Rock art and sacredness of place in the landscape of west-central Colorado as identified by visual and acoustic indicators of prehistoric ritual behavior

Material Information

Rock art and sacredness of place in the landscape of west-central Colorado as identified by visual and acoustic indicators of prehistoric ritual behavior
Williams, Gregory E
Publication Date:
Physical Description:
xi, 227 leaves : illustrations ; 28 cm


Subjects / Keywords:
Petroglyphs -- Colorado -- Uncompahgre Plateau ( lcsh )
Rock paintings -- Colorado -- Uncompahgre Plateau ( lcsh )
Indians of North America -- Colorado -- Uncompahgre Plateau ( lcsh )
Sacred space -- History -- Colorado -- Uncompahgre Plateau ( lcsh )
Antiquities ( fast )
Indians of North America ( fast )
Petroglyphs ( fast )
Rock paintings ( fast )
Sacred space ( fast )
Antiquities -- Uncompahgre Plateau (Colo.) ( lcsh )
Colorado -- Uncompahgre Plateau ( fast )
History. ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )
History ( fast )


Includes bibliographical references (leaves 221-227).
General Note:
Department of Anthropology
Statement of Responsibility:
by Gregory E. Williams.

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:
519534737 ( OCLC )
LD1193.L43 2009m W54 ( lcc )

Full Text
Gregory E. Williams
B.A., University of Northern Colorado, 1979
M.B.A., University of Colorado at Denver, 1994
A thesis submitted to the
University of Colorado Denver
in partial fulfillment
of the requirements for the degree of
Master of Arts

This thesis for the Master of Arts
degree by
Gregory E. Williams
has been approved by
Christopher S. Beekman
Carol Patterson

Williams, Gregory E.
Rock Art and Sacredness of Place in the Landscape of West-central Colorado as
Identified by Visual and Acoustic Indicators of Prehistoric Ritual Behavior
Thesis directed by Professor Tammy Stone
This thesis examines the relationship between the rock art of the Colorados
Uncompahgre Plateau and the seven characteristics of human ritual behavior as
defined by Roy Rappaport (1999). Until recently most rock art research has
focused on identifying and documenting sites and motif styles. Few attempts to
interpret rock art in terms of its relationship to the physical and cultural landscape
have been initiated. This research project employed a landscape approach that
measured numerous variables relating to the context, style, method of production,
local landforms, and the acoustical properties of 22 rock art panel locations in west-
central Colorado. Using Principal Component Factor Analysis on a dichotomized
dataset this research project identified several variables for each of Rappaports
seven characteristics of ritual which together explain a large degree of the variance
in the available data from the study area. The research project established a link
between rock art and ritual and identified several new approaches to rock art
research, including the measurement of the acoustical properties of rock art panel
locations. The results have potential applications for future rock art research and
for the management of the cultural resources in the study area and elsewhere.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.

This project is dedicated to my remarkable assemblage of friends, co-workers, and
family, my communitas in the words of Victor Turner. These are the people who
inspire me, cajole me, reassure me, and who patiently endure me while making life
so wonderfully enjoyable. I hope to someday return the favor.

Special thanks and acknowledgement are due to the following people who
have provided advice, financial support, insight, time, and constructive criticism
relating to my graduate program and research project. Any flaws, errors, or
omissions are mine alone. In alphabetical order I extend my deepest appreciation to
the Colorado Archaeological Society (Alice Hamilton Scholarship Fund), the
Colorado Council of Professional Archaeologists (Ward F. Weakly Memorial
Fund), the Colorado State University (Karen S. Greiner Endowment), Christopher
S. Beekman, Christopher Bevilacqua, Kevin Black, Richard Bond, Jean Clottes,
Larry Cunningham, Glade Hadden, Ed Horton, Brian ONeil, Carol Patterson, Brue
Rippeteau, June Ross, Rich Sanders, Tammy Stone, Steve Waller, and Alan

LIST OF FIGURES............................................viii
LIST OF TABLES................................................x
CHAPTER 1.....................................................1
CHAPTER 2.....................................................8
THEORY AND BACKGROUND.......................................8
Nearly a Century and a Half of Rock Art Research..........8
Rappaport on Ritual......................................28
Central Australia........................................33
Human Experience and Cognition...........................39
Archaeoacoustics and Sound...............................48
CHAPTER 3....................................................56
THE UNCOMPAHGRE PLATEAU....................................56
Physiography and Geomorphology...........................56
Climate and Environment..................................62

Prehistoric and Modern Inhabitants........................70
Present Environment and Threats...........................78
CHAPTER 4....................................................81
Fieldwork and Sampling....................................81
Data Attributes...........................................87
CHAPTER 5...................................................145
CHAPTER 6...................................................176
Specific Conclusions.....................................178
General Conclusions......................................188
APPENDIX A..................................................199
APPENDIX B..................................................202
APPENDIX C..................................................219
REFERENCES CITED............................................221

Figure 3-1 Colorado Plateau................................... 57
Figure 3-2 Confluence of Gunnison River and Escalante Creek... 63
Figure 3-3 Uncompahgre Plateau ............................... 64
Figure 3-4 Vegetation Zones .................................. 66
Figure 4-1 Study Area......................................... 85
Figure 4-2 Roc Creek Establishing Photo....................... 88
Figure 4-3 Roc Creek Viewshed Photo .......................... 89
Figure 4-4 Pecked Images on Several Orientations ............. 96
Figure 4-5 Rock Art Styles ...................................... 102
Figure 4-6 Repeated Imagery ..................................... Ill
Figure 4-7 Abstract and Geometric Imagery ....................... 112
Figure 4-8 Depth Measurement Tool ............................... 115
Figure 4-9 Eagles Nest Site Image .............................. 117
Figure 4-10 Measuring Motif Depth ............................... 118
Figure 4-11 Drawing of Depth Measurements ....................... 118
Figure 4-12 Sound Recorder and Noise Maker ...................... 121
Figure 4-13 Sonograms of Echo Patterns .......................... 124
Figure 4-14 Sound Recording at Eagles Nest Site ................ 127

Figure 4-15 Plan View Drawing .................................... 130
Figure 4-16 Profile View Drawing ................................. 132

Table 2-1 Defined Terms............................................. 10
Table 2-2 Comparison Between Rappaports Seven Features of Ritual
and Ross and Davidsons Field Categories............................ 35
Table 3-1 Geologic Formations of the Study Area.......................... 58
Table 4-1 Expected Associations Between Datasets......................... 93
Table 4-2 Data Types ................................................... 135
Table 5-1 Invariance Components ........................................ 149
Table 5-2 Invariance Cross Tabulations ................................. 150
Table 5-3 Repetition Components ........................................ 153
Table 5-4 Repetition Cross Tabulations ................................. 154
Table 5-5 Special Time Components ...................................... 157
Table 5-6 Special Time Cross Tabulations ............................... 158
Table 5-7 Stylized Form Components ..................................... 160
Table 5-8 Stylized Form Cross Tabulations .............................. 161
Table 5-9 Performance Components ....................................... 164
Table 5-10 Performance Cross Tabulations ............................... 165
Table 5-11 Canonical Messaging Components .............................. 168
Table 5-12 Canonical Messaging Cross Tabulations ....................... 169

Table 5-13 Scree Chart.................................................. 171
Table 5-14Special Place Components ..................................... 173
Table 5-14 Special Place Cross Tabulations ............................. 175
Table 6-1 Final Results By Ritual Characteristic ....................... 179
Table 6-2 Comparisons of Expectations with Results ..................... 187

This thesis addresses one primary and one secondary research question:
Primary Question: What is the structure of prehistoric
ritual behavior in west-central Colorado as demonstrable
in the rock art assemblage?
Secondary Question: What characteristics, if any, of the
rock art assemblage serve as reliable indicators of
prehistoric ritual behavior?
Ritual behavior is one of the few universals that is generally undisputed as a
defining attribute of the human experience (Rappaport 1999, Turner 2009 [1969]).
The purpose of this research project is to link human ritual behavior to rock art.
Ancillary to this, but equally important, will be the establishment of rock art
as something other than a resource that is archaeologically and socially significant
only for its iconography, whether it is decipherable or interpretable by epigraphy or
not. The significance of rock art is derived from its placement in the landscape
which is a direct result of the interaction of prehistoric human cultures with their
physical environment. This, in short, is about how humans make a sacred place out

of a natural space which is a topic that has fascinated anthropologists and
archaeologists for a long time.
Rock art is not of interest just because of the beautiful and enigmatic images
that convey a sense of mystery and antiquity to the viewer. This thesis will expand
the basis for understanding at least some rock art as an indicator of very important
human social activity and interaction, both in groups and individually, in a micro- as
well as a macro-geographic context. This will also counterbalance, to some extent,
the current focus and interest in the iconography or meaning of the individual motifs
and thereby establish the setting of rock art as a relevant and therefore scientifically
and managerially important aspect of the resource itself.
In so doing this project emphasizes the importance of understanding rock art
in its current Native American cultural context. This point is important for two
reasons. First, as Jean Clottes said Here you have the stories behind much of the
rock art in the Uncompahgre Plateau (personal communication 2007). Second,
although the cosmology and myths of present day Native Americans may be
claimed by them as represented in rock art for an era that current archaeological and
physical evidence suggests pre-dates their occupation or their presence in the area
this does not diminish or call into question the spiritual significance of the place as
it is re-integrated into subsequent human activity by more recent cultural groups.

This research approach provides an additional basis for understanding why, and
how, this phenomenon occurs.
Some may say so in this way you dont care about the images themselves
just that they are there and to that the response is a resounding no. When the
images can be linked to a story of human experience they will always be more
interesting to the researcher and to the society in which the researcher does his or
her work. What this project accomplishes is to establish that even without the story
we will know in our minds what we feel in our hearts when we see the magnificent
images; that they were created and placed where we find them for a special reason.
My thesis presumes that all the rock art was created by and for humans.
This is an entirely appropriate yet western Cartesian presupposition and may seem
to be an odd thing to state. It is necessary for scientific hypothetical-deductive
reasoning but it does not in any way suggest that alternative Native American
explanations that some rock art was created by spirits or non-human entities are
invalid. As Ian Hodder said many times and in many ways; it is important to
consider multiple voices and local voices when interpreting prehistory. By utilizing
the western scientific method, by a westerner, to a western audience, in order to
attempt to understand something that is, by its very existence non-western, this
thesis expands our understanding and appreciation of rock art and its place in the
spatial and cognitive landscape of human activity.

This approach will also aid the current managers of cultural resources as
they execute their statutory and regulatory duties as custodians of our public lands.
Finally it will provide an additional approach for consideration when management
decisions are made concerning the establishment and designation of Traditional
Cultural Properties which are of vital contemporary interest to Native Americans.
Chapter 2 (Theory and Background) provides a theoretical basis for the
research project and a background of previous research in rock art. A hundred years
of rock art research are summarized; ritual and the sacred in human behavior are
discussed with particular emphasis on the work of Roy Rappaport and Victor
Turner. Emile Durkheims contribution to their views is addressed, thanks in large
part to a recent (and very readable) translation of The Elementary Forms of
Religious Life by Karen E. Fields. Rappaports seven characteristics of ritual
activity are introduced and the application of these characteristics in an
archaeological setting relating to rock art by June Ross and Ian Davidson is
presented. Rappaports characteristics of ritual are cross-linked to Ross and
Davidsons categories in narrative and chart form. Human cognition from a
neurological perspective is discussed and related to the neuropsychological model
of formal analytical rock art research presented by David Whitley and to the work of
neuroscientist and psychologist Antonio Damasio (specifically the somatic-marker
hypothesis). Finally archaeoacoustics and sound as part of the political aesthetic

of archaeologist Adam T. Smith is presented and discussed with particular emphasis
on the work of Steve Waller in Horseshoe Canyon, Utah and by Chris Scarre in his
recent book entitled Archaeoacoustics. This section sets the groundwork for the
research methods and variables to be tested and which are presented in Chapter 4
Chapter 3 (Study Area) provides an introduction to the Uncompahgre
Plateau study area and its relationship to the Colorado Plateau and Eastern Great
Basin. The physiography and geomorphology, climate and environment, hydrology,
flora and fauna are discussed. An overview of what is known about the
paleoenvironment is presented as is an overview of the prehistoric and modem
inhabitants from 11500 B.C. to the present. A brief summary of the Utes is
presented. This chapter concludes with a discussion of the present environment,
anthropogenic modifications, and threats to the prehistoric resources; particularly to
the rock art.
Chapter 4 (Methods) addresses the sampling procedure, datasets (variables),
field recording methods, and methods of data analysis employed in the research
project. Justifications linking the identified variables to one or more of each of the
seven characteristics of ritual (identified in Chapter 2) are presented and the reasons
for discarding certain variables are enumerated. This section includes a map of the
panel locations within the study area, a list of anticipated results for the research

project is presented in narrative and table form, field results are presented and an
explanation of the procedure for dichotomizing the variables is put forth. Methods
of conducting special data collection (including motif depth and the acoustical
characteristics of the panel locations) are detailed. This section addresses the
validation of the coded data for the 3,190 dichotomized values (in 22 panel
locations). Various rock art styles are presented, different forms of echo readings
are presented and discussed, and photographs of several rock art panels illustrating
the characteristics of ritual behavior that they represent (through the variables
tested) are introduced into the record.
Chapter 5 (Results) reviews the data elements (variables) and presents them
in the form of a brief narrative, frequency distributions, factor analysis results, cross
tabulations, and a detailed matrix listing each of the variables tested for each of the
seven characteristics of prehistoric ritual behavior. This chapter compares the
expected results (from the frequency distributions) with the achieved results (from
the factor analysis) and establishes logical and plausible links (from Chapter 5 on
Methods and Chapter 2 on Theory and Background) to each of the handful of key
variables that are statistically associated with each of Rappaports seven
characteristics of ritual behavior based on the dichotomized values established for
each of the variables tested.

