Professional Documents
Culture Documents
Pozorski, Sheila y Thomas Pozorski 2005 Architecture and Chronology at The Site of Sechín Alto, Casma Valley, Peru
Pozorski, Sheila y Thomas Pozorski 2005 Architecture and Chronology at The Site of Sechín Alto, Casma Valley, Peru
Architecture and Chronology at the Site of Sechín Alto, Casma Valley, Peru
Author(s): Thomas Pozorski and Shelia Pozorski
Source: Journal of Field Archaeology, Vol. 30, No. 2 (Summer, 2005), pp. 143-161
Published by: Maney Publishing
Stable URL: http://www.jstor.org/stable/40024939 .
Accessed: 10/10/2013 15:06
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Maney Publishing is collaborating with JSTOR to digitize, preserve and extend access to Journal of Field
Archaeology.
http://www.jstor.org
Thomas Pozorski
Shelia Pozorski
University of Texas-PanAmerican
Edinburg, Texas
Figure 2. View of the main mound of Sechin Alto from the east.
Alto is300x250m and 35 m in height and is the largest height occupies the center of the mound summit. Conical
mound construction for its time period in the New World. adobe brickswere also occasionallyused to construct stair-
Four large rectangularplazas demarcatedby massive stone cases and the upper portions of stone walls.
walls and smallerstone mounds extend almost 1200 m east
from the main mound. There are also two sunken circular PreviousInvestigations
plazas, one in the second rectangularplaza and one in the Surface survey of Sechin Alto site was undertakenover
fourth rectangularplaza. the past 65 years (Fung and Williams 1977: 116-120;
The general configuration of the mound includes a cen- Kosok 1965: 214-215; Tello 1956: 79-82; Thompson
tral staircasethat leads from a lower atrium flanked by 10 1962: 294) and Collier excavated two test pits there in
m high wings to an upper atrium, which is also bordered 1956 (Collier 1962: 411). Most investigators agreed on an
by wings on the north and south and by the mound sum- "early"date for the site based on the form and layout of its
mit to the west (fig. 8). Between the lower and upper atria surface architectureand the small amount of cultural ma-
there is an increasein elevation of about 11.5 m. A second terial excavatedby Collier.
long staircaseleads to the mound summit at an elevation
some 12.75 m above the upper atrium floor. Most con- Current Investigations
struction is of stones quarriedfrom nearbyhillsides and set Eight field seasons (1995-2002) of excavationsupport-
in silty clay mortar. A solid rectangular "core" made of ed the early dating, and led us to assign the bulk of Sechin
cone-shaped adobe bricksheld in place by silty clay mortar Alto mound construction to the Initial Period (table i).
and measuring some 90 m n-s by 30 m e-w and 9 m in The general form of the main mound clearly indicates an
Figure 3. Plan of the Sechin Alto Complex showing the location of its four component sites, Cerro Sechin, Taukachi-Konkan,Sechin Bajo,
and Sechin Alto as well as A) the south wing of the first plaza east of the main mound, and B) the domestic area of Sechin Alto.
Initial Period construction, but most architecturaldetail re- aid of a Peruvian codirector, 2 to 4 students, and 7 to 20
lated to that construction is unclear. The main difficulty Peruvianworkmen each season. Excavationunits were 2 m
stems from a major Early Horizon (1000-200 cal B.C.)re- squares, and earth moving was accomplished using picks,
occupation of the mound summit after a hiatus in occupa- shovels, and trowels. Excavated material within 10 cm of
tion of about 500 years. These new occupants left volumi- floors and within features was screened through 1/4 inch
nous midden deposits as well as structures built with ma- screen, representativesamples were screened through #10
terial taken from the earlierInitial Period constructions. and #25 geological soil screens, and pollen and radiocar-
Our excavations at Sechin Alto focused on the delin- bon samples were collected from appropriate contexts.
eation of access patterns (staircasesand entrances) of the Workmen readily carried stones weighing up to 130 kg;
Initial Period architecture,stratigraphicexcavation of por- larger stones were moved using ropes and/or solid poles as
tions of the mound to establish chronological controls, de- rollers or skids. Moving the largest stone required 16
tailed examination of specific areasto obtain functional in- workmen.
formation pertaining to the Initial Period occupation, and
horizontal clearing of some Early Horizon architectureto Chronological Sequence at Sechin Alto
better understand that reoccupation. These excavations Six phases of construction and occupation were docu-
were accomplished in eight 10-week field seasons with the mented. Construction of the mound took place during the
Figure 4. Decorated pottery from the Moxeke Phase of Sechin Alto also found at Pampa de las Llamas-Moxeke. Scale is cm.
