Citation
Informing prehistoric foragers' perspectives

Material Information

Title:
Informing prehistoric foragers' perspectives why create a bedrock mine in a material rich environment?
Creator:
Hauser, Neil R
Publication Date:
Language:
English
Physical Description:
xi, 159 leaves : illustrations ; 28 cm

Subjects

Subjects / Keywords:
Quarries and quarrying, Prehistoric -- Colorado -- Uncompahgre Plateau ( lcsh )
Antiquities ( fast )
Quarries and quarrying, Prehistoric ( fast )
Antiquities -- Uncompahgre Plateau (Colo.) ( lcsh )
Colorado -- Uncompahgre Plateau ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 151-159).
General Note:
Department of Anthropology
Statement of Responsibility:
by Neil R. Hauser.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
268661643 ( OCLC )
ocn268661643
Classification:
LD1193.L43 2008m H38 ( lcc )

Full Text
INFORMING PREHISTORIC FORAGERS PERSPECTIVES:
WHY CREATE A BEDROCK MINE IN A MATERIAL RICH ENVIRONMENT?
by
Neil R. Hauser
B.A., University of Denver, Denver, 1974
M.A., University of Colorado, Boulder, 1977
A thesis submitted to the
University of Colorado Denver
in partial fulfillment
of the requirements for the degree of
Masters of Arts (M.A.)
Department of Anthropology
2008


This thesis for Masters of Arts
degree by
Neil R. Hauser
has been approved
by
Tammy Stone
Chris Beekman
Date
Steve Holen


Hauser, Neil R. (M.A. Anthropology)
Informing Prehistoric Foragers Perspectives: Why Create a Bedrock Mine in a
Material Rich Environment?
Thesis directed by Associate Professor Tammy Stone
ABSTRACT
Prehistoric occupants of the eastern slope of the Uncompahgre Plateau in
west-central Colorado mined lithic material from a bedrock mine prior to A.D 66
while also utilizing ubiquitous local secondary, surface sources of comparable
material. The utilization and mobility of lithic materials from the mine and nearby
secondary sources were investigated to inform whether material from a bedrock mine
was perceived by the prehistoric occupants of the area as: 1) equivalent to material
from the secondary sources in cost and benefit, 2) higher quality material than the
secondary sources, 3) had special value such as for use in an important function,
trade, or ritual, or 4) if creation of the bedrock mine was the result of limited access
to other lithic sources. The results of this research indicate that material from the
bedrock mine was more highly reduced near the quarry than material from the
secondary sources and was curated for long distance travel and later use; the cost of
mining the bedrock material may have been perceived to be equal to that of
procuring usable material from the surface sources; and more Flint Cave material
was taken further from the bedrock mine than material from the proximal secondary
source. These findings indicate that the material from the bedrock mine may have
been perceived by prehistoric people as more valuable in the seasonal rounds than
material from the proximal secondary source and of equal or lesser cost to procure.
These results may inform why bedrock mines are not uncommon in areas with other
available surface sources. The mobility of the lithic materials studied in the research
area also reflected segments of the seasonal rounds of groups of prehistoric people
utilizing the eastern slope of the Uncompahgre Plateau, and further work may inform
more details of the landscape use of this area by these peoples.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
Tammy Stone


DEDICATION
This thesis is dedicated to the person most responsible for my interest in
archaeology, my dad, Ray E. Hauser. Over 50 years ago he introduced me to the joys
of time in the field discovering the path of the ancients and the intellectual challenge
of trying to understand their perspectives and ways of life. I also include our
neighbor the late Carlyle Squint Moore, a lifelong student and teacher of
archaeology, in this dedication. He shared with my dad and me his knowledge of the
diverse archaeology and evidence of prehistoric man in Western Colorado and first
introduced me to the Flint Cave.


ACKNOWLEDGMENT
First Id like to acknowledge the understanding and tolerance of my wife,
Teri, for the many trips to the field and hours analyzing data that lead to this thesis. I
would also like to thank ATSAA for allowing me the use of their LIBS system; the
members of the Chipeta Chapter of the Colorado Archaeology Society of Montrose
for their help on surveys and their interest and encouragement; Kim Gerhardt for
going to the field with me and teaching me some geology; Susan Thomas for her
help in my repeated trips to examine curated collections at the Anasazi Heritage
Center; the BLM archaeologists, Julie Coleman and Dr. Glade Hadden, of the
Montrose District for guidance and discussions during the fieldwork done on the
BLM lands, and Leigh Ann Hunt, the USFS archaeologist for the Grand Mesa,
Uncompahgre, and Gunnison National Forests, who allowed me to conduct field
work on USFS lands for this research. I would also like to thank Alan Reed and Dr.
Steve Baker, professional archaeologists in the Montrose area for their support.
Finally, I would like to especially thank the members of my thesis committee, Dr.
Tammy Stone, Dr. Chris Beekman, and Dr. Steve Holen for their support,
discussions, and guidance in this research.


TABLE OF CONTENTS
Figures........................................................................ix
Tables.........................................................................xi
Chapters
1. Introduction.............................................................1
2. Theory...................................................................5
Approach..............................................................7
Expectations from Previous Research...................................8
Factors Affecting Perspective........................................13
Quality of Material..............................................13
Relative Cost of Procurement.....................................18
Special Value....................................................20
Limited Access...................................................25
Synthesis............................................................30
Research Limitations.................................................40
3. Background..............................................................43
Environment..........................................................43
Geology..............................................................46
vt


Cultural History
47
Past Archaeological Work....................................................51
Flint Cave Physical Characteristics....................................53
Flint Cave Chronology..................................................57
Flint Cave Solar Alignment:............................................59
4. Methodology.....................................................................61
Macroscopic comparison.................................................65
Ultraviolet Fluorescence...............................................67
Laser Induced Breakdown Spectroscopy (LIBS)............................67
Lithic Material Data........................................................71
Isolated Finds.........................................................76
Transect Sampling......................................................76
Survey.................................................................78
Excavated Sites........................................................78
Cost Models.................................................................80
Research Parameters.........................................................84
Utilization (RP1 and RP2)..............................................84
Supply Zone (RP3)......................................................84
Reduction (RP4)........................................................85
Distance Decay (RP5)...................................................87
Functional Form (RP6)..................................................88
5. Analysis and Discussion.........................................................90
Utilization Analysis (RP1 and RP2)..........................................90
Reduction Area Tools...................................................90
vii


Research Area Lithic Tools
96
Supply Zone Analysis (RP3)......................................101
Reduction Analysis (RP4)........................................Ill
Distance Analysis (RP5).........................................112
Functional Form Analysis (RP6).................................115
Temporal Analysis...............................................117
Confounding Factors.............................................122
6. Conclusions.........................................................130
Research Results................................................130
Other Observation...............................................136
Recommendations.................................................136
Appendix
A. Reduction Model....................................................139
B. Debitage Weight Estimation Model...................................144
C. Chi Square Statistics for RP1 and RP2..............................147
viii
Bibliography
151


LIST OF FIGURES
Figure
1.1 Location of Area of Interest within Colorado........................1
2.1 Notional Environment of Sources with Contours of Terrain Based Cost.26
3.1 Research Area on Uncompahgre Plateau................................44
3.2: Geological Formations Applicable to the Uncompahgre Plateau
(Gerhardt 2001).....................................................45
3.3 Openings to Flint Cave.............................................54
3.4 Plan View of Flint Cave............................................55
3.5 Interior of the Flint Cave (5MN7429)...............................56
4.1 Locations of Quarries within Research Area..........................63
4.2 LIBS UV Spectra of Flint Cave Material.............................69
4.3 Locations of Sample Areas within Research Area.....................72
4.4 Orientation of Sampled Transects in Reduction Area.................77
4.5 Cost Contours from Flint Cave Quarry (5MN7429).................... 81
4.6 Cost Contours for Quarry 5MN7753...................................81
4.7 Cost Contours for Quarry 5MN7754/55/56.............................82
4.8 Cost Contours for Quarry 5MN4638...................................82
4.9 Cost Contours for Quarry 5MN8384 and 5MN8385...................... 83
4.10 Cost Contours for Quarry 5MN8386..................................83
5.1 Flake Size Distribution for Flint Cave Material in Reduction Area...94
5.2 Flake Size Distribution for 5MN7753 Material in Reduction Area...95
5.3 Expedient Tool of Flint Cave Material with Cortex..................95
5.4 Spatial Distribution of Projectile Points Found on this Project....99
5.5 Blade End Scraper of Flint Cave Material...........................100
5.6 Flint Cave Material Contribution in Lithic Samples Relative to Cost
Contours and Distance Rings.........................................103
5.7 5MN7753 Material Contributions in Lithic Samples Relative to Cost
Contours and Distance Rings.........................................104
5.8 5MN7755 Material Contribution in Lithic Samples Relative to Cost
Contours and Distance Rings.........................................105
5.9 5MN7756 Material Contribution in Lithic Samples Relative to Cost
Contours and Distance Rings.........................................106
5.10 Percent Contributions of Materials by Weight from Flint Cave and
5MN7753 as a Function of Cost Contour from the Quarries.............108
IX


5.11 Percent Contribution of Materials by Weight from 5MN7755 and
5MN7756 as a Function of Cost Contour from the Quarries...........108
5.12 Average Contribution of Materials by Weight from Flint Cave and
5MN7753 as a Function of Distance from the Quarries...............110
5.13 Average Contribution of Materials by Weight from 5MN7755 and
5MN7756 as a Function of Distance from the Quarries...............110
5.14 Percent Flake Count Contribution of Materials from Flint Cave and
5MN7753 as a Function of Cost Contour from the Quarries...........113
5.15 Percent Flake Count Contribution of Materials from Flint Cave and
5MN7753 as a Function of Distance from the Quarries...............114
5.16 Ratio of Flake Count and Weight of Materials from Flint Cave and
5MN7753 as a Function of Cost Contour.............................114
5.17 Ratio of Flake Count and Weight of Materials from Flint Cave and
5MN7753 as a Function of Distance.................................115
5.18 Rosegate Point from 5MN34........................................119
5.19 Cottonwood Triangular Point......................................119
5.20 McKean Point from 5MN2628....................................... 120
5.21 Gypsum Point.....................................................121
5.22 Duncan Point Base................................................121
5.23 Number of Carbon Dates as a Function of Age for Northern Colorado
River Basin (Reed and Metcalf 1999)............................... 126
A1 Model Derived from Fit to Simulated Data...........................143
B1 Fit to Average Flake Weights vs Circumscribing Circle Diameter.....145
Cl Chi Square Analysis of Formal and Informal Tool Composition........148
C2 Chi Square for Informal Cores and Expedient Tool Composition.......149
(Expected Values from Debitage)....................................149
C3 Chi Square for Informal Cores and Expedient Tool Composition.......149
(Expected Values Assuming Equal Access)............................149
C4 Chi Square for Projectile Points...................................150
x


LIST OF TABLES
Table
2.1: Summary of Andrefskys (1994) lithic material quality vs. use results...15
2.2 Test parameters with possible values for each............................36
2.3: Scenarios that inform Hypothesis A................................37
2.4: Scenarios that inform Hypothesis B................................38
2.5: Scenarios that inform Hypothesis C................................39
2.6: Scenarios that inform Hypothesis D................................40
3.1 Dates from Flint Cave................................................58
4.1 Test parameters......................................................61
4.2: Thesis hypotheses...................................................62
5.1: Material composition of transects in the Flint Cave reduction area..91
5.2: Material composition of tools found in Flint Cave reduction area....93
5.3: Material composition of formal tools found in samples of lithic scatter.97
5.4: Material composition of formal tools in isolated finds..............98
5.5: Nominal cost contour (nCAC) and distance for supply zone...........109
5.6: R2 for functional relationships.........................................116
6.1: Summary of findings of this research...............................131
Al. Original flint knapping assemblage data.............................141
A2: Ratio of small to large debitage....................................142
A3: Simulated reduction data and model results..........................142
xi


CHAPTER 1: INTRODUCTION
The eastern slope of the Uncompahgre Plateau in west-central Colorado
(Figure 1.1) has numerous sources of fine grained silicified sandstone that were
heavily exploited by prehistoric peoples over at least the last 9,500 years. This is
evident from the numerous lithic scatter sites at the lower elevations of the
Uncompahgre Plateau. Some of these sites are large and have dense areas of debitage
visible on the surface. The density of debitage exceeds 700 flakes /m in some, and
2
sites with a density of 50 flakes/m are common.
753000IT1E. 431000mE. 565000mE, 753000mE,
WGS84 Zone 12S 767E?9 0
100
200
300 mies
TN
0
100
200
300
400 km
Figure 1.1 Location of Area of Interest within Colorado