Chapter 6 (Conclusion) establishes how the primary and secondary research
questions are answered in narrative and table format. Specific and general
conclusions are reached, shortfalls of the research project are identified, and
suggestions for future research are presented. This chapter also places the results in
a context relative to the preceding chapters in both a methodological and theoretical

Nearly a Century and a Half of Rock Art Research
Rock art has a peculiar saga in the western world. Historically westerners
have stared directly at the images and failed to make note them. It was as if the rock
art was invisible. It was as the rock art did not exist. When rock art was first
brought to the attention of science its authenticity was hotly disputed and it has only
been in the past half century or so that the incredible antiquity of some of the Iberian
Peninsula cave paintings, particularly those recently discovered at Lascaux and
Chauvet has been recognized. This imagery is often cited as an early example of
human symbolic communication and this makes it important by itself. It is,
however, not the oldest such example. The engravings on wood and ochre at the
Blombos cave and elsewhere have been dated to 100,000 BP and possibly much
older (Balter 2008: 711; Mithen 2006: 251). In this regard rock art is one of the
most important tools scientists have at their disposal to study the prehistoric human
cognition and symbolism, but it is not the only tool. Lawrence Loendorf, who has
worked extensively with the rock art of the western United States likens this to a
lens and suggests that the more interpretive approaches (lenses) archaeologists

utilize, the more comprehensive will be the archaeological picture of the past
(Loendorf: 2008: 231).
Defining rock art can be challenging in and of itself. Rock art is
synonymous with landscape art (Whitley 2005: 3) in that it is found on the natural
surfaces of the landscape. It is rich in symbolism, it is almost always the result of
intentional human activity (sometimes natural features alone become culturally
significant so this is not a contradiction) and yet it is often very ellusive.
For the purpose of this project petroglyphs (mostly) and pictographs
(occasionally) comprise the rock art of the region. Petroglyphs are images that are
pecked or incised into a rock face. Pictographs are painted images. The rock art
images that are traditionally recorded are often comprised of discrete lines, dots,
circles and so on (which are often referred to as elements). See Table 2-1 for
definitions. A combination of elements forms a motif which may be an identifiable
image, such as an animal for example. Sometimes the pecking or painting
incorporates natural features of the rock surface into the motif. This is an important
component of rock art and is something that has been overlooked by rock art
researchers in the past who were accustomed to looking for painted or pecked
images or lines and who missed the natural lines, contours, and placement of the
imagery. Not all motifs are recognizable to a westerner as an image and many are
indistinct and abstract or geometric in shape. For this reason 1 will use the term

motif in this paper to describe both complete and representationally recognizable
images as well as other combinations of elements that may not be recognizable as a
physical object. As such motif is more than an element and often synonymous with
a symbol, but it is not necessarily a symbol. This is a general convention among
rock art researchers in Colorado. The formal definition for these terms from the
Colorado Office of the State Archaeologist follows (Office of the State
Archaeologist 2006a).
Table 2-1. Defined Terms
Abrading A method of making rock images by lightly rubbing the rock surface with a coarse, durable stone tool.
Element The smallest definable fragment of a design such as a line, dot, circle.
Incise A method of making rock images by cutting or abrading narrow linear marks into the panel surface; often an outlining technique.
Motif A combination of elements, repeating elements forming an identifiable image.
Panel Any rock face, on bedrock or a free-standing boulder, with one or more rock art elements in spatial association.
Petroglyph Any pictograph made on a cliff face or boulder; in modem usage generally restricted to unpainted rock images made by pecking, incising, abrading, drilling.
Pictograph A sign, symbol or figure made on any substance by any method; in modern usage referring to painted rock imagery.
It is often assumed that iconic or representational images are easy for
outsiders to identify. This is not always the case. One observers phallic human
is anothers tailed lizard (Whitley 2005: 45). A local informant can sometimes
help bring matters to light by offering traditional interpretations. Given the

difficulty in defining rock art in just one region it is not surprising that a universal
definition of the phenomenon is challenging. The term rock art itself suggests a
post-enlightenment renaissance era or Cartesian view of art that is probably not
anything close to an emic view of the phenomenon under study. From the emic
perspective rock art in a cave, rock shelter or overhang is not a wall decoration in
the same sense as poster art for the sake of art is in a modern apartment today.
Art pour l art or art for the sake of art (Lewis Williams 2002: 42) was first
proposed in 1864 to explain the Upper Paleolithic images and for better or for worse
the term art has stuck ever since. It was initially employed to discredit the notion
that early savage man could have engaged in religious symbolic communication
and instead suggested that rock art was a leisure-time decorative activity employed
by primitives to improve their otherwise drab living spaces.
Although this explanation was discredited very soon after it was proposed
we still see vestiges of it in contemporary cave-man cartoons and some popular
literature. Two problems confounded and undermined this explanation. First, most
Upper Paleolithic rock art is in relatively inaccessible dark caves that were not
inhabited and second modem art is made to be looked at. Deep Iberian Paleolithic
cave art was created in dark places that are not readily accessible to the casual
observer. Even if other rock art that has not survived the millennia was more
accessible, and there probably was plenty in that category, it does not negate that

fact that the Upper Peninsula rock art with which we are most acquainted is difficult
and often dangerous to reach, and therefore is not art in the classic sense. It is
something else, and that else can be understood in a context of humans interacting
with themselves and their environment.
Besides, art historians today see art as a social activity that cannot be
understood outside its social context (Lewis Williams 2002: 44). Lewis Williams
quotes Ernst Gombrich in The Story of Art who puts it this way, there is really no
such thing as Art. There are only artists who live and produce their work in a
social context. (Lewis Williams 2002: 45). This is an important point because it
underscores a central tenet of this paper that rock art is a product of human activity
that can be understood in its social and environmental or landscape context. The
archaeological importance of rock art is what it can tell us about the social behavior
of its creators.
The story of the history of rock art research is important to understand. It
began in Europe where caves with prehistoric rock art were known as early as 1458
when Pope Calixtus III condemned Spaniards living in the northern mountains for
performing rites in the cave with horse pictures (Curtis 2006: 47). The cave
Niaux in the Pyrenees has graffiti dating to the seventh-century and the cave
Rouffignac in the Dordogne has eighteenth-century graffiti. So we know that it has
been viewed by westerners on and off again for centuries. Rock art was visited and

known about but except for a line or two in a notebook (Curtis 2006: 48) it was
never taken seriously. The same is true of the rock art that must have been viewed
by Lewis and Clark on their famous expedition up the Mississippi and Missouri
rivers, but that was glossed over in their otherwise meticulous journals of scientific
and natural observation. As said before, it was seen but not seen.
Rock art research involves some of the most famous personages of late
nineteenth and early twentieth century prehistoric studies. Focused and scientific
rock art research began with a Spanish aristocrat and scholar named Marcelino Sanz
de Sautuola who in 1879 discovered the painted ceiling at the cave Altamira, which
his family owned. While visiting the 1879 World Exposition in Paris, Sautuola
had been impressed by the small mobiliary art and stone tools that had been found
in France by Edouard Piette, a noted French prehistorian of the time. Later in the
same year Sautuola was digging in the floor of Altamira and his young daughter
Maria was there with him. While he was concentrating on his digging Maria
reportedly cried out and pointed to the images on the ceiling, exclaiming Look,
Papa, Oxen (Curtis 2006:49). Altimiras great painted ceiling under which
Sautuola had been so laboriously digging had been discovered. Sautuola went on
to suggest that the paintings were the work of the same Stone Age people who made
the artifacts he found in the cave and this caused a significant controversy in
academia that essentially discredited him. He earned the immediate disdain of

Emile Cartailhac, the most respected personage in prehistoric archaeology (Curtis
2006: 52) who said the art was a vulgar joke by a hack artist (ibid), that the
images were too skillful to be genuine. Cartailhac refused to even visit the cave at
Altamira. Throughout the 1880s and 1890s amateurs continued to explore caves in
Southern France and several additional caves were discovered in that time span.
Many of those had been sealed for ages. Excavations in some revealed artwork
below the accumulated soil line, suggesting great antiquity. By 1901 Cartailhac was
convinced that the paintings were original and in 1902 he published a mea culpa
to the now deceased Sautuola. By the beginning of the twentieth century the
scientific community had moved from seeing but ignoring rock art, to seeing but
denying rock art, to seeing and believing that it was real and ancient.
A young assistant of Cartailhac, the priest Henri Breuil, subsequently
conducted decades of work on prehistoric cave art, particularly at Altamira, making
it one of the most studied caves in the Iberian Peninsula. This earned him the
distinction of being one of the fathers of rock art research. Breuil worked in the
days before photography and was a talented illustrator and drawer of cave art. He
was also one of the first scholars to apply ethnographic analogy to the study of rock
art. Breuil studied under Edouard Piette, the same prehistorian who inspired
Sautuola, and in 1906 Cartailhac and Breuil published the first definitive work on
rock art; La Caverne dAltamira a Santillane. It included three chapters on hunter-

gatherers in North America, Africa and Australia which introduced the present-day
hunting magic beliefs, otherwise known as sympathetic magic (Lewis Williams
2002:47), and the ritual shamanism of living hunter-gatherer societies into the realm
of prehistoric rock art. The twentieth century would see these interpretations and
explanations for the cave paintings of antiquity rise, fall into discredit, and rise
Other approaches to understanding rock art have been tried. Max Raphael, a
German Marxist and art historian, in the mid 1940s introduced the concept of
totemism to rock art studies. Raphael observed that different animal species
dominate the rock art in different caves and suggested that these reflected clan-
based distinctions or totems. His view was that the paintings in the caves were not
random but that they had a coherent structure (Curtis 2006: 123) and hypothesized
that the spatial proximity among the represented animals was intended to convey a
meaning which careful analysis would reveal. Raphiels structuralist perspectives
gave rise to other work, most notably that of Annette Laming-Emperaire and
anthropologist Andre Leroi-Gourhan. In 1959 Laming published a short book on
the cave Lascaux, which had previously been discovered in 1940. In her work
Laming rejected Breuils fatally flawed (Curtis 2006: 139) use of ethnographic
analogy which included shamanism and sympathetic magic and suggested instead
that meaning could be derived from the prehistoric imagery through the careful

recordation of the type and location of art in the caves, analysis of its distribution,
and analysis of its context with other archaeological materials, and its content or
form. She was interested in cracking the code (langue) of Upper Paleolithic rock
art and not in the social activity that produced it (Lewis Williams 2002: 55). She
employed the binary opposites of anthropologist Claude Levi-Strauss, the founder
of structuralism, in her efforts. One of her many observations was that the
juxtapositions or superpositioning of apparently discrete images should be
regarded as deliberately planned compositions (Lewis Williams 2002: 57) rather
than as a haphazard montage of randomly placed images.
Although she tragically died in 1978 her colleague and sometime competitor
Andre Leroi-Gourhan followed a similar approach that essentially rejected 50 plus
years of vague and imprecise generalizations concerning rock art (Lewis Williams
2002: 59) using ethnographic analogy. This was a Structuralist approach to
prehistory and was characteristic of the science of the time. Leroi-Gourham
believed that prehistoric art was the expression of ideas concerning the natural and
supernatural organization of the living world (Lewis Williams 2002: 63) and he
speculatively adding a decidedly non-Cartesian view that, the two might have been
one (ibid). He thought that rock art could be interpreted by deciphering the code of
mythograms through a structured analysis of the patterned distribution of the art
within the caves.

