Rediscovering Shanghai modern: Chinese cosmopolitanism and the urban art scene, 1912–1948

Abstract

Republican Shanghai was a renowned art capital. This article is based on a large-scale digital mapping project of the residential locations of 1,349 Shanghai artists. We analysed the transformative spatial distribution patterns of artists in relation to the city's social and urban conditions, and developed an artists’ habitation approach to elucidating the issues of Republican-period Shanghai urban and art history from the perspective of Chinese cosmopolitanism. We mapped areas of high artist concentration and identified a higher percentage of artists residing in the concessions (compared with the Shanghai general population) and the incremental convergence of art clusters in the concessions. We argue that the concessions provided a favourable environment for cultural diversity and the ungovernable, elite spirit of the literati tradition. The mainstream Shanghai art practices, known as haipai, were modern, as they were rooted in the urban modernity of the concessions and embodied Chinese cosmopolitanism.

Shanghai's Republican period was a golden age of art and culture. Located on the eastern end of the Yangtze Delta, Shanghai rose to prominence after the former urban centres of commerce and culture along the Jiangnan Grand Canal declined during the late Qing period.¹ Shanghai inherited the cultural accumulation and vitality from those former centres of literati culture² and developed ties to global cities through Western colonization after the first Opium War (1839–42). Shanghai's reputation as an art capital rose meteorically and was emblematized by iconic artists, such as Wu Changshi, Zhang Daqian, Liu Haisu and Wu Hufan. In its early days, twentieth-century Chinese art history scholarship discussed ‘Shanghai modernity’ through the lens of ‘China vis-à-vis the West’, perceiving art as a part of Chinese society under pressure to achieve national salvation against Western colonial powers. Some of the literature explored the Western features of Chinese artworks in order to uncover the modern components of Chinese art.³ Since the late 1990s, however, the modernity of Chinese art is believed to lie not only in the nature of art itself but also within the art world, an emerging paradigm shift.⁴

In the meantime, the growing field of Shanghai studies celebrated Republican Shanghai as a city of ‘modernity’ and modern culture. The term ‘Shanghai modern’ describes a literary attitude that embraces Western ideas and expressions; it centres on the selves of the literati by appreciating historical progress, popular culture, cultural consumption and material modernity in the context of urban culture.⁵ In contrast, modernity is embodied in the material settings of the concessions, and is expressed within hygienic, security and entertainment facilities.⁶ Samuel Liang suggests that Shanghai modernity includes a spatial perspective, dubbed ‘spatial modernity’.⁷ ‘Modernity entails a new consciousness of space and materiality and a new kind of visual experience’, one linked to an ‘imaginative geography’ of material conditions.⁸ He advocates for a cultural map of urban cultural production and consumption.⁹ ‘Haipai’ or ‘Shanghai style/school of art’, a term originally derived from regional opera, was applied to describe the city's distinctive modern cultural atmosphere and artistic attainments along with the literati tradition in new social and urban conditions, by a mixing of the elite and mass cultures.¹⁰ Can we understand haipai from the perspective of Shanghai's urban history?

Habitation study has attracted rising academic interest within the urban historical research field and has the potential to engender implications for both Chinese urban and art histories. Lu Hanchao notes that past scholarship of Shanghai studies discussed modernity mainly from the ‘occupational or vocational criteria’ perspective. However, in Shanghai, residential location was crucial for labelling people's socio-economic status, thus creating identities and shaping self-expressions. Lu's work reveals a panoramic view of ordinary Shanghai people's daily lives centring on their rowhouse residences and demonstrates the importance of habitation.¹¹

Similarly, the European scholarship of art history has considered habitation as one dimension of art-historical research for decades. This reflects modernist historiographic interest in how artists inhabited cities. Artists’ participation in modern cities’ everyday activities and their embeddedness in the transformation of urban fabrics (e.g., art studios) illuminate inquiries about the artistic zeitgeist in historical communities.¹² A recent digital mapping project of artists’ residential locations in Paris’ eighteenth-century art world, for instance, demonstrates artists’ way of living and the makeup of their communities’ geographic scope.¹³ Drawing on the perspective of urban history, art scholarship demands attention due to its role in urban development.¹⁴ Studying artists’ lives and habitation represents a new approach to narrating urban history, bridging art history and urban history.¹⁵

However, there is a dearth of Chinese historical literature that touches on the topic of artists’ residential locations. The Chinese scholarship relies solely on qualitative methods, and consequently fails accurately to determine the artists’ residential locational patterns.¹⁶ This constitutes a significant knowledge gap because most artists engaged in artistic practices and socializing activities at home; only a small subsection of them worked in institutions in a full-time manner during the Republican period.¹⁷ What kind of urban conditions, geographic facilities and features favoured artists? In which areas of the city did artists choose to live? Answers to these questions will shed new light on the collective characteristics and lifestyles of these Shanghai artists, which in turn will provide insights regarding their artistic expressions.

This article is based on a large-scale digital mapping project of Republican Shanghai artists conducted over the past four years. It is the first computer-facilitated project on twentieth-century Chinese art history that has adopted quantitative spatial analyses using GIS and R. We set the city as the foreground of the Republican Shanghai art scene and we address the correlation between the city's social and urban conditions and the artists’ spatial patterns, whilst exploring issues in both Chinese urban and art history. The main argument is that the Republican-period Shanghai art world blossomed due to the conducive environment of the concessions, and the modernity of haipai essentially embodies Chinese cosmopolitanism. This argument is supported by three points. First, the Shanghai art world featured a prosperous (fanhua) art landscape as demonstrated by identifiable artists’ densely habited areas (art clusters) and a trend of art clusters’ convergence in the concessions. Second, the concessions provided favourable environments for artists and fostered cultural diversity. Third, facilities rooted in the traditional Chinese culture and society (not Western-style facilities) were the contributing factors of artists’ spatial concentration, which indicates that the concessions allowed haipai artists to continue the traditional literati lifestyle and to pursue the unruly, individualistic and scholarly spirit of literati painting.