Figure 5. Decorated pottery from the Haldas Phase of Sechin Alto. It is also found at Las Haldas. Scale is cm.
Figure 6. Polished slate points from the Early Horizon reoccupation of Figure 7. Pottery from the Early Horizon reoccupation of Sechin Alto,
Sechin Alto, also found at San Diego and Pampa Rosario. also found at San Diego and Pampa Rosario.
Moxeke Phase which spanned most of the Initial Period found in association with Moxeke Phase A architecture,in
and can be subdivided into Moxeke Phases A and B be- the overlying construction and fill pertaining to Moxeke
cause two major construction episodes are clearly evident Phase B, and within a domestic areain the se corner of the
(there is no discernible change in associated artifacts).Ra- Sechin Alto Complex (fig. 3B). Architecture dating to
diocarbon dates suggest approximatetime spans of 2150- Moxeke Phase A was exposed in four areas of the main
1500 cal B.C.and 1500-1400 cal B.C.respectivelyfor these mound: the conical core, the depression east of the adobe
subphases.The remainderof the Initial Period, from about core, the north wing of the upper atrium, and the lowest
1400 to 1000 cal B.C., has been designated the Haldas segment of the central staircase.Moxeke Phase B architec-
Phase to distinguish a new cultural presence. These Initial ture was exposed on the mound summit, both wings of the
Period construction phases are followed by the EarlyHori- upper atrium, and on an intermediate-sized mound bor-
zon reoccupation as well as later Middle Horizon and dering the first plaza.
"Transitional"Period ephemeraluses of the mound.
MOXEKE PHASE A
MoxekePhase Moxeke Phase A, the earliest period of occupation yet
The Moxeke Phase as a whole can be dated to between identified at Sechin Alto, saw the building of a significant
2150 and 1400 cal B.C. based on radiocarbon dates from portion of the main mound, possibly as much as two-thirds
Pampade las Llamas-Moxeke,Taukachi-Konkan,Bahia Se- of its 2,000,000 cu m final volume. The adobe core was
ca, and Cerro Sechin. Sechin Alto site dates fall near the likely the tallest part of the mound summit at this time
end of this phase, between 1600 and 1400 cal B.C. (table (fig. 8a-c). It was originally a solid construction standing
i), reflecting the fact that our excavationswere confined to at least 9 m above the adjacentmound surface.Excavation
the upper half of the mound. Moxeke Phase ceramicswere of an intact portion of the west side (fig. 8c) revealedthat
Figure 8. Plan of the main mound of the Sechin Alto site showing: A) North end of the adobe core; B) East
exterior face of the adobe core; C) West exterior face of the adobe core; D) Deep depression east of the
adobe core; E) North wing of the upper atrium; F) Staircasesystem between the lower atrium and the upper
atrium; G) Corridor leading to the summit of the adobe core; H) Summit room; I) Staircasesystem between
the upper atrium and the summit room; J) South wing of the upper atrium; K) Early Horizon architecture
on the mound summit.
the 9 m height included both a bench which is 3.5 m tall ed excavation of a pair of partiallypreservedrooms associ-
and 4.7 m wide and the 5.5 m tall wall segment above the ated with Pampade las Llamas-Moxeke type pottery (figs.
bench. Tracesof square columns were found at the adobe 8b, 8d, 9). One of these is a square-room unit, a modular
core's nw and ne corners. Polychrome friezes decorate building form composed of a square room with round ex-
remnantsof three columns in the nw corner. Each frieze is terior corners that denotes administrative architecture at
distinct, but not enough remains to determine precise mo- Pampa de las Llamas-Moxeke and Taukachi-Konkan(Po-
tifs. Two samples of wood from postholes within the zorski and Pozorski 1994: 53). The square-room unit at
columns yielded dates of 1540 ± 60 and 1410 ± 55 cal Sechin Alto rests on a platform made of conical adobe
B.C. (table i). We believe these square column remnants bricksthat stands at least 5 m above a floor that appearsto
were part of two long colonnades, perhaps containing as be part of a small courtyard or open area. Charcoal from
many as 45 columns each, which lined the east and west this floor yielded a date of 1510 ± 60 cal B.C. (table i).
