe Culture: Chavin Horizon Sites of the Central Peruvian Coast.
estado prı́stino en el antiguo Perú. Instituto Nacional de Cultura, Columbia University Press, New York.
Lima. WILLIAMS, C., AND M. MERINO. 1979. Inventario, catastro y
STANISH, C. 2006. Comment on: The crucible of Andean civilization. delimitación del patrimonio arqueológico del Valley de Supe.
Current Anthropology, 47: 763–764. Report submitted to the Instituto Nacional de Cultura, Lima.
VEGA-CENTENO, R. 2005. Ritual and architecture in a context of WILSON, D. J. 1981. Of maize and men: A critique of the maritime
emergent complexity: A perspective from Cerro Lampay, a Late hypothesis of state origins on the coast of Peru. American
Archaic site in the central Andes. Unpublished PhD dissertation. Anthropologist, 83: 93–120.
University of Arizona, Tucson. ———. 1987. Reconstructing patterns of early warfare in the Lower
———. 2007. Construction, labor organization, and feasting during Santa Valley: New data on the role of conflict in the origins of
the Late Archaic Period in the Central Andes. Journal of complex north coast society, pp. 56–69. In Haas, J., T. Pozorski and
Anthropological Archaeology, 26: 150–171. S. Pozorski, eds., The Origins and Development of the Andean
———. 2010. Cerro Lampay: architectural design and human State. Cambridge University Press, Cambridge.
interaction in the north central coast of Peru. Latin American WULFFEN, J. 2009. Two test excavations at Caballete, Norte Chico,
Antiquity, 21: 115–145. Peru. MA thesis, Northern Illinois University, Dekalb.
VEGA-CENTENO, R., L. F. VILLACORTA, L. E. CÁCERES, AND G. ZECHENTER, E. 1988. Subsistence strategies in the Supe Valley of the
MARCONE. 1998. Arquitectura monumental temprana en el valle Peruvian central coast during the Complex Preceramic and Initial Periods.
medio de Fortaleza. Boletı́n de Arqueologia PUCP, 2: 219–238. Unpublished PhD dissertation, University of California, Los Angeles.

TABLE 1. Andean chronology.

Date Period (Rowe) Stage (Lumbreras)

AD 1440–1532 Late Horizon Inca Empire
AD 1100–1400 Late Intermediate Period regional states and kingdoms
AD 600–1100 Middle Horizon Huari Empire
200 BC–AD 600 Early Intermediate Period regional cultures
1000–200 BC Early Horizon Middle and Late Formative
1700–1000 BC Initial Period Early Formative
3000–1800 BC Late Preceramic Late Archaic
4500–3000 BC Middle Preceramic Middle Archaic
6000–4500 BC Early Preceramic Early Archaic
12,000–6000 BC Early Preceramic hunter-gatherers

64 FIELDIANA: ANTHROPOLOGY
 TABLE 2. Architecture at Fortaleza Valley sites.

Site/Sector Size (m) Height (m) Comments

Porvenir
A 86 3 61 10
Associated circular court 30 n.a.
B 55 3 49 8
Associated circular court 20 n.a.
C 51 3 29 5
D 47 3 45 7
E (north) 28 3 33 4
E (south) 43 3 41 6
F 61 3 49 10
G 102 3 59 8
H concentrated shell and dark soil
I walled enclosure
J looted cemetery
K looted cemetery
Caballete
A 125 3 54 17
Associated circular court 15
B 41 3 39 5
C 59 3 55 7
Associated circular court 25
D 120 3 47 7
E 69 3 53 7
Associated circular court undetermined
F 200 3 200 area bounded by structures/sectors
A, E, and canal
Cerro Blanco 1
Sector A 34 3 33 5
Sector A, South irregular 3
Sector B, 29 3 20 5
Sector C 73 3 42 5
Cerro Blanco 2
A 120 3 68 4
Associated circular court 20
Huaricanga
A1
B1 47 3 41 4
B2 29 3 29 3
C1 285 3 100 15
C2 150 3 41 7
C3 53 3 38 4
Shaura
A2 67 3 52 10
Associated circular court 24

TABLE 3. Radiocarbon dating from Caballete comparing annual plant fiber (APF) and charcoal.

Sector Unit Level/Layer Calibrated RCYBP Material Lab

A 4 pit fill 2320 3870 APF ISGS-5721
A 4 pit fill 2700 4110 charcoal Beta-199777
C 2 11 2120 3720 APF GEO-31584
C 2 11 2640 4040 charcoal GEO-31587
C 2 12 2570 4020 APF Beta-199776
C 2 12 2960 4300 charcoal Beta-199058

CREAMER ET AL: FORTALEZA VALLEY, PERU 65

66
TABLE 4. Radiocarbon dates from Porvenir.

Level/
Sample Provenience Layer Material Lab Cal BC RCYBP ± 13c/12c 68.2—1s range 95.4—2s range
a Sector A, Profile 2 C mixed plant fibers GX-31582 1750 3430 70 210.7 1876–1842 (17.0); 1820–1797 (10.6); 1915–1604 (95.2); 1588–1534 (4.8)
1781–1662 (66.9); 1652–1639 (5.5)
b Sector A, Profile 2 D mixed plant fibers GX-30510 2010 3630 70 211.5 2129–2088 (18.9); 2047–1899 (81.1) 2200–2158 (4.2); 2155–1869 (89.9);
1846–1809 (3.3); 1805–1775 (2.5)
c Sector A, Profile 2 D fiber bag ISGS-5513 2730 4160 70 213.1 2877–2834 (20.6); 2817–2664 (75.9); 2899–2571 (99.2); 2512–2504 (0.8)
2646–2638 (3.5)
d Sector A, TU 3 7 mixed plant fibers Beta-183323 2370 3890 40 211.8 2462–2339 (95.2); 2321–2320 (1.4); 2473–2278 (95.5); 2251–2229 (3.3);
2314–2310 (3.4) 2221–2210 (1.3)
e Sector A, TU 3 7 fiber bag ISGS-5520 2700 4110 70 214.3 2862–2807 (24.4); 2758–2718 (16.6); 2880–2558 (92.0); 2554–2550 (0.4);
2706–2577 (59.0) 2537–2491 (7.5)
f Sector F, Profile 1 mixed plant fibers GX-31586 1830 3500 70 210.3 1916–1741 (100.0) 2021–1993 (2.6); 1982–1662 (96.5);
1652–1640 (0.9)
g Sector F, Profile 1 mixed plant fibers ISGS-5512 2100 3710 70 212.5 2202–2019 (95.2); 1994–1981 (4.8) 2332–2327 (0.2); 2299–1900 (99.8)
h Sector F, Profile 1 fiber bag GX-30630 2210 3780 60 211.7 2297–2132 (90.5); 2084–2057 (9.5) 2457–2418 (3.5); 2407–2375 (3.6);
2367–2361 (0.4); 2351–2032 (92.5)
i Sector F, Profile 1 mixed plant fibers ISGS-5713 2440 3950 80 211.3 2570–2514 (24.2); 2502–2338 (71.5); 2837–2815 (1.3); 2673–2200 (98.6);
2322–2309 (4.3) 2157–2155 (0.1)
j Sector F, Profile 1 mixed plant fibers Beta-183324 3750 4930 70 212.2 3777–3647 (100.0) 3943–3855 (12.5); 3846–3833 (1.0);
3823–3632 (84.7); 3558–3538 (1.9)
k Sector H, TU 1 4 mixed plant fibers GX-30509 1250 3040 80 213.8 1411–1208 (96.9); 1200–1196 (1.4); 1489–1481 (0.5); 1455–1048 (99.5)
1139–1135 (1.7)
l Sector H, TU 1 3 fiber bag ISGS-5735 2580 4020 70 212.4 2851–2813 (9.8); 2743–2727 (3.3); 2872–2800 (10.4); 2793–2785 (0.6);
2695–2684 (2.5); 2680–2458 (84.4) 2780–2331 (86.8); 2328–2299 (2.2)
m Sector H, TU 2 5 shicra Beta-183325 2330 3850 40 212 2451–2446 (2.9); 2436–2420 (8.5); 2461–2205 (100.0)
2405–2378 (15.6); 2350–2276 (50.9);
2253–2228 (15.0); 2223–2209 (7.2)
n Sector H, TU 2 12 charcoal Beta-199060 2840 4250 40 226.7 2911–2870 (80.8); 2802–2778 (19.2) 2926–2849 (65.7); 2813–2739 (27.1);
2730–2693 (6.7); 2687–2679 (0.5)

FIELDIANA: ANTHROPOLOGY
 TABLE 4B

TABLE 5. Porvenir, Sector A, Profile 2.

Phases of occupation (Fig. 4; Table 4a–c)

Phase Layers Description Dated samples
I B Dark soil with small quantities of cultural material.
II B9, C Included a small feature containing charcoal, charred plant remains, ashy soil, 1750 Cal BC (GX-31582), plant
some fragments of shicra (Layer B9). Contiguous with a use surface of fibers (Layer C, Table 4a)
compacted earth that was finer in texture and included more clay than the
adjacent layers. Use surface and feature lay over fill with plant fibers (Layer C).
III D Thick layer of shicra construction fill. 2010 Cal BC (GX-30510), plant fibers;
2730 Cal BC (ISGS-5513), fragments
of a shicra bag (Table 4b, c)

CREAMER ET AL: FORTALEZA VALLEY, PERU 67

 TABLE 6. Porvenir, Sector A, Test Unit 3.

Excavation (Fig. 6)
Level Thickness Description Contents
0 15 Fine, loose soil, some gravel, patches of caliche. Shell, lithics.
1 15 Dry light brown soil and sand, angular rock. Midden, including Charcoal, bone, whole mollusk valves, small hank
small fragments of clay mortar at the base of the level. of hair.
2 15 Loose soil, possible remains of remodeling activities, fragments Shell, lithics, charcoal, botanical remains, bone.
of clay, plaster, plant fiber, burned rock.
3 15 Loose soil, fragments of clay, plaster, plant fiber, burned rock. Shell, lithics, charcoal, botanical remains, bone.
4 15 Dry granular clay and bits of clay mortar mixed with light Low-density cultural materials, burned rock, shell,
brown soil. Similar to Levels 2 and 3. charcoal, botanical remains, textile fragments, hair.
Woven, rock-filled shicra bags were in the S corner.
5 6 Soil, large and small angular rock, gravel, fragments of clay mortar, Shell, lithics, charcoal, botanical remains, bone.
likely construction or remodeling debris. An irregular use surface
over the S part of the unit. Partial shicra bag on E side of use surface.
6 7 Only excavated on the S side of the unit (1 3 1 m). Two use Shell, lithics, botanical remains, bone.
surfaces, the upper a remodeling of the lower.
7 15 Fine gray soil with angular rock, a base layer below the use surface. Shell, lithics, botanical remains, charcoal, hair.

Stratigraphy
Layer Levels Description of stratigraphy Characteristics of layer
A 1–5 Construction fill or possibly trash Low-density cultural material. Rock, gravel,
fragments of clay mortar.
B 3–6 Irregular clay layer/use surface in upper portion of Layer B, Dark brown fill with gravel and rocks up to 20 cm
most clearly visible in N half of unit. S portion disturbed across.
in the past.
C 6 Use surface of wet clay on angular rock. Use surface identified Plant fiber was mixed with the clay of the use
in two levels; half the surface was displaced toward the E surface.
due to subsidence, water, or seismic action. Thus, part of the
use surface was identified in Level 5 on the W side of the test
unit and in Level 6 on the E side.
D 7 Use surface at base of shallow layer. Layer C was not a Angular rock roughly covered with clay.
remodeling of Layer D. D was oriented differently than C.
E 7 Just beneath the lowest use surface. Soil changed from gray to distinctive red-brown,
culturally sterile.

Phases of occupation (Table 4a, b)

Phase Layers Description Dated samples
I A Natural deposition postdates use of this locality; gravelly dirt mixed
with rock.
II B Use surface of yellowish clay over irregular layer of darker fill,
some cultural material. A pit was excavated through this use
surface in the S portion of the unit.
III C Clearly defined use surface, layer of clay over layer of rock.
IV D, E Fill between use surfaces, also the lowest use surface. Lower Shicra fragment from Layer D was dated 2700 Cal BC
surface oriented differently than use surface above it; did not (ISGS 5520). Sample of mixed plant fibers extracted
extend across entire excavation unit. Lower use surface is from crushed fragment of use surface below Layer
oldest in unit, shows construction directly on subsoil layer. D yielded date of 2370 Cal BC (Beta183323)
(Table 4d, e).

68 FIELDIANA: ANTHROPOLOGY
 TABLE 7. Porvenir, Sector F, Profile 1.

Stratigraphy (Fig. 6)
Layer Description of stratigraphy Characteristics of layer
A Material above Floor 1. Loose soil.
B Floor 1. Fine clay.
C Subfloor prepared surface below Floor 1. Gray clay over small rocks.
D Construction fill. Angular rock and shicra.
E Lens of loose soil, indicates period of disuse. Loose brown soil.
F Clay floor with one remodeling (Floor 2). Two clay layers.
G Wall 1. Wall of angular rock, plaster along upper surface and former
mound exterior—side facing away from cleared slope.
H Remodeling of the space created by Walls 1 and 2. Filling of Shicra fill of large and medium angular rock.
a room or plaza area with shicra.
I Floor over Wall 2. Fine clay.
J Wall 2. Angular rock and shicra.
K Collapse associated with looting. Loose brown soil.
L Pocket of burned material. Blackened cane fragments and ash, recent airborne deposit from
adjacent fields.
M Construction fill comprising one phase of mound construction. Large angular rock and shicra.
N Postlooting erosion and collapse of sides of dug-out area. Loose brown soil, angular rock.

Phases of occupation (Table 4f–j)

Phase Layers Description Dated samples
I B, C Occupation of Floor 1, final use of mound. Material above
was product of collapse and wind-blown deposition.
II D Period between use of Floor 2 and construction of Floor 1.
III E, F Construction and use of Floor 2 after Wall 1 was built;
may have extended across the top of Wall 1.
IV G, H Construction of Wall 1 as a retaining wall and the Mixed plant fibers from the clay plaster on Wall 1 yielded
placing of fill behind it. a date of 3750 Cal BC (Beta-183324) (Table 4j).
V I, J Construction of Wall 2, including plastering the wall Wall 2 yielded a shicra bag fragment that was dated 2210
and placing Floor 3 on top of Wall 2. Cal BC (GX-30630) (Table 4h).
VI M Construction fill below Wall 2. Three samples of plant fibers yielded dates of 2440 (ISGS-
5713), 2100 (ISGS-5512), and 1830 Cal BC (GX-
31586) (Table 4f, g, i).

TABLE 8. Porvenir, Sector H, Test Unit 1.

Excavation (Fig. 7)
Level Thickness Description Contents
1 0–30 Sandy aeolian soil, a few small angular rocks, up to 10 cm Lithics, shell fragments, a few small fragments of textiles,
diameter. Lens of ash in NE corner. charcoal.
2 5–16 Silty sandy soil with small rocks. Lens of burned material Lithics, shell, charcoal.
in the SW corner produced most shell fragments.
3 16 Loose soil with gravel on E half and more clayey soil, Lithics, shell, botanical remains, textile fragment.
plant remains, ash, shell in W half. Concentration of
rocks on W side was part of a rough wall, angular rock
in clay mortar, uneven surfaces on both faces.
4 16–24 Mixed small and medium rocks in soil with clay. Area of Shell, charcoal, lithic, plant remains, textile fragments.
collapsed wall in the center, including large rocks and
shicra. Caliche around wall made it difficult to excavate.
5 1–10 Rock, gravel, caliche, difficult to excavate. Center of the unit contained shell, lenses of ash, lithic
fragments, botanical remains, textile fragments.
6 0–14 Rocky soil on E side, caliche throughout. Small quantity of textile, lithics, shell.

