Toolkits and Utility in Australian Lithics kit and discard assemblages

Toolkits and Utility in Australian Lithics kit and discard assemblages
Toolkits and Utility in Australian Lithics
A comparison of a comprehensive woodworking
kit and discard assemblages
John Hayward
October 2010
Revised January 2011
Department of Archaeology
Faculty of Education, Humanities, Law and Theology
Flinders University
South Australia
Thesis submitted in partial fulfillment of the requirements of
the degree of Honours in Archaeology
Table of Contents
1. Introduction
1.1 Introduction................................................................................. 1
1.2 Aims and significance................................................................ 2
1.3 Toolkits and function.................................................................. 2
1.4 Approach................................................................................... 4
1.5 Limitations................................................................................. 6
1.6 Outline of the thesis.................................................................. 6
1.7 Conclusion................................................................................ 8
2. The evolution of the toolkit concept
2.1 Introduction................................................................................. 9
2.2 Early approaches to Australian lithics........................................ 9
2.3 The toolkit in archaeology.......................................................... 10
2.4 Australian Toolkits...................................................................... 18
2.5 Conclusion................................................................................. 22
3. The Lake Hanson Cache
3.1 Introduction .............................................................................. 24
3.2 The Woomera Natural History Society...................................... 24
3.3 The Mungappie sites................................................................. 26
3.4 The Lake Hanson cache and Hewitt’s collecting....................... 28
3.5 Hewitt’s classifications.............................................................. 30
3.6 Cache artefact woodworking types........................................... 34
3.7 Other woodworking artefact types............................................. 38
3.8 Conclusion................................................................................ 44
4. Methodology
4.1 Introduction............................................................................... 46
4.2 Recording systems................................................................... 46
4.3 Data analysis............................................................................ 59
4.4 Volume of artefacts................................................................... 59
4.5 Conclusion................................................................................ 60
iii
5. Results
5.1 Introduction............................................................................... 61
5.2 Artefact types at Mungappie sites.............................................. 62
5.3 Toolkit types .............................................................................. 64
5.4 Raw materials............................................................................. 69
5.5 Utility Units of tulas.................................................................... 74
5.6 Edge angles............................................................................... 83
5.7 Cortex........................................................................................ 87
5.8 Terminations.............................................................................. 89
5.9 Breakages.................................................................................. 90
5.10 Retouch location...................................................................... 91
5.11 Retouch types.......................................................................... 95
5.12 Conclusion............................................................................... 96
6. Discussion
6.1 Introduction................................................................................. 98
6.2 Site comparisons....................................................................... 98
6.3 Lake Hanson cache.................................................................... 101
6.4 Artefact attribute analyses...........................................................103
6.5 Conclusion.................................................................................. 105
7. Conclusion........................................................................................ 106
References........................................................................................... 111
Glossary of terms................................................................................. 126
Mungappie sites drawings................................................................... 131
Mungappie sites photographs............................................................. 140
Appendices........................................................................................... 144
iv
List of figures
Figure 1.1 The Arcoona Plateau and location of the Mungappie sites and the
Lake Hanson toolkit cache........................................................................ 5
Figure 2.1. Cumulative graph of Denticulate, Typical, and Quina Mousterian
as used by Bordes...................................................................................12
Figure 3.1 Ron Hewitt holding an artefact (yodda)........................................26
Figure 3.2 Mungappie sites and Lake Hanson cache site. .......................... 27
Figure 3.3 The Lake Hanson cache. R. Hewitt..............................................29
Figure 3.4 Hewitt’s drawings......................................................................... 33
Figure 3.5 Koondi Tuhla from Yuendumu in Central Australia...................... 35
Figure 3.6 Tula reduction sequence (Campbell and Edwards 1966)............ 36
Figure 3.7. Points Photograph R. Hewitt....................................................... 39
Figure 3.8 Pirrie Point with broken tip........................................................... 40
Figure 3.9 Long distal scrapers. Photograph R. Hewitt................................. 41
Figure 3.10 Geometric Microliths, trapezoid, triangular, and crescent forms 43
Figure 4.1 Recording sheet for Mungappie site assemblages...................... 47
Figure 4.2 Edge angles were taken from three points of the retouched
working edge of the flake........................................................................ 57
Figure 4.3 Ventral and Dorsal faces of flake showing retouch margin
divisions.................................................................................................. 57
Figure 5.1. Percentage distribution of artefact types for each of the
Mungappie sites...................................................................................... 63
Figure 5.2 Comparison of ‘woodworking kit’ artefact distribution between the
Lake Hanson Cache and each Mungappie site...................................... 66
Figure 5.3. Percentage comparison of ‘woodworking kit’ artefact types at
Lake Hanson and all Mungappie sites.................................................... 67
Figure 5.4 Percentage comparison of material types at Lake Hanson cache
and all Mungappie sites.......................................................................... 70
Figure 5.5 Materials for each Mungappie site and Lake Hanson cache....... 71
Figure 5.6 Raw material composition of each toolkit type from A - L at Lake
Hanson Cache and all the Mungappie sites. .......................................... 73
Figure 5.7 Volume and Utility Units of tulas at Mungappie sites................... 76
Figure 5.8 Utility Units for all Munggapie sites compared to individual
Mungappie sites...................................................................................... 77
iii
Figure 5.9 Percentage comparison of tula types at all Mungappie Sites and
Lake Hanson cache................................................................................ 79
Figure 5.10 From Table 9, Utility Units expressed as graph......................... 81
Figure 5.11 The total Utility Units for each site type...................................... 82
Figure 5.12 Average edge angles for all artefacts at all Mungappie sites.....84
Figure 5.13 Number of tulas types for each edge angle range from all
Mungappie sites...................................................................................... 85
Figure 5.14 Average edge angles of tulas from each Mungappie site.......... 86
Figure 5.15 Percentage of tula types from each Mungappie site with cortex.
................................................................................................................ 88
Figure 5.16 Artefact #1060. Distal end break with point............................... 91
Figure 5.17 Comparison between scrapers and tulas - edge angles to
retouch location....................................................................................... 93
Figure 5.18 Percentage amounts of scrapers and tulas for each retouch
location.................................................................................................... 93
Figure 5.19 Comparison between pirrie points and points - edge angles to
retouch location....................................................................................... 94
Figure 5.20 Percentage amounts of pirrie points and points for each retouch
location.................................................................................................... 94
Figure D.1 Type A. Chert tula slug. Artefact #11. Mungappie Creek North. 130
Figure D.2 Type C. Chert semi-discoidal tula #29. Mungappie Swamp..... 130
Figure D.3 Type C. Silcrete semi-discoidal tula #86. Mungappie Creek..... 131
Figure D.4 Type G. Chert tula blank #98. Mungappie Creek...................... 131
Figure D.5 Type PB. Quartzite distal scraper (bulbous) #147. Mungappie
Creek..................................................................................................... 132
Figure D.6 Type P. Quartzite scraper (Distal end fragment) #979. Mungappie
Hut.........................................................................................................132
Figure D.7 Type K. Quartzite flake with serrated lateral margins #1197.
Mungappie Creek North........................................................................ 133
Figure D.8 Type P. Quartzite scraper with serrated/notched lateral margins
and distal end #1223. Mungappie Hut...................................................133
Figure D.9 Type Q. Quartzite core #384. Mungappie Hut........................... 134
Figure D.10 Type Q. Oolitic chert core #1353. Mungappie Creek...............134
Figure D.11 Type PD. Translucent chalcedony discoidal scraper #1013.
Mungappie Creek North........................................................................ 135
iv
Figure D.12 Type PD. Quartzite discoidal scraper #1020. Mungappie Creek.
.............................................................................................................. 135
Figure D.13 Type G. Chert tula blank #258. Mungappoie Creek North...... 136
Figure D.14 Type C. Quartzite discoidal tula #371. Mungappie
Hut. .............................................................................................................136
Figure D.15 Type T. Quartzite Pirrie point #671. Munagppie Creek............137
Figure D.16 Type J. Quartzite Flake with burin #969. Mungappie Hut........ 137
Figure D.17 Type P. Quartzite scraper #1219. Mungappie Creek North..... 138
Figure D.18 Type T. Pirrie, Type N. Point, and Type W. Geometric Microlith.
.............................................................................................................. 138
Figure P. 1 Type C. Quartzite semi-discoidal tula #1004. Mungappie Creek
North......................................................................................................139
Figure P.2 Type K. Oolitic chert flake #1191. Mungappie Creek................. 140
Figure P. 3 Type T. Quartzite pirrie point #869. Mungappie Hut..................141
Figure P.4 Type I. Oolitic chert Micro tula blank #1138 Mungappie Creek
North..................................................................................................... 142
List of Tables
Table 3.1 Cache artefact types and quantities. Codes, types and quantities
are from Hewitt (1976:19)....................................................................... 30
Table 4.1 Colour range codes for stone types at Mungappie sites............... 51
Table 4.2 Showing retouch location combinations and retouch code used on
the recording form................................................................................... 58
Table 5.1 Cores at all Mungappie sites......................................................... 64
Table 5.2 Comparison of tula type quantites between Lake Hanson Cache
and the Mungappie sites......................................................................... 68
Table 5.3 Mungappie Creek and Lake Hanson cache - a comparison
between materials and types.................................................................. 69
Table 5.4 Volumetric averages of tula types from all Mungappie sites in cubic
centimetres (cc)....................................................................................... 75
Table 5.5 Numbers of tulas at each Mungappie site..................................... 78
v
Table 5.6 Calculation of Utility Units by volume (cc) of tulas at Lake Hanson
cache and Mungappie Sites.................................................................... 80
Table 5.7 Quantities of tula types, with percentage of cortex on each, from all
Mungappie sites...................................................................................... 89
Table 5.8 Terminations of all artefacts types from all Mungappie sites......... 90
Table 5.9 Retouch locations for all artefacts with retouch at all Mungappie
sites......................................................................................................... 92
Table 5.10 Retouch types for all retouched artefacts from all Mungappie
sites......................................................................................................... 95
Table 5.11 Summary of differences between Lake Hanson cache and the
Mungappie sites........................................................................................... 97
vi
Abstract
The concept of a toolkit has been used to describe functional aspects of
lithic assemblages since the 1960s, but has proved difficult to define. The
history of the concept, which emerged from the analysis of European
Mousterian assemblages by the Binfords and Bordes, is traced from its
roots to the present day. In Australia it has become a generalised term
which has been used to explain the complete range of technologies
available to a culture, as well as defining strategies for risk management
and mobility. This research investigates the concept and its applicability to
Australian lithic assemblages.
In 1970 a cache of 105 stone artefacts was discovered at the top of a sand
dune in the arid landscape around Lake Hanson on South Australia’s
Arcoona Plateau. Its finder interpreted the cache as ‘a comprehensive
woodworking kit’. This 'tool-kit' is compare with assemblages from four
sites collected from nearby Mungappie Creek by the same person. The
analysis compared the number of artefact types, their sizes, and the
materials used at each of the Mungappie sites, including the Lake Hanson
cache. Using the notion that a functional toolkit would need to have more
potential utility than a discarded one, an assessment of the potential use
life of an artefact in the form of ‘utility units’ was employed to indicate the
possible presence of toolkits at each of the Mungappie sites. The results
indicated that the toolkit cache was a unique collection of artefact and
material types that were scarce at any of the four Mungappie
assemblages. The conclusion being that there are profound differences
between discard assemblages and discrete entities such as caches and
toolkits, suggesting the need for a revision of the toolkit concept from a
generalised to a specific terminology.
vii
‘I certify that this thesis does not incorporate without acknowledgment any
material previously submitted for a degree or diploma in any university;
and that to the best of my knowledge and belief it does not contain any
material previously published or written by any other person except where
due reference is made in the text.’
John Hayward
viii
Acknowledgements
I would like to thank the following people, without whom this thesis would
not exist in its present form.
Academic support - Dr Alice Gorman, who, despite my doubts, insisted
that this was a good idea from the start, and helped bash the rough edges
out during the painful editing process to the final shaping of the finished
product.
Professional support - South Australian Museum staff: Dr Keryn Walshe
(senior collections manager), Gary Toone (collections manager).
Thanks to my cheerful band of willing volunteers who stuck around during
the intense days of measuring and sorting – Adrian Fenech, Trevor Tisdall,
Marie Butler, Michael Slizankiewicz, Scott Jacob, Elizabeth Hartnell, Jo
Thredgold.
My partner, friend, and fellow tango dancer, Dr Susanne Schech who has
supported me both materially and morally for the four years that it has
taken to get to this point. Thanks for sticking with me through to the early
hours of the final morning.
Friends Will Powrie, Charles Davison, and Peter Lindon who feed me
ideas, books, encouragement and cups of coffee; and fellow artist and
writer Ian Hamilton who without any provocation offered to read and
comment on my draft.
Despite ailing health and aging bones, Ron Hewitt was always cheerful
and willing to recount his memories of his collecting days in Woomera.
Along with his wife Joyce, they added a wonderful human dimension to the
process and story.
ix
Toolkits and utility in Australian lithics
1. Introduction
1.1 Introduction
This thesis is a critical analysis of the concept of the ‘toolkit’ as used in
contemporary archaeology to describe stone artefacts. The aim of the thesis
is to investigate if ‘toolkit’ is an appropriate term to apply to Australian
Indigenous lithics, and if so, how toolkits can be identified in lithic
assemblages. The theoretical analysis traces the inception of the concept
from European archaeology in the mid 1960s to its current usage. To
investigate the applicability of the concept, a lithic assemblage from the
Arcoona Plateau region of South Australia, collected in the late 1960s and
early 1970s by amateur archaeologist Ron Hewitt, has been analysed. A
subsection of Hewitt’s private collection was a cache of stone artefacts,
described by him as a “comprehensive woodworking kit” (Hewitt 1977:29).
Whilst there are numerous ethnographic accounts of stone implements being
used for woodworking in Australia, the notion of these tools forming a
discrete toolkit entity is not explicit, and has not been described
ethnographically. This raises the question of whether a stone toolkit actually
existed as a physical entity, and if so, why was it unnoticed by
ethnographers? Either way, the existence of such a grouping could have
significant implications for the study of the adaptation and resource
management of prehistoric cultures (Torrence 1983; Hiscock 1994; Kuhn
1994; Moore 2003), as the changing composition of the woodworking toolkit
through time and space could have a direct relationship to the environment
and cultural behaviour. In contemporary life toolkits do reflect social
behaviour and technological trends, so why should this not be the case with
prehistoric toolkits – if they existed?
1
Toolkits and utility in Australian lithics
1.2 Aims and significance
The impetus behind this research is to gain an understanding of the concept
of toolkits in lithic archaeology. The aims of this research are to:
• Examine the historical origins of the term ‘toolkit’ in archaeology.
• Interrogate the changing concept of the toolkit in contemporary
Australian archaeology.
• Understand how changing notions of the toolkit are related to
theories of mobility and flexibility in Australian archaeology.
• Examine the composition of toolkit structure in comparison to the
composition of lithic assemblages.
The concept of toolkits has been used in many parts of the world to describe
a range of aspects related to lithic assemblages, and in so doing has become
quite elastic and flexible. Most researchers (eg Binford 1966, 1968, 1977,
1983; Bordes 1977; Cahen et al 1979; Hayden 1977, 1979; Hewitt 1976;
Hiscock 1994; Kuhn 1994; McCarthy 1976; Moore 2003; Banks 2009) do not
critically analyse their use of the term and its implications. This thesis seeks
to redress this by focusing on the concept of the toolkit in global archaeology
by tracing its use over the past fifty years, and analysing its contribution to
Australian archaeology. This is the first time that lithic toolkit structure has
been analysed in the Australian context. In doing so, the research will help to
redefine the meaning of the toolkit within Australian archaeology.
1.3 Toolkits and function
The etymology of the term ‘toolkit’ is obscure, and it may not have been
common until the early twentieth century. When the term ‘kit’ was used in the
mid-nineteenth century it was generally understood as the part of the
workshop where cobblers worked and kept their tools, and later in 1881
extended to ‘the kit of tools for a nipple maker’ (Oxford English Dictionary (v)
viii:465). Both definitions suggest a grouping of tools associated with a trade.
Contemporary usage is similar but applied more broadly.
2
Toolkits and utility in Australian lithics
The traditional view in archaeological literature is that toolkits are also
function-specific (Binford and Binford 1966; Freeman 1966:232; Binford
1968). The question of what defines a function-specific stone toolkit is critical
in contemporary lithic studies, with the added consideration of how to identify
such toolkits within an assemblage. What are the distinguishing traits
connecting one artefact to another, that, when combined, make a kit? One
stone tool by itself cannot be a kit, but are two stones of similar type or
function a kit? Are three stone of similar types, or two similar stones and one
different stone a kit? How many stone tools constitute a kit?... and so on.
Contemporary tradespeople carry with them kits of tools that are quite
specific to their trade, and can vary considerably depending upon their area
of specialisation. Toolkits and function are therefore closely aligned; a toolkit
that is comprised of multifunctional tools may cover a range of basic tasks,
but may not fulfill the needs of a specific task; a toolkit that has tools which
are flexible, and can be adapted to specific tasks or tailored for unseen
events could be a good model for some situations but may need to be larger
to accommodate the unexpected; and a toolkit that is designed to complete a
specific task may not be adaptable or comprehensive enough for general
needs. It is possible that at any archaeological site there may be multiple
toolkits present, each with their own function, to facilitate the range of
activities (Binford 1983:147).
Recent research has shown that morphologically diverse assemblages are
not, as previously thought, discontinuous – with each morphological type
having a function-specific design – but in a state of continuous variation
which reflects artefacts in different stages of reduction (Hiscock and
Attenbrow 2003, 2005). The notion that a stone implement can change both
its morphology and function during its use life is no longer contentious,
indicating that a lithic assemblage is not merely a snapshot image of an
archaeological moment where everything is clearly categorised, but rather is
just one frame of an ever-changing, and never-ending epic where the
characters are continually changing roles. This idea not only challenges the
traditional approach to lithic typology but also brings into question the notion
that specific types constitute a toolkit. If toolkits are function-specific (Binford
3
Toolkits and utility in Australian lithics
and Binford 1966; Freeman 1966:232; Binford 1968), as traditional
understanding suggests, then either multi-functional tools cannot be
accommodated in ‘function-specific toolkits’, or the notion of ‘function-specific
toolkits’ need to be re-thought as being multi-functional or trans-functional
toolkits. These issues will be explored through artefact collections from the
Arcoona Plateau of South Australia collected by Ron Hewitt.
1.4 Approach
During the ten years that Ron Hewitt worked in the outback South Australian
town of Woomera, as a technical engineer for the defence establishment, he
collected several thousand Aboriginal stone artefacts from a number of sites
on the Arcoona Plateau that surrounds the town (Figure 1.1). In 1970, the
Woomera Natural History Society, a group of locals interested in
anthropology and natural history, organised a field excursion to the Lake
Hanson area, about 50 kms west of Woomera. During that trip Ron Hewitt
found what he regarded in his 1976 paper as a cache of Aboriginal stone
artefacts (Fig 2). He described the cache as a hoard of “semi discoidal
adzes, discoidal adzes, micro adzes, gouges, engravers, and burins which
together make up a comprehensive woodworking kit” (Hewitt 1976:29;
Hiscock 1988:68). Hewitt’s cache had a similar number of tulas to one
excavated by Hiscock in the Boulia District of Queensland during the 1980s,
which consisted of fifty stone artefacts. These are two of a small number of
assemblages found in Australia that can be classified as caches for later use
(Hiscock 1988:68).
4
Toolkits and utility in Australian lithics
Figure 1.1 The Arcoona Plateau and location of the Mungappie sites and the Lake
Hanson toolkit cache.
Hewitt’s personal stone artefact collection, including the Lake Hanson cache,
has not yet been located, and so direct analysis of the artefacts has not been
possible. Hewitt did, however, document the raw materials and typologies of
the 105 artefacts that constituted the Lake Hanson cache, and these have
been used as a model for the woodworking toolkit as conceived in the 1970s.
This model toolkit was compared with other assemblages collected by Hewitt
at four sites around Mungappie Creek on the Arcoona Plateau in the late
1960s and early 1970s. A total of 1364 artefacts from Mungappie Creek,
Mungappie Creek North, Mungappie Hut, and Mungappie Swamp were
recorded onto an Excel spreadsheet where statistical analysis was done to
test the hypothesis of Hewitt’s toolkit as a recurring discrete item at these
sites.
5
Toolkits and utility in Australian lithics
1.5 Limitations
The assemblages are from surface scatters found in wind-eroded sites of the
Arcoona Plateau which Hewitt described as being “so dense that it is
impossible to not walk on them” (Hewitt 1978:11). These collections may be
more representative of the collector’s bias towards archetypal tool types,
than of the true nature of the scatters. It was expected that Hewitt’s collection
would under-represent unretouched flakes, which at the time would not have
been considered as tools, but debitage (Campbell and Edwards 1966:172).
Much of Hewitt’s Arcoona collection is held at the archaeology store of the
South Australian Museum. However, searches of the museum’s database,
and physical searches through relevant boxes revealed that Hewitt’s Lake
Hanson cache was absent. Hewitt, who lives in Adelaide, was contacted by
collections manager Dr. Keryn Walshe, in the hope that his personal
collection would be intact along with documentation of the artefacts. Hewitt
had contacted the museum some five years ago (before Dr. Walshe worked
there) to donate his personal collection, but the museum declined the offer.
The museum then referred the offer to a local heritage firm who, along with a
representative of the Woomera Indigenous Traditional Owners – Kokatha
Mula, reclaimed the artefacts. Not having the cache for comparative research
has enforced limitations on the type of comparisons that could be made.
1.6 Outline of the thesis
The literature review discusses the inception of the term “toolkit” in the
archaeological literature, and follows the unfolding definitions of the concept
from the 1960s to the 1980s. This chapter also traces some of the
experimental analysis used to determine the composition of curated toolkits,
and how this research relates to Australian archaeology.
An account of Hewitt’s cache and collections, including typologies, material
types, and descriptions of the Mungappie sites, is detailed in chapter 3.
Because Hewitt’s focus on raw materials is significant to the results,
geological, geographical and environmental information is described
6
Toolkits and utility in Australian lithics
including raw material, quarry sources, how they relate geographically to the
sites, and how they affect the significance of the woodworking toolkit. This
chapter also gives an account of woodworking stone tools that have been
ethnographically described in Australia, and how these tool types relate to the
Mungappie sites and the Lake Hanson cache.
Chapter 4 describes the methods used to record the Mungappie Creek
assemblages, including background information about typologies employed,
criteria for definitions, and codes. The spreadsheet analysis including
statistical information about the assemblages; the comparisons made
between the sites, the materials, and the types; metric analysis of each
artefact attribute; and how this relates to the search for toolkits in the
assemblages, will also be discussed.
A complete analysis of the results of all the attributes recorded is given in
Chapter 5. This includes comparisons of the artefact composition for the four
Mungappie sites, relationships between artefact types and attributes, and the
differences between the Mungappie sites and the Lake Hanson woodworking
kit.
