Anth 473L/573L:
Archaeological Measurement and Lab Analysis
Spring 2003
This exercise is
intended to help you integrate the concepts of validity and reliability into your
thinking as evaluative tools through a detailed consideration of the research
of other archaeologists. Specifically,
you will apply the concepts in an evaluation of the measurements, units and
inferences regarding the number of ceramic vessels deposited in the trash mound
at Pueblo Alto as well as “explanations” of the causes of that deposition. Pueblo Alto is a large, 133 -room masonry
pueblo and is considered one of the “Great Houses” in the Chacoan system. It is located on the mesa top north of Chaco
canyon, approximately 50 m above the valley floor where Pueblo Bonito and
several other Great Houses are located.
GENERAL
BACKGROUND
The
archaeological record at Chaco Canyon, especially in a central area along
approximately 16km of Chaco Wash around Pueblo Bonito, has played a seminal
role in the history of Southwestern archaeology. The 14 large ruins along the canyon bottom and on the mesa tops
are impressive (go visit, if you haven’t already!). Because of the dramatic architecture and fine construction of the
ruins, and especially of Pueblo Bonito, Chaco was one of the first locations of
major (semi-) systematic excavation in the Southwest. Pueblo Bonito was excavated by George Pepper and others beginning
in the 19th century. Neil
Judd continued excavations there in the early decades of the 20th
century and Florence Hawley Ellis conducted field schools for UNM at other
sites in Chaco Canyon in the 1930s and 40s.
The National Park Service, in association with UNM Department of
Anthropology continued that work in the 1970s and 80s, examining roads, Great
Houses away from the central canyon (“outliers”), and smaller settlements. Some of that material is still being
processed, analyzed and published.
The
archaeological sequence in Chaco Canyon is very long, but the apex of
development occurred over a 200+ year period, from roughly AD 900 to 1120.
Archaeologists refer to the cultural expressions during this time as the “Chaco
Phenomenon”, or the Chaco regional system, and it is this period that figures
so prominently in considerations of pre-Hispanic cultural complexity in the
Southwest. Depending on how it is
measured, the regional system encompassed at least 65,000 km2 of the
San Juan Basin and adjacent uplands. A
number of traits define this expression, including:
Ø
Shared
architectural styles and similar material culture inventories.
Ø
Two
settlement types – large, architecturally impressive Great Houses and
contemporaneously occupied small sites, usually located near each other.
Ø
Outliner
communities throughout the San Juan Basin and beyond.
Ø A “road system” connecting (at least
symbolically) many of the outliers and some smaller sites to each other and to
the Great Houses in the canyon.
Ø
Extensive
exchange, long distance transport of raw materials for construction and,
perhaps, specialized production of craft items.
Ø
High
quality and exotic materials in the Great Houses in the canyon.
The Great
Houses, especially those within the canyon, appear to have been planned, and
most were multi-storied. The sandstone
core-veneer masonry is extremely well made, and the rooms, by comparison with
others of the same period, are spacious with high ceilings. The material from these Great Houses
includes habitation debris, high quality “art objects” of turquoise and other
materials and exotics, including Macaw skeletons and copper bells from Mexico.
The meaning,
structure, and functioning of the Chacoan system are the subjects of ongoing
debates. Some archaeologists view the
Chacoan system in economic, subsistence, risk management terms. In this interpretation, the Great Houses of
the canyon functioned as both economic redistribution centers (for trade,
etc.), and as storage facilities. The
Great Houses existed to buffer the risks of agricultural subsistence in the
marginal, spatially and temporally variable Southwest climate. More recent explanations of Chaco phenomenon
have emphasized religion and power, with the Great Houses in the canyon
symbolizing real and assumed power (e.g. Renfrew 2001). Rather than villages or towns, the Great
Houses are interpreted as ceremonial centers and pilgrimage destinations with a
small resident population who controlled sacred knowledge and a ceremonial
calendar. The number of Great Houses in the canyon is sometimes explained as
the consequence of competing or sequential power centers (e.g. Sebastian
1992). Steve Lekson (e.g. Lekson 1984;
Stein and Lekson 1992), among others, has argued that much of the Great House
architecture in the canyon is built for display and/ or ritual enactment. Trash mounds, including the large trash
mound at Pueblo Alto, are suggested to have been “platforms” for staging
rituals that included the purposeful destruction of ceramic vessels (Toll 2001;
but see Wills 2001 for a well-reasoned alternative interpretation that was part
of the inspiration for this exercise).