Although bedrock mining is considered a high cost activity and is usually
assumed to be more costly than obtaining material from secondary surface deposits
of material, prehistoric bedrock mines have been identified both regionally and
worldwide (Bamforth 2006; Elston and Raven 1992; Gramly 1980; Holen 1991;
McBryde 1984; Reher 1991). Often those mines were associated with nearby
outcrops of similar lithic material occurring in exposed veins or secondary
deposition. This is also the case in this research area.
Throughout the research area exposures of lithic material were utilized by
prehistoric peoples. One quarry site (5MN7429) is unique among the many sources
of lithic material in this area. Locally this quarry is called the Flint Cave because it
is a bedrock mine that forms a man-made cave with an estimated volume of about 25
j
m Within 2 km of this mine are at least five other utilized quarry sites. The closest
is less than 400 meters from the bedrock mine. With the exception of the Flint Cave,
all the other quarries in the area are secondary deposits of material resulting from the
erosion of the original lenses and veins of material in the canyon walls.
An outstanding question in lithic economy and organization of technology is
why apparently high cost mining for lithic material occurred despite the availability
of more easily accessible lithic material of comparable quality. The lithic
environment on the eastern flanks of the Uncompahgre Plateau provided this
research an opportunity to compare and contrast the utilization and mobility of
materials from a bedrock mine and nearby secondary surface sources of comparable
2


quality materials to inform the cost and value of these lithic materials from the
perspective of the prehistoric forages.
The next chapter of this work defines four hypotheses, two economic and two
social, which might inform the existence of a bedrock mine in an area of ubiquitous
surface sources and reviews the middle level theories of lithic technology and raw
material economy. These middle level theories are then utilized to develop six
research parameters that inform the hypotheses. The research parameters are: use of
material from the bedrock mine (RP1), use of material from the proximal surface
source (RP2), distance of the supply zone of the bedrock mine compared to the
surface sources (RP3), level of reduction (RP4), distance material moved from the
sources (RP5), and the functional form of material as a function of distance (RP6).
The third chapter reviews the geologic and archaeological context of the
research area. In particular Chapter 3 describes the bedrock mine (Flint Cave,
5MN7429) and informs the chronology determined by l4C and optical stimulated
luminescence performed on samples taken from the fill of the bedrock mine.
The methods used to inform the research parameters by sampling lithic
scatter areas throughout the research area, accomplishing tool survey and sampling
along transects in the reduction area around the Flint Cave and quarry 5MN7753,
and use of isolated finds and material from excavated sites are described in Chapter
4. This chapter also describes how the debitage from the samples was quantified for
comparison as well as the methods used to source all the lithic materials. Cost
3


models based on terrain slope and distances from the quarries were generated for
each quarry and are illustrated in this chapter. The cost models and straight line
distances only were both used to evaluate the mobility of the lithic materials from the
quarries as a function of both percent flake count and weight.
Chapter 5 presents the analysis of the utilization and mobility of the lithic
materials. These analyses were accomplished as a function of cost and distance from
the Flint Cave, 5MN7753, 5MN7755, and 5MN7756 quarries, using percent weight
and flake count of the materials, material composition of tools, and isolated finds.
The results from for each quarry were compared and contrasted.
Chapter 6 presents the consolidated results of the analysis and provides the
conclusions of this research. The results of the analysis indicate that material from
the bedrock mine was curated for long distance travel more than the material from
the proximal secondary, surface sources. From this it might be inferred that the
material from the bedrock mine was more valuable for use on the seasonal round
than the materials from the proximal secondary, surface sources. The research results
also indicate that procurement of the material from the bedrock mine was perceived
as equal or less than the cost of prospecting and reducing material from the
secondary, surface sites. Chapter 6 also provides recommendations for future work
that could clarify findings and observations of this research and further inform the
behavior of the prehistoric peoples who utilized the eastern slope of the
Uncompahgre Plateau in west central Colorado.
4


CHAPTER 2: THEORY
Traditionally site assemblages have been used to inform organization of
technology and mobility. It has been argued to what extent economic decisions
influenced the choice and utilization of lithic procurement by prehistoric peoples. In
many cases economic decisions that may have been made by foragers based on
material quality, functional needs, and procurement costs are obscured by the lack of
choice of materials in a region.
In the area chosen for this research, located on the eastern flanks of the
Uncompahgre Plateau in west central Colorado, there are ubiquitous sources of
available materials with a wide range of quality and procurement costs. The
materials include good to moderate quality silcrete (silicified sandstone), chalcedony,
medium to coarse grain quartzite, and some andesite, basalt, and rhyolite. In this
lithic material environment choices based on tradeoffs between material quality,
functional needs, procurement costs, and tool production costs are expected to be
discernible in the archaeological record.
This research focuses on a bedrock mine quarried by prehistoric peoples in
close proximity to many utilized, secondary surface sources of materials of
seemingly similar, if not equal quality. This research investigates the possible
economic and social factors that may have led to the development of this seemingly
5


high cost of procurement bedrock mine in an area where apparently lower cost of
procurement sources of material are common. The author posits four hypotheses, two
economic and two social factors, which could have lead to that behavior. This
research tests these hypotheses against evidence obtained from the archaeological
record of the study area to determine if any of these hypotheses may explain the
reason for the bedrock mine in this lithic rich environment.
From an economic and cost/benefit framework the use of the bedrock
material must provide either a cost/benefit advantage or at least be equal to that of
the material from the secondary surface sources. To achieve that it is posited that
either:
The material from the bedrock mine was perceived as higher quality than
materials from the other surface sources.
Or,
The total cost of procuring material from the bedrock mine was perceived to
be less than or equal to the cost of obtaining material from the secondary
surface sources.
It is posited that social factors could also have resulted in either an increased
value of material from the bedrock mine or socially created shortage of lithic
material. Either could have favored the development of the bedrock mine. The
hypotheses developed from a social factors perspective are:
The bedrock mine material was associated with high value functions such as
trade or ritual use. The importance may or may not have been associated with
material quality.
6


Or,
There was a shortage of available material due to limited access or high cost
to other quarries. This limited access could have been due to group ownership
of territories or individual quarries along traditional seasonal foraging routes.
To test these hypotheses the cost and benefits of the bedrock mine and other
nearby lithic material sources, as perceived by the prehistoric tool manufacturers and
users, as well as evidence of special value, territories, or quarry ownership must be
determined from the archeological record. This research synthesizes a model based
on findings from previous work in lithic technology and organization and territorial
behavior. The model is then tested using the evidence provided by the archaeological
record from the research area to inform what factors may have lead to the
development of a bedrock mine in an area of ubiquitous secondary surface sources.
Approach
In this research an economic, cost/benefit, framework based on optimal
foraging theory (Smith 1983; Ugan, et al. 2003; Winterhalder and Smith 1981) is
used to investigate the development of the bedrock mine in an environment of
ubiquitous surface sources of lithic materials. The currency of cost is assumed to be
energy expended in procurement, transport, reduction of the material to tools, and
maintenance of tools. The currency of benefits are assumed to be such metrics as the
number of tools per unit weight of material, tool efficacy, time between maintenance,
and possibly more ephemeral benefits such as status or maintenance of social
relationships. Although these costs and benefits cannot be quantitatively determined
7


from the archaeological record, relative measures between different materials and
quarries may be informed by observables in the archaeological record. Mid-range
theories from previous work on lithic technology and organization provide a means
of relating these costs and benefits to those expected observables in the
archaeological record. Using these relationships a model was developed to inform
the perceived relative cost /benefit of procuring materials from the bedrock mine and
other nearby surface sources by predicting the cost/benefit dependent observable that
would be expected to occur in the archaeological record.
Expectations from Previous Research
A model for lithic material procurement by foragers was developed by
Binford (1979) from ethnoarchaeological work with the Nunamiut. This model posits
that lithic material procurement was an essentially no-cost activity embedded within
seasonal rounds and was determined by the strategies for obtaining food rather than
any consideration for obtaining lithic material. Although Binford (1979:267) states
that personal gear or even household gear is much more likely to be manufactured
according to quality considerations unaffected by constraints on time or immediate
availability of appropriate material, he holds that the range of quality was
constrained by that available from normal seasonal rounds.
In contrast to Binfords model of low or no cost lithic procurement, Gould
(1978) found that costs were incurred in procurement of lithic material by
aboriginals in the Western Desert of Australia. Special trips to particular quarries
8


were planned and accomplished to obtain particular material. In these cases, costs
were incurred both in the quarrying and transport of the material from sites which
could be up to a days walk away from the seasonal round (Gould 1978). Aboriginal
Australians were forced into these occasional longer forays from their residential
sites during their seasonal rounds due to limited distribution of critical resources
such as water near important or high quality sources of lithic materials (Gould 1978).
In the case of the Nunamiut, Binford does not discuss the quality of materials
available along the seasonal round compared to materials available further from the
seasonal round, for instance within a days walk. Lithic material could also have
been naturally distributed in the region in such a way as to not be a major
consideration. It may well be that the range of materials along the seasonal round
provided a good representation of the material quality available at greater distances
within the region, and the seasonal round had been established prehistorically based
on trades between availability of the needed resources such as water, food, and lithic
material. Like Close (1996), Binford (1979) found that lithic material was carried to
resource areas along the seasonal round that lacked lithic material for future use and
risk reduction. Although transporting raw or partially reduced lithic material may
have been low cost for the Nunamiut at the time of Binfords work as he claims, due
to sleds and mechanized transportation, for prehistoric people there was a cost
associated with transporting the additional weight of raw lithic materials, but this
was apparently balanced by the benefit of risk reduction in the future.
9


To help resolve these contrasting observations by Binford and Gould,
Bamforth (2006) notes that at the time of Binfords work the Nunamiut were using
lithic material as a tertiary back-up to their metal and store bought tools but stone
tools were important components of Aboriginal technology; acquiring raw material
thus mattered to important aspects of peoples lives (Bamforth 2006) during the
time Gould did his work. It may be that since the tools made of lithic material were
backups for the Nunamiut, lithic materials were procured as the materials with the
desired characteristics and costs were available along the seasonal rounds (Bamforth
2006), but that may not have been the case for foragers that depended on tools made
of lithic material such as the aboriginals of Australia studied by Gould or prehistoric
foragers in the case of this research problem. This appears to be a good example of
the pitfalls that can occur when using ethnographic analogy of post contact peoples
in developing theories for the behavior of prehistoric peoples (Bailey 1983).
There is general agreement that lithic procurement occurred as part of the
seasonal rounds of foragers as noted by Binford (1979) and Gould (1978), and that
embedded procurement would reduce the cost of procurement by reducing the cost
of transport. From a cost-benefit framework, it is been shown (Holen 1991) that
obtaining sufficient food and water were the dominant parameters that determined
the path of the seasonal rounds. Both water and food have a much higher use rate
than lithic material and would result in much higher expenditures of energy if they
had to be carried long distances. For example, the average amount of water needed
10


by a person per day is about 3 liters or 3 kilograms, but a kilogram core of reduced
material could be expected to produce at least 100 grams (assuming 10% efficiency)
of tools which is several scrapers or 20-50 projectile points or pre-forms which might
last a person for several weeks. High quality lithic material with its lower
expenditure rate, which could be extended even further by expedient use of lower
quality material when available, would not be expected to dictate the location of
residential camps but could have influenced or been the goal of logistical forays. It
seems likely that seasonal rounds optimized to allow forays to obtain good material
in preparation for future use, as noted by Gould (1978), were the most economical
practice. Increased late stage biface reduction of high quality materials close to the
lithic material sources would also reduce the cost/benefit of transport. Late stage
reduction has been found (Beck, et al. 2002; Newman 1994) to be greater for
materials that were to be transported further.
Although embedded procurement reduced the cost of transport, evidence that
procurement of lithic materials by prehistoric peoples did result in large
expenditures of time and energy and may have even influenced the seasonal foraging
route has come from archaeological evidence of bedrock mining in quarries such as
Windy Ridge in Colorado (Bamforth 2006), Tosawihi in Nevada (Elston and Raven
1992), Spanish Diggings in Wyoming (Reher 1991), and Mt. Jasper in New York
(Gramly 1980). High cost bedrock quarrying activities occurred repeatedly at these
sites over long periods of time. Evidence of camps at the quarries indicates that the
11


activities were multiple days and some may have been residential sites (Bamforth
2006).
In this research, the lithic material sources of interest are close to water and
were likely embedded in seasonal rounds. However, due to terrain, such as canyons,
between sources and the associated cost of moving from one to another, some lithic
sources may have been part of different seasonal rounds or been visited during
different parts of the same seasonal round. The conclusion pertinent to this research
problem is that contrary to Binfords (1979) model of no or low cost lithic material
procurement, it seems that lithic material was procured by prehistoric foragers at
substantial cost. Whether it was strictly embedded or not is nearly impossible to
determine.
Why apparently high cost sources of lithic material such as bedrock mines
were created by prehistoric foragers even when accessible surface sources were
locally available is the focus of this research. In the case of this research, the
proximity of a utilized surface source of lithic material (5MN7753) to the Flint Cave
(5MN7429) provides an opportunity to compare and contrast the mobility and
utilization of materials from these two types of sources as well as from other local
surface sources. For this comparison to be a reflection of prehistoric peoples
perspectives of the cost and benefit of the Flint Cave and source 5MN7753, these
sources both must have been available contemporaneously. The discovery of an
Oshara projectile point dated 8000-5000 B.P. (Irwin-Williams 1973) at quarry
12


