ISLAND INCARCERATION

Today, the island continent of Australia has more than 8,000 smaller islands off its coast.Footnote ¹ As temperatures rose 6,000 years ago, parts of the mainland flooded and islands separated. These events are remembered by many Indigenous communities through “Dreaming” stories. Some islands became bases for fishing, shellfish gathering, and hunting of larger marine animals; others were no longer reachable, but remained part of Indigenous communities’ cultural landscape.Footnote ² When the British colonizers arrived at Botany Bay in 1788, they enforced European concepts of islands as sites of isolation for the first time. Just three weeks after the arrival of the First Fleet, at the first criminal court, convict Thomas Hill was sentenced to spend a week in chains on a rocky island in Sydney Harbour for the crime of stealing from the government stores. The island was named “Pinchgut Island” after the starvation rations that “pinched” Hill’s stomach.Footnote ³ This was the first instance in what became a system of “punitive relocation” to islands off the coast of Australia for much of the colonial period.Footnote ⁴ Between 1788 and 1901, a network of islands surrounding the Australian continent acted as sites of expulsion, punishment, and labour extraction. The islands of Sydney Harbour were the sites of public works completed by convicts, including Pinchgut Island (Ma-te-wan-ye, 1841), Goat Island (Me-Mel, 1833–1839), and Cockatoo Island (Wa-rea-mah, 1839–1869); further down the eastern coast was St Helena Island (No-goon) in Moreton Bay (1867–1932). Off the northern coast, the large island of Melville Island was a penal settlement and military fortification (1824–1829). Off the eastern coast of Van Diemen’s Land (later renamed Tasmania) there was a penal settlement on Maria Island (1825–1832), which later became a convict probation station (1842–1850), as well as Sarah Island in Macquarie Harbour, which was used for secondary punishment of convicts (1821–1833). On various islands in the Bass Strait, most notably at Flinders Island, Indigenous Tasmanians were confined on an involuntary basis for their “protection” from settler violence (1830–1847); off the Western Australian coast, near Fremantle, Carnac Island (Ngooloormayup) held Nyoongar resistance leaders in 1832, until a long-term penal establishment for Indigenous men was established on neighbouring Rottnest Island (Wadjemup, 1838–1931); finally, and most notoriously, the Pacific Island of Norfolk Island (administered first by New South Wales and then by Van Diemen’s Land) was settled by convicts (1788–1814), and then reoccupied for punishment of re-transported convicts (1825–1853). The map below shows the distribution of these colonial-era carceral islands around Australia.

Figure 1

Australia’s Carceral Islands, 1788–1901, with insets of Sydney Harbour and the Western Australian coast.

This article examines the colonial government’s use of Australia’s offshore islands as sites of “punitive relocation” from the late eighteenth to the late nineteenth century. The term “punitive relocation” is well-suited to describe the inter- and intra-colonial movement of prisoners to offshore islands, as the legal sentences varied. Indigenous people were removed to island penal establishments under custodial sentences, including both “transportation” and “imprisonment with hard labour”. On the other hand, recidivists were sometimes “re-transported” to penal settlements and subjected to hard labour.Footnote ⁵ Though these sentences were legally distinct, physical distance and separation by water were key to both.Footnote ⁶

Through their physical boundedness and separation from the mainland, islands were both practical and symbolic sites to incarcerate those who “threatened” colonial society. They acted as “colonial peripheries”, replicating in microcosm transportation from the metropole to the colony. However, punitive relocation to islands was a colonial system of punishment that was distinct from metropolitan transportation, in purpose as well as scale. In particular, it reflected the need to geographically differentiate general convict society and places of secondary punishment for convicts who reoffended in the colony. Relocation to carceral islands was also part of frontier warfare and territorial acquisition, which violently displaced Indigenous Australians from their lands. This, in turn, resulted in racially distinct forms of island incarceration, despite spatial continuities. Since the Australian colonies relied on free labour, islands were also ideal sites for labour extraction, as their isolation allowed limited mobility for extramural labour and they were also proximate to the sea. The convict industries on carceral islands were often maritime, with convicts logging wood and harvesting hemp to build boats, constructing maritime infrastructure – including jetties, seawalls, lighthouses, and docks – or engaging in activities like fishing, shell collecting, and salt panning. The entanglement of punitive and economic motives was directly tied to the natural geography of these island sites, and the need of colonies to be part of imperial networks of trade and communication.

