Settlement and Occupation of the Wellesley Islands,
Southern Gulf of Carpentaria, Australia
Daniel Rosendahl1,2, Sean Ulm2, , Paul Memmott1,
Errol Stock2, Sheila van Holst Pellekaan3
1 Aboriginal
Environments Research Centre, University
of Queensland
2
Aboriginal and Torres Strait Islander Studies Unit, University of Queensland
3
School of Biotechnology and Biomolecular Sciences, University of New South
Wales
Abstract:
The Wellesley Islands
in the southern Gulf of Carpentaria play a key role in our understanding of
Aboriginal lifeways in northern Australia.
Pioneering anthropological studies by Roth and Tindale and later by Memmott,
Macknight, Evans and Trigger document diverse linguistic, kinship and material
culture relationships across the two main island groups owned by Yangkaal,
Kaiadilt and Lardil people and the adjacent mainland Ganggalida. However, the
origins of these cultural phenomena remain poorly understood, with no detailed
published archaeological studies undertaken on any of the numerous islands
which make up the Wellesley
group. Recent archaeological studies to the west on Groote Eylandt and in the Sir Edward Pellew Islands
and in the east in the western Torres Strait have documented major changes in
Indigenous lifeways in the last 1000 years, but it is unknown whether cultural
trajectories in the Wellesley
Islands paralleled or
diverged from these patterns. The Wellesley
Islands have also taken on renewed
importance with the publication of genetic studies that suggest Lake Carpentaria
and its associated river systems to the south, close to the Wellesley Peninsula,
may have been key areas in the colonisation of the continent.
Introduction and Objectives
Recent
studies in northern Australia
identify the Gulf of Carpentaria as a key region in developing our
understanding of initial occupation of Australia and mid-to-late Holocene
cultural dynamics. This research has been conducted in the western Torres
Strait (David and McNiven 2004; McNiven and Hitchcock 2004), western Cape York
Peninsula (Bailey 1999; Beaton 1985), the Sir Edward Pellew Group in the
southwest southern Gulf (Bradley 1998; Prebble et al. 2005), Groote
Eylandt in the west (Shulmeister and Head 1993; Shulmeister and Lees 1992;
Specht and McCarthy 2005), and the Aru Islands in the far northwest (O’Connor
et al. 2005) (Figure 1). All studies document similar late Holocene shifts
in subsistence and settlement patterns but vary in interpretation and
explanation. Data from the Wellesley Islands have not yet been sufficient to contribute to
these debates, leaving a major regional gap in our knowledge of the archaeology
of northern Australia.
Across the Gulf researchers are finding similar results in chronological
trends throughout the mid-late Holocene based on fundamental changes in the
archaeological record which have been summarised by Ulm (2004), including:
- Increased
rates in site establishment and use evident in increased discard of
cultural materials, such as stone artefacts;
- Emergence
of new stone artefact technologies, rock art styles and fishing
technologies;
- Intensified
subsistence with further exploitation of new foods including cycads,
cereals and some marine resources;
- Increased
occupation of marginal environments; evidence for long-distance exchange
networks;
- Establishment
of bounded cemeteries;
- Increased
external contact evident with the emergence of the dingo, fish hooks and
some forms of watercraft.
These changes have prompted a wide range of explanatory frameworks to be
invoked.
- Changes
in social structure with trends towards socioeconomic intensification;
- Differential
site preservation and site destruction;
- External
pressures such as cultural contact and the introduction of new
technologies;
- Environmental
factors, including resource productivity and availability;
- Population
increase and demographic structure (Ulm
2004:189-190).
Similar
questions can be raised for the authors’ study area, the Wellesley Islands.
In fact, the Wellesleys present an ideal case study to investigate these issues
with four Aboriginal Australian populations that have diverged from a common
linguistic ancestor (Evans 2005), living in geographically similar contexts,
but with different material culture repertoires that in part have been adapted
distinctly different ways for subsistence strategies despite a common set of
resources (Memmott et al. 2006). With more detailed archaeological
investigations, the Wellesley
Islands will be able to
contribute to this debate on mid-late Holocene cultural change.
The
research potential of the Wellesley
Islands is highlighted
with the publication of recent genetic evidence (Friedlaender et al 2005; van
Holst Pellekaan et al 2006). Based on mitochondrial DNA (mtDNA), for which
there is a large global database available for comparison, both of these
studies show that several lineages of the major haplogroup P (subgroup of
global ‘N’) are common to both Papua New Guinea (PNG) and Australia. van
Holst Pellekaan et al (2006) identify two subgroups of ‘P’ in north/central
Australia and the Darling River region and note that there is strong evidence
for common ancestry of PNG and Australian ‘P’ matrilines that are
distinct from other haplogroups (‘M’, ‘S’ and ‘O’) identified in their study,
raising the possibility of different colonizing groups entering Australia. This
opens the question of a southern dispersal route along the coast of PNG (then
the northern margins of the Sahul landmass), with an ancestral population entering
from South East Asia and diverging at the Bird’s Head peninsula, subsequently
spreading along the north and south coasts. Colonists on the southern route
would have encountered the lowland, resource-rich plains and associated rivers
of Lake Carpentaria. The evidence offers
multiple points of entry by the first Aboriginal Australians emphasising a need
for research in the Wellesley
Islands and not just the
North West Australia where research has been focused in the past. Further
genetic studies that include participation from Queensland
and the Wellesley
Islands peoples would
facilitate our understanding of these dispersal issues.
