epidemics past and science present: an approach to cholera in nineteenth-century japan

william johnston · wesleyan university

ABSTRACT CHOLERA IN JAPAN: FROM EPIDEMIC WAVES TO ENDEMIC STASIS The history of cholera in nineteenth-century Japan challenges the idea that the disease was primarily episodic. Cholera, at least in the way we understand the Historians and public health specialists have frequently disease today, probably first reached Japan in 1822. One described cholera in terms of global pandemics, with five contemporary Japanese hand-written account quotes Jan occurring in the nineteenth century, one in the early twentieth Cock Blomhoff, the Director of the Dutch trading mission on century, and one having started in 1961 and continuing Deshima in Nagasaki between 1817 and 1824, who claimed today. In Japan, however, cholera remained endemic the disease had spread from Batavia.1 The same manuscript from no later than 1877 until the 1920s, a period during that quotes Blumhoff contains another work that describes which several epidemic outbreaks occurred. Biological and cholera as having spread that year throughout western ecological research on the cholera bacillus from the last two Honshu where it struck old and young, strong and weak and decades explains the reasons why cholera became endemic rapidly caused among many.2 Other contemporary in nineteenth-century Japan as well as the patterns that epidemic cholera took earlier in the century. Furthermore, 1 Korera, part 1 of Bunsei Mizunoeuma Aki Tenkōbyō Korera Mo- cholera’s endemicity in Japan challenges the boundaries of the rubusu Keikin [The Experience of Cholera Morbus in the Fall concept of pandemic as applied to that disease. Epidemic of 1822] (anonymous hand-copied, non-paginated manuscript in the Fujikawa Archive, Kyoto University, not dat- ed). 2 Mizunoeuma Tenkōbyō Setsu [An Explanation of the Epidemic of 1822], part 2 of Bunsei Mizunoeuma Aki Tenkōbyō Korera Moru-

The above image, kindly provided by Waseda University, depicts the main Edo crematory during the cholera epidemic of 1858.

28 HARVARD ASIA QUARTERLY 14.4 (2012) | Epidemics Past and Science Present Epidemics Past and Science Present | HARVARD ASIA QUARTERLY 14.4 (2012) 29 accounts claim that the disease entered Japan from Korea localized or have been a less severe form than the cholera that via Tsushima, first reaching the city of Hagi (in present day appeared during the nineteenth century. The cholera bacillus Yamaguchi prefecture); this followed an active trade route is an evolving organism and the disease-causing form that between Korea and Japan.3 Despite historians’ attempts to spread globally from the early nineteenth century might not establish the correct route of entry, there is no reason that have emerged until then.6 Yet without corroborating genetic cholera could not have entered Japan from more than one material from contemporary cholera bacilli or more detailed route during this or later epidemics. Indeed, multiple entry textual evidence it is impossible to say with certainty whether points would have increased the chances of an infectious the disease reached Japan before 1822. Of interest here is not disease becoming epidemic. the accuracy of this claim but the fact that when cholera did The same manuscript that quotes Blumhoff also arrive in Japan in 1822 it carried a sense of familiarity, at least includes a separate work called Korori Ben Naihen [A Treatise to some observers. on Cholera, Vol. 2] by a physician from Aki (in present-day Hirahira then contradicts his previous assertion when Hiroshima Prefecture) named Hirahira Yūteki. In his account he states, “this disease resembles previous epidemics but is of the epidemic, Hirahira presents contradictory yet intriguing not the same; it resembles kakuran but is not [the same as] assertions about cholera in Japan. The first is a claim of earlier kakuran.”7 The contradiction was probably unimportant cholera epidemics. He writes, “I have heard that some fifty to Hirahira. Many Japanese medical works at that time years ago there was a great epidemic of this disease and those uncritically compiled earlier information on a subject and who contracted it died.”4 Together, the historical and scientific did not attempt to establish what we now consider scientific suggest that this is not completely implausible. There is textual facts even in the sense in which Ludwik Fleck uses the term.8 evidence that cholera was endemic in the Indian subcontinent Kakuran was a disease name for a range of gastro-enteric long before the first pandemic erupted in 1817. European disorders less severe than what we now call cholera. As a seafarers to the Indian subcontinent reported cholera as medical term its use parallels the European use of cholera early as the sixteenth century and during the late eighteenth morbus, which was usually but not universally distinguished century there were reports of the disease among British troops from cholera Asiatica – keeping in mind that nosology at in India and on British ships that had docked there.5 If that the time in both Europe and Japan was a loose science. In were indeed the case, there is some possibility that cholera comparison to later epidemics