Carceral islands fulfilled different roles within the colonial project for colonial governance and imperial expansion. These purposes blurred together and changed over time. First, convicts were sent to colonize remote islands and coastal sites, which were politically and commercially strategic. Second, islands were used alongside other geographically remote locations, as sites of particular punishment for those perceived to be the “worst” kind of convict. Third, Indigenous Australians were forcibly confined on island institutions, which were not always explicitly carceral; yet, by displacing Indigenous people to islands (under sentence or not) the government reduced resistance to European conquest, rendering the land one step closer to terra nullius (nobody’s land). The remainder of the article is structured around this typology of Australian carceral islands.

TERRITORY AND TRADE

The purpose of this article is first to explore the use of distant carceral islands as strategic locations along important trading routes. In Australian historiography, there has been a long-standing debate as to whether convicts were sent to New South Wales simply to empty out Britain’s overcrowded gaols after the American War of Independence (1778–1783) closed the American colonies for convicts, or whether convicts were sent to New South Wales to enhance Britain’s naval power in the Pacific arena.Footnote ⁷ However, unlike the decision to settle Botany Bay, the Colonial Office were explicit that they decided to settle Norfolk Island to harvest flax and pine for naval use, as they were when they decided to settle Melville Island with convicts in 1824. Taking an island perspective allows us to look beyond a binary view – that Australia was settled to dump felons or to ensure British naval dominance – to show that spatial differentiation encouraged multifocal policies.

When the Crown issued Captain Phillip’s instructions for settling New Holland, he was told to survey “the several ports, or harbours upon the coast, and the islands contiguous thereto” for possible settlement.Footnote ⁸ As well as ensuring there was no legal loophole that precluded the British from claiming territory in the region, the instructions also directed Phillip to settle Norfolk Island over 1,500 kilometres to the east of Botany Bay in the Pacific. The island was “contiguous” only in the sense that no land masses interrupted this vast stretch of ocean between Sydney and Norfolk Island. Phillip’s instructions claimed Norfolk Island was “a spot which may hereafter become useful”.Footnote ⁹ Its potential utility was two-fold. First, as a strategic site for commercial expansion. Navigating northwards past New Caledonia put vessels on the South Equatorial Current along the tip of Northern Australia and into the heart of the East India Company’s trading grounds in the South East Asian archipelago. The second attraction was the cultivation of flax for ships’ rigging and felling of timber for masts. The loss of the American colonies not only meant losing Britain’s main convict destination, it also depleted Britain’s naval supplies dramatically, and Britain’s access to flax via Russia was threatened through their alliance with France. On the advice of hydrologist Alexander Dalrymple, a mixed group of twenty convict and free settlers, at a ratio of two to one, were sent to settle the island in March 1788.Footnote ¹⁰ Among them was Joseph Lovell, who was sent to Norfolk Island “for life” as punishment for stealing from the stores (his counterpart, Joseph Hall, was, in turn, sent to Pinchgut Island). This demonstrates that even an island that was settled for economic reasons could go on to have a punitive component, reflecting the changing policies of colonial government.Footnote ¹¹

Governor Thomas Brisbane’s decision to settle the northern coast of Australia in the mid-1820s was also explicitly motivated by commercial interests, though this time to tap into the market for trepang (sea cucumber) in China, and as a gateway to further trade with the southeast Asian archipelago. The British had recently relinquished territories seized from the Dutch in the Indonesian archipelago during the Napoleonic wars, so they were eager to expand their commercial reach.Footnote ¹² A secondary motive was to prevent any European power claiming territory on the unsettled edges of the Australian continent.Footnote ¹³ In 1818, Captain Philip Parker King had surveyed the northern coast and reported back with evidence of abandoned Macassan (Sulawesi) camps for smoking trepang.Footnote ¹⁴ On the basis of this report, trader William Barnes wrote to the Secretary of State for the Colonies, Lord Bathurst, in July 1823 with a proposal to establish a British trepang fishery on the Cobourg peninsula. The chairman of the East India Trade Committee, George Larpent, urged Bathurst to approve a British settlement there for “the greatest benefit to the commerce […] of the United Kingdom [...] [and to] place our flourishing possessions in that quarter of the Globe in greater security”.Footnote ¹⁵ Despite Larpent’s advice to settle on the mainland, the Colonial Office issued Captain Barlow instructions to establish a settlement in the “Apsley’s Channel between Melville and Bathurst Island”.Footnote ¹⁶ Looking at a map in London, the islands may have seemed to Lord Bathurst physically closer to trading routes, even though currents, winds, and reefs actually rendered them almost impossible to access.

