Content uploaded by Robert G. Bednarik
Author content
All content in this area was uploaded by Robert G. Bednarik on Jul 23, 2021
Content may be subject to copyright.
89
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
KEYWORDS: Pleistocene – Petroglyph – Age estimation – Lithology – Australia
AUSTRALIAN ROCK ART OF THE PLEISTOCENE
Robert G. Bednarik
Abstract. The recognition of the occurrence of Pleistocene rock art in Australia is reviewed
in the context of historical developments as well as recent observations. The frequency of
them with the Pleistocene rock art of other continents their close similarity with traditions
elsewhere belonging to Mode 3 lithic industries is noted.
Introduction
A few researchers have long held the view that some
amounts in fact to the earliest intimation of a human
presence in the continent before the Holocene. It is
outside of Europe. Archaeological evidence for this idea
art occurring in Australia is almost as old as the general
acceptance of Pleistocene art anywhere
else (Cartailhac 1902). It was presented
South Australian geologist and medical
and explorer. This polymath pioneered
rock art research by investigating a series
of rock art sites in various parts of his
his participation with the 1903 South
Australian Government North-West
Allen Wells (Basedow [Welch] 2008). He
commenced publishing anthropological
papers in 1904. In commenting on the
petroglyphs of the Yunta Springs (Olary
district) and Red Gorge (Flinders Ranges)
places where it would now be almost impossible
mass must have occurred since the designs were
made (Basedow 1914). He also noted the ubiquity
being familiar with the fossil megafauna found at
animal track petroglyph could represent the extinct
diprotodon.
Aborigines were assumed to be relatively recent arri-
century — although a few other investigators reported
chaeologist and entomologist Norman
Tindale (1900–1993) later also visited
Yunta Springs and speculated that ima-
megafauna. He considered a series of
of Genyornis
229) suggested that large macropod
of Procoptodon. Charles P. Mountford
maze petroglyph at the Panaramitee
Figure 1. Herbert Basedow,
pioneer of Australian rock
art research.
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
90
Crocodilus porosus
no such species has existed in southern Australia for
millions of years. Ronald Berndt (1987) secured a very
detailed indigenous interpretation of the complex
which the image depicts a yarida
representing the spirit body of a human being as well
also reported what they perceived to be depictions
North and Yunta Springs respectively. On that basis
alone they proposed that these images were created
at a time when the ‘sea must have been closer to the
subsequently withdrawn by them (Mountford and
Most Australian megafauna had disappeared by
around 20
of such extinct species or their tracks would have to
proposition that track sizes are intended to be realistic.
reading is known to fail most of the time (Macintosh
more claims of imagery supposedly of Pleistocene
animal species in recent decades. Among them are the
suggestions of the depiction of extinct megafauna by
Chaloupka in Arnhem Land (Murray and Chaloupka
Willing (2009) for the Kimberley.
While it is not possible to conclusively exclude
the possibility that Pleistocene Australians depicted
time most megafauna still existed. The only extinct
been reported from the Pilbara and Arnhem Land
these images are of very naturalistic appearance and
seem to show a good number of presumed diagnostic
Western Australia at least until 3300 years ago (in
beyond that. Its extinction on the Australian mainland
thylacine populations could have existed well into the
late Holocene. For instance the level of repatination
of the many supposed thylacine depictions in the
Dampier Archipelago places them well within the last
three or four millennia. Geological observations are
far more relevant to the question of Pleistocene rock
initial observations concerning geological processes
more pertinent than the subsequent contemplations of
motif interpretations.
The proposition of a Pleistocene age of the cave art in
although here the circumstantial archaeological evi-
dence presented was fairly comprehensive. Since no
rock art covered by Pleistocene sediment was reported
and no direct dating method was applied to rock art
anywhere in the world until that same time (Bednarik
and excessive claims appeared like mushrooms
the Pleistocene age of Spanish and French rock
Figure 2. Petroglyph depicting a yarida object,
Panaramitee North site, the type-site of the
‘Panaramitee style’ (aer Mountford 1929).
Figure 3. Presumed depiction of a pair of thylacines, Tom
Price Site 1, Western Australia, re-discovered and
recorded in 1968.
91
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
evidence in its favour became simply
zoomorph in south-western Europe to the
Upper Palaeolithic period — particularly
equine and bovine images (Bednarik
2009a). Many of these recent petroglyphs
did in fact not even exist when Cartailhac
(Fig. 4) wrote his famous mea culpa (Bed-
narik 2009b). Australia experienced a
similar development: after strenuous
eventual acceptance led to many excessive
claims. It is one of several purposes of this
paper to analyse these together with the
to establish a more reliable knowledge
base for Pleistocene rock art in Australia.
Fallacies about rock art age
Most of the published errors about the age of
Australian rock art seem to be related to simple
misapprehensions concerning geochemical and geo-
morphological issues. Perhaps most consequential
among these is the question of organics found in
rock substrates at or near rock art. The ubiquity of
accretionary deposits and even in the putatively un-
altered rock fabric was demonstrated at petroglyph
sites in the 1970s (Bednarik 1979). This work also
showed that the exponential increase in organic
logical regimes is an open system. This means that
it remains open to contamination by many factors
carbon isotopes. Yet in many subsequent endeavours
Nobbs and Dorn 1988).
Dorn sought to estimate petroglyph ages by ana-
lysing rock varnish covering such rock art in the United
K) are compared with the supposedly more stable Ti
isotope samples for AMS analysis are obtained from
near the petroglyphs in question (the damage caused
by AMS sampling prevents sampling of the varnish on
the actual petroglyph). His work in the Olary district
the reliability of cation ratio analysis be-
and Dorn conceded that it ‘is an inferior
the Olary results on the same motifs (Fig.
eventually led to the retraction by Dorn
author had in 1987 sent him the 1979 paper in which
the openness of the carbon system in rock substrates
and the randomness of carbon presence had both been
demonstrated.
newspaper reported the view of a prominent Aus-
tralian archaeologist that ‘[w]e can’t date the [Dampier]
petroglyphs because there is nothing organic about
Figure 4.
