Socio-technical Systems, Public Space and Urban Fragmentation: The Case of ‘Cybercafés’ in China

Socio-technical Systems, Public Spaceand Urban Fragmentation: The Case of‘Cybercafes’ in China

Gilles Puel and Valerie Fernandez

[Paper first received, September 2009; in final form, February 2011]

Abstract

The geography of Internet access places is constructed by the interaction of thestakeholders, not only by the negotiation of local regulatory policies but also by theuses of these technologies. It differentiates urban space, contributes to urban frag-mentation and the fragmentation into archipelagos of social groups, in a relationshipwhich is not one-to-one. The new socio-technical practices of the city in interactionwith infrastructures linked to mobility and accessibility produce urban forms whichintensify social fragmentation. This work is based on field surveys conducted inChina combining face-to-face questionnaires and in-depth interviews with operators(of cybercafes, local managers, etc.) and their users.

Urban areas are changing at an increasinglyrapid pace and these changes are alteringthe ways in which people live and workthere, create wealth, interact and connectwith the places which make up the urbanspace. Nowadays, these changes are havingan impact on dimensions such as the rela-tions between transport and telecommuni-cations, between virtual communities andthose rooted in a territory, between mobi-lity tools and urban spaces, between theorganisation of the city and cyberspace.Cyberspace does not refer to a virtual spacethat is independent of geographical space

(Dodge and Kitchin, 2004) but to the net-working of places via the Internet, its infra-structures, its services, its content, and alsoits uses (Graham and Marvin, 1996). It istherefore a social construction and not justa technical construction, with an intersec-tion of physical and virtual space (Grahamand Marvin, 1996, 2001), like the digiplaces(Zook and Graham, 2007). This networkingof places equipped with information andcommunication technologies (ICT), socio-technical systems, such as cybercafes, differ-entiates space and therefore contributes tothe production of urban geography.

Gilles Puel is in the Universite de Toulouse, 44, rue Peyrolieres, Toulouse, 31000, France.E-mail: [email protected].

Valerie Fernandez is in Telecom ParisTech, SES, Paris, France. E-mail: [email protected].

1–17, 2011

0042-0980 Print/1360-063X Online� 2011 Urban Studies Journal Limited

DOI: 10.1177/0042098011410333

This paper is part of the global approachto understanding how cities and their inha-bitants interact with socio-technical systemsbased on ICT, how new urban forms emergeand how cyberspace is created in the city.Rejecting any direct or ‘loose’, technical orsocial causal plan and faced with the impos-sibility of separating ‘technical reality’ from‘social activity’ (Castoriadis, 1992), we havedecided to analyse the permanent interac-tions between technology and society(Flichy, 1995). Accordingly, we have usedthe concept of socio-technical systems tostudy ‘cybercafes’, places ‘equipped’ withICT—hybrid spaces (Couclelis, 1996) rely-ing on diverse configurations and emerginguser dynamics. Public Internet access pointsdo not exist in China, or at least only symbo-lically. ‘‘Cybercafes’’ are operated by privateactors; they designate all venues equippedfor Internet access, even those withoutcomputers.

Our research project relies in particularon the thesis advanced by Manuel Castells(1997) and Stephen Graham (2004; Grahamand Marvin, 2001). In this approach, thethesis advanced is that the dissemination ofICT is accompanied by polarisation move-ments which fragment space and reinforcethe notions of limits and separation.

The concept of urban fragmentation, asused by us, refers to the dynamics of theinterdependence and division betweensocial groups or between their relationswith the urban territory. The concept ofsegregation relates to the composition ofurban territories and is sensitive to notionsof homogeneousness and separation. Inthis paper the two notions refer to degreesin a process characterising a space. Thus,when we refer to the recent trend towardsfragmentation in Beijing, we are referringto the transition from a less fragmentedstate to a more fragmented state of urbanspace, but we refrain from discussing theabsolute degree of fragmentation.

These dynamics of fragmentation revealthe connection between activities and theplaces where the activities take place.However, with ICT, activities can, more sothan in the past, be dispersed, be carriedout in different places and at differenttimes: at the office, at home, in a ‘thirdplace’ or on a mobile basis. The standardconnection between an activity and its con-ditions is becoming more flexible, makingthe activity, time and space more flexible(Kwan, 2002; Lenz and Nobis, 2007). Asregards networked services offered by ICT,the fact of differentiating the service pro-vided, creating a situation of inequalitieswith regard to access and uses, contributesto a reduction in pre-existing socio-spatialsolidarity, especially in a context of theweakening of the ‘comprehensive ideal’ ofurban planning (Graham and Marvin,2001). These authors note that today theregulation of access to the Internet isincreasingly governed by the workings ofmarket forces and is de facto reserved forcertain social groups, segmenting themarket and fragmenting urban space:Graham and Marvin (2001) refer to ‘‘splin-tering urbanism’’ (2001). Lenz and Nobis(2007) emphasise that social groups do nothave the same fragmentation potential:their ‘mobile computer fragmenters’, oper-ating in all fields of activities, are very sen-sitive to mobility. David Mangin (2004)refers to franchised cities, characterised bythe development of areas which are not orare no longer under the control of thepublic authorities but are subject to the pri-vate logic of market-driven unintentionaldevelopment (MUD) (Mars, 2008; andHornsby, 2008). Obviously the city is essen-tially composed of somewhat fracturedspaces, and urban fragmentation is not arecent phenomenon, but its spatial organi-sation develops according to a logic of per-manent changes linked in particular totechnological and social changes following

2 GILLES PUEL AND VALERIE FERNANDEZ

a movement which can be segregative orcontribute to integration.

However, the dissemination of ICT isaccompanied by additional homogenisingeffects; this ‘pervasive’ technology con-verges with pre-existing spatial structuresand is being disseminated everywhere.Nature and technology interact closely,linked by new, increasingly invisible tech-nologies which define the positioning ofpeople and property (Sloterdijk, 2004).Software programmes write the city and‘the technological subconscious’ wouldappear to shape urban practices (Thrift andFrench, 2002). Thus, numerous researchershave highlighted the smoothing of limits,between the city and the countryside(Secchi, 1984), between private and publicspaces (Viard, 1994).

