ArticlePDF Available

Serpent's Glen Rockshelter: Report of the first Pleistocene-aged occupation sequence from the Western Desert

Authors:

Figures

Content may be subject to copyright.
Australian Archaeological Association
is collaborating with JSTOR to digitize, preserve and extend access to
Australian
Archaeology.
http://www.jstor.org
Serpent's Glen Rockshelter: Report of the First Pleistocene-Aged Occupation Sequence from the
Western Desert
Author(s): Sue O'Connor, Peter Veth and Colin Campbell
Source:
Australian Archaeology,
No. 46 (Jun., 1998), pp. 12-22
Published by: Australian Archaeological Association
Stable URL: http://www.jstor.org/stable/40287366
Accessed: 06-02-2016 01:54 UTC
REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/40287366?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.
This content downloaded from 130.56.64.29 on Sat, 06 Feb 2016 01:54:55 UTC
All use subject to JSTOR Terms and Conditions
Serpent's Glen Rockshelter: Report
of the
first
Pleistocene-aged
occupation sequence from
the Western Desert
Sue O'Connor1,
Peter Veth2 and Colin Campbell1
In this
paper
we
present
the initial
report
for
the first
Pleistocene
occupation
sequence
to be excavated
in the
Western Desert of
Australia,
from the site
of
Serpent's
Glen
(Figs
1
and
2). We
identify
a three
phase
sequence
with an
earliest unit
dating
to
before
23,500 BP,
an
intermediate unit
comprising
culturally
sterile
sediments and an upper
unit
dating
to less than
4700
BP.
Previous excavations
within
the
Western
Desert have
only
provided
Holocene
assemblages
(Gould
1977;
Smith
1988;
Veth
1993). Indeed,
the
majority
of
these sites
have been dated
to the mid to late Holocene.
Arguments
for
demographic
restructuring
in
the arid
zone
during
the
Last Glacial
Maximum,
due to
resource
stress,
have
been advanced
by
a number
of authors
(Smith
1989;
Veth
1989,
1995a;
O'Connor et al. 1993). More recent
debate has
focused
on the extent
to which
groups
continued to
occupy
marginal
habitats
and the
mechanisms
by
which
they
^incor-
porated
such
habitats
into a broadly
based desert
adaptation
following
climatic
amelioration
(Smith
et al.
1991;
Smith
1993;
Edwards and
O'Connell
1995;
Veth
1995b;
O'Connor and
Veth
1996;
Gorecki
et
al. 1997). With
reference to this de-
bate,
Veth
(1995b:37)
has
noted that 'The need
for further
systematic
survey
and
excavation within
desert
lowlands,
having
uncoordinated
drainage,
is clear'.
The
identification
of a cultural hiatus at
Serpent's
Glen,
falling
within
a bracket
of
uncalibrated dates from
approxi-
mately
23,500
BP
to
4700
BP,
is
consistent with the
settle-
ment
models
which
argue
for
decreased
intensity
of occu-
pation
in
desert
lowlands
during
the last
glacial
maximum.
Equally,
a very
high
rate
of discard
of cultural items
in
the
upper
unit
of
the
site,
and
specifically
within the last
1000
years
of
occupation,
finds
parallels
with
similar efflores-
cences at
sites from the
Rudall
River
area
in
the Great and
Little
Sandy
Deserts
(Veth
1993),
from
the site of
Katampul
in
the
Goldfields
(O'Connor
and
Veth
1996)
and from a
range
of
localities in central
Australia
(Smith 1988).
Environmental
setting
Serpent's
Glen
Rockshelter
lies well within the Western
Desert
Culture
Bloc,
as
defined
by
Gould
(1977). The
site is
located
near
the mouth
of one
of the
larger
valley
systems
that
dissect
the
western face
of the Carnarvon
Ranges,
which
comprise
extensive
uplands
of
quartz
sandstones
on
an
other-
wise
flat
plain
of sand
sheets and
low relief dunes. The
Ranges
fall
in
the Peak Hill
area
of the
Bangemall
Basin,
a
middle
Proterozoic
Group
occupying
an
estimated
17,000
square
kilometres
in
the
northwest of Western Australia.
1 Division of Archaeology and Natural History,
Research School of
Pacific
and Asian
Studies,
The Australian
National
University,
Canberra,
ACT 0200, Australia.
2 School of
Anthropology
and
Archaeology,
Faculty
of Social Sciences,
James Cook University
of North
Queensland,
Townsville,
QLD 4810,
Australia.
The
edges
of the
uplands,
and
particularly
where
drainage
courses
discharge
onto the
surrounding
sand
plains,
are well
vegetated
by
a number of
species
of acacia and
eucalypt.
The
surrounding plains
and dunes
are dominated
by
Mixed
Shrub
Steppe
comprising
Acacia
spp.
and Triodia
spp.
(cf.
Beard and Webb
1974). The
Ranges
lie on the southern
edge
of the Little
Sandy
Desert,
an area known
to
provide
a
multitude of
plant
resources and
particularly
seeds. Seeds
have been
identified as a major staple group
for the
Martu
Aborigines
of
the Little
Sandy
Desert
(Veth
and Walsh
1988;
Tonkinson
1991).
The
Ranges
are
clearly
visible for tens of kilometres
across
the
surrounding
sandplains
and
dunefields.
They
contain
a
number of
spring
sites which were said
by Aboriginal
con-
sultants from
Wiluna
and
Jigalong
to
provide
reliable
and
potable
water.
A number of these
springs
were visited
by
the authors.
Large
rockholes and
soakages
were recorded
near
Serpent's
Glen. The
Ranges,
therefore,
provide
a highly
desirable set of
resources,
in the same manner as the Calvert
and Durba
Ranges
located further north
in
the Little
Sandy
Figure
1 Map
of Western
Australia
showing
arid zone boundaries
and
site
locations.
1 2 A
ustralian
Archaeology,
Number
46,
1
998
This content downloaded from 130.56.64.29 on Sat, 06 Feb 2016 01:54:55 UTC
All use subject to JSTOR Terms and Conditions
O'Connor
et al.
Figure
2 Serpent's
Glen Rockshelter: View
facing
north
showing
test
excavation.
Desert
(see
Veth
1993).
These
major
resource
réfugia
are
sig-
nificant
mythological
referents which also
contain
the most
prolific
rock art
galleries
in
the
region
(see
also Veth in
press).
Summary
of the
excavation
The
site was
excavated over a three week
period
in
July
1995 and
represents
a component
of a larger survey
and
excavation
programme
being
conducted
by
the authors with-
in the arid
zone of
Western Australia. Prior
to
excavation,
consultation
was carried out with senior
Aboriginal
custo-
dians from
Wiluna,
Jigalong
and Punmu
Communities,
all
located in the
vicinity
of the
Canning
Stock Route.
Serpent's
Glen is a south
facing
overhang approximately
20
m
wide and 8 m
deep
with a height
of 2.5 m
above
the
test-pit (Fig.
3). Extensive
panels
of art occur
on
the back
and
side
walls,
comprising
a
range
of
polychrome paintings
of
geometries,
anthropomorphs
with
headdresses,
and a range
of
striking serpentine figures.
A one metre
square pit
was excavated to bedrock
pro-
gressing by
2 to
3 cm
spits,
unless
stratigraphie changes
indi-
cated that
alternate
recoveries were warranted.
All
material
was
dry
sieved
through
5 mm
and
2 mm
nests and accurate
records were made of the
weights
and
volumes of
all
loose
sediments
and
larger grains
in
the matrix. Bulk
samples
of
sediments
were retained for
further characterisations.
Eight
strata
were identified in
the sections
(Fig.
4). Strata
1-6
represent
the
upper
Holocene
unit,
Stratum 7 the cultur-
ally
sterile
unit,
and Stratum
8 contains the
Pleistocene assem-
blages.