this one was relatively mild. reached Japan before 1822. British ships could have passed It spread from Kyushu and western Honshu to Osaka and the disease to the Dutch who in turn could have brought it Kyoto and as far east as Shizuoka but did not reach Edo. It to Japan. There are no records of a cholera epidemic in Japan had dissipated by the beginning of 1823.9 during the 1770s but the disease might either have remained It is notable that the 1822 epidemic brought with it the word korori, which combined both a pun and Japanese busu Keikin. transliteration of the European word “cholera.” The more 3 Yamamoto Shun’ichi, Nihon Korera Shi [A History of Cholera in scholarly-sounding term korera also became widely used, Japan] (Tokyo: Tokyo University Press, 1982), 6-7. especially in medical texts, but throughout the nineteenth 4 ūteki, Hirahira Y Korori Ben Naihen (hand-copied, non-paginat- century the word korori remained common as a word for ed manuscript in the Fujikawa Archive, Kyoto University, dated 1842 in afterword). cholera. In his otherwise medically-oriented text, Hirahira 5 Edmund Charles Wendt, in A Treatise on Asiatic Cholera (New noted two bits of doggerel that played with this word:10 York: William Wood and Co., 1885), 3-5, cites a 1503 descrip- Korori to kokete sora ibiki tion of “Asiatic Cholera” by Gaspar Correa, a member of Vasco [Tumbling down and falling over, pretending to snore.] de Gama’s expedition, as well as other descriptions from the seventeenth and eighteenth centuries including an especially 6 Shah M. Faruque and John J. Mekalanos, “Phage-bacterial Inter- devastating epidemic among British troops in 1782 that reached actions in the Evolution of Toxigenic Vibrio cholerae,” Virulence the British fleet. Robert Pollitzer, in his classic history of the 3 (2012): 1-10. disease, documents numerous accounts of cholera outbreaks be- 7 Hirahira Yūteki, Korori Ben Naihen. fore 1800. See R. Pollitzer, “Cholera Studies, 1. History of the 8 Ludwik Fleck, The Genesis and Development of a Scientific Fact Disease,” Bulletin of the World Health Organization 10 (1954): (Chicago: University of Chicago Press, 1979). 421-427. Biologists continue to find this evidence persuasive. 9 Mizunoeuma Tenkōbyō Setsu, part 2 of Bunsei Mizunoeuma Aki See, for example, Rita R. Colwell, “Global Climate and Infec- Tenkōbyō Korera Morubusu Keikin. See also Yamamoto, Nihon tious Disease: The Cholera Paradigm,” Science, New Series, 274 Korera Shi, 5-13; Nakajima Yōichirō, Byōki Nihon Shi [Disease (1996), 2025; and John J. Mekalanos, Eric J. Rubin and Mat- in Japanese history] (Kyoto: Yūsankaku, 1982), 46-8; and Ann thew K. Waldor, “Cholera: Molecular Basis for Emergence and Bowman Jannetta, Epidemics and Mortality in Early Modern Ja- Pathogenesis,” FEMS Immunology and Medical Microbiology 18 pan (Princeton: Princeton University Press, 1987), 157-60. (1997): 241. 10 Hirahira Yūteki, Korori Ben Naihen.

William Johnston received his BA from Elmira College in Elmira, New York, and his AM in Regional Studies – East Asia and PhD in History and East Asian Languages from Harvard University. He is Professor of History, East Asian Studies, and Sci- ence in Society at Wesleyan University, where he is also a Faculty Fellow at the College of the Environment for the 2012-13 academic year. At present he is working on a book about the history of cholera in Japan. This paper received support from the Japan Foundation.

28 HARVARD ASIA QUARTERLY 14.4 (2012) | Epidemics Past and Science Present Epidemics Past and Science Present | HARVARD ASIA QUARTERLY 14.4 (2012) 29 avoided the second cholera pandemic that scholars now date as having occurred between 1829 and 1851. Japan’s second epidemic, which continued from 1858 into 1859, took place during what is considered the third global pandemic, which dates between 1852 and 1859. This epidemic was severe: it started with an infected American sailor from the USS Mississippi, which had docked in Nagasaki in the sixth month (by the Japanese calendar) of 1858. The disease quickly swept through western Japan and reached Edo the following month. From there it spread throughout most of the country.15 During the seventh and eighth months of 1858, death registers in Edo recorded over 268,000 from all causes, an astounding of approximately 26,000 per 100,000 of population. Because no other epidemic disease was present at the time, the vast majority of this excess mortality would have Figure 1: Kanagaki Robun, Ansei Umanoaki Korori Ryūkō Ki 16 Source: Waseda University. been from cholera. It is of note that according to Yamamoto Shun’ichi, the physician and professor of medicine at Tokyo Yomego no hirune mo korori to se University who wrote a comprehensive history of cholera in [Like a new bride taking a nap, tumbling right over.] Japan, the disease virtually disappeared during the winter of Both point to the rapidity with which cholera claimed 1858-1859, only to reappear in the summer of 1859, then its victims. “Pretending to snore” in the first one possibly again disappearing until the summer of 1860, when a small refers to the death rattle, although because of the dehydration number of cases were reported.17 that cholera causes it did not occur in all cases.11 