Like Raphiel and Breuil he carefully drew, plotted, and analyzed the
imagery. He did so in relationship to the opening of the cave, its central areas and
the recesses and inaccessible areas. He quantified rock art. He found, for example
that 91 percent of the bison, 92 percent of the oxen, 86 percent of the horses, and
58 percent of the mammoths occur in the central portions of the caves (Curtis
2006: 161). His work has been largely set aside due, at least in part, to the
empiricism that he so rigorously employed in his efforts to understand rock art. As
David Lewis Williams points out the diversity of the topography of the caves makes
it impractical if not impossible to compare them with one another in terms of
entrance/central/deep areas and other problems with the structuralist approach arose
as more and more caves were discovered. Eventually it collapsed under its own
weight of metrics and quantitative analysis.
Most recently the work of Jean Clottes, formerly general inspector for the
Division of Archaeology at the Ministry of Culture in France, has dominated the
rock art academic environment in the Dordogne and Iberian Peninsula (and
elsewhere). He is particularly noted in terms of his work on the dating,
photographic techniques, digital recordation, and preservation of prehistoric rock art
sites. Like Sautuola in the late 1800s Clottes endured public and academic
criticism from his peers who claimed his rock art was a forgery. Paul Bahn, a
British prehistorian, and Arlette Leroi-Gourhan, the widow of Andre Leroi-

Gourhan, in 1991 both fanned the fires of a popular media war and said that the rock
art in the newly discovered cave called Cosquer was a fake. Under Clottes
jurisdiction, Cosquer was initially closed to the public pending initial scientific
analysis and dating. The cave was also dangerous; it had to be entered through a
tunnel 130 feet under the surface of the water. Subsequent analysis yielded
preliminary radiocarbon dates for Cosquer suggesting an age of 18,000 years to
27,000 years old (Curtis 2006: 195), making it the oldest known rock art in the
world (ibid) at the time, and the skeptics were quieted. Sites such as Chauvet
which have more recently been discovered have pushed the date for the earliest rock
art to more than 30,000 BP (Chauvet et al., 1996: 122).
Clottes is known for his careful and meticulous recording techniques, his
concern with proper dating, and the preservation of rock art in its complete
landscape context. He has published several hundred articles and books and several
have been co-authored with archaeologist David Lewis Williams who is known for
his work in cognitive archaeology, his ethnographic work on various San ritual
ceremonies, and for the careful application of ethnographic analogy in rock art
This collaboration is an example of how the studies of rock art in Western
Europe have contributed to subsequent rock art investigation worldwide. The
history of rock art research in the Iberian Peninsula and the Dordogne is important

because it has provided much of the theoretical underpinning for subsequent rock art
research on different continents, including North America. It has also provided over
a century of well-documented scholarly work by notables such as Raphael, Breuli,
and Leroi-Gourhan.
In North America there is no rock art as ancient as the deep cave art of
Western Europe and there is little rock art that is actually located in the virtually
inaccessible dark depths of caves. Most rock art is found outside on rock faces, in
rock shelters, and on boulders on talus slopes. This is particularly true in the study
area and in the rock art of the Colorado Plateau in the Great Basin (which
encompasses a vast expanse of land in the Western United States and will be
discussed in greater detail in Chapter 3).
The oldest rock art in the Southwest is probably no more than 3,000 to 4,000
years in age (Schaffsma 1980: 3). The scientific study of this rock art dates only to
the early twentieth century, more recent by several decades than Western European
rock art studies. Some of the earliest correspondence concerning rock art in the
Southwest dates to as late as 1925 (Schaafsma 1980: 7). Rather than to propose
theoretical models to explain the rock art the location of rock art panels was and is a
topic of major concern to rock art enthusiasts and to archaeologists. As recently as
1980 Polly Schaafsma (1980: 32) called for an intensive regional survey of the

rock art of the Colorado Plateau in an effort to identify stylistic complexes which
could be assigned to various culture groups.
The pioneering work of Bill Buckles (1971) Sally Cole (1990) and others in
the last two decades has established a stylistic foundation for describing and
identifying the rock art of the study area. Through their work several stylistic
traditions have been identified and it is largely recognized that the content of rock
art is not random (Schaafsma 1988: 1). It is generally agreed that universal
meanings in rock art symbolism are rare, if they exist at all (ibid) and it is
recognized that other factors in addition to style such as ritual, function, and
location may provide additional insight into the rock art of the region.
By looking to previous studies in Europe and most specifically in Australia
this research project seeks to explore the relationship between rock art and human
prehistoric ritual behavior in a landscape context. This is the first such effort to link
the rock art of the Uncompahgre plateau to a specific type of human behavior,
namely ritual. The work of June Ross and lan Davidson (2006) will be discussed in
greater detail later but it is important to mention here because it is based the
presumption that rock art was an integral part of ritual performance in Australia
(Ross and Davidson 2006: 306), just as it is in the Western United States.
This brings us to the present. As Howey and OShea observed, when
archaeologists encounter patterns in their data that cannot be easily attributed to

such factors as subsistence, warfare, or trade the reflexive explanation is to call it
ritual activity. This evergreen truism (Howey and OShea 2006: 261) is most
often attributed to rock art and monumental structures without much justification or
supporting evidence. This project seeks to at least partially remedy this situation by
applying modem anthropological theory and careful analysis concerning human
ritual behavior to the study of rock art.
Ritual and the Sacred in Human Behavior
Ritual has been defined in many ways. Roy Rappaports definition is that it
is, the performance of more or less invariant sequences of formal acts and
utterances not entirely encoded by the performers (Rappaport 1999: 24). He
identifies seven elements of ritual behavior (which will be discussed later) and notes
that none of the elements, such as performance, invariance, or formality, are unique
to ritual. Instead, ritual particularly religious ritual brings them together into a
unique structure (Rappaport 1999: 26) that is distinctive and out of which
religion grows (Rappaport 1999: 25). The ritual discussed herein is not the
repetitive acts of the neurotic nor is it the mindless repetition of fiddler crabs
engaged in a courtship dance, nor is it the ritual of an athletic contest or a drama
with an audience of passive spectators. The ritual discussed in this research project
is principally (but not entirely) religious in nature. Religious ritual is a social
activity with moral and cosmological significance that is understood by those

performing them and by the communities within which they occur as having
profound sacred meaning. It is performed, not merely observed, by all who have
gathered, and it results in a re-affirmation of community and spiritual values.
According to Roy Rappaport, quoting G. C. Homan, this performance does not
produce a practical result on the external world (Rappaport 1999: 46) and this
lack of material efficacy is one of its defining characteristics. The effect of ritual
is not on the external environment of the world, it is on the internal and social
environment of the participants and the community in which they live. In this
regard the ritual of the Catholic Mass, the Plains Sundance, rites of passage in the
deserts of Central Australia, Papua New Guinea curing rituals and human sacrifice
in Aztec Mexico are all similar human activities (Rappaport 1999: 29).
Victor Turners seminal work on ritual has been so influential to
anthropology and is such an important part of understanding human ritual behavior
that any discussion of ritual would be incomplete without it. Turner was a cultural
anthropologist. He did not write extensively about archaeological applications of
his work or about rock art in particular, although he mentioned it occasionally. One
of Turners great contributions to understanding ritual was his deep appreciation of
performance in the actual practices of people celebrating life in its fullness
(Turner 2009: vii [1969]). Using this approach it is the whole person (Turner
2009: 43 [1969]) that is engaged in the ritual act.

Ritual symbols are multivocal (Turner 2009: 52 [1969]) in the sense that a
single symbol can represent many things simultaneously. Recognizing this is
important to this research project which does not seek to interpret or decipher the
singular or principal meaning of epigraphic symbols in rock art but rather to
establish the overall importance of symbolism and its relevance in a larger socio-
cultural gestalt.
In his writings on liminality and communitas Turner cites as an example the
rites de passage of Arnold van Gennep which accompany every change of place,
state, social position, and age (Turner 2009: 94 [1969]) beginning with separation,
moving into margin (or limen), and ending with aggregation. The liminality phase
in this transition is of interest because it simultaneously gives rise to and rises from
communitas among initiates which is a social relationship wherein the sacred is
acquired through a ritually potent (Turner 2009: 99 [1969]) and symbolically
powerful sequence of human actions. While rites of passage may or may not be part
of any particular rock art panel or location the important point here is that
communitas a strong unifying concept that underscores the importance of a
contextual approach in human social behavior. Communitas also helps address a
sticking point concerning ritual as a group behavior which is simply the observation
that an initiate in the liminal phase of a traditional rite of passage will, of necessity,
be at times alone. The act of being alone does not make his or her activity singular

or solitary because it takes place a larger cultural context over time and through
space, a community of initiates if you will. Vision questing is a known example in
Native American literature where, on occasion, an individual will separate himself
(it is most always male) from the larger social group and engage in a series of ritual
actions with a culturally prescribed goal. Participants in a vision quest are as
capable of creating rock art as they are of doing anything else and undoubtedly
some have done so in the past, even if they were simply a few scratches on a rock
face. The fact that they were alone when they engaged in the ritual act does not
make their ritual behavior solitary; it remains still a part of a group process. This
process is a social process, as Turner said, Society (societas) seems to be a process
rather than a thing a dialectical process with successive phases of structure and
communitas (Turner 2009: 203 [1969]).
Both Turner and Rappaport turned to Durkheim in their analysis and
understanding of ritual behavior, particularly as it relates to religion. Durkheims
dualistic approach saw the concept of religion as the bridge between the sacred and
the profane. While immensely understandable and highly useful, this continued a
western Cartesian dualism (Damasio 1994: 124) that is still largely reflected in
modem science and in the post-enlightenment church.
Durkheims definitions of religion, and of the sacred, are apt. He said, A
religion is a unified system of beliefs and practices relative to sacred things, that is

to say, things set apart and forbiddenbeliefs and practices which unite into one
single moral community called a Church, and all those who adhere to them
(Durkheim 1995: 44 [1912]) and because a church is made up of active practitioners
(not just observers) as such religion is an eminently collective thing (ibid) that is
broader than the supernatural or the gods. According to Durkheims 1995 translator
Karen E. Fields this means that, Sacredness is an aspect of the real that exists only
in the mind but cannot possibly exist as the real in only one mind and, rites are
ways of acting that are bom only in the midst of assembled groups and whose
purpose is to evoke, maintain, or recreate certain mental states of those groups
(emphasis added) (Durkheim 1995: xliii, 9 [1912]).
Rock art seems to invariably lead to a discussion of religion and the
supernatural or spiritual component of religious beliefs and practices. Jean Clottes
puts it this way, One of the rare things on which all specialists agree is that
prehistoric cave drawings have a spiritual purpose (Uncompahgre Journal 2008: 4).
Because ritual behavior and sacredness of place is the focus of this project the
discussion of religion and spirituality is appropriate. The connection between these
elements is necessary in understanding the social context of rock art. To not address
the connection would be to sidestep important contemporary cultural and political
issues and would reduce the relevance of the research project.

Rappaports thesis in Ritual and Religion in the Making of Humanity is that
to convey information symbolically is a great innovation in a world where symbolic
communication does not otherwise exist and perhaps never before existed.
However, the very act of symbolically conveying any information is to potentially
convey misinformation, or to put it another way to convey the possibility of
alternative proposition. It can logically be no other way. It is a universal
proposition that all human societies have culture, language, and symbolism and to
this list Rappaport says should be added religion and ritual in that order; and ritual is
the only way can societies overcome the a priori credibility gap in symbolic
communication that he identified.
The cross, the Star of David, and other religious images arguably serve these
same functions and are demonstrably accepted by contemporary humans as valid
religious brands that we use to interpret the world in general (DAlessandro 2001:
17). These brands, symbols, or images sanctify and structure societal behavior in
fundamental, often violent, seemingly irrational, and incontrovertible ways.
According to Rappaport, religion is as old as language, which is to say
precisely as old as humanity (Rappaport 1999: 16). Religion is the way humans, in
a cultural context modulated by symbolism and language, connect themselves to the
unknown. In his forward to Rappaports book Keith Hart of Cambridge University
invokes Durkheim in this link to religion by saying, for Durkheim, religion was the

organized attempt to bridge the gap between the known and the unknown,
conceived of as the profane world of ordinary experience and a sacred,
extraordinary world located outside that experience (Rappaport 1999: xv). For
Rappaport, according to Hart, ritual is the active ground (Rappaport 1999: xvii)
where language, symbolism, and religion [which includes the sacred] are made
Rappaports thirty years of fieldwork in religion, society, and ecology
developed a concept of religious ritual that did not draw a hard line between the
sacred and the profane. Rappaports ritual, as said above, was the active ground of
religion and he viewed religion as a universal for a species that lives, and can only
live, in terms of meanings it must construct in a world devoid of intrinsic meaning
but subject to physical law (Rappaport 1999: 1). Kus makes a similar point when
she says, humanity creates the world it inhabits by investing it with order and
meaning (1995: 140). See Bell for a contra view (Bell 1997: x). Ritual, according
to Rappaport, is a social act basic to humanity and it is this basic qualitative act
which may, under certain circumstances and in certain regions of the world such as
the Uncompahgre Plateau of western Colorado, be directly and quantifiably linked
to identifiable features of the rock art assemblage.
In The Elementary Forms of Religious Life which was originally published
in 1912 Durkheim wrote an eerily prescient passage, there must necessarily be a

certain number of fundamental representations and modes of ritual conduct that,
despite the diversity of forms that the one and the other may have taken on, have the
same objective meaning everywhere and everywhere fulfill the same functions
(Durkheim 1995: 4 [ 1912]). In Ritual and Religion in the Making of Humanity
which was published in 87 years later Rappaport identified seven features of ritual
behavior which meet the fundamental representations earlier imagined by
Durkheim. These seven features and their relationship to rock art are the focus of
this research project. Durkheim, Rappaport, and Turner view the sacred as part of
the interplay between religion and ritual, in much the same way as length, width,
and breadth are each necessary in order to measure the volume of a three
dimensional cube. While this crude analogy can be stretched too far, it suffices to
make the point that in the realm of human experience, where there is ritual there is
religion and there is also the sacred.
Rappaport on Ritual
Roy Rappaports work has established a useful conceptual bridge between
culture, symbolism, ritual, and the sacred. This connection allows ritual to be
linked with sacred places in the landscape, such as rock art sites. Rappaports great
contribution was the identification of seven fundamental and objective
characteristics of ritual behavior which are as follows: 1) encoding by other than the
performers; 2) formal style and decorum, 3) invariance (more or less); 4)

performance; 5) formality through (emotional) action and repetition; 6) use of a
special medium of communication; and 7) canonical messaging (versus self-
referential). Each of these deserves further elaboration.
Encoding by other than the performers. This first criterion means that the
performers of rituals are not the ones who specify the acts and utterances
constituting their own performances (Ross and Davidson 2006: 311), they follow,
more or less punctiliously, orders established or taken to have been established by
others (Rappaport, 1999: 38). This means that ritual activity is, more or less,
formal and traditional. It is not spontaneously created out of the passions of the
moment but rather is planned and executed, more or less, by performers who do not
control the content of the ceremony or the acts. The more or less is an important
consideration because it recognizes and allow for variations on a theme while at
the same time establishing a convention (Rappaport 1999: 27) and provides for
the sealing of a social contract (ibid). There is room for the rearrangement of
elements, and for discarding some elements and introducing others, but invention is
limited and the sanction of previous performances is maintained (Rappaport;
1999: 32).
Formal style and decorum. Formality is a key aspect of all rituals and this
formality leads to behavior in ritual activities that tends to be punctilious and
repetitive (Rappaport 1999: 33). The gestures and postures in ritual are