sides of the adobe core surface (fig. 8). Moxeke Phase A architecturewas also discovered with-
East of the adobe core a 15 m deep depression facilitat- in the wing north of the upper atrium (fig. 8e). A 3 m
Beta-138056 3240 ± 50 1290 ± 50 1510 ± 55 wood Within a pilaster of an entrance in the wing structurenorth of the upper-
atrium, Haldas Phase
Beta-124948 3240 ± 60 1290 ± 60 1510 ± 60 charcoal Within Sq. 2, level 15, domestic area in se corner of Sechin Alto Complex,
Moxeke Phase B
Beta-150768 3220 ± 60 1270 ± 60 1500 ± 55 charcoal Within fill 85-145 cm below surface of Initial Period gray plaster floor in
and wood wing structurenorth of the upper atrium, Moxeke Phase B
Beta-150766 3170 ± 60 1220 ± 60 1430 ± 50 wood Within a pilaster of an entrance in the wing structurenorth of the upper
atrium, Haldas Phase
Beta-138058 3160 ± 60 1210 ± 60 1425 ± 50 charcoal Within a hearth near floor of Haldas Phase rooms in wing structurenorth of
the upper atrium
Beta-124947 3150 ± 60 1200 ± 60 1420 ± 35 charcoal Fill materialwithin 10 cm of floor of square-room unit on adobe platform
within deep room, Moxeke Phase B
Beta-124945 3140 ± 60 1190 ± 60 1410 ± 55 wood Within a posthole in column in the nw corner of adobe core, Moxeke Phase A
Beta-150767 3110 ± 70 1160 ± 70 1400 ± 70 wood Within 7 postholes in Initial Period gray floor in the wing structurenorth of
the upper atrium, Moxeke Phase B
Beta-164488 3090 ± 70 1140 ± 70 1390 ± 70 charcoal Fill material 0-40 cm below room floor within south wing of first main plaza
east of main mound of Sechin Alto, Moxeke Phase B
Beta-172353 3090 ± 60 1140 ± 60 1390 ± 65 charcoal Fill material 50-60 cm below corridor floor, in domestic area in se corner of
Sechin Alto Complex, Haldas Phase
Beta-110594 3080 ± 60 1130 ± 60 1380 ± 60 charcoal Within Square 1, level 18/19, domestic area in se corner of Sechin Alto
Complex, Moxeke Phase B
Beta-138057 3050 ± 70 1100 ± 70 1305 ± 95 wood Post within corridor in wing structurenorth of the upper atrium, Haldas
Phase
Beta-124946 3040 ± 60 1090 ± 60 1295 ± 95 charcoal Fill material 1.5 m above the floor of the square-room unit on adobe platform
within deep room, Moxeke Phase B
Beta-172354 3010 ± 70 1060 ± 70 1270 ± 125 charcoal Fill material 0-30 cm below floor #4 of room in south wing of first main
plaza east of main mound, Sechin Alto, Moxeke Phase B
Beta-172351 3000 ± 60 1050 ± 60 1260 ± 115 junco plantWithin a column hole of a room within the wing south of the upper
fiber rope atrium, Moxeke Phase B
Beta-138059 2930 ± 60 980 ± 60 1120 ± 110 charcoal Within an intrusive hearth located above Haldas Phase architecturein wing
structurenorth of the upper atrium
Beta-150765 2860 ± 60 910 ± 60 1010 ± 90 wood East post of main south entrance in the wing structurenorth of the upper
atrium, late Haldas Phase
Beta-16448 7 2760 ± 60 810 ± 60 900 ± 70 charcoal Within an intrusive hearth on low platform in south wing of first main plaza
east of main mound, Sechin Alto, late Haldas Phase
Beta-110591 2210 ± 60 260 ± 60 290 ± 70 charcoal Within an intrusive hearth in Early Horizon midden layer overlying Moxeke
Phase B architectureeast of the east end of the upper atrium
Beta-150769 2110 ± 60 160 ± 60 160 ± 75 charcoal Within intrusive hearth overlying Haldas Phase architecturein wing structure
north of upper atrium
* Periods in this
paper are based on calibratedradiocarbondates (Stuiver et al. 1998).
wide staircasesystem made largely of conical adobe bricks to a well-preserved plaster floor that forms part of a large
partiallyunderlies a large room block associated with Las room or courtyard.The northern end of this staircaseand
Haldas type pottery (fig. io). The south portion of the associated room are covered by 3.75 m of rocky construc-
staircaseleads north from the upper atrium up to a land- tion fill dating to Moxeke Phase B, upon which rests the
ing. At the north end of the landing the staircasedescends later Haldas Phase construction.
