Risk, ritual and health responsibilisation: Japan’s ‘safety blanket’ of surgical face mask-wearing

Authors


Adam Burgess, School of Social Policy, Sociology and Social Research, University of Kent, Cornwallis NE, Canterbury, Kent, CT2 7NF.
e-mail: a.burgess@kent.ac.uk

Abstract

This article begins to develop an understanding of surgical mask-wearing in Japan, now a routine practice against a range of health threats. Their usage and associated meanings are explored through surveys conducted in Tokyo with both mask wearers and non-mask wearers. It contests commonly held cultural views of the practice as a fixed and distinctively Japanese collective courtesy to others. A historical analysis suggests that an originally collective, targeted and science-based response to public health threats has dispersed into a generalised practice lacking a clear end or purpose. Developed as part of the biomedical response to the Spanish flu of 1919, the practice resonated with folk assumptions as making a barrier between purity and pollution. But mask-wearing became socially embedded as a general protective practice only from the 1990s through a combination of commercial, corporate and political pressures that responsibilised individual health protection. These developments are usefully understood amidst the uncertainty created by Japan’s ‘second modernity’ and the fracturing of her post-war order. Mask-wearing is only one form of a wider culture of risk; a self-protective risk ritual rather than a selfless collective practice.

Japan’s (all-purpose) ‘safety blanket’?

In the aftermath of the 2011 tsunami the world witnessed the widespread wearing of surgical face masks in Japan, associated with health protection against flu but here employed in the debris against the threat of radiation from the Fukushima reactor. These masks were central to the regional alarm about nuclear risk, as children were made to wear them elsewhere in East Asia amidst concern about nuclear fallout (Kim 2011). Subsequently there has been widespread curiosity about the practice, as it has emerged that masks were already being commonly worn in Japan. Numerous websites initiated debate about ‘Japan’s love affair with the surgical mask’ (Glionna 2011). In foreign eyes, mask-wearing appears to be a unique and curious practice particular to Japan. Its only well known wearer in the developed world was the late Michael Jackson, identifying the practice with eccentric hypochondria.

Observers have pointed out that while it is a more common sight in several East Asian societies than elsewhere in the world, mask-wearing is particularly widespread and normalised in Japan (see Depleted Cranium 2011, Glionna 2011), where masks are commonly worn outside medical or industrial settings. Even before the 2009 swine flu alarm, 72 per cent of manufactured Japanese masks were used at home, compared with 16 per cent in medical applications and 12 per cent in industrial settings (Ohmura 2008). In pilot interviews we conducted with 36 Japanese students studying in the UK, all claimed to use them in Japan – most less than 10 times a year, but some more than 50 times per year. None of these young people expressed criticism towards, or curiosity about, the practice.

Masks are also commonly worn in other East Asian countries such as Thailand and Vietnam, where they have become embedded in everyday practice, even beyond any kind of purpose for protecting health. Japanese women in our surveys reported how, ironically, some Asian women wear them to avoid getting a suntan in an effort to appear more western. They also reported wearing them simply to cover their faces on public transport when they hadn’t had time to put on their make-up. Despite its integration into everyday routine, the covering of the face in this way has significant implications for social interaction given the centrality of bodily perception (Merleau-Ponty 1962). Masks preclude the ‘performance of face work’ that Goffman (1967: 31) saw in his seminal paper as demonstrating ‘willingness to abide by the ground rules of social interaction’. Further, masks are uncomfortable and impractical; restricting breathing and clarity of speech and steaming up the glasses of their wearers. While the pilot interviewees cursorily indicated there were some social circumstances where mask-wearing may be problematic, they did not question what has become a routine practice. That masks are worn despite this discomfort suggests it is done in accordance with powerfully entrenched cultural ideas as well as specific pressures and influences.

Do masks work? Their widespread usage is clearly not driven by evidence of universal effectiveness. While uncertainty remains, an international consensus recognises only some possible effectiveness in reducing disease transmission in healthcare settings (Center for Disease Control and Prevention 2009, Cowling et al. 2010, World Health Organization 2009). While it can be intuitively imagined that filtering breath in this way may help to reduce infection transmission, it is actually touch that is the more important factor. It may be that insofar as masks make it harder to rub noses and thereby to spread an infection to the hands that they have some effectiveness. It is more likely that simply washing hands is more effective in relation to an actual disease threat. Mask-wearing affirms that social responses to disease are rarely driven by scientific evidence alone; as we discuss below, historically, symbolic dimensions can be more important (Alcabes 2009, Tomes 2010).

