Multi-scalar ethnography: An approach for critical engagement with migration and social change

Zhejiangcun in Beijing: Multi-scalar ethnography as an analytical strategy

Zhejiangcun (‘Zhejiang village’) is a squatter-settlement style community in Beijing formed by migrants from Wenzhou prefecture, Zhejiang province in southeast China. Despite the migrants’ lack of legal right to reside in Beijing due to the household registration system and the periodical government campaigns aimed at ousting them, Zhejiangcun grew from six families into a community of 100,000 in less than ten years, in a place 5 km south of Tiananmen Square. Almost entirely based on unregistered informal household production and trading, the community was a supply centre for medium- to low-priced clothes for the whole of north and northeast China. Zhejiangcun was largely self-sufficient, with its own kindergartens, clinics, security guards, vigilante teams, fire brigades, and a range of other service businesses.

My earlier ethnography sought to reconstruct how marginalized migrant peasants became resourceful enough to resist the state. In doing so, I called attention to bottom-up social changes and critiqued the elitist, state-centric discourses about the reform that dominated the 1990s. By following things (garments and raw material), credit, and especially migrants', journeys from home to Beijing and then to other parts of China, my fieldwork was decidedly multi-sited. My basic conclusion was that Zhejiangcun was but a tip of the iceberg that was Wenzhou migrant garment producers’ nationwide networks. As such, the life in the community must be understood through the massive connections between the multiple sites. In contrast to government actions that were strictly confined to jurisdictions, the Zhejiangcun networks knew no administrative boundaries. During the Beijing government’s clean-ups, Zhejiangcun migrants arranged for quick escape to other parts of China, and then an equally prompt regrouping and return once the campaigns died down. While the government campaigns were almost by definition short-lived, the migrants could sustain themselves for quite a while by moving between sites. I described Zhejiangcun as a ‘non-state space’ (Xiang, 1999), using the metaphor of ‘space’ to highlight its multi-sited unboundedness. The ethnography adopted a binary between multi- or un-sited networks and sited institutions to conceptualise the supposed binary between the migrants and the state.

I started questioning this narrative when I noticed a number of significant changes in Zhejiangcun since the early 2000s, after my research was published. The spontaneously developed trading spaces and residential compounds of the migrants were demolished, and thereafter replaced by large marketplaces and high-rise residential buildings that were built jointly by government, real estate developers, and Zhejiangcun businesspeople. Most strikingly, the single most influential migrant in Zhejiangcun, Liu Hefei, who used to be under close surveillance by Beijing Public Security Bureau during crackdowns (Xiang, 2005: ch. 7), is now a member of the Chinese Communist Party and serves as the secretary of the Zhejiangcun committee of the Party. Apart from recruiting new Party members and helping to maintain social order in Zhejiangcun, he is heavily involved in tracking down, detaining, and sending petitioners from Wenzhou back to Wenzhou before they lodge complaints with the central government about the Wenzhou government. These are clearly multi-sited operations based on his informal networks and influence, but are carried out under direct instructions from the Wenzhou and Beijing Party committees. Making money is still Liu’s top priority, and this is what the Party wants him to do. He put it frankly: ‘You won’t have the authority [in the community] if you aren’t successful in your business. Young people want to join the Party because party members are capable people with influence.’ ³ While the migrants’ formal status remain quasi-illegal, their relations to the local state became much closer and more positive. Zhejiangcun is clearly no longer a ‘non-state space’. It probably never was.

With the benefit of hindsight, it appears that the notion of scale is more pertinent and productive than the metaphor of space in understanding Zhejiangcun. First, a multi-scalar perspective can better explain why the multi-sited networks empowered Zhejiangcun migrants from the beginning. The networks and circulations formed emergent scales that generate new capacity. Where did the new scales and the capacity come from? This leads us to reinterpret the everyday life of Zhejiangcun. The basic unit of the everyday life is what I term ‘xi’ – literally ‘bunch’ in Chinese – a cluster of relationships comprised of a relatively stable social circle of kin and friends and a more elastic business circle. Each household had its xi. Relationships in one circle affected those in the other. For instance, a migrant with strong and wide social circles was more likely to develop large-scope and long-distance businesses. This was because their social circles provided back-ups in business dealings, which was particularly important given their lack of access to formal legal means (e.g. in enforcing contracts). Xi was the basic unit based on which the migrants made calculations and decisions in daily life, thus forming a scale of action. Members of a particular xi, both relatives and business partners, were often scattered across China. At the same time, different xis inevitably overlapped. For instance, a relative of household A could be a business partner of household B. Interactions among xis generated larger scales of action. This was how informal leaders in the community, including Liu Hefei, emerged. The leaders had large and powerful xis. Members in their xis tended to be well-connected and able as well, and they developed their own strong xis that overlapped with each other. This provided the central figure with broad bases to exert influence. Operating on larger scales than others, the big players often assumed the role of arbitrator of disputes and guarantor for business undertakings, which in turn provided much needed public goods for the community. The scalar perspective thus brings to light the institutional significance of the daily life, particularly the structuring effects of the seemingly diffusive and amorphous networks.

