Given humans’ ubiquitous desire to know the future, modeling and simulation have arisen as powerful tools for the job. However, the scientific and political aspects of their outcomes—prediction and forecast—can be the target of harsh criticism and dispute. This essay examines recent controversies in the simulation of both seismology and pandemic epidemiology in Japan and elsewhere. We find that disputes over different modalities of perception, as in the intriguing issue of imaging possible alternative worlds versus the singularity of the existing world, may date back to 17th-century philosophy.
In the 1980s, one of the authors conducted field research in Java, Indonesia, on a religious sect headed by a spirit medium who allegedly could communicate with spirits of mythic figures. We visited one of the sacred places in Central Java, where the medium was to serve as an oracle for the guardian spirit of Java for the coming year. Arriving at the village near the place, we were deeply disappointed to hear that we had missed seeing President Suharto and his small company. They had just left the place, allegedly having listened to a similarly high-status spirit through the oracle, probably about the prospects for national politics (Fukushima, 2002).
Our irrepressible desire to know the future is all but universal, and analyses of how people construct knowledge about the future are centrally situated among widely diverse fields, ranging from the anthropology of religion to studies regarding “contested futures” in STS. Against this intellectual background, our research group has published an edited book, Forecasting and Society: How Scientific Narratives Construct Society, a collection of conducted comparative studies of future-oriented scientific discourse, such as prediction and forecasting in diverse fields of science and technology (Yamaguchi & Fukushima, 2019).
Among our topics, seismological prediction (jishin-yochi, in Japanese) has been one of our priorities, given its integral constitution as a complex entanglement of science and policy. Both policymakers and the public in Japan have high expectations for precise predictions of when, where, and how large the impending earthquake will be. In fact, legislation has long been approved for a public action plan when large earthquakes happen (cf. Tomari, 2015).1
Such high expectations, however, have met the reality of actual seismological limitations, which fall far short of providing such a high-precision prediction; all they can provide is an imprecise long-term forecast for earthquakes in specific areas, based on a historiographical analysis of past cases (Suzuki & Koketsu, 2019). In fact, seismologists in Japan have carefully avoided using the term yochi (prediction) among themselves, instead favoring yosoku (forecast), which has a subtly milder connotation; however, such a difference is hardly perceptible to laypeople. The legislation mentioned above was approved specifically on the assumption that scientists would provide precise predictions. It was only in 2017 that the law was changed, partly because seismologists failed to predict the huge earthquake ten years ago in northeast Japan.
Pandemic simulation and its relationship to policy seem to exhibit some different characteristics from simulations for seismology. One of the two authors has long been interested in social simulation, which gives rise to notable differences in policy among countries, eventually leading our attention to pandemic modeling and simulation as a concrete subject of study. This already happened before the global catastrophe caused by the present COVID-19 pandemic.
One of the major characteristics of Japan’s pandemic simulation is that it has had virtually no place in policymaking, in sharp contrast to the ongoing enthusiasm for seismological prediction. Although concerned specialists regard pandemic simulation as a highly useful instrument for understanding both the expansion of infectious diseases and their prevention, the number of such specialists has been considerably small to be visible to policymakers and, consequently, of little concern.
As we faced some trouble in finding a proper example of pandemic simulation being used for policy, we extended our search to Taiwan, which had experienced failed policy on the SARS pandemic, from which, ironically in the end, they gained global recognition for their success in controlling the current coronavirus. Eventually, we found that policymakers their regard the use of pandemic simulation considerably positively, with various instances that foreshadowed the coming confusion manifest in countries’ policy processes at present (Hibino, 2019).
Japan’s management of the present situation has exhibited an intriguing contrast with the Taiwanese case mentioned above. After an initial set of blunders in the case of a cruise-ship infection, Japan appeared to succeed in curbing the expansion of the pandemic until mid-March 2020. Subsequently, in late March, ominous signs of its explosion led to heated disputes in various fields on the proper prevention of viruses. Consequently, the government declared an emergency and asked for an “80% reduction in human interaction,” a number derived from a pandemic simulation by Professor Hiroshi Nishiura, an authority among Japan’s mathematical epidemiologists and a core member of the newly established Action Committee for the Pandemic Cluster in the Ministry of Health. Nishiura even issued a personal message outlining a possible scenario for its expansion: “If no measures are taken, like reducing inter-human interaction, the number of seriously ill patients may reach about 850,000, half of whom will die.” This statement worked to inspire public fear.
In June 2020, when the expansion of the infection seemed to have slowed temporarily, and public opinion appeared to settle down a little, criticism of the foregoing policy measures as excessive rose sharply as the mass media collectively bashed Nishiura, mocking him as “Mr. 80%” by poking fun at his earlier forecast. The reality, however, is not that we succeeded in controlling the pandemic; just as in other parts of the world, we have been hit by second and third waves, which ironically rehabilitated the honors of both Nishiura and his simulation practices.
In terms of policy intervention, one of the visible contrasts between the prediction (or forecasting) of earthquakes and of pandemics is that the earthquake we are concerned with tends to be a massive, single event wherein policy intervention is largely confined to two periods: efforts aimed at disaster prevention beforehand and post-disaster reconstruction from the damage. Conversely, pandemics must be dealt with differently because political decisions have to be made right in the middle of the spread of the disease, and the event itself lasts longer. It follows naturally that the mode of interaction between science and policy may reveal considerable differences as well.
