Plague history, Mongol history, and the processes of focalisation leading up to the Black Death: a response to Brack et al.

Abstract

This essay responds to Brack et al., ‘Plague and the Mongol Conquest of Baghdad (1258)? A reevaluation of the sources’, which is a critique of our 2021 essay in this journal, ‘Plague and the Fall of Baghdad (1258)’. We argue that Brack and colleagues have misunderstood our investigation as an attempt to pinpoint the exact timing of the outbreak of plague connected with the Mongol siege of Baghdad, and so believe that an altered timeframe invalidates our suggestion that plague was involved. Taking this opportunity to revisit the state of plague historiography in western Asia, we address four issues: (1) why Mongol historiography has, until recently, avoided the question of plague’s late mediaeval resurgence within the Mongol Empire and why the ‘new genetics’ of plague now makes the question unavoidable; (2) why reconstruction of the biological processes of ‘focalisation’ is now the most urgent question in plague historiography since it constitutes what we call the prodromal stage of the Black Death pandemic; (3) how a newly informed biological perspective on disease history can allow a more sensitive reading of past observers’ reports of epidemics; and finally, (4) what a plausible scenario might look like for plague’s presence in western Asia and the eastern Mediterranean region in the late-thirteenth and early-fourteenth centuries as an emerging zoonotic disease with occasional epizootic and human outbreaks, before the more catastrophic outbreaks of the 1340s commonly referred to as ‘the Black Death’.

Introduction

In the three years since our essay, ‘Plague and the Fall of Baghdad (1258)’, was published in 2021, approximately 268 major studies have appeared contributing to the late mediaeval history of plague and its causative microbial organism, the bacterium Yersinia pestis. ¹ As we write, at least half a dozen more palaeogenetic studies are moving through peer review or are in press. It is, in short, a very active time for plague history. This has not always been the case. On the contrary, in the final two decades of the twentieth and first decade of the twenty-first century, certain aspects of plague history (like many areas of premodern infectious disease history) were moribund, the victim of the successes of medical history itself as it developed more self-critical perspectives on how information about the past could be gathered and responsibly analysed. On the question of plague in particular, continuing scepticism over whether or not the late mediaeval pandemic commonly known as the ‘Black Death’ had indeed been caused by the bacterium Y. pestis led many historians of medicine to avoid the field, save for certain aspects of social or economic history where the evidence of mass mortality was clear and its effects profound.

Since 2011 (almost a decade before the COVID-19 pandemic), however, the Black Death – and many other aspects of plague history – has seen a resurgence in interest. Two thousand and eleven was a pivotal year because it was then that the genie of molecular approaches to disease history was let out of the bottle by the recovery of Y. pestis from four bodies buried in two mid-fourteenth-century cemeteries in the centre of London. ² These were not simply random fragments sufficient to confirm the presence of Y. pestis by a few diagnostic markers. No, these were nearly complete genomes which could be compared and analysed against every other genetically-sequenced isolate of Y. pestis anywhere in the world. A global history of plague was now possible.

Yet in ushering plague history into the genetics age, this technical achievement in molecular biology has also raised questions about how the physical evidence for microbial pathogens was going to be integrated into medical history. These challenges, far from being insurmountable, we feel, are the very warp and weft of the mantle that the field of medical history now needs to take up and bear. We are indeed in a new era of investigative possibility, and the challenges of a climatically- and microbially-changing future make responsible engagement with that history all the more pressing.

The essay by Jonathan Brack, Michal Biran, and Reuven Amitai offers us the opportunity to revisit the dossier of evidence we compiled in 2021 for the possible presence of plague in western Asia nearly a century before received historical narratives have suggested it was present. ³ We find it curious that on the cusp of a ‘brave new world’ of possibility in integrating physical and textual evidence, Brack and colleagues wish, on the one hand, to evoke one single palaeogenetic study as ‘definitively proving’ the absence of certain phenomena, and on the other, to revert to a decades-old body of evidence and methods that, even among text-based humanists, seems antiquated. In what follows, we draw on the empirical findings of plague scientists, archaeologists, and historians, as well as our own methodological interrogations of how work in disease history can be effectuated, not simply to argue for the validity of our original findings, but to expand them. Indeed, many points that Brack and colleagues make actually support our main argument as we shall demonstrate below.

Our arguments are presented under four headings. First, in the section ‘Plague historiography and the Mongol Empire’, we review the historiography of the late mediaeval plague pandemic as it pertains to western Asia and the Mongol Empire. Remarkably, Mongolists have generally refrained from engaging in plague studies despite clear evidence that the late mediaeval pandemic arose, at least in part, within the Empire’s bounds. These historiographical traditions are then set into the context of the ‘new genetics’ to make two related arguments: that an ascertainment bias, due to the Eurocentric origins of both the field of palaeogenetics and plague history generally, leaves open far too many questions for ‘definitive’ proofs, positive or negative, to be claimed on most plague-related topics right now; and secondly, that the specific ‘definitive proof’ from genetics that Brack and colleagues invoke to claim a fourteenth-century origin of the Black Death has, in fact, already been challenged by plague historians and palaeogeneticists.

In the section ‘Focalisation and the methodological demands of gaps between the biological and documentary archives’, we note that Brack and colleagues have fundamentally not understood that plague is an enzootic disease, meaning it persists evolutionarily in animals, specifically rodents. A central burden of the plague historian is not simply to chronicle human outbreaks, but to establish at least plausible scenarios for the generally undocumented stages of enzootic persistence of plague in animal reservoirs and focalisation of the disease in new landscapes when it is introduced into them. In other words, a comprehensive, global history of plague must explain all aspects of the ‘trophic cascade’ that moves Y. pestis between insect vectors, maintenance hosts (most likely wild rodents), commensal intermediate hosts (such as rats), and (rarely) humans. And it must explain how that process happens not just once, but repeatedly. Most of that history is undocumented, either by the written documentary archive (created by humans, of course) or, currently, by the palaeosciences. ⁴ Nevertheless, as we make clear by drawing on modern plague ecological and palaeogenetic studies, what we call a ‘prodromal’ stage is the necessary antecedent to any sustained plague pandemic, which is precisely what we believe Afro-Eurasia experienced in the thirteenth and early fourteenth centuries.

Having addressed the range of phenomena the plague historian must seek to plausibly claim plague’s presence at any given time or place in the past, we then address, in the section ‘Perceiving plague’s presence: reading the past without a microscope’, the arguments we have drawn from our mediaeval documentary sources. In combining evidence from the natural sciences with the documentary record, the disease historian’s task is to make a plausible account of both the physical disease and the human perceptions of it according to the conceptual frameworks of past societies. For thirteenth-century western Asian societies, which largely functioned within the broader conceptual frameworks of Hippocratic–Galenic medical thinking, the crucial concept was that epidemic diseases (those that struck many people all at once) were due to miasms. As we argued in our original paper, that is precisely the scenario that the collective accounts of contemporary and near-contemporary chroniclers envisioned regarding the events of 1258. Hence, evaluating what happened at Baghdad in the context of what subsequently happened in Syria and Egypt in the following two years is not a wild goose chase, but an attempt to reconcile the historical actors’ observations with what might have been an actual path of spread of plague through new rodent hosts and human populations in the region. In not understanding how we were moving back and forth between historical understandings of the disease and what we now know about Y. pestis to create a plausible path for the displacement of Y. pestis from its long-time reservoir near the Tian-Shan mountains to Western Eurasia, Brack and colleagues mischaracterize our main arguments by fixating on the timing of a given outbreak. In their minds, as long as a disease outbreak took place before the lifting of a Mongol siege, the outbreak could not have been caused by Mongol displacement of the bacterium. They maintain this despite themselves acknowledging at least two disease outbreaks before 1258 in cities under siege by the Mongols. They also incorrectly claim that the ‘linchpin’ of our argument is the account in a Persian chronicle copied by the physician–astronomer, Quṭb al-Dīn al-Shīrāzī (d. 1311) and that we ‘somewhat interchangeably’ use the terms wabāʾ and ṭāʿūn (ostensibly) to mischaracterize individual disease outbreaks as the plague for which, according to them, specific symptoms are not described. ⁵ We of course were very careful in identifying when a text uses wabāʾ (epidemic) and where one uses ṭāʿūn (plague), which is why we used the term ‘plague-like disease’ to broadly refer to lethal disease outbreaks in Egypt and Iraq, for which the more specific word ṭāʿūn is not invoked in the original sources. Our goal was to document that a specific plague outbreak is recorded for Damascus, and how ‘chroniclers could classify the disease outbreak as only an “epidemic” [wabāʾ] … for Baghdad itself but as plague [ṭāʿūn] … for Damascus’. ⁶ We also showed that the Syrian outbreak came to be associated with the fall of Baghdad in a highly influential contemporary account that was repeated by several authors of the next two generations. These claims, which are the central claims of our paper, are not found in the chronicle copied by al-Shīrāzī. Of course, what this Persian chronicle provided us was a plausible mechanism for how the Mongols could have imported Y. pestis into the region from the Tian-Shan without their knowledge nor causing continuous outbreaks along the way, which is by way of grain imports. This is again an example of moving back and forth between our historical sources and their conceptions, and our modern understanding of disease ecologies to arrive at a plausible historical scenario for the displacement and focalisation of the bacterium.

