ABSTRACT 

 

 

Title of Thesis: AN ARCHAEOLOGICAL INVESTIGATION OF 

CLOVIS BLADE TECHNOLOGY AT 

THUNDERBIRD (44WR11), A PALEOLITHIC 

STRATIFIED SITE OF THE FLINT RUN 

COMPLEX, WARREN COUNTY, VA 

 Kurt N. Fredrickson, Master of Professional Studies, 

2024 

 

Thesis Directed By:  Dr. Matthew Palus, Department of Anthropology 

 

The presence of Paleoindians in the Eastern United States at the end of the Pleistocene 

has been a focus of scientific examination for more than a century, resulting in the discovery of 

numerous sites. These sites, occupied more than ten millennia ago, are extremely rare, and even 

more so in an undisturbed context. The Flint Run Complex in Northern Virginia contains not 

one, but several Late Pleistocene and Holocene open-air stratified Paleoindian sites. Thunderbird 

(44WR11) is the main site within the complex with evidence of human occupation in the region 

at around 9,990 BP. Numerous tools were recovered which fit the Clovis technocomplex and 

extensive analysis has been performed on bifacial technology at the site. Additionally, the 

identification of blades at Thunderbird would support previous assertions that the site was an 

important refugia on a migratory pattern where scheduled resource exploitation and toolkit 

refurbishments took place as part of seasonal rounds. How does the analysis of lithic blade 

production at the Thunderbird site (44WR11) refine our understanding of localized seasonal 

migration and exploitation of local resources among Paleoindian people of the Shenandoah River 

Valley. Confirmation of blades and their use would indicate a more robust exploitation of the 

region’s natural resources and reinforce previous assessments of the importance of Thunderbird 



 

 

as a sedentary seasonal base camp. Through the examination of 324 lithic artifacts from the site, 

this study seeks to identify the presence of a concerted blade manufacturing technology where it 

was believed one did not exist, and better understand the behaviors tied to those tools. Blades are 

a known part of the Clovis toolkit and have been found at sites across the United States. The 

identification of blades at Thunderbird will provide an expanded understating of the Clovis 

toolkit, the spread of blade technology, and of Paleoindian lifeways in the Middle Atlantic 

region. 

 

 

 

 

 

 

 

 

 

 

 

 

 



 

 

 

 

AN ARCHAEOLOGICAL INVESTIGATION  

OF CLOVIS BLADE TECHNOLOGY AT THUNDERBIRD (44WR11),  

A PALEOLITHIC STRATIFIED SITE OF THE FLINT RUN COMPLEX,  

WARREN COUNTY, VA 

 

 

By 

 

Kurt N. Fredrickson 

 

 

 

 

 

 

Thesis submitted to the Faculty of the Graduate School of the  

University of Maryland, College Park in partial fulfillment  

of the requirements for the degree of  

Master of Professional Studies 

2024 

 

 

 

Advisory Committee: 

  Dr. Matthew Palus, Chair 

  Dr. Kathryn Lafrenz Samuels  

  Dr. Mark P. Leone  

  Mr. Lyle C. Torp 

  



 

 

 

 

 

 

 

 

 

 

 

 

 

© Copyright by 

Kurt N. Fredrickson 

2024



ii 

 

Forward 

Every artifact provides insight only after finding its place in history, just as research 

contributes to knowledge by building upon what is already known. This study began with a 

prehistoric stone blade, or rather several, and a question about their origin. In 2022 I was 

fortunate to become involved in a private archaeological investigation in the Shenandoah River 

Valley of Northern Virginia supported by Dr. Robert Richards of Jasper Ridge LLC and endorsed 

by the Archaeological Society of Virginia. It was the first such undertaking in decades in the 

Flint Run Complex, with a focus on Paleoindian history in the region. The area is rich in 

archaeological sites dated as far back at the terminal Pleistocene. The site in question lies on a 

ridge overlooking the valley and has evidence of thousands of years of occupation. Recorded as 

44WR506, Jasper Ridge encompasses two acres and sits adjacent to an extensively used jasper 

quarry. It has produced more than 50,000 catalogued artifacts to date. However, its disturbed 

stratigraphy and significant collecting has made dating of the site inconclusive and at times 

frustrating. In the process of cataloging the many lithic artifacts, primarily comprised of local 

jasper, the quantity of blades and blade-like flakes became apparent. The presence of these 

blades and blade-like flakes, like those found in other Clovis contexts, began to raise questions 

about a possible Early Archaic or even Clovis age blade industry at the site. We soon realized we 

would need to look at other collections for comparative insight to better understand the artifacts 

at Jasper Ridge. Fortunately, less than a mile from Jasper Ridge is the Thunderbird site, 

(44WR11).  

  This well preserved Late Pleistocene site underwent excavation for more than a decade 

and produced solid evidence of Clovis habitation from undisturbed stratified levels more than 

three feet below datum. The years of research that came out of Thunderbird would go on to 



iii 

 

establish settlement patterns and ecological understanding for the region and shed light on 

Paleoindian lifeways across the Mid-Atlantic. By the 1980s Thunderbird lost its notoriety, and 

the collections made their way to the archives of the Smithsonian where researchers occasionally 

visited them. Despite being locked away, the work at Jasper Ridge afforded an opportunity to 

access the collection of more than 100,000 artifacts. But mention of blades was absent from all 

literature related to Thunderbird, except for explicit statements that true blades did not exist at 

the site.  

  A complete examination of the collection was a considerable time commitment, and one 

that had no guarantee of paying dividends. Would Thunderbird hold the answers to Jasper Ridge, 

and could it provide the necessary baseline to understand the distinct blade technology being 

employed there? More importantly, were the same people inhabiting Thunderbird also using 

Japer Ridge as an alternate camp as part of their seasonal rounds? One thing remained certain: 

evidence of a Clovis blade industry would only be found in the deepest levels of the Thunderbird 

collection. In this regard, the research done at Thunderbird more than 40 years ago continues to 

support a broader knowledge of history in the region as a legacy collection.  

It is beneficial to anyone who reads this to understand my mindset when tackling these 

broad and complex questions. Each researcher views a problem differently and leverages their 

unique life experiences and academic background to shape the approach to their hypothesis. I am 

new to archelogy, but I am not new to research. I have spent decades studying and working with 

communication theory, specifically related to how it shapes human behavior. The communication 

theories underlying behavior have great parallels with archaeology, which I believe at its core 

seeks to unravel behavior through the examination of artifacts. Communication theories, such as 

the theory of planned behavior (Sussman and Guifford 2019) and diffusion of innovation (Rogers 



iv 

 

2003), leverage the complex combination of internal and external factors to explain behavioral 

outputs, as well as explain how behavior changes in response to inputs. These theories share 

parallels to those found in processual and cognitive archaeology and their examination of 

systemic processes. From weather to culture, everything plays a role in shaping behavior, and I 

believe the output, or artifact, can tell us what types of factors influenced its creation. I view 

every artifact as a behavioral output, shaped in subtle ways by the world around its creator. 

Blades are no exception. Their techniques of manufacture allude to diffusion of innovation 

between peoples, their existence in a particular location implies the type of work done there, use-

wear demonstrates how they were used, and their discard tells another piece of the story. Given 

my view on behavior, I am fully aware that knowing blades exist at a site is only the beginning 

of a larger initiative to understand what influenced their creation in the first place.  

Nowhere in North America is the dispersion of Paleoindian artifacts greater than in the 

Eastern United States, ranging from New England to the Southeast. Research has documented 

many sites with varying degrees of success, creating compartmentalized reports of great detail. 

But we have reached a stage in Paleoindian research where we can begin to share and view 

information about artifacts from these sites on a much larger scale. Comparative analysis of 

artifacts can tell a much bigger story about mobility, material use and transport, social structures, 

and occupation periods. Finding a blade is just the beginning of the story. The existence of a 

blade industry in the Shenandoah River Valley is a necessary addition to a much bigger narrative 

focused on Paleoindians in the East during the Terminal Pleistocene. It is my hope that this 

research provides another subtle addition to the understanding of the role blades played in the 

lives of these people, and thereby a broader understanding of the physical and social 

environment they existed in.  



v 

 

 

 

 

Dedication 

 

 

To the 8-year-old version of myself, who stared endlessly into collection cases at the Peabody 

Museum at Yale, dreaming of lost treasure, adventures, and a really nice hat.  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



vi 

 

 

 

 

Acknowledgements 

 

My gratitude goes to my archaeological mentors, Dr. Mike Johnson, Patrick O’Neill and 

Yvonne French, who have been with me since I decided to commit to the cause of understating 

human origins, and who have provided a constant flow of wisdom and encouragement. My 

thanks go to Dr. Bob Richards for allowing access to Jasper Ridge, and for sharing his 

knowledge of lithic technology. Also, thanks go to Jim Krakker at the Smithsonian Museum 

Support Center for enabling access to the Thunderbird collection. I only got lost in the archives 

once. And I would be coming up short if I didn’t mention the outstanding support of the 

Archaeological Society of Virginia, and the Northern Virginia Chapter of the ASV for setting the 

stage for my entry into archaeology. Finally, my deepest appreciation for my family who 

watched patiently as I sacrificed years of weekends and weekday nights to a quest for knowledge 

that is insatiable and at times compulsive.  

