by Patrick Nunn
Our next stop along the Australian coast is on the shores of its largest indentation, the Gulf of Carpentaria. During the last ice age, when the sea level was more than 120m (400ft) below today’s level, this part of Australia was connected to New Guinea to the north. At the centre of the land bridge lay a great freshwater lake – Lake Carpentaria – around which the people and animals of the area flocked.50 The rising sea level after the ice age ended saw the progressive inundation of this land bridge, the eventual conversion of the freshwater lake to the saltwater gulf, and the appearance of islands within the Torres Strait that today separates Australia from Papua New Guinea.51 The progressive submergence of this well-populated land bridge would of course have driven many groups of people outwards to seek new land from which they might subsist.
The languages spoken by Aboriginal groups in most parts of Australia around the time of its European colonisation all belong to the same language family (Pama-Nyungan), which is thought to have spread across most parts of the continent some time within the past 3,000 years or more. What is key for our purposes is that this dispersal appears to have originated somewhere around the now-submerged land bridge between Papua New Guinea and Australia, and it is plausible to suppose that its outward spread from here was driven by sea-level rise. Some corroboration of the precise and quantifiable linguistic argument may be found in Aboriginal stories about ‘lost lands’, now-submerged homelands from which ancestral mainland populations came. The geographical details about where these homelands were located are typically vague, but one that is told by the Yolngu people of eastern Arnhem Land, who occupy part of the western shores of the Gulf of Carpentaria, states that the ancestral land named Bralgu was located ‘somewhere in the Gulf of Carpentaria’, perhaps therefore a direct memory of the translocation of people from part of the submerging land bridge.52
Our focus is on islands lying off the coast within the modern Gulf of Carpentaria, principally the Wellesley Islands (Figure 3.6), which would have formed only after the sea level rose across the former mainland fringe. The earliest written version of the oral tradition was produced by Dick Roughsey of the Lardil people in the early 1970s, who explained that:
In the beginning, our home islands, now called the North Wellesleys, were not islands at all, but part of a peninsula running out from the mainland. Geologists … thought that the peninsula might have been divided into islands by a big flood which took place about 12,000 years ago. But our people say that the channels were caused by Garnguur, a sea-gull woman who dragged a big walpa or raft, back and forth across the peninsula. 53
The Lardil also remember that the Balumbanda, one of their clan groups, originally came from the west along that peninsula before it was ‘cut up into islands’.
Figure 3.6 The Wellesley Islands, southern Gulf of Carpentaria.
Later research focused on Lardil stories dated to more than 5,000 years ago, about how channels formed that isolated the Wellesley Islands from the mainland.54 A fuller version of the Garnguur (kankurr) story told by Roughsey explains that the seagull woman sought to punish her brother Crane for failing to look after her child, so she dragged her large raft back and forth across the land that connected Forsyth Island to Francis Island until the deep-water channel (more than 10m/33ft deep) that exists there today was formed. The sweet-potato woman, Puri, who was fleeing from Crane, is said to have created the passage (named purikal) between Forsyth and Andrew Islands. The narrow channel between Denham and Mornington Islands is said to have been created by a shark panicked by lightning.
The sheer number of channel-cutting stories here suggests that the gradual submergence of the land connection between the Wellesley Islands and the mainland was witnessed by people who found the process so disruptive, so literally life-changing, that it became a conspicuous element of their oral traditions. Yet for a land connection to be re-established, it would only be necessary for the sea level to be 5m (16ft) lower, although a 10m (33ft) fall would see the peninsula reappear. Exactly how sea-level changes affected human settlement of these islands – something that is key to understanding how old the stories might be – is as yet uncertain. While it is likely that people occupied the area during the last ice age when the ‘islands’ were contiguous with the mainland, we cannot be sure whether people stayed on the embryonic islands as the rising sea level was starting to separate them from the mainland – which is what some channel-cutting stories suggest – or whether they retreated to the relative security of mainland shores at this time. There is evidence favouring both views.55
Moving west from the Gulf of Carpentaria we come to the north-facing coast of Australia, much of which is called Arnhem Land. Perhaps because it is further north than almost every other part of the continent, it has a wealth of Aboriginal traditions and stories that are likely to be some of the earliest in Australia. The four groups of drowning stories in this area all involve offshore islands – Elcho, the Goulburn group, Cape Don and the Tiwi (Bathurst and Melville) Islands. It is perhaps no coincidence that such stories are found predominantly along the coast of Australia where the continental shelf exposed during the last ice age was widest – and where, consequently, the subsequent rise of the sea level was most noticeable. One authority estimated that during this period ‘on the gently sloping northern plains [of Australia] the sea inundated five kilometres of land annually’.56 Not only is this kind of change blatant, but its effects on the way people lived, what they could eat and with whom they had to compete for food and territory undoubtedly left enduring marks on the contemporary evolution of Aboriginal societies in this part of Australia.