Figure 9. View from the sw of the buried Moxeke Phase A adobe platform supporting remains of a
square-roomunit and a rectangularroom.
Two staircasesystems currently ascend the east face of eke Phase B construction was exposed in the adobe core
the main mound along its centralaxis. Only the lower stair- area,the summit room, the centralstaircase,and the north
case system, rising from the lower to the upper atrium and south wings of the upper atrium. Moxeke Phase B ar-
(fig. 8f), was in use during Moxeke Phase A. Two con- chitecture was also encountered during excavation of a
struction phases are evident in this staircaseand they likely small mound bordering the first plaza (fig. 3A), and the
pertain to Moxeke Phases A and B. A short segment of the midden and residentialarchitectureto the se also probably
earlier staircase was exposed at the bottom of the lower date to this phase (fig. 3B).
staircasesystem where four of its steps extend east beyond The adobe core continued to be used during Moxeke
the west wall of the lower atrium. Higher up, this staircase Phase B when all four sides were surrounded and covered
was inset, and its size and configuration were very proba- by stone and mud fill which raised the height of much of
bly like the Moxeke Phase B staircasethat currently over- the mound some 9 m, up to the height of the adobe core
lies it. surface.The few diagnostic ceramicsfound in this fill were
all Moxeke Phase, and two dates for the construction fill of
MOXEKE PHASE B 1420 ± 35 and 1295 ± 95 cal B.C.were crucial in dating
The main mound and its associated four rectangular the upper construction fill of the mound to Moxeke Phase
plazasand two sunken circularplazas attained their present B (table 1). These dates establish a lower bracketingdate
configuration during Moxeke Phase B. Evidence of Mox- for the Initial Period architecturebuilt on the fill.
Figure 11. Plan of the main staircasesystem. A) Staircasefrom the lower atrium to the upper atrium; B)
Staircasefrom the upper atrium to the summit of the main mound.
The second room is entered from the east side of the ve- interaction between the two cultures, however, until about
randa.Here all that remainsof a much largerroom is a par- 1400 cal B.C.when Las Haldas ceramics and cultural ma-
tially preserved bench supporting a low, square platform terial appearedwell north of Las Haldas.
abutting the main mound. South of this platform is a col- The Las Haldas Culture expanded northward along the
umn hole and a remnant of protruding plaster that may be coast and inland into the Casma Valley during the Haldas
the remains of a frieze. phase. Haldas Phase building activity on the main Sechin
On the north wing, a rough floor of gray silty plasterup Alto mound occurs mainly on the north wing above the
to 12 cm thick covers the surfaceof the 3.75 m deep Mox- Moxeke Phase B gray silt floor where a large room and
eke Phase B fill of the Moxeke Phase A room. Charcoaland smaller room block were constructed (fig. 8e). The large
wood in this fill dated to 1500 ± 55 cal B.C.(table i). The room is formed by a wall of large stones that borders the
fill yielded a stylized bird-head pendant of unidentified north wing on the north, east, and south. Excavations
green stone, but no diagnostic pottery. Imbedded in the within this room revealed a 3.4 m wide south doorway in-
gray silt floor is an alignment of seven small postholes, to a corridor bordering a block of contiguous stone-walled
10-12 cm in diameter, that were covered by later Haldas rooms around a small courtyard (fig. 8e, io). Entry into
Phase architecture.Wood fragments from these postholes the courtyardcomplex is by way of two narrow doorways
were dated to 1400 ± 70 cal B.C. (table i). Moxeke Phase in the rearof the structureand through a small room (fig.