The lens of Chinese cosmopolitanism

The scholarship of Shanghai studies provides a unique Chinese cosmopolitan perspective (three dimensions) for our research. To begin with, Chinese cosmopolitanism features prosperous urban culture or lively festive urbanism, known as fanhua (prosperity).¹⁸ Most importantly, fanhua embodies quantitatively thriving commercial, social and cultural activities that were concentrated in focal points that circulated goods and services through trading and merchants. In such places, economic vibrancy and extravagant materiality underpinned the consumption of urban entertainment and artistic production, thus aesthetically connecting wealth to the vitality of the human world. A city of fanhua features flamboyant beauties, decorations, urban amenities for singing and dancing and the display of visual arts, and the luxury consumption of drinks and delicacies.¹⁹ The enclave-centred nature of Shanghai created the conditions for prosperity.²⁰ We applied the perspective provided through the analysis of clusters to determine if Shanghai's art world complied with the cosmopolitan principles in terms of fanhua.

A ‘cluster’ refers to the spatial concentration of actors (individuals or organizations) surrounding focal points within a certain geographic scope. A geographic cluster indicates a higher density of actors due to the convergence of actors with evidently shorter inter-actor physical distances when compared with the density outside of that area. Cluster theory explains the advantages of staying inside clusters.²¹ The main theoretical building blocks are agglomerative economies, endogenous development theory and systems of innovation.²² For instance, the classical Marshallian model explains that firms cluster to share a ‘commons’ of services and a skilled labour pool. The Californian School argues that the spatial concentration of activities enables the minimization of transactional costs. In other arguments, clusters allow actors to have flexible specializations and to create an innovative milieu.²³ An ‘art cluster’ is the clustering of artists housed by residential buildings or workplaces such as studios, cultural institutions and/or facilities. The locational art cluster, often devalued by the scholarship, is the simple geographic co-location of artists without inter-actor connections. Advantageous art clusters feature abundant inter-actor linkages and synergies that enable knowledge/skill sharing, co-learning activities and collective problem solving.²⁴ Vibrant art clusters often attract talent to join the clusters through magnetic cultural effects, leading to the growth of clusters in terms of geographic scale, cultural production and the branding effect.²⁵ In order to identify and classify art clusters, we proposed a new method that consists of four steps, including K density analysis, district-based density analysis, hot spot analysis and social network analysis.²⁶

Second, Chinese cosmopolitanism features multiple centres of influence that diffuse culture into everyday urbanism. This leads to a landscape of cultural diversity and hybridity.²⁷ Shanghai was a cosmopolis of diversity: the foreign settlements were constructed particularly for the colonialists but ended up being inhabited by a mixed population of Chinese and foreigners with diverse ways of expression and vocabularies.²⁸

Furthermore, the lens of Chinese cosmopolitanism helps to explain how the mainstream Shanghai art world, in the literati tradition,²⁹ flourished in the concessions. Meng Yue perceives Shanghai as the refuge of the former Great Canal-based urban centres of cosmopolitanism in the context of globalism.³⁰ Through immigration, the cultural vitality of Yangzhou and Suzhou, the former heartland of Chinese cultural centres, was brought to Shanghai.³¹ Hence, Shanghai modernity retained its Chinese roots and continued its traditional socio-cultural conceptions. Lu Hanchao writes that ‘the persistence of tradition in the life of the people of Shanghai only continues if we are fully aware that such continuity exists’.³² The extraterritorial rights of the concessions empowered the foreign powers to maintain the legal codes and capitalistic devices of their respective home-country governments,³³ staying away from both of the dominant Qing imperial and colonial orders.³⁴ Prior to 1927, Shanghai continued its historical trajectory established during the Qing dynasty. However, it was the absence of Qing hegemony that enabled the city to grow from a peripheral town into the cultural centre of China; it was ‘a de-centric process’ that featured prosperous and unruly cultural practices.³⁵ After 1927, the concessions were exempted from the social order of the Chinese municipality that embodied Chiang Kaishek's ambition to create ‘monopolizing power’ in command of all municipal affairs.³⁶ The foreign authorities of the concessions were determined to maintain the ‘law and order which is necessary to protect life and property in the Settlements’.³⁷ Shanghai's concessions became the haven of outlaws and subversive anarchists.³⁸

Literati culture originated from scholar officials’ leisure activities, known as amateur painters’ art. It was often the expression outlet for underprivileged officials who were excluded from the mainstream ruling circle, and hence it embodied the ungovernable, irrepressible spirit of the cultural elites. Su Shi (1037–1101), recognized as the originator of literati painting, was versatile in painting, calligraphy and poetry, and experienced numerous demotions and hardships in his official career. Other iconic masters in the literati painting tradition, such as Ni Zan (1301–74), Zhu Da (1626–1705) and Shi Tao (1642–1707), were either expelled officials or ruling family members at the end of their dynasties. They practised painting that valued a blend of individuality, literary scholarship and nature-oriented joy-seeking that facilitated an escape from the real-world rigid, hierarchical and hegemonical cultural orthodoxy. Yue writes that literati painting exemplifies ‘the uncontrollable cultural and social practice’ with ‘its own cultural logic embedded in China's late imperial history’.³⁹ As Yue notes, early immigrants to Shanghai were ‘men of letters’ (literati culture), who opted to reside in the foreign settlements as a way to escape conventional career paths serving the imperial Qing; these painters cultivated haipai painting.⁴⁰ The concessions became a haven for the blossoming of the ungovernable elite spirits who expressed the literati cultural logic.