Strata
1-4
are
organic
rich,
containing ashy layers,
hearths and abundant
scattered
charcoal. Strata
5 and
6
are
poor
in
organics,
while
Stratum 7 lacks
macroscopic organics.
The
Pleistocene
unit contains
minor
quantities
of
charcoal,
bone
and,
of
interest,
ochre
fragments.
Initial
particle
size
analysis
of
the
sediments
indicates
a
trinodal
distribution,
suggesting polygenetic
origins.
The
peaks
were
around
0 Phi,
1.5
Phi and 3.0 Phi with most
of
the sediments
falling
in
the
sand fraction. As
the
frequency
curves for
particle
size distribution are
nearly
identical for
all
analysed spits,
the mechanisms
and sources
responsible
for
sedimentation
are
thought
to have been
relatively
constant
through
time
(J.
Magee
1977,
pers.
comm.).
Four
dates have been obtained
so far from
the site
(Table
1). All
dates
are
based
on scattered
charcoal which
was
collected and
plotted
in
situ. Another
sample
of charcoal
from
lower down
in the Pleistocene unit
at
Spit
40 has also
been
submitted for AMS dating.
While charcoal
was recovered
from
Spit
42,
just
above the lowest
cultural
level,
Spit
43,
it
has not been submitted
at this
stage,
as it is
only
associated
with a small number
of stone artefacts.
The
presence
of char-
coal
in seven out of the ten
spits,
beneath
the uncalibrated
date of
approximately
23,500
BP,
affords
the
opportunity
for
cross-checking
of dates.
Stone
artefact
assemblages
The stone artefacts recovered
from the 5 mm fraction
of
Square
A fall into two distinct
stratigraphie
units
(see
Table
2).
The lower unit
(Spits
34-43)
has a minimum date of
23,500
BP
and contains a total of 32 flaked
artefacts,
comprising
flakes,
flake
fragments
and a multiplatform
core. There are no
retouched/utilised
pieces
in
the Pleistocene
unit.
The
upper
unit
(Spits
1-26)
dates
from modern
to
approxi-
mately
4710
BP
and contains
a dense
assemblage
of flaked
Australian
Archaeology,
Number
46,
1998 13
This content downloaded from 130.56.64.29 on Sat, 06 Feb 2016 01:54:55 UTC
All use subject to JSTOR Terms and Conditions
Serpent's
Glen
Rockshelter:
Report
of
the
first
Pleistocene-aged
occupation
sequence
from
the
Western
Desert
Figure
3 Serpent's
Glen Rockshelter: Site
plan
and section.
stone
artefacts,
including
tula adzes and
slugs,
backed
pieces
and blades and a variety
of retouched/utilised
flakes.
A total
of 1642
artefacts,
or 98%
of the excavated
assemblage,
falls
into this
upper
unit.
Approximately
half of these
artefacts
occur
in the
Spits
1-8
(modern
period).
Between
the dates
of
4710 BP and
23,500
BP
(Spits
27-33)
no artefacts
were re-
covered from the 5
mm
fraction.
A similar
pattern
also exists
for ochres and faunal remains
(Table 3).
Despite
the
presence
of two intact basal
grindstones
and
several
fragments
on the shelter floor
(Fig.
3) and numerous
basal
grinding
stones on the sand
plains
surrounding
the shelter
(cf.
Veth and
O'Connor
1996),
only
two
ground fragments
of sandstone were recovered from
the entire excavation.
These
fragments
were
recovered from
Spits
3 and
4 in the
upper
unit.
The raw materials
used for
manufacturing
the artefacts
are available
locally
and are
represented
in
comparable pro-
portions
in
both the lower and
upper
units
(Table 2). The
only preference appears
to be the selective
use of chert and
silcrete for
backed
artefacts.
The
high
densities of artefacts
in
the
upper
unit are
paral-
leled
in
the
2 mm
fraction
(Table
3). The
debitage
in the
small
size
class
appears
to be
largely
the
product
of secon-
dary
retouch.
Faunal
assemblages
The
species
and
genera
identified
from
the
Serpent's
Glen
assemblage appear
to be
representative
of
pre-European
con-
tact
fauna.
All taxa
identified
at
Family
level or
below
occur
in the Holocene
Unit
and,
as
with
the stone
artefact
assem-
blage,
most
occur
in the
Modern
portion
of this
unit.
Frag-
ments of
bone
in the
Pleistocene
unit
were
identifiable
only
as Mammalia.
Taxonomic
assignment
The matrix
was sorted
into two
size
classes
for faunal
analysis,
in
order
to
determine
if the finer
fraction
(2-5
mm)
preserved
more
information
than
the
larger
(greater
than
5
mm).
This
proved
to
be the
case. In
the
fraction
greater
than 5
mm there
was a total
of
331
fragments
with
40
(i.e.
12%) taxonomically assigned
below
class
level. There
are
1476
fragments
in the
2-5 mm
fraction
with 131
assigned
(i.e.
8.8%). In
other
words,
77%
of
the taxonomic
assign-
ments of
vertebrate
remains
derive
from
the
finer matrix
frac-
tion.
Tables
4 and 5 document
the
importance
of examin-
ing
this
fraction
when
finely
comminuted
bone
is
present
in
a
deposit.
At
Puntutjarpa
Archer
(1977:158)
was able
to
assign
9.8%
of
the total faunal
assemblage,
a proportion
not
significantly
14 Australian
Archaeology,
Number
46,
1
998
This content downloaded from 130.56.64.29 on Sat, 06 Feb 2016 01:54:55 UTC
All use subject to JSTOR Terms and Conditions
O'Connor
et
al.
-n
(S*
I
1
|
s.
î
>
different
from
that
achieved
in this
study
and
one
which
precludes
the
application
of con-
ventional
analytic approaches
to
archaeological
faunas such
as
minimum number
indices.
This
analysis
recorded
each individual
fragment
that
could be
assigned
to taxon
per spit
(see
Tables
4
and
5).
Marsupial
carnivores
Family
Dasyuridae
The
larger
dasyurids
are
represented
by
Dasy-
urus
geqffroii
and
possibly
Dasyurus
hallucatus.
It should be
noted that both
could
readily
deal
with the
ecology
of this
locality.
D. geqffroii
can tolerate
high temperatures
and
do without
water
if fresh
meat is available.
They
can also
survive
freezing
conditions
with
periods
of
physiological
torpor
(Arnold
1983:22).
Sminthopsis
sp.
is
probably
represented
by
a single
calcaneum,
with known distributions
suggesting
S. macroura,
the
Stripe-faced
Dun-
nart. This
is a widespread
inland inhabitant
which deals
with climatic
extremes
by
widely
varying
its
body temperature
and
sheltering
in
cracks and
crevices.
It
stores
considerable
fat
in
the tail and
does not
need to drink free
water
(Morton
1983:63).
Four edentulous
mandible
fragments
and
the
coracoid
process
of
a scapula probably
repre-
sent
Dasycercus
cristicauda,
the
Mulgara,
a
small
burrowing
dasyurid
widely
distributed
in
central Australia. It also has the
ability
to
survive without
water and store fat
in its tail
(Woolley
1983:26).
Bandicoots
Family
Peramelidae
A species
of
lsoodon,
probably
auratus
is
represented
by
an edentulous
mandibular
frag-
ment,
an edentulous
maxillary
fragment,
a
worn
lower
M,
or
2,
and two caudal
vertebrae. For-
merly
a
widespread
inhabitant of the northwest
inland,
and now
only
abundant
in limited
areas
of the
northwest
Kimberley
and on Barrow
Island,
it was once found
in all
varieties of
in-
land arid
environments
(McKenzie
1983:98).
Three edentulous mandibular
fragments
be-
long
to
Perameles,
with
only
the extinct
species
eremiana known
to have
had a distribution
en-
compassing
the Carnarvon
Ranges. It was
anatomically adapted
to
extreme
heat,
with
large
ears and
hairy
feet,
and
was at least
locally
abundant into the 1930s
(Gordon
1983:102).