The thesis advanced here is that thegeography of Internet access points isco-constructed by an interaction betweenstakeholders (central government, local sta-keholders, etc.), both as regards the nego-tiation of local regulatory policies and usesof these technologies. In return, this geo-graphy of ‘cybercafes’ differentiates theurban space, contributes to urban fragmen-tation and the splintering of its socialgroups, in a relationship which is not one-to-one since there is a co-construction pro-cess. The new socio-technical practices ofthe city in interaction with infrastructureslinked to mobility and accessibility produceurban forms which reinforce social frag-mentation. So the differentiation of mobi-lity—(the speed as well as the ‘width’ ofaccess (for example, the possibility of con-necting more freely to foreign sites)—is thebasis of the territorial differentiation ofInternet access points in Chinese urbanareas and increases the degree of territorialfragmentation. These processes which arerelatively obscure are transforming cities,places, public spaces and the relationsbetween inhabitants and space.

The presentation of our research workis based around the following sections.Section 1 presents a research model basedon field surveys, followed by section 2 onthe dynamics of the co-regulation of thecybercafes market. Section 3 discusses theuses and appropriation of ICT, the geogra-phy and diversity of the urban forms of‘cybercafes’ and the paper ends with adiscussion.

1. A Research Model Basedon Field Surveys

Some statistics are available on the questionof Internet access points in China, showingthat in 2007 cybercafes represented one ofthe main Internet access points (CNNIC,2007).1 However, little research has beenundertaken into the question of thedynamics of the adoption and appropria-tion of Internet access systems in China(Guo, 2004; Hwang, 2005).

Two waves of field surveys underpin ourresearch.

1.1 An Exploratory Survey

The first field survey focused on Starbucksin Beijing (Puel et al., 2006). The limitedsample of 30 people aged from 20 to 60provides only an exploratory insight intothe profiles of users of these places. In theabsence of objective data on the individualswho use these spaces, we relied on pureobservation in order to construct a samplecombining characteristics such as gender,geographical origins and social profiles.

1.2 Empirical Analysis of the Constructionof Internet Offerings Based on Face-to-faceInterviews

For the second wave, three empirical studiescovering five urban areas (Beijing, Shanghai,Dalian, Tianjin and Chongqıng) carried out

‘CYBERCAFES’ IN CHINA 3

in 2007 to analyse how the interactionbetween the national regulatory policy andlocal stakeholders shapes local markets ofInternet cafes by using marketing segmenta-tion: it corresponds to different cybercafemodels corresponding to value proposalsadapted to each targeted market segment.Thus, this segmentation process assumes aspecial form in that it is the result of co-construction between the state, the regionsand private entrepreneurs. Yet the supplythus constructed will encounter demandwhich does not correspond to Internet ‘con-sumers’ but to technology ‘users’. However,ICT such as the Internet is unique in that,because of its ‘equivocal’ character, it can beused in a very wide range of ways accordingto the relevant social context (Orlikowski,1992). The very different ways in which theInternet is appropriated and used will furtherthe transition from a marketing segmentationreality to an urban fragmentation process.The urban geography of Internet cafes is thesocio-spatial expression of stakeholder inter-actions which demonstrates the considerableinfluence of users in this process.

We selected cities based on their place inthe urban hierarchy of the Chinese Internet:Shanghai and Beijing in the first two rows,Tianjin and Dalian Range 10 and Chongqingin the back row (Wang, 2006). For eachof the cities, comparable test zones wereselected according to pre-established socio-territorial variables (presence of a studentpopulation, ‘ordinary’ residential districtand ‘central’ district—with a strong com-mercial activity).

The surveys were conducted mainly inMandarin and covered 94 managers ofcybercafes and 235 users. The methodologi-cal approach combined in-depth interviewswith the operators of cybercafes and face-to-face questionnaires with users. The objectiveof the interviews was to establish three mainsets of information: first, the sector’s overalldynamics—long-established cybercafes, those

which have recently been created and thosewhich have disappeared; secondly, the entre-preneurship of the owner or manager: —thetype of ownership, management, profitabil-ity, investments, costs, turnover, profits andcatchment area; and finally, a typology ofcybercafes—choice of location, quality ofthe equipment, snapshots of the organisa-tion of the premises and standard consump-tion attitudes, observation of what happensover one or two days at different times of theday, consultation of the browsing history ofusers.

1.3 Analysis of Uses

The questionnaires were distributed in sim-plified Chinese taking into account culturaldifferences. The data collected include socio-economic characteristics (age, gender, levelof education, places where the respondentslive and work, profession, etc.) and variablesrelating to Internet usage (usage reasons,opinions on the services offered, uses, etc.)and the practices of access points (frequency,duration, reason for their choice, type ofsocial interaction, etc.). This corpus was sta-tistically processed and analysed and cross-tabulated with other studies. This work ispart of a research project carried out by amultidisciplinary Franco-Chinese team, inthe socioeconomic area. It relies on the sup-port of a scientific partnership with theUniversity of Chongqıng.1

This paper summarises these first surveywaves (the main statistical results are pre-sented in the Appendix). The following sec-tion presents our analysis of the processes ofthe co-construction of the dynamics of theregulation of cybercafes. The idea advancedis that the established top–down policies areresisted on the ground and compromisedwith the local stakeholders in their enforce-ment (circumventing the law, spatial formsof resistance such as clandestine cybercafes,etc.). The dynamics of the co-construction


of regulation shape the local market ofInternet cafes and mould forms of access(systems, contents, uses, etc.) as well as theurban dynamics.

2. The Dynamics of the JointRegulation of Cybercafes

China had become at the end of 2007 theworld’s leading country in terms of thenumber of Internet users. In January 2008,the number of Internet users in China was210 million people (excluding the SpecialAdministrative Regions). Nevertheless, theInternet remains above all an urban phe-nomenon, concentrated on developing east-ern China. Beijing, Shanghai and theprovince of Guangdong are the undisputedcapitals of the Chinese Internet, being thefocus of domain names, IP addresses,Internet sites, host computers and Internetusers. Concentrating infrastructures, servicesand users, they have penetration rates above30 per cent. These digital inequalities simplyreflect the profound development inequal-ities of contemporary China on which theyare superimposed (CNNIC, 2007; Douzet,2007; Liu et al., 2004; Puel, 2009). Accordingto the CNNIC, Internet users tend to bemale (55 per cent), urban (77 per cent), edu-cated and young (77 per cent are aged under30; only 13 per cent are aged over 40). Theyare often single (58 per cent); they are oftenstudents (36 per cent); and 52 per cent earnless than ?100 a month; peasants users areexceptional (0.4 per cent).