One of the crucial conditions of the concessions was the growth of public spheres. The foreign authorities restricted their roles within public administration, focusing instead on issuing and maintaining regulations.⁴¹ Moreover, the municipal councils in the concessions created a legitimate structure that allowed the free expression of viewpoints and the hosting of public hearings.⁴² This was a necessary precondition to the success of the publishing industry as well as other cultural sectors.⁴³ Although literati painters followed the aesthetic criteria of highbrow culture,⁴⁴ unlike subvention art, they demanded the freedom of self-expression within specific types of public sphere, such as the ‘third realm’.⁴⁵ These public spheres were essential to extra-bureaucratic organizations associated with elites, literati groups, such as poetry and painting organizations, and related organizations dedicated to academic and artistic pursuits.⁴⁶ Accordingly, the public spheres of the concessions expanded to shield the non-institutional networks, including literati art organizations and individual artists.⁴⁷

Digitally mapping and analysing the art worlds in Republican Shanghai

The interdisciplinary nature of urban-historical research often challenges specialized scholars.⁴⁸ We built an interdisciplinary team comprising experts in Chinese art history, Shanghai history, geographic information systems (GIS) and quantitative geography. We created four databases of Shanghai art maps in the first phase, and defined the term ‘artist’ as a practitioner of visual art (e.g., painters, sculptors and calligraphers) each with public profiles.⁴⁹ Historical data were collected from 156 biographies, 22 autobiographies, 9 Republican-period newspapers, 33 types of Republican-period magazines, 652 academic books, 288 academic papers, 94 memories and 57 volumes of archives housed in the Shanghai Archive House. Based on the residential locational information derived from this treasure trove of historical documents, we constructed Database I, a database of historical-locational data for 1,349 artists. Meanwhile, this trove of historical documents provided us with substantial qualitative data about individual living experiences, including clues for analysing artists’ motives for their residential location choices. Database II contains the residential addresses of artists’ ‘social contacts’, defined as persons within/related to the art field that had social ties to Shanghai artists. We identified 1,373 pairwise social connections. Database III contains the records of artists whose annual residential location records were available. Only 45 (out of 1,349) artists’ yearly residential addresses were discovered, and we used these records in our analysis of artists’ relocation trajectories. To explore the environmental attributes that contributed to artists’ spatial distribution patterns, we created the Database IV that contains the addresses of entities/businesses in three broadly defined categories, i.e., social facilities, cultural facilities and cultural organizations,⁵⁰ including 591 addresses of food and beverage venues, 445 addresses of leisure and recreational venues and 1,116 addresses of cultural facilities and art shops. The information comes from 18 types of Republican-period business guides, such as the Da Shanghai Zhinan (Shanghai Business Guide).

To prepare spatial data, we selected the 1937 and 1947 editions of Zuixin Shanghai Jieshi Tu (The Latest Shanghai Neighbourhood Map) at a 1:20,000 scale⁵¹ due to their high resolution. We then performed heads-up digitizing and geo-registered our base maps using 50 control points each. Our digitized 1937 map comprises 1,087 streets, and the 1947 map comprises 675 streets. We also digitized four maps of varying administrative units in five years: 1908, 1910, 1926, 1928 and 1948, and the 1936 Shanghai transportation map. Based on Database I, we created point data shapefiles by locating artists on the base maps. The 1937 and 1947 editions of Hanghao Lutu Lu (Shanghai Commercial Guide) labelled all the residential buildings with street numbers. Based on these maps, we determined artists’ residential locations. For density and hot spot analyses, we created a map (based on the street system and administrative units) that divides the Shanghai territory into 55 sub-districts for the purpose of spatial statistics.

We proposed a new method for identifying and analysing art clusters by combining four statistical methods. The K density estimation (or K function) is the standardized cumulative average number of data points for the typical point within a specified distance. The district-based density of artists maps the density of artists in the 55 statistical districts that we created. The G statistics measure the similarity between values of artist density in statistical districts and their adjacent locations (i.e., ‘hot’ or ‘cold’ spots), defined as spatial autocorrelation. Clusters are scored on the vibrancy of social networks based on the quantity of social contacts. Combining the four dimensions, a cubic model classifies the identified clusters, the scale of their spatial scopes and social network vibrancy (Figures 5–7, 9, 13).

To determine the environmental attributes that contributed to artists’ spatial patterns, we built global and local regression models for each of the three selected years, 1920, 1930 and 1948, based on the variables’ data availability. Geographically weighted regression (GWR) models performed better than others. The GWR models not only establish a statistical relationship between artists’ density (the dependent variable) and multiple causal factors but also weigh local spatial heterogeneity.

Critical realistic and constructivist approaches influenced our interpretation of the triangulated data analysis results. Quantitative approaches scientifically uncover relationships by analysing variables, while qualitative methods describe the actual experience through hermeneutics or case-oriented analyses.⁵² A constructivist perspective suggests the co-existence of different valid yet contradictory explanations of the same reality.⁵³

Fanhua: the formation and transformation of art clusters

Adopting the methods above, we converted our hypothetical conceptualization (from the Chinese cosmopolitan perspective) into a quantitatively testable art world model and the outcome underpins the fanhua hypothesis.

We identified art clusters adopting the K density analysis and visualization. The output (Figure 1) that consists of irregular, concentric polygons with varying colour values of blue delineates the spatial scope of the clusters and the escalated density of artists’ residences (darker blue equals higher density). We found that the 36-year span of history comprises two phases when it comes to the formation and transformation of art clusters. Phase I, from year 1912 to 1936, witnessed the steady increase in the density of art clusters. On the 1912 map, the K value range of the densest cluster is 0.000004656 to 0.000004573,⁵⁴ whilst this range hits 0.000031243 to 0.000035146 in 1936. This K function analytical result aligns with our district-based density analytical result (Figure 2). The artist density range increased from 1.49–2.56 artists per square kilometre in the most densely inhabited districts of artists in 1912 to 11.2–18.91 artists per square kilometre in its 1936 counterpart. This reflects the growing artist population in Shanghai⁵⁵ on one hand and the uneven neighbourhood attractiveness across districts on the other hand. The cluster located in the north-central French Concession, in particular, achieved ascent to prominence and became the leading art hub from the early 1920s onwards. During Phase II, from 1936 to 1948, the density of artists decreased after the outbreak of the Sino-Japanese War in 1937. The K values of the densest art clusters went down from 0.000031243–0.000035146 in 1936 to 0.000022436–0.000025431 in 1948. Accordingly, the district-based density of artists slightly declined from 11.2–18.91 (1936) to 10.15–14.18 (1948) artists per square kilometre in the densest districts. In contrast, the size of artist residents in our database grew slightly from 312 to 319 during this 12-year period.