Family Phalangeridae
The
genus
Trichosurus
is
represented
by
a
fragmentary
LM1
or
2,
an
edentulous
maxillary
fragment
and a scapular
fragment.
The molar
appears
not
to be
vulpecula, leaving
arnhemensis,
the Northern
Brushtail,
as the
likely
candidate.
Australian
Archaeology,
Number
46,
1998 15
This content downloaded from 130.56.64.29 on Sat, 06 Feb 2016 01:54:55 UTC
All use subject to JSTOR Terms and Conditions
Serpent's
Glen Rocks
helter:
Report of
the
first Pleistocene-aged
occupation sequence
from
the
Western Desert
Site Years Depth Spit 613C Sample
code bp below code
surface
SGA1 modern 2
cm 1 -24.0±2.0% ANU
10024
SGA8 modern 19
cm 8 -24.0±2.0% ANU
10025
SGA27 4710±180 67
cm 27 -24.0±2.0% ANU
10026
SGA34 23,550±140
91cm 34 -25.00 ANSTO
0ZB582
Table
1 Radiocarbon dates
from
Serpent's
Glen Rockshelter.
Found
in woodland or
open
forest,
it would not have been
common
in the
Carnarvon
Ranges.
However,
it
could have
found sufficient arboreal
nesting
sites to remain established.
The
kangaroos
Family Macropodidae
Several teeth of
a
very
small
macropodid,
probably
Ony-
chogalea
lunata,
are
present.
Of the
larger
kangaroos
there
are
three dental
fragments
which
compare closely
with
Petro-
gale
cf.
later
alis. There are also tooth
fragments
of the Com-
mon Wallaroo
or
Euro,
Macropus
robustus,
which
is
the most
widely
distributed Australian
marsupial.
A
grazer
favouring
rocky
hills or
stony
rises,
it was a staple
of inland
human
populations
(Poole 1983:250).
Of similar but
more arid and
semi-arid
distribution,
the
presence
of the Red
Kangaroo,
Macropus rufus,
is also to be
expected
as
it occurs over most
of
central
Australia
in
areas
of
less
than 500
mm
average
rainfall
(Sharman 1983:255).
Reptiles
Amongst
lizards,
the
genus
Varanus
(Family
Varanidae,
the
Goannas)
and the
genus
Egernia (Family
Scincidae,
the
Skinks)
are
represented variously
by
teeth and vertebrae.
Specific
determinations have not been
made,
but
at
least
seven
species
of
Varanus and four of
Egernia
have concurrent
ranges
in the Carnarvon
Ranges
area and are common
and
readily
acquired
food
animals
(Cogger
1994).
Snake
vertebrae
repre-
senting
the
Family
Boidae were also
identified.
Bone
reduction
The bone
remains from
the site are
extensively
reduced,
with
the
majority
(1476) being
in
the size
range
2-5
mm,
with far fewer
(33 1)
being
5
mm or over.
In
Spit
1 at
least
there is also
a significant,
although
unmeasured,
fraction
of
bone
fragments
less than
2 mm in size. Identification and
taxonomic
assignments
in
this fraction are
possible.
The lar-
gest
single
bone
specimen
was 35
x
8
mm.
Surviving
whole
bones,
primarily
the vertebrae of
small
reptiles
and
bandicoots,
are
in the size
range
of
the common broken
pieces. Some
bone
bears
apparent
cut marks but these
are not
common.
Burnt
bone is
present
in almost
every
spit
(Table
6,
Fig.
5).
All of
the
63
macropodid
teeth
present
(52
in
the smaller
fraction),
except
for one
(probably
of
the
macropodid
Ony-
chogalea)
were broken. It is
highly
unlikely
that
there is
any
nutrient
value
to
be
extracted
from a broken
tooth and
it is assumed that
they
are broken
incidentally
when
man-
dibles
and
skulls
are reduced.
Index
of Fragmentation
Both size
classes of
sample
have been combined for calcu-
lation of the
Index
of
Fragmentation
(total
bone
weight/#
fragments),
see
Table 6. The
average
value for all
spits
is
0.063
gm/fragment (range
0.020-0.173).
The
consistency
of
these values
from
spit
to
spit strongly suggests
that the
same
mechanism of
breakage
was common
throughout.
Gould's
Puntutjarpa
Holocene fauna
averaged
0.53
gm/fragment,
which is
considerably larger
(Gould 1977). It
is
likely
that
at least
some
of the difference is due
to the
lower fre-
quency
of
large
macropodid
remains
at
Serpent's
Glen.
How-
ever,
the fact that the bone
assemblage
has
been so thor-
oughly
comminuted
suggests
that
the same
hypothesis
of
explanation
applies,
namely
that the
human
population
was
protein-stressed
and
found
it
necessary
to extract
every
vestige
of edible meat and marrow from the
vertebrates
captured.
Taphonomic
factors
Of
the six
taphonomic
factors
postulated by
Gould as
potentially applicable
to
a site
such as
Serpent's
Glen,
namely
the
effects
of
predators,
crushing
due
to
trampling,
crushing
due to
rockfall,
crushing
due
to
weathering
and
soil
compac-
tion,
chemical
factors
and
burning
of
bone,
none would
con-
vincingly explain
the
thorough
and
notably
even
reduction
of bone. There is
no
evidence of
any
native
predators
larger
than
Dasyurus,
and no incremental
increase in
reduction to-
ward
the
younger
strata which
would be
expected
if
dingoes
were to be the cause. Trampling
would
produce
a greater
size
range
of
fragments
and
rarely
so
many
small
fragments,
especially
those less than
2
mm in size. There
are
no
notable
rockfalls in the Holocene
levels,
and the
comminution is
similar
throughout
the section.
Weathering
and soil
compaction
were
certainly
present,
and a
minor
proportion
of the bone
is
soft
and/or whitened.
However,
such
fragments
are no smaller
than others
with
no
apparent decomposition
and with
surviving
sharp
edges
on
the breaks.
Chemical factors
may
have
operated
to soften
and hasten
bone
decomposition
but
again,
the
majority
of
fragments
are
extremely
pristine
in
appearance
and
hardness. As
noted,
burning
is common
(see
Fig.
5),
but
appears
to have acted
on
most
of the bone after
breakage
for in
the
majority
of
in-
stances the
fractured surfaces
as
well as the natural surfaces
are burnt.
The
supposition
of intentional and
systematic
human
breakage
remains as
the
strongest hypothesis
for
the
reduced state
of the vertebrate remains.
Discussion
of
patterning
in cultural
residues
The excavation has
provided unequivocal
evidence
for
a
human
presence
in
the Western
Desert
prior
to
23,500
BP.
The
Pleistocene
unit contains flaked stone
artefacts manu-
factured from
chert,
chalcedony,
silcrete,
banded
sediment,
quartzite
and
quartz;
the
full
range
of
raw
materials
found
in
the
Holocene
unit. The lower unit
also contains small
quan-
tities
of
charcoal, ochre,
bone and bird
eggshell
(Table 3).