Émile Cartailhac in 1872.
Figure 5. Rapidly eroding petroglyph panels at Karolta
1, a site where motifs sampled by Dorn were re-
sampled by Watchman within a few years, yielding
entirely dierent results.
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
92
(1990) reported from what they claimed was blood
no value to dating these motifs. The principal analyst
thoughts’ about these results and returned to Laurie
Creek for more detailed analytical work concerning
the ‘date’ of 20 320 + 3100 / -2300 years . He found
that the reported pigment layer was in fact naturally
precipitated iron oxide of a type frequently occurring
blood residue (Nelson 1993). Although Loy (1994)
continued to claim that mammalian IgG was present
Gillespie’s (1997) subsequent research (see also Tuross
natural organic compounds.
AMS analysis of Australian rock paintings was
introduced by McDonald et al. (1990) who applied it
to charcoal pigments at Gnatalia Creek and Waterfall
Cave in New South Wales. Two results from what is
The most likely explanation for these profoundly
which questions the integrity of all such carbon
isotope results. Alternative explanations for these
Other misunderstandings about the age of rock art
abound in the archaeological literature of Australia.
2002: 133). Yet Dragovich states unambiguously that
her samples were from rock that was not engraved (e.g.
cited above.
for the petroglyphs on a boulder he excavated in
stratigraphy (Morwood 1981: Fig. 7) shows that he
misread the section: the engraved boulder rests on
many petroglyphs of the Dampier Archipelago in
speculating on that basis that some motifs might
repatination curve secured from another part of the
Pilbara region (Bednarik 2009c).
rock art at Dampier was created by Lorblanchet
unexplained. Based on their degree of repatination
far again without presenting testable or credible data.
Unsuccessful endeavours to locate petroglyphs at
Figure 6. Map of Australian rock art sites mentioned in
the text. For inset map (Olary-Flinders Ranges), see
Figure 14.
1 – Murujuga/Dampier; 2 – Depuch Island; 3 – Pilbara;
4 – Spear Hill-Abydos complex; 5 – Tangalma/Carpenters
Gap Shelter; 6 – Kimberley complex; 7 – Jinmium;
8 – Ingaladdi; 9 – Laurie Creek; 10 – Kakadu complex;
11 – Puritjarra and Wanga East; 12 – Carbine Creek;
13 – Saxby Waterhole; 14 – Sandy Creek Shelter;
15 – Early Man Shelter, 16 – Walkunder Arch Cave
17 – Turtle Rock; 18 – Ken’s Cave; 19 – Orchestra Shell
Cave; 20 – Koonalda Cave; 21 – Pimba; 22 – Devon
Downs; 23 – Karlie-ngoinpool, Karake, Prung-kart,
Malangine and Koongine Caves; 24 – Paroong and
Yaranda Caves; 25 – Preminghana; 26 – Sundown Point;
27 – Trial Harbour; 28 – Judds Cavern; 29 – New Guinea
2 Cave; 30 – Gnatalia Creek and Waterfall Cave;
31 – Mt Yengo Rockshelter.
93
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
concentrations of rock art refer to present sea level
(Dortch 2002).
The supposedly oldest dated rock painting in
reported from Carpenters Gap Shelter 1 (the site’s
in the Kimberley region of north-western Australia.
In a deposit yielding occupation evidence of up to
that the coating is of anthropic origin. The shelter
is more likely to account for the deposition of iron
haematite does occur in abundance from the time
of earliest known occupation of Australia onwards
(Bednarik 1994a). Therefore it is entirely reasonable to
but in view of the severe taphonomic truncation of
is extremely improbable that anthropic ochre traces
would survive for tens of millennia in a frequently
water-logged sediment of Tangalma.
The perhaps most spectacularly mistaken rock
the far west of the Northern Territory (Fullagar et
cupules at that site was claimed to date from between
. These
sensational numbers exceed the accepted duration
interior of these fragments are not exposed to light
cupules are a Holocene phenomenon.
Less excessive was the suggestion that a red rock
painting in the nearby Kimberley region is in excess
is also unlikely to be correct. The motif in question
mid to late Holocene antiquity (range 1400–4000
years
emerged (Bednarik 2001a: 133–4). Recent results by
R. Roberts from four sites of the Indian Lower and
of single grains was necessary [the sediments relate
quartz grains most suitable for OSL dating. Selection
of these grains should enable the most accurate and
precise De estimates to be obtained (Fenwick et al.
2008).
distribution of Australian rock art remains to be dis-
cussed in any systematic or comprehensive form. Apart
the principal encumbrance is the wide adoption of
Maynard’s (1979) tripartite model of Australian rock
art and its three consecutive ‘developmental phases’.
It has hampered the establishment of a credible chro-
nology as much as the archaeological confusions about
dating.
partly because of Maynard’s reliance on Edwards’
below. Site corpora were treated as representing
had often contributed to a given site’s repertoire.
is contradicted at countless sites across the continent
Figure 7. Petroglyph at Murujuga, Dampier
Archipelago, Western Australia, re-discovered in
1968, presumed to depict a thylacine.
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
94
whilst the ‘track and circle’ ‘style’ of the Panaramitee
can be the earliest or the most recent component of
rock art traditions around the world. For instance the
author has demonstrated in a blind test that eight
leading Australian specialists of the ‘Panaramitee
style’ cannot distinguish between it and the styles of
rock art sites in all other continents except Antarctica
proportion of perceived motif types. And the same
including large components of the two other ‘styles’.
it was still produced in the 20th century in the form
in contemporary canvas paintings. Her two other
‘styles’ could be any Holocene age and both have been
produced very recently. Maynard did not discount the
possibility that her three styles might to some degree
have emphasised their consecutiveness. Clegg (1992)
mistakenly claims:
There are accepted minimum dates [for the
Panaramitee style] of 13
10
ratio dates range from 31 700 to 1400 years ago
(Nobbs and Dorn 1988: 112–3).