Hunter-gatherer
demography is by and large governed by the availability of freshwater (e.g.
Gale and Carden 2005; Nutley 2005). Intrinsic to occupation, freshwater is a
fundamental constant in site location modelling. Various colonisation models
focus on this resource, such as Tindale’s (1941) ‘Riverine Corridor’ model,
Bowdler’s (1977) ‘Coastal Colonisation’, and Horton’s (1981) ‘Water and
Woodland’s’ hypothesis. With no shortage of freshwater in and around Lake Carpentaria
(Chappell et al. 1982; Chivas et al. 2001; Hope and Aplin 2005;
Torgersen et al. 1983; Torgersen et al. 1985; Torgersen et al.
1988) this region offers an environment rich in subsistence resources and river
systems that extend into the heart of the continent. The regional
archaeological significance of the Wellesley Islands is in its location as
islands that once formed part of a prominent ridge adjacent to major
palaeo-rivers such as the Nicholson, Leichhardt, Albert and Flinders Rivers
that connected Lake Carpentaria to the interior of the continent and
specifically to what are now called the North-west Highlands of Queensland.
Available
datasets for the Gulf of Carpentaria offer a
plethora of information enabling the reconstruction of palaeo-environmental
conditions for the entire Gulf. Representative time-series data on vegetation,
faunal species occurrence, location of freshwater and major rivers, palaeo-lakes
and lagoons can be used to develop predictive models of potential points of
human occupation and possibly dispersal rates and routes into the interior of
the continent. Research in progress on and around the Wellesley
Islands will increase the regional
archaeological knowledge to aid in understanding late Holocene cultural shifts
and provide insights on the initial peopling of Australia.
This research is a component of the ARC Discovery-funded “Isolation, Insularity
and Change in Island Populations: An Interdisciplinary Study of Aboriginal
Cultural Patterns in the Gulf of Carpentaria” project, a joint research
programme between academic institutions and Aboriginal communities to record
sites, provide skills for site conservation techniques and obtain information
about pre- and post-European contact amongst the four main language groups
occupying the Wellesley Islands. The research project has been designed to
systematically research the islands and increase the cultural and physical
knowledge of the region. Research will collate representative data for each
language group, Lardil, Yangkaal, Kaiadilt and Ganggalida. The archaeological
component of the research project is to identify the chronology of occupation,
divergence and transformation of cultures and to identify occupation trends
during times of dynamic coastal change. The background research for this topic
will take a multi-disciplinary approach drawing on anthropological, linguistic,
ethnohistorical, genetic and geomorphological research to aid with fieldwork
preparation and to contribute to the limited available archaeological data.
The
archaeological component of the research will contribute to the overall ARC
Discovery project by investigating cultural and physical environments using
geoarchaeological methods. Working with environmental scientist Dr Errol Stock,
anthropologist and long term researcher for the region Associate Professor Paul
Memmott and archaeologists Dr Sean Ulm and Dr Richard Robins a chronological
sequence will be established for cultural and natural landforms as situating
erosional, depositional and cultural events into a chronological framework. The
data sourced from this research along with available data from environmental and
archaeological literature will be integrated using a Geographic Information
System (GIS) to aid with predictive site location modelling and to present data
coherently to peers and Aboriginal communities.
Literature Review, Synthesis and Significance
This
section presents a description of the study setting and a synthesis of previous
research in the Gulf of Carpentaria with a focus on the southern Gulf and Wellesley Islands. The four Aboriginal language
groups of the Wellesley
Islands are introduced in
an anthropological context. The conclusion will explore the significance of
this research to the community and the local, regional and continental
implications in archaeology and the earth sciences.
The Physical Setting
The Gulf of Carpentaria is an enclave situated in the central north of Australia
(Figure 1). Located between latitudes 10° and 15° south and longitudes 136° and
143° east, the Gulf is framed by Cape York Peninsula in the east, Arnhem Land
in the west, Papua New
Guinea in the north and the Carpentaria
Plain in the south. Oceanic waters breech the Arafura and Torresian sills in
the west and east respectively. The coastline is constituted by numerous
prograded shores, islands and archipelagos with the main groups from east to
west being the Torres Strait, the Wellesley Islands, the Sir Edward Pellew
Group and Groote Eylandt (Figure 1). The next section outlines geographical and
environmental processes that contributed to the formation of the Gulf,
accompanied by a review of flora and fauna.
The Geology and Geomorphology
The
Australian coastline has been shaped by numerous environmental processes
ranging from localised wave winnowing and aeolian depositional events that
increase the continental landmass through progradation, to large-scale erosional
forces of tropical cyclones, storm surges and tsunamis that scour existing
coastlines. Assuming the potential span of human occupation to be the last
50,000 years, it is relevant to describe the major erosional and depositional
events that shaped the Gulf with particular reference to the coastal
geomorphology of the early to mid-Holocene.
|
|
Figure 1. Gulf of Carpentaria showing Lake Carpentaria
and coastal islands
|
The Wellesley Islands lie within the Leichhardt-Gilbert geological
region of northwest Queensland
on the Donors Plateau section of the Carpentaria Plain (Twidale 1966). The
Carpentaria Plain is bordered by the Isa
Highlands to the west,
Inland Plains to the south and Einasleigh Uplands in the east (Twidale 1966).
