That Historical Plague Was Pure Epidemics of Primary Pneumonic Plague

Home , Easterly Winds, List of epidemics

CHAPTER THIRTEEN

GUNNAR KARLSSON’S ALTERNATIVE THEORY: THAT HISTORICAL PLAGUE WAS PURE EPIDEMICS OF PRIMARY PNEUMONIC PLAGUE

Introduction

G. Karlsson presented his alternative theory to the international community of scholars in a paper published in Journal of Medieval History in 1996 where he argues that two late medieval epidemics in Iceland and more generally that plague in medieval Europe were pure epidemics of primary pneumonic plague.1 For his epidemiological interpretation, Karlsson bases his alternative theory directly on Morris’s assertion of the occurrence of pure epidemics of primary pneumonic plague, a modality of plague disease spread by interhuman cross-infection by droplets. Th e concept of a pure epidemic of primary pneumonic plague implies that the origin of the epidemic is not a case of bubonic plague which develops secondary pneumonia but a case in which the fi rst vic- tim contracted pneumonic plague directly by inhalation of infected droplets into the lungs. Karlsson’s paper follows quite closely and draws heavily on a paper he published together with S. Kjartansson in an Icelandic journal in 1994 on two supposed plague epidemics in Iceland in 1402–4 and 1494–5 respectively.2 He is clear about its objective: “Th e present article [as the fi rst in Icelandic] can be seen as a defence of the late Jón Steff ensen against Benedictow’s critique of his conclusions,” namely, that these epidemics, and also the Black Death of 1348–9 in Norway were primary pneumonic plague.3 Th us, the central feature of Karlsson’s (and Karlsson’s and Kjartansson’s) paper on the two supposed plague epidemics in fi ft eenth-century Iceland is a comprehensive and sharp criticism of my doctoral thesis (1993, repr. 1996). Th e apparent objective was to clear the way for his own theory that the epidemics were

1 Karlsson 1996: 263–84. 2 Karlsson and Kjartansson 1994: 11–74. 3 Steff ensen 1973. 40–55.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 494 chapter thirteen pure epidemics of primary pneumonic plague: the epithet “pure” deter- mines that they were without any basis or origin in fl ea-borne plague from rats (or other rodents) or cases of human bubonic plague.4 Th us he thinks that the contagion was the bacterium Yersinia pestis, as in the case of bubonic plague, but the process of dissemination was interhuman cross-infection by droplets and not transmission by rat fl eas at any stage of the epidemic process. Th is immediately presents the insurmountable problem of how primary pneumonic plague could arise in the complete absence of bubonic plague and in the complete absence of rats, a problem he surprisingly does not identify or try to explain (and the editors and consultants of the Journal of Medieval History did not fi nd signifi cant). Alternatively, he might have launched a theory of importation, which he conspicuously fails to do and, as we shall see, for the very good reason that this would disclose another insurmountable problem, which is that there is no place abroad whence plague in any form could have been imported into Iceland in these years (another problem which the editors and consultants of the Journal of Medieval History did not fi nd signifi cant). Th e Icelandic epidemics exhibit such peculiar features that Karlsson, as the fi rst of the advocates of alternative theories, has to invent some suitable mutations in order to defend his theory. He does not consider that his assertion as to the occurrence of these mutations can move from the status as arbitrary or speculative to the status as tenable (at any level of validity) only by meeting some elementary methodological requirements, namely to explain to the scholarly community: (1) why and how and by which process of evolutionary selection this mutant strain or biovar came into being; (2) when and where, in Iceland or abroad, this mutant strain of primary pneumonic plague originated and developed; (3) whether this mutated variant miraculously came into being twice; and (4) where on earth (literally) it had been in the meantime (another problem which the editors and consultants of the Journal of Medieval History did not fi nd signifi cant). Undaunted by these formidable problems, Karlsson expands the perspective of his theory from its tiny and isolated Icelandic territorial base in the middle of the North Atlantic to argue that historical

4 Karlsson 1996: 265.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 495 plague in Europe generally was pure epidemics of primary pneumonic plague. On this basis, Karlsson concludes triumphantly that by having removed the black rat from the medieval plague he has more generally paved the way for asserting that the medieval plague epidemics in general, all over Europe, were primary pneumonic plague, presumably caused by that unknown mutated variant. Th ese assertions will be addressed in this chapter. Th e principal interrelationship between the advocates of alternative theories is the usual one in this case: that they accept his view that there were not rats in Iceland and that the fi ft eenth-century epidemics consequently could not have been rat-borne, which suits their alternative theories, but they fl atly reject his theory that these epidemics or the medieval plague epidemics generally were (pure) primary pneumonic plague.5 Th e gist of these counterarguments is, thus, that there never was a plague epidemic in Iceland, neither bubonic nor pneumonic, and Karlsson’s assertion that there were not rats in Iceland off ers an excellent explanation. Th e obvious implication is that the severe epidemics in Iceland were another epidemic disease or other diseases. To my knowledge, no other scholar has supported Karlsson’s view, at least not outside the Icelandic scholarly community where opinions apparently diff er sharply.6 One should note that there is a remarkable qualitative diff erence between the way Cohn and Scott and Duncan argue their rejections of Karlsson’s theory, and in my opinion, only Cohn makes his case on this point in a scholarly and scientifi cally tenable way.7

Karlsson and Benedictow

I consider Karlsson’s reading of my thesis so bizarre and skewed that it is, in my opinion, grossly misleading and I hope I may be excused for considering it an expression of the weakness of his case.8 In my

5 Cohn 2002: 23, 51; Scott and Duncan 2004: 182. See also Twigg 1984: 68, 161–8. 6 Karlsson’s (and Kjartansson’s) paper’s were discussed at a conference in Iceland and met with much criticism; papers from this conference are published in Sagnir 1997. 7 Cohn 2002: 22–3. 8 Th is is even more the case for the original paper in Icelandic, Karlsson and Kjartansson 1994, but since very few read Icelandic or Icelandic scholarly journals, this is not important in this context, only to the Icelandic scholarly community. Also the off ensive comments on my work can, therefore, be ignored.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 496 chapter thirteen thesis, I presented a detailed account of the scholarly literature on primary pneumonic plague, however, to no avail so far as Karlsson was concerned. Th is highlights another feature of Karlsson’s paper: although he argues very strongly for a theory to the eff ect that the Icelandic epidemics of 1402–4 and of 1494–5 were pure primary pneumonic plague and ends up with a triumphant conclusion with a much wider territorial sweep, linking the nature of “medieval plague” to this disease, he has not made any attempt to acquire scholarly knowledge of primary pneumonic plague. Not a single medical study on primary pneumonic plague can be found in his footnotes. According to traditional academic principles, this ought to mean that he cannot discuss the disease that constitutes the cornerstone of his theory with the competence that only familiarity with the primary research studies and standard works on this disease can provide. Th e reason for this extraordinary approach by a scholar is presumably that the Icelandic epidemics exhibit many epidemiological characteristics or features that are not compatible with the studies on primary pneumonic plague. Th is is where wishful thinking about mutations is introduced into his arsenal of arguments. For the same reason, he cannot avail himself of the fact that I provided a comprehensive summary of all primary studies on primary pneumonic plague and the presentation of this disease in the standard works in my thesis. My endeavours to track down and present the gist of these studies does not earn me any praise, it represents, on the contrary, a signifi cant problem that has to be managed. Since Karlsson cannot use the corpus of scholarly works on primary pneumonic plague in order to form a basis for his theory and likewise cannot use my complete presentation of it, he must fi nd a way out of this problem. His solution is to base his arguments on a secondary paper which is profoundly fl awed, namely Morris’s sharply critical review article of Shrewsbury’s monograph where Morris was the fi rst scholar to launch a theory that the Black Death was an epidemic of primary pneumonic plague. Karlsson starts in this way: “Other diseases [than bubonic plague] are hardly considered by him [Benedictow], and all other forms of contagion are excluded, mainly because they are said to be insuffi cient to explain epidemics of the dimension under discussion here.”9 Firstly,

9 Karlsson 1996: 264.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 497 one should note the logical and factual contradiction in the opening part of the statement, that “other diseases are hardly considered by him” and that “all other forms of contagion are excluded:” since all other forms of contagion unavoidably relate to many other diseases, I cannot have excluded other forms of contagion without having discussed other diseases.10 Secondly, he fails to mention that I made a complete presentation of the corpus of primary research on primary pneumonic plague, and this omission allows him to present a spurious explanation as to why I reject his theory. Th irdly, one should also note that in the second part of the statement he sharply criticises my approach, saying that I exclude all other forms of contagion “mainly because they are said to be insuffi cient to explain epidemics of the dimension under discussion here.”11 Th us, the truth shows through, albeit indirectly and involuntarily, that I have discussed other alternative diseases based on other forms of contagion, but have found that they have, inter alia, insuffi cient mortality-generating properties. Notably, Karlsson acknowledges in his paper that the Icelandic annals do not contain any clinical or diagnostic information that would permit the microbiological identifi cation of these Icelandic epidemics. However, he contends that these chronicles provide information that can be used for producing estimates of the population loss in both epidemics, and that the extreme estimated mortality level could only have been caused by primary pneumonic plague.12 In other words, in prac- tice he avails himself of exactly the same approach that he criticizes me sharply for using (although I do not use it as a fundamental argument but only as an additional argument). Later in Karlsson’s paper, there is another sudden but indirect indication that he has read my presentation of primary pneumonic plague. Th is occurs with a parenthesis within a citation from Morris’s paper to the eff ect that the great pneumonic epidemic in Manchuria would have spread much more, “but for the heroic counter-measures taken by the great Chinese

10 Benedictow 1993/1996: 121–5, 214–27. 11 Benedictow 1993/1966: 16, 125, 266–74. As this statement is inaccurate and could produce misconceptions about my view, I would like to point out that I have discussed not only the size of the epidemics, as Karlsson may be taken to state, but also properties that combine tremendous powers of spread with tremendous lethal powers, and I argue explicitly why that is a rare combination (see also above). 12 Karlsson 1996: 265, 268–84.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 498 chapter thirteen

doctor Wu Lien Teh [one of Benedictow’s greatest authorities on plague] who most fortunately was put in charge.”13 However, Morris continues, unquoted by Karlsson: Wu Lien-Teh’s classic treatise on pneumonic plague is signifi cantly absent from Shrewsbury’s very extensive bibliography.14 Th is treatise is also “signifi cantly absent from” Karlsson’s few footnotes. One could also note the distortion inherent in Karlsson’s combined reference to me and “Wu Lien-Teh [one of Benedictow’s greatest authorities on plague],” since Wu Lien-Teh is the indisputable and uncontroverted towering expert on primary pneumonic plague, not on “plague.” Morris’s aim was to reject Shrewbury’s theory to the eff ect that the Black Death and later plague epidemics in Britain could have caused only a minor part of the national mortality rate, about 5 per cent, and that exanthematic typhus, smallpox and other serious epidemic diseases caused most of the mortality (see above). Because Morris has insuffi cient knowledge of bubonic plague to eff ectively counter Shrewsbury’s arguments, he launched another and apparently relevant alternative theory, namely that plague in Britain was primary pneumonic plague. However, he had poor knowledge also of primary pneumonic plague, and many of his central assertions are unfounded. Th is subject is discussed below aft er the substance of Wu Lien-Teh’s and other researchers’ studies on this modality of plague have been presented, in order to perform a systematic comparison between these studies and Morris’s and Karlsson’s assertions. Karlsson bases his paper and theory on Morris’s ten-page review of Shrewsbury’s monograph instead of on the primary research studies on primary pneumonic plague because they in important respects have the same ambition, namely to prove that the medieval plague epidemics, at least in Britain and in Iceland, were pure epidemics of primary pneumonic plague. However, Shrewsbury’s rejection of primary pneumonic plague as a signifi cant causal factor of mortality is in full accordance with all primary research on this modality of plague, the same research that persuaded me to reject in my thesis the theory that the Black Death or subsequent plague epidemics could have been primary pneumonic plague. Apart from the fact that Morris referred to

13 Karlsson 1996: 283. 14 Morris 1971: 208–9.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 499

Wu Lien-Teh in his text and mentioned his monograph in a footnote, it can hardly be demonstrated that he has read his work or knows more about primary pneumonic plague than Karlsson. Karlsson based his case on the quotation from Morris’s paper for two crucial reasons: (1) because it provided an impression that primary pneumonic plague has great powers of spread, but in this case was halted by the countermeasures introduced by Wu Lien-Teh; (2) because Morris introduced the concept of pure pneumonic plague, which Karlsson makes the cornerstone of his theory. Both assertions are, as we shall see, at complete variance with Wu Lien-Teh’s presentation of primary pneumonic plague, where it is emphasized that epidemics of this disease arise from cases of bubonic plague which develop secondary septicaemia with subsequent lung con- solidation and a bloody cough spreading infected droplets. Th us, Wu Lien-Teh maintains that epidemics of primary pneumonic plague never are pure, never arise de novo from a fi rst or original case of primary pneumonic plague droplet infection. Th e heart of the present monograph is the identifi cation of epidemic diseases of the past. Th is objective requires that some elementary methodological rules are followed in order to ensure the validity of identifi - cation of diseases, otherwise any mention of a severe or disastrous epidemic in the sources can be arbitrarily taken as evidence of plague (or any other serious communicable disease). An illustration of the consequences of an approach unguided by methodological considerations is provided by Scott and Duncan who assert that “there are said to have been 233 outbreaks [of bubonic plague] in China between AD 37 and 1718.”15 Th ey have used Wu Lien-Teh’s general list of epidemics in China, for some reason leaving out the fi rst of 224 BC.16 However, Wu Lien-Teh does not at all suggest that all these epidemics were or could have been bubonic plague. He emphasizes that it is possible to identify only a couple of these 233 epidemics as bubonic plague. He has not succeeded in fi nding clinical or epidemiological evidence of bubonic plague in chronicles or literary accounts, only in three medical works. Two of these medical works were from the fi rst half of the seventh century; their clinical descriptions are unambiguous but cannot be

15 Scott and Duncan 2001: 171. 16 Wu Lien-Teh 1936a: 43–51; Benedictow 2004: 40–2.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 500 chapter thirteen correlated or associated with any specifi c epidemic(s) in China. Only in a medical work of 1642 is a clinical description of bubonic plague given within an epidemic context, obviously referring to the epidemic of 1641–2.17 Identifi cation of an epidemic disease, in this case as plague (either bubonic or pneumonic), requires specifi c evidence of at least one defi n- ing epidemiological or clinical feature or a combination of features that are unique (see above). In weak or inconclusive cases, it must be possible to demonstrate at least a possible link of dissemination from a known epidemic of plague to the epidemic under study which may allow formation of a hypothesis to the eff ect that it may have been plague. Th is does not, of course, constitute any proof that it actually was the plague; the function of a hypothesis is to legitimate research and direct the search for corroborative evidence. Th is requirement is of particular importance in the case of epidemics in a highly isolated island like Iceland. Th is clarifi es the basic methodological, empirical and source-critical requirements that Karlsson should meet in order to justify his theory of the aetiology of the purported fi ft eenth-century plague epidemics in Iceland. Th e Icelandic sources that provide information on these epidemics consist mainly of brief accounts in two chronicles for each epidemic; translations of these accounts into English are supplied in my thesis and by Steff enson. My translation into English is provided in Appendix 2 below.18 As is most oft en the case with this type of source, the quality of the information they provide is quite defi cient (in the eyes of modern scholars). Th ey have nothing to say about whence the contagion arrived in Iceland or about defi ning or weaker indicative clinical features, excepting perhaps on the typical duration of the illness. Th ey contain some comments on spread and mortality that are incidental and sensational in character as would be expected from this kind of source. Since Karlsson asserts that this “information” can be used for