Finally, in the section ‘The case for plague’s presence in the Western Ilkhanate in the late thirteenth and early fourteenth centuries’, we return to the original question that prompted our 2021 essay: if a series of plague outbreaks did indeed occur in the western realms of the newly established Ilkhanate and Ayyubid Syria and Mamluk Egypt in the context of the Mongol conquests of the 1250s, and if physicians were responding to the new presence of the disease in their midst, why was that history forgotten and what does its recovery mean for us now? Brack and colleagues contend it was not ‘forgotten’; they contend it never happened – or if there was any kind of epidemic, it was not plague. Clearly, as we acknowledged in our original work, there is an unevenness in the documentary record that is puzzling. As we point out here, there is still no ‘smoking gun’ of genetic evidence to prove that a new strain of Y. pestis was introduced into western Asia in the mid-thirteenth century. However, there is hardly any genetic evidence for infectious diseases in western Asia for the late mediaeval period, since the region has rarely been included in palaeogenetic investigations to date. ⁷ Nevertheless, the challenge of documenting plague in its pre-pandemic, prodromal phases means documenting it at transmission stages that would likely fall below human awareness. Physicians are an important exception, of course, and it was precisely because of the notes about plague (ṭāʿūn) in the writings of physicians (or scholars with medical training) that our attention was initially drawn to the question of the changed perceptions in the 1250s and after. Using the principle that multidisciplinary approaches must allow arguments to be constructed with corroborating evidence from multiple fields, ⁸ we offer a summary of the case for plague as it stands now. We present new evidence to support something our previous paper only hinted at: that the western Ilkhanate may, in fact, have been one of the regions in which newly introduced strains of Y. pestis focalised in the late thirteenth and early fourteenth centuries, creating the biological foundations for the explosive proliferation of plague in the 1330s and 1340s under the influence of changing climatic conditions.

Plague historiography and the Mongol Empire

In 2015, the late Thomas Allsen – widely hailed as one of the leading Mongolists of the last three decades – published a study on ‘Population Movements in Mongol Eurasia’. In it, he included one and a half sentences on the Black Death, that ‘complicating factor’ in population loss under the Mongol regime. ‘This scourge’, he said, ‘began in eastern Inner Asia in the 1330s, reaching the Crimea in 1346 from whence it spread into the Mediterranean and Europe’. ⁹ Allsen’s epidemiological summary – vaguely conjuring a sort of dark cloud sweeping westward across Eurasia in under two decades – is, in fact, the standard narrative given in virtually all textbooks on mediaeval and global history in recounting the origins of the Black Death. What is striking is that, in support of the historicity of this massive demographic catastrophe, Allsen (a formidable researcher in most of the major languages used in the Mongol realms) could only cite two monographs from the 1970s: William McNeill’s largely outdated Plagues and Peoples (1976), and Michael Dols’ still-indispensable The Black Death in the Middle East (1977). ¹⁰ Allsen’s passing reference to the Black Death was in fact the only time the pandemic was mentioned in the 355-page book where his essay appeared, a volume wholly devoted to ‘nomads as agents of cultural change’. A veritable industry of Mongol Empire studies in the past two decades shows a similar tendency to skirt around the Black Death and, even when acknowledging it, to document it only by reference to dated work. ¹¹ Exceptions are the recent works by Marie Favereau, who presents differing accounts of the pandemic’s origin and route in her two recent books, and Peter Jackson, who, more valuably, reviews many of the standard primary sources already invoked by Dols fifty years ago, though with important additions from Russian and Persian literature not examined by Dols. ¹² Neither Favereau nor Jackson, however, engage with the scientific literature on plague’s biology, ecology, or evolutionary development. The latter is mentioned in passing in the Introduction and Epilogue of the 2023 two-volume, 1522-page Cambridge History of the Mongol Empire, but it is not meaningfully integrated into the narratives; in fact, plague is shunted aside in the chapter on ‘The Climate and Environment of the Mongol Conquest’ as a topic to be deferred to future studies. ¹³

So why has Mongol historiography largely avoided the topic of the Black Death, treating it (if it is mentioned at all) as an inexplicable natural catastrophe that just happened to coincide with the collapse in the mid-fourteenth century of three of the four khanates established by Chinggis Khan’s grandsons? Brack and colleagues, as Mongolists, may be better suited to answer that question, though we, as medical historians, can provide our own speculation. On the one hand, as has now been recognized numerous times, the Black Death itself, as a historical phenomenon, has largely been framed from Eurocentric perspectives. ¹⁴ The idea that it ‘came from the East’ is embedded in many contemporary European accounts, even though what ‘the East’ was varied from vague understandings of the Black Sea to imagined ideas of India or China. Actually investigating the source of the pandemic outside of Europe, let alone reckoning with the impact of the pandemic, however, has not been considered the burden of most plague historians. The most recent revision of a ‘complete’ history of the Black Death persists in its belief (first articulated in the original edition of 2004) that the pandemic originated out of ancient plague reservoirs in the Caucasus and had no appreciable effect on the Mongol world, most of which is literally left off the book’s map. ¹⁵ The continued attraction of McNeill’s half-century-old Plagues and Peoples lies in his willingness to paint a sweeping pan-Eurasian picture that ties together what was known about plague ecologies in the 1970s. ¹⁶ Dols, as we discussed in our 2021 essay, accepted the ‘compromise’ Issyk Kul narrative that the World Health Organization (WHO) plague expert, Robert Pollitzer, had popularised in the 1950s, which placed the pandemic’s origins neither in far east Asia nor the region of the Caucasus, but at an intermediate point in Central Asia. That narrative – which enshrined not simply a geographical origin but a chronological one too (thanks to a well-dated outbreak of mawtānā ¹⁷ in a Christian community of Kara-Djigach, near Issyk Kul in what is now Kyrgyzstan) – was attractive because it told a simple story that lined up inherited narratives (‘It came from the East’) and archaeological support with an absence of any explicitly conflicting evidence. ¹⁸

Getting beyond that simple narrative has been hard, if not impossible, because of the evidentiary disconnect between a rodent disease and human perception. In the section ‘Focalisation and the methodological demands of gaps between the biological and documentary archives’ below, we address the process by which Y. pestis becomes established in a new landscape, embedding itself in a rodent infrastructure that, along with its symbiont flea communities, allows Y. pestis to be perpetuated in a sylvatic cycle. Here, however, there is still a need to address one further aspect of plague historiography: the impact of the ‘new genetics’ and how it ties the Black Death even more firmly to the history of the Mongol Empire.