Figure 1. Visualization of the K function analytical result, 1912–48, by the authors. All the images in this article were produced by Jane Zheng.

Figure 2. District-based density analysis.

Secondly, there was a trend towards art cluster convergence in the concessions. From 1912 to the end of the 1920s, the locations of art clusters were not limited to the concessions. Instead, three areas all contained art clusters (Figure 1): (1) Zhabei, the area extending north to the International Settlement, (2) the walled city,⁵⁶ and (3) the Lao Xi Men area (Old Western Gate) located between the walled city and the south-east border of the French Concession. In 1920, the largest art cluster was located in Lao Xi Men. However, the leading role of this cluster was quickly overtaken by a new one, the north-central French Concession cluster. The cluster in Zhabei disappeared after the Japanese invasion in 1931. By 1936, the cluster in Lao Xi Men incrementally moved towards the French Concession and eventually merged into the new north-central French Concession cluster. The cluster inside the walled city declined by the end of the 1930s after the master Wang Yiting's⁵⁷ (1867–1938) departure. The total number of identifiable clusters was 11 in 1933 (Figure 1); 8.5 clusters were located within the concessions.⁵⁸ By 1936, there were 14 identifiable clusters varying in sizes (Figure 9); 11 of them were located inside the concessions. In the 1940s, most identifiable art clusters were located within the concessions. The same trend is found in the district-based density analysis (Figure 2) and the hot spot analysis (Figure 3). By 1942, the boundary of artists’ densely inhabited areas (0.5 artists per square kilometre and above) aligns with the borders of the concessions (Figure 2). Districts with a continuous density of artists (90% of confidence level and above per hot spot analysis) were all located within the concessions (Figure 3).

Figure 3. Hot spot analysis.

According to Rudolf Wagner, the intrinsic dynamics of Shanghai were embodied in its urban agglomeration and there was a tendency for Chinese people to move into the concessions.⁵⁹ In 1930 for instance, Shanghai's population was 2,980,650 (registered inhabitants), of which 33.8% lived in the International Settlement and 15% lived in the French Concession, whilst 51.18% lived in the Chinese Municipality.⁶⁰ However, we found that there was an evidently higher percentage of artists residing within the concessions. Database I includes 88 artists’ residential records from 1930: 43.18% were located in the International Settlement, and 46.6% were in the French Concession; only 10.22% were in the Chinese Municipality.

The dynamics of art clusters and cultural diversity

The concentration of art clusters in the concessions hypothetically suggests certain unique social conditions of the concessions that favoured a diverse body of artists. Wagner proposes security, conformity and access to urban entertainment as the main contributing environmental attributes.⁶¹ We examined individual artists’ location choices and their relocation trajectories in relation to art clusters, and identified the contributing factors within three historical periods from a political-economic perspective. Our findings also reveal how the concessions’ cosmopolitanism underpinned the cultural diversity of the Shanghai art world.

The emergence of art clusters in the concessions (1910s and 1920s)

When establishing the concessions, the foreign powers envisioned a cosmopolitan style of authority that disrupted the foundational logics of Qing governance while simultaneously maintaining its rule.⁶² Accordingly, they cautiously avoided subversion of the regime and advocated a collective prosperity of commerce.⁶³ We found that economic growth, a booming art market, influx of immigration, commercial and urban amenities and new neighbourhoods constructed under the European standards were the key factors contributing to the emergence of art clusters in the 1910s and 1920s.

The concessions underwent rapid economic growth during the 1910s. Foreign companies were incorporated, and banks responded to the rising monetary demand for active commerce. Increased financial-resource circulation boosted the real estate market. Abundant capital flowed into the concessions, initiated industrialization and nurtured entrepreneurs as well as the middle classes. The 1911 Revolution's dethroning of the Qing government catalysed Shanghai's increasing prosperity.⁶⁴ The concessions adopted an outlook of ‘cosmopolitanism’ and ‘entrepreneurialism’,⁶⁵ and this influenced the Shanghai art market, which prospered and entered its booming age.

Shanghai attracted immigrants from inland cities and underwent rapid population growth from 75,000 (late Qing period) to 500,000 (early Republican period).⁶⁶ Figure 4 maps 500 computer randomly sampled Shanghai artists’ provincial origins. Local artists only accounted for 4.4%; the others came from 13 different Chinese provinces. During the 1910s, several highly influential artists within the field of twentieth-century Chinese art history moved to Shanghai. The leading literati tradition painter Wu Changshi (1844–1927), for instance, used to live in Suzhou. His friend, Wang Yiting⁶⁷ (1867–1938) invited him to move to Shanghai. Wang exemplified the booming art market, and helped Wu to find residence in 1913.⁶⁸ Li Ruiqing (1867–1920), known as ‘Monk Qing’, a calligrapher and former officer of the Qing regime, moved to Shanghai in 1911. He invited his close friend Zeng Xi (1861–1930)⁶⁹ to Shanghai. Both sold their paintings and calligraphy for a living. The leading traditional Chinese-style painter Wu Daiqiu (1879–1949), known as ‘Three Wu and One Feng’ (san Wu yi Feng), moved to Shanghai in 1916, one year after his home art market of Zhejiang Congde started declining.⁷⁰

Figure 4. The provincial origins of artists, 1912–27.