17 Wu Lien-Teh 1936a: 10–3; Benedictow 2004: 35–42. Th is means that also McNeill’s use of the same list is unfortunate, and his choice of a Chinese epidemic of 1331 as the origin on the Black Death that subsequently spread all the way to the Crimea is completely unsupported by clinical or epidemiological evidence; see McNeill 1979: 152–4. Instead, McNeill let himself be victimized as a scholar by “the overwrought imaginings and hopelessly inaccurate quantifi cation of the chroniclers,” to cite yet again Hatcher’s apt description (1977/87: 21). His theory is also contrary to a number of other empirical facts, for instance, the fact that the communication lines between China and the Crimea were severed long before the Black Death broke out in the Crimea or Constantinople. See Benedictow 2004: 44–51. 18 Benedictow 1993/1996: 211–2; Steff ensen 1974: 457, 50–1.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 501 estimates of mortality which are indicative of plague, this point will be discussed below. An intriguing question presents itself immediately and with great urgency: how can it be possible in such evidential circumstances to identify with a (very) high level of certainty the aetiology of the fi ft eenth-century Icelandic epidemics, in this case specifi cally as pure primary pneumonic plague caused by some mutant variant of the contagion Yersinia pestis? Th is problem is exacerbated by the fact that in his paper Karlsson does not perform source-criticism in accordance with scholarly standards, stating only with reference to his and Kjartansson’s paper that: “It would be tedious here to repeat all the arguments and reservations concerning our use of Icelandic sources, let alone all the references to sources and literature that only people who read Icelandic could use.” Inevitably, it means that his paper, on this crucial point, represents incomplete methodology and scholarship. Th ere is also another serious matter of defi cient methodology of source- criticism that will be discussed below. Karlsson’s theory is characterized by at least three independent fatal fl aws of both empirical and methodological nature, in the sense that each represents a suffi cient condition for rejecting it as untenable or invalid. (1) In the circumstances, when the two requirements for identifi cation of epidemic diseases outlined by Shrewsbury cannot be met by the Icelandic sources, it is a serious defi ciency that Karlsson has completely neglected to investigate the possible provenance, the possible territorial origin of the contagion that was introduced into Iceland and that unleashed the purported plague epidemics there in the years 1402–4 and in 1494–5. In reality, this question is composed of two independently decisive parts that must each be satisfactorily resolved in order to keep the theory afl oat: he must for each epidemic be able to identify at least (1) one contemporary plague-stricken port town in Europe which (2) was sending ships to Iceland at the time these epidemics arrived there. Th ese two crucial questions or problems for the tenability of his theory are passed by in silence. In 2002, I published a complete Norwegian plague history from 1348–1654 which covers more than thirty waves of plague epidemics, without fi nding evidence of any epidemic of primary pneumonic plague, let alone of a pure epidemic of primary pneumonic plague; in fact all epidemiological evidence is compatible with bubonic plague.19

19 Benedictow 2002. See also Benedictow 2006.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 502 chapter thirteen

Could Plague Have Come to Iceland from Anywhere?

Only a few ships came to Iceland from abroad at the beginning of the fi ft eenth century, only Norwegian ships or Icelandic ships returning from Norway, normally from Bergen. Th e fi rst English fi shing ship is reported to have arrived in Icelandic waters in 1412.20 It was prohibited by law for the Icelanders to trade with foreigners. In order to illustrate this point one could note that in 1409, there was an intensive discussion at the Icelandic all-moot (‘Alþingi’) about what to do with all the royal incomes that had piled up in the island.21 Iceland was part of the Norwegian kingdom, normally administered by a royal governor (‘hirðstjóri’), and taxes were mostly paid in homespun woollens and, to some extent, in stockfi sh (wind-dried cod). Evidently, the cause of the problem was that too few ships arrived from Bergen, and that not even the commander of the King’s Mansion22 in Bergen who was responsible for collecting the royal incomes in Iceland, regularly sent a ship there to collect them. In the following years the New Annals (Nyi Annáll) occasionally provide telling pieces of information, for instance, under the year 1412 we are told that “No news came from Norway to Iceland,” i.e., no ship arrived from Norway; next year we are told that “A ship came from Norway,”23 a fact that the annalist fi nds worth mentioning. In 1419, the Icelanders wrote a letter to the king complaining that for several years no ships had arrived from Norway according to the old agreement with the Crown, which had been gravely detrimental to their poor country. Th ey therefore considered it unreasonable that the old prohibition against trading with foreigners was still in force.24 British sources tell the same story. Ships from Britain were sailing to

20 Carus-Wilson 1966: 161. 21 Steff ensen 1974: 48. 22 Th e ‘King’s Mansion’ is a translation of ‘Kongsgården,’ the royal administrative centre of western Norway in Bergen. At the time, this was not a castle built in stone, but consisted of wooden buildings surrounded by a wooden palisade and had no real de- fensive capabilities. 23 Annales islandici posteriorum sæculorum. Annálar 1400–1800: 18–9. Also the Icelandic Vatnsfj arðarannáll hinn elzti relates to this epidemic but it is written so long aft er the event and is so infested with source-critical problems that it is not mentioned by Gunnar Karlsson (or Jón Steff ensen) who in other works shows good knowledge of these important types of sources to Icelandic medieval and early modern history. Annales islandici postseriorum sæcolorum, 1933–8: 22. 24 Diplomatarium Norvegicum, vol. 2, no. 651; Diplomatarium Norvegicum, vol. 4, no. 330.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 503

Iceland at the time: the fi rst English fi shing ship arrived in 1412 to be followed by many others in the subsequent years. However, throughout the fi ft eenth century only English fi shing ships and merchant ships sailed to Iceland; Scottish ships do not seem to have been engaged in fi shing in Icelandic waters or trading with Iceland during this century.25 Crucially, there is not a single account of plague in these countries in 1402. Not a single plague epidemic has been registered in England between 1400 and 1405–7, nor was there any outbreak of plague in Ireland or Scotland during this period,26 nor for that matter in Northern Germany between 1396 and 1405–6, nor in the coastal commercial cities of the Netherlands between 1400–1 and 1409.27 Since the epidemic broke out in the autumn of 1402, aft er the return of an Icelandic ship,28 it is also clear that the disease would have to have been imported from a city or region where it was spreading in this year.29 In Norway, the sources provide no certain information of plague between 1391–2 and 1452, although there are some indications that there could have been plague in 1438–9. Norwegian sources are particularly sparse in these decades, and it is likely that one or more plague epidemics may have gone unnoticed. However, it is improbable that plague contagion could have been imported into Norway and re-exported to Iceland in years when there was no plague epidemic in the countries whence plague was imported to Norway, and this was throughout the Late Middle Ages mostly England, and at least once the Netherlands.30 Th ere is no evidence that plague was imported from Northern Germany before

25 Carus-Wilson 1966: 155–82. 26 Shrewsbury 1971: 149–50, mentions that “some time in 1402 there was ‘some disease in Scotland that ‘caused several deaths’ in Dundee.” Evidently, he does not consider it to have been plague, referring to it only as “some disease,” and since the only concrete information on mortality is restricted to the statement that it “caused several deaths” no impression is conveyed of a dramatic epidemic situation. And one must keep in mind that no ships from Britain were sailing to Iceland at the time and that also later in the century only English fi shing ships and merchant ships sailed to Iceland, Scottish ships do not appear to have been engaged in fi shing near or trading with Iceland. Carus-Wilson 1966: 155–82. 27 According to Blockmans 1980: 854, there was an outbreak of plague in Guelders, an inland city situated in the south-eastern part of the Netherlands near the border of Germany. 28 “Nýi Annáll,” in Annales islandici postseriorum sæcolorum, 1922–7: 9–10; “Lögmanns-annáll” in Islandske Annaler indtil 1578: 286. 29 Benedictow 2002: 102–11. 30 Benedictow 2002: 102–11. Evidently the reason is that Denmark’s pattern of international trade was very diff erent from Norway’s.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 504 chapter thirteen

1500 (or more accurately before 1628), and for the year 1402, with a substantial margin of safety of years around it, plague is not recorded in this area.31 According to the extant sources, plague was never imported from Denmark in the whole plague history of Norway, and no plague epidemic has been registered in Denmark from the end of the fourteenth century until possibly in 1405.32 Steff ensen, who recognizes that notions on provenance cannot be dispensed with and that this question therefore cannot be ignored, asserts that the supposed plague contagion came to Iceland with a ship from England. His source is Sticker’s old pioneering standard work on plague of 1908, where plague is mentioned as having occurred in England in 1402, but without indication of his source.33 Probably Sticker has misunderstood a reference to the epidemic in Iceland in the Danish scholar F.V. Mansa’s proto-scientifi c monograph of 1875 on the epidemic history of Denmark.34 It is a mistake or rather a slip of the pen on the part of Sticker that Steff ensen could easily have corrected by using Shrewsbury’s recent monograph on the history of plague epidemics in Britain and, as we shall see, Steff ensen knows the work and has consulted it, which makes his silence on this point the more remarkable. In fact, in another context Steff ensen remarks that Sticker’s mention of a plague epidemic in Denmark in 1402 with reference to Mansa’s monograph is erroneous,35 apparently he knows that this passage in Sticker’s work is not correct. Furthermore, Steff ensen does not attempt to assert that ships sailed directly from Denmark to Iceland at the time which also can be checked and corroborated in Icelandic annals and other sources. Th e staple36 was in Bergen and the Iceland trade a privilege for Norwegians. At the time of this epidemic, Icelanders were living in isolation with only sporadic contacts abroad and, for all practical purposes, abroad was Norway.

31 Shrewsbury 1971: 138, 141, 149–56; Ibs 1994: 206; Blockmans 1980: 836–45; Benedictow 2002: 102–5, 354–5, see especially Table 2A showing all known plague epidemics in Norway, England, Northern Germany and the Netherlands in the period 1348–1500 organized so as to facilitate comparison and discussion of territorial origin. Also see the references on this subject cited just above. 32 Mansa 1873: 92–100. 33 Sticker 1908: 8; Steff ensen 1974: 47–50. 34 Mansa 1873: 99. 35 Steff ensen 1974: 47. 36 Th e Shorter Oxford English Dictionary, vol. 2, p. 2109 gives the following defi ni- tion “staple:” “A town or place appointed by royal authority, in which there was a body of merchants having the exclusive right of purchase of certain classes of goods destined for export; also the body of merchants so privileged.”

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 505

Th us, 1402–4 were years with unusually few outbreaks of plague in Europe and they occurred far away from the northerly parts of Europe and commercial centres that could serve as the point of departure for transportation of plague by ship to Iceland. Th ese fi ndings permit a secure conclusion. No doubt there was a serious or severe epidemic in Iceland in the years 1402–4, but plague in any form based on the contagion Yersinia pestis appears to be excluded: there was no plague epidemic in any of the countries or regions with shipping activity with Iceland whence it could have been transported there. Another point which excludes primary pneumonic plague is the long duration of an epidemic which requires continuous and rapid spread in such a tiny population (probably 30,000–40,000 inhabitants37), and this duration would by far be the longest on the record. At the time, Iceland was a long voyage from any port in the northerly parts of Europe. Th e voyage from western Norway to Iceland would under favourable circumstances and with suitable wind usually take a couple weeks of sailing. However, ships from Bergen or Nidaros/ Trondheim would fi rst sail along the coast approximately to the northwestern point of Stad, which would normally take about fi ve days, before they could turn and with an easterly wind sail westwards to Iceland.38 Waiting periods for useful winds could easily take much more time than the actual sailing.39 Th is would very much be the case for ships waiting for the easterly winds necessary for sailing westwards to Iceland, since Norway is situated in the so-called west wind belt which is dominated by westerly winds. In the twentieth century, easterly winds constituted only 2.2 per cent of all wind measurements in the sailing season 1 April–30 September40 in western Norway.41 Obviously, any ship with primary pneumonic plague on board among the cramped conditions for crew and passengers aft er cargo had been given strong priority would be rapidly heading for disaster.

37 See below, fn. 68. 38 Helle 1982: 68. Cf. Steen 1934: 227. 39 Steen 1942: 306. 40 Th e maximum sailing season is given in the important Norwegian source Kongespeilet [Th e King’s Mirror] written around 1250. Here I have used the recognized translation Kongespeilet 1947: 53, 55. 41 Since the prevailing distribution of wind direction is heavily aff ected by the direction of the earth’s rotation, the distribution of wind directions has probably not varied greatly over historical time. Th e crude data of all wind-force measurements registered by the wind-gauge stations in Bergen are published by the Norwegian Meteorological Institute: www.met.no

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 506 chapter thirteen

Since no epidemic of primary pneumonic plague has been identifi ed in Norwegian sources, and in the year in question (1402) not even a plague epidemic, Steff ensen’s theory is only a speculative fl ight of imagination. In the case of the supposed plague epidemic of 1494–5, the situation is much the same. Karlsson has not attempted to identify the provenance of the contagion and passes in silence by the problem of whence plague contagion could have been transported to Iceland. In the meantime, shipping activities in northern Europe had increased greatly. Although the small island of Iceland with its tiny population situated in the middle of the North Atlantic Ocean on the northwestern out- skirts of Christendom was a destination with limited attraction, it was by now less isolated and to some extent linked to northern Europe through trade in stockfi sh. At the time, English interest in fi sheries in Icelandic waters and commercial interests in the island had been sharply reduced, due largely to strong eff orts by Hanseatic cities to drive the English away and assert their primacy and the importance of the staple of stockfi sh in Bergen. By the end of the fi ft eenth century, Iceland was apparently quite regularly visited by a ship sent there by the commander of the King’s Mansion in Bergen in order to collect the royal taxes, and stockfi sh may at least occasionally also have been pur- chased in order to fi ll up the ship and earn some extra income. isTh pattern can be discerned in a sprinkling of fi ft eenth-century sources and is confi rmed by the records of Bergenhus Castle, the new royal castle in Bergen, which are extant from 1514, and where also the occa- sional Icelandic ship is mentioned.42 Th e sources reveal that Iceland at this time had some regular commercial contacts with Norway and was now more exposed to the importation of contagion than at the beginning of the century. Th is puts the epidemics of 1494–5 in a somewhat diff erent light. However, as can be seen from Table 2A and the comprehensive comments in my Norwegian plague history, there was no plague epidemic in these years in Norway, England, Northern Germany or in the coastal commercial cities of the Netherlands.43 In 1494, epidemics are mentioned in Gouda and Guelders as typical localized outbreaks and in inland cities without shipping contacts or commercial connections