conventionalized, stereotypical, and stylized. Rituals are performed in specified
contexts (ibid) and are regularly repeated at times established by clock, calendar,
biological rhythm, ontogeny, physical condition, or defined social circumstance
(ibid). They often occur in special places. To put it another way, religious rituals
tend to concentrate toward the formal end of the behavioral continuum (ibid) and
that formality requires decorum rather than something akin to the spontaneous and
stereotypical greetings of teenagers or debutantes, for example.
Invariance (more or less). Invariance logically follows from formality and
traditional encoding. Rappaport accords it special significance and its own place in
the list of ritual features because ritual acts tend toward invariance but they are not
necessarily punctiliously invariant. There is room for logically necessary or
deliberate variation and in this regard ritual ceremonies can change over time
(Rappaport 1999: 36). Some rituals allow for more variation in content and
location (such as marriage ceremonies) and others are much stricter in terms of how
they are carried out (such as the coronation of a king, for example) but invariance
(more or less) remains a key element.
Performance. Ritual is performed, unless there is a performance there is no
ritual (Rappaport 1999: 37). Rappaport puts it this way; elaborate liturgical orders
may be inscribed in books but the recipe is not the ritual. They are mere
descriptions or instructions. They are realized, made into res (ibid) only when

they are enlivened by performance but not in the same way that a play is enlivened
by reciting and acting the script. Rappaport provides a good example involving
theater. Those present at a ritual are members of a congregation. Those present at a
play are an audience. In a play audiences and performers are separated from each
other in many formal ways while the audience watches and listens to the
performance. In ritual there is no such separation. In a ritual everyone present is a
performer, to a greater or lesser degree, and members of the congregation leave the
ritual in some way formally transformed (Rappaport 1999: 40). The social status
of audiences at a play are not and cannot be transformed as they are in ritual (such
as a marriage ceremony for example).
Formality through emotional action, not physical efficacy. Ritual is done
in earnest (Rappaport 1999: 46) and as such it is functional physically as
discussed above but it is also functions on an emotional level in a formal social
context. Rituals may not change the external world but they can forever change the
relationships between people and this can be a moving experience. The actors in a
ritual know the efficacy of spells to be different from that of spears (Rappaport
1999: 48) and the efficacy of ritual is derived from what Durkheim called the sacred
and what others may call the supernatural, not by the secular or the material.
Whatever the cause, neuropsychological or otherwise, ritual is often perceived by
the participants to be effective because of the strong emotions (Rappaport 1999:

49) that are persuasive (ibid) to the participants. Rappaports choice to use the
term formality in naming this category is in my view unfortunate because it
confuses the emotional form and spiritual features of ritual with the formal style
of ritual. For this reason I have added emotion to the description of this fifth
Special medium of communication. Rappaport calls this one of rituals
strangest features (Rappaport 1999: 50) that separates it from daily life, that
makes it weird and extraordinary where everyday things take on special meaning
and symbolism. Think of the common words I do (Rappaport 1999: 51) for
example. In day to day life they have one meaning but in a marriage ceremony
those two words, more than any others, seal a legal, social, and spiritual bond
between two individuals that will last for eternity. It is here that Rappaport
recognizes rituals conducted in solitude where the performers presumably do feel
themselves to be communicating with spiritual beings (ibid) in ways that they do
not or cannot in day to day life. To Rappaport ritual communication is a special
medium (Rappaport 1999: 52) that is unique and not interchangeable with other
types or modes of communication (ibid).
Canonical messaging (versus self referential). Ritual messaging, as in most
messaging, often conveys the current physical, psychic, and social status of the
participants but it also does something else. It transcends the present and connects

the performers with those who have come before and by implication with those who
have yet to come. The Canon of the Eucharist which is part of the Catholic Mass is
an example. Rappaport notes that it has remained more or less unchanged for over
a millennium (Rappaport 1999: 53). Canonical messaging is not about the
participants and it is not confined to the present in time or space (Rappaport 1999:
54) and may even be conceived to stand outside the time-space continuum
altogether (ibid). It always includes, words and acts that have been spoken or
performed before (Rappaport 1999: 53) and upon which are based symbols and
icons that are spiritual, conceptual or abstract in nature (Rappaport 1999: 54).
Canonical messages rest ultimately upon symbols (Rappaport 1999: 58) and in
this regard are well-suited to the visual imagery of rock art, particularly to the
abstract imagery contained therein.
Central Australia
The work of Ross and Davidson in central Australia (2006) for the first
time ever applied Rappaports seven ritual forms to the archaeological context of
rock art. They developed a methodology to test the link between rock art
composition and the physical and social context of ritual behavior as defined by
Rappaport, In accomplishing this they set the stage for future work in other regions
of the planet (such as the Uncompahgre Plateau) wherein similar relationships
between rock art and ritual could be examined.

Rappaports (1999) anthropological analysis of the universal form of ritual
show(ing) that an analysis of the structure of art assemblages provides the
best means of identifying ritual activity associated with the production of
rock art. Underpinning this approach to the study of rock art is a set of
assumptions about the structured nature of a corpus of rock art where the
relationships between the motifs themselves and the physical and social
contexts in which they were produced are a cultural construct and therefore
meaningful (Ross and Davidson 2006: 309).
Ross and Davidson note that, it is this concern with the structure of ritual
rather than with the content of the individual rituals that makes his research so
pertinent to archaeological investigation (Ross and Davidson 2006: 311). They re-
categorize and rename Rappaports seven features of the structure of ritual for the
purposes of rock art research as follows:
Invariance can be reflected in the selection of motif type; the method of
producing the motifs; and the physical context of the motifs. Each of the three
attributes can be identified and recorded.
Repetition in rock art can be recorded on three levels: across a region or
area; within a site complex (group of sites closely associated in space and context);
and by repetition of motifs on individual rock panels. Repetition can also be
indicated by reworking of a motif, superpositioning of one motif over another,
greater motif depth, and variations in repatination of different motifs in a panel

Table 2-2. Comparison Between Rappaports Seven Features of Ritual and Ross and
Davidsons Field Categories
Rappaports Seven Characteristics of Ritual
Rapp- aport 1 Rapp- aport 2 Rapp- aport 3 Rapp- aport 4 Rapp- aport 5 Rapp- aport 6 Rapp- aport 7
Endoding by Others Formality as Decorum Invariance Performance Emotional Action Specialized Communi- cation Canonical
Ross and Davidisons Field Categories Invariance X X
Repetition X
Specialized Time X X
Specialized Place X X
Formal or Stylized Behavior or Image X X
Performance and Participation X X X
Motif Form that Can Hold and Transfer a Canonical Messsage X
Determining specialized time is not as problematic in Central Australia as in
the Uncompahgre Plateau of Colorado where occupation debris are not often
associated with rock art. The seasonal migrations of hunter-gatherers in western
Colorado is well-established and serves as an indicator that the utilization of
prehistoric rock art sites would therefore be seasonal as well; as is the occupation of
known habitation sites, for example. Possible indicators include other cultural
resources nearby, azimuth, and distinct image type (such as wildlife).
Specialized place can be established in two ways. First, rock art is often
(surprisingly) not present in places where all the topographic indicators suggest it

should be. The places where it occurs are therefore specialized places by
definition. Second, the addition of motifs over time (as suggested by relative
patination and superposition for example) and reworking are both indicators that the
location was re-utilized over an extended period. Repatination is indicative of a
special place that has been re-visited (and reworked at different times) as well. This
places it into a cultural framework as a locus or magnet for ongoing specialized
activity. Finally, ethnographic evidence suggests that some locations considered
sacred sites in recent history by the Utes can also encompass (through
superposition and other ways such as reworking an image) archaic or post-archaic
rock art motifs and styles. This indicates that the same locations may have served as
specialized or sacred places into the archaic era. Any patterned or unique layout of
the site location as identified by site drawings can also serve as indicators of
specialized place. It will be argued later that special acoustical properties also
contribute to the special nature of a rock art place.
Stylized form is evidenced by invariance in imagery which can encode all
types of messages (from simple to complex). The meaning is culturally specific and
probably not decipherable, but the existence of stylistic traditions as well as
repeated motifs speaks to a recognizable stylized form. Stylized forms are well
known and documented in western Colorado (Cole 1990: 34).

Performance and participation can be identified in several ways. First, the
creation of rock art is a production activity and the existence of the rock art speaks
to a performance ipso facto. Second, the likelihood of group participation
(Rappaport contended that ritual is usually but not always a group practice) can be
estimated based on the nearby local topography, orientation of the rock art panel(s),
access to the panels, and location of the art on the panels. For example some art is
visible from local game trails or water sources, other rock art is not readily visible at
all until the viewer is in close proximity to it, some rock art is not readily visible
from below but its location is suggested by indicator panels (Martineau 2003: 20
[1973]). Is there space near the rock art for groups of ten or twenty people to
congregate (with near being defined as within touching or sound reverberation
distance) and so forth. Secondly, reworking, outlining, and abrading of existing
motifs may provide very strong archaeological evidence for participation by people
other than the original artist. It is worth mentioning that chalking the outline of a
petroglyph for photographic purposes is also technically a performance and
participation indicator (and several panel locations in the study area have evidence
of chalking and plaster castings of the imagery) but archaeologists know that
chalking is not a ritual activity and that is why multiple indicators are needed. It is
also why researchers should not suppose a priori that all rock art is ritualistic in

function simply because a one or a few of the indicators of ritual behavior are
Canonical messaging is defined by tradition and social convention, it can
convey supernatural meaning, and is not profane, personal, or spontaneous.
Invariance, persistence, and repetition in rock art can indicate canonical messaging
and it is irrelevant what the message is in fact the message can change over time
and from one culture group to the next even when the symbol remains constant. In
this regard rock art styles and motifs that are invariant, persistent, and repetitious in
imagery are a priori canonical in its messaging. Unique symbols and motifs cannot,
by definition, be canonical but geometric and abstract motifs (especially when they
are repeated) can lend themselves well to canonical messaging.
Ross and Davidsons application of Rappaports approach to ritual lends
itself very nicely to archaeological investigations because it is thorough, meticulous
and seemingly exhaustive. This holds true provided that the approach is landscape-
oriented rather than site-specific because a single site cannot, by definition, provide
the geographic context or social perspective necessary to address such variables as
specialized place or geographically-distributed stylized imagery, and in this regard
the landscape approach is indispensable to the research question.

Human Experience and Cognition
Taken together the works of Durkheim, Turner, and Rappaport provide a
framework for understanding the relationship between religion, the sacred, and
ritual practices in the totality of the human experience. Recent advances in
neuroscience, neuroanatomy, and neurochemistry have been brought about by brain
imaging technology which allow researchers to investigate the activity of living
brain tissue, often while conducting ingenious experiments involving participatory
research subjects. It is through these methods that the greatest insights into the
neural bases for music and language (Mithen 2005: 31) and expanded knowledge
of human cognition have been achieved.
My purpose for introducing neuroscience into a rock art research project is
twofold. First, I intend to show how current research in neuroscience suggests that
the human brain is so much more than an instrument of Cartesian rationality that
passively processes external stimuli for the purpose of increasing the optimal
reproductive fitness of individual. This is relevant to several points relating to the
gestalt of the human experience associated with the practice of ritual mentioned
above. This pertains directly to the anthropological concept of culture and the
contextual approach to understanding culture and the human experience. I intend to
show how recent advances in neuroscience have lead to a greater understanding of
the role of feelings and emotion in the human decision-making process and how it is

in fact impossible to make informed and rational decisions without them. This
contextual approach corresponds to Ian Hodders characterization of culture as
meaningfully constituted (Hodder 1991: 32; Hodder 1982: 9) which largely
defines symbolic or post-processual archaeology. This is directly relevant to an
understanding of the ritual significance of rock art created, as was said earlier, by a
species that lives, and can only live, in terms of meanings it must construct in a
world devoid of intrinsic meaning but subject to physical law (Rappaport 1999: 1).
Second, I intend to show how current research in neuroscience suggests that the
human brain is receptive to sound and that various acoustic elements of sound such
as pitch, tone, rhythm, and tempo, affect the general mood and overall disposition of
the listener. This is relevant to the acoustical portion of the research project.
To set the stage let me begin by stating that neuroscientific applications to
rock art research are not new. Entoptic phenomena have recently been recognized
as a valid component of the neuropsychological model (Whitley 2005: 110). This
is a formal analytical approach in rock art research that is based on clinical and
ethnographic data relating to the mental imagery associated with trance states and
other altered states of consciousness. These experiences are cross-cultural and are
based on the fact that all modem humans are neurologically hard-wired in the
same way (ibid). While I suspect that this statement is only partially true, I
proceed with it based on a lack of direct evidence to the contrary. Another reason to

move forward with this presumption is based on the recent work by Thomas Wynn
et. al. which presumes that the neural networks of all modem humans are hard w ired
similarly. They suggest that increased frontal lobe activity is largely responsible for
the domain-specific working memory of modem humans that allows for the
creation of imaginary figures like the 32,000 year old half-human and half-lion
Holenstein-Stadel figure (Wynn, et. al. 2009: 79).
Regardless of how the altered state is reached (through ingestion of
hallucinogens, through sleep deprivation, extreme temperatures, or extreme fasting,
for example) the sequence of three distinct stages is the same. The first involves a
variety of entoptic patterns which are spontaneously generated in the optical system.
These patterns recognized as dots, grids, zigzags, nested catenary curves, and
meandering lines (Lewis Williams 2002: 126). They are also associated with
migraine attacks and are well-known to persons suffering from that condition. The
second stage involves the creation of iconic forms (Lewis Williams 2002: 127)
where the subject tries to make sense of the imagery and, for example, an
ambiguous round shape may be illlusioned into an orange if the subject is hungry
(ibid). The third stage is a full-blown hallucination that brings into play powerful
culturally-generated and emotional experiences. The mental images become more
vivid. The images no longer look like they might be something; they are real to the
subject and are dealt with as if they are real, hence the term hallucination. It has

been suggested that much of the geometric, abstract and possibly the theriantropic
imagery (that combines features of humans and animals such as the Holenstein-
Stadel figure mentioned earlier) seen in rock art is the result of subsequent
recordations of this experience. Shamanism has a long history of being based on an
institutionalized altered state of consciousness (Lewis Williams 2002: 133) and as
such the neuropsychological model is closely associated in the archaeological
literature with ethnographic work on shamanism. Whether rock art imagery may be
shamanistic or not will not to be argued here. What is important is that the
neuropsychological approach is well documented and is recognized as a valid
analytical approach to rock art research.
The most recent rock art in the study area is of Ute origin. The Utes have a
cultural history of shamanism and are known to have engaged in vision quest
ceremonies and prolonged dances, such as the seasonal bear dance which can go on
for up to ten days (Cole 1990: 28). This the shamanistic model may be valid in the
study area, but shamanism is not the present concern. The neuropsychological
model, however, remains useful both in the study area and as the foundation for
introducing other aspects of neuroscience to the study of rock art.
The somatic-marker hypothesis (Damasio 1994: 165) in neuroscience is
predicated on a neurobiological principle that rejects the classic Cartesian separation
of the brain from the body and both from the physical and the cultural environment.