In 1824, forty-five settlers – only three of them free men – were shipped aboard the HMS Tamar to the northern coast. The convicts were chosen by the Principal Superintendent of Convicts on the basis of their trades, with the majority skilled in construction, and their ethnicity, as thirteen of the eighty convicts selected were black as officials thought they were better able to withstand hard labour in a tropical climate than white convicts.Footnote ¹⁷ Ultimately, the difficulty of navigating the Apsley Strait – which was shallow, rocky, and subject to strong winds during the monsoon season – meant few British trading ships got through to the settlement and no Macassan vessels at all. On deciding to abandon it in 1829, Governor Ralph Darling suggested the convicts be relocated to Croker Island, a few kilometres off the Cobourg Peninsula. Instead, the convicts were transferred to the existing settlement at Port Raffles.Footnote ¹⁸ It seems that colonial governors and imperial administrators had an island bias even when local experts and East India Company officials suggested better-located mainland sites for settlement.

Underpinning these epistolary exchanges was the idea that islands were interchangeable and universally preferable for convict-built commercial hubs. This is underlined by the comparisons made by East India Company officials and colonial newspapers between the “Australian” islands – Norfolk Island and Melville Island – and Indian Ocean island penal colonies – the Straits Settlements. The Straits Settlements were East India Company penal settlements for Indian convicts at Penang, Malacca, and Singapore, and were united in 1826.Footnote ¹⁹ On 10 March 1825, The Australian colonial newspaper hoped that “[w]hat twenty years have accomplished at Penang, at which period it was a barren sand, it is not unreasonable to suppose that half that time will bring to pass at Melville Island”.Footnote ²⁰ In 1827, an East India Company officer (calling himself “M”) suggested in the Asiatic Journal that Melville Island be reopened to replace “its two rivals”, Penang and Singapore, as the destination for Indian convicts.Footnote ²¹ The anonymous officer concluded that Melville Island should not be abandoned, for “the same reasons that Norfolk Island was reoccupied” as a penal settlement in 1825, namely for “its utility to Australia, as a Northern emporium and naval station”.Footnote ²² Though Norfolk Island and Melville Island were administered by New South Wales, they mapped better onto Pacific and Indian Ocean maritime trading routes. These Australian islands were part of a much wider practice of sending convicts as “empire-builders” to islands that were economically and politically strategic for British imperial interests.

SECONDARY PUNISHMENT

The second purpose of transportation to carceral islands was to discipline convicts who misbehaved or reoffended, through the dual mechanism of distance and labour. In 1817, John Thomas Bigge, former deputy-judge advocate of Trinidad, was commissioned by the British parliament to report on the convict system in Van Diemen’s Land and New South Wales. The two key aspects of Bigge’s convict reforms were to disperse convicts across the countryside under assignment to pastoralists in order to rapidly increase the area of land under cultivation, and to introduce a multi-level system of punishment which isolated convicts undergoing secondary punishment, as well as subjecting them to hard labour. Convicts found guilty of misconduct worked either in road or chain gangs or, for more serious offences, were sent to isolated penal settlements.Footnote ²³ Bigge’s scheme was designed to rapidly expand agricultural and pastoral industries, situated in the coastal and interior regions of New South Wales respectively. In order to fulfil the Colonial Office’s instructions to “separate the convict population from the free population”, Bigge “was naturally led to inquire whether any of the islands in Bass Straits, or upon the eastern coast of New South Wales, were calculated for the reception of convicts”.Footnote ²⁴ However, upon receiving information from surveyors and locals, Bigge complained that Norfolk Island had proved too difficult to access by boat and “no other island […] had the same advantages of soil or climate” to sustain a convict population.Footnote ²⁵ Ultimately, Bigge recommended several sites on the coast of New South Wales as possible locations for secondary punishment stations.

