i 
 
An Analysis of the Symmetry of Large 
Cutting Tools within the South African 
Acheulean 
Raymond Alexander Couzens 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
A dissertation submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in 
fulfilment of the requirements for the degree of Master of Science 
Johannesburg 2012 
 
ii 
 
DECLARATION 
 
I declare that this thesis is my own, unaided work.  It is being submitted for the Degree of Master of 
Science in the University of the Witwatersrand, Johannesburg.  It has not been submitted before for 
any degree or examination in any other University. 
 
 
 
 
_______________________________________ 
 
(Signature of candidate)  
Raymond Couzens 
 
 
 
 
____________________________day of ________________________20________________ 
 
 
 
 
 
 
 
 
 
 
 
iii 
 
ABSTRACT 
 
The use of three dimensional modelling techniques with reference to the study of 
archaeological material is one that is gaining popularity in hominid studies and is already 
being extensively used globally. This research delves deep into the Acheulean period and 
takes a refreshed look at the symmetry of handaxes from two sites,  namely Rietputs 15 (1.4 
ma) which is an early Acheulean site, and the Cave of Hearths, which is estimated to ca 
450/500 000 years old and forms the later Acheulean aspect of this sample. This research 
focuses on creating effective methods for studying symmetry in relation to various variables 
specific to each site, and it aims to gather data using 3D methods that more traditional 2D 
techniques struggle to capture. Ultimately this data provided me with a quantified measure of 
symmetry for handaxes from the two sites. For the Cave of Hearths, statistical evaluation of 
the  measures of left versus right volumes showed strong, statistically significant correlations 
(r = 0.870, p < 0.05), as did measures of left versus right surface areas (r = 0.960, p < 0.05).  
Rietputs provided comparable results of: r = 0.859, p <0.05  for volume, and r = 0.954, p 
<0.05  for area, thus suggesting that good symmetry exists. By using sectoral analysis, this 
study shows that the tip is the most variable sector of the tools for both sites. This result 
supports the assumption that handaxes were designed for varied functions (e.g., cutting, 
skinning, digging roots, or working wood) but ones which required a strong distal end. The 
medial and proximal sectors are both relatively less variable, and their properties may have 
been more constrained by the convergent shape of the tool. Values for the later Acheulean 
sample show only slightly less variability than for the early Acheulean, but this is 
nevertheless an interesting trend which relates to hominid and cultural evolution over ca 1 
million years of time. 
 
 
 
 
 
 
 
 
 
 
 
 
iv 
 
ACKNOWLEDGEMENT 
 
I would like to firstly thank the Palaeontological Scientific Trust (PAST) and the National Research 
Foundation (NRF) for the funding that made a majority of this research possible. I would also like to 
give thanks to the Institute for Human Evolution (IHE) for purchasing the scanning hardware and 
software necessary for this project. I thank all the individuals who helped train me in using the 
associated software and made me feel welcome within the IHE department. I would especially like to 
thank my co-supervisor Dr Kristian Carlson for providing limitless help within the 3D and statistical 
world. I would further like to thank Dr Alexandra Sumner for supporting me during my research and 
being someone with whom I could easily discuss my opinion. I would like to extend a thank you to 
Professor Terrance McCarthy for his expert opinion in helping me identify the raw materials of my 
specimens. Lastly I would like to give enormous thanks to my principle supervisor Professor Kathleen 
Kuman for guiding me through the last two years. It has been quite a journey and I would not have 
completed this thesis without her input. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
v 
 
 
DEDICATION 
 
This thesis is dedicated to my mother and father for getting me interested in Archaeology before I 
could write out the alphabet, let alone complete a Masters thesis. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
vi 
 
 
TABLE OF CONTENTS 
 
Declaration……………………………………………………………………………………………...ii 
Abstract……………………………………………………………………………………………...…iii 
Acknowledgements…………………………………………………………………………………….iv 
Dedication………………………………………………………………………………………………v 
Table of contents……………………………………………………………………………………….vi 
List of Figures………………………………………………………………………………………….ix 
List of Tables……………………………………………………………………………………....xi-xiii 
 
CHAPTER 1: Introduction……………………………………………………………………………..1 
1.1 Introduction…………………………………………………………………………………………1 
1.2 Research Goals……………………………………………………………………………………...3 
 
CHAPTER 2: Literature Review……………………………………………………………………….5 
2.1 The Acheulean and change through time…………………………………………………………...5 
2.2 Handaxe Symmetry and its traditional study……………………………………………………….7 
2.3 Three-dimensional scanning and the future……………………………………………………….11  
 
CHAPTER 3: Methodology…………………………………………………………………………...13 
3.1 Scanning…………………………………………………………………………………………...14 
3.1.1 Cleaning and merging…………………………………………………………………….15 
3.2 Editing……………………………………………………………………………………………..17 
3.2.1 Alignment and Measuresments…………………………………………………………...17 
3.2.2 Division of tools………………………………………………………………………….19 
3.3 Recording………………………………………………………………………………………….21 
3.4 Evaluating Symmetry using Statistics……………………………………………………………..22 
vii 
 
 
CHAPTER 4: Materials……………………………………………………………………………….23 
4.1 Sample and Raw Materials………………………………………………………………………..23 
4.1.1 Rietputs 15 Raw Materials..................................................................................................23 
4.1.2 Rietputs 15 Blank Types………………………………………………………………….25  
4.1.3 Cave of Hearths Raw Materials…………………………………………………………..25 
4.1.4 Cave of Hearths Blank Types…………………………………………………………….27 
 
CHAPTER 5: Results…………………………………………………………………………………28 
5.1 General…………………………………………………………………………………………….28 
5.2 Results……………………………………………………………………………………………..28 
5.3 Interpretation……………………………………………………………………………………....29 
5.3.1 The Coefficient of Variation (CV)……………………………………………………….30 
5.3.2 Interpretation……………………………………………………………………………..30 
5.3.3 Blank Type and Symmetry……………………………………………………………….31 
5.3.4 Raw Material and Symmetry……………………………………………………………..32 
5.4 Volume Scatterplots……………………………………………………………………………….32 
5.4.1 Rietputs 15 Volume………………………………………………………………………32 
5.4.2 Cave of Hearths Volume…………………………………………………………………33 
5.5 Area Scatterplots…………………………………………………………………………………..34 
5.5.1 Rietputs 15 Area………………………………………………………………………….34  
5.5.2 Cave of Hearths Area……………………………………………………………………..36 
5.6 Raw Material Scatterplots…………………………………………………………………………36 
5.6.1Rietputs 15 Area…………………………………………………………………………..36 
5.6.2 Rietputs 15 Volume………………………………………………………………………38 
5.6.3 Cave of Hearths Area……………………………………………………………………..39 
5.6.4 Cave of Hearths Volume…………………………………………………………………41 
5.7 Coefficient of Variation Tables……………………………………………………………………42 
viii 
 
5.7.1 Area……………………………………………………………………………………….42 
5.7.2 Volume…………………………………………………………………………………...43 
5.7.3 Cave of Hearths Symmetry and Blank Type…………………………………………......44 
 
CHAPTER 6: Discussion……………………………………………………………………………...45 
6.1 Discussion…………………………………………………………………………………………45 
6.2 The Question of Handedness……………………………………………………………………...47 
 
CHAPTER 7: Conclusions……………………………………………………………………………48 
 
APPENDICES: CV Tables Showing all Data Presented in this Study…………………………....50-57 
APPENDIX A: All data, Rietputs 15 Area………………………………………………………........50 
APPENDIX B: All data, Cave of Hearths Area………………………………………………….........52 
APPENDIX C: All data, Rietputs 15 Volume………………………………………………………...54 
APPENDIX D: All data, Cave of Hearths Volume…………………………………………………...56 
 
REFERENCES…………………………………………………………………………………….58-64 
 
 
 
 
 
 
 
 
 
 
 
ix 
 
 
LIST OF FIGURES 
 
Figure 1: Map depicting locations of Rietputs 15 and the Cave of Hearths..............................3 
  
Figure 2: Conventional positioning and measurement extraction (from Roe, 1964 and after 
Grosman et al. 2008)................................................................................................................13 
 
Figure 3: Depicting Realscan USB scanner, table and scanning stand………………………14 
 
Figure 4: Manual alignment window………………………………………………………...16 
 
Figure 5: Full pre-merge scan showing incompatible surface……………………………….17 
 
Figure 6: Aligned tool complete with bounding box………………………………………...18 
 
Figure 7: Handaxe scan complete with sectoral divisions, landmarks and bounding box…...19 
 
Figure 8: Division of tools into sectors using a Wacom tablet………………………………20 
 
Figure 9: A complete tool divided into 6 sections…………………………………………...21 
 
Figure 10: Raw materials at Rietputs 15..................................................................................24 
 
Figure 11: Blank Types at Rietputs 15.....................................................................................25 
 
x 
 
Figure 12: Raw materials at the Cave of Hearths.....................................................................26 
Figure 13: Blank Types at the Cave of Hearths……………………………………………...27 
 
Figure 14: Volume scatterplot of left and right sides for Rietputs 15 handaxes………….…32  
 
Figure 15: Volume scatterplot of left and right sides for the Cave of Hearths handaxes……33 
 
Figure 16: Area scatterplot of left and right sides for Rietputs 15 handaxes………………...34 
 
Figure 17: Area scatterplot of left and right sides for the Cave of Hearths handaxes………..35 
 
Figure 18: Area scatterplot of left and right sides of Rietputs 15 handaxes by dominant raw 
material……………………………………………………………………………………….36 
 
Figure 19: Volume scatterplot of left and right sides of Rietputs handaxes by dominant raw 
material……………………………………………………………………………………….38 
 
Figure 20: Area scatterplot of left and right sides of Cave of Hearths handaxes by dominant 
raw material…………………………………………………………………………………..39 
 
Figure 21: Volume scatterplot of left and right sides of Cave of Hearths handaxes by 
dominant raw material………………………………………………………………………..41 
 
 
 
 
 
 
 
 
 
 
 
 
xi 
 
LIST OF TABLES 
 
Table 1: showing relationship between Raw Material type and Blank Type for the site 
Rietputs 15……………………………………………………………………………………24 
 
Table 2: showing relationship between Raw Material type and Blank Type for the site Cave 
of Hearths…………………………………………………………………………………….26 
 
Table 3: Pearson Correlation for volumes of left and right sides of Rietputs 15 handaxes….33 
 
Table 4: Pearson Correlation for volume of left and right sides of the Cave of Hearths 
handaxes……………………………………………………………………………………...34 
 
Table 5: Pearson Correlation for area of left and right sides of Rietputs 15 handaxes………35 
 
Table 6: Pearson correlation for area of left and right sides for Cave of Hearths handaxes…36 
 
Table 7: Pearson correlation for area of left and right sides for Rietputs 15 handaxes for 
hornfels……………………………………………………………………………………….37 
 
Table 8: Pearson correlation for area of left and right sides for Rietputs 15 handaxes for 
ventersdorp lava……………………………………………………………………………...37 
 
Table 9: Pearson correlation for volume of left and right sides for Rietputs 15 handaxes for 
hornfels……………………………………………………………………………………….38 
xii 
 
 
Table 10: Pearson correlation for volume of left and right sides for Rietputs 15 handaxes for 
ventersdorp lava……………………………………………………………………………...39 
 
Table 11: Pearson correlation for area of left and right sides for Cave of Hearths 15 handaxes 
for dolerite……………………………………………………………………………………40 
 
Table 12: Pearson correlation for area of left and right sides for Cave of Hearths handaxes for 
quartzite………………………………………………………………………………………40 
 
Table 13: Pearson correlation for volume of left and right sides for Cave of Hearths 15 
handaxes for dolerite…………………………………………………………………………41 
 
Table 14: Pearson correlation for volume of left and right sides for Cave of Hearths handaxes 
for quartzite…………………………………………………………………………………..42 
 
Table 15: Coefficient of Variation for linear measurements and area in the Rietputs 15 
assemblage…………………………………………………………………………………...42 
 
Table 16: Coefficient of Variation for linear measurements and area in the Cave of Hearths 
assemblage…………………………………………………………………………………...43 
 
Table 17: Coefficient of Variation for linear measurements and volume in the Rietputs 15 
assemblage…………………………………………………………………………………...43 
 
xiii 
 
Table 18: Coefficient of Variation for linear measurements and volume in the Cave of 
Hearths assemblage…………………………………………………………………………..43 
 
Table 19: Coefficient of Variation for linear measurements and volume of end-struck pieces 
for the Cave of Hearths sub-sample (n=9)…………………………………………………...44 
 
Table 20: Coefficient of Variation for linear measurements and volume of side-struck pieces 
for the Cave of Hearths sub-sample (n=10)………………………………………………….44 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1 
 
CHAPTER 1: INTRODUCTION 
1.1 Introduction 
 
The Acheulean period is one of extreme length and absolute importance for researchers 
attempting to understand the thought processes that helped in the evolutionary 
progression from early tool producers such as Homo ergaster to H. sapiens. Within this 
time period a purely unique tool form known as the handaxe was being produced. Due to 
the fact that handaxes are the most purposefully shaped tool in the Earlier Stone Age and 
exhibit significant variations of form, they have caused researchers to invest a much more 
considerable amount of effort into their research than other tool types of the Early Stone 
Age. 
  
In recent literature, symmetry and standardisation of stone tools are thought to have 
improved with time. This study takes an in-depth look at this phenomenal stone type in 
the 3D realm and seeks to apply a methodology that builds upon older more traditional 
techniques in order to study symmetry in a more comprehensive three-dimensional way. 
 
The Acheulean techno-complex follows the Oldowan tradition for East Africa in time, 
first appearing at around 1.76 ma in the Nachukui Formation, West Turkana, Kenya 
(Lepre et al. 2011). A second early date of 1.7 ma for the Acheulean exists for the East 
African site Konso Gardula in Ethiopia (Beyene et al. 2007). Other sites such as Olduvai 
Gorge exhibit dates for the Acheulean that range between 1.6 and 1.4 ma respectively 
(Hay 1976; Isaac 1997). 
 
For southern Africa, the number of sealed Acheulean sites that exist or are able to be 
reliably dated is minimal. Thus we rely mainly on relative dating through faunal 
comparisons with East African assemblages (such as those mentioned above) where 
absolute dating is feasible because of the presence of dateable volcanic sediments (Klein 
2000; Kuman 2007). More recently the cosmogenic nuclide burial method has been 
applied by Gibbon et al. (2009) to the Rietputs Formation (near Windsorton in the Vaal 
River basin, about 1.5km from the modern Vaal river (Figure 1), and dates for the early 
Acheulean in southern Africa now range between 1.69 ± 0.17 and 1.14 ±0.20 ma (Gibbon 
2 
 
et al. 2009; Kuman pers comm.). These dates suggest a well dispersed use of Large 
Cutting Tools (LCT’s) in Africa by ca. 1.6 ma (Gibbon et al. 2009). 
The handaxe sample from Rietputs 15 (n=33 excluding tipless types) makes up the early 
Acheulean part of my sample and is dated to ca 1.4 ma (Gibbon et al. 2009). It was 
retrieved by George Michael Leader from diamond mining operations in 2007 as part of 
his MSc research (Leader 2009). Some earlier work had been done on the site by C. van 
Riet Lowe and the Abbé Breuil  at the time when Windsorton was an active diamond 
mining town (van Riet Lowe 1945). 
 
The Cave of Hearths handaxes represent the later Acheulean aspect of this sample (n=31 
excluding tipless specimens). The site is located in the Limpopo Province some 15km 
from Makopane (formerly Potgietersrus). More specifically the actual cave can be found 
on the southern slope of the Mwaridzi valley which is actually a smaller portion of the 
much larger Makapansgat Valley (McNabb et al. 2004) (see Figure 2). The site was 
discovered by van Riet Lowe in 1937 and excavations were carried out by Guy Gardner, 
James Kitching and Revil Mason between 1947 and 1957. Later one of the most 
important finds, a hominin mandible attributed by Tobias (1971) to Homo sapiens 
rhodesiensis, was discovered in Bed 3 (McNabb et al. 2004). The sample represented in 
this study comes from Beds 1-3. It is typical of the later Acheulean in a number of ways. 
Some researchers group the mandible with Homo heidelbergensis and thus a date of ca. 
400-500 ka for these beds is generally accepted (McNabb et al. 2004). The site is one of 
utmost importance when considering the Stone Age in South Africa, particularly as it 
contains a long sequence including the Earlier, Middle and Later Stone Age (McNabb et 
al. 2004).  
 
The comparison between these two handaxe samples thus relates to an extreme temporal 
scale (between 1.4 ma and ca 0.45 ma) and this remains the main reason for their 
selection in this research. If trends occur in symmetry or standardization, they should be 
easier to determine at these two extremes in time and should help us better understand the 
Acheulean in southern Africa.   
3 
 
 
Figure 1: Map depicting locations of Rietputs 15 and the Cave of Hearths  
 
1.2 Research Goals 
Because most handaxes are considered to be commonly shaped to a tip and are bilaterally 
symmetrical, researchers such as McNabb et al. (2004) argue that some kind of cultural, 
symbolic or functional action must be at play in the minds of the makers. 
  
This study aims to look at the symmetry of handaxes from the above mentioned sites 
mainly because of the explicit features identifiable in each sample. More specifically, 
many early Acheulean handaxes from Rietputs 15 are shaped to a rather crude, robust or 
pick-like tip, while the later Cave of Hearths handaxes seem to be better worked around 
their edges and also are better worked at the tip. The in-depth study of these 
characteristics could hold some functional or stylistic value in their interpretation and thus 
the study of their standardization or lack thereof could be important considering the long 
temporal scale that is available to this study (McNabb et al. 2004; Leader 2009). 
 
The use of three-dimensional modelling techniques with reference to the study of 
archaeological material is a discipline that is gaining popularity in current studies, and 
4 
 
new research is constantly being produced as an effect of the exponential rate of 
improvement in modern day technology. Three-dimensional studies are important 
because they can produce highly accurate measures such as the more common length, 
width and thickness applied in archaeology, as well as provide advanced measures such 
as volume, surface area and centre of mass that have until quite recently been 
understudied. The manufacture of three-dimensional databases is also important to 
archaeology because they provide easily accessible data that can be shared across borders 
without the potential risk of any type of damage occurring to a given assemblage (Sumner 
& Riddle 2008).  
 
This study stands to establish a suitable methodology for assessing symmetry that can be 
applied to complete handaxes through the use of a portable laser scanner. It draws on 
ideas often used during the traditional documentation of tools but primarily aims to build 
upon these in a three-dimensional way, taking into account variables such as the raw 
material and blank types present in the tools for each site. Put more explicitly, it aims to 
show standardization or lack thereof for early and later Acheulean assemblages, as well as 
assist in creating a new method for handaxe alignment within the three -imensional realm 
that is based primarily on what Grosman et al. (2008) calls the intrinsic morphology of 
the tool at hand. The method put forward here is self-contained by the object being 
studied and offers an extremely simple way of aligning handaxes for three-dimensional 
research to obtain values such as volume and surface area, which more classical methods 
struggle to quantify.  The contribution of this research is to portray the nature of the 
Acheulean in greater detail than more traditional methodologies allow, ultimately taking 
us deeper into the minds of the makers of these tools. 
 
 
 
 
 
 
 
 
5 
 
CHAPTER 2: LITERATURE REVIEW 
 
2.1 The Acheulean and change through time 
 In East Africa, the earliest Acheulean begins at 1.76 ma from the Nachukui Formation in 
West Turkana, Kenya (Lepre et al. 2011); a second and more recent early date has been 
established at about 1.7 ma at Konso Gardula, Ethiopia (Asfaw et al. 1992; Beyene et al. 
2007). Other sites in the Rift Valley date between 1.6 and 1.4 ma, namely East Lake Turkana 
(Kenya) and the middle and upper subdivisions of Bed II at Olduvai Gorge (Hay 1976; Isaac 
1997). The latest date for the Acheulean is yet to be firmly established, but Porat et al. (2010) 
suggest that the Middle Stone Age (MSA) replaces the Acheulean at a minimum of 464 ± 47 
ka, according to recent OSR dates from Kathu Pan. For Kenya, Ethiopia and Zambia, reliable 
estimates show that the Acheulean is replaced by the MSA between 250 to 200 ka (Klein 
2000). The Acheulean period thus spans an extensive time period before it is arguably 
replaced by the Middle Stone Age some 250-200 ka according to Klein (2000). 
 
