Cosmopolitan risk community in a bowl: a case study of China’s good food movement

Abstract Ulrich Beck fundamentally transformed our way of thinking about human interdependence through his three core theses on risk, individualisation and cosmopolitanisation. However, two commonly observed deficiencies in Beck’s grand theory were its Eurocentric orientation and a lack of empirical grounding. Based on 5 focus groups and 14 interviews with participants of the emerging Clean Food Movement in China, this paper extends the Beckian discussion outside Europe. Through examining how individuals understand both ‘traditional’ and ‘new’ risks associated with contemporary food consumption, this paper demonstrates that in the face of unpredictable and incalculable harms, risks are not seen as a ‘thing’, but are translated into ‘causal relations’. Subsequently, for Chinese stakeholders, the best way to safeguard food risks is to enact more visible and functioning interdependent relations in the food system. This in turn has given rise to new forms of communities which cut across conventional geographic, socio-economic and political boundaries. The paper deepens a Beckian theorisation in two ways. First, it demonstrates that the ‘enabling’ effect of risk towards a cosmopolitan society is not limited to obvious global crises, such as climate catastrophes and financial meltdown. In fact, the mundane yet intimate concern of putting ‘good’ food in one’s dinner bowl already presses actors to form new social solidarities that are cosmopolitan in nature. Second, it goes beyond Beck’s assertion that the risk society has culminated in a cosmopolitan moment, and explores how a performative cosmopolitan community reshapes the ‘relations of definition’ to mitigate risks on the ground.


Introduction
The idea for this paper came at a lunch hour during my fieldwork on China's 'Good Food Movement' . I was studying a network of civic initiatives in major Chinese cities to combat food risks embedded in the modern food system. Initially organised by middle-class urbanites to find reliable food sources amid aggressive industrialisation and urbanisation, this movement has in effect cultivated new forms of food communities across China. There is no official name for this nationwide campaign and there is no hierarchy of command among this group of closely networked but independently run grass-roots efforts. But I dub the name, the 'Good Food (Liang-Shi) Movement' , for all of these civic initiatives subscribed to the pursuance of 'liang-shi, liang-ren, liang-xin' . Literally translated as 'good food, good people and good heart' , the shared aim for these Chinese urbanites is to improve risk resilience of the contemporary food system through invoking the public's reflections on and reorientations of power relations in the production of food. The success of these civil initiatives is not only due to their tapping into expanding social anxieties over food safety, but also owes to their openness to world experiences. As this paper later demonstrates, not only were a number of key staff in these civic groups overseas-returns, these organisations were also keen, but selective, in borrowing ideas on building alternative food systems from Southeast Asia, Europe, Latin America and North America.
At the time of that particular lunch hour, I was visiting a residential apartment-turned storage house belonging to Wuhan Natur, a highly visible organisation in promoting the Good Food Movement. As I watched volunteers and staff working together in sorting newly arrived raw vegetables into delivery baskets for their fellow community members, one of the volunteers handed me a bowl of lunch. It contained rice and a stir fry of tomatoes, tofu, greens and beansprouts. This was quality eco-friendly food that they assiduously sourced from various independent farms across China, just as the packaged products they imported from like-minded providers in neighbouring countries and from Germany. It dawned on me that in my bowl was, quite literally, produce of what Ulrich Beck would call a 'cosmopolitan risk community' .
Yet, the becoming of a cosmopolitan risk community in China was also different from what Beck has theorised. While Beck fundamentally transformed our way of thinking about human interdependence through his lifelong investigation into the nature of risk communities in a global age, he was vague in articulating how risk brought radical institutional change. This has often been attributed to his lack of empirical engagement and his Eurocentrism (Dingwall 1999;Elliott 2002;Healy 2004;Mythen 2004).
For this special issue on Beck's legacy, this paper extends the Beckian discussion outside of Europe. It is both an empirical demonstration of the cosmopolitan risk community thesis and a criticism of its oversights. Based on 5 focus groups and 14 interviews with participants of the Good Food Movement in China, I argue that China's Good Food Movement is an example of 'rooted cosmopolitanism' (Beck and Sznaider 2006) in response to risks associated with an industrialised food system. The paper deepens a Beckian theorisation in two ways. First, it critically examines the 'enabling' effect of risk towards a cosmopolitan community, which is not limited to just global crises, such as climate change and financial meltdown. In fact, the mundane yet intimate concern of putting 'good' food in one's dinner bowl already presses actors to form new social solidarities that are cosmopolitan in nature. Second, it goes beyond Beck's assertion that the risk society has culminated in a cosmopolitan moment, and explores how a performative cosmopolitan community reshapes the 'relations of definition' (Beck 1995) to mitigate risks on the ground.
The paper first charts a chronological review of how Beck has developed his risk theory, especially how he envisioned subsequent social change. This not only reveals the kernel of his intellectual pursuit and how his ideas evolved in relation to wider academic dialogue, but it also makes visible gaps and limitations. It points out that, given the performative nature of transnational solidarities (Albrow 1996), a lack of empirical engagement is especially constraining for Beck's theorisation of global risk and makes his grand theory vulnerable to conflicting claims. The paper then, in sequence, reviews the irony of a modernisation-induced food crisis in China, the research methodology employed and an in-depth discussion on how the Good Food Movement both substantiates and extends a Beckian examination of the cosmopolitan risk community.

The emergence of cosmopolitan risk communities according to Beck
Beck was arguably one of the most prolific social theorists in recent times. In his ambition to develop a grand theory for the Second Modernity, he published widely on the side effects of a globalised industrial society, drawing on examples from a wide range of contemporary concerns such as work, marriage, religion, global financial crisis and climate change. While his oeuvre is often described as consisting of three conceptual streams, namely risk society, individualisation and cosmopolitanisation, it is safe to say that the concept of risk functions as the thread which connects and syntheses Beck's diverse academic pursuits.
