Introduction
The literature of meteorological science has upheld to date an argument that historians have rarely questioned: telegraphy was the technological prerequisite for the emergence of storm warning systems and, later, national meteorological services. Historians have tacitly or explicitly endorsed a linear narrative so eruditely condensed by the influential meteorologist Cleveland Abbe in 1871 in his history of weather telegraphy: progress in predictive meteorology came only when telegraph lines extended widely over the earth’s surface.Footnote 1 Not only did ‘the telegraph’ make ‘possible the construction of same-day weather maps’, as Frederik Nebeker claims,Footnote 2 but the very ‘evolution of [telegraphic] codes reflected the evolution of the ideas’ of synoptic meteorologists, as Aleksandr Khrgian argues.Footnote 3 The so-called ‘weather telegraphy’ epitomizes this organic relationship. Through telegraphy came the almost instantaneous transmission of information over long distances; through this transmission, which allowed meteorologists to compile records of observations from dispersed stations, came real-time data analysis; and through this capability, the establishment of the world’s first storm warning systems.Footnote 4 If this linear sequence is accepted, if it is inevitable that the telegraphic development was followed by its application to meteorology and later by institutionalization, then science has only an incidental and subsidiary role to play in technological expansion. The technological infrastructure then embodies the continent within which the content of scientific knowledge expands.
One might think, at least at first glance, that the meteorological science cultivated by the Jesuits in Asia has all the hallmarks of this linear sequence: a chessboard of imperial interests at play, expanding telegraph and cable networks, cyclone warning systems, and large-scale meteorological services. The Society of Jesus founded the meteorological observatories in Manila (1865), Calcutta (1866), and Zikawei in Shanghai (1872), the first of their kind in East Asia. In parallel with the laying of submarine cables and telegraph lines, Jesuit scientists established what became the world’s first tropical cyclone warning systems. Later, the Society extended its network of observation stations to the cardinal points of the typhoons’ paths, from the Philippines to the treaty ports of China: the colonial governments and merchant communities thus created the China Coastal Meteorological Service (1882) and the Philippine Meteorological Service (1884), entrusting their direction, organization, and control to the Jesuits.Footnote 5 Such was this confidence in the Jesuits that the Service in the Philippines (then a Spanish colony) predated Spain’s own national meteorological service by three years; and from the late Qing to the Republican era, the Zikawei Observatory embodied the ‘de facto central observatory in China’.Footnote 6 In the decades that followed, the Jesuits became world authorities on cyclone studies, developing their own forecasting methods and leading the most productive centres in East Asia.Footnote 7
Closer examination, however, reveals certain flaws in this linear causal sequence. First, the establishment of a network of observatories—linked to a Jesuit knowledge network—just before the advent of the telegraph, the protection of shipping and maritime trade, and the sharing of functional synergies and nodes in the meteorological and telegraphic networks, all raise questions that show how the relationships between scientific knowledge and communication infrastructures are more complex and bidirectional than the linear narrative has assumed. Second, telegraphic and meteorological infrastructures conceal synergistic processes. The control of meteorological knowledge became critical, even for telegraph and cable companies. And third, and undoubtedly the most significant for its implications for the global history of technology and science, expanding technological networks did not necessarily imply expanding knowledge networks.
It has taken a generation for historians of technology, science, and empire to realize the qualitative leap that telegraphy implied for global communication. It is now well known that the telegraph was central to imperial and commercial expansion and epitomized the fascination with an ‘optimistic faith in progress’.Footnote 8 Global historians owe much to Roland Wenzlhuemer, Amelia Bonea, and Iwan R. Morus for analysing the telegraph’s ability to transform the very nature of telecommunications, and for questioning its supposed capacity to ‘annihilate’ time and space.Footnote 9 The telegraph involved the dematerialization of long-distance information flows. It enabled data to be encoded as electrical impulses and transmitted along a conductor using very little time and energy. As a result, the speed of communication increased in absolute terms. It also enabled the transmission of weather telegrams. Long messages based on flexible telegraphic code systems could be transmitted at low energy cost.Footnote 10 Telegraphy implied the detachment of information transmission from the transport of people and goods. As Wenzlhuemer correctly states, all this ‘impacted pronouncedly on the nature and intensity of nineteenth-century globalization’.Footnote 11 However, given this qualitative change, one might think that the spread of telegraphy by necessity meant the globalization of scientific knowledge. Intriguingly, this was not the case for meteorology in Asia. The supposed causal role of telegraphy, familiar to nineteenth-century meteorologists, reveals the contradiction, surprising to most weather historians, that expanding telegraph networks did not inexorably entail expanding knowledge networks.Footnote 12
The idea that technology develops autonomously and inevitably shapes society in a linear and preordained way came to the fore in science and technology studies (STS) in the 1960s and 1970s, following, among others, Marshall McLuhan’s communicative theory, which postulates that media technologies shape individuals’ thought and behaviour, as well as social structures.Footnote 13 This position was softened years later by advocates of a more lenient form of determinism, echoed by Merritt Roe Smith and Leo Marx in 1994 when they argued for the existence of some agency in technology and acknowledged ‘the human tendency to create a society that invests technologies with enough power to drive history’.Footnote 14 By then, the view of technology as an autonomous, supra-social, self-controlling, self-generating, and self-expanding force had been criticized by at least two lines of thought. On the one hand, critics such as Ruth Finnegan and later David Edgerton denounced historians’ overemphasis on inventions and their overlooking of the circumstances of production, purposes, choices, modes of use, control, and access, among other factors.Footnote 15 Insofar as humans not only invented technology, but also controlled its use and potential effect to some extent, it made no sense to consider it as the prime mover of social change or history. On the other hand, social constructivists questioned whether technology had and followed its own internal logic and determined human action. Quite the contrary, they insisted, it was human action—through social groups, conflicts, and interpretations—that shaped technology.Footnote 16 The present argument builds on the refutation of technological determinism, but takes it forward by analysing how technological and knowledge networks coextend each other through dynamic, hierarchical-free interactions.