The earliest confirmed Acheulean sites in southern Africa occur in two localities, namely the 
Sterkfontein valley, (consisting of sites such as Sterkfontein, Swartkrans and Kromdraai A in 
Gauteng Province) and the alluvial deposits of the Rietputs Formation located in the lower 
Vaal River basin (Kuman & Gibbon, in press). Handaxes have been found at both 
Sterkfontein and Swartkrans, and although no handaxes or cleavers were found at Kromdraai 
A, the site is still regarded as Acheulean (Kuman et al. 1997). The artefacts at these three 
sites are thought to have belonged to the same industry in that they are all very similar 
according to Kuman & Gibbon (in press). Sterkfontein Member 5 contains a fragmentary 
fossil of Homo ergaster and is estimated to be about ± 1.5 ma (Kuman & Clarke 2000), 
Swartkrans Members 2 and 3 are estimated to date ca. 1.5 and 1.0 ma respectively  (Kuman 
& Gibbon in press). Emphasis on the tips of tools from Sterkfontein and Swartkrans seems 
apparent as most exhibit a very sturdy and often pick-like nature which could have related to 
their function (Kuman and Gibbon in press). 
 
For South Africa Gibbon et al. (2009) suggest that the cosmogenic nuclide burial dating of 
the Rietputs Formation is important because it captures very early absolute dates for the 
Acheulean outside of Gauteng Province, where relative faunal dating has been used at the 
6 
 
Cradle of Humankind sites. In this case, dates for the early Acheulean at Rietputs range 
between 1.69 ± 0.17 and 1.14 ± 0.20 ma for individual artefacts or clasts containing quartz. 
The layer is sealed by a fine alluvium lacking artefacts with a date of 1.26 ± 0.10 ma (Gibbon 
et al. 2009). The large error margin in the earliest appearance date conforms to an age of 1.6 
ma. The Cradle of Humankind and Rietputs Formation results therefore suggest a rapid and 
well distributed spread of handaxes and their users within southern Africa and throughout 
Africa by ca 1.6 ma.  
 
Kuman (2007), like Mason (1962), believes that between the early Acheulean (1.7 to ca 1 
ma) and the later Acheulean (0.6 to 0.2 ma), a ''middle'' Acheulean phase exists that may be 
applicable to some South African sites based on associated fauna or on lithic typologies. 
Kuman (2007) points out that many researchers avoid this debate and allocate all sites 
younger than or equal to 1 ma into the ''upper'' Acheulean. However, the concept of a ''middle 
Acheulean'' is important because (and this has become apparent at Olduvai Gorge) it 
increases the resolution of the details related to technological change over time. This change 
is apparent when considering the tools excavated from Olduvai Gorge Beds II to IV (dated 
1.60 to ca 0.8/0.6 ma) and post Bed IV (dated ca. 0.6-0.4). In these deposits, handaxes 
develop from thicker and less standardized forms in Bed II to more regularized shapes (and 
more cleavers) in Bed IV later Acheulean (Roe 1994). After Bed IV in time, artefacts exhibit 
characteristics that indicate increased technological control. A similar pattern occurs in the 
Middle Awash region of Ethiopia, where earlier Acheulean LCT's ca 1 ma (‘’middle 
Acheulean’’) in the Daka Member are followed by more refined examples from the Herto 
Member ca 0.4 ma (later Acheulean). In terms of technology, the progression through time 
resembles that of Olduvai Gorge beds following the early Acheulean (Schick & Clark 2003). 
In addition the long sequence at Olorgesailie in Kenya, dated 0.99 ma to 0.49 ma (Isaac 1977; 
Potts 1989), provides further examples of ''middle'' to ''later'' Acheulean assemblages in East 
Africa. The identification and dating of ''middle Acheulean'' assemblages in South Africa now 
also appears feasible with current research in the Vaal River basin (Kuman, pers. comm). 
 
Large Cutting Tools such as the handaxe, cleaver and pick [Mode II artefacts, described by 
Clark (2001)]signal the beginning of the Acheulean and are often said to be made according 
to a ''mental template'' because they are all commonly shaped to a point (in the case of 
handaxes and picks) or to a broad cutting edge in the case of cleavers (Wynn 1995; Klein 
2000; Kuman 2007). In most cases early Acheulean bifaces (a generic term often used for 
7 
 
handaxes and cleavers) are, according to Klein (2000: 110): ''thicker, less extensively 
trimmed and less symmetric than later ones''. Tools of this time very often have less than 10 
removals, with deep scar patterns indicating a hard hammer technique. Trihedral, pick-like 
handaxes are also common in some early Acheulean assemblages (e.g., Asfaw 1992). This 
contrasts with later Acheulean bifaces which can be very thin, bilaterally symmetrical, 
sometimes intensely worked, and finished with soft hammer percussion resulting in greater 
regularity of form (Klein 2000; Goren-Inbar & Sharon 1999; Goren-Inbar & Sharon 2006). 
Crude bifaces also exist in the later Acheulean, but the general trend is toward a more refined 
technique.  
 
It is uncertain as to why the symmetry of handaxes increased over time, but it has been 
suggested that differences in handaxe morphology could be based on functionality (Clark 
2001), rather than increased cognitive ability as suggested by Klein (2000). The pick-like 
nature of many early handaxes (e.g., Asfaw 1992) could indicate that functions other than 
butchery may have been prominent in the early Acheulean, for example, digging for roots. 
Replicative experiments have revealed that the sharp edges of handaxes could have been 
useful for the cutting and/or skinning of game (Toth 1985); and experimental studies on cut-
marked bone at Swartkrans by Pickering et al. (2008) further confirm this prognosis. With 
regard to variability, Clark (2001) suggests that the selection of raw materials and 
functionality plays the greatest role in the morphological variations among tools, although he 
stipulates that aspects such as environment and age or skill level of the knapper (although 
unidentifiable at present) should also be considered. The opportunistic availability of raw 
material can also lead to variability of tool attributes (Clark 2001; Klein 2000). 
 
Work by Crompton & Gowlett (1993) using mathematically predictive techniques known as 
allometry has suggested that variability can be explained by size of the blank at the time of 
production. They explain that a bigger blank type would lead the producers to make sacrifices 
to other parts of the tool in order to keep the tool balanced in terms of weight thus resulting in 
varied shape (Crompton & Gowlett 1993; McPherron 2000).    
 
2.2 Handaxe symmetry and its traditional study 
 
The handaxes of the Acheulean period exhibit the most classical item of study for researchers 
attempting to understand hominid cognition because they are so unique in form.  In 
8 
 
archaeology the term bilateral symmetry is used to describe an object that appears to have 
equal and opposite sides that mirror each other when viewed on a plane (for 3D) or viewed 
on an axis (for 2D). The study of this sort of phenomenon is one that has received attention 
from many different intellectual sectors including art and science (Gowlett 2011; Saragusti et 
al. 2005). 
 
Traditionally ways to describe handaxes were based on what Grosman et al. (2008) described 
as a ‘language’. This language consisted of two divisions: the descriptive and the metric. The 
former pertains to basic labelling of form, for instance the term tear drop shaped, while the 
latter describes tools in terms of simple measurements such as length, width and thickness. It 
also makes use of ratios such as the index of refinement (which put mathematically is 
maximum thickness divided by maximum width) used by Roe (1968), and the index of 
elongation (width divided by length) cited by McPherron (2000). These ratios can be simply 
used to define certain properties of a handaxe. The index of refinement basically indicates 
handaxe modification so when the ratio is larger the handaxe will be more unrefined 
(Grosman et al. 2008). The same applies for the index of elongation, which indicates how 
long a handaxe is based on the ratio given (McPherron 2000). The use of allometry (which is 
a predictive mathematical model mentioned above) has also been used by Crompton and 
Gowlett (1993) to give functional explanations about the shape of artefacts. Allometry 
basically describes how the physical attributes of a tool react to one another when one 
attribute of form is modified, i.e. what happens to width when length is increased.  
 
In early studies, handaxes were generally considered to be symmetrical based upon the use of 
the methods described above, but it is becoming more obvious that these hypotheses are 
merely based on assumption and that qualitative studies, although useful in certain contexts, 
can only offer very little about a given assemblage in its macro scale. Various authors in the 
realm of lithics posit that qualitative studies do not adequately account for the variability 
shown amongst handaxes that span broad geographical as well as temporal contexts (Archer 
& Braun 2010; Grosman et al. 2008; Clarkson 2001; Saragusti et al. 1998, 2005). One such 
example of a study of symmetry based on loose qualitative methodology is McNabb, Binyon 
and Hazelwood’s By-Eye test that was carried out on the Cave of Hearths handaxes (McNabb 
et al. 2004), on which this present study focuses. Their test for symmetry involved dividing a 
given tool up into six sectors before mentally imagining the tool folded along its long axis 
and then giving each sector a simple yes or no score as to whether or not it was deemed 
9 
 
symmetrical. This method was challenged by Underhill (2007), and while aspects of the 
study were useful such as the explanation of tip types and the increased rate of study for the 
tools, the measure of actual symmetry was a highly subjective one. 
 
Saragusti et al. (1998, 2005), in contrast,provide a method to measure symmetry 
quantitatively, which they use to help explain variability on an inter- and intra- assemblage 
basis. The Continuous Symmetry Measure (CSM) proposed by Saragusti and colleagues 
displays “the minimal distances that the vertices of a shape have to go to attain the desired 
symmetry” through the use of mathematical equations or algorithms (Saragusti et al. 
1998.819). At the same time, the method discards ambiguous descriptive terms that are used 
in more classical works. 
 
With recent progress in modern technology, it is possible to develop an alternative way of 
deriving symmetry. The field of geometric morphometrics, which has been in existence for 
the last 20 years or so, can bridge the gap between the second and third dimensions of 
handaxe analysis using mathematical formulae to replicate aspects of shape (Grosman et al. 
2011). A few different studies exist: the Wynn and Tierson (1990) method involves creating 
22 polar co-ordinates originating from the midpoint of the long axes to the outer reaches of 
the object using a digitizer. Although deemed problematic their method basically aided in 
explaining typological variations of form (Underhill 2007). 
 
Brande & Saragusti (1996) also produced a computer program whereby 2-D linear landmark 
measurements could be converted into 3-D geometric models that better describe shape. 
However, Saragusti et al’s. algorithms are limited because they have been labelled as 
extremely mathematically complex and the software for their use is difficult to obtain 
(Cardillo 2010; Hardaker & Dunn 2005; (Underhill 2007). Brande & Saragusti’s (1996) 
method of alignment was the same as Roe’s alignment.  
 
Lycett et al. (2006) also make use of three-dimensional morphometrics to evaluate the shape 
of Pleistocene cores, combining their study with a biological phylogenetic approach. By 
using multivariate statistical analysis, they were then able to search for larger trends 
pertaining to asymmetry, amongst other variables (Cardillo 2010; Lycett et al. 2006). 
 
Archer & Braun (2010) draw on Lycett and his partners’ work by applying a geometric 
10 
 
morphometric method to the handaxes from the Acheulean site of Elandsfontein. In order to 
show that handaxe variability could be related to a specific reduction strategy based on raw 
material availability or selectivity, they used a 3D digitizer and produced various models 
(Archer & Braun 2010). 
 
Another example of a tool that can be used for quantitative study (although it is not a three-
dimensional one) is Hardaker & Dunn’s (2005) Flip Test. The test is a freely downloadable 
program that was designed primarily for work on handaxes. Using photographs and/or artistic 
sketches, they fold the tool over across its long axis. The overlapping pixel count can be 
calculated to show asymmetry using an index compiled by the authors (Hardaker & Dunn 
2005). The test is useful because it is easy to obtain and easy to apply, but conversely, it does 
exhibit some problems with the alignment of tools, which is something that Grosman et al. 
(2008) explain is problematic in recent lithic studies. More recently academics have been 
delving deeper into the use of three-dimensional modelling techniques as a way of advancing 
lithic studies. Three-dimensional studies are practically limitless in what they can be used for 
and can be applied to many existing artefact types, including stone, bone, ceramics, textiles, 
and even sites can be reconstructed in this way (Sumner & Riddle 2008). Three-dimensional 
modelling is useful because the complete surfaces of an object can be created and viewed 
with simplicity rather than working with digitised co-ordinate values which are numerical and 
can confuse the user during the reconstruction of the object in question (as in Saragusti et 
al’s. work). This section shall briefly describe a few instances whereby newer three 
dimensional techniques are being applied to lithic technologies. It will also discuss some of 
the advantages and disadvantages encapsulated therein.  
 
Sumner & Riddle (2008) used photogrammetry to recreate an exact 3-D replica of an 
Acheulean handaxe, as well as reconstruct the surfaces of refitted Middle Paleolithic cores 
from Egypt in 3-D to examine the core reduction strategy. The methods that the writers used 
in their study are quite similar to the ones presented in the next chapter, although their 
method is more affordable because all that is needed is a laptop, a digital camera and some 3-
D software.  In their photogrammetric study, the authors explain that each object was 
photographed in such a way that an overlap of images could be produced and later merged. 
This step differs significantly when compared to my method in terms of time frame, as 
Sumner & Riddle (2008) explain that the merge process is done manually and could take up 
to 10 hours for an experienced researcher. Turnover rate is one instance where three-
11 
 
dimensional laser scanning seems to have the upper hand for lithic studies.  
 
Regardless of the time implications, Sumner & Riddle (2008) were more focused on the 
advantages of using three-dimensional modelling in archaeology and their study certainly 
proves this in a number of ways. They suggest that the advantages of using 3-D is tenfold 
because models can be accessed anywhere in the world through the internet. They also 
explain that entire assemblages can literally be examined with the touch of a button, and that 
by doing so in this way, real assemblages can be better preserved or not subjected to handling 
damage (Sumner & Riddle 2008). The ability to study cross sections of an original specimen 
is one that should be considered invaluable to modern studies. 
 
 A second important output is that the 3-D programs used to model objects are equipped with 
applications that allow for the direct accurate calculation of a variety of different values. 
Three-dimensional programs such as Avizo 6 (used for this study) make it easy for 
researchers and even the general public to obtain information about an object in a highly 
manageable way. These programs offer tools that can provide the most basic metric 
calculations which traditional approaches struggle to capture, such as surface area, volume 
and centre of mass (Grosman et al. 2008; Sumner & Riddle 2008).  
 
Another recent study by Lin et al. (2010) used three-dimensional scanning to assess whether 
3D measures were more precise for cortex areas than mechanically averaged measurements. 
Their study showed that three-dimensional measures were more accurate when sample sizes 
were larger than those of individually measured specimens furthermore, cementing the fact 
that three-dimensional techniques are applied better in quantitative case studies. 
 
2.3 Three-dimensional scanning and the future 
 
Despite the multitude of practical uses of three-dimensional scanning, there are in fact 
problems within the discipline that need to be overcome through experimentation with 
different techniques. Grosman et al. (2008) suggest that because uneven and/or perhaps 
asymmetric objects offer different metric measures according to how they are aligned, it is 
clear that ambiguities and ‘errors’ will exist from study to study. They go on to further 
explain that: “the intrinsic impossibility to find a ‘true’ or ‘unique’ positioning is a 
mathematical fact that cannot be circumvented” (Grosman et al. 2008: 3102). 
12 
 
 
Grosman and her colleagues designed a computer-formulated algorithm based on complex 
mathematical design to apparently create an unambiguous alignment (or alignments) for 
artefacts, as well as to better document new measures available to 3-D scanning. This study 
in the greater scheme of things produces better information on the application of three-
dimensional modelling techniques to archaeology. It also shows that newer computer 
technologies can reduce human error significantly and rejuvenate traditional paradigms in a 
refreshed light (Grosman et al. 2008). 
 
Although our methodologies differ significantly, the overall concept of the above mentioned 
study to better understand the past through the future of technology is also the underlying 
feature of this study.  The method discussed in the next chapter is one that is extremely 
simple to grasp and apply to any handaxe, without having to wade through complex 
mathematical procedures. It is equally quite affordable and all the data can be easily 
transmitted upon request. It is my expectation that it will greatly add to the pool of 
knowledge for three-dimensional lithic studies. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
  
13 
 
CHAPTER 3: METHODOLOGY 
This section will explain the methodology and equipment I used in order to gauge symmetry 
for the handaxes in my study. There are two phases: the scanning and the editing of scans.  
As previously noted, the alignment or correct positioning of a given tool is not 
straightforward, especially in a three-dimensional analysis. My study seeks to provide 
researchers with a way of aligning handaxes that is based solely on the intrinsic dimensions 
of the tool being studied. Although the alignment is based on two dimensions, the measures it 
creates ultimately provide the most accurate three-edimensional data, such as volume and 
surface area values. The accuracy is greatly improved over the values obtained with more 
traditional measurements in 2D. This study did not use Roe’s (1964) measurement method 
(Figure 2) for studying handaxes, which divides the tool into eight sectors.  Instead I 
separated each tool into six sectors, each representing 16.65%, because I wanted to break 
away from traditional measures. In addition, McNabb et al’s. (2004) by-eye test had already 
been applied to one of my sites (Cave of Hearths) for determining symmetry based on 
sectoring a tool into six sectors. This study seeks to quantify their results in a more accurate 
fashion with 3D scanning.   
 
Figure 2: Conventional positioning and measurement extraction (from Roe, 1964 and after 
Grosman et al. 2008) 
 
14 
 
3.1 Scanning 
The equipment employed is a portable scanner called the Real Scan USB version 5.50.12.15 
COM, which is manufactured by 3D Digital Corp 2008 and was purchased by the University 
of the Witwatersrand at the end of December 2010. 
 
Figure 3: Depicting Real Scan USB scanner, table and scanning stand 
The scanner (shown in Figure 3) is easily assembled and a dimly lit room is more than 
adequate for the scanning process to take place. The scanner is fixed to an adjustable tripod 
and angled either facing the object straight on or at a downward angle. The scanning table 
was covered in black velvet to reduce light reflectivity and minimize background noise, 
because both of these can affect the outcome of the scan adversely. Tools were then fixed to a 
pottery wheel using either Prestik or black modelling clay (the base/circumference of the 
wheel was also covered in black cardboard). The wheel was then marked at every 45 degrees 
to ensure a maximum amount of overlap during the scanning process, as well as to provide 
the researcher with a point of reference when scanning. The scanner is then set up to face the 
tool (see Figure 3).  
The scan process integrates several pre-scanning exercises. The most important is the 
Preview Window, which allows the viewer to manipulate the actual scanner specifications 
and fine tune the limits of the scanner before the formal scanning begins. The extent of the 
laser can be manipulated in the X/Y and Z dimensions. This means that the laser can be set to 
scan certain parameters. For example, the maximum scan area from left to right is the X 
15 
 
value, the Y value is determined by the angle of the scanner on the tripod (which can be 
adjusted if an object to be reviewed is too large), and the Z value represents the maximum 
distance or depth of field of the laser. The intensity of the laser beam can also be set 
depending on the amount of detail that is wanted. For each scan in this study, I scanned all 
tools at the maximum resolution of 255 points and 200 lines per tool (which results in up to 
20,000 triangles of data created). One can do preview scans while manipulating the 
aforementioned variables, and once happy with the settings, a formal scan can be started.  
The tool is rotated and scanned every 45° until it has done a complete 360° circuit, ensuring 
that the rotation is uniform. The scans follow in order, which creates eight scan images, but 
more can be done depending on the amount of detail required.  The base of the tool needs to 
be scanned by holding the butt end straight into the laser and taking multiple scans, or by just 
laying it flat on the podium and turning it over a few times. The former method, however, 
generally produces much better scans. About 10 scans are needed to produce a full 3-D image 
of a handaxe.  
 