B ended before any structures were built on the north io). Within the courtyarda central,rectangular,stone-lined
wing. depression (1.5 x 1.6 m) contains an offset circularhearth
Portions of two rooms forming the south border of the some 50 cm in diameter.Given its large size and prominent
first plaza have interior round corners like most square- position within the complex, this hearth could reflect a lo-
room units of the Moxeke Phase (fig. 3A).There is also calized ritualsomewhat like the ventilated hearth structures
ample gray silt floor and wall plasterthat matches the Mox- of the Moxeke Phase (Pozorski and Pozorski 1996).
eke Phase B gray silt floor of the upper atrium north wing, The courtyard and room complex superficially resem-
and Moxeke Phase sherds were recovered in the excava- bles the layout of intermediate-sizedmounds at Pampa de
tions. Dates of 1390 ± 70 and 1270 ± 125 cal B.C.(table las Llamas-Moxeke (Pozorski and Pozorski 1992: fig. 8,
i) from charcoalrecovered from fill below the floor of one 1994: fig. 3); however, close inspection reveals significant
of the rooms also suggest that these structures date to differences. There are no modular square-room units; in-
Moxeke Phase B. stead, the Haldas Phase rooms are rectilinearwith shared
Excavationsduring 1995 and 2002 within a residential walls and the main entrance is in the rear. Some continu-
area southeast of the main mound uncovered ample evi- ities with earlier architecture can be seen in the use of
dence of midden and some structures dating to the Mox- round corners, raisedthresholds, and paired pilasterswith-
eke Phase (fig. 3B). Only Moxeke Phase pottery was re- in doorways. Las Haldas type ceramics were recovered
covered from these excavations, along with numerous fig- from these rooms, the surrounding corridor,and the south
urine fragments, stone bowl or mortar fragments, twined entrance. A solid ceramic cylinder seal from one room is
and woven textiles, and abundant subsistence remains. similar in form to examples associated with intermediate-
Two radiocarbondates of 1510 ± 60 cal B.C.and 1380 ± sized mounds at Pampa de las Llamas-Moxeke (Pozorski
60 cal B.C. (table i) from charcoal recovered during mid- and Pozorski 1991: fig. 8). The seal and the general con-
den excavations place this occupation within Moxeke figuration of the room complex suggest its primaryfunc-
Phase B. tion was administrative.Haldas Phase people may have at-
tempted to simulate the square-room unit, which symbol-
Haldas Phase ized administrativefunction during the Moxeke Phase, as
Both the Moxeke and Las Haldas culturaldevelopments a means of signaling their administrativepresence.
have long histories within the Casma Valley area.The ear- Four radiocarbon dates come from the Haldas Phase
liest dates for the Moxeke occupation come from the site construction on the north wing (table i). Two dates of
of Pampade las Llamas-Moxeke,which was inhabited from 1510 ± 55 and 1430 ± 50 cal B.C.are from wooden posts
2080 to 1340 cal B.C.The Las Haldas occupation of the within entrance pilasters of the courtyard/room complex.
Casma Valley area is dated to between 1925 and 1410 cal A third date of 1305 ± 95 cal B.C.is from a wooden post
B.C.at the coastal site of Las Haldas (Pozorski and Pozors- embedded in the floor of the corridor south of the court-
ki 1992: table 2). Thus, the Moxeke and Las Haldas Cul- yard/room complex. With two-sigma standard deviation,
tures appeared at about the same time and coexisted for these three dates suggest a date of ca. 1400 cal B.C. as the
several hundred years. There is no evidence of significant dividing line between Moxeke Phase B construction and
later Haldas Phase construction. A fourth date of 1010 ± Early Horizon people razed the taller Initial Period
90 cal B.C.from a post in the south entrance of the stone walls, including most of the large summit room walls, and
surroundingwall probably dates a late Haldas Phase mod- used the resultant loose materialto create a level surfaceas
ification of the north wing. An additional date of 1 120 ± well as to build their own structures.Much of this materi-
110 cal B.C. (table i) is from an intrusive hearth well al was moved to the east, initially via the main entrance of
above the floor level of the courtyard/room complex and the summit room. Then, as material accumulated, addi-
probably represents a squatter presence on the mound tional construction debris was pushed into the inset central
shortly after the Haldas Phase ended. staircaseand over associated landings (fig. 13). In the up-
per atrium, the inset staircasesystem was sealed by a 1.5 m
Early Horizon high stone wall of reused stones. Rubble behind this wall
The Early Horizon occupation is ubiquitous at Sechin leveled the upper atrium by filling the staircase.One date
Alto site and largely obscured the primary construction of 290 ± 70 cal B.C. (table i) came from an Early Hori-
phases on the main mound. This led one investigator to zon hearth in the midden overlying this fill and a date of
misdate the main construction and occupation of the site 160 ± 75 cal B.C. (table i) came from a second intrusive
(e.g. Wilson 1995: 193, 1999: 369) and impeded reliable hearth overlying Haldas Phase architecture in the north
architecturalmapping of the site's Initial Period compo- wing of the upper atrium.
nents. The map by Donald Collier (Tello 1956: fig. 41),
Middle Horizon and "Transitional"Period
showing a regularlylaid out, symmetricalmound, is based
more on knowledge of typical "early55 mound layout than Post-EarlyHorizon occupations of the main mound are
actual evidence. Our first map (Pozorski and Pozorski ephemeral. Surprisingly,almost no Casma Incised ceram-
1987: fig. 46) takes into account the magnitude of Initial ics have been found at the site. The Middle Horizon (a.d.