The nature of the relationship between technological and knowledge networks, therefore, needs to be re-examined. Indeed, in his study of lay weather observers in the United States, Jeremy Vetter has argued that ‘the telegraph was a necessary if not sufficient historical condition for larger-scale forecasting’.Footnote 17 Furthermore, this re-examination is all the more necessary insofar as the expansion of the telegraph is intrinsically linked to the traditional account of the inevitable globalization of weather science. After all, telegraph systems, like postal services and journalism, transcend the local and move into the global—or, as Paul Edwards points out, form the elements of ‘infrastructural globalism’.Footnote 18 In fact, this account of inevitable globalization has also been challenged by Deborah Coen in her book Climate in Motion and other studies, according to which the success of the Austrian meteorologists actually lay in their ability to strategically shift between global and local perspectives.Footnote 19 Finally, recent cyclone meteorology historiography offers case studies in which telegraphy was not really the driving force of cyclone prediction: in her article on Henry Piddington’s cyclonology, Debjani Bhattacharyya shows how ‘a narrative science of cyclone forecasting emerged from the interstices of imperial trade’, namely from multiple textual source accounts;Footnote 20 and in his article on the rise of agrometeorology in Mauritius, Robert Rouphail demonstrates that this rise had much more to do with the role of cyclones in an ecology linked to sugar production than with the advancement of telegraphic-synoptic forecasting.Footnote 21 However, neither of these two studies, nor Vetter’s or Coen’s, have addressed the nature of the relationship between knowledge networks and telegraph networks, which is crucial to understanding this process of globalization.Footnote 22
To elucidate this relationship, the central question is no longer how to analyse science in a technological network but science as knowledge network. The epistemic network thus becomes the subject of enquiry and interaction, a mission and purpose, rarely identified, but analytically very powerful. The epistemic network suggested here is constituted by actors engaged in activities of production and circulation of scientific knowledge. The actors who form this community are not individuals with accidental links or weak affinities, but they all share a set of normative beliefs and principles, a physical world view, a common scientific enterprise, and social recognition—to a greater or lesser degree—as authorities in their science.Footnote 23
To understand the production and circulation of cyclonic knowledge in Asia in the nineteenth century, it is useful to make a first fundamental distinction between two types of actors, ‘researchers’ and ‘doers’—that is, between those who worked in the observatories generating scientific (and often publishable) knowledge and those (usually native personnel) involved in the generation and transmission of data, as well as technical installation and maintenance. Whereas ‘researchers’ were engaged in tasks implying a cognitive process involving the collection, interpretation, analysis, and application of data for predictive and practical purposes, ‘doers’ were concerned with observing and then recording and transmitting organized data without necessarily involving the understanding or application of the information transmitted.Footnote 24 The latter’s personal experience has traditionally been overlooked in favour of historical studies mediated by instruments and publications, as shown by Fiona Williamson’s work on ‘hidden meteorologists’ in Hong Kong and Mark E. Frank’s study of native observers in Tibet.Footnote 25 Furthermore, this distinction applies to the whole of observatory sciences in Asia, as S. Prashant Kumar’s recent study of caste and representation in Madras shows.Footnote 26 The government astronomer at this observatory defined the Brahmin assistants, who were responsible for observations and computations, as ‘machines without the certainty of machinery’.Footnote 27 These are, of course, analytical distinctions applied to cognitive activities and not mutually exclusive categories of cognition. Indeed, many Jesuit researchers at Asian observatories were also involved in observational work, as well as the construction and maintenance of instruments.
What is important from the point of view of the global history of knowledge is the study of the relationship between these network actors and the technological infrastructures that enable their activities. However, insofar as actors define the nature of the network and their actions affect the circulation and production of knowledge, the distinction of actors in observatory sciences is fundamental to elucidating this interaction. In reality, the Jesuit epistemic network involved in the mission observatories consisted of both Jesuit ‘researchers’ and multiple native ‘doers’ (observers, mechanics, calculators, draughtsmen, apprentices, and assistants), whereas those who served in the networks of telegraph and weather stations and customs houses, making meteorological observations as a complementary but remunerated task, were exclusively ‘doers’ (telegraphers and observers).
The recognition that actors (human and non-human) and networks are inseparable entities that co-construct each other is part of the genuinely critical and original side of Bruno Latour’s and Michel Callon’s actor-network theory. However, Latour’s characterization of the observatory as a ‘centre of calculation’ where a heterogeneity of actors converge, while useful to a point, defines observatory sciences in a very limited and rationalistic manner.Footnote 28 The concept of ‘epistemic network’ and the distinction between ‘researchers’ and ‘doers’ seem more precise and appropriate for a broader and more accurate interpretation of the complex observational realities that the critical historian must analyse.Footnote 29
The Jesuit knowledge network had one undeniable advantage over imperial networks: trans-imperial connectivity. The geographical continuity of its areas of influence and the telegraph and cable networks within those areas made it possible to continuously monitor typhoons throughout East Asia, from the eastern Philippines to south-east China. No less significant were the cyclone forecasting methods and the relational perspective adopted at the Jesuit observatories. With the help of the new communications infrastructure, Jesuit scientists in Manila and Zikawei were charged with establishing and operating cyclone warning systems and colonial meteorological services in the Philippines and China, as well as determining the routes and causes of typhoons following cyclone circulation patterns at a subcontinental scale. Intriguingly, other attempts to establish colonial meteorological services based on purely observational networks, either at telegraph stations (such as José Batlle’s meteo-telegraph network in the Philippines) or in maritime customs (Robert Hart’s plan for a meteorological service in China), were not successful. This scientific–technological interplay may be of great interest to the reader because it suggests a new perspective for understanding globalization processes: it is not simply an extension of networks, but globalization processes must be understood as a bidirectional interaction between networks of different types, as shared causal connections between cognitive, technological, and social networks; in short, as a coextension of networks in genere.
In what follows, I first examine the establishment of the Jesuit knowledge network in Asia before the advent of the telegraph, analysing the role of the relational perspective and Angelo Secchi’s meteorograph in the creation of a common scientific enterprise. I then show how the design and expansion of telegraph networks were preceded in the Philippines and partly in Shanghai by the idea of their application to cyclone warning and coastal surveillance. I describe how the Jesuit epistemic network benefited from the expansion of telegraph and cable infrastructure in the region, extending its observing networks and its power by directing colonial meteorological services. Finally, I offer a counterpoint from a network of observers that did not come to fruition, demonstrating that telegraphy was not a sufficient condition for the achievement of weather forecasting and cyclone warnings.