Brack and colleagues never mention the important refutation of a calumny raised against the Mongols with regularity since the nineteenth century: that is, that Mongol warriors ‘initiated’ the Black Death by hurling plague-infected bodies over the walls at the siege of Caffa in 1346, thereby infecting the Italian merchants in their enclave in Crimea, who in turn fled back into the Mediterranean carrying the disease with them. This accusation of bioterrorism has haunted Mongol studies, and may perhaps explain the reticence on the part of Mongolists over the last few decades to engage in plague scholarship. However, this story derives from a single fourteenth-century manuscript, which was not retrieved for modern scholarship until the mid-nineteenth century. In a recent and highly persuasive study, Hannah Barker demonstrates that the account of the siege found in Gabriele de Mussi’s Historia de morbo (History of the Disease, written at some point between 1350 and his death in 1356) conflated different events, invented certain details, and collapsed the timeline of the likely transmission of plague via contaminated grain shipments out of the Black Sea. ¹⁹ Barker’s work has had the effect not simply of vindicating the Mongols of the Golden Horde of the charge of bioterrorism (she demonstrates that the siege ended well before plague was likely transmitted across the Black Sea), but she also links the transmission of plague with the vitally important history of long-distance grain shipping in the late Middle Ages.

Yet, just as Barker was ridding Mongol historiography of one association with plague, genetics was tying the Mongol Empire more firmly to plague’s history with another. Ever since the Aleppan historian and poet, Ibn al-Wardī (d. 1349), wrote his Risālat al-nabāʾ ʿan al-wabāʾ (Essay on the Report of the Plague) just prior to his own death from plague, the major narrative tradition of the Black Death in Arabic connected the fate of the western Islamicate world with that of the Mongol Empire. This connection was further embellished by the fifteenth-century Egyptian historian, al-Maqrīzī (d. 1442), whose account has largely shaped modern narratives as they pertain to the eastern Mediterranean world. As has been established, the ‘out of the East’ narrative that these two authors advanced was enshrined in European plague historiography when an eighteenth-century historian collated Arabic accounts with Chinese encyclopaedic records, finding datable descriptions of epidemics that seemed to fit closely enough with Arabic accounts to infer that all of Eurasia was experiencing a common, swift transmission of a single common threat. When the WHO plague specialist, Pollitzer, was writing his brief summation of plague history in 1951, the European narratives already ensconced in general plague historiography could be combined not simply with a summary of the archaeological discovery of the Kara-Djigach site in the nineteenth century, but with recent ecological reports of plague in marmot populations of the nearby Tian Shan mountains. ²⁰

Most recently, environmental historian Philip Slavin has published a series of studies arguing for what we call the Quick Transit Theory: that is, the idea that a novel strain of Y. pestis, originating in the 1330s, moved swiftly and directly from the region of Kara-Djigach to the northern Caspian Sea and then to the Black Sea (an overland transit of over 3000 km) in the space of about eight years. At the core of Slavin’s work is a 2022 article in the science journal Nature, which he produced with a team of geneticists, anthropologists, and a linguistic historian. Building on the work of the nineteenth-century Russian archaeologists who initially excavated the Kara-Djigach site, Slavin’s team claimed to have identified the moment of origin of the Black Death pandemic – or rather, the Ur-moment, since they argue that the strain they retrieved from the cemetery of Kara-Djigach in Kyrgyzstan represents the stage of the development of Y. pestis just before the Big Bang, the sudden expansion of an ancient lineage of Y. pestis into new ecological niches from which the four modern branches of the organism arose. ²¹ This study is cited by Brack and colleagues as having ‘decisively demonstrated’ that the Y. pestis genotype involved in the Black Death emerged only in the fourteenth century, specifically sometime between 1308 and 1338. ²² Crucially, this dating comes well after the main expansionist phase of the Mongol Empire, under Chinggis Khan (d. 1227) and his sons in the first half of the thirteenth century. We find Brack and colleagues’ absolutist claim surprising since it ignores already published research by other geneticists suggesting that the new lineage of Y. pestis involved in the Black Death (Branch 1) as documented in Europe had already split into two sublineages as much as a century before reported outbreaks in 1347/8 (see Figure 1 for an abstract detail of the Y. pestis phylogenetic tree). ²³

Figure 1. The Big Bang. A schematic phylogenetic tree of Y. pestis evolution from its probable later medieval proliferation in or near the Tian Shan mountains (the Big Bang) up through the documented geographic locations of the four extant lineages, Branches 1-4. Lineages documented solely from ancient DNA (aDNA) are indicated with a black box; those still living are in a white box. The geographic distribution is indicated in the rhomboids.

Source: The authors.

These debates about the evolution of Y. pestis are ongoing, and will likely continue for some time as geneticists incorporate more genomes into their collective databases and refine their phylogeny-deriving and dating methods. However, while geneticists continue to debate within their field, we historians must continue to work within ours. A principal challenge for historians making use of genetic data is that lacking the high-powered computers and bespoke software needed to do computations on many terabytes of data, we are dependent on the tidbits of genetic information and interpretive analyses that scientists choose to share. But as in all cases where historians do comparative work, drawing on the studies of colleagues when we cannot replicate access to their raw data, we apply our skills of inference and logical analysis to determine whether arguments are plausible and whether the collective body of evidence is mutually reinforcing. When we see scientists themselves disputing conclusions that rely solely on computational analyses, that is a signal to us that there is a need for further interrogation. Although we are not quite as sceptical as the research team that has done the most sustained analysis of the phylogenetic divergence of Y. pestis and dating claims, we do note their conclusion: ‘genetic evidence alone is insufficient for accurately reconstructing the timing and spread of short-term plague epidemics’. ²⁴

Our reading of the data from the 2022 Kara-Djigach study, set within the wider context of plague science and historiography, is that, as had been suspected when the first epigraphic evidence from the cemetery was studied in the 1880s, the ‘epidemic’ (mawtānā) reported at Kara-Djigach was indeed caused by Y. pestis, meaning that the excess mortality in that community in 1338–9 was indeed a plague outbreak. However, the claim that the strain involved in these deaths was the ancestral strain to the Big Bang is far-fetched for several reasons. As noted, geneticists themselves have questioned the method of analysis used. Although the method of using predetermined diagnostic SNPs (single nucleotide polymorphisms, essentially ‘typographical errors’ in the genetic code at certain points along the genome) is well-established in palaeogenetic studies, it is not a foolproof method: it allows you to find what you have already deemed to be important, but it may also induce you to ignore variants that you did not expect to find. Subsequent analysis using a more inclusive method showed the Kara-Djigach strain to be firmly a part of Branch 1, more closely related to strains involved in European outbreaks of the 1340s (the historically documented Black Death) and not immediately related to the Big Bang at all (Figures 2A and 2B). ²⁵ Moreover, from the simple standpoint of historical analysis, the Kara-Djigach study excluded comparative data from what must have been earlier evolutionary forms of Y. pestis (Branch 0 in the phylogeny), even though they would have come out of ecologically similar environments in Kyrgyzstan and surrounding regions, and were, it has been argued, involved in the Justinianic Plague pandemic in the early Middle Ages. Most importantly, the researchers did not satisfactorily address how their conclusions fit with the already published evidence for probable plague outbreaks in northern China in the prior century. ²⁶

Figure 2A. Where the Kara-Djigach genome falls in a phylogeny based on SNPs. Above, a map showing the location of key Second Plague Pandemic Y. pestis genomes retrieved from archaeological remains, as well as the location of the port city, Caffa. Source of map: Wikimedia Commons. Below is a detail from the phylogenetic tree for Branch 1 of Y. pestis, structured according to the SNP-based analysis used by Spyrou et al., ‘The Source of the Black Death’ (note 22). In this analysis, the Kara-Djigach genome (BSK001, marked with an arrow) falls basal to (evolutionarily earlier than) all the other early Branch 1 genomes.

Source: Clavel et al., op. cit. (note 23), Figure S5 (detail). The arrow on the map has been added to show Spyrou et al.’s proposed directionality of spread.

To sum up: the Kara-Djigach site is of great importance for documenting the impact of plague on communities in Central Asia. It may eventually help explain the fate of similar Christian communities in central Asia and the history of plague in the Chaghadaid khanate, which would be split in two soon after this outbreak. However, if the phylogenetic critiques are correct, the site has no direct bearing on the question of the temporal origins of the Second Plague Pandemic as a whole. As we will discuss below, right now several European genomes (including that from Nabburg, highlighted in Figure 2B) seem to be more important in that regard.

