1996: In defence of Arnhem Land rock art research

Sty-representrltion and .-l boriginal communities in the .Yorthern Territon. Implications for archaeological resc~arch
Folds, R. 1993 Assimilation by any name: Why the Federal Governments attempts to achieve social justice for indigenous
Australians will not succeed. Australian Aboriginal Studies
l993(l):3 1-6.
Hamilton, A. 1987 Equal to whom? Visions of destiny and the
Aboriginal aristocracy. Mankind 17(2): 129-38.
Handler, R. 1993 Anthropology is dead! Long live anthropology!
American Anthropologist 95(4):99 1-9.
Hodder, I. 1991 Reading the Past: Current Approaches to Interpretation in Archaeology. (Second Edition). Cambridge:
Cambridge University Press.
Hudson,R. 1980 Sociolinguistics. Cambridge: Cambridge University Press.
Jackson, M. 1987 On ethnographic truth. Canberra Anthropology
lO(2):l-31.
Lattas, A. 1992 Hysteria, anthropological discourse and the concept
of the unconscious: Cargo cults and scientisation of race
and colonial power. Oceania 63(1): 1- 14.
Linnekin, J. 1992 On the theory and politics of cultural construction in the Pacific. Oceania 62(4):249-63.
McBryde, I. (ed.) 1985 Who Owns the Past. Melbourne: Oxford
University Press.
Marcus, G. and Fischer, M. 1986 Anthropology ar Cultural Critique.
Chicago: University of Chicago Press.
Muecke, S. 19% Textual spaces: Aboriginality and Cultural Studies.
Sydney: New South Wales University Press.
Myers, F. 1986a The politics of representation: Anthropological
discourse and Australian Aborigines. American Ethnologist
13:138-53.
Myers, F. 1986b Pintupi SeK Pintupi Country: Sentiment, Place
and Politics among Western Desert Aborigines. Canberra:
Australian Institute for Aboriginal Studies.
Myers, F. 1988 Burning the truck and holding the country: Property,
time and the negotiation identity among Pintupi Aborigines.
In T. Ingold, D. Riches and J. Woodbum (eds) Property, Power
and Ideology in Hunting and Gathering Societies. London:
Berg.
Pardoe, C. 1992 Arches of radii, corridors of power: Reflections on
current archaeological practice. In B. Attwood and J. Arnold
(eds) Power, Knowledge and Aborigines. Bundoora: La Trobe
University Press.
Rose, D.B. 1992 Dingo Makes us Human: Life and Land in an
Australian Aboriginal Culture. Cambridge: Cam bridge University Press.
Rowse, T. 1992 Remote Possibilities: The Aboriginal Domain and
the Administrative Imagination. Darwin: North Australia Research Unit, The Australian National University.
Simmons, W.S. l988 Culture theory in contemporary ethnohist o y . Ethnohistory 35(1): 1-14.
Smith, L. 1995 What is this thing called postprocessual archaeology
... and is it relevant for Australian Archaeology. Australian
Archaeology 40:28-32.
Tonkinson, R 1991 The Mardu Aborigines: Living the Dream in
Australia's Desert. Fort Worth: Holt. Reinhart and Winston.
Toussaint, S. 1994 Review of Rose, D. 1992 Dingo Makes Us
Human. Cunberra Anthropology i 7( 1 ):9
-3
Toussaint, S. 1996 Review of Muecke, S. 1932 Textual Spaces:
Aboriginality and Cultural Studies. Ocecnia 66(2):17 1-4.
Trigger, D. 1986 Blackfellas and whitetellas: The concepts of
domain and social closure in tlit: ahalysis of race-relations.
Mankind l6(2):99- 1 1 7.
Trigger, D. 1992 Whitefella comin': Aboriginal Responses to
Colonialism in Northern Australia. Cambridge: Cambridge
University Press.
von Stwmer, J. 1982 Aborigines in the uranium industry: Toward
self-management in the Alligator River region? In R.M.
Bemdt
(ed.) Aboriginal Sites, Rlghts and Resource Development.
Perth: University of Western Australia Press.
Wagner, R. 1980 The Invention of Culture. Chicago: University
of Chicago Press.
In defence of Arnhem Land rock art research
Darrell ~ e w i s ' ~ ~
In 1992 two papers dealing with Arnhem Land rock art
were published by Ivan Haskovec. Both contain previously
unpublished illustrations of rock paintings and increase the
range of published material available for scholarly analysis.
One paper examines a particular style of human figure found
between Oenpelli and Magela Creek, which Haskovec has labelled 'Northern Running Figures' or 'NW' (Haskovec l992a).
He begins his examination with an assessment of the existing
Arnhem Land rock art sequences. Then he moves on to discuss the spatial and temporal distribution of the figures, and to
describe the style and its content. Finally, he suggests absolute
dates for the period during which the figures were produced.
1
2
40 Tiwi Gardens Road, Tiwi, NT 0810, Australia.
Ms. received December 1993. Accepted February 1996. Eds note that
the delay in accepting this ms. arose out of prolonged correspondence
between Eds and author over necessary editorial changes.
His other paper examines the claim made by Chaloupka
(1977, 1979, 1983, 1984, 1985) that large paintings of animals
and human figures in naturalistic style occur near the beginning of the Arnhem Land art sequence (Haskovec 1992b).
This paper may be considered to have two parts. The first
is primarily a reappraisal of Chaloupka's written and illustrative evidence, and an in-the-field reappraisal of the overlay sequences at two 'key' sites at Mount Gilruth - the sites
upon which Chaloupka based his original claim. Haskovec
concludes that evidence for the chronological position claimed
by Chaloupka does not exist.
In the second part Haskovec reinterprets Chaloupka's
evidence and claims the discovery of new overlay sequences
which necessitate a rearranging of the sequence of art styles.
He then goes on to suggest absolute dates for this new sequence and to offer hypotheses to account for some of the
changes in the art.
Australian Archaeology, Number 43, 1996
At face value each of these papers appears to be a substantial contribution to the discipline. In fact, both are so burdened
with serious problems - inaccuracies, misrepresentations,
wrong claims and a recurrent failure to situate the discussion
in existing scholarly discourse - that unanswered they do
serious damage to the discipline. This article seeks to address
some of these problems. Space precludes a comprehensive
listing and reader tolerance must also be considered. I will
therefore limit myself to the most serious issues.
'Northern Running Figures of Kakadu National Park:
A Study of a Regional Style'
Haskovec's 'Northern Running Figures' are a style of paintings first made known by Mountford (l 956). They are found
between Oenpelli and Magela Creek, and are characterised
by a sinuously curved body. Haskovec begins with the claim
that: 'Until recently, classification and grouping of Australian prehistoric art has been attempted only on a very broad
(Maynard 1976; Davidson 1936) or on a very localised scale
(Chaloupka 1977)'. This ignores the regional studies of Crawford (1968), Wright (1968), Brandl (1973), McCarthy (1976),
Trezise (1Wl), Morwood (1979, 1980) and others. He then
moves to a brief assessment of existing style sequences proposed for Amhem Land rock art - those of Mountford, Brandl,
Chaloupka and myself. The intention here was to choose the
best model against which to consider the relative chronological
position and absolute age of the 'Northern Running Figures'.