Jesuit epistemic networking in China, the Philippines, and India
Jesuit scientists adopted a meteorological strategy that sacrificed statistical observation in favour of correlation. Led by Angelo Secchi and Stephen Perry, directors of the Collegio Romano and Stonyhurst College observatories in England, respectively, they sought to equip Jesuit collegiate observatories in colonial Asia with modern instruments, channelling their resources into tropical cyclone forecasting. Their plans could be framed within the transformation of observatory sciences through the use of correlative rather than statistical approaches. In this predictive framework, the relationships between meteorology and terrestrial magnetism were central. As Secchi repeatedly recalled, ‘every great storm in Rome is usually preceded or accompanied by a magnetic disturbance’.Footnote 30 The Jesuit scientists were well aware of the decisive influence of solar activity (whether through radiation, flares or sunspots) on terrestrial climate and magnetism, although they referred to this correlation universally and without making distinctions between regions. Science teachers in some of the Jesuit schools set up after the return of the Society of Jesus, as a result of the educational interests of imperial states, in places such as China, the Philippines, and India, were able to test this approach. In 1865, Spanish Jesuits set up a modest weather station at the Ateneo de Manila High School. Based on barometric curves, mathematics teacher Francisco Colina was able to predict a typhoon that passed near Manila that year. Accordingly, the local merchant and shipowning community was willing to pay for the necessary equipment if the Jesuits undertook to set up a fully fledged observatory.Footnote 31 To this end, Colina met Secchi in Paris at around 1867, and asked him for magnetic apparatuses (via Stonyhurst) and auto-recorders, as well as the regular dispatch of the Bullettino Meteorologico of the Collegio Romano. At Secchi’s own request, Colina also sent him Federico Faura, a young Jesuit who was to be trained at Rome and Stonyhurst before becoming director of the Manila Observatory in 1878. In Calcutta, the influential St Xavier’s College was founded by Belgian Jesuits in 1860 with the ambitious plan of offering a course of study ‘similar to that pursued in the great colleges of Europe’, with advanced science classes and a modest meteorological station erected in 1867.Footnote 32 One of its most distinguished professors, Eugène Lafont, achieved glory for his warning of the famous cyclonic storm of November 1867. Later, Lafont also ordered a self-registering apparatus from Secchi, and his successor, Edward Francotte, sought correlations based on Calcutta’s weather data from 1868 to 1918. So, via Paris, Rome and Stonyhurst provided the instruments; and Manila and Calcutta, the enthusiasm and projects.
One might think that the curriculum and meteorological activities at Collège St Ignace in the Kiang-nan mission would be relatively similar. However, Zikawei found both similarities and significant differences. To begin with, scientific subjects, with the exception of mathematics, were not taught until later.Footnote 33 Unlike the Ateneo’s and St Xavier’s Jesuit scientists, who taught physics and mathematics, those at the Zikawei Observatory were rarely assigned to teaching duties. An important consequence of this situation is the progressive concentration of scientific research at Zikawei, with no or very little effective links to teaching and the Collège St Ignace.Footnote 34 However, the origins of the Observatory does bear similarities to those of Manila and Calcutta. In 1842, the French government, acting as protector of the Catholic missionaries in China, reopened the mission that had been abandoned after the suppression of the Society of Jesus. The first Jesuit meteorological observations were made in 1865 by Henri Le Lec, a science teacher trained at Stonyhurst, from where he brought elementary instruments. The vicar apostolic of Nankin, Adrien Languillat, took up the idea of setting up an observatory and asked Rome to send him one of Secchi’s assistants. In 1869, Fr Auguste Colombel, a former student of Stonyhurst, was posted; and in 1873, at the request of Secchi and Perry, the man who would become the director of the new observatory, Fr Marc Dechevrens, was sent. Dechevrens had trained with Perry at Stonyhurst and was to be responsible for installing Secchi’s self-registering apparatus.Footnote 35 As in Manila, with a view to correlation, Zikawei began by bringing together two sections in the same premises: meteorology and magnetism.
In exploring the reasons why the Jesuit correlative method necessarily imposed its own observational model on the observatory sciences, we must first note that the finding of physical correlations provided empirical evidence in favour of Secchi’s theory of a unified universe. L’unità delle forze fisiche (1864), Secchi’s quintessential essay on natural philosophy, made clear the idea that would guide the purpose of the Jesuit correlative approach: ‘connecting the vast number of phenomena that comprise the manifestations of [natural] forces and showing their mutual connection’.Footnote 36 The dynamistic and holistic view of physics had a relational dimension, as the harmony of physical forces denoted the correlation between them. It must be emphasized that this harmony implied much more than the simple idea that the fundamental forces governing interactions in the universe were interrelated: it meant that the harmony of physical forces reflected the harmony between science and religion. Thus, when the unity and interconnectedness of forces were advocated, a unified universe was defended in which the existence of correlations in no way violated the divine natural order and harmony.Footnote 37
The concept of physical correlation was not only instilled in the minds of the Jesuits trained in Rome and Stonyhurst who were to be posted to the Asian observatories; it also changed how they were to observe phenomena. Secchi invented a colossal, multi-instrument, self-registering apparatus that fused mechanical technology with novel electromechanics and was awarded a prize at the Paris Universal Exhibition of 1867.Footnote 38 Secchi’s meteorograph became the cornerstone of the new tropical predictive meteorology. By 1874, for example, it had been installed in Manila, Calcutta, Shanghai, and Havana. Its invention was a qualitative leap forward in at least two respects.Footnote 39 The first relates to the automation of observing and registering tasks: insofar as it was able to record meteorological elements automatically and continuously, it made it possible to separate this work from the observer’s own tasks, facilitating the organization of centralized data services. The second aspect was the graphical and simultaneous display of results. What distinguished it from other contemporary self-registering devices was its emphasis on simultaneous visualization. Instead of recording parameters on different sheets and in tables, the meteorograph recorded all the elements on a single sheet and in graphs and curves. Thus, the observer could detect ‘at a glance the succession and concatenation of all phenomena’.Footnote 40 The meteorograph transformed the task of recording into an operation aimed at constructing relationships between physical phenomena.
Secchi’s meteorograph not only influenced the modus operandi of the observations conducted by the Jesuits at the Asian observatories; it also changed the way they published and organized results. As Secchi noted in the introduction to the first issue of the Bullettino, the journal was intended to be a window into the precise observation of the weather and the variations of terrestrial magnetism in relation to atmospheric electricity.Footnote 41 The meteorograph marked a turning point in the regular bulletins of these observatories. The Manila Boletín, for example, in its initial period (1865–1869), consisted of a single sheet containing a few tables, six graphs (with daily averages), and a summary of atmospheric phenomena. After the meteorograph was installed, from 1870 to 1883, its four pages (eight from 1880) contained nine tables (each with fourteen meteorological variables), a sheet of magnetic and meteorological curves, and a map of typhoon tracks.Footnote 42 Secchi’s tutelage was well felt: Faura asked him for advice on the plan of observations and publications underway. The scientific bulletins of Manila and Zikawei acquired a space of meaning shaped by Secchi’s meteorograph.Footnote 43
While the culture of correlation suffused Rome and Stonyhurst, nowhere was its force felt more directly and viva voce than in the observatories of the Asian and American missions of the Society of Jesus. In particular, the Asian observatories, founded on the principles of social service, supported by the local merchant and shipping communities, and aiming to create a reliable storm warning system amidst an environment punished by tropical cyclones (whether typhoons or monsoons), proved fertile ground for a marriage between the quest for natural knowledge, the demand for weather forecasting, the need for maritime and agricultural safety, and a network of knowledge that was to be coextended in the following decades by the laying of telegraph and cable networks.