5MN7753 made of material from that source, and the fact material is still available
today at 5MN7753 supports the assumption that material was available at 5MN7753
during the period the Flint Cave was being mined (pre-1939 B.P., see Chapter 3
Background for discussion of Flint Cave chronology).
Factors Affecting Perspective
Although bedrock mining is a high cost method of obtaining lithic material,
there are several possible factors (or combinations of the factors) that could balance
the cost/benefit of the material from the bedrock mine with the cost/benefit of
obtaining material from surface sources. Those factors are:
a. The quality of material from the bedrock mine is much better than the
material from the surface sources.
b. The relative cost of obtaining good material is actually equal to or less for
the bedrock mine than from the surface sources.
c. The material from the bedrock mine has special value, i.e. trade, prestige,
or ritual.
d. Access to sufficient quantities of other material was limited due to social
circumscription or high cost.
Work over the last three decades by archaeologists investigating lithic
technology and organization has created mid-range theories that inform how these
factors might manifest themselves in the observables of the archaeological record.
Quality of Material
It has been posited (Goodman 1944) that specific material characteristics
would be associated with particular functions and associated types of tools. This is
13


supported by archaeological evidence (Andrefsky 1994; Beck and Jones 1990;
Findlow and Bolognese 1984; Gould and Saggers 1985; Jeske 1989; Marks, et al.
1991; Seeman 1994) that indicates material with particular characteristics do appear
to be selected for particular functions.
Defining the quality of material has been approached in several ways, from
physical parameters of rock mechanics such as Youngs modulus of elasticity,
uniaxial unconfined compressive strength, tensile strength and chevron notch
fracture toughness (Domanski and Webb 1992) to more subjective parameters such
as color, grain size, and workability (McElrath and Emerson 2000). Generally fine
grained materials and particularly high silica materials that flake well are classified
as high quality materials. Quality should not be confused with suitability, which is a
function of the intended use of the material. Durability of a blade to keep an edge is
important for scrapers, gravers, and drills but sharpness is more important for knives
and projectile points. The qualities of the material that are best suited for these
functions are quite different. For example, obsidian and high silica cryptocrystalline
materials can break to produce a sharp edge, but they are not durable, suffering
extreme edge fracturing when used as a scraper. Conversely basalt, rhyolite, and
medium grained silcrete and quartzite dont form as sharp an edge but have good
durability.
Informal or expedient tools have been associated with sedentary people and
residential sites, and formal tools have been associated with mobile peoples and their
14


logistical and field sites (Bamforth 1986; Kelly 1988; Parry and Kelly 1987).
However, Andrefsky (1994) found that abundance and quality of material affected
the type of tools more strongly than did mobility. Andrefsky found that in a raw
lithic material environment of ubiquitous good quality sources both informal and
formal tools occurred with nearly equal frequency in both sedentary and short-term
occupation sites. In both cases, the tools were made of the local material greater than
90% of the time. These results were in contrast to his finding that primarily informal
tools were made of the local material when poor quality but ubiquitous materials
were available. He also found that primarily formal tools were made when high
quality material was available but was scarce, and primarily expedient or informal
tools were made when only poor quality material was available but scarce. These
results (Table 2.1) indicate that the functions provided by both informal and formal
tools existed in both sedentary and mobile settings, and that quality and quantity of
available material dictated which types of tools were created.
Table 2.1: Summary of Andrefskys (1994) lithic material quality vs. use results
Good Quality Material Low Quality Material
Material Plentiful Formal and informal tools Informal tools
Material Scarce Formal tools Informal tools
All the lithic materials in Andrefskys (1998; 2007) work came from
secondary sources such as cobbles in stream beds and were likely of equal cost to
15


procure and reduce. Assuming the cost of procurement for the materials was
essentially the same, Andrefskys (1994) results can be interpreted in the cost-benefit
framework as maximizing the benefit obtainable from the available materials while
minimizing the costs. The use of poor quality material primarily for informal tools
extended the use-life of formal tools since they were not used when informal tools
would suffice. It also reduced cost if making formal tools from poor quality material
had lower finished tool production rates or poorer performance (benefit) than tools
made of higher quality materials. When high quality material was plentiful, informal,
expedient tools could have been made to reduce the cost of making formal tools for
functions when informal tools would suffice. However, apparently there were
functions when the performance of formal tools balanced or exceeded the costs of
manufacture, since formal tools were also made when high quality material was
available. Tomka (2001) and Ugan, et al. (2003) argue that the expectation of
lengthy and repetitive processes resulted in the making of formal tools rather than
using expedient tools.
Besides the association of formal and informal tools with material quality and
quantity, specific tool types have been found to be associated with specific materials
(Beck and Jones 1990; Goodman 1944; Gould and Saggers 1985; Marks, et al.
1991). Gould and Saggers (1985) found in the case of aboriginal Australians
specialized tools or functions justified increased procurement costs for specific
materials even when the material appeared no better than material that was more
16


easily procured. In that case the association of a particular material with trade was
what gave it value. In their work in the Great Basin, Beck and Jones (1990) found
that while projectile points were made of most materials, crescents and scrapers were
primarily made of chert rather than the local fine grained basalt. The implication is
that material quality, or possibly suitability is a better descriptor, as perceived by the
tool manufacturer and user depended on the tool or function for which the material
was chosen. Since particular materials were used for specific functions, any
additional cost of obtaining those materials must have been justified by the benefit of
using those materials for those purposes.
Although the archaeological evidence from previous work indicates that lithic
material quality is tied to tool function, none of the work found in the literature
provides sufficient information to explore the relationship between material quality
and suitability and the total costs composed of material procurement, transport,
reduction, curation and maintenance. For example, Andrefsky (1994) does not
provide any information on the types of formal tools or whether there were any
associations between tool types and specific high quality material. Since
procurement costs can be assumed to be the same, any differences would have been
due to the value of the tool function. Similarly, Findlow and Bolognese (1984)
found that procurement of several specific materials with different characteristics
were optimized by sedentary people in central New Mexico but did not address what
tools the materials were used for or relative cost of quarrying the materials.
17


In the case of this research project, if the bedrock material was considered
high quality material and likely to become part of the mobile toolkit, higher levels of
reduction would be expected near the bedrock mine. The same would be true of
material from the surface sources if they were considered high quality materials.
If the quality or suitability of the material from the bedrock mine was
perceived by the prehistoric miners as having been better than the material from the
surface sources, Andrefskys (1994) findings would predict that both expedient and
formal tools would have been made of the material from the bedrock mine. If the
quality of material from both sources was perceived to be equivalent then both
expedient and formal tools would be expected to have been made of both materials.
Relative Cost of Procurement
Potentially the largest contributor to the cost of tools, the cost of quarrying,
can be mitigated if there are choices. It is expected if there was more than a single
source of material of sufficient quality then the source with the lowest cost of
extraction of usable material would have been selected. Although at first
consideration, secondary surface sources of lithic material might seem to have
provided the lowest cost, time and energy expended in searching and testing material
to identify material of acceptable quality in secondary sources may have exceeded
that of bedrock mining. Bedrock mines may also have provided more dependable
and homogeneous material (Elston and Raven 1992; Ericson 1984). To obtain usable
material from secondary surface sources also usually requires additional time and
18


energy to remove cortex. Based on experimental work by Elston and Raven (1992),
if prehistoric people based their lithic procurement decisions on minimizing the
cost/benefit ratio for procuring some usable material in a short time, what Elston
calls efficiency, then encounter procurement, i.e. searching surface quarries, was
always the best strategy. However, Elston and Raven (1992) found quarries in
bedrock provided the highest net return rate, defined as the amount of material
acceptable for processing, produced in some unit of time when the return was
calculated over a day or more. This allowed time for removal of overburden (Elston
and Raven 1992).
Another consideration that may have provided a significant advantage to
bedrock mining over utilizing secondary surface sources is that weathered surface
material does not flake as well as material that is buried and contains more moisture
(Bamforth 2006; Binford and O'Connell 1984; Gould and Saggers 1985). The
desirability of non-weathered material by those using the material is supported by
Binford and O Connell (1984) who noted that Australian aborigines only procured
material they dug out rather than surface materials available at the quarry. A bedrock
mine would have provided reliable quantities of non-weathered material for a
predictable amount of energy expended, but exploring secondary surface sources
would have had an unpredictable expenditure of energy to obtain usable non-
weathered material.
19


If the perceived cost of obtaining material from the bedrock mine was equal
to that of the nearby surface sources, it is expected that the utilization of the material
would have been driven by the perceived quality of the materials as informed by
Andrefsky (1994) and discussed above. If procurement of the material from the
bedrock mine was perceived to have cost less than material from the surface sources,
it is expected that the material from the surface sources would not have been used
except for tools and then only when it was more suitable than the material from the
bedrock mine. Certainly it seems unlikely that the material perceived to be more
costly would have been used regularly for expedient tools.
Special Value
Both economic (technology) and social factors could have affected lithic
material utilization. Some lithic material may have had a higher value and thus
warranted a higher procurement cost than the quality of the material would indicate
if:
a. The material was uniquely, suited, within the local lithic
environment, for tools that were used for an important subsistence
activity, or
b. The material or objects made of the material were used in trade
and potentially considered a prestige item and therefore promoted
and helped maintain social connections with other groups, or
c. The material had ritual importance.
Although ritual is a common catchall when no other explanation seems to fit
the data, the possibility of ritual association must be considered, but with caution.
20


The association of lithic material to ritual importance may be hard to interpret from
the archaeological record when burial and associated mortuary goods are not part of
the record, such as the case in this research. Gould (1978) found that Australian
Aborigines obtained lesser quality material by direct procurement from a particular
source for use in axes even though more suitable material was available locally
because the acquired material was associated with a sacred place. Without knowing
the source of the material was a sacred place this might have been interpreted as
association of a particular function with a particular material, more often associated
with economic efficiency than with ritual.
If the material from the bedrock mine was particularly suited to an important
subsistence activity, there might be evidence in the archaeological record. It is
expected if the material from the bedrock mine was uniquely suited for use in tools
associated with an important subsistence activity, there would be a statistically
significant correlation of that material with that particular tool type when found in
the archaeological record. If that activity occurred in areas of the seasonal round
distant from the bedrock mine, it is expected that tools made of the bedrock material
would have been transported as finished tools to the area where the activity occurred.
In this case, the archaeological record should show a high level of reduction near the
bedrock mine (Beck, et al. 2002) and an absence of debitage from the material
except in the area where the tools were used. In that area it is expected that both
flakes from retouch and exhausted and broken tools will be found.
21


Determining if the material from the bedrock mine was traded or was
considered a prestige item is harder to infer from the archaeological record. If the
material was intended to be traded and was a prestige item it might be expected that
the material would be highly reduced in the vicinity of the bedrock mine (Beck, et al.
2002). The material might be expected to be found at greater distances from the
source than down-the-line trade or special utility (Renfrew 1977).
Renfrew (1977) showed that within a range of less than 5 km from the
source, the supply zone, where direct procurement is likely, the distribution of
material from the source might be expected to drop off linearly with distance and be
given by an expression of the form:
I=To-kx (Eq. 2.1)
where I is the amount of the material of interest found a distance x from the
source, Io is the amount of material at x=0, and k is the rate of decrease in the
amount of material carried from the source. Material might be expected to have been
carried in raw form to camps within the supply zone, but in decreasing amounts
with the distance the camp was from the source, and then further reduced before
being carried longer distances as tools.
To determine what the distribution of material that was associated with
down-the-line trade might be, Renfrew (1977) modeled down-the-line trade as a
process where a constant percentage of the traded material received by a group was
utilized and became part of the archaeological record of that site. The rest, also a
22


constant percentage of the received material, was traded down-the-line to the next
group in a linear path from the source. If the groups are considered to be uniformly
distributed, i.e., the linear down-the-line network composed of equally spaced
groups, then Renfrew (1977) found the amount of that material at a site a distance
x from the reference site can be expressed by the mathematical relationship:
I =loexp (-ax ) (Eq. 2.2)
where I is the quantity of the material left at a site of interest distance x (km)
from the reference site, often the source, Io is the initial amount of material at the
reference site (x=0), and a is a decay coefficient which has units of change per
kilometer, (1/km). When extended to two dimensions, Renfrew (1977) found the
expression relating the quantity of material at a site a distance x from the reference
site became a Gaussian of the form:
1= Ioexp (-ax2) (Eq. 2.3)
where the symbols have the same definition as given above.
Although both these models assume trade by more sedentary groups than the
hunter-gatherers that utilized this research area and that discarded material became
part of the archaeological record at the locations of residential sites, not distributed
along the path of groups seasonal foraging rounds, these models may not be
different from a model that describes use and disposal of a lithic material along the
seasonal round. Brantinghams (2003) neutral model simulates a forager randomly
walking across the landscape using lithic material randomly selected from a finite
23


sized tool-kit at every move. The material is replaced, independent of material type,
from randomly distributed sources of different types of lithic material as the forager
encounters them. This model gives a regression fit for the distribution of any specific
lithic material of:
I=b exp (-ax) (Eq. 2.4)
where I is the amount of material found at a distance x from the reference site
(usually the source) of the material, b is a constant found by fitting the distribution
and is likely related to the amount of material initially in the tool assemblage at the
reference site, and a is the decay coefficient (1/km) of the material.
This expression is the same form as Eq. 2.1 that Renfrew (1977) found for a
linear, one dimensional, down-the-line trade network model. The forager in
Brantinghams (2003) model (Eq. 2.4) is moving in 2-dimensional space and is
constantly using and discarding material, albeit randomly selected from all the types
of lithic material present in the tool-kit. From this it is expected that the distribution,
as a function of distance from the source, of a specific lithic material used in down-
the-line trade will not be distinguishable from the distribution that would result from
a forager using material on a seasonal round. However, if the distribution of material
as a function of distance does not follow an expression of the general form found by
Brantingham (2003) and Renfrew (1977) then it may indicate that the material was
used differently than either down-the-line trading or normal, unbiased use of material
along the seasonal round.
24


Limited Access
Access to material may have been limited by economic factors or by social
factors. As discussed above, embedded procurement reduced the cost of transport but
there is evidence that forays were likely made to sources of material within a days
walk of the seasonal round (Bamforth 2006; Binford 1979; Binford and O'Connell
1984; Gould and Saggers 1985). From the perspective of cost/benefit, it is expected
that sources with high costs, either in travel or procurement were not utilized unless
the benefit of the material balanced those costs. In areas with high relief, such as
mountains and canyons, it is expected that the acceptable travel distances for
procurement of material would be reduced in directions that included traversing
areas of high relief. Thus even though a source might be close in linear distance, the
path to the source could have a high cost due to canyons or mountains that had to be
traversed or circumvented which added cost. If in addition, the cost of procurement
at the source was high, such as proposed to be the case for a bedrock mine, it seems
probable that source would not be utilized except by those that had less costly
physical access. From the perspective of the sources, contours of equal cost, based
on rate of elevation (slope) change along paths radiating from the source, can be
drawn. It is posited that there was a maximum cost that prehistoric peoples would
expend to obtain material from a source. This maximum cost is posited to have
depended on the quality of material, cost of procurement, and value of the material
for special uses such as trade, ritual, prestige item, or other high value functions.
25