This region is a broad corridor c.85km wide with an altitude ranging from
sea-level to c.200m above sea-level (Twidale 1966:31-32). The shoreline is
characterised by progradation and episodic chenier
plain formation, with hundreds of kilometres of chenier ridges and
transgressive dunes parallel to the contemporary coastline (Gillieson 2005;
Robins et al. 1998:81-86). The Carpentaria Plain drains directly into
the Gulf of Carpentaria through major rivers
such as the Leichhardt, Flinders and Gilbert. The plain is divided up into the
following physiographic divisions: Cloncurry Plain, Croydon Plain, Donors
Plateau, Julia Plain, Claraville Plain, Wondoola Plain,
Stirling Plain and Karumba Plain (Twidale 1966:9). This region is characterised
by old, weathered terrain bordered on the east and west by ranges and limestone
formations.
Climate
Climatically the region is characterised as semi-arid and monsoonal
reflecting diverse climatic processes. Regional weather records have been
obtained from the Burketown and Normanton stations and more recently one on Mornington Island. The mean annual maximum
temperature of 29.8°C, a mean annual minimum of 22.1°C, and a mean annual rainfall
of 1214mm (Bureau of Meteorology 1995:201). The southern Gulf is characterised
by a climatic regime of hot wet summers and cool dry winters. On average the Gulf of Carpentaria experiences one cyclone event every
two years (Gillieson 2005:207). Over the last 5000 years it has been argued
that major storm events or super cyclones strike the eastern Gulf episodically
(every 200-300 years)(Nott and Hayne 2001) with extreme intensities capable of
large-scale coastal degradation as documented elsewhere along the tropical
coastline (Bird 1992; Przywolnik 2002).
Hydrology
The Gulf of Carpentaria exhibits a mixed hydrological system with ocean
waters entering from the Arafura Sea and Pacific Ocean through the Torres Strait. The predominant current
flows in a clockwise direction (Forbes and Church 1983) with oceanic processes
depositing sediments along the characteristically prograded shores of the Gulf.
The Carpentaria Plain supports a large river catchment draining north into the
Gulf through a large network of rivers including the Nicholson, Leichhardt,
Flinders, Cloncurry, Saxby, Norman, Clara, Yappar, Gilbert, Einasleigh and
Staaten (Twidale 1966:24). The watershed dissects Cape
York extending around to the south of the Carpentaria plain. The
watersheds to the south either drain into Lake Eyre through the Diamantina River
system or east into the Pacific Ocean via the Burdekin River basin.
Flora and Fauna
The Wellesley Islands are resident to numerous endemic
and exotic plant and animal species, as well as migratory birds and mammals
(Malcolm 1998; Memmott 1979, 1989) that inhabit the diverse seasonally variable
habitats. Although at a macro-scale the Wellesleys
exhibit a largely homogenous environment, each island exhibits micro-variations
in ecosystems that support varying combinations of species of plants, fish,
insects and terrestrial vertebrates.
Sweers Island, for example, is a small island
without permanent surface freshwater or estuarine systems (Gale and Carden 2005:182-183),
restricting the range of ecological habitats. The absence of estuaries has
resulted in minimal mangrove communities and concomitant complex inter-tidal
habitats, rendering the inter-tidal fauna comparatively simple and marine
orientated (Davie
2005:159). The reefs however support a huge range of species, Malcolm (1998)
recorded 126 species of fish, and numerous species of bird including Brown
boobies Sula leucogaster, Greater frigate birds Frigata minor,
Crested terns Sterna bergii, Ospreys Pandion haliaetus and Beach
curlews Esacus magnirostris, in a two week incidental observation survey
of the island cluster. Malcolm (1998) states that the Wellesleys are of high
significance as a habitat for turtles, birds and dugong.
Palaeo-Environment and Environmental Change
For
much of the Pleistocene, Australia
and New Guinea formed part
of the larger Sahul landmass connecting Australia,
New Guinea and parts of
island South East Asia. During this period the
Gulf of Carpentaria was dominated by low plains and a large freshwater lake
which Quaternary scientists have termed Lake Carpentaria.
At this time the Wellesley
Islands extended into the
Gulf as a low hilly range (Robins et al. 1998:77). The mid-to-early
Holocene saw dynamic coastal change with eustatic sea-levels separating the Wellesley Islands from the mainland. Rising
sea-levels (Figures 2a-2d), breached the Arafura land corridor around 10 800
BP, followed by the Torres Strait around 9000 BP with peak sea-levels around
5300 BP and stabilising around 4000 BP (Chivas et al. 2001; Holt 2005;
Reeves et al. 2007).