Period construction evident in major level changes and 600-1000) and the "Transitional"Period, the poorly-
general site configuration but does not clearlydifferentiate known transition between the Middle Horizon and the
between Initial Period construction and later Early Hori- Late Intermediate Period (ca. a.d. 1000-1470) on the
zon alterations. north coast of Peru, are representedprimarilyby intrusive
About 500 cal B.C., people moved back onto the burials.
mound, leveled much of the Initial Period architecture, All but one of the Middle Horizon burials were found
knocked down substantial walls, and converted much of within depressions in the adobe core from which con-
the summit into a level surfaceupon which to build mod- struction materialhad been mined (fig. 8a). These burials
est rooms, plazas, courtyards, and small, free-standing had all been disturbed, some more heavily than others. The
mounds. These new settlers collected stones from wall de- burialswere in extended positions, either on their sides or
bris or stripped Initial Period wall faces to obtain con- face down, and some disturbed llama burials were also
struction material for use in their buildings. Several deep found alongside the human remains.A few vessels painted
Initial Period rooms on the mound summit were initially red-white-blackprovide the primaryevidence for assigning
mistaken for looters' holes (Pozorski and Pozorski 1987: these burials to the Middle Horizon. One other intrusive
73) because facing stones for the walls and fill had been re- burial, found within wall debris at the bottom of the deep
moved. Early Horizon mining of the adobe core for silty depression just east of the adobe core, contained a seated
clay to recycle into mortar and plaster has left a large cen- individual associated with a corroded copper trumpet.
tral depression. As part of this process, most of the poly- Three intrusive bundle burialswere cut into the landing
chrome friezes of the adobe core and the main summit of the Moxeke Phase A staircaseon the north wing. Two
room were destroyed. A cap of 30 to 100 cm of brownish contained seated, flexed infants and one was of an adult fe-
midden that covers much of the mound summit is also at- male. These burials are tentatively dated to the Transition-
tributableto this Early Horizon occupation. al Period based on associated blackware vessels (Carol
The following examples revealthe extent of Early Hori- Mackey, personal communication 2002). Other flexed and
zon alterations of the surface of the main mound. In one extended burials without associated diagnostic grave
35 x 20 m areaon the summit (figs. 8k, 12) there are two goods encountered on the north wing and in the summit
courtyards,three raisedplatforms (1 to 2 m high), and sev- room are also tentatively assigned to the TransitionalPeri-
eralrooms, wall segments, and benches. Near the main en- od based on their stratigraphicpositions, body treatment,
tranceto the summit room (fig. 8h), remains of cane and and proximity (in the case of the north wing examples) to
mud houses as well as stone houses were found within the the bundle burials.A few isolated dog burialsencountered
upper strataof the Early Horizon midden. in the north wing area may also date to this time period.
some 500 years, managing the irrigation system and its highly-visible friezes and large, publicly-accessible plazas,
products, constructing monumental architecture of un- whereas the second functioned as a large warehouse for
precedented magnitude, and functionally integrating a storing comestibles and other valuables (Pozorski and Po-
suite of distinct communities. This unity is reflected in ar- zorski 1991, 1994, 1998). The largest mound at Taukachi-
tifact types, architecturalelements, architecturaltenets, and Konkan is interpreted as a palace because it contains both
foodstuffs that are shared among the communities. At a public space for receiving elite guests and private residen-
very basic level is the square-room unit, a modular archi- tial quarters.We propose that the immense Sechin Alto site
tectural element that signals administrativepresence. First main mound was the overarching administrative center
recognized at Pampa de las Llamas-Moxeke, the square- bringing "churchand state"together by combining the re-
room unit is characterizedby rounded exterior corners, ligious precinct of the adobe core with adjacent square-
8-12 wall niches well above the floor, a carpetof reed mat- room unit administrativeelements. The net result is a col-
ting, raised thresholds, and bar closure mechanisms to re- lection of major inland sites with complementary roles,
strict access. It is used repeatedly at Pampa de las Llamas- that, taken together, comprise a cohesive whole.