Elcho Island is separated from the mainland today by a narrow strait that would be passable, albeit circuitously, were the sea level 5m (16ft) lower than it is today, and fully emergent if the sea level were 10m (33ft) lower. Two extant stories about such a time are known, one about an Elcho resident named Djankawu who tripped while walking along the beach, accidentally thrusting his walking stick into the sand and ‘causing the sea to rush in’.57 The other tells of the Ancestresses, magical beings who, whenever they needed to cross to Elcho, made a sandbar that vanished once they had completed their crossing.58
Further west off the Arnhem Land coast lie the Goulburn Islands, comprising South Goulburn (Warruwi) and North Goulburn (Weyra). One story, which is perhaps an echo of the time when Warruwi was joined to the adjacent mainland, concerns a man named Gundamen who wanted to cross from the latter to the island. Afraid of the deep water, he was helped by a woman with magic powers who called out ‘mubin, mubin, murbin!’, which caused a land bridge to emerge just long enough for him to cross.59 Another story is more detailed and recalls a time when only a small creek named Mandurl-mandurl separated Warruwi from Weyra. The creek teemed with fish, and the local people would set their nets (yalawoi) in it overnight and pull them up full in the mornings. For all this, a man named Gurragag rarely succeeded in catching any fish in Mandurl-mandurl, so one day he cut down a huge paperbark tree (waral). It fell into the creek with such an enormous splash that seawater poured in, creating the ocean gap between the two islands that is today some 7km (4½ miles) wide.60 If such stories recall a time when these islands were in fact closer to the mainland and to each other, then the sea level would have to have been 17–20m (56–65ft) lower than it is today.
The people of coastal Arnhem Land ‘possess names for and maintain intimate knowledge of places far out at sea that are known to have been above sea level 10,000 years ago’, an extraordinary yet undeniable fact that has become the basis for a large-scale marine-protection strategy for this part of the Arafura Sea.61 For example, the inhabitants of the Goulburn Islands know of two submerged islands – Lingardji and Wulurunbu – thought to be the sites of shoals about 20km (12 miles) east of Weyra that now lie in waters 9–11m (30–36ft) deep. How could these people have such intimate knowledge of a submerged landscape unless their ancestors had once walked across it, witnessed its submergence and consigned their memories of it to
posterity?
Moving west along the coast of Arnhem Land, we come to Cape Don, the western extremity of the Cobourg Peninsula, which has been occupied by the Arrarrkbi peoples for tens of thousands of years. They have stories about how one group of their ancestors once occupied an offshore island named Aragaládi that is now underwater. One story tells that a sacred rock (maar) on the island was accidentally bumped one day, causing so much rain to fall that the island became submerged: ‘children and women were swimming about … [they] had no canoe to enable them to cross over in this direction to the mainland at Djamalingi (Cape Don) … trees and ground, creatures, kangaroos, they all drowned when the sea covered them’.62 This story perhaps combines memories of large waves like tsunamis and their effects on coastal peoples in this area with the story of an island being submerged, a common situation with similar oral traditions in the island groups of the Western Pacific Ocean.63
There is another story about the disappearance of Aragaládi that is more entwined with other oral traditions of Arnhem Land Aboriginal peoples, particularly that of the Rainbow Serpent. For this story has Aragaládi as part of a giant snake, one of its coils conveniently (and intentionally) raised above the ocean for people to live upon. But when its inhabitants knocked the maar, the snake drew its coil underwater, swallowing all the people: ‘she made the place deep with sea water. Those first people became rocks. Nobody goes to Aragaládi now.’64
The Mayali people, who now live inland east of the Alligator River (see Figure 3.7), also have a tradition that the first people in this part of Australia – the Nayuyungi – came from the sea between the Cobourg Peninsula and the Goulburn Islands. This is also possibly a distant memory of an inhabited island that was submerged by postglacial sea-level rise, compelling its inhabitants to relocate to the mainland.