Intriguingly, although our concern has been centered on constructing the future, we realize that most of the criticism against meandering pandemic policies often targeted the past, assuming that an untraveled better past has been unrealized because of faulty policy intervention. In fact, as with the criticisms of Nishiura mentioned above, critics seem to claim that measures had not been needed, as if to say that a better world could have been achieved without such measures and that the critics indeed know what it would have been.
We wonder, however, whether we can be reasonably sure of this alternative world wherein allegedly better policies were carried out. Such questions bring to mind Sliding Doors (1997), a fascinating movie directed by Peter Howitt, in which the main character, Helen (actor Gwyneth Paltrow), fails to slide into the closing door of a train in the tube in London in one of the two different worlds. In the other, she does succeed in jumping into the train. This results in two different, but similarly gloomy, consequences for her relationship with her boyfriend.
Commonsensically, we think of the world we have already experienced as unchangeable and the future as being at least somewhat dependent on our choices. However, the power of scientific forecasting makes our future look like a world of necessity, and hence our effort in our edited book referred to above (Yamaguchi & Fukushima, 2019) to deconstruct such a view to leave room for the human will.3 Meanwhile, the rampant criticisms of ongoing pandemic policies—often with rhetoric indicating that things could have been better—remind us of our desperate wish to change even the past or at least to see the other world where we could have slid through the closing door of the tube train.
Obviously, there is no way of conducting a controlled experiment in the real world: at best, either we implement contrasting policies one after another and learn from their consequences, or we scrutinize the outcome of similar policies carried out in other places. Either way, however, things are far from being controlled in terms of ideal procedures in laboratory sciences. Hence, we are obliged to be patient, as the knowledge produced by such a social experiment is fundamentally limited. However, we seem to easily forget such constraints, probably because we are constantly driven to dream of a better possible world, as seen in the harsh criticisms of actual policies in response to either earthquakes or pandemics.
In this context, strikingly instructive is Stewart (2006) on the encounter between Baruch Spinoza and Gottfried Leibnitz in his biographical work that examines their intersecting lives. The pith of this book deals with how Leibnitz tried desperately to attenuate or eventually to annul the destructive impact of the idea of the singular world of necessity advocated by Spinoza, by creating the concept of multiple worlds of possibility. Ultimately, this concept was introduced to save the role of God, who decides upon the best among these possible worlds (cf. Ueno, 2013).
We vaguely understand, in theory at least, what Spinoza insists upon—the need for patience to understand this singular world of necessity owing to our lack of knowledge. However, it is paradoxical and somewhat amusing that we also share the wish to have a glimpse of, or even to jump to, the alternative possible worlds that Leibnitz mysteriously counterposes.4 At the end of his book, Stewart (2006) refers to Spinoza as the first modern philosopher who thought the world as rigorously singular, whereas Leibnitz is the first modern person with a constant craving for possible better worlds. In this sense, we are all descendants of both these ancestors.
The ongoing situation created by the pandemic is a good laboratory for observing the rapid oscillation, in a matter of a few months or even weeks, between two different ideas about the modality of the world(s). It is like the tropical Wayang theater where the shadow pictures of two modalities, both of which reside in ourselves, are endlessly struggling in a manner quaintly reminiscent of the ancient Javanese philosophy (Matsumoto, 1981).
1 It is called the Act on Special Measures Concerning Countermeasures for Large-Scale Earthquakes, 1978.
2 Fukushima (2019) is an experimental essay on the book.
3 Due to space limit, we leave undiscussed the question of how our stance in the book on the issue is related to the argument in the following section.
4 Stewart (2006) notes, however, that their relationship is fundamentally asymmetrical: that is, Leibnitz worked in the shadow of Spinoza’s influence, the former both co-opting and resisting the latter, not vice versa.
Fukushima M (2002) The Religion and Politics of Java: An Ethnographic Memoir of Indonesia under Suharto’s New Order. Tokyo: Hituzi Shobo. (in Japanese).
Fukushima M (2019) A Future Far Away: Forecasting and Society https://www.researchgate.net/publication/335013117_A_Future_Far_Away_An_Essay_on_Forecasting_and_Society
Hibino A (2019) The Ecology of Models in Pandemic Simulation. In: Yamaguchi T & Fukushima M (eds): pp. 113-139 (in Japanese).
Stewart, M (2006) The Courtier and the Heretic: Leibnitz, Spinoza, and the Fate of God in the Modern World. New York: W. W. Norton & Company
Suzuki M & Koketsu K (2019) The Problem of Forecasting Based upon the Past: The Probabilistic Seismic Hazard Maps for Japan. In: Yamaguchi T & Fukushima M (eds): pp. 173-192(in Japanese).
Tomari, J. (2015) 130 Years of the Research on Earthquake Prediction: Form Meiji to The East Japan Earthquake. Tokyo: The Tokyo University Press. (in Japanese).
Ueno O (2013) The Wonderland of Philosophers: On the Seventeenth Century of Modality. Kodansha (in Japanese).
Yamaguchi T & Fukushima M (eds) (2019) Forecasting and Society: How Scientific Narratives Construct Society. Tokyo: The University of Tokyo Press. (in Japanese).