Figure 2B. Where the Kara-Djigach genome falls in a phylogeny based on whole genomes. Above, the same map as in Figure 2A. Below, using a different method of calibration in a second analysis, Clavel et al. (2023) show that the Kara-Djigach genome (again marked with an arrow on the detail of the phylogenetic tree) no longer falls in a basal position to other Branch 1 genomes, but is instead part of a clade (a group sharing a common ancestor) most closely related to genomes from European Black Death sites (marked by a bracket added by the present authors). The two most basal strains are those from Laishevo, on the Volga River, and Nabburg, in Bavaria, both of which are indicated on the map.

Source: Clavel et al., op. cit. (note 23), Figure S6, column E (detail).

It would be well to quickly review the evidence for an alternative narrative that appeared at the same time we published in 2021, along with evidence that has become available since then. As Brack and colleagues have noted, our 2021 essay on Baghdad was the sequel to Green’s single-authored 2020 essay, ‘The Four Black Deaths’. There, Green (1) summarised the evidence already proposed by geneticists and plague ecologists for the phenomenon of the Big Bang, its timing, and its likely geography vis-à-vis the locations of living descendant strains; (2) summarised Robert Hymes’ initial arguments for the apparent association of Mongol sieges and plague-like mass mortality events in Song China in the 1220s and 1230s; and (3) proposed what was explicitly described as a scenario for how plague might have been moved long distances across Central Asia in the 1250s in such a way that would not produce mass mortalities (thereby decreasing the likelihood of being noticed by chroniclers) and yet effectively be delivered (along with suitable insect vectors) to new locales. The timing of the Big Bang could only be suggested along the lines of the time spans proposed by phylogeneticists: sometime between the late eleventh and early fourteenth centuries. No palaeogenetic proof for any of these claims was yet available. ²⁷

The latter reality still holds as of May 2024. With the exception of an anomalous Y. pestis genome from Denmark, announced in 2023, and another from Estonia, both of whose dating is still under debate, no Y. pestis genomes from the thirteenth century have yet been retrieved. ²⁸ More to the point, however, aside from two fourteenth-century genomes from western Russia, ²⁹ there is as yet no late mediaeval genetic data from anywhere in western Asia or North Africa.

This last point is crucial. As we have noted repeatedly, palaeogenetics is a new field and is still developing not only its technical and analytical methods but also its basic databases. Every new genome that is retrieved – whether of pathogens or the human (or animal) hosts that harboured them – expands the field’s ability to make comparative analyses. The most recent study of relevance is, at the time we are writing, still in development, but its survey of data can be referenced. For that study 1313 human remains from across Eurasia were studied, covering a span of 35,000 years up to the present day. In all those samples, thirty-nine Y. pestis genomes were retrieved (Figure 3). Yet when one looks at the database, among the mediaeval samples (of which there are many), there is not a single one that comes from the regions between the Caspian Sea and Egypt that make up the geographic area of our study. ³⁰ The same omission holds true for most other palaeogenetic studies published to date. In short, while we cannot claim any genetic proof for our arguments, neither can the absence of evidence carry any argumentative force for the simple reason that little pertinent data has yet been assembled.

Figure 3. Spatiotemporal distribution of Y. pestis in Eurasia. A map and timeline showing the spatial and temporal distribution of 1313 human remains included in an in-progress study of pathogens found in Eurasian populations, from the Pleistocene to the late Holocene. The coloured circles here indicate remains that yielded traces of Y. pestis; dating is colour-coded (see original for full colour).

Source: Sikora et al., op. cit. (note 30), Figure 3, Y. pestis. Historical notes added by the present authors.

In sum, neither plague nor Mongol historiography nor even the first decade and a half of the era of palaeogenetics have prepared us particularly well for the current moment in which we find ourselves. The lack of any sustained interrogation of the origins of the Black Death within the Mongol Empire by specialists in the field has allowed the unfortunate effects of ascertainment bias to overhype the significance of the Kara-Djigach study. The stalled trajectory of document-based plague history has delayed the creation of even a preliminary dossier of key sources in Arabic, Persian, or other languages of the eastern Mediterranean and western Asia (a topic we return to below). The uneven geographic distribution of archaeological studies that might yield remains from burial sites pertinent to the late mediaeval history of this central region compounds the problem on the side of physical data. ³¹ Be that as it may, recent events in global health have made it clear that ignorance of past pandemics is no longer justifiable. The rare trifectas of having documentary and archaeological and genetic evidence of the presence of Y. pestis for any particular historical place and time (which is what made the 2011 study of the London Black Death Cemetery so distinctive) cannot be held out as the standard that all plague histories can or must meet. ³² More archaeological and palaeogenetic evidence will certainly come to light, but it is likely, for the time being, to be random. In the meantime, historians need to set their sights on documenting and explaining other phenomena. As we have argued, noting anomalies in our sources is an excellent way to start.

Focalisation and the methodological demands of gaps between the biological and documentary archives

In the summer of 2019, public health officials in Colorado (USA) cancelled a camping event planned as a prelude to an outdoor rock concert. Yersinia pestis-infected fleas had been found parasitizing prairie dogs (small burrowing rodents) in the area where campers were expected. Yersinia pestis had arrived on the Pacific coast of California about 120 years earlier, but by the 1940s plague was well-established in various rodent communities in inland Colorado and other western states, where it persists to this day. ³³ This is the process of focalisation. No human infections occurred in Colorado in 2019, but that was because a public health surveillance system for plague was already in place. That is how plague remains controlled in almost all the areas where it persists in the world today.

‘Focalisation’ is a term that historians have recently adopted to explain a historical phenomenon of plague: its tendency to intrude into and then persist in certain regions for decades or even centuries, allowing it to manifest in human outbreaks periodically but then ‘disappear’ for years at a time. Obviously, plague never ‘disappears’ since it is a living organism and must continue to persist through cycles of reproduction. But as evolutionary and ecological perspectives on plague’s history get pushed further and further back into the past, the question has come to the fore of where plague has been able to persist long-term in ecological niches of rodent hosts and insect vectors, or indeed, in the soil itself. Where those niches exist away from human habitation, there will be no human plague outbreaks. But when those isolated niches are disrupted in one way or another and start establishing offshoots in places near human habitation, new possibilities open up for human infection. ³⁴

Plague ecology, the study of plague in its natural habitats among wild rodents, is directly a product of the bacteriological age, when the micro-organism, Y. pestis, could first be tracked in its movements between rodent and flea species. The concept of long-term foyers (hearths or, in Latin, foci) of plague began to be investigated by plague researchers in the late nineteenth century. By the 1950s, developments in the biochemistry of plague allowed a kind of crude differentiation of strains of Y. pestis, and a theory was proposed that connected these different strains with different pandemics of the past. Because the ecological character of Y. pestis was being discovered at exactly the same time as the Third Plague Pandemic was spreading new strains of Y. pestis around the world, ecologists (who coined the term ‘focality’) were able to watch the processes of focalisation in real-time as the plague spread to regions of the world where it had (apparently) never been present before: Madagascar, South America, and the Pacific coast of North America. Much of the most important field research on plague has been done in these newly infested landscapes. Although certain genetic features of the Third Pandemic strains may limit our ability to draw direct parallels between modern field and lab studies and mediaeval presentations of the disease, ³⁵ recognition of the overall genetic similarity of modern and mediaeval Y. pestis allows historians to engage in productive speculation about the climatic and faunal environments in which Y. pestis could have thrived in the past.

To return to our rock concert in 2019, fortunately the detection of plague was made early and although rock fans were disappointed to have their camping plans cancelled, no humans were harmed during what might have turned into an epizootic and even epidemic event. But had such a scenario presented itself, say, 800 years ago, what kind of historical evidence might we expect to find? One camper dies: bad luck. Three campers die: something’s wrong; others flee. A hundred campers die: it would probably only be in this last instance that we’d expect any kind of comment about the event in a local chronicle, and even here it might not rise to levels of higher political (and hence, administrative) importance. Yet in all these scenarios, Y. pestis would have already been present.