Much the same is claimed in Clegg (1987: 241–2).
Yet Rosenfeld maintains on several occasions that her
dating at Early Man Shelter in Cape York Peninsula
refers to petroglyphs that do not belong to Maynard’s
regarded as erroneous. Clegg maintains as recently
as 2009 that the ‘Panaramitee’ petroglyphs of Sturts
Meadows near Broken Hill are ‘more than 10 000
Glacial Maximum and 20 000 years old (Clegg 2009).
entirely undated (Fig. 8). By comparison to other rock
poor lithological supports it is extremely unlikely that
any of the sites’ petroglyphs exceed mid-Holocene
antiquity (see below).
Another issue with the ‘Panaramitee style’ is its
painstaking study of the very core area of this
provides detailed motif counts from seventeen of the
non-iconic marks at Saxby Waterhole painting site.
Some of the petroglyph sites in the core area of the
hence the claim of consistent percentages is as much
a myth as the claim that the ‘style’ has been dated.
in Edwards (1971). Like most other archaeological
antiquity involved fallacies concerning lithological
In this region we know of no engraved surface from
Figure 8. Sturts Meadows ‘Panaramitee’ petroglyphs,
photographed in 1971.
95
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
with a patination identical to that of
the original rock surface. This seems to
mean that in this area it took a minimum
natural color of the patina on the surface
Edwards misrendered this carefully
no engravings have re-weathered to
match the natural dark rock surface.
As some of them are associated with
surface patination to form in the region
He therefore inverted the statement to
by adding her own opinion to an already
erroneous statement:
Trendall’s view [relating to dolerite from Depuch
have not repatinated to match the surrounding rock
(Maynard 1979: 93).
weathering front formation and repatination. Weathering
is the chemical and physical decay of rocks exposed to
colour or composition from both the unaltered rock and
data of the depth of the weathering zone or ‘weathering
His estimate was not only correct in terms of order of
those of Trendall (see analyses of Trendall’s results in
basalt corresponds to 1.1 Ma in central Europe. Trendall
had stated that a weathering zone of 0.2 inches thickness
quotes Edwards’ citation of a statement originally by
but corrected himself [Mori 1974: 89–90] by retracting a
‘quasi scura quanto’ for ‘tanto scura quanto’.)
Maynard’s misunderstandings are in addition to
Bednarik 1979: 22 for correction) of interpreting
Trendall’s unambiguous and impeccably presented
data. These and other misapprehensions have resulted
weathering front) on repatination rates (see Bednarik
2007: 223).
referring to rock art dating have been misunderstood
information has been misquoted or systematically
misinterpreted. In the generic question of Pleistocene
misinformation is such that it may require decades of
patience to displace the structure of falsities and half-
truths that archaeology has created.
Reviewing the empirical evidence
cene) minimum datings of Australian petroglyphs
were secured at Devon Downs (lower Murray
excavated from layers radiocarbon dated to 4920 ± 100
Pleistocene antiquity of Australian rock art was secured
1971). Carbon isotope dates from excavated and
that the cave was not visited in the Holocene. The
Figure 9. Koonalda Cave nger utings, photographed in 1979.
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
96
huge entrance sinkhole renders human access
apparent Pleistocene torches and the con-
siderable ceiling breakdown succeeding
Pleistocene antiquity for the human activity
900 ± 2000
surface deposit in front of the Squeeze (a
places it on top of the huge rockfall deposit
that commences at the Gallus Site. It is
observable through small openings between
the bl
ka probably applies to at least some of the
More secure is the minimum dating of a
series of petroglyphs at Early Man Shelter
sediment covering the lowest examples of
rock art at that site was in the order of 13
reasonably unambiguous Pleistocene
antiquity for Australian rock art. In the
direct dating
(Bednarik 2001a: 124) results from rock art were acquired
From a sequence of three chronologically discrete traditions
carbon isotope estimates of the early Holocene were obtained
suggested in 1982 how conservative these estimates were: the
cave art tradition in question was suggested to be in excess of
arrangements resembling petroglyphs found on the other
ago. Tasmania became sundered from the mainland about 12
into the Pleistocene.
Direct dating via carbon isotope determinations from
laminated calcium carbonate precipitates has been secured
have occurred since it was executed (Fig. 12). Concerning the
only.
Watchman developed the direct dating of rock art by ex-
Figure 11. Petroglyphs (CLMs and isolated
cupules) on the ceiling of Malangine Cave in
1980, the subject of the rst direct dating of
rock art in the world, which yielded 14C and
Th/U results placing the petroglyphs in the
Late Pleistocene.
Figure 10. Early Man Shelter in 1991, with two Australian pioneers
of rock art research: Percy Trezise, who re-discovered the site, and
Andrée Rosenfeld, who minimum-dated some of the rock art to the
nal Pleistocene.
97
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
Although his initial determinations were of the
(from Kakadu
mineral skins that showed no trace of pigment on the
on a flake yielded oxalate ‘dates’ (or minimum
dates) ranging up to about 24
(Watchman 1993). The nano-stratigraphic sequence
700 ±
about 10 ka to 28 ka.
dated stone surfaces suitable for local calibration (e.g.
older than about 200 years are not available. The
method of calibrating the development of micro-
of impacted surfaces of known age from the same
environmental zone (Bednarik 1992). When a cluster
of many dated inscriptions was found amidst one
became available for application to a selection
2002b). The two oldest dates of a randomly
chosen sample of seven in the Spear Hill/
Abydos area were about 20 ka and 27 ka
clear from the relative weathering state
bearing numerous cupules of clearly greater
been reported by many in Australia (and
elsewhere): that the earliest surviving forms
of rock art seem to be dominated by this
Although the existence of Pleistocene
to facilitate its cognisance it is requisite to
explore its formal characteristics. This is also
essential for the creation of a credible formal
more capricious claims similar to those
listed in the previous chapter are bound to appear in
terra rma
must be to counter the ahistorical design of Maynard’s
is at once evident that they relate exclusively to South
Australian sites. In fact nearly all such references
address the general region between the Flinders
The Olary – Flinders Ranges region
Figure 12. Fossilised nger utings in Yaranda Cave,
which predate megafaunal scratch marks, tentatively
aributed to Thylacoleo; photographed in 1993.