However, when the Governor of New South Wales, Thomas Brisbane, and the Secretary of State for the Colonies, Lord Bathurst, came to establish a new penal settlement in 1824, they rejected Bigge’s suggestions and opted instead to settle Norfolk Island in the Pacific. Bathurst believed that, rather than having the “worst description of convicts […] placed in the midst of a thriving and prosperous colony”, Norfolk Island should be occupied “upon the principle of a great Hulk or Penitentiary”.Footnote ²⁶ The penal system that Bigge created relied on distance as the primary mechanism of secondary punishment within the Australian colonies, which translated into officials selecting remote islands. For Norfolk Island to act as an effective deterrent to crime for the convict population, it had to be feared, and a distant island was a powerful image in the minds of the general public. As the Lieutenant-Governor of Van Diemen’s Land, George Arthur, wrote, “being sent to Norfolk Island […] should be considered a place of ultimate limit, and a punishment short of death”.Footnote ²⁷ The fact that Norfolk Island was so distant fed into rumours and myths about the “depravity” of the convicts who were sent there.Footnote ²⁸ Far from being the “worst” convicts, the majority of Norfolk Island’s inmates had been convicted of minor property crimes and a third were serving their original sentence of transportation.Footnote ²⁹ The imaginary of Norfolk Island was so strong in the public mind that insularity became synonymous with isolation in the Australian context, as subsequent prison islands were all understood in relation to their Pacific counterpart.

The other colony that overhauled its convict system along the lines of Bigge’s report was Van Diemen’s Land. Officials here were equally drawn to islands as sites of secondary punishment. At the centre of Macquarie Harbour, a body of water twice as big as Sydney Harbour, was an archipelago of carceral islands. The main settlement, with shipyard, was the vast Sarah Island (also known as Settlement Island), which stretched from the pilot station to the shores of Macquarie Harbour.Footnote ³⁰ Next door was the “detached fort” of Grummet Island (or Small Island), which housed a hospital and penitentiary.Footnote ³¹ In 1826, Lieutenant-Governor George Arthur wrote to the Colonial Office recommending its closure because of the encroachment of free settlers and the high rates of escape.Footnote ³² Between 1821 and 1832, there were 150 escape attempts involving 271 individuals, or nearly one in four of those who had been convicted of a second crime after being transported to the colony.Footnote ³³ Lieutenant-Governor George Arthur’s language mimicked Bigge’s when he stressed that “as the Colony becomes more and more populated, the barrier between these wretched Criminals and the rest of the Community will be decreased, and escape will constantly become more easy”.Footnote ³⁴ Even if convicts were kept on islands overnight, they worked on the mainland, which presented an opportunity for escape. On similar grounds, Arthur criticised the penal settlement on Maria Island, which had been for the punishment of less “serious” secondary offenders a year earlier, in 1825. Situated just four kilometres east of the Tasmanian mainland, Arthur complained that “it is much too near the settled districts on the Main Land to be regarded as a safe depot for very desperate offenders”.Footnote ³⁵

For this reason, Arthur suggested King Island, to the west of the Bass Strait, as a suitable alternative, from which escape would be almost impossible. However, Arthur noted that its warm climate and natural beauty made it more akin to a paradise, than a penitentiary, rendering it in some respects undesirable as a place of punishment. In 1827, Arthur once again put forward a new island penal settlement on Phillip Island – situated off the southern coast of Australian, near modern-day Melbourne. However, Phillip Island was far from a utopia: its dry soil and swampy interior made it economically unviable for convicts to cultivate the land, though Arthur believed it could still be a “viable temporary penal establishment”.Footnote ³⁶ In the same year, Arthur formed an executive committee on the problem of educated convicts, suggesting that they should be segregated from the corrupting influence of the general convict population. Arthur seemed certain that “an island may be found much more convenient and available than any district” to keep educated convicts separate from the rabble.Footnote ³⁷ Similarly, the Colonial Treasurer, Jocelyn Thomas, claimed that “the various islands in the Bass Strait (King, Furneaux, Cape Barren etc. etc.) all afford eligible situations for Penal Settlements”.Footnote ³⁸ Many of these islands were later used for the confinement of Indigenous Australians (as will be discussed in the third section). This demonstrates the enduring appeal of islands as “natural prisons”, though officials used different arguments to explain why a certain population was best suited for confinement there. Islands offered the possibility to protect society from “dangerous” convicts, but the isolation could also protect “gentlemen” convicts from corruption from a society made up of “ex-cons”.

