A silver lining: How Covid forced me to delve deeper into the archives

Saphia Fleury

PhD Student, Falling Through the Net Cluster

Wilberforce Institute, University of Hull


Saphia Fleury reflects on the changes that had to be made to her research methodology as a result of Covid-19 and the questions this raised.

The ink was barely dry on my ethics approval form when the first Covid lockdown was announced. Overnight, my research plan – which involved extensive travel and face-to-face interviews with migrants and practitioners – became about as feasible as a voyage to the Moon. The focus of my investigation was the experiences of children during and after the volcanic eruptions on Montserrat. There was so little information about this period in the archives that conducting interviews with former evacuees had appeared to be the only possible methodology.

Covid forced me to turn the problem on its head and ask why the archival information was so scant. Montserrat is a British Overseas Territory, and its volcanic catastrophe took place less than 30 years ago. Around half of the island’s population had migrated to the UK in the later 1990s – up to 5,000 people. Why, then, did British repositories hold so little evidence as to the nature and impacts of the disaster? I am not a historian by training, but my background is in investigative human rights research. I approached the problem as a detective would, casting the net wide to gather all the extant evidence of what happened on the island during the volcanic crisis. This involved extensive scanning of diverse sources including Hansard, newspaper archives, scientific records, government evaluations and Select Committee hearings, MPs’ correspondence, orders of service for memorial events, children’s poetry, and even self-published autobiographical accounts of the disaster.

I became suspicious when my efforts were frequently thwarted. Key records were missing, or responses to Freedom of Information requests suggested they had never existed to begin with. Even lists of the names of the dead were inconsistent; sources could not agree on who had perished in the disaster. It struck me that this confusion over the identities of the deceased would be unthinkable had the crisis taken place on the British mainland. Following the Kings Cross fire in London in 1987, the British Transport Police spent 16 years conducting painstaking investigations to uncover the identity of the 31st victim. By contrast, the lack of a definitive record of events surrounding the deaths on Montserrat symbolised, for me, a lack of executive interest in the victims of the catastrophe, both living and dead.

This feeling was borne out by my unsuccessful requests to government ministries for official statistics on the numbers evacuated. A former UK Government employee involved in the emergency response on Montserrat helped me to explain the gaps in the statistics, stating that s/he had personally recorded the change in population using information from landing and departure forms:

“From this data it was possible to track the downward movement of the resident population and also small upward trends when Montserratians returned in order to enlist on the government Help schemes for migration to other islands or UK. Nobody requested this information and indeed no-one acknowledged it either. It was distributed to the Governor’s Office, Chief Minister, DFID [UK Department for International Development] and [my] immediate boss in London”.  (Anonymous interviewee)

In spite of their efforts, the information recorded by this individual did not appear in any of the official documents I uncovered. I was forced to conclude that poor record-keeping, including failure to record reported figures, stemmed from a combination of disinterest, ineptitude, and possibly also an attempt to hide certain facts from examination, despite, in some cases, the best efforts of civil servants to keep records. A serious charge, yes, but one upheld by further frustrating attempts to uncover the facts. Instead, I pieced together a picture of the evacuation from snippets of information in around a dozen disparate sources. To this day, it appears that nobody knows the true number, or ultimate destinations, of those evacuated.

Image: Island of Montserrat

Crucial to my research was an Evaluation Report commissioned by DFID in 1999 to evaluate the UK Government’s response to the crisis. Volume I of the Evaluation Report is available online. Volume II, however, was redacted from the online version; only its contents page remained visible. I made three separate requests to DFID, the Home Office, and the Foreign and Commonwealth Office for the redacted information. On each occasion I was told that Volume II was not available. I traced a retired civil servant who was listed in Volume I as having been interviewed for the evaluation. They provided an ISBN number for Volume II, from a photocopy of the back cover of the report in their personal files. This allowed me to call up a copy from the British Library’s Boston Spa repository. It was unredacted. A query to the British Library research service confirmed that I was the first person to request the item, meaning it had not been previously viewed by researchers. It remains unclear to me why Volume II was redacted from the publicly available Evaluation Report, or why the government departments mentioned were unable or unwilling to provide a copy. Interestingly though, the findings of Volume II on the official response were highly critical of both the UK and Montserrat governments. Is this why they were withheld? Or was the omission accidental?