Contours of Equal Relative Cost
Figure 2.1 Notional Environment of Sources with Contours of Terrain Based Cost
For seasonal rounds that passed through contours from a source representing
costs larger than an acceptable value, i.e., outside a contour of acceptable cost
(CAC), material from that source would not be expected to have been utilized and
therefore not be evident in the debitage of camps associated with that seasonal round
In the notional environment shown in Figure 2.1, Group 1 traveling along its
seasonal round would have passed within the CAC for Source 2 (S2) and therefore
would have utilized Source 2 (S2) material. Although they passed through contours
of cost from the other quarries, in both other cases those contours represented costs
for transportation that exceeded the maximum cost the group would spend for those
26


materials. Group 2 on their seasonal round would have passed within the CAC for
both Source 1 (SI) and Source 3 (S3) and utilized those sources.
These contours of cost, as defined, do not represent the effect of the
perceived benefit of materials or the procurement but only the cost of transport of the
material. By determining the average cost contour within which reduction of material
from different sources of the same type, such as secondary sources, occur, a good
approximation of a nominal CAC (nCAC) for a particular type of source can be
derived. This assumes cost of procurement for all quarries of a given type are
approximately equal. If the occurrence of material from another source is compared
to the nCAC, the relative perceived cost/benefit differences for that source can be
informed.
Assuming equal costs of procurement, the supply zone and reduction of
materials with greater perceived value would be expected to occur beyond the nCAC
unless they were going to be transported long distances. In that case, Beck, et al.
(2002) would predict that reduction may have occurred closer to the source. In this
research, all material of comparable quality was being procured for transport along
the seasonal round so it is assumed that the nCAC is a measure of the supply zone
for that activity. If the cost of procurement was perceived to be higher for the source
being compared and the benefit of the materials equal, then the reduction is expected
to also have been closer than the nCAC. If instead the supply zone of the material
was at the nCAC or at a greater cost contour than nCAC, it can be assumed the
27


material had perceived as more valuable. Similarly, if sites are within the nCAC
from a source and yet that material does not occur in the site debitage, it could be
indicative of other factors affecting procurement such as availability and access.
Social factors such as quarry ownership or territories may limit accessibility,
though in the case of foragers, usually considered to be egalitarian, quarries are
generally considered to have been neutral ground (Ericson 1984; Gould, et al. 1971;
Purdy 1975). If territories did exist, ecological models for territorial behavior
indicate an inverse relationship between territory size and resource density
(Cashdan 1983). Dyson-Hudson and Smith (1978) contend, in their economic
defensibility model, that territories are likely to occur in areas of rich resources
because there is less cost involved in defense of those territories. These models
predict, in an environment containing ubiquitous lithic material resources such as is
the case for this research, small territories, possibly consisting of the quarries and a
small amount of surrounding area could exist, however, this model has been applied
only to subsistence resources and not tested for lithic sources.
Cashdan (1983) posits that territories wont exist unless there is competition
for resources due to scarcity or population pressure. In the case of this research area
where sources for lithic material are ubiquitous, pressure on resources related to
population increase seem to be the most likely cause of territories, if they existed.
The area exploited along the seasonal rounds traditionally used by different groups
could have become constricted due to increased pressure on resources as a result of
28


increases in population. In this scenario, access to lithic sources within those areas
would also have been reduced or lost to some groups.
The possible indicators of different ethnic group affiliations in the different
territories might be observable in the archaeological record by:
a. differences in what resources were utilized,
b. different tool kits associated with the utilization of different resources,
c. differences in the style of common artifact types such as pottery, basketry,
clothing,
d. differences in function or style of stone tools such as projectile points,
scrapers, drills, and even formal cores.
There has been extensive discussion regarding the difference between style
and function (Barton 1990; Hegmon 1992; Hurt 2001; Sackett 1977, 1985; Wiessner
1983) and their relationship to artifact variation. For the work discussed here these
arguments are superfluous since the hypothesis of this project is that the synchronic
occurrence of any systemic variation within an artifact type, whether due to
differences in function or style, and correlated with the presence or absence of
material from the bedrock mine inform the presence of different ethnic group
affiliations in different areas. If these occur it is possible that territories may have
existed and may have affected limits on access to the bedrock mine material.
If population pressure and subsequent ethnic territories were not the cause,
quarry ownership might still exist if a material had a special value such as prestige,
trade, or ritual importance. Though this is in conflict with current theory that
29


quarries utilized by foragers were neutral, it should be tested and not summarily
dismissed. If there was quarry ownership, the distribution of material discussed
above in Special Value would be expected and the material would appear only in
areas used by the owners of the quarry.
Synthesis
Many bedrock mines that are discussed in the literature such as Tosawihi
(Elston and Raven 1992), Windy Ridge (Bamforth 2006), Spanish Diggings (Reher
1991) and Mt. Jasper (Gramly 1980) appear to be the only source utilized in the
local area. Decisions that affected the choice to develop these bedrock mines rather
than utilize other available surface sources within the seasonal rounds are not
apparent or have not been addressed in the literature. The bedrock mine, Flint
Cave, chosen for this investigation has a proximal surface quarry within 400 meters.
Within the study area several other secondary deposition surface sources and
outcrops of material were all utilized by prehistoric peoples. These range from 1 km
to 10 km distance from the bedrock mine. By comparing and contrasting the
distribution and utilization of the lithic materials from these sources as they occur in
the archaeological record, the perception of costs and benefits by the prehistoric
people that created the bedrock mine may be informed. Specifically, the following
four hypotheses posited in this research can be tested:
A. The material from the bedrock mine was higher quality than materials
from the other surface sources.
30


B. The total energy of extracting usable material from the bedrock mine was
perceived to be less than or equal to the cost of obtaining material from the
secondary surface sources. The cost may have been more predictable.
C. The bedrock mine material was associated with high value functions such
as trade or ritual use. The importance may or may not have been associated
with material quality.
D. There was a shortage of available material due to limited access to other
quarries. This limited access could have been due to group ownership of
territories or individual quarries along traditional seasonal foraging routes or
unacceptable cost to transport.
If hypothesis A is likely, then the observables in the archaeological record
should show that material from the bedrock mine was used for formal tools and
possibly informal tools as well but material from the nearby secondary surface
source was used for informal tools. If the material from the bedrock mine was
considered higher quality and of limited quantity, it is expected that tools made of
that material are more likely to have been curated and appear at further distances
from the bedrock mine than the material from the secondary surface source. As a
result of preparing for curation, the Flint Cave material will show a higher level of
reduction near the source (Beck, et al. 2002; Newman 1994) than the material from
the surface quarries, and it will have been highly reduced within the supply zone.
The distribution of the Flint Cave material, the amount of material as a function of
distance from the source, is also expected to fit an exponential decay (Brantingham
2003), but the material from the secondary source may stay within the supply zone
(Renfrew 1977) and show a linear decay. However, this could be complicated if the
31


lesser quality material was preferred for a particular tool type. Beck and Jones (1990)
noted that scrapers were most often associated with specific material but that
projectile points were made from a wide variety of materials. If this were the case for
the material from the secondary, surface sources, then a high correlation should be
found between that material and a particular non-projectile point tool type.
Depending on the tool type, its required frequency of use in the seasonal round, and
availability of other suitable material along the seasonal round, the material from the
surface source may occur at distances beyond the supply zone similar to the higher
quality material.
Determining if hypothesis B can be disproved is not as straightforward.
Although it would appear at first that bedrock mining would have been a higher cost
in labor than obtaining material from a surface mine, Elston and Raven (1992) found
that once the cost of removing the over-burden has been expended, bedrock mining
has a higher return of material per energy expended than surface sources. To
determine if that was true for the Flint Cave requires determining the relative quality
of the material from the surface quarry and the bedrock mine. This can be informed
by comparing the CAC for the Flint Cave with the CAC for the secondary surface
quarry. If the bedrock mine material and the material from the nearby surface source
were perceived to be of equal quality, but the cost of procurement of the Flint Cave
material was perceived to be higher, then it is expected the higher procurement cost
would have made the Flint Cave material more valuable, and it would have been
32


conserved and used only for formal tools that were curated. The CAC, distance of
the supply zone to the source, would also be smaller than that of the surface quarry.
This would have resulted in a higher level of reduction of the Flint Cave material
near the source, and the distribution of the Flint Cave material would fit the
exponential decay beyond the supply zone. If the materials were perceived to have
been of equal quality and equal procurement costs, then the two materials would
have been used interchangeably for both formal and informal tools and their level of
reduction and distributions would have been similar, and the CAC for the two
materials would be the same. If the Flint Cave material was of lesser quality than the
surface source material, it is unlikely that the bedrock mine would have been created
unless the procurement costs were less or there were social factors affecting the use
of the Flint Cave material (reference hypotheses C and D).
If the Flint Cave material was of special value, i.e., ritual, trade, or prestige
items, then hypothesis C could be informed by one or more of the following:
correlation of the material with a particular tool type or object, the distribution of
material, and level of reduction near the source. It is expected if the Flint Cave were
of special value then the material would be highly reduced near the source in
preparation for long distance curation (Beck, et al. 2002). If the Flint Cave material
was used for trade, the distribution of material might still fit an exponential decay
(Renfrew 1977), but it might decay slower outside the supply zone and occur further
from the Flint Cave than other materials in the assemblages. If the material was used
33


for ritual or prestige items, the material might be expected to occur in lesser but more
constant quantities once outside the reduction and supply zones, falling off slowly
with distance.
Finally, hypothesis D can be informed if the quality of the Flint Cave
material is less than the material from the secondary surface source. In this case, it is
highly likely that the Flint Cave would not have been created if there was not a
shortage of material available. Since the area has ubiquitous sources, it would imply
that access to those sources was limited. If the Flint Cave material is of higher
quality than material from other surface sources in the area, and sites containing
those other materials fall within the nominal CAC (the average CAC for all quarries
other than the Flint Cave in the area) of the Flint Cave yet there is no Flint Cave
material found in those sites, it is possible that access to the Flint Cave was limited.
From mid range theories six parameters have been identified that can be
determined from the archaeological record and used in this research to compare and
contrast the bedrock mine and the proximal secondary surface source to test the
hypotheses of this research. The research parameters (RPx, where x=l thru 6) are:
use of material from each source (Ml for Flint Cave material and M2 for 5MN7753
material), size of the supply zone (or contour of acceptable cost, CAC), level of
reduction near the source, distance material is found from the source, and frequency
of occurrence as a function of distance from the source. Each of these research
parameters can have more than one value as shown in Table 1. Of the 6804 possible
34


combinations of these values many are contradictory, and many are unlikely to occur
and do not fit current mid range theory. For example, if RP1: informal only, RP2:
formal and informal, RP3: Ml>nCAC and M2=nCAC, RP4: Ml=Yes and M2=No,
RP5: M1>M2, and RP6: M1 =exponential and M2=exponential, then the implication
is that the material from the bedrock mine is of lower quality than the material from
the secondary source yet has a larger supply zone than the high quality material, is
reduced near the source but M2 is not, and travels further but has an exponential
decline in frequency of occurrence. If the material from the bedrock mine had ritual
significance all but the value of RP6 might apply. If it had ritual value it is unlikely
that it would show a pattern of frequency of occurrence similar to material that was
being utilized for normal functions.
Tables 2.3 through 2.6 show some of the scenarios and their parameter values
that could occur in this research and that support each of the hypotheses A thru
D respectively. Many scenarios indicate more than one hypothesis may be
informed.
35


Table 2.2 Test parameters with possible values for each
Value No. RP1: Use of bedrock mine material (Ml) RP2: Use of material from secondary surface source (M2) RP3: Distance of supply zone compared to nominal CAC (nCAC) RP4: High level of reduction close to the source RP5: Distance from the source RP6: Functional form of the frequency of occurrence with distance
1 Formal only Formal only Ml=nCAC M2=nCAC Ml=Yes M2=Y es M1=M2 Ml=Exponential M2=Exponential
2 Informal only Informal only MKnCAC M2=nCAC MKYes M2=No M1>M2 Ml=Exponential M2=Linear
3 Both formal and informal Both formal and informal Ml>nCAC M2=nCAC Ml=No M2=Yes M1 4 Ml=nCAC M2 5 Ml=nCAC M2>nCAC Ml not linear or exponential M2=Linear
6 Ml>nCAC M2>nCAC Ml not linear or exponential M2=Exponential
7 MKnCAC M2 8 Ml=Linear M2 not linear or exponential
9 MKexponential M2 not linear or exponential
36


Table 2.3: Scenarios that inform Hypothesis A
Scenario Configuration RP1: Use of bedrock mine material (Ml) RP2: Use of material from secondary surface source (M2) RP3: Distance of supply zone compared to nominal CAC (nCAC) RP4: High level of reduction close to source RP5: Distance from the source RP6: Functional form of the frequency of occurrence with distance
S(3,2,l,2,2,2) Formal and informal Informal only Ml=nCAC M2=nCAC MKYes M2=No M1>M2 MKExponential M2=Linear
S(3,2,3,2,2,2) Formal and informal Informal only Ml>nCAC M2=nCAC MKYes M2=No MKM2 MKExponential M2=Linear
S(3,2,3,2,2,5) Formal and informal Informal only Ml>nCAC M2=nCAC MKYes M2=No MKM2 MKNot Linear or Exponential M2=Linear
S(3,2,3,4,2,4) Formal and informal Informal only Ml>nCAC M2=nCAC MKNo M2=No M1>M2 MKLinear M2=Linear
S(3,2,7,2,2,2) Formal and informal Informal only MKnCAC M2M2 MKExponential M2=Linear**
S(3,2,7,4,2,2) Formal and informal Informal only MKnCAC M2 S(1,2,7,2,2,2) Formal only Informal only MKnCAC M2 S(1,2,1,2,2,2) Formal only Informal only MKnCAC M2=nCAC MKYes M2=No MKM2 MKExponential M2=Linear ***
S(1,2,3,4,2,5) Formal only Informal only MKnCAC M2=nCAC MKNo M2=No MKM2 MKNot Linear or Exponential M2=Linear
* Could mean material also used for prestige or ritual item
** Possible limited access to other materials due to social factors
*** Possible limited access due to cost
37