Of the different style sequences available, Haskovec decided that Chaloupka's 1977 model was the most useful for
his purposes. This was an extraordinary choice. The 1977
paper, Chaloupka's first publication, was aimed at identifying styles and determining their relative age at two sites near
Mount Gilruth. The resulting model was intended only to
apply to the immediate location of the two Mount Gilruth
sites which are over 40 kilometres from the nearest 'Northem Running Figures', and therefore has no real bearing on
those figures.
Through subsequent research, Chaloupka was able to
expand and refine this original model, culminating in a major
work published in 1993. This was not available to Haskovec at the time he wrote the papers discussed here, so for
present purposes I will refer only to the most highly developed of Chaloupka's published schemes that was available
in 1992 (i.e. Chaloupka 1984, 1985). This l984/85 model,
applicable to the entire Arnhem Land plateau, includes the
particular style which is the subject of Haskovec's paper.
One would expect that Chaloupka's most recent model
would be the appropriate one for Haskovec to use in his consideration of the 'Northern Running Figures', but he chose
instead the earliest, least detailed, and least appropriate model.
The reference to Chaloupka's 1977 model is not a typographical error. Later in his paper (pp. 149-50) Haskovec
refers to Chaloupka's 'Mount Gilruth sequence' several times.
He (apparently) bases a great deal of his argument on this
sequence, trying to apply it as a pan-plateau model for which
it was never intended. Why Haskovec insisted on using the
model which Chaloupka only ever claimed was applicable
to two sites at Mount Gilruth is baffling; to then criticise
the model when it proves inadequate on a pan-plateau basis
is con~pletelyunfair to Chaloupka.
.4 ustralian Archaeology,
Number 43, 1996
Haskovec claims to have based his argument on Chaloupka's 1977 model, but it is clear that, in part, Haskovec
conflates Chaloupka's 1977 model with his later models - and
with the work of other authors. For example, after telling us
(p. 148) that Chaloupka's 1977 model 'is the most usehl for
the present purposes', he immediately goes on to say:
In this model (Table 1), three long geomorphological periods (pre-estuarine, estuarine, post-estuarine)
are subdivided according to number of superimposed styles. The following discussion refers to Chaloupka's model and the names of the styles he used.
It is necessary to note, first, that in his published papers
Chaloupka did not begin to use the terms pre-estuarine,
estuarine, etc, until 198 1 and the table Haskovec refers to
was first published by Chaloupka in 1983. Second, Haskovec goes on to use terms such as 'X-ray I' and 'X-ray 11'
(pp. 148, 149, 150), 'hooked sticks' and 'broad spearthrowers'
(p. 152), 'Oenpelli Figures' and 'Mountford's Dynamic Figures'
(p. 149). Chaloupka does not use any of these terms in his
published papers.
The first two are terms Haskovec and Sullivan (198653)
used in an unpublished report to the then Australian National Parks and Wildlife Service (now Australian Nature
Conservation Agency or ANCA). Their terms appear to relate
to Chaloupka's 'Descriptive X-ray' and 'Decorative X-ray'
(1 984:46), although they are also similar to the terms 'Type I'
and 'Type 11', used by Brandl as alternative names for what
he otherwise referred to as 'Simple' or 'Early' X-ray paintings
and 'Standard' X-ray paintings, respectively (Brandl 1973:
168-9).
The 'hooked sticks', and 'broad spearthrowers' referred to
by Haskovec are my terms (Lewis 19835, 24; 1988:15, 50).
The name 'Oenpelli Figures', which Haskovec appears to attribute to Chaloupka, is also a term used by me in my BA
Honours thesis (1983: 19-22) and my monograph (1 988:36-8),
based on that thesis. If my work was the source of the term
'Oenpelli Figures', Haskovec needs to acknowledge that I used
the term in the context of discussion of the position of these
figures in the sequence of Arnhem Land art styles. In my
monograph my (tentative) conclusion was that the 'Oenpelli
figures' were a regional style in what I termed the "'Hooked
Stick"1Boomerang Period' (for convenience shortened to
"'Hooked Stick" Period'). As such they should be referred
to as 'Oenpelli "Hooked Stick" figures'.
The last term, 'Mountford Dynamic Figures', he attributes
to Chaloupka (1984), but Chaloupka has never used such a
term. In the table of figures on page vii and in the caption
for an illustration on page 35 of his 1984 monograph, and
nowhere else in that volume, Chaloupka used the term, 'postdynamic figures, Mountford's variation' (Chaloupka's emphasis). If this is the source of the term which Haskovec attributes to Chaloupka (1984), then he has left out the word
'post', and in so doing he has altered the position of this style
in Chaloupka's Arnhem Land art sequence, probably by
several millennia.
Another possible source for the term which Haskovec
attributes to Chaloupka (1984) is an unpublished report to
the Heritage Branch of the Conservation Commission of the
Northern Territory (Chaloupka 1990). In this report Chaloupka has (without explanation) changed the name of the
In defince of .4rnhem Land rock art research
style and moved its position forward in his sequence. Instead
of it being a regional style of Post-dynamic figure, in Chaloupka's new sequence the order of styles from oldest to
youngest is: 'Dynamic Figures', 'Post-dynamic Figures', 'Simple
figures with Boomerangs', 'Mountford Figures', and 'Yam
Figures'. Haskovec does not cite this report in either of the
two papers discussed here, but if it was the source of the term
'Mountford Dynamic Figures', inclusion of the word 'dynamic'
would suggest that it was a variant of the much older Dynamic
Figures style, and would therefore represent a major distortion
of Chaloupka's work.
The trouble Haskovec sometimes has in accurately reporting the research findings of others is not confined to
terminology. For example, with reference to Schrire (1982),
he tells us (p. 152) that Schrire had problems 'seeing a continuous uninterrupted transition of an earlier industry ... into
the estuarine/post-estuarine period'. The fact is that Schrire
(1982:23 1) dismissed apparent anomalies in the archaeological
record, stating her opinion that 'if more extensive suites of
radio-carbon dates were available from all sites, many
apparent hiatuses would disappear'. Drawing parallels with
studies elsewhere in Australia, she concluded (1982:252) that,
The current consensus is that the later tradition
contains the elements of the earlier one to some
extent, so that the transition need not necessarily
be viewed as an abrupt takeover, but rather as a
long-term trend with the addition of new elements
around 5000-4000 BP.
A second example: with respect to Ta~on(1987), Haskovec tells us @. 148) that Taqon 'appeared to reject Chaloupka's
notion of "X-ray painting" as a style', a notion which, 1 might
add, was not restricted to or originated by Chaloupka. There
is nothing uncertain about Ta~on'spublished statement - he
did reject the term.