A scalar perspective also sheds new light on the meaning of site. The migrants’ home townships illustrate this. Zhejiangcun migrants went home for every Chinese New Year religiously; they made generous financial contributions to traditional ceremonies and public projects such as road constructions even though the villages were almost empty most of the time due to outmigration. Home townships were important as key sites where business disputes that occurred across China were settled, new business plans discussed, ideas exchanged, and large amounts of money borrowed and previous accounts cleared – and all during the New Year season. The business plans and disputes were national in scope, but the negotiations were intentionally confined to the township level in terms of the networks involved. For instance, fellow villagers were invited as witnesses when a business agreement was signed, and a banquet was offered to the common friends and neighbours when a dispute was settled. The historically accumulated, locally grounded connections and symbols constituted the most important source of trust. The promise that one makes in front of the extended family on the day after visiting the local temple carries more weight than contracts signed elsewhere. The home township was turned from a physical site into a scale that defines the pattern and logic of interactions, which were in turn indispensable for their interactions on larger scales. ⁴

In addition, a multi-scalar perspective leads to more nuanced understandings about the relation between mobility and established institutions. I had earlier attributed the remarkable growth and yet perpetually uncertain status of Zhejiangcun to the place-based compartmentalization in the Chinese bureaucracy. Yet there is more to it when seen from a scalar perspective. The state enforces rigid territorial demarcation because it is useful to ensure centralized control over the vast country. It is by limiting the scale of influence at a subordinate level that the higher level claims its authority. The horizontal fragmentation between places should thus be understood as part of the taxonomical hierarchy. The hierarchy consists of six levels: the centre, province, prefecture, city/district (or county in rural areas), street (jiedao, or township for the rural), residents’/villagers’ committee. While capitalist states of the Fordist era strove to maximize the economies of scale by equalizing socioeconomic developments across the nation, and subjugate the space-bound bureaucracy to the market, the socialist Chinese state of the same period accumulated wealth through localized production instead of cross-space exchange. Socioeconomic unevenness across space is readdressed through centralized redistribution instead of a unified market. This rendered communication between governments of different places particularly complicated. For example, if a street office – the grassroots government organ – in Beijing wishes to enforce family planning with a Wenzhou migrant living there, in theory the street needs to report to the district, the district to the municipality, and then to the centre; the centre passes on the message ‘downwards’ to Zhejiang province, to be relayed to Wenzhou, to the county, and finally to the township and the village where the migrant is officially registered. The scalar composition of the state severely impedes the state from dealing with mobility effectively.

Furthermore, the taxonomical scalar system itself becomes unstable when interacting with the emergent scale. Different levels of bureaucracy react to migration differently. A higher level of government whose purview encompasses both the migrant sending and receiving places tends to be more lenient than that of only the receiving place. For instance, it does not make much sense for the provincial government to be harsh with intra-provincial mobility. It is then far from coincidental that Beijing as the centre of state power became the centre of Zhejiangcun networks. Beijing’s uniquely dense and complex bureaucratic structure provided unusually wide cracks, frictions, and unevenness inside the state that the migrants could exploit. For instance, when migrants ran into trouble with the district government, they may seek support from cadres in the municipality, or may even approach staff in ministries. The migrants’ favourite patrons were senior officials at less powerful, but high level, institutions, such as the Ministry of Health, of Sports, and state cultural associations. Not only were these officials more approachable, but their letters and phone calls were also equally effective for matters such as securing a stall in the prime commercial area in the downtown. Zhejiangcun migrants were also keen to approach various media to expose unjust treatments inflicted on them by the local government. A large number of newspapers in Beijing were associated with national agencies (e.g. The National Youth League or the Ministry of Agriculture), and were therefore more sympathetic to migrants than the local government departments.