In the mid-1830s, policymakers turned away from remote islands to urban islands located in the midst of city harbours as sites that balanced surveillance, security, and labour needs. In the preceding decade, the extraction of convict labour in penal settlements had become increasingly more important than punishment through “internal relocation to the peripheries of New South Wales”.Footnote ³⁹ From the mid-1830s, islands in Sydney Harbour – including Goat Island, Cockatoo Island, and Pinchgut Island – were used as sites of secondary punishment through hard labour. From 1833 to 1839, convicts on Goat Island quarried a gunpowder magazine, soldiers’ barracks, and a wharf to fortify the harbour.Footnote ⁴⁰ Between 1840 and 1841, convicts levelled the top of the island to build a military fortification on the colony’s first prison, Pinchgut Island (now Fort Denison).Footnote ⁴¹ On the largest island in the harbour, Cockatoo, convicts spent over a decade from 1847 quarrying a dry dock directly into the sandstone base of the island, and then built and manned the workshops to repair and outfit ships until 1869.Footnote ⁴² Convicts were sometimes sent to the islands under sentence by magistrates (with powers awarded under the 1830 Offenders’ Punishment and Transportation Act), but more often they were simply transferred from a road gang to an island gang. The Principal Superintendent of Convicts would send convicts deemed dangerous or likely to escape to these islands, which were perceived as sites of increased security despite their proximity to Sydney. In December 1840, for example, a convict found guilty of sexual assault of an Indigenous woman, two convicts suspected of bushranging, and nine convicts who had been re-transported from South Australia were sent to Goat Island (the latter awaiting transfer to Norfolk Island).Footnote ⁴³ When John Carroll committed burglary the convicting magistrate recommended that he be punished “at a distance from Sydney, in consequence of […] [his] desperate character”.Footnote ⁴⁴ With this in mind, Governor George Gipps instructed that he be “sent either to Cockatoo or Pinchgut Island”, rather than mainland stockades that were several hundred kilometres distant from the capital. Clearly, officials viewed the islands of Sydney Harbour as both extra-punitive sites and locales for extra-mural convict labour.

In 1837, the British parliament commissioned a Select Committee on Transportation, which was chaired by Sir Henry Molesworth and comprised anti-slavery abolitionists and evangelicals. Based on testimony by a carefully selected set of anti-transportation witnesses, the committee concluded that the Australian convict system was characterized by excessive violence (flogging and chaining) and many forms of vice (including rape, sodomy, and child molestation).Footnote ⁴⁵ When it became clear that convict transportation to New South Wales would likely cease, the former Secretary of State Viscount Howick issued a memorandum with a list of possible destinations for British and Irish convicts – all of them islands. He rejected the Ionian Islands off the coast of Greece, St Helena in the Atlantic, and the Falkland Islands off the coast of Argentina before settling on Norfolk Island as the best site.Footnote ⁴⁶ This indicates there was a wider British imperial consensus about islands’ suitability as penal colonies, and islands were also favoured as penal colonies in other empires (as is reflected in the essays that comprise this special issue).