This is why, in addition to the concept of focalisation, we now add the concept of the ‘prodrome’ of the pandemic. ‘Prodrome’ is a medical term meaning ‘a symptom, sign, or condition that is characteristic or premonitory of the onset of a disease’. ³⁶ It refers to the state of a patient after they have already been infected but before the distinctive effects of the disease have become manifest: for example, the time between infection (via inhalation) of smallpox and the full eruption of that disease’s characteristic pustules. In our usage, the prodrome of a plague pandemic would mean that the pathogen is already present in a certain environment, just as it was in the Colorado case where the plague was found among the prairie dogs, but has not yet erupted into full-blown epidemics in human populations that would make its presence abundantly apparent. What we have yet to learn are the signs that indicate its ‘invisible’ presence in an age without microscopes or indeed any concept of microbes, sometimes literally under the feet of human observers. In the Colorado case, modern science and medical history have allowed us to reconstruct how Y. pestis came into Colorado’s rodent populations decades before the rock concert in 2019. What we have been lacking for the late mediaeval period is the cooperation of document-based historical investigations and those of the biological sciences to tell us a comparable story. The availability now of well over 100 genomes of Y. pestis from the period of the Second Plague Pandemic, from areas ranging from Central Asia to Iberia and Scandinavia, has opened up the possibility to think about focalisation in deeper time scales. Indeed, the ambitious study examining human remains in Eurasia from across a time span of 35,000 years, referenced above, supports what had already been an emerging impression of the history of Y. pestis: that it arrived on the scene as a human affliction about 6000 years ago, spread within human communities across Eurasia for about the next 3000 years with enough regularity to be called ‘endemic’, and then transitioned to a punctuated state where, after several hundred years of invisibility in the (human) archaeological record, it emerged three times into a ‘pandemic’ phase, moving quickly across landscapes and causing at times massive mortalities in human communities. ³⁷ Only this last, pandemic phase is captured clearly in our documentary records, which have in turn created fixtures in the scientific and historical literature of the past century as the First through Third Pandemics. Yet Y. pestis persisted throughout all the intervals in between. Establishing how to look for it in the absence of the descriptions of mass mortality we have been accustomed to expect (thanks to the memorable accounts of such writers as Procopius [d. 565], Ibn al-Wardī [d. 1349], Boccaccio [d. 1375] and many others) is the task ahead of us as document-based historians.

Perceiving plague’s presence: reading the past without a microscope

Our 2021 essay was precisely an exercise in how to sift carefully through earlier chronicles to catch references to epidemics, odd disease events or mass mortality events to see if they could reveal evidence of the plague’s prodromal stage. Robert Hymes had already found evidence for novel disease outbreaks associated with the Mongol sieges and conquests in early thirteenth-century China. Hannah Barker, as already mentioned, had refuted the Italian de Mussi’s siege of Caffa account of plague transmission and showed instead how grain shipments from the Black Sea region tracked plague outbreaks across the Mediterranean. Building simultaneously upon Hymes’ findings and her own careful reading of results in genetics, archaeology, environmental sciences, and the histories of China, the Mongol empire and Islamic societies, in 2020 Monica Green offered a possible scenario for how Y. pestis could have focalised into four different lineages in her article, ‘The Four Black Deaths’. Using references found in the Persian chronicle copied by al-Shīrāzī to both shipments of grain and epidemic outbreaks (wabāʾ) during Hülegü Khan’s conquests in Iran and Iraq (1256–60), Green identified contaminated grain from the Tian-Shan region as the possible way that Y. pestis may have been introduced unknowingly into new environments. At the same time, Nahyan Fancy identified how physicians in the middle part of the thirteenth century (1242–70) were addressing new symptoms of plague which seemed to be based on experience with the disease in his chapter, ‘Knowing the Signs of Disease’. Armed with a brief reference in Mona Hassan’s text to the outbreak immediately after the fall of Baghdad of an epidemic disease (wabāʾ) – which she mistakenly translated as plague – in 2021 we widened our search to identify any other outbreaks or more evidence about that particular outbreak. ³⁸ What we found was entirely unexpected and profound.

Multiple sources referred to a lethal disease outbreak in areas taken over by Hülegü’s armies or lands adjacent to them, such as Syria and Egypt. The latter two sequences of outbreaks, in particular, took place months after the fall of Baghdad. And the most detailed account of the outbreak in Syria included clear references to it being plague (ṭāʿūn). In their article, Brack and his colleagues deny the presence of plague in Syria, claiming that ‘there is nothing concrete’ and no clear symptoms present in the contemporary chronicler, Ibn Wāṣil’s Urtexte. ³⁹ They even suggest that Ibn Wāṣil’s use of the term ṭāʿūn to describe his patron’s death was ‘not on the basis of symptoms, but rather to assign’ his patron the merit of martyrdom. ⁴⁰ In our reading of the evidence, on the contrary, there are very concrete references to actual bubonic plague in Ibn Wāṣil’s text which concur with some of the new symptoms found in the thirteenth-century medical corpus.

Ibn Wāṣil (1208–98) was a Syrian historian and judge from Hama. In his chronicle, he provides details of his former patron, the Ayyubid ruler al-Nāṣir Dāwūd’s death from ṭāʿūn (plague) on the authority of the ruler’s son and an eyewitness attaché, connecting this specific incident to a widespread plague outbreak in the region in the aftermath of the Mongol conquest of Baghdad. Brack and colleagues have now provided the full text which we could not include in translation in our original because of space limitations. But they have not understood the text correctly. Significantly, they missed the fact that the text describes three delirious dreams – a symptom that is (re-)associated with plague in the Arabic medical tradition only in the thirteenth century. Plague patients across different societies historically have referred to some type of pricking, piercing, or pain in a particular site in the dream, and then woken up to find a plague bubo on that site. ⁴¹ Ibn Wāṣil records that al-Nāṣir Dāwūd dreamt that his left side was speaking to his right side, informing it that soon it would be the latter’s turn to be in pain. And sure enough, later the ruler started experiencing pain on the right. ⁴² Not recognizing the symptoms of plague, Brack and colleagues also err in their translation. First, al-Nāṣir complained of pain not ‘on his right side’ but rather ‘under his right side (taḥta janbihi al-ayman)’. And since janb refers to the part of the body extending from the armpit down to the flank, the translation really should read that al-Nāṣir complained of pain ‘in his right armpit’. Second, the caregivers did not ‘kn[o]w that this was ṭāʿūn’ but rather ‘ascertained/confirmed that it was a ṭāʿūn (taḥaqqaqnā anna dhālika ṭāʿūn)’. The primary meaning of ṭāʿūn is a bubo, and the disease is given the same name in Arabic based on this primary symptom. The clear use of the indefinite ṭāʿūn means that the caregivers confirmed for themselves that what was in his right armpit was a bubonic swelling. This is not an attempt to metaphorically refer to al-Nāṣir Dāwūd’s death as martyrdom from the plague, but rather his actual succumbing to plague after he first complains of a piercing pain in his left side on Wednesday, and then dies on Saturday from the disease. ⁴³