3.1.1 Cleaning and Merging 
The next step in the scan process involves the use of software that is designed to clean the 
images and merge each individual scan to the next one, using Slimwiz software. Each scan 
can be loaded into an erasing wizard whereby unwanted areas of a scan such as background 
noise can be manually removed simply and efficiently. This was particularly useful for 
cleaning up the butt end of a tool because it would often contain noise, such as my hands or 
the Prestik used to secure the tool to its platform. For the merging of the scans, scan files are 
loaded into the Slimwiz program and a reference image is selected as a starting point for each 
scan in the sequence to relate back to should the software experience difficulty during an 
alignment, normally the butt end is chosen because all scans should include some portion of 
that particular scan. Scans are successively merged with one another in the sequence 
beginning with the scan acquired subsequent to the reference image. The merging process 
then shows four windows (see Figure 4). The merge process works on a triangular point 
system whereby similar overlapping landmarks are placed on two adjacent images, each 
shown in a screen of its own (top left and right), and a final merge screen (bottom right) 
creates a best fit aspect, which the researcher can refer to when merging. The last window 
(bottom left panel) indicates the matching ratio between the points during merging; a lower 
16 
 
value means a closer fit between the scans. The software manual suggests that a value under 
5 is reasonable, but for this study I aimed for a value of under 1 (i.e., 0.1 or 0.2), although 
values such as 1.1 or 1.2 were also accepted. The programme allows you to check your 
merges and to make changes before saving your full 3D image in a multitude of formats (in 
this case: .Stl format). Figure 5 provides an example of a full merge with an incompatible 
surface that will need to be changed before it is saved). 
The merging process can be done manually or automatically, although manual alignment was 
preferable in this study. The scanning of tools can be done at a rate of about 15-20 tools per 
day for an experienced user, but the merging process minimally takes about an hour per tool. 
 
Figure 4: Manual alignment window 
 
17 
 
 
Figure 5: Full pre-merge scan showing incompatible surface 
 
3.2 Editing 
The next step in my methodology was the editing of scans to produce measures of surface 
area and volume for each sector of a given handaxe. This step draws on a more traditional 
method of measuring handaxes (Roe 1964) but it adds a third dimension using the Avizo 6 
programme. This programme allows the user to manipulate objects in 3-D space and is 
equipped with a multitude of actions useful to any researcher, including two dimensional 
measures. Avizo 6 provided me with easily extracted measures of volume and area that are 
otherwise impossible to calculate in the two-dimensional realm. The next section will explain 
how I did this. 
 3.2.1 Alignment and Measurements 
First, tools are opened and aligned using an xy button that flips the tool into a downward 
mode with the distal business-end facing downwards. Using the manual rotation wheel 
located in the lower right of the screen the tool must be aligned upright by 180 degrees. A 
bounding box is then placed around the tool automatically, and this box defines the maximum 
dimensions of the tool (maximum width, length, and thickness). This can be seen in Figure 6, 
showing the tool with bounding box represented as triangular mesh. 
18 
 
 
Figure 6: Aligned tool complete with bounding box 
A width measurement is done first using a 2D measure (equivalent to using calipers on a real 
tool). All measurements are reported in mm. Then a mid-point is located using a calculator to 
divide the tool width by two. From the mid-point a perpendicular angle is created to measure 
the maximum length, which extends to the top and bottom of the bounding box. 
Subsequently, maximum length is divided by three to create three proximal-distal arranged 
sectors that are further subdivided into six sectors by the midline maximum length 
measurement (see Figure 7). Each sector receives a landmark point to help align the cutting 
process in the next phase of editing (Figure 7). A thickness measurement is also taken for 
each tool with the help of the bounding box and recorded in MS Excel, traditionally this is 
important for calculating degree of refinement for any given tool and can be done in Avizo by 
simply turning the image sideways (using a shortcut key) and measuring the bounding box. 
These images are then saved as a network of files so the whole process can be opened at any 
time before or even during editing. 
19 
 
 
Figure 7: Handaxe scan complete with sectoral divisions, landmarks and bounding box 
 
3.2.2 Division of tools 
Next, tools are opened in the measures view. The image quality is refined using a transform 
editor, effectively making the number of triangles in the surface smaller (i.e., resampling) and 
increasing the file size. Resampling was performed three times for each tool to improve 
accuracy at cut surfaces.  Otherwise, during 3-D measurements, overlap would have been 
introduced for triangles of the mesh bordering a cut. Any triangles falling across a cutting line 
(i.e., introducing overlap) would introduce noise into volume/area measurements (see Figures 
6 and 7 for a comparison of an unrefined and refined tool). It is important to note that refining 
images in this way does not change area or volume measures for the tool, but serves the 
crucial purpose to protect sectored data from being erroneously skewed. 
Next I used a device called a Wacom tablet which is a touch screen stylus tool and a ruler 
(Figure 8) to divide tools into sectors. 
20 
 
 
Figure 8: Division of tools into sectors using a Wacom tablet 
 
 
 I started by producing images of  the left and right wholes respectively before dividing each 
side into a further three images namely the tip, medial and base for each side independently  
thus making up six sectors per tool. Each of the six sectors and the left and right wholes were 
saved individually and a measure of surface area and volume attached to each sector using a 
statistical tool incorporated within the Avizo package. All sectors were then saved to a 
collective network file, and the entire process of division could be opened at any time. 
21 
 
 
Figure 9: A complete tool divided into 6 sections 
3.3 Recording 
Tools were labelled according to part of tool and side (as it is in 3-D) thus: DS1a= Distal 
Sector 1(Left) a (front) and DS2a= Distal Sector 2 (right) a (front), whereas DS1/2b would 
represent the back of the tool depending on the view in the screen and how it was saved. All 
tools were saved as (a) which refers to images saved in the default forward view so if future 
research is done this paper should serve as a reference for interpretation. Other labels used 
were Left/Right Whole, DS1a/2a, MS1a/2a denotes Medial Sector and lastly PBS1a/2a, 
which denotes Proximal Basal Sector.* 
The editing of a tool measured in this way can take a minimum of about thirty minutes for an 
experienced researcher. 
 
 
*Note all tools were analyzed using this method, regardless of whether they were tipped or tipless, and the data for each tool 
and sector were transferred into an Excel spreadsheet (divided volume/area) so that it could be input by site into the 
statistical program SPSS for further analysis. Tipless tool values were processed in Excel but not included in this study 
because the data for the medial sector using this method would be taken as the tip value. Thus this method is limited only to 
tools that have tips. It is equally important to note that tools were randomly positioned upon initial scanning and tools were 
not scanned with their dorsal surfaces facing forward, for example. This is a minor limit to this research that should be 
rectified in future studies. 
22 
 
 
3.4 Evaluating Symmetry using Statistics 
The data presented in this paper was produced using the program IBM SPSS Statistics 19. 
The methodology used for this study is one that demonstrates considerable resolution when 
we consider the symmetry and standardization of any assemblage. In other words, an 
assemblage can be examined based on the values for whole left and right portions of the 
handaxe (measured using the longitudinal axis), as well as on a much finer scale by analysing 
the values for the distal, medial and basal sectors. My results are shown in terms of both area 
and volume. This is valuable because 2D techniques cannot capture the latter kind of data. 
For this study, volume seems to be more appropriate because it takes into account the entirety 
of the tool by including thickness, which area does not.  
 
 
 
 
 
 
 
 
 
 
 
 
 
23 
 
CHAPTER 4: MATERIALS 
4.1 Sample and Raw Materials 
For this study a total of 150 handaxes was available yet only 66 handaxes were analysed 
using the 3-D scanner. 
The Rietputs 15 sample was selected randomly and tool attributes were taken from an already 
existing database created by Kathleen Kuman. The Cave of Hearths handaxes were also 
selected randomly and raw materials were identified by geologist Professor Terrence 
McCarthy (University of the Witwatersrand, School of Geosciences). The blank types were 
classed by Kathleen Kuman in order to regularise this sample with the Rietputs 15 catalogue 
sample. 
This section will discuss the materials and blank types that were analysed for each site; 
Appendix B provides all the data collected during this study. 
 
4.1.1 Rietputs 15 Raw Materials 
For Rietputs, a total of 34 handaxes was analysed. The materials contained in this sample 
consist primarily of Ventersdorp Lava (n=14) and hornfels (n=16), comprising almost 90% of 
the total. Chert and quartzite make up the remaining ± 11.7%. These values are summarised 
in Figure 10 and Table 1. 
24 
 
 
Figure 10: Raw materials at Rietputs 15 
 
Blank Type Chert Hornfels Quartzite Ventersdorp-
Lava 
Total Percentage 
of  total 
Flake 
Incomplete 
0 5 1 14 20 58.8% 
Chunk 0 0 1 0 1 2.9% 
Cobble 1 6 0 0 7 20.5% 
Indeterminate 1 5 0 0 6 17.6% 
Total 2 16 2 14 34 100% 
Percentage of 
total 
5.8% 47% 5.8% 41.1% 100%  
 
Table 1: The relationship between Raw Material type and Blank Type for the Rietputs 15 site 
 
 
 
 
25 
 
4.1.2 Rietputs 15 Blank Types 
The blank types for Rietputs 15 handaxes consist of: chunks, cobbles, flakes and 
indeterminates. These labels were employed based on a previous database of Rietputs 15 in 
order to maintain regularity. The largest portion is flakes (n=20) comprising 58.82% in total 
although a moderate number of cobbles also exists (n=7 or 20.59%). These values are 
summarised below in Figure 11 and Table 1. 
 
 
Figure 11: Blank Types at Rietputs 15 
 
4.1.3 Cave of Hearths Raw Materials 
For the Cave of Hearths a total of 32 handaxes was examined. There were seven different raw 
materials in this sample as opposed to Rietput’s four; they are classed below in Figure 12 and 
Table 2. Quartzite dominates this sample (n=14) comprising 44% of the total. Dolerite (n=6) 
also makes up a significant portion of the sample, followed closely by quartz (n=5) and shale 
(n=3) which make up a further 40% of the total types. 
 
 
26 
 
 
Figure 12: Raw materials at the Cave of Hearths 
 
Blank Type Chert Dolerite Hornfels Rhyolite Shale Quartzite Quartz Total Percentage 
of  total 
Flake 
Incomplete 
0 0 0 0 0 4 0 4 12.5% 
End struck 
flake 
2 0 2 0 0 4 1 9 28.1% 
Side struck 
flake 
0 3 0 0 0 3 4 10 31.2% 
Corner struck 
flake 
0 0 0 0 0 1 0 1 3.1% 
Slab 0 0 0 0 1 0 0 1 3.1% 
Cobble 0 0 0 1 0 0 0 1 3.1% 
Indeterminate 0 3 1 0 0 2 0 6 18.7% 
Total 2 6 3 1 1 14 5 32 100% 
Percentage of  
total 
6.2% 18.7% 9.3% 3.1% 3.1% 43.7% 15.6% 100%  
 
Table 2: The relationship between Raw Material type and Blank Type for the Cave of Hearths site 
 
 
27 
 
4.1.4 Cave of Hearths Blank Types 
The blank types for the Cave of Hearths include: Corner Struck Flakes, End Struck Flakes, 
Side Struck Flakes, Indeterminate Flakes (flaking axis not visible), Indeterminate, Slabs, and 
Split Cobbles. The most dominant blank type is the End- Struck Flake (n=9) comprising 28% 
of the sample followed closely by Flakes and Indeterminates (n=8 for both), which make up 
25% of the assemblage each. The values are shown in Figure 13 and Table 2. Flaking axis for 
Cave of Hearths was determined for comparative purposes within the sub-types, which will 
be discussed in due course. 
 
 
Figure 13: Blank Types at the Cave of Hearths 
 
 
 
 
 
 
28 
 
CHAPTER 5: RESULTS 
5.1 General 
All data collected during this study can be referred to in the Appendices. Mean sector values 
reported in Tables 15 through 20 are given in percentages, which are easier to interpret than 
the large absolute values. For ease of reference, all figures and tables have been placed at the 
end of this chapter (pg 32-44). 
 
5.2 Results 
In Chapter 2, I stated that symmetry is thought to increase over time, and I aimed to test to 
what degree this is consistent with tools from my two sites of interest. In statistical terms, my 
null hypothesis (Ho) was that: the variables being tested (e.g., left and right wholes) are not 
associated. This implies little or no symmetry. My alternative hypothesis (HA) was that: 
variables being investigated (e.g., left and right wholes) are associated. This would imply that 
symmetry exists, and a correlation value would give some indication of how strong this 
association could be. The second objective was to determine what factors, if any, could 
influence symmetry.  
 
Figures 14 and 15, which are complementary to Tables 3 and 4, illustrate Rietputs 15 (early 
Acheulean) and Cave of Hearths (later Acheulean) volume plots, respectively. Figures 16 and 
17, which are complementary to Tables 5 and 6, illustrate the same, but in terms of area. Each 
scatter plot depicts the relationship between one side/sector and its counterpart, the line 
indicated in each graph is a reference line with equation: y = 1 * x + 0. Each point 
corresponds to a combination of left and right values for specific tools. The closer a point is 
to the reference line, the more symmetrical the values are (y = x). 
  
This relationship can be quantified mathematically by using the Pearson Correlation 
Coefficient (r).  It is shown in the tables below each graph. The coefficient is useful for 
showing the relationship between two variables: a correlation of 1 would represent a perfect 
association, whereas zero would represent that no association exists (Minium 1978). In order 
29 
 
to make sure the Pearson Correlation Coefficient was suitable for this study I tested my data 
using the Komogrov Smirnov (K-S) test to assess variable distributions versus normal 
distributions. I found no differences between them thus, leading me to use the Pearson 
Correlation as opposed to a non-parametric alternative  
  
According to Minium (1978: 143): ‘’the sign of the coefficient may be positive or negative. A 
positive value of r indicates that there is a high tendency for high values of one variable (X) 
to be associated with high values of the other variable (Y), and low values of one to be 
associated with low values of the other.’’ A negative value for r means that high values of 
one variable are associated with low values of the other, and vice versa. It is important to note 
that negative and positive values merely reflect differences in the direction of an association 
rather than the strength of association (Minium 1978). 
We would expect a graph showing a perfect r = 1 value to have points distributed along the 
line and no error resulting in points clustering off the line to a variable degree.  
The second important value when interpreting the tables given below is the ‘Sig. one/two 
tailed value’. This value is also known as the P value, and is a measure of significance. In 
standard scientific research, the general practise when performing statistical testing is to 
reject Ho when the significance value (p) is below a chosen value.  Typically this value is 
chosen to be equal to/less than 0.05 or 0.01.  When Ho is rejected, the alternative hypothesis 
(HA) is accepted. All statistical significance in this study is determined by using a significance 
value (P) of 0.05. 
 
5.3 Interpretation 
Figures 14 and 15 illustrate left and right side volumes of handaxes from Rieputs 15 and 
Cave of Hearths, respectively, Side volumes are highly associated on their own, and also 
individually at Rietputs 15 versus the Cave of Hearths. Their correlation coefficients are close 
to 1 with Rietputs at 0.859 and Cave of Hearths at 0.870 for side volume and 0.954 and 
0.960, respectively for side area (see Figures 16 and 17). Results for both sites are statistically 
significant, meaning Ho is rejected. For both side area and side volume, Cave of Hearths has a 
stronger correlation (which we would expect because it is later Acheulean).  Thus, as one side 
30 
 
increases so does its counterpart. Based on this, we could regard Cave of Hearth’s tools as 
more symmetrical than Rietputs (in terms of left and right). However, the difference in value 
between the two (i.e., strength of correlations) is extremely close, A larger sample size could 
be useful to further draw out these apparent differences. 
  
5.3.1 The Coefficient of Variation (CV) 
The Coefficient of Variation (CV) is statistically a normalized measure of dispersion that 
proved useful to this study with regards to analysing the sector data. The CV is defined as the 
ratio of the standard deviation to the mean: Cv = .  Larger CVs indicate more dispersion 
around the mean (i.e., lots of variability in the sample), while lower CVs indicate less 
dispersion around the mean (i.e., less variability in the sample). 
 
5.3.2 Interpretation 
See the Appendices for complete CV tables that show all variables measured. 
Tables 15 through 18 represent CV and means for each site. 
Means and CVs of linear measures (i.e. width, length and thickness) are very similar across 
assemblages. Cave of Hearths shows better control of shape based on the fact that these CV 
values are more consistent than those for Rietputs (around 0.21-22). The Cave of Hearths 
handaxes also have longer mean lengths than Rietputs. 
By examining mean values and CVs of area and volume for both sites, it is clear that left and 
right sides are very symmetrical in both cases, and that the differences in variability between 
them are minimal. However, when looking at sector data, substantial variability in the distal 
ends of both assemblages compared with the medial and proximal sectors of the tool is 
apparent. In the case of area values, the medial and proximal (base) sectors are less variable 
with almost perfect symmetry displayed (see Tables 15 to 18).  
It should be kept in mind that volumes and their CVs are generally larger than areas and their 
CVs because of the inclusion of thickness in the former (i.e., area scatter plots will appear 
more clustered because thickness has not been accounted for), but regarding our 
interpretation this is a negligible difference. It is also interesting to note that in terms of 
31 
 
volumes, a single edge for both assemblages displays the highest variation in all sectors of the 
tools, especially the tip. It is reasonable to assume from this data that the distal sector as well 
as one edge of the tool was the main focal point during manufacture and use. Unfortunately 
the first face of a tool to be scanned was randomly positioned, rather than having the dorsal 
face of the tool (if made on a flake blank) scanned first. This has made it impossible to 
designate specific sides of reference at this point in time. Future work will need to be 
employed to achieve this.  
The Cave of Hearths seems to exhibit a higher degree of variability for all sectoral divisions  
of a handaxe, but when we look at only left and right it shows a higher degree of symmetry 
between the two sides. This is probably due to continuous edge re-sharpening as a result of 
use, coupled with increased cognitive understanding leading to greater control of form over 
time (McNabb 2004). Rietputs 15 handaxes are generally quite robust and often pick-like, 
and it is probable that the tip was the primary emphasis (Kuman & Gibbon, in press). We can 
see this by looking at Tables 15 - 18 because the tip is more variable than the rest of the tool. 
However, Cave of Hearth’s handaxes are less pick-like and more emphasis was apparently 
placed on the entire edge of the tool and not only the tip. 
 
5.3.3 Blank Type and Symmetry 
With visual study of the Cave of Hearths handaxes, it was easy to distinguish two groups of 
blank types--end-struck and side-struck. When we compare means and CVs of these two 
groups (Tables 19 and 20), we can see that side-struck pieces seem to exhibit the least 
amount of variation in symmetry between left and right sides, as well as in linear 
dimensionality (length, width, thickness). In terms of sector analysis, both groups in the Cave 
of Hearths sample still exhibit similar variations, but overall side-struck pieces seem to 
produce more symmetric handaxes. The reason could be that the broad striking platforms of 
side-struck blanks allow more opportunity to remove thickness, while end-struck blanks are 
more compromised. 
 
 
 
32 
 
5.3.4 Raw Material and Symmetry 
Figures 20 and 21 show a more simplified left/right breakdown for both sites by their most 
dominant raw materials (i.e., dolerite and quartzite). The r and P values for material-specific 
side correlations are given in Tables 11 to 14. All comparisons have very strong correlations. 
 
5.4 Volume Scatterplots 
5.4.1 Rietputs 15 Volume 
 
Figure 14: Volume scatterplot of left and right sides for Rietputs 15 handaxes   
 
 
 
 
 
 
 
 
33 
 
Correlation between volumes of left and right sides of Rietputs 15 
handaxes 
  L Whole R Whole 
L Whole Pearson Correlation 1 .859** 
Sig. (2-tailed)  .000 
N 34 34 
R Whole Pearson Correlation .859** 1 
Sig. (2-tailed) .000  
N 34 34 
**. Correlation is significant at the 0.01 level (2-tailed). 
 
