When we were children my mother always told us there was nothing to fear from thunderstorms. But it was hard not to notice that she did so whilst running around the house disconnecting every electrical appliance. She compared her “sensible” approach to the superstitious excesses of her own mother who, at the sound of rumbling thunder, would gather the cutlery into her apron and hide the saucepans in cupboards. Both my mother and my grandmother had been born and raised in the very urban context of London’s East End, where lightning strikes were presumably pretty rare, but maybe my maternal great grandmother, a rural Irish immigrant of the potato famine era, had come to London imbued with Celtic thunder and lightning folklore. In any case, I always loved the excitement of thunder and lightning as a child. It was not until the age of eighteen that I experienced real, body-shaking fear during a thunderstorm.

I was working as a woefully ill-prepared teacher in a rural community school in the highlands of Swaziland. The school was located high on an exposed mountainside, away from the road and the village, but not far from a working iron ore mine. I lived with another volunteer teacher, not in a warm circular Swazi house, but one built especially from cold concrete blocks, with a corrugated iron roof. We had no electricity, so there was no need to unplug anything during the mind shattering red and blue lightning storms that frequently fizzed around us at close quarters, but I did often worry about the roof: too big to be wrapped in my grandmother’s metaphorical apron. Many things about life in rural Swaziland surprised my small barely post-imperial British mind. Amongst these were the bitter cold, the lightning strikes and the South African police raids across the nearby border. Unannounced and apparently unchallengeable, uniformed South African police would swoop into the classroom, grab one or sometimes two of the older male students, bundle them into a truck and drive away. If the nearby police state was a source of unpredictable danger, so too were the forces of nature. Quite often on winter mornings as we lined up outside to sing the national anthem and raise the flag, I would be told that a primary school pupil had been found dead from hypothermia the previous day, having collapsed on the long trek to school. Just as frequently, during the summer months, we would hear that a child (usually a boy herding cattle) had been struck and killed by lightning. I had no way of calculating the risks, but death by lightning seemed not to be a rare occurrence at all – and people would often remind me of this when they saw me out walking across the hills after school.

I hadn’t thought much more about lightning strikes until quite recently. While researching food systems and agriculture in Malawi, an epidemiologist colleague drew my attention to two articles on the frequency of lightning strikes and fatalities from lightning in a lakeshore area in the north of the country where I had recently been carrying out interviews. Using a community-based survey, the researchers had concluded that the Nkhata Bay district on the western shore of Lake Malawi recorded the highest known rates in the world : 419 strikes per million people per annum and 84 mortalities per million per annum – far outstripping the previous record set by Swaziland (15.5 per million). Around one in four of injuries resulted in death (Salerno, Msalu, Caro, Borgerhoff-Mulder, 2012; Borgerhoff-Mulder, Msalu, Caro and Salerno, 2012). Reading these two papers confirmed with statistics what my former neighbours in Swaziland had been trying to tell me – for some communities, the risk of being struck by lightning was nowhere near as rare as my mother had implied. In the “developed” world an average of 0.3 lightning fatalities occur per million people per annum, whilst in the ‘developing’ world the equivalent number is 6. In part this is because 78% of lightning flashes occur in the tropics, but it is also closely related to urbanisation, work patterns and the quality of housing. These factors account for a reported decline in lightning fatalities over the past century in western European countries and elsewhere in the ‘developed’ world. For example, in the United States in the early twentieth century there were around 6 deaths per million of population per annum; now that figure is estimated at 0.1. (Holle, 2016). So my grandmother really did have more to fear from thunderstorms than I do. Death is not the only thing to be feared from a lightning strike: the authors of the Malawi study also reported on the trauma of survivors and families of those who had died, the shock, lasting pain, hearing damage and recriminations and suspicions. Even in the global capital of lightning strikes the “why me?” question still gets asked.