Example three: Haskovec tells us (p. 150) that Chaloupka
(1985) dates his pre-estuarine period 'between 35,000 and
ten to nine thousand years ago'. Chaloupka's (1985:277) terminal date is actually given as 'between 9-7000 years ago'.
Haskovec also estimates that the change from Chaloupka's
Dynamic to Post-dynamic styles is 'about 8000 BP'. Chaloupka himself places this change at an indeterminate date
towards the upper limit of a time period extending from about
20,000 to 9-7000 BP. Haskovec's attribution to Chaloupka
of a date of 8000 BP for this change is an unacceptable
approximation.
My own work is completely misrepresented. For instance,
Haskovec states (p.148) that, 'Lewis (1988) attempted to categorise Arnhem Land rock paintings in terms of technology
portrayed, rather than by styles'. This is no more correct than
it is to say that Chaloupka categorised the art in terms of
geomorphological periods, rather than by styles. The fact
is that I based my scheme upon that of Brandl (1973), and
expanded Brandl's art sequence by recognising and defining different styles and then placing these within it (1988:
34-8). Let there be no doubt that I used stylistic criteria
throughout my analysis.
Following on fiom the statement above, Haskovec 'summarises' my plateau-wide model and inappropriately contrasts
it with Chaloupka's two-site (1977) model. He claims that
my 'categorisation of the western Arnhem Land art is in
14
some instances identical with Chaloupka's sequence' and 'in
other instances it is a reduction of Chaloupka's sequence'.
This assessment is not correct.
Chaloupka's sequence (1977:247-57) may best be described as consisting of three consecutive 'groups' of paintings
which, in some instances, he refers to as 'styles', and in other
instances as 'periods'. According to Chaloupka, each of these
styledperiods is made up of a number of 'styles', although he
does not offer a definition of 'style'. If we take a broad definition of 'style' as a distinct manner of painting or, in other
words, a particular group of paintings with common characteristics that distinguish it from other such groups of paintings, in some instances Chaloupka's claim is correct However, many of his supposed styles are no more than different
painting techniques or colour combinations. The labels Chaloupka applies to these 'groups' are, from oldest to youngest:
(a) Large Naturalistic Designs (b) Dynamic Figures (c) 'Yam
Figuresf. Chaloupka (1977:245) also mentions the existence
of two 'decorative' X-ray paintings and several beeswax figures
at the Mount Gilruth sites, but these are not formally included in his sequence.
Conversely, my sequence consists of four periods: (a)
Boomerang (b) 'Hooked Stick'/Roomerang (c) Broad Spearthrower, and (d) Long Spearthrower. If each sequence is
considered in terms of periods, mine is not a reduction of
Chaloupka's. Furthermore, if the styles that make up the
periods in the two sequences are compared, no two periods
are identical.
First, I do not yet accept the survival of any pre-Dynamic
style, so 1 have no sequential equivalent to Chaloupka's Large
Naturalistic style, nor do I identify a period anywhere else
in my sequence where only such figures occur. In fact, in
l988 1 came to the following conclusion (Lewis 1988: 7 1-2):
the style of 'large naturalistic animals and humans'
is ill-defined and it may actually consist of a number of similar styles present throughout the entire
sequence of Arnhem Land art. An alternative, and
in my opinion, more likely probability is that most
of the 'large naturalistic style paintings' of the type
that Chaloupka has published post-date 'early' Mirni
[Dynamic] art and pre-date early X-ray art.
Second, in 1977 Chaloupka's Dynamic style contained
Dynamic figures (which are part of my Boomerang Period),
but Chaloupka also included figures which he later reclassified as Post-dynamic figures (i.e. 1977: Figure 7c, Plates 6
and 7). I do not include Post-dynamic figures in my Boomerang Period. Third, Chaloupka's Yam style and my 'Hooked
Stick'/Boomerang Period both contain Yam figures and Rainbow snakes, but Chaloupka appears also to have included
figures with spearthrowers (1977: Figs 18a-f) while I have
included human figures with 'hooked sticks' and boomerangs
(Chaloupka's Post-dynamic and Simple figures with boomerangs), and some Large Naturalistic style animals (1988:72).
Where then are my art periods identical with Chaloupka's?
Finally, Haskovec declares that my model 'has no ability
to isolate culturally identifiable stylistic areas in space'. Again,
this is nonsense. Much of my third chapter is devoted to
outlining regional differentiation of styles in my 'Hooked
Stick'/Boomerang Period (1988:3 1-8), and in chapter 4 (1988:
54, and Figs 24 1, 243-52) I define a regional style of 'long
Australian Archaeology, Number 43, 1996
Lewis
spearthrower figures' located in the Oenpelli-East Alligator
River crossing area. Furthermore, the spatial distribution of
different styles is one of the major keys to my overall analysis
and dating of the Arnhem Land art sequence (1988:79- 105).
In a section titled 'Temporal distribution', Haskovec
(pp. 149-50) discusses the position in the art sequence of the
'Northern Running Figures Style', vis a vis Chaloupka's Mount
Gilruth sequence. He concludes that, 'in the absence of the
Post-Dynamic Figures Style in the [northern Kakadu] region',
the 'Northern Running Figures Style' falls between Chaloupka's Dynamic style and 'X-ray I Style', and may be contemporaneous with either 'Simple-Figures-with-Boomerang Style
or Yam Figure Style'.
This entire discussion is redundant. If anyone should
know where such figures fit into Chaloupka's sequence, it is
Chaloupka. In the most comprehensive published scheme
then available to Haskovec, Chaloupka (1984:34) made it clear
that he believed the 'Northern Running Figures' are the Postdynamic figures in this region. Broadly speaking, the position
suggested by Haskovec is the position already allocated by
Chaloupka, but Chaloupka is somewhat more precise. He
has this style followed, in consecutive order, by 'Simple Figures with Boomerangs', and 'Yam figures', before the appearance of 'X-ray descriptive' (his equivalent of Haskovec's 'Xray 1').
It is difficult to see how Haskovec could have missed
this point. After 'extrapolating' Chaloupka's Mount Gilruth
sequence and placing the 'Northern Running Figures' style
within it, he cites Chaloupka's more recent, more relevant,
expanded, detailed, plateau-wide style sequence. However,
it is not mentioned or used as a model in which to fit the
'Northern Running Figures' - it is referred to solely in terms
of dating.