The intra-state cross-scale fragmentation was particularly ingeniously utilized by Zhejiangcun migrants in coping with the government crackdowns. They called their tactics guerrilla warfare. If the street office launched the clean-up campaign, the migrants simply moved to the neighbouring street for a few days. If the order was from the municipality, they commonly moved to the neighbouring province, especially Hebei, for a couple of months before coming back. The physical scope of the runaway was always one step ahead of the scale of the government campaigns. After each bout of ousting, a few would not bother to return to the original site, and turned instead to the new periphery of Zhejiangcun. Thus, in the arms race over scales, government clean-ups became the direct driving force for the geographical expansion of Zhejiangcun. The community survived and flourished not because it was an autonomous ‘space’ outside of the state, but more because its multiple scales cut across and interacted with the multifaceted bureaucracy.

The intersections between the taxonomical and emergent scales also explain the changes initiated by the Beijing municipal and district governments. While they had previously failed to eradicate Zhejiangcun, they were successful in incorporating it into the new mode of governance in the 2000s. They did so through a combination of market building and Party building. By providing marketplaces that were the centre of Zhejiangcun’s daily socioeconomic life and the hub of its widespread trading networks, the government effectively put its hands on the lifeline of the community. At the same time, the district Party committee in Beijing collaborated with their counterparts in Wenzhou to set up a Party committee specifically for Zhejiangcun migrants. The committee was under the dual leadership of the Beijing and Wenzhou committees, and in turn established a number of cells (dang xiaozu) in Beijing along migrants’ home township. Party cells need no fixed budgets, paid personnel, or designated tasks; and yet they can actively reach out at the grassroots level. Compared to government bureaucracy, the Party is less territorialized and less functionally fragmented, and therefore much easier to initiate cross-jurisdiction communication. After all, revolutionary parties in modern China had their origins in secret societies or reading groups. Because of its scalar flexibility, the Party serves as ‘glue’ that holds the state bureaucracy together; and precisely because of this, the Party has become more important since the 2000s when economic privatization deepened, social life diversified, and government functions became more complex and specialized (Pieke, 2004, 2009). Therefore, there is little irony that commerce and communism go hand-in-hand. Both are adaptable to and are reliant on emergent scales. The reworking of the scalar operation of the party-state is one of the most significant developments in China. This is even more evident in the case of transnational unskilled labour outmigration, to which I now turn.

Transnational labour recruitment: Multi-scalar ethnography as a field method

Unlike site or level, scale is not self-evidently out there. Scale is an abstraction aimed at capturing actors’ strategies and internal dynamics of social change. How can we, then, conduct multi-scalar fieldwork before knowing what multiple scales we may discern? My experience suggests that a multi-scalar perspective is likely to come in halfway through the fieldwork. For a start, we will need to identify the problematics of social and political concerns. As multi-scalar ethnography is ultimately a framework aimed at explaining why and how certain problems arise rather than a descriptive tool, a multi-scalar fieldwork can be confusing if it is not tied to clearly defined problematics. We will then need to identify empirical points of entry – the more or less self-evident, easily observable phenomena or objects around which we can start collecting data. Entry points may lead to ‘empirical anchors’. Empirical anchors are the central subjects through which we can obtain systematic understanding of the problems that concern us. They can be gatekeepers, defining events, quasi-ethnographers in the field such as activists, journalists or lawyers, or any of the six subjects such as those that Marcus (1995) listed in the tracking strategies. From that stage on, if our concerns and the empirical data suggest that a scalar perspective may be helpful, we can deepen fieldwork by following up with questions about scales: At what scales do the key actors operate? How is each of the scales constituted and maintained? What are the inter-scalar relations, and how are the relations reproduced or disrupted?

The first step, identifying problematics, may take a longer time than expected. When I started working on unskilled labour outmigration from China in 2004, I was concerned with the political and social implications of transnational mobility on the Chinese state. I chose northeast China, particularly Liaoning province, as the main research site because I thought the tensions between transnational flows and nation-bounded state power would be particularly acute there. The northeast was the national base of heavy and military industries since the establishment of the People’s Republic. The region represents an archetype of state socialism where the bureaucracy was all-powerful and population mobility was tightly controlled. State-induced privatization in the 1990s, however, led to massive lay-offs, which contributed to the increase of labour outmigration. Japan, South Korea and Singapore were included in the fieldwork not only because they were the three top destination countries of Chinese unskilled labour migrants worldwide, but also because of their deep anxiety of controlling foreign populations, thus providing sensitive testing cases on the receiving side.