 
 
 
 
Table 3: Pearson Correlation for volumes of left and right sides of Rietputs 15 handaxes 
 
5.4.2 Cave of Hearths Volume 
 
Figure 15: Volume scatterplot of left and right sides for the Cave of Hearths handaxes 
 
 
34 
 
Correlation between volumes of left and right sides of the Cave 
of Hearths handaxes 
 L Whole R Whole 
L Whole Pearson Correlation 1 .870** 
Sig. (2-tailed)  .000 
N 32 32 
R Whole Pearson Correlation .870** 1 
Sig. (2-tailed) .000  
N 32 32 
**. Correlation is significant at the 0.01 level (2-tailed). 
 
 
 
Table 4: Pearson Correlation for volume of left and right sides of the Cave of Hearths 
handaxes 
5.5 Area Scatterplots 
 
5.5.1 Rietputs 15 Area 
 
 
Figure 16: Area scatterplot of left and right sides for Rietputs 15 handaxes 
35 
 
Correlation between areas of left and right sides of Rietputs 15 
handaxes 
 L Whole R Whole 
L Whole Pearson Correlation 1 .954** 
Sig. (2-tailed)  .000 
N 34 34 
R Whole Pearson Correlation .954** 1 
Sig. (2-tailed) .000  
N 34 34 
**. Correlation is significant at the 0.01 level (2-tailed). 
 
 
Table 5: Pearson Correlation for area of left and right sides of Rietputs 15 handaxes 
 
5.5.2 Cave of Hearths Area 
 
 
Figure 17: Area scatterplot of left and right sides for the Cave of Hearths handaxes 
 
 
 
 
36 
 
Correlation between areas of left and right sides of Cave of 
Hearths handaxes 
 L Whole R Whole 
L Whole Pearson Correlation 1 .960** 
Sig. (2-tailed)  .000 
N 32 32 
R Whole Pearson Correlation .960** 1 
Sig. (2-tailed) .000  
N 32 32 
**. Correlation is significant at the 0.01 level (2-tailed). 
 
 
Table 6: Pearson Correlation for area of left and right sides for Cave of Hearths handaxes 
 
5.6 Raw Material Scatterplots 
5.6.1 Rietputs 15  Area 
 
Figure 18: Area scatterplot of left and right sides of Rietputs 15 handaxes by dominant raw 
material 
37 
 
Correlation between area of left and right sides of Rietputs 15 
handaxes for hornfels 
 L Whole R Whole 
L Whole Pearson Correlation 1 .952** 
Sig. (2-tailed)  .000 
N 16 16 
R Whole Pearson Correlation .952** 1 
Sig. (2-tailed) .000  
N 16 16 
**. Correlation is significant at the 0.01 level (2-tailed). 
 
 
Table 7: Pearson correlation for area of left and right sides for Rietputs 15 handaxes for 
hornfels 
 
 
Correlation between area of left and right sides of Rietputs 15 
handaxes for Ventersdorp Lava 
 L Whole R Whole 
L Whole Pearson Correlation 1 .949** 
Sig. (2-tailed)  .000 
N 14 14 
R Whole Pearson Correlation .949** 1 
Sig. (2-tailed) .000  
N 14 14 
**. Correlation is significant at the 0.01 level (2-tailed). 
 
 
 
 
Table 8: Pearson correlation for area of left and right sides for Rietputs 15 handaxes for 
ventersdorp lava 
 
 
 
 
 
38 
 
5.6.2 Rietputs 15  Volume 
 
Figure 19: Volume scatterplot of right and left sides of Rietputs handaxes by dominant raw 
material 
 
 
Correlation between volumes of left and right sides of Rietputs 
15 handaxes for hornfels  
 L Whole R Whole 
L Whole Pearson Correlation 1 .833** 
Sig. (2-tailed)  .000 
N 16 16 
R Whole Pearson Correlation .833** 1 
Sig. (2-tailed) .000  
N 16 16 
**. Correlation is significant at the 0.01 level (2-tailed). 
 
 
Table 9: Pearson Correlation for volume of left and right sides for Rietputs 15 handaxes for 
hornfels 
 
 
39 
 
Correlation between volumes of left and right sides of Rietputs 
15 handaxes for ventersdorp lava 
 L Whole R Whole 
L Whole Pearson Correlation 1 .899** 
Sig. (2-tailed)  .000 
N 14 14 
R Whole Pearson Correlation .899** 1 
Sig. (2-tailed) .000  
N 14 14 
**. Correlation is significant at the 0.01 level (2-tailed). 
 
 
Table 10: Pearson Correlation for volume of left and right sides for Rietputs 15 handaxes for 
ventersdorp lava 
 
5.6.3 Cave of Hearths Area  
 
 
Figure 20: Area scatterplot of left and right sides of Cave of Hearths handaxes by dominant 
raw material 
 
40 
 
Correlation between area of left and right sides of Cave of 
Hearths handaxes for Dolerite 
 L Whole R Whole 
L Whole Pearson Correlation 1 .924** 
Sig. (2-tailed)  .008 
N 6 6 
R Whole Pearson Correlation .924** 1 
Sig. (2-tailed) .008  
N 6 6 
**. Correlation is significant at the 0.01 level (2-tailed). 
 
 
Table 11: Pearson Correlation for area of left and right sides for Cave of Hearths 15 handaxes 
for dolerite 
 
 
Correlation between area of left and right sides of Cave of 
Hearths handaxes for quartzite 
 L Whole R Whole 
L Whole Pearson Correlation 1 .963** 
Sig. (2-tailed)  .000 
N 14 14 
R Whole Pearson Correlation .963** 1 
Sig. (2-tailed) .000  
N 14 14 
**. Correlation is significant at the 0.01 level (2-tailed). 
 
 
 
Table 12: Pearson correlation for area of left and right sides for Cave of Hearths handaxes for 
quartzite 
 
 
 
 
 
41 
 
5.6.4 Cave of Hearths Volume 
 
Figure 21: Volume scatterplot of left and right sides of Cave of Hearths handaxes by 
dominant raw material 
 
 
Correlation between volume of left and right sides of Cave of 
Hearths handaxes for dolerite 
 L Whole R Whole 
L Whole Pearson Correlation 1 .972** 
Sig. (2-tailed)  .001 
N 6 6 
R Whole Pearson Correlation .972** 1 
Sig. (2-tailed) .001  
N 6 6 
**. Correlation is significant at the 0.01 level (2-tailed). 
 
 
Table 13: Pearson Correlation for volume of left and right sides for Cave of Hearths 15 
handaxes for dolerite 
 
 
42 
 
Correlation between volume of left and right sides of Cave of 
hearths handaxes for quartzite 
 L Whole R Whole 
L Whole Pearson Correlation 1 .921** 
Sig. (2-tailed)  .000 
N 14 14 
R Whole Pearson Correlation .921** 1 
Sig. (2-tailed) .000  
N 14 14 
**. Correlation is significant at the 0.01 level (2-tailed). 
 
 
 
Table 14: Pearson Correlation for volume of left and right sides for Cave of Hearths  
handaxes for quartzite 
 
5.7 Coefficient of Variation Tables 
5.7.1 Area 
 
 Max 
Width 
(cm) 
Max 
Length 
(cm) 
Max 
Thickness 
(cm) 
Left 
Side 
(mm2) 
Right 
Side 
(mm2) 
Distal 
Left 
(%) 
Distal 
Right 
(%) 
Medial 
Left 
(%) 
Medial 
Right 
(%) 
Proximal 
Left 
(%) 
Proximal 
Right 
(%) 
Mean 60.4 122.6 46.54 18787.1 19583.15 10.86 12.51 19.84 20.4 18.3 18.08 
Standard 
Deviation 
10.17 29.57 9.5 6801.44 6923.52 2.98 2.75 1.2 1.56 2.5 2.65 
CV 0.16 0.24 0.20 0.36 0.35 0.27 0.21 0.06 0.07 0.13 0.14 
 
Table 15: Coefficient of Variation for linear measurements and area in the Rietputs 15 
assemblage 
 
 
 
43 
 
 Max 
Width 
(cm) 
Max 
Length 
(cm) 
Max 
Thickness 
(cm) 
Left Side 
(mm2) 
Right Side 
(mm2) 
Distal 
Left 
(%) 
Distal 
Right 
(%) 
Medial 
Left 
(%) 
Medial 
Right 
(%) 
Proximal 
Left 
(%) 
Proximal 
Right 
(%) 
Mean 59.86 128.12 45.11 19639.89688 20971.74 9.3875 13.34 19.75 20.26 18.85 18.38 
Standard 
Deviation 
12.81 28.88 10.01 8084.37 7451.63 3.09 2.53 1.09 1.2 2.79 1.89 
CV 0.21 0.22 0.22 0.41 0.35 0.32 0.18 0.05 0.05 0.14 0.10 
 
Table 16: Coefficient of Variation for linear measurements and area in the Cave of Hearths 
assemblage 
 
5.7.2 Volume 
 
 Max 
Width 
(cm) 
Max 
Length 
(cm) 
Max 
Thickness 
(cm) 
Left Side 
(mm3) 
Right 
Side 
(mm3) 
Distal 
Left 
(%) 
Distal 
Right 
(%) 
Medial 
Left 
(%) 
Medial 
Right 
(%) 
Proximal 
Left (%) 
Proximal 
Right 
(%) 
Mean 60.4 122.6 46.54 71882.81 96004.14 8.21 12.31 13.42 19.17 21.51 25.36 
Standard 
Deviation 
10.17 29.57 9.5 38258.8 55102.64 3.97 4.34 3.74 4.13 6.16 5.53 
CV 0.16 0.24 0.20 0.53 0.57 0.48 0.35 0.27 0.21 0.28 0.21 
 
Table 17: Coefficient of Variation for linear measurements and volume in the Rietputs 15 
assemblage 
 
 Max 
Width 
(cm) 
Max 
Length 
(cm) 
Max 
Thickness 
(cm) 
Left Side 
(mm3) 
Right Side 
(mm3) 
Distal 
Left 
(%) 
Distal 
Right 
(%) 
Medial 
Left 
(%) 
Medial 
Right 
(%) 
Proximal 
Left (%) 
Proximal 
Right 
(%) 
Mean 59.86 128.12 45.11 79115 103155.97 7.54 12.35 14.43 18.9 21.98 24.81 
Standard 
Deviation 
12.81 28.88 10.01 51638.15 63143.22 4.28 4.46 5.34 5.62 7.07 6.27 
CV 0.21 0.22 0.22 0.65 0.61 0.56 0.36 0.37 0.29 0.32 0.25 
 
Table 18: Coefficient of Variation for linear measurements and volume in the Cave of 
Hearths assemblage 
44 
 
5.7.3 Cave of Hearths Symmetry and Blank Type. 
 
 Max 
Width 
(cm) 
Max 
Length 
(cm) 
Max 
Thickness 
(cm) 
Left Side 
(mm3) 
Right Side 
(mm3) 
Distal 
Left 
(%) 
Distal 
Right 
(%) 
Medial 
Left 
(%) 
Medial 
Right 
(%) 
Proximal 
Left (%) 
Proximal 
Right 
(%) 
Mean 65.8 129.2 46.2 89680.6 108934.7 8.1 10.8 12.9 17.8 24.9 25.2 
Standard 
Deviation 
17.6 29 10.4 78087.2 98225.3 4.1 4.1 4.9 5.2 7.1 6.9 
CV 0.26 0.22 0.22 0.87 0.90 0.50 0.37 0.37 0.29 0.28 0.27 
 
Table 19: Coefficient of Variation for linear measurements and volume of End Struck Pieces 
for the Cave of Hearths sub-sample (n=9) 
 
 Max 
Width 
(cm) 
Max 
Length 
(cm) 
Max 
Thickness 
(cm) 
Left Side 
(mm3) 
Right Side 
(mm3) 
Distal 
Left 
(%) 
Distal 
Right 
(%) 
Medial 
Left 
(%) 
Medial 
Right 
(%) 
Proximal 
Left (%) 
Proximal 
Right 
(%) 
Mean 56.7 134.8 44.8 84029.4 114785.5 8.23 13.6 14.8 21.1 17.7 24.4 
Standard 
Deviation 
10.4 21.3 7.5 49538.8 45794.3 4.7 4.3 5 5.2 4.8 7.3 
CV 0.18 0.15 0.16 0.58 0.39 0.57 0.31 0.33 0.24 0.27 0.29 
 
Table 20: Coefficient of Variation for linear measurements and volume of side-struck pieces 
for the Cave of Hearths sub-sample (n=10) 
 
 
 
 
 
45 
 
CHAPTER 6: DISCUSSION 
6.1 Discussion 
Using this method, symmetry can clearly be quantified in terms of the tool as a whole and 
within separate divisions. Both aspects are equally important when we attempt to interpret 
and compare the Cave of Hearths and Rietputs 15 handaxes. 
McNabb et al. (2004) explain that very little absolute symmetry exists for the Cave of 
Hearth’s Large Cutting Tools because many of the tools display such variable tips. They 
explain that symmetry only rarely occurred in isolated sections of the tools and concluded 
that because of the large amount of variation in the tools the presence of symmetry could be 
taken as either chance or could have been produced when a knapper was trying to produce a 
more elongated tip. Statistically, however, strong correlations in this study support the 
presence of very good symmetry in complete handaxes at the cave (r = 0.870 for volume, r 
=0.960 for area). This study was mathematically based, in contrast with the previous 
subjectively scrutinized method. Results of the present study do not suggest that McNabb et 
al’s. (2004) research is flawed, but simply that the three-dimensional quantification of data 
may lead to more accurate interpretations of a given assemblage. In this study it is shown that 
both sites display similar degrees of symmetry on a holistic level, which may seem surprising 
given the extensive time span between both sites. However, the fact that the values for the 
Early Acheulean sample from Rietputs 15 are also fairly high suggests that the convergent 
template of a handaxe, by its very nature, puts constraints on the bilateral dimensions for this 
special tool type. Overall for the Rietputs specimens, correlations were also strong: r = 0.859 
for volume, and r = 0.954 for area. If blank type is then considered, the lava specimens 
exhibit an even higher value (r = 0.899 v. 0.833 for hornfels). The explanation for this 
difference may be explained by blank type. Lava handaxes in this assemblage are always 
made on flake blanks, whereas hornfels handaxes are made on both cobbles and flakes and 
thus they reflect more inherent variability (K. Kuman, pers. comm.). 
 
The breakdown of these symmetries into sectors is equally important because it allows us to 
focus specifically on what areas of a tool were most significant to their manufacturers.  
46 
 
It is plainly shown that the distal sector in all cases of the aforementioned results is the most 
varied area of these handaxes (even when overall near-perfect left/right symmetry exists). 
This variability is also something that McNabb et al. (2004) stress quite highly in their paper, 
and their analysis of the tip forms is quite rigorous as a result. They also suggest that most 
tips at the Cave of Hearths are asymmetrical or ‘bent’ and that the purpose of this attribute is 
not easily understood. In light of this observation my study proposes that the use of sector 
analysis helps us recognise patterns relating to the whole tool rather than just one aspect. The 
results of this study show that there is a slight trend for the one side of all sectors to be more 
varied than the other, and in future it will be determined if this is consistent when specimens 
are scanned first for the dorsal face, which may tend to have more shaping scars. It could be 
suggested that handaxes at the Cave of Hearths were made in such a way that a unilateral use 
was being implemented, as opposed to an ambilateral use. The trend noted here also occurs 
for Rietputs 15, although the CV values for just the tips at both sites show that the variation is 
larger in the Cave of Hearths assemblage. This could be a functionally driven result. Schick 
& Toth (1995) and Domínguez-Rodrigo & de la Torre (2002) suggest that the edge of a 
handaxe would have been its most important feature for cutting, scraping or butchery; 
Posnansky (1959) also explains that if one side is more asymmetric then we would expect 
more weight to exist behind the cutting edge of the tool thus increasing its efficacy.  
Based on these results it is possible to couple the differences in morphology for both these 
sites with a number of factors. It was stated previously that Rietputs 15 handaxes are 
generally quite robust in terms of tip (Kuman and Gibbon, in press) possibly because they 
were being used for hacking or digging up roots (Kuman pers. comm.). This could account 
for the greater asymmetry in the distal region of the Cave of Hearths sample if this later 
Acheulean assemblage was more dedicated to cutting functions. Or it could simply be that the 
producers could not efficiently shape flatter tips at Rietputs 15 at the time. Both sites have 
handaxes that are made on fine-grained material so it is feasible that the level of technical 
control at the Cave of Hearths was overall greater and the degree of re-sharpening may have 
been of more importance when we consider what activities handaxes were necessary for at 
the cave. It is possible that the greater variation in tips at the Cave of Hearths was because an 
emphasis was being placed on the tips for scraping or skinning and/or cutting (Domínguez-
Rodrigo & de la Torre 2002, Schick & Toth 1995, Posnanksy 1959). In either case the 
difference seems to be functionally based.  
47 
 
By looking at the mean linear dimensions for the Cave of Hearths we can see that handaxes 
are only marginally longer than those at Rietputs 15 and are also thinner. Thinner handaxes 
may have improved cutting ability. This study could not address re-sharpening, but it should 
be considered in the future. 
Blank type also seems to have played an interesting role in handaxe symmetry at the Cave of 
Hearths. Side-struck pieces are thinner, longer and less variable in terms of total symmetry 
than the dominant blank type, end-struck flakes. Although the trend for one side to be more 
varied than its counterpart holds for both types of blank, it would seem that, overall, side-
struck pieces are the least variable and are more symmetric.  
 
6.2 The Question of Handedness.   
Posnansky (1959) explains that handedness is something that has been addressed in lithic 
studies but not much research has been done recently and previous studies were limited, 
isolated cases. In most circumstances modern human proxies are used to create the statistics 
that suggest whether a population is mainly left- or right-handed. This could be something to 
further investigate using the type of methodology presented in this paper and would be 
interesting when combined with a larger sample size than that which was used for this study. 
 
 
 
 
 
 
 
 
 
48 
 
CHAPTER 7: CONCLUSION 
This study has described a few interesting things about Rietputs 15 and Cave of Hearths 
handaxes. First, it has quantifiably measured the correlation between left and right sides for 
each site and shown a good measure of symmetry for each: for the Cave of Hearths, r = 0.870 
for volume, r =0.960 for area; for Rietputs, r = 0.859 for volume, and r = 0.954 for area.  
Secondly, using my methodology I was able to assess variability in tools sector by sector, 
finding that that variabililty was greatest in the tips for both sites. This permitted some 
inferences about function.  Thirdly, by looking at the raw material and blank data, I was able 
to speculate that side-struck tools were easier to make symmetrical than end-struck pieces. 
As our own minds are constantly working to try to make our lives easier by producing 
remarkable new technologies, it seems only suitable to use the technology available to us to 
try and decipher the true beginnings of man and his/her experiments with the earliest stone 
tools of the past. The importance and use of three-dimensional analyses to archaeological 
assemblages of the Acheulean was presented in this dissertation and has thus led to several 
interesting conclusions. The data presented by this paper was useful in producing quantitative 
measures of symmetry for both Rietputs 15 (early Acheulean) and the Cave of Hearths (later 
Acheulean), which were previously unaccounted for at these sites before now. The second 
major finding is that the tip is the most variable portion of the tools at both sites, which could 
be related to function in some way. A third primary conclusion that we could draw from this 
analysis is that blank type and raw material play some role in defining symmetry. 
 The benefits of using three-dimensional techniques for analysis can hardly be downplayed. 
3D scanning is a simple procedure and the data produced can be stored indefinitely quite 
literally in one’s pocket. Scans can be transmitted between research groups and across 
borders with the touch of a button and equally so; the more literature that is produced on 
three-dimensional analysis can only strengthen how it is used and understood. This study 
aimed to do just that, by formulating an entirely new way of analysing and aligning Large 
Cutting Tools, more specifically handaxes. The alignment of tools using this methodology is 
important because the computer program Avizo 6 works with the exact dimensions of the tool 
at hand in a way that manual alignment can take seconds thereafter.  The study described here 
minimizes user biases and lets the tool speak for itself so to speak. Although I realize that 
there are problems in how 3D studies are carried out, for example how tools are aligned, I 
recognize that these types of studies are only the foundations to a much larger framework of 
49 
 
information that is still to come. Thus, one should learn from previous mistakes in order to 
progress in the long-term.  
The method implemented here provides a greater resolution for studying the symmetry of a 
specific type of stone tool. The quantification of symmetrical measures combined with 
statistical analyses was useful for identifying trends within this study. It was shown using this 
methodology that tools should be scrutinized in much finer detail and that by looking at 
symmetry at the most minute levels can tell us a lot about how and why a tool would have 
been made in a particular way can be discovered. Although it is difficult to assume what early 
man was thinking when making these tools, it seems that the differences in symmetry of 
Acheulean tools for the sites Rietputs 15 and the Cave of Hearths could have been 
functionally driven.  
Perhaps the most limiting factor to this study was its sample size and the inability to apply the 
methodology to tipless tools (thus leading to their exclusion). This should be something for 
future researchers to keep in mind when applying or adapting the method proposed here. 
.  
 