Lu Hanchao identifies a transition in the composition of immigrants from the ‘social-elite type of immigrants’ to commoners during the late Qing to Republican transitional period.⁷¹ This also applies to the artist immigrants. In the 1910s, Shanghai's art world was dominated by social elites, i.e., former Qing government officers, including Zheng Xiaoxu (1860–1938), Li Ruiqing and Zeng Xi.⁷² They specialized in calligraphy and poetry. In the early 1900s, a rapid influx of people of high social standing into Shanghai introduced a refined lifestyle, one which appreciated talented singers, writers, painters and sculptors; this new art scene hosted social gatherings of art and cultural activities in the neighbourhoods near Fuzhou Road (‘No. 4 Horse Road’ then) in the International Settlement (eastern coast). These neighbourhoods occupied the bustling core of Shanghai city from the 1860s to the 1910s, as they birthed urban spectacles featuring a mixture of commerce and entertainment, e.g., dining services, publishing houses and brothels.⁷³ Hundreds of newspapers and magazines were established in the concessions, and they embraced the newfound freedom of discussion of public affairs.⁷⁴ Two major art clusters and one medium-scale cluster were located in the International Settlement (Figure 5). The talented poet Su Manshu (1884–1918), for instance, lived in Cluster 2. One medium-scale cluster was located in the walled city, accommodating three influential artists. These were the aforementioned Li Ruiqing, Zhao Shuru (1874–1945)⁷⁵ and Yang Baimin (1874–1924).⁷⁶

Figure 5. Cluster analysis, 1912.

In 1914, the Shanghai art scene prospered by hitching itself to the bandwagon of the third expansion of the French Concession, the same year that the city wall was torn down. Lin Zhuo describes a trend of extending the city's prosperity from the walled city into the Lao Xi Men area through bus line construction.⁷⁷ Our mapping project for the year 1920 affirms Lin's argument: the biggest cluster emerged in the Lao Xi Men area (Cluster 1) in that year, accommodating 13 artists, e.g., Zhang Yuguang (1885–1968), Wang Jiyuan (1893–1975), Ni Yide (1907–70) and Chu Changyan, all living within a seven-block area (Figure 6). Most were Western-style painting instructors from the Shanghai Art College, located within the same district.⁷⁸

Figure 6. Cluster analysis, 1920.

However, we argue that the expansion of the French Concession had far more far-reaching influences on the Shanghai art scene than the wall demolition. The French colonial power structure sought to create its own independent urban autonomy separate from its British rival, from the moment of the signing of the treaty.⁷⁹ The French started to build neighbourhoods according to French construction standards on its newly acquired land, and quickly attracted artists, mostly new immigrants, belonging to the ‘petty urbanites’ class (middle or lower middle income).⁸⁰ To them, those newly built neighbourhoods in the form of shikumen (stone portal or alleyway houses)⁸¹ provided affordable housing. Moreover, alleyway neighbourhoods housed daily life facilities and services, such as commerce, food stores, bookstores and alley schools, supporting residents’ daily lives with material space and everyday conveniences.⁸²

Cluster 3, which later grew to be the leading art cluster or hub, was an emerging cluster of artists in 1920 (Figure 6). A comparison of the 1912 and 1920 maps shows that few artists lived in this area before the French Concession's expansion. Also, it did not constitute a statistical ‘hot spot’ district by 1920 (Figure 3).⁸³ However, the relocation trajectories of the 45 sampled artists (Figure 10) show that, in contrast to none in 1912, 16.7% of new artist immigrants lived in the newly expanded French Concession in 1920. By 1925, 53.8% of the new immigrant artists opted for this residential area. The relocation trajectories of some of the existing artists also matched that of the French Concession's expansion. Feng Chaoran (1882–1954), an established literati tradition painter, for instance, lived at Li Pingshu's (1854–1927)⁸⁴ house near the walled city before 1914. He moved to the Baxian Qiao area within the old territory of the French Concession in 1914. When the bank comprador Gong Ziyu built the Grosse Rue neighbourhood in the expanded French Concession, Feng moved into the building at Gong's invitation (Figure 10).⁸⁵ Wagner notes that the walled city was not comparable to the concessions in terms of cultural innovation, and also social and economic conditions, and that its residents moved to the concessions.⁸⁶ By 1925, Cluster 3 (Figure 6) had become dominant (Cluster 1, Figure 7) after Wu Hufan's (1894–1968) (Figure 14) arrival in 1924. Wu, born in an imperial official's family in Suzhou, inherited extensive antique collections and moved to Shanghai to avoid warfare. He achieved a meteoric rise in the following 10 years and became a leading literati tradition professional painter. He lived next door to Feng Chaoran (Figure 14) on the same lane in Grosse Rue.⁸⁷ In addition to Wu and Feng, this cluster accommodated 20 artists in 1925, such as Xiong Songquan (1884–1961), Qian Shoutie (1897–1967), Xiao Xian (1902–97), He Xiangning (1878–1972), Shen Maishi (1891–1986) and Teng Gu (1901–41).⁸⁸ In 1920, the geographic scope of Cluster 1 (Figure 6) was largely congruent with Cluster 3 (Figure 5), covering 22 blocks. However, its spatial connection to the walled city was reduced. The emergence of this primary cluster also marked the end of the dominance of the former imperial officers in the Shanghai art world, giving way to professional artists.

Figure 7. Cluster analysis, 1925.

In the early 1920s, the Lao Xi Men area (Cluster 3, Figures 6–7) continued a growth trend, but artists began to move into the French Concession.⁸⁹ These artists’ social ties were largely based on their colleagueship at the Shanghai Art College, the city's main Western-style art-education institution under the influence of Cai Yuanpei's (1868–1940), the minister of education, aesthetic education policy.⁹⁰ Some 27 artists, e.g., Zhang Yuguang (1885–1968), Liu Haisu (1896–1994), Wang Yachen (1894–1983), Xie Gongzhan (1885–1940), Ding Song (1891–1972), Zhu Wenyun (1895–1939), Pan Yuliang (1899–1977) and Chen Hong (1898–1937),⁹¹ lived in this cluster. Relative to 1920, this cluster decreased two ranks by 1925, having been outpaced by Cluster 1 (Figure 7) in terms of artist density. The two clusters’ geographic proximity enhanced their legibility as statistically defined ‘hot spots’ (Figure 3).