42 Norske Regnskaber og Jordebøger. See the very good index. 43 Benedictow 2002: 102–11.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 507 with Norway or Iceland.44 Gouda and Guelders are the only places where there may have been outbreaks of plague this year.45 Nonetheless, Steff ensen asserts again that the contagion that triggered the purported Icelandic plague epidemic was imported from England. Th is conclusion is seemingly the outcome of a process of elimination. He admits that plague is not mentioned in “Scandinavia or Denmark” in 1494. He has again consulted Sticker, who in the crucial year of 1494 in the whole of Europe has registered plague only in Nuremberg, and in 1495 more or less probable or possible plague epidemics in Landshut (Bavaria), Erfurt and in Lower Austria.46 Since Steff ensen does not specify these localities he can state rather inaccu- rately that plague “is mentioned in many German towns.”47 Obviously, the contagion that unleashed the Icelandic epidemic could not have originated in any of these localities which were not Hanseatic commercial cities on the North Sea or Baltic Sea. Since Sticker cannot help him with a plague epidemic in England, not even by a slip of the pen, Steff ensen must look elsewhere for support of his theory of provenance. He now mentions that he has consulted Shrewsbury’s recent monograph, thus admitting that he knows it, but was confronted with the fact that Shrewsbury does not consider that there has been any pestilence in England in the last decade of the 15th century.48 Since he is not willing to draw the scholarly conclusion that there was no plague epidemic either in England (or anywhere in the British Isles) or elsewhere in northwestern or northern Europe that could have served as an origin of the contagion which was transported to Iceland in 1494, he looks elsewhere for support. He turns to Creighton’s monograph of 1891, A History of Epidemics in Britain, where “minor

44 Noordergraaf and Valk 1996: 225; Blockmans 1980: 854. 45 In the following year, 1495, there appears to have been a plague epidemic in southeastern Sweden, as the convent of Vadstena and the city of Stockholm on the Baltic Sea is reported in contemporary chronicles as being seriously visited by pestilence. Moseng 2006: 314–5. Since an outbreak of plague in Sweden in 1495 cannot be the origin of plague in Iceland in 1494 and Sweden had no contact with Iceland by ship, this epidemic is not signifi cant in the present context. It is also noteworthy that eastern Sweden was exposed to importation of plague across the Baltic Sea and had, like Denmark, a diff erent temporal rhythm of plague epidemics than Norway which imported plague from England and northwestern Europe more generally, but never from Sweden or Denmark. 46 Sticker 1908: 87. 47 Steff ensen 1971: 53. 48 Shrewsbury 1970: 149, 155.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 508 chapter thirteen epidemics are mentioned in Oxford 1486, 1491 and 1493, and it appears that the pestilence was close to Edinburgh in March 1493.”49 Although Creighton does not use the term “plague epidemics,” but only the unspecifi c term “epidemics,” this statement is taken to constitute justifi cation for an unqualifi ed assertion to the eff ect that Shrewsbury took no interest in minor plague epidemics and “presumably means only large-scale epidemics.” Every reader of Shrewsbury’s highly detailed work will know that this assumption is groundless: no epidemic is too small to escape his acute interest, neither can Shrewsbury’s numerous references to and detailed knowledge of Creighton’s work escape notice. In fact, Steff ensen’s assertion in relation to Shrewsbury work is at variance with Shrewsbury’s presentation of his research pro- gram on fi ft eenth century plague epidemics: “As this work is primarily a history of bubonic plague, an attempt must now be made to pinpoint the possible outbreaks of that disease in the British Isles during this century,”50 making it clear that his objective is to identify every possible outbreak of plague. Th is does not exclude the possibility that he overlooked epidemics, especially local outbreaks.51 Scrutiny of the sources or research literature for more information is always warranted but Steff ensen did not succeed in identifying any plague epidemic and did not accept the consequences. Th e reason Shrewsbury does not mention the epidemics in Oxford in these years is that he did not fi nd any evidence to the eff ect that they were plague. In fact, Shrewsbury has not registered a single epidemic of plague in England “aft er November 1480 until an epidemic erupted at Oxford in 1499” which he considers a “doubtful exception.” Th us, Shrewsbury is clear in his view that in the sources there is no indication of plague in England in the period comprising the epidemics in Oxford and the purported Icelandic plague epidemic. It can also be shown that he was certainly not averse to mentioning small possible outbreaks of plague in Oxford.52 J. Hatcher and P. Slack, prominent English scholars who, in important subsequent studies of plague in England, have fi ne-combed the sources for epidemics, including the sources relating to the last two decades of the fi ft eenth century, did not fi nd any credible information

49 Creighton 1891: 283; Steff ensen 1974: 53. 50 Shrewsbury 1971: 155. 51 See Gottfried 1978: 238–40. 52 See Shrewsbury 1971: 149.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 509 to the eff ect that the small epidemics in Oxford mentioned by Creighton were plague.53 Th e reason may quite safely be assumed to be that there is no such evidence. Another insurmountable problem is how contagion from this small inland outbreak of an epidemic disease should have been moved to Iceland, as the continuous lines of communication presumed by such an event appear to be completely unknown to either contemporaries or to modern historians. Another insurmountable problem is that even the last mentioned epidemic in Oxford occurred in 1493 and cannot be linked to the outbreak of an epidemic in Iceland 1494. Steff ensen appears to assume that he can brush aside Shrewsbury’s recent work, written by a modern bacteriologist, and substitute it with Creighton’s proto-scientifi c work of 1891. Creighton’s monograph is, as noted above, written according to the spirit of an ardent protagonist and defender of miasmatic epidemiological theory in the face of the rise of the new science of bacteriology. Th is means that Creighton denied that each epidemic disease was caused by a specifi c microbiological agent and believed instead that miasma could take on various forms and could develop into plague from other manifestations of miasmatic disease (i.e., other epidemic diseases), especially from typhus and dysentery; he also claimed that plague was due to a poison emanating from putrescent corpses.54 According to miasmatic theory, an outbreak of any epidemic disease could therefore be of interest to a study of plague since it could develop into plague. Th is is also the reason for Steff ensen’s sudden use of the term “pestilence” in the citations above. It refl ects the miasmatic assumptions that have now been inserted at the core of his argument, although nothing is explicitly said about this extraordinary turn of his argument away from modern scientifi c bacteriological and epidemiological premises of analysis to untenable proto-scientifi c miasmatic medical theory. Besides the fact that Creighton has not suggested that the epidemics in Oxford were plague, it must be emphasized that Creighton does not assert that there was an epidemic at Oxford or anywhere else in England in 1494. Nonetheless, Steff ensen sticks to his theory, although it is untrue that Shrewsbury only took interest in major plague epidemics and ignored minor epidemics, so that plague could have been imported from

53 Hatcher 1977/1987: 17–8; Slack 1985: 53–65. 54 See, e.g., Hirst 1953: 74–5, 87, 89, 93–4, 285; Shrewsbury 1971: 150.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 510 chapter thirteen

England to Iceland in 1494 in connection with some minor outbreak of plague which Shrewsbury ignored, Creighton did not know, and which Hatcher and Slack overlooked. According to all accumulated scholarly knowledge on English plague history no plague contagion could have been transported from England to Iceland in 1494. Steff ensen also mentions a possible plague epidemic “close to Edinburgh” in 1493, referring again to Creighton as his source. Shrewsbury takes seriously the possibility of an outbreak of plague in Edinburgh and outlying townships in 1493 which led to the isolation of the town, but thinks that it probably was an epidemic of typhus. He concludes therefore his chapter on plague in Scotland in the fi ft eenth century by stating: “Th ere does not seem to be any outbreak of epidemic disease in Scotland that presents any circumstantial evidence for its identifi cation as an outbreak of bubonic plague, with the doubtful exception of 1499.”55 Signifi cantly, this epidemic did not spread to any other place in Scotland, which makes it a very local outbreak of whatever disease it might have been. Furthermore, an epidemic outbreak in 1493 is out of chronological co-ordination and cannot serve even as a hypothetical origin for the Icelandic epidemic. Edinburgh was not a commercial shipping centre, and it is rather unsurprising that it had no known connection by ship with Norway, and certainly not with Iceland. Again, it is not possible to construct any case for transportation of plague contagion from Scotland to Iceland in 1494. Th us it is not possible to adduce empirical support for the theory that the epidemic outbreak in Iceland in 1494 was primary pneumonic plague or bubonic plague for that matter, since there was no area or city which was visited by plague whence plague could have been shipped to Iceland. Undoubtedly this epidemic in Iceland was a serious outbreak of a contagious disease or multiple contagious diseases taking a high toll of lives. One should take note that there are numerous accounts in the Icelandic annals of serious or severe epidemics, including accounts that pre-date the advent of plague in Europe and also subsequent epidemics that typically are out of rhythm with known European plague epidemics. Some examples may be useful: in 1292, “Came such a big disease over all the country […] and caused great mortality”; in 1306, “Big epidemic in Iceland and high mortality in the south of the country”; in 1309, “Big killing disease in the north of the country”; in 1310, “Killing disease and mortality in all the Vestfj ord area and the

55 Shrewsbury 1971: 151, 155.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 511

Southerners’ area and boil disease”; or in 1380 and 1382: “Huge boil disease all around the country and great mortality […] disease over all the country and huge mortality.”56 Th us, also 1494 was a year with particularly few plague epidemics in Europe and they occurred far away from the northerly parts of Europe and from commercial sea ports that could serve as a point of departure for transportation of plague contagion by ship to Iceland. Steff ensen’s assertion that plague contagion came from England has been shown to have no evidential underpinning, essentially it is an arbitrary assertion, and likewise not a single plague epidemic was registered in these years in Norway. Again, the intensive and comprehensive discussion permits a clear conclusion: no doubt there was a serious or severe epidemic in Iceland in the years 1494–5, but it appears impossible that it could have been due to plague in any form based on the contagion Yersinia pestis. Steff ensen sees the problem of provenance clearly and appreciates the fact that positive identifi cation of at least one case of plague in one sea port with at least some evidential support for possible shipping to Iceland is crucial to instil his theory with any (level of) tenability, however low. However, he cannot come to terms with the the complete lack of supporting evidence on both points for both epidemics. Karlsson has chosen the option of passing the problem by in silence, although this undermines the very foundations of his theory, quite possibly because he recognizes that there is no supporting evidence. Th is exhaustive discussion of a possible provenance of plague contagion (Yersinia pestis) for the purported plague epidemics in Iceland in the years 1402–4 and 1494–5 has been completely negative. Th is fi nding constitutes a suffi cient condition for characterizing the theory as materially untenable and for the conclusion that it must be rejected. Th is does not mean that there cannot be other independent and suffi - cient conditions for rejecting the theory.

Pure Epidemics of Primary Pneumonic Plague: Fact or Fiction?

A thorough and for all practical purposes complete presentation of the medical and epidemiological studies of primary pneumonic plague is presented in my thesis in order to provide the broadest possible basis

56 Annales Islandici 1888: 71, 53, cf. 75, 364. My translation from Icelandic Norse.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 512 chapter thirteen for my discussion of the nature of the late medieval plague epidemics in the Nordic countries. It was also motivated more specifi cally by my desire to provide a background for discussion of Steff ensen’s paper,57 since his theory of epidemics of primary pneumonic plague in Iceland is not supported by a single reference to any medical or epidemiological study of primary pneumonic plague. Karlsson who defends his theory intensively fails to include any reference in his paper to a primary study of pneumonic plague or to the broad summaries of such studies provided in the standard works on plague, not even to K. Meyer’s brief and succinct paper. However, according to the methodological tenets of medical and historical science, their theory of the nature of these Icelandic epidemics as being primary pneumonic plague or some mutant variant of this disease can only be adequately discussed in relation to modern medical and epidemiological knowledge on this modality of plague, implemented in the form of a systematic comparison between historical data and modern data. Th e better and more comprehensive the correspondence or correlation between modern medical data and historical data, the more tenable the inference of the identity as primary pneumonic plague. Fruitful discussion requires a satisfactory empirical knowledge of the medical characteristics, epidemiological structures, lethality and demographic eff ects (population mortality) of primary pneumonic plague. Since Karlsson denies that there were any rats in Iceland, he is obliged to, launch a theory of pure epidemics of primary pneumonic plague, which requires that he consistently avoids the medical literature on primary pneumonic plague and instead makes his case depending entirely on an assertion in a review paper by Morris. Th is assertion occurs in a passage where Morris scathingly and disparag- ingly attacks Shrewsbury for denying the existence of this form of plague: What is the cause of Shrewsbury’s myopia?58 Perhaps it came upon him at the moment when (on page 6) he delivered himself of the dictum “pneumonic plague cannot occur in the absence of the bubonic form and it cannot persist as an independent form of plague.” Th is is untrue […]. Th e Manchurian epidemic was indeed exclusively pneumonic, as was the slightly less destructive outbreak of 1920–1; […] Wu Lien-Teh’s classic

57 Benedictow 1993/1996: 25–31, 214–27. 58 “Myopia” means shortsightedness in a medical sense or fi guratively narrow- mindedness.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 513

treatise on pneumonic plague is signifi cantly absent from Shrewsbury’s very extensive bibliography.59 Th is basic idea and crucial premise of Morris’s theory of the nature of plague in England and Karlsson’s theory of plague in Iceland, the notion of pure epidemics of primary pneumonic plague, can usefully be addressed fi rst. Th eoretically, pure epidemics of primary pneumonic plague are possible in the technical meaning that they could start de novo from an original case of primary infection: hunters could, for instance, be infected by droplet infection when skinning or cutting up a sick animal with plague septicaemia; American veterinary personnel have contracted primary pneumonic plague directly from pet cats which had inhaled contaminated droplets in the process of killing rodents with plague septicaemia. Th ese cases relate to a rodent basis in the form of a plague reservoir where plague circulates continuously, which is excluded in the case of Karlsson’s Icelandic theory. Furthermore, these very few cases have never caused spread to any other person, thus, they have never given rise to an endemic development (with possible conse- quent epidemic forms).60 Morris refers to Wu Lien-Teh’s monograph on primary pneumonic plague for empirical support of the existence of pure epidemics of primary pneumonic plague, and Karlsson does so indirectly by citing parts of the citation including the explicit reference to this work. However, Wu Lien-Teh states the contrary unequivocally: It is generally agreed that the pneumonic form of plague is not directly traceable to the epizootics [i.e., does not originate by cross-infection with plague-infected droplets from rodent to man] but arises from human cases of bubonic plague with secondary lung involvement.61 Wu Lien-Teh states here emphatically, and repeatedly, that primary pneumonic plague develops only from cases of secondary pneumonic plague, i.e., human victims of bubonic plague who have developed septicaemia. He argues that the term ‘pure epidemic of pneumonic plague’ “should be avoided as it is misleading as far as the origin of the outbreaks is concerned.”62 Th is puts Karlsson’s theory of pure

59 Morris 1971: 208. 60 Benedictow 1993/1996a: 215–9. 61 Wu Lien-Teh 1926: 162–4. Wu Lien-Teh, Chun, Pollitzer 1934: 83. 62 Wu Lien-Teh 1936b: 418–9.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 514 chapter thirteen

epidemics of pneumonic plague in Iceland in perspective, disclosing in another way the arbitrary nature of his theory and its lack of empirical underpinning in primary studies on pneumonic plague. It also puts in perspective the fi ctitious support Karlsson draws from Morris who insists that the Manchurian epidemics were pure epidemics of primary pneumonic plague without basis in bubonic plague and incorrectly refers to Wu Lien-Teh’s monograph for empirical support.63 Th e arguments on this point can now be summed up: researchers have not succeeded in identifying a single epidemic of pure primary pneumonic plague. If empirical evidence is for methodological reasons considered essential for establishing factual data with any level of validity or tenability that can be used for construction of explanatory theories, epidemics of primary pneumonic plague do not develop de novo. Th ere is no such thing as pure epidemics of pneumonic plague, and all assertions to the contrary are arbitrary.64 Since there is no such thing as a pure epidemic of primary pneumonic plague, there cannot have been pure epidemics of primary pneumonic plague in Iceland. According to ordinary methodological considerations, the discussion of this point has produced another suffi cient independent condition for rejecting Karlsson’s theory.

Primary Pneumonic Plague in Manchuria: A Model for Iceland?