The development of Internet cafes canbe explained in particular by the relativelylow number of households that have anInternet connection (20 per cent in ruralareas and 50 per cent in cities) and theinsufficient number of access points in uni-versities (CNNIC, 2007). Barely 61 per centof Internet users have unlimited connection(CNNIC, 2007), whereas cybercafes, driven

by fierce competitive pressures, continuallyinvest in increasing the power of theirbandwidth and the size of their computerfleet in order to remain attractive (Zhang,2007).

Finally, and above all, types of utilisation(addiction to on-line games, IP telephony,information searches and social interaction)seem to be the other determining variable.Even if the Internet is used in its traditionalform as a communication tool (mail, messa-ging, etc.) and as a life helper or for life assis-tance (a concept which covers uses such ason-line shopping, job searches, stock markettransactions, etc.), two uses clearly dominatethe Chinese Internet: entertainment andinformation searches (more than one in twousers told us that they consult blogs ‘‘to keepup with the news’’—i.e. independent news).Thus, the entertainment function of theInternet is universal in China: according toCNNIC data, the most frequent Internet uses,in order, are music, on-line videos and on-line games, with utilisation rates considerablyhigher than the American benchmark (68versus 34 per cent for music). And CNNICfigures considerably underestimate the reality.

The structuring of supply is based on amultilevel method of governance (Hoogheand Marks, 2001) which creates layers ofactors, and the layers interact with eachother in two ways: vertically across differentlevels of government and horizontallyamong other relevant actors at the samelevel. Thus, three levels can be distinguished.

2.1 The State’s Regulatory Policy

The rapid take-off in the number of users ofthe Internet and its related markets, includ-ing Internet cafes, has attracted the interestof numerous operators seeking to capitaliseon this boom. Virtually all of China’s publicInternet access systems belong to the privatesector. They offer either Internet accesslinked to a computer, often combined with


services (drinks, meals, library, etc.), orInternet access combined with services butwithout a computer. Since 1998, the centralgovernment has been attempting to stabilise,under its supervision, the business sites ofInternet access services (hulianwang shang-wang fuwu yingyechangsu). This is a legalterm introduced in 2002 and which desig-nates ‘business sites such as Internet cafesand computer lounges’. Netbars, wangbas orcybercafes are synonyms, but ‘cybercafe’ is ageneric term used to designate any Internetaccess place (sometimes offering neithercoffee nor computers!).

In its regulation of cybercafes, the govern-ment is attempting to strike a balance betweenseveral strategic options: the development ofInternet technology among the population,the construction of a national market andthe controlling uses by controlling machines,information, citizens and places. However,this strategy is meeting with strong opposi-tion. The process of controlling Internetaccess, in the same way as access to all othermedia, can be analysed as a form of socialconstruction in which numerous stakeholdersplay a role, often rooted in local territories,but falling to a large extent within the scopeof the macro system.

2.2 Local Public Policies

In agreement with Linchuan and Liunung(2005), we consider that there is not one setof regulations but local regulations of busi-ness sites of Internet access services whosefocus varies according to the balance ofpower and spatial scales, since what may betrue at national level is not necessarily trueat provincial and local levels. If this conflictbetween central government and localauthorities falls within the scope of tradi-tional issues of multilevel governance, it isnot specific to the Internet and is amplifiedby a distance effect due to the immensity ofChina

Between the two spheres (central and local) of

real power . there is a lot of administration

but not much authority (Shue, 1988, p. 152).

A series of administrative or criminal penal-ties to be imposed on officials who fail toapply the state’s policy have not beenenforced to date (Tran Dai, 2007). The mul-tiplication and obscure nature of legal textsis another characteristic. While the penaltiesagainst cybercafe operators range from theconfiscation of equipment to the confisca-tion of their licence, and even criminalcharges, the enforcement of sentencesdepends on the ‘nature of the wrong caused’(qingjie yanzhong de) without any otherexplanation. Likewise, non-adults are notallowed in cybercafes, but interpretationsvary from under 14, to 16 or 18 (Xu, 2007).

To establish its control over the sector,the Ministry of Culture has been promot-ing, since 2003, a model based on nationalchains of cybercafes. The 10 authorisedchains all have connections with ministriesor state interests at the highest level andchains established at provincial or locallevel have all been excluded. However, thelocal authorities blocked the attempt

In some cases, such as the Guangdong, offi-

cials announced that none of the 10 national

licence-holders would be authorised to oper-

ate in the province because they had a duty

to protect local industry. In other cases,

the Xinxiban (Municipal Computerisation

Office) deliberately disregard the instructions

of the provincial authorities of the Ministry

of Culture to promote the chain model

(Linchuan and Liuning, 2005, p. 286).

2.3. The Strategies of Private Operators

To open a cybercafe, an operator must havea licence and must therefore submit a returnto numerous supervisory authorities. Themanager is legally responsible for complying


with legal obligations regarding the controlof information: the Internet cafe’s equipmentmust be ‘secure’ (that is to say, the pages vis-ited by Internet users must be stored in thecomputer’s memory), the identity of Internetusers must be controlled and their particularsmust be kept, filtering software must beinstalled, etc. Cybercafe owners must trans-mit to the public security office a list of theInternet sites which Internet users haveattempted to access. However, most of theserequirements are only implemented to a lim-ited extent; a number of minors visit cyber-cafes (Zhang, 2007; Xu, 2007) and numerousoperators do not have a licence. However, assome operators hinted to us, the guanxi(local social network) often enables them toreceive advance warning of these raids or toreduce the penalties. Attempts to encourageoperators of cybercafes to assume moreresponsibility for self-regulation remain verylimited therefore, especially as the number ofparticipants reduces the scope of these effortsand operators often benefit from the protec-tion of local authorities. Moreover, they aresubject to pressure from their users.

Finally, it should be borne in mind thatthe application of increasingly burdensomeregulations continually increases the eco-nomic constraints on operators andencourages a large number of them to resortto illegal operations. Only the biggest opera-tors can recoup costs, which benefits theother participants in the chain: softwarepublishers, computer manufacturers, theencryption industry, etc. More closely inte-grated in the guanxi, they purchase compul-sory software from the various localregulatory authorities but refrain, as we haveoften noted, from installing it so as not toslow down computers and upset their usersby reducing the quality of access, while stor-ing them carefully in case of an inspection

Who pays the money? Customers! Civil ser-

vants in Beijing levy ever increasing amounts

of taxes . as regards the other [i.e. local civil

servants] it’s the same thing! But at least they

let us work . we have to come to an arrange-

ment (manager in Shanghai).