Palaeoenvironments
of the Gulf of Carpentaria have been reconstructed by Holt (2005), Reeves
(2007), Torgersen et al. (1983, 1985, 1988) and others through
sedimentological and faunal analyses of core samples spanning the past 130ka
BP. Palaeoenvironmental data can also be extrapolated from pollen records and
faunal assemblages from the Aru Islands that identify changing habitats and
environments (Hope and Aplin 2005; O’Connor et al. 2005). Data show an
oscillating environment with periodic oceanic breach and recession of the
Arafura and Torresian sills. Various environmental phases have been identified
ranging from high energy open marine systems to lacustrine freshwater habitats
(Chivas et al. 2001; Holt 2005; Reeves et al. 2007:163; Torgersen
et al. 1988). The basal deposits of one core dated to 130ka BP indicate
sea-levels 3m higher than present, with a recession of water level below the
Arafura sill at c.40 000 BP associated with extensive seasonal monsoons and
meteoric waters (Reeves et al. 2007:175). Between 19 000 -17 200 BP Lake Carpentaria was at its deepest and abundant in
aquatic bivalves and fish species. The first evidence for marine incursion is
at 10 800 BP with present environmental conditions stabilising around 4700 BP
(Figure 2a-2d) (Holt 2005; Hope and Aplin 2005; O’Connor et al. 2005;
Torgersen et al. 1988).
|
|
|
Figure 2a. -63m contour
pre-12.7ka BP (Holt 2005).
|
Figure 2b. Breach of Arafura
sill -53m contour 10.8ka BP (Reeves et al. 2007).
|
|
|
|
|
Figure 2c. Breach of
Toressian sill -12m contour c. 8.7ka cal BP (Holt 2005).
|
Figure 2d. Present day
sea-level established by 4.7ka BP (Holt 2005).
|
Cultural Setting
The Wellesley Islands
each contain a cultural landscape created by the Lardil, Yangkaal and Kaiadilt on
the two island clusters, as well as being supplemented with further cultural
properties by seasonal Macassan trepangers from Sulawesi, Indonesia
and more recent ongoing European contact. The Lardil and Yangkaal occupied the
North Wellesleys, whilst the Kaiadilt occupied the South
Wellesleys with Ganggalida residing on the adjacent mainland.
However, an extensive oral history (including sacred history) connects the
mainland Ganggalida to the Lardil and Yangkaal of the North Wellesleys for
purposes such as camping and hunting territories, with the Yangkaal occupying
the ‘stepping stone’ islands connecting Mornington Island to the mainland at
Bayley Point. As a result the area is characterised by shared cultural
seascapes (Figure 3). There is a parallel argument to evaluate whether the
Ganggalida and Kaiadilt interacted via the ‘stepping stone’ islands of Allen,
Horseshoe, Bessie, Margaret and Albinia
Islands. However
preliminary analyses indicate that the extent of exchange could not have been
nearly as prolific as for the North Wellesleys
(Memmott et al. 2006).
Although
contact with neighbouring groups is noted both in oral histories and historical
accounts, each language group has developed its own unique material culture
while retaining its separate (yet nevertheless closely related) language. A
high level of autonomy is evident amongst the Kaiadilt of the South Wellesley
Islands who appear to
have resisted at times the adoption of ideas and technologies from neighbouring
parties. However, Tindale (1977) argues that the absence of such traits was
related to their isolation. Cultural differences are also evident in the
material culture repertoires of the North and South
Wellesleys, with the Lardil developing a much broader tool kit
with some 65 or more artefact classes as opposed to around 28 for the Kaiadilt
with only a 10% overlap according to preliminary investigations (Figure 4).
Although to date, Kaiadilt exhibit a narrow material culture repertoire, their
density of stone walled fish traps per kilometre of coastline exceeds any other
Australian Aboriginal population (Figure 5) (Memmott et al. 2006).
|
|
|
Figure
3. Cultural landscapes of the Wellesley
Islands
|
North Wellesley
Islands: Yangkaal, Lardil, Ganggalida
The North Wellesley
Islands were occupied by the Yangkaal
and Lardil, with the Lardil traditionally resident on Mornington Island
and its satellite islands. The Yangkaal resided on the ‘stepping stone’ islands
adjoining the mainland with the largest being Denham and Forsyth Islands.
Although having strong trade connections with the Lardil and Ganggalida, the
Yangkaal have a much closer language connection with the Kaiadilt (Memmott,
Evans et al. 2006). One Yangkaal word for country literally means
‘western country’ (Erich Round, Yale
University, pers comm.,
March 2007), implying an oppositional orientation to both the Kaiadilt to the
east and the Lardil to the north-east. The Yangkaal were the trade
intermediaries between the Lardil and visiting communities such as the
Ganggalida and the Yanyuwa from the northwest (Roughsey 1971). Ganggalida have
shared ownership claims to the islands closest to the mainland which fall
within both the North and South Wellesleys (Pains, Bayley and Allen Islands)
and claim hunting territories in those waters. Contemporary Ganggalida
territory extends from Massacre Inlet on the coast in the west, and southeast
to the bottom of the Gulf with recent succession into Mingin country settling
the boundary at the Albert
River; it extends inland
up the coastal creeks such as Cliffdale and Eightmile Creeks (Trigger 1987).