One fascinating aspect of Beck's work is not only his identification of modernity's self-inflicted risks per se, but how he projected social responses to these risks and how he forewarned both social scientists and regulators that cross-boundary changes were at the horizon of our socio-political landscape (Beck 1992(Beck , 183-235, 1995(Beck , 2009(Beck , 2011(Beck , 2015. After clarifying aspects of his risk theorisation that are especially pertinent to this paper, this section reviews how Beck conceptualised risk-invoked social transformations through different stages of his work.

Three characteristics of modern risk
Beck primarily saw risk as the unintended and incalculable side effects of the modern age, embedded in the ambiguities of scientific and technological knowledge (Beck 1992(Beck , 1995. In his later work, Beck further emphasised the 'non-knowing' aspect of modern risks, which includes 'provisional non-knowing, unacknowledged non-knowing, willful ignorance and … conscious and unconscious inability-to-know' (Beck 2009, 115). While some rightly questioned the empirical validity of his segregation of 'new' risks manufactured by industrial society and the 'old' dangers that are defined as identifiable and insurable, this categorisation is instrumental in highlighting a contemporary apprehension that risks are increasingly spinning out of institutional control (Beck 1992;Abbinnett 2000).
Related to this, a second emphasis of Beck's theory is the relational causation of risk (Healy 2004). Not only is risk 'manufactured' through institutionalised human actions, but the failure to respond to modern risks is an epitome of 'organised irresponsibility' (Beck 1995;Giddens 1999). In other words, for Beck, risk is not so much a 'thing' , but a situation produced (and reproduced) by a particular social configuration. As such, the meaning of risk can be 'changed, magnified, dramatised or minimised' and 'are particularly open to social definition and construction' (Beck 1992, 23; original emphasis) with the impact of risk being 'staged' and mediated (Beck 2011). The logic thus follows that relations in the identification and framing of risks become critical in shaping the resilience of a risk society (Beck 1995, 182-183).
Finally, for Beck, risk describes a modern mentality that has an omnipotent presence because one's life becomes a 'risk venture' that hinges on decisions and their unknown consequences (Beck and Beck-Gernsheim 2002, 47;Yan 2012). To be sure, none of the above features can be singled out as uniquely 'Beckian' , as these themes were expounded by a number of contemporary theorists (e.g. Douglas1992; Giddens 1999;Bauman 2000;Zinn 2008). The point is, as the next section demonstrates, that Beck's particular attentiveness to these three aspects of risk led him to his theorisation of cosmopolitan risk communities.

Risk as an enabler for social transformation
In his last book, The Metamorphosis of the World, Beck (2016) seemingly 'reversed' his earlier view that in the face of unpredictable and uninsurable risks, the once supposed positive logic of Western capitalism's distribution of 'goods' had been taken over by a negative logic of the 'social bads' . Instead, Beck discussed the 'positive side effects of the bads' (Beck 2016, 4), namely the innovative formation of new social solidarities across conventional boundaries.
In fact, this 'reversal' of analytical focus on the transformation between the 'goods' and the 'bads' is a key step in Beck's completion of the circle of change he set out in his seminal work, Risk Society, exactly 30 years prior. First published in German in 1986, Beck outlined in Risk Society three basic stages of the dis-embedding and re-embedding effects of modern risk on individuals: (1) the dislodging or 'liberation' of individuals from traditional bonds and support, which in turn generates (2) a loss of stability and security. These 'precarious freedoms' would incentivise (3) a reintegration into new social relations (Beck 1992, 127-137). Through his close collaboration with British sociologists such as Anthony Giddens and Scott Lash, Beck clarified how these risk-invoked new social bonds would work as 'antidotes' to organised irresponsibility (Beck, Giddens, and Lash 1994). Beck called for a collective reflexivity which would reconfigure existing 'relations of definition' to more democratic networks of proof, restraint, control and participation in decision-making (Beck 1995, 129-133).
At the turn of the century, Beck further expanded his theorisation on the effects of risk to the global age (Beck 1996(Beck , 2002. As international exchanges intensified, technocratic controls, such as the use of biotechnology and information technology, increasingly generated unwanted side effects that 'cannot be socially delimited in either space or time' (Beck 1996, 1). Beck (2002) argued the world risk society had culminated to a cosmopolitan moment. By 'cosmopolitan' , Beck acknowledged both the increasing fragmentation of the social (the dis-embedding effect of risk) and the potential instrumentality of a heightened awareness of global interrelatedness (the re-embedding effect of risk). What lies at the heart of a 'cosmopolitan' perspective is how reflexive dialogue with global experience internally transforms social agencies, or what Beck termed 'globalization from within, globalization internalized' (Beck 2002, 23;original emphasis). This has direct socio-political implications. By synthesising 'the co-presence, and co-existence of rival lifestyles' , the 'relations of definition' would expand beyond traditional elites and forge new 'global domestic politics' (Beck 2006, 89;Beck and Sznaider 2011, 418).