My interest in the relation between the transnational and the national was derived from the literature rather than from grounded observations. According to the literature on transnationalism, cross-border flows and connections gain their own momentums and create ‘transnational space’ (Pries, 2001) or ‘transnational field’ (Levitt and Glick Schiller, 2004) that cannot be accommodated by, and therefore pose challenges to, nation-states (Sklair, 1995; Portes et al., 1999; Castles, 2001). Wimmer and Schiller (2002) argued that transnational migration provides a powerful antidote to methodological nationalism that has fundamentally shaped and constrained modern social sciences. Looking through this lens, migration from China is particularly significant given the communist state's long-time isolation. The rapid increase in labour outmigration is certainly one of the most visible indicators of a reformed and more open China: more than 850,000 Chinese were working overseas on relatively long-term temporary contracts by the end of 2012, as compared to less than 60,000 in 1990 (China’s International Contractors’ Association, 2004; Ministry of Commerce, 2013). It is evident from my initial fieldwork that migrants develop dense and expansive transnational networks, and most of them return home on completion of contracts, thus engaging in typical transnational back and forth movements. On the surface, this seemed like the typical transnationalism story.

However, closer observations revealed that the transnational connections do not necessarily empower the migrants, nor do they necessarily challenge nation-states. Migrants’ networks often work against their own interest. For instance, recruitment companies in China effectively utilize migrants’ transnational connections to discipline them through an arrangement called lianzuo, or ‘linked seats’. Originally an invention of the Emperor Qin of 200 BC, lianzuo groups migrants in order to share liability. That is, if one misbehaves, all will be punished. The pressure is imposed not only on the team, but is also leveraged through the migrants’ families back in China. As a manager of a recruitment company recalled, a worker recruited by him went underground in South Korea. The recruiter warned all the families on the ‘linked seats’ that all the migrants would be repatriated to China if the absconder remained on the run. The family of the absconded worker was inundated by visits and telephone calls from other families, and swiftly pressured the worker to return to the implicated employer. As such, recruiters’ transnational coordination severely curtailed migrants’ scalar capacity. This is a type of power relation that emerges from diffusive, wide-ranging circulations involving heterogeneous actors, rather than resulting from a single logic or formal particular institutions.

We need to focus on circulation in order to figure out this type of power relation. But what should we follow, and where should we start? As the migrants’ movements are well planned and organized by recruiters (see Xiang, 2012a), it would not be particularly exciting or profitable to follow the journeys. Instead, I followed what concerned the migrants most: (1) money (paid to the recruiters); (2) information (about overseas job opening); and (3) documents (passports, skill certificates, job contracts, etc.). The three often move in tandem. In following these things, it became clear that transnational traffics were but a small part of many complex flows that took place within each of the countries. The transnational ‘moment’, so to speak, is brief. It is not antithetical to the nation-states, but is positively constituted by social processes within the nations. I found myself not only travelling back and forth between the places of origin, transit, and destination but, more importantly, running up and down between a hierarchy of places in each country. In China, my fieldwork was carried out in places existing on five levels: (1) townships and villages with relatively large numbers of migrant overseas; (2) the county seat to which the townships and villages belong; (3) the capital city of the prefecture; (4) Shenyang, the capital city of Liaoning province; and finally (5) Beijing. In Japan, I interviewed officials and large trade unions in Tokyo and Osaka, and visited small factories hiring Chinese workers in remote places. Money, information, and contracts moved between these sites, and were dealt with differently in different sites depending on which level they were situated in.

Hence, these things and sites became my empirical entry points. They opened windows to the multifaceted processes of migration, and led me to the main subjects for systematic investigation – the ‘empirical anchors’. My empirical anchors were the labour recruitment intermediaries. They demand, collect, and pass on the money. They process as well as possess information. And they acquire, process, and even manufacture documents. Located at different levels, the recruiters can be private companies, public institutions (such as vocational schools or villagers’ committees) or individuals. At the top level are large companies with the exclusive licenses issued by the Chinese government that grant them the right to sign international contracts and certify relevant legal papers. They are thus called ‘window’ companies – the window to the world. Most of these windows are located in big cities, especially Beijing and the provincial capital, and therefore have little connection with prospective unskilled workers in the countryside. As a result, they outsource the task of labour recruitment to middle-level agents in the prefectures, who in turn rely on subagents in local districts or rural townships. These grassroots recruiters, especially the individuals, are referred to colloquially as ‘the legs’. The lower-level recruiters do not receive remuneration from the higher-levels for the recruitment. They take a cut, instead, from the fees paid by would-be migrants. In the late 2000s, migrant workers had to pay an average of USD8,000 to work abroad in Japan, Singapore, and South Korea. Such is the multi-level recruitment chains that shape the process of migration. The system is as much about facilitating as about controlling, and as much about door-opening as about gate-keeping.