A further piece in the puzzle, which took some eighteen months to trace, was the conclusions to the Coroner’s Inquest report into the volcano deaths. I made numerous requests through official channels to government departments and officials, both on Montserrat and in the UK, to access the full report. Most of these were completely ignored. I also contacted numerous individuals, including the former Premier, Governor and Coroner of Montserrat. None were able to track down a copy. The Coroner’s concluding remarks were eventually provided via a further FoI request to the Foreign, Commonwealth & Development Office (FCDO), which took five months to fulfil while officials ‘reached a decision on where the balance of the public interest lies’. My attempts to access the report in its entirety, which reportedly runs to several hundred pages and includes lengthy evidence and witness testimony, has thus far proved futile. The former Coroner who conducted the Inquest expressed to me his surprise that the report was not publicly available, since: ‘one of the purposes of Inquests is to make recommendations to avoid untimely deaths in future [so] their conclusions are meant to be available to all to read’.

Additional records documenting the evacuation and resettlement of Montserratians, held at the National Archives and Bishopsgate Institute, were opened to the public following further requests which I made in 2021. They had previously been closed to scrutiny.

Image: National Archives, Kew, UK.

The unusual lengths to which I had to go to trace documents related to this study are indicative of the elusiveness of information relating to Montserrat and the volcanic crisis, which suggests at best a deprioritisation of Montserrat by the UK authorities, and at worst a deliberate cover-up of policy decisions towards Montserrat. More positively, the success I eventually had in tracing these records is due to the many helpful individuals who assisted me, both in the UK and Montserrat.

Bringing these varied sources together for the first time allows a new story to be told about the evacuation of Montserrat, one which has remained hidden for almost three decades. The voices of the evacuees are still largely excluded, making this an incomplete picture. Nevertheless, it is thanks to the restrictions imposed on my primary research by Covid that the public now has access to a broader range of materials to aid future investigations into a significant period in Montserrat’s – and Britain’s – history.

St Vincent’s volcanic crisis – failing to learn from Montserrat

Saphia Fleury

Falling Through the Net PhD Cluster

Wilberforce Institute, University of Hull


The eruption on 9 April of La Soufrière volcano on the Caribbean island of St Vincent followed months of smoke plumes and other warning signs. At the time of writing, some 20,000 people have been evacuated from their homes, about one fifth of the island’s population. Some are temporarily housed in shelters, others on cruise ships or neighbouring islands. On 14 April, the United Nations launched an emergency appeal after shelter capacity was breached and supplies of drinking water and other essentials ran low. Overcrowding also threatens to spread Covid-19 among the displaced.

To anyone with a long interest in the region or of volcanology, this may all sound familiar. In 1995, La Soufrière’s namesake – the Soufrière Hills volcano – erupted on the island of Montserrat, another former British colony in the same island chain. The southern half of the island, including Montserrat’s capital, were evacuated several times before temporary resettlement eventually became permanent. Around two-thirds of Montserrat’s original population of 10,600 inhabitants eventually relocated overseas, most in the United Kingdom [UK].