Phases of occupation (Table 4k, l)

Phase Levels Description Dated samples
I 1–3 Former ground surface covered by postoccupational fine aeolian
deposits. Distinction between surface–former ground surface
from artifact content and soil texture. No structures/activity
areas.
II 3–6 Construction and occupation of a low platform built on sterile, Shicra bag fragment from Level 3 dated 2580 (ISGS-5735)
gravelly soil, perhaps incorporating a natural rise in the Cal BC and mixed plant fibers in Level 4 1250 Cal BC
landscape indicated by gravelly material in NW corner. (GX-30509) (Table 4l, k).
Wall 1, unplastered retaining wall. Fill on E side of wall
different from W side. Fill included large rocks and shicra
that leveled the platform approximately 50 cm high. Clay
surface identified in Levels 3 and 4 formed platform surface.

CREAMER ET AL: FORTALEZA VALLEY, PERU 69

 TABLE 9. Porvenir, Sector H, Test Unit 2.

Excavation (Fig. 8)
Level Thickness Description Contents
1 0–12 Sandy soil, patches of caliche. No artifacts.
2 1–25 Loose soil mixed with a large quantity of rock. Shicra, shell, lithics, macrobotanical remains.
3 14–18 Compact soil on SW, possibly associated with a Shell, botanical remains, lithics, charcoal, bone.
feature in SW corner, rock in the rest of the level.
4 11–15 Angular rock throughout. E portion is darker and contains Shell, lithics, botanical remains, bone, charcoal.
more botanical remains than W. Fine-textured soil.
5 14–17 Fine, dusty, loose soil; a substantial quantity of rock, Shell, charcoal, bone, botanical remains.
lithics, shicra fragments.
6 15 Fine, dusty, loose soil, small rocks. A large rock in Diverse organic materials, including string, shicra, leaves,
this level was a wall that crossed N to S, 40–50 cm high, seeds; also clay mortar, lithics, charcoal, bone.
Levels 6–12 (Fig. 9).
7 6–10 Ash on the W side, with shell, clay mortar, a few lithics. Abundant textile, fish, charcoal, seeds.
8 4–7 Material from W side of Wall 1 outside of the wall, as Quantity of ash, with charcoal, shell, seeds, cobbles, burned
the rock in Wall 1 was oriented to make a flat W face. rock, fish bone, textiles.
9 15 E side of Wall 1. Rock in a thin layer of clay and plastered Shell, lithics, charcoal, bone.
mortar fragments, fill behind Wall 1, a retaining wall.
10 2 E side of Wall 1, below Level 9 extending to use surface of Shell, lithics, bone, charcoal.
mortar and small rocks.
11 13–17 Below Level 8 on the W side of Wall 1 extending to a use Lithics, charcoal, fauna, botanical remains, shell, coprolite.
surface of mortar and small rocks.
12 26 Wall 1, built on the use surface. Wall was removed, revealing Botanical remains, lithics, shell, charcoal, bone.
black ash, a use surface that went under the wall.
13 20–30 W half of the unit, size decreased due to the quantity of A quantity of charcoal, botanical fragments, burned rock,
material, and the depth, -130 cm bd. Ashy texture, dark color. shell, bone.
14 12–40 To the end of the dark layer in the W half of the Dense ash, charcoal, large shells, shell fragments.
unit. Many rocks mixed with midden.
15 18–21 A depression or hole below the ash layer. Depression/feature filled with ash, organic remains, shell,
charcoal.
16 15–19 Color change from dark brown to light brown gravelly soil.
17 21 Light brown soil, increasingly gravelly through the level. Mollusk valves up to 10 cm long at top of level, no remains
by midlevel. Few lithics, bone, botanical remains.

Stratigraphy
Layer Includes Description of stratigraphy Characteristics of layer
A 1–6 Remodeling debris or remains of collapsed walls. Cultural material in loose, dusty, fine, loose tan soil; occasionally
compact, many rocks.
B 7, 8 Shicra construction.
C 9–12 Levels 9 and 10 were excavated on E side of unit; Large rocklike Layer A, rubble from retaining wall crossed the unit.
Levels 11 and 12 on W side.
D 13 Floor below Wall 1. Distinct dark layer below Wall 1.
E 13 Midden below floor. Rocky section separates upper/lower midden; ash, charcoal,
botanical.
F 13–15 Rocky layer and lower component of midden.
G 15, 16 Former ground surface below midden. Decreasing frequency of artifacts, little charcoal or burned material.
H 16, 17 Gravelly layer. Cultural materials few to none.

Phases of occupation (Table 4m, n)

Phase Includes Description Dated samples
I 1–6 Final occupation in this part of the site. Homogeneous, not very Mixed plant fibers from shicra in Level 5 dated
dense trash over disused platform. 2330 Cal (Beta-183325) BC (Table 4m).
II 7–11 Wall 1 built, formed side of platform 50 cm high. Only the platform in E
half of unit was uncovered. Possible base for residential structures or for
ceremonial activities on Sector A mound. Upper Phase II includes shicra,
lower Phase II includes rock, earth, few shicra fragments, may be from
collapse of structures on platform.
III 12 Wall 1 and associated platform were built. Contents include material Charcoal dated 2840 Cal BC (Beta-199060)
from wall itself (Fig. 9). Radiocarbon sample from this level was (Table 4n).
collected from the materials removed with the rock of the wall.
IV 13–15 Use surface at top, probably created by foot traffic. Surface runs under, thus
precedes Wall 1. During remainder of phase, the area was used for cooking/
heating. High-density carbonized remains. Midden leveled underlying
surface. Layer of rock between midden layers may be construction debris
or an effort to stabilize the midden surface.
V 15 Original ground surface. Not level, little cultural material, grading to no
cultural material in Level 7.

70 FIELDIANA: ANTHROPOLOGY
 TABLE 10. Caballete, Sector A, Test Unit 4—2004.

Excavation (Fig. 12)

Level Thickness Description Contents
Surface 14–21 Loose beige soil, fragments of angular rock, natural Shell, charcoal, plant fiber, lithics.
layer ending in changing texture and color.
1 13–16 Gray soil, fragments of angular rock, Plant remains, shell, lithics, animal bone,
numerous clay mortar fragments. feathers, plaster fragments. Twined textile
fragment at interface of Levels 1 and 2.
2 15–17 Continuation of trash deposit in Level 1. Shell, charcoal, lithics, plant remains, animal bone,
feathers, coprolites, twined textile fragments, clay
mortar.
3 14–16 Deposits in this level change from gray to Shell, charcoal, plant remains, feathers, pieces of
grayish light brown. clay mortar, plaster, rodent and human
coprolites, animal bone.
4 1–5 Use surface prominent in N portion of pit, disappears Material over the use surface included lithics,
at S end, consists of lightly packed fine light beige shell, fiber, coprolites, plaster, bone, feathers.
soil, little archaeological material, a marked contrast
with gray trash deposit both above and below.
5 9–15 Trash deposit similar to that above Level 4, abundant Lithics, shell, coprolites, feathers, bone, plant
cultural remains. A pocket of ash toward N corner. remains, charcoal.
6 13–15 Trash deposit containing a substantial Charcoal, bone, feathers, coprolites, few lithics. Several
amount of plant fiber. well-preserved fish heads (anchovy), twig wrapped
in string (bobbin), fragment of plastered mortar.
7 15 in the portion Removal of midden exposed two walls, the corner of a Construction fill included large quantity of
excavated, Wall 1 low platform. Perpendicular walls (Feature 1) built with medium and small angular rock with chunks
large rocks, clay mortar, meet at the W edge of unit. of clay mortar mixed with plant fiber and
Interior (E) side of wall and below, deposit continued with archaeological trash in secondary context.
mixed rock, clay, plant remains. Platform was oriented NE
to SW with exterior facing E, the interior of the platform
W. S of this structure a stone alignment was identified that
appears to be the other face of the platform, meeting in a
90u corner. Upper surface of platform without distinctive
features or finish. Generally this low platform was covered
by secondary trash as in Levels 5–7.
8 16 in the area Material that filled the interior of the platform was Bone, charcoal, rodent excrement, shell, lithics,
excavated identified as Feature 1. Fill of medium and small feathers, plant remains. Level also yielded
angular rock in matrix of fine beige soil also contained concentration of large leaves, possibly pacae
significant plant fiber and mortar fragments. (Inga feuillei), N of the corner formed by the
two walls.
9 15–30 Fine-grained sediment with fine loose sand visible in Plant fiber, seeds, leaves, charcoal, shell, small
N. This level excavated only in the areas exterior to animal bones, coprolites, clay mortar, lithics.
the platform (Wall 1) identified in Level 7.
10 14, 15 Surface of compact fine-grained beige sand Use surface with quantity of botanical remains,
and soil on the N side of the area excavated. This small pieces of charcoal, shell. Narrow lenses
included only the segment outside the walls, a of ash associated with use of platform may
small area. have been deposited in a pit excavated at the
corner of the platform. Toward S side of
excavated area deposit was less compact and
lighter color, consistent with less active use.
11 1.5–4 Deposit similar to previous level. Compact, fine-grained Charcoal fragments, small shell fragments, animal
beige sandy sediment in W part of pit, extending to bone, plant remains, few very small pieces of
Level 7 platform. Area of packets of pacae leaves (I. quartz. Charcoal and shell fragments pressed
feuillei) in Level 8. Matrix of light beige loose, fine into the use surface at the upper margin of the
sandy sediment. level, defines the start of Level 11.
12 1–3 Use surface below Level 11. Matrix of material similar Lithics, animal bone, charcoal, shell, some
to Level 11, with more ash. Field observations fragments of twined textile.
indicate this was the first use surface in this locality,
used during construction of Level 7 platform and
possibly before it.
Excavation 14 Excavated in 15-cm levels once the base of platform Clay mortar fragments, plant fiber, medium-
of plat- was identified. Matrix within platform was compact sized rocks dispersed irregularly within
form fill— red-brown clayey soil. matrix. Shell, animal bones, coprolites, lithics,
Wall 1 botanical remains to 158 cm below datum.
13 12–18 Dark, fine-grained ashy soil, somewhat uncompacted Burned trash, charcoal fragments, shell
texture across unit under Level 7 platform indicates fragments, lithics, plant fiber, other botanical
prior use. Deposit noticeably more compact at base. remains, bone, feathers, coprolites, a few
chunks of clay mortar.
14 10–18 Compact surface of ash recognized in S wall of unit. Organic inclusions such as plant fiber, fragments
Matrix similar to previous, fine, dark gray ash. Use of cane and wood, textile fragments; lithics
surface preserved in S half of unit. Base of level including a large denticulate tool.
marked by narrow lens of loose gravel running the
length of excavated area.

CREAMER ET AL: FORTALEZA VALLEY, PERU 71

 TABLE 10. Continued.

Excavation (Fig. 12)

Level Thickness Description Contents
15 12–16 Deposit of ash with compact surface. Dark Abundant organic material including cane and
gray to brown matrix, somewhat compact, plant fiber concentrated in center of the unit
fine grained. Varied in depth, appeared (Feature 3). Shell, coprolites, feathers, bone,
to be a badly deteriorated use surface. charcoal, lithics, textile fragments.
Medium-sized gravel lay under the ashy deposit.
16 14, 15 Medium-sized gravel under Level 15. No artifacts recovered. Plant material recovered from
pockets of Level 15 ash intruding into Level 16,
culturally sterile subsoil. One low spot held a
circular bundle consisting of packets of twisted
willow stems (Salix humboldtiana) and achira
(Canna edulis) (Figure 14). These packets mark the
start of occupation in this area, an offering, though
no other materials were included with them.
17 Culturally sterile gravel that extended No artifacts recovered. Two fragments of plant fiber
the length of the unit. at top of level were dated (Table 11d, e).

Stratigraphy
Layer Levels Characteristics of layer Contents of layer
A 0 Thin surface layer of fine-textured, loose beige soil with aeolian Mostly small fragments of shell and charcoal.
sand and natural inclusions, very small fragments of angular rock.
B 0 Thick deposit of cultural material in fine-textured, sandy, dark beige Plant fiber and seeds, fragments of shell, lithics,
soil from somewhat compact to uncompacted. Cultural material animal bone, feathers, chunks of clay mortar,
distributed unevenly. Layer B appears to be the upper portion of a fragments of clay plaster.
deep archaeological trash midden deposited toward end of occupation.
C 1–8 Fine-grained, somewhat compact light beige soil. Cultural materials Primarily plant fiber, seeds and stems, along with
were mixed together, no discernible pattern. Upper half of the layer fragments of shell, small pieces of charcoal,
separated from lower portion toward N part of unit by narrow feathers, rodent excrement, coprolites, animal
layer of fine-textured, compact light beige clayey sediment (3–4 cm) bone, clay mortar, plaster from walls.
that contained very little cultural material. Surface of lens was level,
homogeneous, likely a use surface, excavated as part of Level 4.
The use surface was subsequently covered with more trash.
Fine-textured dark beige to cinnamon colored uncompacted soil.
Excavated as part of Levels 5–12 owing to its thickness. Use surface
in N half of the unit; large number of fish heads, possibly anchovy,
fragments of tree branch wrapped in string very near a layer of ash.
At the beginning of Level 8, a rectangular platform was identified, with
external faces of angular rock and fill of mortar fragments. Wall plaster,
archaeological trash in fine beige clayey sediment was identified.
D 8–12 Outside the platform, the W and N portions of the unit revealed 3 use Plant fiber, stems, leaves, seeds, charcoal
surfaces at the upper margins of Levels 10–12. Each was a thin layer of fragments, animal bones (mostly small species),
fine light beige clayey sediment, somewhat compacted, little cultural feathers, coprolites (mostly rodent), lithics,
material. Use surfaces in Levels 10 and 12 were better preserved than fragments of clay mortar, wall plaster.
Level 11 and included concentrations of ash. Proximity of the use
surface in Level 12 to the platform suggests it functioned when the
platform was in use. Layer D formed over time for disposal of debris and
trash on the use surface that was the base for a stone platform. Platform
activity is indicated by use surfaces in Levels 10–12. Later, platform was
covered by trash. Material was similar but not identical to that
accumulated between the use surfaces. Late Archaic age of deposits
indicated by shicra bags in S wall trash layer.
E 11, 13 Dark gray, fine-grained, uncompacted ash. Surface was lightly Shell, lithics, plant fibers, other botanical remains,
compacted, though not across entire unit. Large quantity of bone, feathers, coprolites, fragments of clay
midden. Area was in use before platform was constructed. mortar distributed irregularly across Level 13.
F 14–15 Fine-textured, dark gray, somewhat compact, ashy soil with botanical Plant fiber, stems, cane, wood, lithics, textile
remains. Surface of Layer F excavated as Level 14 not as thick as other fragments primarily from the N side.
levels (7–8 cm), may have been a use surface, best preserved at S end.
Lower portion of layer marked by narrow lens of medium-textured,
loose gray gravel separating it from ashy layer below. Layer F
represents occupation before the Level 12 platform.
G 16, 17 Thick deposit of dark gray, fine-textured somewhat compact ashy Cane, plant fiber, shell, coprolites, feathers,
sediment. Surface of layer was compact, possible use surface, best animal bone, charcoal, lithics, twined textile
preserved at the S end of unit. Covered with lens of medium-textured fragments.
loose gravel and botanical material, stems, leaves, plant fiber. This
layer filled small pits excavated into underlying coarse gray gravel on
which the site was built. The pits are small, maximum of 15-cm
diameter. Fill mostly plant fiber. One larger pit in SW corner of unit
yielded large number of willow (S. humboldtiana) and achira (C. edulis)
twigs, leaves wound and tied into small packets. As noted above, this
feature was excavated in Level 16. Fragments of plant fiber in Level
17 also part of Layer G.
H 17 Uncompacted coarse gray gravel below the pits dug into the gravel layer. No cultural materials.