To deepen the connections between global risk and cosmopolitan solidarity, Beck returned to his original thesis on 'risk' , which was primarily 'based on casual interpretations' (Beck 1992, 23, original emphasis). Drawing on Benedict Anderson's concept of 'imagined communities' as national constructs, Beck further formulated the emergence of 'imagined cosmopolitan communities' based on the reflexive identification of the causal relations of risks (Beck 2011;Beck et al. 2013). That is, similar to how the emergence of the printing press enabled the forming of a national identity through creating a sense of shared fate, Beck argued that contemporary media's reportage of global risk, such as climate change and financial meltdown, significantly raised an everyday awareness that 'the 'global other' is in our midst' . Global risk creates the condition of 'compulsory cosmopolitanism' which 'glues' various actors into collective action (Beck 2009). Beck briefly considered two opposite effects of modern risk: the sinister 'Schmittian' scenario in which the 'politics of fear' would heighten social conflicts and the 'Hegelian' scenario in which global risk would 'release new sources of legitimation for intensified transnational collaboration' (Beck et al. 2013, 9, 19). Yet, the 'positive side effects of the bads' led him to the optimistic conclusion that 'the young, global generation are members of Homo cosmopoliticus' (Beck 2016, 189).
This conceptual clarification of a risk-led cosmopolitan becoming is the basis for Beck's proclamation of a 'metamorphosis' (Verwandlung) of the social order. It is arguably a research agenda Beck set out for the new century. For he envisioned metamorphosis to be 'a mode of changing the mode of change' , which captures 'the transfiguration of the social and political order' that we are witnessing but 'do not have the word for' (Beck 2015(Beck , 77-78, 2016.

Limitations of Beck's theorisation
Self-inflicted risk with its associated dis-embedding and re-embedding effects forms the spine of Beck's sociological inquiries. Particular attention was given to how risk in the global age gives rise to cosmopolitan communities which radically transform the way we bond. Yet, there are also two major weaknesses to his theorisation, namely its Eurocentrism (Dingwall 1999;Scott 2000) and a lack of empirical engagement (Mythen 2004;Zhang 2015). What adds to the frustration is his 'essayistic' writing style with 'occasion hyperbole in usage of terms and concepts' (Matten 2004, 388). To be sure, one could argue that some of the empirical criticisms miss the point of Beck's theory and put an unfair burden of proof to the theorisation of 'a change in substance' which by nature is an impossible task (Latour 2003, 41;Matten 2004). In fact, Beck opened his Risk Society with the 'disclaimer' that his writing should be read as 'empirically oriented, projective social theory -without any methodological safeguards' (Beck 1992, 9, original emphasis).
However, a lack of empirical examination is especially constraining for a grand theorisation of cosmopolitan communities. Transnational solidarity is performative in nature (Albrow 1996). This is to say that cosmopolitan risk communities are not 'beings' but 'becomings'; we know their existence only through their performative effects (Kogler 2005). An absence of empirical grounding depletes the theory of analytical substance. Instead of examining cosmopolitan risk society as it is, theorisation becomes an exercise of logical patch works of the 'both/and' , which acknowledges rather than explains the conflicting nature of reality. This effectively traps Beck's ambition of developing a new sociology for the Second Modernity 'in a 'laminated' world of the 'multi-layered'' , which captures the final aggregated effect layered by a myriad of social milieus but fails to unfold the dynamic within new social space (Zhang 2015). For example, while Beck delineated a relational causation of manufactured risk, an ignorance to new solidarity performances on the ground left him vague on how risk perception was translated into action (Elliott 2002) and what actual difference 'cosmopolitan' relations made in mitigating actual risks (Antonsich 2010;Zhang 2015).
In his later work, Beck tried to build platforms to extend the application of his theory to non-European, especially East Asian experiences (Beck and Grande 2010;Beck et al. 2013). This intention was most apparent in his unfinished project European Research Council project 'Methodological Cosmopolitanism' .
To extend the legacy of Beck's work, this paper is in line with a number of small sample qualitative studies carried out in Asia with the aim of addressing the gaps in cosmopolitan theorisation (Phuong and Mol 2004;Tyfield and Urry 2009;Zhang 2010Zhang , 2012. In what follows, this paper first demonstrates how the food crisis in China is an epitome of modern risks to which the Good Food Movement sets out to address. It then examines the movement in detail and argues that to understand the enabling factor of risk in the emergence of cosmopolitan communities, one needs to adjust a Beckian focus from the 'scale' to the 'intimacy' of global risk. Relatedly, it also argues that to grasp the 'metamorphosis' of the world, one needs to question rather than take for granted what sustains a 'imagined risk community' and investigate how new 'relations of definitions' bring new social resources in risk resilience.

The irony of China's new food crisis
The industrialisation of the modern food system and its subsequent risks have been key examples in the study of risk society (Levidov and Carr 2000;Beck 2009). China's modernisation of its food system followed a Western path and resembles what George Ritzer termed 'grobalisation' (Ritzer 2007). That is, the pursuance of the 'growth' of profit leads to a global expansion of industrialised food production, which is characterised by centralisation, standardisation and a replacement of human with nonhuman technology (Ritzer 2007, 15, 20-30). Having the world's largest population to feed, the Chinese Government was keen to establish a biotechnology-and agrochemicals-assisted agro-industry so as to 'take good control of its own bowl' (i.e. securing national food security) (Dirlik 2002, 26-28;State Council 2014). This approach has gradually consolidated into what is now known as the 'Big Agriculture, Big Food' framework (MOA 2016;State Council 2016). This entails an 'integration of funding … planning, and the unification of standards, supervisions and regulations' in order to 'upgrade the whole industrial chain of agriculture' (State Council 2016, 1.1, 1.3, 1.4). This approach worked, for China's agricultural productivity has enjoyed 12 consecutive years of growth. To be sure, the modernisation of China's agriculture is still in process. The level of industrialisation in the farming and production of foods, and subsequent socio-economic impacts are often contingent on local eco-political particularities across China (Zhang and Donaldson 2010). But for the focus of this section, it is safe to say that the policy steering of modernising the food system in China mirrors global trends and sees scaled-up industrialisation as the path to good food production for its population.