After 'determining' the position of the 'Northern Running
Figures' relative to other styles, Haskovec goes on to discuss
the appearance, content (including a 'hooked stick'), and absolute age of the style (pp.150-2). Just as he does not mention the fact that Chaloupka has placed this style within his
1984/85 sequence, Haskovec makes no mention of the fact
that I discussed each of these aspects at length in my 1988
monograph. On the one hand, I devoted an entire chapter
to the issue of identifying 'hooked sticks' in the art (1988:
15-29). On the other hand, I argued (1988:36-8) that the style
in question was a regional style in my 'Hooked Stick'h3oomerang Period and (1988: Chapter 7), that this period falls within
9000-6000 BP. I think I can fairly claim that at the time
Haskovec was preparing his paper, my work contained the
most detailed and comprehensive consideration available of
this style, and of 'hooked sticks' depicted in the art (1988:
15-38, 86, Plates 24, 26, Figs 109- 19, 166). Inexplicably,
Haskovec makes no mention of these aspects of my work.
In his conclusion (p. 152), Haskovec claims that his paper
is the 'first attempt to define in space and time a style of
rock picture'. This indicates that Haskovec has an inadequate grasp of the literature. Chaloupka and I have both
described Haskovec's 'Northern Running Figures' as a regional
style located between Oenpelli and Ngarradj Warde Djobkeng, placed it in our respective sequences, and suggested
absolute age limits for its production (Chaloupka 1984: 16,
34; Lewis 1983: 19-22, 1988:36-8, 105). Haskovec's claim
'-1 ustralian Archaeology, Number 43, 1996
also fails to take account of the respective statements of
Chaloupka ( l 984:24-9) and myself (1988:80-6) on the distribution and age of the Boomerang/Dynamic figures, and my
work with respect to paintings of the composite Rainbow
snake and other styles of 'Hooked Stick'Boomerang figures
(Post-dynamic and Simple Figures with Boomerangs) which
I identified ( l 988:34-42).
My final comments on this paper concern Haskovec's
claim (p. 149) that his proposed term, 'Northern Running Figures Style' is 'one which would imply both the general location
of the style in the Kakadu National Park and one of its
major characteristics'. This claim is correct on both counts,
but it is nevertheless an inappropriate name for this style.
If styles were to be named in terms of their compass
location, the relevant unit would be the entire region where
west Arnhem Land rock art is found, not Kakadu National
Park. Kakadu is a European construct of less than twenty
years duration and incorporates a relatively small portion of
the Arnhem Land plateau, on the western side. It has no relevance to the spatial or temporal distribution of west Amhem
Land rock art as a whole. West Arnhem Land rock art extends
From the Wellington Range in the north, at least to the
Cadell River in the east, and to Nitmiluk (Katherine Gorge)
in the south. Taking the upper East Alligator River as the
approximate centre of this distribution, the style in question
is located roughly in the northwest.
As for the 'major characteristic' of 'running', apart £tom the
fact that many of the figures in this style are shown 'standing'
or sitting (e.g. Mountford 1956: Figs 15%22c, 52e; Chaloupka
1984: Fig. 14a), I note that other styles, particularly in the
same locality but also far beyond, are depicted in 'running'
pose (e.g. see Mountford 1956: Figs 14f, 19k, 24,3 1,42e, 46e,
47,52a, 52c; Lewis 1988: Figs 97, 161, 188-9,243-7,249-53).
In some instances Haskovec appears to have trouble correctly identifying his 'Northern Running Figures'. He refers
(p. 150) to a number of paintings reproduced by Mountford
(1956: 116 Fig. 14c) as examples of this style where spearthrowers are present. Stylistically these are different from
the Worthern Running Figures', most noticeably in the absence
of the sinuous body which, by Haskovec's own assertion, is
'The one common characteristic of the style'. They therefore
cannot be uncritically accepted as belonging to that style.
Another example is his Plate 3 (p. 154). This clearly shows
a Dynamic figure, but it is labelled, 'Another example of NRF
Style painting'.
Finally, the chances are high that localised styles will
eventually be recognised throughout the Arnhem Land plateau
and its outliers. If we follow Haskovec's system of nomenclature, we are likely to end up with terms such as 'North,
North-West Fat-legged Figures' or 'Inner Central SouthEastern Big-headed Figures' - somehow I think we can do
better than this. I turn now to Haskovec's other paper.
Mt Gilruth revisited
In this paper, Haskovec's stated objective was to reappraise Chaloupka's claim that large paintings of humans and
animals in naturalistic style precede his Dynamic figures style.
Although somewhat better written than the first paper, proportionally it has just as many problems as the first, and
some of them are actually worse.
15
In defence o/.4rnhem Land rock art research
On the first page (p.61) Haskovec implies, not just that
Chaloupka's model is the 'best for present purposes', but
that it is, at present, the best for any purpose:
So far, the most articulate expression of a perceived order in the Arnhem Land rock art has
been Chaloupka's (1977, 1979, 1983, 1984, 1985);
his chronology and stylistic categories have been
accepted as the most authoritative interpretation
of the subject.
Further on, he appears to contradict this statement:
Some researchers such as Tagon (1987) have tried
to show that the chronology and particularly the
definition of some styles, such as X-ray styles,
are imprecise. Lewis's (1988) review of Chaloup
ka's absolute chronology has largely been accepted
as more realistic.
Then he appears to contradict his contradiction: 'no one
has seriously challenged the validity of Chaloupka's sequence
of relative rock art styles' (much of the remainder of his
paper is an attempt to provide this 'serious challenge').
Whatever Haskovec believes, his last statement is not
correct. I have disputed Chaloupka's claim that prints of
grass, thrown objects and hands, and large naturalistic paintings of animals and humans, are styles with independent
positions at the beginning of the Arnhem Land rock art sequence (1988:67-72). I have argued against the consecutive ordering of his Post-dynamic style, his Simple Figures
with Boomerangs style, and his Yam figures style (ibid:4 l ,
86-95). I have also disputed the existence of Chaloupka's
pre-estuarinelestuarine division (ibid:75-7). I class this as a
serious challenge to Chaloupka's sequence of styles.
Regarding Chaloupka's model, Haskovec states ( p 6 1) that,
'Since first publishing this sequence in 1977, Chaloupka
(1979, 1983,1984 and 1985) has changed the original description very little'. This is also incorrect. I have already explained
that Chaloupka's 1977 sequence consisted of only three named
'styles' - 'Large Naturalistic Animals and Human Beings',
'Dynamic figures', and 'Yam figures'. By 1984-85, he had
named eight styles: those above, as well as 'Hand prints,
grass prints and imprints of thrown objects', 'Post-dynamic
figures', 'Simple Figures with Boomerangs', 'X-ray descriptive'
and 'X-ray decorative'.