PUEBLO ALTO
ARCHAEOLOGY AND THE WESTERN TRASH MOUND
As mentioned
above, Pueblo Alto is considered a Great House. The architectural portion of
the site covers approximately 8 hectares. The trash mound is located at the southern end of the central
plaza and, although badly eroded, is approximately 60 meters long and 26 m
wide. Consistent with the pre-1990’s models of the Chacoan system, the western
trash mound was initially seen as a village dump for household debris that
accumulated during the occupational span of the pueblo, primarily AD 950 –
1200. Parts of pueblo, including some
of the trash mound, were excavated by the Chaco Project (NPS-UNM) in 1976 –
1978. The goals of the excavation
included determining whether Pueblo Alto was an administrative center involved
in redistribution of agricultural products to small communities. The convergence of several road segments at
Pueblo Alto made this hypothesis seem plausible. Unfortunately, evidence to
support the hypothesis of redistribution was not uncovered.
Currently, based
on the analyses following the excavation, Pueblo Alto is interpreted as a
ritual center and, in part because it is the only Great House excavated using
modern methods, this interpretation is extended to other Chacoan Great
Houses. The trash mound is suggested to
have been a constructed platform where, at least during one interval, imported
ceramics were deliberately broken and deposited. As Toll (1985: 196) states: “Translating the estimates into
identifiable imports during the Gallup portion of the Trash mound, at least
49,270 trachyte tempered pots were brought in and deposited [in addition to]
31,310 gray ware jars… 3,834 chalcedonic sandstone graywares… [and] 1,738 San
Juan igneous tempered whitewares.“ The
estimated number of vessels was considered too great to be accounted for by
simple domestic use, leading to the inference that the ceramics were deposited
in a ritual context. This
interpretation was first developed and discussed in two sources:
Toll, H. Wolcott, 1985, Pottery Production, Public Architecture, and
the Chaco Anasazi System. Ph. D.
Dissertation. University of Colorado
and
Toll, H. Wolcott and McKenna, Peter
J., 1987, The Ceramography of Pueblo Alto, in Investigations at the Pueblo Alto Complex, Chaco Canyon, New Mexico
1976-1979. Vol III. edited by Mathien, F. M. and T. C. Windes, National
Park Service, Department of Interior.
Santa Fe.
TRASH MOUND
EXCAVATION
The trash mound
excavation had three components. First,
the investigators dug a long, narrow trench running from northwest to
southeast, separated into 11 grid units of 2 x 2 meters each. A total of 20
levels are listed on their map, and each level was 20 cm deep. Because of slumping
in the mound, exact provenience became a problem. To resolve this problem, the investigators excavated a series of
short stratigraphic columns (termed “booths”) from the southwest wall of the
test trench. 62 of the layers in the
booths duplicated those in the mound.
Third, there were three backhoe tests, although these were not analyzed
in detail and they are hardly mentioned in the reports.
Decorated
ceramics, cross-dated to tree-ring dates, were used to identify temporal
periods in the mounds, but it is important to note that these periods
(deposits) were not really discrete and stratigraphically bounded. The mound was not a simple layer cake. There
was evidence of slumping and various types of mixing (especially bioturbation)
and identifiable fingerlings of later strata projected into the lower
strata. Later sherd types were present
in the earlier unit, albeit in lower frequencies, and each excavated level was
assigned to one or the other temporal/ depositional unit on the basis of the
majority of the ceramics present. The
earlier stratigraphic unit was identified as the “Red Mesa Mound” after the Red
Mesa ceramic series. The date range of
this unit is AD 970-1020. 18.9 cubic
meters of this deposit were excavated.
The second depositional unit is called the “Gallup Trash Mound” after
the Gallup black on white ceramic type.