Yet, the irony is that the more secure Chinese domestic food production is, the less safe its food has become. In fact, some have lamented that China has arrived at the 'Chemical Age' for food (Gong, Cui, and Wang 2012). In addition to severe water and soil contamination as a result of aggressive industrialisation, China is also the world's largest consumer of fertilisers and pesticides (Lam et al. 2013). A 2016 study estimated that around 53,300-123,000 Chinese people are poisoned by pesticide each year (Zhang et al. 2016). Furthermore, the prioritisation of economic growth has encouraged the rise of ''expert' criminals' who are adept in exploiting grey areas to improve the taste, presentation and preservations of food products (Gong, Cui, and Wang 2012;Yan 2012;Lam et al. 2013). China's food safety has also caused international concerns (Thomson and Ying 2007). One example is China's various melamine scandals, which exposed the trade secret of a number of leading milk providers in China who used chemicals, such as melamine, to artificially increase protein percentage in inferior milk (Wang 2013). The subsequent soaring demand from Chinese parents for European-produced milk powders at one time led a number of European countries to temporarily ration such sales in their supermarkets (Tsang and Lucas 2013).
Although China has a full-fledged legal regime for food safety, the enforcement of 'civil liability' for food producers remains low (Asian Development Bank 2007;Ni and Zeng 2009;Cheung 2013). In recent years, China upgraded the Food Safety Commission Office to a ministerial-level agency with a new name of China's Food and Drug Administration (CFDA) and further harmonised its Food Safety Law with international norms and standards (Wu and Zhu 2014). But there are a number of particularities associated with contemporary food risks in China that fall outside of the capacity of traditional regulation.
To start with, health damages done by adulterate or contaminated food are hard to trace and often have latent effects. It could take years for chronicle intoxication or cancer resulting from unsafe food to develop. This makes conventional legal demarcations of liability and regulatory surveillance difficult (Gong, Cui, and Wang 2012;Cheung 2013). As anthropologist Yunxiang Yan concluded in his systematic review on changing patterns of food risks in China over the last six decades, contemporary food safety in China presents 'a typical example of manufactured risk' which is 'incalculable and uninsurable, undermining regulatory governance and having far-reaching social and political implications' (Yan 2012, 715, 720).
More importantly, as Yan further noted, the most worrisome aspect of China's food safety is that one inescapably becomes the victim of 'well-organised and large-scale production and distribution' of unsafe food (Yan 2012, 716). It is extremely difficult to identify who or which group should be responsible for declining food quality. One example is that while urban workers complained about farmers' failure to produce safe food, it is often the pollutions generated by irresponsible urbanisation and industrialisation that caused severe water and soil contamination (Lam et al. 2013). Peasants sometimes were left with no choice but to grow and sell 'trash vegetables' (vegetation that has a high heavy metal content) to urbanites to earn a meagre living (Zhang and Barr 2013). One political science professor at South China Normal University summarises the gloomy picture as 'it seems the Chinese have to get used to poisoning each other' (Tang 2012). Thus, the problem is not so much on how to crackdown on individual criminal activities, but on how to mitigate unintentional side effects of aggressive economic development.
As one's dining bowl increasing became a daily reminder of the precariousness and organised irresponsibility of modern society, it has generated great social anxiety and distrust (Yan 2012, 725). As Beck rightly pointed out, 'through the complicity of policy and technology, every suspected accident becomes a political scandal' (Beck 1995, 161). Similar to many other countries, the profit-driven industrialisation of the food system in China has caused a decline in confidence in government and in professionals (Portinga and Pidgeon 2003;Cebulla 2007). Studies on China's food safety commonly acknowledged that the remedy for this situation calls for a meaningful power sharing supervision system with wider public involvement (Thomson and Ying 2007;Ni and Zeng 2009;Gong, Cui, and Wang 2012;Jia and Jukes 2013). In other words, to truly address risks in the food system, China needs to reform not only its relations of production but also its 'relations of definition' .
In short, the irony of China's new food crisis is that it is a testimony to the triumphs of China's modernisation. In response to growing social anxiety, in recent years, grass-roots explorations of reliable food production started to appear in some Chinese cities. Commonly set out as consumer-sponsored farmers markets, a handful of key groups soon shifted their attention to reorient consumer-producer relations. As a highly vocal activist asserted, 'one could say that we are promoting a social movement' (Interviewee 8). Activists have gone beyond promoting quality production, and have confronted the relational cause of modern food risks. As the following sections demonstrate, these amount to what I call the Good Food Movement which has effectively cultivated the becoming of a cosmopolitan risk community.

Methodology
I first came into contact with the Good Food Movement at the end of 2011. While the focus of this paper is not on state-society relations but on examining the cosmopolitan risk community the Good Food Movement resembles, it is worth pointing out that the Chinese Government is wary of societal organisations' political influence. Consequently, given the high registration standards for non-governmental organisations, it is not uncommon for many organizations to register as social enterprise or to remain unregistered (Zhang and Barr 2013). Thus, the exact data of the Movement are impossible to obtain. On basis of published records (Zuo 2015), fieldwork data and personal contacts in this area, the best estimation is that to date similar type groups have spread to around 20 major cities in China. Fieldwork took place between July 2014 and May 2016 and concentrated on three civil organisations: Beijing Country Fair, Xi'an Farmers' Market and Wuhan Natur.