Like other forms of risk protection, masks are not unconditionally useful and can have the opposite effect to the one intended. Unless there is a complete seal around the mask it is useless against the penetration of a virus, making any protective function illusory. Further, they are uncomfortable and likely to be frequently adjusted by the hands without the wearer even noticing. What’s more, the area covered by the mask can become wet with saliva and the mask itself may become moist, encouraging bacteria. This is not to say that the less than ideal, even counterproductive nature of mask-wearing is unique. Covering the mouth with the hand when sneezing or coughing, as it is polite to do in western countries, is also likely to spread germs rather than contain them. What is defined as healthy behaviour is rarely consistent, being defined and constituted by a wide range of beliefs, pressures and preferences at any one time.

As yet there is no research on contemporary mask wearing, so reflection on this topic is restricted. Numerous explanations have emerged in web discussions, such as the Answer Bank (2011) (‘Japanese are more health conscious and considerate’), Digital Spy (2011) (‘they think other people are more important than themselves’) and akaQA (2011) (they’re polite and ‘they feel it is honorable to protect others from themselves’). Differing accounts share an understanding of mask-wearing as evidence of a relatively fixed, culturally determined disposition; the focus being upon mask-wearing as a form of social etiquette representing courtesy to others when one is feeling sick. Yet there is evidence that patterns of mask-wearing have changed, calling into question any understanding of the practice in fixed, cultural terms. Both historical patterns and more contemporary pressures need to be analysed to make better sense of changing patterns of mask usage in Japan.

The historical emergence of public surgical mask wearing

Mask-wearing is essentialised as a curiously Japanese practice in popular accounts, consistent with the ‘Orientalist’ characterisation of Japanese society by its ‘groupism’ and ‘collectivism’ that lacks empirical foundation (Suzuki et al. 2010, Takano and Osaka 1999). Historical analysis reveals a practice that has some resonance with traditional Japanese folk wisdom but only became embedded in the 1990s, after the bursting of the bubble economy and the intensified social uncertainty that followed.

A study of public reactions during cholera outbreaks prior to 1919 finds no reference to mask-wearing (Kobayashi 2001).There had been established rituals for flu-like illness, however, including the practices of paper stickers (harigami), faith healing, and naming flu in a humanised way (Sakai 2008). It was at the end of the 19th century that masks began to be used in hospitals to prevent the spread of infection (Spooner 1967). More generalised wearing by the public outside medical settings was initiated in the USA during the global Spanish flu of 1918. Several American cities legalised mask-wearing for all citizens in public places and it was recommended in the UK (Crosby 2003, Loeb 2005, Tomes 2010). As Tomes (2010: 48) describes: ‘Faced with a deadly “crowd” disease, public health authorities tried to implement social-distancing measures at an unprecedented level of intensity’. Public-gathering bans were suspended on the condition that masks were then worn instead. Social discipline was not only imposed from above, however. According to Tomes (2010: 50):

Mask-wearing gained considerable popularity as an emblem of public spiritedness and discipline. Newspapers carried instructions on how to make and launder them; women volunteers already accustomed to rolling bandages and knitting socks for the soldiers added making gauze masks to their list of good deeds.

Emphasising the collective character of response, transgression even by officials themselves, was duly punished: one Commissioner of Health was fined for not wearing his mask at a boxing match, for example (Tomes 2010: 51). Some refused to wear them on libertarian grounds and there were critics such as the Detroit health commissioner who dismissed their effectiveness. Yet the evidence-based utility of masks was secondary to their social function, as was made clear in a 1919 publication, Lessons of the Pandemic (cited in Tomes 2010: 51):

[I]f doubt arises as to the probable efficacy of measures which seem so lacking in specificity it must be remembered that it is better for the public morale to be doing something than nothing and the general health will not suffer for the additional care which is given it.

In Japan the Central Sanitary Bureau recommended mask-wearing alongside vaccination and gargling by 1919. While the practice disappeared elsewhere soon after the pandemic, it persisted in Japan. An important factor was the continued influence of pre-modern medical practices against which mask-wearing appeared progressive and worthy of promotion by those intent on driving them out. Yet mask-wearing could also be incorporated into an existing symbolic structure of folk pathology (Rice and Palmer 1993: 403). Despite its biomedical origins, the barrier of the mask was compatible with the traditional symbolic order of purity and pollution and a miasmatic, rather than viral-based understanding of infection.

