Content uploaded by Anabel Ford
Author content
All content in this area was uploaded by Anabel Ford on May 01, 2017
Content may be subject to copyright.
Stone tools and their waste products form the
most plentiful and durable evidence of craft
production in many prehistoric cultures.
Flintknapping is often seen as a likely area of craft
specialization, partly because stone tools include
both essential utilitarian items and elaborate pres-
tige goods but also because, to modern archaeolo-
gists, knapping seems an exotic skill, and high
levels of proficiency are not easy to acquire. In
Maya archaeology, the immense deposits of deb-
itage at Colha (Shafer and Hester 1983) have
become the best- known example of the industrial-
scale production of common utilitarian tools. Lithic
industries at other sites are often expected to be
equally the realm of specialists, and stone tools
may be seen as commodities controlled by the state
or elite members of society. Stone tools have thus
been the focal point of a good deal of argument
LITHIC INDUSTRY IN A MAYA CENTER:
AN AXE WORKSHOP AT EL PILAR, BELIZE
John C. Whittaker, Kathryn A. Kamp, Anabel Ford, Rafael Guerra, Peter Brands, Jose Guerra,
Kim McLean, Alex Woods, Melissa Badillo, Jennifer Thornton, and Zerifeh Eiley
Cahal Tok (Place of Flint) is a limestone rise with some structural evidence, associated with the previously designated LDF
Chert Site, close to the ceremonial center of El Pilar. Excavations uncovered evidence that during the Late Classic period,
specialized flintknappers produced bifaces, primarily chert axes, at the Cahol Tok locus, first on a cleared limestone shelf,
then on a prepared cobble platform. Small flakes remained in situ whereas much of the larger debris was deposited to the
east off the edge of the platform and into the LDF debitage dump. The identification of a specialized manufacturing locale
near the ceremonial precinct of a major center is unusual in Maya archaeology. Central control of an important industry
may be implied, although knapping could equally well be organized more independently. We expect that small industrial
areas are actually present at most large sites, but may often be difficult to recognize.
Cahal Tok (Lugar de Pedernal) es una pequeña elevación de piedra caliza con alguna evidencia estructural asociada al pre-
viamente designado “LDF lugar de pedernal” cercano al centro ceremonial de El Pilar. Las excavaciones han encontrado
evidencia de un taller de especialistas en trabajo de pedernal durante el Clásico Tardío. En este taller se producían hachas
de pedernal, inicialmente sobre una losa de piedra caliza, y posteriormente sobre una plataforma adoquinada. Se encontraron
lascas pequeñas in situ, pero los pedazos grandes habían sido removidos al basurero fuera del límite oriente de la plataforma.
La identificación de un taller de manufactura especializada cercano a un recinto ceremonial prominente es un acontecimiento
raro en la arqueología Maya. Esto puede implicar el control central de una industria importante, pero también puede ser posi-
ble que el taller estuviera organizado de manera independiente. Anticipamos la presencia de pequeñas áreas industriales en
la mayoría de los centros prominentes, pero estas áreas pueden ser difíciles de reconocer.
John C. Whittaker ■Department of Anthropology, Grinnell College, Grinnell, IA 50112 (whittake@grinnell.edu)
Kathryn A. Kamp ■Department of Anthropology, Grinnell College, Grinnell, IA 50112 (kamp@grinnell.edu)
Anabel Ford ■ISBER/MesoAmerican Research Center, University of California Santa Barbara, CA 93106
(ford@marc.ucsb.edu)
Rafael Guerra ■Institute of Archaeology, Belmopan, Belize (guerraal@hotmail.com)
Peter Brands ■ 1050 190th St Ionia IA 50645 (pbrands@gmail.com)
José V. Guerra ■Awe #7 Hospital Street, San Ignacio, Cayo, Belize (joseguerraawe@gmail.com)
Kim McLean ■PO Box 2306 Stateline NV 89449 (akara_2001@hotmail.com)
Alex Woods ■Department of Anthropology, University of Iowa, Iowa City, IA (alexander- d- woods@uiowa.edu)
Melissa Badillo ■ Institute of Archaeology, Belmopan, Belize (mmbadillo@hotmail.com)
Jennifer Thornton ■1140 El Capitan Loop, Grants, NM(thornjenton@yahoo.com)
Zerifeh Eiley ■31 Joseph Andrews Dr., San Ignacio, Cayo, Belize (zerifeh@yahoo.com)
Latin American Antiquity 20(1), 2009, pp. 134–156
Copyright ©2009 by the Society for American Archaeology
134
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 134
about the nature of industry and political organi-
zation in Maya society.
Evidence of areas devoted to industrial activity
within major Maya centers is not commonly
reported, and the situation at the Cahal Tok/LDF
locality at the site of El Pilar in Belize is thus
unusual. At Cahal Tok a small stone platform on a
cleared limestone surface was the actual knapping
area, with debitage deposited nearby to form a
dump, designated the LDF Chert Site during ear-
lier work by Ford and her teams. The debitage from
deposits at Cahal Tok and LDF shows that the knap-
pers made flaked stone celts in great numbers. The
density of deposits and the scale of specialized pro-
duction were comparable to some loci at Colha.
There was some separation of work and discard
over a substantial period of time in the Late Clas-
sic period, using material that was acquired and
partly prepared elsewhere. All of this is evidence
of a considerable amount of organization. The
immediate area around Cahal Tok and LDF seems
to have been used for limestone quarrying, and
shows only sparse signs of habitation. An adjacent
concentration of about 200 intentionally broken
limestone spindle whorls is probably a ritual deposit
that may indicate fabric production in the area as
well (Kamp et al. 2007). This apparent industrial
zone is very close to the ceremonial and civic cen-
ter of the site, and could have been controlled from
it, although a less centralized organization of pro-
duction is equally plausible.
Excavations
The El Pilar Archeological Reserve for Flora and
Fauna is located on the international boundary
between Belize to the east and Guatemala to the
west, approximately 24 km north of San Ignacio
town and 11 km from Bullet Tree Falls Village in
Belize (Figure 1). The site is in the uplands drained
by tributaries of the Belize River. El Pilar is the
largest center of the Belize River area and consists
of 25 mapped plaza complexes, with most con-
struction occurring in the Middle Preclassic period
(950 B.C.–650 B.C.) and Late to Terminal Classic
periods (A.D. 650–A.D. 1000).1
The LDF Chert locus (Ford and Olson 1989:189;
Michaels 1993:227) is a dense deposit of lithic
debris, roughly 150 m northwest of Plaza Copal near
the middle of the El Pilar monumental center (Fig-
ure 2). The densest debris covers an area roughly
15–20 m in diameter in a large depression that is
probably the result of quarrying for limestone,
although Ford estimates the total deposit may be as
much as 50 50 m in extent (Ford 1991). In 1990,
the BRASS/El Pilar project excavated two 1 1 m
test pits, recovering over 1,000 kg of debitage from
a single test pit, according to field notes. Adjacent
to the LDF locus are two low mounds and a plat-
form with lithic debris, referred to as Cahal Tok
(Place of Flint). In the summer of 2004, a team from
Grinnell College, directed by Kamp and Whittaker,
tested this area to understand the context of lithic
production at the previously discovered and tested
LDF Chert Site. Our analysis included the lithic
debitage recovered from Cahal Tok, and a sample
from LDF consisting of material from a 1010 cm
column collected in 15 cm levels from the south part
of the west wall of one of the two test pits excavated
in 1990 (TP2), and bifaces from both LDF test pits.
The entire LDF column had been saved including
dirt and nonartifactual inclusions, allowing us to
calculate the density of each of the elements of the
deposit and permitting identical analytic procedures
to be used on both the LDF and Cahal Tok materi-
als. Debitage resulting from two experimental axes
Whittaker et al.] LITHIC INDUSTRY IN A MAYA CENTER 135
Figure 1. El Pilar and other important Maya sites in Belize.
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 135
136 LATI N AMERICAN ANTIqUITY [Vol. 20, No. 1, 2009
Figure 2. Location of Cahal Tok and LDF in relation to the monumental center at El Pilar.
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 136
knapped by Whittaker provided an experimental
comparison.
The LDF and Cahal Tok deposits do not appear
to be directly associated with house mounds in a
habitation area. The immediate vicinity is an area
of sparse structural evidence with many pits and
edges of exposed bedrock, apparently from lime-
stone quarry activity. The LDF debitage deposit is
adjacent to the southeast edge of a small rise where
a low ledge of exposed bedrock begins a gentle
slope up to the north and west, with apparent struc-
tural remnants in the form of limestone and coarse
chert nodules and slight mounding at the highest
point to the west. We excavated a series of test
trenches to form an almost continuous profile
across the rise, from the limestone edge adjacent
to the LDF deposits on the east, to the edge of the
mound on the top of the rise to the west (Figure 3).
Stratigraphy and Structures at
Cahal Tok and LDF
The stratigraphy at Cahal Tok was not complicated.
All of the trenches were carried to limestone
bedrock, which was encountered at 50–70 cm below
modern surface. Cultural deposits lay directly on the
limestone in all the trenches, and it appeared that
the limestone was exposed and clean at the begin-
ning of occupation. We were unable to tell whether
the limestone surface resulted from intentional
clearing, a common precursor to construction, or
had been bared by erosion following farming or
quarry activities. In Trenches 3, 4, and 5, and in the
western part of Trench 1, the limestone surface was
hard and smooth. We consider this evidence that it
was protected under the platform and the western
mound respectively. Elsewhere the upper few cen-
timeters of the limestone were friable and scaly,
weathering into the bottom of the soil profile devel-
oping on the site. At the eastern edge of the site,
knapping debris had accumulated directly on
bedrock in a layer 50–60 cm thick at the east end
of T3, and progressively thinning westward across
T3 and T4 until it ended as small pockets in bedrock
depressions in T5 and T9. This debitage deposit
was so dense, effectively nothing but flakes with
limey clay filling the voids, that in T3 and 4 we were
Whittaker et al.] LITHIC INDUSTRY IN A MAYA CENTER 137
Figure 3. Cahal Tok site plan.