Telegraph and cable networking
Following the Qing government’s recovery from the Taiping rebellion (1850–1865), the Nian rebellion (1851–1868), and the military defeats of the Opium Wars, secure and reliable international communication became a necessity. Under pressure from the foreign merchant class demanding for concessions, Britain, Russia, France, and the United States insisted on introducing telegraphs and railways into China, as they would later do in Japan, on the grounds that the Chinese communication system was obsolete. Spreading their tentacles throughout the northern hemisphere, the Western telegraph companies wanted to include China in their global plans to link Europe with Asia.Footnote 44
For Westerners, the scene was one of competing interests. Competition was fierce, not only between the countries involved (through ambassadors and diplomats), but also between companies seeking concessions from the Chinese government to lay landlines or undersea cables. Two big companies set the pace. The Great Northern Telegraph Company (GNTC), founded in Denmark in 1870, sought to extend its undersea cables to the Asian coast of the Pacific Ocean, from Siberia to Osaka, Yokohama, or Nagasaki in Japan, and to Shanghai, Fuzhou, and Hong Kong in China.Footnote 45 The Eastern Extension, Australasia, and China Telegraph Co., a merger of three British companies, aimed to extend the great chain of submarine cables—being built between England, India, and Singapore—from Singapore to Hong Kong, and then to Shanghai, and then to other treaty ports.Footnote 46
For the Chinese authorities, however, the scene was very different. For many and varied reasons, the Chinese central government and provincial authorities in the 1860s showed little interest in the use of telegraph and consistently refused any request for a concession. Examples abound. In 1865, for instance, Robert Hart, the Inspector-General of the Chinese Maritime Customs, and Thomas Wade, a consul in the British Legation, submitted separate proposals to the Zongli Yamen, requesting the adoption of Western technological systems. In a discussion with coastal officials, the Zongli Yamen described telegraphs and railways as ‘forbidden enterprises’.Footnote 47 Among the varied reasons adduced by the Chinese authorities was the threat of telegraphs and railways as a cause of large-scale social dislocation. Almost all officials argued that telegraph and railway lines would negatively alter the local landscape and undermine the legitimate interests of local people.Footnote 48 The threat to national security was another reason for Chinese reluctance. Placing communication technology at the disposal of Western powers would jeopardize Chinese sovereignty in times of war. The threat was no less in peacetime, as the laying of telegraphs would encourage the expansion of foreign influence beyond the treaty ports. As one Chinese official pointed out, ‘Westerners were obviously not just intended [to introduce telegraphs] to coastal areas.’Footnote 49
However, a great deal of the Chinese unwillingness to grant concessions to telegraph companies or to adopt Western communications technology was entirely rational and not merely bound by culture and geopolitics: the introduction of telegraphs brought with it telegraph networking—as did the railways. The Chinese government soon realized that, in a country on the brink of economic collapse, there would be no way of controlling the expansion of the telegraph network if it allowed Western companies to promote its construction. In fact, the Chinese government did not systematically and per se reject all Western technologies but rather showed a particular reluctance to accept telegraphs and railways because of their very network character—that is, the fear that telegraph networks, rather than individual telegraph lines, would be built.Footnote 50
Chinese reluctance notwithstanding, Western companies managed to demolish this ‘Great Wall’ through a unique tactic. Like all other network technologies, the extensibility of the telegraph depended on the insertion and concatenation of wedges (i.e., short lines). The very existence of a wedge was per se a powerful reason and invitation for its extension. Once one link was opened, Western traders thought, the way was clear for other wedges. This was especially true where the penetration of the ends and the introduction of the short lines were so gradual as to raise little popular opposition or objection from the Chinese government. Once a wedge had been successfully established and socially accepted, the Chinese authorities would understandably be induced to endorse its extension or to undertake such a venture at their own expense and risk. In either case, the establishment of a telegraph network would be unstoppable.
The introduction of telegraph wedges in China was closely intertwined with the laying of submarine cables. In the 1860s, on behalf of the China Submarine Telegraph Company, Wade negotiated with the Zongli Yamen to explore the possibility of laying submarine cable-based telegraph lines along the Chinese coast. Following Hart’s advice, he made a clear distinction between telegraph lines on land and those at sea. According to this plan, the cables would remain at the bottom of the sea, hidden from view, to avoid sabotage by local people, and only the cable ends would be visible and floating, anchored in some way outside the harbour.Footnote 51 Their tactics began to take effect in 1870. Prince Gong came to recognize that if the cables did not reach the shore, there was no need for much negotiation.Footnote 52 He did, however, demand that the cable should ‘terminate in a hulk to be anchored outside the port’ and that the Chinese authorities should not be responsible for any damage caused to the cables. As a result, the construction of the first telegraph wedge in China went underway.Footnote 53 China’s access to the international communications network came at the price of a significant relaxation of its official policy of prohibition. The Qing government agreed to regard the cable as ‘a thing to be protected’.Footnote 54 In this context, and partly to pre-empt its rival, the China Company, the Danish GNTC rushed to lay a cable between Hong Kong and Shanghai without asking for permission from the Chinese government and presupposing that Wade’s case would apply to the GNTC cable. Then, in 1870, the GNTC and the China Company signed an agreement to operate in Asia.Footnote 55 While the GNTC completed its cable between Hong Kong and Shanghai in 1870, and between Shanghai and Nagasaki and Nagasaki and Vladivostok the following year, the China Company laid a cable between Singapore and Hong Kong at around the same time.Footnote 56 By 1873, China was connected to the international telegraph network by coastal lines.