72 FIELDIANA: ANTHROPOLOGY
 TABLE 10. Continued.

Phases of occupation (Table 11a–g)

Phase Layers Description Dated samples
I A–C Aeolian and recent deposits over midden.
II C Clay floor appears to be last occupation of this part of the
site in the middle of Layer C. Floor created on deposits
of rock, gravel, cultural materials. Alternatively, this
layer may be a flood deposit as cultural materials above
are similar to those below.
III D Wall 1 and associated low platform, floors, and remodeling Dates of 2390 (ISGS-5521) and 2730 Cal BC (Beta-199062)
of floors that took place while Wall 1 platform was in use. (Table 11a, b) obtained from plant fibers.
IV E, F Black, ashy midden deposits were among the earliest A date of 2220 Cal BC (GX-31583) (Table 11c) from the base of
deposits in this area, clearly midden and hearth contents. the ashy midden deposit; dates of 2320 (ISGS-5721) and 2700
Cal BC (Beta-199777) (Table 11f, g) obtained from the
feature/midden-filled irregularity at layer’s base.
V G, H Includes the base of Test Unit 4, gravelly deposits. A date of 1890 (GX-31581) (Table 11d) obtained from the
material intruding into Layer H. A date of 2300 Cal BC (GX-
31580) (Table 11e) obtained from a sample from the same level
that was pit fill or midden concentrated in the NW corner.

CREAMER ET AL: FORTALEZA VALLEY, PERU 73

74
TABLE 11. Radiocarbon dates from Caballete.

Sample Provenience Level/Layer Material Lab number Cal BC RCYBP ± 13c/12c 68.2—––1s range 95.4—––2s range
a Sector A, pit profile 8 fiber from wall plaster ISGS-5521 2390 3920 70 212.8 2546–2544 (1.1); 2488–2294 (98.9) 2579–2200 (99.8); 2158–2155 (0.2)
b Sector A, TU 4 10 mixed plant fibers Beta-199062 2730 4160 70 225.7 2873–2833 (18.7); 2818–2662 (75.0); 2895–2569 (98.4); 2516–2500 (1.6)
2649–2635 (6.3)
c Sector A, TU 4 15 mixed plant fibers GX-31583 2220 3770 70 218.4 2294–2124 (81.6); 2092–2043 (18.4) 2458–2417 (4.0); 2410–2022 (95.3);
1992–1983 (0.7)
d Sector A TU 4 17 mixed plant fibers GX-31581 1890 3550 60 227.6 1964–1867 (60.7); 1848–1774 (39.3) 2112–2102 (0.6); 2036–1738 (98.7);
1708–1697 (0.7)
e Sector A, TU 4 17 final mixed plant fibers GX-31580 2300 3840 70 226.7 2456–2419 (15.5); 2407–2376 (13.0); 2476–2129 (97.1); 2088–2047 (2.9)
2351–2203 (71.4)
f Sector A, TU 4 pit fill mixed plant fibers ISGS-5721 2320 3870 70 212.6 2464–2283 (91.9); 2248–2232 (6.9); 2562–2534 (2.1); 2493–2139 (97.9)
2217–2215 (1.2)
g Sector A, TU 4 pit fill charcoal Beta-199777 2700 4110 70 211.3 2862–2807 (24.4); 2758–2718 (16.6); 2880–2558 (92.0); 2554–2550 (0.4);
2706–2577 (59.0) 2537–2491 (7.5)
h Sector B, Profile 1 Layer E fiber bag Beta-183312 1940 3590 70 226.2 2110–2104 (2.2); 2036–1877 (88.5); 2135–1753 (100.0)
1841–1824 (5.4); 1795–1783 (3.9)
i Sector B, Profile 1 Layer G fiber bag GX-30511 2120 3720 70 211.0 2270–2259 (3.3); 2206–2023 (94.2); 2340–2313 (1.4); 2310–1920 (98.6)
1991–1984 (2.5)
j Sector B, Profile 1 Layer G mixed plant fibers ISGS-5523 3070 4450 290 213.5 3624–3602 (2.0); 3524–2861 (91.6); 3893–3883 (0.1); 3798–2338 (99.7);
2808–2756 (5.0); 2719–2705 (1.3) 2322–2309 (0.2)
k Sector B, TU 1 4 mixed plant fibers ISGS-5537 2280 3810 70 215.8 2427–2426 (0.3); 2400–2382 (6.3); 2467–2119 (92.7); 2096–2040 (7.3)
2347–2140 (93.4)
l Sector B, TU 1 5 mixed plant fibers GX-30512 2070 3680 70 214.7 2192–2179 (5.0); 2142–1962 (95.0) 2284–2248 (3.3); 2233–1887 (96.7)
m Sector B, TU 1 6 fiber bag Beta-183313 2350 3890 80 226.7 2475–2274 (85.2); 2255–2208 (14.8) 2618–2608 (0.4); 2595–2594 (0.2);
2584–2128 (97.6); 2088–2046 (1.8)
n Sector B, TU 1 7 charcoal GX-31588 2450 3960 60 227.2 2570–2514 (33.8); 2502–2433 (41.4); 2828–2824 (0.2); 2625–2284 (99.1);
2423–2402 (9.6); 2380–2348 (15.2) 2248–2234 (0.7)
o Sector B, TU 1 8 mixed plant fibers GX-31590 2620 4030 70 211.5 2834–2817 (5.8); 2663–2647 (5.5); 2868–2803 (10.2); 2777–2429 (85.6);
2636–2468 (88.7) 2425–2400 (1.8); 2382–2347 (2.5)
p Sector B, TU 1 9 fiber bag ISGS-5710 2340 3870 70 210.8 2464–2283 (91.9); 2248–2232 (6.9); 2562–2534 (2.1); 2493–2139 (97.9)
2217–2215 (1.2)
q Sector B, TU 1 9 fiber bag ISGS-5712 2320 3860 70 212.8 2459–2280 (87.7); 2250–2230 (8.7); 2562–2535 (1.5); 2492–2135 (98.5);
2219–2212 (3.5) 2067–2066 (0.05)
r Sector B, TU 1 9 mixed plant fibers Beta-199057 2500 3980 80 210.2 2619–2607 (3.1); 2599–2593 (1.4); 2858–2810 (4.1); 2750–2723 (1.6);
2585–2392 (83.1); 2385–2346 (12.4) 2700–2277 (92.2); 2252–2228 (1.4);
2222–2210 (0.6)
s Sector C, Profile 2 N face fiber bag Beta-183314 2400 3920 50 29.6 2474–2339 (95.5); 2321–2319 (1.5); 2568–2519 (7.5); 2499–2281 (90.1);
2315–2310 (3.0) 2249–2231 (1.9); 2218–2212 (0.5)
t Sector C, Profile 2 N face mixed plant fibers Beta-199059 2320 3860 70 211.9 2459–2280 (87.7); 2250–2230 (8.7); 2562–2535 (1.5); 2492–2135 (98.5);
2219–2212 (3.5) 2067–2066 (0.05)
u Sector C, Profile 2 E face tillandsia GX-31589 2160 3750 70 212.7 2275–2254 (10.7); 2227–2224 (1.3); 2335–2324 (0.7); 2307–2302 (0.2);
2209–2124 (61.1); 2091–2043 (27.0) 2301–2020 (98.3); 1994–1982 (0.8)
v Sector C, TU 2 8 fiber bag ISGS-5516 2490 3980 70 214.2 2616–2612 (1.4); 2580–2431 (80.1); 2850–2813 (2.9); 2743–2728 (0.6);
2424–2402 (7.4); 2381–2348 (11.1) 2695–2686 (0.4); 2680–2284 (95.5);
2248–2234 (0.6)
w Sector C, TU 2 9 charcoal ISGS-5719 2320 3860 70 224.7 2459–2280 (87.7); 2250–2230 (8.7); 2562–2535 (1.5); 2492–2135 (98.5);
2219–2212 (3.5) 2067–2066 (0.05)
x Sector C, TU 2 11 junco GX-31584 2120 3720 70 212.2 2270–2259 (3.3); 2206–2023 (94.2); 2340–2313 (1.4); 2310–1920 (98.6)
1991–1984 (2.6)
y Sector C, TU 2 11 charcoal GX-31587 2640 4040 60 225.3 2832–2820 (5.3); 2657–2654 (1.2); 2867–2803 (10.8); 2777–2459 (89.2)
2632–2474 (93.6)
z Sector C, TU 2 12 mixed plant fibers Beta-199776 2570 4020 90 211.7 2833–2818 (4.9); 2662–2649 (3.7); 2864–2805 (8.4); 2760–2341 (91.6)
2635–2465 (91.4)

FIELDIANA: ANTHROPOLOGY
 TABLE 11. Continued.

Sample Provenience Level/Layer Material Lab number Cal BC RCYBP ± 13c/12c 68.2—––1s range 95.4—––2s range
aa Sector C, TU 2 12 charcoal Beta-199058 2960 4300 60 224.1 3011–2977 (23.6); 2974–2948 (14.5); 3096–2856 (90.5); 2811–2747 (7.5);
2944–2880 (62.0) 2724–2698 (2.0)
bb Sector C, TU 2 13 fiber bag Beta-183315 2350 3890 70 29.3 2471–2286 (97.4); 2247–2243 (1.5); 2568–2518 (5.3); 2499–2195 (92.5);
2238–2235 (1.1) 2172–2145 (2.2)
cc Sector C, TU 2 15 mixed plant fibers ISGS-5511 2660 4050 80 213.3 2847–2844 (1.2); 2840–2813 (10.0); 2878–2451 (97.1); 2445–2438 (0.4);

CREAMER ET AL: FORTALEZA VALLEY, PERU

2737–2734 (0.9); 2692–2689 (0.9); 2420–2405 (0.9); 2378–2350 (1.6)
2678–2473 (87.1)
dd Sector C, TU 2 16 charcoal Beta-199775 3620 4840 70 223.9 3702–3627 (55.6); 3591–3527 (44.4) 3775–3501 (94.4); 3428–3380 (5.6)
ee Sector D, Profile 3 fiber bag GX-30513 1620 3330 90 214.1 1735–1714 (8.1); 1694–1508 (91.9) 1877–1841 (3.4); 1825–1795 (2.2);
1782–1430 (94.4)
ff Sector D, Profile 3 fiber bag Beta-184859 2150 3740 50 225.6 2263–2261 (1.4); 2205–2117 (61.9); 2294–2015 (98.5); 1997–1980 (1.5)
2098–2039 (36.8)
gg Sector D, Profile 3 mixed plant fibers ISGS-5532 2390 3920 70 210.7 2546–2544 (1.0); 2488–2294 (99.0) 2579–2200 (99.8); 2158–2155 (0.2)
hh Sector E, trench charcoal GX-30514 3120 4440 40 228.3 3322–3272 (20.6); 3269–3235 (17.7); 3333–3213 (34.1); 3189–3154 (6.9);
3171–3162 (3.6); 3116–3017 (58.1) 3134–2927 (59.0)
ii Sector E, trench mixed plant fibers ISGS-A0477 2520 4000 70 213.1 2831–2821 (2.6); 2630–2457 (91.2); 2857–2811 (5.2); 2749–2723 (1.8);
2418–2407 (2.7); 2375–2367 (1.7); 2699–2296 (93.0)
2362–2353 (1.9)
jj Sector F, TU 1 4 (Feature 1) mixed plant fibers ISGS-5724 940 2870 80 213.3 1190–1178 (4.0); 1159–1144 (5.4); 1290–1280 (0.7); 1270–842 (99.3)
1131–926 (90.6)
kk Sector F, TU 1 7 (Feature 2) mixed plant fibers ISGS-5729 680 2580 70 223.3 817–746 (47.1); 688–664 (13.2); 896–507 (98.3); 459–453 (0.3);
646–552 (39.7) 439–419 (1.4)

75
 TABLE 11B

76 FIELDIANA: ANTHROPOLOGY
 TABLE 12. Caballete, Sector B, Profile 1—2003.

Stratigraphy (Fig. 15)

Layer Description of stratigraphy Contents of layer
A Shicra, medium-sized and small rocks. Radiocarbon samples were collected during profile
clearing. No other.
B Clay floors, three episodes of remodeling. Floor made of clay mixed with plant fiber. Materials were collected.
C Construction fill above a wall, including some shicra.
D Floor on top of wall that formed the face of the structure.
E Plaster-coated wall at base of pit.
F Plastered floor on S side of W wall and along S profile. Wall visible in S profile
of unit rests on this plastered floor.
G Construction fill, shicra bags filled space behind plastered wall.

Phases of occupation (Table 11h–j)

Phase Layers Description Dated samples
I A, B Debris, soil from collapse of last structures to be built;
postdates abandonment of mound.
II B Final occupation of mound; intensive use shown by multiple
remodelings of the floor.
III C Phase during which fill was used to cover previous floors to A date of 1940 Cal BC (Beta-183312) (Table 11h) obtained
raise or level the area. from shicra fragment recovered below floors at top of
Layer C.
IV D Two floors constructed on top of earlier wall, associated
construction fill.
V E, G E is the wall of the buried room. G is the construction fill A date of 2120 Cal BC (GX-30511) (Table 11i) obtained
behind that wall, consisting of shicra bags. from shicra. A date of 3070 Cal BC (ISGS-5523)
(Table 11j) obtained from plant fibers mixed into clay
plaster from wall in Layer E.
VI F Floor on which plastered wall was built.

CREAMER ET AL: FORTALEZA VALLEY, PERU 77

 TABLE 13. Caballete, Sector B, Test Unit 1—2003.

Excavation (Fig. 16)

Level Thickness Description Contents
0 7–9 Aeolian sediment and soil, recently disturbed.
1 0–15 Sandy soil, ash, charcoal, small area of floor in NE Burned shell, one bone bead.
corner.
2 0–12 Sandy soil with ash, charcoal. Burned shell, heat-fractured rock, lithics.
3 0–10 Natural level including the rest of the deposit of Charcoal, burned shell, heat-fractured rock, botanical remains.
dark ashy soil.
4 12–18 Fragment of clay floor containing a large Construction debris, including clay fragments, plant fiber, lithic
quantity of plant fiber. remains, shell, charcoal.
5 12–17 Fill, large quantity of shicra. Shicra fragments, individual fibers, braided plant stems, reed
fragments. Clay mortar, angular rock in the SW portion.
6 12–18 Light brown, powdery soil, gravel, rock. Lithics, shell fragments, charcoal, bone, shicra, plant remains,
including pacae, peanut, squash, and cotton seeds, stems and
bolls, textile fragments.
7 13–18 Dry, powdery soil, gravel, rock. Plaster wall or Shicra, charcoal, a few strands of cotton, pacae, squash.
floor, fragments of clay with surface of gray
paint or fine plaster.
8 8–15 Dry, powdery light brown soil. Plant fiber, preserved shicra bags, one beside a large rock at the base
of unit.
9 16–26 Medium and small angular rock in dry, powdery light Very little cultural material.
brown soil. A layer of shicra bags was part of
construction fill above loose medium-sized rock.
Below was an irregular and rough use surface.
10 4–16 Orange-brown, culturally sterile soil. First use of No cultural material.
this area occurred when wet clay formed the use
surface, possibly an effort to level loose, powdery soil.

Stratigraphy
Layer Levels Characteristics of layer Contents of layer
A 1–5 Ashy layer across surface of unit associated with filling Perhaps a pachamanca, or pit cooking feature.
a large pit feature.
B 2 Patch of yellow-brown soil preserved between two Heat-fractured rock with burned and unburned plant fiber,
areas of disturbance, the intrusive pit in A and botanical remains, shell, and charcoal intruding into
intrusive pit D. Layers B and E.
C 1 Fragment of hearth or burned surface. Fire-reddened soil.
D 2–4 Pit features. Uniform fill of gravelly soil.
E 3–7 Midden fill predating features in A and D. Abundant plant remains.
F 6–8 Shicra deposit.
G 7, 8 Shicra used to construct a low platform.
H 9, 10 Use surface of clay and plant fiber, not uniform; 20 cm thick on
N side, thinning toward S end, an effort to level the area.
J 2–8 Intrusive pit feature excavated and refilled with
gray gravelly soil.