These three sites were chosen for their complimentary insights on the Good Food Movement. As the national capital, Beijing is a well-resourced and highly globalised north-eastern city. Wuhan is the moderately well-off southern provincial capital of Hubei and is a main hub for agro-industry research. Xi'an is the capital of north-western Shaanxi province, which is less innovation-driven and financially less advantaged than Beijing or Wuhan. These differences led to different complimentary organisation profiles. Beijing Country Fair is the one of the most vocal and influential civil organisations in advocating an alternative modernisation path for China's agriculture (He 2015). Being closely engaged with transnational academic and policy debates, Beijing Country Fair is arguably the 'idea hub' for this campaign. In fact, Xi'an Farmers' Market, begun in 2011, is one of the many organisations that was initially inspired by the Beijing model. Xi'an offers a good example of how to translate national and international experience to a relatively conservative local culture. Wuhan Natur, established in 2008, is one of the first civic initiatives to seek reliable food production and is perhaps the most successful in building a cohesive food community. Each group has an estimate of 500 to 2,000 subscribers on their mailing list, with Beijing the largest and Xi'an being the smallest and youngest.
The data-set consists of 14 one-hour semi-structured interviews with key activists and 5 90-minute focus groups with interested publics (2 in Beijing, 2 in Wuhan and 1 in Xi'an). Focus groups have a particular advantage in studying collective efforts, for they allow interaction among participants, which provides the opportunity to observe how collective meaning is negotiated (Morgan 1996). I also choose group discussions as participants potentially 'enjoy the opportunity of taking on the teaching role' as they 'individually and collectively explain their situation' and elaborate tacit knowledge that may otherwise be difficult to obtain (Gaskell 2000, 51).
Calls for focus group participants were sent through the three organisations' online social media (e.g. Chinese versions of Twitter and Whatsapp) and emailing list. A key concern for selection was to balance between having sufficient common concerns and having diverse opinions (Olofsson and Ohman 2007). Thus, only people who were responsible for family food shopping were invited to apply. Interested individuals were asked to submit information on their age, gender and how long they had known their local Good Food Movement organisation. Based on these data, invitations were then sent out to create mixed groups. The resulting 5 focus groups consisted of 6-9 people each. In total, 37 people from18 different professions participated. The average age for participants was 39.5, with the youngest being 23 and the oldest 63. Overall the groups had a 8:3 female to male gender ratio. This is not surprising, as in China it is mostly women who manage family meals and all three organisations claim to have a higher appeal among women. Semi-structured discussions focused on what qualified as 'good food' , perceptions of food safety and the roles of consumers, producers and technology in food production.
Interviews were conducted to understand the rationales and objectives of the Good Food Movement. Interviewees were identified through a snowballing sampling. Written consent was obtained from both interviewees and focus group participants. All data were recorded and transcribed. Transcripts were sorted into themes according to reoccurring concepts (Corbin and Strauss 1990). Coding was subsequently developed to further chart findings into analytical focus of this paper (Maxwell 2005, 97).

The good food movement as a cosmopolitan risk community
The Good Food Movement is an illustrative example of 'globalization internalized' (Beck 2002), for its development was fed by synthesising a range of relevant international experiences. Many interviewees recounted that practices, such as community support agriculture (CSA) in North America, the participatory guarantee system (PGS) in Europe and the idea of solidarity economy from Latin America, gave them initial inspiration on how to reform the food system (Interviewees 1,9,8,14).
Sustaining an exposure to global alternatives was vital in the success of these organisations. A number of key organisers of these local initiatives were once Chinese-expats in Europe or North America. Importantly, all three organisations host and participate in lively transnational debates. This is especially apparent in Beijing, where Country Fair frequently invites international academics and practitioners for public seminars. It also sends staff to short-term training and exchange programmes to other (mostly Asian) countries and to present at international conferences. This access to global experience is further shared domestically through inter-organisational visits, exchange and co-sponsoring of events.
What makes the Movement cosmopolitan is not only their global vision and open-mindedness, but the fact that the organisers and affiliated producers are highly reflexive and selective in what to take from these international lessons. During my fieldwork, I noticed that not only do these organisations welcome people with different farming ideals (be it permaculture or natural farming) but they also regularly vary their public engagement focus (for example, from CSA to PGS).
This 'fluidity' of their intellectual orientation is revealing. One interviewee confessed that sometimes they felt they were in a race with the food industry, and they 'had to' rebrand their ideas in order to accurately communicate what they aimed to achieve.
A practical concern is that we had to continuously ditch some of the popular concepts that are 'polluted' by industrialists. Businesses are very smart in adopting the latest buzzwords, but once they adapted the PGS and CSA to their old market model, it deprived what these concepts are really suggesting … So we need to find new words to explain what we want to do clearly. (Interviewee 3) For civil activists in China, to cosmopolitanise is not just about blindly absorbing global trends, but about pragmatically screening ideas to address their particular cause. This is a highly reflexive process in that they recognise that words alone can be deceptive, and can be hijacked or 'polluted' by repackaging into the old bottles on the industrial production belt. Thus, the normativity in their practices resides not in a particular ideal, but in commitments to continuous solicitation and synthesisation of different approaches. This sometimes entails, as suggested in Interviewee 3, quite a literal search for new vocabularies from the global food community to articulate what they have set their minds to achieve.
The goal shared by all three organisations is to reset social relations within the food system and to re-establish community resilience to food-borne risks. Currently, all three organisations operate in the following three areas: (1) each civil organisation functions as an independent vetting and sales channel for ecologically farmed food; (2) they are communication platforms for consumers and producers with a specific aim of community building. Through events such as farmers markets, workshops and weekend farm visits, they open up 'intersubjective communications' by re-connecting urbanites with the food sources, and invite them to bear witness to the challenges of sustainable farming (Healy 2004, 292-293). It gives rise to ideas such as 'participation is procurement' which emphasises a contribution from both the consumers and the producers to improve risk resilience in food production; and (3) working with international food organisations and agricultural scientists, they also offer training and aid to farmers in their region. But, as one of the interviewees hastened to add, they 'are not 'purist'' (Interviewee 8). That is, they encourage discussions and experiments with diverse methods as long as they are not associated with chemical farming or food processing.