Of the two environmental considerations (physical and cultural) it seems that the
cultural may be the most relevant according to neuropsychologist Antonio Damasio
who said, To understand in a satisfactory manner the brain that fabricates human
mind and human behavior, it is necessary to take into account its social and cultural
context (Damasio 1994: 260). However, before engaging in a discussion of the
brains cognitive abilities in a cultural sense it is necessary to realize that the brain
and the body are indissociably integrated by mutually targeted biochemical and
neural circuits (Damasio 1994: 87). One route is through sensory and motor
peripheral nerves and the other is through the bloodstream. Both involve electronic
stimulation and chemical signals such as hormones and neurotransmitters and there
is no clear cut boundary between the body and the brain. Nearly every part of the
body, every muscle, joint, and internal organ, can send signals to the brain
(Damasio 1994: 88) and they do so millisecond by millisecond.
The brain in the vat (Damasio 1994: 227) thought experiment is a case in
point. Imagine a brain removed from the body and placed in a nutrient bath
stimulated through its dangling nerves by some external computer program. Would
such a brain have a normal human mental experience? Some rational people,
according to Damasio, would say yes. His point is that neuroanatomy suggests the
opposite. There is a constant electro-chemical and physical modulation occurring
between the brain and the body that is so pervasive it is for all practical purposes

impossible to separate the two. As Damasio says, for millions of years, brains
have been first about the organism that owns them (Damasio 1994: 229) to such a
degree that the body is not so much a tool of the brain as the brain is the bodys
captive audience (Damasio 1994: 160).
The somatic marker hypothesis brings this full circle and suggests that the
purpose of reasoning is deciding (Damasio 1994:165) and deciding has more to do
with caring what to decide about than with ordered rational efficient decision
making of how to optimally do something (anything really). To put it another way
(into a management perspective), the first step toward effectiveness is to decide
what are the right things to do (Drucker 1992: 198). The somatic marker
hypothesis, in a nutshell, presumes that the various segments of the brain (frontal,
temporal, occipital, parietal, and limbic, for example) act in concert to reach a
conclusion about an appropriate goal for the organism faster than the frontal cortex
can process data, faster than the person can think. Thinking is what helps the
organism do what it decided to do in the first place.
This decision-making process involves a complex interplay of feelings and
emotions appropriate to the desired somatic state of the organism. It may not seem
to be a rational decision-making process but it can be, depending upon the definition
ofrational. One of the best working definitions so far comes from economic
theory which suggests that rational agents are engaged in multiplying the

probability of getting what they want by the amount of pleasure (utility) that getting
what they want will bring (Lehrer 2009: 100) thereby maximizing happiness,
which is what rational agents are always supposed to do (ibid).
Damasios theory is that progress toward this goal is initiated through
emotions such as Happiness, Sadness, Anger, Fear, and Disgust (Damasio 1994:
149). Neuroscience demonstrates that these very emotions can be expressed and
perceived through universal facial expressions (Holden 2008: 705) that are not
controlled consciously. They are controlled through the autonomic nervous system
(Frank 1988: 118). In a social context they occur spontaneously through the
contraction and relaxation of facial muscles that are not under the conscious control
of the frontal cortex such as the corrugators supercilli, the occipito frontalis, the
pyramidal nasi, and the quadrates menti (Frank 1988: 121-122). Many of them are
expressed and conveyed through fleeting expressions or gestures (Mithen 2006:
89) or microexpression (Frank 1988: 125). Darwin recognized and elaborated on
this in The Expression of the Emotions in Man and Animals in 1873 (Frank 1988:
124). Many of the muscles responsible for these telltale expressions are not
physically linked to the same cognitive brain centers as the facial muscles under
conscious control but are rather linked to and controlled by portions of the brain
responsible for emotional activity such as the fight or flight (Frank 1988: 118)

If emotions usually precede thought and most often serve to direct the
conscious decision-making process of the human biocultural organism then it stands
to reason that the experience of the emotion of ritual process in-and-of-itself has a
strong and well-established place in the whole person (Turner 2009: 43 [1969])
not just the mind (ibid) and any research on ritual that does not take the emotional
state of the subjects into consideration would be incomplete. Emotion then becomes
a necessary, and possibly sufficient, cause for human ritual behavior to occur. This
relates to what Rappaport said about ritual being the active ground (Rappaport
1999: xvii [1967]) where language, symbolism, and religion are made whole. It also
stands to reason that the emotional results of ritual behavior, whether symbolically
generated or occurring as the result of group ceremonial activity involving
neuropsychological stimuli such as song, dance, rhythm, natural stimulants such as
adrenalin, sleep or sensory deprivation, or hallucinogens or the combined effect of
some or all of the above, produce a state of consciousness in which the subject
perceives himself/herself as playing an important and prescribed social role in the
ceremony. This can be viewed either as Durkheims social or Turners
communitas, and it provides the profound emotional effect of ritual. It is the rock
art of ritual expression, or put another way, the expression of ritual through rock art,
that this project seeks to address. In this way perception becomes manifest reality.

Storing images and recalling images in the mind is not a process of mentally
locating a facsimile or photocopy and retrieving it. The brain does not use the
hundred billion neurons connecting to each other through a hundred trillion
synapses (Mithen 2006: 28) to mentally create and file digital photos or films of our
lives. Visual perception involves constructing an image in the visual cortex that is
an approximate representation of what the retina is exposed to and other sections of
the brain then interpret, make sense out of, and determine a utility for that image.
Memory involves the reconstruction of not only images but also of the patterns that
were once experienced and then associated with that image (Damasios somatic
markers). Recalled images are held in memory fleetingly and their activation
results in a topographically organized representation (Damasio 1994: 101) that
engages many of the same regions of the brain that were involved in the initial
perception of the image. These regions include emotional centers and are not
limited to just the visual cortex. Positron emission tomography (which measures
blood flow in the brain) and functional magnetic resonance imaging (which does the
same thing by measuring minute changes in the magnetic properties of blood
hemoglobin molecules) have shed some light on this. For example in a subject
viewing a cross or a square a pattern emerges in the early visual cortices of the
subject that conforms somewhat to the shape the subject is viewing. An external
observer viewing the cross or the square, and the neuroimage of the subjects

cortical activity, will see similar shapes in both locations (Damasio 1994: 104). To
complete the process of mentally reconstructing the image however, the subject
will engage other portions of the brain such as the occipital and temporal regions.
Recalling an image of a friend might result in stimulation of the early visual cortex,
but milliseconds later the auditory cortex will fire back an approximation of that
persons voice, and so on, involving many centers of the brain including those
associated with emotion such as the amygdala and the anterior cingulate of the
limbic system. The recall of a mental image instantly becomes a complex human
experience involving many centers of the brain and the associated feelings (or
somatic representations of the state(s) associated with that image), it becomes a
cognitive gestalt because the feelings are just as cognitive as any other perceptual
image (Damasio: 1994: 159). This process of emotionallymarkingspecial
people and events in our minds can also be applied to marking significant ritual
events in the past, special (sacred) things which make up our material culture, and
the special places associated with those events. The places where ritual is
performed would be expected, therefore, to be places with strong emotional ties
involving more than one of the five senses.
Archaeoacoustics and Sound
In his article on the political aesthetic archaeologist Adam T. Smith observed
that humans are both sensate and sensuous (Smith 2000: 138) creatures and

criticized materialist interpretations that rely on invented psychological universals
where emotion merely serves rationality (ibid). Smith called for archaeology to
involve actors, whose emotions are as involved in the production of the material
culture as their subsistence requirements (Smith 2000: 139). In her comprehensive
review of archaeological approaches to emotion Sarah Tarlow echoes this view
when she says, emotion cannot be separated from other aspects of social and
cultural meaning and experience (Tarlow 2000: 713). In their preface to
Archaeoacoustics Chris Scarre and Graeme Lawson introduce archaeoacoustics as a
field of study responding to recent calls for an archaeology of the senses (Scarre
2006: vii) and suggest that there is plenty of opportunity for research into the
historic and prehistoric role of sound in human behavior. It is through
archaeoacoustics that an archaeology of the senses becomes real (also see
Pallasmaa on an architecture of the senses including acoustic intimacy [p. 34])).
The sound component of the human lived experience is fundamental to ritual and if
rock art is to be credibly linked to ritual it must probably be done through at least
one sense other than the visual one. A study of the acoustical properties of rock art
sites lends itself well to this endeavor.
Early interest in sound and music in archaeology can be traced to the work
of John Coles on Bronze Age horns. Quite a bit of the work in archaeoacoustics
involves identifying and reconstructing ancient musical instrumentation ranging

from simple percussion instruments made of stretched hide on a wooden frame to
much more complex and novel items. The cognitive and behavioral implications of
music archaeology are apparent and quickly lead to deeper questions that build
upon what archaeology already knows about tool-based and environmental
phenomena in prehistory. This approach is called cognitive archaeoacoustics
(Scarre 2006a: viii).
Scarre and Lawson point out that the most ancient sound-producer of all is
the human body; feet, hands and voice (ibid) and they suggest that future work in
cognitive archaeoacoustics may help shed light on the deep-rootedness (ibid) of
rhythmic and musical behaviors in the modem lived human experience, such as the
ritual activity described by Turner and Rappaport.
Ritual is a participatory thing according to Turner and Rappaport and if it is
anything it is noisy. Humans communicate in many ways but auditory verbal
communication is our principal method of exchanging information with one another
and of expressing our wants and dislikes. Being effective at expressing and inducing
emotion through the proper use of pitch sequences may provide a selective
advantage to individuals because they would promote greater degrees of cooperative
behavior (Mithen 2006: 136). According to archaeologist Steven Mithen, Music
induces emotional states both in those who perform and in those who simply listen
(Mithen 2006: 94). Language and music processing seems to involve a large

portion of the brain including part of the interior frontal lobe, Brocas area (Mithen
2006: 66) and a homologous area in the right hemisphere, as well as portions of the
temporal lobes, the right superior temporal gyrus, and the right cerebellum. The
neural networks for processing sound are apparently not as discrete and identifiable
as those for processing visual imagery. It might be that neural networks vary
dramatically between individuals. Evolution may have programmed the brain to
develop neural networks for music without specifying where they are to be located
(Mithen 2006: 64). Research suggests that the frontal cortex is also active in
processing music but the focus varies, for rhythm it was the superior region of the
frontal cortex that was most strongly activated; for melody it was the inferior
region; and for harmony it was somewhere in between (Mithen 2006: 67).
What this suggests is that sound, like visual imagery, involves a complex
interplay between various regions of the brain that begin with sensory perception
but rapidly (within milliseconds) extend into several disparate regions of the brain
for analytical processing and emotional assimilation. Because the human
phenomenological experience is multi-sensory it stands to reason that special
physical places encountered in the environment will be sought out based on their
visual and acoustic features and that they will serve as choice locations for the
multisensory phenomena associated with ritual ceremonies and events.