The volcanoes themselves also seem to be following similar patterns: eruptions of ash, debris and gases following a period of seismic activity and heavy ash falls that blight everyday life and pose a risk to health. Montserrat’s Soufrière Hills lava domes continue to rumble to this day, although the last major eruptions occurred between 1995 and 2010. Likewise, La Soufrière seems set to pose a threat to Vincentians for some time to come. Yet in my own research into the disaster on Montserrat, I found that neither the UK Government (Montserrat is a British Overseas Territory) nor the Government of Montserrat had been prepared to respond to an eruption, despite warnings from scientists, eruptions on neighbouring volcanic islands, and increased seismicity. Considering that St Vincent suffered three major eruptions in the twentieth Century, and like Montserrat is subject to other natural disasters such as hurricanes, the lack of preparedness and international support appears inexcusable.

To minimise disruption from displacement, and prevent more serious harms such as trauma, exploitation and poverty among displaced people, temporary accommodation must provide for an adequate standard of living. Despite St Vincent’s history of volcanic activity and months of warning signs, it appears that the government did not have enough shelters, hygiene kits or beds, and no adequate contingency for providing clean water and sanitation. The Montserratian government made similar oversights 30 years ago; as a result, 19 people who had returned to the relative comfort of their homes in the exclusion zone were killed by a pyroclastic flow on 25 June 1997.

The inadequate response of the UK Government to an unfolding catastrophe in its overseas territory caused further hardships on Montserrat. My analysis of Hansard records shows little interest in the situation among UK parliamentarians and government until the tragedy of June 1997 forced the issue into the spotlight. By this time, islanders had struggled on in cramped and insanitary conditions with limited access to food, education and health care, for almost two years. Eventually, the British government opened an assisted passage scheme to help Montserratians leave the island, with several thousand ultimately relocating to the UK. Yet the privations and indignities continued when many of the evacuees, who were supposed to be treated as British citizens, were denied National Insurance numbers and other basic support in the UK. Meanwhile the impacts of the late-1990s eruptions continue to be felt on the island to this day, with many Montserratians still lacking adequate housing and employment opportunities.

As the UN launches its emergency funding appeal, Britain, the former colonial ruler of St Vincent, has pledged just £200,000 of support. It is inevitable that the crisis on St Vincent, like La Soufrière itself, will rumble on for some time to come. To prevent a severely reduced quality of life for the relocatees, and potential tragedy if people opt to return home rather than suffering the misery of inadequate shelters, the British government must do more. The international community too should consider routes for safe and legal passage off the island for those who choose to leave, either temporarily or permanently, to reduce pressure on the limited available services. Such relocations should be carried out with dignity and with respect for the wishes of the evacuees.

Montserrat and St Vincent and the Grenadines sit in an increasingly vulnerable position as climate change warms the neighbouring seas and air. More forceful hurricanes and rising sea levels put the islanders at greater risk from future natural disasters. Heavy storms triggered by climate change can also increase the risk of, and dangers from, volcanic eruptions in the region. The sad inevitability that natural catastrophes will become more severe and frequent should prompt policymakers in the Caribbean nations and their former colonial rulers to step up their disaster preparedness and response. Sadly, for the currently displaced population of St Vincent, the window of opportunity may already have been missed.

The volcanic eruption plume of La Soufrière on St Vincent, during explosive eruptions on 9 April 2021. ©ItsMeOni (from @_ItsMeOni on Twitter) 

Criminology, climate change and the ‘useable past’

Saphia Fleury, ‘Falling Through the Net’ PhD Research Cluster

Wilberforce Institute, University of Hull


What are the harms inherent in human migration? Who are the victims and who is responsible? Does it make a difference whether somebody is fleeing environmental catastrophe rather than persecution or conflict? These are some of the questions I am grappling with in my PhD research, which seeks to understand the difficulties faced by migrants, particularly children, in the context of climate change.

My approach is somewhat unorthodox. I am trying to answer these questions by looking to the past, using case studies that are not directly connected to changes in the Earth’s climate. The first case study concerns the ‘boat people’ migration from post-war Vietnam (c.1975-1992), which is generally ascribed to political causes but also had environmental roots. The second is that of Montserrat, specifically the evacuation of more than half the island’s population following a series of catastrophic volcanic eruptions between 1995 and 1998. In what sense can these histories be considered ‘useable’, and provide an insight into future climate migration? For me, the answer lies in a perspective called eco-global criminology (EGC).