 

 

 

 

 

 



vii 

 

Table of Contents 

 

 

List of Tables .................................................................................................................................. ix 

List of Figures ................................................................................................................................. x 

Chapter 1 ....................................................................................................................................... 1 

1.1 The Blade - An Essential Companion Throughout Human History ......................................... 1 

1.2 Origins and Ice .......................................................................................................................... 4 

1.3 A Story Written in Stone ........................................................................................................... 4 

1.4 Overview ................................................................................................................................... 6 

Chapter 2 ....................................................................................................................................... 9 

2.1 Virginia - A Paleolithic Hotspot in a Cold World...................................................................... 9 

2.2 Paleoindian Timeline and Climate ............................................................................................ 9 

2.3 Searching for the American Paleolithic .................................................................................. 12 

2.4 A Paleoindian Complex in Virginia ........................................................................................ 17 

Chapter 3 ..................................................................................................................................... 22 

3.1 Driving Theories and Models in Paleoindian Research .......................................................... 22 

3.2 Defining Blades ...................................................................................................................... 27 

3.3 Blades as a Key Component of Paleoindian lifeways ............................................................ 32 

Chapter 4 ..................................................................................................................................... 34 

4.1 Research Methods and Methodology...................................................................................... 34 

4.2 Collection, Location, and Condition ....................................................................................... 35 

4.3 Artifact Selection .................................................................................................................... 39 

4.4 Measurement Tools ................................................................................................................. 41 

4.5 Measurements and Recording ................................................................................................. 41 

4.6 Visual Documentation ............................................................................................................. 47 

4.7 Cataloging ............................................................................................................................... 49 

Chapter 5 ..................................................................................................................................... 51 

5.1 Results ..................................................................................................................................... 51 

5.2 Artifact Wholeness .................................................................................................................. 52 

5.3 Maximum length, Width, and Thickness ................................................................................ 52 

5.4 Platform and Bulb Dimensions ............................................................................................... 53 



viii 

 

5.5 Platform Angle ........................................................................................................................ 53 

5.6 Ventral Ripples ........................................................................................................................ 54 

5.7 Modification and Edgewear .................................................................................................... 54 

5.8 Blade Cores and Debitage ....................................................................................................... 55 

5.9 Proximal and Distal Terminus Mends ..................................................................................... 56 

5.10 It’s All In The Numbers ........................................................................................................ 57 

Chapter 6 ..................................................................................................................................... 59 

6.1 Analysis ................................................................................................................................... 59 

6.4 Attribute Analysis.................................................................................................................... 61 

6.2 True Blades at Thunderbird .................................................................................................... 65 

6.3 Macro Blades .......................................................................................................................... 69 

6.5 Crested Blades ........................................................................................................................ 72 

6.6 Blade Cores and Debitage ....................................................................................................... 74 

6.6 Modified and Edgeworn Blades.............................................................................................. 79 

6.7 A New Outlook on Blades at Thunderbird .............................................................................. 82 

Chapter 7 ..................................................................................................................................... 84 

7.1 Observation and Future Opportunities .................................................................................... 84 

7.2 Final Thoughts ........................................................................................................................ 90 

Appendix 1 ................................................................................................................................... 93 

Bibliography ................................................................................................................................ 94 

 

 

 

 

 

 

 

 

 

 

 



ix 

 

List of Tables 

 

Table 1. Artifacts from the Paleoindian-Early Archaic levels at Thunderbird. ............................. 36 
Table 2. Artifacts examined and documented ............................................................................... 51 
Table 3. Blade-like flake wholeness levels 6,7, & 8 ..................................................................... 52 
Table 4. Mean maximum length, width, thickness for all blade-like flakes ................................. 53 
Table 5. Mean platform width, thickness/bulb total thickness and thickness from midline ......... 53 
Table 6. Mean platform angles in degrees .................................................................................... 54 
Table 7. Ventral ripples ................................................................................................................. 54 
Table 8. Blade like flake modification and edgewear ................................................................... 55 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



x 

 

List of Figures 

 

  

Figure 1. A fluted Clovis point and double crested blade from Thunderbird ................................. 3 
Figure 2. The extent of the Virginia coastline at the end of the Late Pleistocene ..........................11 
Figure 3. A map of selected Paleoindian sites in the East ............................................................. 16 
Figure 4. Thunderbird site location map ....................................................................................... 17 
Figure 5. Collections storage units at the Smithsonian Museum Support Center ........................ 37 
Figure 6. Artifacts in drawers at the Smithsonian Museum Support Center ................................ 38 
Figure 7. The author at the Smithsonian archives and associated workspace .............................. 38 
Figure 8. A selection of blade like flakes pulled for further analysis ........................................... 40 
Figure 9. Visual representation of an arris on a small blade like flake ......................................... 40 
Figure 10. Tools and associated materials used for data capture .................................................. 41 
Figure 11. Calipers capturing terminus measurements. ................................................................ 44 
Figure 12. Contour gauge used to capture the thickness of the bulb of percussion ...................... 45 
Figure 13. Graphic showing medial, ventral, and dorsal angle measurement positions ............... 46 
Figure 14. Description of line drawing method ............................................................................ 48 
Figure 15. Example of an artifact with imaging tags .................................................................... 48 
Figure 16. Microsoft Access database data entry form ................................................................. 50 
Figure 17. Broken blades mended using proximal and distal terminus measurements ................ 63 
Figure 18. Medial section of a blade with modification and use-wear ......................................... 62 
Figure 19. Dorsal platform angles represented as a graph ............................................................ 63 
Figure 20. The dorsal platform angle graph .................................................................................. 64 
Figure 21. Bulb of percussion prominance graph ......................................................................... 65 
Figure 22. TB-340, a double crested blade form level 6 of Feature 68 ........................................ 67 
Figure 23. Double crested blade recovered from Feature 68 ........................................................ 68 
Figure 24. A large blade from Thunderbird .................................................................................. 69 
Figure 25. Prismatic macro blade ................................................................................................. 70 
Figure 26. Macro blade with evidence of prior blade removal ..................................................... 71 
Figure 27. Macro blade TB-209 with modification and edge wear .............................................. 71 
Figure 28. Crested blade TB-207 with rough bifacial preparation ............................................... 73 
Figure 29. Left, crested blade TB-248 from level 6 at Thunderbird ............................................. 74 
Figure 30. Depleted blade core TB-314 recovered from Feature 68 ............................................ 75 
Figure 31. Blade core TB-314 recovered from Feature 68 ........................................................... 76 
Figure 32. Blade core debitage from Feature 68........................................................................... 77 
Figure 33. Expended polyhedral core with removal of flakes ...................................................... 78 
Figure 34. Expended polyhedral core with removal of flakes ...................................................... 78 
Figure 35. Expended wedge core with removal of flakes from one face ..................................... 79 
Figure 36. A prismatic micro blade with modification and use-wear ........................................... 81 
Figure 37. Modified and used blade TB-347 ................................................................................ 82 
Figure 38. Scraper on a double crested macroblade ..................................................................... 88 



1 

 

Chapter 1 

 

1.1 The Blade - An Essential Companion Throughout Human History  

Research at sites around the world (Apel and Knutsson 2006, Otsuka 2017, Takakura 

2020,), to include those in North America (Carr et al. 2010, Collins 1999, Eren and Redmond 

2011, Roux et al. 2024, Sain 2012; 2016) has demonstrated the importance of blades and blade 

technology in the toolkit of Paleoindians. These highly mobile people relied on complex seasonal 

migratory patterns supported by a refined lithic toolkit which supported a range of behaviors. 

The identification of blades at Thunderbird would support previous models that the site was an 

important refugia on a migratory pattern where scheduled resource exploitation and toolkit 

refurbishments took place as part of seasonal rounds. How, when, and where people decided to 

manufacture certain types of tools is directly related to the migratory patterns of those people, 

which is also highly influenced by environmental stressors (Sain 2012:15-16). How does the 

analysis of lithic blade production at the Thunderbird site (44WR11) refine our understanding of 

localized seasonal migration and exploitation of local resources among Paleoindian people of the 

Shenandoah River Valley.  

Of the few Paleoindian sites in North America, even fewer have intact stratified levels, 

making Thunderbird a genuine snapshot in time (Gardner, 1974). Numerous excavations at 

Thunderbird identified a wide variety of lithic tools, as well as post holes outlining structures 

which place artifacts in context of a stratified Late Pleistocene habitation (Gardner, 1974; 1983; 

1994). The low number of finished tools found compared to the amount of waste from 

production would indicate it was where toolkits were refitted before these small hunter gatherer 

bands moved on to other locations (Gardner 1974). The presence of any blades at Thunderbird 



2 

 

would suggest a complex toolkit like other locations in the East. Since blades were frequently 

produced and transported from one location to another for future use (Boldurian & Hoffman 

2009), the presence of blade cores without blades is noteworthy as it would suggest cache 

behavior for anticipated migration. Thunderbird has been the subject of numerous studies and 

research papers, several of which focused on the common biface technology of the Clovis period 

(Boyer 1974, Callahan 1979). Blades on the other hand are unique in their function, only being 

found at certain sites, and in varying forms. Blades are recognized as a prominent tool in Upper 

Paleolithic tool kits around the world. Yet their presence in research has been underrepresented. 

This is due in part to blades being lumped in with debitage (Sain 2012). This has led to an 

increasing reexamination of North American sites for blades especially in the East, such as 

Shoop (Cox 1986), Shawnee Minisink (Iceland 2013), Adams (Hagg and Jones 2022), Topper 

(Sain 2012), Carson-Conn-Short (Jones 2018), and Cumberland (Gramly 2013). 

It is widely accepted that people of the Clovis Era, to include those in Virginia were 

highly skilled in their ability to create distinctive stone tools from high quality toolstone. The 

most notable is the Clovis fluted point which is a product of bifacial reduction and takes on the 

appearance of a rigid and durable tool (Figure 1). And while these complex lithics are a wonder 

of craftsmanship, this was not the only tool in the toolkit of early people. Likely their daily lives 

involved a variety of organic and stone tools necessary to survive the harsh environment at the 

end Pleistocene. Functional utility was key and tools were often multipurpose by design to 

accomplish a variety of tasks. Research has demonstrated that a key tool within the Clovis toolkit 

was the blade for its masterful design and utility (Collins 1999, Boldurian & Hoffman 2009), 

which at first glance does not look durable enough to be significant when comparted side by side 

with bifacially knapped tools (Figure 2). Despite their delicate appearance the physical structure 



3 

 

of blades makes them incredibly durable and versatile tools which made them the “tools of 

choice” for delicate tasks like slicing and shaving of soft materials, and heavy work like shaping 

bone and wood (Boldurian & Hoffman 2009:169). Clovis blade technology is unique in that the 

desired output is a specific tool type with distinct characteristics not created by chance. Blades 

by their design serve a vastly different purpose than that of large, carefully curated points used 

for repeated heavy use or for taking large game. What could be called the Paleolithic version of 

paring knives would have complemented the larger cleavers of their day. Knowing the types of 

tools used by Paleoindians in a specific region and time opens a window, however small, to their 

daily lives. 