The date range of this unit is A D 1020-1120, and 35.5 cubic meters were
excavated. Based on Tom Windes’
estimates of the total volume of the mound and on the assumption that the
proportions of each depositional unit in the trench were equivalent to their
proportions in the entire mound, the investigators estimated that 2.2% of each
depositional unit was excavated.
Therefore, the sample size of each stratigraphic/ cultural unit was
2.2%.
TABLED DATA
Tables 1 and 2
present the data that led the excavators to the interpretation that ceramic
consumption at Pueblo Alto was extraordinary. Table 1 presents the counts of
ceramics, volumetric measurements and estimations and projected consumption of
ceramic vessels at Pueblo Alto. In
order to arrive at vessel counts, the excavators assumed (without clearly
stating their assumption) that each rim sherd that could not be refitted with
another in the sample represented one vessel.
Table 2 shows estimates of per annum pottery consumption at Pueblo Alto
and other settlements. Estimates of
number of families (and therefore consumption rate of ceramics) are often based on extrapolation from ethnography
and archaeological data including room sizes, types of rooms and the presence
of certain features, especially hearths, in some rooms. In this case, the investigators simply
counted the number of rooms facing the plaza.
Note that a smaller proportion of the midden was excavated at Pueblo
Alto than at other sites and that the projection of total number of ceramic
vessels consumed is much larger, especially in the Gallup Trash Mound. Again, the estimated consumption rates are
the basis of the inference of ritual breaking and deposition of ceramic
vessels.
As Toll
recognizes, the final inference of ritual destruction of ceramic vessels at
Pueblo Alto required a series of assumptions, inferences, calculations and
estimates. He states, “Estimating the
ceramic population at a site and arriving at per annum and per family use rates
is a procedure fraught with guessing, assuming, fudging, and leaping at any
site; at Alto the degree of these gerunds is larger than usual. Two major reasons lie at the root of this
exaggeration: the small relative amount of the site dug, and the greater than
usual uncertainty as to the use and resident population of the site through
time.”(1985: 176 – 177) The
(simplified) outlines of their argument can be restated as follows:
1. Assume that each rim sherd represents 1
vessel.
2. Assume that the vessels in the mound were
used and broken at Pueblo Alto.
3. Estimate the total size of the mound.
4. Estimate the total size of each
temporal-stratigraphic unit based on the assumption that the proportion of each
in the excavated area is the same as the proportion in the unexcavated portions
of the mound.
5. Assume that the density of sherds (#/
cubic meter) is constant throughout the mound.
6. Extrapolate from the sample of rim sherds
to the number of vessels in each temporal-stratigraphic unit.
7. Estimate the number of years represented
by each temporal-stratigraphic unit based on the dominant ceramic types and
calculate number of vessels deposited per year.
8. Create an estimate of the number of
families who lived at Pueblo Alto.
9. Estimate the consumption rate of ceramic
vessels per family per year.
10. Infer that the excess of ceramic
consumption above what would be expected is due to ritual.
ACTIVITIES
Before leaving
the lab, study the tables and the information we have provided until you understand
how Toll moved from potsherds in a trash mound to ritual killing of
vessels. If anything is unclear, ask questions. Chances are it is unclear to others as well.
WRITE UP
This exercise is
not particularly well suited to the standard lab report format. Instead, the assignment consists of two
parts. The first is a series of
questions and brief exercises intended to get you thinking about the
reliability, empirical validity and abstract validity of the unit structures
created and described by the Chaco Project participants. Write a concise answer (not more than half
of a page, with the possible exception of question 11) to each question. For the second part of the assignment, write
a short paper (~5 pages) about the validity and reliability of the measurements
and units used in the investigation at Pueblo Alto. Try to create a logical flow of ideas in your paper and
incorporate and synthesize the issues raised by the questions. (There will naturally be some redundancy
between the two parts of the assignment.)
Ultimately, the point is not to trash the Chaco Project but to carefully
evaluate the research of others so that, when you build your own research
projects, you will be able to think clearly about the reliability and validity
of the unit structures you create and the measurements you perform.