The second Contemporary textual works also attest to the extreme compares a person with cholera to a young bride, many of mortality this epidemic caused. One, Ansei Umanoaki Korori whom in Japan at that time had such demands placed upon Ryūkō Ki [A Record of the Cholera Epidemic of 1858] by them by their mothers-in-law that they were constantly the well-known satirist and journalist Kanagaki Robun tired and could instantly fall asleep. Both suggest another (1829-1894), provides a multi-dimensional view of the contemporary euphemism for cholera, korobine, or “fall- epidemic. It recounts how it started in Nagasaki, explains over-asleep.” A related expression, mikka korori or “three- some contemporary European ideas about the disease, and day-tumble-down [and die],” which points to the rapidity details how it spread across the country. In a black-and-white with which cholera caused death, also originated at this time line illustration, it shows a parade of pallbearers carrying and remained common into the early twentieth century.12 caskets to the Edo crematory, which appears in the center Physicians wrote some texts that advocated the use of of the upper left half of the image and was located next to traditional Chinese-based medicine to treat the disease. the Kotsugappara execution grounds (see Figure 1). The Others translated or adapted European texts about cholera. accompanying text notes that from the seventh month bodies Many included both European therapeutic measures along had started to accumulate at the crematorium. This onslaught with traditional herbal remedies.13 Although there is no let up briefly only to increase so that between the fourth and record of cholera between its first outbreak in 1822 and its sixth days of the eighth month the number of bodies exceeded next epidemic, which started in 1858, the disease had taken capacity by 20 or 30; after the tenth of the month over 600 root in Japanese medical discourse and popular culture alike bodies remained to be cremated. By the second or third day as a deadly disease and a danger imported from abroad. of the ninth month so many bodies had accumulated that Japan’s first cholera epidemic occurred during what no matter how much money a family offered to pay to the scholars later in the nineteenth century considered the first crematorium it was impossible to receive a timely .18 world pandemic of the disease.14 The Japanese archipelago The page that follows the image of the pallbearers shows one of the most famous images from the 1858 epidemic, an 11 F.A. Burrall, for example, in his description of the disease, anonymous but well executed full-color woodblock print claimed that the death rattle did not occur in cases of cholera. entitled Yakiba Konzatsu no Zu [Crematory, Picture Showing See Frederick Augustus Burrall, Asiatic Cholera (New York: Wil- Confusion], a close-up of the crematorium that appeared liam Wood & Co., 1866), 79. Eyewitness accounts of the Ham- in the distance on the previous page. (This is the image at burg epidemic in 1892, however, clearly describe it. See Richard J. Evans, Death in Hamburg: Society and Politics in the Cholera the beginning of this essay.) It shows exhausted pallbearers Years, 1830-1910 (Oxford: Oxford University Press, 1987), 332. carrying no fewer than five caskets in the foreground; they 12 Nakajima, Byōki Nihon shi, 46. are surrounded by officials trying to maintain some degree 13 Hirahira, in Korori Ben Naihen, includes both European and tra- ditional remedies; Mizunoeuma Tenkōbyō Setsu contains transla- 15 Yamamoto, Nihon Korera Shi, 16-24. tions from unidentified European texts on cholera morbus. 16 Ibid., 21-2. 14 Pollitzer describes the discrepancies over dating cholera pandem- 17 Ibid., 14. ics among scholars during the late nineteenth and early twenti- 18 Kanagaki Robun, Ansei Umanoaki Korori Ryūkō Ki (Edo: eth centuries. Pollitzer, “Cholera Studies,” 427. Tenjudō, 1858).

30 HARVARD ASIA QUARTERLY 14.4 (2012) | Epidemics Past and Science Present Epidemics Past and Science Present | HARVARD ASIA QUARTERLY 14.4 (2012) 31 of order and to keep track of the incoming caskets by giving accounts from the New York Times concur that a cholera them numbers and recording those in a record book. Two epidemic followed one of measles, and that British sailors in crematory laborers face the incoming caskets with expressions Yokohama also had died of cholera.22 Yamamoto Shun’ichi, of exasperation and fatigue. Thirteen caskets, numbered and however, doubts that a cholera epidemic even occurred in labeled, are waiting to be burned; one, with “No. 170” written this year but does not explain his skepticism.23 It seems to be on it sits atop one inscribed “No 70.” No fewer than seventeen based on a view of cholera that sees the disease as episodic and urns containing ashes can be seen in the middleground; the whose causal organism, the cholera bacillus or Vibrio cholerae, crematory fire roars in the background. Lit lanterns and dark has no hosts other than humans. According to this view, lighting imply that the scene is unfolding in the middle of when the causal organism disappears from a population it the night. needs to be re-introduced by humans who carry the bacillus. While