This is not to say that they were yet worn during the 20th century to the same extent that they are today. Although no reliable data exist, it appears to have remained primarily one of a range of minority alternative treatments. Yet against this backdrop mask-wearing was able to reappear and become increasingly fixed as a familiar part of the national response to flu pandemic. An indication of this is the next references that were made between 1949 and 1950 to masks in the major newspapers, Asahi and Yomiuri, coincident with the Italian flu and then between 1966–1968 with the so-called Hong Kong flu. Significantly, masks were also employed in relation to a new problem in the late 1970s, with discussions of increased hay fever attributed to the post-war planting of cedar trees. Not only did this establish the notion of masks as an appropriate response to a threat beyond flu, they could now be, at least potentially, worn into the spring and summer pollen season. Virtually no articles on masks were published during other years in between these moments, but since the beginning of the 1990s there has been a slow but steady increase in such references. References to surgical masks in Asahi increased from one in 1992 to 98 in 2008 and from two to 105 in Yomiuri. High-profile risk events followed in quick succession as masks became news in 2003 when severe acute respiratory syndrome (SARS) broke out, and then in 2004 when an avian flu pandemic was feared. It was as if there was scarcely a chance to take off the masks between the two alarms. But it was the anticipated swine flu epidemic of 2009 that saw mask-wearing really become a focus for public discussion. Asahi published 1162 articles referencing masks in 2009 and Yomiuri 1282, reflecting and reinforcing the practice.

Mask-wearing was officially recommended in reports on both the 1958 and 1968–1969 flu outbreaks but it was the internationally feared avian flu of 2004 that triggered a more concerted official response. A series of guidelines published since 2005 by the Ministry of Health, Labour and Welfare (MHLW) recommended their use by infectious individuals and key public sector workers. There was a wider process of amplification that closely bound the threat of flu pandemic to wearing a mask in a context where, as in other countries, the threat persisted without clarification over a prolonged period (Alcabes 2009, Pidgeon et al. 2003). Further, the small number of casualties did not temper concern but, in the context of heightened anxiety, served both as a focus and opportunity for masks to become associated. On 18 May 2009 the first cases of swine flu were confirmed in Kobe, where three children were infected. Television coverage showed their school’s headmaster talking to reporters wearing a mask and the reporters then broadcasting from outside the hospital, also now masked. On the following day train operators instructed their workers to wear masks while on duty and asked passengers to do the same. Other businesses, such as restaurants and banks, encouraged their workers to serve customers wearing masks as they became available in high street shops and children were instructed to wear them at school.

Sales increased enormously; after May by almost 50-fold at Sugi pharmacy stores, for example (Nikkei Weekly 2009), as masks became ‘de rigueur’ (Reynolds 2009). It became a social norm to wear masks in public, especially on commuter trains; in a shift of bodily perception non-wearers becoming an object of attention (Merleau-Ponty 1962). In retrospect, May 2009 was the tipping point when masks both helped stimulate a degree of panic and established their role at the frontline of protection.

Workplace developments were central to the 2009 normalisation of masks. The confidence of employers in so readily demanding that employees cover their faces, and the employees’ resignation in complying are rooted in the post-war Japanese corporate social bargain around which society was structured. Japanese workers were ‘guaranteed a stable life course as long as they are employed as full-time workers by corporations’. Workers were given family and housing allowances, among other benefits of a company-centred society where ‘welfare by private corporations was a substitute for public welfare’ (Suzuki et al. 2010: 519). One distinctive process is unusually extensive company specific in-house training designed to ‘encourage an employee’s emotional attachment and loyalty to his or her company’. In exchange, the individual accepted an unusual degree of commitment to the corporation, most classically in the expectation to unquestioningly devote extremely long hours to work. That bargain is now fundamentally fractured, broken by prolonged recession in the 1990s and leading to a ‘second modernity’ of greater insecurity and individualisation (Suzuki et al. 2010). Yet the pressure of obligation upon employees remains and can even be more intense, given the greater insecurity of the labour market. Under these circumstances even those reluctant to don masks felt compelled to do so, as one employee complained:

My company is forcing everyone to wear them all day from the moment one leaves the house to when one gets back home. It feels like I’m suffocating, all the while keeping a microbe-culture adhered to my face and feeding it warmth and moisture with my own breath, so naturally, I hate it. A lot. But when my pay’s on the line, I have no choice but to comply . . . at least while the bosses are looking. (Japan Probe 2009)

Around 12 per cent of 58,000 respondents in a survey that was held after the anxiety had receded in September 2009 reported regular wearing of a mask, while half mentioned hand-washing and gargling (Sankei et al. 2009). The figure of around 10 per cent of regular users – with more during winter – accords with the impression gleaned from everyday encounters, at least before 2009. It was, however, unclear whether this was a distinct minority of wearers set against a majority of non-wearers, and this was one reason for conducting the surveys reported below. The more recent association of mask-wearing with the quite different threat of radiation may now further consolidate the practice as an all-purpose personal protection device.