Chapter 6 discusses the results of the Mungappie Creek sites analysis and
relates them to current theories of mobility, and flexibility in Australian
archaeology. The results are also used to compare Hewitt’s analysis of the
Lake Hanson cache on material and typological terms. The comparison of
the cache with the Mungappie site assemblages on a range of criteria,
generated a number of measurable differences between them, which is
discussed in detail.
The significance of the outcome of the research, and its implications for the
understanding of toolkit structure is summed up in the conclusion chapter.
Some points that can be used to define toolkits are outlined, as well as
indicators of future research ideas.
7
Toolkits and utility in Australian lithics
1.7 Conclusion
This thesis responds to Hiscock’s call for a “more sophisticated
conceptualisation of tool use and toolkits” (Hiscock 2005:53). Archaeologists
have tended to use the term toolkit as a heuristic device that has even been
used to refer to the entire range of technologies available to a culture. This
thesis attempts to take a more reflective and individualised view of the toolkit
by considering its role in prehistoric culture. The aim of the research is to
increase understanding of the meaning of toolkits within Australian lithics,
and to offer definitions for toolkit structure that have not yet been examined.
8
Toolkits and utility in Australian lithics
2. The evolution of the toolkit concept
2.1 Introduction
Toolkits, having been defined by their functional attributes by Lewis and Sally
Binford (1966, 1968) in the mid 1960s, have continued to be the subject of
both theoretical and experimental research in archaeology, including
contemporary ethnoarchaeology (Hayden 1977, 1979; Cane 1984, 1992),
use-wear analysis (Kamminga 1985; Grace 1989; Fullagar et al 1996; Banks
2009), reduction sequences (Kuhn 1990; Hiscock 2005) and risk
management theories (Hiscock 1994; Kuhn 1994; Moore 2003). These
studies, which have not only raised issues related to the technological and
functional aspects of stone artefact production, but also their cultural
associations, will be described in relation to current debates in Australian
archaeology.
2.2 Early approaches to Australian lithics
Stone tool typology was introduced to Australia from Britain by settlers such
as Robert Brough Smyth, anthropologist Baldwin Spencer and geologist
Walter Howchin in the late nineteenth century. The cartographer and
ethnographer Brough Smyth, who arrived in Melbourne in 1852 initially as a
miner, on his retirement from the Board of Protection of Aborginies in 1878
produced a classification of Australian stone implements which used
European terms such as hatchets, knives, and adzes, to describe the way
tools were used by Aboriginal people (Hoare 1976:161 Mulvaney 1977:263;
Roddom 1997:16).
This publication influenced Kenyon and Stirling’s stone artefact classification,
including nearly 70 variations of particular types determined by the tools’
shape, which was used as the basis of the display of Aboriginal stone
artefacts at the new National Museum of Victoria established by Baldwin
Spencer in 1899 (Kenyon and Stirling 1900; Howchin 1934:22; Mulvaney
1977:264). Spencer had been trained at Oxford and was closely associated
9
Toolkits and utility in Australian lithics
with Edward Tylor, first Professor of Anthropology at Oxford (Mulvaney
1977:264).
Howchin had also published on British stone tools before he arrived in
Adelaide in 1881, and his typology of the stone implements of the Adelaide
area is probably the earliest South Australian typology of its kind (Mulvaney
1977:264). This typology, based upon his work in Britain, is described in his
1934 publication ‘The Stone Implements of the Adelaide Tribe of Aborigines’.
He found marked differrences between European and Australian stone
chisels, chipped-back knives, and hollow scrapers, but drew favourable
comparisons between the general scrapers, and some drill type points
(Howchin 1934:31- 68). These early researchers had a significant influence
on Australian archaeologists who were mostly concerned with typologies and
cultural sequences until the 1960s when the advent of processual
archaeology allowed other discaplines such as anthropology to shape its
ideology.
2.3 The toolkit in archaeology
The idea that prehistoric sites may contain toolkits had been discussed by
the anthropologists Howell and Freeman, both from Chicago University, in
1964 and 1966 (Howell 1964, Freeman 1966). Howell’s concentration on the
differences manifest at Acheulian sites from the UK raised questions about
the nature of artefact variation within assemblages (Howell 1966:181), and
Freeman’s analysis of Mousterian sites from Spain and France advanced the
possibility of assemblage variation being more than just industrial evolution or
cultural influence, but rather toolkits which reflected different activities
(Freeman 1966:232). Lewis and Sally Binford’s reassessment of François
Bordes’ Mousterian facies of the 1950s utilised the same concept and
determined that the assemblages did not represent cultural variations, as
argued by Bordes, but were toolkits aimed at performing different tasks at
different times and places, depending upon seasonal variations (Binford and
Binford 1966:292; Holdaway and Stern 2004:50). The Binfords’ toolkit
categorisations, later defined in 1968 by the artefacts’ presumed functional
attributes, went on to become a mainstream archaeological term used to
10
Toolkits and utility in Australian lithics
describe stone tool assemblages, whether their functional attributes are
being defined or not.
Bordes (1953) introduced standards of description and comparison into
archeological analysis that were previously unknown. Dibble and McPherron
(2006:783) describe this standardization of the typology for lithic artefacts as
one of the most important contributions to palaeolithic archaeology during the
middle of the last century. Bordes’ methodology made objective comparison
between basic types of Mousterian assemblages possible for the first time.
Bordes initially classified these assemblages, in 1953, into six groups (Dibble
1991:239), which he later condensed into four typological units in 1961 (cited
in Binford and Binford 1966:238). They are:
The Mousterian of Acheulian Tradition (M.T.A.): characterised by the
presence of handaxes, numerous side-scrapers, denticulates, and
backed knives
Typical Mousterian: like M.T.A., but with reduced frequencies of
handaxes and knives.
Denticulate Mousterian: 80% denticulates and notched tools; no
handaxes or backed knives.
Charentian Mousterian: subdivided into two types - Quina and Ferrassie;
characterised by few or no handaxes or backed knives, as well as very
high frequencies of side-scrapers.
These typological units, the Mousterian facies, are still used to this day
(Dibble and McPherron 2006:783). Bordes use of cumulative frequencies
(Figure 2.1) allowed him to trace the relatively minor fluctuations of artefact
types in the different Mousterian assemblages based upon the presence,
absence, and frequency of particular artefacts – typical backed knives, Quina
scrapers, and handaxes (Bordes 1970:61).
11
Toolkits and utility in Australian lithics
Figure 2.1. Cumulative graph of Denticulate, Typical, and Quina Mousterian as used
by Bordes.
The Binfords took a different approach. The Binfords’ use of statistics to
determine toolkit structures was derived from a theory of ‘multivariate
causation’ which states that the ‘determinants of any given situation are
multiple and may be linked’ (Binford and Binford 1966:241). Initially each
artefact was given a functional interpretation based on implement types,
which they called ‘variables’ (Binford and Binford 1966:243). The factors used
in this statistical method were groups of ‘variables’ linked together through
frequencies of occurrence, that when isolated through the aid of a factor
analysis programme, showed consistent patterns of mutual covariation
(Binford and Binford 1966:245; Mellor 1970:82). This method measured the
variability present in the Mousterian assemblages, which was interpreted by
the Binfords as ‘functional variability’ (Binford and Binford 1966:292). The
analysis produced five factor groups, each with their own list of tool types,
suggested activity, and relation to Bordes’ types (Binford and Binford
12
Toolkits and utility in Australian lithics
1966:259), which the Binfords stated were ‘sub-units of artefacts which we
can infer were used in a related set of activities’ (Binford and Binford
1966:292). Binford and Binford (1966:292) made two main points about their
analysis of the Mousterian complex:
The use of multivariate statistics allowed the partitioning of Mousterian
assemblages into subunits of artifacts which could reasonably be interpreted
as representing tool-kits for the performance of different sets of tasks.
These subunits of artifacts vary independently of one another and may be
combined in numerous ways.
These five factors were later refined into functionally distinct toolkits by
Binford (1968:52). They are (abbreviated here):
• Activities carried out around the home-based manufacturing of
secondary tools and hide finishing. Borers, becs, end scrapers, burins,
and naturally backed knives.
• Hunting and butchering tools. Points and side scrapers of all types.
• Food preparation. Backed knives, naturally backed knives, end
notched pieces, typical and atypical Levallois flakes, and unretouched
blades.
• Processing of plant material. Denticulates and notched tools, scrapers
with abrupt retouch, raclettes, truncated flakes.
• Specialised hunting and butchering. Elongated Mousterian points,
discs, scrapers on the ventral surface, typical burins, and unretouched
blades.
The obvious differences in approach between Bordes’ cultural assemblages
and the Binfords’ multivariate factors can be summarised as typological
frequencies versus functional groupings. Bordes’ assemblages were purely
determined by the frequency of particular tool types (mainly handaxes),
whereas the Binfords had grouped tool types and given them a functional
interpretation (Binford and Binford 1966:244).
Another disputed difference in approach emerged later, when Bordes used
an analysis of reindeer teeth and antlers from Mousterian sites to conclude
13
Toolkits and utility in Australian lithics
that they were permanently occupied all year round (Bordes 1961:806;
Bordes and Bordes 1970:66). Binford’s toolkit variability rationale was based
upon his view of the character of Mousterian life, which relied on mobility as a
major survival strategy (Binford 1973). Mobility assumes seasonal activities
at different places, ensuring the need for a more complex set of tools related
to the spatial and temporal variations (Binford 1973:238). Despite Binford’s
resolute denial that Mousterian culture could affect assemblage variability,
Bordes never denied the existence of toolkits as a component of lithic
assemblages (Bordes 1961:804). It seems that their differences lay in their
theories about Mousterian life, with Bordes advocating assemblage variation
due to changing cultures, living sedentary lifestyles, but responding to
seasonal and environmental changes, whereas Binford saw a culturally
continuous group, perpetually on the move, and responding to environmental
changes which reflected their varying toolkits (Binford 1973:227).
Lewis Binford went on to study contemporary Nunamiut people in Alaska
during 47 hunting trips (Binford 1977). The data collected from these trips
demonstrated the important principle that archaeological remains refer
directly to the organisation of behaviour, with Binford drawing correlations
between which tools were taken on hunting trips, the distance travelled, and
the amount of time spent away from home.
The distribution, association between, and relative frequencies of tools are
greatly affected by the character of the technological organisation. No simple
equation between tool and task, or frequency and population is possible.
Before one can make meaningful statements as to the significance of patterns
of observed variability in the archaeological record, one must consider the
causal determinants of the patterning (Binford 1977:36).
Binford was referring to temporal patterning as a result of technological
organisation and the importance of the activity for which a tool is used. This
relates to theories he developed around expedient and curated technologies,
in which toolkits are generally composed of curated tools. This concept,
which was developed during Binford’s ethnographic research with the
Nunamiut suggested that expedient tools (non toolkit types) were discarded
at a higher rate than curated tools, consequently forming the majority of
14
Toolkits and utility in Australian lithics
debris at lithic assemblage sites (Binford 1977; 1979). Bordes (1977:36), on
the other hand, described a generalised Mousterian level of evolution as
being defined by a basic flake toolkit to which more specialised tools could be
added to make more specialised toolkits which, in his opinion, reflected
cultural differences over time. In this sense ‘toolkits” are not used to define
functions or tasks, but as cultural markers; although he concluded that if each
cultural group had different toolkits then they must have had different
technologies and economies (Bordes 1977:39). In some cases it is “the
proliferation of only one tool which highlights the individuality of the
assemblage” (Bordes 1977:39).
Five years later Binford (1983:147) defined his meanings of ‘toolkit’, ‘activity’,
and ‘activity areas’:
A toolkit is a set of tools used in the execution of a task.
An activity is an integrated set of tasks, generally performed in a temporal
sequence and in an uninterrupted fashion.
Activity areas are places, facilities, or surfaces where technological, social, or
ritual activities occur.
We can readily imagine activities which make use of a number of toolkits and,
conversely, different activities, which make use of one or more identical
toolkits.
Binford refers once again to the task-specific nature of toolkits, but then
introduces some flexibility into the concept in saying that identical toolkits can
be used for multiple tasks. Function in this sense seems to be dependent
upon the interactive nature of the tool and its user, and not necessarily
specific to one task (Binford 1989:185).
Binford used anthropological observation to determine the layout of activity
areas and associated toolkits (Binford 1983:149), but it has been argued that
these concepts were not useful for archaeologists because toolkits were not
abandoned at their place of use (Simek 1984:3; Binford 1989:234) which
15
Toolkits and utility in Australian lithics
could be the reason why others have suggested that he never offered any
empirical evidence for their existence (Dibble 1991:240). Some have argued
that his analysis was misguided, because not only had he misinterpreted the
statistics, but more importantly because he had used data (Bordes’ types)
which were never intended to be analysed as functional types (Grace
1989:155). Independent use-wear analyses on Middle Palaeolithic
assemblages by Beyries (1988) and Dibble and Rolland (1992) revealed that
the tool types described by Binford and Binford had no direct relationship to
their use (Holdaway and Stern 2004:51). A high number of tested artefacts
showed no use at all, and most of the used tools, regardless of type, had
been used to work wood, although a few had been used for butchery and
hunting (Beyries 1988:214). Further analysis of stone tool edge resharpening
by Dibble and Rolland (1992) revealed that many of Bordes’ types were
actually the same tool in different stages of the resharpening, and therefore
not different cultural types at all.
However, whilst evidence against the existence of archaeological toolkits was
building in some areas, other experimental analysis on toolkit composition
was adding legitimacy to the concept. In the late 1970s Cahen, Keeley and
Noten (1979) re-fitted a late Upper Palaeolithic stone assemblage at Meer,
Belgium, in order to analyse toolkits within lithic assemblages. This consisted
of reassembling various artefacts, including tools, flakes, and fragments that
had been knapped from the same block at the same time, with the outcome
of being able to calculate more accurately the ratio between ‘expedient’
artefacts (opportunistic implements made for immediate tasks) and ‘curated’
ones (tools kept for many tasks) (Cahen et al. 1979:662). Their argument that
curated artefacts are tools commonly used together, thus representing a
toolkit (Cahen et al. 1979:662) was dependent upon Binford’s (1973)
‘functional’ argument which was developed during his research with
Nunamiut communities (Binford 1977; 1979). Cahen et al’s (1979) refitting
demonstrated several stages of utilisation, transformation (reduction) and
series preparation for specialised activities in particular locations. It also
showed that the predominant artefacts at the site were expedient tools,
suggesting that assemblages in the Upper Palaeolithic contained toolkits that
were not necessarily comprised of only curated tools (Cahen et al 1979:671).
16
Toolkits and utility in Australian lithics
Banks (2009) used cast resin models of assemblages from Solutré in France
to determine if use-wear on experimental models could be matched to the
originals. Solutré is a complex site that covers a range of eras including
Aurignacian (around 35,000 bp), Gravettian (28,000 bp), Solutrean (17,000
bp), and Magdalanian (12000 bp), and each layer had a different tool
assemblage structure. He conducted analysis of use-wear signatures to
determine how artefacts were used, and how the composition of curated
toolkits are structured over time. Banks concluded that:
Curated toolkits are dominated by tools that can perform a variety of tasks, are
easily and efficiently maintained, are made in advance of their anticipated use,
are transported from place to place over the course of their life, and may well
be recycled into functionally different tools prior to their eventual discard. One
must keep in mind that a toolkit may have started out as one that was
composed of small functionally specific tools. However the actual realised
needs and uses of the toolkit may not have mirrored those that were originally
anticipated, and functionally specific tools had to be used in an improvised
manner in some scenarios over a long period of time (Banks 2009:31).
In this sense the toolkit is a flexible entity, which started life as a planned
grouping of functionally specific tools, but was superseded by tools that were
capable of being multi-functional or improvisational. Banks (2009:31)
hypothesised that ‘If a curated toolkit was to be abandoned or lost shortly
after its creation, and recovered archaeologically, then one might be able to
demonstrate that Kuhn's predictions are accurate’.
Kuhn (1994) used ideas about efficient tool design to predict how mobility
affects the kind of toolkit a person took with them. He argued that mobile
populations needed tools that were not only light and efficient to transport but
were also multifunctional, having a ratio of size to effectiveness (mass/utility)
suggesting they should be small and versatile (Kuhn 1994). He proposed that
a common behavior of mobile populations is the need for expediently
manufactured and curated tools, including cores, that can be cached at
frequently visited places (Kuhn 1994:427). Hewitt’s cache from Arcoona
Plateau, and Hiscock’s from the Boulia District of Queensland, are two of a
17
Toolkits and utility in Australian lithics
few caches that have been reported by ethnologists and archaeologists in
Australia (Hewitt 1976; Hiscock 1988:67). Both caches fulfill Banks’ (2009)
scenario of being cached hoards that contain unused artefacts, and therefore
reveal stages of manufacture which Hiscock claims are often lost to
archaeologists who find mainly used artefacts (Hiscock 1988:60).
2.4 Australian toolkits
The function-specific toolkit hypothesis put forward by the Binfords in 1966
coincided with a new approach to the problem of archaeological
interpretation in the 1960s which became known as processual archaeology,
where quantitative techniques, scientific method and anthropology were
applied to archaeological data (Willey and Phillips 1958; Binford 1962).
Processual archaeology and the toolkit had found their way into Australian
archaeology by the 1970s (Gould 1977:163).
Nowhere in Australian ethnographic literature is there a description of an
Aboriginal toolkit; one reason possibly being that the term ‘toolkit’ did not
appear to be in the vocabulary of archaeologists and ethnographers until the
1970s. Norman Tindale (1965:161) used ‘stock-in-trade’ as a collective noun
for an assemblage, and Howchin oscillated between ‘stock in trade’ and
‘outfit’ (Howchin 1934:v). Aiston (1928:130) regarded tools such as the knife,
the pirrie, the scraper, and the tuhla as part of ‘a series’ that showed an
evolution from the simple flake. We only see the term toolkit as a reference to
stone tool assemblages being used in archaeological texts after Bordes/
Binford debate in the 1960s. One prime example of this is McCarthy’s book
on Australian Aboriginal stone implements. The 1967 edition does not
mention the toolkit concept, as it was probably written before the critical 1966
Binford paper, but by 1976 a paragraph describing a toolkit used by the
Aborigines in Western Australia had been included (McCarthy 1976:19).
McCarthy’s (1976:19) description of the toolkit used by Western Desert
Aborigines included “a crude handaxe for detaching slabs from trees and for
rough shaping of them into weapons and other objects; a hafted stone chisel
18
Toolkits and utility in Australian lithics
for general woodworking; large hafted knives for domestic use; wooden
wedges and stone hammers; and two sacred tools used by men – a hafted
flake for engraving spearthrowers and ritual boards, and a knife for body
scars and circumcision”. This ‘kit’, which covers both everyday domestic, and
occasional specific tasks, would not fit into any of Binford’s function specific
toolkit categories, begging the question – is it a toolkit, or just the range of
technologies available to Aboriginal desert people as perceived by a
European?
Balfour (1951) describes a much more specialised mobile kit that is carried
by the men of the Worora tribe from the Kimberley district of Western
Australia when they go hunting. A wallet made of paper bark and lined with
bird down contained a hardwood stick, a small piece of sandstone, four
pointed kangaroo bones, a partially made biface point, a number of
completed spear points, a sinew, gum from the bloodwood tree, and string
made from the inner bark of a boabab tree (Balfour 1951:273). This is a
maintenance kit for repairing spears and sharpening the stone tips damaged
during a hunting trip. Mulvaney (1969:69, 1975:74) briefly described
Aboriginal hunting spears, spear throwers and boomerangs as “a small but
remarkably adaptable tool-kit which consisted chiefly of wooden implements”,
and by defining weapons as tools severely stretches the definition of what is
a tool.
This kind of personal kit, which an Aboriginal person would carry on a hunting
trip, has been of interest in countries such as America and Australia where
ethnographic accounts of lithic artefact use and manufacture have been
recorded. Despite this, ethnographic accounts of Indigenous toolkits as
discrete entities are not common. Balfour’s (1951) short but succinct
description of a hunter’s spear maintenance kit is one of the clearest
examples of personal gear in Australia. Another study that may help us
understand the paucity of personal toolkit recordings is Hiscock’s (2004)
compelling observations of Slippery and Billy’s stone knapping process in
Amaroo, Central Australia. Hiscock concluded that an explanation for
assemblage variation in Australia is still an archaeological challenge (Hiscock
2004:76), because knappers like Slippery and Billy initially make flakes, not
19
Toolkits and utility in Australian lithics
predefined tools, and therefore not toolkits. Toolkits must, in this case, be
made later.
The other challenge for later researchers has been how to fit European toolkit
theory into their work on Aboriginal Australians. The ‘new toolkit’ (Johnson
1979:144 cited in Hiscock 1994:271) of the mid-Holocene was defined by the
introduction of a range of small tools such as points, backed blades, and
tulas commonly known as the ‘small tool tradition’. This toolkit included
elements that have been contentiously interpreted as non-functional or
stylistic, as well as being more efficient in food procurement (Mulvaney 1977;
Hiscock 1994:272). Jones (1977:202) described the Tasmanian toolkit as the
simplest in the world, being the ‘irreducible minimum for the long term
survival of a human society in Australian conditions’. This toolkit also included
non-tools such as spears, throwing sticks and trapping devices, as well as
grinding and pounding stones. Cane’s (1984:276) interviews with Aboriginal
people of the Western Desert about their stone artefacts proved to him that
unretouched flakes, of which 98.9% were considered rubbish, were an
unimportant component of the Aboriginal toolkit. Law (2005:92) used Kuhn's
(1990) geometric index of unifacial (stone tool) reduction (GIUR) and
perimeter-reduction index (PRI) to characterise provisioning strategies of late
Holocene populations as risk reducing: mobile foraging groups utilised a
mobile toolkit consisting of backed artefacts, tulas, points, and high quality
retouched flakes. Hiscock and Attenbrow (2005b:139) have suggested
specific stone engineering strategies that prolong the usefulness of the
material at hand as ways of enhancing the readiness of the toolkit.
The above examples demonstrate that the toolkit in Australian archaeology is
a flexible concept. Despite the term ‘toolkit’ being traditionally used to
describe an organised, linked, and specialised set of tools that have been
produced for specific tasks, each of these applications of the toolkit concept
are quite different, suggesting a range of ideas and functions which reflect
both specific and general use.
20
Toolkits and utility in Australian lithics
Toolkits have been used to describe stone artefact assemblages in Europe
(Binford and Binford 1966; Freeman 1966; Bordes 1977; Banks 2009) that
have been found in stratified layers, correlated to clear temporal and cultural
phases, and can therefore arguably be used as representative models for
these phases. In Australia the issues are not so clear. Many Australian
assemblages are surface scatters that may have built up over long periods of
time (Mulvaney and Kamminga 1999:19), and therefore do not represent
clear temporal and cultural phases. The toolkits that Binford saw in the
Mousterian assemblages in the 1960s are not the same toolkits that
archaeologists have identified in the Australian desert (Binford 1973:244).
Gould’s (1971:167) comparison of Australian stone artefact types against
similar Mousterian types showed some similar use-wear patterns on some
scrapers, and some marked differences on other types, and so given the
difference in environments the outcome was inconclusive.