The reference to the right armpit takes on more meaning in light of other medical works from this time. As Fancy has shown, a key anomaly of the thirteenth-century medical discussions of Syrian authors, Ibn al-Quff and al-Sāmirī (both of whom composed their works after 1258), is that they restrict plague buboes to the armpits. We also know that al-Nawawī (d. 1277), a Syrian religious scholar writing after 1258, describes for the first time the existence of secondary cutaneous pustules, a key symptom of bubonic plague which is absent in all prior medical and religious literature. Focusing on post-1348 works, Michael Dols had mistakenly assumed that Ibn Khātimah (d. 1369), writing in Almeria in 1349, was the ‘only writer’ to distinguish clinically between the secondary pustules of the skin and buboes. Al-Nawawī not only records this clinical symptom but also associates the secondary pustules with the primary bubonic site of the armpits. ⁴⁴ This restriction of the bubo to the armpits is an anomalous feature of the post-1258 plague discussions stemming from Syria since this restriction is absent in the work of Ibn al-Nafīs’s earlier 1242 commentary, where the information on the plague is derived from his friend who visited Ethiopia. In the subsequent plague outbreaks of 1348, buboes were observed in all three primary sites again (neck, armpits, and groin), and pneumonic plague and the spitting of blood were associated with plague possibly for the first time. The memory of this restriction to armpits though is evident in the Mamluk author al-Ṣafadī’s (d. 1363) account of the plague of 749/1348–9 that he wrote and sent to the religious scholar and poet, Bahāʾ al-Dīn al-Subkī (d. 1372). Here, al-Ṣafadī provides a couplet for each of the symptoms, buboes behind the ears, under the armpits, and in the groin as well as the spitting up of blood. ⁴⁵ He clearly wanted to emphasise that buboes can break out in three different primary sites and not just one. Similarly, in the post-1348 commentary on the Canon of Medicine, al-Shīrāzī’s student, Āmūlī (d. 1352) too (re)-associates the three sites (armpit, groin, and the ear/neck region) with bubonic plague, despite citing his teacher’s claim that plague buboes are restricted to armpits. ⁴⁶

Having established that a plague outbreak had indeed occurred in Syria three months after the fall of Baghdad to Mongol forces, we used the earlier publications (including our own) to explain how the outbreak could have been caused by the Mongols unwittingly bringing Y. pestis into this region. We followed the reports of disease outbreaks following the arrival of the Mongols in the region, particularly those that took place after the Mongol conquest of Baghdad. Of course, our original article is careful enough to state that plague (ṭāʿūn) is almost exclusively used to describe the outbreak in Syria (especially the death of the ruler), while the outbreaks in Egypt, Iraq, Iran, and Anatolia generally use the term for epidemic disease, wabāʾ (Syr. mawtānā). However, it is worth emphasising that wabāʾ is the term used to describe the Black Death outbreaks of 1348 by at least two plague treatise writers: Ibn al-Khaṭīb (d. after 1374), from al-Andalus, and Ibn al-Duqmaq (d. 1407), from Mamluk Egypt. ⁴⁷ Even the famous Moroccan traveller, Ibn Baṭṭūṭa (d. 1368/9), who was travelling through the region during the summer and fall of 1348, used wabāʾ preferentially to ṭāʿūn (which appears only once in the entire text of his Riḥla). ⁴⁸ Similarly, mawtānā is the term used in Syriac chronicles to document the sixth and seventh-century outbreaks of the first pandemic ⁴⁹ and is inscribed on the Christian gravestones from fourteenth-century Central Asia. We now know from recovered aDNA that Y. pestis was the pathogen involved in the first and second pandemic, and also the one found in the 1338/9 Central Asian graves. This should open up the possibility of treating any mention of wabāʾ/mawtānā as plague when there are other indicators that Y. pestis may have been involved.

Having said that, we did not use the terms ‘interchangeably’ ourselves in our 2021 essay (it is why we referred to plague and plague-like disease). ⁵⁰ We knew (and documented) that our historical authors understood ṭāʿūn to be a type of wabāʾ, both caused by putrefaction of air such that a miasma stemming from one place, i.e. destruction of Baghdad, could be seen by our historical authors as causing a wabāʾ in Iraq, but intensify into a ṭāʿūn by the time it reached Syria. We also documented that the reports about the specific plague outbreak in Syria, and its connection to the fall of Baghdad, disappeared from the more widely-read fifteenth-century chronicles. Neither the Syrian plague outbreak, nor its connection to the fall of Baghdad is found in al-Shīrāzī’s work, so his account is not the ‘linchpin’ for our 2021 essay, as Brack and his colleagues mistakenly claim. ⁵¹ Of course, what leads them to investigate the timing of a specific outbreak in a city that was besieged by the Mongols are two aspects: (1) they are unfamiliar with disease history, in particular, how plague focalisation and ecology work; and (2) they are looking for primary sources to directly connect the arrival of the disease to the Mongols. As they state bluntly while reporting on epidemic disease outbreaks during Mongol sieges of forts and cities, ‘there is no evidence that the disease was brought by the besiegers’. ⁵² The fact that they are looking for evidence in our sources for the Mongol besiegers bringing a disease or ‘serv[ing] as vectors’ ⁵³ shows that they do not fundamentally recognize that there is no germ concept of disease in the thirteenth century (at least not in Islamic societies or the Latin West). And this is in fact what we stated in our article:

It should be stressed that for none of these authors, whether historians or physicians, were the Mongols seen as bringing the disease to the region (as we would now understand plague transmission in a biomedical framework). For Ibn Wāṣil, the anonymous author of Ḥawādith, and those who followed them, the Mongols caused an epidemic to erupt by the havoc they wrought, the dead bodies of the slaughter giving off smells that penetrated and corrupted the temperament of the ambient air. ⁵⁴ (emphasis added)

Brack and colleagues’ emphasis on whether or not a disease outbreak impacts a besieged city during or after the seige is lifted suggests an unfamiliarity with disease history, and especially the disease ecology of Y. pestis. Setting aside the discrepancy in the sources over the timing of the Baghdad epidemic (whether before or after the Mongols besiege the city), Brack and his colleagues claim that an epidemic outbreak even in the midst of long Mongol sieges in Lanbasar (1257), or even Girdkuh (1253), does not suggest the possibility of a ‘disease … brought by the besiegers’. ⁵⁵ They feel only if the outbreak were to take place immediately following the end of the siege would that somehow suggest the possibility of Mongols bringing the disease.

Obviously, lacking genetic evidence we cannot yet say what strains of Y. pestis might have been involved in any of these outbreaks. It is certainly possible that, as has been documented for the Caucasus, much older strains of Y. pestis had survived in the regions the Mongols passed through and these outbreaks (assuming they are in fact plague) were caused by disturbances of pre-existing reservoirs. ⁵⁶ However, Green already documented, in 2020, using the account found in the Persian chronicle copied by al-Shīrāzī, that the Mongols imported their own grain supplies from the Tian Shan region – precisely the ecological circumstance that would allow the importation of rodents and their fleas. As Green noted with respect to the Baghdad siege,

This would also account for a further detail that al-Shīrāzī adds. It was not simply the besieged inhabitants of the city who became ill, but also the Mongol army itself: ‘and many died of the disease.’ Among those stricken (though he later recovered) was Hülegü himself. ⁵⁷

Given that Hülegü and his troops arrived at their base camp at Hamadan in 1257 (after having spent a somewhat erratic nine months regularly moving their base camp), ⁵⁸ and given that, according to the chronicle copied by al-Shīrāzī, the grain supplies had been placed along the army’s path in advance of their arrival, if indeed there were any sacks of contaminated grain in those supplies, they would have already been present in the region well before the attack on Baghdad in January 1258. As we have been insisting, plague is a zoonotic disease, whose hosts are rodents and whose vectors are fleas. Neither rodents nor fleas are prevented from crossing into cities by means that generally prevent or slow down considerably the movement of humans. So, if Y. pestis has arrived outside the city walls, the siege does not need to be lifted in order for the besieged city to suffer a possible outbreak.