The Ganggalida have endured numerous phases of frontier violence beginning in
1864 with the “pastoral invasion” which remained commercially unstable until
1874. In 1889 the Queensland Mounted Police were brought to the area bringing
two decades of violence through the 1890s and the first decade of 1900s
(Trigger 1987). In 1914 the Presbyterian Church established a Mission
on Mornington Island adjacent to the Appel Channel,
whilst in 1933 the Christian Brethren Mission was established at the site of
Doomadgee (‘Dumaji’) on the coast in Ganggalida country.
|
|
|
Figure 4. Recorded material
culture repertoires of the Wellesley
Islands, with a 10%
overlap in Lardil and Kaiadilt repertoires. Yangkaal and Ganggalida have
still to be recorded.
|
|
|
|
Figure 5. Densities of stone
walled fish traps of the southern Gulf of Carpentaria
|
South Wellesley
Islands: Kaiadilt
The
larger of the South Wellesley Islands
consist of Bentinck, Sweers and Allen Island, While Bentinck and Sweers are occupied by the
Kaiadilt, Allen Island is, according to surveys during
the Native Title Claim, shared between the Kaiadilt, Yangkaal and Ganggalida.
At the time of European contact, the Kaiadilt supported the highest population
density of any Aboriginal Australian group and exhibited a specialised tool kit
for marine subsistence which included the walpa (raft), oyster stone (used to
prise oysters from rocks), a range of barbed and pointed spears, a handheld
small fishing net and a prolific range of stone walled fish traps. The stone
walled fish traps of the southern Gulf are a striking feature on aerial
photographs and satellite imagery with Bentinck
Island supporting the highest density
of fish traps in Australia
per kilometre of coast line. The first recorded account of European contact
with the Kaiadilt was in 1802 by Mathew Flinders (Flinders 1814). Some 140
years later, the Kaiadilt endured a decade of disruptive social and
environmental events between 1940-1948 including famine, storm surge/tidal
wave, internal conflict, and missionary intervention (Memmott 1982; Tindale
1962a:270) the population was left decimated. The last occupants of Bentinck Island
were relocated by missionaries to Mornington
Island in October 1948
(Tindale 1962b:300).
Macassan Contact
Prior to European contact, Macassans from
pre-colonial Sulawesi, Indonesia, embarked on seasonal voyages to
northern Australia,
at times planting groves of Tamarind trees to supplement their episodic visits
(Gale and Carden 2005:185; Macknight 1976; Macknight and Thorne 1968; Memmott
1979). There is
evidence that Fowler Island in the South Wellesley group, was a Macassan
beche-de-mer processing site which is supported by a grove of Tamarind trees
(Memmott 1979:238-239) as well as associated fragments of ceramics (Richard
Robins pers comm.) and numerous unidentified ship wrecks believed to be
Macassan prau with associated burials (Memmott 1979). Macassan contact in north
Australia
is widely accepted, what is unknown however, is exactly how far they ranged to
the east and west and the extent of their interaction with Aboriginal
Australians (Mulvaney 2002). If they ranged as far east as the Wellesleys, was it seasonally or only
sporadically? If Fowler Island can be confirmed as a Macassan site then it
would represent the most easterly recorded Macassan site in Australia.
Previous Research
Introduction
The
sparsely-known archaeological record of the Wellesley Islands, is supplemented
by ethnographic, linguistic and geomorphological literature (David 1995; Evans
1985, 1995b; Flinders 1814; Jackson 1902; McKnight 1967, 1979, 1986, 1995;
Memmott 1982, 1986, 1989, 1992, 1998; Memmott et al. 2006, Memmott,
Lilley et al. 2006; Memmott and Robins 1984; Memmott and Trigger 1998;
Roth 1901, 1903; Stubbs 2005; Stubbs and Saenger 1996; Tindale 1962a, 1962b).
These sources denote a diverse and rich archaeological record for the Wellesley Islands, with records and accounts from
navigators, missionaries and pastoralists as well as anthropologists, including
frequent accounts of the large array of stone walled fish traps.
Pioneer
Research
Roth
(1901, 1903) and Tindale (1962a, 1962b) were pioneers of research in the
Wellesleys with Roth conducting at least three visits over three consecutive
years during 1901-1903, whilst Tindale spent two seasons in the Wellesley
Islands in the early 1960s. Roth’s fieldwork consisted largely of cataloguing
and collecting material culture with observations and notes of the application
and/or manufacture of artefacts. Tindale’s work was mainly anthropological in nature, however, he also drew on archaeological research and
like Roth, collected material culture items for museum reference. Tindale’s research
on the cultural geography of Bentinck Island (Tindale 1962a, 1962b) led him to speculate on
the longevity of Kaiadilt occupation on the Wellesley
Islands and further on the Pleistocene
pathways into Australia.
Tindale supported his claim, attributing the differences in material culture
and physical anthropology as a result of isolation from other Aboriginal
communities.
Linguistic
Linguistic research has been carried out on the Wellesley for the past
few decades with a well recorded data set from the Kaiadilt and Lardil, but
patchy for Yangkaal and Ganggalida (Evans 1985, 1995a, 1995b; O’Grady et al.
1966). Linguistic analyses have demonstrated that the four language groups –
Lardil, Yangkaal, Ganggalida and Kaiadilt – of the southern Gulf
of Carpentaria diverged at various stages from a linguistic
ancestor founding group who spoke a language coined Proto-Tangkic by the
linguists (Evans 1985, 1995b; Keen 1983; O’Grady et al. 1966). Applying
the available linguistic data for the region, Evans (in Memmott et al.,
2006:35) has proposed a model for cultural divergence stemming back to the
Proto-Tangkic ancestor, with Ganggalida the first to diverge. Kaiadilt split
from Ganggalida around 1000 BP severely challenging Tindale’s (1962a) statement
that Kaiadilt speakers have been isolated for millennia and probably represent Australia’s
first colonists.