Figure 13. Traditional custodian Monty Hale requested that the age
of the curvilinear petroglyphs he is seated next to be estimated by
the author. The rock art, at Woodstock site 65B, Western Australia,
is between 16 000 and 26 500 years old. Older rock art occurs a few
metres from it.
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
98
sence in Australia. Basedow’s evidence
comprises four parts: the presence at two
of detached ‘tumbled blocks of rocks found
the other portion of which remained in situ
the presence of dark patina or glaze on
being of extinct megafauna species. The
the evidence Basedow cites. The high motifs
he mentions occur on the western side of the
The site consists principally of sub-schistose
is also a sub-horizontal sandstone lens of
about 40 m length and up to a few metres
thickness. Of poor mechanical strength
and it is the removal of blocks of this
sandstone by the creek that has rendered
Basedow’s petroglyphs on the phyllite above in-
accessible. Unless the makers of the high motifs
has to postdate the event of rock art production.
of dating the time of exposure. The probability that
art at these sites found on vertical surfaces occurs at
levels of convenient access from rock ledges. The
were the rock ledges of sandstone destabilised by the
creek and claimed by gravity. Although that remains
antiquity of the high motifs.
Figure 14. Map of the Olary-Flinders Ranges rock art region, with
some of the key petroglyph sites shown.
1 – Sturts Meadows/Eight Mile Creek, Euriowie; 2 – Karolta 1;
3 – Morialpa; 4 – Winnininnie 3; 5 – Panaramitee; 6 – Tiverton;
7 – Yunta Springs; 8 – Manunda Springs; 9 – Stone Chimney Creek;
10 – Orroroo; 11 – Yanyarrie Creek; 12 – Sacred Canyon;
13 – Pertawurtina; 14 – Moolooloo; 15 – Deception Creek/Red Gorge.
Figure 15. Yunta Springs, South Australia, view of the
western clis. The sandstone lens in the lower half of
the image has been truncated since the petroglyphs,
occurring on the overlying phyllite, were made; hence
the face of the sandstone postdates this rock art.
99
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
in October 2009 failed to locate any motifs
some rock climbing skills are required in
a few instances. A detailed examination
of two ‘Genyornis
that this attribution cannot be upheld
Genyornis newtoni is thought to
Creek ‘large bird tracks’ are certainly of
microscopic examination suggests very limited surface deterio-
indication of surface retreat. Individual peck-marks remain well
(Fig. 17). In this lithological and erosive regime it is unlikely
that these motifs would exceed an age of two or three millennia.
Red Gorge would also favour a late Holocene age.
fauna that have so far been examined seem to predate the late
‘Procoptodon tracks’ at Tiverton (Fig. 18). Like all of the region’s
site complex with its thousands of motifs has no prospects of
petroglyphs of the Yunta-Mannahill region occur on metamorphic
mudstones and tillites — among the exceptions being one rock at
localised facies reaching the schist phase. What all of these rocks
have in common is that it is easy to produce both percussion
to form soluble bicarbonate. As the dark-brown accretionary
substrate becomes physically unstable and exfoliates.
Other relevant information derives from sites where petro-
abrasive wear by suspended load. This has occurred for instance at
Figure 18. ‘ Large macropod tracks’, among several found at the
Tiverton main site.
Figure 16. ‘Large bird tracks’ at Deception
Creek Site. Photograph by Livio Dobrez.
Figure 17. Microphotograph of one peck
mark forming part of a ‘large bird track’,
showing the perfect preservation of the
fracture edges. Note manganese deposit,
preferentially forming in the impact pit.
Deception Creek Site.
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
100
Such wear can be by bedload abrasion or the impact
Hartshorn et al. (2002) have shown that small grains
of diameter d
far in excess of what can be expected in the study area.
petroglyph sites between the Flinders Ranges and
orders of magnitude more erodible than quartzite or
whereas the active sediment at the Stone Chimney
Creek site east of Burra does contain some quartz.
with a regime of much greater kinetic energy and far
old.
on a single outcrop of well-metamorphosed siltstone
that contains occasional but rare quartz grains of the
sand fraction. These are rounded and frosted (pre-
A
Figure 19. Petroglyph right on the thalweg
of the Yanyarrie Creek, showing typical
wear by suspended load, with a Degree of
Erasure of 25%.
Figure 20. Circle, ‘bird track’ and ‘macropod track’ petroglyphs at
Manunda Springs Site. The locations of three analysed quartz
grains in the ‘bird track’ and one in each of the two parts of the
‘macropod track’ are indicated by the markers. Three of these
grains yielded microerosion data.
Figure 21. Microphotograph of one of the fractured
quartz grains in the ‘bird track’ motif at Manunda
Springs, showing the curved edge of the fractured
grain that provided ten micro-wane widths A (on the
far le in Fig. 20).
101
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
curve (relating to reasonably similar climatic condi-
a tentative estimate of these motifs’ absolute ages.
track’ is approximately E1440 +180 / -280 years (Fig.
‘macropod track’ are E1870 +440 / -480 years and
E1890 +420 / -270 years respectively (Fig. 23). In
centuries younger.
Only one other rock art site in NE South Australia
Sacred Canyon in the Flinders Ranges (see front cover).
1867 W W
Burke and Wills expedition. (Having been made third-
A
basis of the Spear Hill calibration would correspond to
regarded as being too imprecise to allow a meaningful
curve itself is of such a short range.
coarse sand fraction do occur. Another form at the
site’s two main panels has individual particles of about
400 μm set in a well-sorted matrix uniformly in the
Figure 22. Manunda Springs ‘bird track’ motif,
microerosion analysis, using the Spear Hill
calibration curve.
Figure 23. Manunda Springs, microerosion analyses
of the two parts of ‘macropod track’, with Spear Hill
calibration curve.