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 137
forced to collect a sample only. This was usually in
the form of a 20 20 cm column collected from
the southeast corner of each 1 1 m subdivision
of the trenches, for each 10 cm arbitrary excavation
level. All other fill from the debitage layers was
screened and sorted for pottery, obsidian, retouched
tools, and other artifacts. To keep our samples com-
parable, we maintained the column sampling in T3
and T4 for all levels, including the upper ones that
were not dense debitage.
West of the debitage deposit of T3 and T4, the
material on bedrock was a dark clayey loam that
filled the cracks, root holes, and animal burrows in
the top few centimeters of the limestone, and con-
tained a moderate number of artifacts. Approach-
ing the surface, in profiles the fill appeared darker
and more humic, with stronger development of soil
peds, and much recent biological activity in the
form of roots and leaf litter in the upper few cen-
timeters. There was little to distinguish fill in one
part of the site from another, and there was no
ceramic evidence of chronological distinctions.
Generally, the limestone bedrock was covered with
30–60 cm of cultural fill weathering to a shallow
soil formation.
In T3 and T4, a dark, rocky layer about 20 cm
thick overlay the debitage. It was discontinuous
and presented an uneven surface, and it was not
until we found the wall line and a better surface in
T5 that we were convinced it was the top of a low
platform. The wall in T5 was 20–30 cm high, usu-
ally two courses of unshaped limestone and chert,
forming a rough edge for a platform built of smaller
lumps of limestone and chert. The base of the wall
appeared to rest directly on bedrock in places, how-
ever, directly adjacent to the wall was a depression
in the bedrock filled with fine debitage, and the
rocky fill of the platform overlay debitage to the
east in T4 and T3. There was no evidence that there
was ever a plastered floor.
Above the rocks of the platform, the dark fill was
similar to the rest of the site. Only in the eastern
half of T3, closest to the LDF deposit, was there a
further level of dense debitage exposed at the sur-
face. We regard this as evidence that the knapping
continued after the platform surface was built.
At LDF, the two test trenches showed 70–100
cm of fill that was almost entirely debitage, with
minor portions of soil and limestone rocks, and
only a few sherds and other artifacts (Figure 4). The
surface today is well below the adjacent surface of
our trenches at Cahal Tok, and the debitage appears
to be filling a depression.
Artifact Analyses
The artifacts of Cahal Tok are dominated by lithic
production debris, on which our analyses concen-
trated. We recovered a limited assemblage of tem-
porally diagnostic ceramics that provided the best
evidence for dating activities at the locus.
Ceramics
The Cahal Tok ceramic assemblage included 4,144
sherds, of which 229 (5.5 percent) were temporally
diagnostic, a low proportion for contexts at El Pilar,
where construction fill averages about 12 percent
and midden contexts up to 25 percent temporally
diagnostic ceramics. The ceramic assemblage over-
all was sparse and poorly preserved. As no sealed
deposits were recovered, the ceramic assemblage
represents mixed contexts; nevertheless, the tem-
poral range of the ceramics is limited, showing that
the platform on the site was in use in the late facet
of the Late Classic period, corresponding to the
Spanish Lookout Phase, from A.D. 700 to A.D.
950 (Gifford 1976), and to Tepeu 2 at Uaxactun
(Smith 1955). Large open jars and bowls of fine
paste dominated the assemblage, which lacked the
coarse paste cooking vessels common in domestic
sites in the region (Ford 2000; Lucero 2001). Our
conclusion from the limited ceramic evidence is
that Cahal Tok was not a residential area, but a spe-
cial activity site dating to the Late Classic period.
Lithic Tools
Table 1 shows the distribution of major tool types
at Cahal Tok compared to the assemblage from
Chiik Nah, a small habitation mound somewhat fur-
ther south from the central plazas of El Pilar (Whit-
taker et al. 2002). The assemblages are broadly
similar, and comparable to those from other habi-
tation sites in the area. Chiik Nah, ranked at the very
bottom of the residential scale at El Pilar, has rel-
atively few obsidian prismatic blades, and rela-
tively many gravers. Another site in the El Pilar
sphere, 272–136, a group of larger house mounds
around a plaza 4.5 km south of El Pilar, had an
unusual assemblage dominated by obsidian blades
138 LATI N AMERICAN ANTIqUITY [Vol. 20, No. 1, 2009
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 138
and debris from their manufacture, plus large num-
bers of chert gravers or drills and chert flakes with
battered edges (Hintzman 2000; Michaels 1993).
Such specialized tools are rare at Cahal Tok.
Tools of any kind were especially scarce in the
debitage- dominated deposits in the eastern trenches
of our excavation (Table 2). Throughout the assem-
blage, most tools are casually retouched flakes.
About 37 percent of the tools were made on biface
thinning flakes, 25 percent on relatively thick flakes
struck from cores, and the rest on other pieces,
including flakes of indeterminate origin. The tool
Whittaker et al.] LITHIC INDUSTRY IN A MAYA CENTER 139
Figure 4. Profile face in LDF excavations, 1990, to show density of debitage.
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 139
assemblage suggests that besides biface manufac-
ture, there was a low level of the normal craft and
maintenance activities found at nearby residential
sites. The LDF tool assemblage, in contrast, is
almost entirely biface manufacturing failures and
broken biface tools.
Bifaces
Biface fragments, failures, and discards would be
expected in the manufacturing contexts suggested
by the debitage at Cahal Tok and LDF. There were
no bifaces in our Cahal Tok column samples,
although we recovered one from the debitage in T3
and several others from the trenches to the west.
The two test pits in the LDF deposits produced a
number of biface failures among the debitage. The
sample is not very large, and there are no apparent
patterns by depth. The following analysis consid-
ers only the LDF specimens.
All told, there were 57 biface pieces from the
LDF test pits, plus a dozen or so flake tools and an
obsidian blade segment that will not be considered
here. The knappers at Cahal Tok and LDF empha-
sized axes, but there are also 12 pieces of thin
bifaces (Figure 5), sometimes referred to as laurel
leaves, that would have been knives or spear points.
In finished form, these are sometimes associated
with burials, elite residences, and other special con-
texts. Even at an early stage they can be recognized
as different from the axes by their thinness and
flake scars that pass the centerline of each face. Two
of the LDF specimens are very finely flaked and
finished by pressure retouch; a third, less well fin-
ished, shows signs of heavy use. The other nine
pieces are from middle or late stages in manufac-
ture, and were still being thinned when they were
broken.
The chipped stone axes made at Cahal Tok were
ovoid bifaces, usually with a thick biconvex cross
section. At El Pilar, the bit may be either flaked or
flaked and ground. Common labels for these and
related tools include “general utility bifaces,” “oval
bifaces,” and “flaked celts.”2Such tools were all-
purpose necessities of Maya life, as ubiquitous as
the machete today. Although we gloss them as
“axes,” experiments and use- wear studies show that
such bifaces are effective for a wide range of chores
including forest clearance, woodworking, agricul-
tural tillage, and quarrying and shaping limestone
(Aldenderfer et al. 1989; Aoyama 2007; Lewen-
stein 1987; Shafer and Hester 1986a; Stemp 2004;
Woods and Titmus 1996). Examination of speci-
mens from other areas at El Pilar shows a wide
range of macroscopic use- wear, including small
stepped flake damage, heavy battering and dulling,
and glossy polishes. Many El Pilar specimens have
been reflaked, often crudely, suggesting that here
as elsewhere owners maintained tools made by spe-
cialists and recycled chert (Moholy- Nagy 2003;
Shafer 2000).
The importance of general utility bifaces and
related forms as a basic Maya tool is underlined by
140 LATI N AMERICAN ANTIqUITY [Vol. 20, No. 1, 2009
Table 1. Comparison of Tool Types
Tool Type Frequency Percent
Cahal Tok
Small Tools Scraper 108 38.0
Notch 11 3.9
Tool Edge 47 16.5
Multi-EdgedTool 4 1.4
Gravers 21 7.4
Obsidian Blade 55 19.4
Large Tools Axe 17 6.0
Chopper 1 0.4
Core 9 3.2
Hammer 1 0.4
Thin Biface 2 0.7
Rough Biface 8 2.8
Total 284 100.0
Chiik Nah
Small Tools Scraper 71 29.6
Notch 26 10.8
Tool Edge 11 4.6
Multi-Edged Tool 9 3.8
Gravers 55 22.9
Obsidian Blade 14 5.8
Large Tools Axe 13 5.4
Chopper 14 5.8
Core 20 8.3
Hammer 0 0.0
Thin Biface 0 0.0
Rough Biface 7 2.9
Total 240 100.0
Table 2. Comparison of Tool Concentrations in Debitage
Area and Rest of Cahal Tok
Debitage Area Rest of Whole
(T3, T4) Site Site
Number of Small Tools 29 217 246
Number of Large Tools 10 27 37
Total Tools 39 244 283
Percent Tools 13.8 86.2 100
Percent Excavation Area 29 71 100
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 140
their wide distribution in space and time. They
occur in contexts ranging from residential quarters
to architectural fill. Oval bifaces began to be made
at Colha in the Middle Preclassic period, with mass
production beginning by the Late Preclassic period
(Shafer and Hester 1986b). They were manufac-
tured in enormous numbers at Late Classic period
contexts at Colha (Roemer 1991; Shafer and Hes-
ter 1983), but many other sites also produced bifa-
cial axe forms (Black and Suhler 1986; Hyde 2003;
Meadows and Hartnett 2000; Potter 1993; Thomp-
son 1991), and related tools. At El Pilar, Ford esti-
mates that bifaces occur at a frequency of
approximately one per cubic meter in excavated
midden and fill.