 
 
 
 
 
 
 
50 
 
APPENDICES: CV TABLES SHOWING ALL DATA PRESENTED IN THIS STUDY 
APPENDIX A: All data, Rietputs 15 Area 
 
To
ol 
ID Site 
Area/Vol
ume 
Ma
x 
Wid
th 
(cm
) 
Max 
Leng
th 
(cm) 
Max 
Thick
ness 
(cm) 
Blan
k 
Type 
Raw 
Mate
rial 
Left 
Who
le 
Right 
Whol
e 
Total 
Tool 
value 
DS
1a 
(%
) 
DS
2a 
(%
) 
MS
1a 
(%) 
Ms
2a 
(%
) 
PBS
1a 
(%) 
PBS
2a 
(%) 
Perc
ent 
whol
e 
Compl
ete 
Distal 
(%) 
Compl
ete 
Media
l (%) 
Compl
ete 
Proxi
mal 
(%) 
Absol
ute 
DS1a 
(mm²) 
Absol
ute 
DS2a 
(mm²) 
Absol
ute 
MS1a 
(mm²) 
Absol
ute 
MS2a 
(mm²) 
Absol
ute 
PBS1
a 
(mm²) 
Absol
ute 
PBS2
a 
(mm²) 
Absol
ute 
Distal 
(mm²) 
Absol
ute 
Medi
al 
(mm²) 
Absol
ute 
Proxi
mal 
(mm²) 
 
R 
99 
Rietp
uts Area 63.8 
135.
6 38.7 Flake 
Hornf
els 
1931
3 22439 41752 11 
11.
3 20.3 
22.
5 15 19.9 100 27.1 40.7 32.2 
24719
.7 
3480.
7 
3215.
9 
5031.
9 
5032.
9 
3862.
6 
28200
.4 
8247.
83 
8895.4
3 
 
R 
10
2 
Rietp
uts Area 48.3 
115.
7 39.1 Indet Chert 
1634
3.1 
17571
.3 
33914
.4 
10.
2 
11.
4 20.4 
22.
8 17.6 17.7 100 23.5 42.2 34.2 37864 
3173.
7 
5738.
6 
7829.
6 
8155.
2 
5694.
1 
41037
.73 
13568
.15 
13849.
37 
 
R 
10
7 
Rietp
uts Area 57.7 
104.
6 48.6 
Cobb
le 
Hornf
els 
1680
6.8 
16206
.7 
33013
.5 8.6 
10.
8 18 
16.
1 24.2 22.1 100 20.7 39.3 40 
32856
.5 
2381.
3 
4416.
9 
6392.
2 
6529.
3 
6806.
7 
35237
.75 
10809
.18 
13336.
01 
 
R 
10
8 
Rietp
uts Area 77.5 
144.
4 53.2 Indet 
Hornf
els 
2592
7.4 
25381
.5 
51308
.9 
10.
8 
12.
8 23.4 
21.
4 16.3 15.3 100 24.2 42.3 33.5 
32412
.2 
3767.
5 
4065.
2 
7020.
5 
6693.
2 
5623.
6 
36179
.71 
11085
.68 
12316.
81 
 
R 
11
0 
Rietp
uts Area 45.9 27.8 35.1 
Cobb
le 
Hornf
els 
1058
3.5 
9569.
5 20153 
13.
9 6 20.4 
18.
9 18.2 22.6 100 21.5 43.2 35.3 
3451.
5 
3851.
6 
6924.
1 
7722.
4 
5967.
5 
5997.
3 
7303.
09 
14646
.5 
11964.
83 
 
R 
17
2 
Rietp
uts Area 80.1 
165.
5 64.9 Flake Qzite 
3078
4.7 
33069
.5 
63854
.2 11 
17.
3 20.5 
21.
6 16.7 12.9 100 19.5 34.2 46.4 
2847.
6 
3575.
6 
5957.
6 
5328.
8 
8001.
5 
7302.
2 
6423.
21 
11286
.46 
15303.
75 
 
R 
25
1 
Rietp
uts Area 61.9 
110.
6 60.4 
Chun
k Qzite 
2019
1.8 
18195
.9 
38387
.7 15 
11.
4 19.8 18 17.8 18 100 23.6 44.8 31.6 
5563.
4 
6546.
3 
11993
.5 
10990
.9 
8370.
6 
7844.
3 
12109
.69 
22984
.4 
16214.
82 
 
R 
26
2 
Rietp
uts Area 51.8 98.6 36.2 Flake 
Hornf
els 
1109
7.3 
11544
.9 
22642
.2 6.7 
10.
5 20.3 
22.
2 22 18.3 100 19.9 39.3 40.8 
2808.
6 
1200.
1 4103 
3809.
7 
3671.
8 
4559.
6 
4008.
72 
7912.
78 
8231.4
3 
 
R 
30
2 
Rietp
uts Area 63.4 
102.
1 50.5 
Cobb
le Chert 
1506
4.9 
16561
.1 31626 9.6 
11.
9 18.5 
21.
1 19.6 19.3 100 28.2 42.1 29.7 7002 
11030
.6 
13089
.1 
13798
.1 
10693
.6 
8240.
8 
18032
.62 
26887
.2 
18934.
41 
 
R 
32
8 
Rietp
uts Area 78.4 161 57.1 
Cobb
le 
Hornf
els 
2821
8.8 
31343
.2 59562 
12.
3 
13.
4 20.8 
21.
3 14.3 17.9 100 26.5 37.8 35.8 
5762.
8 
4394.
1 
7611.
5 
6895.
2 
6817.
5 
6906.
5 
10156
.91 
14506
.74 
13724.
08 
 
R 
37
5 
Rietp
uts Area 66.6 
126.
3 48 Flake 
V-
lava 
2155
6.5 
19700
.5 41257 
15.
1 9.2 19.6 
20.
2 17.5 18.3 100 17.2 42.5 40.3 
1518.
8 
2382.
5 
4590.
3 
5024.
8 
4988.
3 
4137.
6 
3901.
22 
9615.
13 
9125.8
9 
 
R 
37
8 
Rietp
uts Area 64.4 
156.
1 44.9 Flake 
V-
lava 
1891
2.2 
24236
.5 
43148
.7 5.2 
15.
9 20 22 18.7 18.3 100 21.5 39.6 38.9 
3035.
2 
3779.
2 
5837.
7 
6684.
4 6192 
6097.
6 
6814.
37 
12522
.03 
12289.
57 
 
R 
41
2 
Rietp
uts Area 56.1 
132.
1 45.1 Flake 
V-
lava 
1782
1.1 
18382
.3 
36203
.4 
16.
3 9.2 20.4 
22.
2 12.6 19.3 100 25.7 42.1 32.2 
7303.
8 
8004.
7 
12385
.3 
12667
.4 
8529.
7 
10671
.1 
15308
.47 
25052
.7 
19200.
82 
 
R 
46
6 
Rietp
uts Area 50 
113.
9 54.8 
Flake 
Th 
V-
lava 
1551
7.7 
19837
.1 
35354
.8 9.2 
14.
7 16.9 
20.
8 17.8 20.6 100 24.3 39.9 35.9 
6215.
3 
3792.
7 
8102.
4 
8350.
5 
7238.
9 
7557.
3 
10007
.94 
16452
.85 
14796.
23 
 
R 
46
9 
Rietp
uts Area 51.2 
100.
2 27.9 Flake 
Hornf
els 
1087
6.7 
12222
.1 
23098
.8 9.8 
12.
8 18.7 
18.
7 18.7 21.4 100 21.1 42 36.9 
2225.
9 
6873.
6 
8636.
7 
9482.
9 
8049.
6 7880 
9099.
5 
18119
.57 
15929.
61 
 
R 
50
6 
Rietp
uts Area 56.4 
126.
1 53.8 
Cobb
le-S 
Hornf
els 
1963
0.8 
18985
.3 
38616
.1 8.1 
16.
1 20.4 
19.
4 22.4 13.6 100 25.5 42.6 31.9 
5883.
5 
3342.
7 
7376.
5 
8043.
9 
4561.
2 
6995.
7 
9226.
14 
15420
.32 
11556.
96 
 
R 
53
9 
Rietp
uts Area 62.1 
123.
6 42.5 Flake 
V-
lava 
1974
1.1 
19329
.7 
39070
.8 
13.
5 
12.
1 19.3 
19.
9 17.7 17.5 100 23.9 37.8 38.3 
3251.
4 
5200.
3 5987 
7362.
9 
6279.
3 7274 
8451.
62 
13349
.83 
13553.
38 
 
R 
55
9 
Rietp
uts Area 52.2 
107.
4 43.5 Indet 
Hornf
els 
1392
8.7 
14559
.4 
28488
.1 
12.
6 
12.
7 18.5 
20.
6 17.8 17.9 100 22.5 37.4 40 
2252.
9 
2953.
8 
4315.
6 
4330.
7 
4308.
1 
4937.
6 
5206.
72 
8646.
31 9245.8 
 
R 
56
6 
Rietp
uts Area 49.6 
100.
3 38.5 
Flake 
Co 
Hornf
els 
1442
9.9 
12649
.9 
27079
.8 
15.
1 10 20.6 
18.
8 17.6 17.9 100 24.2 39.8 36 
3113.
8 6228 
7866.
2 
7507.
3 
8650.
9 5250 
9341.
83 
15373
.44 
13900.
86 
51 
 
 
R 
67
2 
Rietp
uts Area 60.5 
117.
9 47.9 Indet 
Hornf
els 
1700
8.9 
18277
.6 
35286
.5 
10.
5 
10.
9 19.8 
19.
5 18 21.4 100 25.6 39.2 35.3 
5271.
6 
4713.
3 
7536.
6 
7764.
3 
6932.
9 
6852.
1 
9984.
84 
15300
.94 
13784.
99 
 
R 
74
6 
Rietp
uts Area 50.8 87.4 43.8 Flake 
Hornf
els 
9717.
2 11149 
20866
.2 4.9 
13.
9 19.2 
19.
4 22.4 20.1 100 25.2 39.1 35.7 
3586.
3 
3604.
2 
5283.
1 5858 
5059.
4 
5097.
1 
7190.
48 
11141
.11 
10156.
51 
 
R 
77
0 
Rietp
uts Area 55.4 
114.
8 37.4 Flake 
V-
lava 
1400
5.4 
14610
.6 28616 8.8 
11.
7 21.7 
19.
4 18.5 19.9 100 25.1 39.3 35.6 
4091.
6 
2704.
1 
5566.
5 
5085.
5 
4771.
9 
4860.
3 
6795.
71 
10651
.95 
9632.1
8 
 
R 
77
2 
Rietp
uts Area 77.6 
168.
6 34.6 Flake 
V-
lava 
2889
8.5 
32769
.9 
61668
.4 6.6 
19.
8 19.8 
22.
5 20.5 10.8 100 21.4 39.2 39.3 
3701.
9 
3851.
5 
6969.
3 
6880.
3 
6337.
8 
7545.
8 
7553.
34 
13849
.64 
13883.
6 
 
R 
77
3 
Rietp
uts Area 52.4 
122.
8 57 Flake 
V-
lava 
1694
4.7 
17199
.6 
34144
.3 
13.
9 
14.
8 19.3 
21.
7 16.4 13.9 100 18.8 38.6 42.5 
1029.
7 
2896.
2 
4007.
1 
4055.
7 
4680.
4 
4197.
2 
3925.
83 
8062.
79 
8877.6
3 
 
R 
77
4 
Rietp
uts Area 62.1 
139.
6 42.8 Flake 
Hornf
els 
1977
7.4 
21222
.2 
40999
.6 
10.
6 
11.
3 19.9 
20.
1 17.8 20.3 100 20.5 41.1 38.4 
2507.
6 
3350.
3 
6203.
9 
5563.
7 
5293.
9 
5696.
6 
5857.
92 
11767
.58 
10990.
49 
 
R 
78
8 
Rietp
uts Area 65.2 
118.
9 41.2 
Flake 
Co 
Hornf
els 
2010
1.1 18335 
38436
.1 
10.
5 
10.
3 19.3 
19.
1 22.5 18.3 100 26.3 42.3 31.3 
4046.
8 
12198
.2 12213 
13883
.1 
12638
.7 
6688.
7 
16244
.96 
26096
.1 
19327.
43 
 
R 
79
3 
Rietp
uts Area 70.9 
150.
2 48.6 Flake 
V-
lava 
2737
6.2 
29769
.5 
57145
.7 9.1 
15.
4 20.8 
20.
2 18 16.5 100 28.7 41 30.3 
4753.
7 
5061.
2 
6587.
8 
7401.
8 
5603.
2 
4736.
6 
9814.
86 
13989
.66 
10339.
75 
 
R 
80
4 
Rietp
uts Area 83.6 
202.
2 71.7 Flake 
V-
lava 
4273
8.3 
38467
.1 
81205
.4 
14.
9 9.4 19 
17.
5 18.7 20.5 100 21.9 40 38.1 4339 
4643.
6 
8153.
4 
8256.
9 7285 
8321.
7 
8982.
56 
16410
.35 
15606.
68 
 
R 
85
2 
Rietp
uts Area 61.8 133 53.2 Flake 
V-
lava 
2129
0.5 
23005
.2 
44295
.7 
11.
6 
15.
2 18.3 
21.
2 18.2 15.6 100 20.8 38.3 40.9 4040 
3943.
5 7399 
7339.
9 
8662.
1 
7051.
6 
7983.
48 
14738
.89 
15713.
66 
 
R 
96
1 
Rietp
uts Area 52.1 96.9 53.8 Flake 
V-
lava 
1309
2.9 
12448
.4 
25541
.3 
12.
8 9.1 18.5 
22.
2 20 17.5 100 24.5 41 34.5 
5205.
4 
8800.
5 
11912
.5 
11536
.3 
10258
.3 
9432.
7 
14005
.88 
23448
.8 
19690.
96 
 
R 
1 
02
9 
Rietp
uts Area 47.7 
109.
2 44.7 Flake 
V-
lava 
1237
5.2 
12344
.5 
24719
.7 
14.
1 13 20.4 
20.
4 15.6 16.6 100 24.4 36.5 39.1 
12133
.3 
7671.
9 
15450
.3 
14171
.9 
15154
.7 
16623
.3 
19805
.2 
29622
.2 31778 
 
R 
1 
03
1 
Rietp
uts Area 66.5 
121.
2 31.4 Indet 
Hornf
els 
1669
7.5 
21166
.5 37864 8.4 
15.
2 20.7 
21.
5 15 19.2 100 26.7 39.5 33.8 
5131.
5 
6711.
7 
8112.
4 
9374.
9 
8046.
6 
6918.
5 
11843
.18 
17487
.31 
14965.
19 
 
R 
1 
05
2 
Rietp
uts Area 56.4 
112.
5 48.3 Indet 
Hornf
els 
1558
0.2 
17276
.3 
32856
.5 7.2 
13.
4 19.5 
19.
9 20.7 19.3 100 21.8 40.7 37.5 
3263.
5 2315 
4725.
6 
5668.
5 
5103.
8 
4464.
8 
5578.
49 
10394
.16 
9568.6
3 
 
R 
1 
06
0 
Rietp
uts Area 53.2 
121.
6 43.4 Flake 
V-
lava 
1641
1.6 
16000
.6 
32412
.2 
11.
6 
12.
5 21.7 
20.
7 17.4 16.2 100 22.3 42.8 34.9 
4572.
9 
4733.
2 
8471.
2 
9395.
3 
6268.
9 
8310.
5 
9306.
11 
17866
.46 
14579.
41 
                               
Mean 
   
60.4 
122.
6 46.54 
  
1878
7.1 
19583
.15 
38370
.26 
10.
86 
12.
51 
19.8
4 
20.
4 18.3 
18.0
8 
 
23.37 40.24 36.38 
7581.
28 
4798.
74 
7376.
61 
7838.
53 
7083.
19 
6777.
53 
12380
.01 
15215
.14 
13860.
74 
Std.Deviation 
   
10.1
7 
29.5
7 9.5 
  
6801.
44 
6923.
52 
13565
.41 
2.9
8 
2.7
5 1.2 
1.5
6 2.5 2.65 
 
2.76 2.17 3.77 
9398.
5 
2467.
96 
2991.
16 
2807.
35 
2430.
97 
2359.
96 
9390.
48 
5722.
45 
4505.4
2 
Co-efficient of 
variation 
   
0.16 0.24 0.2 
  
0.36 0.35 0.35 
0.2
7 
0.2
1 0.06 
0.0
7 0.13 0.14 
 
0.11 0.05 0.1 0.8 0.51 0.4 0.35 0.34 0.34 0.75 0.37 0.32 
 
 
 
