I am what Chinese would call a waihang (outsider to the guild) in this discussion, neither a specialist in European urban history nor up-to-date on the body of critical social theory that informs much of the discussion in the roundtable. Therefore, I see my role here as primarily presenting a comparative case – China – and only incidentally engaging with the theoretical aspects of the discussion.

Let me get the latter out of the way first, offering some general comments about the question of ‘urban agency’ as a tool of analysis. First, of course, ‘agency’ implies reification, something we all ought to be at least cautious about. I see this as akin to such ambiguous generalizations as ‘feudalism’ and ‘capitalism’; in invoking such terms, are we talking about presumed empirical reality, or simply a narrative device imposed by the historian? Such authorial interventions are necessary and desirable in some cases, but we need to be constantly aware that all such narrative choices bring with them liabilities as well as strengths – what is obscured or left out by such choices? Second, do we conceive of a possible, perhaps even preferable, middle ground between understanding the city as active ‘agent’ and as passive ‘site’? Would we settle, say, for such a notion as ‘catalyst’ or enabler for processes that were in fact larger than urbanism alone?

Moreover, if the city is to be taken as ‘agent’ for major historical transformations, it would seem incumbent upon us to specify carefully just which presumed transformation(s) we are talking about. The transformation to capitalism? To industrialization? To some kind of Weberian ‘rationality’? To individualism? To political self-determination? By failing to specify, we leave ourselves open to the presumption that we are talking about the sum total of historical transformations that occurred in the early modern and modern west – thus universalizing the local and effectively closing off comparative discussion with other regions of the world.

In fact, there is a long tradition, dating back to Max Weber but including scores of western sinologists influenced by him, of imposing the motif of urban agency on the historiography of China.Footnote ¹ It is a narrative of lack. Lacking urban agency, China failed to become ‘modern’, like the west. China here becomes either the failed ‘other’ to Europe or, yet more perniciously, the passive recipient of necessary western guidance. On the other hand, if we argue, as I have for years, that China's lack here was more apparent than real, are we not facing the danger of simply reaffirming the European master-narrative but adding one new player?Footnote ² How can we effectively disentangle China, and other non-western societies from this imposed master-narrative? Or, better yet, create a new, more satisfactory narrative, within which the west is not so privileged? It is not an easy task.

There are several possible ways to approach the question of urban agency in Chinese history. The first and arguably most important of these would be to look for economic agency. Analogously to E.A. Wrigley's analysis of London's agency in early modern England, Robert Hartwell years ago posited the role of Kaifeng's ‘metropolitan demand’ in shaping North China's regional coherence during the Northern Song (980–1126 CE). Focusing on construction materials and household fuel to support the newly declared imperial capital and its suddenly inflated population, Hartwell argued that higher volume and a new nodal direction of trade flows, and increasingly specialized local production, effectively articulated an organized and efficient metropolitan region where none had existed before. But this process was cyclical, not secular: regional integrity could not and did not survive the subsequent movement of the capital to another urban site.Footnote ³ It could be argued that similar processes of regional formation took place with the declaration of Nanjing (in 1368) and then of Beijing (in 1403) as imperial capitals, with the declaration of Guangzhou as the sole permissible port for maritime trade with Europe (around 1760), with the breaking of that monopoly and the rapid shift of the centre of the western trade to Shanghai (after 1842), and, conceivably, with the declaration of the multiprovince ‘Shanghai Economic Zone’ under the post-Mao People's Republic in 1982. It is worth noting that all of these processes of urban ‘agency’ were in fact initiated by deliberate decisions of the central state.

Culturally, urban agency in China may be detected in the city's role as taste-maker, stimulus and guide to burgeoning consumerism – affecting such areas as artistic performance (opera, storytelling), fashion, cuisine and even tourism. Recent research has underscored the fact that this was not simply a modern phenomenon but was well underway in the early modern period as well. It was not generically associated with the influence of ‘the city’, but rather with specific, even competitive, urban places – in the Ming and early Qing ‘Yangzhou-style’ vs. ‘Suzhou-style’, in the twentieth-century ‘Shanghai-style’ vs. ‘Beijing-style’ and so on.Footnote ⁴

In the political realm, other papers in the roundtable raise the issue of ‘governmentality’ – the regulation of truth-regimes, the cultivation of the self-disciplining subject. Imperial China was in fact a consummate master of this, via state Confucianism and its success in implanting the civil service examination as the orthodox avenue of social advancement. But how specifically ‘urban’ was this? In fact, while Confucian thought has much to say about commerce and the profit-motive, it is nearly totally silent on the problem of the city. This does not mean, of course, that administrators who thought of themselves as good Confucians did not routinely grapple with issues of urban governance in their daily practice.