The LDF axes include pieces at all stages of
manufacture. There are 10 early pieces on which
little has been done except edging and decortica-
tion of a chert nodule (Figure 6). The debitage
analysis suggests that this is the form in which most
of the material must have arrived. The available
cherts and chalcedonies around El Pilar are not
large enough to produce the macroflake blanks used
at Colha. Fourteen pieces from LDF show middle
stage work, thinning and shaping the preform. A
late stage, in which final shaping is done and edges
are finished is represented by 11 pieces (Figure 7).
Almost all the biface specimens show the kind of
bending fracture common to manufacturing failures
(and heavy use in the case of finished axes). In
most cases, the fracture initiated at a weak spot in
flawed material. Although as noted above, most of
the stone is perfectly adequate for axes, a few pieces
are so obviously bad that one wonders why a knap-
per bothered with them. We might speculate that
poor material could be used for training apprentices
at such a workshop site (Ferguson 2003; Grimm
2000; Milne 2005; Stout 2002). The relative
scarcity of failures compared to Colha, and the fre-
quent hard use and recycling of axes elsewhere on
the site suggest that better quality stone was some-
what valuable. Good pieces should not have bro-
Whittaker et al.] LITHIC INDUSTRY IN A MAYA CENTER 141
Figure 5. Thin bifaces from LDF. The upper row are finished pieces with carefully refined edges; the specimen to right
shows damage and reworking. The two lower specimens are typical of pieces broken at a middle stage in manufacture.
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 141
ken often in later stages of axe manufacture, and
when they did, they may have been removed from
the deposit and reworked.
Among the fragments are several from axes that
had been finished and used, including two with
ground bits, and seven proximal (butt) end pieces
with heavily dulled edges or ground surfaces (Fig-
ure 8). One other axe is complete, but worn out by
resharpening and reworking of edges and the bro-
ken butt. These pieces suggest that the knappers on
site not only made new implements but reworked
old tools. The discarded butt ends are often
expected to have been discarded as new blades
were fitted to old hafts, and McAnany (1992) argues
that these mark retooling sites, whereas bit ends are
more common in field areas.
Few knapping tools were recovered from Cahal
Tok and LDF. Antler would probably not have sur-
vived in these deposits, but the evidence of the flake
platforms suggests that work was done with stone
hammers, which would. Limestone hammers are
common at Colha, and although we might have
missed some among the other limestone lumps in
fill at Cahal Tok, we were alert to this possibility
and did look, finding only one recognizable ham-
merstone in all the Cahal Tok trenches. This was a
chert specimen, not ideal for a knapping tool. Tools
of any kind were readily noted in the homogeneous
debitage deposits of LDF, which produced one
good quartzite or sandstone knapping hammer.
Debitage
The debitage analyses allow us to further discuss
the process and location of manufacturing activi-
ties at Cahal Tok and LDF, and provide a basis for
estimating the scale of production. Debitage was
analyzed by size class (maximum dimension < 2
cm, 2–4 cm, 4–6 cm, and > 6 cm), presence of cor-
tex, coarseness of material, evidence of burning,
and type of flake (Tables 4 and 5). Thick flakes with
large bulbs of percussion and large flat platforms
were identified as the result of hard- hammer per-
cussion. Thinner flakes with diffuse bulbs of per-
cussion, lipped platforms, and multiple flake scars
142 LATI N AMERICAN ANTIqUITY [Vol. 20, No. 1, 2009
Figure 6. Early stage bifaces from LDF, discarded after bending fractures during decortication and initial thinning. They
are probably close to the form in which material was brought to Cahal Tok. The piece at bottom left is the only one vis-
ibly made on a macroflake. The other two are nodules with cortex on both faces.
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 142
on the exterior surface were classified as biface
thinning flakes. Flakes with missing platforms
and/or lacking other recognizable indicators were
classified as indeterminate. An additional category
of shatter described angular pieces that appeared
freshly broken but lacked identifiable flake fea-
tures. Chert pieces that were clearly broken by fire
and showed no evidence of knapping were not
recorded.
The Cahal Tok column samples were weighed,
screened with one- sixteenth inch (16 mm) mesh to
remove the loose fill, and the flakes washed in the
field lab. Each sample was then screened through
a quarter- inch (64 mm) mesh. Rocks, and the mate-
rial that passed through the screen, mostly
microdebitage,3was also saved and weighed but
not analyzed. To compare our excavated material
with that from the LDF deposit itself, we analyzed
five column samples collected from the LDF
deposit in 1990. The LDF samples had not been
screened and contained all the material that was
excavated, so they could be subjected to the same
treatment as the samples from Cahal Tok. The only
major difference is that initial weights for Cahal Tok
were obtained while there was still some moisture
in the sediment. The LDF column samples were a
10 10 cm area from each 15 cm level and our
Cahal Tok column samples varied from 20 20
cm to 10 10 cm of each 10 cm excavation level.
Our field experience demonstrated that rocks and
other inclusions, as well as the inherent difficulty
of taking a consistent size sample even within the
nominal boundaries, created variability in the accu-
racy of the samples that was revealed in some dif-
ferences in weight and volume for theoretically
identical samples. Because of this variability, the
comparisons presented in Table 3 are by propor-
tion, rather than absolute quantities.
Whittaker et al.] LITHIC INDUSTRY IN A MAYA CENTER 143
Figure 7. Late stage bifaces from LDF. The bottom biface suffered a bending fracture through a flaw in the otherwise
good chert while the final edge was being trimmed and gives a good idea of the desired final form. The upper left speci-
men was discarded because of flawed material at the bit, but is otherwise close to completion. Upper right is a fragment
earlier in the process of thinning and shaping.
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 143
With the exception of a few pieces of obsidian,
mostly small fragments of prismatic blades, all the
lithic material from the site can be glossed as chert.
The chert varies from opaque brown or grey stone
to translucent browns, greys, purples, and colorless
material. Some other projects in Belize would prob-
ably call the more translucent material chalcedony
(Cackler et al. 1999; Oland 1999; Shafer 2000;
Shafer and Hester 1983). Most of the debitage is
good- quality stone, although not up to the stan-
dards of Colha. Quality varies from coarse and
gritty to very fine and glassy, although material of
all qualities may have voids or other flaws. There
are some pieces so grainy that quartzite is proba-
bly a better term, and a few flakes are of hard lime-
stone with a silicious component. We do not make
a distinction between chert and chalcedony because
in the El Pilar material, a high proportion is deeply
patinated to a dead white, only revealing original
color when broken. Moreover, in local nodules, the
translucent and opaque often intergrade in the same
nodule. There may be some distinction in source
locations, which we cannot currently specify. The
cortex is usually a rough white hard limey surface
over a grey or brown cherty rind. There is little evi-
dence of river rolling, but most of the stone does
not appear to come from the limestone at El Pilar
itself, although without further survey for chert
sources we cannot be too sure.
Production Process. Although LDF has been
described as a “quarry and reduction site” (Ford and
Olson 1989), and “adjacent to chert quarry holes”
(Michaels 1993:227), we do not consider these ini-
tial assessments to be correct. First, although there
are many flakes with minor amounts of cortex, the
masses of decortication flakes produced in the early
stages of knapping and expected at quarry sites are
lacking at LDF and Cahal Tok. Table 5 presents per-
144 LATI N AMERICAN ANTIqUITY [Vol. 20, No. 1, 2009
Figure 8. Finished and reworked biface axes from LDF. The lower left specimen has a ground and polished bit with sub-
sequent damage, and a broken and slightly reworked butt. Upper left and lower right are snapped butt ends with edges
finished by dulling. The specimen upper right shows damage and reworking on bit and butt.
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 144
centages of “cortical” flakes (defined as those with
any cortex on them), and percentages of “> 25%
cortical” (flakes with more than 25 percent of their
exterior surface showing cortex). It appears that
knappers brought partially decorticated nodules to
Cahal Tok for further reduction. Second, although
some chert nodules are found on the site, and appar-
ently derived from the quarry holes, we made a
point of testing a number of these. The chert nod-
ules on the surface of El Pilar today are flawed,
tough, and grainy, suitable for building stone but
not for systematic tool making. Most of the deb-
itage is much better, although a minor part of the
LDF material is very poor and could be from lime-
stone quarries at El Pilar.