Table 2.4: Scenarios that inform Hypothesis B
Scenario Configuration RP1: Use of bedrock mine material (Ml) RP2: Use of material from secondary surface source (M2) RP3: Distance of supply zone compared to nominal CAC (nCAC) RP4: High level of reduction close to source RP5: Distanc e from the source RP6: Functional form of the frequency of occurrence with distance
S(3,3,l,l,l,l) Formal and informal Formal and informal Ml=nCAC M2=nCAC Ml=Yes M2=Yes M1=M2 M KExponential M2=Exponential
S(3,3,6,l,l,l) Formal and informal Formal and informal M1 >nCAC M2>nCAC Ml=Yes M2=Yes MKM2 Ml=Exponential M2=Exponential
S(3,3,5,3,3,l) Formal and informal Formal and Informal Ml=nCAC M2>nCAC Ml=No M2=Yes M2>M1 Ml=Exponential M2=Exponential
S(3,3,l,4,l,4) Formal and informal Formal and informal M1 =nC AC M2=nCAC Ml=No M2=No M1=M2 MKLinear M2=Linear
S(l, 1,7, U,l) Formal only Formal only MKnCAC M2 S(1,1,1,4,1,4) Formal only Formal only Ml=nCAC M2=nCAC Ml=No M2=No M1=M2 MKLinear M2=Linear
S(2,2,1,4,1,4) Informal only Informal only Ml=nCAC M2=nCAC Ml=No M2=No MKM2 MKLinear M2=Linear*
S(2,2,7,4,l,4) Informal only Informal only MKnCAC M2 * Possible limited assess to other material
38


Table 2.5: Scenarios that inform Hypothesis C
Scenario Configuration RP1: Use of bedrock mine material (Ml) RP2: Use of material from secondary surface source (M2) RP3: Distance of supply zone compared to nominal CAC (nCAC) RP4: High level of reduction close to source RP5: Distance from the source RP6: Functional form of the frequency of occurrence with distance
S(1,3,1,1,2,6) Formal only Formal and informal M1 =nC AC M2=nCAC Ml=Yes M2=Yes M1>M2 Ml=Not linear or exponential M2=Exponential
S(l,3,3,l,2,l) Formal only Formal and informal Ml>nCAC M2=nCAC Ml=Yes M2=Yes M1>M2 Ml=Exponential M2=Exponential
S(3,3,3,l,2,6) Formal and informal Formal and Informal M1 >nC AC M2=nCAC Ml=Yes M2=Yes M1>M2 Ml=not Linear or Exponential M2=Exponential
SO,3,U,1,6) Formal only Formal and informal Ml=nCAC M2=nCAC Ml=Yes M2=Yes M1=M2 Ml=not Linear or Exponential M2=Exponential
S(2,l,7,l,l,6) Informal only Formal only MKnCAC M2 S(2,3,1,3,2,6) Informal only Formal and informal Ml=nCAC M2=nCAC Ml=No M2=Yes M1>M2 M1 =not Linear or exponential M2=Exponential
* Possible limited access to materia
39


Table 2.6: Scenarios that inform Hypothesis D
Scenario Configuration RP1: Use of bedrock mine material (Ml) RP2: Use of material from secondary surface source (M2) RP3: Distance of supply zone compared to nominal CAC (nCAC) RP4: High level of reduction close to source RP5: Distance from the source RP6: Functional form of the frequency of occurrence with distance
S(2,2,6,4,l,4) Informal only Informal only Ml>nCAC M2>nCAC Ml=No M2=No M1=M2 MKLinear M2=Linear
S(2,l,5,3,3,3) Informal only Formal only Ml=nCAC M2>nCAC Ml=No M2=Yes M2>M1 MKLinear M2=Exponential
S(l,1,7,1,1,1) Formal only Formal only MKnCAC M2 S(l, 1,1,1,1,1) Formal only Formal only Ml=nCAC M2=nCAC Ml=Yes M2=Yes M1=M2 MKExponential M2=Exponential
S(1,2,3,2,2,2) Formal only Informal only Ml>nCAC M2=nCAC Ml=Yes M2=No M1>M2 MKExponential M2=Linear
S(3,3,7,l,1,1) Formal and Informal Formal and Informal MKnCAC M2 Research Limitations
Although Table 2.3 thru 2.6 provide a decision matrix for teasing apart the
economic and social factors that resulted in a bedrock mine in the midst of an area
with ubiquitous secondary sources, the ability of this investigation to gather all the
information needed to inform these parameters is limited by the 15 km radius area
around the bedrock mine investigated in this work. Differentials in the travel distance
of curated, formal artifacts due to trade, prestige items, or ritual importance are
expected to occur beyond this distance. Therefore, the interpretations possible from
this work will only be those that can be obtained from differences in supply zone,
level of reduction, distance the materials moved from the sources within the 15 km
40


limitations of the research area, composition of debitage assemblages, and
differences in frequency of discarded informal and formal tools made from the
materials.
This research is a synchronic analysis of the distribution and utilization of
lithic material from primarily two proximal quarry sites, one a bedrock mine and the
other a surface quarry. It is assumed in this work that 1) the secondary surface
quarries were available for use over the period of time that the bedrock mine was
being utilized and 2) the majority of lithic scatter sites sampled and analyzed in this
research reflect the same lithic economy. There are two possible scenarios that could
apply to use of the Flint Cave and proximal surface source (5MN7753) lithic
material in the research area: 1) both sources were used during the period the Flint
Cave was utilized, and the surface source was utilized before and after that period,
and 2) only the Flint Cave material was utilized during the period of mining and the
surface quarry was utilized before and after that. The first scenario is that assumed
by this research. Choices and perceptions affecting the utilization of the material
from 5MN7753 and the Flint Cave should be evident in the Flint Cave material but
might be obscured for the use of surface source material due to the periods of
exclusive use of the 5MN7753 material both before and after the Flint Cave.
However, if the second scenario applies instead, then it is expected that the material
from both sources would have been used for all tool types during the periods of their
respective, exclusive utilization; evidence of long distance travel would be indicated
41


for both materials; contribution of both materials should exponentially decline with
distance from the respective source. It might also be expected that there would be
more sites with material from the surface source than the Flint Cave, and deflated
sites that were utilized both before, during, and after utilization of the Flint Cave
would be expected to have more material from the surface source due solely to the
likely longer period of utilization of the surface source than the Flint Cave.
Occurrence of materials within stratified sites may also inform whether the
assumptions of this research were true. However, there are a limited number of
excavated sites within the study area and further excavations in support of this work
were beyond the scope of this project.
A project was funded through the Colorado State Historical Society General
Fund with the support of the Bureau of Land Management (BLM) and the Montrose
Chapter of the Colorado Archaeological Society that informed the time period when
at least some activity at the Flint Cave ended. Projectile point typologies and limited
dating from the excavations in the research area provided some temporal
information. However, due to the limited area investigated and lack of temporal
controls, this investigation will only provide indications of the economic and social
factors that resulted in the utilization of both surface quarries and bedrock mine in
the study area and identify lines of investigation for future work.
42


CHAPTER 3: BACKGROUND
Environment
The research area for this work was a roughly circular area of 15 km
radius centered on the Flint Cave quarry, the focus of this research, located on the
eastern sloping flank of the Uncompahgre Plateau in west-central Colorado (Figure
3.1). The Uncompahgre Plateau is an uplift running northwest to southeast
terminating at the Uncompahgre River on the eastern side and cut by several
canyons. Five of these canyons cut through the research area; Shavano Canyon on
the extreme east and south; Dry Creek Canyon, Cushman Canyon, and Coal Bank
Canyon cut through the middle of the research area; Roubideau Canyon on the west
and north. Currently both Roubideau Canyon and Shavano Canyon have perennial
streams but the recent drought period has resulted in the stream in Dry Creek Canyon
being intermittent. Historically Cushman Canyon has only been seasonal and Coal
Bank Canyon has only localized sources.
The average temperature range at the lower elevations on the Uncompahgre
Plateau is from -2 C to -3 C in January to 18 C to 210 C in July. Though not
applicable to this research, the growing season is 140-150 days at the lower
elevations and decreases to less than 50 at the higher elevations. The annual
43


precipitation at the lower elevations ranges from 200 mm to 310 mm with half
occurring during the growing season (Reed 2001).
Vegetation across the research area currently ranges from salt brush and
sagebrush flats at the lower elevations to pinon-juniper at the middle elevations and
Gambles Oak and Ponderosa pine at the highest elevations. The vegetation at the
highest elevations of the Uncompahgre Plateau, which were outside the research
area, includes aspen, spruce, and fir. The drought of the last several years and an
infestation of pine beetles have resulted in juniper dominating the lower-middle
elevation and pinon occurring only at the higher-middle elevations on the plateau,
higher than they were historically.
733000rr>E. 745000mE. 761 000lWG3S84 Zone 13S 260E.
10/07/07
Figure 3.1 Research Area on Uncompahgre Plateau
44


Prehistorically the climate in the research area was significantly cooler during
the early Holocene (13,000-10,500 B.P.) than today but by 8000-4000 B.P., during
the Middle Holocene, the climate may have been hotter and drier than today (Reed
and Metcalf 1999; Reed, et al. 2001). By the Late Holocene (after 4000 B.P.) cooler
conditions were again dominating and pine became more dominant at lower
elevations than today (Tucker 1987). This was followed by a warming period and by
2000 B.P. conditions similar to modem conditions had developed.
Kim Gefhardt. 4/28/01
ManCOS Shale (Km), eroded from Uncompahgre Plateau
Dakota Fm. (Kd)
Laminated / Nodular Chert Quartzite I
Burro Canyon Fm. (Kb)
7 Brushy Basin Mbr.
X (Jmb) E
Salt Wash
Mbr. (Jms)
Junction Creek
Sandstone (Jj)
Figure 3.2: Geological Formations Applicable to the Uncompahgre Plateau
(Gerhardt 2001)
45


Geology
The Uncompahgre Plateau is composed of Mesozoic era formations,
primarily from the Jurassic and Cretaceous periods (Taylor 1999). The Uncompahgre
Plateau provided plentiful sources of lithic materials for the prehistoric hunter-
gatherer peoples that utilized that area from the Paleoindian to the Protohistoric Ute
who were moved from the area in 1881. The sources of lithic materials were
primarily from two formations, the Brushy Basin member of the Morrison Formation
in the Jurassic Formation and the Burro Canyon Formation which lies between the
Morrison Formation and the Dakota Formation (Figure 3.2). Within these formations
silicified sandstone, silicified volcanic ash, and nodular chert are the primary sources
of the fine grained materials suitable for knapping (Reed and Metcalf 1999).
The Flint Cave bedrock mine, which is the focus of this research, is located
on the west facing rim of Dry Creek Canyon in a vein of fine-grained silicified
sandstone or silcrete in the Burro Canyon formation. Other lithic sources that were
utilized by prehistoric people occur as secondary deposition, surface sources on the
slopes of the canyons. These are the result of erosion of lenses and veins of material
in the canyon walls. This includes one source a few hundred meters to the north-
northwest of the Flint Cave and several others across Dry Creek Canyon, in
Cushman Canyon and Coal Bank Canyon, all within 1-2 km of the Flint Cave. These
sources are clasts from the Brushy Basin member of the Morrison Formation that
have eroded from the walls of the canyons.
46


Cultural History
Within the research area evidence of Paleoindian, Archaic, Formative, and
Proto-historic cultural periods are present. In this region, the Northern Colorado
Plateau, the Paleoindian period extended from 13,000 to 7500 B.P. (cal) with Clovis
thought to have extended from 13,000 to about 12,700 B.P., followed by Folsom
from 12,700 to about 11,500 and then the Foothill-Mountain complex extending
from 11,500 to 7500 B.P. (Reed and Metcalf 1999). The Clovis point is characterized
by large fluted lanceolate projectile points suited for hunting megafauna. Smaller,
finely crafted fluted lanceolate points for hunting now-extinct species of bison are
attributed to Folsom. The Foothill-Mountain complex had unfluted, restricted stem
and indented bases lanceolate points. It has been suggested (Frison 1992) that the
greater regional variability seen in these point types may be due to more localized
specialization. Pitblado (2003) found that these point types are more likely to be
manufactured from local materials, particularly quartzite. Evidence of this period in
the research area has primarily been from surface finds. Only one excavated site near
the research area, 5DT2 (Buckles 1971; Reed, et al. 2001), has yielded artifacts
associated with the Paleoindian period.
The Archaic period in the region extended from 8350 B.P. to about 1950 B.P.
and has been broken into four periods (Reed and Metcalf 1999), the Pioneer (8350-
6450 B.P.), Settled (6450-4450 B.P.), Transitional (4450-2950 B.P.), and Terminal
(2950-1950 B.P.). Projectile points from the Archaic period occurred in a wide
47


variety of styles including lanceolate, stemmed, side-notched, and comer-notched. It
is generally believed that they were for atlatl dart points but spear points and early
experimenting with bow and arrow cannot be dismissed (Reed and Metcalf 1999).
Lanceolate styles, Deception Creek, Humboldt, and McKean, occurred between
about 8000 B.P. and 4000 B.P. (Holmer 1986; Reed and Metcalf 1999), and most
predate 5600 B.P. Stemmed points in the region include Gypsum, Elko contracting
stem, Gatecliff contracting stem, Pinto, Duncan-Hanna points of the McKean
complex, and a wide range of other straight to convex and rounded base, stemmed
points. Generally, these contracting stemmed points occurred in the region between
7800 B.P. and 2200 B.P., and most date prior to 3000 B.P. (Reed and Metcalf 1999).
Points of the McKean complex occurred between 4600 and 3000 B.P. (Reed and
Metcalf 1999). Side notched points include Elko Side-notched, Bitterroot, Northern
Side-notched, Hawken, Mallory and Mt Albion and other unnamed styles. These
side notched points include both low and high side notches and have basal shapes
that include concave, straight, and convex. Side notched points range from 7000
B.P. to 1200 B.P. with most predating 3500 B.P. Comer-notched points show a great
deal of variety in size and basal shape and are generally included in the Elko Comer-
notched typology. Comer notched points occur from 7800 B.P. to about 500 B.P.,
and most occurred between 3000 B.P. and 2000 B.P. (Reed and Metcalf 1999).
The Formative period occurred between 400 B.C. and A.D. 1300 in the
region. This consisted of the Fremont tradition to the north and west of the research
48