 
52 
 
APPENDIX B: All data, Cave of Hearths area 
 
 
To
ol 
ID 
Si
te 
Area/
Volum
e 
M
ax 
Wi
dt
h 
(c
m) 
Ma
x 
Le
ngt
h 
(c
m) 
Max 
Thic
knes
s 
(cm) 
Bl
an
k 
Ty
pe 
Ra
w 
Mat
eria
l 
Left 
Whol
e 
Right 
Whol
e 
Total 
Tool 
value 
DS
1a 
(%
) 
DS2
a 
(%) 
MS
1a 
(%) 
Ms
2a 
(%) 
PB
S1
a 
(%
) 
PBS
2a 
(%) 
Per
cen
t 
wh
ole 
Com
plete 
Dist
al 
(%) 
Com
plete 
Med
ial 
(%) 
Com
plete 
Prox
imal 
(%) 
Abs
olut
e 
DS1
a 
(m
m²) 
Abs
olut
e 
DS2
a 
(m
m²) 
Abs
olut
e 
MS
1a 
(m
m²) 
Abs
olut
e 
MS
2a 
(m
m²) 
Abs
olut
e 
PBS
1a 
(m
m²) 
Abs
olut
e 
PBS
2a 
(m
m²) 
Abs
olut
e 
Dist
al 
(m
m²) 
Abs
olut
e 
Med
ial 
(m
m²) 
Abs
olut
e 
Prox
imal 
(mm
²) 
 
coh 
11 
C
o
h Area 
52.
7 147 40.6 
Es 
Fla
ke 
Qua
rtz 
17605
.7 
19426
.8 
37032
.5 8.7 13 20.3 
20.
9 
18.
5 18.6 100 21.8 41.1 37.1 
323
8.1 
482
3.5 
750
3.5 
772
8.6 
686
4.1 
687
4.7 
806
1.57 
152
32.1
2 
1373
8.82 
 
coh 
203 
C
o
h Area 
53.
1 
131
.3 37.2 
Es 
Fla
ke 
Qua
rtz 
14401
.1 
16653
.8 
31054
.9 6.2 13.2 20.1 
21.
9 20 18.5 100 19.4 42 38.5 
193
1 
410
7.7 
625
5.1 
680
0.1 
621
5 
574
6 
603
8.68 
130
55.2
2 
1196
1.02 
 
coh 
206 
C
o
h Area 
41.
3 
36.
2 67.7 
Ss 
Fla
ke 
Qzit
e 
16747
.3 
15150
.4 
31897
.7 
12.
1 8.9 20.1 
19.
2 
20.
3 19.4 100 20.9 39.3 39.8 
384
4.7 
282
9.6 
641
2.5 
612
3 
649
0 
619
7.8 
667
4.32 
125
35.4
5 
1268
7.86 
 
coh 
287 
C
o
h Area 
56.
8 
136
.4 50.4 
Cs 
Fla
ke 
Qzit
e 
20660
.7 
22341
.7 
43002
.4 
12.
6 15.3 21.1 
19.
7 
14.
4 16.9 100 27.9 40.8 31.3 
541
6.8 
657
9.9 
906
8.8 
848
0.2 
617
5 
728
1.5 
119
96.7
3 
175
49.0
7 
1345
6.53 
 
coh 
304 
C
o
h Area 
68.
9 
154
.1 59.9 
Ind
et 
Qzit
e 
27617
.7 
28919
.1 
56536
.8 12 14.2 18.7 20 
18.
1 17 100 26.2 38.7 35.1 
679
6.1 
802
5.7 
105
95.4 
112
81 
102
26.1 
961
2.4 
148
21.8
1 
218
76.4 
1983
8.49 
 
coh 
340 
C
o
h Area 
57.
4 
110
.6 42.2 
Es 
Fla
ke 
Qzit
e 
17308
.6 
15727
.9 
33036
.5 
11.
6 14.2 19.7 
18.
7 
21.
1 14.7 100 25.8 38.4 35.8 
382
7.6 
469
1.5 
651
9.4 
616
8 
696
1.6 
486
8.4 
851
9.06 
126
87.3
6 
1183
0.05 
 
coh 
365 
C
o
h Area 52 
137
.8 39.1 
Es 
Fla
ke 
Qzit
e 
18689
.5 
20791
.7 
39481
.2 9.3 16.4 19.2 
19.
4 
18.
8 16.9 100 25.7 38.5 35.8 
368
3.8 
646
6.9 
756
9.2 
764
5.6 
743
6.6 
667
9.2 
101
50.6
5 
152
14.7
6 
1411
5.83 
 
coh 
419 
C
o
h Area 
71.
6 
149
.6 49 
Ind
et 
Dol
erite 
25060
.4 
28467
.5 
53527
.9 
10.
1 16 19.9 
21.
1 
16.
8 16.1 100 26.1 41 32.9 
538
2.2 
857
5.6 
106
72.4 
112
88.8 
900
5.8 
860
3.2 
139
57.8
1 
219
61.2 
1760
8.91 
 
coh 
423 
C
o
h Area 
75.
4 
144
.6 53.6 
Fla
ke 
Qzit
e 
30748
.8 
29927
.7 
60676
.5 
14.
6 14.3 20.1 
18.
7 16 16.3 100 28.9 38.8 32.3 
886
5.5 
865
1.7 
121
90.7 
113
70 
969
2.6 
990
6 
175
17.2
2 
235
60.7 
1959
8.52 
 
coh 
440 
C
o
h Area 
68.
8 
119
.5 36.9 
Fla
ke 
Qzit
e 
18574
.4 
19700
.3 
38274
.7 8.9 11.3 19.6 
19.
4 20 20.7 100 20.2 39.1 40.7 
340
5.5 
433
4.7 
752
0 
743
4.9 
764
8.9 
793
0.8 
774
0.22 
149
54.8
2 
1557
9.68 
 
coh 
441 
C
o
h Area 58 
116
.7 32.2 
Ind
et 
Shal
e 
16191
.1 
17370
.3 
33561
.4 7.8 12.1 19.6 
20.
8 
20.
8 18.9 100 19.9 40.4 39.8 
261
9.5 
405
0.8 
658
3.8 
696
6.5 
698
7.9 
635
3 
667
0.27 
135
50.2
3 
1334
0.93 
 
coh 
444 
C
o
h Area 
50.
2 
124
.2 46.9 
Ind
et 
Dol
erite 15925 
17363
.7 
33288
.7 11 13.4 20.4 
19.
5 
16.
5 19.2 100 24.4 39.9 35.7 
365
1.1 
446
7.5 
678
1.9 
650
3.1 
549
2 
639
3.1 
811
8.64 
132
85.0
4 
1188
5.03 
 
coh 
445 
C
o
h Area 
60.
5 
139
.8 44.5 
Sp
lit 
Co 
Che
rt 
21302
.8 
22307
.7 
43610
.5 
11.
8 11.1 18.3 
18.
9 
18.
7 21.1 100 22.9 37.2 39.8 
516
1.1 
483
1.2 
797
9.6 
826
3.3 
816
2.1 
921
3.3 
999
2.32 
162
42.8
4 
1737
5.37 
 
coh 
447 
C
o
h Area 
50.
2 
118
.2 38.1 
Ind
et 
Qzit
e 
15715
.1 
14901
.8 
30616
.9 
13.
4 10.2 21 
19.
7 
16.
9 18.7 100 23.6 40.7 35.6 
411
3.7 
312
2 
642
5.4 
604
5 
517
5.9 
573
4.7 
723
5.77 
124
70.4
5 
1091
0.6 
 
coh 
450 
C
o
h Area 
65.
8 
80.
4 45.4 
Sla
b 
Qua
rtz 
13845
.1 
14325
.3 
28170
.4 7.1 9.4 16.8 
18.
9 
25.
2 22.6 100 16.6 35.6 47.8 
201
2.9 
265
1.9 
472
9 
531
2.5 
710
3.3 
636
0.9 
466
4.73 
100
41.5 
1346
4.21 
 
coh 
453 
C
o
h Area 
63.
5 
134
.6 55.6 
Ind
et 
Qzit
e 
16455
.1 23588 
40043
.1 8.7 14.2 21.9 
23.
9 
10.
5 20.8 100 22.9 45.8 31.3 
349
0.4 
569
1.9 
876
0.9 
956
6.2 
420
3.8 
832
9.8 
918
2.29 
183
27.1
7 
1253
3.63 
53 
 
 
coh 
454 
C
o
h Area 47 
103
.8 32.2 
Fla
ke 
Dol
erite 
10280
.8 
14405
.5 
24686
.3 1.8 17.8 17.4 
20.
6 
22.
5 19.9 100 19.6 38 42.4 
452.
3 
438
9.1 
428
4.3 
509
7.2 
554
4.2 
491
9.3 
484
1.41 
938
1.46 
1046
3.48 
 
coh 
455 
C
o
h Area 
65.
3 
155
.5 56.8 
Es 
Fla
ke 
Rhy
olite 
25645
.6 
27461
.3 
53106
.9 8.7 11.9 20.1 
20.
9 
19.
4 18.9 100 20.6 41 38.4 
462
2.1 
633
2.9 
106
97.4 
110
80.6 
103
26.1 
100
47.8 
109
55.0
2 
217
78 
2037
3.9 
 
coh 
458 
C
o
h Area 
57.
5 
118
.2 33.6 
Es 
Fla
ke 
Dol
erite 
15472
.1 17606 
33078
.1 3.5 14.4 21.1 
19.
8 
22.
2 19 100 17.9 40.9 41.2 
115
0.6 
477
3.8 
696
5 
654
7.8 
735
6.4 
628
4.4 
592
4.39 
135
12.8 
1364
0.85 
 
coh 
460 
C
o
h Area 
65.
6 135 46.2 
Es 
Fla
ke 
Dol
erite 
22704
.5 
21889
.2 
44593
.7 12 14.9 21.6 
19.
6 
17.
3 14.6 100 26.9 41.2 31.9 
535
2 
665
8.2 
963
8.5 
872
0.6 
771
4 
651
0.4 
120
10.2 
183
59.0
9 
1422
4.45 
 
coh 
462 
C
o
h Area 66 
125
.8 43.9 
Sla
b 
Dol
erite 
19082
.2 
23901
.9 
42984
.1 9.5 18.3 20.5 
20.
5 
14.
4 16.8 100 27.8 41 31.2 
408
2.9 
786
3.8 
881
5 
879
6 
618
4.2 
724
2.1 
119
46.7
5 
176
11.0
2 
1342
6.35 
 
coh 
463 
C
o
h Area 
62.
1 
130
.9 55.3 
Fla
ke 
Hor
nfel
s 
20887
.4 
20731
.1 
41618
.5 
10.
7 13 19.1 
21.
5 
20.
4 15.3 100 23.6 40.7 35.7 
443
3.3 
539
4.1 
796
1.8 
896
2.2 
849
2.3 
637
4.8 
982
7.39 
169
23.9
9 
1486
7.04 
 
coh 
480 
C
o
h Area 
37.
2 
91.
3 35.3 
Es 
Fla
ke 
Qzit
e 
7706.
7 
10780
.9 
18487
.6 4.8 17.6 19.6 
21.
6 
17.
3 19 100 22.4 41.2 36.3 
886.
9 
326
2.8 
361
9 
399
9.2 
320
0.8 
351
8.9 
414
9.69 
761
8.21 
6719
.68 
 
coh 
802
2 
C
o
h Area 
53.
2 
122
.8 35.7 
Fla
ke 
Qua
rtz 
15182
.7 
15816
.6 
30999
.3 8.7 13.9 19.6 
20.
9 
20.
6 16.2 100 22.6 40.6 36.8 
269
1.5 
430
9.2 
609
0.2 
648
9 
640
1 
501
8.3 
700
0.75 
125
79.2
4 
1141
9.31 
 
coh 
806
9 
C
o
h Area 
64.
9 
174
.5 50.4 
Ind
et 
Qzit
e 
22695
.9 
29608
.9 
52304
.8 3.4 14.4 20.1 
22.
9 20 19.4 100 17.7 42.9 39.3 
175
6.1 
751
6.1 
104
94 
119
61.8 
104
45.9 
101
30.9 
927
2.22 
224
55.8 
2057
6.8 
 
coh 
808
9 
C
o
h Area 
10
9.2 
184
.4 66.2 
Fla
ke 
Qzit
e 
53586
.1 
50014
.7 
10360
0.8 
12.
3 10.5 18.3 
19.
4 
21.
1 18.4 100 22.8 37.7 39.5 
127
91.8 
108
57.4 
189
30.8 
200
77.9 
218
63.6 
190
79.4 
236
49.2 
390
08.7 
4094
3 
 
coh 
900
1 
C
o
h Area 
68.
8 
154
.3 54.2 
Es 
Fla
ke 
Che
rt 
26400
.1 
25915
.7 
52315
.8 
10.
4 10.1 19.5 
20.
1 
20.
5 19.3 100 20.5 39.6 39.9 
542
6.1 
530
0.9 
102
27.7 
104
99.5 
107
46.3 
101
15.3 
107
26.9
6 
207
27.2 
2086
1.6 
 
coh 
e22 
C
o
h Area 
49.
9 
125
.1 33.6 
Ind
et 
fl 
Hor
nfel
s 
14077
.3 
16462
.4 
30539
.7 6 14.2 18.7 
20.
8 
21.
4 18.9 100 20.2 39.5 40.3 
182
2.8 
434
4.4 
571
7.9 
635
8.2 
653
6.6 
575
9.8 
616
7.13 
120
76.1
6 
1229
6.37 
 
coh 
kj1
3 
C
o
h Area 
42.
7 
87.
2 34.3 
Cs 
Fla
ke 
Qzit
e 
9130.
2 
10504
.7 
19634
.9 8.4 15.2 19.9 
20.
5 
18.
2 17.8 100 23.7 40.4 35.9 
165
8.5 
298
8.8 
390
5.2 
402
8.1 
356
6.5 
348
7.9 
464
7.24 
793
3.32 
7054
.38 
 
coh 
n17
- 
C
o
h Area 
57.
8 
169
.1 48.5 
Fla
ke 
Qua
rtz 
22067
.5 
21701
.2 
43768
.7 
10.
2 8.6 20.2 20 20 21 100 18.8 40.3 40.9 
447
4.4 
376
7.5 
885
8.8 
876
1 
873
4.4 
917
2.6 
824
1.87 
176
19.7
6 
1790
7.04 
 
coh 
pr2
4 
C
o
h Area 
64.
4 
124
.6 31.3 
Fla
ke 
Qzit
e 
19078
.1 
18921
.4 
37999
.5 
12.
4 11.4 19.7 
19.
7 18 18.7 100 23.8 39.4 36.8 
472
0.3 
432
1.8 
750
0.9 
747
5.4 
685
7 
712
4.3 
904
2.04 
149
76.2
1 
1398
1.31 
 
coh
461 
C
o
h Area 58 
116
.6 46.8 
Ind
et 
Hor
nfel
s 
17631
.1 
18410
.6 
36041
.7 
11.
7 13.5 19.9 
18.
9 
17.
3 18.7 100 25.3 38.7 36 
423
0.7 
488
1.2 
715
7.4 
679
5.6 
624
3 
673
3.9 
911
1.89 
139
53 
1297
6.91 
                               
Mean 
   
59.
86 
128
.12 
45.1
1 
  
19639
.8968
8 
20971
.7437
5 
40611
.6406
3 
9.3
87
5 
13.3
406
3 
19.7
531
3 
20.
262
5 
18.
85 
18.3
843
8 
 
22.7
3 
40.0
1 
37.2
4 
399
9.75 
533
9.5 
801
3.48 
820
7.09 
762
6.65 
742
5.15 
933
9.25 
162
20.5
7 
1505
1.81 
Std.De
viation 
   
12.
81 
28.
88 
10.0
1 
  
8084.
37 
7451.
63 
15378
.7 
3.0
9 2.53 1.09 1.2 
2.7
9 1.89 
 
3.2 1.82 3.7 
239
8.1 
194
3.09 
290
1.19 
304
0.04 
321
3.02 
280
7.28 
403
7.67 
591
5.89 
5909
.45 
Co-
efficie
nt of 
variati
on 
   
0.2
1 
0.2
2 0.22 
  
0.41 0.35 0.37 
0.3
2 0.18 0.05 
0.0
5 
0.1
4 0.1 
 
0.14 0.04 0.09 0.59 0.36 0.36 0.37 0.42 0.37 0.43 0.36 0.39 
 
54 
 
APPENDIX C: All data, Rietputs 15 Volume  
 
T
oo
l 
I
D Site 
Area/V
olume 
Ma
x 
Wi
dth 
(c
m) 
Ma
x 
Len
gth 
(cm
) 
Max 
Thic
kness 
(cm) 
Bla
nk 
Typ
e 
Raw 
Mat
erial 
Left 
Who
le 
Righ
t 
Who
le 
Total 
tool 
value 
DS
1a 
DS
2a 
M
S1
a 
M
S2
a 
PB
S1a 
PB
S2a 
To
tal 
(
%
) 
Com
plete 
Dista
l (%) 
Com
plete 
Medi
al 
(%) 
Com
plete 
Prox
imal 
(%) 
Abs
olute 
DS1
a 
Abs
olute 
DS2
a 
Abs
olute 
MS1
a 
Abs
olute 
MS2
a 
Abs
olute 
PBS
1a 
Abs
olute 
PBS
2a 
Abs
olute 
Dist
al 
Abs
olute 
Med
ial 
Abso
lute 
Prox
imal 
 
R 
99 
Riet
puts 
Volum
e 
63.
8 
135
.6 38.7 
Flak
e 
Horn
fels 
7758
7.7 
9496
5.5 
1725
53.2 
10.
5 9.8 
17.
9 
20.
3 
16.
5 25 
10
0 20.3 38.2 41.5 
1812
7.1 
1690
2.4 
3095
0.3 
3494
2.4 
2851
0.3 
4312
0.7 
3502
9.5 
6589
2.7 
7163
1 
 
R 
10
2 
Riet
puts 
Volum
e 
48.
3 
115
.7 39.1 
Inde
t 
Cher
t 
5758
2.4 
7478
1.9 
1323
64.3 7.7 
11.
3 
14.
6 
21.
4 
21.
2 
23.
8 
10
0 19 36 45 
1013
0.9 
1501
2.6 
1936
2.7 
2831
9.9 
2808
8.8 
3144
9.4 
2514
3.5 
4768
2.6 
5953
8.2 
 
R 
10
7 
Riet
puts 
Volum
e 
57.
7 
104
.6 48.6 
Cob
ble 
Horn
fels 
6909
0.2 
6645
1.7 
1355
41.9 4.7 7 
11.
7 9.3 
34.
6 
32.
7 
10
0 11.8 21 67.3 
6398
.1 
9536
.5 
1579
8.2 
1261
0.5 
4689
3.8 
4430
4.6 
1593
4.69 
2840
8.7 
9119
8.4 
 
R 
10
8 
Riet
puts 
Volum
e 
77.
5 
144
.4 53.2 
Inde
t 
Horn
fels 
1168
56 
1292
33 
2460
89 
13.
4 
17.
5 
20.
6 22 
13.
5 13 
10
0 30.9 42.6 26.5 
3293
7.1 
4303
6.4 
5067
5.2 
5417
2.9 
3324
3.6 
3202
3.9 
7597
3.5 
1048
48.1 
6526
7.5 
 
R 
11
0 
Riet
puts 
Volum
e 
45.
9 
27.
8 35.1 
Cob
ble 
Horn
fels 
3574
5 
3041
9.8 
6616
4.8 
10.
7 3.4 
18.
3 
12.
3 25 
30.
3 
10
0 14.2 30.5 55.3 7110 
2265
.7 
1207
9.5 
8122
.5 
1655
5.5 
2003
1.7 
9375
.65 
2020
1.96 
3658
7.2 
 
R 
17
2 
Riet
puts 
Volum
e 
80.
1 
165
.5 64.9 
Flak
e 
Qzit
e 
1069
90 
2089
20 
3159
10 5.6 
18.
1 9.8 
24.
8 
18.
5 
23.
2 
10
0 23.7 34.6 41.7 
1765
3.3 
5727
3.6 
3102
5.6 
7826
0.8 
5831
1.1 
7338
5.8 
7492
6.9 
1092
86.4 
1316
96.9 
 
R 
25
1 
Riet
puts 
Volum
e 
61.
9 
110
.6 60.4 
Chu
nk 
Qzit
e 
9399
6.5 
7697
4.4 
1709
70.9 
15.
7 
10.
9 
17.
4 14 
21.
9 
20.
1 
10
0 26.6 31.4 42 
2677
7.3 
1865
9 
2975
2 
2395
9.5 
3746
7.3 
3435
5.9 
4543
6.3 
5371
1.5 
7182
3.2 
 
R 
26
2 
Riet
puts 
Volum
e 
51.
8 
98.
6 36.2 
Flak
e 
Horn
fels 
3937
6.4 
3593
9.7 
7531
6.1 2.6 3.8 9.8 
12.
3 
39.
9 
31.
7 
10
0 6.4 22.1 71.6 
1968
.8 
2827
.8 
7372
.3 
9237
.9 
3003
5.3 
2387
4 
4796
.57 
1661
0.17 
5390
9.3 
 
R 
30
2 
Riet
puts 
Volum
e 
63.
4 
102
.1 50.5 
Cob
ble 
Cher
t 
5319
9.9 
7812
1.1 
1313
21 4.9 7.3 9.4 
20.
6 
26.
2 
31.
7 
10
0 12.2 30 57.8 
6478
.3 
9552
.4 
1235
3.8 
2700
2.6 
3436
7.8 
4156
6.1 
1603
0.73 
3935
6.4 
7593
3.9 
 
R 
32
8 
Riet
puts 
Volum
e 
78.
4 161 57.1 
Cob
ble 
Horn
fels 
1200
28 
2017
44 
3217
72 7.9 14 14 
20.
8 
15.
5 28 
10
0 21.8 34.8 43.4 
2531
5 
4490
8 
4495
7.7 
6689
1.4 
4975
5.5 
8994
4.2 
7022
3 
1118
49.1 
1396
99.7 
 