For comparative material, Whittaker, an expe-
Whittaker et al.] LITHIC INDUSTRY IN A MAYA CENTER 145
Table 3. Microdebitage and Other Fractions
Total % Lithics Microdebitage
weight based on as % all Number Number
UNIT Level (grams) weight debitage of tools of ceramics
LDF
TP2 2 2505 86 700
TP2 4 3295 76 221
TP2 6 2899 74 500
TP2 7 3288 76 420
TP2 8 840 26 24 00
All Analysed LDF 12827 74 541
Cahal Tok
T3E 1.0 4228 23 27 05
T3E 1.1 5570 5 43 04
T3E 3.0 6448 8 48 00
T3E 3.2 2095 26 38 00
T3E 3.3 3165 38 41 01
All T3E 21506 16 38 0 10
T3C 1.0 3630 7 41 12
T3C 1.1 8871 6 46 34
T3C 3.0 7313 34 42 0 11
T3C 3.2 1586 63 40 04
T3C 3.3 1741 25 83 00
All T3C 23141 21 46 4 21
T3W 1.0 5046 8 36 00
T3W 1.1 7800 1 80 00
T3W 3.0 2061 25 41 05
T3W 3.1 1143 64 27 01
T3W 3.2 1277 58 28 04
T3W 3.3 1981 33 58 11
All T3W 19308 16 39 1 11
T4E 1.0 17869 1 90 00
T4E 1.1 1838 3 74 03
T4E 2.0 2230 3 31 04
T4E 3.0 1881 6 47 04
T4E 3.1 1216 28 47 0 22
All T4E 25034 3 56 0 33
T4C 1.0 7590 5 62 0 11
T4C 1.1 2564 4 76 0 12
T4C 2.0 1321 3 97 02
T4C 3.0 1293 6 46 07
T4C 3.1 956 26 58 0 13
All T4C 13724 6 63 0 45
T4W 1.0 5410 4 58 09
T4W 1.1 1210 4 98 00
T4W 2.0 1158 3 88 01
T4W 3.0 1600 17 65 0 14
All T4W 9378 6 66 0 24
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 145
146 LATI N AMERICAN ANTIqUITY [Vol. 20, No. 1, 2009
Table 4. Debitage Comparisons: Technological Categories
% Hard % Biface
Type N hammer thinning % Indeterminate % Shatter
Experimental Axes
All Decortication Debitage 430 27 7 46 20
All Non-Decortication Debitage 1490 14 30 49 7
All Debitage 1920 18 24 48 10
Chiik Nah (272-32)
All Debitage 1471 29 9 25 37
LDF Columns
All Debitage 4980 4 45 46 5
Cahal Tok
All Non-Column Lithics (T1,2,5-9) 3677 18 28 42 12
Above Platform Columns (T3,4) 2995 5 54 36 5
Platform Level Columns (T3,4) 11 3 50 43 3
Below Platform Columns (T3,4) 11208 4 54 39 3
Comparisons East to West of LDF and
Cahal Tok Debitage
LDF 4980 4 45 46 5
3E 3846 4 56 36 4
3C 5446 3 53 40 3
3W 3935 4 55 38 2
4E 470 13 45 40 2
4C 321 9 41 43 6
4W 196 8 41 48 3
Table 5. Debitage Comparisons: Size and Cortex
%%% % % % > 25%
Context N <2 cm 2-4 cm 4-6 cm > 6 cm Cortical Cortical
Experimental Axes
Decortication Debitage 430 61 21 10 8 49 50
Non-Decortication Debitage 1490 73 19 52 6 5
All Debitage 1920 70 20 7318 13
Chiik Nah (272-32)
All Debitage 1471 21 60 17 2 54 NA
LDF Columns
All Debitage 4980 63 33 5024 10
Cahal Tok
All Non-Column Lithics (T1,2,5-9) 3677 34 53 12 1 29 NA
Above Platform Columns (T3,4) 2995 90 10 1010 6
Platform Level Columns (T3,4) 11 94 60015 9
Below Platform Columns (T3,4) 11208 94 60022 5
Comparisons East to West of LDF
and Cahal Tok Debitage
LDF 4980 63 33 5024 10
3E 3846 92 70012 6
3C 5446 94 60030 5
3W 3935 93 60013 4
4E 470 90 10 1010 5
4C 321 88 11 2015 5
4W 196 85 13 2018 9
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 146
rienced knapper, produced two axes and collected
and analyzed the debitage (Figure 9). One axe was
made of fine opaque “Chert Bearing Zone” chert
collected from riprap on the waterfront in Belize
City, derived from modern quarries north of the city
between Sand Hill and the Belize River (T. Hester,
personal communication 2007). The second axe
was made from a nodule from road gravel coming
from the gravel quarries around Spanish Lookout.
This last was very comparable to site material, hav-
ing a thin layer of translucent purplish material
under the cortex, followed by opaque banded grey
and purple material of variable quality toward the
center. The debitage from these axes was treated
as much as possible like that from the archaeolog-
ical contexts: passed through a quarter- inch (64
mm) screen to remove microdebitage and dust, and
then categorized according to the same criteria.
Different stages of the knapping process were col-
lected separately. The decortication debitage was
all produced during initial preparation of the intact,
corticated nodules by percussion with hammer-
stones. The nondecortication debitage comes from
subsequent thinning using a large antler billet, and
thinning, shaping, and platform preparation using
hammerstones. Tables 4 and 5 present summary sta-
tistics based on counts for the experimental axes
and the site debitage.
The debitage comparisons support our initial
suspicions that the Cahal Tok knappers were pro-
ducing chipped stone axes from previously decor-
ticated blanks. In addition to the generally low
proportions of flakes with cortex, the large flakes
with hard- hammer traits and most or all of the exte-
rior covered with cortex that were produced by the
experimental nodule preparation are almost nonex-
istent in the Cahal Tok and LDF debitage. Knap-
ping poor material and removing cortex also tends
to produce much angular shatter, material without
recognizable flake features. There was little shat-
ter in the assemblages from Cahal Tok, and most
of it was small and thin, probably indistinguishable
bits of flakes. The lack of shatter is primarily a
reflection of the generally high quality of the raw
material, and of the removal of cortex and flawed
pieces in early stages of knapping that were largely
performed away from the Cahal Tok area.
Our anal ytical dist inction between hard-
hammer and biface thinning flakes was intended to
distinguish biface manufacture from core reduction
Whittaker et al.] LITHIC INDUSTRY IN A MAYA CENTER 147
Figure 9. Experimental axes. Large flakes to right result from hard hammer decortication of upper axe and the lower
flakes are from thinning and shaping with an antler billet. The original nodule weighed 4470 g, the axe weighs 790 g,
measures 205 x 79 x 40 mm, and produced 884 pieces of debitage. The lower axe’s nodule weighed 3730 g, the axe weighs
282 g, measures 140 x 70 x 27 mm, and produced 748 pieces of debitage.
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 147
to produce tool blanks, and in these contexts con-
veys the wrong impression. In particular, it should
be stressed that almost all the flakes in the column
samples probably derive from biface production,
and there are very few that appear to be struck off
other types of cores. Most of the flakes classified
as “hard hammer” show large platforms with dis-
tinct bulbs and points of impact, but also are rela-
tively flat and expanding, and have facetted
platforms and numerous dorsal scars (Figure 10).
In other words, they are also flakes from the pro-
duction of bifaces, and their hard- hammer traits
reflect the probable use of limestone hammers on
the edges of relatively thick bifaces, that is, axes.
A similar phenomenon is visible at Colha (Shafer
1985; Shafer and Hester 1983:Figure 6), and in
Whittaker’s experimental knapping. Most of the
“indeterminate” category flakes are also fragments
of probable biface thinning flakes that lack plat-
forms. LDF and Cahal Tok differ from Chiik Nah
(272–32), a nearby house mound (Whittaker et al.
2002), where the higher percentage of hard-
hammer flakes (29 percent) compared to biface
thinning flakes (6 percent) does indeed reflect the
presence of larger flakes struck off cores for use as
tool blanks. The differences are also reflected in the
relative paucity at Chiik Nah of small flakes that
would indicate on- site knapping. Evidence for the
production of chert blades as seen in some of the
Late Classic period workshops at Colha (Roemer
1991) is also completely lacking at Cahal Tok and
LDF.
Knapping occurred on the Cahal Tok platform
and the debris there is primary refuse. Some of the
lithic waste was removed from the knapping area,
however, by dumping it into the adjacent aban-
doned limestone quarry area, forming the LDF
Chert locus. The distributions of flake types and
sizes east to west in the column samples (Tables 4
and 5) relate Cahal Tok to the LDF deposits. Biface
flakes predominate throughout, visible especially
in the high proportions of biface thinning flakes
compared to very low proportions of flakes with
hard- hammer platforms. The main difference is
that LDF has higher proportions of larger flakes.
We interpret this to mean that knapping was actu-
ally done on the limestone surface and subsequent
platform above LDF, with larger waste being pref-
erentially removed and dumped over the edge to
form the LDF deposit. The lithic assemblage from
the column samples at Cahal Tok looks the same
whether it comes from levels above or below the
148 LATI N AMERICAN ANTIqUITY [Vol. 20, No. 1, 2009
Figure 10. Selected flakes from the LDF deposit to show range of materials and platforms. Those to the left above the
scale have relatively large facetted platforms and prominent bulbs, probably a result of knapping with limestone ham-
mers. Those to the right have small platforms, often lipped, and reduced bulbs. All flakes result from biface manufacture.
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 148
platform remnants, showing that the knapping
activities continued throughout the sequence.
Debitage Composition. T3 and T4 are repre-
sented only by the column samples. For the other
trenches (T1, T2, and T5–T9), which were off the
platform and did not have the concentrated lithic
debris, we analyzed the complete lithic collections.
These assemblages look different. Although biface
manufacture is still strongly represented, there are
more tools, more hard- hammer flakes (incl. a few
that are definitely not from biface manufacture),
more shatter, more cortical flakes, and larger flakes.
We interpret this as a more typical lithic assemblage
for El Pilar (compare to the Chiik Nah house-
mound) but with some of the biface industry from
LDF mixed in.
Microdebitage is often considered a marker of
the actual locations of knapping activities. As men-
tioned, we screened all column samples from LDF
and Cahal Tok T3 and T4 through quarter- inch (64
mm) screen before analysis. The fraction that
passed through the screen was not analyzed, but the
relative proportions of microdebitage by weight
(grams) are useful. Table 3 is arranged in spatial
order, from the LDF sample in the east, through T3
and T4 moving west. As noted, we found it
extremely difficult to take a sample of consistently
accurate dimensions, so all comparisons are by pro-
portion, although weights in grams of each sample
are given. The samples from LDF had little non-
lithic material, being about 20 percent soil and rock
by weight. The comparison to Cahal Tok 3 and 4
is not quite fair because the LDF material was dry
when weighed, and that from Cahal Tok was still
moist, but the differences between the areas are
still clear. Even in the denser levels of lithic debris
in the Cahal Tok trenches more of the fill was rock
and soil.