Like all branches of criminology, EGC is concerned with harms, victims and perpetrators. It seeks to predict future risks (to the environment, humans and animals) and develop solutions to prevent the worst environmental crimes occurring. Unlike many other branches of criminology, however, it is not limited to harms which are illegal. It also scans the global horizon for acts which are ‘lawful but awful’, which hurt the environment and by extension humanity, but are not strictly criminal. An important aspect of EGC is the transnational nature of these harms, and climate change is a truly transnational problem.

The two case studies, Vietnam and Montserrat, represent geographical regions with a history of major human migration and environmental degradation. In both cases, the people who fled their homes in the twentieth century, including large numbers of children, experienced human rights violations at the hands of the state and other actors.

Many of the children who fled volcanic eruptions in Montserrat came to the UK. Their arrival here often occurred after multiple relocations on their home island and in the Caribbean region. On arrival in the UK, some spent months, even years, in insecure and poor quality accommodation. The policies designed to prevent them leaving Montserrat in the first place, and later to protect them on their migratory journey, often failed to uphold their human rights. As a result, many of the children faced issues including poverty, insecurity, racism, trauma, family separation and a lack of educational attainment.

For all its failures, the evacuation of people from Montserrat was, to some extent, an example of planned and assisted relocation. For the boat people on the other hand, grave uncertainties, including a high chance of death, attended their irregular departure from Vietnam. Thousands were turned away by neighbouring states, resulting in many migrants perishing at sea. Others faced serious human rights abuses in camps and holding centres as they awaited resettlement.

By understanding the patterns of risk and harm that affected these migrants, EGC can help us to predict the risks that displaced persons may face in the future. Importantly, it also gives us the opportunity to prevent harm, by putting in place policies and programmes today that allow people to adapt to their changing environment, and/or protect them if they are forced to move. Vietnam and Montserrat are already experiencing climate change-related degradation and are forecast to experience worsening impacts in the coming decades. It is therefore possible that both countries will see a significant uprooting of their populations in the near future.

Today, Montserrat faces an increasing risk of strong hurricanes and, thanks to the destabilising effects of heavy rainfall, further volcanic activity. EGC can use the lessons of the past to plan for the future; to propose better policies to help Montserrat’s current generation of children remain in their homes, or in the worst case, to migrate safely and with dignity. Similar comparisons and lessons can be drawn from the Mekong Delta in Vietnam, where wartime environmental destruction led to massive food insecurity and was one push-factor in the boat people migration. Today, as flooding, drought and salinisation sweep through the Delta, similar issues threaten to uproot and scatter the rural poor. These examples of the ‘useable past’ provide the benefit of hindsight, and EGC compels us to anticipate and mitigate future harms to prevent another human tragedy.

In 1959, the environmentalist Peter Farb suggested that ‘life is like a delicate fabric’, presenting a romanticised vision of the interdependencies of the human and natural worlds. But his ultimate conclusion had a more ominous overtone:

The wonder is not that so many threads are necessary in the fabric, but rather than the fabric manages to exist at all. (P. Farb, Living Earth, 1960: 164)

When environmental harms pull at these threads, there is a grave risk that the structure will ultimately disintegrate. Both Vietnam and Montserrat have faced historical periods when the fabric of life certainly appeared to be falling apart, with both the natural world and human society hurtling towards a dangerous threshold. Climate change represents a similar existential threat today. Using an EGC approach may help prevent repetition of some of humanity’s past mistakes, as a small contribution to our collective battle against the gravest risk we face.

15 May 1984 – Vietnamese refugees wait to be rescued by the USS Blue Ridge from a 35-foot fishing boat 350 miles northeast of Cam Ranh Bay, Vietnam, after spending eight days at sea