Moxeke in modules. Thirty-eight square-room units form Contemporary coastal sites were integrated into the
the storage modules that create a huge warehouse on Hua- Sechin Alto polity through a subsistence exchange system.
ca A, one of the two main mounds at the site. A cluster of Bahia Seca, Tortugas, and Huaynuna were semi-au-
compounds, formed of single square-room unit modules, tonomous fishing villages during the Late PreceramicPeri-
lie on either side of Huaca A where they probably served od (Pozorski and Pozorski 1987: 12-16, 1992: 848-850).
to monitor or regulate access to and from the mound. With the introduction of irrigation and ceramics and the
Square-roomunit modules form the centralrooms of over rise of large inland centers during the subsequent Initial
100 structures,aligned with the site axis and facing the site Period, these sites became satellites providing much-need-
center, that comprise the east and west sides of the site. ed animal protein in the form of fish and shellfish. These
These intermediate-sized mounds are interpreted as ad- were exchanged for cultigens such as beans, lima beans,
ministrative structures because their architecture is well- potatoes, sweet potatoes, peanuts, cotton, gourds, and
built, clean, and similar in layout to Huaca A and because squash grown near the inland centers. Evidence of control
of the presence of occasional stamp and cylinder seals of of these sites by the inland centers is seen at Bahia Seca
fired clay.The few intermediate-sizedmounds that deviate which has an intermediate-sized mound with a square-
from the prevailing alignment face roads that enter the site room unit module at its center, just like the mid-level ad-
from the east and west and probably monitored intersite ministrative mounds at Pampa de las Llamas-Moxeke,
movement of goods and people. In this context the inter- Taukachi-Konkan,and Sechin Alto (Pozorski and Pozorski
mediate-sized mounds may have provided a third adminis- 1992: fig. 2).
trative tier to regulate the movement of products among We estimate that approximately 18,000 people inhabit-
Huaca A, the support communities, and other contempo- ed the Sechin Alto Complex at its apogee and that outly-
rary Casma Valley centers (Pozorski and Pozorki 1991, ing sites integrated into the polity had an additional popu-
1994). At Pampa de las Llamas-Moxeke, small square- lation of almost 5000 people. These estimates are based on
room unit modules were also discovered within areas of data from Pampade las Llamas-Moxeke,the best-preserved
residentialarchitecturethat yielded ample evidence of cot- Casma Valley Initial Period site. Two-thirds of the site lies
tage industries such as textile production. In this context, outside areasof modern cultivation, and it is relativelyun-
these administrative modules may have controlled this affected by later occupation and natural forces. We deter-
facet of the local economic system. At Taukachi-Konkan mined that approximately7% of the total site areawas oc-
and Sechin Alto, square-room units were also found in cupied by residential architecture comprising some 500
probable administrative contexts within monumental ar- structures. Using a conservative estimate of five persons
chitecture, and Cerro Sechin consists of at least three con- per structure ( = 2500 persons) and correctingfor the one-
centric modules. third of the site under modern cultivation (= 1250 per-
The pervasivenessof the square-room unit at Sechin Al- sons), we calculated the residential population of Pampa
to polity sites reveals that the component communities del las Llamas-Moxeke to be approximately 3750 people.
shared a common administrative"language,"expressed ar- Extrapolating to the Sechin Alto Complex, which is 4.77
chitecturally.Varied functions for major mounds at other times larger than Pampa de las Llamas-Moxeke,we get a
sites provide additional evidence of interdependence figure approaching 18,000. The remaining estimated pop-
among the sites. At Pampa de las Llamas-Moxeke, one of ulation of 1200 persons, is based on the combined areasof
the two main mounds functioned as a temple, based on its the coastal sites which were almost totally residential.