In this regard, the special acoustical properties of some rock art sites are
beginning to be investigated by archaeologists. This additional line of evidence
adds credence and legitimacy to the various ethnographic reports that extensively
document the use of rhythm and dancing in ritual. If there are special acoustics
related to the locations of some (if not all) rock art panels then the use of auditory
stimuli (such as chanting, singing, speaking, drumming, or rhythmic dancing, for
example) at these special places can be logically inferred.
An example of this can be found in the Upper Paleolithic painted caves of
the Iberian Peninsula. Research conducted there by Iegor Reznikoff suggests that
red dots made of ochre are related closely to the resonance of the part of the cave
where they are located (Reznikoff 2006: 79). Resonance, for the purposes of his
experiment, was defined as the point of the acoustical main antinode. According to
Reznikoff, these red dots are usually about 2 centimeters in diameter and they can
be found at various locations throughout the dark interior of painted caves,
including locations where there are no paintings and there is no space available for a
painting due to the narrow confines in which they are located. This suggests that
prehistoric dot makers were possibly responding in some way, for some purpose, to
the acoustical qualities of the dark chambers and crawlspaces in which they found
themselves. Reznikoff calculated the odds of these occurring by chance at the
acoustical main antinode and determined that they were something on the order of

a million to one (Reznikoff 2006: 80). His research brings to mind some of
Raphiels and Leroi-Gourhans early structuralist approaches to cave topography
but in this case there seem to be some defined associations between the imagery and
the location.
At Horseshoe Canyon in Utah, which is a high-walled sandstone canyon,
archaeologist Steve Waller has digitally measured the echo intensity at regular
intervals throughout the canyon and identified a correlation between rock art panel
clusters and spikes in his measurements. Waller was able to link the rock art to
the acoustical properties of the canyon. He also succeeded in locating a previously
unknown rock art panel and thereby showing that acoustical measurements can be
useful in predicting the location of heretofore undiscovered rock art panels (Waller
2006: 33; Whitley 2005: 128). This predictive value may be as important as the
observed correlation between echoes and rock art locations.
Waller suggests that since echoes seem to emanate from the very surface of
the rock, and a Durkheimian animistic spiritual genesis could be inferred as their
cause. He compiled an exhaustive list of echo myths from around the world
beginning with the Greek nymph Echo (Waller 2006: 37) that relied heavily on
ethnographic and historical research. This research suggests that the acoustical echo
phenomenon has a strong place in the mythology and spirituality of a wide variety
of cultures throughout the world, both living and ancient. Part of the attraction of

testing for echoes is their unique acoustical signature. However, not all echoes can
be perceived by the human brain. A time delay of at least 1/1 Oth to 2/10th of a
second between the initial sound percussion and the return sound is required in
order for it to be perceived by the human brain as a separate acoustical event. I will
expand on this in the methods section.
Waller pioneered the development of careful field measurements that test the
relationship between echoes and rock art. He has recently expanded his own work
from Horseshoe Canyon in Utah to Hieroglyphic Canyon in Arizona (Waller 2006:
35) and other sites in the arid southwest thereby establishing echo identification as a
legitimate component of archaeoacoustical research. The proximity of the
Uncompahgre Plateau in western Colorado to previous acoustical research in the
southwest and the climactic and geomorphological similarities make it an ideal
location to further test this approach.
Nearly a hundred and fifty years of rock art research (mostly in Western
Europe) has taken scientists in many new directions and has greatly enhanced our
knowledge of the corpus of rock art sites. Until recently most rock art research in
the Western United States has focused on identifying and documenting sites and
motif styles and this effort has produced a large and useful body of literature that
continues to grow. Other efforts to attach decipherable and perhaps mythological

meanings to the imagery have proven less successful (particularly in areas where
local informants are not available) but a few recent attempts to interpret rock art in
terms of its combined location in the landscape and its cultural relevance (Bradley
2002; Loendorf 2004; Lewis-Williams 2002; Ross and Davidson 2006; Waller
2006) have been promising. This contextual approach (Ross 2003) suggests that a
complex interplay of the socio/cultural, physical, and cognitive aspects of human
prehistoric behavior are collectively responsible for the production and modification
of rock art sites and site complexes over time. In this way persistent places,
connected places, and culturally sacred places can be and likely have been created
(and re-created) for ceremonial or ritual purposes out of otherwise isolated spaces in
the landscape (Crumley and Marquardt 1990: 78).
The intent of this research project is to identify a handful of key variables
(out of two dozen or so possibilities) that can be linked to each of Rappaports seven
characteristics of ritual behavior and thereby demonstrate and define the structure of
ritual as it relates to rock art in the study area. The procedure for accomplishing this
will be discussed in Chapter 3 (Methods).

Physiography and Geomorphology
The study area is geologically defined as the Uncompahgre Plateau which is
located in the northeast quarter of the Colorado Plateau in west-central Colorado
and is characterized by very rugged terrain with high flat plateaus and mountain
peaks interspersed by deeply cut ravines and canyons with vertical drops several
hundred feet deep (Figure 3-1).
The Colorado Plateau is a large geological area in the Eastern Great Basin
and is the second-largest plateau in the world (Childs 2006: 449). The Colorado
Plateau encompasses large portions of the Four Comers states (Colorado, New
Mexico, Arizona, and Utah) and two local subdivisions of the plateau are
recognized; the Uintah Basin and Canyonlands (Cordell 1997: 2). The Uintah Basin
is to the north of the Uncompahgre Plateau and is not part of the study area. The
Canyonlands subdivision is just south of the Tavaputs Plateau (otherwise known as
the Roan Plateau in Colorado) and encompasses the majority of the Uncompahgre
Plateau (ONeil 1984). The Uncompahgre Plateau is a 25 to 30 mile wide anticline
which extends northwest from the San Juan Mountains approximately 100 miles

into eastern Utah and encompasses approximately 3,500 square miles (ONeil 1993:
Figure 3-1. Colorado Plateau and Uncompahgre Plateau. Drawn by the author.
The Canyonlands have been largely shaped by the Uncompahgre, Gunnison,
Colorado, San Miguel and Dolores Rivers (Figure 3-2) that have eroded the
relatively soft Upper Cretaceous Period Mancos Shale (that caps the Dakota
Sandstone) to form deep, wide valleys. Mesozoic Era formations consist primarily
of sandstone, shale, and coal formations. Dakota Sandstone is a Mesozoic Era
formation that common in the study area (Reed and Metcalf 1999: 10) and is often

associated with rock art. This is an Upper Cretaceous Period deposit that is a tan to
light-brown sandstone with a Lower Cretaceous white conglomerate at the base
(Bishop and Collins 1985) that often erodes faster than the exposed Dakota
Sandstone and creates large hollow depressions below the Dakota Sandstone rock
face (Table 3-1). More recent Cenozoic sedimentary formations in the study area
include siltstone, marlstone, claystone, musdstone, sandstone, and conglomerates.
Tertiary-aged rocks include basalt flows, andistic lavas, tuffs and breccias, and ash-
flow tuff. Felsic composition intrusive rock outcrops are not uncommon.
Table 3-1. Jurassic through Tertiary Period Geologic Formations of the Study area (from
Bishop Collins 1985).
Age (million years) Formation Feet Thick Description
0-1.6 Alluvium and colluviums 0-50 Unconsolidated sand, gravel, silt and debris.
1.6-66 Extrusives and related intrusive 0-200 Flows, dikes, sills and intrusions of igneous rocks.
Wasatch Formation 0- 1,500 Variegated clay, shale, sandstone with limestone and conglomerate.
66-98 Mesaverde Formation 0-2,600 Tan and brown sandstone, siltstone, mudstone, coaly shale, coal.
Mancos Shale 0-4,000 Gray and black marine shale. Especially thick in Uncompahgre Valley.
DAKOTA SANDSTONE 100-210 Tan, yellow, light-red, and light-brown sandstone with conglomerate lenses and coarse white conglomerate at base.
98 144 Burro Canyon Formation 0-210 White, gray, and red sandstone and conglomerate.
144-208 Morrison Formation 200-500 Brushy Basin Member is interbedded red, purple, blue, green, and gray mudstone and siltstone. Salt Wash Member is light-brown, gray, and rusty-red, cross-bedded sandstone and mudstone.
Wanakah Formation 60-130 Greenish-gray and reddish-brown sandstone.
Entrada Sandstone 80-250 Massive, white, tan, pink or salmon-colored cross-bedded sandstone.

Elevations in the study area range from 1311m (4300 ft) to 3139 m (10300
ft). Very little area is below 1524 m (5000 ft) or above 3353 m (11000 ft). The
middle elevation zones (Reed and Metcalf 1999: 8) between 1829 m (6000 ft) and
2438 m (8000 ft) make up 50 percent of the study area. The mean elevation is
approximately 2395 m (7859 ft). This is a mere 41 m below the conventional
cutoff for a high altitude region which is generally defined as 2500 m (8000 ft)
(Moore, Niermeyer, Zamudio 1998: 27).
The formal Uncompahgre Plateau is a high domed upland rising from the
Colorado River and peaking at 3139 m (10,300 ft) at the top of Horsefly Peak. The
summit ridge of the plateau is relatively flat with an average elevation of 1981 m
(9500 ft) (Uncompahgre Plateau Project 2009). This area shares some similarities
with extant high altitude environments occupied by humans in areas of the Peruvian
Andes and the Tibetan Himalayas. Environmental features such as ambient cold
temperature, high ultraviolet solar radiation, limited vegetation, rugged
geomorphology, and low oxygen conditions (Beall and Steegmann, 2000; Weinstein
2007: 36) are typical in these higher elevation zones. This altitudinal variation is an
extremely important part of the seasonal activities, settlement patterns, migrations
and hunting and gathering cycles of the prehistoric and proto historic human
occupants of the region.

Stone resources in the study area of interest to archaeology include
sandstone, volcanic extrusives, rhyolite, and cryptocrystalline silicates. Sandstone
naturally spalls into thin tabular sections of up to a few inches in thickness and has
been used in the region in human prehistoric masonry structures without much
modification (Reed and Metcalf 1999: 10). Vertical sandstone outcrops are also
the preferred location for rock art motifs. Cryptocrastalline silicates (chert,
chalcedony, jasper, and petrified wood), quartzite, and rhyolite sources have been
documented in the study area. Some limited utilization of a Cochetopa Dome
obsidian source south of Gunnison, Colorado have been documented in the study
area although the surface nodules are generally too small for reduction (Reed and
Metcalf 1999: 10). Many of these lithics have been used primarily as tools for
hunting and hide processing although they have also been identified in association
with petroglyph locations and petroglyph production. In addition, quartz nodules
are sometimes associated with rock art sites in the study area. Quartz has a unique
quality in that it is a triboluminescent. This means it will generate light when
mechanically shocked by rubbing or striking two quartz cobbles together (Whitley
2005: 11). The cognitive and religious implications of this phenomenon in
association with rock art production are potentially important and therefore the
association of quartz manuports in the vicinity of rock art panels is potentially as

important as the identification of the lithic tools which were utilized in the
production of the petroglyph motifs.
Soils in the study area vary by elevation. Above 2438 m (8000 ft) in
elevation the surface soils tend to be light-colored and loamy. Between this
elevation and above 1829 m (6000 ft) the soils are loamy clays and can be found on
moderately sloping sagebrush- and grass-covered plateaus. Above 1372 m (4500 ft)
the soils in the Gunnison and Uncompahgre valleys are loams and clays that support
shadscale and greasewood. Soils below this elevation are shallow loamy but sandy
soils with clay. They are shallow, dark-colored, friable and lack distinct horizons.
The hydrology of the region is dominated by the Colorado River which
drains most of the study area. The confluence of the Colorado and Gunnison rivers
is in Grand Junction which is on the northern periphery of the Uncompahgre
Plateau. The Gunnison River has several large tributaries, including the
Uncompahgre River which joins the Gunnison River at Delta. The Uncompahgre
River drains the northern slope of the San Juan Mountains and the eastern flank of
the Uncompahgre Plateau. The Dolores River in the southwestern portion of the
area drains the San Miguel Range and the salt anticlines near Paradox. Its major
tributary is the San Miguel River which extends northeasterly from Telluride and
skirts the western side of the Uncompahgre Plateau. The drainages formed by these
rivers and their tributaries often have eroded steep sandstone valleys several

hundred feet high and the confluences of these drainages physiographically pre-
determines the migratory patterns of large game and their hunters. Put another way,
due to the rugged nature of the Canyonlands the hydrological drainage patterns of
the region have apparently pre-established the natural migration routes for game and
for the prehistoric human populations. The geologically distinct Unaweep Canyon
cuts across the northern portion of the Uncompahgre Plateau and is not home to a
river at the present as a result of the gradual uplift of the Uncompahgre Plateau over
the last several million years.
Climate and Environment
Western Colorado has low humidity and high evaporation rates. Moisture
moves into the area from the west and is usually driven by air currents originating in
the Pacific Ocean although localized factors can significantly affect climatic
conditions (Reed and Metcalf 1999: 15) within a distance of a few kilometers.
This results in distinct micro-climatic variations throughout the area. Both
precipitation and temperature tend to increase with elevation.
Relatively warm days and cool nights are characteristic of the region.
Precipitation levels are highest in the winter months and lowest in June. Flash
floods produced by intense thunderstorms in the summer are common. Although
spring floods are uncommon the spring snowmelt raises the volume of the rivers and
the velocity of the water flow significantly. This seasonal runoff pattern is well-

established and several river crossings are not usable during the spring runoff
Figure 3-2. Confluence of Escalante Creek (background) and the Gunnison
River showing high water spring runoff.