According to Steff ensen and Karlsson, the Icelandic epidemics were large-scale, sweeping away a large part or even most of the Icelandic population. Th is view should be seen in the light of the fact that only one fairly large epidemic of primary pneumonic plague has so far been identifi ed with certainty in the history of plague, namely the epidemic of 1910–11 in Manchuria and adjacent regions that took around 60,000 lives, with 10,000 deaths occurring in the Russian Vladivostok area. Th is is by far the largest epidemic of this modality of plague ever recorded, and was followed by the second largest epidemic in the same area a decade later. Th e fi rst of these epidemics are cited enthusi- astically by protagonists of the importance of large-scale epidemics of

63 Morris 1971: 208–9. 64 See especially Wu Lien-Teh 1926: 162–95; Wu Lien-Teh 1936b: 401–3.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 515 primary pneumonic plague, also by Steff ensen and Karlsson. Th is is only possible because they avoid taking into account: (1) the tiny rate of population mortality in these epidemics which has no explanatory potential for the Icelandic case; (2) the exceptional circumstances of these two epidemics; (3) the medical and epidemiological evidence provided by scholarly studies of primary pneumonic plague when it does not suit their case. A brief summary here may therefore be of use. A fresh presentation of the central structural circumstances that gave these two Manchurian epidemics of primary pneumonic plague their exceptional dynamics is needed in order to permit a concrete comparison of basic and epidemiologically relevant social structures between Manchuria at the beginning of the twentieth century and late medieval Iceland (or Europe more generally): (1) At the time, Manchuria’s population underwent exceptional growth through the immigration of adult males from China for work, and increased from 12.7 million according to the Chinese Census of 1912 to 22 million in 1922.65 Th e great majority of the Manchurian population was Chinese coolies working in the mines and along the rivers Sungari and Amur. Some 10,000 hunted large rodents called tarabagans (marmots) for their valuable fur, and since plague infection circulated among these rodents, hunting was the origin of quite frequent cases of bubonic plague that were the origin also of the primary pneumonic epidemics. Th is means that the social structure and the age and gender dimensions of the population were extremely skewed and entirely unrepresentative of any normal population. When the epidemic of 1910–11 broke out, there was no anti-epidemic organization which could combat it. When the second epidemic broke out in 1921, an organization had been built to deal with such an event and thus much better records are available which provide more detailed knowledge also of the population’s social composition and its extremely skewed character compared with normal populations: for instance, among the patients at the hospital in Harbin there were 1139 labourers out of a total of 1252 patients (91 per cent), 67 women as against 1243

65 Wu Lien-Teh 1922–3: 262.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 516 chapter thirteen males (5 per cent), and 21 children of ages 0–10 constituting 1.67 per cent of the patients.66 Th e immense inrush of poor migrant workers created a gigantic market for miserable accommodation. Th ey stayed mostly in incredibly overcrowded inns and lodging houses of a kind completely unheard of in normal populations whether in Iceland or elsewhere in medieval Europe: Th e accommodation in both types of inns is based upon much the same plan. Two, and sometimes three, tiers of berths for the lodgers are present, there being just room enough between the tiers for a man to sit up […]. Dr Ch’uan visited some of these inns in the winter. His report showed the conditions to be bad in the extreme: “in one hut, 15 ft . square and 12 ft . high [20.9 sq. metres or 76.4 cu. metres], there were packed in three tiers of berths, one above the other, more than 40 people,” thus, each having at his disposal less than 1.9 cu. metre of space. For protection against the cold in the winters, when the temperature “as a rule reaches −30°C to −40°C,” the lodging-houses were constructed either entirely underground or at least partly underground. Th e windows were closed during the winters, and “even in May the rooms were ill-ventilated and stuff y.” “Th e men eat, sleep and very oft en cook in the same room. Th ey sleep in rows with the head against the wall and the feet towards the central passage. Th ere is no partition between adjoining berths so that they can easily breathe and cough into each other’s faces.67 Th ese men were living in incredible density in extremely ill-ventilated housing, breathing, sneezing and coughing in the immediate proximity of the face of the next man. For a very large proportion of the population, the social circumstances were extreme, in the sense of extremely favourable to propagation of disease by droplet infection, much more favourable than in the case of any normal medieval population. Th e contrast to Iceland is extreme: an isolated island society with slight contact abroad, with very sparsely and highly dispersed set- tlements and a largely stable population distributed on (probably) 5000–6000 detached tenancies or free holdings.68 Th e inhabitants were

66 Wu Lien-Th e 1926: 82; Chun 1936: 320–1. Th e change in the total number of patients from 1252 to 1310 in the information on the gender incidence is due to a cat- egory of 60 persons with unknown age which means that the total number of patients actually was 1312, but these inaccuracies do not signifi cantly aff ect the estimates. 67 Wu Lien-Teh 1913: 248–9. 68 Benedictow 1996b: 182–4; Benedictow 2003: 34; Orrman 2003: 438. According to Orrman, the maximum number of peasant holdings in operation in Iceland at the time was 3800–4500, but according Jón V. Sigurðsson 5000–6000. In my view, Orrman’s use

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 517 overwhelmingly making their living by subsistence economy, mostly extensive domestic husbandry more or less in combination with some fi shing, gathering and small-scale hunting, which means that the agricultural units needed large territories and the farmsteads were greatly dispersed. Th ere was, consequently, relatively little social exchange and commercial contact between these constituent parts of the Icelandic population, quite likely less than among any other European population. (2) Th e railway played a pivotal dynamic and entirely unmedieval role in the process of disseminating the epidemic: it was fi rst spread from Manchouli69 “along the railroad, westwards into Siberia again and eastwards into China” and “the Chinese Eastern Railway conveyed the infection broadcast.”70 Because primary pneumonic plague is spread by interhuman dissemination, Wu Lien-Teh, points out that “such spread will be slow in sparsely populated territories […] an epidemic will cover wide distances with almost lightning rapidity as soon as modern means of communication are available.”71 Th e last observation refl ects the process of dissemination in the Manchurian epidemics and ought to put the dynamics of the Manchurian epidemics in an instructive perspective, particularly in relation to medieval Iceland. (3) Th e epidemic ran its course from April to the end of 1910 without signifi cant countermeasures. On Christmas Eve, the Cambridge- trained Chinese physician Wu Lien-Teh arrived in the provincial capital of Harbin under orders from the Imperial administration to

of these fi gures refl ects incomplete source-criticism: the actual fi gure must be higher and Jón Sigurðsson’s estimate seems more credible. Th ere is no evidence on contemporary household size and thus on household multiplier, except some indications that multiple households were not unusual. Th e Icelandic Census of 1703 shows an average household size of 6.5 persons. If this can be taken as fairly representative also of early fourteenth-century household structure, and in the meantime there had been little change in the economic basis or structures regulating population size according to Malthusian principles, Jón Sigurðsson’s estimate indicates a population size of 30–40,000. It is highly unlikely that medieval population was at any time larger than recorded by the Census of 1703, namely 50,400 inhabitants, which also is the fi gure suggested by Orrman, implying a much higher average household size than that found by Jón Sigurðsson. See the discussion in Benedictow 1996b: 182–4. 69 Situated in northern Manchuria at the junction of the Chinese Eastern Railway and the Trans-Siberian Railway. 70 Wu Lien-Teh 1926: 71; Wu Lien-Teh 1936a: 32. Th e Manchurian Plague Prevention Service was established in 1912. Manchurian Plague Prevention Service. Memorial Volume. 1934. 71 Wu Lien-Teh 1926: 182–3, cf. 74.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 518 chapter thirteen take charge of combating of the epidemic. One should recognize that what this excellent scholar could achieve was quite limited. At the time, very little was known about primary pneumonic plague. He had at his disposal a minimal sanitary administration or medical staff which was responsible for a vast region with c. 12 million inhabitants spread over 939,000 sq. km (363,000 sq. miles), more than six times the size of England and Wales, with the possible support of only a rudimentary Chinese Imperial administration and with military occupation of size- able areas by Russian and Japanese forces.72 Wu Lien-Teh had at his disposal only the most rudimentary medical facilities, a situation that was much improved only aft er the establishment of the Manchurian Plague Prevention Service in 1912 with him as director and the appro- priation of substantial funds for hospitals, laboratories and staff .73 Importantly, Wu Lien-Teh does not in any of his copious and detailed writings on the Manchurian epidemic make any claim to the eff ect that his actions had signifi cant eff ect on the further course of this epidemic. Morris’s claim to the contrary on his behalf is groundless. Nonetheless, by April 1911, the epidemic had faded away.

Th e Spontaneous Decline of Epidemics of Primary Pneumonic Plague

Wu Lien-Teh, who led the campaigns against both Manchurian epidemics of primary pneumonic plague, has written most of the medical and epidemiological studies on this disease, including the only monograph, a fact that also refl ects the relative rarity of this disease. He has clear opinions on the reasons or mechanisms causing the decline: the main reason for the fading away of the epidemic was not his personal eff orts but the characteristic tendency of epidemics of primary pneumonic plague to undergo “spontaneous decline.”74 In a later work, Wu Lien-Teh relates that when he fi rst came to Harbin, “many of the inhabitants with whom I conversed informed me that this epidemic was quite an ordinary visitation, to which they had been accustomed. It never killed many, and would stop of itself, if no attention were paid to it.”75

72 Wu Lien-Teh 1913, 1926. 73 Wu Lien-Teh 1934a: 1–3. 74 Wu Lien-Teh 1926: 188–92; Wu Lien-Teh, Chun, Pollitzer 1934: 85–6. 75 Wu Lien-Teh 1934b: 16.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 519

Obviously, something new had happened that gave the outbreak of primary pneumonic plague much enhanced powers of spread, namely the immense infl ux of coolies living in exceptional density in the lodging houses or inns, the extreme cold that kept them indoors, and the dis- seminative eff ects of the railway. Nonetheless, also in the new conditions with much strengthened epidemic dynamics and higher rates of spread and mortality the disease’s tendency to self-limitation and spontaneous decline still prevailed. Th e fact that spontaneous decline can be such a prominent feature of epidemics of primary pneumonic plague despite exceptionally favourable conditions for respiratory transmission compared with the medieval social scene explains why population mortality in much the largest of all known epidemics of primary pneumonic plague was only a minute fraction of the rates ordinarily wrought on populations by medieval plague epidemics. Th e Manchurian epidemic of 1910–11 claimed about 50,000 lives in Manchuria, which means that population mortality was only a tiny 0.4 per cent, not at all compatible with or comparable to the mortality rates assumed by Steff ensen and Karlsson for the Icelandic epidemics under discussion or by Morris for England. Since Karlsson had read parts of my doctoral thesis, he saw the threat that these facts presented to his theory and argues that the same problem, or “riddle” as he calls it, of minimal mortality also prevailed in the bubonic plague epidemics in India around the turn of the previous century, and asserts that this must have escaped my notice.76 Th at is not correct. As pointed out above, the mortality in the Indian epidemics of bubonic plague was small because they were from the outset combated intensely with effi cient anti-epidemic countermeasures, which conspicuously was not the case with the Manchurian epidemic of primary plague of 1910–11 or with the Black Death or subsequent plague epidemics.77 If Wu Lien-Teh’s observation of a marked tendency of self-limitation and spontaneous decline of epidemics of primary pneumonic plague is correct, it implies that it is a characteristic feature of such epidemics to be small. Th is proposition can be tested by collecting and collating all evidence on all known epidemics of this mode of plague, a task Karlsson

76 Karlsson 1996: 283. 77 See above: 194–202.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 520 chapter thirteen neglected. Th is material makes it clear that the two Manchurian epidemics are the only instances of fairly large epidemics of primary pneumonic plague, and that even they caused miniscule population mortality rates. Th e corpus of scholarly studies on primary pneumonic plague contains only one probable and one possible instance of epidemics designated primary pneumonic plague (other than the two in Manchuria) with as many as one thousand cases: (1) in 1905, an epidemic with around 1000 cases occurred, according to Wu Lien-Teh, in the Bukeëf Horde of the Kirghiz Steppes of the (Imperial Russian) Astrakhan Government (“guberniya”); however, according to Klimenko, there were 659 cases with 621 deaths, which means that the epidemic was considerably smaller, included about190 bubonic cases,78 and should rather be designated a mixed epidemic. In 1920, 1056 pneumonic cases were registered in the northern Punjab province of India.79 Scholars have registered a small number of epidemics designated primary pneumonic plague with over one hundred cases, the largest comprising 416 cases.80 Th us, the overwhelming majority of known outbreaks of epidemics designated primary pneumonic plague are quite small or even tiny, most of them restricted to the social level of household, relatives or neighbours, the great majority of outbreaks comprising from a couple of cases to a few tens of victims. Th is is a general feature covering all areas where epidemics designated primary pneumonic plague have been medically verifi ed, whether in Russia, Central Asia, India, China, Manchuria, Egypt, Madagascar, Vietnam or Indonesia, or elsewhere.81 Arguably, this could be taken to mean that primary pneumonic plague in the overwhelming majority of outbreaks assumes an endemic character, becoming extinguished without having developed an epidemic form comprising a signifi cant number of cases. Similarly, it appears impossible to identify any certain or likely historical epidemic of primary pneumonic plague. Sticker, who was an impressive pioneer in the fi eld of plague history, attempted to collect all

78 Th e estimate is based on assumptions of a 100 per cent lethality rate in cases of primary pneumonic plague and 80 per cent in cases of bubonic plague. 79 Wu Lien-Teh 1913–4: 244; idem, 1926: 56; Klimenko 1910: 660; Wu Lien-Teh 1927–8: 58. 80 Klimenko 1910: 660. Cf. Wu Lien-Teh 1913–4: 244. 81 Studies presenting epidemics of primary pneumonic plague include Klimenko 1910: 657–63; Wu Lien-Teh 1913: 237–52; Wu Lien-Teh 1926: 9–99; Wu Lien-Teh 1927–8: 55–92; Tieh, Landauer, Miyagawa et al. 1948: 52–8; Pollitzer and Li 1943: 212– 6; Brygoo 1966: 51–2; Trong, Nhu and Marshall 1967: 93–7; Narkevich, Onishchenko, Naumov et al. 1991: 32.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 521 relevant historical evidence on pneumonic plague, but managed to fi nd only a few interesting possible cases of such epidemics, six before 1800. Importantly, he cautiously designates them as being characterized by “predominant [“vorherrschende”] lung disease,” which presumably should be taken to mean that, in his view, they were mixed epidemics that also included signifi cant proportions of bubonic cases.82 One should also note that he does not (because the sources will not permit it) diff erentiate between secondary and primary pneumonic plague.83 Wu Lien-Teh, who took a strong interest in plague history, succeeded in adding a few epidemics to the list, but also in his case all epidemics are mixed.84 In his excellent study of plague and other epidemic diseases in Early Modern Central Europe (1560–640), E.A. Eckert discusses some evidence based on parish registers relating to plague epidemics in seventeen communities in southern central Europe and the Rhenish region in the years 1563–4. Th e clinical evidence provided by contemporaries does not discern between secondary and primary pneumonic plague, but the fact that 11 of 17 communal epidemics reached their peaks of burials in the winter months merits attention. However, the author concludes his discussion by stating that these and other outbreaks of winter epidemics probably were mixed epidemics, pointing to the Madagascan epidemics for an analogous model.85 In my monograph on the Black Death, doing my very best to collect and col- late all relevant studies for the whole of Europe, I similarly found no evidence implying epidemics of primary pneumonic plague, only two possible cases of mixed epidemics, namely in Avignon and only (just) possibly in Novgorod. And this outcome was strongly confi rmed by the seasonality and pattern of spread of the Black Death.86 Despite much eff ort, it has so far not been possible to identify a likely historical case (before c. 1850) of an epidemic of primary pneumonic plague. Th e notion of pure epidemics of primary pneumonic plague is, as shown above, fi ctitious, an epidemic of this type has never been