Kanagaki’s work is relatively well known, another Yamamoto implies as much when he writes: “In 1877, cholera in the same genre is Chimata no Yume [Dreams along the appeared again for the first time since the great epidemic of Street] by the author Hata Ginkei. Hata recounts numerous 1858, nineteen years earlier.”24 Ann Bowman Jannetta also anecdotes that detail various popular understandings of subscribes to this view when she asserts that humans are the the epidemic, understangings that reflect current religious sole host of the cholera bacillus and that the 1862 epidemic and scientific thinking that people used to interpret the of cholera had to have been caused by people who brought it phenomena they observed. He begins the text with the to Japan from mainland Asia.25 Current scientific evidence, lament, “Friends I had saké with yesterday are on their way however, which I discuss in detail in the following section, to the underworld today,” and dedicates it to those who had contradicts this view. died in this epidemics. This work includes a large number The episodic view of cholera in Japan starts to break of anecdotes recounting, for example, how villages had been down after 1877, when the Japanese government began saved from cholera due to local deities, how a talisman- recording annual cases and deaths attributable to the disease. carrying ferryman had prevented cholera demons from From that year until 1923, no year passed when there were crossing a river, how a comet had brought on the epidemic, no cases of cholera somewhere in Japan (see Table 1). Major and how a farmer whose good karma accrued by his refusal to epidemics occurred in 1876 and 1879, when there were over kill locusts when they had plagued his fields kept him from 100,000 deaths each year from cholera, and in 1882, 1890, dying of cholera once he had contracted the disease.19 These and 1895, when it caused more than 10,000 deaths annually. kinds of popular works have done much to focus attention on Yet in every other year between 1877 and 1923 there were cholera as an episodic disease, one that people associate with cases or deaths reported, implying that cholera remained pandemics and epidemics. Because these works were written endemic in Japan for nearly a half century. Of course this in when epidemics occurred they reinforced the impression runs counter to the view that cholera is an episodically that cholera is present only in those years. epidemic disease that storms through societies, bringing Where in the aftermath of the previous epidemic with it panic and chaos, and not an endemic disease that can physicians had created a handful of works on cholera, the remain in a population for years, nearly without notice, and 1858 epidemic produced dozens of texts that reflected only occasionally break out in widespread epidemics. Yet the contemporary medical thought. Some physicians adamantly Japanese government’s official statistics leave no doubt that stood by Chinese medical interpretations of cholera, usually cholera had become endemic by 1877, if not before. calling it by the archaic names of kakuran or shabyō [diarrheal While the data in this table is problematic, with more disease] instead of korera, while others integrated various deaths reported for some years than the number of reported strands of European thought in their writings about it.20 cases, it leaves no doubt about cholera’s endemicity in Japan Subsequently, interest in the disease remained high through the early 20th century. Regional data, moreover, although it is not clear if there were outbreaks of cholera shows that it tended to occur in coastal more than inland between this and the next epidemic, which occurred in regions during years when there were small numbers of 1877. Indeed, between the 1858-1859 epidemic and 1877, cases. Hiroshima prefecture, which was both coastal and on after which annual statistics for cholera becomes available, a main travel route between Kyushu and cities on Honshu the historical record of cholera in Japan remains somewhat including Osaka, Kyoto, and Tokyo, consistently reported a murky. Some scholars including Fujikawa Yū, the author number of fatalities every year between 1881, when figures of an encyclopedic history of Japanese medicine, assert that the next outbreak of cholera occurred on the heels hon Shi, 49-50, 52; Jannetta, Epidemics and Mortality in Early 21 Modern Japan, 168. of a measles epidemic in 1862 and 1863. Contemporary 22 “Interesting from Japan ... The Empire Desolated by Measles, &c.,” New York Times, dated Yokohama, September 22, 1862, 19 Hata Ginkei, Chimata no Yume (Edo: Gotokudō, 1858). printed December 14, 1862, 2; “Affairs in Japan: Ravages of the 20 A typical example of the former is Okuda Mitamon, Shabyō Shin- Cholera,” dated Kanagawa, October 23, 1862, printed January ron [A New Theory of Diarrheal Disease] (Edo: Jōyō, 1858); 12, 1863, 2. an example of the latter is Shingū Ryōmin, Korerabyōron [A 23 Yamamoto, Nihon Korera Shi, 14. Theory of Cholera] (1858). 24 Ibid., 27. 21 Fujikawa Yū, Nihon Igaku Shi [A History of Medicine in Japan] 25 Jannetta, Epidemics and Mortality in Early Modern Japan, 156, (Tokyo: Keisei Sha, 1972), 612-3; see also Nakajima, Byōki Ni- 168.