Employer instructions and the media amplification of the desirability of mask-wearing was reinforced by commercial pressures that reflected and reinforced the masks’ more individuated role as personal protector. From the early 2000s the gauze mask was replaced by the nonwoven mask which now comprises 90 per cent of the market (Ohmura 2008). According to its developer, Unicharm (2008), it arose in response to the ‘heightened health-care awareness of people to protect one’s health by oneself’. Their advertising campaign of ‘self-healthcare promotion’ was fronted by a cartoon character called Super Three-Dimensional Mask Man, who preaches the right self-healthcare method for the winter. The characters’ opponent, Dr Sukimar, introduces interesting symbolism, reflecting how the fight against infectious disease can be represented as a national mission (Martin 1990). His name denotes a narrow gap which, at one level, simply describes the perfect fitting of the mask. But Unicharm’s campaigning slogan ‘Save Porous Japan!’ (sukima darake no nihon o sukue!) suggests that the bodily boundary is also a national one requiring each individual’s vigilance.

Significantly, Unicharm launched new campaigns after the epidemic, from September 2009, now part of the goal of a wider promotion of personal hygiene. These campaigns also built upon national themes, extending the association of mask-wearing to the family and economy. One of the two poster advertisements was called ‘Family masks of Japan’ and showed a three-generational family where mask-wearing embodies care and love between its members. The other campaign translates as ‘Virus Guard for those who cannot have a rest’ and shows people in different occupations all wearing masks on a commuting train, symbolising the sense of responsibility and duty of workers towards their job. By guarding their bodily boundaries these workers ensure that their contributions to the economy are not interrupted by catching flu. In short, a good Japanese worker and family member wears a mask.

Unicharm’s campaign shows how the character of mask-wearing is now radically different from that in the past, separate from the threat of flu pandemic. The practice now appears to be shaped by a focus on taking responsibility for personal and family health, a task demanding perpetual engagement (Uesugi 2008). But it is important to recognise a larger imperative here, that constitutes the fourth actor promoting contemporary mask wearing. State promotion of a more individually oriented lifestyle health risk agenda began in the 1990s with, for example, the healthy city movement (Nam et al. 2006). By 2000 the government launched its third initiative for health promotion ‘Healthy Japan 21’, emphasising individual responsibility for healthy lifestyles. The Health Promotion Act of 2002 asserts that the active monitoring of health and engagement in health promotion are the duty of each citizen. Legally based state responsibilisation of individuals has extended further with the anti-obesity ‘Metabo Law’ which stipulates maximum waist sizes, with those over 40 subject to annual measurement (Oda 2010). Back to the threat of flu, the 2009 campaign poster of the MHLW insists that ‘The spread of the influenza must be prevented by each individual!’

Overall, it is difficult to distinguish the relative importance of the different actors encouraging contemporary mask wearing and their coincidence with moments of epidemic alarm. It is arguably their very combination and interaction that helps account for the consolidation of the practice. What unites and shapes them, and also creates an environment of individual susceptibility to them, is the intensified uncertainty of the 1990s, rooted in economic stagnation and the breakdown of the post-war Japanese order. Suzuki et al. (2010) describe Japan’s consequent second modernity when key social institutions of the family and company lost their stabilising function (Suzuki et al 2010). This notion of Beck (1992, Beck and Beck-Gernsheim 2002) is related to reflexive modernisation and refers to the erosion of institutions and assumptions of classical modernity and their capacity to absorb risk on behalf of individuals: instead they intensify the now more individualised sense of mistrust and risk.

There has been a dramatic decline of trust in public institutions from the 1990s, leading to an increasingly self-protective tendency amongst the Japanese (Kikuchi 2008, Nomura Research Institute 2001). Hook and Takeda argue (2007: 109) that ‘economic stagnation of the 1990s and early 2000s has shaped profoundly the discourse of self-responsibility in terms of how individual citizens organize their everyday lives’. In the terms of Beck and Beck-Gernsheim (2002), the individual has been left to negotiate greater uncertainty, at least partially retreating back into more privatised concerns and anxieties, often focused upon keeping the body free of risk. The notion of self-responsibility (jikosekinin) has arisen to mediate a range of risks.

It is impossible to say whether SARS and then the successive flu epidemic threats of 2004 and 2009 alone could have stimulated routinised public mask-wearing. Yet it is difficult to imagine they could have done so alone, apart from the distinctly individuated insecurity of post 1990s Japan, the associated ideology of health responsibilisation and the pressure of the different actors instrumental to mask promotion. Mask-wearing appears as a distinctive form of health anxiety characteristic of the global risk society of which Japan became a part from the 1990s (Suzuki et al 2010).