One factor that has frequently been adduced to explain the lack of urban agency or urban autonomy in early modern China is the absence of formal urban administrative discreteness: cities were administered as part of their surrounding agrarian counties, by a single local official. In the early twentieth century, the limitations inherent in this were recognized, and important cities were granted administrative discreteness as ‘special municipalities’ (tebie shi), with county-equivalent status of their own. But in the late imperial period, it is conventionally argued, most urbanites of middle-class status and above, both merchants and literati, were sojourners, identifying most directly with their (usually rural) native places (laojia) and unlikely to form a transcendent community attachment to their host city.

In my own early work, I argued that this seeming lack of urban community was regularly overcome in practice, with groups of sojourners from disparate native places co-ordinating their activities to serve their adopted urban home. This was true in daily circumstances for issues of philanthropy and poor relief, but also more spectacularly in cases where the host city was threatened: in Chongqing, for example, the ‘Eight Guilds’ of extra-provincial merchants together contributed funds to ransom the city, first from a renegade Taiping army in the 1860s and then again a decade later from a French gunboat seeking to exact reprisals for the murder of a French missionary. The establishment of a gentry-merchant run municipal administration in Chinese Shanghai in the first decade of the twentieth century, I argued, though ostensibly based on observation of the European model enacted in the contiguous International Settlement, built on these wholly indigenous precedents.Footnote ⁵ My train of logic here has been disputed by Bryna Goodman, who argues that among Shanghai merchants native-place identity was never trumped by host city solidarity. Native-place identity was joined in the twentieth century by a new Chinese national consciousness, but this was built directly on native-place sentiments, bypassing entirely the middle level of trans-urban solidarity.Footnote ⁶ The truth, as it often does, probably lies somewhere in between.

A related line of argument concerns the emergence of what I and others have termed a ‘public sphere’ of elite activism in eighteenth- and (especially) nineteenth-century Chinese cities, a realm of collegial elite activism that was legitimated by service to the public interest (gong), in deliberate distinction, though not usually in overt opposition, to the sphere of bureaucratic administration.Footnote ⁷ Let me cite but three examples. In Foshan, a commercial-industrial town not far from Canton, a de facto town hall was established in the 1790s by local gentry and sojourner merchants, to manage the city's various public enterprises (schools, granaries), based on non-bureaucratic sources of funding such as self-imposed wharfage fees and ferry service tolls.Footnote ⁸ In the Shandong treaty port of Yantai in 1889, a single municipal ‘benevolent hall’ (shantang) co-ordinated all the city's infrastructural services, independent both financially and managerially of the local representatives of the Qing imperial administration. And in the central China commercial entrepôt of Hankou, where the scores of merchant-gentry ‘benevolent halls’ followed the more usual pattern of neighbourhood scale, all the shantang of the town officially amalgamated in the first decade of the twentieth century into a municipality-wide alliance that gave formal articulation to the pattern of collective collaboration that had emerged informally over the preceding half century.

Let me now attempt to historicize this process a little more effectively. In terms of periodization, Bert De Munck's essay here argues for two breakpoints in European urban history, around the years 1500 and 1800. In China, give or take a few decades, it is not difficult to accept the pivotal significance of the first of these breakpoints as the onset of ‘early modernity’, the start of a fertile period of commercial intensification, demographic growth, social and geographical mobility and social change. This would include processes of urbanization as well, although we now believe that the rise in China's urbanization rate in the early modern era was centred not on the largest cities, as seems to have been the case in Europe, but instead on the proliferation and expansion of smaller cities and market towns. It is worth stressing that these changes signalling China's transition to early modernity were indigenously generated and not significantly the result of European influence, although the Columbian exchange – with its introduction of American silver and New World crops – clearly played an important role in China, as it did at the other end of the Eurasian land mass.

As for the other suggested breakpoint, marking the transition from early modern to modern, I think the year 1900 works much better for China than does 1800. For one thing, it is this year (or rather, the aftermath of the Treaty of Shimonoseki of 1895) that signalled the onset of Chinese urban industrialization. And here, not in the earlier period, we do see the compelling influence of western models. Indeed, one hallmark of this era is a frenetic ‘remaking’ of Chinese cities, to bring about catch-up ‘civilization’ (wenming) under the gun – as many Chinese elites of the time saw it, in this new Social Darwinist universe their country must either ‘civilize’ (i.e. westernize) or face extinction.