QUESTIONS:
There are three
variables that form the basis of estimates of rates of ceramic consumption:
pots, people and time. The estimates of
all three are necessarily based on assumptions regarding the archaeological
record, sampling and the people (“cultural system”) that created the
record. In answering the following
questions, you will identify some of these assumptions. Then you will substitute other plausible
assumptions and use them to derive different estimates for each variable. This is one way of evaluating the
reliability and validity of the unit structures employed at Alto and it should
clarify the logic behind the inferences and conclusions that were published by
the excavators.
sherds and
vessels
1. What synthetic unit does Toll use
when he is building his arguments about rates of consumption? What does he actually measure to create his
synthetic unit (what is his observational
unit)? Evaluate the linkage
connecting the two in terms of precision, accuracy, empirical validity, and/or
conceptual validity.
2. Toll and McKenna (1987: 37 – 39, tables
p.43, 44) note that there is a difference in the proportions of graywares to
decorated wares between sites and (at Alto) when measured by total sherd counts
versus estimated numbers of vessels represented. They attribute between- site differences to differential
post-occupational disturbance. At Alto,
they note the confounding effects of the greater durability of trachyte
tempered vessels (52% of the graywares and 12% of the decorated wares) causing
the graywares to persist as larger sherds (fewer sherds/ vessel) coupled with
the larger average size of grayware vessels (more sherds/ vessel). It is a convoluted argument. The graywares in the Alto sample are 99.8%
jars and the decorated wares are 63% bowls, 16% jars and 21% other forms. How could vessel form affect their estimates of number of vessels present, given
their methods? Given the different
distributions of forms, how would this affect the ratio of graywares to
decorated wares in bulk sherd counts vs. vessel estimates? If the “target” is an estimate of the number
of vessels, are their methods precise and accurate, given the effects of vessel
shape?
3. Toll and McKenna calculate that their
numbers imply an average of 16.8 sherds/ vessel (1987: 206). How could you empirically evaluate whether
this is a reasonable number of sherds per vessel for this sample? In his excavations at Mug House (a roughly
contemporary site in the Chaco area with a similar estimated population size)
Rohn estimated 138 sherds per corrugated jar and 63 sherds per whiteware
(decorated) vessel (cited in Toll 1985: 187 - 188). Although there is no particular reason to believe Rohn’s
estimates have any bearing on Pueblo Alto, use the tabled data to calculate the
number of grayware and whiteware vessels present using his estimates.
4. Looking at the problems you identified in
questions 1 - 3, are counts of potsherds a reliable (accurate and precise) way
to estimate the number of vessels in an archaeological deposit? Is there
another unit of measurement that would avoid some of the problems of
comparability between sites, differential breakage of ceramic types, different
shapes of ceramic vessels etc? What is
the unit, and how might you use it to approach the “target” of estimating
number of ceramic vessels?
samples and
populations
5. In order to extrapolate from the
excavated sample to the total number of vessels deposited in the mound, the
investigators performed calculations that used an estimate that 2.2% of the
mound was excavated. Toll (1985: 187 –
189) and Toll and McKenna (1987: 206-207) assert that this estimate is
problematic, and throw out a guess that as much as 10% of the mound might have
been excavated. Using the information
in the tables (and accepting for now all of the other assumptions) how many
vessels were deposited if 10% of the mound was excavated? What is the difference between the estimates
of vessels present based on different estimates of how much of the mound was
excavated?
6. In passing, Toll and McKenna note “the
low ceramic content of the backhoe cuts” (1987: 39) in the trash mound. The sample of sherds recovered from the
trench (and not from the backhoe
cuts) was used to estimate the total number of vessels in the mound. What does Toll implicitly assume regarding
the representativeness of the two different samples? What does this suggest about the reliability (both precision and
accuracy) of his estimates?
7. Toll and McKenna state that, “the numbers
in the estimates generated in this way are frankly alarming – it seems almost
inconceivable that so many pots were disposed of at this one site.” (1987:
206) Given your answers to previous
questions, should we be “alarmed” by the number of pots or is there something
else to be concerned with? Briefly
discuss the reliability and validity of their methods for estimating the number
of vessels present in the mound.
derived
estimates
8.