Survey results

We carried out surveys on the streets of Tokyo about mask-wearing in January 2011. The aim was to investigate the extent of normalisation of mask-wearing and any indications that it can be regarded as a more individual, self-protective act, rather than the more selfless one suggested by cultural assumptions. In particular, we sought to establish whether there was a clear distinction between those who wear masks and those who do not and how any difference might be perceived. We were interested to see if, as in our first pilot, the practice was uncritically regarded as just a part of everyday routine, even when respondents were invited to reflect, tending to affirm its ritualistic character. We anticipated that uncertainty about the practice would be minimal if, indeed, it had become a ritual. We also hoped to gain some insight into its relationship to generalised individual risk by looking at the reasons given for wearing one, particularly whether it was a matter of self-protection against a range of threats. We were interested if any pattern emerged in relation to those who did and those who did not wear masks and whether the respondents identified a type of person who regularly wore them. In this question and in relation to recognising mask-wearing as a distinctively East Asian practice, would we see any indication that mask-wearing marked a cultural boundary? And would we find conscious awareness of direct responsibilisation and be able to identify its source?

Surveys were conducted amongst 120 passers-by, with a roughly proportionate range of ages (from those in their twenties to fifties) and gender (= 66 male and 54 female) in groups of both those wearing masks at the time (= 44) and those not (= 76). We asked:

  •  if, and how often they wore them
  •  whether they felt under any pressure to do so, and from where
  •  whether they thought they should be worn more often
  •  whether wearing one ‘told you anything about the person’
  •  whether it was ever inappropriate to wear one and where
  •  whether they were aware that masks were worn particularly in Asia
  •  whether there was anything else interesting or significant about them

In sum, 39 of our sample identified themselves as regular users, while 36 ‘rarely’ wore one. Only six people said they did not wear masks (at all). They reported a combination of different, rather arbitrary reasons: ‘too warm’, ‘irritating’, ‘haven’t done it for long time’, ‘it’s not something I do’; ‘I don’t go out often; when I do I gargle rather than wear a mask’; ‘because I didn’t wear one as a child, I don’t have the habit’. One answer stood out here, opposing the practice on principle, and saying: ‘I question the whole business of mask-wearing – nowadays the Japanese overreact’. A total of 22 respondents said they ‘always’ wore one, 19 of whom were wearing one when surveyed. There was no clear distinction in any category (wearing them ‘sometimes’, ‘often’ or ‘very often’) in the extent and purpose of usage between those asked while they were wearing a mask, and those without one on. Only those saying they always wore one indicated significant discrepancy, with half (22) of these being mask wearers, but less than a tenth (seven) of non-wearers. This suggests mask-wearing does not mark a clear internal boundary, at least as a self-conscious practice. Three of our sample overall elaborated a general discomfort, regret and even embarrassment at what they perceived as being rather curious conformity.

For most respondents, masks had lost any specific purpose of limiting a flu pandemic and had become more general protective devices. There were only two comments that reminded respondents of the more specific use and history of masks, both of which refer to the swine flu epidemic. One said, ‘I wore a mask in 2009, when we had a new influenza’, and another respondent recalled, ‘In 2009, when we had the new influenza, the atmosphere at work make me feel like I had to wear a mask’. These comments tended to affirm the importance attributed to this period in the consolidation of mask-wearing.

This seemingly collective behaviour appears individualistic in practice. Most respondents (98 of the 120) identified threats to themselves as the reason for their usage, primarily the seasonal threats of flu and pollen. This also suggests that, however routinised it has become, mask-wearing cannot function entirely independently without an ostensible risk from which it offers protection. On the other hand, the fact that it is donned in relation to mundane and everyday complaints – the common cold and hay fever – is also revealing. It is suggestive of the rhetorical implication of the risk society; one where we protect ourselves even from the ups and downs of life itself. This is underlined by some arbitrary explanations such as: ‘I don’t wear it normally, but I am wearing it because it’s windy today’.

The survey indicated a ritual that still involves perceived pressure, with 62 reporting they had been ‘told to wear one’: 20 at work, usually expressed as from ‘seniors’, 10 by doctors, five by their mother, four by their parents and four by their wife, and two by government agencies, with the remaining answers a combination of these sources as well as school, their grandmother and the family. Around a quarter of the sample also gave examples of where and when ‘it can be difficult’ to wear one, including speaking to ‘seniors’ (work ‘superiors’), at funerals and shrines, during interviews, when serving customers and in some social settings like restaurants, and while talking, swimming and having sex. The most common supplementary comment was the wish to see more user-friendly, attractive and even fashionable masks. This suggested the strength of its absorption into routine, tolerated despite its inconvenience and likely to be used even more routinely once a better design is found.

Asked whether people should wear masks more often, the predominant reaction was that individuals should remain free to choose, suggesting to us the difficulty of reconciling any collective dynamic of risk rituals with their focus upon the individual and their choices. There was, however, a mild form of wider ‘ought’, where mask-wearing constitutes a marker for a considerate person who respects etiquette. There was a general consensus that one should wear a mask in public if one has a cold or flu, especially when coughing or sneezing, and particularly in crowded spaces like a train.