The difference in Australia, and one of the reasons that researchers from
overseas such as Gould and Hayden came to Australia, was that stone tools
were being used by Australian Indigenous populations for a number of years
after colonisation, and consequently the opportunity to study stone tool use
first-hand through ethnography was available; but herein lies a problem.
Ethnographic insights only relate to the state of stone tool technology since
the arrival of Europeans, whereas most of the collected archaeological
assemblages date back much earlier. Ethnographic descriptions, as well as
having a Eurocentric bias (Sheridan 1979:12) also tend to describe a highly
flexible and unstructured approach to tool use and function (eg. Gould et al
1971:163. Hayden 1977:178), whereas archaeologists who study reduction
sequences in stone tool assemblages see knapping events as being highly
structured (Moore 2003:24). Another factor is that many Australian Aboriginal
people were seasonally mobile (Hiscock 2008;200, Gould 1980), and so the
notion of an Australian “toolkit” has to include both sedentary and nonsedentary behavior, and become something that is portable, flexible, and
therefore multifunctional (Hiscock 1994, 2006), whereas Mousterian
assemblages came from sedentary sites that were occupied almost all year
round (Lev et al 2005:475; Bordes 1970:65).
21
Toolkits and utility in Australian lithics
2.5 Conclusion
The concept of archaeological toolkits that define function within stone
artefact assemblages emerged from Europe in the mid 1960s from a
Mousterian site occupied by Neanderthals for several thousand years.
Binford and Binford’s (1966, 1968) thesis was that variability within those
assemblages could be described as ‘functional variability’, which they
formulated by measuring the frequency of tool types within each factor
(toolkit). The term ‘toolkit’ was subsequently adopted in many parts of the
archaeological world, and arrived in Australia in the mid 1970s, where it has
been used to describe a range of attributes relating to Aboriginal stone
artefact studies ever since. In the ten years it took to arrive here the original
concept had evolved into a term that encapsulated all tool types regardless of
function, material, or gender. Ethnographic accounts of stone tool
manufacture and use in Australia have emphasised a flexible approach by
Aboriginal people which allowed for artefacts to have multiple functions. The
flexible nature of lithic use in Australia, and the increased understanding of
stone reduction methods that document both form and function variation
during the life of an artefact, means that Australian lithics do not fit
comfortably into the functionally-specific toolkits as described by the Binfords.
The Binfords’ initial analysis of toolkit structure was achieved through the use
of multivariate statistics, not through looking at individual artefacts. Their
characteristation of toolkits as ‘subunits of artefacts which vary independently
of one another and may be combined in numerous ways’ (Binford and
Binford 1966:292) indicates that there are no set parameters that can be
applied to toolkit recognition, and that sophisticated computations may be
needed to find them. This is possibly the reason that toolkits have been seen
as large-scale elements that can be used broadly to paint big picture
theories.
Use-wear and residue analysis has been widely used to reduce the functional
ambiguity of artefacts, in so doing demonstrating the misuse of functional
22
Toolkits and utility in Australian lithics
typologies, and in one case discrediting toolkit theories (Beyries 1988). The
idea that functional, personal toolkits are comprised mainly of curated
artefacts, suggests that they will not normally occur in discard or expedient
assemblages (Binford 1979:269), or be found at activity sites because their
residual usability was valued for future use (Simek 1984:3; Odess and Rasic
2007:691). It is therefore not surprising that everyday toolkits which are
elusive in the archaeological record have had little critical attention given to
them. One such assemblage that could come under the category of a
personal toolkit is the Lake Hanson cache.
23
Toolkits and utility in Australian lithics
3. The Lake Hanson cache
3.1 Introduction
The stone artefact material collected by Hewitt was within a limited radius of
Woomera where he worked as a technical officer. Woomera is almost
centrally located on the quartzite gibber plain known as the Arcoona Plateau,
and the isolation of the Woomera region made the area ideal for the
development of a defence institution. Both the Lake Hanson cache site and
the Mungappie sites are located within the geological region of the Arcoona
Plateau, and therefore had comparable characteristics. This chapter
describes the study area from where the artefacts were collected, both
geographically, geologically and socially.
3.2 The Woomera Natural History Society
Woomera was established as a rocket range in 1947 to accomodate Britain’s
defence plans. Various missiles and rockets were launched towards the north
west, where numbers of Aboriginial people were known to live in reserves,
particularly in the Westen Desert (Morton 1989:80).
In 1947 the Minister for Defence established the Native Patrol Officer (NPO)
scheme, to protect Aboriginal people during the proposed rocket tests at the
range (Morton 1989:81). Walter MacDougall, who was recruited as the first
Native Patrol Officer, was also one of the founding members of the Woomera
Natural History Society (Morton 1989:81; Edwards 2000:201). During the
British nuclear tests at Emu Field and Maralinga after 1953 it was
McDougall’s job to discourage Aboriginal people from traveling across the
desert (Morton 1989:81; Gara 2010).
One of Ron Hewitt’s (Figure 3.1) earliest memories of arriving in Woomera in
the mid 1960s was seeing two Aboriginal men walking across the desert in
the direction of Port Augusta, and never seeing them again (Hewitt 2010 pers
comm). In fact after that he never saw another Aboriginal person around
Woomera, and like many Europeans in Australia during the mid-twentieth
24
Toolkits and utility in Australian lithics
century, witnessed the inevitable expansion of settlers at the expense of
traditional Aboriginal culture (Griffiths 1996:188; McGrath 1991:121). The
fascination with the relics of the ‘primitive’ stone age, which were plentiful in
the area, became the focus of interest for those who established the
Woomera Natural History Society (Gorman 2009:137). They conducted
monthly excursions to various locations to learn more about the local
environment, including wild flowers, sand dune age, ochre pits etc (Gibber
Gabber 1968:52) and to collect geological and cultural objects (Gorman
2009:137). They also invited visiting speakers such as palaeobotanist Heinz
Amstberg, S.A. Museum curator Bob Edwards, and archaeologist/
anthropologist Norman Tindale to give lectures and slide shows to the local
community and members (Gibber Gabber 1969:43, 1968:52; Gorman
2009:137). Hewitt was an active member of the society while he lived in
Woomera, adding to the collection of stone artefacts that became the centre
of the annual exhibitions that were initially held at the United Church until a
more permanent space was made available at the rear of the new clubrooms
in August 1969 (Gibber Gabber 1969:44).
In 1965 the South Australia Aborigines and Historic Relics Preservation Act
was enacted, and the role of Protector of Relics was established. This role
was given to the Director of the South Australian Museum to whom local
wardens under the scheme reported (Allen 1975:70). The Woomera Natural
History Society had members who were wardens who sought to protect
Aboriginal sites and relics from vandalism and tourist damage, such as the
Eucolo Creek engravings (Gibber Gabber 8.5.1969:11). Hewitt was
appointed a warden by the Governor in 1971 and remained active in this role
until he left Woomera a couple of year later to live in Adelaide (Hewitt 2010
pers. comm).
25
Toolkits and utility in Australian lithics
Figure 3.1 Ron Hewitt holding an artefact (yodda) that was later given to Norman
Tindale after his talk to the Woomera Natural History Society. Photo unknown.
3.3 The Mungappie sites
The sites around Mungappie Creek fall into four areas identified by Hewitt (as
written on the artifacts). They are Mungappie Creek, Mungappie Creek North,
Mungappie Hut, and Mungappie Swamp.
Mungappie Creek runs into Lake Koolymilka, which is approximately 35km
north west of Woomera, and Mungappie Hut is about 10km south of the lake.
The Mungappie Creek sites are north of the Hut heading towards Lake
Koolymilka. Mungappie Swamp is near a perennial lake that lies
approximately 7kms east of Lake Koolymilka. Lake Hanson is situated about
23 km west of Lake Koolymilka (Figure 3.2).
26
Toolkits and utility in Australian lithics
N
0
!
10
20 km
Figure 3.2 Mungappie sites and Lake Hanson cache site. From Google Earth
images.
The sites, on the traditional land of the Kokatha people, are in an area of
South Australia known as the Arcoona Plateau. The Plateau is roughly
triangular in shape and extends from Lake Torrens in the east to the northern
tip of Lake Hanson in the west, and from the northern shores of Island
Lagoon in the south to Andamooka at the northern tip of Lake Torrens in the
north, with Woomera roughly in the center. Geologically, the Arcoona Plateau
is an undisturbed platform of Upper Proterozoic sediments, capped by a
resistant quartzite called the Arcoona Quartzite Member, which is an evengrained sandstone quartzite of fine to medium quartz grains (Johns
1968:21,31). The plateau is composed of strata of sandstones, quartzites,
shales, clays and grits, with a cap rock stratum of hard rock called Arcoona
quartzite, and a surface of deep red clay soil (Johns 1968:21; Hewitt 1978:6).
This surface soil has been protected by a cover of lag gravels, pebbles and
boulders called gibbers, and a strong growth of saltbush and other arid land
species (Hewitt 1978:6; Bourman and Milnes 1985:229). The gibber areas in
the region are from mixed origins, with the most common being quartzite and
silcrete clasts (Bourman and Milnes 1985:229).
27
Toolkits and utility in Australian lithics
Hewitt described the sites as seasonal hunting bases, some 12 to 15
kilometers apart, that could easily be reached in an unhurried day’s walk
(Hewitt 1978:12). The Mungappie Creek campsites were located at the base
of sheltered sandhills near the watercourse, which were densely covered with
implements exposed by the action of wind (Hewitt pers comm 2010).
In a letter to the senior curator of anthropology at the South Australian
Museum on 16.8.1977, Hewitt described his comparative classification of the
archaeological site sizes, ranging from large, medium to small. Large sites
covered an area up to 100 metres or more across with heavy concentrations
of implements so dense as that ‘it is impossible not to walk on them’, with
water availability offering nearly permanent occupation (Hewitt 1978:11).
Medium to smaller sites have a location with a poorer water supply
suggesting a more seasonal use. Small sites are often isolated and generally
devoid of all but very short-term water, suggesting only casual short term
occupational possibilities. Mungappie Hut is a medium size site and the
Mungappie Creek sites are small but dense.
3.4 The Lake Hanson cache and Hewitt’s collecting
Lake Hanson is a salt lake situated on the western edge of the Arcoona
Plateau approximately 55 kms north-west of Woomera and 220 km northwest of Port Augusta (Figure 3.2). In October of 1970 the Woomera Natural
History Society held a field excursion to the Lake Hanson area, and it was
during the trip that Ron Hewitt, a fastidious collector and researcher of stone
artefacts for some years, discovered the cache of stone artefacts on the
upper part of a sand dune which was set apart from the main sites in the area
(Hewitt 1976:16; Figure 3.3). Hewitt’s (1976:29) assessment of the cache
was:
The implement types represented in this hoard are semi-discoidal adzes,
discoidal adzes, microadzes, gouges, engravers and burins, which together
make up a comprehensive woodworking kit. These implements are all
effective woodworking tools when used in a hafted state, with the possible
28
Toolkits and utility in Australian lithics
exceptions of the burin and engraver types, some of which may have been
used unhafted. A distinction is made between these implements and those
commonly referred to as distal scrapers, end scrapers, women’s knives and
other forms of implements used for light woodworking and other general
purpose functions, but having differences in the method of use, the nature
of the working edge and the shape of the lower surface.
The reasoning behind Hewitt’s decision to classify the cache as a
woodworking kit seems to lie in the unambiguous woodworking function
attributed to these implements by writers like McCarthy, who Hewitt
referenced. The current thinking on other tools at the time such as scrapers
and knives that he mentions, was that they were thought to have been used
for other purposes as well as woodworking. All 105 items of the cache were
found within a one square metre area, some still neatly piled up and partially
covered by sand (Hewitt 1976:20).
Figure 3.3 The Lake Hanson cache. R. Hewitt
Hiscock (1988:67) has defined an archaeological cache as ‘the underground
and concealed storage of objects’, with the qualification that the objects may
not necessarily have re-use potential. The excavation of a cache of fifty tulas
at Mucklandama Creek in western Queensland’s Boulia District, was
however, interpreted by Hiscock as a cache ‘destined for barter’ (Hiscock
1988:60). The ochre trade route from Cape York and Queensland to the
South Australian Lake Eyre region (Jones 2007:353) passed through the
29
Toolkits and utility in Australian lithics
Boulia District, making it a ceremonial and trading centre (Hiscock 1988:66).
Hewitt also considered the possibility of the Lake Hanson cache being a
‘trade parcel’ but suggested that they could also be a ‘tribal stockpile’ (Hewitt
1976:48). The homogeneity of the material (not common in other
assemblages in the area, and of unknown provenance) and the minimal
reduction of most of the implements implied to Hewitt that the cache was the
work of one person (Hewitt 1976:31).
The Lake Hanson cache became a part of Hewitt’s private collection, which
he kept until 2004, when it was handed back to Aboriginal Liaison Officer with
the Defence at Woomera, Andrew Starkey from Kokotha Mula.
3.5 Hewitt’s classifications
Hewitt did not publish a recording sheet for the cache, but has listed the
artefact and material types in his 1976 paper (summarized in Table 3.1).
Hewitt classified the cache into twelve morphological types. Hewitt used the
term adze to refer to tula, which has been put in brackets here.
Artefact Types
Qty
% of total
A. Tula: Slug
3
2.9
B. Tula: Part Reduced
6
5.7
C. Tula: Semi-discoidal
43
41.1
D. Tula: Asymmetric
6
5.7
E. Tula: Discoidal
2
1.9
F. Tula: Side Trimmed
5
4.7
G. Tula: Blanks
9
8.6
H. Micro Tula
4
3.8
I. Micro Tula Blank
2
1.9
J. Burin/Engraver
14
13.3
K. Nondescript Flake
5
4.7
L. Nondescript Lump
6
5.7
105
100
Total
Table 3.1 Cache artefact types and quantities. Codes, types and quantities are from
Hewitt (1976:19).
30
Toolkits and utility in Australian lithics
Hewitt’s descriptions and coding of the cache tool types have been
abbreviated here (in italics) from his 1976 paper:
A. Adze (tula) Slugs
Reduced to the point that no percussion point is evident. Two of the three
have been trimmed to make a graving point (Hewitt 1976:20).
B. Adze (tula): Partially reduced
The cutting edge has been trimmed for re-use in five. Two have one end
trimmed as an end scraper. One has been broken across the middle, and
then trimmed on the broken edge for re-use. One has had the platform
trimmed for reverse re-use (Hewitt 1976:20).
C. Adze (tula): Semi discoidal
The platform and /or point of percussion is diametrically opposed to the
worked edge, and the worked edge is curved and extends to points that
agree roughly with the extremities across the width of the flake. More than
two thirds have cortex material on some part of the flake, and on several the
cortex extends to within a millimetre of the trimmed edge. Three or four are in
need of sharpening, the rest are considered to be “ready for use”. Two in the
group have the shape of small cores (Hewitt 1976:20-23).
D. Adze (tula): Asymmetric
Similar characteristics to type C except that the working edge is about 45º to
the platform. One has two working edges, the third has cortex. One has
cortex over most of the dorsal and the platform, and has a fossil leaf imprint
(Hewitt 1976:23).
E. Adze (tula) Tula: Discoidal
Nearly circular with a delicately prepared working edge that extends for three
quarters of the perimeter. 5.8 cm is the largest in the group. High standard of
workmanship (Hewitt 1976:23; Figure 3.4).
31
Toolkits and utility in Australian lithics
F. Adze (tula): Side Trimmed
Four have both margins trimmed but not the front (distal). Two of the four
have a hinge fracture at the front edge produced by the initial flake removal.
One of the other two is triangular in section and has been trimmed at both
sides. One specimen is the shape of a large convex/concave implement but
with cortex on the entire dorsal face and platform with both sides trimmed to
“conventional adzing edges”. The last in this group has cortex on two faces
either of which could have been the platform. (Hewitt 1976:25; Figure 3.4).
G. Adze (tula): Untrimmed Blanks
All have cortex on the dorsal face and seven of these have cortex on the
platform too. Several have a few small experimental flakes removed, and
some seem to be discoidal (Hewitt 1976:25).
H. Micro adze (tula)
A variety of forms, but all have micro scraper or chisel-like edges. One is a
stout double-edged micro-adze with the worked edges terminating in a point
at one end and a flat butt at the other. No point of percussion is evident. One
is a conventional micro adze shape except that one third of the working face
is untrimmed and appears unfinished. The platform is not opposite the
working face. Two have a micro chisel edge on the thick edge of a small flake
(Hewitt 1976:25).
I. Micro adze (tula) blank
One has a convex ventral surface and a concave dorsal face with secondary
retouching on both lateral edges. The other is of a similar shape to the first
but quite thick and has no secondary work, but has a curved sharp edge with
a convex lower face that would already be suitable for use (Hewitt 1976:28).
32
Toolkits and utility in Australian lithics
Figure 3.4 Hewitt’s drawings. Fig 12 and 13 are type f ‘side trimmed’ and Fig 14 and
15 are type e ‘discoidal’.
J. Burin/engraver
All have at least one sharp projection or point with evidence of spalling and
appear suitable for engraving work. Retouched concave edges on several.
Four pairs fit together. The fracture surfaces fitted together exactly leaving no
doubt that each pair was originally a single piece (Hewitt 1976:28).
33
Toolkits and utility in Australian lithics
K. Nondescript Flake
Two show no secondary work and no potential as working tool. The other two
have sharp edges and have some attrition indicating that they may have
been used for cutting (Hewitt 1976:29).
I. Nondescript Lump
Some have sharp edges and points, but no methodical trimming. Use is
possible in some cases (Hewitt 1976:29).
It is possible that the last two categories were not actually associated with the
cache as Hewitt has alluded to some other artefacts in the immediate area as
nondescript flakes and lumps.
Hewitt noted the high frequency of oolitic chert in the cache (47%) compared
to other local assemblages, and considered the nature of the stone as a
prime factor in its selection for the tulas in particular.
3.6 Cache artefact woodworking types
Stone tools used for working wood have been described ethnographically
since the nineteenth century (i.e. Roth 1897; Spencer and Gillen 1899) and
during the twentieth century (Roth 1904; Spencer and Gillen 1904; Horne
and Aiston 1924; Basedow 1925; Howchin 1934; Tindale 1932; McCarthy
1946; Mitchell 1949; Campbell and Barrett 1958; Hayden 1977, 1979; Cane
1984; Bates 1985). These observations indicate a range of products being
made, such as spears, spearthowers, clubs, throwing sticks including
boomerangs, bowls, digging sticks, hafted tools, shields, and ritual objects.
The Lake Hanson woodworking types identified by Hewitt, and recorded
ethnographically, include tulas, micro tulas, burin/engravers and flakes, which
are described here. Other types not present in the Lake Hanson cache but
known to be at other sites in the Arcoona Plateau region are also discussed
in a similar manner (type codes in brackets).
34
Toolkits and utility in Australian lithics
Tulas (A - G)
Tulas were the archetypal Aboriginal Australian woodworking tool, developed
during the mid-Holocene in arid zones for working hardwood (McCarthy
1977:260). This tool survived as one of the main woodworking implements
until the arrival of steel with Europeans (O’Connell 1977:277). The tula is one
part of a composite tool made up of three components: the stone cutting
implement, the shaped wooden handle, and the resin to cement the stone to
the handle, which when combined has been called Koondi Tuhla, (Figure 3.5)
from the Central Australian Wonkonguru name Tuhla (Horne and Aiston
1924:29; Sheridan 1979:10; Hiscock 1997). The cutting end of the tool, the
chipped stone flake, has particular characteristics that make it distinct and
therefore easily recognisable in the archaeological record. When found
archaeologically, it has often been sharpened many times and has taken on a
different form to the original tula flake that would have been initially inserted
into the resin (Figure 3.6). This heavily reduced form is called a tula slug
(Gould 1977:165; Holdaway 2004:255).
Figure 3.5 Koondi Tuhla from Yuendumu in Central Australia, hafted by Minjena on
22.8.1951. Collected by N. Tindale. J. Hayward.
Hiscock and Attenbrow (2005:110) have described the tula slug as a wide
flake with a pronounced bulbar curve on the ventral surface, that is
characterised by its short length created by low angle retouch onto the dorsal
across the entire width of the flake at the distal end. It is not uncommon for
35
Toolkits and utility in Australian lithics
tula slugs to be inverted in the haft and the dorsal margin of the striking
platform used as a scraping edge, or sometimes rehafted with the point
exposed and used as graving tools (Cane 1984:277).
Many early ethnographers have described the tula as an adze (Roth
1897:101; Spencer and Gillen 1899:594; Horne and Aiston 1924:80;
Campbell and Edwards 1966:193) and even more recently (Gould 1977;
O’Connell 1977:277; Cane 1984:277). This is unusual because tulas do not
resemble the traditional European or New Guinea adze, which has a blade
that is fixed at almost right angles to the handle and is used standing up
(Sheridan 1979:16). Hiscock (1997) states that most prehistoric specimens
are thought to have been used primarily for wood scraping, but advises that
the functional description of this class as "tula adze" is to be avoided, since
adzing (rough primary shaping) was almost certainly a minor function. The
fact that the tula has also been described as functioning as a chisel, a gouge,
and scraper indicates that it was principally a woodworking tool, but has also
been recorded as having other functions (Roth 1897:101; Horne and Aiston
1924:90). For example, when hafted to the woomera or spear thrower
(O’Connell 1977:277) the tula flake then became a butchering tool (Spencer
and Gillen 1899:584; Campbell and Barrett 1958b).
Figure 3.6 Tula reduction sequence (Campbell and Edwards 1966). Top, blanks –
bottom, slugs.
36
Toolkits and utility in Australian lithics
Micro Tulas (H - I)
Micro adzes or micro tulas are morphologically similar to tulas but smaller.
The examples from Lake Hanson illustrated by Hewitt are no more than 25
mm long. Some archaeologists categorise them as scrapers as they conform
more closely to thumbnail scrapers than tulas (Holdaway and Stern
2007:255) but they could have been used for either purpose (Gould 1977).
Hewitt’s division of these two types is confusing, and his drawings do not aid
clarification.
Burin/Engravers (J)
Burin type artefacts have been recorded in European, Middle Eastern and
Indian sites dating back to Middle Palaeolithic periods. Debénath and Dibble
(1993:36) have described them as artefacts “where a lateral (or occasionally,
a transverse) edge is removed with one or more blows that follow the ridge
formed by the sharp margin of the flake resulting in a surface that is more or
less perpendicular to the original flat surface of the flake”. The removal of one
or two the burin spalls leaves a very sharp edge on the burin margin where
the flake scars meet (Holdaway and Stern 2007:241). The primary function of
these types in the western Australian desert was for engraving of sacred
hardwood ceremonial boards, which were engraved at sacred locations and
hidden at special places in the landscape (Hayden 1979:168; Cane
1984:236).
Flakes (K)
Flakes are generally considered as the product of knapping events. Flakes
have been variously described by ethnologists as instant tools that can be
opportunistically picked up from debitage to be used for a task at hand as
expedient tools (Binford 1973; Hayden 1977:179; Cahen et al 1979:161;
Gould 1980:72) in order to distinguish them from curated tools (Kuhn 1994;
Shott 2009:94).