Figure 24. Part of the 1867 inscription in Sacred
Canyon, by W. W., a few metres from the main panel.
The marker at the numeral ‘1’ indicates the location of
the fractured quartz grain subjected to microerosion
analysis.
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
102
weathering that may have facilitated the production of
the petroglyphs on this very hard rock. Of particular
because the middle part of a profusely engraved sur-
densely engraved (see front cover). Therefore the two
the petroglyphs covering them. This is starkly evident
in their condition of preservation: those on the lower
mination of the panel demands a Pleistocene antiquity
support that contention. Microscopic examination of
the petroglyphs so high above the ground has been
span and can be subdivided further. For instance a
but there seems to be also recent
cially smallish circles (also clearly
the numerous older petroglyphs (most of which are
they are on the leeside of the distinctive barrier across
the canyon.
of the Pleistocene rock art reached the high locations
of most of the older motifs at this site before the
detachment of part of the cliff face. Of the two
which might suggest that he was not aware of this site.
closely the monumental activity traces sometimes ob-
cave sites in the Mount Gambier district (Bednarik
distinctive cupule panels at Sacred Canyon (and most
other sites of the region) could be seen as excluding
Figure 26. Portion of the upper circle
of the Holocene section, north-
facing main panel of Sacred
Canyon, with marker indicting
the location of the quartz grain
analysed.
Figure 25. Comparison of the (lower) Pleistocene panel with the (upper)
Holocene panel at Sacred Canyon. One of the circles in the upper panel
is estimated to be about 6400 years old (see colour scale). There is very
limited similarity between the two panels, in both style and behavioural
production paern.
103
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
present age estimate for the early palaeoart phase
at this remarkable site is 10–20 ka. It seems feasible
to secure reliable microerosion dates from Sacred
of the upper petroglyphs (Fig. 27). Although much of
the Pleistocene palaeoart at Sacred Canyon is too faint
multiple arcs.
Discussion
such as those found in limestone caves or on granitic
style. The only obvious component they share with that
phase towards ‘variant circles’ or ‘curviform mazes’
is apparent. One of the most surprising aspects of the
with the actual ‘type-site’ of this purported style (see
there are the usual circles and ‘tracks’ at Panaramitee
yarida motif and
typical to represent the construct of the Panaramitee
fervent advocates of the ‘Panaramitee style’ have
Figure 27. The author, an expert rock climber, pointing
to some of the high petroglyphs of Sacred Canyon in
1984. The arrow indicates the circle motif estimated to
be 6400 years old. Photograph by Elfriede Bednarik.
Figure 28. One of many complex designs at Panaramitee North
defying the denition of a ‘Panaramitee style’.
Figure 29. A pisciform petroglyph at Panaramitee
North, apparently early and fairly naturalistic;
recordings by (a) Mountford and Edwards (1962)
and (b) showing individual peck marks and exfoliated
areas.
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
104
about its age — but most importantly to the lumping
because they occur at the same sites. The idea of the
notio
piction of extinct species. Since there is not a shred
art has
independently.
to the ‘Panaramitee style’ was Maynard’s conviction
which must therefore have arrived on the island prior
to its sunderance towards the end of the Pleistocene.
Tasmanian petroglyph sites are either coastal (Sims
and typically not on the coast or on mountains. But the
Tasmanian tradition closely resembles Pleistocene site
early phase (e.g. arcs at Sundown Point) or the Karake
tradition of the Mt Gambier caves (virtually matching
sites such as Preminghana). As in Tasmania and in the
many similarities with the late Karake genre.
The two fundamental errors that were made in the
of the ‘Panaramitee style’ are the following:
1. The most elementary methodological tool of the
archaeologist is the separation of chronological
provided by excavation. No archaeological purpose
of any kind would be served if the excavator of
a site lumped together into one single lot all the
stone implements of an entire deposit spanning
pronounced them as being of the ‘Stone Age’.
archaeological’ approach is precisely what has
rock art to the same chronological separation as
types and pronouncements will be and must be
falsities (Fig. 30).
2. To appreciate the relative longevity of petroglyphs
needs to be understood. The time it takes natural
processes of erosion and weathering to eace a petroglyph
is proportionally similar to the time it takes to create it,
relative to rock hardness and density. Thus if it takes
a thousand times as long to create a 12-mm-deep
cupule on fully metamorphosed quartzite (Kumar
2007) than it takes to make an identical cupule on
weathered sandstone (Bednarik 1998c) — as is
indeed roughly the case — it will take in the order
of a thousand times as long to expunge it on the
Concerning the second point: to create a ‘standard
cupule’ (Bednarik 1998c) on very hard quartzite with
a hammerstone requires in excess of 30
several days of pounding — as shown by the diligent
it also provides a measure of the equally profound
case of the ‘Panaramitee’. What is particularly dis-
turbing about these misapprehensions is that at some
provide a good measure of rock marking ages. Such
that petroglyphs of the same site are of the 20th century
insist that they must be Palaeolithic.
Figure 30. Superimposition of two petroglyph traditions
of greatly diering ages (compare dierence in
accretionary deposits) at Panaramitee North, the
purported type-site of the ‘Panaramitee style’.
105
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
One more observation concerning the taphonomy
of the Broken Hill – Flinders Ranges petroglyph
occur at sites where a stream broke through a rock
barrier. Such barriers range from the perhaps most
pronounced examples at Euriowie and Sacred Can-
somewhat harder than the region’s general lithology.
It may then be that such sites have facilitated selective
preservation of petroglyphs (Bednarik 1994b).
In the case of the professed Pleistocene antiquity
recently investigated such petroglyphs at two central
demonstrated that they are of mid-Holocene ages.