The key point here is that along the Arnhem Land coast, likely to be the longest continuously inhabited part of Australia, as well as the one with the broadest continental shelf to be drowned by postglacial sea-level rise, there are numerous extant stories (and probably many more forever lost) that tell of islands off the coast that were once inhabited but are now submerged. This is just what would be expected in such a geographical situation that experienced rapid sea-level rise over several millennia, yet it is impossible to know the minimum depth that the sea level would have to have been in order for these islands to have been emergent. If we assume, without any real evidence, that Aragaládi Island was once the shoal some 150km (93 miles) east of the Cobourg Peninsula, then the sea level would have had to be 9–15m (30–50ft) lower than it is today for habitable land to be exposed there.65
The next stop on our tour of the Australian coast is a group of sizeable offshore islands – the Tiwi Islands – where traditions tell of the apparent isolation of their populations as the sea level gradually rose. Better known as Bathurst and Melville Islands (Figure 3.7), they lie a minimum of 25km (16 miles) off the Australian mainland near the northern city of Darwin. It is to the water gap known today as Clarence Strait that Tiwi stories about the apparent cutting off of these islands from the mainland refer.
Figure 3.7 Location of Bathurst and Melville Islands (the Tiwi Islands), Cape Don and the Cobourg Peninsula.
The stories begin in the Tiwi ‘time of darkness’, before Bathurst and Melville were ‘born’ as islands, a time when the land ‘contained no geographical features, animals or humans’.66 Then, from within the earth – meaning from what may have been becoming the mainland (rather than the islands) – came an old blind woman named Mudangkala (or Murtankala), who crawled with her three children along what is now Clarence Strait until they reached Melville Island. Critically, water followed the group as it moved northwards: ‘the flow of water continued to increase and is today known as Clarence Strait … [Mudangkala] continued to move over the land known as Bathurst Island till finally water flowed on to form what is now known as Apsley Strait’.67 These stories may be memories of one of the last significant human migrations from the Australian mainland to the Tiwi Islands, the references to blindness perhaps signifying the uncertain nature of Mudangkala’s destination, and those to crawling indicating the difficulty of crossing a partly submerged land bridge.
For anyone in the past attempting to travel from the mainland to the Tiwi Islands without the use of watercraft, the Vernon Islands (Potinga to the Tiwi people) would have been key ‘stepping stones’. One Tiwi story recalls that the Vernon Islands were once attached to the southern part of Melville Island at a place called Mandiupi, but then, ‘as a result of an earthquake in the distant past’, the connection was broken.68 Such a story illustrates the issue of the plausibility of oral traditions well. In order to expect a tradition – like the instinctively improbable ‘breaking-off’ of islands – to be duly repeated through future generations, it may need some explanation. Not associating the story with a sea-level rise, perhaps the storyteller added the detail about an earthquake simply to make it credible. For an earthquake could not possibly have caused the Vernon Islands to break off from Melville, but a rising sea level would have swamped the connection, giving the same apparent result.