The connivance between operators and thelocal authorities can be explained by thecommunity of economic interests whichthey share and which threatens the govern-ment’s policy.

Linchuan and Liuning compare thenational regulatory policy to a structurewithout roots

The local execution of the national policy has

long been a challenge for the central govern-

ment, given the size of the country, internal

differences and the relatively weak level of

standardisation of administrative operations.

This is a crucial point . since the vast major-

ity of Internet cafes produce local operations,

by locals and for locals’’ (Linchuan and

Liuning, 2005, p. 274).

As reported by a manager of a Xinxiban,‘‘National and provincial regulators giveonly the policy, but not money’’. And opera-tors of local netbars are better placed to ben-efit from their relations with the guanxi toevade regulations with which they do notagree and which threaten their profitability.The owner of a cybercafe gave us this exam-ple: ‘‘If the water is not of a good quality, it’spointless to check the tap too often’’. Here,the water is the content and the tap is thecybercafe. However, certain Internet users(the silent majority) and pressure groups(parents and teachers) are also importantstakeholders as regards cybercafes or theprovision of the Internet services more gen-erally. They play an effective intermediaryrole for the central state, which slightlymodifies Lichuan and Liuning’s assertion.

Despite regulation at national level,owners/managers therefore have consider-able leeway to create their own value


proposal and the related business model oftheir cybercafe according to the type of cus-tomers they target.

3. The Geography and Diversity ofUrban Forms of Cybercafes: Usesand Appropriation of ICT.

The surveys reveal several urban forms ofpublic access corresponding to differentbusiness models consistent with the socioe-conomic spaces where these cybercafes areinstalled and shaped by the way in whichtheir particular groups of users appropriatethe Internet. We distinguish three models:the Starbucks models (model 1), neighbour-hood cybercafes (model 2a), the multiser-vices model (model 2b), the ‘community’model (model 2c) and alternatives (model3). Of the five cities studied, we construct amodel of the geography of cybercafes atwork in Chinese cities by eliminating thecontingent aspects: Tianjin, located intoBeijing’s orbit, or Chongqing with its spe-cific internal geography and low rate of for-eigners, for example.

3.1 Urban Centres and Pericentres

There are absolutely no cybercafes in thecore of the city; however, it does have sev-eral Starbucks and up-market Chinese cafes.All have a wi-fi connection and are used byexpatriates and affluent Chinese citizenswho work in this district; they go there withtheir laptop, their mobile phone and con-nect to the Internet while talking withfriends over a cup of coffee. Thus, theStarbucks occupy the urban space of pre-mium users. The selection determined bythe very high prices2 and the strong symbo-lism offers individuals roaming access, athird access point between their office andhome, where professional and/or privateinteractions remain uninterrupted or simul-taneous. This nomadism is the result of a

provisional but intentional immobilitywithin mobility, popular with businessmen

Every morning, I start by coming and having

a coffee, even though I have of course had one

at home; I don’t stay long, just long enough to

organise my day, then I generally come back

after lunch and at that time I manage to make

a business appointment. On Fridays, I come

at the end of the day, to relax and meet my

colleagues or some customers.

This territory, a strong symbol of theAmerican way of life in China, represents aspatial form that is renewing the urbanlandscape and a territorial extension of theglobalised economy, in the image of luxuryhotels (Sanjuan, 2001), a ‘census agora’ towhich the local and foreign social eliteflock, often travelling long distances.

I come here because it’s a good place; there

are Starbucks nearer my office and I some-

times go there to work. But I come here with

friends and it’s really popular. It’s clean. . I

like the atmosphere.

Yet it was a European woman who declaredthat the place ensured that she saw

respectable and clean Chinese people . not

like those in the street. Here, this is the real

China, a lot of artists, people from the film

industry or television.

Starbucks are accessible and visible: as theyare not considered cybercafes, they are notsubject to the regulations. Starbucks hasbecome a trendy place adapted to the veryhigh spatial capital (Levy, 2003) of mem-bers of the local elite and expatriates.

Along my route I visited Jianguo Gate, the

Parkson Store, Financial Street, the HQs of

the biggest Chinese banks, and then I visited

the western commercial district of Xidan .


The reason I wanted to visit all the Starbucks

was that I wanted to see the city. Starbucks

helped me to choose the best spots in the city.

Their market research is excellent. Anywhere

there’s a Starbucks you can be sure of two

things:

—there’s a pool of good customers

(managers or tourists)

—the location offers a good view of the

city

. Just look at a map, and you’ll see it’s true

(Wang Jianshuo; quoted in Puel et al., 2006,

p. 40).

Their carefully selected situation in a space offlows is proof that geography matters, espe-cially as the logic behind their location leads toa geographical concentration which impactson the production of urban space and whichis the focal point of the spatial habits of inha-bitants. Generally speaking, in Chinese cities,central areas inhabited or used by affluentsocial groups are the territories marked by thefree wi-fi access of trendy places (62 per centcome here for social interaction and 90 percent use the Internet as a ‘communicationtool’) (see Figure 1).

Everything around the pericentre wherethe Chinese middle classes live is domi-nated by the emerging model of multiser-vices cybercafes backed by the chains andlarge organisations.

Halei, created in July 2005 in Dalian, has300 flat-screen computers with webcams.The 10-megabyte Chinatelcom andChinanet double optical fibre network pro-vides rapid and stable Internet access. Yet itoffers far more elaborate services than asimple cybercafe; it has a bar, a restaurant,an electronic games room and a billiardsroom. This type of cybercafe is the trend inDalian as in other Chinese metropolises:the idea is to multiply services, to create a

space where all customers can find whatthey are looking for. The aim of this posi-tioning is to attract more customers thantraditional cybercafes based on a multiser-vices approach. However, they also requirea far more substantial financial standing.

However, competition is fierce and isincreasing with the development of majorbar groups that are opening premises nearthe most profitable cybercafes; these bars(such as those owned by the Hoju chain)also offer a free wi-fi service to customerswith their own laptop and base their busi-ness model on the consumption of relatedproducts, adapting the Starbucks model.Some expatriates from Beijing told us thatthey sometimes visit these places in orderto view certain websites to which access hasbeen temporarily blocked by the authoritiesin the embassy district in which they work.

Figure 1.