Providing two internally consistent series of 14
from both sedimentary charcoal and calcium oxalate
petroglyph corpora assigned to the ‘Panaramitee’ tend
advocates of this ‘style’. This follows several previous
but probable mid-Holocene age of the numerous circle
petroglyphs in Mt Yengo Rockshelter (McDonald
still produced in central Australia in the 20th century
it not been for the intransigence of the Panaramitee
such as how does one detect Pleistocene rock art in
be assumed to have survived in this country. The
Panaramitee mythology is relinquished. Several cor-
pora or distinctive types of rock art have high prospects
of being of the Pleistocene.
Cupules
These are among the earliest rock art known in the
(Bednarik 2008b). This does not necessarily mean that
that they are then unlikely to be the oldest made.
the presence of cupules as such is not necessarily
sandstone) and no doubt elsewhere. The occurrence
of Lower Palaeolithic cupule panels in southern Asia
Cave petroglyphs
this tradition was certainly continued well into the
‘cave version’ of the archaic linear petroglyph tradition
which matches not only the linear petroglyphs of Tas-
across the continent (Fig. 32).
Figure 31. Production of ‘Panaramitee-style’ petroglyphs
in the 20th century, central Australia. Photograph by
Charles P. Mountford in 1937 at Thompson’s rock-
hole, near The Granites, Northern Territory.
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
106
Archaic linear petroglyphs
tradition or group of traditions of rock art
seems to occur widely across Australia
and is dominated by curvilinear mazes
(the main motif at Sandy Creek Shelter in
perhaps) circles with internal barring and
divided circles. Another apparently late
addition to the repertoire of this ‘tradition’
(hence called ‘trident’ by Rosenfeld) and in
tradition is entirely free of both human
and other animal track-like forms as well
conveys the impression that the concept of individual
motifs is of limited relevance to its manifestations. Its larger
compositions could be seen as combinations of many
Because these archaic linear petroglyphs have long been
neglect of Pleistocene rock art in Australia. Apart from
where a few randomly chosen circular and curvilinear
motifs have provided dates of up to about 27 ka (Bednarik
Pictograms
Watchman’s pioneering work of detecting paint resi-
dues in sequences of oxalate accretions has soundly
established the presence of early ochre applications in
of this Pleistocene tradition of pictograms remains largely
candidates for such age are hand stencils in deep limestone
such antiquity remains to be demonstrated satisfactorily.
Certainly there is ample evidence that haematite has been
previously in three other continents since the Lower
Palaeolithic (Bednarik 1994a).
Figure 32. Cave petroglyphs of the Karake genre, which generally
resembles Pleistocene linear petroglyphs at open sites. Karlie-
ngoinpool Cave, near Mt Gambier, in 1985.
Figure 33. Typical curvilinear maze design of the
archaic linear tradition, surviving in a well-
protected location under a thick silica skin;
Sandy Creek, Cape York Peninsula. The traces of
chalking predate 1982 (see Flood 1987: Pl. 8).
107
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
The Tasmanian issue
any indication of Pleistocene age for rock
dence from cave pictograms (Cosgrove and
One of Maynard’s contentions concerning
the proposed antiquity of her ‘Panaramitee
style’ is its purported similarity with Tasma-
nian petroglyphs. Although that similarity
the case: if one eliminates the Panaramitee
concept from early mainland petroglyph
evidence was not.
Other aspects of early traditions
Pleistocene rock art is the integration of natural
topography) into the arrangement of motifs. This is
aspect also widely observed in the Lower and Middle
that it is sometimes perceived as illustrating cognitive
reactions to pre-existing conditions of a support sur-
its marking strategies. It is much less pronounced in
Another distinctive feature of Pleistocene rock art is the
it seems to be the act of mark production rather than
Quantifying Australian Pleistocene rock art
In the absence of large-scale dating programs it may
be premature to assess the frequency of Pleistocene
of rock paintings or other pictograms having survived
environmental conditions can at open sites survive for
periods of tens of millennia. Taphonomic logic decrees
that this applies especially on very hard and dense
cut petroglyphs survive longest (Bednarik 1994b). The
earliest period seems to be dominated by cupules and
the Americas by cupules and linear grooves (Bednarik
Age’ tradition. Australia is presumed to have been
initially settled by Middle Palaeolithic seafarers
presence of this rock art tradition in India have been
suggested to have imported it with their initial arrival
The Middle Palaeolithic stone tool technology they
also introduced continued in Australia to the mid-
Tasmania up to the British destruction of traditional
rock art in Australia is necessarily of Mode 3 (‘Middle
Palaeolithic rock art traditions. Tasmanian rock art
is dominated by cupules (although they have been
divided circles and circles with internal barring that
are so prominent in the ‘Karake genre’ of the caves
Figure 34. Tasmanian petroglyphs at Trial Harbour, west coast.
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
108
discrete features not intended to depict tracks.
It is possible to speculate about the extent of
Pleistocene rock art in Australia by resorting to the
following reasonable assumptions. Deeply ham-
iconic petroglyphs on particularly erosion-
resistant rock types are probably of the
found in limestone caves. At open sites
these petroglyphs occur usually in arid or
types such as granites and other igneous
well-metamorphosed quartzites. At a
rough estimate the proportion of motifs
that should be expected to fall into this
of the total Australian inventory. Since it
is reasonably estimated that there are at
it follows that over a million petroglyphs
could be expected to have survived from
This may well be higher than the combined
number of surviving Mode 3 petroglyphs
from the rest of the world (few are known
the total number of motifs so far reported from pre-
sumed Upper Palaeolithic or Mode 4 traditions in the
almost exclusively a western European phenomenon
according to present knowledge — although that pro-
position also needs to be tested.
Two fundamental observations follow on from
surviving Middle Palaeolithic (or Mode 3) than Upper
to be appreciated that this could well be a sampling
art (Bednarik 1994b). All surviving Mode 3 rock art
can be regarded as being of the greatest taphonomic
longevity. It should therefore logically be seen as a
which the less deterioration-resistant forms have all
of the Mode 3 petroglyphs should be regarded as
sense as the perceived preference of cave locations for
the production of Mode 4 rock art is almost certainly
as remnant populations that have experienced massive
evident from the composition of the surviving sample
of Upper Palaeolithic mobiliary art: it consists almost
entirely of materials that survive preferentially in high-
Figure 35. Circles with internal barring in Paroong Cave, near Mt
Gambier, in 1983, shortly aer their re-discovery by G. Aslin.