Haskovec (pp.62-4) next turns his attention to the various definitions and descriptions of the 'Large Naturalistic style'
offered by Chaloupka (1977, 1983, 1984, 1985). He is rightly
critical of the various inadequacies and inconsistencies in
Chaloupka's work, many of which I pointed out in my 1988
monograph, and notes that Chaloupka includes stick figures
in his 'Large Naturalistic' style. The main problem here relates
to Haskovec's observations and discussion of depictions of
a hooked artefact associated with this style. Every point
Haskovec makes here is a point I made in my 1988 monograph (pp.15-29), but the only time Haskovec cites my research is where he states that 'it is difficult to distinguish
thus between a spearthrower and a fighting pick (see also
Lewis 1988)'. This particular statement oversimplifies the
situation. In my 1988 monograph I explained that this difficulty arises only in some contexts and only with respect to
the cylindrical form of spearthrower. In many instances
contextual clues enable identification of hooked artefacts as
cylindrical spearthrowers. However, in the few published
instances where hooked artefacts are associated with Large
Naturalistic style paintings (e.g. Chaloupka l977:248, Fig. 2;
Lewis 1988:245, Fig. 94), they are portrayed in an ambiguous manner.
After reviewing Chaloupka's descriptions of the Large
Naturalistic style, Haskovec (pp.64-6) examines Chaloupka's
(1977, 1979) Mount Gilruth evidence for the sequential position of these paintings. He begins this section by pointing
out the difficulty in assessing the sequence of superimpositioning of rock paintings (although he then goes on to use
visual assessment of superirnpositioning as his primary methodology). He does not qualiQ this statement in any way, yet
it is primarily a problem where red ochre designs are superimposed on each other. Superimpositions are usually quite
clear where bichrome or polychronle art - normally containing
some relatively coarse-grained pigment - is superimposed
on fine-grained red art. Brand1 (1973: 169, 17 1-4; 1977:
226), Chaloupka (1 977:246) and I (19883-5) have addressed
this problem, but our respective discussions are not referred
to by Haskovec.
Haskovec next lists various other methods for establishing a sequence of styles (pp.64-5). Among these, he mentions
the use of technology portrayed, and cites my work. I actually
use technological items portrayed w establish periods rather
than a sequence of styles (Lewis 1988:13). He also mentions
analysis of resources and economic activities portrayed, but
does not give Chaloupka credit for this. He does not even
mention Brandl's (1973: 166, 172, 177) analysis of artistic conventions, and their changes through time, as a means to
establishing a style sequence.
According to Haskovec (p.65), Chaloupka's chronological table starts with object imprints at 'about 30,000 years BP'.
It is quite clear on Chaloupka's table, which Haskovec reproduces on page 62, that Chaloupka's chronology starts at
35,000 years BP. Elsewhere (p.73), Haskovec tells us that
Chaloupka dates the minimum age for Dynamic figures at
9000 years BP (although in the previous paper he put it at
about 8000 years). This is another misreporting of Chaloupka's work. A glance at Chaloupka's table shows that he has
the beginning of his 'X-ray descriptive' at 7-9000 BP, i.e.,
possibly as early as 9000 BP or as late as 7000 BP. Preceding
his 'X-ray descriptive' he has, in order of increasing age, Yam
figures, Simple Figures with Boomerangs, and Post-dynamic
figures. Then come the Dynamic figures.
If Chaloupka had ciairned that Dynamic figures persisted
until 9000 years ago, as Haskovec states, there would either
be no time for his other styles to be painted, or a maximum
of 2000 years. Chaloupka does not seem to have ever been
explicit on the minimum age of Dynamic figures, but I do
not think it likely he actually contends that his Post-dynamic
figures, Simple Figures with Boomerangs, and Yam figures
replaced each other in a matter of a few years, or even two
thousand years. Given the time period of 11-13,000 years
between Chaloupka's Large Naturalistic and 'X-ray descriptive' styles, and even allowing for the Dynamic style to persist for longer than the others, a more reasonable assumption
is that Chaloupka allowed several thousand years for each
of the post-dynamidpre-'X-ray descriptive' changes to take
place.
Australian Archaeology, Number 43, 1996
Lewis
On page 67, Haskovec begins a section titled, 'Additional
evidence'. This deals with Chaloupka's data from other sites,
published after 1979. Referring to superimposed paintings
of a snake, Dynamic figure, flying fox and hand stencil that
both Brandl and Chaloupka reproduce, Haskovec (p.67) says
'Neither researcher interprets the painting superimposition sequence'. This claim, repeated on page 7 1, is incorrect. Brandl
(1 973: 197, documentation for Fig. 89) states that the snake
underlies the 'early' Mimi (Dynamic) figure and that the
flying fox is superimposed on the spear held by a 'kangarooman'. The 'kangaroo man' in question is a recurrent type which
is well known to be an integral component of the Dynamic
style (e.g. see Brandl 1973, Fig. 73; Chaloupka 1977:253
Fig. 8a; Lewis 1988, Figs 24 and 25).
Haskovec ends this section with two conclusions: Chaloupka has not presented any evidence to explain why stick
figures should be included in the Large Naturalistic style, and
he has presented no evidence that this style precedes his
Dynamic figures. Ignoring the fact that human stick figures
have been found directly associated with large 'naturalistic'
animal depictions (e.g. Lewis 1988: Figs 90, 94, 185) and that
Brandl (1977:237) has argued against stick figures being
considered a style, Haskovec contends that Chaloupka's Large
Naturalistic style is really two separate styles - 'naturalistic'
figures (Ll) and stick figures (L2). He proposes nine possible
sequences of L l , L2, Dynamic figures and Yam figures as
alternatives to Chaloupka's Large Naturalistic style-Dynamic
style-Yam style Mount Gilruth sequence. In the following
section (pp.68-71), he sets out to establish which of his nine
possible alternative sequences is correct.
Haskovec begins by noting 'a puzzling similarity between
the Yam and the Large Naturalistic style paintings'. He points
out that what he interprets as a Large Naturalistic Human
figure (his plate 7) is painted over a Yam style bird and that
it is painted in the same style as Large Naturalistic Human
figures elsewhere. In particular he notes the similarity of
the head with those of Yam figures. On this basis he states
(p.68): 'I believe this to be clear evidence that justifies a rearrangement of the Mt. Gilruth sequence, placing L1 style
after Yam style' (Haskovec's emphasis). Haskovec presents
no evidence for placing the L l style after the Yam style rather
than in it. I suggest that the evidence is more compelling
for the two styles - L 1 and Yams - being contemporary.
I can make my case with reference to Haskovec's Plate
7. First, it would have assisted readers if he had advised
them that the figure in Plate 7 and the figure in his Plate 9,
are part of a single large panel of paintings reproduced by
Chaloupka (1984:38, Fig. 17). Chaloupka captions his photograph as follows: 'Pre-estuarine period, yam figures: multiple
superimpositions of paintings in this style from simple yam
images to human figures including a number of ibises' (Chaloupka's emphasis). This caption indicates that Chaloupka
never considered the particular human figures that Haskovec
reproduces in his Plates 7 and 9 as anything other than Yam
figures.
Second, Haskovec's photograph shows marked differential weathering from left to right. The two 'yam-birds' at
the right are relatively well preserved, but an identical 'yambird1, said by Haskovec to be under the human figure, is
faded. Lines which extend from the 'yam-bird' at top right
Australian Archaeology, Number 43, 1996
fade away as they extend to the shoulders of the human figure.