82 Mixed epidemics comprising (for social and cultural reasons) an increased incidence or level of pneumonic plague have been studied by modern scholars in Madagascar 1924–40 (20 per cent), Upper Egypt 1904–22 (31 per cent), Java (4–8 per cent) and episodically in Vietnam. Brygoo 1966: 45–7; Wu Lien-Teh 1926: 33, 23; Trong, Nhu and Marshall 1967: 93–7. 83 Sticker 1910: 246. 84 Wu Lien-Teh 1926: 7. 85 Eckert 1996: 83, 85–6. For the Madagascan plague epidemics, see fi ne and comprehensive overview by Brygoo 1966. 86 Benedictow 2004: 214, 233–41, 238–9.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 522 chapter thirteen observed in the research history of plague. Steff ensen’s and Karlsson’s assertions that there were epidemics of primary pneumonic plague or pure epidemics of primary plague in Iceland and Karlsson’s view that that the Black Death in general was primary pneumonic plague, are without empirical support. Th is may explain that there is no indication in Karlsson’s paper of any eff ort to obtain relevant information on the matter. Th is discussion has established another independent suffi cient condition for rejecting their assertions. Th e epidemiological eff ects that could be expected if Wu Lien-Teh’s observation is correct, that the course of epidemics of primary pneumonic plague are characterized by “spontaneous decline,” have now been tested on the complete record of registrations of epidemics of this modality of plague and the outcome has been found to conform entirely to his prediction: generally, outbreaks in normal populations are quite small or tiny, rarely exceeding a hundred cases. Neither in the past nor in recent history is there any instance in the records of an epidemic of primary pneumonic plague in a normal population comprising the number of cases (mortality) or sweeping away the proportion of the population (population mortality) corresponding to Steff ensen’s and Karlsson’s assertions as to the mortality eff ects of the Icelandic epidemics of 1402–4 and 1494–5, and by very large margins. Th is outcome of the discussion of this epidemiological aspect, the highly limited powers of spread of primary pneumonic plague, suffi ces as another independent suffi cient condition for rejecting Steff ensen’s and Karlsson’s theory.

It would, however, be unsatisfactory to leave this subject without identifying the main reasons for the three very important characteristic epidemiological features of epidemics of primary pneumonic plague, namely to undergo “spontaneous decline,” be small or tiny, and to give rise to a relatively small number of outbreaks—three epidemiological manifestations of the disease that are, presumably, closely causally linked. Th ere are three main reasons: (1A) Contrary to what is asserted surprisingly oft en, primary pneumonic plague is not strongly contagious and does not easily “recruit” new infective agents who assure the continuation and dynamic development of the epidemic process, except, as under- scored above, under exceptional social circumstances of extreme overcrowding. Again, in the words of Wu Lien-Teh:

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 523

Inasmuch as, in our opinion, an overwhelming majority of cases receive infection through being in the direct range of a cough, we wish to lay stress upon overcrowding, and he goes on to illustrate the point: At Dalinor mines, though the population consisted of 4,000 Chinese and 2,000 Russians, the 1921 epidemic claimed only four Russians out of a total mortality of 1,017. Th e Russians were better housed on level ground, while the Chinese lived in crowded semi- underground dwellings.87 Th e crucial importance of close physical proximity is still the prevailing view. Th e American scholar J.D. Poland makes this point: Primary plague pneumonia occurs primarily in persons in close and prolonged contact with another person with pneumonic plague. Hence, respiratory transmission occurs most frequently to medical personnel or household contacts who are directly involved with the care of the patient.88 (1B) Presumably, this feature is at least to a large extent due to the fact that plague is a bacterial disease and bacteria, being hundreds of times larger than viruses, need much larger and heavier droplets for aerial transportation and will, consequently, normally move much shorter distances through the air. Hirst summarizes recent research showing that droplets of sputum ejected from the mouth tend to fall rapidly towards the ground and that, even when people cough, a surprisingly small number of bacterial colonies develop on culture plates placed only a foot directly opposite the mouth. Th is corresponds well with the fact that most patients give a history of association with a previous case for a period of hours, or even days.89 (2) Another crucial reason for the weak powers of transmission and dissemination of primary pneumonic plague is that this disease is an extraordinarily fast killer, leaving little time for interhuman spread. In 1921, in Harbin (Manchuria) mean duration of the

87 Wu Lien-Teh 1926: 180–1, 298–302. 88 Poland 1983: 1230. See also Meyer 1961: 252. 89 Hirst 1953: 226–7.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 524 chapter thirteen

disease in 1,128 patients was 1.8 days.90 Th is characteristic is confi rmed by observations made during smaller epidemics, for instance, during an epidemic in Vietnam in 1915 comprising twenty-four persons including three medical personnel, all of whom died. It was documented that “death occurred within forty-eight hours of the onset of primary symptoms,” which must be taken as a more inaccurate way of expressing an identical observation. Also in a study of another plague epidemic of about the same size in Vietnam in 1941 it was documented that all cases were dead within forty-eight hours of the onset of the symptoms.91 Importantly, it takes, on average, twenty-four hours before the contagious cough with bloody expectoration develops, which leaves only 0.8 day for transmission.92 Also this clinical feature explains why outbreaks of this disease normally comprise only a small number of cases and will tend to be quite easily extinguished in the incipient or early endemic phase or, quite rarely, early in the epidemic process. Arguably, this clinical feature also represents an independent suffi cient cause. (3) However, the causation of “spontaneous decline” is more com- plicated; it results from the combined eff ect of three usual clinical forms of the disease that do not (or only slightly) lead to development of the cough with bloody expectoration which is the principal vehicle of dissemination of primary pneumonic plague: (3A) As suggested by the extremely rapid course and brief duration of the disease, one such form is quite a frequent incidence of a fulminant course of ordinary cases of primary pneumonic plague leading to the rapid demise of the patient before the contagious cough with bloody sputum has had time to develop. (3B) When larger droplets are inhaled (and as pointed out above, this will quite oft en be the case, see 1B), they tend to impinge on the upper parts of the respiratory tract where they infect the tonsillar region, causing pharyngeal and primary septicaemic plague with extremely rapid death of the patients who will not become contagious since the droplets are not

90 Chun 1936: 318–9. 91 Trong, Nhu and Marshall 1967: 93. 92 Wu Lien-Teh 1922–3: 274; Wu Lien-Teh 1926: 297; Wu Lien-Teh, Chun, and Pollitzer 1934: 80.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 525

deposited in the lungs and cannot give rise to the cough with bloody sputum. (3C) Wu Lien-Teh stresses in particular the importance of a third type of fulminant case called pulmonary plague. Dis- section of lungs of such cases did not show the usual pneumonic foci [which produce the cough with bloody sputum: my insertion] but presented—though undoubtedly infected through the respiratory tract—only hyperaemic93 and oedematous94 changes in the lungs as well as marked septicaemic features. We designated these cases pulmonary plague because, though they were anatomically similar to the septicaemic cases encountered in bubonic plague, they were etiologically diff erent from them, the infection having entered through the respiratory tract. Conspicuously, the incidence of pulmonary plague increased sharply towards the end of the epidemics, while pneumonic cases, i.e. cases with pneumonic foci in the lungs (which give rise to the cough with bloody sputum), became relatively scarce. In Wu Lien-Teh’s opinion, these cases were probably due to an enhanced virulence of the B. pestis developing through passage from lung to lung. We assumed that through such repeated passages the invading organism fi nally became so virulent as to cause mainly pulmonary cases. Th ese—though very fatally infected—are comparatively little infectious, because the disease runs a rapid course. As a consequence, the principal medium of infection, namely, the sputum, is absent. In other words, our post-mortem fi ndings at the end of the 1920–1 outbreak give a scientifi c reason in favour of a spontaneous decline of pneumonic- plague epidemics […]. Being thus rendered less infectious, the “spontaneous decline” of the epidemic, even in foci unattended by sanitary organizations, might be easily explained.95 R. Pollitzer, who was among the medical staff working in the epidemic of 1920–1 and made “a fairly large number of postmortems on lung pest victims,” later confi rmed that he made exactly the same observations as

93 Denoting hyperaemia: an excess of blood in a part, engorgement. 94 Pertaining to, or of the nature of, oedema, aff ected by oedema. Oedema: the presence of abnormally large amounts of fl uid in the intercellular tissue spaces of the body, usually applied to demonstrable accumulation of excessive fl uid in the subcutaneous tissues. 95 Wu Lien-Teh 1926: 189–92; Wu Lien-Teh, Chun, Pollitzer 1934: 86; Wu Lien-Teh 1936b: 420–1.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 526 chapter thirteen

Wu Lien-Teh. Importantly, this was also the case with respect to the rhythm of developments, that dead bodies showing no pneumonic foci were mainly met with at the end of the epidemic. It was therefore tempting to assume that there might be a causal connection between the preponderance of this “non-pneumonic” form of lung pest and the decline of the outbreak.96 In 1942, Pollitzer and Li conducted a complete investigation of an outbreak of quite a usual size of primary pneumonic plague, comprising sixteen cases, which had taken place in a district in North Hunan (China). Th ey made the following observations: (1) “Th e last patient in each of the two fi rst aff ected families had no more bloody sputum.” (2) “In fi ve of the six instances where the infection was carried from these two households into other families, no further spread took place, evidently because the secondary victims had no bloody sputum, in three instances even no cough.” (3) “Further inquiries showed that infection never resulted from contact with patients having neither bloody sputum nor cough. In only two out of fi ft een instances was infection due to contact with patients having cough but no bloody sputum.” (4) Of particular importance in the present context are the following assertions: (a) the outbreak came to an end before measures taken by us could have any eff ect: (b) this decline took place during a rainy spell, when the temperature was comparatively low. Th e fi rst point of the last citation confi rms the spontaneous decline at an early stage that characterizes outbreaks of primary pneumonic plague in which no countermeasures are applied, the second point is of relevance to some extraordinary assertions by Steff ensen and Karlsson as to the importance of the climate in Iceland (see below). Pollitzer and Li make the following general conclusion: More or less rapid evolution of a “non-pneumonic” type of lung pest, characterized by absence of the usual vehicle of infection, the bloody sputum, and oft en also of cough, appears to be of causal importance in the decline of pneumonic plague epidemics.97

96 Pollitzer and Li 1943: 215. 97 Pollitzer and Li 1943: 212–6.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 527

Outbreaks of primary pneumonic plague have been shown above to be generally small or tiny, comprising only a paltry fraction of the number of cases or level of population mortality envisaged by Steff ensen and Karlsson. Th e medical and epidemiological reason for this feature are also clear: the high frequency of various types of fulminant non- contagious forms that do not contribute to further dissemination of the epidemic and therefore break the chain of transmission. Most important among these forms is the appearance at an early stage of outbreaks of a strong increase in the incidence of cases of pulmonary plague that leads to a rapid decline and disappearance of the epidemics, a phenom- enon designated “spontaneous decline.” Th ese non-contagious forms arise from a combination of the virulence and size of the plague bacterium and the mode of respiratory transmission. Th us, they constitute a basic property and characteristic feature of this modality of plague which mould all known outbreaks in the past and in modern times into clearly defi ned medical, epidemiological and demographic patterns. Th ese patterns diverge radically and decisively from the patterns envisaged by Steff ensen’s theory that the two Icelandic epidemics of 1402–4 and 1494–5 were primary pneumonic plague and Karlsson’s theory to the eff ect that these two Icelandic epidemics were pure epidemics of primary pneumonic plague and that this was a general feature of the Black Death in Europe. Th eir theories are incompatible with the corpus of knowledge of the medical and epidemiological causes of the rise and course of the outbreaks of primary pneumonic plague. Th is explains also why there is no epidemic of primary pneumonic plague on the record with a duration of two or three years as in the case of the Icelandic epidemics of 1402–4 and 1494–5: the clear and systematic tendency of spontaneous decline would prevent this from happen- ing. Th us also the duration of the Icelandic epidemics is a strong argument against their being primary pneumonic plague; that the notion of pure epidemics of primary pneumonic plague is fi ctitious is another matter. Finally, I would also like to point out that in the plague history of Norway from 1348–654, covering more than thirty waves of bubonic- plague epidemics, there is not a single instance of an epidemic that lasted so long in the same region, which is also a strong argument against this Icelandic epidemic having been bubonic plague.98

98 Benedictow 2002.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 528 chapter thirteen

Th e Icelandic Climatic eoryTh of Primary Pneumonic Plague

Steff ensen, who bases his theory on the hypothesis that the supposed Icelandic plague epidemics started from bubonic plague victims who developed secondary pneumonia, accepts the fact that there were rats in Iceland and thus avoids the impossibilities and absurdities of Karlsson’s theory as to the origin of epidemics of pure primary pneumonic plague. However, he cannot avoid the issue of the exceptional powers of spread that are not observed anywhere else, and must construct a unique Icelandic case based on assumptions of climate. Steff ensen argues that bacteria harboured in the mucus of the respiratory tract derive from it some protection against dryness. In damp and cold air, such drops of moisture can preserve living bacteria for months on end, especially if there is frost. Under Icelandic conditions, therefore, it is quite possible for living plague bacteria to remain in tiny drops of moisture on, for example, bedclothes and personal clothing, long aft er the patient’s death, and for drops to be stirred up again and carried down into the lungs of the person, for example, who is about to bury the body […] there is, therefore, no apparent biological reason why an epidemic of primary pneumonic plague could not have occurred in Iceland.99 Steff ensen does not consider this statement a composite working hypothesis constituted by several hypothetical elements each depend- ent on empirical support or corroboration for its viability. In his understanding of scholarly or scientifi c work the mere proposal of a hypothesis confers tenability on it without empirical support or references to research literature. Th is is obviously fallacious. Th e research literature on primary pneumonic plague contains no support, not even a unique case or sporadic case histories of this type of dissemination. No case has been observed of a person who has contracted primary pneumonic plague from plague-contaminated droplets in clothing or bedclothes left by patients at any time or in any climatic zone. This might constitute a pure epidemic of primary pneumonic plague which, as shown above, has never been observed or ascertained in any scholarly study of this disease. “Damp and cold air” is very far from being a unique Icelandic type of climate, as the circumstances of the epidemic of primary pneumonic plague studied by Pollitzer and Li in Hunan Province show (see above). Th e coasts of western and northern Norway