Ethnologists have reported that the process of flake selection for further
retouch can be both socially complex, and surprising (Hiscock 2004:72;
Hayden 1977:179). Cane’s (1992) research with Indigenous people in the
37
Toolkits and utility in Australian lithics
Western Desert produced some interesting results. Of 4478 unretouched
flakes shown to Indigenous people that Cane had collected from a site 1.15%
(51) were considered usable, but only 0.4% were actually sharp enough for
butchering or scarring, and 0.35% were good enough to be hafted (Cane
1992:19). Hayden (1977:179), on the other hand, was surprised how often
these flakes were utilised. Others have considered the flake as an important
part of stone assemblages, being deliberately manufactured items for use
(Hiscock 1998:261; Dibble and McPherron 2006:777).
3.7 Other woodworking artefact types
The tool types described above relate directly to the cache; the types
described below were not components in the Lake Hanson woodworking kit,
and have not therefore been described by Hewitt, but have been recorded as
woodworking tools by ethnologists and researchers. Others, such as pirrie
points were not still in use when Europeans arrived in Australia so have not
been ethnographically described, but have since been associated with
woodwork through use-wear and residue analysis.
Points
This research project distinguishes points from pirrie points, as they appear
to be morphologically, and functionally different. Pirries have a specific
morphology (described in next section), and have not been recorded
ethnographically, whereas points came in a variety of forms and were still in
use when Europeans arrived.
Howchin (1934:62) noticed that points vary in size from a fine needle-like
form to the thickness of a heavy nail. He described analogous ‘awls or
‘borers’ from Europe which are “usually thick and clumsy implements” but
thought that these fine points were unique to Australia. Campbell and
Edwards (1966:197) also describe similar types they called asymmetric
points, which are usually long and thin, triangular in section, and with a sharp
point at their distal end (Figure 3.7).
38
Toolkits and utility in Australian lithics
Figure 3.7 Points Photograph R. Hewitt.
There are ethnographic descriptions of skins being made flexible, after they
have been scraped and cleaned, by scoring across the fleshy side with sharp
stones classified as points (Howchin 1934:63), but what type of points were
used for that purpose is unclear. Ethnographic films made by the Board for
Anthropological Research (BAR 1935) in the Western Australian Warburton
Ranges shows a split spearthrower being repaired using a stone point
opportunistically fashioned with a couple of skillful chips to form a point on
the distal end of a flat flake. Holes were made with the point through the side
of the broken spearthrower, which was then ‘sewn’ together with kangaroo
sinew through the holes.
Pirrie Points
These stones are tear or leaf-shaped, triangular in section, longer than broad
with a sharp point at the distal end, and are unifacially trimmed. The dorsal
face can be finely chipped or pressure flaked to produce a curved surface
that has the dorsal ridges removed (Figure 3.8). In some examples the
proximal end is worked to produce a thin rounded margin where the flaking
can extend onto the ventral face. They were first reported, named and
illustrated by Horne and Aiston in 1924, although they had probably first been
collected from the Arcoona Plateau region in the late nineteenth century
where they have been found in large numbers (Campbell and Edwards
39
Toolkits and utility in Australian lithics
1966:199). Those from the northern parts of South Australia were termed
‘Eyrean’, being distinct from the southern smaller versions termed ‘Fulham’
by Campbell and Noone in 1943 (Campbell and Edwards 1966:199).
Mulvaney (1975:221) speculates that these items could have been spear
heads. Horne and Aiston (1924; Aiston 1928:125) had already asserted that
pirries were never used as spear tips, but when hafted in slim handles were
used for fine engraving on timber artefacts like women’s carrying vessels or
boomerangs, and making holes by drilling. Howchin agreed, stating that in
the Adelaide region at least all the spears had wooden tips (Howchin
1934:64), and Spencer and Gillen (1899:575) had already recorded that
spears with stone tips were not encountered south of Powell Creek near
Tenant creek in the Northern Territory (Lourandos 1997:291).
Plate 3.8 Pirrie Point with broken tip. (Holdaway and Stern 2004:268)
Scrapers
Tools described as scrapers or end scrapers are found commonly in
prehistoric sites worldwide (Andresfky 1998:193). Bordes included a number
of different types of scrapers or racloirs in his Lower and Middle Palaeolithic
typology, which were classified by the amount or positioning of retouch on the
40
Toolkits and utility in Australian lithics
flake edge (Dibble 1987:109). In Australia, scrapers, along with axes,
choppers, and pounder/grinders, have maintained a continuity of tradition
from the Pleistocene through to the arrival of Europeans (Pretty 1977:10).
Howchin’s (1934:52) Australian typologies have differentiated between a
number of scraper types including: straight, horseshoe, circular, side, duckbill, kite shaped, oblique, flat, hollow, pointed or scalloped, and thumb
scrapers. This differentiation, which is based solely on the morphology of the
flake, does not reflect Bordes’ Palaeolithic scraper classes according to
retouch, but might have similarities in that both Bordes’ and Howchin’s types
could be trans-functional tools in differing stages of reduction (Dibble
1987:110). The reduction process which changed both the size, angle and
form of the working edge could have affected their functional capabilities, but
according to Hayden (1977:182) they would still have been associated with
woodworking.
Campbell and Edwards (1966:188) reduced central and southern Australian
scraper classifications to two basic types, asymmetric and symmetric, without
referring them as scrapers. The asymmetric types refer to side trimmed and
irregular flakes which include concave faces and points on nosed forms,
whereas the symmetric types are broad or long distal trimmed flakes
(Campbell and Edwards 1966:189-191. (Figure 3.9). They also acknowledge
a miscellaneous group of trimmed flakes.
Figure 3.9 Long distal scrapers. Photograph R. Hewitt.
41
Toolkits and utility in Australian lithics
Geometric Microliths/ Backed Artefacts
Geometric microliths have been associated with cultural industries in South
Africa from 70,000 years ago (Wymer 1982:223; McBrearty and Stringer
2007:793) Excavations in European Magdalenian sites have produced a
relative abundance of these microliths which were to become common and
typical in the Mesolithic (Bordes 1973:223). Backed artefact assemblages
similar to Australian types have been found in India and Ceylon (Sri Lanka)
from 7000 BP (McCarthy 1977:255). They are known to have appeared in
north east Australia in the late Pleistocene and become abundant in the
south east from 3500 – 1500 BP (Lourandos 1997:289; Robertson et al
2009:296).
Geometric microliths are backed at either one or both ends, and sometimes
on one lateral edge. They include crescent, segment, triangle and trapezoid
shapes and are classified as being less than 3 cm or less on their maximum
dimension (Campbell and Edwards 1966:204; Holdaway and Stern 2007:262
Figure 3.10).
The ‘backing’ which describes the artefact is a form of steep retouch that is
formed on one or two margins of the stone, usually on the opposing side to
the working edge or chord, which acts as either a strengthening element to
the blade, or a preparation of the stone to receive hafting resin (McCarthy
1976:45; Holdaway and Stern 2004:159). Because manufacture of these
items had ceased before European settlement in Australia, no ethnographic
record of their use exists, and so their function has been speculative
(Robertson et al 2009:296). McCarthy (1967, 1976) thought that they must
have been points for throwing spears, and many archaeologists have
adopted that view (Mulvaney 1975:108; Kamminga 1982).
42
Toolkits and utility in Australian lithics
Figure 3.10 Geometric Microliths Trapezoid, triangular, and crescent forms. R.
Hewitt.
Robertson et al’s (2009) use-wear analysis research of these artefacts
refutes this idea, arguing that the microscopic indications are that they were
used predominantly for working wood or other craftwork, whereas Boot’s
(1993) residue analysis suggest that a small percentage were hafted and
used for cutting plant material (Boot 1993:11; Robertson et al 2009). This
outcome is contrary to previous speculation that these implements, which
were produced in great abundance, were primarily used in ritual and
ceremonial activities and on human flesh (Robertson et al 2009:297).
Burren
Burrens were sometimes categorised as adzes in that they were hafted, and
reduced to a slug form (Kamminga 1978:28; Cane 1992:22; Holdaway and
Stern 2007:257). They differ from tulas, whose retouch is mainly on the distal
end, by having steep retouch on one or both of the laterals (Holdaway and
Stern 2007:257). The lateral margins were continually used and sharpened,
43
Toolkits and utility in Australian lithics
resulting in a slug which was then sharpened at each end, hafted and used
for wood boring or graving (McCarthy 1976:31).
Hand Choppers
Campbell and Edward (1966:184) have described these implements in
Australia as large trimmed flakes which have the size and weight suggestive
of a hand held chopping tool. Whilst these types of implements have been
described as heavy duty butchering tools (Holdaway and Stern 2004:53) they
have also been ethnographically recorded in primary woodworking activities
such as the initial shaping of artefacts like spears, gigging sticks, bowls, and
spearthrowers (Hayden 1977: 183; 1979:30-37).
3.8 Conclusion
Hewitt’s definition of the Lake Hanson cache as a woodworking kit was
based on artefact types that when hafted would be effective woodworking
tools. The range of different tula types, being archetypal Aboriginal
woodworking tools (McCarthy 1977:260), from unworked blanks to slugs
found in the cache does suggest that the assemblage was a kit with a
specific function, not solely for trade, but made for a special purpose from
materials that were not commonly found in lithic scatters around the camps.
The tool types from A – J, including all tulas, micro tulas and engravers, are
classic examples of Binford’s curated tools types with a high degree of
production investment, and relatively long use life (Binford 1973:215;
1979:269). The storage and re-use of tools in anticipation of future needs is
also a characteristic of curation that can be directly applied to the Lake
Hanson cache (Shott and Nelson 2008:24). The high quantity of tula blanks
and partially reduced tulas in the cache fulfills Bank’s (2009:31) requirements
for a curated toolkit by having tools that are efficiently maintained, are made
in advance of their anticipated use, and may well be recycled into functionally
different tools prior to their eventual discard. The tula slugs had signs of
reworking and could still have been functional as fine engraving tools. The
variants in this woodworking toolkit appear to be the nondescript flakes and
44
Toolkits and utility in Australian lithics
lumps. The flakes could have been opportunistic or expedient tools, as Hewitt
recorded some use wear on a few of them, and the lumps did also show
signs of attrition. The other possibility is that both the lumps were a source of
good quality raw material for further shaping, or not a part of the actual cache
at all, but rather unassociated lithic scatter exposed by the wind. The fact that
they were similar materials to the cache, such as oolitic chert, jasper and
translucent chalcedony, favours the former.
45
Toolkits and utility in Australian lithics
4. Methodology
4.1 Introduction
Hewitt’s collections from various sites near Woomera are numerous, and
occupy many museum boxes at the South Australian Museum’s archaeology
collection. After an initial search through some of the boxes it was decided to
concentrate on one site for analysis; a site with quantities that were
manageable within the time frame of the research, and a site that had a good
range of artefact types. Another criteria was to work with a site that was in
close proximity to the Lake Hanson cache, but there were no collections on
the museum database attributed to Hewitt from the Lake Hanson area.
Mungappie Creek and associated sites were chosen because they were only
35kms from Lake Hanson at the closest point and seemed to have a variety
of type concentrations which provided the opportunity for comparisons with
each other as well as with the Lake Hanson cache. The artifacts were
collected by Hewitt from August 1966 to June 1971.
All of the Mungappie site collection boxes were sorted through to list the
accession numbers that were represented in each box. These were recorded
on an Excel database sheet that had been copied from the museum’s
database.
Volunteers who helped with recording the artefacts at the museum were
briefed on attribute standards such as colour, type and material in order to
maintain recording consistency.
4.2 Recording systems
The recording form for the Mungappie sites was designed to facilitate not
only a basic comparison with Hewitt’s Lake Hanson cache, but also to
accommodate a more complex level of analysis (Figure 4.1). The four main
attributes that Hewitt had used for the cache were used as the basic criteria;
they included artefact type, material, colour, and retouch location. Other
46
Toolkits and utility in Australian lithics
attributes were added to allow for a more detailed analysis of the collections,
and to make projected analyses about the Lake Hanson cache based upon
the main four criteria.
Figure 4.1 Recording sheet for Mungappie site assemblages.
47
Toolkits and utility in Australian lithics
The fields on the recording form included: site name, box number (the
museum collection box number that appears on the exterior of the box for
ease of storage and identification, and relates to the database), collector (in
this case Hewitt), date (date of recording), and recorder (the name of the
person recording). The recording form itself has eighteen attributes for each
artefact as well as a comments column:
1. Accession number.
These numbers were given to the artefact collections by the South Australian
Museum when they were formally accessioned. They were originally
recorded into a ledger and then onto microfilm; some were then transferred
onto a database.
2. Site name.
The site names had been identified by Hewitt on each stone, and were
recorded on the sheets as he had written them on the artefacts.
3. Artefact number.
A unique artefact number was given to each artefact as it was recorded.
These numbers are used for identification purposes in analysis, drawings and
photographs.
4. Raw material.
Hewitt gave a list of raw material types that he had found in the region
(Hewitt 1976:17). This list was used and added to with other materials that
were found amongst the collection. These materials are:
Oolitic chert
Translucent chalcedony
Silcrete
Banded jasper
Chert
Jasper
Quartzite
Quartz
Indurated Mudstone
Porcellanite
Rock crystal
48
Toolkits and utility in Australian lithics
Hewitt’s definitions of the stone types found in the Lake Hanson cache may
be idiosyncratic, but they correspond well to textbook definitions, and were
descriptive of the types found within the Mungappie assemblages. They have
therefore been used for the purpose of classifying the assemblage artefacts,
and for direct comparison with the cache.
A jeweler’s 30x magnifying glass with a light was used to help identify some
of the material types, which could only be confirmed by close examination.
The following is a list of material types encountered and identification
methods used.
Oolitic chert is distinguished by its “regular sized rounded concretions that
loosely float in a chalcedonic matrix” (Hewitt 1976:17) that can easily be seen
under magnification. Ooliths are spherical rock particles which have grown by
accretion around an inorganic or organic nucleus (Whitten and Brooks
1972:321). Translucent chalcedony (as classified by Hewitt), a mix of fine
grained quartz and the silica moganite (Heaney and Post 1992:441) differs
from most other crypto-crystalline material in the assemblages in that it has
at least some translucent properties as well as, in some cases, faint oolitic
markings (Hewitt 1976:18). Because of this it was sometimes difficult to
distinguish this material from oolitic chert and banded jasper, which can also
have oolitic markings.
The banded jasper in the assemblages were recognised by the rich brown
bands of colour that laced through it. Jaspers tended to be recognised by a
single pure colour throughout the artefact and have a very soapy feel, and
are a fibrous form of micro-crystalline quartz and silica, as is chalcedony
(Howchin 1934:15; Heaney and Post 1992:441; Luedtke 1992:6). The chert,
being a non-fibrous form of micro-crystalline quartz (Heaney and Post
1992:441; Luedtke 1992:6) was predominately dark grey and sometimes had
a chalky cortex.
Silcrete forms much of the duricrust that covers the Arcoona plateau in the
form of gibbers (Stephens 1964:1407; Bourman and Milnes 1985:229). The
silcrete clasts that make up the duricrust contain loosely packed water49
Toolkits and utility in Australian lithics
rounded quartz grains which were deposited in the silica when it was laid
down (Stephens 1966:497), making it difficult to sometimes differentiate it
from quartzite, which is also prominent in the region (Johns 1968:31,
Stephens 1966:497; Smale 1973:1084; Hewitt 1976:16). Some coarse
grained varieties of quartzite also occur, with rounded grains that can look
much like silcrete, and the very fined grained varieties had a waxy
chalcedonic feel with no visible quartz grains. Hewitt had omitted quartzite,
quartz, and rock crystal from his list, because they were not represented in
the cache. However, in the other assemblages, quartzite came in a range of
colours from light grey to light brown/green with some pink and reds. Quartz
is not common in the region and was not expected in the assemblages,
whereas rock crystal, a glass like transparent form of pure quartz (Whitten
and Brooks 1972:412) that occurs as a sparse natural material amongst the
gibbers, has been recorded at other campsites (Hewitt 1978:17).
The indurated mudstones described by Hewitt were from a limonite or
haematite base (Hewitt 1976:18), which includes various amorphous iron
oxides providing the red colour in ochres (Whitten and Brooks 1972:269).
Porcellanite, which Hewitt has described as a very fine quartzite (Hewitt
1976:18) was not recognised in the assemblage, but could have been some
of the very fine quartzites that were occasionally recorded, or could also have
fallen into the jasper-chalcedony classification; all being forms of microcrystalline silica (Heaney and Post 1992:441).
5. Colour.
Colour was estimated within a range that seemed to cover most materials in
the assemblages (Table 4.1). Artefacts that had more than one colour, or
seemed to be a mix of two colours, were given a dominant colour, and a
secondary colour ie. a buff stone that had a light grey tendency would be
coded bu/lg, or a principally dark grey chert with white bands would be dg/w.
50
Toolkits and utility in Australian lithics
Table 4.1. Colour range codes for stone types at Mungappie sites.
It was intended to use this information for artefact identification purposes,
and for analysis of materials that might have come from specific sources. For
example, there was a red-brown quartzite quarry located near Pernatty
Lagoon 20 km SE of Woomera (Johns 1968:27). Colour could also relate to
heat treatment, which can result in a systematic colour alteration although it
is not a reliable method for detecting heat treatment (Rowney and White
1997:653).
6. Artefact type.
As Hewitt’s toolkit types are the model for this research, the occurrence or
non-occurrence of these types in each Mungappie site will determine the
presence of toolkit members there. Hewitt’s classifications and typologies,
which are based on McCarthy (1967), have been used as the comparative
types. Hewitt had described the toolkit artefacts from A - L in enough detail to
be able to use the descriptions to code types from Mungappie Creek sites
and do a direct comparison with the cache.
51
Toolkits and utility in Australian lithics
Some artefacts fall between types, or are not easily reconciled using the
classification system. For example, Hewitt has a burin/engraver type – a
classification which describes both the technology used to make the item and
it’s supposed function. This can be confusing; even though it is documented
that burins were used for specialised engraving purposes (Cane 1984:236),
many classifications have engravers which are also described as pirries or
points (Aiston 1928:125). Hewitt’s drawings of this type are ambiguous and
do not seem to illustrate classic burin technology, where spalls are taken off
to make a sharp point, or pirries, which have a quite distinctive morphology.
This raised the issue of whether Hewitt’s burins are really burins even though
he describes them as having burin spalls (Hewitt 1976:28). It was decided to
include only artefacts with burin spalls in this category and to make a
separate category for other engraving type tools such as pirries and points.
Points were classified as flakes with point-type terminations or manufactured
point protrusions on the lateral edges, but were clearly different to the more
clearly defined pirries that demonstrated a classic uniform tear/leaf shape
morphology.
Hewitt’s nondescript flakes from Lake Hanson have been discussed in the
previous chapter. The flakes recorded in the Mungappie assemblages were
52
Toolkits and utility in Australian lithics
identified by their lack of obvious retouch. Some had signs of use-wear, or
use damage, and these were recorded accordingly. Some were broken or
had no obvious proximal or distal ends, and were recorded as medial flake
pieces (Kp).
7 - 9. Artefact dimensions.
The length, width, and thickness measurements were based upon the
method of measuring the length on the ventral surface from the PFA (Point of
Force Application - see item 12) to the mid point of the termination. The width
of the flake was measured at right angles to its axis at the mid point, and the
thickness being measured at the same axis as the width (Andrefsky 1998:97;
Smith and Burke 2004:214). Where there is no distinct termination or
platform (which is quite common due to retouch) the length is measured
along the flake path. Many flakes have varying widths due to their irregular
shape, and therefore the width of a flake at the mid point, or the medial axis
(Holdaway and Stern 2004:139) is considered to be a medium width that is
less than the widest point, but more than the narrowest; likewise with
thickness. The equipment used for measuring the artefacts was a 150mm
vernier caliper capable of 0.02mm calibrations. All measurements were in
millimeters with 0.5 of a millimeter being the smallest calibration.
10 - 11. Platform size
Much information about manufacture and reduction history can be
determined from the platform (Holdaway and Stern 2004:120), but because
this project is not principally concerned with stone tool reduction sequences,
the basic information about its size only was recorded as it can relate to
some retouched tool types such as tulas. Platform width was measured at
the widest point between the lateral edges, and thickness was measured at
right angles to the width at the widest point between the ventral and dorsal
faces, which is commonly at the impact point (Andrefsky 1998:93 Holdaway
and Stern 2004:124).
53
Toolkits and utility in Australian lithics
12. The Point of Force Application (PFA)
The PFA is the point where the hammer was struck on the core platform to
detach a conchoidal flake. The PFA, which is characteristic of a Hertzian-type
initiation, is defined by a semicircular protrusion that is visible on the platform,
and which radiates down the ventral face as ring cracks to form a bulb of
percussion (Holdaway and Stern 2004:109 Cotterell and Kamminga
1983:685). The width of the PFA can be measured across its diameter on the
platform. The size of the PFA can have characteristic relationships to
material, hammer type, and initiation processes (Cotterell and Kamminga
1983:687). The PFA is often very small and in some cases not visible. In
these instances it was recorded as indeterminate.
13. Cortex.
The cortex is the outside skin of the stone or pebble that the artefact was
struck from, and accordingly cortex will be found mainly on the dorsal face
and proximal end of a flake, but can occur on the distal if the termination was
plunging towards the rock’s interior. The amount of cortex on the artefact
gives an indication where in the reduction sequence the artefact was. A
dorsal surface consisting entirely of cortex indicates the flake being struck at
the early stages of a knapping sequence, and consequently no cortex
indicates that the artefact was struck at a later stage nearer the centre of the
core. One way of measuring cortex is to divide the dorsal face into quadrants
and record its presence in the number of quadrants (Holdaway and Stern
2004:144). The diverse range of shapes and sizes of the artefacts in the
collections necessitated the need for the cortex on the dorsal face to be
estimated in 5% ordinals.
14. Termination
The termination of a flake is at the end opposite to the platform where the
flake last separates from the core. The four basic termination types were
recorded: feather, hinge, step, and plunging (Grace 1989:14; Andrefsky
1998:88; Cotterell and Kamminga 1987:685). If the termination had been
chipped off the artefact during the process of shaping and sharpening, it was
therefore not recordable.
54
Toolkits and utility in Australian lithics
15. Breakage
Flakes can be either intentionally or unintentionally broken by the maker,
accidentally broken during use, trampled by animals or people after discard,
damaged through agricultural processes, or naturally broken by heat or fire
(Hiscock 1985; Holdaway and Stern 2004:111). Artefacts that have either no
distal end or proximal end, or both, could have been broken transversally.
These fragments are called proximal, medial or distal flakes depending upon
which portion of the flake they represent. Flakes can also be broken
longitudinally and may therefore have both proximal and distal ends
(Holdaway and Stern 2004:169). Breakage is an important factor in large
assemblages which have high percentages of debitage if accurate statistical
representations are required. In this project where the assemblage has
already been collected and debitage is not recorded, breakage only affects
artefact identification.