The case of the GNTC illustrates well how telegraph and weather warning networks interacted dynamically. In June 1873, G. H. N. Dreyer, the GNTC’s superintendent at Shanghai, agreed with the harbour masters of Hong Kong, Shanghai, and Amoy, and the Dutch chemist Antonius J. C. Geerts of Nagasaki, to telegraph daily observations on the state of the weather and the strength and direction of the winds.Footnote 57 The GNTC’s deployment in the China Sea greatly facilitated the enterprise, as the stations in these harbours were the hubs for the undersea cables laid by the Danish company. Not coincidentally, the GNTC decided to provide this service free of charge. To understand the reason for this measure, it is necessary to realize that the GNTC’s dominant position was not supported by the local merchant community. As a foreign company operating in China, it was viewed with suspicion by large sections of the business community, who felt that it prioritized its economic interests over those of the local population. Hence, the free telegraph-meteorological service seemed to be an obvious and fundamental way of reducing this hostility and facilitating the renewal of future concessions and obtaining permits to expand its operations. According to historian Marlon Zhu, this favour was ‘one of the Company’s measures to improve its poor public relation with the mercantile communities in these ports’. Moreover, this favour ‘was unique and unparalleled in other areas suffering from tropical storms around the world’.Footnote 58 The decision soon had a positive impact. In August 1873, Hong Kong’s influential English-language newspaper, the China Mail, began publishing a daily wired weather column called the China Coast Meteorological Register (or simply Register).Footnote 59 A few days later, the Shanghai editor of the North-China Daily News (NCDN) urged the harbour master to emulate his Hong Kong counterpart, and send records for publication in the local press.Footnote 60 By the 1880s, other companies such as the Eastern Extension Telegraph Co. and the Chinese Telegraphic Administration were freely transmitting meteorological telegrams in East Asia, in imitation of the GNTC. Meteorological information was the tool used by the telegraph companies to expand into the mercantile community.
With the establishment of the first telegraph connections in China, Japan, and Cochin China in the early 1870s, the Philippines emerged as a key node in the economic chessboard of South-East Asia. Evidence of this interest was a series of applications for cable concessions that were rejected by the British government, which did not allow cable to be laid from Hong Kong and Singapore. The public tender of 1879 was the first opportunity for the Spanish government to relax the concession conditions and offer a closed tender, as requested by many foreign agents.Footnote 61 Two companies with proven track records, the Eastern Extension and the Telegraph Construction and Maintenance Co., won the tender to lay a cable between Hong Kong and the Philippine cape of Bolinao, and then extend a land link to Manila. The first direct telegram between the Philippines and Spain was transmitted on 2 May 1880.Footnote 62
Actually, this telegraph connection was part of a wider process of modernization of the telecommunications infrastructure, which had taken its first steps at the domestic level in the 1860s. The colony had already installed an optical telegraph line to announce the arrival of ships. In 1867, the engineer José Batlle was commissioned by the Spanish government to study the introduction of the telegraph in the Philippines, and replace the optical system with an electrical one.Footnote 63 As a result, Batlle submitted a general plan of domestic telegraphic communications, which soon took shape.Footnote 64 In 1872, the Escuela Práctica de Telégrafos was established in Manila and the first telegraph line was opened between Cavite and Manila. Interestingly, Batlle’s plan for a domestic telegraph network was largely based on strictly meteorological criteria. The available evidence is very reliable—the testimony of Batlle himself, who made his aims and ambitions public. In his plan, published in the Revista de Telégrafos in 1873, Batlle explained the ‘ideas that had preceded the proposal’ as follows:Footnote 65
In the execution of all these studies, we prioritized the idea of giving telegraphy the multiple applications that it has for uniting localities, for meteorology and weather forecasting, and for coastal vigilance and, to that purpose, the projected [cable] lines terminating in a semaphore located at the most suitable point along the coast, with its observatory and instruments required for discovering the presence, direction and velocity of the meteors, and communicate the appropriate warnings for the lines, in order to predict disasters that might occur from distant points.
The imperative need to implement a system of storm warnings and weather forecasting dominated the design of Batlle’s national telegraph network plan and even prompted the overlapping of functions and interaction of personnel sharing resources, practices, and objectives that were useful for the transmission of meteorological information. This is evidenced by the composition of the telegraph network: of the fifty-nine stations established in the Philippines between 1873 and 1881, forty-one were electro-telegraphic, five semaphores, and thirteen electro-meteorological.Footnote 66 However, during this period, the desired outcomes were not achieved. The network of telegraphers—made up of ‘doers’ (namely, observers and mechanics who handled instruments and operated telegraph machines)—was incapable, by itself, of making forecasts or storm warnings. Batlle’s expectations were only fulfilled when this network of ‘doers’ was incorporated into the Jesuit network in 1884, as shown below.
Coextending the Jesuit network of meteorological knowledge
The telegraph infrastructure plans promoted by the Spanish government in the Philippines sought ‘meteorological’ colonial security. They sought to put technology at the service of predictive science and predictive science at the service of technology. Telegraph communication had to be encouraged because weather forecasting depended on it, and weather forecasting had to be encouraged because the prevention of cyclonic disasters depended on it. In fact, by 1873 the government had already laid more than a thousand kilometres of lines on the island of Luzon and installed three semaphore service stations with meteorological instruments to guide ships in the port of Manila. Therefore, meteorology had contributed to the expansion of telegraph networks. However, the mere laying of lines and cables did not carry an overt extension of the Jesuit knowledge network. The new infrastructure was used by both telegraph observers and semaphore observers. The telegraphic influence on the Jesuit network had to be made explicit. How did then the Jesuit epistemic network benefit from the telegraph network?
At this point, the Philippine Meteorological Service helped to furnish an answer. Created by Royal Decree on 28 April 1884, the Philippine Meteorological Service was the first official service of its kind to be established in Spain. In fact, its genesis sheds a lot of light on this question. In 1880, following the installation of the undersea cable between Hong Kong and Manila, the British government wanted to extend its meteorological service to the entire China Sea. To this end, the governor of Hong Kong, John Hennessy, invited his Philippine counterpart to a ‘daily and reciprocal exchange of meteorological observations’ between the ports of the two cities.Footnote 67 The Spanish government accepted the invitation and extended the network of telegraph stations run by Batlle. Interestingly, the government decided to articulate this network not around the telegraphists but around a particular nerve centre: the Jesuit-run Manila Observatory, where all the functions of coordinating and recording observations were centralized. The Observatory came under the Spanish civil administration, and its main purpose was to forecast cyclones and the weather in general. It also assumed control of the thirteen electro-meteorological stations operated by the telegraphists. The commission responsible for organizing the new service embodied an institutional triad formed by the three sensitivities at stake: the Telegraph Corps, the Jesuits, and the Navy.Footnote 68 Their first proposal was to establish a network of secondary stations in various ports and provincial capitals, to be chosen on the basis of telegraphic accessibility from Manila. The network thus constituted would have a telegraphically well-communicated structure and would be of great practical utility to the military and merchant navies, which was precisely what the Spanish authorities were looking for.Footnote 69 The Philippine Meteorological Service thus sheds much light not only on how the meteorological network and the telegraph network shaped each other, but also, from a broader historiographical perspective, on how the epistemic and the technological co-evolved in this period.