It is highly likely that the three 'yam-birds' are contemporary,
providing yet another warning on the use of visual assessment
of superimpositioning to determine the sequence of paintings.
Third, examination of the figure in question reveals that
its head does not just 'resemble' the head of a Yam figure it is a 'yam' head. Whether this 'yam-headed' human being
is superimposed on or by the 'yam-birds' is of little consequence. A yam-headed human figure superimposed on (or
by) 'yam-birds' constitutes nothing more than a panel of
superimposed Yam style paintings. Chaloupka's caption is
completely accurate.
On the basis of his supposed overlay evidence, Haskovec
(p.70) proposes a reinterpretation of Chaloupka's Mount
Gilruth sequence as follows:
l . Dynamic style
2. Yam style
3. Large Naturalistic style (Ll)
4. Stick figures (L2)
The only evidence he provides for placing the Stick figures
(L2) as a separate 'style' at the recent end of this sequence is
the hooked stick held by one of Chaloupka's Mount Gilruth
stick figures. After pointing out anomalies in Chaloupka's
interpretation and positioning of this artefact, he states:
It is clear now from the data presented here, that
the painting of the tool in the Stick figure man's
hand and the painting of the man are themselves
from a much later Yam (?) and post-Yam periods,
where there is plentiful evidence for the use of
both the fighting pick and the spearthrower, as well
as the three-pronged fishing spear. (p.7 1)
Haskovec does not acknowledge that most of these points
were made in my l988 monograph (Lewis 1988:17,41,42).
The only points which I did not make, and which I dispute,
are that there is 'plentiful evidence' for the use of the fighting
pick anywhere in the Arnhem Land art sequence or that indisputable spearthrowers occur in the Yam period. In my
1983 BA Honours thesis I argued that the 'hooked sticks' associated with boomerangs in Arnhem Land art were fighting
picks. However, in my 1988 monograph I reconsidered at
length the issue of identifying 'hooked sticks' in the art (Lewis
1988:15-29) and found that it is extremely difficult to identify
reliably any 'hooked stick' as a fighting pick. In fact, my professional opinion since 1988 has been that most and possibly
all of the 'hooked sticks' associated with boomerangs represent
a prototype spearthrower. A final point: in 'Hooked Stick'
Period art, the three-pronged spear, described as a fishing
spear by Haskovec, is sometimes shown piercing a mammal
(e.g. Brandl 1977: Fig. 17; Lewis 1988: Fig. 94), but has not
yet been found piercing or directed at a fish.
Haskovec rounds off this section with the claim that he
has found 'the first evidence for a technology associated with
the Yam style paintings' (a 'hooked stick' which he identifies
as a fighting pick). This is yet another unsound claim. First,
Brandl (1977:232 Fig. 17) reproduced a composition of Yam
style paintings where multi-pronged, multi-barbed spears and
'hooked sticks' are depicted. Second, Chaloupka (1984:34)
mentioned that human figures in the Yam style possess
weapons and he reproduced such a figure with a multipronged, multi-barbed spear (p.39, Fig. 18f). Third, in my
17
In defence of A rnhem Land rock art research
1988 monograph I published five examples of Yam style
motifs where weapons, including 'hooked sticks' and multipronged multi-barbed spears, are portrayed (Figs 129 and
148-51). Largely on the basis of these weapon types l argued
(1988:41-2) that Yam paintings were part of my 'Hooked
Stick'IBoomerang Period.
The next section (p.7 1) is titled 'Archaic paintings'. Haskovec begins by claiming to have found the first example of
a Dynamic style painting superimposed on an earlier painting.
The 'earlier' painting, a crudely painted red figure outlined
in white, is described by Haskovec as 'probably a silhouette
of a person' and he says it 'clearly shows that paintings prior
to the Dynamic style were executed in red and white.' I have
already mentioned that Brandl claimed to have found such
an overlay sequence over twenty years ago (Brandl 1973:
197, documentation for Fig. 89).
Problems inherent in assessing the sequence of superimposed red ochre art aside, Haskovec's assumption here is
that where there is superimposed art, there must be a great
time difference between the acts of painting. The fact is that
there may be minutes or millennia between the first painting
and the second. If a long interval was involved there may
also have been other styles in vogue between the two painting episodes. He notes the possibility that paintings in this
proposed 'Archaic style' might be contemporaneous with the
Dynamic style, but that 'evidence for that is not available yet'.
I would suggest that there is no evidence that the perceived
earlier painting substantially pre-dates the other. Furthermore, the likelihood of any white pigment art surviving from
the pre-Dynamic period is virtually nil - it is far more likely
that the white pigment was added much later or that the
entire painting considerably post-dates the dynamic style.
Haskovec does not define this 'style'. In any case, a new
style cannot credibly be claimed on the basis of only one
painting and a style sequence cannot be deduced on the basis
of only one example of superimpositioning.
Similar criticisms apply to Haskovec's subsequent claim
to have found a Dynamic figure superimposed on a hand
stencil which suggests to him 'that the hand stencils were
already being executed some time before the Dynamic style'.
In addition, it would hardly be surprising if a Dynamic style
figure was found superimposed on a hand stencil as Brandl,
Chaloupka and I have previously noted that hand stencils
are commonly associated with Dynamic figures. Indeed, I
provide a number of illustrations where hand stencils and
Dynamic figures are superimposed (1 988: Plates 12, 14, Figs
3, 8, 21, 27, 38), but I do not claim to be able to reliably
determine the sequence of the motifs. I would suggest that
in both of Haskovec's examples - dynamic figure on red and
white painting and dynamic figure on hand stencil - the
evidence is just as strong for contemporaneity as it is for
either of the motifs to pre-date the period when Dynamic
style paintings were being produced.
On page 72, Haskovec begins a section titled 'Working
hypotheses'. This section reports other people's research inaccurately. In one instance, Haskovec (p.73) refers to the
art of the most recent period as 'a complex of substyles' and
appears to be crediting Tacon (1987) with labelling this complex 'X-ray art'. The fact is that Taqon explicitly rejected
'X-ray art' as a general label for this 'complex of substyles',
18
some with X-ray features and some without, and suggested
instead that it be referred to as the 'Complete Figure style'
(Taqon 1987:43-4).
In another instance (p.72j Haskovec claims that 'Lewis
(1988) estimates the beginning of his Boomerang ... period
to be 12,000 years ago'. This is wrong and it is not just a
misreading of a different date. Nowhere do I claim that
the upper age is 12,000 years, or any other specific age. On
page 86 of my monograph I state that: 'The upper age limit
of the Boomerang figures now seen on the rocks cannot yet
be defined', and on my chronological table (1 988: 105) 1 deliberately left the upper age open-ended.