 

Figure 1. (Left) Fluted Clovis point from Thunderbird. (Right) Double crested blade from 

Thunderbird. The Clovis point was created with hundreds of carefully planned flake removals. 

The crested blade on the other hand was created after other blades were removed laterally from 

the left and right, followed by a single strike to remove the blade as seen above. Both are created 

with clear intention with very different outputs. Photos by the author. 

 



4 

 

1.2 Origins and Ice 

The Shenandoah River Valley in Northern Virginia was a vastly different 13,000 years 

ago. The landscape was possibly covered by boreal woodland with open grasslands and bog-like 

marshes (Carbone 1974). This was home to mammoth, bison, giant sloth, and saber tooth tigers. 

Braided rivers, forests and marshland gave refuge to a multitude of species long extinct (Gardner 

1974). The small bands of people inhabiting this landscape used the world around them to create 

the tools necessary to survive and left behind only shadowy traces which produce more questions 

than answers. With organic remnants all but gone, scientists rely on stone tools and their 

byproducts of manufacture to tease out some small understanding of this early chapter of human 

history in North America. Paleoindian research is contentious, and findings claiming to have 

unraveled a piece of the enigmatic story must be absolute and irrefutable because of the weight 

they carry. No origin story is something to be taken lightly. However, there are things we do 

know about these people, largely tied to their tools and their distribution across the landscape.  

1.3 A Story Written in Stone 

Humans have been reducing stones to an increasing level of complexity for hundreds of 

thousands of years, and the shaping of stone by our distant cousins’ dates back nearly 2,000,000 

years. If read properly and with skill, stone can peel back the millennia and give us subtle 

insights into daily activities. The study of lithic technology shows us the progress of these 

people, their abilities, and movements across landscapes (Crabtree 1974). The process of 

flintknapping demonstrates the foresight of early people to induce and control the fracture of 

stone to create tools which progressed in complexity and technical skill through the ages, based 

on the “systematic knowledge of forming stone into useful cutting, chopping and other functional 

impairments… Flintknapping was not a haphazard art, but rather, a carefully planned process of 



5 

 

making stone tools to suit a specific functional purpose,” (Crabtree 1972:2). The core of this 

research is based on pure intentionalism; the concept that an individual’s intentions and 

conceptualization of something have a direct role in its creation. Therefore, it could be said that 

lithic analysis is the validating and deciphering of the intentions of prehistoric knappers.  

As a scientific process lithic analysis has its roots in physics and effects of applied force on 

various materials. Different materials have different properties which produce noticeably 

different attributes. These attributes are predictable and repeatable, providing great insight to a 

trained analyst. One way analysts do this is by dividing the process into specific stages, such as 

preliminary modification, blank, preform and final implement (Bradley 1975). Being able to read 

the attributes of one type of material is akin to speaking one language, but recognizing 

similarities to others is like translating across many. It could take a present-day lithic analyst a 

lifetime to become well versed in the physical attributes and variations of many different 

material types. But early people were also lithic analysis, well versed in the workings of stone. 

Evidence from Virginia demonstrates that Clovis people were multi lithists (Johnson 1989). At 

the core of this study are the lithic attributes of jasper and the technological process of creating 

blades and bifacial tools.  

Blades have been noted for their optimization of cutting edge, and as a foundation for 

creating curated tools. Many definitions exist for blades (Collins 1999, Sorensen 2006, Sain 

2012), however a blade in its simplest form is a flake, intentionally removed for its lateral cutting 

edge, and which is often longer than it is wide upon removal. However, the definition of a blade 

remains fluid, and under increased scrutiny and debate. Unfortunately, blades have taken a 

backseat to more diagnostic and refined lithics like the Clovis point, which was used as a 

temporal marker when early archaeologists saw more value in establishing chronology through 



6 

 

artifacts than what those artifacts could reveal about ancient human behavior. Further 

exasperating the issue is the fact that Paleoindian tools are rare, often found far from their point 

of manufacture and discarded when worn, misshapen though curation, and no longer of value to 

the user. Blades are no exception. Blades, because of their less-than-formal nature and ill-defined 

definition can be easily misconstrued and overlooked. This research will examine artifacts 

previously considered as waste due to overly stringent typologies, and different archaeological 

priorities at the time of initial collection. In short, previous researchers may have been biased 

toward “finished” tools with formal complexity, thereby completely overlooking the value of 

more simplified and expedient tools which likely played a key role in the everyday life of 

Paleoindian people in Virginia and elsewhere.  

1.4 Overview 

This study highlights the importance of Paleoindian research to the history of North 

America and calls out the blade for its specific role in that journey. To understand the context of 

blades it is first important to know more about the Paleolithic world of the Middle Atlantic and 

Susquehanna River Drainage ancestral to the Chesapeake Bay as it existed 13,000 years ago. 

Chapter two presents an overview of the unique geography of Virginia and how it rapidly 

changed at the end of the Pleistocene but fortunately preserved evidence of Paleoindian 

settlements. The heated quest to unravel the Paleo-American story is addressed, with major 

historic milestones explained, leading up to the discovery of Thunderbird and the Flint Run 

Complex in Northern Virginia.      

Chapter three discusses the theories that have been used to investigate Paleolithic sites in 

the Eastern United States, starting with some of the first major investigations to note blades and 



7 

 

making comparisons to artifacts of the Old-World Upper Paleolithic. These discoveries and 

debates have led to refined theories focused on the role of blades, and more importantly active 

discussion on what blades are and how narrow typologies and biases can limit research and 

understanding.  

Chapter four outlines the techniques and methods used to carry out this study after 

addressing the development and progression of relevant theory, a less restrictive definition of 

blades, and common methodologies for analysis. Over the course of several months the 

necessary data was collected from the solitude of the Smithsonian collection repository. Many of 

the methods for analysis are drawn from best practices of other researchers, while some were 

created through trial and error to create an original analysis of lithics recovered from Clovis-

period contexts at the Thunderbird Site. Ultimately a sizable dataset of more than 300 artifacts 

were analyzed in the context of trying to better understand blades and blade like flakes. 

The tabulation of data is presented largely in raw format in Chapter five and highlights 

the value of the data in answering some of the key questions of this study. In total more than 

18,000 measurements as datapoints were available for comparison. This required the narrowing 

and careful organization of datasets into those that are meaningful to blade research.  

The analysis presented in Chapter six finds unique attributes in both the quantitative and 

qualitative analysis of the collection, not only raising questions about the role of blades at 

Thunderbird but highlighting the value of answers yet to be found in legacy collections. By 

selecting several key artifacts from the analysis, multiple lines of evidence are presented on the 

role of blades at Thunderbird within the Clovis context. 



8 

 

Like many research projects, the facts uncovered created more questions than answers 

and illuminated future research opportunities. In Chapter seven, a holistic look back on the 

research for its broad implications is paired with recommendations for additional research. The 

opportunities in blade research are numerous, with multiple collections, like Thunderbird, never 

cataloged and just waiting for analysis. That is provided someone has an immense amount of 

patience, a good soundtrack, and an affinity for rocks. 

 

  



9 

 

Chapter 2 

 

2.1 Virginia - A Paleolithic Hotspot in a Cold World 

Virginia has a unique set of physiographic provinces due to its long and complex 

geologic history that complements Paleoindian research. The state contains five distinct 

physiographic regions. On the easternmost side of the state, bordering the Atlantic Ocean, is the 

mostly flat Coastal Plain. Further west the low and unobtrusive elevation of the Coastal Plain 

gives way to the rolling hills of the Piedmont which steadily increase westward until they meet 

the Blue Ridge mountains. To the west of the Blue Ridge Mountains is the Valley and Ridge 

region which is notable for its repeating ridges and valleys increasing from east to west. Finally, 

to the far southwest of the state is the Appalachian Plateau comprised of rugged, heavily 

dissected hills. These regions and their drainages are of notable importance when we consider 

models developed by researchers, who connected Paleolithic movement, resources, and 

settlement patterns to the landforms which acted as corridors and barriers to movement. But the 

Virginia of the late Paleolithic, more than 10,000 years ago during the Late Pleistocene looked 

very different than it does today. 

2.2 Paleoindian Timeline and Climate 

It is believed that the greatest expansion of humans in North America took place during 

the Middle Paleoindian period between 13,450-12,900 cal. BP (Anderson, 2001:154). This was a 

time of climate warming at the end of the Pleistocene. Clovis technology is a key marker of this 

spread. Around 11,500 BP during the Late Glacial period the Shenandoah Valley had many 

microhabitats which varied by geography. Uplands were likely covered in alpine tundra, 

coniferous forests on slopes, mixed forests on the foothills and valleys and deciduous forests in 



10 

 

the valleys (Carbone 1976 as cited in Carr et al 2013:163). By 11,000 BP groups employing 

fluted points were in the Shenandoah region (Carr et al. 2013). But this time of significant 

warming was followed by a sharp reversal around 10,900 BP which occurred abruptly within 40 

to 100 years and returned to a cold and dry glacial like climate known as the Inter-Allerod period 

Younger-Dryas (Isarin and Bohncke 1999 as cited in Carr et al. 2013:162). This event was 

catastrophic and likely caused the extinction of megafauna in Eastern North America such as 

mammoth, camel, and giant sloth (Anderson, 2001:152). This drastic impact on the environment 

may have also killed off early humans who had adapted to these resources for survival. As 

explained by Edwards and Merrill (1977) as cited in Egghart (2020), during this period of 

glaciation and cooling so much water was locked in ice that sea levels were as much as 130 

meters lower than present day. Much of the Northern Hemisphere was pressed beneath massive 

ice sheets 3,500 to 4,000 meters thick. It’s fortunate for archaeologists that Virginia escaped the 

scouring effects of these massive glaciers, however the region would not escape significant 

change driven by sea level rise when the sudden cooling reversed once again at the end of the 

Younger-Dryas (Egghart, 2020).  