The Servicio’s pioneering nature notwithstanding, telegraph technology came to the Philippines at a price. As in almost all other modern infrastructures, the country’s capital occupied centrality. All domestic telegraphic traffic had to pass through Manila at some stage. But when it came to international communications, Manila became the terminal node or endpoint in a regional network in which the Hong Kong–Manila line segment was a mere wedge with no continuity. This was neither accidental nor coincidental, but the very consequence of Philippine reliance on British cable technology. The problem was compounded by the poly-island and labyrinthine nature of an archipelago made up of 11 main islands and more than 7,000 islets. The Spanish government had neither the financial resources to extend its domestic network to other islands by laying costly submarine cables, nor the technical know-how to maintain and repair them.Footnote 70 This resulted in serious structural limitations to the meteorological service. Thus, the founding decree of 1884 stipulated that the secondary stations would be established on the island of Luzon, as there was no telegraph infrastructure in the rest of the archipelago.Footnote 71 Elsewhere, little progress was made. The network was not extended until 1897, when a twenty-year concession was granted to the Eastern Extension, which undertook to lay a cable from Luzon to the islands of Panay, Negros, and Cebu.Footnote 72 This telegraph expansion brought not only the establishment of new secondary stations in Cebu, Iloilo, and Capiz, but also an increase in ‘doers’ within the Jesuit epistemic network, with the hiring of more observers, mechanics, draftsmen, calculators, and assistants (all of whom were native Filipinos).Footnote 73
In China, the meteorological activities of the Zikawei Jesuits slowly began to have some impact on Shanghai society in the late 1870s. Marc Dechevrens, breaking with the purely observational practice of Robert Hart, head of the Chinese Imperial Maritime Customs, published a pamphlet entitled Le Typhon du 31 Juillet 1879, which included a map of its track.Footnote 74 This earned him the appreciation of many local merchants, who had access to the weather forecast notes published in the trade newspapers. Ironically, Dechevrens had drawn heavily on customs observations, supplied by captains of treaty ports, for his typhoon analysis. For Dechevrens, the collaboration between his observatory and the Customs Service’s observers in ports and lighthouses seemed to provide a firm basis for empirical prediction, the implementation of which would be of use to captains preparing to leave port by informing them of the winds and weather conditions they would encounter en route.Footnote 75 But Dechevrens’s tacit proposal was objected to by Hart, who completely rejected the underlying idea of giving Zikawei central status. Both the mercantile community and the local trade press attached great importance to the announcement of typhoon warnings. The work of Dechevrens, and Zikawei in general, was appreciated more for its predictive potential than for its more fundamental theoretical perspectives.
Insofar as Dechevrens’s technical analysis of the 1879 typhoon entailed a predictive view of meteorology that differed significantly from the observational view of Hart and the Customs Service, the nationwide adoption of his method was promoted by the Shanghai General Chamber of Commerce in September 1881, shortly after a strong typhoon had struck the region. Its chairman, Francis B. Forbes, convened a meeting to discuss the ‘feasibility of organizing a system of meteorological reports from the China coast and the interior’. The purpose was to improve ‘the knowledge of the origin and direction of storms, and to warn mariners of their approach’.Footnote 76 Dechevrens was entrusted to lead the project and the Chamber formed a committee of delegates made up of traders, the GNTC, and shipping companies, which endorsed the plan.Footnote 77 In addition, there was the special position occupied by Hart and the Customs Service, which initially were against it, but eventually caved in to growing pressure from the Shanghai press and mercantile community.Footnote 78 Forbes had urged Dechevrens to think of ‘a system of meteorological observations to be organized on board the Shanghai merchant vessels and the coast of China’.Footnote 79 This reformulation provided an explicit statement of the practical issues to be addressed by the new Chinese meteorological service, issues that concerned Zikawei, either directly or indirectly, insofar as the Jesuit knowledge network became a reference for them. These included, among others, the transmission without delay of weather records from ships and customs stations to Zikawei, the preparation of weather reports and forecasts by Dechevrens, the provision of the necessary instruments to ships and the Customs Service, and the free transmission of weather messages by telegraph companies. Moreover, the timing was very opportune. Within months of the plan’s approval, the GNTC was laying the first cable along the East Asian coast, the Chinese government was approving the construction of an inland telegraph network, a telegraph landline was being laid from Shanghai to Canton, and the Eastern Extension was laying a submarine cable between Hong Kong and Shanghai via Foochow.Footnote 80 The Chamber delegates’ formulation of the role of telegraphy demands particular attention, not only because it made communication possible, but also because it enabled the very extension of the network of meteorological stations.
The decision to entrust the Jesuits with running what became known as the China Coast Meteorological Service (CCMS), founded in 1882, cannot be understood unless the roles not only of the telegraph, insurance, and shipping companies, but also of the local newspapers in Shanghai and Hong Kong, are taken into account. With the success of the Jesuit prediction of the 1879 typhoon, their readers envisaged a world free of cyclonic threats, whose early warning would bring security and prosperity to the inter-port community. In their wake came a quest, almost a race, to organize and define this new science, and to create an accurate and timely warning system (the CCMS): readers acted as the main shapers of inter-port telegraphic meteorology, reading and comparing the different warning announcements, sending praise and criticism to local newspapers, and forcing shipping and insurance companies to introduce improvements in the weather service. In fact, it was public criticism in the press that caused Hart to relent and to change his mind about not cooperating with Dechevrens and Zikawei; and it was the inter-port press system that, through its editorials, forced the Customs Service to incorporate the CCMS plan.Footnote 81 The Zikawei Observatory, which provided both a scientific method and a transnational knowledge network, became the shibboleth of the inter-port press.
The Zikawei Observatory became the centrepiece of the CCMS in the following decades. The Jesuits made advances in several areas of the weather system. In all of them they faced very different opponents. The first breakthrough concerned the installation of a weather semaphore and a time-ball on the Shanghai Bund in 1884. Their weather forecasts and time service attracted some criticism in the newspapers, as the Jesuits switched to presenting their weather data in the metric decimal system instead of the English imperial system to which the population was accustomed.Footnote 82 The second development, which was also controversial, concerned the standardization of methods and codes. In 1897, Dechevrens’s successor, Stanislas Chevalier, proposed a code for weather telegrams to the Far East, which was welcomed by his confreres in Manila but rejected by the director of the Hong Kong Observatory, William Doberck, on the grounds that codes referring to winter storms in the north were irrelevant to the British colony.Footnote 83 Finally, there was Zikawei’s daily weather bulletin, which from October 1882 was inserted in local newspapers, such as the NCDN, and generated considerable enthusiasm. If allowance is made for the great difference between the services provided before and after 1882, Zikawei’s expansion was truly remarkable.