In what is probably the most serious inaccuracy in either
of the two papers discussed here, Haskovec (p.72) claims that:
Lewis envisages that this painting period is a final
product of people of a desert culture, who, presumably forced by the rising sea levels, diminishing availability of the land and consequent exposure to coastal people, started painting. This is a
difficult proposition to accept. Such an interpretation does not account for the lack of any stylistic
intecedents [sic] to, or the overwhelming homogeneity of, the Dynamic style.
Dear readers, I would be the first to agree that this is a
difficult proposition to accept. Haskovec's statement is accurate only as far as the first comma - From that point on it
bears no relationship to my work. Let me make it clear that
nowhere do 1 suggest that a combination of rising sea level,
loss of land and 'consequent exposure to coastal people' in
any way caused plateau people to suddenly begin painting
Dynamic figures on the rocks, or in any other way stimulate the production of Dynamic style art. In fact, I do just
the opposite, explicitly stating my belief that the production
of Boomerang (Dynamic) figures, 'is only likely to have occurred before the fill effects of post-glacial sea level rise
began to be felt among the societies then living in the Arnhem
Land plateau area' (1 988:8 1, my emphasis). It is bewildering that he should attribute to me the opposite of what 1 have
written.
In my 1988 monograph I devoted over 2000 words and
more than forty-five illustrations to an analysis of Boomerang
(Dynamic) figures, and I cannot hope to do more than summarise my findings here. My interpretation of the Boomerang
Period is that, until convincing evidence to the contrary is
produced, the Boomerang figures must be considered the
earliest paintings in Arnhem Land rock art (1988:80). I
believe that the most likely reason for the apparent lack of
stylistic antecedents to the Dynamic style is, not that there
were none, but rather that few, if any, earlier paintings have
survived on the rock face. My analysis of the homogeneity
of the style throughout the plateau is that the paintings were
executed during a more arid climatic period when sociolinguistic territories were much larger and when people used
art style to emphasise similarity rather than difference - that
is, when people lived within and sustained more open social
networks (1988:80-6). When the effects of sea level rise began
to be felt with sufficient force among plateau societies, groups
began to emphasise social differences by adopting new art
styles - regional styles in what I have termed the 'Hooked
Stick'IBoomerang Period.
Australian Archaeology, Number 43, 1996
Lewis
After completely misstating the effect of my work, and
then criticising my supposed statements, Haskovec goes on
to suggest a hypothesis which he says accounts for the 'sudden
appearance' and homogeneity of the Dynamic style and the
similarity of Dynamic figures with Kimberley Bradshaw figures. According to this hypothesis, the Dynamic style artists
were 'emigrants', displaced by the rising sea, and who (1 presume he is suggesting) brought their painting style with them.
The similarity between Dynamic figures and Bradshaw figures
in terms of painting techniques used, and hair styles and
tools depicted, suggests to Haskovec, 'A common origin of
the styles in the now submerged Arafura river basin' (p.72).
First, Haskovec makes no mention that anyone else has
ever discussed the similarity between early Arnhem Land art
and early Kimberley art. This similarity and apparent cultural
connection has in fact been noted by, among others, Crawford (1968:82), Brandl (1973: 185-7), Chaloupka (1984:55),
and most recently and comprehensively by myself (1988:
84-5, 93-5, 109, 1 1 1).
Second, Haskovec's topography is puzzling when he
suggests a common origin for the two styles in the ancient
'Arafura River basin', now submerged beneath the A r a b Sea.
The 'Arafura River basin' lies between Arnhem Land and New
Guinea; it is difficult to see what connection this has with the
Kimberley. It is the Timor Sea, or more specifically, the Joseph
Bonaparte Gulf, that lies between Amhem Land and the Kimberley. Predictably, Haskovec makes no mention of my discussion of the implications of the Pleistocene 'Bonaparte catchment', a giant Pleistocene river system which directly connected the Kimberley Bradshaw art area and the Amhem Land
Dynamic art area (Lewis l988:85, 88, l l l , maps 3 and 4).
Third, a hypothesis should have at least some supporting
evidence; anything less is unsubstantiated speculation. Haskovec offers no evidence in support of the migration theory which
he claims explains the similarity of style and content in both
regions. His suggestion that a supposed and sudden appearance of Dynamic Bradshaw paintings is best explained by a
migration theory, does not consider alternative possibilities.
I have already stated my belief that the apparent lack of
antecedents to Dynamic art (and Bradshaw art) is a function
of preservation - there is a physical limit to how long rock
paintings can last and the two styles are merely the oldest
surviving artworks in each region. My analysis of the sirnilarity between early Arnhem Land art and early Kimberley
art, based on European models of Palaeolithic information
systems, is that the two regions were originally encompassed
by an extended information network. The change from relatively open to less open social networks, reflected in the change
fYom plateau-wide Boomerang figures to regionalised 'Hooked
Stick' figures in Arnhem Land, is directly paralleled in Kimberley art. As sea level continued to rise, the information
network connecting Arnhem Land and the Kimberleys was
disrupted and the art of the two regions began to diverge.
There are yet other problems which I could address. In view
of what has gone before most of these are not pressing and,
as I said at the beginning, reader tolerance must be considered.
Summary
Both of Haskovec's papers are presented as original and
ground breaking research. The focus of one paper is to define
Australian Archaeology, Number 4 3 , 1996
the 'Northern Running Figures' in time and space. This
issue had been addressed previously by myself and Chaloupka.
The other paper sets out to examine Chaloupka's evidence
that large paintings in naturalistic style precede his Dynamic
figures style. This issue also had also been addressed previously, by myself.
I have shown that in both of Haskovec's papers, many
of the arguments and conclusions presented replicate arguments and conclusions in previous analyses produced by other
scholars. Similarly, I have shown that most of the 'first
discoveries' he claims to have made had already been made
by others.
In both of Haskovec's papers, the work of a number of
other Arnhem Land researchers has often not been given
due credit or it has been inaccurately reported, and I have
illuminated numerous other problems. The end result is
not scholarship at a level we have the right to expect.
References
Brandl, E. 1973 Australian Aboriginal Paintings in Western and
Central Arnhem Land. Canberra: Australian Institute of
Aboriginal Studies.
Brandl, E. 1977 Human stick figures in rock art. In P.J. Ucko
(ed.) Form in Indigenous Art: Schematisation in the Art of
Aboriginal Australia and Prehistoric Europe, pp.220-42.
Canberra: Australian Institute of Aboriginal Studies.
Chaloupka, G. 1977 Aspects of the chronology and schematisation of two prehistoric sites on the Arnhem Land Plateau. In
P.J. Ucko (ed.) Form in indigenous Art: Schematisation in the
Art of Aboriginal Australia and Prehistoric Europe, pp.
243-59. Canberra: Australian Institute of Aboriginal Studies.
Chaloupka, G. 1979 Djarwanmbi - Mt Gilruth. Unpublished
report to the Australian Heritage Commission, Canberra.
Chaloupka, G. 1981 Man's first mark. Territoty Digest, pp.29-33.