3.2 Intermediate, University andResidential Districts

The second district, where research centres,most of the universities and a high-techindustrial zone are located, is the areawhere the city’s cybercafes are concen-trated: their customers are virtually all stu-dents. It should be remembered that themassive presence of cybercafes contributesto the Chinese education system. Despitethe efforts made by universities, networkinfrastructures are inadequate. As all stu-dents are housed on campuses within theuniversities, those who do not have anInternet connection or want more freedomopt for a neighbourhood cybercafe. Theircriticisms concern the lack of speed due tosharing bandwidth with numerous users,but also the restrictions on the scope ofaccess due to a shortage of computers, lim-ited access hours and the surveillance towhich they are subject. It is in ‘student’cybercafes that information-related usesand social interaction are the lowest (seethe Appendix); entertainment and mostlygames (81 per cent) dominate.

The third district is a popular districtwhere ‘ordinary’ citizens live. It has threetimes fewer cybercafes, but these satisfy thebasic needs of local residents. As a generalrule, the quality of service that they provideis superior to that of the student district. Thedecoration is more elaborate and the envi-ronment is also more welcoming. They tendto be located in the hubs where populationflows are concentrated, such as shoppingmalls. These intermediate districts (residen-tial or student areas) stand out for theirdiversified offering linked to the user area,combining neighbourhood services, placesdedicated to games or social interaction, andsometimes they even affect the cybercafe’sorganisation. Thus, the Ming Ren Jin Dian issub-divided into four areas, plus VIP rooms.One area, the ‘reserved’ area, is devoted to

entertainment, the second to games, one tochat and the last one to downloading

I am trying to offer multiple services while

creating a venue that is welcoming and com-

fortable for everyone. Also I have to make

optimum use of the equipment and avoid

disputes between customers.

Obviously, the biggest area is that reservedfor games.

It is also in these urban territories thatvery localised and community niche mar-kets are developing as community cyber-cafes. Cybercafe operators adapt to marketdynamics and to their catchment area—that is to say to the characteristics of theirterritory.

Tian Xi is a Korean cybercafe in the univer-sity district in Tianjin; access is not restrictedbut, as the computers are in Korean, it is usedsolely by Koreans. Its services are fairly com-prehensive with a rest room (with a noticeboard allowing the exchange of information,news, offers, etc.), a payphone room (the linesare, of course, connected only to Korea),mobile phone chargers and evening newspa-pers; an ‘after-sales service’ option even offerscustomers a computer home-repair service.The majority of customers are members:there is a high degree of social interactioneven if they come for the games, the chat,information searches, the stock market andfilms (Xu, 2007).

3.3 The Urban Fringe

Finally, in the urban fringe areas, there are,on the one hand, traditional neighbourhoodInternet cafes and, on the other hand, clan-destine places specialising in games. It is dif-ficult to gain access to these places for ourresearch for obvious reasons. In the face ofthe recent tightening of regulatory controls

a section of Chinese youth falls back on clan-

destine gaming rooms, sometimes installed


in cars abandoned in car parks of apartment

blocks in Beijing.3

The strict application of the regulations isvery expensive and therefore almost impossi-ble for independent operators; as a result,they have no choice other than to join achain, sell their licence on the Black market,or go underground and specialise in the on-line gaming market. This market integratesin a highly competitive and global frame-work, alongside real players, urban studentsworking in small units located in the urbanfringe or in farms. Brokers sell the gamingelements that they collect to mainly NorthAmerican and European ‘players’ (Ge, 2006).

At international level with multinationalsand at national level, these forms of resis-tance are officially prosecuted, with recur-ring campaigns for the development of a‘healthy Internet’, but at local level suchaction is limited, except in very specific caseswhere the political interaction of local stake-holders involves responding and reacting toan incident for the sake of appearances.

These case studies clearly demonstrate theextent to which the government’s strategy ofcontrol and homogenisation is handicappedby the problem of the social practices ofusers. Contrary to the indications of theempirical studies on uses observed in Europe(Puel and Vidal, 2004), Chinese users do nothesitate to travel long distances to circum-vent this technocratic territorialisation (Xu,2007; Zhang, 2007; Linchuan and Liuning,2005). Some are interested only in high-speed access (for leisure activities), secureaccess (for stock market transactions), acommunity goal (for social interaction) orunsupervised access (to read or disseminateinformation), but many others simply wanta public place, a meeting place. The existenceof common interests shared by local authori-ties and users is the key favourable element.According to Guo (2004), this does not existin small towns. On the contrary, in Italy,

where the new regulation making it compul-sory since 11 September 2001 to present anID card is applied, communities of illegalimmigrants have disappeared from cyber-cafes which have had to diversify the servicesoffered (VOIP, money transfers, fax, photo-copies, etc.) or close (Thiam, 2009).

4. Discussion

Over and above the contextual differencesbetween five major cities, a form of modelcan be detected and a number of pointsemerge.

First of all, according to the operators, thecybercafe sector is apparently still promisingbecause of the extremely rapid breakthroughof the Internet in society and the addictionto certain uses (games and information),combined with the ever-increasing attrac-tion of cybercafes as a public space for socialinteraction (37 per cent of Internet usersdeclared that they used them in 2007 versus27 per cent in 2005; CNNIC, 2008). Thisindustry has become fiercely competitive.Thus the 2007 government ban on any newcybercafe openings has fuelled the develop-ment of a Black market for the resale oflicences which has become highly specula-tive. It is subject to a process of consolida-tion encouraged by the government and itsregulatory policy, which is ushering in majorchange: a decline in traditional cybercafesand the elimination of clandestine cyber-cafes in the pericentre where new models areemerging, combining a technological offer-ing, related services and increasingly refinedopportunities for social interaction.

In this ecosystem under (re)construction,the central government is assembling behindit a network of allies composed of nationaland foreign ICT-sector stakeholders (com-puter manufacturers, game software pub-lishers, encryption specialists, Internet andtelecoms operators, etc.), major cybercafe


operators and certain pressure groups (par-ents, teachers, etc.).

However, this model is encountering thereservations of local authorities, operatorsthat wish to retain control over the cyber-cafes and the misgivings of users who wantan open Internet. These protean alliances ofopponents, combining small operators, thechains and local authorities, users who wantan open Internet, ‘minors’ and the under-ground economy, are resisting and con-structing a system of public Internet accesswhich, although official, is not subject tonational regulations on the one hand, withalternative urban forms in the urban fringeon the other hand.

The resultant segmentation of publicInternet access points goes hand-in-handwith the fragmentation of an increasinglysegregative urban space.