99 Steff ensen 1974: 41.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 529 are also characterized by “damp and cold” climate for a large part of the year, but no epidemic of primary pneumonic plague can be ascertained from the sources; all Norwegian plague epidemics conform to the pattern of spread and seasonality of bubonic plague. Th ere is also another crucial point that Steff ensen has failed to consider and which undermines or even invalidates his theory, namely, the eff ect of extraneous conditions on the virulence of plague bacteria. As stated by Pollitzer and Meyer: it is generally agreed that, although […] plague-contaminated objects may remain infected for some time, the chances of their proving infective, thus becoming responsible for the spread of the disease, are usually remote. Plague […] in its pneumonic form […] is almost always the result of a direct infection contracted in the immediate vicinity of patients suff ering from secondary or primary lung involvement.100 Similarly, Wu Lien-Teh states specifi cally in relation to primary pneumonic plague that: the unsuitability of the plague bacillus for prolonged existence as a sapro- phyte and the unlikelihood of rodent or man contracting direct infection from contaminated inanimate objects, rule out the free existence of the causative organism as a means for preservation of the disease.101 Th is quotation is also given in my doctoral thesis102 in the discussion of this matter, however, to no avail, as Karlsson simply ignores it. Steff ensen argues his case for primary pneumonic plague in Iceland in terms of what is possible under particular Icelandic climatic conditions: his concluding sentence is couched in the formal language of hypothesis formation. But he has built a model which cannot be tested, because no Icelandic material is available that allows the identifi cation of the form of plague or epidemic disease affl icting the population. For this situation the following methodical rule applies: the inability to test an explanatory model does not invalidate it, but does render it useless as a scientifi c hypothesis. Th e justifi cation for constructing this explanatory hypothesis, model or working hypothesis would have been strengthened if at least a sprinkling of this type of endemic cases had been observed elsewhere. But Steff ensen’s is based on pure specula- tion, it has no evidentiary basis, no validity, no tenability. Its use in the

100 Pollitzer and Meyer 1961: 434–5. 101 Wu Lien-Teh 1936a: xxviii. 102 Benedictow 1993/1996: 216.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 530 chapter thirteen construction of an alternative theory of plague in Iceland can only serve to undermine the theory, and claims that such a theory of plague is superior to other empirically supported theories are obviously invalid.103 Conspicuously, later in his paper, Steff ensen argues that the clinical elements in the account of the Lawman’s Annal on the Black Death in Bergen in 1349 show that it was an epidemic of primary pneumonic plague, an argument that will be addressed below and shown to be untenable. Here the following point shall be made: by turning to Norway, Steff ensen greatly extends the scope of climatic conditions that allow large-scale epidemics of pneumonic plague. Whilst he origi- nally stresses the uniqueness of the Icelandic climate, making a claim for Iceland as a unique case, the extreme weakness of his case has led him to a line of climatic argument which includes a much wider range of territorial relevance, covering quite large parts of northerly Europe including northern England and Scotland. Th is makes the rarity of epidemics of primary pneumonic plague, if indeed any occurred, even more hard to explain, as no such epidemic, as mentioned above, has been confi rmed in any pre-industrial society. Steff ensen’s and Karlsson’s assertions are without empirical medical or epidemiological support in the most accurate meaning of the terms. It has been possible to identify several independent suffi cient causes for rejecting their theories.

Mortality Rate of the Purported Plague Epidemics in Iceland

Karlsson makes two central assertions with respect to the mortality caused by the Icelandic epidemics: (1) that the mortality rate in the epidemic of 1402–4 was 60 per cent and in the next epidemic of 1494–5 it was 53 per cent, and (2) that these mortality rates constitute defi nite proof that the epidemics were primary pneumonic plague.104 Th e massively documented fact is that no epidemic of primary pneumonic plague is known which has caused a population mortality exceeding 1 per cent. Th is is the case for very concrete and strong

103 Benedictow 1993/1996: 223–5. 104 Karlsson 1996: 268–76. In the second case, Karlsson quite arbitrarily moderates his mortality estimate for the second epidemic to 30–50 per cent, since he recognizes that the evidence is very poor to put it mildly.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 531

medical and epidemiological reasons. A level of population mortality of 60 per cent thus constitutes certain proof that the Icelandic epidemics must have been a diff erent disease or conjunction of diseases. However, Karlsson’s estimates of mortality must be rejected. Th e texts of the Icelandic annalists or chroniclers on which these estimates are based are cited in English translation in Appendix 2. Th e sugges- tion that medieval chroniclers provide accurate information on distribution, mortality rates and areas aff ected cannot be taken seriously, as shown repeatedly above and again below.105 Hatcher disregards the “overwrought imaginings and hopelessly inaccurate quantifi cation of the chroniclers.”106 Just to take a couple of illustrating examples of the nonsense of fi gures in Icelandic annals: in the Annal of Skálholt, it is related that in 1305 that King Håkon V of Norway went to war with Sweden with an army of 40,000 warriors, which is without any basis in the known facts on Norwegian military manpower at the time, to put it mildly. Compared with England’s contemporary population of around six million, it would correspond to an assertion that the king of England went to war with 700,000 warriors. Th e Lawman’s Annal asserts that only fourteen persons survived the Black Death in London.107 Instead, as Hatcher states, “advances in knowledge and technique have enabled us to proceed […] to a position where the majority of historians are agreed upon a reasonably narrow range of probabilities.” Only mortality fi gures based on population records that can be examined and eval- uated according to the principles of demography are used in my monograph on the Black Death.108 Chroniclers’ assertions cannot be accepted, especially their assertions of numerical and demographic data, if source-criticism cannot be applied and the possible tenability at any level can be ascertained by independent material evidence. In this perspective it could be of interest that there are numerous

105 See, for instance, above: 78–84, 219–22, below: 589–92. 106 Hatcher 1977/78: 21. 107 Islandske Annaler 1888: 200, 275. 108 One partial exception is the chronicle of Giovanni Villani. which can be considered usable, because it can be shown that he takes real interest in collecting statistical information on his beloved city of Florence and that the city’s authorities produced statistically usable population registers, for instance of taxable householders, that he has real competence with regard to statistical data, that his estimates are based on empirical material of various kinds, and because his estimate of mortality in the Black Death can be tested against various types of population records. His approach is proto- scientifi c, refl ecting his central background in the Early Renaissance in Florence, and therefore contains valuable empirical and statistical material.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 532 chapter thirteen demographic studies showing that epidemics of bubonic plague produce mortality rates at the level of 60 per cent.109 Karlsson does not mention that Steff ensen, on the basis of the same sources, conjectures a population mortality of one-third,110 about half of Karlsson’s estimate. Since Steff ensen honestly characterizes his estimate as a conjecture, i.e., a guess or guesstimate, the margins of error are obviously huge. Th us, Karlsson asserts that he has found solid ground for an accurate mortality estimate on the basis of sources that Steff ensen thinks permit only a conjecture or qualifi ed guess of a much lower level of population loss. Th is puts in perspective Karlsson’s con- tention that the margin of error could be a ludicrously low 10 per cent and the fact that this has been accepted by the editors and consultant(s) of the Journal of Medieval History.111 Th e diff erence between Steff ensen’s and Karlsson’s estimates is 100 per cent, representing a corresponding margin of error. Karlsson incredibly maintains that the annals’ indication of area aff ected, distribution and level of mortality rates are accurate within margins of a couple of per cent, which soundly beats the margins of error associated with modern population censuses based on statistical data systematically collected by trained personnel. In view of the fact that margins of uncertainty associated with the various elements that constitute estimates relate multiplicatively to each other, the margins of error are much higher. Any scholarly critical examination of the annals and other evidence cited by Karlsson will demonstrate the justifi cation for at least a modicum of prudence as shown by Steff ensen. In my opinion, they permit no useful quantifi cation of a specifi c mortality rate at all, only the impression of serious and locally also severe mortality (see below, Appendix 2). Th is perspective is corroborated in the most recent evaluation of population losses in the two fi ft eenth-century epidemics based on agrarian studies relating to northern and eastern Iceland where it is stated that they indicate a limited abandonment of farms in 1420–50 and 1495–1550 when some 20–25 per cent of the farms were deserted. Th e eff ect was particularly notable on the coast but it was temporary and probably caused by

109 Benedictow 2004: 245–384. 110 Steff ensen 1971: 54. 111 Karlsson 1996: 276.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 533

outbreaks of plague in 1402–4 and 1494–5. In the fi ft eenth century climatic deterioration should also be taken into account, the westernmost parts of Iceland, on the other hand, saw population growth and settlement expansion from about 1300 and show no signs of recession at all in the three following centuries.112 Clearly, Karlsson’s “estimates” fl y in the face of the sources and empirical studies on settlement developments. Th e settlement studies suggest a population loss which is much lower; this shows that Karlsson’s “estimates” are wildly exaggerated, the predictable outcome of a basically arbitrary approach. Mortality at the level of 15–20 per cent can be caused by a number of epidemic diseases or conjunction of more or less concomitantly spreading epidemic diseases, especially in the epidemiological circumstances characterizing Iceland at the time.113

Summary: Why Th ere Never Was a Plague Epidemic in Iceland

Th e presentation of the characteristic medical and epidemiological features of primary pneumonic plague based on the complete corpus of relevant research constitutes a broad basis for comments on Morris’s alternative theory of late medieval plague as being primary pneumonic plague and on Karlsson’s use of this theory in his paper on the epidemics that ravaged Iceland in 1402–4 and 1494–5. It has been shown that their assertions are unfounded and erroneous on the following main points: (1) Th ere is no such thing as pure epidemics of primary pneumonic plague, i.e., independent of an infected murine basis. No case of an epidemic of pure primary pneumonic plague has been identifi ed in the entire corpus of historical studies of plague. All assertions to the contrary are empirically unfounded and arbitrary. (2) Since there were no rats in Iceland, according to Karlsson, no epidemic of plague of any kind could have arisen in the island, pure or not. (3) It is not correct that assertions in the sources to the eff ect that plague victims typically died in the course of three days of illness are indicative of primary pneumonic plague. Th e average duration

112 Vahtola 2003: 575–6. 113 Cf. Cliff and Haggett 1985; McArthur 1968.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 534 chapter thirteen

of the course of primary pneumonic plague is 1.8 days, and patients normally die within 48 hours. One would also expect that the frequent incidence of even much shorter fulminant cases would attract attention by their frightening character and tend to acquire prominence in contemporary accounts. Th ree days are within the normal range of the course of illness of bubonic plague, namely three to fi ve days,114 and an average duration of plague disease within this range is therefore a clear indication of bubonic plague.115 (4) It is not correct that all victims of primary pneumonic plague develop infectious cough with bloody expectoration. On the contrary, epidemics of primary pneumonic plague are characterized by large and increasing proportions of cases of several clinical types which do not cause development of a cough with bloody expectoration in the diseased, who therefore do not pass on the infection.

114 Chun 1936: 313. Above: 279–80. 115 Morris 1971: 206–7, maintains erroneously that the average duration of cases of primary pneumonic plague is three days and of cases of bubonic plague is fi ve days. Since he does not support his view with a footnote, the basis or source of this assertion cannot be ascertained and examined. He refers to Gui de Chauliac’s account of the Black Death in Avignon where this contemporary observer discerns between a fi rst phase, in which the course of illness lasted three days, and a second phase where it lasted fi ve days, as proof that the fi rst phase was primary pneumonic plague. However, this is not evidence to this eff ect, since both indications of duration of illness are within the normal range of bubonic plague. Since cough and bloody expectoration of secondary pneumonic plague cases normally constitute a signifi cant but variable proportion of cases of bubonic plague, there cannot have been any phase without a notice- able incidence of this clinical feature, but there can have been substantial variation or change in the incidence. It is also not correct, as Morris contends or implies, that primary pneumonic plague can be the origin of epidemics of bubonic plague; the in- correctness of this assertion is demonstrated by the experience in the Manchurian epidemics. Although there were 50,000 cases in Manchuria itself and the dwellings of the Chinese coolies that constituted the great majority of cases swarmed with fl eas and other blood-sucking parasites, bubonic cases that could have been caused by fl eas infected by pneumonic cases were extremely few and isolated and only one case can be considered quite certain; in the words of Wu Lien-Teh: “practical experience has shown that pneumonic plague rarely caused the bubonic form,” and “Not so much the existence as the rarity of such cases calls for comment.” 1926: 186–7. Th e rstfi phase of plague in Avignon was, therefore, probably a mixed epidemic in which the primary pneumonic part faded away due to the characteristic “spontaneous decline” of this modality of epidemic plague and a more usual epidemic of bubonic plague emerged. Benedictow 2004: 236–8. No instance of a primary plague epidemic that develops into a bubonic plague epidemic has ever been recorded.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 535

(5) It is therefore not correct that epidemics of primary pneumonic plague have great powers of spread. On the contrary, epidemics of primary pneumonic plague are characterized by low diff usibility and “spontaneous decline.” Outbreaks of primary pneumonic plague are small or tiny, population mortality rates are consistently minuscule, and even the largest epidemic of this type, the fi rst Manchurian epidemic of 1910–11, occasioned a population mortality rate of only 0.4 per cent. Mortality rates of the size asserted by Morris for England and by Karlsson (and Steff ensen) for the Icelandic epidemics of 1402–4 and 1494–5 are unknown in the complete corpus of research on primary pneumonic plague, and by wide margins. Th eir assertions to the eff ect that the English or the Icelandic epidemics swept away large parts of the respective populations are powerful arguments against the notion that they could have been primary pneumonic plague. (6) No epidemic of primary pneumonic plague or of bubonic plague has been identifi ed in the northerly parts of Europe 1402 or 1494, the years when the two epidemics broke out in Iceland or in the preceding years. Th ere is therefore no way plague contagion in any form could have been transported to Iceland in these years. (7) Karlsson’s theory of a mutant version of plague disease specifi - cally designed to suit his arguments is arbitrary, and it is also fallacious, since it is not developed according to the requirements of evolutionary theory and the obligation to uncover and present the mechanisms of selection. Th e territorial origin of a specifi c mutant version of primary pneumonic plague of the purported Icelandic epidemics is a crucial question left unanswered; its origin is also therefore arbitrary. Also unexplained is the crucial question of whether the same mutant strain really could have arisen twice, or where on earth it took refuge in the meantime. No such mutant strain has been identifi ed by modern DNA studies of genetic material taken from plague graves. All of these crucial assertions by Stephenson and Karlsson are arbitrary and untenable. According to the normal methodological and empirical considerations and assumptions of scholarly work, these seven medical and epidemiological points settle the case, as they contain as many independent suffi cient grounds for invalidating the theory. Th ere never was a plague epidemic without rats in Iceland. Th ere never was a plague epidemic in Iceland.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 536 chapter thirteen

Was the Black Death in Bergen (Norway) 1349 Primary Pneumonic Plague?