Darwin: Northern Territory Information Service.
Chaloupka, G. 1983 Kakadu rock art: Its cultural, historic and
prehistoric significance. In D. Gillespie (ed.) The Rock Art
Sites of Kakadu National Park - Some Preliminary Research
Findings for their Conservation and Management. Canberra:
Australian National Parks and Wildlife Service. Special Publication 10.
Chaloupka, G . 1984 From Palaeoart to Casual Paintings. Darwin:
Northern Territory Museum of Arts and Sciences. Monograph Series 1.
Chaloupka, G. 1985 Chronological sequence of Arnhem Land
Plateau rock art. In R. Jones (ed.) Archaeological Research
in Kakadu National Park, pp.269-80. Canberra: Australian
National Parks and Wildlife Service. Special Publication 13.
Chaloupka, G. 1990 Rock art survey 1990. Arnhem Land
Plateau - Southern Margins. Report to the Heritage Branch,
Conservation Commission of the N.T. Government. Commonwealth of Australia and the Northern Temtory Government.
Chaloupka, G. 1 993 Journey in Time: The World's Longest Continuing Art Tradition. The 50,000-YearStory ofthe Australian
Aboriginal Rock Art of Arnhem Land. Sydney: Reed.
Crawford, 1. 1968 The Art of the Wandjina: Aboriginal Cave
Paintings in Kimberley, Western Australia. London: Oxford
University Press.
Davidson, D. 1936 Aboriginal Australian and Tasmanian Rock
Carvings and Paintings. Memoirs of the American Philosophical Society, Vol. 5 . Philadelphia.
Haskovec, 1. 1992a Northern Running Figures of Kakadu National
Park: A study of a regional style. In J. McDonald and
I. Haskovec (eds) State of the Art: Regional Rock Art Studies in Australia and Melanesia, pp.148-58. Melbourne:
19
In dejence oJArnhem Land rock art research
Australian Rock Art Research Association. Occasional AURA
Publication No. 6.
Haskovec. 1. 1992b Mt Gilruth revisited. Archaeology in Oceania
27:61-74.
Haskovec, 1. and Sullivan, H. 1986 Nujombolmi: The Li/e and
Work of an Aboriginal Artist (2 Vols). Canberra: Report to
the Australian National Parks and Wildlife Service.
Lewis, D. 1983 Art, archaeology and material culture in Arnhem
Land. B.A.(Hons) thesis, Department of Prehistory and
Anthropology, The Australian National University, Canberra.
Lewis, D. 1988 The Rock Paintings of Arnhem Land, Australia:
Social, Ecological and Material Culture Change in the PostGlacial Period. Oxford: British Archaeological Reports,
International Series 4 15.
Maynard, L. 1976 An archaeological approach to the study of
Australian rock art. Unpublished MA thesis. University of
Sydney, Sydney.
McCarthy, F. 1976 Rock Art of the Cobar Pediplain in Central
Western New South Wales. Canberra: Australian Institute
of Aboriginal Studies.
Morwood, M. 1979 Art and stone. Unpublished Ph.D. thesis,
Department of Prehistory and Anthropology, The Australian
National University, Canberra.
Morwood, M. 1980 Time, space and prehistoric art: A principal
component analysis. Oceania 15(2):98-109.
Mountford, C. 1956 Art, Myth and Symbolism: Records of rhe
American-Australian Scientific F~peditionto Arnhem Land.
Vol. l. Melbourne: Melbourne University Press.
Schrire, C. 1982 The Alligator Rivers: Prehistory and Ecology in
Western Arnhem Land. Canberra: Department of Prehistory,
Research School of Pacific Studies, The Australian National
University. Terra Australis 7.
Ta~on,P. 1987 Internal-external: A re-evaluation of the 'X-ray'
concept in western Arnhem Land rock art. Rock Art Research
4(1):36-50.
Trezise, P. 1971 Rock Art of South-East Cape Z'ork. Canberra:
Australian Institute of Aboriginal Studies.
Wright, B. 1968 Rock Art of the Pilbara Region, Nor-tlz-West Alcstralia. Canberra: Australian Institute of .Aboriginal Studies.
Occasional Papers in Aboriginal Studies No. I l . Prehistoty
and Material Culture Series N o . 2 .
2$
A preliminary analysis of basal grindstones from the Carnarvon Range,
Little Sandy Desert
Peter ~ e t h and
'
Sue
conn no?
Australian prehistorians have devised morphological categories for grindstones in order to separate those assumed to
be of a generalised knction fiom those used more specifically for grinding seeds (cf. Cane 1984; Smith 1985, 1986).
This division is a functional one, with grindstones, or 'arnorphous grindstones', being multipurpose tools used to grind not
only plants but also other materials such as animals and minerals. In contrast, seedgrinding implements are described as
being more specialised tools. Their primary, if not exclusive,
function is to process edible seeds. Amorphous grindstones
are characterised as having a flat grinding surface, while formal grindstones used in the wet milling process have one or
two deeper grooves.
In challenge to these morphological categories, it has been
argued that the two extreme types of form are functional variants of the same implement, representing its successive reduction through time, often presumably to its final discard state
(Cane 1989:113; Gorecki et al. forthcoming). It is noted that
grindstones are far more efficient and versatile in earlier stages
of reduction than when worn (cf. Gorecki et al. forthcoming).
Tindale (1974:lOO) suggested that the discard behaviour
surrounding these implements not only relates to the degree
of wear but also to the availability of new stone supplies.
1
2
Dept of Anthropology and Archaeology, Jarnes Cook University of
North Queensland, Townsville, QLD 48 1 1, Australia.
Division of Archaeology and Natural History, Research School of
Pacific and Asian Studies, The Australian National University,
Canberra, ACT 0200, Australia.
When the raw material used for grindstones is derived from
distant sources it is more likely to be used to the maximum
of its functional life.
We believe that the dichotomous classification between
amorphous grindstones and millstones is questionable and to
illustrate the point we present data recently collected from
basal grindstones at a major occupation site in the Little Sandy
Desert, Serpent's Glen (Fig. 1). The site complex lies on the
southern edge of the Carnarvon Range and contains an extensive and diverse flaked stone assemblage, large quantities of
grinding material and a painted rockshelter, which has recently
been dated to the Pleistocene (O'Connor et al. in prep.). Stone
supplies for grindstones are abundant at the site. Relevant
attributes were recorded from a sample of the first 100 intact/
near-intact basal grindstone platforms located within a 150 m
radius fiom the main rockshelter at Serpent's Glen. Attributes
included:
a. type of raw material,
b. length of maximum axis,
c. width of maximum axis,
d. thickness normal to plane of other axes, and
e. depth of ground depression below original rock face.
Data for these attributes are presented in Table 1. Additional
notes were also taken of such features as rejuvenation and
pecking. The context of the site and typical grindstones are
shown in Figures 2 and 3.
Australian Archaeology, Number 43, 1996