The Malawi study seemed to me to be interesting and important in its own right, but it is also a lesson in the construction of a global health object. What is required for such an object to come into being? There are a few stages. First is the need for quantification where limited data has previously existed. Quantification leads to demystification and a kind of disenchantment, allowing us to see that there is no such thing as an act of god, and or an unmediated act of nature. The subject is attached to a developmental agenda. This is where lightning researchers persuade us that their work is relevant to a much larger set of global issues. An anthropological audit is conducted of “local beliefs and practices”, with a view to revealing which of these are possibly to blame for poor outcomes and which are harmless. Some recommendations are made.

First, quantification. Though metrics sometimes get bad press in the critical global health literature, it’s hard to argue against the more accurate reporting of health conditions and hazards that go hidden in the world’s poorest communities and are sometimes wrongly assumed to be infinitesimally rare. Having gathered the data to establish that rates of consequential lightning strikes are very high in northern Malawi, the authors of the papers go on to place these statistics within a longstanding, though contentious, model of development, arguing that “Where the risk of extrinsic mortality is great, there are diminishing returns to parental effort, which selects for high fertility and lower investment in offspring,”(Borgerhoff-Mulder, Msalu, Caro and Salerno, 2012: 2). Pointing to the evidence that “giving birth at an early age is linked to high crime rates in the USA and high rates of HIV in South Africa”, they push the argument further: “We might therefore expect less diligent child care, higher fertility and earlier age of first reproduction in areas of especially high CLS than in areas with lower lightning activity…”(2).

Let me be clear that I do have a lot of sympathy for the lightning researchers. If you live in a community where lightning strikes are terrifying and not so rare events, it doesn’t matter to you that the risk is still much smaller than dying of an infectious disease. But it does appear that to get this relatively niche research acknowledged in the global health literature involves some fairly extreme acrobatics linking lightning fatalities to an early age at reproduction, neglect of children and crime rates and a generally apocalyptic scenario. Even the authors themselves seem to feel a little uncomfortable with this, issuing a kind of health warning at the end of their paper and stressing that in Malawi “rates of mortality in infancy from diseases including malaria, TB and HIV/AIDS are far greater threats to community welfare…”(Borgerhoff-Mulder, Msalu, Caro and Salerno, 2012:3).

If consequential lightning strikes are to a large degree avoidable, then of course it makes sense to investigate what circumstances give rise to higher rates of injury and mortality and which groups within a population are most at risk. The forensic and epidemiological literature from both South Africa and Swaziland makes for grim and disturbing reading. As my Swazi neighbours had warned me, standing outside on an exposed hillside during a thunderstorm is not a good idea. Children walking home from school and herding animals are, indeed, particularly vulnerable. It’s a scientific literature full of the bodies of children and adolescents, struck down, clothing seared. And though being indoors is some protection, it’s not much protection if your house or shelter is poorly grounded. My belief that the corrugated iron roof might attract lightning was just as unfounded, it seems, as my grandmother’s fear that her cutlery would act as a magnet. It might not offer much protection, but at least it would not burst into flames as a thatched roof might.

Despite the widespread recognition that it’s peoples’ living conditions that make them vulnerable to injury and death from lightning strikes, the impulse of much global health research is to teach people a lesson. So, whilst clearly acknowledging that telling people they should seek shelter during thunderstorms is of only marginal benefit if those buildings are inadequate (in a Swazi study most deaths actually occurred indoors), nevertheless one study in South Africa also concluded that by walking in open ground during a storm “the victims have made themselves prime targets,”(Hunt, Blumenthal, Nixon and Gomes, 2020:13). In the Swazi study the author similarly concludes that:

“Although lightning fatalities can be related to coincidental circumstances, they are still mainly determined by the activity and location of the individual or group of individuals during the thunderstorm… This implies that fatalities are mainly the result of the choices that people make and thus are potentially preventable,” (Dlamini, 2009 : 188; my emphasis).