We can apply this periodization to the issues of ‘visuality’ and ‘materiality’. The early modern package of changes in Chinese cities visually was fairly modest. Cities continued to look pretty much the same as they had before: they were walled, enclosed a dense network of warren-like lanes, and, except for the imperial capital, tended to lack monumental architecture. What did characterize this new era of intensified long-distance exchange was the spread of the city extramuros, the growth of suburbs with commercial rather than administrative functions.

Changes in urban visuality in the early twentieth century were much more dramatic.Footnote ⁹ A new aesthetic of urban openness and grandness took hold. City walls were torn down, an act signifying, counter-intuitively, a greater rather than lesser urban retreat from the countryside. Broad, straight boulevards were carved out of existing neighbourhoods, and a new monumental architecture was implanted: government buildings, railroad stations, banks and most notably department stores (baihuo gongsi) and the new consumerism that came with them. Long-standing squatter colonies were forcibly removed. In China, this was emphatically western-derivative, influenced especially by the American ‘City Beautiful Movement’. Changes in urban visuality were central to advancing a new joint elite/state agenda of governmentality: what Chris Otter calls ‘liberalism’ and the historian of China, Michael Tsin, sees as a distinctive combination of ‘emancipation’ and ‘discipline’ of the urban citizen-subject.Footnote ¹⁰

Let us consider two especially well-studied Chinese cases, Tianjin and Chengdu. In the former, the largest and most important seaport in North China and a city periodically under foreign or hybrid rule, a series of successive municipal governments in the first decades of the twentieth century each pursued their own version of ‘hygienic modernity’. The city's ‘Sanitation Bureau’ (Weisheng ju) was extremely activist: sending out uniformed crews (often made up of homeless persons) to clean the streets, tearing down the old city walls in order (ostensibly) to remove stagnant water and dilapidated housing, and enforcing new zoning regimes – removing repellent businesses such as dyeshops and tallow-makers out of the city's new ‘downtown’. In Chengdu, the ancient capital of inland Sichuan province and never a western treaty port, early twentieth-century ‘civilization’ campaigns headed by the warlord Yang Sen tossed street beggars into workhouses, confined prostitution to less visible districts, installed gas lamps for public lighting, set up benches along major streets, expropriated temple grounds for repurposing as ‘public’ parks and demolished large tracts of the city centre to create a central business district of department stores and tree-lined boulevards.Footnote ¹¹

This two-stage model of change can also be applied in the Chinese case to two phases of emergent urban consciousness and urbanites’ political activism (I will leave to my readers to decide whether or not this constitutes urban ‘agency’). In the final decades of the Ming (1620s to 1640s), historians have found an emergent, class-transcendent ‘city-people's mentality’ (shimin sixiang), linking patrician and plebian elements in resistance against fiscal and other intrusions on the part of the agrarian-based imperial state.Footnote ¹² In the second, early twentieth-century, phase, we see the coalescence of what has been called the ‘urban reformist elite’ – not strictly speaking a ‘bourgeoisie’ (that is, not primarily new industrial capitalists), but a diverse middle-class interest group including businessmen, certain classically trained literati and various emerging white-collar professionals. These elites pushed a developmental and constitutional agenda that led them to become, in many cases to their own surprise, the engine of the 1911 Republican Revolution. Though they were seldom explicitly urban in their professed goals, their common residence in large cities was central to their cohesion and outlook.Footnote ¹³ Then, the ascendance of the Nationalist Party (Guomindang) in the 1920s offered a new, forceful ideology of the nation-state, rooted in an unprecedentedly urban class-basis and policy orientation.Footnote ¹⁴

Let me conclude by offering a brief alternative or modification (a friendly amendment?) to the ‘urban agency’ model of historical change in China. It has often been noted that the root of the Chinese word ‘civilization’ (wenming) has nothing to do with ‘civil’ or ‘urban’, as in the west, but derives instead from the notion of written language (wen). It might analogously be argued that the real agent of transformation in early modern and modern China was not the city per se but instead the growth of literacy, print media, and the print culture, expanding progressively downward in the social scale to include middle-brow audiences of greater numbers and less classically informed tastes.Footnote ¹⁵ Although cities were disproportionately the location of this expending reading public, and (especially in the modern period) its subject, it might be harder to identify the city as its agent.