Genetic Research
The peopling of Australia has attracted much debate with hypotheses
including Tindale’s (1941) well-known ‘Riverine Corridor’ model, Birdsell’s
‘Rapid Colonisation’ model (1957), Bowdler’s (1977) ‘Coastal Colonisation’
model and Horton’s (1981) ‘Water and Woodlands’ model. Archaeologists and
prehistorians have scoured Australia,
South East Asia and Papua
New Guinea for clues to shed light on the
origins of the first Australians with no conclusive results or closure on the
matter. In light of recent results from genetic research and increased
archaeological investigations in Indonesia, Insular Southeast Asia
and Wallacea, the occupation of Sahul and the dispersal of people has once again emerged as a key research focus.
Genetic research of the Wellesley
Islands was pioneered in
the 1960s by Simmons, Tindale and Birdsell (1962). They hypothesised vast
differences in the populations of the North and South
Wellesleys with the Yangkaal and Lardil genetically
indistinguishable, with only an ancestral link to the Kaiadilt. The Kaiadilt
showed evidence of long-term genetic isolation from the North
Wellesleys and with genetic markers not found elsewhere in
Australian Aboriginal populations. Simmons et.
al. (1962) proposed an external genetic influence. European genetic input
has been ruled out largely on the accounts of early European explorers’
documentation of the unsocial behaviours of the Kaiadilt, whilst Malay or
Macassan influence has been deemed unlikely due to lack of Malay words in the Wellesley syntax. Simmons
et.al offered the explanation of mainland influence, possibly from
western Cape York.
This early work has been superseded by recent studies of mtDNA in the
broader region. Using these data, van Holst Pellekaan (2006) proposes a
migration from New Guinea
(van Holst Pellekaan 2006b), at that time the northern section of Sahul. This
evidence shows a genetic split or two migration routes following the north and
south coasts of present-day New Guinea
with the southern route taking the peoples onto the plains of Lake Carpentaria
and the numerous river systems that fed in from New
Guinea and Australia. The Gulf of Carpentaria
is recognised by researchers as potentially being one of the first colonised
habitats in Australia,
coined as a “sweet spot” (O’Connell 2006) or place of refuge, capable of
supporting large numbers of people.
An Archaeological Survey of the Southern Gulf of Carpentaria
A number of archaeological surveys have been carried out since the 1980s
by Richard Robins along with Traditional Owners, anthropologists Paul Memmott
and David Trigger and coastal geomorphologist Errol Stock. In total, 66
archaeological sites have been recorded with 12 reported in published sources
(Robins et al. 1998:87-122) and six with an established chronology
(Table 1). This next section provides details on the sites that have an
established radiocarbon age and a summary of the archaeology recorded in the
surveys.
Archaeological Excavations
Wurdukanhan shell deposit
This site is also known in the literature as ‘Wudukanan’ and/or Birri,
on the central north coast of Mornington
Island in Lardil country. It is located not far from the outstation of Birri
and within the Birri estate of a Lardil claim. Wurdukanhan is a site complex
consisting of a series of oyster shell deposits along a hypersaline mudflat
adjacent to Wurdukanhan Creek. In 1996 Robins conducted an excavation on a low
mound of shell approximately 8m in diameter and 30cm high. The substrate for
the site is moist, sticky, shelly clay. The excavation was a 25cm x 25cm test
pit with a maximum depth of 32cm comprising 10 excavation units (XUs). The
deposit comprised predominantly of the large black-lipped oyster Saccostrea sp.
although oyster was absent in XU 10. During the wet season this site is
inundated with freshwater and at times, saltwater. A basal and cessation date
has been obtained from a black lipped oyster from XU1 (5100±90 BP - Beta-100241)
and XU9 (5180±80 BP - Beta-100242) making it the oldest recorded site in the
southern Gulf of Carpentaria. This site
remains unpublished and is currently being analysed by the candidate.
Kanbaa
Kanbaa is situated along the central east coast of Mornington Island.
The site is located on the edge of the crest of a shelly ridge (3-4m high) 300m
west of the coast. The ridge has been dissected by a creek. Freshwater swamps
lie to the south whilst mangroves and an estuarine ecosystem lie to the north.
The ridge has little vegetation on the crest with open woodland on the flanks.
A single test pit of 25cm x 25cm was excavated to a maximum depth of 37cm. The
sediment matrix was dark shelly sand at the surface increasing to light compact
sandy shell towards the base. Two dates were obtained: 1710±80 BP (Beta-100240)
and 1920±80 BP (Beta-100239) dated material vertical reference unknown. This
site remains unpublished and is currently being analysed by the applicant.