Phases of occupation (Table 11k–r)

Phase Layers Description Dated samples
I C, D Last use of this area includes Layers C and D, part of small intrusive hole filled
with gravelly soil. Layer C may be fragment of floor identified in Level 1. Thin
layer of red sand may be remains of hearth/burned area. Layer D is intrusion of
dark gray compact soil, hard and gravelly, coarser than other material in unit.
Small to medium-sized rocks, gray color from ashy material, no plant remains.
Layer D is within an earlier intrusive deposit, Layer J. No indication of specific
function. Layer D could be remains of feature below Layer C floor. Phase I may
be late occupation barely sampled by this test unit.
II A Ashy spot in NW corner could have been an activity area for cooking with hot
stones. Heat-fractured rock, burned and unburned plant fiber, botanical
remains, shell, charcoal, intrusive into Layers B and E.
III B Soft yellow-brown soil, a few rocks, very little plant fiber, may have been used to
level, smooth, or clean the space, possibly prior to food preparation. Not
compact enough for a floor, little cultural material. Layer B was put in place
after Layer J intrusion was filled but before pit was excavated that became
Layer A.
IV J Intrusion of tan soil mixed with small and medium-sized rocks. Removal of soil
disturbed shicra bags, moved aside in a pile forming Layer F.

78 FIELDIANA: ANTHROPOLOGY
 TABLE 13. Continued.

Phases of occupation (Table 11k–r)

Phase Layers Description Dated samples
V E Extensive midden with high concentration of plant remains, divided into Four samples yielded dates of 2450–2070
upper/lower segments by layer of small rocks (,5 cm), a rough floor or use Cal BC (ISGS-5537; GX-30512; Beta-
surface. Upper portion of Layer E was more compact, with a few small 183313; GX-31588) (Table 11k–n).
rocks. Lower portion includes more medium-sized rocks. Layer E is bounded
on N and S by intrusive deposits; only a short stretch of profile shows layers
of deposition.
VI G, H Low platform built over Layer H. During this phase a platform was built A sample dated to 2620 Cal BC (GX-
orming surface for working reeds. Large rocks formed side of platform. 31590) (Table 11o).
H Layer H was the first use surface in this locality; clay was used to level Three samples dated 2500–2320 Cal BC
the area. (ISGS-5710; ISGS-5712; Beta-199057)
(Table 11p–r).
VII I Original ground surface of gravel and bedrock fragments.

TABLE 14. Caballete, Sector B, Test Unit 2—2004.

Excavation (Fig. 17)

Level Thickness Description Contents
0 0–15 Loose sediment over a thin layer of ash. Shell, lithics, fish bone, charcoal fragments, plant fiber.
1 12–16 Compact light gray soil with patches of caliche. Base of level Fragments of shell, charcoal, plant fiber and lithics.
had a lens of compact brown soil with some rocks. SW corner
included patch of light brown soil lacking cultural material.
2 15 Gray gravel in the center of the unit appeared to be part of A compact surface associated with charcoal fragments, plant
a compact beige surface. fiber and hair along upper edge. The rest of the level was
without cultural material; the patch of cultural materials
may be a flood deposit.
3 92 Upper portion included remains of light gray surface. A few plant fibers, charcoal fragments.
Gravel layers were natural rather than cultural deposits.

Stratigraphy
Layer Levels Characteristics of layer Contents of layer
A 0, 1 Shallow surface layer formed by fine-grained, loose-textured Little cultural material, principally small shell fragments.
sediment mixed with fine aeolian sand. Small natural inclusions of large-grain angular rock.
B 1 Fine-grained, dark brown, semicompact sediment extending Cultural materials include shell fragments, charcoal, plant
across the unit. Natural inclusions of angular rock distributed fiber, lithics, fish bone, in concentrations associated with
irregularly suggest layer was accumulated wind-blown material thin layers of ash.
mixed with existing surface sediment following site abandon-
ment, compacted by humidity. Upper part of layer may be a
poorly consolidated use surface.
C 1 Fine-textured, loose, light beige sediment across entire unit. Natural and cultural materials similar to those in Layer B.
In places sediment mixed with medium grayish gravel.
D 2 Fine gray partly compacted sediment, not as thin as previous Cultural materials similar to Layers B and C include
layers. Layer extended across entire unit, slight tilt from charcoal, plant fiber, hair concentrated in upper part of
N/S clearly visible. Inclusions of angular rock, as in Layers layer; greater density than Layers B and C. Density,
2 and 3. position suggest this was a use surface highly deteriorated
at time of excavation.
E 2, 3 Medium-sized, slightly pinkish somewhat loose gray gravel Culturally sterile.
constituted the upper portion of thick deposit of coarse gravel
present in all sectors of the site underlying cultural deposits.
Gravel becomes coarser with depth, losing the pinkish tint.
F 2 Portion of use surface on E side of unit. A few shell and charcoal fragments.

Phases of occupation
Phase Layers Description Dated samples
I A–D Layers of different textures, compositions, include cultural
material; may represent occupation of this part of the site.
II E, F Includes base of Level 2, upper portion of Level 3. On E side,
possible use surface was identified. Below was what appeared
to be a shallow pit. Layers of fine sand and clay cutting across
the gravel deposits appear to be natural. Evidence of activity
confined to the small use surface.

CREAMER ET AL: FORTALEZA VALLEY, PERU 79

 TABLE 15. Caballete, Sector C, Profile 2—2003.

Stratigraphy (Fig. 18)

Layer Description of stratigraphy Contents of layer
A Postoccupation disturbance. Loose soil, rock fill, and debris.
B Section of clay floor.
C Layer of fill below floor, showing leveling before the clay floor was laid.
D Collapsed construction. Wall 1 visible at extreme N of profile. Large rock and rubble.
E Clay floor at base of Layer D floor appears constructed from Wall 2 toward center of mound.
Wall 2 appears to have been a retaining wall within lower levels of mound.

Phases of occupation (Table 11s–u)

Phase Layers Description Dated samples
I A, B Debris from looting, erosion, collapse, with dates long after final A date of 2400 Cal BC obtained from a
occupation. At base of Phase I is floor that probably indicates use of fragment of shicra bag (Beta-183314)
the mound area after abandonment. (Table 11s).
II B Construction fill and rubble from collapse of structures. Wall 1 was A date of 2320 Cal BC obtained from the N
identified at one end of the profile, with large rocks, fragments of face (Beta-199059) (Table 11t).
shicra dispersed throughout.
III C Floor dating to occupation, with retaining wall at N end; floor segment may A date of 2160 Cal BC (GX-31589)
have been part of original construction of floor and associated elements. (Table 11u) obtained from Phase III
Excavation ended at base of damaged area, though cultural materials con- materials.
tinued into lower levels. Base of cleared area was well above base of
mound. Cleared section shown is only a portion of the Sector C mound
deposits.

80 FIELDIANA: ANTHROPOLOGY
 TABLE 16. Caballete, Sector C, Test Unit 2—2003.

Excavation (Fig. 20a, b)

Level Thickness Description Contents
1 10–16 Dry, sandy, light-brown soil, rock; a few compacted patches. Shell fragments.
2 10–15 Dry, sandy, light-brown soil, patches of clay, ash. Shicra, clay mortar fragments, a few large rocks, probably
remains of collapsed wall.
3 14–20 Compact sandy soil around large rocks first observed in Level 2. Cotton, lithics, charcoal, plant remains, shell, feather,
apparently midden.
4 13–15 Light brown soil with some clay lumps, though soil was not Textile fragments, bone, shell, lithics, charcoal, plant
clayey. remains.
5 15–17 Few rocks in compact soil, fragments of clay mortar mixed More artifacts including lithics, shell, animal bone, human
with plant fiber. Cross section of wall visible in W wall hair, charcoal, botanical remains. Large quantity of plant
but not in E wall. Ash, burned soil, burned rock. fiber in every level.
6 15–22 Botanical remains with clay or clay mortar. Use surface Many botanical remains with clay and clay mortar
at base of level, Floor 1, neither level nor plastered. fragments.
Plant material on top of the surface. Small rocks in
use surface, other materials, suggest surface formed
by use or traffic across area, not a prepared surface.
7 0–8 Dry sandy soil with ash, clay mortar fragments. Numerous Lithics, shell, botanical remains, charcoal.
fragments of charcoal just underneath use surface.
8 15–18 Loose soil with numerous reed fragments, angular rock. Shell, botanical remains, fragments of clay floor.
A mix of clay and plant fibers was used to make a
surface, though the floor, Floor 2, was fragmentary.
9 2–8 Compact soil. Little charcoal though a small area of ash. Many shell fragments and botanical remains.
A use surface with evidence of several renovations, Floor 3,
made it difficult to distinguish among layers of use.
10 9–26 Compact soil with several use surfaces, making it difficult to Trash deposit, burned rock, ash, charcoal, dark soil, with
follow out the individual components of this level. shell, plant remains, relatively less shell. Quantity of
botanical material, lithics, burned rock.
11 2–17 Sandy soil with rock, burned soil, fragments of clay floor. Botanical remains, fish bone, a few shell fragments.
S end of unit contained concentration of rock and ash.
12 12–15 Construction debris such as clay floor and wall fragments, Many botanical remains, little charcoal and shell. Some
with plastered surfaces visible. Wall or bench finished lithics, patches of ash, large rocks in S half.
with plaster identified on W side of level’s base.
13 11–15 Loose soil with shicra bags on S side, a plastered wall on Botanical remains, lithics, shell.
W side; large rocks with fragments of construction
material on E side. Wall along the W side made it difficult
to remove soil, small size of test unit (1 3 2 m) made
excavating awkward.
14 2–15 Loose fill along the side of Wall 2 including fragments of shicra. Shell, lithics, botanical remains including leaves, stems.
15 5–10 Continued shicra fill from previous levels. Shicra Shell, botanical remains including wood and leaves.
bags from Levels 13–15 appear to have been lined
with leaves. Level ended at a plastered floor.
16 9–13 Clay and fiber floor or bench attached to Wall 2. Charcoal, plant remains.
Floor was broken in center of unit, as was the bench;
beneath were loose earth and gravel, no cultural material.
Charcoal and burned soil in center of unit.
17 10–15 Only a narrow area could be excavated without removing Small quantity of shell, botanical remains, charcoal in center
bench and plastered floor. Excavation reached culturally where floor and bench were broken.
sterile sandy soil in lower portion of level. Area excavated,
sample, both small.

Stratigraphy
Layer Levels Characteristics of layer Contents of layer
A 1 Aeolian deposit of sand. Narrow layer, little cultural material.
B 1 Fill deposited or accumulated during final period of site use.
C 1, 2 Compact clay layers, use surfaces associated with Wall 1. Short segments of clay surface.
D 1–5 Feature associated with former ground surface in Level 1. In profile, Pit D is a distinctly later feature than adjacent
Floor C and Wall 1.
E 2–5 Mixed fill postdating occupation of Floor 1. Cotton, feathers, other materials.
F 2–6 Mixed fill postdating occupation of Floor 1. Thick layer of plant material.
G 7, 8 Ashy midden between Floors 1 and 2. Charcoal, ash, floor fragments.
H 10, 11 Below Floor 3, cultural fill over midden.
I 10–12 Dark gray, ashy midden. Ash and burned rock with less shell than other levels.
J 13, 14 Wall 2. Wall or bench with smooth plaster surface. Top of feature thought to be wall but upper edge was
plastered. Likely bench considering condition and depth.
K 13–17 Shicra fill inside broken section of Wall 2. Floor built over original ground surface, making it the
lowest constructed element found.
L 16 Floor or bench attached to Wall 2. Base of feature identified in this layer.
M 17 Sandy fill beside and below Wall 2 and associated floor or Sandy soil may be base of this construction, reached beside
bench. Excavation ended when the unit became too bench. Area was narrow; difficult to see whether
narrow to work. construction ended here.

CREAMER ET AL: FORTALEZA VALLEY, PERU 81

 TABLE 16. Continued.

Phases of occupation (Table 11v–dd)

Phase Layers Description Dated samples
I A, B Excavation of pit feature at end or after final occupation.
Contents of pit similar to materials in upper levels.
II Wall 1, C, D Included Wall 1 and subsequent occupation as shown by
surfaces built up against base of wall.
III E, F Period between use of Floor A and construction of Wall 1.
Deposition followed abandonment of Floor A, creating
inclined surface; later deposits largely leveled the area.
IV G, H Period of intensive use, construction of floor or compact use A date of 2490 Cal BC (ISGS-5516) (Table 11v) from a
surface of clay mixed with plant fiber was remodeled/ fragment of shicra bag in Level 8. A sample of charcoal
replastered at least three times. A thicker deposit, from Level 9 (Floor 3) dated 2320 Cal BC (ISGS-5719)
intentional fill or fill representing a period of disuse, was (Table 11w). A sample of reed, junco, from Level 11
capped by Floor A. dated 2120 Cal BC (GX-31584); a sample of charcoal
dated 2640 Cal BC (GX-31587) (Table 11x, y).
V I Layer of ash with plant remains in gravel, contrasting with A sample of mixed plant fiber from this level dated to
material below. May represent distinct change in 2570 Cal BC (Beta-199776); a sample of charcoal dated
occupation. Instead of a structure or floors, this phase to 2960 Cal BC (Beta-199058) (Table 11z, aa).
was trash disposal, predominantly charred.
VI J–M Earliest phase of occupation identified in this area. Includes A sample of shicra from Level 13 dated to 2350 Cal BC
Wall 2, wall or bench more than 20 cm wide across the top (Beta-183315) (Table 11bb). Plant fibers from Level 14
and completely covered by fine, smooth plaster. Top of dated 2660 Cal BC (ISGS-5511) (Table 11cc). A
feature was 2 m below present ground surface. Depth charcoal sample from Level 16 dated to 3620 Cal BC
suggests this is not top of a wall but a bench. Excavations (Beta-199775) (Table 11dd).
in surrounding area rarely reached 2 m. In this narrow test
pit (1 3 2 m) it was impossible to continue excavating.

TABLE 17. Caballete, Sector D, Profile 3—2003.

Stratigraphy (Fig. 21)

Layer Description of stratigraphy Contents of layer
A Soft fill, plant remains, crumbly rock.
B Wall 1, retaining wall of rock with clay plaster, changed direction slightly at E end. Wall Wall 1.
plastered on N side with rough surface of clay plaster.
C Fill of rock, clay, compact plant remains, some shicra. Construction fill below Wall 1.
D Fill of rock, clay, little plant fiber. Some shicra. Burned rock and shell.
E Large rocks, clay mortar in compact mass.
F Large angular rock with clay mortar. Shicra, loose soil. Burned rock.

Phases of occupation (Table 11ee–gg)

Phase Layers Description Dated samples
I A, B Construction of Wall 1 divides deposits into two phases. A date of 1620 Cal BC (GX-30513) (Table 11ee) obtained
Phase I included period following construction of Wall 1, from a shicra fragment at the base of Wall 1.
material in Layer A and two subsequent deposits of fill.
II C–F Deposits prior to construction of Wall 1, including Shicra fragment from this layer dated to 2150 Cal BC (Beta
D–F. These were either different sources of fill or 184859) (Table 11ff). Mixed plant fibers provided a date
different episodes of construction. of 2390 Cal BC (ISGS-5532) (Table 11gg).