Other aspects of Tiwi society, from its language to its uncommon matrilineal society, make it distinct from mainland Aboriginal societies, something that is ascribable to physical and cultural isolation imposed by postglacial sea-level rise.69 In addition, there are various aspects of Tiwi Island terrestrial ecosystems that also point to the effects of isolation.70 So for how many years have the islands been isolated, and how long is it since Mudangkala made her last desperate crossing of Clarence Strait? We have to wait until the end of Chapter 4 for a precise answer to these questions, but for now it seems that isolation would have occurred when the ocean around the Tiwi Islands was at least 12m (40ft) lower than it is today, a time in the past when – through a combination of walking, wading and short swims – it would probably have been possible to reach them from the mainland. When the sea level was 20m (65ft) lower, it is likely that someone could have crossed without getting their feet too wet, although the likely route would have been tiresomely circuitous.71
Moving further west into north-west Australia, it is surprising, given this region’s demonstrably long history of Aboriginal occupation and its extraordinary legacy in rock art (throughout the Kimberley and Pilbara regions in particular), that there are not more stories about coastal drowning here than there are. In fact, there is just the one group of stories – about Brue Reef (discussed below) – but there are also two intriguing studies that demonstrate the massive impact which postglacial sea-level rise had on the Aboriginal inhabitants of this region.
The first of these studies concerns the baobab tree (Adansonia gregorii), which is native to north-west Australia as well as parts of Africa. In Australia it is usually called the bottle tree, and for thousands of years it has been prized by savannah-dwelling Aboriginal groups for food, medicine and shelter from the sweltering conditions in which it thrives, typically from the coast to the desert fringes. Parallel lines of research into baobab gene flows and into the routes along which the various words for baobab among Aboriginal groups passed allow it to be plausibly shown that these trees, once confined to the (now-submerged) continental shelf of north-west Australia, were probably carried inland and intentionally dispersed as the sea level rose in this area.72 The clever idea behind this scenario is that the oldest (genetically less diverse) types of baobab are found along the coast, where the people also have the fewest words for the tree (and its component parts). Further inland there is greater genetic baobab diversity and a far greater range of names for the tree. These two observations suggest that people living with baobabs along the now-drowned continental shelf, recognising their great value, carried their fruit pods with them as they were gradually forced inland by the rising waters. In addition to deliberately planting the baobab inland, they also introduced it to inland Aboriginal groups that, quickly convinced of its value, dispersed it even more widely.
The second study used Aboriginal rock art to plot the course of postglacial sea-level rise in the Dampier Archipelago (Murujuga). The earliest rock art here dates from the time during the last ice age when the sea level was so low that today’s islands were part of a mountain range some 160km (100 miles) inland from the coast. With the rise of the postglacial sea level, the area was gradually transformed – from inland to coastal to offshore islands – but the people stayed put, adapting their livelihoods to the changes that the sea level (and climate) imposed on them. They recorded these changes in their rock art that is fortuitously etched through a veneer of rock varnish into the weathered surfaces of the gabbro and granophyre boulders that are scattered about the Murujuga landscape in vast numbers. They look like scree but have in fact weathered out of the underlying bedrock in the places where they are found now, unmoved for perhaps millions of years. Thus, the rock art dating from when the area was far inland shows examples of inland fauna like macropods (kangaroos and wallabies), but in the rock art from the time the shoreline had reached the area, engravings of fish and other sea creatures are dominant.73
Stories about Brue Reef are well known among the Bardi and Jawi peoples of the Kimberley coast.74 Lying about 50km (30 miles) off the mainland and uncommonly isolated, Brue Reef is said once to have been a habitable island named Juljinabur; today it is awash at high tide. The gist of the various stories is that Juljinabur was inhabited by a person called Jul and his greedy kinfolk (munjanggid), who had become cannibals. One day, in fruitless pursuit of a turtle, a Jawi family from Tallon Island (Jalan), 100km (60 miles) to the south, found themselves drifting in a swift current (lu) towards Juljinabur. To save the Jawi family from being eaten, Jul hid the people on the island for a few days, and when his kinsfolk’s attention was diverted, sent them back to Jalan in a double-hulled canoe (inbargunu). Unfortunately, the munjanggid got wind of what had happened, and there was a big fight that precipitated the sinking of the island and its conversion to the inter-tidal reef it is today.