16. Edge angles
Edge angles on stone tools have been used to describe many aspects of tool
use and manufacture including use life, where the retouch angle relates to
the amount of reduction (Kuhn 1990; Hiscock 1994; Collins 2007); and
function (Wilmsen 1968; Cantwell 1979; Collins 2008). A common theory is
that the working edge angle of a tool may be representative of the type of
work that the tool is used for, with a lower edge angle being better for cutting,
slicing or shaving, and a higher edge angle being more suited for scraping
(Wilmsen 1968:156; Grace 1989:93). Modern woodworking tools represent
these differences quite well. A general hand plane has a working bevel of
around 30º but is set in the plane body at between 45º-50º (Hayward
1973:38; Sheridan 1979:13; Kingshott 1992:66). Chisels come in a variety of
edge angles but the harder working mortise chisels would generally have a
higher angle of around 50º than finer finishing and carving tools which would
be around 30º-40º (Goodman 1964:197). Scrapers used for fine grain
finishing are used at an angle of around 80º-85º to the timber (Hayward
1973:119).
55
Toolkits and utility in Australian lithics
Dibble and Bernard (1980) have identified four methods for measuring stone
tool edge angles. Methods a) and b) use a polar coordinate graph (Burgess
and Kvamme 1978:482). In method a) the artefact edge, when seen in cross
section, is set against a polar coordinate graph which has angle increments
as it’s index to measure the angle. In method b) a clay impression of the
working edge is measured against the graph index. These two methods were
deemed less flexible, and accurate than the c) goniometer, and d) caliper
methods. The d) caliper method requires a specially modified set of calipers
with a calibrated depth gauge which measures two triangulated lengths of the
working edge which are then entered into a formula: ø = 2[Tan -1 (.5T ÷ D)] to
calculate the angle of the edge ø (Dibble and Bernard 1980:861). Given the
inaccuracy of methods a) and b) and the complexity of method d), it was
decide to use method c), the goniometer, for this project. A goniometer is an
engineering tool for measuring fine angles. It is similar to a protractor, but has
an additional moveable arm which moves against the protractor angle gauge
when the artefact is placed between it and the instrument base (Burgess and
Kvamme 1978:482). When used for stone tools it gives a reasonably good
indication of the angle even though there needs to be some user
interpretation due to the inconsistency of the stone surface. The middle of the
working edge, the apex, can generally be a sufficient indication of the overall
angle (Grace 1989:75) but in this instance (depending upon retouch extent)
two, three or four measurements were taken on the retouched section of a
working edge, or edges, to give an average measurement for the whole
working edge. When there was more than one working edge, all the working
edges of the principal retouch type were measured. The average of these
measurements was calculated to give a single composite angle for recording
purposes (Figure 4.2).
56
Toolkits and utility in Australian lithics
Figure 4.2 Edge angles were taken from three points of the retouched working
edge of the flake. J. Hayward
17. Retouch location
The retouch location was calculated by dividing the dorsal face of the flake
into four segments by drawing diagonal lines from top left corner to the
bottom right corner and visa versa for the other side (Figure 4.3). The
margins are named proximal, distal, left lateral, and right lateral. There are 15
possible combinations of the margins, which are coded from A – O (Table
4.2). Proximal retouch is only recorded on the ventral face. Retouch on the
dorsal face at the proximal is not recorded as this may be core overhang
preparation prior to flake removal, and has no influence on artefact
functionality or type (Dibble and Bernard 1980:92).
Figure 4.3 Ventral and Dorsal faces of flake showing retouch margin divisions. J.
Hayward.
57
Toolkits and utility in Australian lithics
Table 4.2 Showing retouch location combinations and retouch code used on the
recording form.
18. Retouch type and use wear
The retouch type column recorded the method used for retouching the
working edge. These were serrating, crushing, pressure, chipping, and
backing. Serrating is a notching that forms a series of small or fine
projections like a saw edge. Crushing is a shattering of the tool edge, either
by damage or attrition, as result of use wear rather than deliberate shaping
(McCarthy 1976;102; Hayden 1979:19; Holdaway and Stern 2007:167), and
so is not retouch as such, but does give an indication of intense use wear.
Pressure flaking is a method of retouch where pressure is applied to the
artefact by a tool often made of wood or bone, which results in a fine control
over the removal of flakes (Holdaway and Stern 2004:12). Chipping, in this
instance, is a generic term for a range of retouch which can either be steep
or flat, stepped or scalar, but is usually done with a hammer stone. Backing is
defined as steep, short retouch of more than 80º, which can be either uni- or
bi-directional shaping of the flake, usually found on the opposite edge to the
working edge, or chord, and is initiated from the ventral platform of the flake
(Hiscock 2004:97 Holdaway and Stern 2007:111,159). Flakes with more than
one type of retouch had all the retouch types recorded.
58
Toolkits and utility in Australian lithics
4.3 Data analysis
The data was taken from the handwritten recording sheets and entered into a
Apple spreadsheet programme called ‘Numbers 08’ which can be exported to
a Microsoft Excel format. The hand written sheets were the primary data
which did not change, and have been referred to a number of times when
checking the Excel sheets for discrepancies. ‘Sorts’ were used to arrange the
data according to attributes into separate worksheets, so that refined ‘sorts’
could be achieved. For example, type A artefacts were put into a separate
sheet which could be sorted for colour, size, retouch type, retouch location
etc. For most of the size attributes, mean (average) calculations were used
as this was used by Hewitt, and was therefore comparable. Percentage
calculations were used for artefact quantities because the ranges were too
high for accurate charting.
The data from each of these attributes were recorded in ordinals of 5 from 0 100, other than termination type, breakage, retouch location, and retouch
type, which were recorded by codes. Each ordinal range states the number
of artefacts that appear in that ordinal range. The average volume was
calculated using the length, width, and thickness data, which indicated
comparative volumes between artefacts. This information will be used to
calculate a reduction index for tula blanks and other tula types. Knowing how
much an artefact has been reduced indicates the amount of use it has had,
and how much potential use an artefact has (Shott and Nelson 2008; Shott
2009). These reduction calculations feed into theories related to ‘curation’
and ‘mobility’ (Cahen et al 1979; Kuhn 1994; Banks 2009).
4.4 Volumes of artefacts
The average volume of each artefact type from the Mungappie sites was
calculated using the length, width, and thickness data (Table #4 in Appendix
metrics). Even though this can never be an accurate calculation of the
volume of the artefact due to the amorphous nature of a flake, it does
represent a comparative measurement as all the flake’s volumes were
calculated using the same method. The artefacts were measured in
59
Toolkits and utility in Australian lithics
millimetres, but the volumes have been converted into cubic centimetres as a
standard volumetric measure throughout this project.
The volumes of all tulas from blanks (G. tula blanks) to fully utilised tulas (a.
tula slugs) from the all of the Mungappie sites were used to calculate a
reduction index for each tula, using Kuhn’s (1994) theory of the potential
utility of a core being based on its volume. The utility of a core is assessed on
its potential to produce fresh usable edges, with the larger core expressing
more potential than a smaller one (Kuhn 1994:429). The potential utility of a
tula is its ability to be resharpened a number of times, until it becomes
unusable. Tula slugs were used as the minimal usable, or irreducible, artefact
size to calculate utility, instead of an arbitrary figure of 3 as employed by
Kuhn (Kuhn 1994:430).
4.5 Conclusion
The principle reason for some of the recording methodologies used was to
allow for a direct comparison with the Lake Hanson toolkit cache by
reproducing the methods used by Hewitt in the 1970s, which were concerned
with artefact type, material, colour, and retouch location. Further attributes
were used to facilitate a more sophisticated analysis of the Mungappie
assemblages, which may not have necessarily related to the Lake Hanson
cache. Bivariate attribute comparisons were possible with ‘sort’ options that
could generate new tables; percentage calculations were generated with
formulas; averages, means, and ranges were automatic calculations; and
complex graphs could be generated directly from tables.
60
Toolkits and utility in Australian lithics
5. Results
5.1 Introduction
In this chapter, the Mungappie Creek sites are compared to the Lake Hanson
cache in order to determine the frequency that cache artefact and material
types appear. The total number of artefacts recorded from the Mungappie
sites was 1364. The number of artefacts types from each site was:
Mungappie Creek
866
63.5%
Mungappie Creek North
226
16.6%
Mungappie Hut
213
15.6%
Mungappie Swamp
59
!
!
4.3%
Attributes of the artefacts recorded including material type, colour, artefact
size/volume, cortex, termination, breakage, edge angles, retouch type, and
retouch location are compared and analysed. The artefact sequence of toolkit
types from A - L is used to match Hewitt’s sequence.
One of the aims of this thesis is to examine the composition of toolkit
structures compared to lithic assemblage structures. Hewitt’s Lake Hanson
cache is the model toolkit which can be tested against the four Mungappie
lithic assemblages. The four Mungappie sites as one entity increase the
sample size for toolkit type comparison. In cases where there are significant
differences with one or more of the Mungappie sites, this will be indicated.
This comparison focuses on toolkit types represented in the Lake Hanson
cache, and therefore non-toolkit types and non-woodworking types such as
hammerstones, and grinding stones are not included in the analysis. Whilst
they are important components of the Mungappie assemblages and other
toolkit types, they are not immediately relevant to the cache comparison.
61
Toolkits and utility in Australian lithics
5.2 Artefact types at Mungappie sites
The percentages of all artefact types from A – X recorded at the Mungappie
sites was calculated for each site and compared on schematic graphs (Figure
5.1). These graphs illustrate the distribution frequency of artefact types at
each site, and allow for an immediate comparison between sites. Lake
Hanson woodworking kit types are represented between A – L as recorded
by Hewitt; the sequence from M – X is arbitrary.
All sites had similar rates of scrapers, being P. scrapers 20 – 28%, and PL.
long scrapers around 10%. All sites also had medium percentages of W.
geometric backed artefacts with Mungappie Hut and Mungappie Creek North
having approximately 20%, and Mungappie Swamp and Mungappie Creek
having between 10 – 15%. In most other respects, the rates of artefact
frequencies differ from site to site.
Mungappie Swamp had the smallest sample size of the four sites (58
artefacts), and was only visited once by Hewitt – on 26/07/1970. It had no G.
tula blanks, B. part-reduced tulas, J. burin/engravers, and no K. unretouched
flakes but did have the highest percentage of A. tula slugs of all the sites, as
well as a high proportion of all scraper types and cores, and the highest
percentage of R. thumbnail discoidals. Unusually, it was the only site to have
no T. pirries and N. points.
62
Toolkits and utility in Australian lithics
Figure 5.1. Percentage distribution of artefact types for each of the Mungappie sites.
Woodworking kit types range from A - L.
63
Toolkits and utility in Australian lithics
Cores at each site
Core samples at each site were similar in percentage terms (Table 5.1).
Mungappie Creek had the largest range of material types including a rare
banded jasper and oolitic chert sample (Figure D.1). Unsurprisingly, quartzite
was the most common material at all sites. Mungappie Hut had the largest
cores (Figure D.9) with an average volume of 115cc with a range from 53 –
231cc, but had the lowest cortex count, except Mungappie Swamp which had
none. The cortex range was similar at all three sites.
Table 5.1 Cores at all Mungappie sites.
5.3 Toolkit types
The quantity of toolkit types A – L found at all the Mungappie sites was 319.
The numbers from each of the Mungappie sites are:
Mungappie Creek
204
63.9%
Mungappie Creek North
47
14.7%
Mungappie Hut
46
14.4%
Mungappie Swamp
22
6.9%
The only Mungappie site to have a full contingent of all toolkit types was
Mungappie Creek, which also had the highest sample size of artefacts (866)
64
Toolkits and utility in Australian lithics
collected from the four sites. The other three sites all had at least two
components of the kit not represented. All sites had samples of types A. tula
slugs, C. semi discoidal, D. asymmetric and E.discoidal. Mungappie Creek
had a small number of G. tula blanks and Mungappie Swamp had none. All
sites had more than 1% of H. micro adzes and I. micro adze blanks except
for Mungappie Hut, which had 0.5% of H. and no I. All sites had between
3-6% of K. unretouched flakes, except for Mungappie Swamp which had
none.
Comparison of Mungappie sites and Lake Hanson cache
A frequency graph of toolkit artefact types A – L from each Mungappie site
was overlaid on the Lake Hanson cache (Figure 5.2). The pattern of
woodworking types from each site show areas of difference against the
cache. No site matches the cache, but the overlaying areas indicate the
degree of difference that each site has. Each site also has artefact frequency
variations; the lower the sample number the less the artefact range, with
Munagappie Swamp showing distinct peaks and troughs. Mungappie Creek,
with the largest sample, has a similar make-up to the combined Mungappie
sites.
65
Toolkits and utility in Australian lithics
Comparison of ʻWoodworking Kitʼ Artefacts at Lake Hanson Cache and Mungappie Sites
A -!
B -!
C -!
D -!
E -!
F -!
G -!
H -!
I -!
J -!
K -!
L -!
A
Tula: slugs
Tula: part reduced
Tula: semi discoidal
Tula: asymmetric
Tula: discoidal
Tula: side trimmed
Tula: blank
Micro adze/chisel
Micro adze blank
Burin/engraver
Nondescript flake
Nondescript lump
L
B
40%
31
23
16
10
K
C
5
1
J
D
0
I
E
Lake Hanson Cache
F
H
Mungappie Sites
G
!
!
!
All Mungappie sites (319)
A
L
A
B
40%
L
31
31
23
16
10
5
1
K
J
B
40%
23
16
10
K
C
5
1
D
0
I
J
F
E
F
H
G
!
G
Mungappie Hut (46)!!
!
!
Mungappie Creek (204)
A
A
L
L
B
40%
31
D
0
I
F
Mungappie Creek North (47)!
D
0
E
F
H
G
!
J
C
I
E
H
23
16
10
5
1
K
C
16
10
5
1
J
B
40%
31
23
K
D
0
I
E
H
!
C
G
!
Mungappie Swamp (22)
Units are a percentage of ʻwoodworking kitʼ type artefacts from each site (quantities in brackets).
Figure 5.2 Comparison of ‘woodworking kit’ artefact distribution between the Lake
Hanson Cache and each Mungappie site.
66
Toolkits and utility in Australian lithics
Artefact quantities in the toolkit range A – L at the four Mungappie sites were
compared to the Lake Hanson cache (Figure 5.3). The cache has high
quantities of C. semi discoidal tulas, (41.1%) and G. tula blanks (8.6%),
which is not matched in the Mungappie sites, which have 10.6% and 1.6%
respectively. The other irregularity is the J. burin/engraver category where the
cache has 13.3% compared to 4.5% at all Mungappie sites.
Figure 5.3. Percentage comparison of ‘woodworking kit’ artefact types at Lake
Hanson and all Mungappie sites
Tula Types at all sites
The range of tulas in the cache, from blanks to slugs, represents all stages of
the reduction sequence; therefore the quantities of all tula types were
compared with the Munagappie sites (Table 5.2). The cache has a total of 9
type G. tula blanks, whereas the combined assemblages of 1364 artefacts
produced only 5 of the same type. Likewise with type C. semi-discoidal tulas:
the cache had 43, the assemblages had 33. The Mungappie sites had 50
type A. tula slugs (Figure D.1) compared to 3 in the cache. The lack of tula
67
Toolkits and utility in Australian lithics
types in two of the Mungappie site are apparent, with Mungappie Creek
North having no type F, and and Mungappie Swamp having no type G and
both with no type B. Both Mungappie Creek and Mungappie Hut had the full
range of tulas.
Tula types
Cache
Mungappie Mungappie Mungappie Mungappie Mungappie
sites
Creek
Creek North
Hut
Swamp
A. Tula slugs
3
50
35
4
5
6
B. Part
reduced
6
7
4
0
3
0
C. Semi
discoidal
43
33
16
7
7
3
D. Asymmetric
6
39
19
2
12
6
E. Discoidal
2
26
14
8
3
1
F. Side
trimmed
5
5
3
0
1
1
G. Blanks
9
5
2
2
1
0
Totals
74
165
93
23
32
17
Table 5.2 Comparison of tula type quantites between Lake Hanson Cache and the
Mungappie sites.
Mungappie Creek and the Cache
A direct comparison between Lake Hanson cache and Mungappie Creek (the
site that has all cache types) highlights one main criteria that differentiate
these two assemblages: artefact types. Mungappie Creek had nearly twice
the number of cache types as the cache (204/105); it did not, however, have
quantities of types such as G. blanks and C. semi discoidal tulas and J. burin/
engravers to match the cache, but had much higher amounts of flakes
(35.2%) and tulas slugs (17.2%) than the cache (4.8% and 2.8%) (Table 5.3).
The other criteria which gives the cache its distinctive composition is its raw
materials.
68
Toolkits and utility in Australian lithics
MC
total
A. Tula: slugs
B. Tula: part reduced
C. Tula: semi discoidal
D. Tula: asymmetric
E. Tula: discoidal.
F. Tula: side trimmed
G. Tula: blank
H. Micro adze/chisel
I. Micro adze blank
J. Burin/engraver
K. Nondescript flake
L. Nondescript lump
Totals
%
CACHE
A. Tula: slugs
B. Tula: part reduced
C. Tula: semi discoidal
D. Tula: asymmetric
E. Tula: discoidal.
F. Tula: side trimmed
G. Tula: blank
H. Micro adze/chisel
I. Micro adze blank
J. Burin/engraver
K. Nondescript flake
L. Nondescript lump
Totals
%
OC
%
35
4
23
19
14
3
2
6
11
9
72
6
17.2
2
11.3
9.3
6.9
1.5
1
2.9
5.4
4.4
35.2
2.9
1
204
100
100
8
4
3
6
43
6
2
5
9
4
2
14
5
6
2.8
5.7
40.8
5.7
1.9
4.8
8.7
3.8
1.9
13.4
4.8
5.7
105
100
100
TC
S
2
3
1
4
4
4
1
1
1
1
1
1
3
2
1
3
8
4
2
31
4
1
3
5
2
1
49
46.6
14
13.3
C
J
17
1
6
3
3
1
1
1
3
4
9
2
M
9
3
3
1
8
2
33
16
3
1
50
25
27
13
1
2
2
3
1
1
1
1
1
6
1
3
2.9
11
10.5
QU
P
3
2
11
11
7
2
1
3
10
2
1
2
1
1
8
2
2
BJ
2
1
37
1
1
0
74
37
0
0
1
1
1
2
4
2
1
1
1
2
2
1
13
12.4
5
4.7
8
7.6
0
0
2
1.9
Table 5.3 Mungappie Creek and Lake Hanson cache - a comparison between
materials and types.
5.4 Raw materials
The range of raw materials recorded at all the sites was:
Oolitic chert
Translucent chalcedony
Silcrete
Banded Jasper
Chert
Jasper
Quartzite
Mudstone
Porcellanite
There was no porcellanite recorded at the Mungappie site, but there were
some items in the cache from this material. The cache had no quartzite
artefacts.
The Mungappie sites had high percentages of the common quartzite and
silcrete materials found in the environment (58% and 15% respectively). The
less easily obtained mudstone and jasper are present, but in lesser amounts
69
Toolkits and utility in Australian lithics
(0.15% and 6.1%, Figure 5.4). The antithesis of this is found in the Lake
Hanson cache, which has no quartzite and very few silcrete artefacts, but has
a very high percentage of oolithic chert (47%), and moderate amounts of
translucent chalcedony (12.4%), banded chert (10.5%), and mudstone
(7.6%). The only area of commonality is shared by chert (13%) and jasper
(5.5%; Figure 5.4).
Figure 5.4 Percentage comparison of material types at Lake Hanson cache and all
Mungappie sites.
Material distribution for each of the Mungappie sites shows similar patterning
(Figure 5.5), except for higher quantities of silcrete and chert at Mungappie
Creek North and Mungappie Swamp. The high percentage of quartzite at
each site is apparent, with Mungappie Hut having the highest (76.3%) and
Mungappie Swamp the lowest (29%). Quartzite at Mungappie Hut was used
principally for P. scrapers, W. microliths, and T. pirrie points (32%, 17% and
5.6%). The only types not made from quartzite at any of the sites were G.
tula blanks and H. micro tulas which were made from mainly cherts and
70
Toolkits and utility in Australian lithics
jaspers. All the Mungappie sites had a low incidence of oolitic chert and high
incidence of quartzite.
Figure 5.5 Materials for each Mungappie site and Lake Hanson cache.
Raw Material colour
The colour of each artefact was recorded (appendix database sheets 8 and
3), but did not appear to have any significant correlations with raw material or
artefact types. Further research into quarry sites in the area would be useful
in determining the provenance of particular materials and colours. In brief,
the most common material was a light grey quartzite (21.7%) which was used
for 10.5% of all scrapers and 3.5% of pirrie points, followed by buff quartzite
(7%) which was also used for 3.5% of scrapers and 2.7% of geometric
microliths, and buff silcrete (3.8%) which made up 2.1% of scrapers,
predominantly at Mungappie Creek. The cache had no quartzite and only 3
silcrete artefacts (Appendix database sheet 6).
71
Toolkits and utility in Australian lithics
Raw material composition of each toolkit type
Comparing the material make up of each toolkit type from A – L for each site
with the cache shows that some material types which are used for a wide
range of tool types in the cache are hardly used in any of the Mungappie
sites (Figure 5.6). Banded Jasper, translucent chalcedony and oolitic chert
combined are used for 11 out the 12 types at the cache, but these material
are found in only small pockets at any of the other sites. The J. burin/
engravers, composed of either banded jasper and translucent chalcedony,
are not repeated at any of the sites including Mungappie Creek which has a
reasonable number of the types. The large amounts of types C. G. and J.
cannot be found at any of the other sites.
72
Toolkits and utility in Australian lithics
Mungappie Creek
70
60
Lake Hanson Cache
50
50
40
40
30
30
20
20
10
10
0
0
A B C D E F G H
I
J K KP L
A B C D E F G H
I
J K KP L
Mungappie Hut
20
Mungappie Creek North
10
10
0
A B C D E F G H
I
J K KP L
Mungappie Swamp
10
0
A B C D E F G H
I
J K KP L
0
A B C D E F G H I
MATERIALS
OC - Oolitic Chert
TC - Translucent Chalcedony
S - Silcrete
BJ - Banded Jasper
C - Chert
J - Jasper
M - Mudstone
P - Porcellanite
QU - Quartzite
OC
S
C
QU
P
TC
BJ
J
M
J K KP L
TOOL TYPES
ATula: slugs
BTula: part reduced
CTula: semi discoidal
DTula: asymmetric
ETula: discoidal
FTula: side trimmed
GTula: blank
HMicro adze/chisel
IMicro adze blank
JBurin/engraver
KNondescript flake
KP Flake piece
LNondescript lump
Figure 5.6 Raw material composition of each toolkit type from A - L at Lake Hanson
Cache and all the Mungappie sites. Actual artefact numbers are compared.
73
Toolkits and utility in Australian lithics
5.5 Utility Units of tulas
The Utility Units of tulas is a measurement of the amount of potential utility
that each tula type has. This is calculated by using the volume (length x width
x thickness) of each artefact.
Hewitt’s cache of 105 artefacts was composed of 70% tulas, with seven
different types being recorded. These types ranged from tula blanks, a flake
that had been shaped to a classic tula format, but not trimmed or retouched
for use (Andrefsky 1998:161), to tula slugs which were at the end of their
usable life (Gould 1977:165; Holdaway 2004:255). This range of artefacts,
from G. tula blanks to A. tula slug, also recorded in the Mungappie sites
assemblages, represents the only full reduction sequence that can be
compared between the sites and the cache. The tulas were given the same
typology as the Hewitt types so that the two could be compared.