R 
37
5 
Riet
puts 
Volum
e 
66.
6 
126
.3 48 
Flak
e 
V-
lava 
7397
1.4 
9974
4.9 
1737
16.3 7.2 7.1 
12.
2 
20.
6 
23.
2 
29.
7 
10
0 14.3 32.8 52.9 
1243
0.9 
1241
1.5 
2123
3.7 
3575
5.9 
4030
6.8 
5157
7.5 
2484
2.4 
5698
9.6 
9188
4.3 
 
R 
37
8 
Riet
puts 
Volum
e 
64.
4 
156
.1 44.9 
Flak
e 
V-
lava 
7737
8.3 
9528
4.9 
1726
63.2 4.8 
15.
6 
17.
6 
17.
9 
22.
5 
21.
7 
10
0 20.4 35.4 44.1 
8285
.4 
2698
7.5 
3030
5.1 
3085
4.9 
3878
7.8 
3744
2.4 
3527
2.94 
6116
0 
7623
0.2 
 
R 
41
2 
Riet
puts 
Volum
e 
56.
1 
132
.1 45.1 
Flak
e 
V-
lava 
6592
0 
7274
6.5 
1386
66.5 18 
11.
6 
17.
4 
21.
2 
12.
2 
19.
7 
10
0 29.6 38.5 31.9 
2493
7.8 
1609
5.2 
2410
7.9 
2932
8.1 
1687
4.3 
2732
3.2 
4103
3 
5343
6 
4419
7.5 
 
R 
46
6 
Riet
puts 
Volum
e 50 
113
.9 54.8 
Flak
e Th 
V-
lava 
5925
3 
1028
85 
1621
38 5.8 
13.
3 
10.
6 
18.
9 
20.
1 
31.
2 
10
0 19.1 29.6 51.3 
9458
.3 
2158
3 
1722
9.2 
3072
0.8 
3256
5.5 
5058
0.9 
3104
1.29 
4795
0 
8314
6.4 
 
R 
46
9 
Riet
puts 
Volum
e 
51.
2 
100
.2 27.9 
Flak
e 
Horn
fels 
2202
0.1 
4100
6.5 
6302
6.6 4.3 
13.
6 8.2 
18.
1 
22.
4 
33.
4 
10
0 17.9 26.3 55.8 
2709
.1 
8578
.8 
5190
.9 
1140
5.5 
1412
0.2 
2102
2.2 
1128
7.89 
1659
6.36 
3514
2.4 
 
R 
50
6 
Riet
puts 
Volum
e 
56.
4 
126
.1 53.8 
Cob
ble-
S 
Horn
fels 
6893
7.2 
1031
40 
1720
77.2 4.4 
15.
6 
10.
9 
23.
1 
24.
8 
21.
3 
10
0 20 33.9 46.1 
7492
.6 
2688
4.7 
1873
7.5 
3967
7 
4270
7.1 
3657
8 
3437
7.27 
5841
4.5 
7928
5.1 
 
R 
53
9 
Riet
puts 
Volum
e 
62.
1 
123
.6 42.5 
Flak
e 
V-
lava 
9493
8 
6566
8 
1606
06 
13.
7 7.9 21 
11.
9 
24.
4 
21.
1 
10
0 21.6 32.9 45.5 
2208
2.4 
1267
2.4 
3373
6 
1904
7.8 
3911
9.6 
3394
7.8 
3475
4.8 
5278
3.8 
7306
7.4 
 
R 
55
9 
Riet
puts 
Volum
e 
52.
2 
107
.4 43.5 
Inde
t 
Horn
fels 
4400
6.6 
6109
4.9 
1051
01.5 
11.
3 
13.
9 
11.
5 
21.
5 
19.
1 
22.
8 
10
0 25.2 33 41.8 
1187
6.2 
1460
6.4 
1207
6.6 
2256
6.4 
2005
3.8 
2392
2.1 
2648
2.6 
3464
3 
4397
5.9 
55 
 
 
R 
56
6 
Riet
puts 
Volum
e 
49.
6 
100
.3 38.5 
Flak
e 
Co 
Horn
fels 
3565
7.3 
6131
7.2 
9697
4.5 9 
10.
2 
11.
9 
21.
2 
15.
9 
31.
8 
10
0 19.1 33.2 47.7 
8701
.5 
9859
.9 
1158
0.8 
2059
1.3 
1537
5.1 
3086
6 
1856
1.34 
3217
2.1 
4624
1.1 
 
R 
67
2 
Riet
puts 
Volum
e 
60.
5 
117
.9 47.9 
Inde
t 
Horn
fels 
5882
5.9 
7485
1.9 
1336
77.8 7.3 
11.
1 
14.
8 
17.
2 
21.
9 
27.
8 
10
0 18.4 31.9 49.7 
9777
.7 
1477
3.6 
1973
6.8 
2293
3.5 
2931
1.4 
3714
4.7 
2455
1.32 
4267
0.3 
6645
6.1 
 
R 
74
6 
Riet
puts 
Volum
e 
50.
8 
87.
4 43.8 
Flak
e 
Horn
fels 
2750
6.2 
3786
0.8 
6536
7 2.3 9.5 
10.
2 
18.
8 
29.
6 
29.
6 
10
0 11.8 29 59.2 
1485
.7 
6225
.2 
6644
.2 
1229
1.5 
1937
6.3 
1934
4.1 
7710
.87 
1893
5.68 
3872
0.4 
 
R 
77
0 
Riet
puts 
Volum
e 
55.
4 
114
.8 37.4 
Flak
e 
V-
lava 
3872
3 
6094
5.6 
9966
8.6 4.6 
10.
4 
12.
5 
17.
9 
21.
8 
32.
8 
10
0 15 30.4 54.6 
4589
.5 
1040
2 
1241
6.3 
1784
5.1 
2171
7.1 
3269
8.6 
1499
1.53 
3026
1.4 
5441
5.7 
 
R 
77
2 
Riet
puts 
Volum
e 
77.
6 
168
.6 34.6 
Flak
e 
V-
lava 
1574
93 
1794
23 
3369
16 5.2 
19.
4 
15.
7 
21.
3 
25.
8 
12.
5 
10
0 24.6 37 38.4 
1762
5 
6535
9.4 
5285
4.7 
7182
3.2 
8701
2.9 
4224
0.6 
8298
4.4 
1246
77.9 
1292
53.5 
 
R 
77
3 
Riet
puts 
Volum
e 
52.
4 
122
.8 57 
Flak
e 
V-
lava 
4454
7.2 
8038
1.6 
1249
28.8 
15.
8 
23.
6 
10.
1 
23.
1 9.7 
17.
7 
10
0 39.4 33.2 27.4 
1976
7.9 
2947
7 
1263
3.5 
2882
8.3 
1214
5.8 
2207
6.3 
4924
4.9 
4146
1.8 
3422
2.1 
 
R 
77
4 
Riet
puts 
Volum
e 
62.
1 
139
.6 42.8 
Flak
e 
Horn
fels 
6159
3.3 
1130
56 
1746
49.3 8 
16.
4 
14.
1 
21.
8 
13.
1 
26.
5 
10
0 24.5 35.9 39.6 
1402
5.4 
2870
9.8 
2461
6.5 
3813
5.4 
2295
1.5 
4621
0.4 
4273
5.2 
6275
1.9 
6916
1.9 
 
R 
78
8 
Riet
puts 
Volum
e 
65.
2 
118
.9 41.2 
Flak
e 
Co 
Horn
fels 
8228
2 
8205
5.3 
1643
37.3 9.6 11 
14.
8 
16.
5 
25.
7 
22.
4 
10
0 20.6 31.2 48.1 
1582
2.5 
1809
1.3 
2424
7.9 
2710
4.9 
4221
1.6 
3685
9 
3391
3.8 
5135
2.8 
7907
0.6 
 
R 
79
3 
Riet
puts 
Volum
e 
70.
9 
150
.2 48.6 
Flak
e 
V-
lava 
1068
60 
1972
78 
3041
38 3.6 
14.
5 12 
22.
1 
19.
5 
28.
3 
10
0 18.2 34.1 47.8 
1108
6.4 
4411
7.7 
3653
5.2 
6710
3.5 
5923
8.1 
8605
6.5 
5520
4.1 
1036
38.7 
1452
94.6 
 
R 
80
4 
Riet
puts 
Volum
e 
83.
6 
202
.2 71.7 
Flak
e 
V-
lava 
2024
96 
2652
68 
4677
64 
11.
8 
11.
6 
11.
6 
16.
4 
19.
8 
28.
6 
10
0 23.5 28.1 48.5 
5530
9.1 
5445
1.5 
5443
0.8 
7691
6.9 
9275
5.9 
1338
99 
1097
60.6 
1313
47.7 
2266
54.9 
 
R 
85
2 
Riet
puts 
Volum
e 
61.
8 133 53.2 
Flak
e 
V-
lava 
7758
6.1 
1397
17 
2173
03.1 8.4 
18.
1 9.8 
24.
4 
17.
5 
21.
8 
10
0 26.5 34.3 39.3 
1819
2.7 
3931
4.9 
2139
7.4 
5309
1.2 
3799
6 
4731
0.8 
5750
7.6 
7448
8.6 
8530
6.8 
 
R 
96
1 
Riet
puts 
Volum
e 
52.
1 
96.
9 53.8 
Flak
e 
V-
lava 
3965
2.9 
5976
9.8 
9942
2.7 
10.
8 
10.
1 10 26 
19.
1 24 
10
0 20.9 36 43.1 
1071
6.2 
1004
3.3 
9904
.7 
2589
7.8 
1903
2 
2382
8.7 
2075
9.5 
3580
2.46 
4286
0.7 
 
R 
1 
02
9 
Riet
puts 
Volum
e 
47.
7 
109
.2 44.7 
Flak
e 
V-
lava 
2587
8.2 
5449
1.6 
8036
9.8 7.4 
14.
1 7.4 
24.
1 
17.
4 
29.
6 
10
0 21.6 31.5 47 5980 
1134
7.2 
5941
.1 
1935
9.4 
1395
7.1 
2378
5.1 
1732
7.22 
2530
0.48 
3774
2.2 
 
R 
1 
03
1 
Riet
puts 
Volum
e 
66.
5 
121
.2 31.4 
Inde
t 
Horn
fels 
9398
5.3 
7045
1 
1644
36.3 8.5 
10.
1 21 
11.
9 
27.
6 
20.
9 
10
0 18.6 32.8 48.6 
1402
2.5 
1658
8 
3449
9.5 
1948
8.7 
4546
3.3 
3437
4.4 
3061
0.5 
5398
8.2 
7983
7.7 
 
R 
1 
05
2 
Riet
puts 
Volum
e 
56.
4 
112
.5 48.3 
Inde
t 
Horn
fels 
5495
0.6 
7662
2 
1315
72.6 3.8 
12.
3 
11.
4 
18.
6 
26.
6 
27.
3 
10
0 16.1 30 53.9 
4978
.6 
1623
4 
1503
0.8 
2443
6.7 
3494
1.2 
3595
1.2 
2121
2.57 
3946
7.5 
7089
2.4 
 
R 
1 
06
0 
Riet
puts 
Volum
e 
53.
2 
121
.6 43.4 
Flak
e 
V-
lava 
5910
1.9 
7153
0.3 
1306
32.2 10 
14.
5 
16.
1 
19.
8 
19.
1 
20.
4 
10
0 24.5 35.9 39.6 
1305
5.2 
1899
6.6 
2104
2.7 
2584
4.4 
2500
4 
2668
9.3 
3205
1.8 
4688
7.1 
5169
3.3 
                               
Mean 
   
60.
4 
122
.6 46.54 
  
7188
2.81 
9600
4.14 
1678
86.95 
8.2
1 
12.
31 
13.
42 
19.
17 
21.
51 
25.
36 
 
20.53 32.59 46.88 
1403
8.36 
2217
0.15 
2295
4.62 
3278
1.42 
3488
9.8 
4105
2.5 
3620
8.52 
5573
6.04 
7594
2.3 
Std.De
viation 
   
10.
17 
29.
57 9.5 
  
3825
8.8 
5510
2.64 
9013
2.44 
3.9
7 
4.3
4 
3.7
4 
4.1
3 
6.1
6 
5.5
3 
 
6.25 4.3 9.58 
1051
3.37 
1615
3.27 
1346
5.21 
1967
1.82 
1873
3.73 
2348
2.6 
2404
5.91 
3127
8.87 
3974
6.34 
Co-efficient of 
variation 
  
0.1
6 
0.2
4 0.2 
  
0.53 0.57 0.53 
0.4
8 
0.3
5 
0.2
7 
0.2
1 
0.2
8 
0.2
1 
 
0.3 0.13 0.2 0.74 0.72 0.58 0.6 0.53 0.57 0.66 0.56 0.52 
 
56 
 
APPENDIX D: All data, Cave of Hearths Volume 
 
To
ol 
ID 
Si
te 
Area/V
olume 
Ma
x 
Wi
dth 
(c
m) 
Ma
x 
Len
gth 
(cm
) 
Max 
Thic
kness 
(cm) 
Bl
an
k 
Ty
pe 
Raw 
Mat
erial 
Left 
Who
le 
Right 
Whol
e 
Total 
tool 
value 
DS
1a 
DS
2a 
M
S1
a 
M
S2
a 
PB
S1a 
PB
S2a 
To
tal 
(
%
) 
Com
plete 
Dista
l (%) 
Com
plete 
Medi
al 
(%) 
Com
plete 
Prox
imal 
(%) 
Abs
olute 
DS1
a 
Abs
olute 
DS2
a 
Abs
olute 
MS1
a 
Abs
olute 
MS2
a 
Abs
olute 
PBS
1a 
Abs
olute 
PBS
2a 
Abs
olute 
Dist
al 
Abs
olute 
Med
ial 
Abso
lute 
Proxi
mal 
 
coh 
11 
C
o
h Volume 
52.
7 147 40.6 
Ind
et 
fl 
Quar
tz 
5535
7.4 
8545
1.5 
1408
08.9 5.4 
13.
4 
12.
5 
20.
9 
21.
5 
26.
4 
10
0 18.8 33.4 47.8 7581 
1892
5.7 
1755
6.9 
2942
2.6 
3021
9.6 
3710
3.2 
2650
6.65 
4697
9.5 
6732
2.8 
 
coh 
203 
C
o
h Volume 
53.
1 
131
.3 37.2 
Fla
ke 
Quar
tz 
2738
4.3 
8009
7.2 
1074
81.5 3.1 
16.
5 8.3 
31.
8 
14.
2 
26.
2 
10
0 19.5 40.1 40.4 
3279
.9 
1773
1.4 
8879
.1 
3420
7.6 
1522
5.3 
2815
8.2 
2101
1.27 
4308
6.73 
4338
3.5 
 
coh 
206 
C
o
h Volume 
41.
3 
36.
2 67.7 
Fla
ke 
Qzit
e 
4785
6.1 
6576
4.8 
1136
20.9 8.6 9.9 
11.
5 
18.
8 
22.
1 
29.
2 
10
0 18.5 30.2 51.2 
9748
.7 
1130
0.5 
1302
5.3 
2132
3.8 
2508
2.1 
3314
0.5 
2104
9.24 
3434
9.1 
5822
2.6 
 
coh 
287 
C
o
h Volume 
56.
8 
136
.4 50.4 
Fla
ke 
Qzit
e 
8961
0.8 
1056
72 
1952
82.8 
17.
2 
20.
3 
18.
9 
19.
1 9.8 
14.
7 
10
0 37.5 37.9 24.5 
3357
2.5 
3970
2.7 
3687
7.9 
3721
5.2 
1916
0.4 
2875
3.7 
7327
5.2 
7409
3.1 
4791
4.1 
 
coh 
304 
C
o
h Volume 
68.
9 
154
.1 59.9 
Cs 
Fla
ke 
Qzit
e 
1305
63 
1637
35 
2942
98 
10.
4 
13.
8 
13.
3 
19.
5 
20.
6 
22.
4 
10
0 24.2 32.8 43 
3061
9.1 
4069
6.6 
3924
3.1 
5724
3.4 
6070
0.5 
6579
5.5 
7131
5.7 
9648
6.5 
1264
96 
 
coh 
340 
C
o
h Volume 
57.
4 
110
.6 42.2 
Es 
Fla
ke 
Qzit
e 
7557
4.2 
3157
1.7 
1071
45.9 
17.
7 9.5 
25.
4 
10.
8 
27.
5 9.2 
10
0 27.2 36.1 36.7 
1896
4.6 
1015
5.6 
2717
6.6 
1154
2.5 
2943
3 
9873
.5 
2912
0.2 
3871
9.1 
3930
6.52 
 
coh 
365 
C
o
h Volume 52 
137
.8 39.1 
Fla
ke 
Qzit
e 
6371
7.3 
8730
0.4 
1510
17.7 8 
20.
3 
15.
1 
19.
1 
19.
1 
18.
4 
10
0 28.3 34.2 37.5 
1205
3.1 
3064
4 
2281
7.7 
2886
7.9 
2884
6.6 
2778
8.5 
4269
7.1 
5168
5.6 
5663
5.1 
 
coh 
419 
C
o
h Volume 
71.
6 
149
.6 49 
Ind
et 
Dole
rite 
9587
0.7 
1566
20 
2524
90.7 6 
14.
9 
12.
1 
22.
9 20 
24.
2 
10
0 20.8 35 44.2 
1505
5.6 
3755
2.9 
3043
6.4 
5786
3.9 
5037
8.7 
6120
3.2 
5260
8.5 
8830
0.3 
1115
81.9 
 
coh 
423 
C
o
h Volume 
75.
4 
144
.6 53.6 
Fla
ke 
Qzit
e 
1856
46 
1660
33 
3516
79 
14.
7 
12.
1 
18.
5 
14.
7 
19.
6 
20.
4 
10
0 26.8 33.3 39.9 
5154
8.1 
4271
8.5 
6523
2.6 
5172
8.1 
6886
5.3 
7158
6.8 
9426
6.6 
1169
60.7 
1404
52.09 
 
coh 
440 
C
o
h Volume 
68.
8 
119
.5 36.9 
Es 
Fla
ke 
Qzit
e 
5476
9.1 
6693
5.9 
1217
05 5.6 8.2 
10.
7 
15.
6 
28.
8 
31.
3 
10
0 13.7 26.2 60 
6786
.3 
9923
.8 
1297
3.5 
1896
3.8 
3500
9.3 
3804
8.3 
1671
0.13 
3193
7.3 
7305
7.6 
 
coh 
441 
C
o
h Volume 58 
116
.7 32.2 
Sla
b 
Shal
e 
5164
1.3 
6768
7.9 
1193
29.2 6.2 
10.
1 
14.
2 
21.
4 
22.
9 
25.
2 
10
0 16.3 35.6 48.1 
7434
.7 
1207
3 
1693
7.7 
2553
5.5 
2726
8.9 
3007
9.3 
1950
7.7 
4247
3.2 
5734
8.2 
 
coh 
444 
C
o
h Volume 
50.
2 
124
.2 46.9 
Ind
et 
Dole
rite 
5343
5.4 
7742
7.3 
1308
62.7 6.7 
11.
3 
13.
5 
15.
5 
20.
6 
32.
4 
10
0 18 29 53 
8767
.5 
1477
4.9 
1770
5.6 
2024
4.5 
2696
2.3 
4240
7.9 
2354
2.38 
3795
0.1 
6937
0.2 
 
coh 
445 
C
o
h Volume 
60.
5 
139
.8 44.5 
Es 
Fla
ke 
Cher
t 
7158
7 
1045
61 
1761
48 
11.
5 
14.
3 12 
19.
8 
17.
1 
25.
2 
10
0 25.9 31.8 42.3 
2033
9.9 
2527
0.1 
2118
3.7 
3487
4 
3006
3.4 
4441
7.3 
4561
0 
5605
7.7 
7448
0.7 
 
coh 
447 
C
o
h Volume 
50.
2 
118
.2 38.1 
Es 
Fla
ke 
Qzit
e 
4327
8.7 
7341
7.8 
1166
96.5 8.1 
11.
5 
12.
1 
21.
2 
16.
9 
30.
3 
10
0 19.6 33.3 47.2 
9414
.6 
1341
3.1 
1415
3.9 
2468
9.2 
1971
0.2 
3531
5.5 
2282
7.72 
3884
3.1 
5502
5.7 
 