area and the Anasazi tradition to the west and south of the research area. Although
some Formative era pottery sherds were found within the research area as part of this
work, there is currently no evidence of horticulture occurring in the research area or
any structures usually attributed to the Formative period peoples. Reed (1997) has
defined a Gateway tradition within the Formative Period for east-central Utah and
west-central Colorado, particularly the western side of the Uncompahgre Plateau,
dating from 2350 B.P. to 700 B.P. Reed (2001). This tradition is characterized by
Reed (2001) as having limited reliance on horticulture, small comer-notched
projectile points, ceramics procured through trade with Anasazi and Fremont, lack of
their own ceramics, short term use of circular and rectangular masonry surface
habitation structures, possible pit habitation structures, granaries and storage cists in
rock shelters, and rock art with both Anasazi and Fremont influence. Within the
research area a small number of Puebloan pottery sherds have been found, but no
surface habitations, pit houses, or storage cists have been discovered, indicating it is
unlikely the Gateway tradition was present in the research area. Reed and Metcalf
(1999) have also defined an Aspen tradition, concurrent with the Gateway tradition,
as an adaptation of full time hunter/gatherers associated with intensified subsistence
practices coincident with a dramatic population increase in the region (Reed, et al.
2001). This tradition is characterized by Reed and Metcalf (1999) as devoid of
horticulture or masonry habitation structures but having small quantities of ceramics
obtained by trade with other groups, intensified use of game drives and storage
49


structures in rock shelters, and small comer-notched projectile points consistent with
adoption of the bow and arrow. The Aspen tradition was present in the research area
until the ancestors of the Utes, who occupied the area in the Protohistoric era, arrived
and may have even overlapped with the ancestral Utes for a time.
The Protohistoric era is the period between the end of the Formative Period
and the removal of Native Americans to reservations which occurred in the region in
1881. Although Utes were the dominant presence in the research area during the
Protohistoric period, there is limited evidence, mostly rock art and some ceramics
(Reed, et al. 2001), that indicates the Navajo were also in the area at times. It is
speculated that the ancestral Utes immigrated from southwestern Nevada and
southeastern California sometime after A.D. 1100 and were primarily
hunter/gatherers. The Utes are characterized as having utilized brush structures and
wickiups for habitation, manufactured fingertip-impressed and plain, micaceous
Uncompahgre Brown Ware (Buckles 1971; Reed, et al. 2001), and used Desert side
notched and Cottonwood triangular projectile points. The Protohistoric Ute tradition
has been broken into two phases, the Canalla between A.D. 1100 and approximately
A.D. 1650 which included the adoption of the horse (Reed and Metcalf 1999), and
the Antero phase from A.D. 1650 until A.D. 1881. Traditional artifact types and
habitation types continued into the Antero phase but the horse allowed use of tipis
and increased interaction with more distant groups of Native Americans such as the
Navajo and Arapahoe as well as Euroamericans. Euroamerican trade items such as
50


glass beads, metal cone tinkers, and tin cans are common in Antero phase sites
(Reed, et al. 2001).
Past Archaeological Work
Archaeological work has been sporadic on the Uncompahgre Plateau and has
been largely confined to the eastern and western flanks of the plateau. Most of the
archaeological work on the western flank has concentrated on sites with masonry
architecture (Crane 1977; Hurst 1940, 1941, 1942, 1943, 1944, 1945, 1946, 1947,
1948; Kasper 1977) located 80 km to 120 km west of Flint Cave.
The first formal archaeology on the eastern flank of the Uncompahgre
Plateau (Wormington and Lister 1956) was excavation of several rockshelter sites in
1937 to 1939 and 1951 and 1952 resulting in the definition of the Uncompahgre
Complex, a variant of the Desert Culture (Reed and Metcalf 1999). Those several
excavated sites are located from approximately 16 km to 100 km north of the Flint
Cave and at somewhat lower elevations than the research area for this project. After
that, little was done until the cultural historical Ute Prehistory Project (Buckles
1971). That project resulted in limited excavation of some of the thirty-nine sites
identified by the project, ranging from 1 km to 80 km from the Flint Cave. These
sites nearly circumscribe the bedrock mine, missing only the south to west quadrant
completely, and the east to south quadrant was sparsely represented. Artifacts from
nine of the sites excavated by Buckles were included in this research. Buckles work
fell short of its goal of establishing the prehistory of the Ute in the region but did
51


provide additional data on the Uncompahgre Complex (Reed and Metcalf 1999)
particularly in ceramics and rock art and added to the database available for the area.
Although the data was relatively sparse, Buckles also attempted to develop a
projectile point typology and associated chronology for the area.
The next significant effort in the area was the result of cultural resource
management work associated with the Trans-Colorado Natural Gas Pipeline (Reed,
et al. 2001). This effort excavated two additional sites that produced artifacts
included in this research, 5MN3859 and 5MN2628. These sites are 1 km northwest
and 15 km southwest of the Flint Cave, respectively. The archaeologist from the
Bureau of Land Management Office in Montrose, Colorado has also been working
with local residents and members of the local avocational archaeology group, the
Chipeta Chapter of the Colorado Archaeological Society, since 2001 in an
exhaustive effort to record prehistoric and historic sites across the whole region,
particularly on the Uncompahgre Plateau. This effort has significantly improved the
knowledge of quarry sources and site locations in the area pertinent to this research
effort. Finally, between 2005 and 2007 an investigation and documentation of the
physical characteristics and chronology of the Flint Cave was made possible by a
Colorado Office of Archaeological and Historic Preservation (OAHP) General Fund
grant, 2005-M2-030, to the Chipeta Chapter of the Colorado Archaeological Society
(Montrose, Colorado) in cooperation with the Montrose District of the Bureau of
Land Management (BLM). That work, accomplished by Chuck Richey, principal
52


investigator for the Chipeta Chapter, and this author did establish that the Flint Cave
was of prehistoric origin and informed some chronology for activity at the Flint
Cave, which was critical to this thesis work.
Flint Cave Physical Characteristics
The Flint Cave is at an altitude of 1825 meters (m). It is approximately 70 m
above the floor of Dry Creek Canyon and approximately 30 m below the rim in the
Burro Canyon Formation. Flint Cave has two openings (Figure 3.3), the larger
opening is not accessible today except by use of technical climbing techniques. It
may have been accessible in the past by an extension of the current ledge that
provides access to the smaller opening, but fell away some time in the past leaving
the original entrance isolated on the sheer cliff. The smaller opening, approximately
60 centimeters (cm) in diameter, provides access to the main room of the Flint
Cave.
It is unknown when the smaller opening was made but likely was made after
initial use of the larger opening and possibly after mining of material from the
bedrock mine had ended. This possibility is inferred from the observation that the
silcrete ore body that the smaller entrance penetrates was not mined to its furthest
extent as might be expected if it was an original opening and was created during the
time mining of the silcrete was in progress. In contrast, the ore body around the
large opening is more fully mined out. It appears likely that existing fractures and
53


cracks in the ore body and surrounding sandstone provided a starting point for the
mining activities.
Figure 3.3 Openings to Flint Cave
54



\
Figure 3.4 Plan View of Flint Cave
The interior of the Flint Cave is composed of three rooms, designated as the
Main Room, Back Room, and View Room (Figure 3.4). The silcrete material
appears to have been a vein that composed the lower meter of the Flint Cave
measured from the existing floor. The mining activity left depressions in each room,
pits, now covered by fill in the Main Room and View Room and partially filled in
the Back Room. Figure 3.5 is taken from the small opening in the Main Room
looking towards the Y. The large opening is to the right and entrance to Back
Room to the left. White areas in the black are where pieces of sandstone have been
removed by vandals.
55


The sandstone ceiling and to a lesser extent the silcrete walls of the Flint
Cave have a black coating. This was originally thought to be soot from repeated fires
used to facilitate the mining process or from later use. Dr. Alan Watchman, a world
recognized authority on rock coatings and dating, analyzed the coating to determine
its composition and to determine if l4C methods could be used to date it. It was
determined that the coating did originate from an organic pathway, likely nano-
bacteria, and was datable (Richey and Hauser 2007).
Figure 3.5 Interior of the Flint Cave (5MN7429)
56


Flint Cave Chronology
Two sets of OSL samples were collected from the fill in the Flint Cave in 2.5
cm increments which extended from the surface of the fill to bedrock. One set of
samples was taken in the Y, the convergence between the View Room, the Main
Room, and the Back Room (Figure 3.4). This set of samples went to a depth of 45
centimeters (cm). The second set of samples was taken in the Back Room and went
to a depth of 37.5 cm. All the samples were collected in the dark to allow optical
stimulated luminescence (OSL) dating to be accomplished on the samples. Charcoal
found in these sample increments was separated and several were dated using 14C
along with a sample of the organically derived black coating from the wall near the
opening of Back Room into the Y. As indicated in Table 3.1, the dates obtained
from the l4C of the wood found at the deeper levels (Y14, Y16, Y17), 14C of the
carbonate from the wall, and single grain OSL from the deepest levels (Y17, BR14)
are quite consistent. It appears from these results that mining activity in the Y
ended ca 2600 years ago (2600 B.P. average) and in the Back Room at least 500
years later. Because the depressions where the material was obtained were mined
from the top down, these dates likely represent the time of the last mining activity in
those pits. The nano-bacteria that left the carbonate (black coating) on the wall
may not have started as soon as the quarrying activity ended but was not disturbed
since its growth, therefore the 2350 rcybp obtained from the l4C of the carbonate
indicates that mining activity in the Y had ended by at least that time and no
57


further quarrying was done at that location in the Flint Cave after that (Richey and
Hauser 2007).
Table 3.1 Dates from Flint Cave
Source l4C Dating Single Grain OSL
Y Level 17 (42.5 cm) 2520 +/- 40 rcybp (cal 656 B.C.) 2440 +/- 40 rcybp (cal 502 B.C.) 2598+/- 186 ya (593 B.C.)
Back Room Level 14 (35 cm) 1939+/- 190 ya (A.D. 66)
Wall Carbonate 2350 +/- 200 rcybp. (cal 415 B.C.)
It is not possible from the archaeology done on the Flint Cave to date to
determine when the mining first started. The chronology of the Main Room, the
largest area mined in the Flint Cave, was not determined and may predate or postdate
that of the Back Room. Because the Y had to be mined first to gain access to the
material that was mined to create the Main Room, it is likely mining in the Main
Room ended after the activity in the Y. From the dates obtained in this
investigation, which likely represent the end of the Flint Cave mining activity, it is
likely that mining in the Flint Cave occurred over a long period of time, finally
ending late in the Archaic or possibly early in the Aspen tradition of the Formative
Period.
58


Flint Cave Solar Alignment:
To inform if the Flint Cave might have had some function that could have
given the material from the mine special importance, the possibility of a seasonal
calendar was investigated by Richey and Hauser (2007). The comparison of the sun
tracks by month, with the azimuth and elevation constraints defined by the large
entrance of the Flint Cave to both 1.5 meters inside and from the back wall of the
Flint Cave were determined (Figure 3.6). From this graphic is can be seen that the
inside of the Flint Cave is directly illuminated only during the period between mid-
September and mid-March. During the summer months the sun does not directly
illuminate any part of the interior of the Flint Cave. The last direct illumination
occurs near the spring equinox and starts again near the fall equinox. However, more
precise measurements than those done in the initial investigation carried out by
Richey and Hauser (2007) would have to be accomplished to determine how close to
the spring and fall equinox the direct illumination starts and ends. From this
preliminary work, some function of the Flint Cave as a solar calendar cannot be ruled
out, however, it is likely the alignment is coincidental.
59


June
Figure 3.6 Solar Illumination of Flint Cave
60


CHAPTER 4: METHODOLOGY
In the previous theory chapter, six parameters (Table 4.1) were identified that
in concert could be used to test the four hypotheses proposed in this thesis (Table
4.2) to inform why a prehistoric bedrock mine exists in an area of ubiquitous
secondary surface sources of lithic materials. This chapter describes the
methodology used to obtain the data needed to inform each of these parameters from
the archaeological record of the project area.
Table 4.1 Test parameters
Parameter Description
RP1 Use of the material from the bedrock mine
RP2 Use of the material from secondary surface source 5MN7753
RP3 Distance of the supply zone for the bedrock mine compared to the nominal contour of acceptable cost (nCAC) for other quarries in the area.
RP4 Level of reduction of material close to the source
RP5 Distance of material occurrence from the source
RP6 Functional form of the frequency of occurrence of the material with respect to distance
61


Table 4.2: Thesis hypotheses
Hypothesis Description
A The material from the bedrock mine was higher quality than materials from the other surface sources.
B The total energy of extracting usable material from the bedrock mine was perceived to be less than or equal to the cost of obtaining material from the secondary surface sources. The cost may have been more predictable.
C The bedrock mine material was associated with high value functions such as trade or ritual use. The importance may or may not have been associated with material quality.
D There was a shortage of available material due to limited access to other quarries. This limited access could have been due to group ownership of territories or individual quarries along traditional seasonal foraging routes or unacceptable cost to transport.
Data to inform the parameters for this work were obtained by sampling
surface lithic scatter sites, isolated finds, limited tool survey, and artifacts and
debitage from previously excavated sites that occurred within the research area. A
permit, C-67660, for limited surface collection within the research area was obtained
from the Colorado office of the U.S. Bureau of Land Management (BLM) for
calendar years 2004-2007 and was monitored by the archaeologists at the Montrose
District of the BLM, Julie Coleman and Dr. Glade Hadden. A letter of agreement
was obtained from the Grand Mesa, Uncompahgre, and Gunnison District of the
USFS in Delta, Colorado to allow research on US Forest Service land which was in
the extreme western part of the research area. This was monitored by the USFS
district archaeologist, Leigh-Ann Hunt. A curation agreement was granted by the
62