Mask-wearing by healthy individuals is generally thought to be unnecessary by those wearing them less frequently, but most respondents, including less frequent wearers, still see healthy wearers as indicating a positive and ‘respectful’ attitude towards one’s health. At this level, responsibilisation seems to have been effective, and mask-wearing has become an exemplar of Japanese citizens’ legal duty. Of mask wearers, 19 agreed that wearing one did tell you something about the type of person, as did 44 non-wearers – 12 in more critical than positive terms. Of additional comments on this topic, the meaning given to mask-wearing was relatively neutral in a few cases (for example, indicating whether an individual is ‘socially oriented or not’, ‘aware of etiquette’, ‘wanting security’) and positive in 13, denoted someone ‘being considerate’, ‘who cares’, ‘who understands others’ feelings’ and who does ‘not cause trouble to others’. On the other hand, 10 respondents saw habitual mask-wearing as an indication of ‘taking things too far’, of people who ‘care too much’ or are ‘over-sensitive’, for example. There were also a few more directly moral dismissals of those who, by not wearing masks, displayed their selfish refusal to put collective good ahead of their own inconvenience, such as the comment that non-wearers are therefore ‘lazy people who don’t want to buy and put up with discomfort’. Yet these few explicitly judgemental comments tended to stand out as exceptions to a general rule of, at most, only implicit moralisation.

There were few responses to the suggestion that mask-wearing was a particularly East Asian practice. This tended to confirm that it has become an unrecognised ritual but also that it doesn’t appear to act as an, at least conscious, boundary between Japanese culture and others.

Ethical issues were considered. The nature of the study was explained to participants and their consent gained. Among other limits, there remain potential issues of translation (but we are not aware of any specific possible problems).

The study was subject to self-reported behaviour but probably not in a way that could significantly alter the results. General interviewer effects were unlikely to be powerful, given the absence of a clear agenda attributable to our questioning to which respondents might pitch their answers. Respondents may have felt restrained in expressing moral condemnation of non-wearers and non-Japanese, however. We have little idea whether answers to the more open-ended questions were in any way representative and aim to carry out further studies for this reason, including interviews and focus groups, as the public circumstances of street questioning may discourage reflective answers. It may be relevant, in this regard, that we elicited only one response suggestive of the alienation highlighted in some media coverage of mask-wearing (Parry 2011). There also remain considerable issues of interpretation, but no data are self-evident and rely upon interpretation, argument and further contestation and testing, and we intend this research to be an opening attempt to understand an interesting contemporary practice, rather than the final word.

Collective and risk ritual in individualised Japan

Mask-wearing is not only an externally imposed demand but appears to have become an embedded social practice as part of the armoury of individual responsibility for good health. An originally collective, targeted and science-based response to a public health threat has dispersed into a generalised practice lacking a clear end or purpose. Thinking further, it is useful to draw upon approaches in social risk research, the three principal ones being derived from the work of Beck, Douglas and Foucault (Zinn 2008). The Foucauldian perspective is useful in thinking how the state and corporate actors have responsibilised individuals through mask-wearing (Hook and Takeda 2007). The contours of Beck’s risk society (1992) are well known and have been applied to Japan by Suzuki et al (2010), who usefully highlight the dimension of individualisation. Greater structural insecurity creates both the problem and the opportunity, as while the individual is freer of social obstacles, this freedom has to be negotiated more individually and with less guidance from established custom and assumption. This directs us once again towards the responsibilised individual, with greater receptivity towards anxieties that threaten the self. These hazards have been democratised, in perception at least, appearing as an equal threat to all, even though, in themselves, the risks are likely to be intangible and insignificant by historical standards. Yet their uncertain character and projection into the future can make contemporary risk appear beyond resolution, necessitating continual (personal) risk management. Anxious routines such as mask-wearing make social sense in this environment, but can also further reinforce a sense of individuated insecurity.

Douglas (1990) and and Douglas and Wildavsky (1983) delineated the changing language and terms through which culturally ascribed dangers evolve, from understanding in terms of a (broken) taboo to (transgressed) sin, to the more contemporary language of (avoiding) risk. This latter transition is because of risk’s suitability to a modernity founded upon scientific, numerical authority, and because with risk, ‘the slope is tilted . . . away from protecting the community and in favour of protecting the individual’ (Douglas 1990: 7). Further, as Giddens (1991: 145) observes, the amoral language of risk accords with the substantially demoralised character of the second modernity. Maintaining good health can become a partial substitute for moral ends or at least an unquestioned imperative (Brandt and Rozin 1997). On the other hand, while the probabilistic language of risk more effectively addresses the modern individual, its relationship to establishing a clear cultural boundary is more problematic. Protection from risk is, at best, a negative good and self-protection is limited in its capacity to generate shared norms. While vigilant policing of personal health has been asserted as the duty of each individual in Japan, this is scarcely a compelling imperative and is even a jarring one in an age where individualism more generally concerns choice not duty.