Some “choices”, the literature implies, are related to “traditional beliefs”. This is another familiar component of the construction of a global health subject: identifying local (meaning pre-modern, superstitious) beliefs and practices that increase vulnerability to harm or are actively harmful in themselves. Rarely are such practices found to be protective. And not infrequently practices thought locally to be protective are found to be harmful. So, in the case of Swaziland, where there exists a dense and powerful set of ideas around lightning, the belief that certain trees species are protective clearly has the potential to lead to harm (Dlamini, 2009) – though how often in a thunderstorm one has time to locate a specific tree for shelter is not very clear. In one South African study a tire placed on the roof of the house as protection was clearly not protective in any scientific sense (Trenglove and Jandrell, 2015) – but I think you could argue that by encouraging people to stay beneath the roof rather than out of doors, it might offer some protection, depending on the nature of the shelter. Along with a belief in the protective powers of “muti” or medicines, these ideas are assumed to lull people into a false sense of security. In other words, the subjects of these studies are insufficiently fearful of lightning and therefore less likely to follow scientifically based advice. Some of these ideas are interpreted as “cultural coping mechanisms” employed to make sense of the otherwise unfathomable. In the most thorough study of “lightning myths’ in Southern Africa (which included a sample of dual-thinking electrical engineering students) the authors conclude that some false beliefs can be regarded as harmless, as long as they do not put others at risk. However, the belief that witches could harness lightning presented more of a dilemma. Though recognising this belief as mistaken and misleading, in the end, it seems, there has to be a limit placed on science and its disenchantments. Arguing that it is “not the place of those who do not believe in witchcraft to dissuade those that do”, the authors recommend that lightning education “should focus on protecting oneself against natural lightning…”, implying that another kind of lightning does indeed exist (Trenglove and Jandrell, 2012:10).

Somewhat predictably, in 2019/20, the lightning strike has become a frequently cited comparator for the risk of dying from COVID 19. The airline industry argues that an individual (which, where?) has a much greater chance of being struck by lightning over a twelve month period than they do of catching COVID 19 during a commercial flight. In South Africa it is pointed out that children under 15 years have a 1/35 million chance of dying from COVID19, compared to their 1/350000 chance of being struck by lightning. The global pandemic and climate change either separately or together have given further fuel to lightning research. As the frequency and distribution of lightning inducing storms are affected by climate change and air pollution, so COVID-19 presented an opportunity to study the effects of ‘lockdowns’ on lightning activity (Chowdhuri et al. 2020).

My most recent near encounter with lightning occurred while kayaking on a river. I followed my instincts and got out of the river pretty quick – then sheltered from the torrential rain under a tree. Some of us, it seems, never learn.

---

Megan Vaughan is Professor of African History and Health at University College London. She has written on the histories of medicine and psychiatry in colonial Africa, on gender, nutrition and food systems and on slavery in the Indian Ocean. She heads a Wellcome Trust funded project on the critical histories and anthropology of ‘NCDS’ in Africa and is writing on food, environment and metabolism in Malawi and Southern Africa.

---

Borgerhoff-Mulder, M., Msalu, L., Caro, T., and Salerno, J., 2012, ‘Remarkable Rates of Lightning Strike Mortality in Malawi’, PLoS ONE doi: 10:1371/journal.pone.oo29281

Chowdhuri, I., Pal, S.C., Saha, A., Chakrabortty, R., Ghosh, M., Roy, P.,2020, ‘Significant decrease of lightning activities during COVID 19 lockdown period over Kolkata megacity in India’, Science of the Total Environment, 747. doi: 101016/j.scitotenv.2020.141321

Dlamini, W.M, 2009, ‘Lightning Fatalities in Swaziland, 2000-2007’, Natural Hazards, 50:179-191

Holle, R.L., 2016, ‘A summary of recent national-scale lightning fatality statistics’, Weather, Climate and Society, 8:35-42

Hunt, G.P., Blumenthal, R., Nixon, K.J., and Gomes, C., 2020,  ‘A multidisciplinary forensic analysis of two lightning deaths in South Africa’, International Journal of Disaster Risk Reduction, doi: 10.1016/j.i.drr.2020.101814

Salerno, J., Msalu, L., Caro, T. and Borgerhoff-Mulder, M. (2012), ‘Risk of injury and death from lightning in Northern Malawi’, Natural Hazards, 62:853-862

Trenglove, E., and Jandrell, I.,2015, ‘Lightning Myths in Southern Africa’, Natural Hazards, 77:101-110