The Sechin Alto political and economic system, with ar- mounds face inwards, was seen at spatially-distinct loci and
chitectural accomplishments that make the Casma Valley maintained through severalcenturies.Variedmounds with
unrivaledin the entire New World for its time period, be- distinct functions at distinct sites document interdepen-
gan to weaken near the end of the Moxeke Phase. Two sig- dence among the major sites. Within these sites, the
nificant events evident in the archaeological record may square-room unit architecturalform persists as an emblem
help explainthis decline. One is the great battle or massacre of political power and bureaucraticcontrol. Such demon-
depicted in the Cerro Sechin carved-stone facade, where stratedlongevity of a preconceived plan arguesfor a lineage
the warriorsand the victims can be clearlydistinguished by of rulerswho maintainedtheir hegemony acrossmany gen-
their dress. Significantly, the victims in the Cerro Sechin erations. The "masterplan" conceived by these rulers was
carvingswear the same pleated skirts and scalloped tunics executed with precision. At Pampa de las Llamas-Moxeke
as the immense god-like or priest-like figures that adorn where preservation is optimal, for example, the mound
the facade of the temple mound at Pampa de las Llamas- fronts that define the longest alignment are still within 15
Moxeke (Kauffmann 1983: 179; Pozorski 1987: 27; Tello cm of a perfectlystraightline across a distance of more than
1956: 154, 159). In the absence of evidence of attack by 700 m, even after over 3500 years of exposure to the ele-
external enemies at this time, the Cerro Sechin iconogra- ments (Pozorski and Pozorski 1989, 1994). A clear five-
phy suggests internal strife. More specifically,it shows that level hierarchyin the settlement pattern is reflected in de-
insurgence by a faction from the southern Pampa de las creasing site magnitude, ranging from the largest capital
Llamas-Moxekebranchhad been quelled and its leadersex- site of Sechin Alto, through the secondary centers of Pam-
ecuted. Second, about 1400 cal B.C., a large-scaleEl Nino pa de las Llamas-Moxeke and Taukachi-Konkan,tertiary
event struck the Casma Valley area. This is clearly docu- centers of Sechin Bajo and Cerro Sechin, to the lesser
mented at Cerro Sechin and was observed on Huaca A at coastal sites- including the administrativecenter of Bahia
Pampade las Llamas-Moxeke.El Nino damage on one ma- Seca with its subsidiaries at Tortugas and Huaynuna. No-
jor mound was partially repaired, but the site was aban- ticeably lacking, however, are burials rich in grave goods
doned soon after this event. indicative of ascribed status. It seems, rather,that conspic-
The Sechin Alto polity responded during Moxeke Phase uous consumption, ideology, and artistic achievement
B by consolidating its leadership at Sechin Alto site and by were manifested through monumental architectureand its
altering the main mound to tangibly represent this politi- accouterments because metallurgy was not yet known and
cal restructuring.The main mound was expanded by rais- ceramic, stone, and textile technologies were still develop-
ing its summit elevation some 9 m to the height of the sur- ing.
face of the adobe core. A huge square-room unit was built This detailed reconstruction of early political and eco-
east of the core, and the four rectangularplazas and two cir- nomic development within the Casma Valley is based on
cular plazas and their bordering mounds were also con- our excavationsat nine Preceramic,Initial Period, and Ear-
structedat this time. The east face of the summit room was ly Horizon sites plus data compiled by other investigators.
covered with an immense polychrome frieze readilyvisible The result provides an in-depth look at the Sechin Alto
from the plazas below. The two circular sunken plazas polity, the grandest development in the New World from
within the row of plazas, like amphitheaters,may have held ca. 2150 to 1400 cal B.C.We have documented featuresof
localized rituals or spectacles as a means of bringing cere- one of the earliest Andean states, including the essential
monies closer to the people. This would seem to give the characteristicsof its sites, the economic and population
main mound a more public orientation, perhaps to reas- base, and the internaland externalvariablescontributing to
sure uneasy citizens. its demise. These data allow comparison between the Cas-
A number of characteristicsof Sechin Alto site and the ma Valley and other world areas where complex society
polity it headed argue convincingly for the emergence of evolved.
state-level political organization in the Casma Valley by
Initial Period times. The sheer magnitude of construction Acknowledgments
involving building phases represented by hundreds of Funding for the investigations of the Sechin Alto site
thousands of cubic meters of stones and earth documents was provided by National Science Foundation grant SBR-
the leaders'capacityto mobilize and support a huge labor 9806833, the H. John Heinz III Fund of the Heinz Fam-
force drawn from a sizable population. Careful planning, ily Foundation, the Curtiss T. Brennan and Mary G. Bren-
typified by Pampa de las Llamas-Moxeke, Taukachi- nan Foundation, the American Philosophical Society, and
Konkan, and Sechin Alto where huge central mounds es- the Faculty Research Council of the University of Texas-
tablish the sites' axes and rows of aligned, subsidiary Pan American. Permission to excavatewas granted by the