There
is
a considerable volume of
deposit
in the
Pleistocene
unit,
between
Spits
34-43,
though
all
categories
of
cultural
material are
sparse
(Table
3). While
further
dating
is
required
to evaluate
the
period
of
time
represented by
this
unit,
it
is
difficult to
interpret
the
low
density
and nature of
cultural
materials
as
representing anything
more than
occasional
forays
1
6 A
ustralian
Archaeology,
Number
46,
1
998
This content downloaded from 130.56.64.29 on Sat, 06 Feb 2016 01:54:55 UTC
All use subject to JSTOR Terms and Conditions
O'Connor
et
al
Spit Age Chert Chalcedony Silcrete Banded Quartzite Quartz Total
sediment
1 Modern 42 2 20 9 8 1 82
2 53 5 10 6 27 6 107
3 34 1 11 3 12 5 66
4 28 2 3 5 12 3 53
5 10 0 6 0 8 0 24
6 42 1 4 2 5 1 55
7 143 4 31 3 2 3 186
8 Modern 201 24 16 12 32 8 293
9 187 15 50 4 11 9 276
10 78 12 31 2 7 14 144
11 27 3 22 0 6 18 76
12 8 0 3 0 28 16 55
13 14 2 12 0 23 12 63
14 22 0 9 1 11 8 51
15 20 2 1 1 10 10 44
16 24 3 11 0 7 3 48
17 15 3 10 0 0 6 34
18 7 3 10 2 1 3 26
19 2 0 4 0 4 2 12
20 8 0 2 0 1 0 11
21 0 0 0 0 10 0 10
22 3 0 0 0 0 0 3
23 5000005
24 0000000
25 0 0 0 0 0 0 0
26 0230207
27 4710
BP 0 0 0 0 0 0 0
28 0000000
29 0 0 0 0 0 0 0
30 0000000
31 0 0 0 0 0 0 0
32 0000000
33 0000000
34 23,500
BP 3 0 0 0 2 2 7
35 0 0 0 0 10 1
36 0 0 10 4 0 5
37 4 1 1110 8
38 4 0 0 0 0 15
39 0000303
40 10 0 0 0 0 1
41 0 0 0 0 10 1
42 0000000
43 10 0 0 0 0 1
Total 986 85 271 51 239 131 1763
Table
2 Serpent's
Glen: Numbers of artefacts
from
>5 mm
fraction.
into
this
arid
landscape. While the breadth of lithic
raw
materials
parallels
those from
Holocene assemblages,
and
might imply
a level of
familiarity
with the
landscape
not con-
sistent with
ephemeral usage,
all are
actually
available from
the
uplands
catchment.
The form
of
site use
appears
to more
ephemeral
than the Holocene levels,
lending
support
to
argu-
ments that
Pleistocene land-use
systems
in
the
Western Desert
are
likely
to have
been
different
in
nature
(Veth 1995b).
The
culturally
sterile middle
unit
is bracketed
by
the lower
date of
23,500
BP
and an
upper
date
of
4700 BP. These dates
only
provide
an envelope
for the
age of the
unit,
however.
It
may
have formed
at
a relatively
constant
rate
or, instead,
experienced
a number of
episodes of
rapid
sedimentation.
Initial
characterisations of the
sediments and
larger grains
in
the matrix
suggest
that
neither
the
sources
nor mechanisms
of
deposition
have changed significantly
between
adjacent
units.
Importantly,
there is no
evidence for a lag
event
at
the
boundary
of this and the
Pleistocene
unit,
although
an in-
crease in
large
roof fall
fragments
is
apparent.
At
this
stage
we interpret
the unit to represent
a cultural
hiatus
where
sedimentation
has occurred
without
occupation,
and not a
stratigraphie unconformity
and hiatus.
Australian
Archaeology,
Number
46,
1
998 1
7
This content downloaded from 130.56.64.29 on Sat, 06 Feb 2016 01:54:55 UTC
All use subject to JSTOR Terms and Conditions
Serpent's
Glen
Rockshelter:
Report
of
the
first Pleistocene-aged occupation sequence
from
the
Western
Desert
Spit Stone artefacts Ochre Charcoal Bone Emu
egg Plant Bird
egg
5 mm 2 mm 5 mm 2 mm 5
mm 5
mm 2 mm 5 mm 2
mm 5
mm 2 mm 2 mm
1 77.0 9.5 0.9 0.1 63.3 - 0.2 0.2 0.1 6.3 6.2
2 94.8 10.3 0.5 0.2 107.5 0.7 1.0 1.0 0.1 21.3
3 84.0 10.6 - 0.2 102.4 0.2 1.9 - 0.1 -
4 50.2 11.1 0.1 0.4 51.5 0.8 1.7 0.3 0.5 - - 0.2
5 30.5 9.0 - - 11.3 1.0 2.9 0.8 0.2 - - 0.2
6 31.7 10.5 - - 8.2 4.8 3.0 0.7 ...
7 153.4 34.3 - 0.2 1.2 13.2 20.3 .....
8 226.8 79.1 - 0.1 0.3 29.1 34.8 0.1
9 183.2 60.9 - 0.3 0.2 23.3 23.3 .....
10 125.5 25.6 0.2 0.3 0.1 9.9 24.2 .....
11 104.6 10.3 0.2 0.5 2.3 5.6 9.0
12 97.6 7.3 2.8 0.7 - 1.4 5.6 - - -
13 54.4 9.2 0.6 0.4 0.2 0.5 0.7
14 50.8 7.6 3.0 0.1 0.3 - 0.6
15 81.5 2.3 - 0.1 .....
16 67.4 3.8 - - - - - - ...
17 24.4 5.2 ----- .... 0.1
18 24.6 1.7 - - - - - .....
19 10.3 0.8 - 0.2 - - - .....
20 6.1 0.9 - - - - - .....
21 4.4 1.0 . 0.5 - - - - ...
22 1.7 0.5 - - 0.6 .....
23 8.1 1.6 - - 0.6 - - .....
24 - 0.2 - 1.3 - 0.1 - - ...
25 2.4 0.8 .....
26 92.8 0.1 - - 0.8 .....
27 0.4 - - 1.0 .....
28 ............
29 ............
30 ............
31 -...-.-.....
32 ............
33 ............
34 *28.0 0.4 - 0.1 7.3-0.3 - - ...
35 0.9 0.1 - - 1.1 - 0.1 - - ...
36 7.1 0.4 - - 4.1 - - .....
37 5.4 0.2 - 0.2 6.0 - - .....
38 3.8 0.5 - 0.2 0.3-0.1 - - ...
39 7.8 - 1.7 .....
40 0.2 0.2 - - 0.6 .....
41 0.3 0.2 - - - - 0.1
42 0.1 - 0.7 .....
43 0.1 - - - - .....
44 ............
Table 3 Weights
of
all
cultural
materials
recovered in
5 mm
and 2 mm
fractions.
(Weights
are
uncorrected for
volume). (Note:
*
Majority
of
total
weight
from
one
large
core).
A stratigraphie
unconformity
describes a surface of
non-
deposition
that
appears as a break
between
two sediment
units
in
a sequence. Unconformities
arise
from
changes
in
depositional conditions,
to conditions of erosion or non-
deposition.
The
unconformity
must
be
stratigraphically
identi-
fiable as a surface.
This
is
not the
case with
the
sequence
at
Serpent's
Glen
(O'Connor,
Veth
and
Barham in
press).
Some
Pleistocene-aged
sites
in
the
Australian
arid
zone,
and specifically
the
nearby
site
of
Katumpul
in
the
Gold-
fields
(Fig. 1),
have
units
with
late
Pleistocene
dates
directly
overlain
by units
with
mid- to late Holocene dates. This
represents
a chronological
hiatus
where the break
is not
repre-
sented
stratigraphically.
Without evidence
for
stratigraphie
unconformities
it
has been
argued
that it is unacceptable
to
invoke
geomorphological
explanations,
and that the
hiatuses
can be more
plausibly
explained through
changes
in
settle-
ment
dynamics
and
types
of
occupation
in
shelters
(O'Connor
and Veth
in
press).
We argue
that
the
culturally
sterile
unit
provides
evidence
for the
passage of time
where
occupation did not occur.