Karlsson states explicitly, as mentioned above, that the central objective of his paper is to defend Steff ensen’s paper against my criticism of it and his theory that plague in Iceland was primary pneumonic plague. He is especially outraged at my criticism of Steff ensen’s presentation and analysis of the clinical description of plague cases given in the Lawman’s Annal’s account of the Black Death in Bergen in 1349 and of his conclusion that it describes an epidemic of primary pneumonic plague. Th is account and another given in the so-called Fragment of the Annal of Skálholt, which is closely related but not identical,116 are normally and reasonably assumed to be based on accounts of two prominent Icelandic prelates who stayed in Bergen at the time of the Black Death and returned in 1350 and 1351, respectively.117 However, other Icelanders who returned, some quite likely in 1350,118 may have added independent information. Karlsson aggressively defends Steff ensen’s views; I consider some of his arguments personally off ensive, but will refrain from commenting on them, as the facts speak for themselves clearly and loudly. One should note that there are crucial problems not only with regard to the interpretation of the clinical descriptions of the Black Death in Norway provided by these Icelandic chronicles. Tacitly, Steff ensen assumes that it is unproblematic to use the Lawman’s Annal’s account of the Black Death in Bergen in 1349 for the purpose of corroborating his views on the nature of the epidemics that spread in Iceland in 1402–4 and 1494–5 as primary pneumonic plague (see Appendix 2). Th is inevitably implies, as pointed out above, a sudden abandonment of his argument of the specifi city of Icelandic climate as a necessary condition for the spread of large-scale epidemics of primary pneu-

116 Islandske Annaler 1888: 224: “þat uar kyn sottarinnar at menn lifdu iij dægr med hardan stinga þa toku menn blodspyu ok for þar med onndin.” Th e text is cited without indicating the expansion of abbreviations. Cf. below: 551 and fn. 151. 117 Benedictow 2006: 88, 101–2. 118 In the Lawman’s Annal it is mentioned that Th orlak’s ship at quay in Bergen was ready to leave for Iceland with a number of named Icelanders on board, but was so severely hit by the Black Death that it was prevented from putting to sea. Quite likely, this ship returned to Iceland in the shipping season of 1350 with survivors and other passengers and crew members on board who had their own stories to tell. Benedictow 2002: 67–72.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 537 monic plague. In addition, his approach is problematic in several other important respects that he does not address: (1) Th e Icelandic annals give no clinical information about the two purported fi ft eenth-century epidemics of primary pneumonic plague. Consequently, there is no basis for valid comparison with the Black Death in Bergen or elsewhere, they are not in pari mate- ria. Steff ensen’s use of comparative analysis is fallacious. (2) Steff ensen gives a highly problematic translation of a crucial part of the text which in a decisive way seemingly strengthens his case. Th is can be demonstrated by citing both the original text from the Lawman’s Annal,119 as I did in my thesis,120 and Steff ensen’s and my translations, italics showing the crucial bit of contested translation relating to the two original Norse words “hordum stinga,” with the basic meaning of sharp, pricking, painful sting (literally “hard sting”), which is here associated with the clinical manifestations of the Black Death in Bergen.

– Benedictow: Th e disease was such that men did not live for more than a day or two with sharp pangs of pain, then they began to vomit blood, and then they expired. – Steff ensen: Th e nature of the sickness was such that men lived not more than one day or two, with great pleuritic pain, whereupon they began to vomit blood and then the spirit passed on its way.

Steff ensen uses the medical word “pleuritic” in his translation, although no word with this meaning occurs in the source. To add one or more crucial words to the translation of a text, without making it clear to the reader and providing a persuasive reason for it, violates one of the basic methodological conventions of historical science. Conceivably, it could be a matter of interpretation or connotations, but then, according to the conventional principles

119 Islandske Annaler 1888: 275–6: “þat var kyn sottarinnar at menn lifdu eige meirr en eitt dægr edr tuo. med hordum stinga. eft ir þat sætte at blod spyiu ok for þar ondin med sinn vegh.” Th e text is cited without indicating expansion of abbreviations. 120 Th e text is almost identical with the text of the Fragment of the Annal of Skálholt; see above: fn. 116, and below: 551 and fn. 151.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 538 chapter thirteen

of historical methodology and source-criticism, the reason should be explained and discussed openly, and not, as in this case, introduced tacitly or covertly. I simply cannot understand how Karlsson or other Icelandic scholars can defend this procedure. It is pre- cisely the addition of the medical term “pleuritic” which permits him to identify the Black Death in Norway as an epidemic of primary pneumonic plague. He believes that he can assert that primary pneumonic plague is directly and normally associated with great pleuritic pain without any reference to the medical literature on primary pneumonic plague which confi rms or supports that this is the case. (3) Steff ensen and Karlsson do not consider it necessary to base assertions regarding the clinical and medical manifestations of (primary pneumonic) plague on the primary research literature or the standard textbooks on plague. Th eir unfl inching confi dence in their assertions that cases of primary pneumonic plague are clinically characterized by “great pleuritic pain” stand completely unsupported by any reference to the medical literature on primary pneumonic plague. As such, these assertions are arbitrary according to the standard conventions of methodology of historical, social and medical sciences. If empirical facts count, the medical research literature on primary pneumonic plague should be consulted. In his excellent treatise on primary pneumonic plague, the only monograph on this disease, Wu Lien-Teh discusses relevant points in this connection at some length. He does not state that cases of primary pneumonic plague are characterized by “great pleuritic pain.” He points out that there may initially be “no pain in the chest, only a feeling of constriction when breathing and coughing” and that “later on, these pains are usually more marked,” but that the pain is “never so severe as in ordinary pneumonia.”121 Th us, according to Wu Lien-Teh, some pain, but not great pain, is characteristic of cases in advanced stages of primary pneumonic plague. However, importantly, this element of pain is not associated with pleuritic pain caused by infl ammation of the pleura, the membrane of the thorax and lungs. Th is description, some pain but not pleuritic pain, conforms and accords with other relevant comments. He underlines also that in cases of primary pneumonic plague there is

121 1926: 255, 259.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 539

a “marked disproportion” between slight lung changes and serious general signs (of illness) which is “of great diagnostic value.” Th e crucial point is that pleurisy is characterized by “infl ammation of the delicate membrane of the thorax and the lungs, causing severe pain in the chest or sides,” which is not a characteristic clinical feature of primary pneumonic plague.122 Instead, Wu Lien-Teh associ- ates the small elements of pain in primary pneumonic plague with dyspnœa (diffi culty of breathing) caused primarily by “the action of plague endotoxins upon the respiration” and “changed blood circulation.”123 In the most important general standard work on plague which is co- authored by Wu Lien-Teh, Chun, Pollitzer and C.Y. Wu, undoubtedly the most competent group of experts on primary pneumonic plague ever to come together on the same plague-related project, Chun, who wrote the part on clinical manifestations, states: “Th e patient complains of pain and a restricted feeling in the chest, but this is not severe […] he appears to suff er little,” and goes on to summarize the point clearly and unambiguously: Th ere is not much pain in connection with plague pneumonia in contrast to ordinary pneumonia, though a few exceptions were observed.124 Chun stresses also, in accordance with Wu Lien-Teh’s description, that the “physical signs in the lungs are oft en slight, even in cases well advanced in the disease,” which, as pointed out, contrasts sharply with the condition of pleurisy. Th ese statements by Chun must be assumed to have been accepted as adequate by his co-authors and are compatible with Wu Lien-Teh’s statements. Th is is confi rmed by Pollitzer who in his own general standard work on plague of 1954 explicitly agrees with Chun’s views on these points, and even cites him quite extensively.125 As shown above, Pollitzer had a leading role in the combat of the second Manchurian epidemic of primary pneumonic plague and also performed numerous autopsies. He also carried out studies of later outbreaks of this disease. Th e fact that there are rare victims of pneumonic plague who suff er considerable pain cannot be assigned signifi cance in this context in connection with

122 Oxford Advanced Learner’s Dictionary of Current English 1989: 950. 123 Wu Lien-Teh 1926: 259. 124 Chun 1936: 318–20. 125 Pollitzer 1954: 441–2.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 540 chapter thirteen the quite generalized accounts of the Icelandic annals. Furthermore, this pain does not at all accord with the character of the two Icelandic scholars’ assertion with regard to “great pleuritic pain.” Th e slight signifi cance of pain associated with lungs or the chest in normal cases of primary pneumonic plague fi nds also an indirect expression in the fact that several of the leading standard works on plague do not mention this subject at all, presumably because it is not a signifi cant clinical feature.126 Consequently, Steff ensen’s addition of the word “pleuritic” and Karlsson’s defence of this addition are not only fl agrant breaches of the basic principles of source-criticism and scholarly work, but their inter- pretations that this pain indicates the presence of primary pneumonic plague are factually wrong according to the corpus of primary studies on primary pneumonic plague and the summaries given in standard works. I do not understand on what grounds Karlsson accuses me of not recognizing that “stinga” can refer also to strong pleuritic pain caused, for instance, by pneumonic tuberculosis or pneumonia.127 My point is that the clinical picture of primary pneumonic plague is not characterized by great pleuritic pain or sharp pain in the lungs or chest more generally, and that the term “hordum stinga” for this reason cannot refer to primary pneumonic plague and “great pleuritic pain” caused by this disease. “Stinga” just means “sharp pain,” and can refer to such pain in any part of the body. It is the same word as the English “sting” and has in medieval Norse much the same meaning as in modern English. Sting or “stinga” can be felt any place on the body that is stung by an insect, for example. Karlsson mentions a source that refers to “stinga” “under the arm,” which certainly could refer to a plague bubo in the second most usual site, namely an axilla. However, since the source is earlier than the arrival of the Black Death in Europe, this cannot be the case. Instead, this source refl ects the elementary fact that infections of almost any origin can in special circumstances be drained through a lymph tract to one of the sites of clusters of lymph nodes and can there give rise to sharply painful or intensely tender swollen lymph-nodes of the type called bubo or boil. It is the epidemic incidence of swollen lymph-nodes in the form of buboes that is a defi ning feature of bubonic plague.128 Clearly, according to the principles of source-criticism, only

126 See for instance Hirst 1953, Langen and Lichtenstein 1936. 127 Karlsson 1996: 281. 128 Above: 203–05.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 541 the use of the word “stinga” in sources produced aft er the arrival of the Black Death or subsequent plague epidemics can be of value for understanding the meaning and connotations of this word associated with plague disease. Karlsson has found only one instance, which occurs in a fi ft eenth-century Icelandic manuscript in a text about the treatment of plague which refers to “stinga” with “extreme pain and a boil/bubo.” Karlsson passes easily, even hastily, by it and prefers to emphasize the signifi cance of the few pre-plague uses of the word.129 However, this is strong evidence to the eff ect that extreme pain is associated with the buboes of bubonic plague and that the word “stinga” at the time was considered suitable in Iceland for the description of the pain in plague buboes. Karlsson is extremely critical of my translation of “hordum stinga” with an expression using the plural form “pangs” of pain when the original is in the singular form. He accuses me of making a deliberate mistranslation in order to construct a false basis for a theory that this pain could be associated with other parts of the body than the lungs and, thus, that the clinical descriptions of the Icelandic annals could refl ect bubonic plague.130 I am not a native speaker of English, but a foreigner who has worked hard to achieve a usable knowledge of academic English for the purpose of communicating with the international community of scholars in the fi eld of the history of disease and demography. I may not make accurate translations in all contexts which successfully incorporate the right connotations and nuances whatever my ambition and endeavours. However, the point for me in using the plural form was to indicate continuous sharp pain in order to avoid what I assumed would be the implication in English of using the singular form, namely that there was one sharp pang of pain which then disappeared. Notably Cohn, who is an (American) English speaker, uses the same plural form of the term when describing the type of pain associated with plague buboes, namely “sharp pangs.”131 However, some central points still await full clarifi cation. Now for the hard evidence. Steff ensen and Karlsson ignore standard medical works on plague, although the crucial question of whether or not the descriptions provided by the Icelandic annals of the clinical manifestations of the Black Death in Bergen in 1349 is compatible with primary pneumonic plague

129 Karlsson 1996: 282. 130 Ibid. 131 Cohn 2002: 24.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 542 chapter thirteen or bubonic plague and buboes can only be decided according to the information provided there. It has been shown that the expression “hordum stinga” cannot relate to pleuritic pain in cases of primary pneumonic plague, and Steff ensen’s and Karlsson’s assertions to this eff ect are arbitrary and untenable. However, it remains to be seen whether or not the standard works on plague which summarize the great mass of clinical evidence on bubonic plague confi rm my opinion that plague buboes are extremely painful in a way that corresponds with the clinical descriptions given in the Icelandic annals on the Black Death in Bergen 1349 and can be responsibly translated by “sharp pangs of pain” indicating repeated or continuous sharp pain as a normal clinical manifestation. Th e following citations are arranged according to the chronology of modern scientifi c standard works on plague and represent to my knowledge a complete and therefore also a representative collection of statements on the matter: (1) Simpson 1905: If a bubo appears it is usually ushered in by intense pain in the groin, arm-pit, or neck which is increased by movement or pressure. Th e pain at fi rst is so severe that the attention of the patient is mainly directed to it, all other symptoms being considered insignifi cant compared to the suff ering experienced in the gland aff ected. Th e pain is followed by a swelling which constitutes the bubo.

Pain, tenderness and swelling are the general characteristics of the bubo. Th e pain may be dull and aching or sharp and stabbing […] and is independent of the size of the bubo. Th e smaller the size of the bubo the more painful is it likely to be […]. Sometimes the tenderness is so acute that pressure over the bubo will cause wincing and moaning from pain even when the patient is in a comatose condition.132 Simpson describes buboes as generally intensely painful to a degree that dominates completely the patient’s attention. Obviously, this is a clinical manifestation which could be observed by ordinary medieval persons. In fact, the pain is so intense that even light pressure on the bubo will make patients in coma and therefore deeply unconscious wince and moan. Th e pain is characterized by being intense, sharp and stabbing and is triggered by movement or pressure, which, in my opinion, justifi es the use of

132 Simpson 1905: 263, 274.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 543

the word “sharp” and, because it almost inevitably will by triggered repeatedly, use of the plural form “pangs” of pain is also justifi ed. (2) Chun 1936: Th e glandular mass [of a bubo: my insertion] is very painful and tender to the touch. For this reason, in order to avoid pain caused by movement of the limbs “If the bubo is in the groin, the leg on the aff ected side is drawn up, somewhat abducted and twisted outwards; if it is in the axilla, the aff ected arm is abducted from the trunk; during examination it is oft en supported by the sound arm.”133 According to Chun, plague buboes are “very painful.” His description of how patients keep arms or legs out from the trunk of the body in order to avoid pain by movement justifi es my translation in plural, “pangs” of pain, since it means that patients will experience strong pain every time a limb associated with a bubo is moved, as will inevitably occasionally occur. Th e characteristic position of limbs associated with buboes would easily be noted also by medieval contemporaries and induce questions as to the reason. (3) De Langen and Lichtenstein 1936: Th e buboes are usually visible on the second or third day and are as a rule very painful. If the patient is lying down, we can oft en see from a distance whether there is a bubo or not, and where, because of the typical but unusual position (according to the place of the bubo) of the arm, head or leg, which is so held as to give the least pain. […] it is usually most painful as long as it is small and hard and the expressive movements of the patient during examination betray his fear that the bubo will be touched […]. Th e pain of the bubo is greatest during the fi rst days when it is still small.134 Th us, these two Dutch scholars with their background in the plague epidemics in Indonesia entirely agree with Simpson and Chun who draw on research in India and China respectively. Again the point is made that repeated intense pain of buboes is caused by movement of associated limb or by touch, which justifi es my use of the plural form “pangs” (of pain).