Microdebitage (less than 64 mm) makes up 2–7
percent of the LDF chert, except for Level 8 where
the bottom of the deposit mixes with the underly-
ing clay. In the column samples from T3 and T4 at
Cahal Tok, the microdebitage is mostly around 40
percent of the total debitage, with some samples,
especially smaller ones, much higher. In compari-
son, the microdebitage from the two experimental
axes makes up 3 percent of all the debitage, if the
extensive material from hard- hammer decortication
is included. As noted, this stage of work was appar-
ently underrepresented at LDF and Cahal Tok. If
we exclude the decortication material, and use only
the biface thinning stages (BTF) about 8 percent
of the debitage from the experimental axes is less
than 64 mm.
Cahal Tok thus has very high proportions of
small debitage and lower proportions of larger
flakes than LDF. The situation is best accounted for
by a sequence of activities. Our interpretation is that
primary materials were prepared before arrival at
Cahal Tok. Then at Cahal Tok, the knappers actu-
ally shaped the tools. Finally, they swept their waste
off the edge of the platform into the adjacent LDF
deposit when it mounted up inconveniently. Exam-
ples of similar associations between knapping areas
and debitage have been recognized at Colha (Hes-
ter and Shafer 1992; Roemer 1991; Shafer and Hes-
ter 1983). At Colha, some Late Classic period
debitage deposits are described as “talus deposits
of varying thickness, trailing off the edge of resi-
dential platforms” (Shafer and Hester 1983:529).
Cahal Tok is not, however a habitation area. The
lithic deposits are particularly sparse of sherds and
other household debris. Even in deposits west of
the dense debitage, the sherd count is low, and the
lack of utility forms for storage and cooking sug-
gests that a full household ceramic assemblage was
not present.
Scale of Production
Some of the arguments about Maya lithic craft have
focused on the scale of production. Calculating
productivity is inevitably based on many assump-
tions, and estimates based on different samples,
procedures, and theoretical expectations are noto-
riously difficult to compare. Mallory’s critique of
the evidence for workshops at Colha (Mallory
1986; Shafer and Hester 1986b) seems to have ini-
tiated a fondness for comparative density mea-
surements. Density measurements are a perfect
example of the problems of quantifying produc-
tivity and the assumptions on which calculations
are based. The vast range of variability in density
estimates reflects among other things the difficulty
of collecting standardized samples, variation in
packing and the admixture of fill, and the minimum
size of the flakes counted. Nevertheless, Table 6
compiles available Maya data to demonstrate two
points: (1) the Cahal Tok/LDF material is compa-
rable to densities at the better known Colha work-
shops; (2) it is far beyond what is normal at
Whittaker et al.] LITHIC INDUSTRY IN A MAYA CENTER 149
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 149
residential sites.
We used debitage density to try two methods for
estimating the number of axes that might have been
produced at Cahal Tok, one based on debitage
weight and the second based on number of pieces
of debitage. In both cases, we estimated total quan-
tities for the combined Cahal Tok and LDF deb-
itage, then used the data from the experimental axe
production to calculate a total number of axes.
Using the most conservative dimensions, the
size of the deposit is about 225 m3(15m 15 m
1 m) in the densest area, the LDF debitage dump,
with an additional 15 m3on the platform at Cahal
Tok. Using a weighed sample of known dimen-
sions, we calculated the weight of the LDF/Cahal
Tok deposit as 1,069 kg/m3. To put this in per-
spective, the weight of a sample of solid quartz the
same size would be 2,643 kg/m3, and loose dry
earth weighs 1,249 kg/m3(SImetric Co. 2006).
Shafer and Hester (1986b)4estimated that their
debitage deposits at Colha weighed about 960
kg/m3.
Estimating a density of 1,069 kg/m3and a
deposit volume of 240 m3, we calculate 250,000
kg as a conservative estimate of the total weight of
the LDF/Cahal Tok debitage. Our two axe replica-
tions approximated the size range for El Pilar axes,
and were made from nodules that were probably
larger than the already reduced blanks implied by
the Cahal Tok debitage. Not counting the hard-
hammer decortication material and microflakes,
the large axe produced 1,396 g of analyzable biface
reduction debitage, and the smaller axe 1,460 g. If
we convert this to an estimated— but, again con-
servatively broad— range of 1,000 to 1,500 g per
axe, this suggests that somewhere between 166,000
and 250,000 axes were produced at El Pilar in the
Cahal Tok/LDF area.
The number of pieces of flake and shatter 64 mm
(quarter inch) or greater for 10 10 15 cm col-
umn samples from levels 2, 4, 6, and 7 averaged
1,211. Rounding this to 1,200 we calculate an
approximate average density of 800,000 flakes/m3.5
Again using our estimate of 240 m3of debitage,
this density produces a figure of 192 million total
pieces. The two experimental axes each produced
about 600 pieces of biface thinning flakes and
related shatter after the initial decortication that we
believe took place away from the locus. These fig-
ures allow an estimate of about 320,000 axes pro-
150 LATI N AMERICAN ANTIqUITY [Vol. 20, No. 1, 2009
Table 6. Comparisons of Debitage Density
Density of Weight
Debitage Debitage
Site Site Type Date (pieces/m3) (kg/m3) Comments Source
LDF Workshop Late Classic 800,000 1069 > 1/4" This project
LDF "Quarry" Late Classic 65,690 Not Available > 1/4" Michaels 1993
LDF "Quarry" Late Classic 550,000 Not Available > 1/4" Ford and Olsen 1989
Yaxox "Quarry" Late Classic? 1,541,600 Not Available > 1/4" Ford and Olsen 1989
272-229 "Quarry" Late Classic? 6,994 Not Available > 1/4" Ford and Olsen 1989
Colha Op2024 Workshop Late Preclassic 603,000 642 > 6 cm Drollinger 1989; Shafer and Hester 1986b, 1991
Colha Op2007 Workshop Late Classic 4,956,125 960 > 3 cm Roemer 1991; Shafer and Hester 1986b, 1991
Colha Op2037 Workshop Early Post Classic 575,443 876 No size information provided Michaels 1989
Copan Valley Residential Late Classic 3,500 8.9 Obsidian, densest site, no size Mallory 1986
information provided
Copan Valley Residential Various Late Classic mean=97 Not Available Calculation process not described Mallory 1984 in Ford and Olsen 1989
range 3-482
Belize River Residential Various mean=430 Not Available > 1/4" Ford and Olsen 1989
Valley range 13-6,994
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 150
duced. Using the average of total debitage pieces
(816, incl. decortication), gives around 235,000
axes.
We have tried to be conservative in our esti-
mates, and obviously, these numbers should be
taken with a grain of salt. For one thing, in manip-
ulating the data in terms of “axes,” we have not con-
sidered the unknown but possibly substantial
contribution of the smaller thin bifaces. However
imprecise our estimates certainly are, they do
demonstrate one critical point: the Cahal Tok work-
shop produced very large numbers of axes. If the
locus was in use for the whole 250-year span of the
Late Classic period, more than 1,000 axes a year
could have been manufactured there. With this long
an occupation span, knappers at the locus would
have produced at least several axes a day. Less con-
servative estimates of debitage quantities would
elevate production values, as would shortening the
hypothetical use- life of the platform, which we
have no way of accurately estimating. In rough
terms, the workshop at El Pilar is comparable to
some individual loci at Colha, although at Colha
there are many larger deposits and a longer time
span, and the concentration of many loci implies a
much larger overall productivity.
Conclusions: Cahal Tok in Social Context
In Maya archaeology, much discussion of lithic
industries, and craft production in general, has
attempted to come to terms with the question of
occupational specialists and the organization of
production in Maya centers (Aldenderfer et al.
1989; Aoyama 2001; Clark 1987, 2003; Fedick
1991; Ford and Olson 1989; Fowler 1991; Mallory
1984, 1986; Masson 2001; McSwain 1991;
Michaels 1989, 1993; Moholy- Nagy 1997; Potter
1993; Shafer and Hester 1983, 1986b, 1991; Sheets
1978). Basic to these discussions is the problem of
identifying workshops and deposits resulting from
them. Many studies have pointed out that debitage
concentrations should not be uncritically inter-
preted as representing workshop loci; many are
more likely to be secondary “dumps” (Clark 1986;
Mallory 1986; Moholy- Nagy 1990). Although
some of these studies appear to stem from a bias
against the idea of industrial scale lithic production
in prehistoric times, the underlying point is still
valid. So do Cahal Tok and LDF represent a lithic
production locus, or merely a dump of hazardous
waste?
Modern discussions of lithic waste disposal do
indeed relate to our feelings that sharp flakes were
hazardous to prehistoric feet, and the observation
that in some ethnographic and archaeological sit-
uations, they were removed from underfoot (Clark
1991; Moholy- Nagy 1990; Whittaker and Kaldahl
1996[CE2]). This makes sense when knapping is
done at home, but may be less critical in a work-
shop context. Furthermore, it is often assumed that
“debitage in archaeological context can be regarded
as having no potential value for those who dis-
carded it” (Moholy- Nagy 1990:272). In fact, even
in some “dump” contexts, waste flakes may remain
a useful resource, a handy supply of cutting edges
and blanks for tool making (Healan 1992; Whit-
taker and Kaldahl 2001). In any case, a great many
cultures regarded debitage with indifference, to
judge by the quantities scattered around and the lack
of facilities for segregating it. The Maya compli-
cated matters for archaeologists by moving great
quantities of fill around their centers during con-
struction, demolition, and remodeling of monu-
ments. This fill included debris from households
and apparent specialist workshops, including lithic
debris (Moholy- Nagy 1997), as well as quarried
limestone, clay, and other material. It was also not
uncommon to include large deposits of debitage
around high status tombs as at Tikal (Moholy- Nagy
1997) and other sites, including those in Belize
(Healy et al. 2004).