1
8 Australian
Archaeology,
Number
46,
1
998
This content downloaded from 130.56.64.29 on Sat, 06 Feb 2016 01:54:55 UTC
All use subject to JSTOR Terms and Conditions
O'Connor et al
Spit
Number
Taxon 1 2 3 4 5 6 7 8 9 10 11 12 13 14 24 34 35 38 41
TELEOSTEI ...54-3
REPTILIA
Lacertilia
(Sauna) 1 1-2
Agamidae 1 3
Varanidae
Varam/s
sp. 111
Scincidae 2 - - 1
Egernia sp. 1
Serpentes 1
Boidae 12 2
Aves 5
Emu
(eggshell) 6-1931
MAMMALIA 5 24 35 39 56 30 316 167 200 - 228 - 13 7 - 2 2 2 1
Small
mammal 32-2
Medium
mammal 2
Marsupialia ... 1 6 - - - 1
Dasyuridae 1 1 2 1
Dasycercus
chsticauda - - 1 2
Dasyurus geoffro
1
Dasyurus
halluca 1
tus
Sminthopsis sp. 1
Peramelidae .... 1
Isoodon
sp. 1 - 1 - 1 1 ....
Isoodon auratus 1
Perameles
ssp.?eremiana 1 - 1 - 1
Phalangeridae , - - - 1
Trichosurussp.
1-2
Macropodidae 3
Macropodid, large 1 13-4
Macropodid,
small 1 - - - 1 - 4
Lagorchestes
hirsutus 2 2-1
Macropus
spp. 2 - 12 - 1
Macropus
agilis ...... 1 . 1
Macropus
robus-
tus .... 1 . 1 1
Macropus
rufus 11
Onychogaleasp.
1-22-1
Petrogale
cf.
lateralis 1 .-13
EUTHERIA
Muridae 2324
Rattussp.
Table
4 Serpent's
Glen
vertebrate
fauna,
Square
A,
2-5
mm
fraction;
number of
fragments
assigned
to
each
taxon.
While
such
a break
in one site
may
not reflect
regional
trends,
we
would
note that
it is consistent
with models
which
argue
for
significant
demographic
shifts
during
the Last Glacial
Maximum
and
significantly
decreased
intensity
of
site occu-
pation
and
possibly
regional
abandonment
(Veth 1995b).
The
upper
unit
contains
the
majority
of
the cultural assem-
blages,
for
example
98% of
all flaked stone
artefacts.
In
fact,
much of the
assemblage
can
probably
be assigned
to the late
Holocene,
with almost 50% of
the site's
artefacts
dating
to
the
modern
period.
Further
charcoal
samples
are
being
submitted
to
provide
better
chronological
resolution between
Spits
8-27.
Tula
adzes
and
slugs,
as well as backed
pieces
and
blades,
are found
only
in
the
upper
half of the
unit,
a distribution
con-
sistent
with
its
likely
mid-
to late
Holocene age (cf.
Hiscock
Australian
Archaeology,
Number
46,
1
998 1 9
This content downloaded from 130.56.64.29 on Sat, 06 Feb 2016 01:54:55 UTC
All use subject to JSTOR Terms and Conditions
Serpent's
Glen Rocks helter:
Report
of
the
first Pleistocene-aged
occupation
sequence
from
the
Western
Desert
Spit
Number
Taxon 1 2 3 4 5 6 7 8 9 10 11 12 13 14 24 34 35 38 41
Aves
Emu
(eggshell) 2-525821
MAMMALIA 1 - 6 28 - 37 30 8 3
Small
mammal 1 1 2 - - 2
Medium mammal 4
Marsupialia
Phalangeridae
Trichosurus cf.
vulpecula 1 1
Macropodidae 1 3 3 12 31 49 79 1
Macropodid,
large 12
Macropodid,
small 1
Macropus
robust 1 2 4
Macropus
rufus
Onychogalea sp. 1
EUTHERIA
Muridae 1
Table
5 Serpent's
Glen
vertebrate
fauna,
Square
A,
>5
mm
fraction,
number of
fragments assigned
to each taxon.
Spit
number Total
bone Burnt bone Unburntbone Total number
of Number of burnt Index of
gm gm gm fragments fragments fragmentation
1 0.23 0.00 0.23 10 0 0.023
2 1.70 0.27 1.43 29 6 0.059
3 2.10 0.48 1.62 37 10 0.057
4 2.53 0.23 2.30 25 8 0.101
5 3.87 0.89 2.98 69 15 0.056
6 7.79 1.24 6.55 65 21 0.120
7 33.50 9.83 23.67 414 87 0.081
8 44.56 16.29 29.52 258 86 0.173
9 46.51 16.63 29.88 517 143 0.090
10 34.15 11.69 22.46 642 206 0.053
11 14.60 11.69 2.91 294 247 0.050
12 6.95 1.59 5.36 175 55 0.040
13 1.20 0.84 0.36 18 8 0.066
14 0.55 0.10 0.45 11 2 0.050
24 0.04 0.00 0.04 1 0 0.040
34 0.29 0.00 0.29 5 0 0.058
35 0.06 0.00 0.06 2 0 0.030
38 0.04 0.00 0.04 2 0 0.020
41 0.03 0.00 0.03 1 0 0.030
Table
6 Data for
combined bone
fractions,
Serpent's
Glen,
Square
A.
and
Veth
1991).
Variously
retouched/utilised
flakes
are found
throughout
the
Holocene
unit.
The
obviously
significant
increase in
discard
rates
of
stone
artefacts
and
ochres
from
Spits
15-20,
which
most
likely
date
within
the
last
2000
years,
is
consistent
with
efflores-
cences
in
these
categories
as
witnessed
from
other
Western
Desert
and
central
Australian
sites
(Smith 1988;
Ross
et al.
1992;
Veth
1993). Here
these
changes may
be linked to
the
rapid
and
widespread
expansion
of
the
Western
Desert
language
which
has
been
recently
assigned
by
a number
of
authors
to
this
period (McConnell
1996;
Veth in
press).
Mechanisms for
the
expansion
are
likely
to
have
comprised
a combination
of
changes
in
social
organisation
and
tech-
nological
strategies
(cf.
Evans and
Jones
1997).
Conclusion
The
early
assemblages
from
Serpent's
Glen
provide
the
first
firm data
for
Pleistocene use of
the
Western
Desert and
specifically
the
sandy
deserts.
Given
the
unique
place
of the
Western
Desert as one
of the
most
marginal
arid
landscapes
to
have
been
occupied
(Veth
1995a),
this
early
phase
of
occu-
pation
must be
viewed as
significant.
Equally
of
interest
is
the
fact that
the
evidence for
ephemeral
occupation
is
not
repeated during
the
critical
phase
of
the Last
Glacial Maxi-
mum. A substantial unit
of
culturally
sterile
deposits
attests
to
at
least
one
period
of
time when
occupation
did
not occur
at
all. This
evidence
for a cultural
hiatus is
consistent with
models
for
demographic
restructuring during
the
heightened
aridity
of
the
Last Glacial
Maximum
(Veth
1993).
In
contrast,
20 Australian
Archaeology,
Number
46,
1998
This content downloaded from 130.56.64.29 on Sat, 06 Feb 2016 01:54:55 UTC
All use subject to JSTOR Terms and Conditions
O'Connor et ai
Figure
5 Proportions
of burnt to
unburnt
bone,
Serpent's
Glen Rock-
shelter,
Square
A,
combined
size fractions.
the
central
Australian
site of
Puritjarra,
whose catchment
shares
many ecological
features
with the
sandy
deserts,
has
some
evidence
for
repeated
occupations
during
this
period
(Smith
1989).
The difference
in
occupation
is
likely
due to
the fact that
the central
Australian
ranges
are
much more
extensive
than
the Carnarvon
Ranges
and
that
Puritjarra
is
proximal
to coordinated
drainage systems
and networks of
major
and
permanent
artesian
water sources.
These could
have
mitigated
the effects
of resource
stress
by
providing
nodes
of
access
to,
and retreat
from,
the desert
lowlands
from
the
central
Australian
ranges.
We
would note
that the indices
of bone
fragmentation
from
Serpent's
Glen are
considerably
higher
than
those at
Puntutjarpa,
in the Warburton
Ranges,
further
emphasising
the
comparatively
protein-depauperate
nature
of
the
surrounding
desert
lowlands of the
Little
Sandy
Desert.
In
conclusion,
we
argue
that the
apparently
ephemeral
nature
of the Pleistocene
assemblages
at
Serpent's
Glen and
the
presence
of a clear cultural
hiatus
highlight
the
degree
to
which
groups
employed
highly
flexible settlement
strategies
to
cope
with resource
stress
at different
scales
of time
(cf.
Gould
1991). We
advocate
continued
characterisations
of
early
settlement
behaviours
and their
comparison
to those
of
the late
Holocene,
when the
Western
Desert
linguistic
and
technological
unity
evolved.
Acknowledgements
We
would
like to thank
Dr Ken
Aplin,
Western Austra-
lian
Museum
for assistance
with
identifications
of the faunal
material.
Professor
Richard
Gould,
Department
of Anthro-
pology,
Brown
University,
USA,
kindly provided
us with a
draft of
his
paper
'Faunal
reduction
at
Puntutjarpa
rockshelter,
Warburton
Ranges,
Western
Australia',
prior
to
its
publication.
References
Arnold,
J.M. 1983 Dasyurus
geoffroil
In R. Strahan
(ed.) The
Complete
Book
of
Australian
Mammals,
p.22. Sydney:
Angus
and
Robertson.
Archer,
M. 1977 Faunal
remains
from
the
excavation
at Puntut-
jarpa
Rockshelter.
In R.A. Gould's
Puntutjarpa
Rockshelter
and
the Australian Desert
Culture,
pp.
158-65.
New
York:
American
Museum
of
Natural
History.
Anthropological Papers
of
the
American
Museum
of
Natural
History
54.
Beard,
J.S. and
Webb,
J. 1974 Vegetation
Survey of
Western
Australia:
Great
Sandy
Desert.
Perth:
University
of
Western
Australia
Press.
Cogger,
H.G. 1994
Reptiles
and
Amphibians
of
Australia.
Sydney:
Reed
Books.
Edwards,
D.
A.
and
O'Connell,
J.F. 1995 Broad
spectrum
diets in
arid
Australia.
Antiquity
69:7 '69-83.
Evans,
N. and
Jones,
R. 1997 The cradle of the
Pama-Nyungans:
Archaeological
and
linguistic
speculations.
In P.
McConvell
and
N. Evans
(eds) Archaeology
and
Linguistics:
Global
Perspectives
on Ancient
Australia,
pp.385-417.
Melbourne:
Oxford
University
Press.
Gordon,
G. 1983 Perameles eremiana. In R. Strahan
(ed.) The
Complete
Book
of
Australian
Mammals,
p.
102. Sydney:
Angus
and Robertson.
Gorecki, P., Grant, M., O'Connor,
S. and
Veth,
P. 1997 The
morphology,
function
and
antiquity
of
grinding implements
in Australia.
Archaeology
in Oceania
32:141-50.
Gould,
R.A. 1977 Puntutjarpa
Rockshelter and the
Australian
Desert Culture.
New York: American
Museum of Natural
History.
Anthropological
Papers of
the
American Museum
of
Natural
History
54.
Gould,
R.A. 1991 Arid-land
foraging
as seen from
Australia:
Adaptive
models and behavioural realities.
Oceania 62:
1
2-33 .
Gould,
R.A. 1996 Faunal reduction at
Puntutjarpa
rockshelter,
Warburton
Ranges,
Western Australia.
Archaeology
in
Oceania
31:72-86.
Hiscock,
P. and
Veth,
P. 1991 Change
in the
Australian desert
culture:
A reanalysis
of tulas from
Puntutjarpa.
World Arch-
aeology
22:332-45.
Mabbutt,
J.A. 1971 The Australian arid
zone
as a prehistoric
environment.
In
D.J.
Mulvaney
and J. Golson
(eds)
Abori-
ginal
Man and Environment in
Australia,
pp.66-79.
Canberra:
Australian
National
University
Press.
McConvell,
P. 1996 Backtracking
to Babel:
The
chronology
of
the
Pama-Nyungan
expansion
in
Australia.
Archaeology
in
Oceania 3
1(3):
125-44.
McKenzie,
N.L. 1983 Isoodon auratus. In R.
Strahan
(ed.)
The
Complete
Book
of
Australian
Mammals,
pp.98. Sydney:
Angus
and Robertson.
Morton,
S.R. 1983 Sminthopsis
macroura.
In
R. Strahan
(ed.)
The
Complete
Book
of
Australian
Mammals,
p.
63. Sydney:
Angus
and Robertson.
O'Connor,
S. and
Veth,
P. 1996
A
preliminary archaeological report
on recent
archaeological
research
in
the semi-arid/arid belt of
Western Australia. Australian
Aboriginal
Studies
1996(2):42-50.
O'Connor,
S.,
Veth,
P.
and
Barham,
A.J. in
press
Cultural versus
natural
explanations
for lacunae
in
occupation deposits
in
northern Australia.
Quaternary
International
Poole,
W.E. 1983 Macropus
robustus.
In R. Strahan
(ed.) The
Complete
Book
of
Australian
Mammals,
pp.250-
1. Sydney:
Angus
and Robertson.
Ross,
A.,
Donnelly,
T.
and
Wasson,
R. 1992 The
peopling
of
the
arid zone: Human-environment
interactions.
In J.
Dodson
(ed.) The Naive Lands: Prehistory
and Environmental
Change
in
Australia
and the Southwest
Pacific, pp.76-
1 14.
Melbourne:
Longman
Cheshire.
Sharman,
G.B. 1983 Macropus rufus.
In R. Strahan
(ed.) The
Complete
Book
of
Australian
Mammals,
pp.255-7.
Sydney:
Angus
and Robertson.
Smith,
M. A. 1988 The
pattern
and
timing
of
prehistoric
settlement
in
Central Australia.
Ph.D.
thesis,
Department
of Archae-
ology
and
Palaeoanthropology,
University
of
New
England,
Armidale.
Australian
Archaeology,
Number
46,
1998 21
This content downloaded from 130.56.64.29 on Sat, 06 Feb 2016 01:54:55 UTC
All use subject to JSTOR Terms and Conditions
Serpent's
Glen Rockshelter:
Report
of
the
first
Pleistocene-aged
occupation
sequence
from
the Western
Desert
Smith,
M.A. 1989 The case
for a resident
human
population
in
the
Central
Australian
ranges
during
full
glacial
aridity.
Arch-
aeology
in Oceania 24:93-105.
Smith,
M.A. 1993 Biogeography,
human
ecology
and
prehistory
in
the
sandridge
deserts.
Archaeology
in
Oceania
37:35-50.
Smith, M.A.,
Williams,
E. and
Wasson,
RJ. 1991 The
archaeology
of the JSN
site: Some
implications
for
the
dynamics
of human
occupation
in the
Strzelecki
desert
during
the late
Pleisto-
cene.
Records
of
the South Australian
Museum 25:175-92.
Strahan,
R.
(ed.) 1983 The
Complete
Book
of
Australian
Mammals.
Sydney:
Angus
and
Robertson.
Tonkinson,
R. 1991 The Mardu
Aborigines:
Living
the
Dream in
Australia's
Desert.
Fort Worth:
Holt,
Rinehart and
Winston.
Veth,
P. 1989
Islands
in the interior:
A
model
for the colonisation
of Australia's
arid zone.
Archaeology
in Oceania
24:81-92.
Veth,
P.M. 1993 Islands
in the Interior:
The
Dynamics
of
Pre-
historic
Adaptations
within the Arid Zone
of
Australia.
Ann
Arbor,
Michigan:
International
Monographs
in
Prehistory.
Archaeological
Series
3.
Veth.
P. 1995a
Marginal
returns
and
fringe
benefits:
Characterising
the
prehistory
of the
lowland deserts
of
Australia
(a reply
to
Smith).
Australian
Archaeology
40:2-38.
Veth,
P. 1995b Aridity
and settlement
in
northwest
Australia.
Antiquity
69:733-46.
Veth,
P. in
press
Language
on
the
move:
Archaeologically
gen-
erated
models
for the
timing
and direction
of
the Western
Desert
Language.
Origins
of
the
Western
Desert
Language.
Veth,
P.M. and
Walsh,
F.J. 1988 The
concept
of
'staple1 plant
foods
in the
Western desert
region
of
Western
Australia.
Australian
Aboriginal
Studies
1
988(2):
1
9-25.
Woolley,
P.
A. 1983 Dasycercus
cristicauda.
In
R.
Strahan
(ed.)
The
Complete
Book
of
Australian
Mammals,
p.26. Sydney:
Angus
and
Robertson.
22 Australian
Archaeology,
Number
46,
1998
This content downloaded from 130.56.64.29 on Sat, 06 Feb 2016 01:54:55 UTC
All use subject to JSTOR Terms and Conditions
... At this time the first detailed recording was made of the Serpents Glen rock art assemblage and the nearby Wirrili Shelter (named Bella Vista on that trip). We undertook direct-dating of the art, because we were particularly interested in trying to constrain the rock art production to the lithic assemblage which had been dated to the indistinguishable-from-modern period (O'Connor, Veth & Campbell, 1998: Table 2). There was also the evidence that custodian Willy Ward had retouched the iconic snake motif in 1967 ( Figure 6). ...
... Direct dating of Karnatukul's most recent pigment art phase revealed an apparent disjunct between the art production and the most recent dated occupation evidence. The most recent art at Karnatukul (white and black bichrome motifs) was produced between 300-900 years ago (McDonald et al. 2014: Table 3), while the most intensive last phase of occupation was indistinguishable-from-modern (O'Connor, Veth & Campbell 1998: Table 4). The 2014 excavation aimed at directly testing the contemporaneity of rock art and occupation evidence at this site. ...
... Early technological innovation registered in this assemblage highlights the resourcefulness of the first Australian's adaptive culture. The revised Karnatukul sequence provides the first unequivocal evidence for occupation of the Western Desert during the Last Glacial Maximum (contra O'Connor, Veth & Campbell, 1998;Smith 2013a), as well as an occupation pulse during the late Pleistocene/early Holocene transition. Unfortunately, the recent excavations revealed too small a faunal assemblage to refine the questions raised (Codding 2011: 254) about an apparent decline in macropod abundance in the late Holocene. ...
Article
Full-text available
Recent work at Serpents Glen ('Karnatukul') in the Carnarvon Ranges (Katjarra) of the Western Desert has changed our archaeological understanding of both deep time occupation and more recent arid-zone social geography. Mobilising rock art evidence into earlier models for how arid zone peoples have entered, settled and known Country has allowed us to project people into cycles of human mobility. Our understanding of the deep time and more recent engagements with Country ('ngurra') has changed significantly since Richard Gould wrote 'Yiwara' and 'Living Archaeology' in the late 1960s. Early ethno-archaeological studies portrayed the desert as harsh and precarious, and the lifeways of arid zone peoples as marginal and conservative. Fifty years of archaeological endeavour working with traditional custodians in the Western Desert, has changed this view of the ‘dangerous desert’. ‘Risk-minimisation’ and the ‘dietary stress hypothesis’ have been replaced with models that consider human mobility, social geography and information exchange theory as ways of understanding how arid-zone peoples have been successfully on country since the earliest human occupation of this continent. 'Karnatukul'’s record rewrites the deep history of the arid zone, as well as refining our understanding of social complexity by combining late Holocene arid zone art and occupation evidence.
... The archaeological sequence commences between 47,130 and 42,140 cal BP (i.e., the lower modelled Bayesian age estimate for base of deposit) and continues into the recent past. Importantly, it documents repeated visits to the rockshelter during the peak of the LGM, providing new detail on human movements in a place where previously held views were of a hiatus in human occupation over this time (e.g., Hiscock, 1988;Veth, 1993;Thorley, 2001;O'Connor, 1995;O'Connor and Veth, 2006;O'Connor et al., 1998Przywolnik, 2005a. At least for parts of the Hamersley Plateau, there is now a growing body of archaeological evidence supporting intermittent visits by Aboriginal people during the LGM (Cropper and Law, 2018;Marwick, 2002;Morse et al., 2014;Reynen, 2019;Reynen et al., 2018;Slack et al., 2017;Slack et al., 2018). ...
... Dating studies and Bayesian modelling at Juukan 2 support visitation of the site during the height of the LGM and refute the commonly held view that there was a hiatus in occupation in the semi-arid and arid zones of Sahul at this time (e.g. Hiscock, 1988;Veth, 1993;Thorley, 2001;O'Connor, 1999;O'Connor et al., 1998;Marwick, 2002;O'Connor and Veth, 2006;Przywolnik, 2005a;Morse et al., 2014;Cropper and Law, 2018;Slack et al., 2017Slack et al., , 2018Reynen et al., 2018). A distinct stratigraphic change at 24,200-23,360 cal BP, tightly bracketed by radiocarbon dates on in situ charcoal, along with the lithics, bone discard and combustion features supports a human presence at peak glacial aridity. ...
Article
Our paper provides the first in depth scientific analysis of the world significant Juukan 2 rockshelter site. It provides a revision and update of the original 2008 test excavation results, alters the regional archaeology of the inland Pilbara, and Australia based on new methods and excavation data. The results substantially revise the known antiquity of human occupation for the region and with that our understanding of the timing and processes of human settlement of the arid interior of Australia.
... As such, zooarchaeological measures of relative abundance can be used to evaluate archaeological assemblages in order to determine the relative acquisition risk and division of labor experienced by past populations. As a preliminary examination, here we draw on faunal remains recovered from two sites occupied in the Late Holocene of Western Australia that were evaluated using fine-grained excavation techniques: Serpent's Glen Rock Shelter (O'Connor et al., 1998) andWindjana Gorge Water Tank Shelter (O'Connor et al., 2008). ...
... In this way, grinding stones acted as 'site furniture' for processing lower ranked foods such as hard nuts and seeds (that require more intensive processing) at a location that was visited more frequently and predictably during drier times 23,75 . Animal materials were also processed on grinding stones for the first time during the LGM (Phase 4), perhaps for reasons of nutritional stress to reduce faunal wastage, mirroring the practice of extensive bone processing for protein extraction documented in arid economies in Central Australia in the recent past 76,77 . ...
Article
Full-text available
Grinding stones and ground stone implements are important technological innovations in later human evolution, allowing the exploitation and use of new plant foods, novel tools (e.g., bone points and edge ground axes) and ground pigments. Excavations at the site of Madjedbebe recovered Australia’s (if not one of the world’s) largest and longest records of Pleistocene grinding stones, which span the past 65 thousand years (ka). Microscopic and chemical analyses show that the Madjedbebe grinding stone assemblage displays the earliest known evidence for seed grinding and intensive plant use, the earliest known production and use of edge-ground stone hatchets (aka axes), and the earliest intensive use of ground ochre pigments in Sahul (the Pleistocene landmass of Australia and New Guinea). The Madjedbebe grinding stone assemblage reveals economic, technological and symbolic innovations exemplary of the phenotypic plasticity of Homo sapiens dispersing out of Africa and into Sahul.
... Several taxa were utilised in the experiments including brushtail possums (Trichosurus vulpecula), eastern grey kangaroos (Macropus giganteus) and emu (Dromaius novaehollandiae), as they are common taxa within contemporary Australian faunal and archaeofaunal assemblages (e.g., Barker 1987;O'Connor et al. 1998). Brownstripe red snapper (Lutjanus vitta) and green turtle (Chelonia mydas) were also included as these genera were identified in the Boodie Cave assemblage (Veth et al. 2017b). ...
Article
Burning characteristics of bone are dependent on taxa, bone condition, and heating method. Until now, heating characteristics of Australia’s unique fauna were undocumented. Here we present heating experiments of Australian taxa using a laboratory kiln, hearth fires and earth ovens, and apply the results to faunal assemblages from Boodie Cave, northwest Australia. Results indicate comparable burning characteristics in Australian faunal bone to previous experimental research, with clear differences in the temperature at which burning characteristics occur between taxa and in particular marine and terrestrial fauna. Earth oven cooking causes very limited burning of the extremities of faunal bones, whilst hearth fires produce a high degree of non-uniform calcination and fracturing. In the Boodie Cave assemblage, low levels of heating throughout the deposit are consistent with post-depositional burning and fleshed cooking, though the exact heating process remains unclear in the absence of any identified combustion features.
... On the opposite side of the valley is a small creek line. Several small rock holes and soakages can be found to the east at the head of the valley (McDonald et al., 2018a;McDonald et al., 2018b;O'Connor et al., 1998). ...
Article
Despite environmental factors being at the forefront of socio-ecological models in Australian archaeology, detailed local environmental and vegetation datasets are uncommon. Such data is important in assessing, for instance, if and how shifting climatic conditions influenced and conditioned hunter-gatherer movements and choices. Archaeological re-excavation of Karnatukul (Serpents Glen) in Katjarra (the Carnarvon Ranges) provided an opportunity to undertake anthracological (archaeological wood macro-charcoal) analysis. This data offers an insight into the earliest uses of firewood and collection strategies in the Australian Western Desert. This study aimed at testing global anthracological methodologies to examine the problems and potentials offered by this important sub-discipline which is currently developing in Australian archaeology. This study makes an important contribution to international anthracological studies, given these are rarely applied to arid contexts, especially with an occupation record spanning almost 50 ka. The study demonstrates the presence and persistence of Acacia (sens. str.) woodlands from the Pleistocene, through the Last Glacial Maximum, and into the Holocene with the case made that this productive plant makes an essential contribution to the habitability of this arid landscape.
Chapter
Full-text available
Mainland Australia was connected to New Guinea and Tasmania at various times throughout the Pleistocene and formed the supercontinent of Sahul. Sahul contains some of the earliest known archaeological evidence for Homo sapiens outside of Africa, with a growing record of early complex social, technological, and artistic life. Here we present an overview of the oldest known sites in Australia along with key evidence pertaining to the dynamic cultures of early Aboriginal peoples. We review debates surrounding the age of rst settlement and present evidence for the earliest technology, economy, and symbolism in Australia, emphasizing maritime skills, a large founding population size, novel technology, and adaptation to a wide range of environments.
Article
Mainland Australia was connected to New Guinea and Tasmania at various times throughout the Pleistocene and formed the supercontinent of Sahul. Sahul contains some of the earliest known archaeological evidence for Homo sapiens outside of Africa, with a growing record of early complex social, technological, and artistic life. Here we present an overview of the oldest known sites in Australia along with key evidence pertaining to the dynamic cultures of early Aboriginal peoples. We review debates surrounding the age of first settlement and present evidence for the earliest technology, economy, and symbolism in Australia, emphasizing maritime skills, a large founding population size, novel technology, and adaptation to a wide range of environments.
Article
Full-text available
Until a few years ago it seemed possible that the arid interior of the continent was unoccupied prior to about 10,000 to 12,000 years ago (cf. Bowdler 1977; Horton 1981). Enough archaeological evidence has now accumulated to show that the major desert uplands and river systems were occupied during the late Pleistocene (Brown 1987; Lampert and Hughes 1988; Maynard 1980; Smith 1987; Smith et al. 1991) and the debate has moved on to questions about the nature of this occupation (Hiscock 1989; Smith 1989) and whether or not there were significant differences in the timing of settlement in different parts of the arid zone. In an influential paper Veth (1989b) drew attention to the lack of evidence for occupation of the major sandridge deserts prior to 5000 years ago and argued that these regions would have been first colonised in the mid-Holocene. Given the paucity of evidence about the prehistory of the sandridge deserts, Veth has put forward what is essentially an a priori argument for their mid-Holocene colonisation, hinging upon the premise that before this time, there were persistent biogeographic barriers to the human settlement of these regions. The 'barrier desert' theory, as I have dubbed it, warrants critical examination because it is the cornerstone of a model explicitly proposed as a new framework for desert prehistory (Veth 1989a and 1989b). The purpose of this paper is to examine the 'barrier desert' theory in some detail. In turn I hope this will open up for discussion a range of issues concerning the biogeography, human ecology and prehistory of the arid zone. The paper is in four parts. The first considers whether the major sandridge deserts, the Great Sandy, Great Victoria and Simpson Deserts, form a coherent biogeographic unit. The second part asks whether the 'barrier' deserts are fundamentally different to anything prospective colonists would have previously encountered in other parts of the arid zone. The third part of the paper deals with social and technological prerequisites for the colonisation of these regions. The final section reviews the archaeological evidence.
Article
A characteristic feature of human subsistence as the last glaciation ended was the turn towards new food sources, in a ‘broad spectrum’ transformation. Australia took an unusual course, and the trajectory in its arid zone is especially striking. What were the broad spectrum diets in arid Austalia? Why did they arise so late? Did they arise late?
Article
The Holocene faunal assemblage from Puntutjarpa Rockshelter in the Western Desert has always been problematic owing to the extreme fragmentation of bones and teeth there. This study examines alternative explanations for this fragmentation by looking closely at the Puntutjarpa material and also by controlled comparisons with late Holocene faunal materials from Intitjikula Rockshelter (James Range East), Northern Territory. The results of this analysis indicate that certain taphonomic factors probably account for extreme fragmentation of animal bones at both sites. But gross differences in fragmentation of macropod teeth at these two sites point to behavioral explanations related to dietary shortfalls of meat and meat-related products resulting from variability in the biogeography of hunting in these two arid-zone localities. An ethnographic-behavioral model is proposed to account for differences in macropod tooth reduction at these two sites.
Article
The full glacial climate was marked by enhanced aridity, suggesting that there would be no human occupation of the interior of the continent at the glacial maximum. However, evidence from Puritjarra Rockshelter shows that the Central Australian Ranges continued to be occupied between 22,000 and 13,000 BP. The repeated use of Puritjarra, together with its location away from any natural corridor for travel into the region, indicates the presence of a resident local population. The archaeological evidence is complemented by a model of human ecology in Central Australia at 18,000 BP, showing that there is no a priori reason for expecting the region to have been totally abandoned during the last glacial maximum.
Article
In the Holocene the Pama-Nyungan language family expanded to cover most of Australia, replacing languages previously spoken in many regions. Most phases of this expansion in the west of the continent have occurred in the last 3000 years as a result of migration, and can be correlated with archaeological evidence of new activity. Intermeshed with the migration of peoples and language expansion are cultural diffusion events which can be assigned a chronology relative to each other and to the migration events by linguistics. These relative chronologies have the potential to become absolute chronologies with the assistance of archaeological evidence. Even without many such clues, however, the method of ‘backtracking’ described here allows us to construct detailed chronologies for the late Holocene. Known stages of expansion and cultural diffusion are sequenced making reasonable inferences about the time needed between each stage and ensuring that sequences in different regions which link together also connect chronologically, leading to an estimate of 6000 BP for Proto-Pama-Nyungan. Particularly in focus in this paper is the chronology of diffusion of the western section and subsection forms of social categorisation in the last 2000 years, and how this interacts with the later phases of Pama-Nyungan expansion.