The volume of the tula blanks averaged at 22.34cc, with a range of 14.2cc to
35.42cc (Table 5.4). The tula slugs had a volumetric average of 5.34cc, with
a range of 1.56cc to 13.57cc. By subtracting the average volume of the A.
tula slug 5.34cc from the G. tula blank 22.34cc, 17cc becomes the average
Utility Unit (UUs) for the overall life of a tula. Subtracting the average
volumes of the other tula types from the slug volume gives a utility unit for
each tula type. Utility Units in this context represent the amount of
resharpening potential that a tool possesses, with larger volumed tools
having more potential that smaller ones (Kuhn 1994:429).
74
Toolkits and utility in Australian lithics
Table 5.4 Volumetric averages of tula types from all Mungappie sites in cubic
centimetres (cc).
Other tula types have a range of Utility Units (UUs) depending upon the
amount of reduction they have already received, with type B. part reduced
still having 16.21 UUs remaining whereas type F. side trimmed have only
5.72 units of utility left.
The comparative volumes of each of the tula types generated a reduction
sequence that was calculated based on reducing volumes from G. tula
blanks to A. tula slugs. The reduction sequence was: G. tula blanks, B. part
reduced, D. asymmetric, C. semi discoidal, E. discoidal, F. side trimmed, to
A. slugs (Figure 5.7). The expected sequence would have E. discoidal
changing places with D. asymmetric. The inclusion of an uncharacteristically
high volumed D at 126.63cc has skewed this expectation.
The amount of usable life, or Utility Units (UUs) in each artefact was
calculated on volume and followed the volume reduction sequence. Tula
slugs are assumed to have been discarded with zero UUs, therefore all other
75
Toolkits and utility in Australian lithics
tula types will have a UU of 5.34 less than their volume. These figures
demonstrate that the tula blanks that were identified in the assemblages did
have the largest volumes that could have been reduced into the various other
tula types.
30.0
22.5
15.0
7.5
0
A. !
B. !
C. !
D.!
E. !
F. !
G. !
G
B
slug
part reduced
semi discoidal
asymmetric
discoidal
side trimmed
blank
D
C
Volume
E
F
A
Usable Units
Figure 5.7 Volume and Utility Units of tulas at Mungappie sites.
The volume of each tula type is used to calculate its remaining Utility Units that
follow the reduction sequence from G. blanks to A. slugs.
76
Toolkits and utility in Australian lithics
Mungappie Hut - 122.34cc
All Mungappie Sites - 75.43cc
32 cc
24
16 cc
16
8
8
0
0
G
B
A. !
B. !
C. !
D.!
E. !
F. !
G. !
D
C
E
F
G
A
B
D
slug
part reduced
semi discoidal
asymmetric
discoidal
side trimmed
blank
C
E
F
A
UU index
Mungappie Creek - 67.39cc
Mungappie Creek North - 44.94cc
16
8
0
G
B
D
C
E
F
A
G
B
D
C
E
F
A
Figure 5.8. Utility Units for All Mungappie site compared to individual Mungappie
sites.
Total UU index for each site is displayed next to name.
77
Toolkits and utility in Australian lithics
Utility Units of each Mungappie site.
The Utility Units of each site, other than Mungappie Swamp which did not
have a complete A – G sequence, were calculated using the same formula
for the combined sites of subtracting the average volume of A. tula slugs from
the average volume of each tula type (Figure 5.8). The reduction sequences
for these three sites varied significantly from the combined sites due, in part,
to two sites each lacking a type.
Mungappie Swamp had no G. tula blanks or B. part reduced, making a
complete reduction sequence impossible. Mungappie Hut had no F. side
trimmed, but had the highest volume tulas with E. discoidal the highest.
Mungappie Creek North had no B. part reduced. The variations in these
graphs are representative of the average volume of all tula types at each site.
Mungappie Hut has some large G. D. and E. tulas.
Table 5.5 Numbers of tulas at each Mungappie site.
The Utility Unit graphs in Figure 5.8 were generated from these numbers of
tulas for each site (Table 5.5). Large numbers of all tulas do not necessarily
represent high UUs; it is the number of high value UU artefacts that that
generates the high UUs at a site. ie. slugs have zero UUs, therefore a high
quantity of slugs at Mungappie Creek does not raise the overall UUs.
78
Toolkits and utility in Australian lithics
Mungappie Hut, which had the highest UUs of all the sites, also had the
highest number of quartzite tulas (58.1%). All the high UU rating tulas from
this site, without exception, were made from quartzite, which was not the
case for the other sites. Quartzite frequency for the other sites are:
Mungappie Creek had 32.6%, Munappie Creek north 50%, and Mungappie
Swamp 29%.
Utility Units of the cache.
Calculations for the Lake Hanson cache UUs are a theoretical projection
based solely upon tula type quantities which are known, and not upon the
actual volumes of the tulas, which are unknown. The Lake Hanson cache tula
percentages from A – G were compared to the Mungappie site tulas (Figure
5.9). The cache has peaks of C. semi discoidal and type G. tula blanks,
which have medium to high UUs. The Mungappie sites have a peak of type
A. slugs, which have zero UUs.
50.0
37.5
25.0
12.5
0
A
B
C
Lake Hanson Cache
A -!
B -!
C -!
D -!
Tula: slugs
Tula: part reduced
Tula: semi discoidal
Tula: asymmetric
D
E
F
G
Mungappie Sites
E -!
F -!
G -!
Tula: discoidal
Tula: side trimmed
Tula: blank
Figure 5.9 Percentage comparison of tula types at all Mungappie Sites and Lake Hanson
cache.
79
Toolkits and utility in Australian lithics
The percentages of tula types A – G, from both the combined Mungappie site
assemblages and the cache assemblage, were used to calculate the overall
UUs of each tula type, and then both tula assemblages. The percentage of
each tula type from each assemblage was multiplied by the UUs index for
each tula type to give the overall UU value for each type at each site. The
formula is: (UU index) x (% of tulas) = overall UU for each type. The total
UUs of all the types for the Mungappie sites was 910.1 whereas the total
UUs for the Lake Hanson Cache was 1221.4, a difference of 311.3 indicating
that the tula component of the Lake Hanson woodworking kit had 25.5%
more use value than the tulas in the four Mungappie sites (Table 5.6).
Table 5.6. Calculation of Utility Units by volume (cc) of tulas at Lake Hanson cache
and Mungappie Sites.
80
Toolkits and utility in Australian lithics
Figure 5.10 From Table 9, Utility units expressed as graph
Utility units for all sites
Looking at the UUs of each Mungappie site indicates distinct locational
variations (Figure 5.11). Once again Mungappie Swamp is not included due
to its lack of tula blanks, meaning that it would have an incomplete reduction
range, and would therefore not have been comparable to the other sites. The
total UUs are calculated for each site using the range and quantity of tulas at
each site. The remaining three sites are compared with the Lake Hanson’s
UUs as calculated in the previous section. In all of the sites the highest UUs
are found in the D – E range. The expectation would be to find the highest
rate in the G. blanks, but only Mungappie Creek North and the cache show
high levels. The UU rates are an indication of the utility of each item, as well
as the quantity of each types at each site. The high numbers of types D – E
and low numbers of G at the Mungappie sites account for these anomalies.
81
Toolkits and utility in Australian lithics
Figure 5.11 The total Utility Units for each site type. Mungappie Swamp is not
included due to an incomplete tula range.
82
Toolkits and utility in Australian lithics
5.6 Edge angles
The average edge angle for each artefact type was generated for all
retouched artefacts.
Figure 5.12 indicates the average edge angle of each type at the time of
discard. Average angles vary from a low 28º (chord) for geometric microliths
to high 79º for handchoppers. Most artefacts had angles in the mid 60º
range, which could be interpreted as either being an efficient working angle
for most scraping and planing tasks (Wilmsen 1968:157), or that the working
edge had been exhausted and was no longer useful. Others such as W.
geometric microliths, V. knife blades, and T. pirrie points, had much lower
angles of between 28º – 50º. Some anomalies, such as F. side trimmed tulas,
suggest some other reason for a consistently lower angle.
Tulas have angle ranges from 35º to as high as 90º with the mode
concentrating in the 60º-75º range. The micro tulas had a lower range from
45º - 85º but peaked a little higher at between 70 – 74º. All the scrapers had
similar ranges but there were much higher concentrations of numbers in the
middle range from 50º – 80º, maybe indicating a wider range of uses. The
geometric microliths had mid range concentrations between 20º – 35º, but
varied from as low as 10º to 65º.
83
Toolkits and utility in Australian lithics
Figure 5.12 Average edge angles for all artefacts at all Mungappie sites.
The range for type A. tula slugs was between 45º – 90º, with the majority
clustering between 60º – 75º (Figure 5.13). The lower angle would account
for 49% of these types having retouched laterals. C. semi discoidal tulas had
a lower range from 30 – 89º, but E. asymmetric’s range was reduced to 50 –
80º. The five tula G. blanks in the assemblages had three between 55 – 59º,
with the highest at 90 – 95º.
84
Toolkits and utility in Australian lithics
Figure 5.13 Number of tulas types for each edge angle range from all Mungappie
sites.
Figure 5.14 shows the edge angle of each tula type. These graphs support
the expectation that more reduced types such as A. slugs would have higher
edge angles than less reduced types such as E. discoidal. Variations to this
expectation occurs at Mungappie Creek North having type A. with low edge
angles, and D. asymmetric with relatively high readings.
85
Toolkits and utility in Australian lithics
Figure 5.14 Average edge angles of tulas from each Mungappie site. Only tula types
represented at each site are graphed.
86
Toolkits and utility in Australian lithics
5.7 Cortex
The flakes that are the first struck from the outside of a core with cortex will
have more cortex on the dorsal face than those struck later in the knapping
sequence. Hewitt indicated the number of tula types with cortex in the Lake
Hanson cache but did not record the amount of cortex on each artefact.
Therefore the tula types with cortex at Lake Hanson and the Mungappie sites
can only be quantitively compared.
Individual tula types at each Mungappie site and Lake Hanson cache were
compared (Figure 5.15). The order of tula types from G – A was set using the
UU reduction sequence G. B. D. C. E. F. A. so that cortex could be compared
with tula UUs. All types in the cache, except for types A. slugs, and B. part
reduced, were described by Hewitt as having cortex (Hewitt 1976:20). The
Mungappie sites, despite having far fewer types with cortex, had some types
with mid to high percentage rates such as D. asymmetric at Mungappie
Swamp (24%) and Mungappie Hut (16%,) resulting in Ds having the highest
cortex rate for all sites (11%). The two blanks from Mungappie Hut both had
cortex and 66.6% of types C. semi discoidal and D. asymmetric from
Mungappie Swamp also recorded cortex. Type C. semi discoidal was the
cache type with the highest cortex rate of 42%.
It was expected that there would be some correlation between reduction
amounts and remaining cortex on tulas, with more reduced types like A. tula
slugs having little or no remaining cortex and less reduced types such as G.
blanks having more cortex. However, all of the Mungappie sites, except for
Mungappie Creek North, recorded G. blanks with no cortex, and A. slugs with
cortex. In this sense Mungappie Creek North follows the expectations to
some degree, whereas the others did not. The larger percentage of artefacts
with cortex from Lake Hanson cache supports the result that the cache also
has tulas with higher UUs than the Mungappie site tulas.
87
Toolkits and utility in Australian lithics
Figure 5.15 Percentage of tula types from each Mungappie site with cortex.
Sequence G - A is the same as UUs.
88
Toolkits and utility in Australian lithics
Table 5.7 Quantities of tula types, with percentage of cortex on each, from all
Mungappie sites.
Table 5.7 shows all tula types from the Mungappie sites with percentage of
cortex in 5% ordinals. The majority of tulas had cortex on between 5 – 20%
of their surface, and most of those were C. D. and E., supporting the
individual site graphs (Figure 5.17). Of the five type G. blanks, two had no
cortex and one had 90%. Types C. and D. had the most artefacts with cortex.
5.8 Terminations
Most flakes in the Mungappie assemblages with retouch on the distal end
had the termination chipped away. Those with no retouch on the distal end,
such as K. flakes, N. points, P. scrapers and T. pirrie points, had terminations
remaining which could be recorded. Feathered terminations represented 83%
of all terminations, but were only 6.9% of all artefacts recorded from the
Mungappie sites (Table 5.8). Of 150 T. pirrie points from all Mungappie sites
131 (87%) had feathered terminations. Hewitt did not record terminations of
the Lake Hanson cache so no comparison can be made.
89
Toolkits and utility in Australian lithics
Termination
Feather
Stepped
Hinged
Plunging
Totals
% of
total
% of all
sites
A. Tula: slugs
B. Tula: part reduced
C. Tula: semi discoidal
D. Tula: asymmetric
1
3
4
1.5
0.3
1
2
0.4
0.7
0.1
0.2
4
50
1
1.5
18.5
0.4
0.3
3.7
0.1
1
28
0.4
10.3
0.1
2.1
40
14.8
2.9
7
2.6
5
2
0.7
0.2
131
48.3
87
271
100
100
19.868
E. Tula: discoidal.
F. Tula: side trimmed
G. Tula: blank
H. Micro adze/chisel
I.
Micro adze blank
J.
Burin/engraver
K.
Nondescript flake
KP. Flake Piece
L.
1
2
3
36
1
2
1
10
2
Nondescript lump
M. Burren
N. Points
1
28
O. Hammerstone/pounder
P.
Scraper
15
4
7
14
PB. Bulbous Scraper
PD. Discoidal Scraper
PL. Long Scraper
Q.
5
2
Core
R.
Thumbnail discoidal
S.
Stone block
T.
Pirrie points
U.
Flake core
V.
Cutting blade/knife
1
1
131
W. Geometric/backed microliths
X.
Hand chopper
TOTAL
% of terminations
225
83
7
2.6
23
8.5
16
5.9
Table 5.8 Terminations of all artefacts types from all Mungappie sites.
5.9 Breakages
Breakages from the Mungappie sites totaled 6% of all artefacts, of which
70.7% (58 artefacts) were scrapers. Fifty seven of the broken scrapers were
distal ends which still displayed retouch. The majority of proximal breaks (7
out of 8) were recorded as flakes as they displayed no retouch. Many scraper
breaks displayed similarities in length (4.5 cm average) and break angle (45 60%) that could have been deliberate; in some cases the breaks were
retouched into fine points (Figure 5.16). Hewitt’s collecting strategy, to look
for recognisable artefact types (letter to SA museum 1969), suggests that his
intention was not to collect broken artefacts, resulting in the low incidence of
90
Toolkits and utility in Australian lithics
breakages in the assemblages. Those he did collect would have looked like
complete artefacts.
Figure 5.16 Artefact #1060. Distal end break with point. J. Hayward.
5.10 Retouch location
Nearly 76% (1033) of all artefacts from all Mungappie sites had retouch that
could be allocated a code. The 206 type W. geometric microliths were not
given retouch location codes due to the proximal and distal being chipped off.
Of those 1033, C. distal retouch accounted for 43% and L. left + right laterals
+ distal retouch for 14% (Table 5.9).
91
Toolkits and utility in Australian lithics
Table 5.9 Retouch locations for all artefacts with retouch at all Mungappie sites
A complete chart of all retouch locations for all artefacts is in appendix
metrics sheet 10. Because of the morphological similarities, and occasional
confusion between tulas and scrapers, and points and pirrie points, a
comparison was made correlating edge angles and retouch location between
these types from the Mungappie sites. The pattern of retouch location in
relation to the average edge angles was similar for both tulas and scrapers
(Figure 5.17). The same comparison for pirrie points and points indicated a
marked pattern difference in retouch location (Figure 5.219).
The percentages of each artefact type for each retouch location were
calculated (Figure 5.18 and 5.20). Scrapers had a larger number of type C
retouch whereas tulas had a larger percentage of type L retouch. Pirrie points
had over 30% of type I and M – which were also some of the lowest edge
angle measurements – whereas points had a larger range of type B, D, and I
retouch locations.
92
Toolkits and utility in Australian lithics
Figure 5.17 Comparison between scrapers and tulas - edge angles to retouch
location.
Figure 5.18 Percentage amounts of scrapers and tulas for each retouch location.
93
Toolkits and utility in Australian lithics
Figure 5.19 Comparison between pirrie points and points - edge angles to retouch
location.
Figure 5.20. Percentage amounts of pirrie points and points for each retouch
location.
94
Toolkits and utility in Australian lithics
5.11 Retouch types
Of all the artefacts collected by Hewitt from the Mungappie Creek sites, 90%
had some kind of retouch (Table 5.10) The remaining 10% was accounted for
with types such as hammerstones, grindstones and some nondescript flakes.
Just over 41% of these flakes had had some sort of minimal retouch and
69% of these was type C. distal retouch. Chipping was the most common
type (53.8%) of retouch recorded for the whole site. W. geometric backed
microliths accounted for 99% of all backing, and T. pirrie points had received
90% of all pressure flaking. A combination of crushing and chipping,
indicating both use-wear and resharpening was found on 43% of tula slugs,
and 15% of scrapers.
RETOUCH TYPE
A. Tula: slugs
CH
CR
21
B. Tula: part reduced
SR
PR
BA
7
CR/
CH
21
2
5
C. Tula: semi discoidal
20
D. Tula: asymmetric
30
7
E. Tula: discoidal.
25
1
F. Tula: side trimmed
5
G. Tula: blank
2
1
CH/
PR
33
1
5
2
4
1
1
11
1
3
4
J.
Burin/engraver
2
2
K.
Nondescript flake
9
4
Nondescript lump
1
4
1
15
6
4
7
2
2
2
54
9
286
3
PB. Bulbous Scraper
40
1
PD. Discoidal Scraper
4
108
R.
Thumbnail discoidal
15
T.
Pirrie points
70
U.
Flake core
3
V.
Cutting blade/knife
1
W. Geometric/backed
microliths
X. Hand chopper
6
2
2
9
2
31
1
4
7
1
1
27
361
48
3
1
8
13
1
129
3
3
4
41
21
1
3
147
3
1
2
206
3
39
1
25
PL. Long Scraper
13
4
N. Points
Scraper
19
2
M. Burren
P.
38
26
13
L.
totals
12
Micro adze blank
5
SR/
CH
7
I.
KP. Flake Piece
PR/
CR
49
H. Micro adze/chisel
27
BA/
CH
1
206
4
8
734
24
7
30
207
150
64
2
2
11
1231
Percentage of recorded
59.6
1.9
0.6
2.4
16.8
12.2
5.2
0.2
0.2
0.9
100.0
Percentage of total (1364)
53.8
1.8
0.5
2.2
15.2
11.0
4.7
0.1
0.1
0.8
90.2
Totals
CH – chipping. CR – crushing. SR – serrated. PR – pressure. BA – backed.
Table 5.10 Retouch types for all retouched artefacts from all Mungappie sites.
95
Toolkits and utility in Australian lithics
5.12 Conclusion
Comparisons between the four Mungappie sites of all the artefact types
indicated similar patterns of assemblage composition, with some variation
due to either sample numbers, collector’s bias, or an actual difference of site
use. Inter-site comparisons, such as cortex distribution and Utility Unit
dispersal, revealed some data that could be significant to determining site
use. Mungappie Hut in particular has a number of features such as highest
UUs for tulas of all sites including the Lake Hanson cache, highest volume for
cores, and the highest percentage of quartzite artefacts.
The use of Utility Unit calculations to identify potential utility in tulas can be
used as an indication of toolkit signatures in assemblages. The high UUs at
Mungappie Hut, for example, correlates with the average artefact and core
size, suggesting that this site could be a source of high utility quartzite tools.
Comparing the Lake Hanson cache with the Mungappie sites highlights
significant differences between the assemblage structures (Table 5.11). The
typological and material makeup of the cache is distinctive in its use of less
common materials such as oolitic chert and banded jaspers, and in its range
of tula types that show little retouch and the high percentage of engraver type
tools. The UUs of the cache are higher than the average UUs for all the
Mungappie sites, but not higher than Mungappie Hut as an individual site.
Comparisons of other artefact attributes such as breakage and retouch
location has produced results that can be used for technological analysis
within the Mungappie sites, but are not comparable with the Lake Hanson
cache as Hewitt did not include these types of analyses in his reports. The
number of artefacts with breakages at the sites are low suggesting that
Hewitt did not generally collect incomplete items. Comparisons of retouch
location and edge angles of types with similar morphologies demonstrated
that tulas and scrapers share similar edge angle ranges but differing retouch
locations, whereas pirrie points and points had a less compatibility between
the two.
96
Toolkits and utility in Australian lithics
Item
Lake Hanson cache
Mungappie sites
Raw Materials Principally oolitic chert, banded
Mainly quartzite and silcrete
jasper, and translucent chalcedony.
Artefact Types 105 artefacts in all.
Types range from A - L.
74% tulas.
13% engravers.
Woodworking typologies.
1364 artefact at 4 sites.
Types range from A - X.
14.5% tulas
1% engravers.
Mainly woodworking types but other
types too.
Toolkit types
More blanks and curated types.
Larger quantity of high UU types.
More discard and expedient types.
Lower UUs for all sites.
High UUs for Mungappie Hut
Cortex
Higher percentage of tulas with
cortex - particuarly G. and C.
Oolitic chert, banded jasper, and
translucent chalcedony all have low
cortex rates.
Edge angles
Not recorded
Lower cortex rates caused by either
more retouch and more use, or
different material.
Quartzite and silcrete have highest
cortex rates.
Types C - E have highest cortex
rating.
Type A slugs have generally higher
edge angles than Type E. discoidal.
Mungappie Creek north has low
angles for slugs but high angles for
D. asymmetric. Mungappie Hut has
the lowest edge angles for type E.
Table 5.11 Summary of differences between Lake Hanson cache and the
Mungappie sites.
97
6. Discussion
6.1 Introduction
The results in the previous chapter provide detailed information about the
physical characteristics of the artefacts at the four Mungappie sites.
Comparisons between these sites demonstrate compositional differences
which may relate to their past function. The differences between the
Mungappie sites and the Lake Hanson cache are highlighted in the
combination of artefact types and materials that constitute the cache; and the
amount of usability (Utility Units) that differentiates the cache tulas from the
Mungappie tulas.
6.2 Site comparisons
The Mungappie sites lithics
An abundance of microliths in the lithic assemblages, which were prevalent
from about 3500 BP to 1500 BP, indicates the possible antiquity of the sites
(Boot 1993: 11; Robertson et al 2009:296). Hewitt attributes the abundance
of lithic material found in such places to higher populations that flourished
during a moister climate (Hossfeld 1966:87), and the possibility of artefacts
being hidden due to sand dune movement during periods of vacation and
then being exposed again in later times (Hewitt 1978:19). The full extent of
the density of lithic material has only become apparent in more recent
historical times by the action of strong winds that have blown the sand that
once covered the sites several kilometers away onto the surrounding areas
of drifting dunes (Hewitt pers. comm. 2010).