82 FIELDIANA: ANTHROPOLOGY
 TABLE 18. Caballete, Sector D, Test Unit 3—2004.

Stratigraphy (Fig. 22)

Layer Description of stratigraphy Contents of layer
A Loose, fine beige sediment, a little coarse sand, few rocks distributed Cultural materials not abundant, included shell fragments,
irregularly within layer. Slightly less compact on N end. Toward lithics, a few plant remains, some ceramic fragments.
center, small compact areas approximately 10 cm diameter were
present, possibly caused by humidity. Likely recent formation of
wind-blown material and rubble.
B Medium-gained, compact beige sediment containing small Shell fragments, botanical remains, lithic fragments.
concentration of ash in S half of unit, 8 cm in diameter.
A whitish material was dispersed through the layer,
concentrated in center, possibly decomposed rock with high
carbonate content. Layer was a use surface, now
highly deteriorated.
C Loose light pink gravel with little cultural material; may have Three small pit features originated in Layer D, extended into
been use surface. Near N corner of pit, concentration of insect Layer E. F.2 materials included leaves, charcoal, lithic
casings. Intrusive pit filled with sediment, trash was labeled fragments, fish vertebrae. F. 4 included leaves, plant stems,
Feature 1. Contents included botanical remains, fragmented charcoal.
shell, one fragment of plain pottery. Feature 2, near the N
corner of the unit, 13 cm in diameter. F.2 was filled with loose
gray medium to coarse-grained sediment. Feature 3, pit feature
located SE of F. 2, filled with loose gray medium to
coarse-grained sediment. Contained few cultural remains,
11 cm deep. Feature 4, S of F. 2, small pocket of plant remains.
D Cultural fill deposited prior to formation of use surface in
Layer C. Three pits, Features 2, 3 and 4, intruded into Layer
D from Layer C. Deposit of compact dark beige fine-grained
sediment with little cultural material. Compact use surface
formed along upper surface of Layer D. Shallowness of
features associated with Layers C and D, suggests they were
natural variations in the surface that filled with trash rather
than areas specifically used for disposal, or postholes.
E Use surface formed by series of activities that took place Feature 5, infant burial, intruded from Layer C through Layer D
on the sterile ground surface. to 20 cm below the surface of Layer E. The matrix was
outlined with medium-sized stones that appeared to surround
the bundle (22 3 32 cm). Bundle wrapped in 1-2 layers highly
deteriorated fabric. Innermost layer highly deteriorated, tied
around middle with cord of twisted plant fiber. Infant was
seated facing east. Cultural materials included stems, leaves,
plant fiber, marine shell fragments, few lithics, charcoal
fragments. Material below Layer E natural sediments. Unit
extended below the layers of clear human occupation to
examine possible deeply buried occupation; none found.

Phases of Occupation
Phase Layers Description Dated Samples
I B–C Hard packed soil possibly associated with occupation and
use, cemented by evaporation and salt. Layer C seals pit in
which infant was interred.
II D, E, Feature 5 Use surface at top of Layer D. A fine layer of sand sealed No dates have been processed from this unit.
the top of Layer D, including burial. Surface in use when Fabric that wrapped the infant was simple flat
infant was buried could not be identified. weave, suggesting the burial was from the end
of the occupation. No artifacts associated.

CREAMER ET AL: FORTALEZA VALLEY, PERU 83

 TABLE 19. Caballete, Sector E, Trench 1—2003.

Stratigraphy (Fig. 24)

Layer Description of stratigraphy Contents of layer
A Visible on W side of trench, narrowing toward center. Matrix very fine Lithics, shell fragments, plant remains.
sand grading toward clay.
B A mix of fine gravel with sand. Lithics, shell, bone, textile fragments;
beans, lucuma, squash, pacae.
C Rock, cultural fill, gravelly deposits. Cultural fill may indicate former
ground surface; rock and gravel appear natural.

Phases of occupation (Table 11hh, ii)

Phase Layers Description Dated samples
I A Fine clay deposited by wind and water; fill postdates active
use of area. Postabandonment accumulation in low
points of ground surface.
II B May be flood deposits that filled depression excavated Two dates were obtained from Layer B. Charcoal
into Phase III. sample dated to 3120 Cal BC (GX-30514);
mixed plant fibers dated to 2520 Cal BC (ISGS-
A0477) (Table 11hh, ii).
III C Natural deposits of gravel and rock with layer of cultural fill at
top may indicate former ground surface.

84 FIELDIANA: ANTHROPOLOGY
 TABLE 20. Caballete, Sector E, Test Unit 5—2004.

Excavation (Fig. 26)

Level Thickness Description Contents
0 9.5–17 Uniform loose gray soil. Few small stones.
1 9–14 Sandy, fine-grained light gray soil. A compact irregular layer Shell, lithics, botanical remains, small fragments of
over deposit of loose soil. charcoal.
2 13–15 Pale compact, clayey deposit, some small angular rocks Shell, plant fiber, botanical remains, charcoal fragments,
within matrix. Below was deposit of brown gravelly soil, lithics, feathers, lumps of clay that appear to have been
increasing proportion of gravel with depth. mortar.
3 24–32 Fine pale gray gravel. As the level was excavated the Shell, plant fiber, botanical remains, a few animal bones
gravel was somewhat coarser and the color a darker gray. from the upper portion. Toward the base, the matrix
appeared to be culturally sterile.
3A 0–4 Materials associated with the bottom of Feature 3. Some botanical remains, shell, charcoal.

Stratigraphy
Layer Levels Description of stratigraphy Contents of layer
A 0 Fine, loose, beige sediment with aeolian sand. Narrow, Little cultural material other than shell.
from 3–10 cm thick. Angular rock forms irregular natural
inclusions. Layer inclined slightly from E to W.
B 1 (upper) Fine-grained slightly clayey light beige to orangey compact Plant remains, fragments of shell, small pieces of
earth, tilted like the layer above, almost 5 cm at the W charcoal, some lithics. Botanical material concentrated
end of the unit and practically invisible at the E end. in large pit (Feature 1) in SW corner began in Layer B,
Small angular rocks concentrated toward W side of unit. continued into next layer. Layer B may be mix of
existing deposits and aeolian material during a humid
period with sporadic human occupation.
C 1, 2 Fine to medium-textured grayish brown to light beige Shell, plant fiber, other botanical remains, charcoal
uncompacted soil. Like Layer B this layer was tilted fragments, lithics, feathers, clay fragments that appear
from E to W. to be mortar. Most material concentrated in intrusive
pit on W (Feature 2) with several large angular rocks.
D 2 Fine to medium-textured loose dark brown soil mixed with Shell, botanical remains, some animal bone, most from
sand and medium-sized gravel. Like Layer C, matrix pit visible in N profile (Feature 3) intruded into sterile
included small angular rock. gravel.
E 3, Medium to coarse light gray gravel underlies occupation No cultural material recovered.
Feature 3 layers. Within layer are numerous thin layers of caliche,
indicating episodes of high humidity or pooled surface
water prior to site occupation.

Phases of occupation
Phase Layers Description Dated samples
I A, B Surface of compact soil, pits including Feature 2 excavated into surface. Feature
2 contents not distinctive; either natural and filled with material that washed
or blew in, or man-made pit.
II C, D Remains of first occupation on this surface, assumed related to main occupation.
Feature 2 intruded into Layers C and D.
III E A large intrusive pit in W side of unit (Feature 3), with several large angular
rocks. With Layer D, this comprises first occupation in this locality since
Feature 2 was excavated into the surface. Contents of Feature 3 not distinctive
despite being intentionally excavated pit.

CREAMER ET AL: FORTALEZA VALLEY, PERU 85

 TABLE 21. Caballete, Sector F, Test Unit 1—2004.

Excavation (Fig. 27)

Level Thickness Description Contents
0 0–15 Loose, sandy soil. Small rocks.
1 0–12 Fine, loose sediment mixed with fine gravel; Shell fragments, textile, lithic fragments, a few charcoal fragments, human
disturbed surface materials, thick layer remains left by looters.
of ashy sediment in NE.
2 16–19 Fine loose sediment mixed with small gravel Shell, textiles, lithics, fish bone, botanical remains, burned rock, human
and rocks up to 10 cm; mixed natural and hair, charcoal, clay with embedded plant fiber.
cultural material.
3 12–13 Fine, loose soil in Level 2 replaced on N by Small rocks, shell fragments, textiles, fish bone, seeds, hair, Tillandsia sp.,
loose, light gray gravel without cultural charcoal. On N, a few small pockets of charcoal were associated with
content. In S, compact clay associated unburned botanical remains.
with carbonized material was identified.
4 1–29 Compact, fine light brown clay soil with a Numerous textile fragments in S part of pit in compact sediment containing
number of small stones concentrated shell, plant fiber. Textile concentration suggests textiles were wrapped
toward S side of level. N portion culturally around burial (Feature 1). Remains of infant excavated separately from
sterile fine gravel. rest of unit, deposited in fine light brown clay soil containing few small
rocks, textiles, botanical remains, some shell. Burial placed on
concentration of plant fiber, wrapped in textiles. Cranium in pieces
though majority of bones recovered intact. Infant in fetal position, right
side, lightly flexed. S wall profile indicated burial excavated into gravel,
filled with light brown soil.
5 14–23 After burial removal, Level 5 excavated across Near base and to the S, some clayey soil may indicate presence of
unit. Loose gray gravel with few larger rocks another grave.
on N side. In S end, small quantity of clay
containing charcoal, shells, fish bone, plant
remains, seeds, quartz fragments was recovered.
In N side of unit, gravel was natural without
cultural content; level was excavated only on
S side.
6 12–20 Excavated only in E half of unit. Contained Charcoal, lithics, bone, seeds, textiles, Tillandsia sp. remains of a basket
gravel mixed with light brown clay recovered. The basket suggested another burial, probably related to infant in
similar to infant burial in Feature 1. Feature 1, entire layer was excavated, mostly culturally sterile.
7 8–10 Natural deposit of culturally sterile gray Under infant burial, layer of culturally sterile gravel found in N side of unit
gravel on N. On S side, deposits changed continued, showing infant burial pit was intrusive. S side was mix of clay and
to brown. In S end some large stones gravel, Tillandsia sp., other fibers. Basket inverted over gourd bowl and cranium
were recorded just below level of basket. in pit just N of large rocks (Feature 2). Adult female wrapped in textile, on bed of
plant fibers, wrapped in mat of tied reeds. Individual reclined on her back with
knees to chest and arms at sides. Disintegrating textiles covered bundle. Funerary
objects included three gourd bowls, one found near basket covering cranium,
two by feet. A few metal beads found during removal of burial began were
necklace of numerous pyrite and one gold bead (see text for description). On S
under large rocks was layer of reeds with other offerings, four obsidian blades,
one obsidian fragment, four shell beads, one small spindle whorl with faces
incised around circumference, one whole mollusk valve, implements of wood and
bone 5–10 cm long, reeds presumably associated with mat or textile production.
8 37–47 Culturally sterile gray gravel visible when
Feature 2 was removed. Excavation continued
50 cm, testing equivocal evidence of cultural
remains. Cultural deposits did not continue.

Stratigraphy
Layer Levels Description of stratigraphy Contents of layer
A 0 Gray aeolian soil mixed with fine sand, loose texture. Small, angular rocks Small quantity of shell.
irregularly distributed in matrix. Deposits inclined slightly N/S,
averaging 3 cm thick, apparently recent, wind-borne deposit.
B 1–7 Compact, fine-textured, light beige sediment, some sand. Most evident in NW Mollusk valves, animal bone, small
corner of pit, 7 cm thick, with small angular rocks. Intrusive pit held Feature fragments of charcoal, irregularly
1 and Feature 2 burials. Characteristics and contents indicate long process of dispersed.
wind accumulation mixing natural and cultural soils after site occupation and
abandonment. Humid conditions may have produced compacting of this layer
representing a long period when Caballete was not occupied.
C 1, 2 Surface deposit of midden from occupation. Uncompacted fine-grained Charcoal fragments, shell, textile
beige soil darker than previous layer. A little charcoal, a few small and fragments, lithics, bone, in
medium-sized rocks. Like Layer B, Layer C is present only in N half of unit, heterogeneous distribution within
7–10 cm thick. Deposits formed similarly to Layer B. matrix.

86 FIELDIANA: ANTHROPOLOGY
 TABLE 21. Continued.

Stratigraphy
Layer Levels Description of stratigraphy Contents of layer
D 3–7 Fine-grained beige uncompacted soil with large quantity of small and medium More cultural material than previous levels,
sized rock, few large rocks, distributed irregularly through matrix. Greatest shell, deteriorated textile fragments, lithics,
concentration of rock on N side of pit. Noticeably thicker than previous layers, fish bone, botanical remains, burned rock,
20–30 cm, deepest on S end of excavated area, 36 cm thick with N/S tilt. Layer hair, charcoal, clay with plant fiber
was major deposition of rubble following principal occupation of Sector F. inclusions. Gourd bowls, basketry, metal,
Materials dispersed from N to S, filling slight declivity that determined the form shell beads, ceramic fragments associated
of subsequent layers. Near end of formation of this layer, intrusion at S end of pit with the adult. After the burial, the pit was
was created, burial pit for two individuals, Features 1 (infant) and 2 (adult). covered with mixed earth from the
Burials associated with layers of Tillandsia sp. disturbed layers, suggesting the interments
date to a period after the Late Archaic and
postdating the main occupation.
E 2, 3 Fine dark brown somewhat compact sediment, with ash in several places, Shell, textile fragments, fish bone, seeds,
covered in others by narrow layer of loose gray gravel. Layer extends across Tillandsia sp., animal hair, small
entire unit except for extreme S, burial pit. Top of layer at time of excavation was charcoal fragments. Most charcoal
compact use surface. Former surface, used during main occupation of site, was fragments pressed into layer’s surface.
created of soil trodden over gravelly terrain. After final abandonment, layer was On N side, layer filled two small holes
covered with loose aeolian gravel, in turn covered with rubble from collapse of in gray sterile gravel beneath, some
structures adjacent to excavated pit. charcoal and botanical remains.
F 3–8 Deep deposit of coarse light gray gravel, loose to slightly compacted, narrow No cultural material.
lenses of pale beige in successive superimposed layers.

Phases of occupation (Table 11jj, kk)

Phase Layers Description Dated samples
I A, D, E Includes material from occupation of Caballete in Layer
E through abandonment of site.
II Includes intrusive burials Features 1 and 2. A sample of plant fiber associated with the infant burial dated 940
Cal BC (ISGS-5724) (Table 11jj). A radiocarbon date from plant
fiber associated with the Feature 2 burial dated 680 Cal BC
(ISGS-5729) (Table 11kk).

CREAMER ET AL: FORTALEZA VALLEY, PERU 87

88
TABLE 22. Radiocarbon dates from Cerro Blanco 1.

Sample Provenience Level/Layer Material Lab number Cal BC RCYBP ± 13c/12c 68.2—1s range 95.4—2 s range
a Sector A, Profile 1 U(21) mixed plant fibers ISGS-5526 1690 3370 80 228.4 1747–1602 (76.9); 1590–1533 (23.1) 1883–1495 (100.0)
b Sector A, Profile 1 F(6) fiber bag GX-30515 1360 3110 70 210.8 1486–1485 (0.4); 1453–1294 (99.6) 1523–1193 (99.1); 1172–1168 (0.3);
1142–1133 (0.6)
c Sector A, Profile 1 L(12) fiber bag Beta-183317 1330 3080 70 225.6 1427–1265 (100.0) 1496–1187 (95.4); 1183–1154 (2.8);
1146–1130 (1.8)
d Sector A, Profile 2 fiber bag GX-30516 1190 2960 70 226.4 1291–1279 (4.0); 1270–1108 (78.3); 1393–996 (99.9); 982–981 (0.1)
1104–1056 (17.6)
e Sector A, south hole 2 fiber bag Beta-183316 1190 2960 70 224.2 1291–1279 (4.0); 1270–1108 (78.3); 1393–996 (99.9); 982–981 (0.1)
looters’ hole 1104–1056 (17.6)
f Sector B, Profile 3 fiber bag Beta-183318 1340 3090 70 224.6 1432–1287 (92.6); 1284–1269 (7.4) 1500–1189 (96.8); 1180–1157 (1.9);
1145–1130 (1.3)
g Sector C, TU 1 3 fiber bag ISGS-5716 1950 3600 70 227.9 2120–2095 (8.4); 2041–1879 (89.7); 2140–1754 (100.0)
1837–1832 (1.8)
h Sector C, TU 2 3 mixed plant fibers Beta-183319 1730 3420 70 225.5 1873–1844 (12.7); 1814–1800 (5.8); 1894–1598 (92.9); 1595–1531 (7.1)
1778–1629 (81.5)

FIELDIANA: ANTHROPOLOGY
 TABLE 22B

CREAMER ET AL: FORTALEZA VALLEY, PERU 89

 TABLE 23. Cerro Blanco 1, Sector A, Profile 1.