So drastic was this change that sea levels rose by several hundred feet. This resulted in 

movement of the Atlantic coastline westward as water inundated the gently sloping portions of 

the continental shelf. Estimates place this change at 510 horizontal meters inundated for every 

0.30 m rise in sea level (Milliman and Emory 1968 as cited in Egghart 2020). What this means 

for Virginia’s early Paleoindian coastal sites is that many are now submerged and have been for 

millennia. Prior to this inundation Paleoindians could have ranged an additional 70 miles east of 

the current coastline, as shown in Figure 2. The Cinmar site is reported to have been found 

approximately 40 nautical miles off the Virginia coast where a biface and mastodon remains 



11 

 

were recovered by a scallop trawler (Stanford and Bradley 2012), however this find remains 

under scrutiny (Eren, Boulanger, and O’Brien 2015). This does not negate the fact that the 

Virginia coastline was much further out than it is today. But Paleoindians were not confined to 

these not submerged coastal regions. Sea level rise had a mixed impact, most severe in the larger 

floodplains that had been scoured deeply during spring melts in the mountains (Johnson 2012). 

Sites such as Cactus Hill, Blueberry hill and Rubis-Pearsall, the first two containing pre-Clovis 

components, are all in the present day inner Coastal Plain and avoided the erosive effects of full 

glaciation as they sit on smaller rivers (Johnson 2012).  

 

 

Figure 2. The extent of the Virginia coastline at the end of the Late Pleistocene would have 

extended approximately 70 miles from today’s coastline where the current depth is 

approximately 100 meters, as shown in light blue. Map produced by the author using the 

National Oceanographic and Atmospheric Administration, Bathymetric Data Viewer, 2024.    

 



12 

 

Following the Younger Dryas, around 10,000 BP, temperatures rose markedly into the 

Archaic Period, ushering in the spread of warm-adapted flora and fauna (Egghart 2020:24). But 

this may not have been a constant process across all regions. By 9000 BP the Wisconsin Glacier 

remained as far south as Lake Superior and Ontario. The cooling effect from the corresponding 

periglacial lakes in the north would have had a profound impact on winter climate in the East. 

The residual effect of this being the persistence of cold adapted microenvironments in the 

Plateau, Ridge and Valley and Blue Ridge regions where alpine grasses and spruce parkland 

remained well into the later warming period (Mike Johnson, personal communication July 2024). 

Virginia’s reprieve for the crushing ice and partial inundation of its eastern coastline 

means that very old prehistoric deposits, although rare, do exist in Virginia. These Paleoindian 

sites are limited to a few regions, notably the Coastal Plain, the Northern Shenandoah Valley, and 

the Valley and Ridge physiographic region in Southwest Virginia (Boyd 2020:39). As noted by 

Boyd (2020:39) as of 2020, the State of Virginia database listed only 88 Paleoindian sites in 

Virginia, but there are 240 in total with Paleoindian components; the majority lacking 

provenience. However, those that have been found are of great importance.  

2.3 Searching for the American Paleolithic 

Extensive research has been conducted on Paleoindian cultures of the Mid-Atlantic, with 

notable focus being placed on the sourcing of lithic materials and the production of bifacial tools, 

such as the well-known lanceolate fluted Clovis point. The quest for evidence of an American 

Paleolithic in the East dates back to the 1850s, as excavations in Europe spurred a hunt on the 

part of American scientists for artifacts of similar antiquity in America (Meltzer, 2015). 

However, artifacts did not present themselves as readily as they did in European caves and rock 



13 

 

shelters, forcing scientists to leverage findings of limited credibility. The controversy 

surrounding these questionable artifacts raged for decades before the quest for a Paleolithic 

presence in America was written off in the early 20th century (Meltzer, 2015:4). However, in 

1927 the find of a fluted point in situ with extinct bison bones in Folsom, New Mexico changed 

that belief and rekindled the study of Paleoindian sites in the United States. Following the find at 

Folsom, discoveries of fluted points in the Eastern Woodlands began to shed light on the 

possibility of a separate and distinct group of Late Pleistocene people in North America. 

Although few and far between, the presence of fluted points in the East spurred analysis of the 

lithics found at these Eastern sites. By the 1950s several sites had been located which allowed for 

collective comparative analysis. One early analysis was performed by Witthoft (1952) who 

examined the collections from Paleoindian sites in North Carolina, Virginia, and Pennsylvania. 

Through that analysis we begin to see terms of interest to this thesis such as triangular and 

trapezoidal cross section, “flake knife fragments”, and polyhedral core begin to emerge in 

writings, all of which are indicative of blade technology being present in the East. Witthoft’s 

(1952) study focused on a small site, surface collected in the 1930s on a farm in Eastern 

Pennsylvania. The Shoop Site as it became known produced 48 fluted points, and a variety of 

formal tools on flakes. Through his analysis we learn that the flakes and corresponding flake 

scars begin to paint a unique picture of the methods used at the site to form and reshape tools. 

Interestingly, he noted a high percentage of long, flat flakes terminating in hinge fractures. He 

noted that flakes at the site and resharpening scars were more “conspicuous in blade-making and 

tool shaping” (Witthoft, 1952:474).   

  The fluted point blanks found at Shoop show a triple channel flake pattern, whereby two 

longitudinal flakes are taken prior to removal of the central channel flake (Witthoft, 1952, 481). 



14 

 

This is of particular relevance when trying to understand blade technology in the East, as 

Witthoft explains, “In so preparing the blank for the removal of the central channel flake, the 

flint knapper was applying core and blade technique to a bifaced tool (Witthoft, 1952, 483).”  

Witthoft’s analysis of the flakes and tools at Shoop concluded that the chert industry at the site 

“is a remarkably clear-cut blade industry, of a pattern which is totally new for the Eastern 

Woodlands and perhaps for the whole United States and Canada” (Witthoft, 1952: 475). He 

elaborated, “Typical scrapers represent a blade tradition otherwise unknown in the Northeast but 

very familiar in many Old-World cultures” (Witthoft, 1952:478). Depleted cores found at the site 

further strengthen this argument, as do pointed prismatic spalls which Witthoft (1952:476) 

believed were the by-product of “flake knife manufacture.” Witthoft pointed out that cores found 

at Shoop were so expended that they could no longer produce bladelets. This work led Witthoft 

to look at the Williamson site in Virginia and the Hardaway site in North Carolina and conclude 

they were all part of the same complex (Witthoft, 1952:466). “The people who came at Shoop 

were the carriers of a blade industry which is best illustrated by scraper forms; in America, their 

blade technique was extended to the basal thinning of projectile points (Witthoft, 1952:493). 

What is most interesting about this claim by Witthoft is that the flutes on Clovis points are 

facilitated by an understating of blade technology. “The way of life and the technology of the 

Shoop site were probably not too different from that of the upper paleolithic of the Old World” 

(Witthoft, 1952, 494). But like many conclusions in archaeology, this too was quickly challenged 

by later researchers who did not feel Witthoft’s assessment fit the typology of a true blade and 

core technology (Krieger 1954; Wilmsen 1970; Cox 1986). Witthoft’s theories have bene neither 

proven nor disproven despite the controversy, and a definitive answer would be best served by 

further comparative studies (Cox,1986). Herein lies the issue with blades. The definition of a 



15 

 

blade can vary widely as can the interpretation. This shows the need for comparative study and a 

further pool of research regarding blade technology in the East.  

At the time of his analysis in 1952, Witthoft noted that only one Eastern site with a fluted 

component had been excavated, and even then, it was not stratigraphically isolated from a Late 

Archaic component. Now, as then, stratigraphy is a key component of temporal understanding. In 

these early days of looking for Paleoindian sites in the East, Witthoft alluded that other sites with 

a fluted component may be found in high mountainous country. Today we know this to be true. 

As summarized by Boyd (2020:34-35), investigations into Paleoindians in Virginia began in the 

1930s primarily through fluted point surveys and the earliest Paleoindian artifacts in Virginia 

come from the Cactus Hill site (44SX202) and have been dated to approximately 15,000 years 

ago (Boyd 2020). In fact, nowhere in the Americas are the number of Paleoindian sites higher 

than in Eastern North America (Figure 3), yet few published reports exist, and most 

documentation is reserved to special publications or hard to find dissertations and master’s theses 

(Gingerich, 2013). One such site of critical importance to this research, which, like Shoop did 

not achieve more than local notoriety, is the Thunderbird site (44WR11) which is part of the Flint 

Run Complex. The site lies an hour and a half west of Washington D.C. along the South fork of 

the Shenandoah River (Figure 4). Fully aware of the importance of Thunderbird, and 



16 

 

understanding the publication vacuum surrounding its data, Gingerich noted that the “need for 

more work within the Thunderbird assemblage has long been apparent, as well as the need to  

further refine the models generated from the work at Flint Run” (Gingerich, 2013:3). To 

Gingerich’s point, to date there have been three master’s theses (McNamara 1980; Melton 1978; 

Rubenstein 1980) and one PhD (Verrey 1986) with a focus on Thunderbird (Carr et al. 

2013:174). Moreso, Carr et al. (2013) notes that in no documents is there a collective analysis, 

nor an all-encompassing list of cataloged artifacts. This is not to say the work at Thunderbird was 

not done well, but rather that it established a sound foundation of knowledge for which to build 

more refined understating of Paleoindian lifeways.  

Figure 3. Selected Paleoindian sites in the East mentioned in early and current research. 