Piyal Burial Site
Piyal burials constitute nine individual burials partially eroded from
the shifting sands on the crest of a sand dune. The burials are located on the
northwest coast of Mornington
Island on a low and
sparsely vegetated sand dune close to the beach. A single date of <700 BP (laboratory
reference number unknown) has been obtained from a tooth from one of the
individuals (Memmott and Robins 1984). The site was investigated owing to
concerns made by the Moyenda Association of the Wellesley
Islands (an incorporated Aboriginal
association which comprised of Elders representing all of the traditional
landowners of the Wellesley
Islands) in regards to
the preservation of the skeletal remains. An interesting fact about the burials
is that the Moyenda Association did not consider it to be a Lardil site for
numerous reasons such as the location, posture of buried individuals, type of burial and associated material culture or burial
goods. For instance one of the burials contained what appeared to be a
dentalium shell necklace, an item of material culture not recorded in the
Lardil repertoire. Further archaeological investigation could be taken on this
site to determine the origins of the Piyal burials if the Elders agreed.
Bayley Point or Gaabula
Bayley Point is located within Ganggalida country on the mainland in an
area characterised by diverse botanical ecosystems comprising numerous species
of tree, shrub, grass and creepers (Robins et al. 1998:100). Bayley
Point is a headland but is essentially an island as it is encompassed on three sides
by estuaries and on the fourth by the sea. The area is characterised by diverse
ecosystems and land systems such as saltpans, chenier
and transgressive beach ridges. The headland is an inlier of Tertiary laterites
(Aurukun Surface) in which shallow soils have developed allowing for vegetation
to become established. Ethnographically-recorded significances of the site
include it being part of a Dugong Dreaming estate in the vicinity. Two surveys
were undertaken by Richard Robins with Traditional Owners in this area in 1983
and 1988. Archaeological features at Bayley Point included numerous large
stone-walled fish traps and a concentration of sites along the large chenier system to the west of the fish trap. A single
excavation was carried out with radiocarbon dates obtained showing a consistent
chronological sequence of 440±70 BP (Beta-61791), 550±70 BP (Beta-28747),
850±80BP (Beta-37835) and 1100±70 BP (Beta-26903) (Robins et al.
1998:104-107).
Gunamula
Located on the mainland, Gunamula is situated on the eastern side of the
mouth of Cliffdale Creek approximately 25km west of Bayley Point. The central
feature of the site is a large freshwater lagoon situated in a dune swale
between Holocene beach ridges around 200m south of the beach and 400m east of Cliffdale
Creek. The area is characterised by transgressive beach ridges that terminate
on the salt pans approximately 1.2km south (Robins et al. 1998:91).
Gunamula is ethnographically significant as a permanent source of freshwater
and the focus of at least two Dreaming sites. There is a Devil Dreaming
pertaining to a single line of trees growing in the lagoon, and a Barramundi
Dreaming place (Robins et al. 1998:91). Archaeological evidence at this
site includes eight discrete shell scatters and shell lenses eroding from
parallel dunes extending from the coast. Two dates have been obtained for this
site from the eroding face of a dune that exhibited two distinct cultural shell
layers; the dates are 140±60 BP (Beta-12849) and 1300±80 BP (Beta-12850).
Old Doomadgee
Old Doomadgee is a transitional site with evidence of pre and post
contact occupation with an abandoned mission station built in an area of
pre-contact camps. Accompanied by several senior Ganggalida men, including
Willie Doomadgee and Major Walden, Richard Robins surveyed the area and
recorded cultural material such as corrugated iron, rusty tins and drums, glass
bottles and wire mesh as well as pre-contact camp sites. Two camp sites where
excavated with one radiocarbon age of 310±80 BP (Beta-28748). The location of
the excavation was described by one of the senior men as the site of the main
camp prior to the missionary period commencing in 1933. This site was revisited
by Robins six years after the initial survey in 1982 and he noted that all
remnant materials from the Mission
had been recycled in the construction of the new outstation (Robins et al.
1998:99).
Other sites
The southern Gulf of Carpentaria is
more than missions, middens, burials and shell mounds. Various anthropological
and archaeological surveys document a range of site types. Incorporating pre-
and post-contact ages, sites include stone walled fish traps, water holes,
grinding grooves, wells, inscribed trees, Macassan camps, sacred sites,
hearths, ship wrecks, plane wrecks, a stone walled pig
pen among others (Flinders 1814; Memmott 1998; Robins 1998; Robins et al.
1998).
Summary of Research, Conclusions and Future
Directions
With the exception of two unpublished dated excavations on Mornington
Island (Wurdukanhan and Kanbaa), archaeological research in the southern Gulf
of Carpentaria is concentrated along the mainland coastline or confined to the
numerous stone walled fish traps (Robins et al. 1998). However, many as
yet uninvestigated sites have been recorded in extensive field surveys carried
out by Paul Memmott and Richard Robins and Traditional Owners. This summary of
research shows a need for a more systematic and detailed approach to the
archaeology of the region to provide insight into various fields such as linguistics,
colonisation and human adaptive responses and strategies to large scale
environmental and climatic change and internal social pressures. The dated
occupation sequence shows consistent occupation in the southern Gulf of Carpentaria over the past 5000 BP. With further
archaeological investigations this chronology will be refined to represent the
central southern Gulf of Carpentaria seas to contribute to an archaeological
model of human occupation of the Wellesley
Islands.
|
Table 1. Technical
data on excavated sites of the southern Gulf of
Carpentaria.