At this point, the fieldwork could proceed in different directions. I could, for instance, focus on migrants’ experiences of having transnational connections ‘under siege’, or on companies who have developed strong transnational recruitment business. I could emphasize differences among the recruiters, among the destination countries, or among the migrants along their place of origin or occupation, and thus focus on one stream of multi-sited transnational flows. But the question that concerns and intrigues me the most is: how do processes of transnational mobility contain migrants’ agency and reinforce state power? The chain of recruitment mirrors the state bureaucracy. This resemblance is by no means coincidental and is crucial in controlling migration. Yet most recruiters are commercial entities who seek maximum profits through market transactions. As such, a narrow focus would be insufficient to examine such multi-faceted relations between the market and the bureaucracy. This phenomenon must be examined in its entirety, with the view of disentangling the interactions among the multiple parties systematically.

To this end, scalar questions can be useful in identifying the types of detailed data I need to collect. What is the scalar position of the different recruiters? What is the scalar division – the differences and complementarities – of labour among them? What keeps the recruiters bounded at each scale – for instance, ‘the legs’ do not go international and circumvent the higher levels – and how are the migrants prevented from ‘scale jumping’? These questions translate the empirical anchor into an analytical subject matter.

The multi-scalar ethnographic fieldwork revealed more intricate relations between the market and the state than I had appreciated. The window companies are powerful partly because of their international scale of operation. But they are certainly not at a taxonomical scale. The windows do not encompass, let alone encapsulate, the legs. They are business partners. The windows tend to have an upper hand in actual negotiations and earn higher margins because they have the exclusive access to the international scale. In other words, what matters is not the scale per se, but the monopoly over the scale. The windows’ unique position is a result of the commercialization of the taxonomical scalar power of bureaucracy. Labour outmigration from China has its origin in foreign aid projects in the 1950s. When carrying out aid projects overseas, the ministry of international economy required its provincial branches (bureaus of international economy) to organize project teams, and the province government may in turn have tasked a municipality office to do the same. It was thus a typical taxonomical scalar structure. At the end of the 1970s, the central government started turning foreign aid into international commercial projects, and reorganized the relevant ministerial sections into state-owned companies called ‘central companies’, which then became window companies. The window companies continued with the old practice of tasking local government companies – transformed from local bureau of international economy – with the recruitment of workers to be sent overseas. Administrative command chains were thus turned into commercialized agent chains. Subsequent economic liberalization did not undermine the position of the windows (Xiang, 2012b). For instance, the government allowed private companies to enter the business of international labour placement in 2002, but set the barriers to entry very high through licensing. The catch is that most licensed private companies are privatized former state-owned enterprises. Most other private companies are unqualified for these licenses and have to continue working as subagents.

The relations between the taxonomical and emergent scales are equally deep and complex at the grassroots level. The legs in villages and townships operate on a local emergent scale. They utilize various local networks, institutions, and knowledge to mobilize as well as discipline migrants. They sometimes require would-be migrants to identify one or two local civil servants as guarantors, who will have to pay the legs for the migrant’s wrongdoing overseas. The payment is often referred to as ‘compensation’ since the legs will be fined by higher-level recruiters for the migrant’s violation of regulations. Civil servants are usually the most influential figures in an extended family, and moral pressure from them is more powerful than the concern about financial loss in keeping the migrants in line. In order for the legs to assess whether a particular civil servant is related to a particular migrant, local knowledge is essential. The legs also need close connections with the local government in order to enforce the guarantee. It is this local scale that makes the legs indispensable for other recruitment companies. While the local scale is emergent from the legs’ actions, the scale is effective because it entangles with the grassroots level in the bureaucracy taxonomy.