Figure 3-3. Uncompahgre Plateau Hydrology. Drawn by the author.
Temperature varies by elevation and the mountains are usually cooler than
the valleys and lowlands. Mountain summits have an average annual temperature of
only 32F. Mountain temperature variations can be extensive; temperatures can
reach roughly 81F in the summer daytimes and exceed -51F in the winter on clear,

still nights (Reed and Metcalf 1999: 15). Temperature relates to the length of the
growing season and influences plant and animal distributions and productivity. The
study area has a relatively short growing season and in well-watered areas the
biomass for human and animal consumption peaks at roughly the same time which
is an attractive situation for both humans and animals (ibid).
A wide variety of plant species can be found in the study area and these tend
to be generally associated with elevation, except for the riparian eco zones which
follow stream and river beds that drop thousands of feet in elevation as they flow
toward the Colorado River. General plant communities include riparian, saltdesert
shrub, sagebrush, pinyon/juniper, montane, subalpine, and alpine communities.
Microenvironmental settings, such as slope, sun and wind exposure, soil warmth,
and available water (Reed and Metcalf 1999: 16) can greatly affect the distribution
of plant types within the local floral community. This patchy distribution of plant
resources is common throughout the Uncompahgre Plateau.
Common floral species in the study area beginning at the lower elevations
and moving to the higher elevations include cottonwood, willow, and Tamarisk (an
introduced species), saltbush, rabbitbrush, galleta grass, Indian ricegrass,
greasewood, piny on, juniper, wheatgrass, bluegrass, needlegrass, forbs, Gambel
oak, Douglas-fir, blue spruce, white fir, aspen, fescue, muhly, bluegrass, shrubs,

spruce fir, lodgepole pine, aspen, sedges, willow, and birch. In the higher elevations
willow and grasses dominate streamside communities.
S_lev. (ft ,i Distributiion on South-facing Slopes \K Distribution on North-facing Slopes
Figure 3-4. Vegetation Zones of west-central Colorado. Adapted from ONeil 1993. Drawn by
the author. Note the locations in the study area fall between 4,760 ft and 6,401 ft. and are all at
the Upper Sonoran/Foothills Transition Life Zone.
The wildlife includes mule deer, elk, bighorn sheep, and pronghorn. Some
whitetail deer can be found in the study area but mule deer predominate. Bison
ranged into the Uncompahgre Plateau in historic times but were never very
common. Localized populations of moose occur to the east and mountain goats can
be found in some high mountain ranges, but these two species were recently

introduced and were unavailable to prehistoric peoples during the Holocene. These
animals spend winters in the lower to middle elevation zones, generally in
pinyon/juniper zones and in the spring most migrate to the higher elevations to take
advantage of the abundant vegetation, and possibly to avoid swarming insects.
Much of this wildlife including deer, elk, and sheep are depicted in the rock art of
the study area.
Larger carnivores and omnivores follow the same seasonal movements.
Black bear, grizzly bear, mountain lion, and gray wolf ranged across this region in
the nineteenth century. Black bear and mountain lion are still present. Smaller
carnivores including coyote, lynx, bobcat, wolverine, badger, long- and short-tail
weasel, mink, marten, ringtail raccoon, swift fox, gray fox, red fox, and kit prey
upon smaller species that do not migrate. These smaller species include marmot,
various rats, mice, voles, gophers, squirrels, prairie dogs, snowshoe hare, cottontail,
black-tailed jackrabbit, and pika. River otter, beaver, and muskrat were once
common along streams and rivers in the study area. Bats are common as are several
hundreds of species of birds. Common species include Canada goose, ducks,
sandhill crane, snipe, blue grouse, sage grouse, sharp-tailed grouse, wild turkey,
ptarmigan, and Gambles quail. Chucker and ring-necked pheasant and other
popular game birds have been recently introduced. Reptiles include snakes and
lizards, frogs, toads, and salamanders. Insects probably comprise more of the

faunal biomass than all of the other species (Reed and Metcalf 1999: 20) and were
likely consumed by prehistoric peoples.
Paleoenvironmental models for the study area vary greatly and are largely
dependent upon seasonal summer monsoon moisture and monsoonal flow. When
the monsoon is weak its effect diminished northward and when it is strong its effect
extends into the study area. Topographic variability has a large effect on localized
microclimates. Paleoenvironmental reconstructions on the broad scale provide a
general backdrop for large areas and spans of time of the study area and include
glacial sequences, solar radiation measures, general circulation models, fossil insect
studies, pack rat midden studies, and other datasets (Reed and Metcalf 1999: 21).
Local records for particular places have been successfully reconstructed based on
pollen records and bog stratigraphy.
The broad models suggest a series of changes in late Pleistocene-Holocene
climates. About 18 ka most data indicate full glacial conditions in the region with a
cold, dry climate. Estimates of deglaciation in the Rocky Mountains vary from
about 15 ka to 13.5 ka, depending on the mountain range (Reed and Metcalf 1999:
23). By 12 ka conditions were warmer and wetter than before although they were
an estimated 5-9 F cooler than today and a good deal wetter than today. After 9 ka
the Southern Rocky Mountains, including most of the Colorado Basin, do not

experience the drought conditions characteristic of the Middle Rocky Mountains
and the Great Basin. The last 6,000 years are not modeled on the same grand scale
but in general the record shows substantial variability around the modern norm
By 6 ka there was less seasonality in temperatures, with a weaker overall
summer monsoon, and the northern part of the study area was as dry as it is today.
There probably was not an Altithermal over most of the upper Colorado Basin
(ibid). The Dansgaard-Oeschger cyclical variations that have been deduced from
ice cores, ocean cores, and larger continental lake cores suggests cycles of climate
change operating on 1,000 to 2,000 year intervals. Each cycle is initiated by a rapid
rise in temperature followed by a gradual return to moderate conditions over about
1,000 years and the cycle ends with a rapid return to very cold temperatures just
prior to the start of a new abrupt warming trend that begins the next cycle. The
amplitude of these fast (decade or two) temperature shifts is on the order of 9-14 F
during the Pleistocene and about 2-5 F during the Holocene (Reed and Metcalf
1999: 24). These abrupt millennial shifts (ibid) suggest a stepwise climactic
change resulting is greater adaptive stress for prehistoric peoples than do the more
gradual models.
Paleoclimatic estimates based on pollen, plant macrofossil, fossil insect,
glacial, and sediment data suggest that after 4000 years B.P. the upper timberline

dropped 100-200 m (328 650 ft) and subalpine fir dominated the upper elevations.
This was caused by slightly drier conditions and a shift to winter-dominated
precipitation. Upper timberline has been near its present position for the last 2000
years, with essentially modem conditions present (Reed and Metcalf 1999: 25).
Brief paleosol-forming periods of stability are noted about 2200 B.P. and 1500 B.P.
with deterioration again after 1000 B.P. and a period of higher effective moisture
after 600 B.P.
There is compelling evidence that occupation of the study area did not begin
until considerable warming had occurred and the strong seasonality inherent to high
altitude interior continental settings such as the Uncompahgre Plateau likely placed
severe limits on winter use of the higher elevations in prehistory just as it did in
protohistoric times.
Prehistoric and Modem Inhabitants
The prehistory of the study area is comprised of four eras, each with several
subdivisions referred to as traditions, periods, or phases. The oldest is the
Paleoindian Era dating from approximately 11500 B.C. to 6400 B.C. This was
followed by the Archaic Era from 6400 B.C. to 400 B.C. Next came the Formative
Era dating from 400 B.C. to A.D. 1300. The final era was the Protohistoric era
which lasted from A.D. 1300 to A.D. 1881. (Reed and Metcalf 1999: 6). All of
these dates are approximate.

Human occupation of the Northern Colorado Basin commenced with the
Paleoindian Era. No evidence of a Pre-Clovis occupation has been found in the
study area. The Paleoindian Era reflects adaptations of the earliest immigrants to
the New World during the time of great environmental transformations at the end of
the Pleistocene. There is no known rock art associated with the Paleoindian Era in
the study area.
The Archaic Era encompasses a long relatively stable period when a broad-
based hunter-gatherer lifeway was practiced. It contrasts with the preceding
Paleoindian Era in that the Archaic lifeway was less mobile and more focused on
the use of local resources on a scheduled seasonal basis (Reed and Metcalf 1999:
71). The main technological marker is a transition from the use of a lanceolate
projectile point to the use of stemmed and notched point varieties and an increase in
the overall variability in point styles. There are numerous rock art sites in the study
area that can be dated to the Archaic Era; two in particular can be done so based
largely on the presence of rock art depictions of atlatl technology.
The Formative Era in the study area saw the continuation of the relatively
stable and effective Archaic adaptations with some subtle changes. Subsistence
efforts were intensified, a limited use of cultigens such as com, beans, and squash
occurred in those (small) areas at lower elevations with sufficient growing seasons,
and there was a more intensive use of processed plant foods. The major

technological marker was the adoption of the bow and arrow, a decrease in
projectile point variability, and the limited use of ceramics. The cultural affiliations
of sites in this era include Fremont, Gateway, and Ancestral Puebloan (formerly
Anasazi). An Aspen tradition has been proposed but has recently been disputed and
will not be discussed here. The Ancestral Puebloan tradition is defined by
substantial habitation structures, pottery styles, and distinctive rock art styles. The
Fremont tradition is characterized by four main attributes: one-rod-and-bundle
basketry, moccasins made of deer or mountain sheep skin (not woven), trapezoidal
anthropomorphs in rock art and as clay figures (often with elaborate ornamentation),
and a distinct coiled pottery tradition. The associations between Fremont rock art
figures and Fremont archaeological deposits are difficult to establish (Reed and
Metcalf 1999: 109). The Gateway tradition is proposed as a separate, indigenous
group that was considerably influenced by both the Fremont and the Anasazi
traditions (Ancestral Puebloan) (Reed and Metcalf 1999: 131) and is characterized
by the following attributes: limited com horticulture, small corner-notched
projectile points; procurement of a small amount of Ancestral Puebloan and
Fremont ceramics, habitation of circular and rectangular masonry surface structures
and possible habitation of pit structures, relatively short-term use of habitation
structures (as evidenced by shallow middens), construction of granaries and storage
cisterns in rockshelters, and rock art with both Ancestral Puebloan and Fremont

influences. It was thought that there was a lack of inherent ceramic production in
this Era but this has been recently refuted (Greubel, Andrews, and Reed 2006: 87,
Protohistoric Era aboriginal occupation occurred between the end of the
horticultural-based subsistence practices of the Formative Era and the final
expulsion of the Ute to reservations (Reed and Metcalf 1999: 146). The dividing
line between the end of the Formative era and the beginning of the Protohistoric Era
is not well-defined. The Ancestral Puebloans migrated to New Mexico and Arizona
by A.D. 1300, at about the same time that the Gateway peoples either emigrated out
of the area or altered their lifeways to the point that they became undistinguishable
in the archaeological record. The Fremont tradition began to contract at about A.D.
1250. There is evidence of the immigration of a new hunting and gathering group
into the area at or shortly before A.D. 1300 but it is unclear as to whether this group
assimilated or displaced their predecessors (Reed and Metcalf 1999: 146).
Protohistoric-era peoples in the study area were highly mobile hunter
gatherers who constructed wickiups for shelter, manufactured brown ware ceramics,
and hunted with bows and arrows. Spanish explorer Juan Rivera first recorded
reports concerning the presence of the Utes in Colorado in his diary in 1765 and
they were later documented throughout the region by Fathers Dominguez and
Escalante during the famous Dominguez-Escalante expedition of 1776. This was

the first European excursion into the rugged and uncharted areas of western
Colorado. By 1680 (Cassells 1997: 244) the Ute in the study area had begun
acquiring enough horses to transition to an equestrian lifeway thereby expanding
their territory. Although records indicate that the Ute were the primary occupants of
the Northern Colorado Basin (including the Uncompahgre Plateau) since the late
eighteenth century the Shoshone may have occupied the extreme northwestern
portion of the state (beyond the Uncompahgre Plateau). The 1776 Dominguez-
Escalante Expedition journal reported Comanche in the study area although they
were not directly encountered by the Dominguez-Escalante party (Reed and Metcalf
1999: 146). Virtually all subsequent historic records from the region indicate
occupation by the Ute.
The Utes are a Numic speaking peoples who have received the research
attention of such anthropological notables as Julian Steward (Baker 2007) and
Edward Sapir (Bright 1992). Sapir conducted an exhaustive study of the Ute
language and ethnography in the early twentieth century for which he is famous.
For this work Franz Boas called him one of the most brilliant scholars in linguistics
and anthropology in our country (Bright 1992: 2). Julian Steward studied the Utes
extensively in the early twentieth century and was involved in an infamous legal
battle which pitted him against his student Omer Stewart in a U.S. Indian Claims
Commission court case designed to establish the relationship of the Ute bands to

certain geographical territories in western Colorado. Omar Stewarts interpretations
of Ute prehistory prevailed and helped settle a landmark 1957 case in favor of the
Utes. This settlement proved to the courts satisfaction that the Utes had ancestral
claims to lands in Colorado that had been unlawfully taken from them by the
United States (Baker 2007: 37). Omer Stewart noted that the Utes share an
emotion charged belief that they are alike and united in opposition to other Indians
and non-Indians (Baker 2007: 45) in their claims to western Colorado.
The Utes refer to themselves at the Nuche or Nuutsiyu which translates to
the Sun Mountain People (Office of Community Services, Fort Lewis College
2003: 2). According to the The Utes of West Central Colorado they are a land
based culture, a tribal society that cannot be understood without knowing about their
attachments to their mountain environment emphasis added (Office of Community
Services, Fort Lewis College 2003: 2). Their habitation areas typically extended
from the winter zones or warm basins and plateaus at 1542 m (5000 ft) to 3048 m
(10,000 ft) in the summer when the mountain meadows became lush and attracted
game (Office of Community Services, Fort Lewis College 2003: 25). The first
treaty between the Ute people and the United States government was signed in
1850. The first removal occurred in 1874 which restricted their access to portions
of the San Juan Mountains. That same year Ernest Ingersoll wrote in his journal
that Utes resided on their eastern Utah acre reservation only in the winter (Baker

2008: 83) and continued to occupy their high altitude ancestral homelands during
the spring, summer, and fall. Upon their final removal in to reservations, after the
Meeker Massacre (Cassells 1997: 244) the Utes were assigned to three
reservations based on their group affiliation. One is located in eastern Utah and and
two are located in southwestern Colorado. Todays Utes routinely leave the
reservations to visit sites and locations in western Colorado that were important and
often sacred to their ancestors.
The Ute trails through these same mountains are still used today. The Utes
are, in regard to trails, also known as fleet-footed runners (Baker 2008: 32) who
traveled extensively by foot along native trails that followed mountain ridgelines
and crossed the high mountain passes. Many of these trails were unsuitable for
horses (op cit) and partly for this geomorphological reason the Utes were one of
the last groups of Native Americans to adopt horses. Ethnographically the high
mountain trails were sacred to the Utes and when moving from place to place they
would not leave the trail even though a shorter and better road is very perceptible
(Office of Community Services, Fort Lewis College 2003: 21). Many of the
original high altitude trails became the basis for roads and others are visible in
the landscape or have otherwise been identified and today serve as physical
reminders and evidence supporting the Ute claim as the mountain people.