Figure 36. Mode 3 circle and cupules on quartzite at
Klipbak Site 1, southern Kalahari.
109
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
pH sediments and is always found in such sediments.
which also implies a massive taphonomic bias.
Summary
A review of secure or potential candidates of
Australian Pleistocene rock art suggests not only that
it also implies that it consists entirely of a repertoire
having been produced by societies possessing what are
popular notion of art-like productions commencing
with the famous Upper Palaeolithic traditions of the
has marred practically all discussions of the origins
of symboling. Rather than seeing the emergence of
graphic exograms (single entries in an extra-cranial
-
stimuli as a biophysical or biochemical change in neu-
that emerged miraculously in France with the advent
marks modernity in the human lineage — is a gradual
and mostly outside of the Franco-Cantabrian theatre.
remaining Pleistocene palaeoart of the world has been
only in public appreciation but also in preservation
priorities. While sites of Franco-Cantabrian cave
corresponding to their perceived relevance to the
known corpus of Pleistocene rock art occurs but has
a result of archaeological misconceptions has not even
the discovered body of early palaeoart has prompted
of Australia remains severely neglected and most
Pleistocene rock art. There can be no doubt that the
research based on fallacies has contributed to this
Australian Pleistocene palaeoart is considerably more
from an earlier technological context and therefore
can tell us far more about the beginnings of symbol
use and human cognitive evolution. Although similar
material of comparably early nature does occur else-
indications harbours far more Mode 3 rock art than
the rest of the world combined. Large corpora of such
the world’s Mode 3 rock art has survived in Australia.
Instead of celebrating and promoting this incredible
has contributed to its neglect and ongoing destruction:
no European site of Pleistocene rock art would be
sites have consistently experienced.
Acknowledgments
RAR referees of this paper: Dr Livio
Robert G. Bednarik
Australia
E-mail: auraweb@hotmail.com
REFERENCES
-
tinct marsupial Thylacoleo
The Beagle, Records
of the Museum and Art Gallery of the Northern Territory
14: 117–121.
2009. An ancient rock painting
Thylacoleo carnifex-
Antiquityhp://antiquity.
ac.uk/projgall/akerman319/
Journal of Geology
Palestine before the Hebrews
New York.
-
-
Tectonics, cli-
mate, and landscape evolution
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
110
in South Australia. Journal of the Royal Anthropological
Institute
Notes on some native tribes of central
Australia. Compiled and edited by D. M. Welch. Australian
and E. 1998. Ambiguities
in direct dating of rock surfaces using radiocarbon
measurements. Science
Wiener
Neustädter Nachrichten
analysis in Australian archaeology — part 1. The Artefact
4: 14–38.
dating. hp://mc2.vicnet.net.au/home/date/shared_les/
dating 1980.pdf.
Fingerlinientradition. Anthropologie 23: 73–79.
Rock Art Research
2: 80–82.
and Australia. Rock Art Research
Zeitschri
für Ethnologie
Rock Art Research
Australian Aboriginal Studies
La Pintura
17(3): 8–9.
Archaeometry 34(2): 279–291.
Rock Art Research 10:
138–139.
Anthropos
Antiquity
style. AURA Newsleer
Lower Palaeolithic. Current Anthropology
methodology of direct rock art dating. Archaeometry
38(1): 1–13.
Bolleino
del Centro Camuno di Studi Preistorici
Sahul. Anthropos
Journal of
the Royal Society of Western Australia
cave petroglyphs. Geoarchaeology 13(4): 411–418.
Rock
Art Research
The Artefact 22:
Rock art science: the scientic
study of palaeoart
Rock Art
Research
Anthropos
Records of the Western Australian Museum 20: 414–429.
Journal of Archaeological Science
Cave Art Research
part 1. Rock Art Research
art in the world. Cave Art Research 8: 1–12.
Rock Art Research
100.
the question. Rock Art Research
Journal of Archaeological
Science
of petroglyphs. Rock Art Research
and E.
Rock Art Research
1999. Nale Tasih: eine
Floßfahrt in die Steinzeit
and R. G.
Rock Art
Research 22: 147–197.
in prep. Forensic studies
Records of the South
Australian Museum
and S. 1991.
Precision of rock-varnish chemical analyses and cation-
ratio ages. Geology
rock paintings. Archaeology and Physical Anthropology in
Oceania 7(1): 24–30.
changing cultural traditions in a changing environment.
Rock art and prehistory:
papers presented to Symposium G of the AURA Congress,
Darwin 1988
Advances
in dating Australian rock-markings
L’Anthropologie
weathered basaltic crust in absolute chronology. Nature
World
Archaeology
State of the
art: regional rock art studies in Australia and Melanesia
111
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
Sturt’s Meadows guide
guide for 2009 AURA Inter-Congress Symposium in
Broken Hill.
Australian Institute of
Aboriginal Studies Newsleer
Rock Art
Research
Report on the
Aboriginal engravings and ora and fauna, Depuch Island,
Western Australia
distribution of pecked faces in Australia. Memoirs of the
Queensland Museum
and M.
northern Australian prehistory. In R. G. Bednarik and
NEWS 95 Proceedings, Symposium 1A.
ROM).
Origins of the modern mind: three stages in
the evolution of culture and cognition. Harvard University
-
pacities. Behavioural and Brain Sciences
A mind so rare: the evolution of human con-
sciousness
age-determination technique. Quaternary Research 20:
49–73.
Aeolian
geomorphology
art perspective. La Pintura
engravings. International Newsleer on Rock Art 2: 10–14.
New light on old art
Angeles.
La Pintura 23(2):
10–11.
14C ages for petroglyphs
Archaeology in
Oceania
(Portugal) engravings with radiocarbon dating. Antiquity
and
D. S. 1992. New approach to the radiocarbon
Annals of the Association of American Geographers 82(1):
1984. Chronometric and
relative age determination of petroglyphs in the western
United States. Annals of the Association of American
Geographers 42: 308–322.