Thunderbird is centrally located between several of these prominent Eastern sites.  Map created by 

author using ArcGIS basemaps. 

 



17 

 

 

Figure 4. The Thunderbird site lies approximately 80 miles west of Washington, D.C. It is part of 

the Flint Run Paleoindian Complex located in the Shenandoah River Valley of the Blue Ridge 

Mountains. Map created by the author using ArcGIS basemaps.   
 

2.4 A Paleoindian Complex in Virginia 

It is not possible to talk about Paleoindians in Virginia without mentioning Dr. William 

Gardner of The Catholic University of America. As noted by Carr et al. 2013, Gardner and his 

multidisciplinary team began excavations at Thunderbird in 1971 and continued to conduct 

research for more than 15 years. Initial investigations at Thunderbird were focused on surface 

collections, but the realization that there were intact stratified despots drove further investigation. 

Prior to this work there were no known stratified Paleoindian sites discovered in the East (Carr et 

al 2012) Gardner’s work at the Flint Run Complex in Northern Virginia would become a primary 

source of understanding Paleoindian occupation in the region (Boyd 2020:35). Gardner used the 

data from Flint Run to rethink local, regional, and pan-regional Paleoindian lifeways (Carr et al 

2012). This was a dynamic time in archaeological theory. Gardner placed an emphasis on 



18 

 

cultural ecology and environmental reconstruction resulting in a more detailed understanding of 

the world in which Paleoindians existed (Carr et al 2012:157). Previous archaeological theories, 

which I discuss later in greater detail, not only approached archaeology differently, but are likely 

to be blamed for the absence of documented blades through their stringent adherence to typology 

and normative views.  

There are more than 20 sites within the Flint Run Paleoindian complex, Thunderbird 

being just one. What makes Thunderbird notable is that it “contains practically all known 

components of the Flint Run Complex in a sealed stratified context” (Carr et al. 2013:171). As 

noted by Gardner (1974:2), so diverse were the sites that they represented a full span of pre-

contact history covering the Ridge and Valley province. In addition to Paleoindian artifacts 

within the deepest levels, the site encompasses cultural and natural features such as stockaded 

villages, hunting and fishing camps, as well as buried flood plains, swamps, and extinct streams 

from many cultural periods. The amount of foundational research and model building conducted 

at Thunderbird and the surrounding sites is of great benefit to anyone studying nuanced aspects 

of early occupations in the Mid-Atlantic region. The area of continuous Paleoindian occupation 

was estimated at 3,600 by 200 feet. By 1973 excavations at Thunderbird had opened 20 ten-foot 

squares, a lesser number of 5-foot squares, to a maximum depth of 3.5 feet to the identified 

living floor. Not only was the site meticulously investigated archaeologically (Gardner, 1974), 

but the entire region was studied from geologic (Segovia, 1974; Foss, 1974), ecologic, and 

climatologic (Carbone, 1974) perspectives. As summarized by Carr et al. (2013:156) the Flint 

Run Complex was defined based on years of work by Gardner at several sites in the northern 

Shenandoah Valley of Virginia. The complex is primarily comprised of six sites near one another, 

and in close proximity to sources of cryptocrystalline toolstone (Carr et al. 2013:157). 



19 

 

Thunderbird is only one of two sites in the complex with stratified Paleoindian components (Carr 

et al., 2013:202). Gardner’s work opened the door for the hypothesis that the Paleoindian 

lifeways were in fact not based on highly mobile big-game hunting as previous models had 

suggested, but rather extended occupation of sites (Carr et al. 2013:156). Gardner notes that 

“lithic parsimony and tool kit curation are necessary requisites for any system which insists on 

maintaining a raw material preference unless that system is virtually sedentary and/or surrounded 

by large quantities of the preferred material” (Gardner, 1983:51). This means that we should 

expect to find a wide variety of tools at sites, including those constructed using blade and blade 

like reduction methods. A wide variety of lithic and non-lithic tools would have been required to 

accomplish the daily tasks of surviving during the late-Pleistocene. This typical Paleoindian 

toolkit would have included “scrapers, with or without gravers, wedges, unifacial flakes with 

varying kinds and degrees of edge modification, expediently and briefly used flakes, and other 

types of tools which defy description other than crude such as banging, smashing, chopping and 

hacking tools” Gardner (1994:19) Gardner’s (1974) lithic assessment of the site names several 

tool forms to include fluted, corner notched, and side notched points, prepared unifacial tools, 

wedge or bifacial tools, utilized flakes, and associated tools to produce those previously noted. 

He concluded that “all tools were made from either flakes or cores. Deliberately manufactured 

blades were absent” (Gardner, 1974:5). “Completely lacking in the Shenandoah Valley sites are 

true blade cores and burins. Flakes conforming to blade length-width ratios do occur but are so 

rare as to be considered accidental” Gardner (1994:19). A breakdown of artifacts by Carr et al. 

(2013:193) from the Fifty site (44WR50), also within the Flint Run Complex, shows a 

generalized tool kit, more than half of which was comprised primarily of utilized and retouched 

flakes, with just over 1% being classified as projectile points.   



20 

 

  Thunderbird was noted for its high quantity of debitage, or waste material, and a low 

percentage of expended tools. Most tools recovered appear to have broken during various stages 

of manufacture and were therefore discarded. The technology of Paleoindians is especially 

characterized by a preference for high quality cryptocrystalline tool stone in the manufacture of 

tools suited for bifacial reduction strategies, repeated rejuvenation and repurposing adapted to 

settlement mobility (Goodyear 1979; Carr et al. 2012.) These “sources of cryptocrystalline lithics 

were viewed as an essential organizing principle in settlement patterns within territories that 

were relatively small to moderate in size. There appears to be evidence that Thunderbird had 

extended occupation, not the highly mobile characteristics typically placed on Paleoindian sites” 

(Carr et al 2012). During the Paleoindian occupation at Thunderbird, high quality jasper was the 

stone of choice, making imported stone remnants scarce on the site (Gardner, 1974). Jasper is a 

high silica content isocratic material extremely suitable for tool manufacture. So prevalent was 

jasper that its disappearance in the stratigraphic record serves as a terminus for the abandonment 

of the Flint Run Complex by Paleoindian people (Gardner, 1974). While jasper was a preferred 

material by Paleoindians, it was used through later time periods due to its high workability as a 

high grade toolstone. But the abandonment of Thunderbird likely coincides with a change in 

climate between 9-8000 BP when a warm and dry period resulted in periodic flooding and 

seasonal food variation (Gardner 1974). This also causes a reduction in the water table which 

causes a decrease in surface water making the location less attractive for habitation (Carr et al 

2012:165). Climate change in the late Pleistocene and Early Holocene effectively left once 

resource rich areas “high and dry” (Gardner, 1994:27). 

  The revisiting of previously excavated sites like Thunderbird, and summary examinations 

of Paleoindian research, has provided a platform for which to launch focused and pertinent 



21 

 

Paleoindian research in the East. Some notable synthesis of this subject includes Gardner’s 

(1994) overview of Paleoindian research in Virginia. In this research Gardner details the many 

issues with typology and periods and uses examples of morphologic change and similarities to 

make the point that periods overlap, regional site contents vary, and above all else our 

understanding does and should change as new information is presented. Gingerich’s (2013) 

summary of Paleoindian research over the past several decades shows the broad and expanding 

examination of Paleolithic sites in the East. Through this synthesis he develops a holistic picture 

of studies in the East to expose opportunities for continued research into Paleoindians in Eastern 

North America. “Crucial to our understanding of the spread of populations and Paleoindian 

material culture is the recognition and understanding of sites that may represent the earliest 

occupations in eastern North America” Gingerich 2013:5). 

It should also be noted that the type of tools found at sites is essential to deciphering 

them. Analysis of tool kits, like those at the Fifty site, and Thunderbird do not differentiate or 

note blades, despite a variety of blades being highlighted by Bradley et al. (2010:10) as integral 

to Clovis life. Bradley et al. (2010:12) note that blades represent a far more efficient technology 

when compared to biface manufacture. Despite this, archaeologists have devoted little attention 

toward Clovis blade technology and have frequently failed to capture the existence of blades in 

site records and analysis (Bradley et al. 2010:14); Collins (1999; Sain 2012). To counter this 

trend, this research explicitly examines the Thunderbird site for the presence of blade-like 

technology to determine if it played a role at that location during the Late Pleistocene.   

  



22 

 

Chapter 3 

 

3.1 Driving Theories and Models in Paleoindian Research 

For more than 50 years researchers have been examining the sites that make up the Flint 

Run Complex. The predominant guiding theories underlying these investigations have used the 

processual theories of Lewis Binford as foundations for teasing out cultural, social, and 

economic behaviors. This approach makes sense as Lewis Binford sought to step beyond the 

simple classification of artifacts and culture by the cultural-historical archaeological theorists, to 

fulfill a need for archaeologists to contribute to the understanding of cultures from an 

anthropological perspective. Binford’s systemic approach in the evaluation of artifacts was 

concerned with all the subsystems within the broader cultural system, as there is a systematic 

relationship between people and the environment, with culture being the item that brings them 

together. If this theory remains true, individuals, regardless of location will have similar 

adaptations to their environment, as well as hindrances (Binford, 1962:218). Based on this idea 

we should see similarities in technologies, such as blade manufacture at Thunderbird, among 

groups facing similar circumstances shaped by their immediate environment (Binford, 

1962:218).  

Recent research has effectively used processual models to test hypothesis related to 

Paleoindians, namely that of Gardner (Boyd 2020:35). The work of Gardner (1974, 1983, 1994) 

is a prime example of applying the processual theoretical framework of Binford in formulating a 

model for the Paleoindian settlement patterns in the Mid-Atlantic region.  As noted by Carr et al. 