Raw material
Quartzite was dominant in the Mungappie assemblages (57.2%), while oolitic
chert was rare (3.2%). The cache had 47% oolitic chert and no quartzite: the
complete antithesis of the sites. The provenance of oolitic chert was not
precisely known to Hewitt. It is not a common material in other assemblages
98
Toolkits and utility in Australian lithics
in the area, as Hewitt had recorded other campsites around Lake Hart which
had only 2 oolitic chert out of 117 artefacts (1.7%) (Hewitt 1976:31). Hewitt
asked the South Australian Museum to identify the material, and they
reported a source in the Flinders Ranges, some 250 kms directly to the east.
Another possible source could be near Mt Harvey or Mt Toondina, north of
Coober Pedy, about 330 kms north of Lake Hanson, where small quantities
of this unusual form of silica has been observed (Smale 1973:1084). A more
local source would not account for the rarity of the material in other
assemblages, unless it was reserved for culturally specific use (Taçon 1991).
If the raw material had been brought in from a distance to be made into
implements for a special purpose, possibly by one person, there would have
been a time and effort cost that added value to the cache as a trading
stockpile, or toolkit for specialisation.
The more common material of the Mungappie sites, quartzite, is found in
gibber form across most of the Arcoona Plateau, but also occurs as boulder
outcrops in a number of locations, including Mungappie Hut (Hewitt 1978:15).
Each Mungappie site
The difference in the artefactual composition of each site assemblage could
be the result of differing site specialisations, similar to Binford’s Mousterian
factors. It could also be a result of Hewitt’s collecting methods, other
taphonomic interventions, or the different sample sizes which range from 866
at Mungappie Creek to 59 at Mungappie Swamp.
Mungappie Creek
The Mungappie Creek assemblage was the only one of the four Mungappie
sites that contained the complete range of toolkit types from A – L, replicating
the cache. It was also the only site to have banded jasper and oolitic chert
cores. For these two reasons Mungappie Creek would appear to be the most
obvious place to look for replica cache toolkits, other than the fact that it also
has the lowest UU rate of all the sites. This tends to suggest that this site,
despite having a full complement of toolkit types, has the highest rate of
discarded toolkit types. Despite having the rare core types it did not have as
99
many other implements of the same rare materials, suggesting that they were
either not collected by Hewitt or removed from the site some time before.
Mungappie Creek North
Mungappie Creek North had one characteristic to directly compare to the
cache. Its cortex distribution (quite different to the other Mungappie sites)
was the one most like the cache, with blanks having a high cortex rate, and
slugs with none. It also had a UU index that most closely followed the
combined Mungappie sites index, other than not having any type B. part
reduced tulas. Having a similar, but slightly higher, site UU rating as
Mungappie Creek would put it in the same category as being a site that may
have more depleted implements than usable ones.
Mungappie Hut
This site had the largest cores made from quartzite and silcrete; the largest
tulas in the types C. D and E; the largest tula blank (chert); and was the site
with the highest UUs, albeit the most erratic. The combination of these
features, along with the site being close to a quartzite outcrop, suggests that
it could have been the place closest to the quarry where raw material was
procured and knapped into blanks. It is for these reasons that this site is the
one of all the Mungappie sites to have the highest potential of providing
artefacts that could be formed into a functional toolkit.
Mungappie Swamp
The lack of tula blanks at Mungappie Swamp meant this it was impossible to
do a complete UU index for its artefacts that could be compared with the
other sites. The absence of a number of other types from Mungappie
Swamp, which were abundant at other sites, needs explanation. These types
are B. part-reduced tulas, T. pirrie points, N. points, J. burin/engravers and K.
unretouched flakes. It did, however, have the highest percentage of A. tula
slugs and R. thumbnail discoidals of all the sites, and a high proportion of all
scraper types and cores. Types such as pirrie points and tulas would have
been high on Hewitt’s list of collectable items, suggesting that they were quite
possibly not there. Even nondescript flakes would have been collectable at
this small site, as their abundance at other assemblages testifies. Their
100
Toolkits and utility in Australian lithics
absence from the site suggests that this was not a knapping site (Cane
1984:198), but a place where finished artefacts were brought in, or taken
away, for a particular reason.
Each site has its own unique characteristics, suggesting that different
activities were occurring at each place. Sites like Mungappie Hut with high
UUs will be more likely to have useable toolkit types than Mungappie
Swamp. The total UU rate for all tulas at Mungappie Hut, being much higher
than Lake Hanson cache, suggests that not having a full sequence of tula
types does not necessarily equate to not having a toolkit. These types,
though, would make a quite different toolkit to the one in the cache.
Mungappie Creek, whilst having all toolkit types available, has a lowest UU
index of all the sites indicating the usability of the toolkit types may not be
satisfactory.
Lake Hanson cache seems to be a particular type of toolkit designed to fulfill
a specific task. The Mungappie sites may not have had all the requirements
to replicate the cache but could have supplied some high UU tools from
some sites, in particular Mungappie Hut.
6.3 Lake Hanson cache
The distinctive structure of the cache such as the different use of raw
materials for particular tool types and the high percentage of blanks and
other high utility tools, becomes clear in comparison with the Mungappie
sites. Hewitt thought the cache was the work of one craftsperson, due to the
homogeneity of the material, and the same amount of minimal reduction on
most of the implements (Hewitt 1976:33). It is an assemblage with personal
attributes, as opposed to the main Mungappie assemblages which are a
result of communal activities. Tools made by one person from rare materials
and cached in a remote location are all attributes of a personal toolkit not
found in general sites.
101
The high percentage of type J. burin/engravers are of interest in the cache.
They have been ethnographically recorded as being specifically used for
engraving sacred ceremonial hardwood boards in the Western Desert
(Hayden 1979:168; Cane 1984:236). The fact that the cache has a large
component of these types made from particular raw materials could indicate
that its use was for cultural and ceremonial purposes. The symbolic use of
particular stone types has been noted by some researchers (Hayden
1977183; Taçon 1991). Engraving of cultural motifs on wooden artefacts is an
important aspect of cultural transmission (Shennan 2003), and the
craftsperson needs appropriate tools made from culturally appropriate
materials to achieve this.
The Utility Unit calculations from the Mungappie sites support the idea that
the cache contains tula types of high utility. The high percentages of tula
types and engraver type tools also supports Hewitt’s claim for it being a
function specific woodworking kit. This combination of high utility, function
specificity, and caching suggest this assemblage was systemic rather than
discarded; it was still in use, or had potential use.
Caching
Caching in archaeology can represent a range of behaviours, from ritual,
dedicatory or votive deposits, to banking or insurance caching (Schiffer
1987:79). The former type of caching has symbolic meaning, and the latter a
more functional purpose. Banking refers to the hiding of objects with some
value (monetary or functional) for safekeeping, and future use (Schiffer
1987:79). The reason they become part of the archaeological record could be
due to the hoarder forgetting the location, or being prevented from return to
the cache. Binford’s (1976) description of Nunamiut Eskimos leaving multiple
insurance caches in strategic locations around the territory for emergency
use might also account for some caches being forgotten or lost.
Hiscock’s (1988:67; 1989:38) definition of caching as ‘the underground and
concealed storage of objects’, ruling out grave burials and most artefacts
102
Toolkits and utility in Australian lithics
found in stratigraphic excavations unless they show signs of concealment,
differentiates caches from discard. Discard assumes that the item will no
longer be needed, whereas the association with caching is that the items will
be recovered and presumably re-used (Hurst 2006:102). Although Hiscock’s
definition limits the number of caches claimed to be found in Australia to quite
a low number, Hewitt’s claim for the Lake Hanson cache does appear to be
legitimate (Hiscock 1998:68).
Cache finds in the Southern Plains of North America, on the other hand, have
been documented at over 100 in 2006 (Hurst 2004:101). Most of these were
uncovered in isolated contexts, with the minority being discovered within
campsites (Hurst 2004:101). If caching is a concealing behaviour, then it
would be logical to conceal a cache in a place where other people are not
expected to go. Hewitt’s cache was found approximately 50 metres away
from the main Lake Hanson campsites (Hewitt 1976:16), whereas Hiscock’s
Mucklandama Creek 6 cache seemed to be found within a surface artefact
scatter (Hiscock 1988:61). Hiscock had the benefit of being able to directly
compare measurements of cache and surface tulas, with the result being that
the cache items have larger dimensions than those from surface scatters
(Hiscock 1988:65), indicating higher volume and utility values (Kuhn 1994).
The loss, or abandonment, of cached artefacts that represent toolkits is not
commonly reported, but the notion that specialised tools can be cached for
safekeeping at frequently visited workshop sites is not unheard of (Baker
1975; Schiffer 1987:93; Barkai et al 2002:673; Odess and Radic 2007:691).
6.4 Artefact attribute analyses
Cortex
Hewitt (1976:31) thought that the reason for the high percentage of cortex on
the cached tulas was due to the raw material occurring in small nodules. It is
also a characteristic of blanks, or less reduced artefacts, to have more cortex
than highly reduced or discarded artefacts (Andrefsky 1998:181). The cortex
analysis (Fig 26), which indicates the quantity of tulas with cortex at all
assemblages, confirms that significantly more cached stones had cortex on
103
some part of their surface than their counterparts at the Mungappie sites. The
fact that the dominating cache materials are different from the Mungappie
sites suggests that the difference in cortex could also be related to the
difference in material types and their sources. Quartzite artefacts at
Mungappie Creek with cortex accounts for 57.4% of all material with cortex,
whereas oolitic chert is 4.2%. Cortex signatures for each material indicate
that quartzite had a much higher occurrence rate than oolitic chert,
translucent chalcedony, and banded jasper on all Mungappie artefacts. This
being the case, one would expect there to be a higher cortex rate at the
Mungappie sites, not less, suggesting that the artefacts at the Mungappie
sites were, in general, more reduced than the cache.
Edge angles
The average edge angles of the tula types measured at the Mungappie sites
were generally in a narrow range of 65º – 71º except F. side trimmed tulas, at
58º. The mean edge angle for all scrapers from all Mungappie sites is 65º
(range 23 – 90º) and for tulas is similar at 66º (range 29º – 90º). These
results imply that these two types might have had similar applications, as
edge angle variants can suggest particular use, but may also be indicative of
the most efficient angle for prolonged use life (Collins 2008:2168). Side
trimmed tulas at Mungappie could have been developed for different
woodworking techniques that required a sharper, lower angle.
The site with the highest toolkit supply potential, Mungappie Hut, had larger
tulas such as type E. discoidal with higher edge angles suggesting that they
were less reduced (Figure 5.14). Consequently, this site also had tula type A.
slugs with higher angles, which correlates with higher reduction. This was the
model site; others like Mungappie Creek North had opposing trends with type
A. having lower angles than any of its larger less reduced counterparts. One
explanation for this could be that the high angled types D. and C. were more
reduced at discard, and the lesser angled A. slugs had been recycled into
finer graving tools prior to discard.
104
Toolkits and utility in Australian lithics
Retouch Location
One result that can be extrapolated from the tula/scraper edge angle to
retouch analysis graph (Figure 5.18) is that there was a strong correlation
between the two types over most retouch locations other than A. proximal
and K. Proximal and Left Lateral and Distal. These retouch locations,
including M. Right Lateral and Proximal and Left Lateral, and N. Proximal and
Right Lateral and Distal which did not rate, were either sole proximal retouch
or combination proximal retouch locations, indicating that not many tulas had
proximal retouch, whereas scrapers did. This could be an important
distinction between the two types which are sometimes difficult to distinguish.
6.5 Conclusion
The attributes of the Lake Hanson cache and the Mungappie sites, either as
a group or as individual sites, are quite divergent. The choice of raw
materials that the maker of the Lake Hanson cache used indicates a special
toolkit that possibly had a significant use. It is a systemic assemblage and
therefore will not be found in surface scatters like the Mungappie sites. This
is confirmed by the difference in artefact types, numbers and materials at all
Mungappie sites compared to the cache.
Mungappie Creek, the one site that had the artefact range to provide all the
toolkit types, also had the lowest utility units of all the sites, indicating a more
used or depleted range of tool types. Mungappie Hut, on the other hand, had
Utility Units higher than any other site, including the cache, and therefore is
the one Mungappie site with the potential to provide a range of tulas with high
UUs that could become a working toolkit. This is supported by a range of
edge angles in some tula types, which suggest less reduction, and therefore
more potential utility.
105
7. Conclusion
Bordesʼ insight into the way that variation within Mousterian facies in France
could be visualised through distribution analyses stimulated debate about
assemblage variation amongst archaeologists in North America. Howell
(1966) speculated on the variation amongst Acheulian assemblages in
Northern Europe and Freeman discarded the possibility that variation
represented differences in industrial and cultural evolution, and speculated
that they therefore signified different activities (Freeman 1966:232). It was
Binfordʼs (1966) re-evaluation of Bordesʼ Mousterian lithic assemblages,
using computer technology and multivariate statistics, that launched a
controversial theory that has since become standard archaeological thinking.
Now, both typological and technological approaches to lithic analysis use
toolkits to categorise and explain assemblages. Prehistoric peoples from all
over the world, from northern Alaska (Binford 1977) to northern Belgium
(Cahen et al. 1979) and from the Oldawan tradition of Africa (Clark 1992:201)
to the central deserts of Australia (McCarthy 1976:19), have been postulated
to use toolkits. Extensive analysis of prehistoric lithic toolkits (Cahen et al
1979; Grace 1989; Kuhn 1994; Odess and Rasic 2007; Banks 2009)
supports use of the concept as a global approach for describing aspects of
risk management (Torrence 1983; Myer 1989; Hiscock 1994), mobility (Kuhn
1994), curation (Shott 2009), and even functionality in prehistoric stone tool
assemblages (Banks 2009).
Archaeologists have predominantly used toolkit as a generalised term that
can quite often refer to an entire range of technologies available to a culture,
ie. the Tasmanian toolkit, a Mousteroid toolkits, or the Aboriginal toolkit
(Jones 1977:202; Bordes 1977:37 Cane 1984:276). This generalisation of the
term has overridden the original idea of a toolkit being a personal collection
of ʻtools of the tradeʼ and as a result there has been very little specific
analysis of prehistoric toolkits at the personal or functional level. There are
three reasons why this might be the case: the difficulty of finding toolkits that
fit the definition of personal in lithic assemblages, mainly as a result of there
106
Toolkits and utility in Australian lithics
being few personal traits that can be attributed to stone tools when they are
analysed statistically; another is that stone tools can be multifunctional; and
lastly, personal toolkits fit into a systemic rather than an archaeological
context, meaning that toolkits were useable entities that had not yet been
discarded. Sites that display Pompeiian-type assemblages will have systemic
toolkits (Schiffer 1985). The few lithic assemblages that have been identified
as personal toolkits have tended to be discrete entities that have been
cached or stored away for future use (Baker 1975; Schiffer 1987:93; Hurst
2004; Odess and Radic 2007:691)
The Lake Hanson cache was buried and stacked at the top of a wind-swept
sand dune, away from the campsites, for a particular purpose. The
predominance of two rare material types, and uniform retouch, indicates that
it was possibly made at the same time, by the same person (Hewitt 1976:48).
It represents a systemic assemblage, and was a function-specific collection
of tools which could have been stored for future use or for trading. It was
distinctly different from the general lithic assemblages in a number of ways,
as the results of the assemblage analyses have shown.
The generalised approach would be to describe each Mungappie site
assemblage as a different toolkit. In this sense each assemblage is the
toolkit, and the variations they display can reflect different activities or
behaviours over time. The “assemblage as toolkit” approach can account for
big picture theories, but does not necessarily address individual behaviour.
Interpreting the Lake Hanson cache as a personal toolkit can offer insights
into both the individual and the culture. It can also be used a measure of
toolkit structures in other assemblages.
The analysis undertaken in this project has looked at four lithic scatter
assemblages within ten kilometres of each other, each with different
attributes, and characteristics. Comparing these sites with the cache has
raised issues about the nature of toolkit structures in lithic assemblages. The
combination of specific tool types with a high utility rating made from special
materials represents a unique entity. When looking for this entity in the
107
Mungappie assemblages it was found that none of the sites could replicate
the entity as a whole, on both typological and raw material criteria.
Mungappie Creek could provide all the types but in limited numbers, and
could not match the material types. Mungappie Hut, whilst not having a full
range to match the cache, did have high Utility Units that suggest a high
probability of toolkit type components being there. They will not be cache like
in their material composition but they will have potential usability.
The use of Utility Units to measure the potential utility of an artefact is a
method that can identify toolkit signatures in an assemblage. Signs of high
utility in a range of tools could suggest a functional toolkit, and low utility
signatures will indicate a higher rate of depletion. Depleted tools are the
remains of toolkits. Evidence of artefacts at the early stages in the reduction
sequence will be a sign of tools being renewed, and therefore of toolkit
replacement. Tool replacement indicates a toolkit maintenance programme
that is the signature of preparedness, and the ability to ameliorate risk. In this
respect the Lake Hanson cache is the model of a highly maintained toolkit as
it has a full range of tula types, from blanks to slugs. The antithesis is a site
like Mungappie Swamp, with a stockpile of tula slugs and few other tula
types, which does not indicate toolkit renewal.
Lewis and Sally Binford’s initial observation, when they began their
investigation of lithic toolkits, was that there are no set ‘types’ of toolkits.
“These subunits of artefacts vary independently of one another and may be
combined in numerous ways” (Binford and Binford 1966:292). This suggests
that a toolkit cannot be recognised by the individual artefacts of which it is
composed. There has to be something about the combination of the
individuals that, when grouped into subunits, forms this elusive entity. The
applicability of this to Australian lithic archaeology lies in the ability to analyse
assemblages with a conceptual framework that allows for the type of
flexibility that has been attributed to Indigenous tool use.
The scarcity of caches which can be interpreted as toolkits in Australian
archaeology compared to other continents such as North America can be
explained by the nature of the landscape and it current usage. Many of the
108
Toolkits and utility in Australian lithics
places where cached entities of tools have been discovered in America have
been in developed or agricultural landscapes where the sites have been
unearthed by heavy digging or ploughing (Mallouf 1982:79; Rogers 1985:116;
Wyckoff 1996:289; Hucknell et al 2011:965). These cache sites have usually
been away from known archaeological occupation sites (Hurst 2004:101),
reinforcing the notion that caching is an activity that is carried out in places
where the cache will not be found easily. The arid nature of much of the
Australian landscape has not been conducive to such development.
Australian archaeology tends to focus on occupation and workshop sites;
places where toolkits will not be found (Simek 1984:3). The paucity of toolkit
type entities in Australian archaeology could be that they have either not
been found, or have not been recognised. Other reasons could be that they
are very rare cultural items, or do not exist. The latter is supported by the lack
of ethnographic recordings of toolkits in Australian Aboriginal culture, other
than the mobile type hunting maintenance toolkits as demonstrated by
Balfour; and references to these are scarce too. The added problem of
surface scatters, which are common archaeological sites in Australia, is that
they were deposited over long periods of time and the possibility of being
able to identify discrete entities such as toolkits belonging to one individual at
one time, is all but impossible. The conclusions are that; complete entities
which can be classified under traditional and common definitions of a toolkit
will not be found in discard lithic assemblages; cached items are more likely
to include artefact types with high utility, and therefore possibly be a toolkit, or
items for toolkit replacement; and the use of the term toolkit to describe a
complete range of artefact types in an assemblage is inaccurate and
misleading.
There are also broader conceptual implications of this research project,
which was initially inspired by the desire to enquire into what seemed a rather
inadequate and inaccurate use of the toolkit concept by archaeologists.
Having found that there are advantages in conceptualising the toolkit as a
specific set of tools for personal or trade use, I suggest that archaeologists
should strive to tighten the use of this term. Some of the benefits of such a
tighter concept for Australian archaeology include: a more sophisticated
109
understanding of the ways in which Indigenous cultures used technologies
and resources in their everyday life; understanding that sites with differing
functions such as quarries, workshops, storage places etc. have different
assemblages which reflect these variations, but do not necessarily contain
toolkits; and, how we as archaeologist can identify these variations without
deferring to heuristic devices and generic terms such as ʻtoolkitʼ without more
in depth analysis.
110
Toolkits and utility in Australian lithics
References
Allen, J. 1975 Report on the Conference on Historical Archaeology
and the National Estate. Australian Archaeology 2: 62-97.
Andrefsky, W. 1998 Lithics: Macroscopic Approaches to Analysis.
Cambridge: Cambridge University Press.
Baker, C. M. 1975 Site Abandonment and the Archeological Record: An
Empirical Case for Anticipated Return. Arkansas Academy of Science
Proceedings 29:10-11.
Balfour, H. R. 1951 A Native Tool Kit from the Kimberley District, Western
Australia. Mankind 4(7):273-274.
Banks, W. E. 2009 Toolkit Structure and Site Use: Results of High Powered
Use-wear Analysis of Lithic Assemblages from Solutré (Saône-et-Loire),
France. British Archaeological Reports, International Series 1970. Oxford:
Archaeopress.
Barton, H. 2008 Expedient Technologies and Curated Tools within a System
of High Residential Mobility: An Example of using Mass Analysis of Flakes
from the Simpson Desert, Central Australia. Lithic Technolog. 33(1):51-71.
Basedow, H.1925 The Australian Aboriginal. Adelaide: Preece.
Beyries, S. 1988 Functional Variability of Lithic Sets in the Middle Paleolithic.
In H. L. Dibble and A. Montet-White (eds) The Upper Pleistocene Prehistory
of Western Eurasia. pp. 213-223. Philadelphia: University Museum.
Binford, L. R. 1962. "Archaeology as Anthropology". In M. Leone (ed)
Contemporary Archaeology. pp. 93-101. Carbondale: Southern Illinois
University.
111
Toolkits and utility in Australian lithics
Binford, L.R. and Binford, S.R. 1966 A Preliminary Analysis of Functional
Variability in the Mousterian of Levailois Facies. American Anthropologist
68:238-295.
Binford, S. R. 1968 Near Eastern Mousterien of Levallois Facies. In Binford,
S. R. and L. R. Binford (eds) Archaeology in Cultural Systems. Chicago:
Aldine Publishing Co.
Binford, L. 1973 Interassemblage Variability: The Mousterian and the
‘Functional’ Argument. In C Renfrew (ed) The Explanation of Culture Change:
Models in Prehistory; Proceedings of a Meeting of the Research Seminar in
Archaeology and Related Subjects held at the University of Sheffield. pp.
227-254. London: Duckworth
Binford, L. R. 1977 Forty Seven Trips. In R. V. S. Wright (ed) Stone Tools as
Culture Markers: Change, Evolution and Complexity. pp. 24-36. Canberra:
Australian Institute of Aboriginal Studies.
Binford, L. R. 1979 Organisation and Formation Process; Looking at Curated
Technologies. Journal of Anthropological Research 35 (3): 255-73.
Binford, L. 1983 In Pursuit of the Past: Decoding the Archaeological Record.
New York: Thames and Hudson.
Binford, L. 1989 Debating Archaeology. San Diego: Academic Press Inc.