The Colonial Office, eager to reform convict discipline along rehabilitative lines, offered Captain Alexander Maconochie the command of Norfolk Island to trial his system of penal reform on newly arrived convict transportees. Maconochie’s “mark system” incentivized convicts to work hard and behave well by allowing them to earn time off their sentence through good conduct and labour.Footnote ⁴⁷ However, Maconochie contested the choice of island, complaining that it was “too remote”, “inaccessible”, and “tropical” for profitable industry. In Maconochie’s view, successful rehabilitation relied on prison labour reflecting real world economies as much as possible, making proximity to urban settlement more desirable. Instead, Maconochie suggested dividing the convicts – according to their behaviour – between two peninsulas of Van Diemen’s Land and Maria Island off the eastern coast. He also put forward a similar spatial configuration of punishment whereby King Island in the Bass Strait would house the majority of convicts working in agriculture, and recalcitrant convicts would be sent to two small islands (New Year Island and Christmas Island), which would act as “penitentiaries for separate imprisonment […] with little expense of masonry”.Footnote ⁴⁸ Maconochie was so confident that his scheme could incentivize good behaviour that he was eager to trial it on mainland road gangs, but Governor Gipps knew there would be public uproar if the scheme were trialled within the vicinity of free settlers: for Maconochie’s “experiment” an island laboratory was needed.Footnote ⁴⁹ In his letter to the Colonial Office in 1840, Gipps commented that all the natural geographical features that made Norfolk Island a good carceral island were the features Maconochie complained about: “namely, its remote situation, its insular character, its limited extent”.Footnote ⁵⁰

It was incumbent upon Gipps to find a new penal settlement for secondarily transported convicts who needed to be removed from Norfolk Island. However, since convict transportation to New South Wales had ceased, Gipps could no longer transport convicts to penal settlements within the colony, leading him to pass legislation to remove convicts from penal settlements to any “site of hard labour”.Footnote ⁵¹ Thus, relocation to islands continued, under a different legal sentence, as a regional practice after transportation between the metropole and colony had ceased. In February 1840, Gipps proposed that either Tasman’s Peninsula or King Island in the Bass Strait replace Norfolk Island as “a new penal colony”.Footnote ⁵² However, Governor John Franklin refused to accept secondarily transported convicts within the limits of Van Diemen’s Land. Franklin, for his part, proposed Auckland Island, off the coast of New Zealand. In 1841, Lord Russell suggested Goat Island in Sydney Harbour, but Governor Gipps adapted his instructions to send convicts to another harbour island, Cockatoo Island, because it was not safe to send convicts to a “place already occupied by a magazine of gunpowder”.Footnote ⁵³ Despite being separated from Sydney’s shore by just a few kilometres, Gipps insisted it was “the place of greatest security within the colony, not actually a prison”.Footnote ⁵⁴ Indeed, Gipps asserted that proximity was preferable to isolation when it came to secondary punishment, claiming that “stations for doubly convicted men, seem to me to have been erroneously placed at great distances from the seat of Government […] [so they] have rarely, if ever, been visited by the Governor of the Colony, or by any person high in authority”.Footnote ⁵⁵ Cockatoo Island, in the midst of Sydney Harbour, was both secure and easy to survey; or, as Gipps put it, Cockatoo Island was “surrounded […] by deep water, and yet under the very eye of authority”.Footnote ⁵⁶

Over the next five years, secondarily transported convicts were transferred from Norfolk Island to Cockatoo Island under a scheme that more than halved the terms of their remaining sentences. They were joined by the Superintendent of Agriculture, Charles Ormsby, who became Superintendent of Cockatoo Island from 1841.Footnote ⁵⁷ The movement of both the Superintendent and a large body of convicts from one to the other led convict James Laurence to remark that Cockatoo Island was the same as Norfolk Island in every respect, except for the fact that Cockatoo was a “small island”.Footnote ⁵⁸ This marks a decisive shift away from isolation as punishment, which was replaced with hard labour for the public benefit, but with the added value of the security offered by water and walls to keep the felons in. Despite Cockatoo Island’s proximity to Sydney, Godfrey Charles Munday described it as a “natural hulk”, using the same descriptor as Lord Bathurst had for Norfolk Island.Footnote ⁵⁹ Long after the majority of secondarily transported convicts had left, and Cockatoo Island effectively operated as a local gaol, it retained its associations with the convict system through its Pacific predecessor. In an 1857 inquiry, Cockatoo Island was dubbed a “worse hell-on-earth even than Norfolk Island”, and Henry Parkes claimed that the superintendent “Mr. Ormsby is so isolated, as much indeed as if he were a thousand miles off in the Pacific”.Footnote ⁶⁰ Despite their clearly opposite geographies in relation to the mainland – the former just one-and-a half kilometres and the other 1,500 kilometres away from Sydney – they were considered comparable due to their insularity. Though from the mid-1830s a clear shift had taken place in favour of proximate urban islands, in the public mind islands were by definition “isolated” – an idea dating back to Robinson Crusoe’s “desert island” (1719), which was further entrenched in the Australian colonies through Norfolk Island’s mythology. This led officials to believe that Australian islands were a better deterrent and were more suitable for the “worst” offenders.