For Douglas, the concern to preserve purity is at the heart of society, and nowhere is this more apposite than Japan. Maintaining purity is prior to any more specific concern with health and it is not literal; the ‘dirt’ that threatens purity is not intrinsically dirty, but is, rather, ‘matter out of place’ (Douglas 2002: 35). This definition directs us to the spatial classification important to traditional Japanese folk pathology and its ‘boundary between the “inner” sanctum of the household and the “polluted” outside’ (Palmer and Rice 1992: 323). In this context, influenza was often called kaze which means ‘wind’ but implies a foreign origin for the infection (Sakai 2008). Harigami is believed to protect the purity of the inner sanctum of the household from the invasion of foreign wind from the polluted outside. Outside space remains symbolically associated with dirt in Japan; the Japanese expression ‘people dirt’ (hitogomi) meaning a crowd of people in public. In this context: ‘it is the bodily dirt of others, perceived and expressed as “germs” that the Japanese avoid inhaling by wearing a mask’ (Ohnuki-Tierney 1984: 31). This is not to suggest there is a fixed cultural embrace of masks, or that all mask-wearing remains animated explicitly in these terms. Folk pathology in Japan, as elsewhere, remains residual. Nonetheless, such assumptions can lend weight to, and be reanimated by, other pressures such as Unicharm’s advertising campaign.

The distinctive character of mask-wearing is usefully highlighted through thinking of it as a risk ritual (Burgess et al. 2009, Moore and Burgess 2011). In a sense, they originate in the breakdown of conventional habitual practice and assumption, accompanying wider insecurity. Routines revolve around ritual and de-routinisation is at the heart of the insecurity of the risk society where self-protection predominates (Giddens 1991: 175). Traditional paraphernalia and etiquette (such as, perhaps, the handkerchief) tend to be undermined as mundane, inadequate responses to risk; even regarded as positively risky themselves. Less explored is how new kinds of rituals might evolve and be shaped in response to insecurity, rituals that might begin to construct their own routine that can help manage or displace uncertainty (Burgess et al. 2009, Moore and Burgess 2011). These risk rituals may characteristically involve new, more specialised products like the now ubiquitous hand-wash gels or, in this case, face masks.

Rituals are unquestioned social practices whose practical effects are secondary to their markedly symbolic character (Bell 2010). Risk concerns future possibilities, particularly negative ones, and their elevation in society invites the individual to manage their possible consequences. Such risks are characteristically ongoing and have no clear end, inviting continual policing. Managing them can be an aggravated process for individuals already coping with greater social uncertainty and insecurity. In turn, this invites ritualistic behaviour – albeit behaviour likely to be more disconnected from the more collective ends traditionally associated with symbolic behaviour (Bell 2010). Classical rituals are also finite and involve a transformative dimension (such as in rites of passage). Risk rituals are more open-ended; at least in one respect they could be called risk routines rather than rituals. They also tend to lack any transformative aspect. Instead they provide comfort and channel anxiety into mundane activity. We return here to the journalistic tag given to masks, the ‘comfort blanket’.

A distinction between mask-wearing and other risk rituals is that while mask-wearing channels uncertainty into a simple act, this remains a relatively inconvenient and uncomfortable one. It also differs in that the act of wearing a mask itself is not necessarily self-absorbed and distracting, the discomfort involved perhaps more reminiscent of collective self-sacrifice. As we have seen, mask-wearing can, and has, been part of a more collective public health effort where individual choice is put before the public good. Most recently, they were deployed as part of collective response in Hong Kong during the SARS crisis of 2003, in a Chinese society that has not undergone the individualisation experienced by Japan. Baehr (2008: 150–1) notes how, under these circumstances, even the form of mask-wearing appears suited to its collective rather than individual purpose:

Mask-wearing became the quickly improvised, if obligatory, social ritual; failure to don one was met with righteous indignation, a clear sign of ritual violation. The mask symbolised a rule of conduct – namely an obligation to protect the wider community and an expectation regarding how one was to be treated by others. More simply the mask was the emblematic means by which people communicated their responsibilities to the social group of which they were members. By disguising an individual’s face it gave greater salience to collective identity. By blurring social distinctions, it produced social resemblance.

As in the earlier American experience, this did not preclude, and indeed was underlined by, an occasional official transgression. Hong Kong Health Secretary Yeoh refused to wear one, explaining that the virus was transmissible only through intimate contact. Not contesting the veracity of his claim, Baehr describes his action as a ‘social gaffe of the first order’ (2008: 150–1), reflecting that it is the social, not the bio-medical, function of health rituals that is paramount in this case.