Although we can probably rule out the possi-
bility of a tomb at LDF, because instances of deb-
itage around tombs are always in high status
architectural contexts and built tomb structures,
there are a number of possible interpretations of
Cahal Tok and the LDF chert deposit. It could be
that stoneworkers all over El Pilar brought their
waste to one spot. Mallory (1986) suggested this
for Colha, but as at Colha (Shafer and Hester
1986b) in that case we should expect more admix-
ture of humic layers and other trash. At Cahal Tok
the debitage is increasingly mixed with soil and a
limited amount of trash as one moves west away
from LDF, but the LDF deposits contain almost
nothing but debitage. Our material also has little of
the evidence of flake tools and their manufacture
common on residential sites or domestic middens;
it is a specialized deposit. Nor is the debitage in our
Whittaker et al.] LITHIC INDUSTRY IN A MAYA CENTER 151
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 151
case architectural fill. It could be suggested that the
original deposit was the LDF material, and some
of that was then scooped up and used as founda-
tion material for the platform at Cahal Tok. How-
ever, the Cahal Tok deposits include less of the
larger debitage, the platform rests at least partly on
bedrock, and the debitage continued to be deposited
on top of the platform. It makes more sense to inter-
pret the edge of the limestone rise at Cahal Tok and
the platform built on it as the actual workshop, and
the deposit at LDF as the associated dump.
The presence of microdebitage is often cited as
evidence of on- site knapping (Behm 1983; Clark
1986; Moholy- Nagy 1990; Shafer and Hester
1986b), although it can be argued that knappers
could work over a container or sweep clean a plas-
tered floor (Clark 1986). At Cahal Tok and LDF,
small flakes and microdebitage are common. The
extremely high proportion of Cahal Tok debitage
that is less than 64 mm is a strong indication that
knapping was done on site. As debris accumulated,
it was swept off the edge of the limestone ledge
(and later the built platform), where it formed the
deep deposit recognized as the LDF chert site.
Late Classic period workshops at Colha, which
were also producing biface axes as part of their out-
put, are usually, but not always, associated with res-
idential platforms (Masson 2001; Roemer 1980,
1991; Shafer and Hester 1983). The obsidian blade
production site at 272–136 not far from El Pilar is
part of a large residential compound housing crafts-
men (Ford and Olson 1989; Hintzman 2000; Olson
1994). At Cahal Tok, we extended our trenches
west beyond the debitage areas to test this expec-
tation. Although there are possible platform rem-
nants and some general artifactual debris, the
densities, especially of normal household ceramic
assemblages, are very low compared to other El
Pilar contexts, and we found none of the manos and
metates that are also typical of habitation sites. We
interpret the area as lacking well- established resi-
dential areas. The LDF deposit appears to be fill-
ing an old limestone quarry depression. Although
close to the monumental center, the site is in a zone
of quarry activity with few house mounds, so the
knapping work and its resulting debitage would
have been out of the way, but accessible.
The next question then, is who did the knapping
and who controlled production? Cahal Tok and
LDF do not fit the usual expectations for large Maya
sites. Central sites are often seen as including res-
idential compounds whose inhabitants engaged in
craft activities (Becker 1973), often making pres-
tige goods, but the centers are usually believed to
be more involved in distribution than in production,
especially when considering common tools (Fedick
1991). Potter (1993:201) goes so far as to say
“Without exception, lithic workshops occurred dur-
ing the Classic period at sites that were not occu-
pying primary or even secondary positions within
their regional settlement hierarchies,” and Moholy-
Nagy (1997) argues that surface debitage dumps
occur at small sites like Colha,6whereas concen-
trated debitage tends to be found at large sites like
Tikal as buried aggregates. In our area, Ford (1991,
2004; Ford and Olson 1989) has used lithic and
other evidence to argue that the occupants of
smaller sites in less productive foothill zones relied
in part on craft specialization to integrate them-
selves into the larger economic system, and simi-
lar models are often applied elsewhere, for example
in the Three Rivers region (Hyde 2003).
The debitage at Cahal Tok and LDF indicates a
very specialized operation. Virtually all the mate-
rial results from biface production, focusing on
thick “axe” forms made from preforms that had
been quarried and prepared elsewhere. Systematic
manufacture of bifaces, especially thin bifaces and
carefully made axes, implies a high level of knap-
ping skill. Some of the stone was selected for high
quality, although the presence of some poor qual-
ity stone also suggests that the knappers were not
always fussy about raw material. Perhaps access
was sometimes difficult, or perhaps material qual-
ity did not matter too much for utilitarian axes, or
as Thompson (1991) suggests for Becan, there were
different levels of skill within the workforce. In any
case, the restricted forms and use of great quanti-
ties of material that had to be carried in implies a
fairly organized system.
Cahal Tok and the LDF chert deposit are prac-
tically “downtown.” The central area of monu-
mental architecture at El Pilar would have been
visible from Cahal Tok; Plaza Copal is only 150 m
to the southeast, and Plazas Duende and Faisan are
about 50 m east. About 50 m south of Cahal Tok is
the large causeway or sacbe running west from the
El Pilar center toward another monumental group
called Pilar Poniente in Guatemala, and a major
aguada is about 100 m northwest. This whole sec-
152 LATI N AMERICAN ANTIqUITY [Vol. 20, No. 1, 2009
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 152
tor of El Pilar is dotted with house mounds and must
have been densely inhabited during much of the
occupation. However, in the immediate area around
LDF and Cahal Tok mounds are relatively scarce
and the topography is broken by small limestone
outcrops and quarry pits. Evidently, this was an area
of industrial activities, set close to the heart of El
Pilar.
Proximity to elite centers could indicate that the
knappers at Cahal Tok were under direct elite con-
trol. The distribution of common goods such as
obsidian blades or axes produced by specialists at
a limited number of loci is often taken as evidence
of a centralized distribution system in which the
lower levels receive goods as rewards for partici-
pation, and the ability to distribute these goods bol-
sters elite prestige (Ford 2004:248). However, less
centralized entrepreneurship, including the market
economies being considered by some Mayanists
(Dahlin et al. 2007; Hirth 2000; Jones 1996), could
produce similar patterns. Usually the evidence is
equivocal. Lithic workshops are often associated
with habitation units or plazuela groups at Colha.
At El Pedernal, some were (Black 1987), but oth-
ers appeared isolated, and one overlay a limestone
quarry as is apparently the case at LDF. At Cahal
Tok, the knappers apparently did not live on site.
This could imply that knappers who lived else-
where were employed at a state “factory,” or merely
that knapping by individuals was best done coop-
eratively, as in some ethnographic situations (Stout
2002), and in some locale where the debris was not
a nuisance. The knappers produced mostly com-
mon tools, although thin bifaces may have had
some more prestigious uses. There is no evidence
that the knappers had access to exotic material. The
scale of production, although far beyond that of any
single household’s needs, is not so grand as to sug-
gest a major manufacturing enterprise, like that
seen at Colha.
In our opinion, it is likely that many major cen-
ters have such small industrial zones. We would
expect that many tools were produced for local
consumption in both large and small sites. In fact,
lithic workshops in and near centers, comparable
to Cahal Tok/LDF at El Pilar, have been recog-
nized at sites such as Chan Chich (McDow 2000;
Meadows and Hartnett 2000), and others. Smaller
centers like El Pedernal near Rio Azul (Black and
Suhler 1986) have workshops that sound very like
Cahal Tok/LDF, with a limited range of large bifa-
cial tools made on mediocre material, probably to
serve local agriculture. Thompson (1991) inter-
prets mounds at Becan, Campeche, Mexico, as
workshops where chert nodules were accumulated
and decorticated for flaked celt manufacture. Even
in the “Northern Belize Chert Bearing Zone” near
Colha, local manufacture, especially of expedient
tools, complicates models of centralized produc-
tion (Speal 2006).
Evidence of small industrial operations will be
hard to discern, if it takes the form of small artifact
concentrations in densely forested sites. Colha was
discovered and studied in part because debitage
was visible on the surface of mounds in cleared pas-
ture (McNatt and Cox 1979; Wilk 1976), and the
LDF locus was noticed in 1990 after it had been
recently burned. As Moholy- Nagy (1997) points
out, masses of debitage in architectural fill and else-
where are evidence of large- scale tool production,
even if not of workshop locations, and such evi-
dence is common in large centers like Tikal.
Specialized production of bifaces was certainly
the focus of the Cahal Tok workshop. The debitage
on the Cahal Tok platform and in the LDF dump
represents the work of many knappers in one place
and the sustained and continuous use of a knapping
area over a long period of time, using a limited
range of material to produce and rework a couple
of tool types. All of these factors are consistent
with specialized production. Whether the location
of the workshop close to the high- status monu-
ments indicates centralized control or whether the
location is merely an expedient one, close to the
sacbe and near a quarry pit convenient for waste
disposal, depends more on our biases about the
Maya than the evidence of this site.
References Cited
Aldenderfer, Mark, Larry R. Kimball, and April Sievert
1989 Microwear Analysis in the Maya Lowlands: The Use
of Functional Data in a Complex- Society Setting. Journal
of Field Archaeology 16(1):47–60.