Anasazi Heritage Center, Dolores, Colorado for curation of all material collected for
this research.


o
A
0 25 5 10 Kilometers
1 ' i i l ' l
Legend
* Flint Cave (5MN7429J
+ 5MN7756
5MN7755
A 5MN8385
G 5MN8384
5MN7753
5MN7754
5MN8386
0 5MN4638
Project Area Elevation
Value
m High : 2946 m
Low: 1624 m
Figure 4.1 Locations of Quarries within Research Area
The research area in which all data was collected was defined as a circular
area with a radius of 15 kilometer (km) centered on the bedrock mine 5MN7429
(Flint Cave). The circular area was truncated starting about 4 km to the east of the
Flint Cave by agricultural and private land, and there were smaller pieces of private
land scattered within the remaining area where data gathering was limited. Within
the remaining area composed of BLM and US Forest Service land nine quarry sites
were identified (Figure 4.1) and evidence of other nearby sources were indicated in
63


the archaeological record, but their locations were not identified. Other than the
Flint Cave, all of the quarries were secondary surface sites.
One of the secondary surface quarry sites, 5MN7753, is approximately 400 m
NNW of the Flint Cave on the same canyon rim as the Flint Cave. The proximity of
these two dissimilar type quarries provided an opportunity for this thesis to compare
and contrast the utilization and distribution of the material from these two quarries
which had the same costs of travel on the landscape, e.g. contours of acceptable cost
(CAC), but seemingly different costs of extraction. The data necessary to determine
utilization and distribution of the material from the Flint Cave (5MN7429),
5MN7753 quarry, 5MN7755 quarry, and 5MN7756 quarry, and to a lesser extent
from the other five remaining quarries identified in this research, were developed by
obtaining samples of lithic materials and artifacts from across the research area and
sourcing the materials to these quarries.
Sourcing
Unlike obsidians and other volcanic materials which are locally
homogeneous within each eruption event, the composition of individual silicified
sandstone and chert outcrops are more heterogeneous in macroscopic properties and
chemical composition as a result of the diagenesis processes that created them. This
makes sourcing artifacts and debitage composed of these materials more difficult.
Traditionally, comparison of the crystal structure using thin section petrology has
been the method used to attempt to source quartzites, silicified sandstone, and cherts.
64


Use of ultraviolet (UV) fluorescence has also been successful in some cases for
sourcing cherts (Luedtke 1978, 1979; Lyons, et al. 2003) and recently Pitblado
(2007) has had success using ICP-MS and LA-1CP-MS to characterize quartzites
from quarries in the Gunnison, Colorado area, about 90 km east of this research area.
Three methods of sourcing materials were attempted in this work;
macroscopic comparison of color, texture (grain size), and inclusions using direct
comparison with representative hand-samples; short and long wave UV fluorescence;
elemental spectra analysis by laser induced breakdown spectroscopy (LIBS).
Macroscopic comparison was used to initially source virtually all the artifacts and
debitage, and fluorescence was utilized on quarry samples and samples from lithic
scatter sites. LIBS is an experimental method that required extensive procedural and
analytic development for use in this research. Although the plan had been to use it on
material from all quarries, artifacts, and debitage, hardware problems limited its use
to some quarry samples and some artifacts.
Macroscopic comparison
Because the outcrops of the silicified sandstone that composed all the
quarries identified in this study area were assumed to be heterogeneous, a minimum
of fifty hand-samples were collected from each of the nine quarries. These hand-
samples were typically debitage left from the quarrying activity. A criterion for
selecting a hand-sample was, that whenever possible, at least one face was a scar
from reduction or testing by the prehistoric people who utilized the quarry. An
65


attempt was made to collect hand-samples with multiple flake scars since that
indicated the material was most likely utilized by the prehistoric peoples rather than
discarded after testing.
Two procedures were used to select hand-samples. Hand-samples were
selected that represented the spatial variability within the quarry as well as the
macroscopic variability. Because diagenesis results in differences in composition
across the ore body, hand-samples were collected across the length and breadth of
each quarry in an attempt to obtain samples that fully represented the spatial
variation of the quarry material. Since these hand-samples were to be used for
macroscopic comparison as well as chemical composition comparison, it was
important that all the variability in color, grain textures, and inclusions present in the
quarry also be represented in the collected samples.
From all the hand-samples collected, a smaller subset of hand-samples
representative of the macroscopic variability from each quarry, particularly from the
Flint Cave and 5MN7753, was created. This subset of hand-samples was taken to the
Anasazi Heritage Center to use in initial, macroscopic sourcing of the artifacts and
debitage from the excavated sites. A larger subset, more representative of all the
macroscopic variations, was used in the lab for sourcing of artifacts and debitage
collected from the sampled lithic scatter sites in the research area. Only if an artifact
or piece of debitage matched the color, grain, and inclusions of a hand-sample was it
sourced to that quarry and included in the analysis of this research. It is the belief of
66


the author that this provided a high confidence, greater than 80%, sourcing of the
material used in the analysis.
Ultraviolet Fluorescence
Typically silica, the major component of most of the lithic material in the
research area, is not fluorescent but some impurities that may be associated do
demonstrate fluorescence. To be useful for sourcing lithic material to a quarry the
usable silicified sandstone and cherts had to be fluorescent, not just carbonates and
calcites or other minerals that might be part of the composition of the cortex of the
material.
The hand-samples collected from each quarry were examined using both long
and short ultraviolet light to determine if any of the lithic materials from the quarries
within the research area were fluorescent. The quarry, sample number, type of
material (chert, chalcedony, silicified sandstone), inclusions or bulk material that
fluoresced, visible color, and fluorescence color were logged for each sample that
did indicate some fluorescence. While many of the samples showed some
fluorescence, the fluorescence was not distinctive or consistent enough between
quarries to serve as a discriminate.
Laser Induced Breakdown Spectroscopy (LIBS)
A new, experimental method of sourcing material was developed as part of
this research. This method, laser induced breakdown spectroscopy (LIBS), uses a
high powered Nd:Yittrium (YAG) laser to vaporize a small amount of the material to
67


be tested. Those atomic elements composing the vaporized material, plasma, emit
light characteristic of each element. That light is then spectrally decomposed using a
spectrometer for the ultraviolet wavelength range (200 nm to 400 nm) to produce a
UV spectra of the material (Figure 4.2). The relative intensities of the spectral lines
are proportional to the quantity of the elements present in the sample. Those
proportions of elements, particularly trace elements, associated with the sample are
often uniquely characteristic of the materials source. The material composition
obtained by LIBS is similar to that obtained from mass spectroscopy (MS)
techniques commonly used in 1NAA, LA-ICP-MS, and ICP-MS. The advantage of
LIBS is that it is essentially non-destructive since it vaporizes only a pinpoint
amount of material in each laser shot. On most of the silicified sandstone,
chalcedony, and chert samples tested in this research using LIBS, the locations of the
laser shots were not discernible with the naked eye. This allows artifacts to be
sourced without destroying them or removing significant pieces.
Since silicified sandstone, chalcedony, and chert materials are heterogeneous,
a spectra from a single sample of the material is not sufficient to characterize the
source quarry. For this research a minimum of thirty spectra were obtained from
several locations on hand-samples after cleaning the surface of weather altered
material at each location using a minimum of five laser shots. The spectra from each
hand-sample were averaged and spectral lines and their intensities determined.
Several methods of analyzing the spectral data from the quarry samples were tested
68


by the author including probability of spectra occurrence, Baysian probability, and
root-mean-square deviation from spectral vectors. These were applied using both
averaged spectra from single hand-samples from a quarry and averaged spectra of
many hand-samples from a quarry. The results from these methods were compared
with those obtained using discriminate analysis. Discriminate analysis using
averaged spectra from many hand-samples provided the best results. Using averaged
spectra from seven hand-samples from each of four of the nine quarries, Flint Cave
(5MN7429), 5MN7753, 5MN7754, 5MN4638, the material from the quarries were
discriminated from each other with better than 90% success using leave one out
discriminate analysis in SPSS 15.0.
Wavelength, nm
Figure 4.2 LIBS UV Spectra of Flint Cave Material
69


To source artifacts and debitage using LIBS, a minimum of forty spectra was
collected from at least two locations on the artifact or debitage. The first five spectra
at each location were discarded so that surface contamination and effects due to
surface weathering were not included in the analysis. The resulting spectra were
averaged and spectral peaks identified. The Fischer coefficients obtained from the
discriminate analysis of the spectra from the four quarries discussed above were then
used to determine the most likely quarry source. Using Fischer coefficients for
sourcing will result in all samples being associated with one of the quarries in the
training case. It does not recognize the possibility of an unknown source. To allow
for unknown sources, which undoubtedly existed for artifacts and debitage in this
research area, a maximum distance of association (MDOA) between a sample and a
quarry in the component space was defined. This maximum distance was equal to
1.5 times the largest distance between quarry centroids in the discriminate analysis
component space. Any sample with a distance from all quarries greater than the
MDOA was sourced as unknown.
The LIBS system was made available for this research by ATSAA, a small
company applying advanced technologies to archaeology. Hardware problems with
the LIBS system caused large variability in the spectral intensities which required
significant preprocessing of the data and ultimately stopped collection of LIBS data
in support of this research. Only four of the nine quarries were characterized and
some materials from the excavated sites were the only artifacts sourced using the
70


LIBS system. The hardware problems were corrected but not in time to allow further
use of LIBS in this research.
Lithic Material Data
The data used in evaluating utilization and distribution of lithic material from
the Flint Cave and 5MN7753 as well as the other quarries were obtained by four
different methods: samples from lithic scatter sites, isolated finds, site survey, and
previously excavated sites from within the research area.
Surface Lithic Scatter Sites
Lithic material was collected and analyzed from sixty-eight sample areas
(Figure 4.3). Three of these sites had ceramics and two of those sites had Flint Cave
material even though ceramics occurred in the area after the latest Flint Cave activity
dated at 1940 B.P. in the chronology study of the Flint Cave (Richey and Hauser
2007). These occurrences could have been due to deflation of the sites, use of
debitage or discarded artifacts from an earlier period, mining activity in the Main
Room of the Flint Cave which was not dated by Richey and Hauser (2007), or
material extracted from narrower portions of the same vein which do show evidence
of utilization.
The criteria used whenever possible in selecting sample areas were: 1) obtain
lithic material from sites at a range of distances and directions from the Flint Cave
and 5MN7753 quarries, 2) obtain enough pieces of lithic material from each sampled
area to provide a measure of materials composing 10 percent or more of the sample
71


that was not sensitive to small differences (+/- 2 pieces or less) in the number of
pieces of that lithic material, 3) the sample area should be representative of the
materials present in the site and not contain an obvious single chipping station, 4) the
lithic material density within the site had to allow for collection of at least 100 pieces
of material, and 5) an area no smaller than a circle of 1 m diameter would be
collected. However, at higher elevations the size of lithic scatter sites was found to
be generally smaller forcing samples with less than 100 pieces to be used in some
cases in order to obtain samples from those areas.
Legend

Sample d Lilh ic Areas
Project Area Elevation
Value
Ugh 2946 m


Low 1624 m
0 5 10
20 Wlometers
Figure 4.3 Locations of Sample Areas within Research Area
72


The lithic scatter sites large enough to meet the sample area criteria above
were primarily along ridges or near rims of canyons or draws and occasionally
occurred in blow-out areas in the sagebrush flats. A problem inherent to using
surface sites is the lack of chronological control. The fact that surface sites occurred
in deflated areas implies that materials from these surface sites could be a mixture of
materials from significantly different time periods. It is the assumption of this
research that if the Flint Cave was being utilized, i.e. Flint Cave material occurred in
the site, the economic choices influencing the procurement and utilization of Flint
Cave and 5MN7753 materials were the same. However, it is also possible that
material from 5MN7753 was used before and after the Flint Cave mine was
producing material which could inflate the proportion of 5MN7753 material in
deflated sites and artifacts made of that material occurring as isolated finds.
Since the collection of material at a range of distances and directions from the
Flint Cave and 5MN7753 quarries was the driving requirement for sampling, each
search area, typically a section (lmi x 1 mi), was walked until a lithic scatter area or
site with sufficient lithic material density was located. A scatter area was defined as
an area with at least 10 pieces of lithic material showing utilization within a 5 m x 5
m area. The site was examined and a sample area within the site was chosen that had
a representative density of material and did not contain an obvious single chipping
station. This later criterion was formulated to obtain the best measure of materials
present in the site and to minimize biasing material representation due to collection
73


of a single flaking station. Once a sample area that appeared to have sufficient
density was identified, all lithic material from the surface of that sample area was
collected for further analysis using a constant radius circle (dog lease approach) to
define the area. The size of the sample area and the datum (UTM or Lat, Long) were
recorded for each sample area. Generally one sample area was collected from a
given site. In two cases, two sample areas were collected from a single site to get an
indication of how much variation might be expected between samples within a site.
Of course that required the sites to have relatively high flake density or be large to
fulfill the flake count criteria for two samples.
The lithic materials collected from each of the sample areas were then
separated using macroscopic characteristics of color, texture, and inclusions into
possible different materials. The number of possible different materials in the
samples ranged from fifteen to forty-five.
The size of each piece of debitage assigned to each of the different material
groups was measured. Size was determined using the smallest circumscribing circle
method. The range of sizes of circle templates used (3.5, 3, 2.75, 2.5, 2.25,2, 1.875,
1.75, 1.625, 1.5, 1.375, 1.25, 1.125, 1, 0.9375, 0.875, 0.8125, 0.75, 0.6875, 0.625,
0.5625, 0.5, 0.4375, and 0.375 in.) were the same for all samples and ranged from
3.89 cm (3.5 in) diameter to 0.9525 cm (0.375 in) diameter.
Macroscopic visual comparison of the hand-samples from the quarries was
then used to determine which of the different material groups was likely associated
74


with one of the quarries of interest. The association of quarries with debitage was
made at two levels of confidence, high confidence and possible. The possible
level of association indicated that the debitage could be a combination of the
variations of material seen at one of the nine quarries identified in this research but
was not a match to any of the hand-samples. The high confidence level was given
to those samples that were good macroscopic matches in color, grain texture, and
inclusions to hand-samples from one of the quarries. Only the high confidence
associations were used in the analysis presented here. Once the pieces of debitage in
each sample area were associated with quarries of interest, the corresponding
percentages of the total number of pieces of debitage in the sample were calculated
for each of the quarries of interest.
Patterson (1990) found that a log-linear relationship between the proportion
of flake size in an assemblage and flake size could be used to determine type of
reduction, i.e., bifacial or core, and that this method was insensitive to flake
fragments and angular pieces included in the measurements. Using this approach, the
type of reduction associated with each material that had been sourced to a quarry was
determined. The size distributions of these materials were also used to inform the
level of reduction and retouch that was occurring. In this case, flake fragments
included in the distributions likely skewed the size distribution to smaller sizes, but it
was not expected to be large enough to cause misinterpretation.
75