This expansion was not without its difficulties. In the course of his decade-long directorship of the Observatory, Dechevrens at times expressed himself with varying degrees of optimism or pessimism as to whether Zikawei could guarantee an effective typhoon warning service. It is apparent from his studies that Dechevrens made very limited use of the telegraph network. In his notes in the Zikawei Bulletin Mensuel of 1885 and 1886, he lamented the poor telegraph connections to places outside Shanghai.Footnote 84 For the practical purposes of cyclone forecasting, this became almost synonymous with inefficiency and inability to foresee. The Franco-Chinese War (1884–1885) had partially damaged the telegraph infrastructure. In those years, none of the telegraph lines along the coasts and inland were directly connected to the Observatory. Most of the information gathered by stations in the seas east of Shanghai and inland, with the exception of Amoy and Hong Kong in the south, was excluded from Zikawei’s scope. Consequently, the storm warnings transmitted by semaphore were not based on external information but on the Jesuits’ own barometric observations.Footnote 85 The Zikawei warning service embodied a limited version of the new CCMS, which aspired to bring forecasting and telegraphy into an indispensable alliance in the future.
Counterpoint: Failed weather services
In the 1860s and 1870s, the establishment of telegraph and cable infrastructures led Hart to promote plans for meteorological services. These plans were of two kinds. On the one hand, the plan for a domestic meteorological service in China, to be articulated through the CCMS, began to be considered of great interest. On the other hand, the intercolonial and transnational weather plan, to be coordinated from a central observatory in Beijing, took advantage of the undersea cables being laid along the east coast of Asia. As shown below, neither plan came to fruition, despite being sustained by the extension of communication infrastructures. The dreamed-of Central Weather Bureau in Beijing never materialized and the Customs Meteorological Service was not expanded until the twentieth century, when it came under the control of the Marine Department.Footnote 86 Hart’s failed plans are a counterpoint to the Jesuits’ achievements.
Hart’s plan to establish a customs-run meteorological service was proposed (in a circular issued in November 1869) with a dual objective, both scientific and maritime. Through the systematic collection of meteorological data, he aimed at ‘throwing light on natural laws’ and furnishing data of ‘practical value’ to the ‘seafaring men and others on the Eastern Seas’.Footnote 87 To this end, he proposed to set up a network of weather stations (about fourteen in all) in the customs houses of all ports, both coastal and riverine, and to instruct the customs commissioners (all Europeans) to make observations. In a subsequent memorandum of 1873 to his deputy in London, James Duncan Campbell, Hart extended his plan to the network of coastal lighthouses, whose keepers, he said, ‘were fairly intelligent Europeans, steady men who can read and write’ and would be suitable for observational duties.Footnote 88 He urged Campbell to order twenty sets of meteorological equipment to be shipped to China and distributed to stations.Footnote 89 Hart envisioned his network being organized around a meteorological office in Beijing, where all coordination and recording functions would be centralized. The stations would be linked to a chair of astronomy at Beijing College (Tongwenguan). Part of the plan came to fruition: of the thirty-two stations in China reporting data to Zikawei in 1880, twenty-seven belonged to the Customs Service observer network (thirteen harbour stations, twelve lighthouses, and two lightships).
Hart’s meteorological plan was actually part of a wider modern infrastructural development then taking shape in China with the coastal lighthouse building programme (from 1867), the courier services, and the hydrographic department project. Behind this drive was a truly unique institution, the CCMS, an official agency of the Qing government, albeit within the British sphere of influence.Footnote 90 Founded on paper to collect revenue, in practice it performed a wide range of other tasks, such as facilitating navigation along China’s coastline, providing hazard warnings, assisting with local tides and currents, and ensuring order and safety at anchorages. The CCMS acted as a shoehorn for Western powers’ interests in China.Footnote 91
At about the same time that Western companies were opening the first submarine cable links to the east of Asia, a more ambitious intercolonial and international meteorological scheme was capturing Hart’s imagination. In 1873, he wrote to the governors-general of Hong Kong and Singapore (Britain), Manila (Spain), Saigon (France), Eastern-Siberia (Russia), and Java (the Netherlands) and to the governors in Nagasaki, Yokohama, and Bangkok, inviting them to cooperate in his plan for the publication and exchange of meteorological data.Footnote 92 The plan envisaged that the costs of equipment, personnel, press, and telegrams would be shared by the governments involved if the telegraph companies so requested. It also provided for the establishment of twelve stations between Possiette (the port of Vladivostok where the cable landed) and Batavia, at various points between 100º and 140º east longitude and 10º south and 50º north latitude, in a first phase, and eight more at southern hemispheric points in a second phase. In doing so, Hart sought to fill the gaps existing in the Pacific and provide a comprehensive network in the part of the globe between 100º and 180º east longitude.
Ironically, the desire for control and centrality also exposed serious weaknesses in Hart’s domestic and international meteorological plans. In the 1873 memorandum, Hart proposed that the customs stations should be linked to a chair of astronomy at Peking College, and that a customs central observatory should coordinate the whole project. But despite his best efforts, he was unable to find anyone qualified to lead such an enterprise.Footnote 93 There was widespread suspicion among the Western governments involved that Hart’s plans lacked scientific leadership. For example, when the governor-general of the Philippines received his plan, he appointed a commission, including the Jesuits of the Manila Observatory, who unanimously supported the project.Footnote 94 However, according to the Jesuits, Hart’s project ‘soon failed as a consequence of lacking a somewhat defined direction’.Footnote 95 Indeed, Hart was unable to find a research astronomer for the chair, and worse still, one of the four candidates in contention sued him when he discovered that Hart’s promises regarding the library, observatory, and teaching load were false. Hart had to pay him damages. This legal scandal caused tensions between Hart and the Shanghai mercantile community, damaging his public image in the local press.Footnote 96 In other words, the network of weather observers provided no guarantee of reliable analysis to predict the threat of any typhoon or storm that might form in the waters of the China Sea. This kind of consideration gradually dampened the enthusiasm for his meteorological plans among many, for whom a network of mere informers with no leadership or authority held little appeal.
Conclusion
Meteorologists and historians have considered the telegraph infrastructure a key factor in enabling large-scale weather forecasting and storm warnings. As early as 1871, the American meteorologist Cleveland Abbe argued that it was only possible ‘to study with advantage the progress of atmospheric changes when the telegraph lines had become widely extended over the earth’s surface’, extolling the power of ‘weather telegraphy’.Footnote 97 More recently, historian Mark Monmonier embraced that idea: the synoptic weather map appeared only ‘when the commercial electric telegraph made possible the timely collection of synchronous observations of surface weather’.Footnote 98 The case of the Jesuits and tropical cyclones in colonial Asia suggests that the relationship between communications infrastructure and forecasting services was not so simple or linear. As the telegraph and cable network was laid in East Asia and the Jesuit scientific network spread across the region, concomitances between the two were as remarkable as monocausal relationships. Moreover, as the cyclone warning systems in China and the Philippines suggest, the interaction between the two networks was fully bidirectional, as necessary for science as for trade, and just as beneficial for meteorologists as for telegraph companies.