The relation between the emergent and the taxonomical is, however, not a neat one-on-one correspondence. The role of the middlemen, who are structurally positioned between windows and legs, clearly illustrates this. The middlemen can be found in every site, but are most active in prefectural capital cities. They are typically entrepreneurial individuals who have inside knowledge about labour migration – most of them worked for state-owned window companies before. The reasons for the existence of their niches are multiple. For instance, they are indispensable for some staff at state-owned window companies to make extra money. As the window companies are subject to government regulations about the maximum fees that they can charge, window staff sometimes ‘sell’ job orders to trusted middlemen secretly at high prices, though the window remains as the nominal recruiter in formal documents. The middlemen in turn seek higher fees from the migrants. The legs also need well-connected middlemen to find placements for the would-be migrants, especially because the legs often over-recruit due to their lack of accurate information about overseas demands. A particular middleman goes between large numbers of intermediaries instead of confining him/herself to working with particular windows and legs. Some middlemen have strong connections with overseas recruiters or even employers and cut deals with them directly. The middlemen thus create their own emergent scales. But the middleman’s emergent scale does not undermine the hierarchical structure of the recruitment chain. Even with international deals in hand, the middlemen still have to pay a window company (called ‘window fees’) to package the deals as the window’s in order to obtain the required paperwork. The middlemen also rely on the legs to mobilize and collect money from the would-be migrants because the migrants trust only the local legs. It is also in the middlemen’s interest to maintain such a hierarchy. After all, the middlemen profit not by developing an open market for smooth migration, but by capitalizing on their special connections with the windows and the legs.

Multi-scalar ethnographic fieldwork also calls attention to migrants’ specific strategies of resistance. The deep entanglement of the recruitment market and the state bureaucracy renders it difficult for the migrants to develop emergent scales that disrupt the system from within. The migrants are fully aware that they are constrained not only by state policies, but more importantly by the transnational, multi-scalar structure: the problems that they face overseas are often directly tied to the hometown, and irregularity on a particular level may be caused by another level. Facilitators are also exploiters. But this does not mean that the migrants are powerless victims. A common strategy of the migrants is to stage protests in the capital city of the destination country, often in front of the Chinese embassy. They also actively seek support from the public media and civil society in the host country. They do so in order to turn their contestations into public events of an international scale and to maximize its visibility to the states; otherwise the issue may simply disappeared in the multi-scalar relations. In a sense, the migrants’ strategy is strictly ‘sited’ – by concentrating on the square in front of the Chinese embassies for instance – but they choose this site because it generates scalar effects. Such physical sites provide concrete points for engagement and intervention.

Multi-scalar ethnography as critical engagement

Multi-scalar ethnography in itself does not provide substantive theories. But as a middle-range framework, it gives structure to dispersed daily activities, and at the same time unpacks the workings of established institutions. Centred on the intersection between taxonomical and emergent scales, multi-scalar ethnography helps to relate immediately observable phenomena to larger institutional setups and structural changes. Multi-scalar ethnography simultaneously ‘studies up’ to interrogate the centre of power (Nader, 1972), ‘studies down’ to examine everyday interactions, and most importantly ‘studies through’ (Reinhold, 1994: 477–9; cited in Shore and Wright, 1997: 14) to ‘trace ways in which power creates webs and relations between actors, institutions, and discourses across time and space’ (see also Hannerz, 2006; Nyiri, this issue).

Multi-scalar ethnography does not aim to provide comprehensive accounts about migration. Its main concerns are instead about logics of actions, emerging capacity, and possible changes. Multi-scalar ethnography follows flows and connections, but more importantly traces people’s concerns, calculations, and strategies. It seeks to explain why certain changes take place and others don’t, and identify the interfaces between mobility and institutions where interventions are feasible and productive. Multi-scalar ethnography does not at all discount the importance of sites, but articulates the meanings of sites to the actors.

If multi-sited ethnography constructs rather than merely reflects reality, multi-scalar ethnography goes further to construct the reality according to the logics of how the reality is internally constituted in practice. If multi-sited ethnography resonates with the Russian avant-garde artists, multi-scalar ethnography may have ideological affinity with Mao’s philosophy. In one of his most celebrated philosophical pieces, ‘On Contradiction’, Mao (1966 [1937]) asserted that internal contradictions are the fundamental driving force of changes. Everything is born out of contradictions, and everything contains in itself contradictions. Internal contradictions may be transformed into external contradictions, and vice versa; but external contradictions must work through the internal ones to affect social changes. Scholars can endlessly debate whether internal forces or external influences drive revolutions; but Mao was absolutely convinced that one has to grasp and build on internal contradictions in order to make changes. Multi-scalar ethnography, with its emphasis on the intersection between emergent and taxonomical, provides us with a method to explain how the new is born out of the old, and how cracks within established systems can be widened to lead to changes. Multi-scalar ethnography certainly does not claim to be a supreme way of doing research. But it can be useful for those who see research as a means of critical engagement.