On the one side, a first universe of cyber-cafes corresponds to the new social controlsthat the government wants to impose: thesesupervised spaces (Lianos, 2001) are fre-quented by users who, irrespective ofwhether or not they are informed of the realaccess conditions, are either not really both-ered (they view the Internet as a range ofopportunities to improve their own life), orsupport the values defended by the authori-ties (the Internet must be ‘safe and healthy’)and defend them actively (as proven by thesuccess of the campaign to denounce crimi-nal acts committed on the Internet). Theirsuccess is also based on the strong desire forsocial interaction and the sign of high socialdistinction that it represents. On Sundaysand certain weekend evenings, these placesare frequented by the ‘chiburbanites’ wholive in walled urban enclaves (compounds,villa communities for the most affluent, orother gated communities). The social prac-tices of this type of cybercafe add to thefragmentation of the city into archipelagos,into networks of places cut off by materialor symbolic barriers

The chiburbanite goes from one specific loca-

tion to the next inside a car. There is no contact

with the transition—with the city itself. This is

only a time gap (Hornsby, 2008, p. 206).

In these intermediate spaces, the differentforms of access have in common low socialdiversity and therefore a high segregativevalue. According to our surveys, cybercafes,whether they are community models oramong the most refined models, lock downcustomers, whether the barriers are theprice, keyboard design or language or uses.Even neighbourhood cybercafes, despite thefact that they focus on the needs of custom-ers in their catchment area, are threatenedby the urban transformation which is lead-ing to the disappearance of siheyuans andtheir social mixity.

At the other end of the scale, there are‘fringe’ spaces, clandestine or semi-legalplaces (sometimes temporary), frequentedby young people, ‘non-adults’ or students,and individuals who aspire to an openInternet. The main characteristic of theseplaces is perhaps their spatial fluidity, elud-ing the control of the authorities andconstantly on the move. They differ funda-mentally from other illegal premises, suchas those of the sex industry (karaoke bars,massage parlours or hairdressing salons),which are found everywhere in the city,because they are directly controlled by theauthorities (Farrer, 2008). In compensationfor this spatial precariousness, types of usesthere are freer and access costs are lower.These ‘low-cost’ access points contribute ina two-fold way to the dynamics of mobility:on the one hand, thanks to the spatial capi-tal4 of their users, they are fully integratedinto the globalised cyberspace of players;on the other hand, they interact with theurban space, switching continuously to freeterritories, the underworld of the Internet,driven away by urbanisation and regulationto be reborn elsewhere, in the image of


temporary dwellings and floating villages.Beyond the periphery, far more rare, theymix with non-permanent and semi-legaldwellings, continually pushed back byurban extension. Their low societal diver-sity (young and students or ‘mingongs’) isanother of the characteristics that theyshare with wi-fi territories.

The latter, which escape most of therestrictions—since regulation of them ismore flexible and technological monitoringtools less effective—continue to be reservedfor the foreign or local urban elite, the ‘pre-mium users’. Although the range of accessthere is wider (in particular with regard toforeign sites), they are frequented by the‘right kind of people’, with the selectionprocess being based on price, social codesand social capital. If they share with otherplaces of consumption, such as trendy res-taurants and designer boutiques, similarsocial codes and related representations(effects of belonging and exclusion), theydiffer from them by their high degree ofsocial interaction.

5. Conclusion

Several forms of reasoning seem to inter-twine the dynamic of urban fragmentationfavoured by cybercafes in China: those ofplanners and public authorities, those ofoperators and the market, and finally thoseof inhabitants and their uses.

The state’s strict Internet controls lead tocybercafe models which target wealthy cus-tomer segments (the obligation to hold anoperating permit and other restraints whichincrease the cost of the cybercafe businessmodel; the official channels charge highrates). Another consequence is the locationof illicit cybercafes on the outskirts of citiesbecause that puts them ‘out of reach’ of stateagents. Another consequence is an emergingneed among users for a ‘free’ Internet.

Yet state regulation is losing its power atlocal levels, which enables private operatorsto meet the latent demands of users by diver-sifying cybercafe business models: the valueoffered is not Internet access but a rangeof ‘Internet’ services and therefore diverseInternet uses. The different offerings corre-spond to a marketing segmentation targetingdifferent geographical areas of the city andtherefore contribute to the city’s geography.The social groups, among whom there is astrong demand for differentiated urban prac-tices and uses, in turn reinforce the fragmen-tation movement: geographical and socialfragmentation, because it is furthered byforms of social homophily.

The socio-spatial fragmentation trendwhich we have noted in the case of cyber-cafes corresponds to a more general butrecent territorial differentiation movementin China (Chang, 2006). The social diversitywhich characterised all districts is becomingless perceptible. The geography of cybercafesfollows the MUD, increasingly separatingthe rich from the poor as Chen Xiaofeng5

underscores. ‘‘Communist China has beenconverted into a bastion of individuals fend-ing for themselves.’’6

Thus Chinese metropolises appear as suc-cessions of networks of places appropriatedby social groups that create differences andare segregative. Socio-technical systems,such as the cybercafes studied here, supple-ment these networks and possibly even rein-force them. They correspond to a two-foldabsence, the public place as a place of con-tact and privacy,7 and serve as a bridge,ensuring continuity in the urban space ofthe elite.

Notes

1. The CNNIC is a government agency whichoperates under the supervision of theMinistry of Information. Every six months itpublishes a statistical report on the Internetin China. Its methodology corresponds to


international standards. To collect data ituses metrology software, on-line question-naires and telephone surveys. The latter datacollection method is undoubtedly the mostdistorting because the Internet users areidentified: thus it underestimates the users ofgames, victims of moral disapproval whooften take refuge in illegal activities. AnInternet user, according to the CNNIC is aChinese citizen aged over six who uses theInternet at least one hour a week or (sincethe June 2007 survey) a Chinese citizen whohas used the Internet in the past six months.

2. A cup of coffee at Starbucks costs around e2.The average salary in Beijing in April 2006was estimated to be e120, or 60 cups ofcoffee.

3. See: http://www.01net.com/article/320277.html(consulted on 16 October 2007).

4. Series of capitalised resources that can bemobilised by an operator to benefit from theutilisation capacities of the territories andtheir spatial differentiation.