This ability to work with diverse stakeholders and to consolidate overlapping interests into community building is a significant feature of the Good Food Movement. For example, in 2016, a main organiser of the Beijing Country Fair and her Shanghai colleague went on a month-long trip to France, Germany and the US, visiting sites such as the Greenmarket in New York, the Organic Valley in Wisconsin and the farmers market in Strasburg. She said the trip was inspiring but was also affirming, for compared to a few years ago, she realised that the Good Food Movement had grown beyond its previous Western role models into something quite original: Originally we thought we may be still behind international experiences … now one starts to realise that: Hey! Actually we are doing really well in China. For example, many Western farmers' markets still function as a sales platform, but we are more than that … In addition to improving eco-friendly farming, we also consider if there is anything we need to do collectively as a community in response to local or national policies … Very few food organisations brings together consumers and producers as close as we are. [What we are doing] is small but it's original. (Interviewee 8) Thus, in addition to its reflexive internalisation of global experience, what makes the Good Food Movement cosmopolitan is its efforts in breaking down previous social boundaries, and bringing actors from diverse social spheres (e.g. consumers, producers, researchers, civic activists and policy-makers) into the same community. This is further demonstrated below.
So far this paper has demonstrated that Beck's theory has been effective in capturing the empirical contour of what is emerging in China's social response to industrial food risks. Yet, as the following two subsections demonstrate, a Beckian perspective may have overlooked the actual dynamic between risk and a cosmopolitan outlook and how cross-boundary community building can improve risk resilience on the ground.

A closer examination of risk-invoked cosmopolitan outlook
As discussed earlier, in theorising the becoming of a cosmopolitan risk community, Beck emphasised the 'scale' of contemporary risks, in both temporal and spatial terms. This is to say that as risk becomes global and as the consequences of risk can no longer be contained by national boundaries, it creates a sense of 'shared fate' which enforced a 'compulsory cosmopolitanism' , prompting social change (Beck 1996(Beck , 2011).
Yet, an examination of the Good Food Movement shows that it may not be so much the 'scale' but the 'intimacy' of risk that impelled the coming together of a cosmopolitan community. Media framings of food safety issues as nationwide problems in China can be traced to 2004 (Wang 2013). While both interviewees and focus group participants claimed that media staging of risk contributed to their early awareness of this problem, the scale of risk itself did not prompt them to seek change. Rather, it often took a 'personal' encounter to make the pervasive nature of modern risk sink in and subsequently incentivise them into action. One participant in Beijing said although she had long been sceptical of large-scale farming and commercial food production, what really registered systematic food risks with her was an incident in which a clump of food additives in her snack caused a severe throat reaction. 'It was then I realised' , she said, 'eating was no longer a matter of taking in nourishment, but of taking in less poison' (Participant 9, FG1). One consumer-turned amateur farmer in Xi'an described her participation in the Good Food Movement as quite simple and personal: 'it's a matter of self-protection and protecting one's own family' (Participant 8, FG5). A number of interviewees also echoed deep personal reasons as to what prompted them into food activism, such as having new babies in one's family or extended family (e.g. Interviewees 9, 4), personal health issues (e.g. Interviewee 5, 8, 13) and a chanced experience with farming (e.g. Interviewees 3, 6, 7, 11).
Thus, although it may not be the immediacy of identifiable danger and hazard for individuals to act, findings suggest that it is the intimacy of possible risk experienced at their personal level that urged these urbanites from different walks of life to come together to seek collective remedy. This is not a Chinese phenomena, as previous empirical studies in developed countries have shown similar results (Cebulla 2007;Olofsson and Ohman 2007). To some extent, it could be argued that it is these diverse personal connections prior to the coming together of a society that nurtured a cosmopolitan openness and the intellectual 'fluidity' of the Good Food Movement.
This acute awareness of the fragmentation of risk is also exhibited in how these civil organisations see their role in this movement. That is, just as the movement itself is a synthesis of a number of international experiences, organisers were keen to spread ideas but sceptical of a simple extension of their practice to other cities (Interviewee 3, 8, 10). The point of networking with like-minded groups and individuals was to generate mutual inspiration and facilitate 'rooted' expertise, as the founder of Wuhan Natur explained: Anyone is welcome to have a look at how we operate, but they are not obliged to adopt anything. For example, Beijing's experience may not work in Wuhan, and vice versa. Thus the most rooted experience is the best experience … Once uplifted from the root, it loses its value. (Interviewee 9) Focus group discussions echoed the importance of maintaining context-sensitivity. A number of participants in focus group 5 raised concerns that if their local food community were to be expanded into a national or regional collective, the movement itself would 'turn sour' (bianwei) for it missed the point that reliable food production necessarily needed to be kept diverse to speak to local particularities (Participants 1, 2, 3, 7, 8, FG5). One participant in Beijing drew on the Chinese saying, 'single spark can light the prairie afire' , and highlighted that 'the most important mission' of Good Food Movement lies in its offering of a flexible space of idea exchange, so as to 'make more people to see and try out alternative ways of procuring food themselves' (Participant 8, FG2). Xi'an Farmers Market is an excellent example to demonstrate the necessity of context-sensitivity. As noted earlier, it originally started as a copy-cat group of Beijing Country Fair. But soon the organisers realised that in Xi'an 'consumer expectations were different' , 'farmers were much less educated and more financially stressed' , and the local political environment was also different (Interviewee 4, 5). In other words, while the Xi'an public was also concerned with the same food adulteration and contamination, they experienced the problems quite differently.