The historical and archaeological record suggests that the Utes occupied the
mountains of western Colorado for at least several centuries (Office of Community
Services, Fort Lewis College 2003: 10) and perhaps much longer if a continuum can
be established linking historic Ute to the existing prehistoric record of the
Uncompahgre Plateau which extends at least to the middle Archaic era ca. 3,500-
4,500 B.P (Cassells 1997: 113; Reed and Metcalf 1999: 6). It is of note that at least
one researcher questions the continuity of a year-round proto historic presence of
the Ute in stating that the highest elevations were, not inhabited permanently until
150 years ago (Niermeyer 2001: 43) although their seasonal presence in the study
area over a longer period of time is undisputed. The historical and archaeological
evidence suggests that Colorado was occupied on more than a seasonal basis for
many thousands of years. As Ute linguist James Goss notes their adaptation to
mountain living should be treated as a cultural focus (Baker 2007: 36). Some
archaeologists do not see a discontinuity (Kaestle and Smith 2001: 2) in the material
culture record that would suggest a wholesale replacement of the prehistoric
population by recent arrivals and they have proposed that the Numic presence in the
Great Basin is possibly quite ancient (ibid). Recent analysis of mtDNA among
Native Americans shows that the haplogroup distribution of the ancient inhabitants
of the Great Basin is most similar to those of some of the modem Native American
inhabitants of California which tends to support the linguistic origin theories,

however, the same mtDNA study observes that a random admixture (Kaestle and
Smith 2001: 10) between the ancient pre-Numic residents of the area and Numic
immigrants could have produced the haplogroup distribution that we see today.
The descendants of the Utes now live on three reservations: the Southern
Ute Reservation of Colorado, the Ute Mountain Ute Reservation of Colorado, and
the Uintah and Ouray Reservation of Utah and collectively number approximately
7,000 enrolled tribal members on the three reservations (Office of Community
Services, Fort Lewis College 2003:53).
Present Environment and Threats
The study area extends slightly beyond the confines of the 3,500 square
miles that comprise the Uncompahgre Plateau. Several federal and state agencies as
well as private landowners control a mosaic patchwork of surface access and
subsurface mineral rights in the area. The United States Forest Service and the
United States Department of the Interior, Bureau of Land Management, each control
approximately 37% of the land area, or about 545,000 acres each. Private
landholders control 25% of the land area, or about 365,000 acres, and the Colorado
State Land Board and the Colorado Division of Wildlife control approximately 1%
of the land area, or about 8,600 acres (Bureau of Land Management, 2009). In
addition to these stakeholders, the United States Park Service administers the
Colorado National Monument (located at the northwest end of the Uncompahgre

Plateau near the confluence of the Gunnison and Colorado Rivers) which is in the
northern portion of the study area. Of note is the Tabeguache Area which is
comprised of 17,240 acres in the heart of the Uncompahgre Plateau. This area is
part of the Dominguez Escalante National Conservation Area which was established
by Congress this year (2009). Recreational, hunting, energy development, and
natural resource utilization activities all contribute to the anthropogenic
modifications of the Uncompahgre plateau. The major recreational activities on
public lands in the study area include hiking, backpacking, horseback riding,
fishing, river rafting, and wildlife viewing. The area is popular for seasonal hunting
and fishing. Energy development and natural resource utilization is not as extensive
as in the areas to the north such as the Roan Plateau but include uranium mining (a
new uranium mill is under construction in the west side of the study area near the
Utah boarder), natural gas drilling, logging, cattle grazing, and subsurface coal
mining. The remnants of small and large-scale nineteenth and twentieth century
prospecting and mining enterprises can be found throughout the landscape and some
are of historic significance. Portions of the western section of the study area, such
as Nucla, have undergone remedial mitigation by federal agencies such as the
Environmental Protection Agency in an effort to contain radioactive tailings from
uranium mining and processing activities in the mid twentieth century. Both the
Forest Service and the Bureau of Land Management engage a philosophy of

multiple resource management which promotes the co-existence of recreational and
commercial interests on the public lands throughout the region. All of these
activities potentially affect the prehistoric resources of the region, including the rock
A case in point is the recently approved uranium mill slated to be built near
Nucla and Naturita by Energy Fuels, Inc. This project is estimated to generate 85
new jobs in two communities with populations of approximately 700 each (Finley
2009). The increased industrial and mining activity in the region will likely have
both direct and indirect effects on the local prehistoric resources.
As mentioned in the introduction, one of the outcomes of this research
project will be an increased understanding of the placement of rock art imagery in
the overall physical (and cultural) landscape of the region. This will assist
archaeologists, cultural resource managers, and others as they make land use
decisions in the future and weigh the relative costs and benefits of preservation and
development in the region. The next chapter discusses the methods employed in the
research project and identifies 20 variables that were subsequently subjected to
quantitative analysis. Some of these variables, for example, include measurements
that go beyond recording and categorizing the images themselves and capture data
concerning the relationship of the rock art panels to other physiographic features
including viewshed, acoustical characteristics (echoes), and visibility.

This section will address the field recording methods, sampling procedures,
datasets, and methods of data analysis employed in the research project.
Justifications will be provided linking the data attributes to one or more of each of
the seven characteristics of ritual (identified in the theory section above) and the
expected limitations of the data elements will be identified.
The fieldwork goal was to acquire data from at least twenty rock art panel
locations in the Uncompahgre Plateau. This data included standard locational
information, photographs, drawings, and measurements for twenty different kinds of
variables that could potentially provide information linking each panel location to
one (or more) of Rappaports seven characteristics of ritual behavior.
Fieldwork and Sampling
Fieldwork was conducted over an eight month period between August 2,
2008 and March 1, 2009. At the time that fieldwork commenced there were
approximately 122 known rock art sites in Delta, Dolores, Mesa, Montrose, Ouray,
and San Miguel counties which are the political subdivisions that encompass the

study area. The vast majority of known sites are located in three of the counties;
Delta, Mesa, and Montrose. These three large counties also account for the vast
majority of the land area in the Uncompahgre Plateau and are generally under the
jurisdiction of the Bureau of Land Management or the United States Forest Service.
Not all of the known sites were on record with the Office of the State Archaeologist.
A stratified purposive sampling strategy (Patton 2002: 240) was selected in
order to identify between 20 and 40 sites out of the known population for
recordation. Three criteria were identified for selecting sites: 1) the sites should be
geographically dispersed broadly over the study area; 2) the sites should vary in
elevation as much as practicable so that multiple bio-zones are represented in the
dataset (this proved impracticable); and 3) the sites should be accessible or
relatively accessible from existing public vehicular right-of-ways. The first
criterion was selected in order to ensure that the database represented as much of the
study area as possible and to prevent a large portion of the data from being collected
in just one small geographic section of the Uncompahgre Plateau. The rationale for
this was to ensure that unique local features did not predominate the data, such as
might be the case if all the sites were from the same east-west drainage system, for
example. The second criterion was selected because the elevation of the study area
can vary dramatically and it was thought to be important to attempt to collect data
from as many different elevations as possible (this proved to be impractical as

explained below). The third criterion was selected because of a limited field season
and the fact that it was important that each site could be safely accessed, fully
recorded, and exited during the daylight hours. The drive time on paved and dirt
roads from the home base in Montrose could reach four hours each way, plus two
hours on foot each way, plus two to three hours on site and this meant the possibility
that each site could require up to 16 hours of travel and recording time. This third
constraint proved to be the deciding factor in the selection of many of the sites in the
sample area. Sites more than a two hour walk each way from the nearest vehicular
access point, or which required permission to cross private land in order to access
were excluded from the study.
The final sample size was made up of 22 rock art panel locations. This was
in the low end of the target sample range. These 22 locations were located in 17
separate archaeological sites. The data from these locations makes up the complete
database for this project. This represents approximately 14% of the known rock art
sites in the Uncompahgre Plateau. As a preliminary research project this sample
size was deemed sufficient. Clearly adding more sites and locations to the database
would have increased the potential of providing additional information, both in
terms of identifying patterns and unique features. However, by the time 22
locations had been surveyed an observable degree of redundancy (Patton 2002:
264) had been achieved in the data and the summer field season had come to an end

and winter had set in. The final site locations were recorded with snow on the
ground in January, February, and early March of 2009.
All research was conducted in compliance with the ethical guidelines of the
Colorado Council of Professional Archaeologists and the Colorado Rock Art
Association (Colorado Council of Professional Archaeologists 2009, Colorado Rock
Art Association 2009). The primary purpose of the researcher while on site was the
protection of the resources under study. Many of the rock art panel locations in the
study area are considered to be sacred by the Ute and proper and dignified respect
was demonstrated during the visitation and recordation process. Preliminary
research and file checks were conducted at the Denver office of the Colorado State
Archaeologist and in the BLM field offices in Grand Junction and Montrose. Park
Service access permission was secured from the Colorado National Monument
Headquarters. All sites were accessed via existing roadways, paths, public trails or
game trails. While on site every effort was made to ensure that social trails were not
created during the recordation process and no sites or panel locations were
approached when unidentified people were within view. This occasionally delayed
site access, particularly on those sites close to river raft launch locations and hiking
and horseback trails. In areas where there was a high level of pedestrian traffic the
footprints of the research party were erased after recording the sites by gently
brushing them out with a dead tree branch. Most of the sites were visited and

Figure 4-1. Study Area and Panel Locations. Drawn by the author.
recorded in the company of a Registered Professional Archaeologist. Permission
was obtained from the BLM and from the NPS to take careful depth measurements
of the motifs because this involved actually physical contact with certain petroglyph
images. Local informants occasionally volunteered to help the researcher locate
several sites and not all of the sites had been previously recorded, or had been
incompletely recorded. None of the informants were part of the research project.

No human subjects were involved in this research project and therefore a review by
the University Human Subjects Research Committee was not necessary.
Some of the rock art sites in the study area are quite extensive and comprise
many separate rock art panels and panel groupings. It was often a matter of
individual judgment and sometimes the result of unique field circumstances that
ultimately determined what the site boundary was. Add to this the fact that some of
the sites were initially recorded 40 or more years ago (two sites in the data set are
the first sites ever recorded in their respective counties, 5DTI and 5MN1) when site
recording standards were just developing and it becomes clear that the definition of
a site, for the purpose of rock art, was and is more subjective than objective. For
this reason it was decided at outset to use panel locations as the unit of study for the
research project, particularly where several panel locations could potentially be
identified within one site boundary. This was particularly useful in terms of one
site, the Shavano site (5MN5), which yielded five separate panel locations. The
designation of separate panel locations within a site was based principally on the
distance between panels, and was reinforced if the panels were located on separate
and distinct rock faces or rock formations (such as boulders) that were facing in
different directions and/or located at different elevations within the site area.
Because site recordation forms and site data in the available literature is
sparse, inconsistent, and in some cases is limited to short verbal site descriptions it

was not possible to build a dataset for this research project from the existing data on
file with the Colorado Office of the State Archaeologist. If this had been feasible
then the sample size could have been much larger and could have possibly been
statistically randomized. This, however, was not possible.
Data Attributes
Several dozen data points were collected in the field at each panel location
utilizing a standard data recording form was developed for this purpose (Appendix
A) in order to ensure that the field data were obtained in a consistent fashion from
site to site and panel location to panel location. This data form was initially
developed in June, 2008 and field checked in July, 2008 before being put into
service in August, 2008. The field form was retooled slightly again between August
7, 2008 and August 9, 2008 (after seven panel locations had been recorded) in order
to expedite field recording. This retooling process removed the elevation view
sketch of the rock art site which was deemed to be redundant due to the site
photographs. Plans to videotape the site locations were also set aside. Videotapes
proved impractical because of the difficulty of transporting the videocamera in
addition to all the other equipment (by one person on foot in one trip). Also,
panoramic videotapes did not provide imagery or context or scale that was
fundamentally different from the digital photos which could be stitched together
using Photoshop or similar software. Because of the subjective nature of

identifying and interpreting some rock art motifs and panels a special general
notes section was included on all the forms. When quartz nodules or worked lithics
were located they were photographed and drawn. These items were infrequently
located and are not part of the data analyzed in this research project. Digital
photographs were taken of the panels, the motifs, and of the viewshed looking away
from the site location (aspect), as well as an establishing photograph which provides
a general overview of the site (see Figure 4-2 and 4-3).
Figure 4-2. Roc Creek Establishing Photograph (Panel Location 19)

Figure 4-3. Roc Creek Viewshed photograph (Panel Location 19)
Motif photos were taken using a metric scale that was visible in the
photograph in order to provide scale (see Figure 4-9). Sometimes this was not
practicable. Two digital cameras were used for this purpose, one was a Sony Cyber-
shot 8.0 megapixel camera with 7x optical zoom and a 2-2.8/7.1-51 Carl Zeiss lens.
Backup photographs were obtained using a Hewlett Packard Photosmart R927 8.2
megapixel camera with a 3x optical zoom and a 2.8-5.0 lens. On-site sketch maps
were drawn in both profile and plan view which illustrated local physiographic
features, locational elements, and the locations of acoustical and other
measurements which were taken in the field.