To estimate
consumption rates, the investigators had to use estimates of the length of time
represented by each deposit. They
recognize (Toll and McKenna, 1987: 117) that these estimates may be
flawed. Returning to the data in the
tables, how do the estimated consumption rates change if each of the two time
periods is longer by a decade? By fifty
years? By a century?
9. The third estimated value (in addition to
number of vessels and time) used to derive rates of consumption (vessels per
year per family) is the number of people that inhabited Pueblo Alto. Based on the number of rooms facing the
plaza, the investigators estimated 20 family units. A common pattern in ancestral Puebloan sites is for each family
to occupy suites of 2 to 4 rooms (one for living, and another one to three for
storage). Given 133 rooms at Alto,
calculate the number of family units present if each occupied 4 rooms and if
each occupied 2 rooms. Another approach
used by some archaeologists to estimate population size is to apply an
ethnographically derived mean value of approximately 10 square meters of roofed
area per person (Narroll 1962). Using
the map of Alto, measure (don’t be too
concerned with the accuracy of your measurement, but do tell me what it is)
the roofed area (excluding kivas) and, assuming 5 people per family, estimate
the number of families that lived there.
What new estimates of number of vessels per family per year can you
derive with the new estimates of number of families?
10. In this question, you will combine the
effects of beginning from different assumptions than those of the
excavators. Basically, you will derive
a new estimate of ceramic consumption that represents a “lowest plausible
value” for comparison with the estimates derived by Toll and McKenna. Take the number of vessels you calculated in
question 3, then assume that 10% (instead of 2.2%) of the mound had been
excavated like you did in question 5.
What is the estimated number of vessels in the mound using these
assumptions? Now assume that each time
period was 50 years longer than stated by the investigators, and add in your
highest population estimate for the site.
What is the use rate of vessels per family per year that you estimate
using these assumptions? How does it
compare to ceramic consumption rates at other sites shown on the table?
ceramics and
ritual
11. The ultimate conclusion of the Pueblo
Alto excavation was that there was a very high rate of ceramic deposition there
due to ritual gatherings at the site.
Given your previous answers, is this a secure inference? Were the excavations of the trash mound at
Alto and the calculations of ceramic use rates a valid means of measuring ritual?
Were they designed to be?
Briefly trace the line of reasoning from potsherds to interpreting Chaco
as a “Location of High Devotional Expression” (Renfrew 2001). Using the concept of abstract validity,
suggest where there are major weak points in the argument.
extra credit
12. You are (miraculously) given permission
to conduct excavations at a Great House in Chaco Canyon to investigate whether
it was, in fact, a pilgrimage destination and ritual center. What would you measure, how would you
quantify it, and how is it related to what you want to know?
REFERENCES:
Narroll, R.
1962
Floor Area
and Settlement Population. American Antiquity 27: 587 – 589.
Lekson, Stephen
1984
The
Standing Architecture of Chaco Canyon and the Interpretation of Regional
Organization. In Recent Research on Chaco Prehistory, edited by W.J. Judge and J.D.
Schelberg, pp. 55 – 73. Reports of the
Chaco Center No. 8, National Park Service, Albuquerque.
Renfrew, Colin
2001
Production
and Consumption in a Sacred Economy: The Material Correlates of High Devotional
Expression at Chaco Canyon. American Antiquity 66 (1): 14 – 25.
Sebastian, Lynne
1992
The Chaco Anasazi: Sociopolitical
Evolution in the Prehistoric Southwest. Cambridge
University Press, Cambridge.
Stein, J. and S.
Lekson
1992
Anasazi
Ritual Landscapes, in Anasazi Regional
Organization, edited by D. Doyel.
Maxwell Museum of Anthropology Papers No. 5, Albuquerque. p. 87 – 100.
Toll, H. Wolcott
(see references
in text above)
2001
Making and
Breaking Pots in the Chaco World. American Antiquity 66(1): 56 – 78.
Wills, W.H.
2001
Ritual and
Mound Formation during the Bonito Phase in Chaco Canyon. American
Antiquity 66(3): 433 – 452.