Stratigraphy (Fig. 31a, b)

Layer Description of stratigraphy Characteristics of layer
A Collapse and erosion on outermost margin of Angular rock in loose earth.
cut into Mound A.
B Additional material from collapse. Lens of loose soil with few small rocks.
C Wall 1, a retention wall. Angular rock in clay mortar, smooth, unpainted plaster surface.
D Wall 2, a retention wall. Angular rock in clay mortar covered with clay plaster, finished with smooth
layer of yellow plaster.
E Fill between Walls 1 and 2. Walls 1 and 2 are close together, probably constructed about the same time. Fill
between walls consists of angular rock, plant fiber, earth.
F Construction fill of shicra. Shicra made with braided plant fiber (stripped or beaten stems); braided
untreated whole plant stems. Substantial construction episode includes all
material between Wall 2 and Wall 3.
G Deposit running under Walls 1, 3, and 4 may be Fine, loose gravel, angular rock, plant remains.
among oldest deposits recorded in this unit.
H Wall 3, E side of rubble-filled structure. A row of large and medium-sized rocks set in midden and soil. Exterior (E) side
of wall covered with thick clay plaster.
I Fill between Walls 3 and 4. Large and medium-sized rocks with midden and soil.
J Base of Wall 3. Dark-colored clay and rock forming hard surface.
K Wall 4, W side of a rubble-filled structure. Large rocks in clay mortar with unpainted plaster on W side, toward center of
mound. Walls 3 and 4 were part of single large structure; Wall 3 plastered on E
side only, Wall 4 plastered on W side only, forming two sides of rubble-core wall.
L Shicra fill. Shicra, two types similar to Layer F above.
M Clay floor associated with the plaster surface of Wall 4. Floor on top of layer of fine gravel. Materials in Layer L deposited after Layer
M no longer in use.
N Fill to level area of shicra fill. Fine gravel leveled shicra layer over Floor M.
O Floor. Clay and gravel, burned patches visible. First of several occupation floors made
over gravel in Layer N.
P Deposit creating level surface over O. Lens of gravel laid down between floors.
Q Floor. Floor of untempered clay (no added plant material).
R Floor. Floor of untempered clay, extending upward onto Wall 5.
S Floor, remodeling of Floor R. Clay and gravel floor, also associated with Wall 5. Evidence of final occupation
on Mound A.
T Wall 5. Row of stones set in clay mortar with plastered face part of Floor R.
U Fill over Floor S. Rock, loose soil over Floor S, intentional or result of postabandonment
collapse of structures.
V Fill layer. Rock and plant fiber in dark soil.
W Wall W. Row of stones set in dark-colored hard clay mortar.
X Fill inside Wall 6. Large rocks in dark soil.
Y Floor Y. Dark-colored clay floor with yellow plaster surface at base of Wall 4. This may
be small piece of Floor M.
Z Floor Z. Dark-colored clay with yellow plaster surface.
AA Floor AA on top of G. Deteriorated clay floor. Floor AA touches Wall 4, does not appear to run under
the wall, no evidence that floor curves up along inside of Wall 4.
BB Prefloor above Layer G between Walls 2 and 3. Dark-colored gravel layer between floors.
CC Floor surface. Badly deteriorated clay floor, over prefloor BB.

Phases of occupation (Fig. 31b)

Phase Layers Description Dated samples
I O, P, Q, R, S, T, U, V, Wall 5 Final phase of occupation includes Wall 5 (T, V), associated Mixed plant fibers from Layer U
floors and layers O, P, Q, R, S. Layer S was later covered dated 1690 Cal BC (ISGS-5526)
by fill in Layer U. (Table 22a).
II F, J, M, Y, BB, CC, Phase IIa construction of double-sided, rubble-filled wall, Walls 3 One radiocarbon sample from
Wall 3, Wall 4 and 4, and fill between them, raising Mound A 2 m. Includes Layer F dated 1360 Cal BC (GX-
period when space created by plastered face of Wall 4 and floor 30515) (Table 22b). One sample
M in use. Wall 1, W of Layer F built at same time as Walls 3 of a shicra bag from Layer L
and 4. Layer J a footing for Wall 3. In addition to floor M, floor dated 1330 Cal BC (Beta-183317)
segments BB, CC, and possibly Y were in use in rooms on either (Table 22c).
side of double-faced wall. Phase IIb construction filled space
between Wall 4 and center of mound after room with Floor M
abandoned. Shicra (Layer L) raised mound substantially, in place
before seismic event toppled double faced wall (Walls 3 and 4).
Layer F perhaps deposited at this time. Layer N natural aeolian
layer or added after seismic event. Floor O created and then
covered by P, a layer of fine gravel apparently for leveling.

90 FIELDIANA: ANTHROPOLOGY
 TABLE 23. Continued.

Phases of occupation (Fig. 31b)

Phase Layers Description Dated samples
III G, Wall 2, Wall 6 Underlies extensive wall building (Walls 1, 3, and 4) episode prior to
seismic event. Wall 2 much shorter than Wall 6, topped with wall
plaster, indicating that it stands at finished height. Wall 2 may be
more recent than Wall 6. Alternatively, Wall 6 may originally have
been low, enclosing an extensive level floor area with Wall 1. Wall
6 may have been raised later. Wall 6 remains earliest construction
visible. Wall 1 may be original wall forming side of room with Wall
4 with Layer AA as floor. Wall 2 may be bench or retaining wall
added after Wall 1 to combat erosion or support Wall 1.
IV G, J, Y, AA, BB, CC Includes earliest use of area cleared, when Mound A consisted of
low platform. A floor was built on this mix of gravel and rock.
Several fragments extend across profile, including Y, AA, BB,
CC.

TABLE 24. Cerro Blanco 1, Sector A, Profile 2.

Stratigraphy (Fig. 33)

Layer Description of stratigraphy Characteristics of layer
A Thick layer including Wall 1. Large rocks, clay mortar.
B Two use surfaces. Floors 1 and 2 were compact, made of solid clay, without added fiber or
gravel, probably formed by sprinkling water on clay, stamping or
pressing the surface to flatten and smooth it.
C Leveling area prior to making Floors 1 and 2. Compact gravel leveled construction in its final stages, possibly over a
course of large rocks at base of gravel.
D Construction fill. Shicra bags filled with angular rock.

Phases of occupation
Phase Layers Description Dated samples
I A Final use of mound, when walls of angular rock, clay mortar were built on top of floors
constructed during Phase II. From limited area cleared, cannot determine whether
walls along side of looters’ pit were part of Phase II construction and occupation or
subsequent construction.
II B Clay floors built over layer of gravel. Floor 1 is remodeling of Floor 2, using same clay
and with no intervening deposits; indicates continuous or closely spaced occupation.
III C Leveling of shicra construction in final stages.
IV D Layer of shicra bags used in construction of body of mound, oldest activity identified. A date of 1190 Cal BC (GX-30516)
(Table 22d) obtained from Layer
D shicra fragment.

CREAMER ET AL: FORTALEZA VALLEY, PERU 91

 TABLE 25. Cerro Blanco 1, Sector B, Profile 3.

Stratigraphy (Fig. 34)

Layer Description of stratigraphy Characteristics of layer
A Disturbed material from looting. Angular rock in loose brown soil.
B Use surface or Floor 1. Compact clay.
C Layer under floor in Layer B, perhaps to level surface. Thin layer of gravel.
D Construction fill. Angular rock, shicra in substantial construction.
E Floor 2, possibly side of a room, as clay layer turns Compact clay.
upward onto side of unit.
F Prefloor or leveling event. Thin layer of gravel just below floor in Layer E.
G Floor 3. Compact clay, without gravel below in contrast to Layer E and
B floors.
H Construction fill. Thick layer of shicra and rock.

Phases of occupation
Phase Layers Description Dated samples
I A, B, C, Floor 1 Occupation of Floor 1, gravel layer below floor presumably used to
level surface. After use ended, Floor 1 gradually covered by wind-
borne material disturbed in recent times by looting.
II D Construction episode represented by thick layer of fill.
III E, F, G, Floors 2 and 3 Multiple phases of use or remodeling of floors. Corner of a room may
be exposed in Phase III.
IV H Construction episode. Deposits continue below area cleared. Layer H was dated 1340 Cal BC
(Beta-183318) (Table 22f).

TABLE 26. Cerro Blanco 1, Sector C, Test Unit 1.

Excavation (Fig. 35)

Level Thickness Description Contents
1 0–30 Large rocks, dark soil, reddish patch in NE that included Ceramics, lithics, shell, textile fragments, bone, heat-
burned ceramic fragments and no shell. fractured rock.
2 10–13 Soil with ash and angular rock. Botanical remains, textile fragments, lithics, shell.
3 19–29 Rock and soil, with depression or hole in S side of level Lithics, botanical remains, shell, two sherds, textile
holding textile fragments, charcoal. Culturally fragments, charcoal.
sterile soil in NE corner.
4 7–15 Rocks in loose earth. Lithics, shell, botanical remains in upper portion of level.

Stratigraphy
Layer Description of stratigraphy Characteristics of layer
A Concentration of rocks. Rock mixed with orangey soil lacking cultural material.
B Brown, gravelly soil. No cultural material.
C Similar color to A. Gravel and rocks with orangey soil lacking cultural material.
D Hard, compact soil. More compact than other layers, lacked cultural material.
E Midden deposit may have filled in a pit. Concentration of plant material, textiles, shell, loose earth, rock.
F Soil with ash, may be recent from adjacent fields. Concentration of shell, textile fragments, no other botanical remains.
G Midden deposit. Ash mixed with soil and shell, more than in Layer F. Some textile
fragments.
H Gravel and rock layer. Loose orangey soil with rock and gravel lacking cultural materials.

Phases of occupation
Phase Layers Description Dated samples
I A, F, G Uppermost layers identified by texture and color. Layers A and G may be
of similar origin with Layer F an intrusive pit or where wind and water
deposited material in subsidence. Each layer included soil with ash, large
quantity of shell, few rocks, midden, evidence of occupation nearby.
II E Intrusive deposit predates Phase I. Distinctive form deposit, as though a pit Level 3 Samples of plant fibers
were excavated and then filled with debris and additional material spread dated 1950 (ISGS-5716) and
around it, perhaps during process of refilling. Loose soil with ash, rock, 1730 Cal BC (Beta-183319;
burned textile fragments. Provides only evidence of human presence in unit. Table 22g, h).
III B, C, D, H, I Varied deposits of gravel in loose soil (B, H); rock in orangey soil lacking
cultural material (C, I); compact earth, culturally sterile (D).

92 FIELDIANA: ANTHROPOLOGY
 TABLE 27. Cerro Blanco 2, radiocarbon dates.

CREAMER ET AL: FORTALEZA VALLEY, PERU

Sample Provenience Material Lab number Cal BC RCYBP ± 13c/12c 68.2—1s range 95.4—2s range
a Trench, under Floor 2 mixed plant fibers from clay GX-30629 1700 3390 70 215.4 1861–1852 (2.9); 1771–1607 (91.9); 1880–1836 (8.1);
mortar 1570–1561 (3.3); 1546–1541 (1.9) 1835–1523 (91.9)
b Trench, Layer M fiber bag Beta-184860 1980 3630 80 210.1 2132–2083 (20.6); 2058–1893 (79.4) 2204–1766 (99.6);
1764–1758 (0.4)
c Trench, Layer H base mixed plant fibers from clay ISGS-5538 2130 3720 90 226.7 2280–2250 (9.0); 2230–2219 (2.8); 2456–2419 (2.2);
of Wall 2 mortar 2211–1012 (81.7); 1999–1978 (6.6) 2406–2376 (2.0);
2351–1889 (95.9)

93
 TABLE 27B

94 FIELDIANA: ANTHROPOLOGY
 TABLE 28. Cerro Blanco 2, Trench 1, main mound.

Stratigraphy (Fig. 38a)

Layer Description of stratigraphy Characteristics of layer
A Extensive deposit/construction phase without shicra bags or large rock. .1-m deposit of compact, gray gravelly soil, small rocks.
B Layer of rock intruding into Layer A, possibly part of Layer A Compact fine gray gravel with medium-sized rocks.
though different material.
C Distinct gravelly layer. Compact gravelly soil with medium-sized rocks, inclusions of
gray plant fiber.
D Thin layer of ash. Ash layer in fine, loose soil, possibly wind deposit.
E Layer of soil and cultural materials. Powdery beige soil, few cultural remains. Small pocket of dark
brown burned earth with powdery texture within Layer E.
F Thin layer of soil and caliche. Powdery light brown soil, few cultural remains. Patch of soil
hardened by salt deposits.
G Thin layer lacking cultural material. Fine gravel, rock.
H Includes all construction elements associated with Wall 3, Area of caliche, rock, patches of compact gravelly light brown soil,
running E/W along profile. patches of rock with clay mortar, fiber inclusions. Some mortar
adhered to rocks used in construction with rock and caliche.
I Layer may be continuation of Layer A, though radiocarbon Gravelly soil, medium-sized rock, inclusions of plant fiber.
date from lower deposits suggests Layer I was debris/fill
postdating construction of mound terraces formed by Walls 1–3.
At base of Layer I was Floor 1, poorly preserved, but evidence
of at least one replastering.
J Layer of soil over a floor of clay tempered with plant fiber. Medium-sized rocks in powdery soil. At base use surface, Floor
2, clay mixed with plant fiber, applied while wet. Floor 2
remodeled at least once.
K Layer below Floor 2. Compact light brown gravelly soil with shell, lithics, plant remains.
L Layer below Floor 2. Loose gray gravelly soil with shell, lithics, plant remains.
M Layer of construction material. Compact light brown soil, rock, shicra, cultural materials.

Phases of occupation (Fig. 38b)

Phase Layers Description Dated samples
I A, B, C During Phase I a retaining wall, Wall 2, was constructed and filled with
medium-sized rocks. Lowest level of fill included gravel, clay, plant fiber.
Caliche formed in the fill (Layers A and B). Wall 2 extended SSW from
Wall 3, creating a corner of mound with Wall 1, remnant of which was
identified in SE corner of area cleared. This area appears to have been
corner of terrace along side of mound visible from valley. No further large-
scale construction in area after this phase.
II D, E, F, G Layers E, F, G, narrow layer of ash and cultural remains above them (Layer
D). Layers tilted under Wall 1 and predate all subsequent construction
cleared in Trench 1.
III H, I, J, K, L, M Occupation during which two use surfaces (Floor 2) and two plastered A sample extracted from Layer J,
floors (Floor 1) were constructed, alternating with layers of fill Floor 2 dated to 1700 Cal BC (GX-
(Layers J–M) including shicra at lowest level. These surfaces and 30629) (Table 27a). Shicra from
floors lie beneath Layer I. Layer M under Floor 2 dated to
1980 Cal BC (Beta-184860)
(Table 27b). Plant fibers from the
clay mortar at base of Wall 2, Layer
H dated 2130 Cal BC (ISGS-5538)
(Table 27c).