133 Chun 1936: 316. 134 De Langen and Lichtenstein 1936: 192–3.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 544 chapter thirteen

(4) Hirst 1953: Th e size of the bubo varies from that of an almond to an orange, and, until it begins to discharge, may give agonizing pain.135 Th us Hirst, with his enormous experience with cases of bubonic plague, gives a description of buboes that justifi es the description “pangs of pain” in the plural form, since Hirst clearly has not in mind the single sharp sting as that of a wasp or gadfl y but continuous agonizing pain. De Langen and Lichtenstein and Hirst, who do not mention pain associated with pneumonic plague, presumably because they considered this clinical feature insignifi cant, both emphasize that buboes are continuously very painful or ago- nizingly painful which they consider a signifi cant clinical point. (5) Pollitzer 1954: As is unanimously stated in the literature and has been borne out by the fairly ample fi eld observations made by the present writer, manifestations on the part of the primarily-aff ected lymph-nodes appear, as a rule, quite early in the disease, simultaneously with or even before, the onset of fever […]. Even when seen within the fi rst hours of illness, as a rule the patients already complain of discomfort or even of some pain at the site of the future buboes. Th e subjective sensations produced in the patients by the presence of plague buboes may vary considerably, some complaining of constant dull or even stabbing pains, others feeling little or no local discomfort as long as they do not move and keep in a position which lessens pressure on the aff ected parts of the body. Patients with a groin bubo keep for this purpose the corresponding thigh fl exed, those with axillary buboes lie on their back and hold the aff ected arm away from the trunk, and those with cervical buboes hold their head still and inclined to the aff ected side. Generally speaking, pain is more marked in the case of smaller buboes than in that of larger ones […] during the acute stage of the disease the buboes are invariably sensi- tive to touch, oft en to such a marked degree that even comatose patients react when the aff ected lymph-nodes are touched. Th is marked tenderness to touch, which as a rule, appears in the earliest stage of illness […] forms one of the most outstanding clinical features of bubonic plague.136 Pollitzer uses the plural form “stabbing pains,” which puts in perspective Karlsson’s off ensive accusation. Also Pollitzer describes

135 Hirst 1953: 29. 136 Pollitzer 19954: 421–2.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 545

buboes as intensely painful, and says that as a result patients keep limbs associated with buboes out from the trunk of the body in order to avoid as much as possible pain from movement, and he also makes the same comment as Simpson to the eff ect that the pain is so intense that patients in coma will react when their lymph-nodes are touched. Again my use of the plural for “pangs [of pain]” is justifi ed. (6) Cantey 1974: Th irteen [Vietnamese] patients were hospitalized with bubonic plague […] Th e [lymph] nodes were exquisitely tender to palpation […].137 Th is refers to studies of Vietnamese patients during the Vietnam War, when the last large-scale epidemics of plague (with as many as 10,000 deaths a year138) gave American physicians the last opportunity up to now of studying great numbers of patients. Butler also has his personal experience from this context. (7) Butler 1983: Patients are typically aff ected by sudden onset of fever, chills, weakness, and headache. Usually, at the same time, aft er hours, or on the next day, patients notice the bubo, which is signaled by intense pain in one anatomic region of lymph nodes, usually the groin, axilla, or neck. A swelling evolves in this area that is so tender that the patients typically avoid any motion that would provoke tenderness of the aff ected lymph nodes. For example, if the bubo is in a femoral area, the patients will characteristically fl ex, abduct, and externally rotate the hip to relieve pressure on the area and will walk with a limp. When the bubo is in an axilla, the patients will abduct the shoulder or hold the arm in a splint. When a bubo is cervical in location, patients will tilt heir heads to the opposite side. Th is extreme tenderness of the buboes naturally causes patients to resist physicians’ attempts to palpate and aspirate their buboes […]. Palpation will typically elicit extreme tenderness.139 (8) Butler 1993: At about the same time [the onset of disease], patients notice the bubo, which is accompanied by intense pain in the aff ected lym- phatic region […]. Palpation typically elicits extreme tenderness.140

137 Cantey 1974: 281. 138 Butler 1993: 887. 139 Butler 1983: 73, 75. 140 Butler 1993: 884, 885.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 546 chapter thirteen

Th is complete collection of evidence taken from all leading standard works on bubonic plague demonstrate that relevant statements on the extraordinary painfulness and tenderness of buboes can be readily found in the scholarly literature on plague. Th is material has been published over a period of eighty years, covering the whole period of scientifi c studies of plague, and given by authors who have great personal experience with cases of bubonic plague and great knowledge of the scholarly plague literature available at the time, has great evidential force in the matter under discussion. It is highly signifi cant that they exhibit very similar clinical descriptions of the intense pain associated with buboes: it is great, agonizing, sharp, stabbing, will occur repeatedly with the movement of the associated limb or neck, and is so intense that even comatose patients wince and moan at the touch of the bubo. Th is intense pain is considered to form “one of the most outstanding clinical features of bubonic plague,” which can be taken to mean a defi ning clinical feature of bubonic plague. Only bubonic plague is characterized by the usual clinical manifestation of buboes which are also intensely or excruciatingly painful, and characterized by sharp “stabbing pains.” In my thesis I also refer to more recent primary studies on bubonic plague, citing the scholars’ descriptions of buboes as “painful and exquisitely tender” or characterized by “extreme tenderness.”141 Th is presentation of the subject is therefore complete and updated. Th us, all the prominent authors of the standard works on plague who also have great personal experience with plague patients emphasize that plague buboes are characterized by intense, piercing, stabbing and agonizing pain and that this feature is readily observable and can readily be clinically characterized as repeated sharp pangs of pain. Th is justifi es my translation of the element in the Icelandic annals’ clinical description of the Black Death in Bergen referring to intense pain, “hordum stinga” as “sharp pangs of pain.” It corroborates also my understanding of it as referring to the characteristic sharp, stabbing, agonizing pain of plague buboes and, consequently, that this is a strong indication that the Black Death in Bergen in 1349 was bubonic plague.

141 See, for instance, Reed, Palmer, Williams et al. 1970: 479; Welty, Grabman, Kompare et al. 1985: 641–3; Owens 1990: 155.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 547

If the Icelandic scholars Steff ensen or Karlsson, who reject this understanding of the Lawman’s Annal’s clinical description of the Black Death in Bergen (see Appendix 2) had, according to conventional standards of methodology and scholarly work, consulted any of the standard works on plague, they would readily have ascertained the central facts of this topic. Instead, their will to ignore all the medical standard works on plague constituted the basis of off ensive accusations against my view which is squarely based on the scholarly studies and standard works of plague. Th e authors of the standard works on plague consider the intense pain of plague buboes to be a predominant clinical feature of bubonic plague, so conspicuously characteristic that it in itself provides a strong indication of bubonic plague to observers. It should be easily observable also to persons without scholarly training and should be expected to be mentioned by medieval chroniclers when they choose to mention the Black Death and give some clinical detail, including the informants of the Icelandic annals. In fact, there is a surprising number of contemporary descriptions of the extraordinary painfulness of plague buboes. It must again be underlined that historical accounts must be seen in historical and source-critical perspectives. Th ey are normally given by chroniclers without real scientifi c education and with rather confused ideas about disease, generally prone to accept notions of other-worldly intervention or misleading miasmatic notions and to express their views with great rhetorical exaggeration. As in the case of the Icelandic annals, ecclesiastics would oft en avoid specifi cation of anatomical sites of buboes with sexual connotations like the groin.142 Th ere are quite frequent contradictions, misunderstandings and other indications in the texts of contemporaries, including those who are most oft en cited, which show that they have quite likely not observed persons suff ering from or having died from plague. Certainly, there is no indication that any of them have performed anything like a medical examination of plague victims. Th e usual impression is that they render more or less generalized second-hand or third-hand descriptions or accounts, rumours or hearsay. Th is explains why blood spitting is mentioned probably as oft en as buboes and that buboes also can be indicated only indirectly by reference to the intense or excruciating pain associated

142 See for instance Benedictow 2004: 192–4.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 548 chapter thirteen with them (see below). Th is is the case, for instance, with the Icelandic annals’ descriptions. Th e Franciscan friar Michael of Piazza’s (Michele da Piazza) account of the Black Death’s arrival and spread in Sicily is one of the most frequently cited, but features calling for source-criticism are generally overlooked. Th e central clinical features of the disease as it appeared in Catania, according to Friar Michael are that there developed gland boils on the sexual organs, the thighs, the arms, or on the neck. At fi rst they were of the size of a hazel nut […]. Soon the boils grew to the size of a walnut, then to that of a hen’s egg or a goose’s egg, and they were exceedingly painful, and irritated the body, causing the suff erer to vomit blood […]. Th e sickness lasted three days, and on the fourth, at the latest, the patient succumbed.143 As can be readily seen, Friar Michael or rather his informants emphasize the sharp penetrating pain, the formation of “boils” or “gland boils” (= buboes) in the ordinary places for plague buboes, the ordinary size of plague buboes in various phases of development, the “vomiting” of blood, and the ordinary duration of the course of bubonic plague disease, that is three to fi ve days.144 Th ese pieces of clinical information, together with the additional terrifi ed statements of extremely high mortality, constitute a very strong case for identifying the disease that spread in Messina and Sicily with bubonic plague. Sticker cites Gabriel de Mussis’s account of the clinical features of the Black Death in Piacenza in northern Italy. In accordance with modern standard works on bubonic plague, de Mussis starts by stating that completely sound persons suddenly were hit by piercing pains. Next, they were overwhelmed by great frost that made them shiver with fever and they felt stinging spikes, as if they were pierced by arrowheads. Some were cruelly hit under the clavicle [a phrase that probably means “in the armpits:” my insertion], others where the belly and thigh meet [a euphe- mism for the groin or nearby femoral area]. Here grew a smaller or larger lump and, then, soon fulminant fever with head pains […]. Many died on the fi rst day of illness or on the second, most on the third or fi ft h day. Th e spitting of blood was always deadly […]. Th ose who had hardened glands died unless they did not turn soft […].145

143 Cited aft er Nohl 1961: 7–8. 144 In secondary septicaemic cases, it will take a couple of days before the condition with bloody expectoration develops; a duration of this phase of three days indicates an ordinary duration of the whole course of this type of illness of fi ve days. 145 Sticker 1908. 52–3. My translation from German.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 549

De Mussis’s use of the plural forms “piercing pains, “stinging spikes,” “pierced by arrowheads” is interesting. De Mussis provides much the same clinical information as Michael of Piazza regarding the sudden onset of disease, the piercing pains associated with buboes in the armpits and the groin, fulminant fever, the severe headache, that secondary pneumonic plague accompanied by blood spitting which was always deadly, which conforms with modern observations, and information on the duration of illness that fi ts the modern standard works on bubonic plague perfectly. In his excellent monograph on the Great Plague in London in 1665, W.G. Bell provides a succinct summary of the clinical features mentioned by contemporaries, inter alia: “the rising of plague buboes was attended by such severe pain, and a feeling as of intolerable burning as the time for suppuration approached that the suff erers oft en became raving mad.” Other usual features were that people became comatose,146 which accords with the descriptions given in modern standard works on plague cited above.147 Th is analysis can be completed by citing the impressive proto- scientifi c presentation of the plague epidemic in the Swedish city of Norrköping in 1710–1 given by the local physician M.G. Block who for his time was well educated and certainly had a clear and observant mind. Over three pages, he gives a detailed clinical description of bubonic plague, and towards the end of his book he provides numerous detailed case histories. Th is comprehensive proto-scientifi c clinical information conforms closely to the sections on clinical features given in modern textbooks on bubonic plague, and it is much more detailed than that given by late medieval chroniclers or physicians but also provides information which is similar to that given by late medieval informants. He mentions the headache, shivering and strong fever, and a number of other physical manifestations of the disease, but under- scores that some features are more important than others, among them: Th at it pricks like needles everywhere in the glands and particularly under the shoulders and in the groin. Swellings in those places, swellings that eventually, though quite slowly mature into buboes […].148

146 Bell 1951: 127. 147 Bell 1951: 125. 148 Block 1711: 21. My translation from Swedish.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 550 chapter thirteen

Th e plural form should also be noted. Th us, also in connection with the last wave of epidemics in Scandinavia (only Sweden and Denmark) the same clinical information is met with again, this time as systematically and acutely observed by a dedicated proto-scientifi c physician.

Summary and Conclusion

(1) It has been documented that modern studies of plague emphasize that primary pneumonic plague is normally associated with little pain and is not characterized by pleurisy or pleuritic pain. Th is proves that Steff ensen’s understanding of the term “hordum stinga” as refl ecting pleuritic pain associated with primary pneumonic plague is erroneous and arbitrary. It also proves that his addition of the word “pleuritic” in his translation of this term, “great pleuritic pain,” is unjustifi ed and misleading. (2) It has been demonstrated that, according to the corpus of scholarly studies of plague, bubonic plague is consistently associated with intensely painful buboes, and that movements of an associated limb or the neck or touch will cause sharp pangs of pain. Th is justifi es and vindicates my translation of “hordum stinga” with “great pangs of pain” and the use of plural form. Plural form is used also by authors of standard works on plague. Th is means that it is justifi ed to associate the Annal’s clinical description of plague cases as characterized by sharp pangs of pain with the characteristic sharp pangs of pain of plague buboes and take this as a strong indication that the Black Death in Bergen was bubonic plague. Karlsson’s accusations are unfounded. (3) In addition it has been shown that contemporary chronicle(r)s emphasize the intense or excruciating painfulness of buboes, which are characterized by piercing, sharp, stabbing pains, which justifi es my translation of the intense stabbing painfulness associated with plague cases in Bergen in 1349 in the plural form and its association with buboes. (4) It has been shown that the Icelandic annals’ clinical description of plague cases as associated with “vomiting” of blood conforms to contemporary and modern medical descriptions of bubonic plague cases with secondary pneumonia.149

149 See also above: 7–8; below: 491–3.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access gunnar karlsson’s alternative theory 551

(5) No evidence supports the notion that the epidemics in Iceland 1402–4 or 1494–5 spread in the primary pneumonic modality. On several independent grounds it can be ascertained that these epidemics were not caused by plague in any form. Th ere never was a plague epidemic in Iceland. Th e clear outcome of this discussion constitutes an important element in the process of identifying the epidemic disease that ravaged Bergen and Norway in 1349. Next, the other clinical elements of diagnostic value provided by the Icelandic annals must be considered in the light of modern knowledge of bubonic plague in order to ascertain whether or not they conform to the clinical manifestations of this epidemic disease. Clinical descriptive elements of diagnostic value are provided in the independent contemporary Icelandic fragment of the Annal of Skálholt.150 Th e disease was such that men lived for three days with sharp pangs of pain, then they began to vomit blood and next they expired.151 As can be readily be seen, this clinical description is so closely related to that given in the Lawman’s Annal that it could come from the same informant but deviates on one point of importance: (1) Both annals testify that the sharp (pangs of) pain set in from the beginning of the outbreak of the disease in humans. As copiously documented above, this is characteristic of the pain of plague buboes that starts before or at the beginning of the formation of buboes. Th us, this feature also supports the case that the onset of intense sharp or piercing pain refl ects the onset of bubonic plague disease and heralds the development of one or more buboes. Th e point can be illustrated with a Norwegian case history associated with a plague epidemic in the region of Hardanger in 1600. A young girl who was soon to be married returned one evening from the cowshed; suddenly she let down her buckets and said: “Now the plague stung me,” and the next day she was dead.152

150 Islandske Annaler 1888: XVIII–XX; Benediktsson 1970: 390–1: “it is quite independent, even though some similarities with other annal redactions occur. It is an important primary source for the period it covers” [namely 1328–72]. 151 Islandske Annaler 1888: 224. Cf. above: 536, fn. 116. My translation from Icelandic Norse. 152 Benedictow 2002: 238. My translation from Norwegian.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access 552 chapter thirteen

(2) Th e duration of the course of illness accords fully with modern medical descriptions of the development of bubonic plague in victims who also develop secondary pneumonic plague, which is normal, occurring in about a quarter of all plague cases. Th e bubonic phase lasts three days before the development of secondary pneumonia manifests itself with the onset of bloody expectoration, and then the patient (almost) invariably dies. (3) Jón Steff ensen’s rejection of Shrewsbury’s assertion that “primary pneumonic plague cannot develop independent epidemics to any degree” is not supported by documentation from the scholarly literature on primary pneumonic plague, and must therefore be considered arbitrary and untenable. Shrewsbury’s view stands vindicated. Th e evidence supports unconditionally and without exception a conclusion to the eff ect that the Black Death in Bergen and Norway more generally was bubonic plague.

O. Benedictow - 9789004193918 Downloaded from Brill.com09/29/2021 07:22:21AM via free access