On the other hand, we might say that mask-wearing in response to SARS was a ritual only in the most general sense of the term, rather than in more clearly sociological and anthropological terms as a practice with both latent and manifest functions (Bell 2010). Masks were used as part of a wider public health strategy in a society faced with the emergency of a new disease. Their efficacy against SARS may have been open to question but this was within a range of bio-medically based contestation, indicating it had not yet passed into unquestioned practice. Their significance was not really latent, such as in how the rain dance brings a community together at a time of social strain. On this basis we also suggest that the use of masks in attempting to combat Spanish flu did not become ritualistic, and was quickly dispensed with in most countries. Only in Japan did it then take on this character, particularly from the 1990s amidst intensified uncertainty and individualisation. Rather than promoting solidarity and offering transformation – as social rituals are understood to have done in the past – mask-wearing appears to have become a permanent part of the social fabric, tenuously linking the individual to a health-focused risk society.

Final reflections

This is an initial inquiry into mask-wearing that has delineated its history and character. Further studies aim to develop a deeper sociological sense of the experience of mask-wearing and how it affects social interaction. The actual efficacy of face masks seems to be rarely considered and is largely irrelevant to a practice that has been absorbed into common practice. Masks are worn through a combination of pressures but these influences appear to be more effective in the more individualised environment of post-1990s Japan where, in addition, traditional notions of purity and danger persisted and were even reanimated. Under these circumstances masks give wearers a sense of protection independent of any knowledge of their effectiveness and, to some extent, even of the source of pollution and danger.

We argue that understanding mask-wearing in Japan as a risk ritual begins to capture the character of this social practice better than generalised cultural accounts and can be usefully counterposed to classical, collective ritual. These risk rituals are not transformative and, in this sense, are more likely to embed rather than resolve the anxieties around which they are organised, even creating a ‘spiral of anxiety’ (Crawford 2004: 505). Existential health risk anxieties are not easily managed and this also goes to the heart of the problem of responsibilisation. Individuals are made rhetorically responsible for problems over which they actually have little control, making the assumption of responsibility as illusory as the protective effects of the rituals that can accompany them.

Our findings do not, however, concur with the dramatic prognosis of masks as a means of ‘taking refuge from the world at large’ (Parry 2011: 48). Individualisation is indeed important, as such heavy-handed accounts imply, but social processes and individual behaviour within these patterns are more complex and conditional; greater individualisation does not determine such straightforward alienation, conveniently embodied by hidden faces. Individualisation does determine a greater focus upon the self-driven by a fracturing of shared norms and assumptions, but this shapes and modifies rather than determines absolute breaks in behaviour and outlooks (Beck and Beck-Gernsheim 2002). People still marry and have children in the individualised risk society, for example, arguably on a more durable, now chosen basis than in the past. Yet marriage has also lost its compelling character and become an act chosen for different, individual reasons. Likewise, in Japan, people have not completely changed their view of the world and how to survive within it, but there is a tendency towards a reordering of priorities and shift of attention. Changes can be both relatively subtle but also significant, in this case perhaps shifting from asking ‘why wear a mask?’ to ‘why not?’ A more individual act, it is not suddenly an anti-social one, but the balance has shifted. In the abstract, people endorse mask-wearing as a social act but in reality the imperative of self-protection predominates.

In focusing upon ‘exotic’ behaviour we are open to the charge of a sociology of error, where only ‘irrational’ behaviour needs explanation (Hamnett 1973). Miner (1956) highlights this well, focused upon the act of washing, deemed to be hygienic and therefore rational. In our defence, this study should be seen as one of a series into risk rituals identified with different parts of the world (Burgess et al., 2009, Moore and Burgess 2011). We have emphasised that neither the behaviour, pressures nor context are unique to Japan. Individualised health promotion and responsibilisation are part of an international trend, albeit one that takes different forms and with different targets. The process of contemporary individualisation shares characteristics seen elsewhere but takes a particular form in East Asia (Hansen and Svarverud 2010, Suzuki et al. 2010). At the same time, the implication confirms the more general analysis of Japan arriving at its second modernity and becoming a risk society. This study may contribute towards empirically fleshing out and substantiating what remain unusually arid concepts even many years after Beck first elaborated them. On the other hand, research needs to be carried out in other East Asian countries that have not experienced their second modernity, but where mask-wearing is also common. Clearly, a number of different dynamics and motives can inform the same practice of covering the face with a mask ostensibly intended to avoid the spread of flu.

Acknowledgements

The study was funded by the Sasakawa Foundation. Principal thanks go to Yuji Mizoguchi, without whom the surveys would not have happened, and who also carried out useful media research. We also thank interviewers Yukiko Shirahama, Seiji Matsui, Reia Sato, Rei Takara, Erika Takahashi and Asuka Nakane. Thanks also for comments from Peter Baehr, Tony Walter, anonymous reviewers and numerous colleagues.

Ancillary