Archaeology and Physical Anthropology in Oceania 11: 32–
44.
Antiquity
Devil’s Lair, a study in prehistory. Western
Australian Archaeology
Aboriginal rock engravings: a review of problems and
prospects. Archaeology in Oceania
Search
Australian Archaeology
Search
Search
Mankind
in South and central Australia. Transactions of the Royal
Society of South Australia 90: 33–38.
Aboriginal man and
environment in Australia
2008. Optically stimulated luminescence
dating of quartz sediment associated with Middle Palaeolithic
assemblages in India. Poster presented at December 2008
Noosa Heads.
Rock Art Research
Rock art of the Dreamtime
Sydney.
Geological
Society of America Bulletin, Part II 91: 2189–2213.
1997. Mode 3 technologies and the
evolution of modern humans. Cambridge Archaeological
Journal
Rock art and prehistory:
papers presented to Symposium G of the AURA Congress,
Darwin 1988
and L. M.
human occupation of northern Australia: archaeology
Northern Territory. Antiquity
Helictite
Archaeology
of the Gallus Site, Koonalda Cave
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
112
Form in indigenous art
markings in Europe and Australia’. Rock Art Research
Science
278: 220–222.
oxalate: further studies on organic carbon content and
radiocarbon age of materials relating to Australian rock
art. Antiquity 71: 430–437.
1930. Notes on some human
Records of the South Australian Museum
and G. F.
Records of the South Australian Museum 9(4).
and R. L.
2002. Climate-driven bedrock incision in an active
mountain belt. Science
Archaeological research in Kakadu National
Park
Canberra.
in India. Paper presented to the First International Cu-
Australian rock art: a new synthesis. Cam-
State of the art: regional rock art
studies in Australia and Melanesia
Continent of hunter-gatherers: new
perspectives in Australian prehistory. Cambridge University
Australia. Helictite
Australia: a reply to Nelson. Antiquity
and R. 1990. Accelerator radiocarbon
dating of human blood proteins in pigments from Late
Pleistocene art sites in Australia. Antiquity
A guide to the geology and mineral resources
of South Australia
Adelaide.
region: context and theory in the analysis of a dual-
Rock art and prehistory: papers presented to Symposium G
of the AURA Congress, Darwin 1988
and
B. 1990. Investigating 14C AMS: dating prehistoric
Rock
Art Research 7: 83–92.
Northern Territory. Oceania
Form in indigenous
art
Canberra.
Exploring the visual art of Oceania
Archaeology of the Gallus Site, Koonalda Cave
Canberra.
Archaeology at ANZAAS 1983
and L. K. 1999.
Pleistocene extinction of Genyornis newtoni: human impact
on Australian megafauna. Science
Tadrart Acacus
Antiquity 48: 87–92.
highlands: the stone component. Archaeology in Oceania
Rock Art
Research
Visions from the past: the archaeology of
Australian Aboriginal art
Sydney.
Flinders University of South Australia.
rock engraving at Panaramitee North. Transactions and
Proceedings of the Royal Society of South Australia
248.
engravings of extinct creatures in South Australia. Man
Transactions
and Proceedings of the Royal Society of South Australia
Australia. Anthropos
The prehistory of Australia
Melbourne.
Walbiri iconography. Cornell University
1984. The Dreamtime
animals: extinct megafauna in Arnhem Land rock art.
Archaeology in Oceania
Antiquity
Rock Art Research
1: 91–118.
1988. Age determinations for
rock varnish formation within petroglyphs: cation-
Australia. Rock Art Research
Australian Archaeology
Australia. Journal of the Royal Society of Western Australia
113
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
and M. A. 1990.
occupation site in northern Australia. Nature
and M. A. 1993. Optical dating
Australian Archaeology
Nature 393:
and
G. M. 1999. Optical dating of single and multiple
Archaeometry
and I. 1997. Luminescence dating of
rock art and past environments using mud-wasp nests
in northern Australia. Nature
2000. Rock-picture chronologies and palaeo-
environmental records from fossil mud-wasp nests: pre-
liminary investigations using optical dating. In G. K. Ward
Advances in dating Australian rock-
markings
Australian Institute of Aboriginal Studies Newsleer 3: 37–
40.
Early
Man in north Queensland
Rock art and prehistory: papers
presented to Symposium G of the AURA Congress, Darwin
1988
Oxford.
Form in indigenous art: schematisation in the art
of Aboriginal Australia and prehistoric Europe
Rock Art Research
1998. River longitudinal pro-
Rivers over rock: Fluvial processes in bed-
rock channels
2009. Direct dating
indicates a mid-Holocene age for archaic rock engravings
in arid central Australia. Geoarchaeology 24(2): 191–203.
2000. Landscape response to tectonic forcing: digital ele-
-
Geological
Society of America Bulletin
and D.
1999. Australia’s oldest human remains: age of the Lake
Mungo 3 skeleton. Journal of Human Evolution
of giant bird tracks at Pimba. Records of the South
Australian Museum 9: 381–2.
Depuch Island
Dream road: a journey of discovery. Allen and
blood residues on experimental stone tools. Journal of
Archaeological Science
rich crusts in Australia. Rock Art Research
bration curve: evidence from South Australia. Rock Art
Research
at Australian Aboriginal sites: evidence from rock
Retouch: maintenance
and conservation of Aboriginal rock imagery
in northern Australia. Geoarchaeology
Dating and the earliest known rock art
methods for dating carbon in silica skins and oxalate
Advances in
dating Australian rock-markings
Rock Art Research
and C.
1997. AMS radiocarbon age estimates for early rock
results. Rock Art Research
Archaeology and
Physical Anthropology in Oceania
RAR 27-964
Rock Art Research 2010 - Volume 27, Number 1, pp. 89-113. R. G. BEDNARIK
114 Rock Art Research 2010 - Volume 27, Number 1.