(2013:156), Gardner’s work established a new settlement pattern model in the region. In this 

Gardner (1974, 1983) identifies six site types. These are quarry, reduction stations, base camp 



23 

 

maintenance stations, hunting sites, and isolated fluted point finds. Under the processual theory 

these sites are part of a system. The types of lithic artifacts found vary depending on the site type 

and provided the evidence necessary to deduce what type of site the artifacts are related to. 

Gardner also created the model of “lithic determinism” which he used to explain Paleoindian 

settlement patterns (Boyd, 2020:35). This model states that given their preference for high 

quality toolstone, groups would center settlements around quarries capable of providing this 

material, while smaller camps for other purposes, hunting, foraging, etc., would be established at 

distance from the main camp. In this context the lithic source determines the central location of 

the basecamp, versus a group that ranges to collect resources. The concept of lithic determinism 

is key to the evaluation of the role of blades in this thesis.  

As researchers continue to refine conclusions on the influence of environmental factors, 

we can better explain the behavior of Paleoindians, among others. A unique aspect of behavior is 

the output of artifacts. To further refine the applicability of individual artifacts to the larger 

cultural system, Binford pointed out that in the sea of thousands of artifacts, we must have the 

wherewithal to be able to identify those of relevance to the social, technological, and ideological 

sub-systems of the total cultural system. At Thunderbird, where tens of thousands of artifacts 

were recovered, this rings true. Similarly, we cannot look at differences between these items for 

an explanation of material culture, as items alone cannot expose intangible social constructs.  

Gardner (1994:33) believed “that sites like Thunderbird or Williamson served a social as well as 

a technological function.” Binford adds, these artifact assemblages must be paired with their 

contextual elements to create a picture of the extinct cultural system in its totality (Binford 

1962:219). He suggests culture be “viewed as a system composed of subsystem, and it is 

suggested that differences and similarities between different classes of archaeological remains 



24 

 

reflect different subsystems and hence may be expected to vary independently of each other in 

the normal operation of the system” (Binford,1965:203).  Binford noted that those archaeologists 

using normative methods are essentially acting as cultural-historians or paleo-psychologists, or 

merely typologists, neither of which they are trained to perform (Binford, 1965:204). But while 

Binford criticized normative processes, his quest for systemic understanding has received equal 

criticism, for its dry application of analysis through the lens of arbitrary conditions (Schmidt and 

Kehoe 2019: 21-22). As a refinement to Binford’s systemic processes Schmidt and Kehoe (2019) 

stress that there is a human element focused on knowledge transmission that has long been 

ignored by anthropologists brought up under scientific methods laced with colonialism.  

Schmidt and Kehoe (2019: 21-22) discuss the benefits of ethnographic interaction saying, 

“archaeologists and archaeology as a discipline benefit from an interchange with local and 

descendant communities through which their deep experience and historical knowledge broaden 

our base for inference to the best explanations” (Schmidt and Kehoe, 2019:2). Despite the 

criticisms, Binford’s Processual approach tops the creation of multiple avenues in Paleoindian 

research, with a focus on cultural systems, rather than technologies. Having the theoretical 

foundation to look at sites as more than collections of artifacts in a cultural timeline has changed 

how these Paleoindian sites are treated and examined. But while processual theory satisfies some 

requirements, the cognitive theory that developed in the post-processual movement also has 

notable applications to the analysis of past cultures and gets at the issue Schmidt and Kehoe 

(2019) point out regarding Binford’s reluctance to get into the mind of those he was studying. 

Abramiuk (2012) notes that all humans interpret the world around them, and archaeologists are 

doing the same type of interpretive process when examining the archaeological record, rather 

than using an agreed upon set of standards or methodology. Abramiuk adds that cognitive-



25 

 

processual archaeologists argue that even if numerous interpretations do carry seeds of truth, 

such as that found in normative and processual theories, method rooted in comparison and 

generalization is still needed to help determine the more truthful interpretations (Abramiuk 

2012:14). “Cognitive archaeology should accommodate not only archaeological data, but any 

form of accessible data in pursuit of understanding the mind in the past” (Abramiuk 2012). 

Cognitive archaeology does not consist of a single approach, rather it is a discipline using 

multiple approaches.  (Currie and Killin 2019, 263-279) note that human cognition and things are 

intimately linked and that much can be gleaned about mental and cultural lives from material 

objects. Relative to this thesis’s focus on lithic technologies, Currie and Kiln (2019:268) note 

that knapping, the process of manipulating stone for tool production, takes significant cognitive 

ability. Aspects include forward planning, capacity, good working memory, focus and impulse 

control, impressive manual dexterity, and task specialization. Sellet (1993, 106-112) elaborates 

on this idea of tasks by stressing that blade manufacture is a process, not an accident. And while 

there may be accidental blade-like flakes, the intentional creation of true blades as a specific 

technology requires a process.  

This process, or chaine opératoire, has long been highlighted by researchers as a means 

of understanding the blade production process (Collins 1999, Sorensen 2006). This highly 

cognitive approach to lithic analysis and study looks at knapping as “a succession of mental 

operations and technical gestures, in order to satisfy a need (immediate or not), according to a 

preexisting project” (Sellet 1993:106). With very similar outcomes to processual methods, 

chaine opératoire originated in ethnology, the other in processual archaeology (Sellet 1993:107). 

Chaine opératoire addresses five subsystems. Raw material procurement, reductions sequences, 

use, maintenance and discard (Sellet 1993:108-110), all of which are highly relevant when 



26 

 

conducting lithic analysis or experimental processes. This is because not only are diagnostics 

important, but perhaps more so are the distinct byproducts of their creation. “The chaine 

opératoire also integrates a conceptual level and, thus, cannot be understood without reference to 

the technical knowledge of a group” (Sellet 1993: 106). While both the systemic and cognitive 

approaches provide insight into behavior, the advent of computer aided spatial analysis has 

added a new level of refinement to models supporting Paleoindian studies.  

The space between spaces and the interplay of artifact positioning has been a key part of 

archaeological analysis for decades. Today, models of spatial analysis are taking previous 

systemic models, like those of Binford, and applying computer aided statistical analysis to create 

nuanced pictures of human behavior. Studies which once focused on site structure are now able 

to delve deeper into the behaviors likely to have created features at those sites. As noted by Clark 

(2022:8), behavior cannot always be explained by the “configuration of space”. Of relevance to 

this study, Clark (2022) points out how the location of formal tools is no longer the focus, but 

rather the distributional patterns of artifact types, which when examined together and for their 

distinct attributes can have more meaningful outcomes for understanding how a site was used. At 

the Topper site (38AL23) in South Carolina, we see the implementation of spatial analysis as a 

modified model for behavioral study. Like Thunderbird, Topper is a stratified Clovis occupation 

with natural variations from terrace, hillside escarpments, and uplands. Smallwood et al (2023) 

apply spatial analysis to compare the distribution and variation of artifact types across the zones 

of the site to explore intrasite variation. Similarly, Gingerich (2022) reexamined the Shawnee-

Minisink site (36MR43) in Pennsylvania, which represents another undisturbed and stratified 

Clovis occupation.  The spatial examination of artifact mass, raw material types, artifact density 

and refitting distances has the potential to increase our understating mobile forages as well as 



27 

 

sedentary groups at similar sites worldwide (Gingerich 2022).  Studies like these demonstrate 

how models have evolved to account for intrasite variation and the spatial position of artifacts 

and what it can tell us about past human behavior.  

The process of examining and interpreting artifacts share similarities to the periods in 

which analyses were conducted. When normative archaeologists examined blades, the confines 

of statistical analysis and typology created a clear picture of what blades were. Processual 

processes made blades part of system but did little to refine what encompassed blades. Now that 

we have reached a period encompassing many theories and models the definition of blades is 

beginning to expand to include valuable statistics, systemic processes, as well as environmental 

and cognitive influences on their creation and use. Like the blades of the Old World or 

Mesoamerica, Clovis blades have their own vein of research, and while widely varying at times, 

the role of blades in the Paleoindian toolkit is increasingly being examined next to some of the 

more prominent tool forms of the time. 

3.2 Defining Blades 

While the principle diagnostic artifact for Clovis classification is the lanceolate, fluted 

Clovis point, other objects are also synonymous with Clovis culture such as ivory points, large 

prismatic blades, and polyhedral blade cores (Collins, 1999:35). Endscrapers have also been 

diagnostic of Clovis in Pennsylvania (Whitthoft 1952; McNett, McMillan and Marshall 1977). 

Blades are a well-known technology, especially in Old World contexts, and have received their 

own level of examination concurrent with more widely recognized biface and fluted point 

technology. In Collins (1999) we see a concerted effort to examine the specific characteristics of 

Clovis blades and blade like flakes, not to confine them, but to statistically examine them for 



28 

 

variation. But a huge hurdle exists on how to define blades. There has been little agreement 

among archaeologists on how to define blades, and a wide variation of definitions of what makes 

a blade a blade. Several definitions have been proposed over the decades, notably that of Bordes 

(1961) as cited by Collins (1999) which defines blades simply as “any detached piece that is 

twice as long as it is wide” (Collins 1999). But there are also “true blades” which Bordes 1967; 

Bordes and Crabtree 1967, as cited by Collins (1999), elaborate as involving technique of 

production, not just length to width ratio. By refining definitions based on technique it has 

constricted what can be interpreted as a true blade and subjected all but the clearest of examples 

to classification as elongated flakes produced by chance (Collins, 1999). We see a clear 

demarcation between true blades and flakes by Crabtree (1982:16), who defined blades as: 

Specialized flake with parallel or sub-parallel lateral edges; the length is equal to, or more 

than, twice the width. Cross sections are plano convex, triangulate, subtriangulate, 

rectangular, trapezoidal. Some have more than two crests or ridges. Associated with 

prepared cone and blade technique; not a random flake.      

 

In comparison, Crabtree (1982:36) defines flakes as: 

Any piece of stone removed from a larger mass by the application force, either 

intentional, accidentally, or by nature. A portion of isotropic material having a platform 

and bulb of force at the proximal end. The flake may be of any size or dimension, 

depending on which technique was used for detachment. 