CONFINEMENT OF INDIGENOUS AUSTRALIANS

The third purpose that the colonial government used carceral islands for was to confine Indigenous people. In the 1830s, the colonial government established “Aboriginal Settlements” on a series of islands off the coast of Van Diemen’s Land. During the Black War, Lieutenant-Governor George Augustus Robinson convinced Indigenous Tasmanians fleeing from settler violence to voluntarily go to islands for their own “protection”. These temporary measures became permanent establishments that Indigenous inhabitants were not allowed to leave, and where they were subjected to poor living conditions, restrictive routines, and punishment. Thus, they operated like carceral institutions, despite the evasive language of the archive. According to N.J.B. Plomley, for the colonial administration it was always a question of which “island [was] suitable for aboriginal settlement”.Footnote ⁶¹ The “Aborigines Committee” was charged with finding the best site for the reserve and considered Maria Island, King Island, Bruny Island, and the Hunter Islands in the Bass Strait as possible locations.Footnote ⁶² The committee were looking for an island large enough for the Indigenous inhabitants to roam freely and with an abundance of game for them to hunt.Footnote ⁶³ This showed the government’s intention for Indigenous people to live as “hunter-gatherers”, though they ignored both the ecological and cultural connection between the many different communities represented on the island and their particular homelands. This was only partially recognized by the committee’s fear that if the island were in sight of the mainland then the Indigenous Tasmanians would “pine away”, meaning that homesickness would cause their health to deteriorate. The coercive nature of these island reserves is clear as the committee repeatedly insists that the island could not be too close to the mainland, otherwise the Indigenous Tasmanians would swim across and escape. In 1831, the committee noted that a benefit of choosing Maria Island, formerly a penal establishment for European convicts, would be re-using the prisoners’ barracks and having the police crew on Lacklan’s Island sweep the water for escapees.Footnote ⁶⁴ The overlapping of carceral spaces for European convicts undergoing punishment and Indigenous Australians under government “protection” suggests how malleable and persistent islands were as sites of incarceration and coercion, though ideas of “race” shaped how these policies were presented and understood.

It was George Augustus Robinson who actually surveyed these islands for their suitability as a settlement. After convincing the first party of Indigenous Tasmanians to join him on Swan Island in November 1830, they were transferred to different islands – including to Clarke Island and Preservation Island – as he inspected them before settling on Gun Carriage Island in May 1831.Footnote ⁶⁵ A lack of fresh water and poor access for ships led to the abandonment of Gun Carriage Island and a move to Flinders Island in 1833 due to its good anchorage, warm weather, abundant game, and access to fresh water.Footnote ⁶⁶ In reality, extremely poor conditions prevailed on the island, leading to the death of half of the Indigenous population of the island due to neglect, malnourishment, trauma, and disease.Footnote ⁶⁷ These conditions were resisted by the community on the island. In March 1847, eight inhabitants petitioned Queen Victoria, complaining about being treated as prisoners on Flinders. They wrote that they “freely gave up our country to Colonal [sic] Arthur […] after defending ourself” and that they were “a quiet and free people and not put in gaol”.Footnote ⁶⁸

These failures were explained away by Robinson in his 1837 report to the Colonial Office through the idea that Indigenous people were “weak” and would inevitably become extinct after their encounter with the superior white race. On these islands, they were at least not victim to settler violence, and they were “civilized” by being taught both farming techniques and Christian principles. The Colonial Office readily accepted this fiction because Robinson’s island settlements seemed to align with the 1835–1837 British Parliamentary Select Committee on Aboriginal Tribes, which recommended appointing “protectors” to prevent settler violence and encourage “civilization” through Christian teachings.Footnote ⁶⁹ Islands epitomized the Colonial Office’s justification for imperial conquest of Indigenous lands through “humanitarian” governance. Sending Indigenous people to islands effectively cleared the way for European settlement, but without visibly imprisoning them, offering instead limited mobility in a natural environment.