30 HARVARD ASIA QUARTERLY 14.4 (2012) | Epidemics Past and Science Present Epidemics Past and Science Present | HARVARD ASIA QUARTERLY 14.4 (2012) 31 spread appeared either in the urban areas of Nagasaki, Osaka, and Tokyo, or on the Inland Sea, with Hiroshima Prefecture being one. It is not surprising that these urban areas were “hot spots,” in that many foreigners visited them, sanitation was something of an issue with nightsoil being a prime commodity in each, and many Japanese visited all of those cities with great frequency for both business and pleasure. All were also located in maritime environments. In addition, the Inland Sea areas considered “hot spots” were well suited to the sustenance of endemic cholera and as a consequence in epidemic years would have acted as dissemination points for the disease, reinforcing the spread that was occurring possibly for unrelated reasons. What we see is that in both epidemic and non-epidemic years, cholera tended to be a problem more in coastal areas than inland. Even this limited data delineates a clear pattern that tends to counter widely held ideas about cholera. If cholera is indeed more of an endemic disease than one caused by pandemic waves that sweep through populations it forces us to change not only our way of conceiving its impact on populations in history but challenges the value of thinking about it primarily in terms of pandemics. This brief look at Japan’s earliest epidemics of cholera and its becoming endemic during the nineteenth century reveals several patterns that in turn raise questions. What are Table 1: Cholera Cases and Deaths for all Japan the characteristics of the cholera bacillus and how does it cause (Mainland), 1877-1934 Source: “Patients and Deaths of Infectious Diseases and Food Poisoning disease and the symptoms described in the historical sources? (1876-1999),” accessed September 12, 2012, www.stat.go.jp/english/data/ Are there various kinds of the bacillus that in turn make some chouki/24.htm. epidemics more severe than others? Why did cholera spread out of the Indian subcontinent for the first time only in the become available for this prefecture, and 1899. Tokushima early nineteenth century? What is the incubation period prefecture, which was also coastal but not on a major travel for cholera – the time between infection and symptoms? route, reported fatalities from cholera every year between If humans are not the sole carriers of the cholera bacillus 1883 and 1900, with the exception of 1889. Between 1879 how else might it be spread? Are there people who carry and 1899, however, inland Gifu prefecture reported fatalities the bacillus in their bodies without exhibiting symptoms? almost exclusively in epidemic years, with only the occasional What are the environmental conditions under which cholera death from cholera occurring in non-epidemic years (see becomes endemic? What is the role of immunity in limiting Table 2). This geographic pattern of cholera deaths strongly its spread? The scientific literature provides answers to most suggests that endemicity depended on a maritime or aquatic of these questions and in turn illuminates the history of environment, something that the science discussed in the cholera, the disease that did more than any other to instigate next section strongly supports. the establishment of public health institutions in nineteenth- It is well known that cholera is a water-borne disease, century Japan. but this distribution of both cases and deaths from the disease CHOLERA’S PATHOGENICITY, TRANSMISSION, AND seem to suggest that there is more to the story. Two papers HISTORY by Chun-Lin Guo and Hiromichi Fukui address some of these complexities. They have written a GIS-based study comparing the spread of cholera in Fukushima prefecture that Historians are by no means the only scholars who shows the largest number of cases in the 1882 epidemic as have continued to assert that cholera is an epidemic disease having occurred in places that were closest either to rivers or whose pathogen travels only between human hosts. The to estuaries.26 In a second piece, they analyzed national data award-winning science writer David Quammen, in his best- for the epidemic that occurred in 1890, and found that the selling book Spillover: Animal Infections and the Next Human “hot spots” that were responsible for much of the epidemic’s Pandemic, describes cholera in just those terms: By a strict definition, zoonotic pathogens (accounting for about 60 percent of our infectious diseases, as I’ve mentioned) are 26 Chun-Lin Kuo and Hiromichi Fukui, “Geographical Structures those that presently and repeatedly pass between humans and and the Cholera Epidemic in Modern Japan: Fukushima Pre- fecture in 1882 and 1895,” International Journal of Health Geo- other animals, whereas the other group of infections (40 percent, graphics 6.25 (2007), doi:10.1186/1476-072X-6-25. including smallpox, cholera, measles, and polio) are caused by