Understanding that the 'transformative force' of risk lies in its intimacy rather than its scale is important. This leads to a critical reflection on Beck's premise on the 'risk society's cosmopolitan moment' (Beck 2009(Beck , 2011. While it is true that 'global risk' and the emergence of 'cosmopolitan communities' may coincide, the causal relation is much more nuanced and cannot be taken for granted. To be sure, in his early work, Beck argued that risk is always contextualised as a 'social self-encounter' and 'must be interpreted through the inner and personal experience of social ways of life' (Beck 1995, 7, 158). Yet, this point got lost in his later work on social transformations brought by 'global risk' . The phrase 'global risk' may be somewhat misleading at it implies a global homogenisation of risk. It must not be forgotten that the same event, be it climate change, financial meltdown or industrial food production, poses different risks to different people in different places. For social actors, it is not so much that risks are 'perceived' differently, but that they actually are different.
Contrary to Beck, knowing the scale of risk embedded in industrial society itself may not lead to action, for these 'risks are rather associated with hopelessness and being despondent' (Olofsson and Ohman 2007, 192). As my data suggest, it may take an intimate (and thus fragmented and individualised) comprehension of risk to impel collective initiatives.
A possible explanation to why the development of a cosmopolitan solidarity may not conveniently map onto a 'imagined society' as Beck speculated is that, unlike traditional identities, cross-border solidarity in a global age is by nature 'performative' (Albrow 1996;Zhang 2015). Manufactured risks and intensive globalisation have not only de-territorialised societies into 'regimes of living' that are beyond national and institutional boundaries (Lakoff and Collier 2004, 420) but also nurtured performative collectives that draw on resources from different colonised communities (Albrow 1996, 175-80). The coming together of a cosmopolitan community resembles what Isin calls 'activist citizenship' (rather than 'active citizenship'), which is not to fulfil a readily acknowledged vision (be it a national or risk imaginary) but is to author a new imagination by 'break[ing] habits and act [ing] in a way that disrupts already defined orders, practices and statuses' (Isin 2009, 384).
To push this argument further, it is not solely the presence of risk itself that holds together a trans-boundary community, but also the continuous search, innovation and sharing of solutions that sustains a collective as a community. The 'world of contingency' opened up by industrialised modernity imposes an obligation together with a newly discovered freedom for the individual to construct new forms of belongings (Thomassen 2014, 7;Zhang 2015). This is further demonstrated below through examining how the movement re-anchored social resources to respond to risk.

Transforming 'relations of definition' and risk resilience
One focus group comment in Xi'an is highly illustrative of what the Good Food Movement identified as the root of China's food safety problem: 'I feel there is no one that really cares about our lives anymore, really, no one cares, all they [government and industries] care about is profits and profits' (Participant 8,FG5). This alienation between consumers and producers seems to correspond to the food industry's increased ability in churning out well-packaged and chemically adulterate 'imagined consumer preferences' (Participant 1, FG1). Chinese consumers are not part of the 'relations of definition' in the food system to identify either what risk or what good food is. Thus, to respond to the intimate risk the public felt in their everyday diet, all three civil organisations addressed the relational cause in modern food risks. To start with, they urged urbanites to take the initiative and to play a bigger role in the production of food through face-to-face interactions with producers. This is an innovative effort for it is not simply a campaign for more public supervision in the food industry or to democratise the 'relations of domination' by giving the consumers a 'voice' (Beck 1995;Ni and Zeng 2009;Jia and Jukes 2013). On the contrary, the idea of having consumers be more engaged in the production process is not oriented towards an expression of concerns (or worse, blame), but an expression of interest. The point is to make both consumers and producers see that they are not at the opposite ends of a long industrial chain with conflicting interests, but that they are interdependent parts of the food community.
For example, all three organisations have semi-regular farm visit days with their key suppliers, mostly small independent farms. This is an open-house occasion in which visitors are invited to inspect the farm, to have hands-on experience with farm work and to ask farmers questions. It is an eye opener for both sides, as one organiser commented: The farmers would feel very much respected, for urbanites are willing to listen to what they have to say … it's a way to inform our consumers, but it is also make the farmers realise that what they are doing is worthwhile and is valued. (Interviewee 4) This is much welcomed by urban consumers, as it 'adds humanness to food' and gives them a sense of 'self-assumed control' amid all the scandals (Participants 2, 6, FG1). Even the simplest act of having a conversation with farmers may be empowering for consumers, as the bits of knowledge about the production process helps urbanites be more conscious of one's own consumption behaviour and empathise with challenges for sustainable production (Interviewee 3, 5, 11. Participant 1, 7, FG5). Whereas for the farmers, this socialising process also helps attach a human face to consumption, and adds an 'emotional incentive' to adhere to good practice (Interviewee 5).
Re-connecting consumers and producers, the urban and the rural, the goods and the soil, not only promotes a collective commitment to responsible production and consumption. More importantly, through engaging with different stakeholders in the negotiation of what constitutes 'good' food, this widened relations of definition brought corresponding changes to the framing of risk resilience.
In fact, practices such as eco-farming and short supply chains have their own risks. One common problem is that they can't compete with industrial farming in product consistency. Yet, many focus group participants made an interesting distinction between 'reliable food' and (commercially) 'quality food' . One revealing example is from Wuhan: Participant 1, FG3: The product they [producers associated with Wuhan Natur] sale is reliable, but in fact its quality fluctuates (a number of participants nodded) … sometimes the vegetable is a bit stale.
Participant 6: That's right, a couple of days ago I bought cabbage from them but found the heart of the cabbage already start to rot, would you call that 'good quality'? (many laughed in agreement) Participant 3: Yes, but it is still reliable food.