The golden age of Shanghai's art world (1930s)

From 1927 to 1937, Shanghai stood at the apex of its prosperity. We argue that the prosperity of the art landscape within the concessions was enabled by the city's fragmented governance framework, the low intervention level of the authorities, the accumulated cultural atmosphere, modern urban amenities, public transport and personal security during wartime.

In 1927, Chiang Kai-shek came to power and aspired to advance the national construction programme. The Kuomingtang regime manipulated boycotting and resistance, cracked down on activists’ movements and undermined the unruly character of the city.⁹² The foreign authorities’ low intervention level into personal lives and expressions inside the concessions contrasted with the Kuomintang regime's new policies of censorship.⁹³ Also, the unique extraterritoriality of the concessions created ‘fissures’ or asylums caused by the interplay among different political forces.⁹⁴ The concessions, labelled as ‘the special municipality’ in 1927, became the resort for rowdy cultural and social activities.⁹⁵ Some artist-related cases exemplify this concept. Wang Yiting, for instance, was deemed a wanted fugitive by Yuan Shikai (1859–1916) as he was opposed to Yuan's restoration of the imperial system. He fled into the International Settlement.⁹⁶ Similarly, Pang Xunqin (1906–85) (Figure 14)⁹⁷ was considered a wanted delinquent by Chinese police in the 1930s. He was alerted to their intentions by his art friends and then he fled into the French Concession.⁹⁸

Also, the material conditions of the concessions were improved during this time period. During the 1930s, the presence of public transport was influential in artists’ residential site selection process. We analysed the tramway map of 1936 and its relation to artists’ residential location, using ArcGIS to create 300-metre and 500-metre bubbles surrounding the tramway and verified a positive relationship: 59.39% of the whole artist population lived within a 500-metre distance from the tramway; 35.37% of the artist population lived within a 300-metre distance (Table 1 and Figure 8). We also examined the correlation of the art clusters and the tramway and found that 9 out of 14 clusters were located adjacent to the tramway. Most of the artists (66.7% to 100%) residing within the clusters were located within a 500-metre distance of the tramway. Hence, access to mass transit explains 64.3% of the clustering pattern of artists.

Figure 8. Art clusters and the tramway in 1936.

Table 1. The number of artist residents in 14 clusters (1936) within a 300-metre and 500-metre distance of the tramway; percentage of artists within a 500-metre distance to the total number of artists in the clusters

Warfare was another factor that caused the influx of artists into the concessions. In 1926 and 1927, warlords forced intellectuals and artists to seek refuge in Shanghai, thereby contributing to its population growth.⁹⁹ In 1932, the Wusong-Shanghai War of 1932 (known as the ‘128’ war) broke out on 28 January. Japanese armies invaded Zhabei, devastating art schools and cultural facilities among thousands of buildings.¹⁰⁰ As a result, the former art cluster that was centred on the Lida Art Academy in Zhabei had disappeared by 1933. Refugees, including some artists, poured into the concessions that provided shelters. The concessions warded off turmoil and disruptions to businesses¹⁰¹ and were found to be ‘a precarious haven for the refugees’.¹⁰²

From the late 1910s onwards, European- and Japanese-trained Chinese artists later returned to China. Many of them chose to reside in Shanghai's concessions and were actively engaged in Western-style urbanism. Although this group of artists only accounted for 10.2% of the total artist population in our databases, they contributed to the cultural diversity of the concessions. Frederic Wakeman notes that Republican Shanghai, as a metropolis, featured modern urban and entertainment facilities, such as movies, theatres and ballrooms, sheltered by the concessions.¹⁰³ These amenities, with low statistical significance,¹⁰⁴ were favoured mainly by Chinese artists dedicated to Western-style painting and design. We found evidence for this amongst our collected historical documents. One essay documented Zheng Zhenduo (1898–1958) and Feng Zikai (1898–1975) who enjoyed drinking brandy at the Jinlong Western-style Restaurant in the 1930s.¹⁰⁵ Zheng Mantuo (1888–1961), a commercial artist and cartoon designer, had an adventurous personality and a broad range of hobbies. He often visited ballrooms and photo shops in the International Settlement where he lived (Figure 9).¹⁰⁶ As Bergère notes, cityscapes resplendent with neon lighting and cars were overflowing with symbolic capital and intoxicating delights that inspired writers and artists, or at least a subsection of them. The concessions facilitated cultural transmission as well as original creation.¹⁰⁷

Figure 9. Cluster analysis, 1936.

Our analysis of the relocation trajectories of the 45 sampled artists, from three different historical periods,¹⁰⁸ indicates that the late 1920s (c. 1928) was a pivotal turning point from which artists began to recognize the desirable environmental conditions of the concessions. The territorial concessions became the prestige destination for artist relocation. From 1912 to 1925, 7 out of 45 artists relocated, and their settlement patterns revealed no particular preference towards the concessions (Figure 10). At that time, only three out of the seven artists relocated to the International Settlement (Li Ruiqing, Zhu Wenyun and Feng Zikai); two moved out of the concessions (Huang Binhong and Zhao Shuru) (Figure 10). In contrast, the 1930s saw a high frequency of relocation: 62.8% of the artists moved at least once, and all the destinations were located within the concessions (Figure 11).

Figure 10. Relocation trajectories of artists, 1912–25.

Figure 11. Relocation trajectories of artists, 1928–36.

Furthermore, the art clusters were the major attractors for artists. A high-resolution map of the largest cluster (located in the north-central part of the French Concession)¹⁰⁹ shows that over 60% of the artists living there in 1936 moved there from other parts of the city after 1928 (Figure 11). For instance, Ni Yide (1901–70), a Japanese-trained Western-style painter, moved from Lao Xi Men into the French Concession. Zhang Chongren originally lived in Xu Jiahui and settled in the major French Concession cluster after returning from Europe. Traditional-style artists also opted for the major art cluster. Yu Jianhua (1895–1979) moved from Lao Xi Men into Jiangwan (North Shanghai, outside of the concessions) in 1931 and then moved to No. 32 Xi Cheng Li in 1932. Two years later, he moved again, into No. 66 Xi Cheng Li.¹¹⁰ In 1928, Huang Binhong moved from Zabei to the International Settlement, and then moved again into Xi Cheng Li in 1929. This trajectory mapping reveals a dynamic process of cluster growth through the inward flow of artists (Figure 12).