 

The definitions of blades have been more of a hindrance than a help, possibly eliminating blades 

from site catalogs for decades. To this point, Collins (1999) noted that the reporting of blades 

from Clovis contexts has likely been inadequate.  

Regarding the more objective positivist classifications of lithics that have occurred over 

the decades, Sorensen (2006:280) notes that “nothing seems more subjective than to force lithic 

assemblages into some subjective chosen metrical categories. The lithic analyst misses the 



29 

 

opportunity to recognize different technologies, to specify blade concepts and cultural groups or 

to interpret social interactions through technology.” Sorensen (2006) proposes rethinking the 

definition of lithic blades, from the perspective of technological use variation. Sorenson (2006: 

278) notes that blades are often classified as distinct groups situated between tools and waste in 

the traditional artifact hierarchy.  In his Old World research, he examined Paleolithic and 

Mesolithic blades and found that items once classified as waste were in fact blades. For this 

reason the examination of new sites for blades and revisiting previously excavated sites for the 

presence of blade technology is relevant and warranted, especially on sites where heavy 

collecting is believed to have stripped the surface of easily identifiable Clovis points. A growing 

focus on blade manufacture, relationships to source material, and regional variation is evident in 

research focused on the Southeast (Boulderain and Hoffman 2009, Bradley et al. 2010, Collins 

1999, Eren and Redmond 2011, Goodyear, 2006, Gramly 2013, Hagg and Jones 2022, Hagg et 

al. 2014, Sain and Goodyear 2016, Sain 2016: 2012, Tune et al. 2022). Sain & Goodyear note 

that “continued efforts to systematically assemble and share data on Paleoindian artifacts have 

generated new insight into patterns of land use, demographic trends, and raw material utilization 

for the Paleoindian period” (Sain & Goodyear, 2016: 128). Case in point is the experimental 

work related specifically to blade technology by Sorensen (2006) on Maglemosian blade 

technology in the Old World, whereby he postulates broader definitions of blades, thereby 

removing some of the confining typological standards and methodology. Sorensen’s examination 

is very much a culmination of decades of theory building and refinement, with data collection 

elements of normative, systemic constructs of processual, and cognitive examinations of process. 

In his experimental research Sorensen identifies not one type of blade, but four traditions or 

groups (Sorensen, 2006:287). Typology aside, each has its own set of distinct replicable traits 



30 

 

that facilitate dating of blades, type of cores, and tools used. From this Sorensen presents a new 

definition of blades:  

“A blade is a serially produced removal made with the intention of being a tool or 

preform for a tool. Blades in the same industry are produced by the same technique, 

method and mental representations and are characterized by a similar morphology and 

that same set of diagnostic attributes” (Sorensen, 2006:289) 

 

From a behavioral standpoint this makes sense as all outputs are a result of behavior, and there 

are countless variations that shape behavior - environmental factors being key. On those grounds 

blades should also be variable. Sorenson (2006) goes on to outline two benefits of this defining 

approach; focus can be placed on prehistoric human behavior, and archaeology’s traditional ways 

of approaching culture can give way to examining behavior though traditions defined by 

technological applications of the time. 

When we examine blades found at one geographic location through this lens as Sorensen 

describes it, we can localize behavior by not immediately lumping all artifacts sharing a common 

typology into the same technocomplex. Clovis blades as described by Collins (1999) are very 

much defining of a culture, but finding blades with varying morphology in the East does not 

imply Clovis blade culture was not present. Merely, we see the same culture applying varied 

technology to meet environmental and situational needs as explained by Sorensen (2006). By 

examining the artifacts from a technological approach of how they were applied to the immediate 

environment, we can understand the tools needed and used by a distinct group. These functional 

attributes define why blades are a key part for the Clovis toolkit. Blades could be produced and 

easily transported and future use. Blades could be used as created for their incredibly sharp 

cutting edge or modified to create a wide variety of tools such as scrapers, burins or as preforms 

for points. Their function as burins is noteworthy as it extends an individual’s ability to 



31 

 

manipulate features of the surrounding environment to create more complex tools which help 

them accomplish a particular task. This makes these types of tools essential for complex tasks 

and demonstrates higher order thinking on the part of the user. This goes well beyond blades 

being views solely for their sharp edge and immediate use as a cutting implement as it. This 

variation can speak volumes about the environmental factors affecting a people at a given time. It 

was Witthoft (1952) who said that a definitive blade technology existed in the East, like that of 

the Old World, and by the definition of true blade as defined by Collins (1999), this statement 

was accurate. However, if we remove the confines of typology and look at the function of blades, 

combined with Witthoft’s observations, why would a blade technology by a looser definition not 

exist at Thunderbird as blades serve the same function regardless of geographic region, and 

independent from the function of biface technology. The contents of a toolkit are not created to 

match a predetermined idea of what it should contain. Rather, it contains what is needed to meet 

the needs and expected tasks to come. And as people change locations those tasks may also 

change, making the Paleoindian toolkit a constantly evolving system effected by everything from 

season to geography. We must examine each toolkit in relation to its location and consider tools 

not only for their presence but also their absence and decipher what that means about the 

activities being taken at that time and place. 

Experimental archaeology has played a vital role in understanding the variability of 

toolkits in both form and function. Undoubtedly, the creation of lithics through experimental 

processes by archaeologists, or 13,000 years ago by Paleoindians in Northern Virginia, is a group 

processes rooted in the sharing of knowledge. Callahan (1979) applied experimental archaeology 

to investigate lithic variability in the early stages of reduction in the creation of Virginia fluted 

points. He noted that “experimental replication and techno functional activity are indispensable 



32 

 

aids to the lithic analyst in defining variability” (Callahan 1978:172). Callahan believed that 

there was significant variability in Virginia Clovis fluted points, and that several types can be 

isolated and defined. If this is the case for points, should we not consider variability when 

examining the creation of blades? For this reason, variability must be considered when 

developing and executing methodologies of examination, study, and comparison.  

3.3 Blades as a Key Component of Paleoindian lifeways 

While blades may not be at the forefront of the broader examination of Paleoindian 

lifeways, the many detailed excavations at sites across the East have amassed significant data 

and conclusions about chronology, subsistence, and tools that can support future study in the 

region. Numerous blade focused studies have been conducted in the past decade such as, (Hagg 

& Jones, 2022; Hagg et al. 2014; Tune et al. 2022; Sain 2012; Sain and Goodyear 2016), with 

many old sites being revisited from a new perspective.  These types of studies are beginning to 

create a clearer picture, and more importantly a new dataset, specifically related to Paleoindian 

blade technology in the East. Blades have been noted in Clovis contexts throughout the East.  

“Most sites are quarry-related reduction or temporary camp locales, where lithic extraction and 

tool manufacture were significant activities. Assemblages are dominated by locally available raw 

material, as sites are generally located adjacent to or near river drainages” (Sain 2012). 

These studies also highlight effective methodologies for conducting these types of blade centric 

research projects. Many site collections are housed at the Smithsonian, where these legacy 

collections are providing valuable datasets decades after the fieldwork that produced them. In 

many cases these legacy collections contain larger samples, which help place in context the 

limited data collected in more recent excavations (Joyce, 2022:28). This also affords researchers 



33 

 

an opportunity for intensive analysis while addressing the ethical considerations of excavations 

and the challenges of creating and housing new collections (Joyce, 2020). Significant 

excavations at the Flint Run Complex ended in the 1980s, but the value contained in their well-

documented collections is immeasurable. Considering the availability of data, and the challenges 

of new excavations, this thesis is tapping into the legacy collections housed at the Smithsonian 

while building upon decades of thoughtful modeling and research methodologies from many 

schools of thought.   

  



34 

 

Chapter 4 

 

4.1 Research Methods and Methodology 

By its nature lithic analysis is both a quantitative and qualitative process dedicated to the 

examination of artifacts, not the behavior that created them, as Binford (1989:3) noted. However, 

what the study of artifacts allows us to do is infer the actions that caused their creation, i.e., the 

behavior. In the case of Thunderbird, the goal should be to understand what took place there and 

why. “If the goal of science is to know something about the past that is dependent upon 

inference, then one must develop a reliable methodology that permits reliable inferences to the 

phenomena of interest” (Binford 1989:4). To this point, how others have chosen to define and 

examine blades provides significant guidance in the methods of lithic analysis. As shown in 

Chapter 3 there is not only great variability in how lithics are measured, but ongoing 

disagreement of how to define their attributes. Regardless, it is important that the methods of 

collection meet the goals of respective research, while remaining cognizant of proven best 

practices. This research uses some of the most common variables to ensure effective and 

meaningful data collection and make data comparable to that of previous research. The 

examination of blades within the Thunderbird collection leans heavily on the attribute analysis 

methods and variables explored and refined by Collins (1999), Sain (2012), Sain & Goodyear 

(2016), and Pargeter et al. (2023). The goal of many of these studies was to identify the 

analytical processes most likely to return consistent results when analyzing blades. Consistency 

in data is essential to analysis, which can span hundreds or thousands of artifacts from a site, and 

require comparison to other site collections regionally, nationally or across continents. Two 

approaches were used to examine artifacts withing the Thunderbird collection for the presence of 

blades and blade-like flakes. The first was a qualitative assessment requiring a visual inspection 



35 

 

based on the experience of the lithic analyst and comparison to blades documented in previous 

studies. The second method was quantitative and used both common measurements as well as 

those specific to this study.  

4.2 Collection, Location, and Condition 

The Thunderbird collection is located at the Smithsonian Museum Support Center (MSC) 

in Suitland, Maryland and is comprised of thousands of lithics, soil, and other organic samples. 