Moreover, once the bacterium is introduced into a new environment, it needs to find appropriate hosts and vectors. The mere introduction of the bacterium does not necessarily entail a massive epizootic or human outbreak, since the bacterium needs to be able to proliferate in a sustaining animal population (rodents or other animals) and needs sufficient vectors (like fleas or lice) to transmit it into humans. ⁵⁹ Which conditions are met at which time in which place are not easy to predict nor historically reconstruct, especially from sources that do not even associate the disease with germs, fleas and rodents. Additionally, we need to be aware of the biases of our sources. The contemporary chronicles that we used are not going to be concerned with deaths amongst low-ranking Mongol soldiers as much as they are with the vast population centres of the Islamic world, and occasionally some high-ranking Mongol officials. Given these constraints and biases, massive mortality from disease can be recorded for besieged cities even while the siege is ongoing with negligible reports on outbreaks amongst the besiegers. And given how the disease functions (i.e. that it needs vectors to transmit the bubonic plague from one person to another and is not otherwise directly contagious), it is not at all surprising that people could have travelled to Baghdad to enrich themselves upon the misfortune of those who died from it without contracting the disease. ⁶⁰

If our reconstruction of the circumstances of the arrival of Y. pestis in the region is correct (that is, it was brought to Iraq and its environs in Hülegü’s grain supplies, arriving in or near Hamadan – Iraq al-ʿAjam – no later than 1257 and moving further westward thereafter), then the progression of the reported human outbreaks follows logically. We have evidence for a very clear outbreak of plague in Damascus three months after the Mongol destruction of Baghdad, in addition to other reported disease outbreaks in Egypt, Iraq, and Anatolia all taking place months after the fall of Baghdad. We see a continuing engagement with plague in medical works, religious texts and chronicles throughout the second half of the thirteenth and early fourteenth centuries. The fact that later chroniclers were relying on Ibn Wāṣil or other first-generation chroniclers is a fact we acknowledged in our original article, so the discussion by Brack and colleagues on Urtextes is unnecessary since a glance at our table reveals the dependencies. As for the Ibn Furāt (d. 1405) text that they provided, it may or may not hearken back to a work that was available to earlier historians. However, it should be obvious that Ibn Wāṣil provides testimony for a plague outbreak in Syria on the authority of those in the entourage of his former patron, al-Nāṣir Dāwūd, who died of plague. He explicitly links the Syrian outbreak using a miasmatic explanation with the destruction of Baghdad, since he begins the account of al-Nāṣir Dāwūd’s death by stating: ‘When the epidemic and plague intensified following the Mongol conquest of Baghdad’. ⁶¹ His linking was accepted by later scholars until the Syrian plague outbreak itself disappeared from the accounts of the fifteenth-century Mamluk authors, al-Maqrīzī, Ibn Ḥajar al-ʿAsqalānī (d. 1449) and Ibn Taghrībirdī (d. 1470), that have informed modern scholarship. Speculating over whether or not Ibn Wāṣil himself was relying on a further unidentified source is a moot point and does not change what we found. After all, given that Ibn Furāt does not cite sources in this passage but otherwise is, according to Brack et al., ‘scrupulous’ about that, ⁶² it is just as possible that he is the one piecing together information from the works of well-known, earlier Mamluk historians and religious scholars, such as al-Dhahabī (d. 1348), and al-Yūnīnī (d. 1326), rather than relying on some lost, original work.

The case for plague’s presence in the Western Ilkhanate in the late thirteenth and early fourteenth centuries

Plague-related concerns and mass mortalities did not disappear after 1258. The concern for plague is most evident in the Muslim convert, physician, and vizier at the Ilkhanate court, Rashīd al-Dīn (d. 1318). In his introduction to the translation of a Chinese medical work into Persian at the Ilkhanid court, Tānksūq-nāma, Rashīd al-Dīn praises Chinese medical techniques and their ability to treat hot diseases. He gives three examples, all of which entail swellings and fever, and for the first and the third, delirium is also a symptom: birsām ⁶³ (inflammation of the diaphragm), dhāt al-janb (pleurisy), and ṭāʿūn (plague). ⁶⁴ In other words, plague is listed alongside two very well-known and regularly discussed (and encountered) diseases across mediaeval Islamic medical literature. Such matter-of-fact references to ṭāʿūn suggest plague was a disease to be feared and reckoned with in post-1258 Ilkhanid Western Asia. The outbreaks may have been small, though we would have to sift carefully through all the various records from the intervening years till 1348 to be sure. After all, the Syrian plague outbreak had dropped out of memory, too, until recently. Moreover, references to major epidemic disease outbreaks in the region in the intervening years exist. The Coptic historian Jirjis ibn al-ʿAmīd al-Makīn (d. 1293) describes an epidemic outbreak (wabāʾ) in Mardin in 658/1260, which occurred after a terrible siege of the city by Hülegü. In 694–5/1295–96, al-Birzalī (d. 1339) reported an epidemic (wabāʾ) in Cairo and Lower Egypt which killed up to 1500 people a day and upwards of 100,000 in total. ⁶⁵ Intriguingly, this Cairo outbreak is mentioned by the Andalusian scholar and poet who settled in Cairo, Ibn Abī Ḥajala (d. 1375) in his plague list, but is left out of the subsequently more influential plague list of Ibn Ḥajar. ⁶⁶

This latter incident also points to another matter already attested by one of the contemporary observers discussed above. Brack and colleagues describe Bar Hebraeus as saying that ‘there was a very severe famine and a pestilence (māwthānā) in the land of Babylonia (Senʿār), Assyria (Āthūr), Mesopotamia (Bēth Nahrīn), Syria (Sūrīya), and Anatolia (Bēth Rūmaye), so a young pigeon for a sick man was sold in Damascus for 12 Nasiri dirhams’. ⁶⁷ This linking of famine (Syr. kepnā marrīrā) with pestilence became a very common element of catastrophic mortality events reported in Anatolia, Egypt, and even Europe, up to and including episodes considered Black Death events. In looking for epidemics that have plague-like mortality effects, a comprehensive history of this postulated prodromal stage of the Second Plague Pandemic under the Ilkhanate would look at changes in the landscape environment of the region, including signs of economic decline.

That task involves looking for factors that we know have contributed to the plague’s eruption and persistence in other times and places. We already know that the region around the Caucasus Mountains was ripe for plague infestation. It still harbours living strains that must have emerged in the Bronze Age; palaeogenomes of even older strains have been retrieved from Bronze Age sites; and one sub-branch of another product of the later mediaeval Big Bang, Branch 2, has colonized the entire area around the Caspian Sea in modern times. ⁶⁸ Which, if any, of these strains were involved in the extinction of the marmot populations that used to inhabit mountainous areas to the south of the Caucasus is unclear, though populations of the animal were still numerous enough in the later thirteenth century to induce Hülegü’s son, Abaqa (r. 1265–82), to build a new palace at the Mongol summer camp between Lake Urmia and Hamadan because of its excellent hunting grounds. It was called Sughurluq, ‘the place abounding in marmots’. ⁶⁹ How many other species of rodents and insects may have been involved in the plague’s new focalisation in the region, we can only guess.

Rather than dismissing references to bad harvests and infestations of mice as unrelated to the plague (as Brack and colleagues do), we should instead recognize them as conducive to the plague’s focalisation. ⁷⁰ Bad harvests, whatever their cause, will put stress on rodent populations that rely on gleanings as a source of food. Changes in rodent movements will also mean changes in flea transmission. Once plague has entered a terrain, any new environmental stressors can change a relatively stable enzootic plague infestation into an epizootic one.

As for the reasons for the erasure of local epidemics from the historical record, we acknowledge that more research is needed. We have already stressed that our modern understanding of plague in the thirteenth and fourteenth centuries has been shaped by the later fifteenth-century historians, who were living in a post-Black Death world, witnessing major outbreaks of plague across many regions regularly. Michael Dols’ 1977 narrative of the Black Death in the Middle East was built almost entirely from these late witnesses. That they should have frequently used the more clinically precise term ṭāʿūn (plague) is hardly surprising, given the ubiquity of the disease by the time they were writing. Indeed, already by the 1360s a qadi (judge) in the Levantine town of Safed was drawing on medically precise terminology to describe the symptoms of ṭāʿūn. ⁷¹ Nevertheless, post-Black Death writers continued to use the term wabāʾ for what clearly must have been plague outbreaks. If we are serious about investigating plague outbreaks during the prodrome of the pandemic as well as its crescendo in the 1340s, we need to be alert to the fact that not every human outbreak will be widespread, and not every outbreak would have provided real, trustworthy eyewitness reports of the actual course of the disease.