|
|
Site
|
Location
|
Sample description
|
Age
|
Reference
|
|
Wurdukanhan
|
Mornington Island
|
XU1 – 0-4
|
5100±90
|
Beta-100241
|
|
XU9
|
5180±80
|
Beta-100242
|
|
Kanbaa
|
Mornington Island
|
NA
|
1710±80
|
Beta-100240
|
|
NA
|
1920±80
|
Beta-100239
|
|
Piyal Burial Site
|
Mornington Island
|
tooth
|
<700
|
NA
|
|
Bayley Point
|
Bayley Point
|
XU1 – 0-17
|
440±70
|
Beta-61791
|
|
6.5-9.7m
|
550±70
|
Beta-28747
|
|
31.4-34.8
|
850±80
|
Beta-37835
|
|
72.4-76.7
|
1100±70
|
Beta-26903
|
|
Gunamula
|
Cliffdale Creek
|
175
|
140±60
|
Beta-12849
|
|
201
|
1300±80
|
Beta-12850
|
|
Old Doomadgee
|
Old Doomadgee
|
Base
|
310±80
|
Beta-28748
|
Field Work Objectives
The research objectives for the first archaeological field season is to
commence establishing regional archaeological and geomorphological datasets for
the southern Gulf of Carpentaria. A flexible
research design will be adopted based on previous studies in the Gulf of
Carpentaria such as the Sir Edward Pellew Group (Bradley 1998; Prebble et
al. 2005), Groote Eylandt (Clarke 2002; Shulmeister and Head 1993; Specht
and McCarthy 2005) and the Torres Strait
(Barham et al. 2004; McNiven 2003; McNiven and Feldman 2003). The notion
behind this approach is to incorporate archaeological patterns of change and
stability in Aboriginal Australian populations (David 1995:362). Incorporating
ethnoarchaeological, ethnohistorical, ethnographic research along with the
well-recorded and documented oral histories (Barham et al. 2004) to
identify site types and old camps along with geomorphological data used to
understand various landform processes, this project
hopes to produce a predictive site location model for the Wellesley Islands.
Projected results for this research include: (1) an increase in the body
of archaeological knowledge available for the area, (2) establishing a
chronology of occupation for the research area, (3) to potentially reconstruct
past socio-cultural dynamics and inter-island relationships, (4) evidence of
Macassan contact (5) providing skills and knowledge to Traditional Owners on
site conservation techniques, (6) develop a multi-layered secured GIS database
to manage datasets such as, (a) various site locations such as fish traps,
fishing, freshwater, wells, hearths, middens, trade routes, tracks etc, (b)
palaeo-geographic data including palaeo-rivers, channels, landforms, (c)
bathymetry data that aid in plotting the effects of rising sea-level on the
Islands, and to extrapolate data on known flooding events to see potential
visualise the full extent of the impact, (7) as discussed the Wellesley Islands
are an ideal region to investigate models of mid-late-Holocene cultural change
and the colonisation of Australia.
Reference
List
Bailey, G. 1999 Shell mounds and coastal archaeology in northern Australia. In H.J. Hall and I.J. McNiven (eds), Australian Coastal Archaeology, pp.105-112. Canberra: ANH Publications.
Barham,
A.J., M.J. Rowland and G. Hitchcock 2004 Torress Strait
bepotaim: An overview of archaeological and ethnoarchaeological
investigations and research. In I.J. McNiven and M. Quinnell (eds), Torres Strait Archaeology and Material Culture,
Memoirs of the Queensland
Museum, Cultural Heritage
Series, pp.1-72. Brisbane.
Beaton, J.M. 1985 Evidence for a coastal occupation time-lag at Princess Charlotte
Bay (North
Queensland) and implications for coastal colonization and
population growth theories for Aboriginal Australia. Archaeology in Oceania
20(1):1-20.
Bird,
M.K. 1992 The impact of tropical cyclones on the
archaeological record: An Australian example. Archaeology in Oceania 27(2):75-85.
Birdsell,
J.B. 1957 Some population problems concerning
Pleistocene man. Cold
Springs Harbor
Symposia on Quantitative Biology 22:47-69.
Bowdler,
S. 1977 The coastal colonisation of Australia. In G. Golson, J. Allen
and R. Jones (eds), Sunda and Sahul: Prehistoric
Studies in Southeast Asia, Melanesia and Australia, pp.205-246. London: Academic Press.
Bradley,
J.J. 1998 'We always look north': Yanyuwa identity and the maritime
environment. In N. Peterson and B. Rigsby (eds), Customary
Marine Tenure in Australia. Oceania Monographs,
pp.125-139. Sydney: Oceania
Publications.
Bureau of Meteorology 1995 Climatic Averages Australia. Canberra: Australian Government Publishing
Services.
Chappell,
J., E.G. Rhodes, B.G. Thom and E. Wallensky 1982 Hydro-isostasy and the
sea-level isobase of 5500 B.P. in north Queensland, Australia. Marine
Geology 49(1-2):81-90.
Chivas,
A.R., A. Garcia, S. Van Der Kaars, M.J.J. Couapel, S. Holt, J.M. Reeves, D.J.
Wheeler, A.D. Switzer, C.V. Murray-Wallace and D. Banerjee 2001 Sea-level and
environmental changes since the last interglacial in the Gulf of Carpentaria,
Australia: An overview. Quaternary International 83-85:19-46.