coh 
450 
C
o
h Volume 
65.
8 
80.
4 45.4 
Es 
Fla
ke 
Quar
tz 
5711
7.2 
4948
7.8 
1066
05 4.4 4.9 
10.
3 
11.
9 
38.
9 
29.
7 
10
0 9.3 22.1 68.6 
4681
.5 
5238
.5 
1092
7.5 
1264
0.8 
4150
8.2 
3160
8.5 
9919
.98 
2356
8.3 
7311
6.7 
 
coh 
453 
C
o
h Volume 
63.
5 
134
.6 55.6 
Fla
ke 
Qzit
e 
9368
3.5 
8288
6.6 
1765
70.1 14 
14.
3 
28.
1 
16.
8 11 
15.
8 
10
0 28.3 44.8 26.8 
2468
4.4 
2533
3.2 
4952
8.3 
2961
7.4 
1947
0.7 
2793
6.1 
5001
7.6 
7914
5.7 
4740
6.8 
 
coh 
454 
C
o
h Volume 47 
103
.8 32.2 
Ind
et 
Dole
rite 
3278
6.3 
4500
5.7 
7779
2 0.7 
14.
1 
11.
1 
16.
2 
30.
3 
27.
6 
10
0 14.7 27.3 57.9 
527.
2 
1094
4.8 
8655
.1 
1261
8.7 
2360
4 
2144
2.1 
1147
1.96 
2127
3.83 
4504
6.1 
 
coh 
455 
C
o
h Volume 
65.
3 
155
.5 56.8 
Sla
b 
Dole
rite 
1092
34 
1503
46 
2595
80 5 
10.
4 
13.
7 
19.
7 
23.
4 
27.
9 
10
0 15.4 33.3 51.3 
1301
8.1 
2695
6.9 
3546
7.4 
5107
7.2 
6074
8.2 
7231
2.3 
3997
5 
8654
4.6 
1330
60.5 
57 
 
 
coh 
458 
C
o
h Volume 
57.
5 
118
.2 33.6 
Ind
et 
Dole
rite 
5738
4 
6717
0.7 
1245
54.7 1.9 
13.
2 17 
17.
5 
27.
1 
23.
2 
10
0 15.1 34.5 50.3 
2426
.4 
1641
7.7 
2118
0.8 
2181
9 
3377
6.9 
2893
4 
1884
4.05 
4299
9.8 
6271
0.9 
 
coh 
460 
C
o
h Volume 
65.
6 135 46.2 
Spl
it 
Co 
Igne
ous 
7953
2 
1059
15 
1854
47 9 
15.
6 
14.
4 
20.
4 
19.
5 
21.
2 
10
0 24.6 34.8 40.6 
1668
3.1 
2891
8.3 
2672
4 
3774
0.1 
3612
4.9 
3925
6.9 
4560
1.4 
6446
4.1 
7538
1.8 
 
coh 
462 
C
o
h Volume 66 
125
.8 43.9 
Ind
et 
Dole
rite 
7963
5.4 
1142
22 
1938
57.4 
10.
2 
20.
5 
17.
9 
20.
6 13 
17.
9 
10
0 30.7 38.5 30.8 
1979
3.2 
3973
8.3 
3470
3.5 
3984
0.9 
2513
8.6 
3464
2.7 
5953
1.5 
7454
4.4 
5978
1.3 
 
coh 
463 
C
o
h Volume 
62.
1 
130
.9 55.3 
Es 
Fla
ke 
Horn
fels 
5940
4.9 
1164
75 
1758
79.9 5.9 
16.
7 8.5 
26.
9 
19.
4 
22.
6 
10
0 22.6 35.4 42 
1033
3.7 
2936
8 
1490
7.5 
4732
4.2 
3416
3.7 
3978
2.4 
3970
1.7 
6223
1.7 
7394
6.1 
 
coh 
480 
C
o
h Volume 
37.
2 
91.
3 35.3 
Fla
ke 
Qzit
e 
1432
93 
1987
60 
3420
53 5.9 7.4 
21.
9 
24.
9 
14.
1 
25.
9 
10
0 13.3 46.8 40 
2026
6.6 
2517
2.7 
7482
3.9 
8508
9.1 
4820
2.3 
8849
7.9 
4543
9.3 
1599
13 
1367
00.2 
 
coh 
802
2 
C
o
h Volume 
53.
2 
122
.8 35.7 
Fla
ke 
Quar
tz 
5335
7.3 
4883
1.6 
1021
88.9 
10.
6 
14.
7 
17.
6 
17.
8 
24.
1 
15.
4 
10
0 25.3 35.3 39.4 
1083
5.8 
1499
5.2 
1794
3.3 
1814
0.4 
2457
8.1 
1569
6 
2583
1 
3608
3.7 
4027
4.1 
 
coh 
806
9 
C
o
h Volume 
64.
9 
174
.5 50.4 
Ind
et 
Qzit
e 
6822
0.9 
1444
49 
2126
69.9 0.9 
11.
6 9.2 
25.
4 
21.
9 31 
10
0 12.5 34.6 52.9 
1959
.3 
2461
3.6 
1959
6.5 
5400
1.7 
4666
5.1 
6583
3.7 
2657
2.93 
7359
8.2 
1124
98.81 
 
coh 
808
9 
C
o
h Volume 
109
.2 
184
.4 66.2 
Es 
Fla
ke 
Qzit
e 
2902
69 
3591
94 
6494
63 7.3 9.2 
10.
5 
16.
9 
26.
9 
29.
2 
10
0 16.4 27.5 56.1 
4708
6.5 
5960
3.4 
6842
3.3 
1098
74 
1747
59 
1897
17 
1066
89.9 
1782
97.3 
3644
76 
 
coh 
900
1 
C
o
h Volume 
68.
8 
154
.3 54.2 
Es 
Fla
ke 
Cher
t 
1120
32 
1135
85 
2256
17 7.1 7.2 
13.
9 15 
28.
7 
28.
1 
10
0 14.3 28.9 56.8 
1603
7.2 
1621
2.9 
3134
8.6 
3392
6.2 
6464
6.5 
6344
5.9 
3225
0.1 
6527
4.8 
1280
92.4 
 
coh 
e22 
C
o
h Volume 
49.
9 
125
.1 33.6 
Es 
Fla
ke 
Horn
fels 
4309
3.7 
6518
4.4 
1082
78.1 6 
16.
5 
13.
3 
22.
3 
20.
5 
21.
4 
10
0 22.4 35.7 41.9 
6447
.9 
1781
9.2 
1444
2.5 
2417
8.8 
2220
3.4 
2318
6.5 
2426
7.09 
3862
1.3 
4538
9.9 
 
coh 
kj1
3 
C
o
h Volume 
42.
7 
87.
2 34.3 
Ind
et 
Qzit
e 
4563
2.4 
1592
8.7 
6156
1.1 7.4 0.5 
25.
6 1.2 
41.
2 
24.
2 
10
0 7.9 26.8 65.3 4567 
291.
1 
1573
1.7 
741.
9 
2533
3.7 
1489
5.7 
4858
.07 
1647
3.6 
4022
9.4 
 
coh 
n17
- 
C
o
h Volume 
57.
8 
169
.1 48.5 
Cs 
Fla
ke 
Quar
tz 
4334
0.9 
1210
15 
1643
55.9 3.5 
10.
2 5.7 
26.
2 
17.
2 
37.
3 
10
0 13.7 31.9 54.5 
5765
.3 
1673
2.8 
9296
.7 
4305
2.4 
2827
8.9 
6123
0 
2249
8.1 
5234
9.1 
8950
8.9 
 
coh 
pr2
4 
C
o
h Volume 
64.
4 
124
.6 31.3 
Ss 
Fla
ke 
Qzit
e 
7799
0.5 
5874
7.1 
1367
37.6 
10.
5 7.6 18 
11.
7 
28.
6 
23.
7 
10
0 18 29.7 52.3 
1429
2.8 
1032
4.5 
2457
8.7 
1604
8.1 
3911
9 
3237
4.5 
2461
7.3 
4062
6.8 
7149
3.5 
 
coh
461 
C
o
h Volume 58 
116
.6 46.8 
Ind
et 
Horn
fels 
3938
2 
1115
12 
1508
94 1.9 
14.
8 7.1 
22.
5 
17.
1 
36.
6 
10
0 16.7 29.6 53.7 
1033
3.7 
2936
8 
1490
7.5 
4732
4.2 
3416
3.7 
3978
2.4 
3970
1.7 
6223
1.7 
7394
6.1 
                               
Mean 
   
59.
86 
128
.12 45.11 
  
7911
5 
1031
55.97 
1822
70.98 
7.5
4 
12.
35 
14.
43 
18.
9 
21.
98 
24.
81 
 
19.88 33.32 46.78 
1452
8.29 
2259
1.62 
2616
8.33 
3558
6.78 
3904
4.07 
4513
2.68 
3711
9.9 
6175
5.12 
8417
6.76 
Std.De
viation 
   
12.
81 
28.
88 10.01 
  
5163
8.15 
6314
3.22 
1110
37.84 
4.2
8 
4.4
6 
5.3
4 
5.6
2 
7.0
7 
6.2
7 
 
6.64 5.11 10.19 
1210
1.45 
1291
2.16 
1725
9.84 
2202
7.58 
2843
3.64 
3240
4.23 
2351
1.54 
3642
5.79 
5977
6.83 
Co-
efficien
t of 
variati
on 
   
0.2
1 
0.2
2 0.22 
  
0.65 0.61 0.6 
0.5
6 
0.3
6 
0.3
7 
0.2
9 
0.3
2 
0.2
5 
 
0.33 0.15 0.21 0.83 0.57 0.65 0.61 0.72 0.71 0.63 0.58 0.71 
  
58 
 
REFERENCES 
 
Archer, W. and Braun, D. R. 2010. Variability in bifacial technology at Elandsfontein, 
Western Cape, South Africa: a geometric morphometric approach. Journal of Archaeological 
Science 37:  201-209. 
 
Asfaw, B. Beyene, Y. Suwa, G., Walter, R. C. White, T. D. Woldegabrieal, G. and Yemane, 
T. 1992. The earliest Acheulean from Konso Gardula. Nature 360: 732-735. 
 
Beyene, P.Y. Suwa, G., Katoh, S. Asfaw, B. 2007. The beginning of the Acheulean culture in 
its environmental context at Konso, Ethiopia. Abstract. In: de Lumley, H. (ed.), Les Cultures 
à Bifaces du Pléistocene Inferieur et Moyen dans le Monde. Émergence du Sens de l' 
Harmonie. Nice: Edudud. 
 
Brande, S. & Saragusti, I. 1996. A Morphometric Model and Landmark Analysis of 
Acheulean Handaxes from Northern Israel. In: Marcus, L. F. et al (eds), Advances in 
Morphometrics. Plenum Press, New York. pp. 225-243. 
 
Cardillo, M. 2010. Some applications of Geometric Morphometrics to archaeology. In: 
Elewa, A. M. T. (eds), Morphometrics for Nonmorphotericians. Lecture Notes in Earth 
Science 124. Springer, Dordrecht, pp. 325-341. 
 
Clark, J. D. 2001. Variability in primary and secondary technologies of the Later Acheulean 
in Africa. In: Milliken, S. & Cook, J. (eds) A Very Remote Period Indeed:1-18  Oxford, 
Oxbow Books. 
 
59 
 
Clarkson, C. 2006. Quantifying flake scar patterning on cores using 3D recording techniques. 
Journal of Archaeological Science 33: 132-142. 
 
Crompton, R. H. and Gowlett, J. A. J. 1993. Allometry and multidimensional form in 
Acheulean bifaces from Kilombe, Kenya. Journal of Human Evolution 25: 175-199. 
 
Domínguez-Rodrigo, M. and de la Torre, I. 2002. The ST Site Complex at Peninj, West Lake 
Natron, Tanzania: Implications for Early Hominid Behavioural Models. Journal of 
Archaeological Science 29, 639-665. 
 
Gibbon, R. J., Granger, D. E., Kuman, K. and Partridge, T. C. 2009. Early Acheulean 
technology in the Rietputs formation, South Africa, dated with cosmogenic nuclides. Journal 
of Human Evolution 56: 152-160. 
 
Goren-Inbar, N. & Sharon, G. 1999. Soft percussor use at the Gesher Benot Ya'aqov 
Acheulean Site? Journal of The Prehistoric Society 28: 55-79. 
 
Goren-Inbar, N. and Sharon, G. 2006. Invisible handaxes and visible Acheulean biface 
technology at Gesher Benot Ya'aqov, Israel. In: Goren-Inbar, N. and Sharon, G. (eds), Axe 
Age: Acheulean Tool-making from Quarry to Discard: 111-136. London, Equinox. 
 
Gowlett, A., J. 2011. The Vital Sense of Proportion: Transformation, Golden Section, and 1:2 
Preferences in Acheulean Bifaces. Paleoanthropology, 174-187.  
 
Grosman, L., Goldsmith, Y., Smilansky, U. 2011. Morphological Analysis of Nahal Zihor 
Handaxes: A Chronological Perspective. Paleoanthropology. 203-215. 
60 
 
Grosman, L., Smikt, O. and Smilansky, U. 2008. On the application of 3-D scanning 
technology for the documentation and typology of lithic artifacts. Journal of Archaeological 
Science 35: 3101-3110. 
 
Hardaker, T. & Dunn, S. 2005. The Flip Test – a new statistical measure for quantifying 
symmetry in stone tools. Antiquity 79 (306)> File:///C:/Documents%20 
and%20settings/09400149/ 
My%20Documents/OLD%20C%20drive%20Files/flip%20test%20antiquity,%20Project%20
Gallery%20Hardaker%20&%20Dunn.mht 
 
Hay, R. L. 1976. Geology of the Olduvai Gorge: A Study of Sedimentation in a Semi-Arid 
Basin.  Berkeley: University of California Press. 
 
Isaac, G. L. 1997. Koobi Fora Research Project, Vol 5, Plio-Pleistocene Archaeology. 
Oxford: Clarendon Press. 
 
Klein, R. 2000. The Earlier Stone Age of southern Africa. South African Archaeological 
Bulletin 55: 107-122. 
 
Kuman, K., Gibbon, R. J. in press. The early Acheulean of South Africa: Sterkfontein valley 
and the Vaal River Basin. In: de Lumley, H (eds), Les Cultures à Bifaces du Pleistocene 
Inferieur et Moyen dans le Monde. Émergence du sens de l’Harmomie. Nice: Edusud. 
 
Kuman, K. 2007. The Earlier Stone Age in South Africa: site context and the influence of 
cave studies. In: Pickering, T. R. Schick, K. and Toth, N. (eds), Breathing Life into Fossils: 
Taphonomic Studies in Honour of C. K. (Bob) Brain. Bloomington (Indiana): Stone Age 
Institute Press, pp 181-198. 
61 
 
Kuman, K., Field, A., S. & Thackery, J., F. 1997. Discovery of new artefacts at Kromdraai. 
South African Journal of Science 93: 187-193. 
 
Kuman, K. & R.J. Clarke 2000. Stratigraphy, artefact industries and hominid associations for 
Sterkfontein, Member 5. Journal of Human Evolution 38: 827-847. 
 
Leader, G. M. 2009. Early Acheulean in the Vaal River basin, Rietputs Formation, Northern 
Cape Province, South Africa. MSc Dissertation. University of the Witwatersrand. 
 
Lepre, C. J., Roche, H., Kent, D. V., Harmand, S., Quinn, L. R., Brugal, J., Texier, P., 
Lenoble, A., Feibel, C. S., 2011. An earlier origin for the Acheulian. Nature 477. 82-85. 
 
Lin, S. C. H., Douglass, M. J., Holdaway, S. J., Floyd, B. 2010. The application of 3D laser 
scanning technology to the assessment of ordinal and mechanical cortex quantification in 
lithic analysis. Journal of Archaeological Science 37: 694-702. 
 
Lycett, S. J., von Cramon-Taubadel, N., Foley, R. 2006. A crossbeam co-ordinate calliper for 
morphometric analysis of lithic nuclei: a description, test and empirical examples of 
application. Journal of Archeological Science 33. 847-861. 
 
Mason, R. J. 1962. Prehistory of the Transvaal. Johannesburg: Witwatersrand University 
Press. 
 
Minium, E., W. 1978. Statistical Reasoning in Psychology and Education (2nd ed). New York, 
John Wiley & Sons. 
 
62 
 
McNabb, J. Binyon, F. Hazelwood, L. 2004. The Large Cutting Tools from the South African 
Acheulean and the Question of Social Traditions. Current Anthropology 45: 653-677. 
 
McPherron, S. P. 2000. Handaxes as a measure of mental capabilities of early hominids. 
Journal of Archaeological Science 27: 655-663. 
 
Pickering, T. R., Egeland, C. P., Domínguez-Rodrigo, M., Brain, C. K., Schnell, A. G. 2008. 
Testing the ‘’shift in the balance of power’’ hypothesis at Swartkrans, South Africa: 
Homonid cave use and subsistence behaviour in the Early Pleistocene. Journal of 
Anthropological Archaeology 27: 30-45. 
 
Porat, N. Chazan, M. Grun, R. Aubert, M. Eisenmann, V. and Horwitz, K. L. 2010. New 
radiometric ages for the Fauresmith industry from Kathu Pan, southern Africa: implications 
for the earlier to Middle Stone Age transition. Journal of Archaeological Science 37: 269-
283. 
 
Posnansky, M. 1959. Some Functional Considerations on the Handaxes. Man 61. 42-44. 
 
Potts, R. 1989. Olargesailie: new excavations and findings in Early and Middle Pleistocene 
contexts, southern Kenya rift valley. Journal of Human Evolution 18: 477-484. 
 
Roe, D., A. 1964. The British Lower and Middle Paleolithic: some problems. Methods of 
study and preliminary results. Preceedings of the Prehistoric Society 30. 245-267. 
 
Roe, D.,A. 1994. Summary and overview. In M.D. Leakey and D.A. Roe (eds): Olduvai 
Gorge, Vol. 5. Cambridge: Cambridge University Press.  
63 
 
Saragusti, I. Karasik, A. Sharon, I. and Smilansky, U. 2005. Quantitative analysis of shape 
attributes based on contours and section profiles in artifact analysis. Journal of 
Archaeological Science 32: 841-853. 
 
Saragusti, I. Sharon, I. Katzenelson, O. and Avnir, D. 1998. Quantitative analysis of the 
symmetry of artefacts: Lower Paleolithic handaxes. Journal of Archaeological Sciences 25: 
817-825. 
 
Schick, K. and Toth, N. 1995. The importance of Actualistic Studies in Early Stone Age 
Research: Some Personal reflections. In Schick, K. and Toth, N. (eds), The Cutting Edge: 
New Approaches to the Archaeology of Human Origins 3, 267-344. 
 
Sumner, T. A. and Riddle, T. R. A. 2008. A virtual Paleolithic: assays in photogrammetric 
Three Dimensional Artifact Modelling. Paleoanthropology: 158-169. 
 
Tobias, P., V. 1971 Human skeletal remains from the Cave of Hearths, Makapansgat, 
Northern Transvaal. American Journal of Physical Anthropology 34: 335-367. 
 
Toth, N. 1985. The Oldowan reassessed: a close look at early stone artifacts. Journal of 
Archaeological Science 12: 101-120. 
 
Underhill, D. 2007. Subjectivity inherent in by-eye symmetry judgements and the large 
cutting tools at the Cave of Hearths, Limpopo Province, South Africa. Papers from the 
Institute of Archaeology 18: 1-12. 
 
van Riet Lowe, C. 1945. The evolution of the Levallois Technique in South Africa. Man 
XLV: 37-51. 
64 
 
Wynn, T., Tierson, F., 1990. Regional comparison of the shapes of later Acheulean handaxes. 
American Anthropologist 92. 73-84. 
 
Wynn, T. 1995. Handaxe Enigmas. World Archaeology 27: 299-322
0