32 HARVARD ASIA QUARTERLY 14.4 (2012) | Epidemics Past and Science Present Epidemics Past and Science Present | HARVARD ASIA QUARTERLY 14.4 (2012) 33 pathogens descended from forms that must have made the leap to human ancestors sometime in the past.27 Later in the same work he repeats this claim: “Thousands of people, young and old, die of cholera and other diarrheal diseases and pneumonia and tuberculosis and measles. Note that none of those afflictions is newly emergent, mysterious, or zoonotic.”28 While the inaccuracy of this claim does not detract from Quammen’s central argument in this book regarding the emergence of new human diseases from animal afflictions, it does show how deeply rooted this view of cholera remains. There is good reason for the persistence of this perspective. In part, it is probably because the cholera bacillus is a highly complex organism and its relationship with its environment was not well understood until relatively recently and even now many questions remain unanswered. As a recent study noted: “Centuries of research on Vibrio cholerae has revealed it to be a constantly evolving bacterium with numerous reservoirs, routes of transmission, and variations in the manifestation of subsequent epidemics.”29 Historians and others who are not science professionals have much to learn from the current scientific literature about the disease and its causal organism. To begin with, not all Vibrio cholerae cause disease. To the contrary, there are Table 2: Cases and deaths from Cholera over 200 serogroups of Vibrio cholerae few of which cause (Hiroshima, Gifu, and Tokushima prefectures) any disease symptoms whatsoever. Only two serogroups, the Source: Compiled from prefectural statistical yearbooks (fuken tōkei sho). O1 and O139, have been associated with epidemic cholera but others have been implicated in causing isolated cases of been understood only during the last two decades. Advances severe disease in places as disparate as the United States and in the deciphering of cholera’s genetic code led researchers Japan.30 The O139 serogroup became widespread only from to focus on what they called its pathogenicity island, the 1992; all previous pandemics have been attributed to the O1 sequence of genes within the overall cholera genome that serogroup. In addition, the O1 serogroup is subdivided into facilitated the production of the cholera toxin (also called at least two biotypes, the classical and the El Tor. The seventh CTX) and the pili (called the toxin-coregulated pilus or TCP) pandemic, which started in Indonesia in 1961, is of the El used to attach the bacillus to intestinal cell walls. This was a Tor biotype, and researchers are confident that the classical major step, but an even more important development came biotype both caused all previous pandemics of the disease and with the discovery in 1996 that the genes for the cholera toxin is now extinct.31 were inserted by a bacteriophage, a virus that transferred its The bacillus causes disease by attaching itself to the cell genetic content into the genes of the Vibrio to help create the walls of the intestines using pili, small filamentous structures, pathogenicity island.32 In other words, non-toxic forms of the and then producing the cholera toxin which stimulates the cholera bacillus had become pathogenic due to the transfer of cells to secrete large amounts of fluid and drain the body of the genetic code for the cholera toxin from virus to bacillus. electrolytes. The toxin itself was discovered only in 1959 and The group that had made that discovery also hypothesized the mechanisms by which the bacillus is enabled to attach that TCP also had emerged on the toxic forms of the bacillus itself to the intestinal walls and produce the cholera toxin have due to horizontal gene transfer, and within the decade they had been proven correct.33 It was further discovered that the 27 David Quammen, Spillover: Animal Infections and the Next Hu- genetic code for TCP could either become integrated into the man Pandemic (New York: Norton, 2012), 264. 28 genetic code of the bacillus itself or remain free-floating as a Quammen, Spillover, 381. 34 29 Ricardo Izurieta et al., “Re-emergence of Cholera in the Ameri- plasmid inside the bacillus. cas: Risks, Susceptibility, and Ecology,” Journal of Global Infec- The disease-causing cholera bacillus is a true chimera, tious Diseases 3 (2012): 162. made up of genetic material from the Vibrio bacillus and two 30 Rita R. Colwell, “Infectious Disease and Environment: Chol- phages. The order of infection is important: TCP phage must era as a Paradigm for Waterborne Disease,” International Micro- biology 7 (2004): 287; Melissa Tobin-D’Angelo et al., “Severe 32 Mekalanos et al., “Cholera,” 245-6. Diarrhea Caused by Cholera Toxin–Producing Vibrio cholerae 33 Ibid., 247; David K.R. Karaolis et al., “A Bacteriophage Encod- Serogroup O75 Infections Acquired in the Southeastern United ing a Pathogenicity Island, a Type-IV Pilus and a Phage Receptor States,” Clinical Infectious Diseases 47 (2008): 1035, 1039. in Cholera Bacteria,” Nature 399 (1999): 375-9. 31 Faruque and Mekalanos, “Phage-bacterial Interactions in the 34 Faruque and Mekalanos, “Phage-bacterial Interactions in the Evolution of Toxigenic Vibrio cholerae,” 1. Evolution of Toxigenic Vibrio cholerae,” 1.

32 HARVARD ASIA QUARTERLY 14.4 (2012) | Epidemics Past and Science Present Epidemics Past and Science Present | HARVARD ASIA QUARTERLY 14.4 (2012) 33 come first and establish pili on the surface of the bacillus to change in the early 1980s when researchers discovered because the CTX phage can infect the bacillus only by both that the cholera bacillus can transform itself into a entering it through the pili. They allow the bacillus to attach viable but non-culturable state and that it can attach itself to itself to the cell walls of the intestine, where the cholera toxin zooplankton and phytoplankton.39 The first discovery made causes the symptoms of disease. Recent research has shown it clear that cholera bacilli existed in marine environments that there are both multiple forms of the CTX phage and the but could not be discovered using earlier methods to test for TCP phage as well a numerous other phages that infect Vibrio their presence. It was also shown that ingestion of this viable cholerae.35 The study of cholera genetics and