Figure 12. Four artists’ re-locational trajectories, 1928–36.

During the 1930s, a favourable cultural atmosphere was forged in the concessions along with a prosperous outlook for art schools and organizations.¹¹¹ Starting from the late 1920s, we identified 17 spatial clusters where artists were affiliated solely with one or two art organizations. For instance, commercial artists were highly concentrated in the eastern International Settlement. In 1928, 10 artists lived in this district and all of them were included in the Shangye Minglu (Business Directory). In 1933, the south-eastern part of the French Concession accommodated 21 artists, most of them members of the Zhongguo Huahui (China Painting Association). In 1936, the International Settlement's north-central part (Cluster 3) was affiliated with the Zhongguo Nüzi Shuhuahui (China Female Artist Association). Prominent female artists such as Chen Xiaocui (1907–68), Li Qiujun (1899–1973) and Gu Qingyao (1896–1978)¹¹² lived there. A number of art organizations committed to advocating Western-style art were established, exemplified by the Heaven Horse Society (1919)¹¹³ and the Storms Society (1932).¹¹⁴

The end of the concessions and the decline of the Shanghai art scene (1940s)

The Sino-Japanese War, formally declared in 1937, heralded the decline of Shanghai's concessions. The Japanese army made incursions into the concessions and caused the partial loss of the foreign powers’ authoritative control over these territories. The concessions were cut off from the rest of Shanghaiese territory, became isolated islands and entered a harrowing period, despite an apparent transient and fragile prosperity in 1938 and 1939.¹¹⁵ Artists’ clusters were mostly located in the concessions, which were the last resort for the security and expression of the artists who remained in Shanghai.

Between 1937 and 1944, intellectuals’ social gatherings partially ceased. The artist population dwindled as many fled Shanghai. Feng Zikai (1898–1975), for example, took his family to Hangzhou and described Shanghai as a city filled with an atmosphere of ‘killing’.¹¹⁶ In 1937, Qian Juntao (1907–98), a staff member at the Kaiming Bookstore, resigned from his post and moved inland with his family.¹¹⁷

From 1938 to 1941, the concessions were able to shelter artists and some of their anti-Japanese, patriotic art practices.¹¹⁸ The International Settlement, in particular, was transformed into a ‘concentration camp’.¹¹⁹ Jiang Danshu (1885–1962) returned to Shanghai in 1938, and resided in the upper floor of a candle and incense shop. The lodging was small, shabby and noisy; but often served as a gathering spot for artists stranded by the war. Chen Baoyi (1893–1945), Wang Yachen, Zhu Qizhan (1892–1996) and Qian Ding (1896–1989), all former Xinhua Art College faculty members and Western-style painters, remained for the purpose of book and historical-document preservation. In 1939, they hosted a collective exhibit at the International Settlement's Shanghai Da Sun Corporation.¹²⁰ Traditional-style Chinese artists organized art exhibits displaying preserved masterpieces that accentuated national patrimony and spirit.¹²¹ These activities ceased after 1941, and artists again fled inland.¹²² Art schools and organizations were dismissed, destroyed or relocated inland.¹²³ Artistic gatherings yielded to dispersed, individual, artistic creation.

On another level, the wars brought a new tide of refugees into the concessions and the real estate sector prospered in response. By the early 1940s, most newly constructed dwellings in the Nanjing Road area of the International Settlement were built in the new alleyway house configuration.¹²⁴ A critical mass of new housing led to the emergence of another big art cluster in 1948, i.e., Cluster 2 (which emerged from 1936’s Clusters 3 and 4) (Figure 13). Zhang Daqian and his associates as well as a group of professional and commercial artists lived in this area. Except for this Cluster 2 emergence, the remainder of the artists’ spatial distribution pattern did not show significant change during this period. The primary cluster (north-central French Concession cluster) remained the centre of Shanghai's art world. The convergence of Clusters 1 and 9 (Figure 9) is clearly detectable. Following Wang Yiting's departure in 1937, the spatial influence of the walled city's artists had completely collapsed by the early 1940s; he had been the keystone artist whose presence sustained that district's last remaining influence.

Figure 13. Art cluster analysis, 1948.

The literati tradition and Chinese cosmopolitanism

Bergère's work suggests close connections between prolific literary production and a cosmopolitan urban environment.¹²⁵ Lu Hanchao observes that the traditional modes of Chinese people's daily lives persisted inside the concessions; this phenomenon is indicative of Shanghai's multifaceted modernity.¹²⁶ A combination of these perspectives engenders a hypothesis that the concessions’ environment was conducive to the growth of traditional Chinese art, as the perspective of Chinese cosmopolitanism informs us. We have verified this hypothesis. We argue that most Chinese artists favoured the facilities embedded in the Chinese social tradition and continued their literati lifestyle and artistic pursuits within the concessions, which embodies the ‘Chinese cosmopolitan’ conception.

We hypothesized 21 variables constituting a creative milieu and created combinations of 12–14 variables (according to data availability) in order to model data for our three selected years. We found that the presence of most Western-style facilities and urban amenities are statistically insignificant to artist density; alternatively, the facilities that facilitated literati lifestyle are significant.

In the category of cultural facilities, ‘library and bookstore’ is a significant variable (MLR and GWR) for the models of the years 1930 and 1948. The impact of this variable on artist density is ranked #2 among variables in 1930 and #1 in 1948. The French Concession's central districts are the most highly correlated to libraries and bookstores, indicating the continuation of prior reading habits.¹²⁷ In contrast, museums were negatively correlated to artist density in 1920, suggesting that artists relied on private networks for borrowing paintings and books.¹²⁸ This explains why some leading artists like Wu Hufan and Zhang Daqian, who had lavish collections of antiques, became the core of the art cluster, not museums.