5. See Courrier International No. 925, July2008, pp. 36–37.

6. See http://sat.burb.tv/view/1._project_outline(consulted on 05/01 January 2009).

7. The terms privacy or intimacy have noequivalent in Chinese.

Funding Statement

This work was partially supported by fundingfrom the National Research Agency and theInstitut des sciences de la communication duCNRS (Paris)

References

Castoriadis, C. (1992) Article technique, Ency-clopædia Universalis, 22, p. 125.

Castells, M. (1997) The Power of Identity, TheInformation Age: Economy, Society and Cul-ture, Vol. II. Cambridge: Blackwell.

Chang, L. (2006) The internal logic of spatialproduction of Beijing (http://burb.tv/view/POLICY_SPRAWL?_session=lghvto9re3l3ta4mogaldij7e1).

CNNIC (China Internet Network InformationCentre) (2007) 20th statistical survey report

on the internet development in China (http://www.cnnic.net.cn/en/download/2007/20thCNNICreport-en.pdf).

CNNIC (2008) 21st statistical survey report onthe internet development in China (http://www.cnnic.net.cn/uploadfiles/pdf/2008/2/29/104126.pdf).

Couclelis, H. (1996) Editorial: The death ofdistance, Environment and Planning B, 23,pp. 387–389.

Dodge, M. and Kitchin, R. (2004) Flying throughcode/space: the real virtuality of air travel,Environnment and Planning A, 36, pp. 195–211.

Douzet, F. (2007) Les frontieres chinoises del’Internet [The Chinese frontiers of the Inter-net], Herodote, 125, pp. 127–142.

Farrer, J. (2008) Play and power in Chinesenightlife spaces, China: An InternationalJournal, 6(1), pp. 1–17.

Flichy, P. (1995) L’innovation technique [Techni-cal innovation]. Paris: La Decouverte.

Ge, J. (2006) Chinese gold farmers in the gameworld I, Consumer Studies Research Network,7(2) (https://netfiles.uiuc.edu/dtcook/www/CCCnewsletter/7-2/jin.htm).

Graham, S. (2004) The Cybercities Reader.London: Routledge.

Graham, S. and Marvin, S. (1996) Telecommuni-cations and the City. London: Routledge.

Graham, S. and Marvin, S. (2001) SplinteringUrbanism. London: Routledge.

Guo, L. (2004) Internet development in Chinesesmall cities (http://www.markle.org/news/case_studies.pdf).

Hooghe, L. and Marks, G. (2001) Multilevel Gov-ernance and European Integration. Oxford:Rowman and Littlefield.

Hornsby, A. (2008) Hey fuck! Where’d the citygo?, in: N. Mars and A. Hornsby (Eds) TheChinese Dream, pp. 198–248. Rotterdam: 010Publishers.

Hwang, J. (2005) Coffee goes to China: an exami-nation of Starbucks’ market entry strategy,Journal of Undergraduate Research, 6(8)(http://www.clas.ufl.edu/jur/200507/papers/paper_hwang.html).

Kwan, M. P. (2002) Time, information technol-ogies and the geographies of everyday life,Urban Geography, 23(5), pp. 471–482.

Lenz, B. and Nobis, C. (2007) The changing allo-cation of activities in space and time by the


use of ICT—‘fragmentation’ as a new conceptand empirical results, Transportation ResearchA, 41, pp. 190–204.

Levy, J. (2003) Capital spatial, in: J. Levy and M.Lussault (Eds) Dictionnaire de la geographie etde l’espace des societes, pp. 124–126. Paris: Belin.

Lianos, M. (2001) Le nouveau controle social [Thenew social control]. Paris: L’Harmattan.

Linchuan, J. and Liuning, Z. (2005) Through theprism of the Internet cafe: managing access inan ecology of games, China Information, 19,pp. 261–297.

Liu, G., Lu, H. and Qian, L. (2004) Spatial-temporal analysis of regional disparities ofinternet development in China, Chinese Geo-graphical Science, 14(4), pp. 314–319.

Mangin, D. (2004) La ville franchisee: formes etstructures de la ville contemporaine [The fran-chised city: forms and structure of contem-porary cities]. Paris: Editions de la Villette.

Mars, N. (2008) The Chinese city: a self-contained utopia, Architectural Design, 78(5),pp. 40–43.

Orlikowski, W. (1992) The duality of technology:rethinking the concept of technology in orga-nizations, Organization Science, 3, pp. 398–427.

Puel, G. (2009) Geographie des lieux d’acces aInternet: les conditions de l’acces public etles modeles d’usages dans les grandes villesde Chine [The geography of Internet accessplaces: conditions of public access and usagemodels in large Chinese cities], Espace geo-graphique, 1, pp. 17–29.

Puel, G. and Vidal, M. (2004) Les lieux d’acces aInternet en Midi-Pyrenees [Internet accesspoints in Midi-Pyrenees]. University of Tou-louse (http://www.ardesi.fr/IMG/pdf/syntheseEPN_ARDESI-GRESOC_2004.pdf).

Puel, G., Pons, B. and Xiaoting, J. (2006) Newsforms of mobility and social practices: theStarbucks cafes in Beijing, China Perspectives,67, pp. 34–44.

Sanjuan, T. (2001) Lieux et nouvelles sociabi-lites en Chine aujourd’hui: les grands hotelsa l’avant-garde [Places and new sociabilitiesin China today: forward-thinking luxuryhotels], in: R. Hagelstein and P. Servais (Eds)Perception et organisation de l’espace urbain:une confrontation Orient–Occident [Percep-tion and organisation of urban space: an

East–West comparison], pp. 145–163. Lou-vain-la-Neuve: Academia Bruylant.

Secchi, B. (1984) Il racconto urbanistico. Torino:Einaudi.

Shue, V. (1988) The Reach of the State: Sketches ofthe Chinese Body Politic. Stanford, CA: Stan-ford University Press.

Sloterdijk, P. (2004) Spahren III, Schaume, Suhr-kamp. Paris: trad. L’Ecume.

Thiam, K. N. (2009) Les usages et les conditionsd’acces aux TIC par les emigres senegalaisvivant en Italie: le cas de Brescia [Uses andconditions of access to ICTs by Sengaleseemigrants living in Italy: the example of Bres-cia]. MA thesis, University of Toulouse.

Thrift, N. and French, S. (2002) The automaticproduction of space, Transactions of the Insti-tute of British Geographers, 27(4), pp. 309–355.

Tran Dai, C. (2007) L’Internet en Republiquepopulaire de Chine: vers un outil de democrati-sation ou de renforcement de l’Etat-parti? [TheInternet in the People’s Republic of China:towards a democratisation tool or the rein-forcement of the State-party?]. Fondationpour la Recherche Strategique (http://www.frstrategie.org/barreFRS/publications/rd/RD_20070126.pdf).