There is, finally, an additional issue: the failure to mention epidemics in a way we readily recognize. Here it is necessary to weigh not simply the likelihood that human outbreaks during the plague’s prodromal stage may have been very small, but also the possibility that they were not seen as a ‘disease’ at all but some other malign force. Based on what we currently know about the geographic distribution of the pre- and post-Big Bang lineages of Y. pestis, it is possible that the Mongols had no long-term experience with plague prior to their expansion out of Mongolia and may not have had a pre-existing vocabulary to deal with it. ⁷² The speed with which the disease kills would have been shocking and they conceivably would have tried to understand this experience within their existing shamanistic frameworks. In three studies published in the past decade, the Bulgarian researcher Maximin Golev describes episodes of witchcraft accusations in the Mongol realms. Golev never makes any connection with plague, and only obliquely mentions disease as being a factor in one case. (In another case, the suspicious deaths are talked about as ‘poisoning’.) But in all cases, the reactions are extreme, involving the killing of the alleged protagonists and the execution or chasing away of a large number of the protagonists’ followers or associates. In other words, in no case are these responses small or insignificant. ⁷³ One involves Hülegü chasing Jochid army members through the Caucasus: exactly the route that the plague travelled if we could assume that the plague evidence for Baghdad, Syria, and Anatolia is the same lineage that the geneticists have documented in Black Death-related sites from the Volga region. ⁷⁴

As we have indicated above, at the present moment genetics cannot yet confirm any particular narrative about when or via what route the strains of Y. pestis involved in the Black Death reached western Eurasia and North Africa. ⁷⁵ Yet as Monica Green and André da Silva have recently suggested, there is mounting evidence that some new strains reached Europe before the Black Death as conventionally dated (that is, for Western Europe, 1347). ⁷⁶ One example will suffice. In Figure 2B above, we flagged that the Bavarian town of Nabburg had yielded a Y. pestis genome that seemed to be evolutionarily earlier than that retrieved from the Kara-Djigach cemetery in Kyrgyzstan, dated to 1338–9. Although referred to by the researchers who originally sequenced it as a ‘Black Death’ genome, the Nabburg remains have only been dated radiologically, which produced the broad range of 1292–1392. ⁷⁷ The nearby town of Amberg also has mass burial graves; these have not yet been tested for the presence of Y. pestis nor have they yet been radiologically dated, but the burials themselves show consistent signs of a severe cultural unease elicited by the manner in which these individuals died, which was almost certainly an epidemic since no signs of violence are apparent. In the case of Amberg, these abnormal burials are clearly the earliest in the site, which can itself be dated to the time of the hospital’s founding in 1317. ⁷⁸ A comprehensive assessment of this Bavarian evidence and other documentary and bioarchaeological evidence for the plague’s possible presence in pre-Black Death Europe is underway. If these initial findings hold, then this will add weight to our general contention that strains now identified as part of Branch 1 of the Y. pestis phylogeny did indeed pass through both western Asia and Europe (and probably North Africa, too) several decades before the outbreak at Kara-Djigach in 1338–9.

It remains, of course, to explain why the Black Death wave of the plague would have been so catastrophically worse, with a far greater range of transmission, than any ‘prodromal’ phase we have yet been able to document. That is a question we hope to address in future work. For the moment, we would emphasise how much remains unknown about the crucial events between the 1330s and the mid-1340s. Beyond looking for the bacterial strains, insect vectors, and rodent hosts involved, we have yet to gather all the evidence in plain sight. Michael Dols, for example, writing in 1977, asserted that the earliest mention of plague in Armenian sources comes from 1348. In fact, he overlooked the clear references to ‘pestilence’ and ‘plague’ (in Sanjian’s translation) in Armenian colophons from 1321, 1329, and 1337. ⁷⁹ The earliest of these references came from the hand of a scribe from Ani, the Armenian city whose elites succeeded in accommodating well to Mongol rule, only to see the city collapse into financial (and it seems, demographic) ruin by the 1330s. ⁸⁰ The city has lain in ruins ever since. Indeed, there are many puzzles about the collapse of both the Ilkhanate and trade in the region that have never been satisfactorily explained. Philip Slavin, arguing for a ‘quick transit’ of plague to the north of the Caspian Sea, himself notes that the first chronicle reports of human plague mortality come from 1343 (two years before other reports from the Golden Horde) and these are from the southern shores of the Caspian Sea, not the north; he declined to offer an explanation. ⁸¹ It is unlikely that any single piece of evidence – even a newly sequenced genome – will settle a question so complex as the processes that led up to the events of the 1340s. We may need to look for several foci and several separate epidemics, rather than one single epidemiological eruption. The Black Death was the crescendo of slowly amassing forces: climatic as well as faunal and bacterial. The events of 1258 are only part of this much more expansive story.

Conclusion

We thank Drs Brack, Biran, and Amitai for engaging with our 2021 essay, and with some of our other work on the history of plague which, as they know, has been many years in development. ⁸² We have argued here, however, that they have overlooked our larger body of work and our ambitions in the history of medicine. Our intention has always been to understand the historical events that led to the dispersal of the bacterium from the Tian-Shan mountains and its focalisation in regions much farther away, giving rise to four distinct lineages, at least one of which (Branch 1) wreaked havoc upon major population centres in Western Eurasia for centuries. As stated by Green in 2020, ‘this is not about assigning “blame.” It is about documenting humans doing what humans do’. ⁸³ It is also, increasingly, about documenting what bacteria, insects, animals, and even the earth itself do. We agree that there is room for valuable dialogue in reconstructing the textual documentation for this vital period in human history and we are as committed to philological rigour in these endeavours as Brack and colleagues are. Indeed, we are quite excited about the prospects of renewed attention to medical-historical questions made possible by long-standing and new digital humanities projects focusing on the mediaeval Islamicate world, which we have listed in the attached Appendix. ⁸⁴ Such resources helped us discover an error in our original essay in citing the Syrian, Mamluk historian, Ibn Shākir al-Kutubī’s (d. 1363) chronicle; this, in turn, opened our eyes to the likelihood that there was yet another unexplored wave of plague prior to the Black Death, in this case in 1057–8 in Samarqand and central Asia as recounted by the Ayyubid preacher and historian, Sibṭ ibn al-Jawzī (d. 1256). ⁸⁵

Of course, for both these unexplored waves of plague (or ‘plague-like mortality events’) in the eleventh and thirteenth centuries, we acknowledge that currently there is no palaeogenetic evidence for Y. pestis having been present in Iran, Iraq, Syria, or Egypt. But that is in fact true for most infectious diseases, for the simple reason that very little sampling has yet been done in this region. ⁸⁶ Hence, saying categorically that something did not happen more than 700 years ago, when that ‘something’ would have been happening at the level of microbes, insect vectors, and small rodents, seems more absolutist than is wise at the present juncture. As we documented in our original essay, four thirteenth-century contemporary observers described actual epidemic outbreaks that occurred after the Mongol conquest of Baghdad in 1258, to which we can now add a fifth, Jirjis ibn al-ʿAmīd al-Makīn. What none of them commented on – nor should we reasonably expect them to comment on – were changes in the larger environment. Climate history has overwhelmingly documented the global effects of the 1257 Samalas volcanic explosion (in Indonesia) in terms of sulphur emissions and temperature changes; yet we are only beginning to learn how to perceive its effects on human actors and natural landscapes in subsequent years, the most volcanically active century of the Common Era. ⁸⁷ Changes in flea (or lice) ecologies may be perceptible when scribes complain about their ‘writing companions’, but these sources, too, demand careful consideration within the constraints of their genres and vocabularies. ⁸⁸ Our expectation is that additional evidence for the dramatic climatic and environmental changes in the thirteenth and fourteenth centuries will continue to accrue, and as these physical changes are more precisely dated, we will be better able to link them to political, economic, and social changes. We also expect that additional evidence will document what we have proposed as a prodromal phase of the Black Death pandemic. In the meantime, we look forward to increased dialogue and methodological understanding among all the scholarly disciplines that contribute to these investigations. And we look forward to an openness in addressing our ethical obligations in bringing balance and care to how we talk about the epidemiological challenges that past societies faced – including, as we have demonstrated, the Mongols themselves.