Isolated Finds
Artifacts found as isolated finds during the search for lithic scatter sites were
photographed, sourced using macroscopic characteristics, and the location recorded.
Only diagnostic projectile points and ceramics were collected for curation at the
direction of the Bureau of Land Management (BLM) Montrose District
archaeologists (Julie Coleman and Dr. Glade Hadden). One occurrence of ceramics,
Ute micaceous brown ware, and thirty-eight projectile points were curated as isolated
finds out of a total of 325 isolated finds documented in this work.
The artifacts sourced to the quarries in the research area were used to inform
both the movement of the prehistoric people that utilized the quarries and correlation
between tool types and materials. The typology of the projectile points associated
with Flint Cave material informed the periods when that material was utilized.
Transect Sampling
Material from both the Flint Cave and 5MN7753 were reduced on the rim of
Dry Creek above the Flint Cave. No reduction area is evident directly above
5MN7753, but material from 5MN7753 does occur in the reduction debitage above
the Flint Cave. There also are large pieces of material from 5MN7753 evident along
a trail that connects 5MN7753 with the reduction area above the Flint Cave.
Therefore, it appears that material from 5MN7753 was carried from 5MN7753 to the
area above the Flint Cave for reduction.
76


To obtain a statistical comparison of the level of reduction of material from
the Flint Cave and 5MN7753 a systematic sampling strategy was used in the
reduction area common to both quarries. Three transects that ran perpendicular to the
rim of the canyon, roughly on an E-W line, and were separated by 15 meters were
sampled every 15 meters along their length (Figure 4.4).
Figure 4.4 Orientation of Sampled Transects in Reduction Area
The samples consisted of all debitage within a 1 meter diameter circle
centered every 15 meters along each transect. The length of each transect ran from
the rim of Dry Creek Canyon until the density of debitage in the 1 meter samples
was less than 5 pieces for four consecutive samples. This latter criteria was to avoid
77


premature termination of a transect due to bare rock outcrops that occurred within
the reduction area.
The debitage from each sample area was analyzed in the same manner as
described above for samples from lithic scatter areas. All pieces of debitage from
each sample were grouped according to material and sized using circumscribing
circles as described above. Material was sources as Flint Cave, 5MN7753, or other.
Survey
A survey of tools was accomplished in the reduction area adjacent to both the
Flint Cave and 5MN7753 quarries with the help of members of the Chipeta Chapter
of the Colorado Archaeological Society. This area has a high density of debitage so
examining all the flakes for expedient use was not attempted. The survey identified,
photographed, and recorded location, tool type, and probable source of the material
for all formal tools, e.g. side and end scrapers, bifaces, projectile points, drills,
hammer stones, and formal cores, regardless of size and all informal cores and
utilized edges on flakes larger than 5 cm in maximum dimension. The probable
source of material for each tool was assessed in the field, by the author, using
macroscopic properties only. In the lab, the tools were compared to the hand-samples
from the quarries to further inform the source of the materials.
Excavated Sites
Excavated sites were used as a source of data to further inform utilization and
chronology of Flint Cave material and to a lesser extent 5MN7753 lithic material.
78


The chronology was informed using the limited number of l4C dates associated with
the excavated sites as well as projectile point typology.
Data was obtained by examining lithic materials from nine excavated sites
from within the research area (5MN17, 5MN33, 5MN34, 5MN35, 5MN38, 5MN40,
5MN57, 5MN2628, 5MN3859) curated at the Anasazi Heritage Center, Dolores,
Colorado. Seven of the sites were excavated by Buckles (1971) and had curated
artifacts and utilized flakes but only a small amount of the debitage. Therefore,
insight into the use of Flint Cave and 5MN7753 material at these sites was limited.
The other two sites (5MN2628, 5MN3859) were excavated by Alpine
Archaeological Consultants, Inc. (Reed, et al. 2001) and had curated all materials
including the debitage. Visual examination of the debitage from these sites indicated
a very low occurrence of possible Flint Cave material in 5MN3859, near Coal Bank
Canyon, and possibly large contribution in 5MN2628 located at the very west edge
of the research area. LIBS spectra were taken on debitage from these sites for
sourcing but could not be used due to the problems discussed above. Detailed
analysis of the debitage such as that done on the samples from the lithic scatter areas
was not possible due to time constraints. Further analysis using the materials from
5MN2628 might inform how Flint Cave material was being utilized at this high
altitude site.
79


Cost Models
Cost models associated with moving across the terrain of the research area
from each quarry were calculated using ArcGIS 9.2. Relative cost is calculated in
ArcGIS 9.2 based on the change in slope of the terrain and straight line distance of
those slope contours on the terrain from a point of interest. The units of cost were
arbitrary but the same for each quarry so comparisons of costs associated with
material mobility from different quarries could be compared.
For this research, the contours of slope were created from digital elevation
model (DEM) data obtained for the non-agricultural land in the research area. The
DEM data used in this research had grid points at 10 meter resolution. Each quarry
was used as the point of interest to create a cost model for that quarry, and contours
of equal relative cost were created for each quarry (Figure 4.5-4.10). The cost
models were used to compare the cost of mobility to the distribution of material for
material from each quarry.
80


0 3 6 12 Kilometers
I_i_i_j_L_i_i_i_J
Figure 4.5 Cost Contours from Flint Cave Quarry (5MN7429)
0 3 6 12 Kilometers
L_i__i_i_L_i__i_il
Figure 4.6 Cost Contours for Quarry 5MN7753
81


w
0 3 6 12 Kilometers
1 i i i__L_i__i i I
Figure 4.7 Cost Contours for Quarry 5MN7754/55/56
0 25 5 10 Kilometers
1 i i i I i i I
Figure 4.8 Cost Contours for Quarry 5MN4638
82


Figure 4.9 Cost Contours for Quarry 5MN8384 and 5MN8385
0 25 5 10 Kilometers
1 i i i i i i l
Figure 4.10 Cost Contours for Quarry 5MN8386
83


Research Parameters
The data collection, sourcing, and cost models described above were used to
inform the parameters defined to test the hypotheses of this research. How these data
were analyzed to inform each parameter is discussed below.
Utilization (RP1 and RP2)
The parameters RP1 and RP2 relate tool classes, formal and expedient, to
quarry material source, Flint Cave (5MN7429) and 5MN7753 respectively, and
could have values of informal tools, formal tools, or both. Several analyses
were performed to inform RP1 and RP2.
A chi square of material source and formal and informal tool classes was
calculated to determine if there was significant association between the Flint Cave or
5MN7753 material and formal and expedient tool classes. The data obtained from
the survey informed these parameters for the major reduction area located near both
Flint Cave and 5MN7753 quarries. The isolated finds data and those from the
excavated sites provided information on tool type versus material over the larger
research area.
Supply Zone (RP3)
The supply zone is that area around the quarry where direct procurement
occurred from campsites. The maximum cost contour of the supply zone, RP3, for
the Flint Cave compared to the nCAC (nominal contour of acceptable cost) was
informed by examining the relationship between the percentages of the materials
84


from the Flint Cave, 5MN7753, 5MN7755, and 5MN7756 that occurred in the
samples collected from the lithic scatter sites and the cost models for those quarries.
This analysis was limited to those quarries because only material from those four
quarries was represented in enough of the samples to perform this analysis.
Since the absolute cost value is not important for comparisons between
quarries, the supply zone for a quarry was assumed to occur within the cost contour
where the contribution of material from that quarry was greater than 50% of its
highest contribution by weight. The number of cost contours between the quarry and
the cost contour where the percent contribution was reduced to 50% of its peak value
was considered the maximum cost for direct procurement for that material or the
CAC. The maximum contour of acceptable cost (CAC) for the Flint Cave material
was compared to the average or nominal cost contour of the other quarries, the
nCAC, to inform RP3. Although cost contours and distance from the quarries are
related in this analysis by a weighting due to the slope of the terrain, this same
analysis was also done using distance from the quarries rather than the cost contours.
Reduction (RP4)
The level of reduction for both Flint Cave and 5MN7753 close to the sources,
RP4, was informed using data derived from the samples from the lithic scatter sites.
The level of reduction was measured by calculating the ratio of the amount of
debitage with dimension larger than or equal to 2.54 cm to the amount of debitage
with dimension less than 2.54 cm for materials from the Flint Cave and 5MN7753 in
85


each sampled lithic scatter site. Smaller ratios, i.e., more large pieces relative to
small pieces, were interpreted to indicate less reduction before material was utilized
or removed from the sites and larger ratios were interpreted as indicating higher
levels of reduction.
The ratios from the samples were compared to ratios obtained from debitage
resulting from the creation of biface cores and Clovis points made of chert and
obsidian by Bob Patten and Jeff Ferguson respectively, two knappers in the Denver
area. The artifacts were considered to be a good representation of the level of
reduction that would have occurred if the material was being prepared for long
distant transport. Utilization of larger pieces of debitage for tools and subsequent
removal from the assemblage was accounted for by applying an empirically derived
logarithmic reduction equation developed by the author before taking the ratios (see
Appendix A). From this analysis ratios larger than six were considered higher levels
of biface reduction (see Appendix A). Other types of reduction such as unifacial
blade or flake reduction were not modeled in this project.
The ratios for material from the two quarries were compared for all samples
taken within the supply zone of the quarries. The significance of the differences of
the means of these ratios for the two quarries was tested using the t-test to determine
if one material had a significantly higher level of reduction than the other. The same
debitage ratio was calculated for both the Flint Cave and 5MN7753 using the total of
the debitage obtained just from the sampled transects in the reduction area and
86


sourced to each of the quarries. Chi square was used to determine if they were
significantly different. As an additional measure of the amount of bifacial reduction
in the reduction area, Pattersons (1990) log-linear relationship between flake size
and percentage of the total number of flakes was applied separately to debitage of
Flint Cave material and 5MN7753 material obtained from transect sampling. The
more linear the relationship the more bifacial reduction occurred. The more non-
linear the relationships the more core reduction is indicated.
Distance Decay (RP5)
The distance materials from the Flint Cave and 5MN7753 moved from their
quarries, RP5, was determined from the samples of lithic scatter sites. This was
addressed from two perspectives. The percentage of the number of pieces of debitage
of each sourced material to the total number of pieces of debitage in each sampled
lithic scatter site was used to compare the distributions of the materials as a function
of distance and cost from their sources. This perspective assumes that each material
was being used in a similar way and that number of pieces of debitage is a measure
of the amount of each material present and could indicate curation of that material,
i.e., biface reduction and retouch. A comparison of the weight of each material
present in the samples was also accomplished (See Appendix B). Since weight is
proportional to the cost of transport, the more material by weight present the more
cost incurred in transporting that material and potentially the more valuable the
87


material. If one material or the other is consistently represented by more weight it is
possible that one material had more value than the other.
To compare the distributions of materials from the Flint Cave and 5MN7753,
the ratio of the average number of pieces of debitage of the two materials and the
ratio of average weight of the two materials were calculated as a function of both
cost contour and distance. If the two sources were declining at the same rate, the
ratios should stay constant. If the amount of one material decreased faster than the
other, then the ratio will either increase or decrease as a function of cost and
distance. In this analysis the ratio was Flint Cave material/5MN7753 material so a
decrease in the ratio was an indication that Flint Cave material was decreasing faster
than 5MN7753 material and an increase in the ratio was an indication that 5MN7753
material was decreasing faster than Flint Cave material.
Functional Form (RP6)
The functional form of the frequency of occurrence of the materials with
distance and cost contour, RP6, was determined using data from the samples of the
lithic scatter areas. The average percent contribution of material from Flint Cave and
5MN7753 to the assemblage from all sample within a distance band of 2 km or a
cost contour were plotted. Those data for each material were fit with linear,
2
quadratic, power, log, and exponential functions. The function with the highest R
above 0.70 was interpreted as being the best representation of the decay of the
material with distance. If functional forms other than those discussed in Chapter 2
88


were the best fit, this indicated the possibility of an alternative relationship to that
predicted by theory.
Only the debitage that had high confidence of association with hand-samples
from one of the quarries were used in the analysis to inform the research parameters
of this project. The results of the analysis that informed each of the research
parameters are presented and discussed in the following chapter.
89