Board for Anthropological Research Film 1935 Warburton Range, Western
Australia. South Australian Museum archives AA 346/9/10. Associated with,
Tindale, N, B. July-September 1935 'Anthropological Expedition to Warburton
Range, Western Australia by. Journal. Adelaide, S. A. 1935 (AA 338/1/14).
Boot, P. 1993. Analysis of Resins and Other Plant Residues on Stone
Artefacts from Graman New South Wales. In B. Fankhauser and J. Bird (ed)
112
Toolkits and utility in Australian lithics
Archaeometry: Current Australasian Research. pp 3–12. Canberra:
Australian National University.
Bordes, F. 1953 Essai de Classification des Industries “Mousterian” Bulletin
de la Societe Préhistorique Française 50: 457-466.
Bordes, F. 1968 The Old Stone Age. London: Weidenfeld and Nicholson.
Bordes, F. 1973 On the Chronology and Contemporaneity of different
Palaeolithic Cultures in France. In C Renfrew (ed) The Explanation of Culture
Change: Models in Prehistory; Proceedings of a Meeting of the Research
Seminar in Archaeology and Related Subjects held at the University of
Sheffield. pp. 217-226. London: Duckworth
Bordes, F. 1977 Time and Space Limits of the Mousterian. In R. V. S. Wright
(ed) Stone Tools as Culture Markers: Change, Evolution and Complexity. pp.
37 - 39. Canberra: Australian Institute of Aboriginal Studies.
Bordes, F. and D. de Sonneville-Bordes 1970 The Significance of Variability
in Palaeolithic Assemblages. World Archaeology 2(1): 61-73
Bourman, R. P. and A. R. Milnes 1985 Gibber Plains. Australian Geographer
16:229-232.
Bowdler S. and J. Smith 1999 Identifying Style in Australian Stone Artefacts:
An Attempt to Find a Theoretical Basis. Australian Archaeology 49:1-6.
Brough Smyth, R. 1878 Aborigines of Victoria. Melbourne: Government
Printer.
Burke, H. and C. Smith 2004 The Archaeologist’s Field Handbook. Crows
Nest: Allen and Unwin.
Burgess, R. J. and K. L. Kvamme 1978 A New Technique for the
Measurement of Artifact Angles. American Antiquity 43(3): 482-486.
113
Toolkits and utility in Australian lithics
Cahen, D. Keeley, L. H. and F. L. Van Noten 1979 Stone Tools, Toolkits, and
Human Behavior in Prehistory. Current Anthropology 20(4): 661-683.
Campbell T. D. and M. J. Barrett 1958a The Boomerang. South Australian
Museum Archives AA346/9/15/4.
Campbell T. D. and M. J. Barrett 1958b The Woomera. South Australian
Museum Archives AA346/9/15/5.
Campbell, T. D. and R. Edwards 1966 Stone Implements. In C. Cotton (ed)
Aboriginal Man in South and Central Australia, part 1. pp. 159-220. Adelaide:
British Science Guild Handbooks.
Cane, S. 1984 Desert Camps: A Case Study of Stone Artefacts and
Aboriginal Behaviour in the Western Desert. Unpublished Phd thesis, ANU.
Cane, S. 1992 Aboriginal Perceptions of their Stone Tool Technology: A Case
Study from the Western Desert, Australia. Australian Archaeology 35:11-31.
Cane, S. 2002 Pila Nguru: The Spinifex People. Fremantle: Fremantle Arts
Press.
Cantwell, A. 1979 The Functional Analysis of Scrapers: Problems, New
Techniques and Cautions. Lithic Technology 8(1): 5-9.
Clark, J. D. 1992 African and Asian Perspectives on the Origins of Modern
Humans. Philosophical Transactions: Biological Sciences 337(1280):
201-215.
Collins, S. 2008 Experimental Investigations into Edge Performance and its
Implications for Stone Artefact Reduction Modelling. Journal of
Archaeological Science 35: 2164-2170.
114
Toolkits and utility in Australian lithics
Cotterell, B. and J Kamminga 1987 The Formation of Flakes. American
Antiquity 52(4): 675-708
Debénath, A. and H. L. Dibble 1993 Handbook of Paleolithic Typology.
Philadelphia: The University Museum.
Dibble, H. L. 1987 The Interpretation of Middle Palaeolithic Scraper
Morphology. American Antiquity 52(1):109 - 117.
Dibble, H. L. 1991 Mousterian Assemblage Variability on an Interregional
Scale. Journal of Anthropological Research 47(2):239-258.
Dibble, H.L. and M. C. Bernard 1980 A Comparative Study of Basic Edge
Angle Measurement Techniques. American Antiquity 45(4): 837-865.
Dibble, H. L. and N. Rolland 1992 On Assemblage Variability in the Middle
Palaeolithic of Western Europe: History, Perspective and New Synthesis. In
H. L. Dibble and P. A. Mellars (eds) The Middle Palaeolithic; Adaption,
Behaviour, and Variability. pp. 1-28. Philadelphia: University of Pennsylvania.
Dibble, H. L. and S. McPherron (2006). The Missing Mousterian. Current
Anthropology 47(5): 777- 803.
Edwards, W. H. 2000 MacDougall, Walter Batchelor (1907 - 1976), Australian
Dictionary of Biography 15:201-204.
Evans, J. 1987 Ancient Stone Implements, Weapons, and Ornaments of
Great Britain. 2nd ed. London: Longmans Green and Co.
Freeman, L. G. Jr 1966 The Nature of Mousterian Facies in Cantabrian
Spain. American Anthropologist 68(2): 230-237.
Gara, T. 2010 Walter MacDougall and the Emu and Maralinga Nuclear Tests.
Online, available at: http://www.samuseum.sa.gov.au/page/default.asp?
site=1&id=1596&ArchPosted=true (5.9.2010)
115
Toolkits and utility in Australian lithics
Gibber Gabber newspaper. 1967 1968 1969. Woomera: Woomera Board.
Goodman, W. L. 1964 The History of Woodworking Tools. London: G. Bell
and Sons Ltd.
Gorman, A. 2009 The Archaeology of Space Exploration. Bell, D and M.
Parker (eds) The Sociological Review: From Apollo to Space Tourism. pp
132–145. Malden: Blackwell Publishing.
Gould, R. A. Koster, D. A. and R. L. Sonzt 1971 The Lithic Assemblages of
the Western Desert Aborigines of Australia. American Antiquity 36(1):
149-169.
Gould, R. A. 1977 Ethno-Archaeology: or Where Do Models Come From?: A
Closer Look at Australian Aboriginal Lithic Technology. In R. V. S. Wright (ed)
Stone Tools as Culture Markers: Change, Evolution and Complexity. pp.
162-168. Canberra: Australian Institute of Aboriginal Studies.
Gould, R. A. 1980 Living Archaeology. Cambridge: Cambridge University
Press.
Grace, R. 1989 Interpreting the Function of Stone Tools: The Quantification
and Computerisation of Microwear Analysis. British Archaeological Reports,
International Series 474. Oxford: Archaeopress.
Griffiths, T. 1996 Hunters and Collectors. Cambridge: Cambridge University
Press.
Fullagar, R. Furby, J. and B Hardy. 1996 Residues on Stone Artefacts: State
of a Scientific Art. Antiquity 70(270): 740-746.
Hayden, B. 1977 Stone Tool Function in the Western Desert. In R.V.S. Wright
(ed) Stone Tools as Cultural Markers: Change, Evolution and Complexity. pp.
178 - 188. Canberra: Australian Institute of Aboriginal Studies.
116
Toolkits and utility in Australian lithics
Hayden, B. 1979 Paleolithic Reflections: Lithic technology and Ethnographic
Excavations among Australian Aborigines. Canberra: Australian Institute of
Aboriginal Studies.
Hayward, C. 1973 Tools for Woodworking. Revised edition. London: Evans
Bros.
Heaney, P. J. and J. E. Post 1992 The Widespread Distribution of a Novel
Silica Polymorph in Microcrystalline Quartz Varieties. Science 255: 441-443.
Hewitt, R. 1976 Aboriginal Adze Stone Hoards Found on the Arcoona Plateau
near Woomera, South Australia. Australian Archaeology 4: 14-52.
Hewitt, R. 1978 The Arcoona Plateau: An Outline of Aboriginal Habitats and
Relics. Journal of the Anthropological Society of South Australia.16(6): 6-19.
Hiscock, P. 1988 A Cache of Tulas from the Boulia District, Western
Queensland. Archaeology of Oceania 23: 60-70.
Hiscock, P. 1994 Technological Response to Risk in Holocene Australia.
Journal of World Prehistory 8(3): 267-292.
Hiscock, P. 1998 Revitalising Artefact Analysis. In T. Murray (ed) Archaeology
of Aboriginal Australia. pp.257-265. Sydney: Unwin and Allen.
Hiscock, P. 2004 Slippery and Billy: Intention, Selection and Equifinality in
Lithic Artefacts. Cambridge Archaeological Journal 14: 71-77.
Hiscock, P. 2006 Blunt and To the Point: Changing Technological Strategies
in Holocene Australia. pp. 69-95. In I. Lilley (ed) Archaeology in Oceania:
Australia and the Pacific Islands. Blackwell.
Hiscock, P. 2008 The Archaeology of Ancient Australia. Oxon: Routledge.
117
Toolkits and utility in Australian lithics
Hiscock, P. & V. Attenbrow 2003 Early Australian Implement Variation: A
Reduction Model. Journal of Archaeological Science 30: 239-249.
Hiscock, P. and V. Attenbrow 2004 A Revised Sequence of Backed Artefact
Production at Capertee 3. Archaeology in Oceania 39: 94-99.
Hiscock, P. & V. Attenbrow 2005a Australia's Eastern Regional Sequence
Revisited. Technology and Change at Capertee 3. British Archaeological
Report, International Series 1397. Oxford: Archaeopress.
Hiscock, P. & V. Attenbrow 2005b Reduction Continuums and Tool Use. In C.
Clarkson, and L. Lamb (eds) Lithics Down Under: Australian Perspectives in
Lithic Reduction Use and Classification. pp 43-55. British Archaeological
Report, International Series 1408. Oxford: Archaeopress.
Hoare, M. 1976 Smyth, Robert Brough (1830 - 1889). Australian Dictionary of
Biography. Melbourne: Melbourne University Press 6:161-163.
Holdaway, S. and N. Stern 2004 A Record in Stone: The study of Australia’s
Flaked Stone Artefacts. Melbourne: Museum Victoria.
Horne, G. and G. Aiston 1924 Savage life in Central Australia. London:
MacMillan and Co.
Horton, D. (1991). Recovering the Tracks: The Story of Australian
Archaeology. Canberra: Aboriginal Studies Press.
Hossfeld, P. S. 1966 Antiquity of Man in Australia: Origins of the Early
Inhabitants. In B. C. Cotton (ed) Aboriginal Main in South and Central
Australia: Part 1. pp. 59-96. Adelaide: British Science Guild Handbooks.
Howchin, W. 1934 The Stone Implements of the Adelaide Tribe of Aborigines,
Now Extinct. Adelaide: Gillingham and Co.
118
Toolkits and utility in Australian lithics
Howell, F. Clark. 1964 Observations on European Middle Acheulian Tool-kits.
Unpublished Paper. Detroit: American Anthropological Association.
Howell, F. Clark. 1966 Observations on the Earlier Phases of the European
Lower Paleolithic. American Anthropologist 68(2): 88-201.
Hucknell, B. B. Kilby, J. D. Boulanger, M. T. and M. D. Glascock 2011
Sentinel Butte: Neutron Activation Analysis of White River Group Chert from
a Primary Source and Artifacts from a Clovis Cache in North Dakota, USA.
Journal of Archaeological Science 38:965-976.
Hurst, S. 2006 An Analysis of Variation in Caching Behaviour. Lithic
Technology 31(2): 101-126.
Johns, R. K. 1968 Geology and Mineral Resources of the AndamookaTorrens Area. Department of Mines SA Bulletin No. 41.
Jones, R. 1977 The Tasmanian Paradox. In R.V.S. Wright (ed) Stone Tools
as Cultural Markers: Change, Evolution and Complexity. pp. 189-204.
Canberra: Australian Institute of Aboriginal Studies.
Jones, P. 2007 Ochre and Rust. Adelaide: Wakefield Press.
Kamminga, J. 1982 Over the Edge: Functional Analysis of Australian Stone
Tools. Occasional Papers in Anthropology. No. 12. University of Queensland
Keely, L. 1980 The Experimental Determination of Stone Tool Uses: A
Microwear Analysis. Chicago: University of Chicago Press.
Kenyon, A.S. and D. L. Stirling 1900 Australian Aboriginal Stone Implements,
A Suggested Classification’. Proceedings of the Royal Society of Victoria. 13:
191-197.
Kingshott, J. 1992 Making and Modifying Woodworking Tools. Lewes: Guild
of Master Craftsman Publications Ltd.
119
Toolkits and utility in Australian lithics
Kuhn, S. L. 1990 A Geometric Index or Reduction for Unifacial Stone Tools.
Journal of Archaeological Science 17: 585:593
Kuhn, S. L. 1994 A Formal Approach to the Design and Assembly of Mobile
Toolkits. American Antiquity 59(3): 426-442.
Law, Boone. W. 2005 Stone Artefact Reduction, Mobility, and Arid Zone
Settlement Models: A Case Study from Puritjara Rockshelter. In C. Clarkson
and L. Lamb Lithics Down Under: Australian Perspectives in Lithic Reduction
Use and Classification. pp.81-94. British Archaeological Reports,
International Series 1408. Oxford: Archaeopress
Legge, R. W. 1929 Tasmanian Stone Culture: Some Notes on Distinctive
Types, Spokeshaves, Borers, and Chipping Tools, and Their Probable
Usages. Royal Society of Tasmania, Proceedings and Papers 1927-1929 pp.
39-43.
Lev, E. Kislev, M. E. and O. Bar-Yosef 2005 Mousterian Vegetal Food in
Kebara Cave, Mt. Carmel. Journal of Archaeological Science 32:475-484.
Lourandos, H. 1997 Continent of Hunter-Gatherers: New Perspectives in
Australian Prehistory. Melbourne/Cambridge: Cambridge University Press.
Luedtke, B. E. 1992 An Archaeologist’s guide to Chert and Flint.
Archaeological Research Tools 7 - Institute of Archaeology. Los Angeles:
University of California.
Mallouf, R. J. 1982 An Analysis of Plow-damaged Chert Artefacts: The
Brookeen Creek Cache (418HI86), Hill County, Texas. Journal of Field
Archaeology 9(1):79-88.
McBrearty, S. and C. Stringer 2007 Palaeoanthropology: The Coast in
Colour, Nature 449: 793-794.
120
Toolkits and utility in Australian lithics
McBryde, I. 1977 Determinants of Assemblage Variation in New England
Prehistory: Environment, Subsistence Economies, Site Activities, or Cultural
Tradition? In R.V.S. Wright (ed) Stone Tools as Cultural Markers: Change,
Evolution and Complexity. pp. 224-250. Canberra: Australian Institute of
Aboriginal Studies.
McCarthy, F. D. 1967 Australian Aboriginal Stone Implements, including
Bone, Shell and Teeth Implements. Sydney: Australian Museum.
McCarthy, F. D. 1976 Australian Aboriginal Stone Implements, including
Bone, Shell and Teeth Implements. 2nd edition. Sydney: The Australian
Museum Trust.
McCarthy, F. D. 1977 The Use of Stone Tools to Map Patterns of Diffusion. In
R.V.S. Wright (ed) Stone Tools as Cultural Markers: Change, Evolution and
Complexity. pp. 251-262. Canberra: Australian Institute of Aboriginal Studies.
McGrath, A. 1991 Travels to a Distant Past. Australian Cultural History 10:
113-124.
Mellars, P. 1970 Some Comments on the Notion of 'Functional Variability' in
Stone-Tool Assemblages. World Archaeology 2(1):74-89
Mitchell, S. R. 1949 Stone Age Craftsmen. Melbourne: Tait Book Co.
Moore, M. W. 2003 Flexibility of Stone Tool Manufacturing Methods on the
Georgina River, Camooweal Queensland. Archaeology in Oceania 38:23-36.
Morton, P. 1989 Fire Across the Desert: Woomera and the Anglo-Australian
Joint Project 1946 - 1980. Canberra: Australian Government Publishing
Service Press.
Mulvaney, D. J. 1969 The Prehistory of Australia. London: Thames and
Hudson.
Mulvaney, D. J. 1975 The Prehistory of Australia. Harmondsworth: Penguin.
121
Toolkits and utility in Australian lithics
Mulvaney, D. J. 1977 Classification and Typology in Australia. In R. V. S.
Wright (ed) Stone Tools as Culture Markers: Change, Evolution and
Complexity. pp. 263-268. Canberra: Australian Institute of Aboriginal Studies.
Mulvaney, J. and J. Kamminga 1999 Prehistory of Australia. St. Leonards:
Allen and Unwin.
Myer, A. 1989 Reliable and Maintainable Technological Strategies in the
Mesolithic of Mainland Britain. In R. Torrence (ed) Time, Energy and Stone
Tools. Cambridge: Cambridge University Press.
Nilsson, S. 1838 Skandinaviska Nordens Urinvanare. In J. Lubbock The
Primitive Inhabitants of Scandinavia. 1868. 3rd edition. London: Longman
Green and Co.
O’Connell, J. F. 1977 Aspects of Variation in Central Australian Lithic
Assemblages. In R. V. S. Wright (ed) Stone Tools as Culture Markers:
Change, Evolution and Complexity. pp. 269-281 Canberra: Australian
Institute of Aboriginal Studies.
Odell, G. H. 2004 Lithic Analysis: Manuals in Archaeological Method, Theory,
and Technique. New York: Kluwer Academic.
Odess, D. and J. T. Rasic 2007 Toolkit Composition and Assemblage
Variability: The Implications of Nogahabara 1, Northern Alaska. American
Antiquity 72(4): 691-717.
Pretty, G. L. 1977 Archaeology in South Australia. S.A. Year Book 1977. D. J
Woolman, Government Printer South Australia.
Roddom, W. 1997 Like, But Oh How Different: Stone Point Variability in the
Top End, N.T. Unpublished B.Sc. Honours, Northern Territory University.
122
Toolkits and utility in Australian lithics
Robertson, G. Attenbrow, V. and P. Hiscock 2009 Multiple Uses for Australian
Backed Artefacts. Antiquity 83: 296-308.
Rogers, R. A. 1985 The Leppke Site: Early Archaic Hunters in Central
Kansas. Transactions of the Kansas Academy of Science 88(3-4):116-120.
Roth, W. E. 1897 Ethnological Studies of the North West Central Queensland
Aborigines. Brisbane: Government Printer.
Roth, W. E. 1904 Domestic Implements, Arts and Manufactures. North
Queensland Ethnology Bulletin #7. The Home Secretary’s Department.
Department of Public Lands, Queensland.
Sackett, I. R. 1982 Approaches to Style in Lithic Archaeology. Journal of
Anthropological Archaeology 1: 59-112.
Semenov, S. A. 1964 Prehistoric Technology. London: Cory, Adams and
MacKay.
Schiffer, M. B. 1972 Archaeological Context and Systemic Context. American
Antiquity 37(2): 156-165
Schiffer, M. B. 1985 Is There a Pompeii Premise in Archaeology? Journal of
Anthropological Research 41: 18-41.
Schiffer, M. B. 1987 Formation Process of the Archaeological Record.
Alburquerque: University of New Mexico Press.
Shennan, S. 2003. Genes, Memes and Human History: Darwinian
Archaeology and Cultural Evolution. London:Thames and Hudson.
Sheridan, G. 1979 Tulas and Triodia: A Multidisaplinary Investigation of the
Mechanics and Antecedents of the Koondi Tulha and their Implications for
Prehistory. Unpublished M.A. thesis. Australian National University.
123
Toolkits and utility in Australian lithics
Shott. M. J. 2009 Systematic Properties of Stone Tool reduction: Curation
Analysis of Palaeoindian Bifaces and Unifaces. In J. Wilkins and K. Anderson
(eds) Tools of the Trade: Methods, Techniques and Innovative Approaches in
Archaeology. pp. 91-106. Calgary: University of Calgary Press.
Shott, M. J. and M. C. Nelson 2008 Lithic Reduction, Its Measurement, and
Implications: Comments on the Volume. In W. Andrefsky Jr. (ed) Lithic
Technology: Measures of Production, Use and Curation. pp 23-48.
Cambridge: Cambridge University Press.
Simek, J. F. 1984 A K-Means Approach to the Analysis of Spatial Structures
in Upper Palaeolithic Habitation Sites. British Archaeological Reports,
International Series 205. Oxford: Archaeopress.
Smale, D. 1973 Silcretes and Associated Silica Diagenesis in Southern Africa
and Australia. Journal of Sedimentary Petrology 43(4):1077-1089..
Spencer, B and F. J. Gillen 1899 The Native Tribes of Central Australia.
London: MacMillan.
Stephens, C.G. 1964 Silcretes of Central Australia. Nature 203(4952): 1407
Stephens, C.G. 1966 Origins of Silcretes of Central Australia. Nature 209
(5022): 496-497.
Taçon, P. S. C. 1991 The Power of Stone: Symbolic Aspects of Stone Use
and Tool Development in Western Arnhem Land, Australia. Antiquity.
65:192-207.
Thorley, P. 2001 Uncertain Supplies: Water Availability and Regional
Archaeological Structure in the Palmer River Catchment, Central Australia.
Archaeology in Oceania 36(1): 1-14.
124
Toolkits and utility in Australian lithics
Tindale, N. B. and B. G. Maegraith 1931 Traces of an Extinct Aboriginal
Population on Kangaroo Island. Records of the South Australian Museum. 4
(3): 275-289. Adelaide: Museum Board.
Tindale, N. 1941 The Hand Axe used in the Western Desert of Australia.
Mankind. 3:37-41.
Tindale, N. B. 1965 Stone Implement Making among the Nakako,
Ngadadjara and Pitjandjara of the Great Western Desert. Records of the
South Australian Museum 15(1): 131-164.
Tindale, N. B. 1972 The Pitjandjara. In M. G. Bicchieri (ed) Hunters and
gathers Today. pp.217-268. New York: Holt, Reinhardt and Winston.
Torrence, R. 1983 Time Budgeting and Hunter Gatherer Technology. In G.
Bailey (ed) Hunter-Gatherer Economy in Prehistory. pp 11-22. Cambridge:
Cambridge University Press.
Willey, G. R. and P. Phillips 1958. Method and Theory in American
Archaeology. Chicago: Univ. of Chicago Press.
Wilmsen, N. E 1968 Functional Analysis of Flaked Stone Artifacts. American
Antiquity 33(2): 156-161.
White, Isobel. (ed) 1985 Daisy Bates: The Native Tribes of Western Australia.
Canberra: National Library of Australia.
Whitten, D. G. A. and J. V. R. Brookes. 1972 The Penguin Dictionary of
Geology. Harmondsworth UK: Penguin.
Wyckoff, D. 1996 The Westfaht and Engle Bifaces: Isolated Finds of Large
Bifaces on the Southern Plains. Plains Anthropologist 41(157): 287-296.
Wymer, J. 1982 The Palaeolithic Age. New York: St Martins Press.
125
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement