Cuddie Springs critique

[Tim Barrows]

Dear listmembers,

The following message is from Bruno David, please reply directly to 
him or to the list for discussion.

cheers,

Tim.

_______________________________________________________

Dr. Bruno David
Dept. Geography & Environmental Science
Monash University
PO Box 11A
Clayton, 3800
Australia
Bruno.David at arts.Monash.edu.au
Phone W (03) 9905 2938
Fax (03) 9905 2948

Extract from book, from Chapter 8 ('Seeds of Change'):

Bruno David. In press.  Landscapes, Rock-art and the Dreaming: an
archaeology of preunderstanding.  Continuum Press, London.

Yet a single locale with reported Pleistocene seed grinding stones is
left:  the central northern New South Wales spring site of Cuddie
Springs.  Little has so far been published of the geomorphological or
taphonomic contexts of these finds.  As few as 21, and possibly as many
as 26, Pleistocene grinding stones have been claimed from this site.
Based on a combination of the stones' morphologies and on the presence
of certain classes of use wear and residue, at least five of the
grinding stones are treated as plant processing millstones or mullers by
the excavators.  All are claimed to be around 30,000 years old, and all
of the claimed Pleistocene specialised seed grinding stones illustrated
are very small fragments, less than 7.2cm maximum length (Fullagar &
Field 1997, fig. 3b).  However, only one of these (artefact CS6034) was
found below a chrono-stratigraphically critical 'deflation surface'
(Stratigraphic Unit 5, Archaeological Level 4) that is overlain by mixed
mid Holocene and Pleistocene sediments.  This 'deflation surface' has
been reported as a key to the interpretation of the site's stratigraphic
integrity.  Field and Dodson (1999, p. 284) report that the 'bone
recovered from this unit was fragmented, and the degree of
mineralisation varied considerably 
 Since some of the bone was very
dense and black, while other bone was brown in colour and friable, the
bone assemblage may be mixed'.  Consisting of an up to 5cm thick mass of
interlocked rocks and bones (including stone artefacts), Stratigraphic
Unit 5 'marks the upper limit of the in situ megafauna deposits' (Field
& Dodson 1999, p. 283), and with this the upper limit of the in situ
stratigraphy generally (see Photo 12 at
http://artalpha.anu.edu.au/web/arc/resources/photos/cuddie/cuddie97.htm).
Said to be continuous across the excavation squares, this 'deflation
surface' is reported as a 'pavement of stone formed by the lag deposits
[that] seals the overlying clays from the archaeology and faunal record
below' (Field & Dodson 1999, p. 284), an 'archaeology and faunal record
below' that threatens to revise our understanding of human-megafaunal
relations as well as of the antiquity of systematic seed grinding in
Australia.  I ask:  where do the Stratigraphic Unit 5 rocks come from in
this stone-poor riverine plain?  The modern lake floor itself consists
of grey alluvial soils, and the surrounding plains of red soils.  The
other excavated strata below and above this unit are clayey
(Stratigraphic Unit 4), clayey and silty (Stratigraphic Unit 6), clayey
and sandy (Stratigraphic Unit 1), silty (Stratigraphic Unit 3) or
gravely (Stratigraphic Unit 2).  Field and Dodson (1999, pp. 291, 292-3)
write:  'The low relief in the area around Cuddie Springs, the
fine-grained sediments that comprise the deposit and the absence of any
stone outcrops for at least 4km supports the notion that the stone
present in the archaeological levels must have been brought on site by
people'; 'The closest and most accessible stones are silcrete cobbles
from a gibber plain (a plain covered in cobber sized stones) c.4km west
of the site'.  How did the rocks come to be deposited in Stratigraphic
Unit 5, and what is their source?  And has the 'pavement' truly sealed
the underlying sediments since the late Pleistocene, given the evidence
of extant (but presumably dead) tree roots immediately below it (see
Photo 14 at
http://artalpha.anu.edu.au/web/arc/resources/photos/cuddie/cuddie97.htm).
Geomorphological investigation is required to answer whether or not the
pavement is anthropogenic, perhaps constructed by farmers during the
late 19th or early 20th century to create a firm footing for people or
cattle (a European well was dug in 1876 but 10m from the excavation; did
the rocks come from the well sediments, as one possibility, or was it
brought in from the gibber plain 4km away?).  The pavement's possible
post-European antiquity is only strengthened by the presence of cattle
bone in the stratum immediately overlying it:  'Archaeological Level 5
is considered to be significantly disturbed as the bone from this level
contains skeletal elements of both megafauna and modern cow' (Field &
Dodson 1999, p. 295). Tracing the extant of the pavement through
excavation or other means, including an assessment of its relationship
to the recent well, may be warranted to resolve such questions.
The stratigraphic integrity of underlying cultural sediments at Cuddie
Springs is also problematic.  First, the site is a ground-fed spring
according to Flannery (1997), although this is not clear from the
literature as it is also reported as an 'ephemeral freshwater lake' by
the excavators (Fullagar & Field 1997, p. 300; also Field & Dodson 1999,
p. 279).  As a site that is periodically inundated, doubts will remain
of the stratigraphic integrity of sediments until geomorphological
results appear in press.  The question of whether or not deposits become
fluid or viscous when wet requires address.  The presence of
stratification within the deposits is not in itself sufficient to dispel
this potential problem, for viscosity of the sediment matrix may affect
the stratigraphic integrity of certain particle masses but not others.
In particular, the question needs to be asked as to whether or not
stone, bone, large pieces of charcoal and other relatively large items
had a tendency to settle onto a firmer but mixed stratigraphic level
during times of inundation when sediments were muddy, creating a
contaminated deposit in the critical Stratigraphic Unit 6A.  Secondly,
stone artefacts occur in direct chronological association with extinct
Pleistocene fauna (particularly in Stratigraphic Unit 6B), including
Sthenurus, Diprotodon and Genyornis (although only Sthenurus is claimed
to have been directly used by people, as evidenced by a burnt femur)
(Dodson et al. 1993; Field 1999).  The giant flightless bird Genyornis
in particular has recently been the subject of intensive study in the
Lake Eyre basin to the west of Cuddie Springs, where its extinction has
been dated to 50,000±5000 years ago (Miller et al. 1999), at least
15,000 years before their apparent presence at Cuddie Springs.
Furthermore, Field and Dodson (1999, pp. 294-5) note that 'The presence
of a Pallimnarchus sp. tooth in AL2 [Stratigraphic Unit 6A,
Archaeological Level 2] is considered intrusive and may have been
derived from an exposure in another part of the site, for example well
digging.  The tooth is exfoliated and heavily mineralised and the
preservation is not consistent with other bone and teeth from this
level'.  And again:  'The incidence of heavily mineralised bone is
higher [in Stratigraphic Unit 6A = Archaeological Levels 2 and 3] than
in AL1 [Stratigraphic Unit 6B] and increases again in AL4 [Stratigraphic
Unit 5, the 'pavement'].  It is not yet clear whether differences in
mineralisation of bone represents differential preservation or the
presence of intrusive material from other horizons' (Field & Dodson
1999, p. 295).  In these critical units there is thus evidence of
intrusions.  Never before have such megafaunal remains been found in
direct chronostratigraphic association with cultural materials in
reliable stratigraphic contexts , fuelling scepticism over the integrity
of deposits.  Indeed, all of the megafaunal species represented at
Cuddie Springs are commonly suspected of having become extinct more than
40,000 years BP, eight thousand years or more before the apparent age of
the deposits in which they are found.  In this context, it must also be
asked why so many stone artefacts (including grinding stones) are found
in sediments suggesting past 'marshy conditions', and 'shallow, still
freshwater environments' of the earliest 'cultural' levels (Field &
Dodson 1999, p. 285, and elsewhere).  And if 'a substantial proportion'
of the megafauna in the earliest 'cultural' sediments 'died in situ',
why are there no articulated megafaunal bones in those strata?
There are other signs that there may be critical taphonomic problems at
Cuddie Springs, not least from the distribution of radiocarbon dates.
Of note are the 10 AMS and conventional dates from the critical and
supposedly more or less intact Stratigraphic Unit 6 below the 'pavement'
(Field & Dodson 1999, table 3).  From the lowest part of this layer,
Stratigraphic Unit 6B (=Archaeological Level 1), come six radiocarbon
dates averaging 30,720±160 years BP. Stratigraphic Unit 6A
(=Archaeological Levels 2 and 3) above has four dates averaging
29,553±170 years BP (averages calculated on Calib 4.1).  Despite this
apparent near-contemporaneity, the faunal and stone artefact remains in
the two strata show major differences.  Of special concern is that the
bone distribution implies a sequential stratigraphic structure, with an
increasing incidence of modern fauna and decrease in megafauna in the
upper Stratigraphic Unit 6A, despite the apparent near-contemporaneity
of Units 6A and 6B as indicated by the radiocarbon dates:  'The bone
assemblage from AL2 was more fragmented compared to AL1, as well as
containing a higher percentage of extant species. 
 The species
composition of AL2 shows a greater number of animal species represented
in the deposit compared to AL1, with four species of extinct fauna
identified.  Genyornis newtoni is now absent 
' (see also Field & Dodson
1999, table 5 for its presence in Stratigraphic Unit 6B, although we are
also told that 'By the time people had arrived at Cuddie Springs 

Genyornis newtoni 
 are all extinct' [Field & Dodson 1999, p. 293]).
These are worrying signs:  why is there a sequential change from much
modern fauna and few megafauna in the upper cultural levels of
Stratigraphic Unit 6, to increasing proportions of megafauna in the
lower parts of this Unit, if these strata are near-contemporaneous?  Are
we to believe that we are faced with the moment of extinction?  More
likely is the possibility that there has been some significant degree of
mixing of likely pre-human megafaunal deposits with more recent cultural
deposits.  But if such mixing has taken place, it becomes difficult to
determine which artefacts relate to which periods of time within the
cultural sequence, especially if the overlying strata contain very
recent materials (including cattle bone).
Also of concern is the apparent association of Pleistocene dates and
Tula adzes or Tula adze-like artefacts, a stone tool type shown by Peter
Hiscock and Peter Veth (1991) to be restricted to the mid to late
Holocene elsewhere in the arid zone.  John Dodson et al. (1993, p. 97)
have thus noted of Cuddie Springs:  'from sediments dated between 30,000
and 19,000 BP, grindstones, ochre fragments and woodworking tools with
identical usewear and similar morphology to tula adzes (Kamminga 1982)
were recovered'.  While the Cuddie Springs sediments are reported to be
stratified, the absence of post-depositional movement of materials
within and between layers must be demonstrated empirically via
geomorphological studies before this otherwise important site that
threatens a need to revise established wisdom on human-megafaunal
relations as well as on the antiquity of seed grinding in Australia can
be reliably accepted.