Viard, J. (1994) La societe d’archipel ou les terri-toires du village global [The archipelago soci-ety or the territories of the global village]. LaTour d’Aigues: Editions de l’Aube.

Wang, M. (2006) The network advantage ofcities: the evolving urban geography of Internetin China. Centre for Modern Chinese CityStudies, East China Normal University(http://web.mit.edu/dusp/chinaplanning/publications.htm).

Xu, C.-X. (2007) Usages et conditions de l’accespublic a Internet a Pekin et Tianjin [Uses andconditions of public access to the Internet inBeijing and Tianjin]. MA thesis, Universityof Toulouse.

Zhang, M. (2007) Usages et conditions de l’accespublic a Internet a Shanghai, Dalian et Chongq-ıng [Uses and conditions of public access tothe Internet in Shanghai, Dalian and Chongq-ıng]. MA thesis, University of Toulouse.

Zook, M. and Graham, M. (2007) The creativereconstruction of the Internet: Google andthe privatization of cyberspace and digiplace,Geoforum, 38, pp. 1322–1343.


Tab

leA

1.

Dat

afr

om

surv

eys

under

take

nin

2006

and

2007

Star

buck

sm

odel

(1)

Mu

ltis

ervi

ces

mod

el(2

a)

Nei

ghbo

urh

ood

mod

el(2

b)

Com

mu

nit

ym

odel

(2c)

Stu

den

td

istr

ict

Oth

er

Dat

afr

omin

terv

iew

sw

ith

man

ager

sC

yber

cafe

sfr

om

the

two

surv

eys

(144

)a50

2156

143

Co

mp

ute

rs/c

yber

cafe

sN

on

e20

016

517

513

0A

rea

insq

uar

em

etre

s20

040

025

036

023

0

Cu

stom

erse

gmen

tsac

cord

ing

toth

ein

terv

iew

edn

anag

er(p

erce

nta

ges)

Stu

den

tsb

No

resp

on

se5

8010

5N

eigh

bo

urs

6515

8040

Oth

erci

tize

ns

200

055

Min

gon

gc10

510

0

Cos

tin

yuan

/hou

rN

orm

alF

ree

bu

tco

nsu

mp

tio

nex

pen

sive

(25

yuan

sm

inim

um

)

2.5–

31.

63

Fo

rm

emb

ers

2–2.

51.

22.

1F

or

on

en

igh

t9

98

Dat

afr

omqu

esti

onn

aire

s

Use

rs(2

35)d

Stu

den

ts(1

13)

730

9020

10M

ingo

ngs

(18)

00

515

0O

rdin

ary

citi

zen

s(4

2)0

405

6585

Up

per

soci

o-p

rofe

ssio

nal

cate

gory

(62)

9330

00

5

Dis

tan

ce-t

ime

from

cybe

rcaf

e(h

ome,

wor

k)le

ssth

an10

min

ute

s(w

alki

ng,

car,

publ

ictr

ansp

ort)

(per

cen

tage

s)

6370

9010

030

(con

tin

ued

)


Tab

leA

1.

(Conti

nued

)

Star

buck

sm

odel

(1)

Mu

ltis

ervi

ces

mod

el(2

a)

Nei

ghbo

urh

ood

mod

el(2

b)

Com

mu

nit

ym

odel

(2c)

Stu

den

td

istr

ict

Oth

er

‘‘Wh

yd

oyo

uch

oose

this

cybe

rcaf

e?’’

(per

cen

tage

s)e

Pro

xim

ity

3360

9090

10T

ech

nic

aleq

uip

men

t10

3580

2020

Am

enit

ysi

te52

4510

835

Soci

alin

tera

ctio

n66

425

2070

Inte

rnet

acce

ss(p

erce

nta

ges)

Ho

me

100

5420

3030

Wo

rk90

7810

014

15

Use

s(p

erce

nta

ges)

e

En

tert

ain

men

t5

3490

3490

Lif

eh

elp

er10

5616

5616

Co

mm

un

icat

ion

too

l90

7490

7490

Soci

alin

tera

ctio

n62

2812

1460

Du

rati

on

/vis

it(a

vera

ge)

(min

ute

s)O

ver

6030

–60

180

30–

60O

ver

60F

req

uen

cy/v

isit

(ave

rage

)Se

vera

lti

mes

/wee

kT

wic

ea

wee

kSe

vera

lti

mes

/wee

kO

nce

aw

eek

On

cea

wee

k

Do

you

use

itas

am

eeti

ng

pla

ce?

(per

cen

tage

s)70

3010

1035

Hav

eyo

um

etn

ewp

eop

leh

ere?

(per

cen

tage

s)35

50

55

a 94cy

ber

cafe

man

ager

sw

ere

inte

rvie

wed

;50

Star

bu

cks

wer

evi

site

d,

bu

tth

em

anag

emen

tre

fuse

dto

be

inte

rvie

wed

.bT

his

cate

gory

isco

mp

ose

do

fst

ud

ents

and

hig

hsc

ho

ol

pu

pil

s.c p

easa

nt,

mın

+u

nsk

ille

dw

ork

er,

g� on

g:P

ejo

rati

veex

pre

ssio

nu

sed

tore

fer

tom

igra

nt

wo

rker

sin

Ch

ina.

dT

his

typ

olo

gyis

bas

edo

nin

com

e(a

nd

wit

hre

fere

nce

toth

eC

NN

IC).

Itco

ntr

asts

the

low

est

wag

es(s

tud

ents

,m

ingo

ngs

)an

dth

eh

igh

est

wag

es(a

bo

ve50

00R

MB

per

mo

nth

).T

he

inte

rmed

iate

cate

gory

(‘o

rdin

ary

peo

ple

’)co

mb

ines

emp

loye

es,

laid

-off

and

reti

red

wo

rker

s.e Q

ues

tio

nta

ken

fro

mth

eq

ues

tio

nn

aire

.U

sers

can

give

seve

ral

answ

ers.

Not

e:It

was

no

tp

oss

ible

toco

llec

tan

yem

pir

ical

dat

afo

rm

od

el3,

for

the

reas

on

ssp

ecif

ied

inth

ete

xt.


Socio-technical Systems, Public Space and Urban Fragmentation: The Case of ‘Cybercafés’ in China

Documents