the phenomenon but non-culturable form of Vibrio cholerae could cause disease of horizontal gene transfer has gone so far as to challenge the among humans. As a consequence it became certain that the way scientists until now have categorized the bacillus using bacillus can exist in marine environments for long periods serogroups. Because horizontal gene transfer creates new and still cause the outbreak of epidemics.40 Later research has forms of the bacillus that are not the result of phylogenetic shown that while attached to plankton the cholera bacillus mutations between generations it has become necessary to can travel long distances on ocean currents and in the process devise new taxonomies, a project that is still underway.36 infect shellfish and fish that feed on the plankton. Arguably, these recent discoveries raise more questions The movement of cholera-infected plankton in ocean with regard to the history of cholera than they shed light on currents was associated with the outbreak of epidemics in the past. Although the previous descriptions of cholera from Peru and Mexico in 1991.41 Similarly, it is easy to imagine the eighteenth century and earlier point to the symptoms that plankton carrying the cholera bacillus could easily have of the disease including nausea, severe diarrhea, , contributed to the spread of the disease in Japan during the delirium, and rapid onset of death in many cases, it remains epidemic of 1858-1859 and to the endemicity that cholera unclear why the disease never spread beyond the Indian achieved after 1877. Ocean currents in the Pacific flow into subcontinent before the nineteenth century, or if it did why it Japan’s inland sea from the east and from the south. Currents never caused widespread disease. Yet the fact that the bacillus that flow from the east past Osaka move toward Okayama, is constantly evolving, incorporating genetic material from between Honshu and Shikoku. Currents that flow from the both viruses and surrounding microorganisms, raises the south between Kyushu and Shikoku then split, with those possibility that it changed sufficiently at some point in the on the west then flowing toward Shimonoseki and those on early nineteenth century to become a new and more deadly the east continuing in that direction between Hiroshima form that rapidly increasing numbers of ships and people and Tokushima. The opposing currents meet at the line moving overland transported first across Asia from 1817 and where Hiroshima meets Okayama prefecture.42 Assuming into Europe and the New World by the 1830s. Indeed, recent that the currents were similar during the Meiji Period, they research points to how this might have been the case: there probably contributed to the relatively large number of cases are numerous phages that result in horizontal gene transfer of cholera in Hiroshima prefecture at that time by circulating to Vibrio cholerae and thereby influence its evolutionary water-carrying plankton within the Inland Sea while not development.37 It is reasonable that evolutionary changes carrying it out to the Pacific to the west. Indeed, a geographic in Vibrio cholerae occurred in the late eighteenth or early information systems (GIS) analysis of the 1890 epidemic nineteenth century that led to the spread of epidemic cholera showed that the Inland Sea coastline of Honshu was a hotspot beyond the Indian subcontinent. Without material evidence for cholera in that year. Fish in these areas would also have this remains conjecture but not unrealistic. In other words, become carriers of the cholera bacillus, something that might Japan’s nineteenth-century epidemics could easily have been have been observed in Japan as early as the 1858 epidemic, the result of biological as well as social and technological when prevention methods recommended the avoidance change. The ecology of the cholera bacillus also has been poorly understood until the past two decades, during New York Academy of Sciences 740 (1994): 44. 39 which scientists have created a much clearer picture of how Rita R. Colwell et al., “Viable but Non-Culturable Vibrio chol- erae and Related Pathogens in the Environment: Implications it survives in the absence of epidemics and how climatic for Release of Genetically Engineered Microorganisms,” Bio/ and other environmental conditions help determine the technology 3 (1985): 817; Anwarul Huq et al., “Ecological Re- timing and location of epidemics. From the time Robert lationships Between Vibrio cholerae and Planktonic Crustacean Koch confirmed in 1884 that Vibrio cholerae caused cholera Copepods,” Applied and Environmental Microbiology 45 (1983): 275-83. scientists continued to believe that the bacillus could not 40 38 Rita R. Colwell, in “Global Climate and Infectious Disease: The survive for long outside the human intestine. This started Cholera Paradigm,” Science, 274 (1996), 2025-31, provides a clear overview of changes in the science concerning the ecology 35 Ibid., 2-7. of cholera up to that time. 36 Yong-Joon Cho, “Genomic Evolution of Vibrio cholerae,” Cur- 41 Izurieta et al., “Re-emergence of Cholera in the Americas,” 163. rent Opinion in Microbiology 13 (2010): 646-51. 42 Maps showing currents in the Inland Sea can be found at “Seto 37 Faruque and Mekalanos, “Phage-bacterial Interactions in the Naikai no chōryū” [Currents of the Inland Sea], Kankyōshō, Evolution of Toxigenic Vibrio cholerae,” 1-10. accessed November 19, 2012, http://www.env.go.jp/water/ 38 Rita R. Colwell and Anwarul Huq, “Environmental Reservoir of heisa/heisa_net/setouchiNet/seto/kankyojoho/sizenkankyo/cho- Vibrio cholerae: The Causative Agent of Cholera,” Annals of the ryu/choryu.htm.