In cyclone science, the Jesuits developed forecasting methods and typhoon warning systems that were unrivalled in the commercially competitive East Asian scenario. They pioneered tropical cyclone warning systems, while applying a relational perspective to the study of seemingly disparate physical phenomena. Their production rate was also extraordinary. From 1865 to 1898, the Manila and Zikawei Observatories published 154 and 58 typhoon studies, respectively, while the Hong Kong Observatory produced only 14.Footnote 99 Yet this Jesuit network was not made up of mere observers. New knowledge was generated by analysing, interpreting, and applying observational data. Last but not least, it produced research monographs on atmospheric phenomena and events (up to twenty-five in Manila, from 1880 to 1900), which contrasts with the pattern followed by most neighbouring observatories, where production was more closely linked to periodicals.Footnote 100 Interestingly, by extending a network to analyse the course of typhoons on local, regional, and international scales, the Jesuits fostered observational standardization. They provided experts and informants alike with instructions for making observations that were conceived as hybrids, embedded in both universal orientations and colonial specificities. To what extent were these instructions important? What did they represent? These questions underlay the routine activities of the meteorological services in this period, but they also reflected the disparity of networks.
Indeed, the contrast between experts and informants is historically instructive. It foreshadowed the most significant dividing line between knowledge networks and observer networks in the 1870s and 1880s. This division was rooted in different meteorological functions. On the one hand, there were those experts whose conception of observatory science was imbued with the Jesuit empirical culture and for whom the meteorologist’s functions were the collection and organization of data and their subsequent analysis and interpretation for predictive and practical purposes. These experts from the Jesuit observatories were accompanied by native employees in positions such as observers, calculators, mechanics, draughtsmen, and orderlies, who handled and repaired instruments, operated telegraphs, and kept records.Footnote 101 On the other hand, there were informants for whom meteorological observations were part of an additional, explicitly paid activity, the aim of which was not to generate knowledge from the data, but to fulfil the tasks assigned to the telegraph and semaphore stations. This division was marked by attempts to provide observers with technical instructions in terms acceptable to scientific standards, instructions that sought to systematize observations, unify protocols, and use instruments correctly to avoid measurement errors. In the Philippines, for example, in 1881 Faura published practical instructions for secondary stations, describing local peculiarities and how to send daily messages to the central station in Manila.Footnote 102 This contrasts with the book of instructions given to telegraphers in the Philippines in 1872, with general descriptions of meteorological instruments similar to those given to observers in Spain’s educational institutions.Footnote 103 This contrast was largely determined by the fact that in China and the Philippines the Jesuit knowledge network had become familiar with local atmospheric conditions. Those rulers and traders who relied on the Jesuit network appreciated the real challenges of meteorological services, which, of course, involved much more than strictly observational matters. In this respect, the networks of meteorological-telegraphic and customs stations promoted by Hart and Batlle were identified by those rulers and traders as networks of mere observers and therefore unlikely to form the basis of national meteorological services.
Regardless of Hart and Batlle’s attempts to instruct observers at telegraph and customs stations and their apparent lack of rigour in observational standards, it is clear that the recognized Jesuit scientific authority contributed to their meteorological leadership in China and the Philippines. Compared to the project which Hart embarked on, the Jesuit network seemed to have the full confidence of the mercantile communities in Shanghai, Hong Kong, and Manila, even though—ironically—Hart had ambitions to head the weather service in China, despite his reluctant commitment to Dechevrens’s plan. The establishment of the Jesuit network and the way it expanded in the Philippines and Shanghai, I suggest, were shaped by the contingencies of telegraph and cable companies, the interplay of geopolitical interests, and the security needs of mercantile communities in the face of cyclone and storm threats, just as the expansion of telegraph infrastructure in those regions was shaped, in part, by considerations of cyclone warnings and maritime and commercial security.
Finally, an examination of the relevant meteorological literature does not lead to the conclusion that the development of telegraph infrastructure per se was the direct causal factor that enabled the establishment of cyclone warning systems in Asia. Cyclone threats were well known and had already justified the constitution of a limited form of cyclone warning by the Jesuits in Manila, Calcutta, and Shanghai even before the installation of the telegraph in the early 1870s. The telegraph and cable networks laid in the 1870s undoubtedly catalysed these warning systems, and it is clear that rulers and mercantile communities had to choose between the Jesuit network and the observer networks conceived by Hart and Batlle. That is what happened in China and the Philippines with the establishment of the CCMS and the Meteorological Service, respectively. But the expansion of the Jesuit network through the national or colonial centralization of meteorological observations was not a causal effect in light of the expansion of the telegraphic infrastructures that existed at the time. This process can only be understood in terms of the coextension of knowledge networks and technological networks in which both were mutually beneficial. Telegraph stations—like semaphore and customs houses—often housed meteorological instruments and observers, making them nodes in both networks.Footnote 104 Telegraph authorities introduced specific telegraph codes to transmit meteorological data.Footnote 105 In turn, the Jesuit knowledge network provided its own conceptual framework, standardized methods, and system of observation. Such bidirectional interaction redefines the origins of meteorological services and hence reformulates the relationships between knowledge networks and technological infrastructures in globalizing enterprises. Cyclone forecasting was less a result of an efficient telegraph infrastructure than a fruit of epistemic networks interacting with it.
Acknowledgements
I would like to thank the librarians and archivists of the Ateneo de Manila University and Observatory, the Bibliotheca Zi-ka-wei in Shanghai, the University and Observatory of Hong Kong, the Archivio della Pontificia Università Gregoriana and the Archivum Romanum Societatis Iesu in Rome for securing copies of materials.
Financial support
Research for this article was supported by the Basque government’s funding (IT1441–2022) and the Ministry of Science, Innovation and Universities, Spain (PID2023-147611NB-100), financed by MICIU/AEI/10.13039/501100011033/ and by FEDER/UE.
Competing interests
The author declares none.
Aitor Anduaga is Ikerbasque Research Professor at the Department of Contemporary History and the Basque Museum of the History of Medicine and Science at the University of the Basque Country, 48940 Leioa, Biscay, Spain. He works on the history of modern physics (in particular, geophysical sciences), sociology of knowledge, colonial science, and science and religion. He has published several books with Oxford University Press and Routledge and nearly a hundred articles.
Open access