CREAMER ET AL: FORTALEZA VALLEY, PERU 95

 TABLE 29. Cerro Blanco 2, Profile 1.

Stratigraphy (Fig. 39)

Layer Description of stratigraphy Characteristics of layer
A Aeolian surface deposit. Fine, powdery tan sediment.
B Mixed deposit mostly gravel, some soil. Materials that accumulated over time after final use of site. Light
brown to tan, no cultural material.
C Surface and lens of hard-packed soil, possible use surface Tan, fine-grained soil, charcoal flecks.
associated with adjacent looted zone.
D Gravel and aeolian deposits between compact soil layers. Appears to represent hiatus in use of burial zone.
E Thick layer of compact soil associated with use of Light brown, few cultural materials.
adjacent burial zone.
F Possible use surface, distinct layer below E. Tan gravelly soil.
G Gravel layer distinguishable from surroundings. Tan gravel.
H Gravel below ash Layer I. Tan gravel.
I Ash lens between gravel layers. May indicate burning associated with burial activities.
J Use surface. Brown soil with few cultural remains, possibly associated with
adjacent burials.
K Gravel matrix. No cultural material.

Phases of occupation
Phase Layers Description Dated samples
I C, D, E, F, G, H, I Surface associated with burial event, Layer D, representing final use No samples were dated from this
of area; no layers above it contained cultural material. Layers G profile.
and I also associated with a burial event. Both layers would postdate
occupation of mound structure. Material in Layer I may be remains
of fire associated with use surface or from later use.
II H, J, K Earliest use of area, indicated by brown soil, artifacts in Layer J use surface.

96 FIELDIANA: ANTHROPOLOGY
 TABLE 30. Huaricanga, Sector B, Profile 2.

Stratigraphy (Fig. 43; Fig. 44)

Layer Description of stratigraphy Characteristics of layer
A Construction fill. Angular rock, rectangular adobes, gravel, rock chips in matrix of
partly compacted beige soil with shell, plant fiber.
B Remodeling of floor associated with Wall 1. Clay floor extending along portion of wall and onto floor C.
C Floor associated with Walls 2 and 3.
D Base for Floor C. Fine gravel, compact soil.
E Construction of Wall 2. Partly compacted soil, fine gravel, medium-sized angular rock.
F Floor associated with Walls 2–4. Fine gravel, medium-sized rock with plant fibers.
G Fill between Wall 5 and Wall 6.
H Construction fill. Medium and large-sized rock with fine gravel and partly
compacted soil.
I Remodeling fill. Large rocks in fine gravel, soil, dried clay mortar fragments.
J Remodeling of floor associated with Wall 5. Section of clay floor extends from base of Wall 5 on top of Floor K.
K Floor associated with Wall 6; floor covers Wall 8. Clay floor starts at base of Wall 6 and top of Wall 9, extending across top
of Wall 8 forms thick floor layer, ending in rock fall part of Layer I.
L Construction fill from Wall 6; same fill covered Walls 7 and 9.
M Base of Floor K. Ash, fine gravel.
N Floor associated with Wall 8.
O Fill layer. Ash, fine gravel.
P Thick use surface. Clay.
Q Remodeling layer. Fine, gravelly fill.
R Floor. Clay.
S Remodeling fill. Fine gravel, small rocks, soil.
T Fragment of floor associated with Wall 8
U Floor fragment. Clay.
V Construction fill. Fine gravel, partly compacted soil.
W Floor associated with Wall 9.
X Floor associated with hearth.
Y Base of Floor X. Fine gravel.
Z Floor probably associated with Wall 7.
AA Layer below Floor Z. Mix of charcoal, fine gravel.
BB Fill below floor Z. Loose, light brown soil with fine gravel, small rocks.
CC Construction fill. Rock, clay mortar fragments.
DD Floor fragments. Clay.
EE Construction fill. Earth, fine gravel.
FF Remains of fire or material spread on floor surface. Ash, fine gravel.
GG Use surface.
HH Use surface.
II Construction fill. Fine gravel, coarse sand.
JJ Use surface. Clay.
KK Material between use surfaces, possibly to level surface. Fine gravel, compact clay.
LL Use surface.
MM Use surface or natural surface hardened by exposure to air and mist. Partially compacted soil with fine gravel, small stones.
NN May have been material used to level surface before construction Composed of partly compacted fine gravel, medium-sized stones.
of use surface (Layer MM).
OO Base layer of material on which construction began. Loose soil with fine gravel, small stones.
PP Below all construction. No cultural material, appears to be natural deposit.

Phases of occupation
Phase Layers Description Dated samples
I A Occupation following deposition of Layer A. Rock from collapsed structures
covers mound surface, walls do not extend into profile.
II B Layer A, over 1 m thick in most places. Construction episode represents
complete change in layout of structures on mound, possibly lapse in
occupation separating Phases I and II.
III C Construction and remodeling of series of structures. Initially, a room Layer G sample of plant fibers dated 2360
extended from Wall 7 to Wall 8, remodeled to extend between Wall 6 and Cal BC (Beta-AMS 183320)
Wall 4. Later remodeling reduced room to that visible between Wall 3 and (Table 31a), predates Walls 1 and 3,
Wall 4 with addition of fill layers G and L, later between Wall 1 and Wall 2, postdates Walls 5, 7, and 9.
reducing size of room from maximum 4 m to less than 2 m. Construction of
multiple floors visible in three other places along profile. Adjacent to the
rooms added to the N, Layer I ends in Wall 10; N of Layer I, Layers CC–LL
extend between Walls 10 and 11; N of this is less intensively occupied area
includes Layers KK–OO. One other set of floor construction, remodelings
visible in profile to the S (Y, AA, BB). These occupations should be divided
into a more detailed sequence.

CREAMER ET AL: FORTALEZA VALLEY, PERU 97

 TABLE 30. Continued.

Phases of occupation
Phase Layers Description Dated samples
IV Earliest phase exposed in profile, Phase IV can be subdivided into four units Sample of charcoal from Layer NN dated
extending across base of mound above layer of sand lacking artifacts, to 2510 Cal BC (ISGS 5525)
occupation. Horizontal stratigraphy includes Layer BB at S end of profile, (Table 31b).
thick layer of clay just above sand. Overlying N portion of BB is a floor,
Layer W, extending from Wall 6 over top of Wall 8, ending at base of
Wall 10. Lowest levels discontinuous at base of Wall 10; to N Layer LL
runs below Wall 11. At northernmost end of profile, Layer OO includes
additional fragments of floor running below other deposits.

TABLE 31. Radiocarbon dates from Huaricanga.

Sample Provenience Level/Layer Material Lab number Cal BC RCYBP ± 13c/12c 68.2—1s range 95.4—2s range
a Sector B, G mixed plant Beta-AMS 2360 3870 40 220.8 2456–2418 (26.6); 2467–2274 (88.7);
Profile 2 fibers 183320 2408–2374 (23.5); 2256–2208 (11.3)
2368–2361 (3.8);
2356–2292 (46.1)
b Sector B, NN charcoal ISGS-5525 2510 3970 110 224.6 2829–2823 (1.0); 2870–2802 (6.1);
Profile 2 2626–2291 (99.0) 2779–2197 (92.9);
2167–2149 (1.0)
c Sector B, A charcoal Beta-199061 2570 4020 70 225.0 2833–2818 (4.9); 2864–2805 (8.4);
TU 1 2662–2649 (3.7); 2760–2341 (91.6)
2635–2465 (91.4)
d Sector B, 4 charcoal GX-31585 2260 3810 60 225.6 2396–2395 (0.3); 2463–2130 (95.7);
TU 1 2391–2385 (2.0); 2086–2049 (4.3)
2345–2191 (80.8);
2180–2141 (16.9)
e Sector B, 5 charcoal Beta-184861 2700 4110 70 224.0 2862–2807 (24.4); 2880–2558 (92.0);
TU 1 2758–2718 (16.6); 2554–2550 (0.4);
2706–2577 (59.0) 2537–2491 (7.5)
f Sector B, 7 charcoal GX-30507 3550 4780 50 224.5 3639–3621 (16.5); 3654–3498 (87.0);
TU 1 3607–3522 (83.5) 3435–3378 (13.0)
g Sector B, Layer C, charcoal ISGS-A0481 2430 3940 40 213.3 2548–2539 (5.3); 2567–2520 (13.9);
Trench Sample B 2489–2427 (50.8); 2498–2332 (80.7);
1 2426–2400 (18.4); 2327–2299 (5.4)
2382–2347 (25.5)
h Sector B, Wall 4 mixed plant GX-30506 1040 2850 80 213.1 1128–901 (100.0) 1260–1226 (3.5);
Trench fibers 1223–835 (96.5)
1
i Sector C, mixed plant Beta-183322 2340 3860 40 223.6 2457–2418 (22.4); 2462–2269 (83.7);
Profile 1 fibers 2407–2375 (21.0); 2259–2206 (16.3)
2367–2362 (3.2);
2352–2285 (47.5);
2247–2235 (5.9)
j Sector C, fiber bag ISGS-5531 2440 3950 70 210.4 2568–2518 (24.0); 2830–2822 (0.5);
Profile 1 2499–2344 (76.0) 2629–2268 (95.3);
2260–2206 (4.2)
k Sector C, fiber bag ISGS-5514 2620 4030 70 210.4 2834–2817 (5.8); 2868–2803 (10.2);
Profile 1 2663–2647 (5.5); 2777–2429 (85.6);
2636–2468 (88.7) 2425–2400 (1.8);
2382–2347 (2.5)
l Sector C, mixed plant GX-30508 2800 4230 90 213.3 2918–2835 (36.3); 3086–3061 (1.3);
Profile 1 fibers 2817–2665 (62.0); 3029–2570 (98.1);
2645–2639 (1.7) 2514–2502 (0.6)
m Sector C, C charcoal ISGS-5519 2220 3770 70 225.1 2294–2124 (81.6); 2458–2417 (4.0);
Profile 3 2092–2043 (18.4) 2410–2022 (95.3);
1992–1983 (0.7)
n Sector C, G charcoal Beta-183321 2390 3910 40 221.7 2468–2390 (65.6); 2549–2538 (1.0);
Profile 3 2385–2345 (34.4) 2490–2284 (97.8);
2248–2234 (1.2)
o Sector C, 4 charcoal ISGS-5518 2400 3940 70 220.5 2565–2532 (15.1); 2620–2604 (1.1);
TU 2 2528–2525 (0.7); 2601–2592 (0.6);
2496–2338 (79.6); 2588–2205 (98.3)
2322–2309 (4.5);

98 FIELDIANA: ANTHROPOLOGY
 TABLE 31B

CREAMER ET AL: FORTALEZA VALLEY, PERU 99

 TABLE 32. Huaricanga, Sector B, Test Unit 1.

Excavation (Fig. 45)

Level Thickness Description Contents
1 4–15 Loose light brown soil; layer tilted slightly toward NE. Small rocks, plant fiber.
2 41–50 Initially designated Layer A, a natural layer consisting of compact gray
soil with small rocks. In some places fine gravel was part of layer,
and in other places were concentrations of marine shell and patches
of clay mortar. Field designation changed to Level 2, excavation ending
at 65 cm below datum on S side and 84 cm below datum on E.
3 0–16 Excavated to level the test unit, 16 cm in S and W of unit, very shallow in rest Dark patches of ash within level; cultural
of unit. material included shell, lithic
fragments, bone, botanical remains.
4 9–11 Compact light brown soil. Fragments of clay mortar, small and Despite the color change, contents of level
medium-sized rocks with some gravelly soil, especially in E corner. were similar to previous level.
On SW side soil appeared darker gray color than in NE corner.
Rock proved to be part of wall crossing the excavation unit.
5 28–43 Loose fine gray gravel with small stones and lumps of baked earth. At base, sterile soil was visible at SW end,
Wall running SE/NW divides level. distinctly lighter reddish color than
material above it, without any cultural
contents.
6 19–34 Excavated on NE side of wall that crosses short axis of unit, narrow Two large, shaped rocks uncovered beside
lens of fine brown gravel distinct and more compact than Level 5 Level 4 wall. Rocks likely fallen from
soil. Two other large stones uncovered in N corner may be part of wall.
another wall segment.
7 26–33 Uncompacted dark brown trash. Distinct level of trash deposition Marine shell, charcoal, botanical remains,
contained more cultural material than preceding levels. No additional very small rocks.
stones uncovered. Trash in Level 7 deposited over culturally sterile subsoil.

Stratigraphy
Layer Levels Description of stratigraphy Characteristics of layer
A 1, 2 Postdates us of area. Small quantities of shell, charcoal, botanical remains.
B 2–4 Postdates Wall 1. Layer B resulted in filling and leveling Marine shell, plant fiber, fragments of clay mortar.
area S of Wall 1, not clear whether this was intentional or
natural buildup from collapse and erosion.
C 2 Rock and mortar. Concentration of rock used to create upper surface of
platform ending at Wall 1.
D 2 Compact dark soil. Fill used to create platform incorporating Wall 1.
E 4 Mixture of rock and earth. Rocky soil used to create platform ending at Wall 1.
F 2 Dark soil and rock. Platform fill.
G 6, 7 Layer of fill distinguishable from Layer H below. Trash and fill, possibly placed to level area before
construction of platform.
H 4–7 First occupation of locality, resting atop distinct pale Distinctive layer of trash, thicker at E (downhill) side.
tan-orange soil with no cultural material. Higher-density cultural material than in layers above.

Phases of occupation
Phase Layers Description Dated samples
I A, B Layer B deposited over platform, completely covering it. Volume A sample of charcoal from Level 2 dated 2570 Cal
of fill suggests intentional covering/ leveling effort. Layer A is BC (Beta-199061) (Table 31c). A charcoal
natural accumulation postdating occupation of site. sample from Level 4 dated 2260 Cal BC (GX-
31585) (Table 31d).
II C–F Construction and occupation of platform. Occupation began with
accumulation of gravel, compact soil on ground surface created
by Layers G and H; may have been connected with construction
of Wall 1 or were earlier use surface. Wall 1 was retaining wall;
there are no associated plaster floors, the uppermost associated
fill is predominantly rock.
III G, H A period when this area was first used, with trash deposited on top A charcoal sample collected from screened
of culturally sterile soil. There are two components of this phase, contents of Level 5 dated 2700 Cal BC (Beta-
trash deposited in Layer H before Wall 1 was constructed may 184861) (Table 31e). A charcoal sample
have been very general trash disposal since layer shows evidence collected from screened material dated 3550 Cal
of sliding downhill. Later, more compact trash deposit in Layer BC (GX-30507) (Table 31f).
G may have served to level area prior to building Wall 1. Leveling
effect visible in contours of Layer E.

100 FIELDIANA: ANTHROPOLOGY

 TABLE 33. Huaricanga, Sector B, Trench 1.

