When examined closely, the backgrounds of some of the most widely cited origin stories for zoonotic disease outbreaks have been found to be irreconcilable with empirical data. Stated simply, these explanatory landscapes do not appear to have existed. Here, I present a detailed case study of one such fictional landscape, that of a monkey-filled forest which was identified as the source of a suspected zoonotic outbreak in the Brong-Ahafo Region of Ghana in 2010. Taking my approach from cultural epidemiology, I elucidate the mechanisms by which this fictional landscape was constructed and transmitted among the professionals involved in the response.
In early 2011, an epidemiological investigation was undertaken into an apparent outbreak of B virus, a monkey-borne simian herpes virus, in children in the then Brong-Ahafo Region (BAR) of Ghana. The investigating team reported that the affected children were all from communities bordering a forest-belt and that the children had likely become infected interacting with the wild monkeys that inhabited it. Over the following two years, a variety of research coalitions, comprised of both Ghanaian and foreign doctors and researchers, investigated the outbreak. When subsequent laboratory testing cast doubt on B virus being the causative agent, these projects pivoted to consider other zoonotic diseases associated with animals believed to inhabit the forest-belt. These later projects were never fully realized and attention around the outbreak petered out without any generalizable knowledge having been generated nor any public health intervention taking place. This series of events is curious for a number of reasons, not least because there was no forest-belt in the affected area. In fact, the majority of affected children come from the sprawling, largely barren, city of Techiman and none of the children nor their families had encountered any wild monkeys. In short, these protracted public health responses were organized in response to a landscape that did not exist.
This odd situation is not necessarily unique to the public health responses to the BAR outbreak. In recent years, some of the most widely circulated accounts of zoonotic disease outbreaks have featured environmental and social backdrops that are not easily reconcilable with empirical evidence. For example, popular narratives around the origin of the 2014 West African Ebola epidemic rendered Meliandou, the town in Guinea where the epidemic was believed to have started, a site of deforestation and therefore a place primed for spillover of the Ebola virus from local bat populations into human ones. However, anthropologists James Fairhead and Dominique Millimouno, who were already familiar with the history of the area, challenged this rendering of south-east Guinea, suggesting that the idea it was a site of significant deforestation was a “misreading” of the landscape, one which exaggerated how forested the area had previously been (Fairhead et al. 2017). Similarly, the “cut hunter hypothesis,” perhaps the predominant account of the origins of the HIV/AIDS epidemic, described a setting whereby a rural Cameroonian population, newly armed and mobile courtesy of late nineteenth century European colonizers, embarked on hunting non-human primates. Through archival research and the collection of oral histories a group of historians showed, however, that the cut-hunter hypothesis misunderstood both the preexisting interactions between non-human primates and the general mobility of the local population (Rupp et al. 2016).
In both examples, while the series of events explaining the outbreaks were presented as speculative and open to revision, the backdrops of these zoonotic tableaus and activities ascribed to the local populations were communicated as being certain. These backdrops were not marked for scrutiny in the same way as other components of these explanatory tableaus, such as the implicated animal hosts or activities that enabled transfer of the pathogen to a human population. Though, as both groups of researchers noted, these backdrops underwrote the reasonableness of the other components of the tableau and their speculated connections while precluding alternative explanations. These aberrant landscapes were therefore imbued with explanatory power and have direct bearing on what will be identified as a “driver” of zoonotic emergence and robust target for public health intervention.
The BAR outbreak provides a useful case study through which to examine how such aberrant landscapes or settings can come to be promulgated within these epistemic communities and in doing so infiltrate both models of specific zoonotic disease outbreaks and of zoonotic emergence more generally. Prior to the absence of the forest-belt and its non-human inhabitants becoming apparent, the responses to the BAR outbreak were already the subjects of an ethnographic study I was conducting and as such there was an abundance of observations, interviews, and primary documents, available to me through which to trace the genesis and spread of the “forest-belt” among the responders to the outbreak. In analyzing this material, I will borrow an approach from the anthropology of rumor. Specifically, I will draw on Dan Sperber’s “epidemiology of representations” as adapted by Julien Bonhomme for use on empirically developed case studies (Bonhomme 2016; Sperber 1997). This is not because I think that this phenomenon can be understood as a type of rumor. Certainly, it lacks many of the usual trappings of a rumor, in that it was not primarily spread through word-of-mouth, it does not appear to be entirely oriented toward stigmatizing deviant behavior, and the idea that there is a forest in the BAR of Ghana is not an “inherently irrational belief” (Sperber 1985:35). I primarily chose this quasi-epidemiological approach to investigating rumors, because dryly describing the “dynamics of the distribution in space and time of phenomenon within a population” and “identifying the causal factors that can explain said distribution” (Bonhomme 2016:13) seems an effective way to circumvent fixating on overly symbolic interpretations of the “rumor” itself – that of a fantastical giant monkey-filled forest – and instead focus on elucidating the underlying mechanisms and drivers of it. That said, I also chose this approach because it is more familiar terrain to me as a sometimes field epidemiologist working on zoonotic outbreaks in Africa who came to anthropology in the hopes of leveraging its insights in order to better understand the deficiencies of such responses.
I will break my analyses into three sections. In the first section, I will describe the temporal and spatial features of the “rumor” (i.e., where the idea of a forest-belt first arose and how it was communicated and transformed over time). In the second section, I will ask to what extent those who were propagating the rumor were cognizant that what they were communicating was inaccurate. In the third section, I will consider why the idea of a monkey-filled forest in particular gained the traction it did within this specific epistemic community. I will then attempt to draw these insights together in order to describe how and why this phenomenon might have occurred in relation to the BAR outbreak and what this might reveal more broadly about transnational knowledge production around suspected wildlife-derived zoonoses in sub-Saharan Africa. Before embarking on all of this, however, I will first give a more detailed account of the BAR outbreak responses and how the presence of the forest and its human and animal occupants came to be disputed.