34 HARVARD ASIA QUARTERLY 14.4 (2012) | Epidemics Past and Science Present Epidemics Past and Science Present | HARVARD ASIA QUARTERLY 14.4 (2012) 35 of fish.43 The bacillus also has been found on the feet and substantially to its dissemination.47 feathers of waterfowl that have been known to migrate from coastal to inland areas, providing yet another possible route of CONCLUSION infection.44 Although there is no way to demonstrate whether this might have happened in Japan during the nineteenth As a historian, it is impossible to say simply that century, in an area where the disease is already endemic, as it scientific knowledge, as it stands today, is beyond question. was for the last half of the century in Japan, birds easily could The historical literature on scientific change that emerged have provided yet another route of transmission making the over the past century, from Ludwik Fleck’s Genesis and control of the disease all the more difficult. Development of a Scientific Fact to Thomas Kuhn’s The Structure Recent discoveries in the ecology of cholera also help of Scientific Revolutions to Bruno Latour’s The Pasteurization explain the episodic pattern of the disease when it is both of France, to name just a few examples, have shown that we epidemic and endemic. In temperate climates the disease cannot assume any scientific fact is immutable. The scientific starts to become more widespread in summer months, which evidence presented in the contemporary journal literature also is when both phytoplankton and zooplankton blooms does illuminate aspects of the history of cholera in Japan that occur. Because plankton carry the bacillus its reservoirs in otherwise would remain obscure at worst or alluded to at marine environments can contribute most intensely to its best. But just as the concepts of cholera that the physicians spread at this time. Later in the season an increase in the and other observers of the disease used in nineteenth-century density of viruses that attack Vibrio cholerae in turn inhibits Japan are now dated, some of the concepts that researchers its spread, helping to slow the spread of the disease and giving today accept as providing the best possible explanations to it a self-limiting nature.45 observed phenomena quite likely will fall from use in years One possible explanation for endemic cholera is the to come. Already, the idea that serogroups are the best way presence of long-term carriers—persons who are infected for to understand the phylogenetic structure of the cholera many years with the disease but who remain asymptomatic. bacillus is facing serious challenge. Nevertheless the scientific Over a century of clinical observations indicate this is not literature is useful for just that reason: it provides the best the case.46 In addition, the incubation period for cholera possible explanations to date of observed phenomena. usually ranges from approximately eighteen hours to five Based on that literature, we can conclude that the days, after which a person usually either abruptly develops the episodic pattern that cholera epidemics displayed in Japan was symptoms of extreme diarrhea and nausea or remains infected the result of natural phenomena and common in other areas but symptomless for up to three to five days. Only between where the disease is linked to marine ecologies. The textual 1 in 3 to 1 in 100 persons who ingest sufficient bacilli to literature from the early nineteenth century demonstrates become infected ever develop symptoms, depending on age, beyond doubt that cholera epidemics swept through Japan in immunity, nutrition, and blood type, and as a consequence 1822-1823 and in 1858-1859. Evidence of later epidemics is when the disease is widespread it can be expected that both textual and numerical, with strong numerical evidence these infected but asymptomatic individuals contribute for the entire country after 1877. The most important conclusion to be drawn from the historical and scientific evidence presented here is that 43 See, for example, Oku Kōkō, Shabyō ruiki (hand-copied manu- cholera became endemic in Japan by the 1870s, and as a script, Fujikawa Archive, Kyoto University Library) 1858. Pro- consequence challenges the utility of pandemicity as a tool hibitions on fish continued into the Meiji Period. See the broad- sheet entitled Korera-byō Yobō no Kokoroe [Teaching on Prevent- for understanding the spread of cholera worldwide. Japan’s ing Cholera] (1877), which explicitly forbids eating sardines, natural environment, and especially the extent of marine flying fish, shrimp, clams, and octopus, among others (Univer- ecosystems in the Japanese archipelago, also contributed sity of California, San Francisco, Special Collections); and Oda to the epidemiological patterns that cholera reflected there Kōsaku, Korera Yobō Kokoroegusa [Reminders on the Prevention during the nineteenth century, especially with regard to the of Cholera] (Tokyo: Honda Masakuni, 1880), n.p. Biologists disease emerging in its endemic form. Because endemic have found Vibrio cholerae in the guts of some fish. See Yigal Senderovich, Ido Izhaki, and Malka Halpern, “Fish as Reservoirs cholera in Japan encompassed the fourth (1863-1879), fifth and Vectors of Vibrio cholerae,” PLoS ONE 5.1 (2010): e8607, (1881-1896), and sixth (1899-1923) pandemics of cholera, doi:10.1371/journal.pone.0008607. the concept of pandemic itself is open to challenge. If it 44 Malka Halpern, Yigal Senderovich, and Ido Izhaki, “Waterfowl signifies a period during which a disease spreads throughout – The Missing Link in Epidemic and Pandemic Cholera Dis- the world, it should be safe to assume that cholera is absent semination?” PLoS Pathogens 4 (2008): e1000173, doi:10.1371/ journal.ppat.1000173. in “off” years. Yet that was not the case. Although there is no 45 See Shah M. Faruque et al., “Seasonal Epidemics of Cholera In- doubt that the idea of cholera pandemics will not die soon versely Correlate with the Prevalence of Environmental Chol- it is equally clear that its boundaries are indefinite at best. era Phages,” Proceedings of the National Academy of Science 102 Then again, that is also the case with many of the most useful (2005): 1702-7; Shah M. Faruque et al., “Self-limiting Nature of concepts in history and science alike. Seasonal Cholera Epidemics: Role of Host-mediated Amplifica- tion of Phage,” Proceedings of the National Academy of Science 102 (2005): 6119-24. 46 David A Sack et al., “Cholera,” Lancet 363 (2004): 224. 47 Ibid.

34 HARVARD ASIA QUARTERLY 14.4 (2012) | Epidemics Past and Science Present Epidemics Past and Science Present | HARVARD ASIA QUARTERLY 14.4 (2012) 35