Stratigraphy (Fig. 47a, b)

Layer Description of stratigraphy Characteristics of layer
A Stone collapsed from structures combined with aeolian deposits of soil, ash from Somewhat compact light brown soil, irregular
burning cane fields. thickness, large rocks, fragments of clay
mortar, plant fiber.
B, C Materials collapsed from Wall 1 over Floor 1, base of Layers B and C. Wall 1 Lumps of clay, plant remains in soft to
located in W corner. Wall I built of large regular-shaped stones without evidence slightly compact soil with little rock.
of polishing, though stones were shaped. Exposed part of wall was corner of
platform directly associated with Floor 1. Platform bounded by Wall 1 later
expanded to remaining edge of Floor 1. Floor 1 does not extend to base of
Wall 1, yet remains of plaster lip covering floor/wall joint were visible, even
though wall is absent. At E side of Layer C are Walls 3 and 6. Wall 3 located in
central part of trench near E profile. Only surface of wall was cleared, revealing
it was constructed of large cobbles held together with small quantity of mortar.
Wall 3 was remodeling of Wall 6, located in center of trench near E profile, an
alignment of angular rock in clay mortar. Wall 3 built up against Wall 6, both
partly covered by Floor 1.
D Area filled in where wall formerly stood adjacent to Wall 4. Wall 4 built in Gravel with a small quantity of plant fiber,
central part of trench near unit’s W margin. This consisted of wall made of few lithics.
angular rock, held together with mortar of clay mixed with plant fiber. No
remains of associated floor were identified, though this may have been in
location of Layer D.
E Material above Floor 2. Hard soil with very little cultural material, a
few mollusk shells, lithic fragments.
F Floor 2 uncovered in N end of trench between Wall 5 and Wall 2. Fire-reddened Compact, medium-fine beige soil with a few
in NW corner, only small portion of floor was cleared. Further excavation shell fragments, botanical fragments.
could show whether Floor 2 directly associated with Wall 5 or Wall 2. Wall 5
was located in N half of Trench 1, running WNW to ESE. Upper part of wall
made up of large stone blocks, some already collapsed. Many stones in walls
were near original position, leaning toward N. Wall continued across entire N
face of mound, all stone blocks forming wall are tilted same direction, suggests
a single event could have caused uniform displacement.
G Floor on platform created by Wall 7, retaining wall in extreme N end of trench, Very hard beige soil with numerous angular
perpendicular to and below Wall 2 at foot of mound. Wall 7 constructed of rocks of all sizes. A few shell fragments
large stones in clay mortar. Wall 2 at N end of trench, retaining wall along present.
lower part of mound. Wall was constructed of medium and large slabs mortared
with clay. Stones appear to have fallen from upper course, revealing flat,
unpolished faces. Wall was associated with fill in layers G and H, and Wall 7.
H Fill below Floor G. Highly compact layer contained a few specks
of charcoal.
I Fill below Layer H. Soil without cultural material.

Phases of occupation
Phase Layers Description Dated samples
I A Corner of low platform bounded by low wall or row of rocks.
II Wall 1, Floor 1, B, C Floor 1, associated with Wall 1 and edge of platform. Surface of A sample of charcoal from Layer C
floor highly deteriorated, very irregular, N side completely was AMS dated to 2430 Cal BC
destroyed. Gap between Floor 1 and Wall 1 may be remodeling (ISGS-A0481) (Table 31g).
or erosion.
III Wall 3, Wall 6 Wall 6 formed step or low platform parallel to Wall 1. Wall 3
appears to be later remodeling of Wall 6. Floor 1 originally
extended from
Wall 1 over Walls 3 and 6.
IV Layer D, Wall 4 Layer D between end of Walls 1 and 6, ending at Wall 4. A sample of plant fibers from N
side of Layer D near Wall 4
dated to 1040 Cal BC (GX-
30506) (Table 31h).
V Layer E, Wall 5 Wall 5 and collapsed stones, extend to Wall 2, margin of area cleared.
VI Layers F–I Floor 2 and material below it including Wall 7 and Floor G.

CREAMER ET AL: FORTALEZA VALLEY, PERU 101

 TABLE 34. Huaricanga, Sector C, Profile 1.

Stratigraphy (Fig. 48)

Layer Description of stratigraphy Characteristics of layer
A Fill at the top of unit. Steep slope minimally cleared of loose soil and rock. Area composed of fill, rock in dirt
matrix, no evidence of shicra bag construction. Clearing moved to lower, more stable area.
B Series of clay use surfaces, floors, Hillside’s irregular slope made small sections of floor visible in stepped arrangement, including
remodeling efforts. Floors a–f, each fragmentary. Some overlapped, revealing at least four layers of remodeling.
Tilt to floors as area eroded made it difficult to recognize separate fragments that may be part
of single floor collapsed unevenly. At base of floors were a series of rocks, remains of Wall 1.
C Construction fill. Shicra bags.
D Wall 2 at the base of Layer C. Potentially retaining wall for entire construction, only a portion preserved. Some rocks
very large, reminiscent of wall construction at Initial Period sites (e.g. Cerro Sechin),
where monoliths were incorporated into wall with intervening space filled by smaller
rocks, clay mortar.
E Series of use surfaces below Wall 2. As in Layer B, these all fragmentary, suggest period of construction, remodeling of use
surfaces and floors.
F Fragmentary floor surfaces 6a and 6b.
G Construction fill. Small rock, beige soil.
H Ashy soil below the floors in Layer G. Trash deposited as part of mound construction.
I Material around Wall 3. At top of Wall 3 a group of seven distinctive small clay tablets were uncovered (Fig. 50).
Fragments were found together in rubble, each carefully shaped and finished with
yellowish-cream or self slip. Tablets were unmarked, unique.

Phases of occupation
Phase Layers Description Dated samples
I A Final period of construction. With additional excavation, area
might reveal more detail about final phases of occupation.
II B, Wall 1 Occupation of mound over period during which Wall 1 and
adjacent floors constructed and remodeled at least six times.
Floors seem to have been supported by Wall 1, though the
remains of wall are scanty.
III C, D, Wall 2 Construction of Wall 2 and material between Walls 1 and 2. Radiocarbon dates of 2440 Cal BC (ISGS 5531)
and 2620 Cal BC (ISGS 5514) (Table 31j, k)
were obtained from Layer C. A date of 2340
Cal BC (Beta-183322) (Table 31i) was
obtained from Layer D below Wall 2.
IV E, F, G, H Construction of Wall 3 and material between Wall 2 and A sample of plant fiber from Layer F dated 2800
Wall 3. Cal BC (GX-30508) (Table 31l).
V I, Wall 3 Construction from base of mound to Wall 3.

102 FIELDIANA: ANTHROPOLOGY

 TABLE 35. Huaricanga, Sector C, Profile 3.

Stratigraphy (Fig. 52)

Layer Description of stratigraphy Characteristics of layer
A Layer of surface material from collapse of Medium, large-sized angular rock, loose dark gray soil, wind-blown sand, marine
structures built on mound. shell fragments, botanical remains. Contemporary trash in uppermost layer
comes from proximity of highway and grazing goats.
B, Wall 1 Construction of Wall 1 and fill behind it. Small stones, clay mortar with plant fiber, a mix often used as plaster, though
Wall 1 appears to be retaining wall filled material is no longer attached to wall. Wall 1 built of medium-sized and large
with Layers B and C as part of mound rocks with flat exterior facing NW, clearly visible from highway (Fig. 53).
construction. Visible wall runs SW to NE 5 m, 60 cm high.
C Construction fill with a lens of compact clay Medium-sized rocks, clay mortar.
mortar.
D Layer of rocky fill. Rocks of different sizes in loose beige soil.
E Compact soil layer with fine gravel, small rocks. Shell fragments, plant fiber.
F Construction fill. Medium-sized rocks, clay mortar fragments.
G Series of floors, use surfaces, fill between floors, Ash, fine gravel in alternating layers.
considered single layer, appear to be successive
remodelings or treatments of surface.
H Original ground surface. Soil with no cultural material.

Phases of occupation
Phase Layers Description Dated samples
I A Postdates final occupation of mound, consisting of material collapsed from
mound itself, structures that were on top of it.
II B, C, Wall 1 Period of time when Wall 1 was in use. A date of 2220 Cal BC obtained
from charcoal sample in Layer C
(ISGS-5519) (Table 31m).
III D, E, F Construction episode associated with fill in Layers D–F, prior to construction
of Wall 1.
IV G Earliest occupation of area visible at base of roadway, Layer G. At least seven A charcoal sample from this layer
floor fragments identified in profile, patches of ash may indicate presence of dated 2390 Cal BC (Beta-183321)
hearths. Area may have been residential prior to start of mound construction. (Table 31n).

CREAMER ET AL: FORTALEZA VALLEY, PERU 103

 TABLE 36. Huaricanga, Sector C, Test Unit 2.

Excavation (Fig. 54)

Level Thickness Description Contents
0 0–8 Surface disturbance included aeolian loose grayish brown soil. Shell, lithics, modern trash.
1 16–21 Loose soil with numerous rocks, some charcoal, ash. Substantial quantity of marine shell comprised
cultural material. Level not uniform, reddish soil
in NW portion, compact earth with rock in S half.
2 3–17 Reddish soil appeared to be result of burning. Associated with ash, plant remains, shell, one ceramic
fragment.
3 0–20 Compact surface associated with broken shell, probably use
surface, as shell was more finely broken than in previous
levels. Inclined surface of deposit also noted.
4 8–18 A zone of ashy soil and cultural material that was less compact Base of unit was reached over portion of area where
than the material in Level 3. soil was reddish and without cultural content.
5 18–33 Included base of pit noted as Phase VI. Gravel with some charcoal, few lithics.

Stratigraphy
Layer Levels Description of stratigraphy Characteristics of layer
A 0 Contemporary trash, corn, animal dung from seasonal Loose, powdery soil on surface, including shell, lithics.
encampment of herders.
B 1–3 Loose soil with abundant medium-sized rocks, large quantity Fire-cracked rock, charcoal, shell, very few botanical
of shell, some ash. Several different activity areas, or corner remains. Within Layer B, Feature 4 was pocket of burned
of burned area with associated work area of unburned, organic material, dark brown color, fragments of burned
compacted material. These activities could be historic or fiber visible. Layer of reddened soil above suggests
recent due to location of unit beside highway. Feature 4 was remains of a hearth.
C 1–3 Partially compacted soil containing plant remains, lithics. Some stone mixed into this layer along with smaller fragments
Feature 3 within this layer was an area of compact clay, of charcoal, shell suggests a structure. Form was not
possibly mortar. distinctive, test unit not extensive enough to be certain.
D 1–3 Distinct layer of concentrated ash, charcoal. Layer was darker than those adjacent, contained greatest
variety of cultural material, charcoal, shell, lithics, botanical
remains. Some lithics appeared to be heat-fractured. Layer
covered entire unit, closest to surface on N, deepest on S.
E 3–5 Deposit of fill, appears to have been borrow pit, pit feature Rock, plant fiber, shell.
filled with debris, trash for disposal or to fill and level
area. Features 5 and 6 are thin lenses of ash in Layer E
extend into deposits suggesting they were individual deposits
of ash in pit filled with mixed debris. Material in Layer E
not dark, making lenses of ash stand out.
F 3–5 Material appears to have been cut by pit feature in Layer E, Consistency of the soil different from other levels, highly
this layer found only on N side. compact, many fragments of fractured rock. Feature 7 is
thin layer of ash within Layer F. Contents of feature
similar to Layer D, organic material, shell, lithics, ash.
G 4, 5 Material below pit included as Layer E. At N end, layer consisted of orangey, fine sand, no cultural
material. At center and S side pit, layer into which most of
Layer E was cut, no cultural material, but soil was dark gray.

Phases of occupation (Fig. 53)

Phase Layers Description Dated samples
I B, C Accumulation of rock, soil some cultural material after area ceased use for trash (Layer D).
At S end of profile, Layer C, materials in Feature 3 may indicate construction, but no
distinctive structural elements found. Phase ended with accumulation of fill including
cultural material, possible debris from collapse of nearby structures, or from recent road
work.
II D Renewed use of locality for disposal of organic material, household trash, rich in This layer dated to 2400
charcoal, botanical remains, distinctive dark color. Material in Layer D deposited on Cal BC (ISGS-5518)
flat, sloping ground surface, top of Layer E. (Table 31o).
III E Period when borrow pit was filled with debris, trash associated with habitation. Surface
was used long enough to form distinct, even surface, suggests lapse between Layer E
formation and all Phase I and II deposits.
IV F, G During first use of locality, borrow pit was excavated into original ground surface (Layer
G). Layer F may represent activity prior to pit, but small portion uncovered that not
sufficient to show extent of deposits and whether Layer F represents even earlier activity.

104 FIELDIANA: ANTHROPOLOGY

 TABLE 37. Radiocarbon dates from Shaura.

Sample Provenience Level/Layer Material Lab number Cal BC RCYBP ± 13c/12c 68.2–1s range 95.4–2s range

CREAMER ET AL: FORTALEZA VALLEY, PERU

a Trench 1 Layer 2 charcoal ISGS-5522 1330 3080 70 226.4 1427–1265 (100.0) 1496–1187 (95.4);
1183–1154 (2.8);
1146–1130 (1.8)
b Trench 1 Layer 3 charcoal Beta-183327 2050 3660 60 223.9 2134–2078 (34.0); 2201–1889 (100.0)
2062–1953 (66.0)
c Test Unit 1 Level 3 charcoal Beta-183326 1880 3540 60 225.6 1950–1862 (54.0); 2031–1737 (98.6);
1851–1772 (46.0) 1711–1695 (1.4)

105
 TABLE 37B

TABLE 38. Shaura, Trench 1, main mound.

Stratigraphy (Fig. 57)

Layer Description of stratigraphy Characteristics of layer
A Modern disturbance. Medium-sized rocks, cobbles in light brown soil.
B Fill, some cultural material. Medium-sized cobbles, small rocks in light brown soil.
C Wall. Large cobbles mixed with loose dark brown soil.
D Use surface extending from top of Wall 2 to base of Wall 1. Compact light brown soil.
E Wall. Large cobbles, some melted adobe may be remains of mortar.
F Use surface created above single layer of cobbles. Fine soil over cobbles included shiny particles, probably mica.
Polished boulder located on top of this surface.
G Culturally sterile layer. Orangey-brown hard-packed soil.

Phases of occupation
Phase Layers Description Dated samples
I A–F Trench 1 revealed one phase of occupation of main mound. Segment of wall A date of 1330 Cal BC (ISGS-5522)
with broad, plastered surface extended out in Layers E and F. Segment (Table 37a) from sample of charcoal
appears to be base of step in overall construction of step-sided structure. from Layer B. Charcoal from Layer C
Irregular, polished boulder on step, its black, fine-grained volcanic rock dated 2050 Cal BC (Beta-183327)
distinct from rock used in wall construction, indicating an intentional (Table 37b).
placement (Fig. 58).

106 FIELDIANA: ANTHROPOLOGY

 TABLE 39. Shaura, Test Unit 1.

Excavation (Fig. 59)

Level Thickness Description Contents
0 5–20 Aeolian soil, highly disturbed. Fine ash, charcoal.
1 7–28 Section of wall identified bordering SE side (Wall 1). Second wall segment Lithics, shell fragments, charcoal.
crossed NW side (Wall 2), cobbles, angular granite. Remainder of level loose
soil, river cobbles, remnants of clay mortar on some stones in wall. Level 1 ended
with patch of thick mortar in irregular deposit, melted off Wall 2. Wall built on
top of ashy layer, did not extend below base of Level 1.
2 8 Fine gray ash across entire unit 4–5 cm thick, also filled depression in surface of Shell, charcoal.
level. May be remains of hearth, but very limited cultural contents suggest ash
was dumped rather locality a living or work space.
3 20 Dry soil mixed with ash. Cultural materials decreased toward base of level. Shell, lithics.

Phases of occupation
Phase Levels Description Dated samples
I 1 Deposit of fill above ashy layer identified in Level 2, associated walls including
construction of Wall 2.
II 2, 3 Includes ashy layer, materials below layer. Top of Level 2 ash may have been use Charcoal from upper portion of
surface. Below it, cultural materials in Level 3 diminished and excavation ceased. level dated to 1880 Cal BC (Beta-
183326) (Table 37c).

CREAMER ET AL: FORTALEZA VALLEY, PERU 107

 TABLE 40. Calibrated date intervals from all Fortaleza Valley sites tested in 2003–2004.

108 FIELDIANA: ANTHROPOLOGY