In Western Australia, Governor Hutt pursued a policy based on the Committee’s suggestions by appointing Protectors of Aborigines who would administer British law on behalf of, and more often against, Indigenous peoples to “protect” them from settler violence.Footnote ⁷⁰ In this view, punishment was actually protection. So even when the failure of Flinders Island was known, the Colonial Office still “aggressively pursued a policy […] that attempted to replicate Robinson’s fantasy island”, including in the newly settled colony of Western Australia.Footnote ⁷¹ As early as 1830, barrister and Western Australian colonist George Fletcher Moore said he feared violent conflict unless members of the Indigenous Nyoongar community were “removed wholesale to some island”.Footnote ⁷² In 1832, Carnac Island, off the coast of Fremantle, was used to detain a group of Nyoongar resistance leaders, including Yagan and Midigoroo. They were treated as prisoners of war and had their capital sentences commuted to confinement on Carnac Island at the recommendation of surveyor John Septimus Roe. After just a month, the prisoners escaped to the mainland on an unattended dinghy and were eventually shot and killed by the authorities.Footnote ⁷³ In July 1838, the government established a permanent and “humanitarian” prison for Indigenous men on neighbouring Rottnest Island. It believed that the eighteen kilometres that separated the island from the mainland made escape so difficult for the Indigenous convicts that they could be worked without chains and be allowed to hunt and roam regularly without compromising security. This was necessary because, as was stated in the 1840 Act to Constitute Rottnest a Legal Prison, “the close confinement of a gaol […] [had] been found to operate most prejudicially to their health”.Footnote ⁷⁴ In the early years, the policy pursued on the island reflected those on missions, as prisoners were taught agriculture and allowed to roam and hunt on the island on Sundays.Footnote ⁷⁵

Yet, underlying these official humanitarian reasons was deterrence: as Rottnest was “winnaitch” (or forbidden) for Nyoongar Whadiuk as a realm for bad spirits.Footnote ⁷⁶ Thus, the colonial administration argued that transportation to Rottnest elicited a particular kind of dread that could not be replicated by local imprisonment or even capital punishment.Footnote ⁷⁷ In the Tasmanian context, the island was seen by the Colonial Office as a sliver of land to replace what had been conquered. On Rottnest Island, in contrast, the cultural meaning of the island was used as a deterrence. As late as 1884, a Nyoongar prisoner named Bob Thomas told a commission that “Natives do not like the sea voyage […] Rottnest is dreaded by the natives”.Footnote ⁷⁸

In 1847, George Augustus Robinson described Rottnest Island in a way that showed clear parallels with its predecessor Flinders Island, though he made no explicit comparison.

This shows that the colonial administration was intent on presenting islands as “boons” and “blessings” to the Indigenous populations who were (in Robinson’s own words) “banished” from their country. This encapsulates the ambiguity of colonial governance that justified territorial acquisition and economic gain through their presumed superiority. Studying islands is an important part of recognizing the spatial trajectories of the criminal justice system as applied to Indigenous Australians. In particular, the political and social imperative to eliminate Indigenous communities – conceptually, physically, or politically – in order to clear “space” for colonizers. Since the majority of prisoners were serving sentences for theft (mostly of livestock), and were often prosecuted as a group, transportation to Rottnest effectively dispossessed Indigenous communities, just as the Tasmanian reserves had.Footnote ⁸⁰ A key difference between the two was that no women were incarcerated on Rottnest, though by removing so many men it still effectively disrupted Indigenous communities and weakened resistance to European conquest of “country”. The colonial government briefly considered a scheme for incarcerating Indigenous women, making a deal with James Reid on Garden Island to confine short-sentenced Indigenous women at a cost to the Treasury of nine pence per person per day.Footnote ⁸¹ No more mention of this scheme appears in the Colonial Secretary’s correspondence and it did not become an institutionalized practice. Thus, we see Flinders Island was the model for a constellation of island sites in Western Australia, which by virtue of being “natural prison hulks” satisfied a contradictory logic of “punishment” and “protection” in the colonial context.