Facilitator: If this happened to any of your supermarket purchases, would you ring 315 (the consumer-rights hotline) to file a complaint and demand compensation?
Everyone: Yes, yes, of course! Interestingly, while all the participants readily acknowledged 'quality' as a problem and admitted that they would file official complaints if this were the case with supermarket merchandises, none of the participants ever rang 315 about products (including rotten products) they purchased from the Good Food Movement's providers. When asked why he decided to 'give up' his 'consumer rights' , Participant 6 corrected me that he did exercise his consumer prerogative, for he immediately sent out a message to Natur's organiser: But my point is not so much on getting compensation for my lost, my point was firstly to alert the farmers that there is something between harvest and delivery that they or we need to work out to avoid similar incidents in the future and secondly, I was worried about other consumers and wanted to warn them to check their deliveries. (Participant 6,FG3) Many other participants in that focus group concurred with Participant 6's view. That is, it is not so much that being part of the food risk community made them more 'tolerant' of quality issues. But being a contributing part of an interdependent community rather than a receiving end of an industrial chain reoriented how they saw their role as consumers in dealing with production inconsistency and uncertainties. Identifying risks in one's dining bowl becomes not act of punitive whistle-blowing but a call for collective monitoring and problem-solving. To put it in another way, the Good Food Movement is not about instituting a 'third-party' supervision but about cross-sectoral communication and negotiation of practices and associated risks that are mutually acceptable.
This example echoes Beck's emphasis that good governance hinges on the social relations in the identification and framing of risks (Beck 1995, 182-183). Risk associated with food is both social and real. The higher 'reliability' associated with the Good Food Movement is not because different relational 'framings' have made risk 'disappear' , but that it transforms contentious relations into cooperative ones, which widens social resources in responding to potential risk. Many studies have pointed out that the growing social anxieties towards food risks in China call for a 'relational' change in policy framing, in which both the producers and the consumers are respected as stakeholders rather than subjects and beneficiaries of regulations (Tang 2012;Jia and Jukes 2013). More specifically, as demonstrated by the above-cited example, seeing oneself as a contributing actor of a local food community reorients social actors' risk perception, and consequently adopts more constructive (rather than simple punitive) coping strategies.
To some extent, one could argue that the Good Food Movement facilities a 'bricologe' in which different stakeholders are re-socialised through to integrate their overlapping interests into a workable new collective meanings (Innes and Booher 1999). It is interesting how, in re-configurating social relations, the Good Food Movement focused on simple and seemingly mundane activities of face-to-face community events. The Chinese experience seems to suggest that it doesn't necessarily take an emancipatory urge or a social catharsis of self-guilt to activate a collective reflexivity (Beck 2011(Beck , 2015. What may be more powerful in making changes happen is the knowledge of and the subsequent ability to speak to the intimate perspectives of formerly distant (or even opposing) social realms.

Concluding words
For activists and the general public involved in the Good Food Movement, their objective in engaging with various transnational food movements, seeking diverse agricultural experts and communicating with local and regional producers is a simple one: to put safe ingredients in one's dining bowls. This collective effort to mitigate China's new 'food crisis' in itself is a testimony to the inescapability of risk associated with industrial modernity and to Beck's assertion that to cosmopolitanise is no longer an option but a pre-condition for modern survival (Beck 1992(Beck , 2006. It is important to put both the success and limitation of the Good Food Movement in perspective. Similar to other Western consumer-based alternative food initiatives, the Good Food Movement remains chiefly an urban middle-class campaign (Nonini 2013). In addition, as I have discussed elsewhere, in an authoritarian state, the survival of a civil organisation in China is subject to the tacit approval of the authority in relation to their perceived political antagonism. Thus, the impact of civil groups in China often cannot be measured in immediate policy outcomes. Instead, they must be seen in the wider social-political behaviour they cultivate (Zhang and Barr 2013). As such, the movement's impact on China's agri-economy lies not so much in the volume and scope of its production and distribution, but in its initiation of a collective reflection over the causal relations of modern food risks in major Chinese cities. As in the words of a movement staff cited previously, the social experiments initiated by each of the associated civil organisations may be 'small but it's original' (Interviewee 8). This is where a key value of the movement lies. It demonstrates the emergence and spread of a risk-invoked cosmopolitan openness among Chinese publics which leads to new forms of food communities.
The Good Food Movement in China is illustrative of 'rooted cosmopolitanism' (Beck and Sznaider 2006). It is rooted in its attentiveness to and respects for directing world experience into addressing the intimacy and particularities of food risks at the local level. It is also rooted in bringing and sustaining relational changes across conventional boundaries in the food system. The Chinese experience is both a demonstration of the validity of Beck's general theorisation, and an exposure of its conceptual gaps. It puts a question mark on the taken-for-granted logic that the shear scale of 'global' risk necessitates a 'cosmopolitan' community. In fact, cosmopolitan community is not simply an 'imagined community of fate' but a 'performative community of actions' (Zhang 2015).
It may be true that we are at the dawn of an 'epochal change of horizons' which brings 'a change in the frame of reference of change' (Beck 2015, 77). Refreshed understandings of interdependence and interconnectedness illuminated by systematic risk in the modern age would reconfigure both relations of production and relations of definition. But it is easy to lose sight of the actors and myriads of contingencies in grand theorising. More important questions are to elucidate how newly found cross-boundary solidarities reorient actors' prerogatives and legitimise a re-anchoring of social resources in collective response to risk, to which this paper only provides some initial answers.

Disclosure statement
No potential conflict of interest was reported by the author.

Funding
This work was supported by the Economic and Social Research Council [grant number ES/L009803/1].