In the category of urban amenities, only Chinese-style restaurant is a statistically significant variable; Western-style food and beverage venue is not.¹²⁹ Similarly, in theory, parks ¹³⁰ fostered a modern bohemian lifestyle to showcase the visions of globality advanced by the Western bourgeois.¹³¹ Activities such as strolling in public parks, leisure complexes, amusement parks and a craze for cafés and bars reflect Western customs. Private gardens, in contrast, represent the literati tradition of the elegant gathering (yaji) of poetry, calligraphy and antiques within the private circles of scholars.¹³² Garden spaces largely escaped the purview of the Western imperialists.¹³³ Our research indicates that the presence of private gardens is a statistically significant variable for increasing the density of artists (Figure 15), whilst the presence of public parks is not. We found examples in various historical documents. One case was Wang Yiting's walled-city residence, named ‘Zi Garden’, that included a building complex and a traditional-style Chinese private garden (built in 1682). Wang hosted receptions for political and cultural elites, including Albert Einstein in 1922.¹³⁴ A number of traditional-style artists lived in the surrounding neighbourhood of the Zi Garden.

Figure 14. Mapping artists of various artistic genres.

Figure 15. GWR results: private garden and bookstores.

In the category of entertainment venues, only ballroom is statistically significant for the year 1930;¹³⁵ others (e.g., cinema, theatre) are not. This lack of significant variables contradicts the allure of the modern city, and its attendant charms, and bolsters the evidence for a literary milieu where Chinese literati lifestyles continued. Traditional literati artists found quiet spaces for reading, reflection, art practice and socialization with literati painters and elites in private gardens. They shunned the glamour of the concessions’ urban festivities and were less enthusiastic about embracing novel urban experiences.

With the continuity of literati lifestyle, the haipai artists found a sanctuary for their pursuit of the literati aesthetic ideals. The resultant artistic production was the essence of tradition, purity and individuality, that sought transcendency and escape from the real world. The three primary genres of literati painting, viz., landscape, figure and flower and bird, continued to dominate the art practices of haipai. Being unresponsive to the vicissitudes of the socio-political environment indicates an oppositional attitude towards the hegemony of both Chinese and Western authorities, and suggests philosophical insights into the essence of the world. Distancing oneself from orthodoxy enables these painters to centre on human, individual and literati aesthetics. The leading haipai artist, Wu Hufan, rarely stepped out of this residence; however, he received many visitations from his peers on a daily basis. He said that ‘a scholar does not leave his house but knows everything of the world’.¹³⁶ Wu believed that the same held true for the landscapes of the world.¹³⁷ Their artistic endeavours focused on a fusion of poetry, calligraphy and painting; these disciplines marked the signature features of literati painting traceable to the master Su Shi (Figure 16).¹³⁸

Figure 16. Paintings by Wu Changshi, Wu Hufan, Huang Binhong and Zhang Daqian, 1920s and 1930s (left to right).

Past scholarship on individual haipai artists recognizes the continuity of the literati tradition, but it interprets the modernity of haipai paintings either from the perspective of nationalism or from Shanghai commerce.¹³⁹ Our research approaches this issue from the perspective of habitation by artists and it shows how Chinese cosmopolitanism enabled the scholar artists – who relocated from former cultural heartlands (like Yangzhou and Suzhou) to take refuge in the Shanghai concessions – to pursue literati aesthetics and express the persistence of the elites’ ungovernable spirits. In other words, the mainstream artistic practices of the haipai were modern because the fundamental philosophy of literati painting, viz., individual-centred, irrepressible, scholarly spirit in opposition to the material world, was sheltered and advanced by the material, locational and social modernity of the concessions.

Conclusions: towards a new approach to twentieth-century Chinese urban and art history

This article develops an artists’ habitation approach to understanding twentieth-century Chinese urban and art history. Artists’ residences were an essential component of the urban modernity of the city, whilst their artistic production contributed to the imaginative geography of Shanghai and the modernization of Chinese art.

The implications for Chinese urban history scholarship are substantial. Our research shows that the urban art landscape, which consisted of a series of spatially concentrated areas of artists, was a crucial dimension embedded in the urban fabrics of Shanghai's urban and cultural cosmopolitanism, known as fanhua. The Republican period witnessed the incremental congregation of art clusters within the concessions with an evidently higher percentage of artist habitants in the concessions (compared with the overall Shanghai population). We identified the exact locations and spatial scopes of the art clusters with varying degrees of social contact vibrancy, as well as their transformative patterns over that historical period. This research also reveals the environmental attributes of the concessions that contributed to the influx of a diverse body of artists: including a thriving economy (growing trade and commerce, increasing job opportunities, a booming art market), favourable urban conditions (accessible transport, decent housing, modern amenities), relative security in wartime, low intervention levels from the authorities (the ‘third realm’ public sphere), cultural atmosphere (abundant inter-artist social networks, accessibility to art organizations and schools). Our research also uncovers a literati tradition-dominated art landscape possessing cultural diversity: while the Western-trained, returning artists actively engaged with modern urban amenities, most traditional-style painters located their residences in proximity to Chinese-style amenities that sustained a literati style of living.

In relation to the totality of twentieth-century Chinese art history, our perspective on urban history engenders insights regarding the modernity of haipai. We argue that haipai was modern because it essentially reflected what Meng Yue called ‘Chinese cosmopolitanism’, an insubordinate elite spirit that underscored individualistic and scholarly souls who embodied a profound sense of disconnection from the Western imperialist orthodoxy and their imperialist semi-colonies. Ironically, such a spirit was enabled to flourish by the urban modernity of the concessions.

Additionally, we pioneered methodological innovations, using large-scale computer-aid mapping and quantitative spatial analysis (e.g., K function, hot spot, GWR) to test hypotheses derived from the Chinese urban history scholarship (e.g., fanhua, diversity, social network). Although these methods are not new in GIS or quantitative geography, they are utilized to analyse the art world for the first time.