Aoyama, Kazuo
2001 Classic Maya State, Urb anism, and Ex change:
Chipped Stone Evidence of the Copan Valley and Its Hin-
terland. American Anthropologist 103:346–360.
2007 Elite Artists and Craft Producers in Classic Maya
Society: Lithic Evidence from Aguateca, Guatemala. Latin
American Antiquity 18(1):3–26.
Baumler, Mark F., and Leslie B. Davis
2004 The Role of Small- Sized Debitage in Aggregate Lithic
Whittaker et al.] LITHIC INDUSTRY IN A MAYA CENTER 153
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 153
Analysis. In Aggregate Analysis in Chipped Stone, edited
by Christopher T. Hall and Mary Lou Larson, pp. 45–64.
University of Utah Press, Salt Lake City.
Becker, Marshall J.
1973 Archaeological Evidence for Occupational Special-
ization among Classic Period Maya at Tikal, Guatemala.
American Antiquity 38(4):396–406.
Behm, Jeffrey A.
1983 Flake Concentrations: Distinguishing Between Flint-
working Activity Areas and Secondary Deposits. Lithic
Technology 12(1):9–16.
Black, Stephen L.
1987 Settlement Pattern Survey and Testing, 1985. In Rio
Azul Reports Number 3, the 1985 Season, edited by
Richard E. W. Adams, pp. 183–221. Center for Archaeo-
logical Research, University of Texas, San Antonio.
Black, Stephen L., and Charles K. Suhler
1986 The 1984 Rio Azul Settlement Survey. In Rio Azul
Reports Number 2, the 1984 Season, edited by Richard E.
W. Adams, pp. 163– 192. Center for Archaeological
Research, University of Texas, San Antonio.
Cackler, Paul R., Michael D. Glasscock, Hector Neff, Harry Ice-
land, K. Anne Pyburn, Dale Hudler, Thomas R. Hester, and
Beverly Chiarulli
1999 Chipped Stone Artefacts, Source Areas, and Prove-
nance Studies of the Northern Belize Chert- Bearing Zone.
Journal of Archaeological Science 26:389–397.
Clark, John E.
1986 From Mountains to Molehills: A Critical Review of
Teotihuacan’s Obsidian Industry. In Economic Aspects of
Prehispanic Highland Mexico. Research in Economic
Anthropology. Supplement2, edited by Barry L. Isaac, pp.
23–74. JAI Press, Greenwich.
1987 Politics, Prismatic Blades, and Mesoamerican Civi-
lization. In The Organization of Core Technology, edited
by J. K. Johnson and C. A. Morrow, pp. 259–284. West-
view Press, Boulder.
1991 Flintknapping and Debitage Disposal among the
Lacandon Maya of Chiapas, Mexico. In The Ethnoar-
chaeology of Refuse Disposal. Arizona State University
Anthropological Research Papers 42, edited by E. Staski
and L. D. Sutro, pp. 63–78. Arizona State University,
Tempe.
2003 A Review of Twentieth- C entury Mesoamerican
Obsidian Studies. In Mesoamerican Lithic Technology:
Experimentation and Interpretation,edited by Kenneth G.
Hirth, pp. 15–54. University of Utah Press, Salt Lake City.
Dahlin, Bruce H., Christopher T. Jensen, Richard E. Terry, David
R. Wright, and Timothy Beach
2007 In Search of an Ancient Maya Market. Latin Ameri-
can Antiquity 18:363–384.
Drollinger, H. D., Jr.
1989 An Investigation of a Late Preclassic Chert Workshop
for Colha, Belize. Unpublished Master’s thesis, Depart-
ment of Anthropology, Texas A&M University.
Fedick, Scott
1991 Chert Tool Production and Consumption among Clas-
sic Period Maya Househol ds. In Maya Stone Tools:
Selected Papers from the Second Maya Lithic Conference,
edited by Thomas R. Hester and Harry J. Shafer, pp.
103–118. Prehistory Press, Madison.
Ferguson, Jeffrey R.
2003 An Experimental Test of the Conservation of Raw
Material in Flintknapping Skill Acquisition. Lithic Tech-
nology 28(2):113–131.
Ford, Anabel
1991 Economic Variation of Ancient Maya Residential Set-
tle ment in the Uppe r Beliz e River Area. Anci ent
Mesoamerica 2:35–46.
2000 Unpublished Lab Manual. BRASS El Pilar Project.
University of Santa Barbara.
2004 Integration among Communities, Center, and Regions.
In The Ancient Maya of the Belize Valley: Half a Century
of Archaeological Research, edited by James Garber, pp
238–256. University Press of Florida, Gainesville.
Ford, Anabel, and Kirsten Olson
1989 Aspects of Ancient Maya Household Economy: Vari-
ation in Chipped Stone Production and Consumption. In
Prehistoric Maya Economies of Belize. Research in Eco-
nomic Anthropology. Supplement 4, edited by Patricia
McAnany and Barry Isaac, pp. 185–211. JAI Press, Green-
wich.
Fowler, William
1991 Lithic Analysis as a Means of Processual Research in
Southern Mesoamerica: A Review of Recent Research. In
Maya Stone Tools: Selected Papers from the Second Maya
Lithic Conference, edited by Thomas R. Hester and Harry
J. Shafer, pp. 1–20. Prehistory Press, Madison.
Gifford, James C.
1976 Prehistoric Pottery and Ceramics of Barton Ramie in
the Belize Valley. Peabody Museum Memoir 18. Harvard
University, Cambridge.
Grimm, Linda
2000 Apprentice Flintknapping: Relating Material Culture
and Social Practice in the Upper Paleolithic. In Children
and Material Culture, edited by J. S. Derevenski, pp.
53–71. Routledge, London.
Healan, Dan M.
1992 A Comment on Moholy- Nagy’s “The Misidentifica-
tion of Lithic Workshops.” Latin American Antiquity
3(3):240–242.
Healy, Paul F., Jaime J. Awe, and Hermann Helmuth
2004 Defining Royal Maya Burials: A Case from Pacbitun.
In The Ancient Maya of the Belize Valley: Half a Century
of Archaeological Research, edited by James Garber, pp.
228–237. University Press of Florida, Gainesville.
Hester, Thomas R.
1985 The Maya Lithic Sequence in Northern Belize. In
Stone Tool Analysis: Essays in Honor of Don E. Crabtree,
edited by M. Plew, J. Woods, and M. Pavesic, pp. 187–210.
University of New Mexico Press, Albuquerque.
Hester, Thomas R., and Harry J. Shafer
1992 Lithic Workshops Revisited: Comments on Moholy-
Nagy. Latin American Antiquity 3(3):243–248.
Hintzman, Marc William
2000 Scarce- Resource Procurement and Use: The Techno-
logical Analysis of an Obsidian Blade Workshop in the
Lowlands of Belize. Unpublished Master’s thesis, Uni-
versity of California, Riverside.
Hirth, Kenneth
2000 The Economics of Distribution: Marketplaces, Mar-
ket Systems, and Long- Distance Trade. In Ancient Urban-
ism at Xochicalco: The Evolution and Organization of a
Pre- Hisapanic Society, Archaeological Research at Xochi-
calco, Vol. 1, edited by Kenneth Hirth, pp. 181–209. Uni-
versity of Utah Press, Salt Lake City.
Hyde, David Michael
2003 Lithic Technological Organization in the Three Rivers
Region of the Maya Lowlands. Unpublished Master’s the-
sis, Washington State University.
Jones, Christopher
1996 Tikal Report No. 16: Excavations in the East Plaza
of Tikal.University Museum, University of Pennsylvania,
Philadelphia.
154 LATI N AMERICAN ANTIqUITY [Vol. 20, No. 1, 2009
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 154
Kamp, Kathryn A., John Whittaker, Rafael Guerra, Kimberly
McLean, Peter Brands, and José V. Guerra Awe
2007 A Ritual Spindle Whorl Deposit from the Late Clas-
sic Maya Site of El Pilar, Belize. Journal of Field Archae-
ology 31(4):411–423.
Lewenstein, Suzanne M.
1987 Stone Tool Use at Cerros. University of Texas Press,
Austin.
Lucero, Lisa J.
2001 Social Integration in the Ancient Maya Hinterlands:
Ceramic Variability in the Belize River Area. Anthropo-
logical Research Paper 53. Arizona State University,
Tempe.
Mallory, John K.
1984 Late Classic Maya Economic Specialization: Evi-
dence from the Copan Obsidian Assemblage. Unpublished
Ph.D. dissertation, Department of Anthropology, Penn-
sylvania State University.
1986 “Workshops” and “Specialized Production” in the
Production of Maya Chert Tools: A Response to Shafer
and Hester. American Antiquity 51(1):152–158.
Masson, Marilyn
2001 The Economic Organization of Late and Terminal
Classic Period Maya Stone Tool Craft Specialist Work-
shops at Colha, Belize. Lithic Technology 26(1):29–49.
McAnany, Patricia
1992 Agricultural Tools and Tasks: Patterns of Stone Tool
Discard Near Prehistoric Maya Residences Bordering Pull-
trouser Swamp, Belize. In Gardens of Prehistory: The
Archaeo logy of Settle me nt Ag ricul tu re in Great er
Mesoamerica, edited by Thomas W. Killion, pp. 184–213.
University of Alabama Press, Tuscaloosa.
McDow, David A.
2000 Analysis of Stone Tools from Chan Chich. In The
1998 and 1999 Seasons of the Chan Chich Archaeologi-
cal Project. Papers of the Chan Chich Project 4, edited by
Brett Houk, pp. 105–118. University of Texas, Austin.