The collection is not publicly accessible and is available upon request to those who provide 

research-based justification. Access to the Thunderbird materials was approved by Torbin Rick, 

Smithsonian curator for Virginia archaeology collections. The repatriation policy for the 

Smithsonian National Museum of Natural History is set out in the National Museum of the 

American Indian Act of 1989, and the Smithsonian Museums are excluded from the Native 

American Graves Protection and Repatriation Act of 1990 (NAGPRA). Consultation under 

NAGPRA was not required for access to the collection (James Krakker, personal communication 

July 26, 2024). Documents related to the site are kept in the Smithsonian Archive separate from 

the collection. It is estimated that more than 100,000 artifacts were mapped during the 15 years 

of excavation (R. Verrey, personal communication, as cited by Carr et al. 2013). When the 

collection was accessioned “a preliminary, but incomplete count of selected artifacts and 

diagnostic flakes from the Paleoindian and Early Archaic levels were compiled by Smithsonian 

staff” as shown in Table 1 (Carr et al. 2013:174). 

 

  



36 

 

Table 1. Preliminary Counts on Artifacts from the Paleoindian-Early Archaic Levels at 

Thunderbird (Carr et al. 2013:174). 

Artifact     Count 
Bifaces      449 
Cores     98 
Scrapers     120 
Hammerstones     79 
End thinning flakes     35 
Unifacial absence     209 
Overshot flakes     35 

 

McNamara (1980) created the most complete rough classification which included artifacts like 

cores, utilized and unutilized flakes, prepared tools, unprepared tools, and unclassified items for 

a total of 98,965 artifacts from Areas 1B and 4 as well as 40 features. Evidence of previous 

examinations was apparent as items were sorted, boxed and or bagged, but there was no obvious 

indication of why catalogued. Many artifacts were meticulously labeled with unit and artifact 

number.  

Each MSC Unit is comprised of a metal cabinet with a right and left door, and a varying 

number of drawers on each side (Figure 5). The MRC term “unit” does not relate to the same 

term used in archaeology to identify and area of excavation. For the purposes of this study MU 

refers to storage units themselves, while AU refers to archaeological excavation units. The 

Thunderbird collection fills approximately 18 MUs. Drawers are primarily organized by AU and 

level, or by feature. Tools were separated from the collection and consolidated into a specific 

MU. These were grouped by AU, level, or feature. Although organized in this way, analysis of 

the collection required opening every MU and observing the contents of each drawer as the 

collection has never been catalogued. Each drawer contains individual trays of roughly classified 

materials, or those organized to various extents by previous researchers (Figure 6). All artifacts 



37 

 

were examined and measured at the MSC with the permission of MSC curation staff. Workspace 

was provided ajacent to the collections to setup and condust analysis (Figure 7).  

 

Figure 5. Collections storage at the Smithsonian Museum Support Center in Suitland, Maryland. 

Each two-door cabinet is designated a MU number, unrelated to the term used to describe an AU. 

The Thunderbird collection filled 18 MUs which is nearly all the cabinets on one side of the 

lighted hallway. Photograph by the author. 



38 

 

 

Figure 6. Drawers within MU at the Smithsonian housing artifacts from the Thunderbird 

collection. Photos by the author.  

 

 

Figure 7. The author at the Smithsonian archives and associated workspace.  



39 

 

4.3 Artifact Selection 

Given that the aim of this study is specific to answering questions about the Clovis 

period, only those drawers with artifacts from levels 6, 7, and 8 were examined. Previous 

research found that these levels correspond to the Clovis period, while levels 5 and above are 

Archaic and later (Carr et al., 2013). During the initial cursory examination of the collection, 

notes were applied to each drawer to indicate the levels and AUs it contained. Once the 

appropriate levels were identified, each artifact in those drawers was visually examined for 

attributes related to blade technology and pulled for closer examination (Figure 8). For this study, 

this means any lithic with one or more arrises originating at the proximal or distal, and the 

appearance of flake removal running parallel to the arrises. Arrises refer to the sharp crests on the 

surface of the stone where two or more flat or curved surfaces meet, or literally in this case 

where previous flake scars terminate (Figure 9). Flake removal perpendicular to the proximal and 

distal ends would be indicative of biface manufacture. The notable exception would be in the 

creation of formal crested blades and potential core preparation blades. For this reason, attention 

was paid to the direction of flake removal when selecting blade like flakes to analyze further. 

This was not always obvious. These criteria apply regardless of length, as blades may be 

represented in a collection as shortened, intentionally snapped, or used segments (Sorensen, 

2017). For this reason, many of the blade like flakes examined would likely be classified as 

debitage and not true blades by most classification systems as they lack the minimum necessary 

2:1 length width ratio. 



40 

 

 

Figure 8. A selection of blade-like flakes pulled from boxes of debitage for further analysis at the 

Smithsonian archives. Photo by the author. 

 

 

Figure 9. A single arris on a small blade-like flake originating at the proximal platform and 

extending to the distal end. Photo courtesy of Mike Johnson.  

 



41 

 

4.4 Measurement Tools 

The following tools were used to collect measurements and imagery of each artifact. 

• Sliding caliper 0-80mm 

• Digital scale 0.01-200g 

• Contour gauge consisting of 145, 1mm blades. 

• Digital angle ruler 

• iPhone 12 

• Photo booth with lights 

• Rubber pad – anti slip and impact proception for artifacts 

• Foam pad – anti slip and tool storage  

• Toothbrush 

• Data collection forms 

• 1.8X head mounted magnifier 

 

 

Figure 10. The workspace as setup during the study containing a portable LED lighted photo 

booth, forms and measurement tools used to capture artifact data. Photo by the author. 

4.5 Measurements and Recording 

To answer the questions of this study it is necessary to efficiently document both the 

qualitative and quantitative data captured. Following the methodology of Pargeter (2023), this 

study segmented data into three classes. Ratio-scale attributes, examples of which include length, 



42 

 

width, and weight. These attributes have equal intervals between neighboring points and scale 

from true zero. Discrete-scale attributes count variables on whole numbers such as arrises or 

flake scars. And Nominal attributes which are non-numeric such as thermal alteration or 

intentional surface polishing. By combing these variables artifacts can be effectively isolated by 

the core attributes related to blade production. This by no means verifies if an artifact is a blade 

but allows for further segmentation based on the available data during analysis. Blades, however, 

have both quantitative attributes, such as length-to-width ratio, and qualitative attributes such as 

the number and orientation of arrises and flake scars, which when combined are telltale signs of 

an intentional blade technology. If blades were at Thunderbird the combination of qualitative and 

quantitative data should separate blades from blade like flakes as well as from accidental blade 

like flakes. All data was captured by hand on a blade attribute form created for this study. The 

form was organized into four sequential sections: Administrative, Measurements, Blade 

Overview, and Modification (Appendix 1). 

Administrative 

Administratively, each artifact was assigned a date of analysis, a unique analysis 

identification starting with TB-001, the Smithsonian MU, MU side (Left, or Right), Drawer 

number, artifact location in the drawer (front, middle or back). Any original Thunderbird 

provenience or recording information either on the artifact or on any associated tags were 

captured for each artifact.  

One thing that must be noted is that several drafts of the blade attribute form were created 

between May 2023 and March 2024 while the analysis was ongoing. And while each iteration 

refined the data being collected, it produced an increasingly complex dataset. This means that 

earlier artifacts have fewer data points than those collected later. However, many of the core 



43 

 

attributes were present from the start of the analysis and samples were adjusted accordingly 

when performing more complex relational calculations later, as will be demonstrated in Chapter 

5.  

It became apparent during analysis that by taking measurements first, before conducting 

any visual analysis, the opportunity to study the object was increased. Although analytical 

examination is considerably more focused, by conducting measurement first it doubles the 

amount of time the object is held and viewed thereby allowing the analyst greater opportunity 

during handling to examine the artifact before entering more subjective analytical data such as 

use-wear or thermal alteration. It remains to be seen if this has an impact on the outcome of 

analysis, but it is this researcher’s opinion that it improved the confidence of analysis. The 

average processing time per artifact was 15 minutes using this method. Processing time averaged 

more than 20 minutes when visual analysis was conducted first followed by measurement.  

Measurement 

Weight was captured in grams. Each segment was classified as whole, medial, distal, 

Proximal/Medial, Medial Distal, or Unknown. Proximal and distal segment terminus 

measurements were captured.  

  As a point of clarification, while researchers have commonly taken proximal, medial, and 

distal measurements, I was not able to find in my research a term specifically describing the 

broken span of a blade, from one lateral edge to another. Most blades are broken, leaving a 

proximal section with a break, a medial section with two breaks, or a distal section with one 

break. I propose the term proximal terminus, and distal terminus to specifically identify the span 

from one lateral edge to another. Proximal is commonly used to describe the top portion of a 

flake, closest to the platform. In this case I am using proximal terminus to indicate the break 



44 

 

closest to the proximal end. Likewise, distal terminus refers to the break furthest from the 

proximal end. Thereby, a proximal segment would only have a distal terminus while a medial 

segment would have both a proximal and distal terminus. Likewise, a distal segment would have 

one proximal terminus. The purpose of the terminus measurement is to be able to compare in a 

dataset those segments with matching distal terminus and proximal terminus measurements for 

the potential refitting of broken blades. While these measurements have not been proven useful, 

it is a measurement of opportunity that may pay dividends across a large dataset already 

undergoing significant measurement (Figure 11). It also removes some ambiguity of the 

measurement of lower 1/3 measurements which do not specify where the measurement should be 

taken.  

 

Figure 11. Capturing terminus measurements in the event two broken flakes have matching 

widths, possibly resulting in a mend.    

 

 



45 

 

Blade Overview 

  The morphology of the Proximal end was classified as hinged, stepped, or snapped. 

Distal morphology was classified as hinged, stepped, snapped, feathered, or overshot. Maximum 

dimensions of length, with and thickness were taken with calipers in cm, to the hundredth if 

needed. Width/Thickness measurements were taken at the Proximal 1/3, Medial 1/3, and Distal 

1/3. Platform width and thickness were measured in cm, to the hundredth if needed. Bulb tota