McNatt, Logan, and Waynne Cox
1979 A Note on Mapping at the Site of Colha, 1979. In The
Colha Project, 1979: A Collection of Interim Papers,edited
by Thomas R. Hester, pp. 138–144. Center for Archaeo-
logical Research, University of Texas, San Antonio.
McSwain, Rebecca
1991 A Comparative Evaluation of the Producer- Consumer
Model for Lithic Exchange in Northern Belize, Central
America. Latin American Antiquity 2(4):337–351.
Meadows, Richard K., and Kristin M. Hartnett
2000 Archaeological Investigations at Group H: Investi-
gating Craft Production and Domestic Architecture at Chan
Chich, Belize. In The 1998 and 1999 Seasons of the Chan
Chich Archaeological Project. Papers of the Chan Chich
Project 4, edited by Brett Houk, pp. 15–39. University of
Texas, Austin.
Michaels, George H.
1989 Craft Specialization in the Early Postclassic of Colha.
In Prehistoric Maya Economies of Belize. Research in
Economic Anthropology. Supplement 4, edited by B. L.
Isaac and Patricia A. McAnany, pp. 139–183. JAI Press,
Greenwich.
1993 Evidence for Lithic Craft Specialization by the Clas-
sic Period Maya of the Upper Belize River Valley, Belize.
Unpublished Ph.D. dissertation, University of California,
Santa Barbara.
Milne, S. Brooke
2005 Paleo- Eskimo Novice Flintknapping in the Eastern
Can adian Arcti c. Journal of Field Arch aeology
30(3):329–345.
Moholy- Nagy, Hattula
1990 The Misidentification of Mesoamerican Lithic Work-
shops. Latin American Antiquity 1(3):268–279.
1997 Middens, Construction Fill, and Offerings: Evidence
for the Organization of Classic Period Craft Production at
Tika l, Guat emala . Journal of Field Archaeol ogy
24(3):293–313.
2003 Tikal Report No. 27, part B: The Artifacts of Tikal:
Utilitarian Artifacts and Unworked Material. University
of Pennsylvania Museum of Archaeology and Anthropol-
ogy, Philadelphia.
Oland, Maxine H.
1999 Lithic Raw Material Sources at the Southern End of
the Freshwater Creek Drainage: A View from Laguna de
On, Belize. Lithic Technology 24(2):91–110.
Olson, Kirsten A.
1994 Inclusive and Exclusive Mechanisms of Power: Obsid-
ian Blade Production and Distribution among the Ancient
Maya of the Belize River Area. Unpublished Master’s the-
sis, University of California, Los Angeles.
Potter, Daniel R.
1993 Analytical Approaches to Late Classic Maya Lithic
Industries. In Lowland Maya Civilization in the Eighth
Century A.D., edited by Jeremy Sabloff and John S. Hen-
derson, pp. 273–298. Dumbarton Oaks Research Library
and Collection, Washington, DC.
Roemer, Erwin
1980 Operation 2007: A Preliminary Report on the Exca-
vation of a Late Classic Maya Lithic Workshop. In The
Colha Project Second Season, 1980 Interim Report,edited
by Thomas R. Hester, J. D. Eaton, and Harry J. Shafer, pp.
87–104. Center for Archaeological Research, University
of Texas, San Antonio.
1991 A Late Classic Workshop at Colha, Belize. In Maya
Stone Tools: Selected Papers from the Second Maya Lithic
Conference, edited by Thomas R. Hester and Harry J.
Shafer, pp. 55–66. Prehistory Press, Madison.
Shafer, Harry J.
1985 Two Maya Lithic Workshops at Colha, Belize. In
Stone Tool Analysis: Essays in Honor of Don E. Crab-
tree,edited by M. G. Plew, J. C. Woods, and M. G. Pavesic,
pp. 277–315. University of New Mexico Press, Albu-
querque.
2000 The Lithic Artifacts of the Pulltrouser Area: Settle-
ments and Fields. In Pulltrouser Swamp: Ancient Maya
Habitat, Agriculture, and Settlement in Northern Belize,
edited by B. L. Turner and P. D. Harrison, pp. 212–245.
University of Utah Press, Salt Lake City.
Shafer, Harry J., and Thomas R. Hester
1983 Ancient Maya Chert Workshops in Northern Belize,
Central America. American Antiquity 48(3):519–543.
1986a An Ancient Maya Hafted Stone Tool from Northern
Belize. Working Papers in Archaeology 3. Center for
Archaeological Research, University of Texas, San Anto-
nio.
1986b Maya Stone- Tool Craft Specialization and Produc-
tion at Colha, Belize: Reply to Mallory. American Antiq-
uity 5(1):158–166.
1991 Lithic Craft Specialization and Product Distribution
at the Maya Site of Colha, Belize. World Archaeology
23(1):79–97.
Sheets, Payson
1978 From Cra ftsman to Cog: Quantitative Views of
Mesoamerican Lithic Technology. In Papers on the Econ-
omy and Architecture of the Ancient Maya. Monograph 8,
edited by Raymond V. Sidrys, pp. 40–71. Institute of
Archaeology, University of California, Los Angeles.
Whittaker et al.] LITHIC INDUSTRY IN A MAYA CENTER 155
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 155
156 LATI N AMERICAN ANTIqUITY [Vol. 20, No. 1, 2009]
SImetric Co.
2006 Density of Materials. SImetric Co. UK webpage. Elec-
tronic document, http://www.simetric.co.uk/si_materi-
als.htm, accessed January 6, 2006.
Smith, R. E.
1955 Ceramic Sequence at Uaxactun, Guatemala. Publi-
cation 20. Middle American Research Institute, Tulane
University, New Orleans.
Speal, C. Scott
2006 Regional Economic Integration in the Coastal Maya
Lowlands: The Lithic Assemblage of Saktunha, Belize.
Lithic Technology 31(1):3–26.
Stemp, W. James
2004 Maya Coastal Subsistence and Craft Production as San
Pedro, Ambergris Caye, Belize: The Lithic Use- Wear Evi-
dence. Lithic Technology 29(1):33–73.
Stout, Dietrich
2002 Skill and Cognition in Stone Tool Production: An
Ethnographic Case Study from Irian Jaya. Current Anthro-
pology 43(5):693–722.
Thompson, Marc
1991 Flaked Celt Production at Becan, Campeche, Mex-
ico. In Maya Stone Tools: Selected Papers from the Sec-
ond Maya Lithic Conference, edited by Thomas R. Hester
and Harry J. Shafer, pp. 143–154. Prehistory Press, Madi-
son.
Whittaker, John, and Eric Kaldahl
2001 Where the Waste Went: A Knapper’s Dump at
Grasshopper Pueblo, Arizona. In Lithic Debitage: Context,
Form, Meaning, edited by William Andrefsky, pp. 32–60.
University of Utah Press, Salt Lake City.
Whittaker, John, Kathryn Kamp, Melissa Badillo, Zerifeh Eiley,
Jennifer Thornton, and Alexander Woods
2002 Excavations at Chiik Nah, A Maya Housemound at
El Pilar, Belize. Unpublished report, Institute of Archae-
ology, Belmopan, Belize.
Wilk, Richard
1976 Work in Progress at Colha, Belize, 1976. In Maya
Lithic Studies: Papers from the 1976 Belize Field Sympo-
sium. Special Report 4, edited by Thomas R. Hester and
Norman Hammond, pp. 35–40. Center for Archaeologi-
cal Research, University of Texas, San Antonio.
Woods, James C., and Gene L. Titmus
1996 Stone on Stone: Perspectives on Maya Civilization
from Lithic Studies. In Eighth Palenque Round Table,
edit ed by M. G. Robertson, pp. 47 9–489. The Pre -
Columbian Art Research Institute, San Francisco.
Notes
1. The Belize River Archeological Site Survey Project
under Annabel Ford began work at El Pilar in 1983. Our exca-
vations at Cahal Tok in June 2004 were under permit from the
Institute of Archaeology, Belize (permit number
I/A/H/2/1/04(07) granted to Anabel Ford). The project was
funded by a grant from the Henry Luce Foundation to
Kathryn Kamp, David Campbell, and John Whittaker of
Grinnell College. The Belize River Archaeological Site
Survey under Anabel Ford (University of California, Santa
Barbara) provided a home base and logistical support.
2. Some projects have attempted to define some or all of
these as distinct types, and there are probably temporally and
regionally meaningful variants of these large bifacial tools.
For instance, the “oval biface” as defined at Colha and
Pulltrouser Swamp, made from the unusually good Colha
chert, is thinner and broader than the “general utility biface,”
and characteristic of the Late Preclassic period (Hester 1985;
McAnany 1992). Elsewhere, as in the Three Rivers region
(Hyde 2003; Meadows and Hartnett 2000:25) it appears that
these types intergrade or are used with some overlap of defi-
nition. The specimens we see at Cahal Tok/LDF and other
loci at El Pilar are relatively small and thick, and would prob-
ably be called “general utility bifaces” by most.
3. Definitions of microdebitage vary, as discussed by
Baumler and Davis (2004). Here microdebitage means all
debitage that passed through quarter- inch (64 mm) screen.
4. The discrepancy between Colha and LDF weights
reflects the difficulty of comparing sites where different mea-
surement standards and assumptions were used, because the
debitage at Colha is if anything denser than that at LDF.
5. We cannot explain why Michaels’s (1993) estimate of
65,690 pieces/m3is so grossly different from our estimates,
although it was apparently made rapidly in the field.
6. Colha was actually a regional center for part of its his-
tory.
Submitted June 4, 2008; Accepted June 20, 2008.
LAQ20(1) Whitaker:Layout 1 3/3/09 9:37 AM Page 156
